HARVARD UNIVERSITY

Library of the
Museum of
Comparative Zoology

I

k

V,

f

The Wilson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. 108, NO. 1 MARCH 1996 PAGES 1-204

(ISSN 0043-5643)

Thk Wii^oN Oknithoi.ogical Society
Founded Decemheu 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.
President-Keith L. Bildstein, Hawk Mountain Sanctuary, RR 2, Box 191, Kempton, Pennsylvania
19529-9449.

First Vice-President— Edward H. Burtt, Jr., Department of Biology, Ohio Wesleyan University,
Delaware, Ohio 43015.

Second Vice-President— John C. Kricher, Biology Department, Wheaton College, Norton, Mas-
sachusetts 02766.

Editor— Charles R. Blem, Department of Biology, Virginia Commonwealth University, Richmond,
Virginia 23284-2012.

Secretary— John A. Smallwood, Department of Biology, Montclair State University, Upper Mont-
clair, New Jersey 07043.

Treasurer— Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana
46556.

Elected Council Memhers— Donald F. Caccamise and Laurie J. Goodrich (terms expire 1996),
Carol A. Corhat and William E. Davis (terms expire 1997), and Margaret C. Brittingham and
Herbert T. Hendrickson (terms expire 1998).

Membership dues per calendar year are: Active, $21.00; Student, $15.00; Family, $25.00; Sus-
taining, $30.00; Life memberships $500 (payable in four installments).

The W11.SON BUEt.ETIN is sent to all members not in arrears for dues.

The Josseeyn Van Tyne Memoihai, Libkary

The Josselyn Van Tyne Memorial Library of the Wilson Ornithological Society, housed in the
University of Michigan Museum of Zoology, was established in concurrence with the University
of Michigan in 1930. Until 1947 the Library was maintained entirely by gifts and bequests of
books, reprints, and ornithological magazines from members and friends of the Society. Two mem-
bers have generously established a fund for the purchase ol new books; members and friends are
invited to maintain the fund by regular contribution, thus making available to all Society members
the more important new books on ornithology and related subjects. The fund will be administered
by the Library Committee, which will be happy to receive suggestions on the choice of new books
to be added to the Library. William A. Lunk, University Museums, University of Michigan, is
Chairman of the Committee. The Library currently receives 195 periodicals as gifts and in ex-
change for The Wilson Bidlelin. With the usual exception of rare books, any item in the Library
may be borrowed by members of the Society and will be sent prepaid (by the University of
Michigan) to any address in the United States, its possessions, or Canada. Return postage is paid
by the borrower. In(]uiries and requests by borrowers, as well as gifts of books, pamphlets, reprints,
and magazines, should be addressed to: The Josselyn Van Tyne Memorial Librarv', University of
Michigan Museum of Zoology, Ann Arbor, Michigan 48109. Contributions to the New Book Fund
should be sent to the Treasurer (small sums in stamps are acceptable).

'Die Wh.son Buei.etin
(ISSN 0043-5643)

HIE II Util ,1 .E I IN (ISSN (K)4.t-.'>64;5) is [luhlislu'rj (jiiaricriy in March. June. Septcnilx'r. ami Deccniln'r by the Wilson
Dniilhologiral Society. f!IO Ea.sl lOtli Street. Diwrence. K.S WtO'W-HfWT. The subscription price, both in thi' United States and
elsewhere, is $40.00 per year. Second-class postaj;e paid at I jiwrence, KS. I’OSTM AS'l'Elt: Send address changes to THE WllisON
nUl.I.EriN. I’.O. Box IB07. Ijiwrence. KS 66(H4-8H07.

Back issues or single I'opies are available for $12. IK' etich. Most back issues of tbe liuUelin an- available and may be oniered
from tbe Treasurer. Special prices will 1m- (pioted for rpiantity orders.

.■\ll articles and cnmmnnications for publications. b(H>ks and publications for revit-ws should be addresst-d to tin- Editor. Ex-
chang.-s should he addressed to The josselyn Van Tyne Memorial l.ibrarv. Museum of Zoologv. Ann Arbor. Michigan 48100.

Suhsi riptions. changes of adilrcss and claims for undelivered copies should be sent to the DSNA. I’.O. Box 1807. Ijiwrence.
KS W)0H-8807. I’hone: (01.4) IH.4-1221; FAX: (01.4) (H.4-1274.

© Coftyright 1996 by ihe WiLson Ornithological Society
Printed by Allen Pre.ss, Inc., Lawrence, Kansas 6604-4, U.S.A.

@ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

MCZ

LIBRARY

MaR 1 5 1996

HARVARD

UNIVERSITY

Frontispiece. The Chiribiquete Emerald, Chlorostilhon olivuresi. Male (above right) perched
on branch of Tepiiicmthus savannarunr, female (below left) feeding from flowers of Deccigon-
ocarpus cornutiim, on a mesa of the Siena de Chiribiquete, southeastern Colombia.

THE WILSON BULLETIN

A QUARTERLY MAGAZINE OE ORNITHOLOGY
Published by the Wilson Ornithological Society

VoL. 108, No. I March 1996 Pages 1-204

Wilson Bull., 108(1), 1996, pp. 1-27

A NEW SPECIES OF EMERALD HUMMINGBIRD
(TROCHILIDAE, CHLOROSTILBON) FROM THE
SIERRA DE CHIRIBIQUETE, SOUTHEASTERN
COLOMBIA, WITH A REVIEW OF THE
C. MELLISUGUS COMPLEX

E Gary Stiles'

Abstract. — The Chiribiquete Emerald {Chlorostilbon olivaresi sp. nov.) is described
from the Sierra de Chiribiquete, an isolated range of table-top mountains rising from the
flat Amazonian lowlands of the Departments of Guaviare and Caqueta, SE Colombia. This
hummingbird is a common inhabitant of the edaphic scrub and adjacent forests of the middle
and upper levels of the Sierra, but evidently does not occur in the suiTOunding lowlands. In
its morphology, the new species shows closer affinities with C. gibsoni of the Magdalena
Valley than with the adjacent cis-Andean populations of C. mellisiigus, but it is much larger
than all related forms. The Chiribiquete Emerald probably originated through the dispersal
of gibsoni-lypc birds to the Sierra during a dry period of the Pliocene or early Pleistocene,
perhaps in conjunction with hybridization with the local form of inelli.sugus\ large body size
probably evolved subsequently in the population as a response to its peculiar, insular habitat.
Variation in the Chlorostilbon mellisiigus complex in NW South America is described and
analyzed, and I conclude that the various forms are best treated as comprising a single
superspecies; melanorhynchus (including pumilus) of western Colombia and western Ec-
uador is sufficiently distinct from the adjacent assimilis and gibsoni. as well as from the
eastern forms of C. mellisugus to deserve (allo)species rank, and I suggest for it the English
name of West Andean Emerald. I recommend recognition of the following allospecies (from
north to south): auriceps, forficatus, canivetii, assimilis, melanorhynchus. gib.soni, olivaresi,
and mellisugus. Received 16 Feb. 1995, accepted 10 June 1995.

Abstracto. — Se describe Chlorostilbon olivaresi sp. nov. de la Sierra de Chiribiquete,
una serie aislada de mesetas en la planicie amazonica de los Departamentos del Guaviare y
del Caqueta, SE Colombia. E.ste colibri es comtin en las sabanas casmofitas y los bosques
aledanas de la parte media y superior de la Sierra, pero evidentemente no ocurre en los
bo.sques basales circundantes. Por sus caracteres morfologicos, C. olivaresi probablemente
tiene mas afinidad con C. gibsoni del Valle del Magdalena, que con las formas cisandinas

' Institute de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, Bogota, Colombia.

I

2

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

vecinas de C. mellisugiis, pero es un ave mucho mas grande que cualquier forma de estas
especies. La poblacion de C. olivaresi posiblemente origino a traves de la dispersion de
un(os) individuo(s) del tipo gibsoni a la Sierra durante un pen'odo seco del Plioceno o
Pleistoceno temprano, tal vez incluyendo hibridizacion con la forma local de mellisugus',
subsecuentemente, el tamano corporal grande evoluciono en esta poblacion en respuesta a
su habitat peculiar y aislado. Se describen y se analizan los patrones de variacion en el
complejo de Chlorostilbon mellisugus en el NO de Suramerica, y se concluye que melan-
orhynchus (incluyendo a pumilus) del O de Colombia y Ecuador es suficientemente distinta
de gibsoni y assimilis, las formas adyacentes, como para ser considerado como una
(alo)especie. Recomiendo que el complejo de C. mellisugus se considere como una sola
superespecie, constituida por las aloespecies (de norte a sur) auriceps, forficatus, canivetii,
assimilis, melanorhynchus, olivaresi, y mellisugus.

The small hummingbirds of the genus Chlorostilbon are widespread in
the Neotropics from central Mexico to northern Argentina and the West
Indies. The genus is quite uniform in coloration: males of all species have
flashing green underparts, while females are pale gray below, nearly al-
ways with a distinctive facial pattern and dusky malar auricular area bor-
dered above by a white postocular stripe. Several species are easily dis-
tinguished by the distinctive form and color of the rectrices of the males
(e.g., poortmanni, alice, stenura), very bronzy or coppery coloration {rus-
satus), or extensively red bills {aureoventris). However, a number of
mostly or entirely allopatric forms in which the males have blue, more
or less forked tails, have long been a source of taxonomic confusion.
These forms, which collectively might be called the "'mellisugus com-
plex”, cover most of the range of the genus from Mexico to Bolivia and
eastern Brazil. Variation among them involves bill color and, in males,
the presence or absence of a glittering crown, the depth of the tail fork,
the shape of the outer rectrices, and the color of the underparts and, in
females, the shape of the rectrices and the amount of gray in the outer
rectrices (cf Zimmer 1950). Geographic variation in these characters pre-
sents something of a mosaic pattern, with similar forms often separated
by others of rather different appearance, making the determination of
species limits controversial. The discovery of a new form in the C. mel-
lisugus complex, here described as a new (allo)species, makes it desirable
to review the patterns of geographic variation in this complex in north-
western South America, and to reevaluate the relationships among the
various members of the complex as a whole in the light of recent studies.

The avifauna of Colombia has received at least as much attention from
ornithologists, both native and foreign, as has that of any large South Amer-
ican country. The birds of Colombia have been comprehensively mono-
graphed no less than three times (Chapman 1917, Meyer de Schauensee
1948-1952, Hilty and Brown 1986). Nevertheless, many parts of Colombia

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

3

have been visited only briefly or sporadically by ornithologists, and a num-
ber of areas remain ornithologically unexplored. Until very recently, one
such area was the Sieira de Chiribiquete, a small, isolated mountain range
in the Departments of Guaviare and Caqueta. Because of its topographic
uniqueness and pristine character, this area had been set aside as the Parque
Nacional Natural Chiribiquete in September 1989, but aside from a visit
by botanist Richard E. Schultes in 1943-1944, the Sierra remained biolog-
ically unexplored until the present decade. The only previous ornithological
collections from this entire region of Colombia were made by H. Romero
for the Institute de Ciencias Naturales of the Universidad Nacional de
Colombia at Araracuara, over 100 km S of the Sierra de Chiribiquete on
the Rfo Caqueta, in August-September 1977.

Between December 1990 and December 1992, three expeditions to the
Sierra de Chiribiquete were organized by the Agencia Espanola de Coop-
eracion Intemacional, the Instituto de Ciencias Naturales, and the Institute
de Recursos Naturales Renovables (INDERENA) of the Colombian govern-
ment. The first two expeditions, in December 1990 and August 1991, were
devoted to botanical, archaeological, and geological studies. During the first
expedition, the Colombian botanists P. Palacios and P. Eranco obtained a
specimen of an unusual hummingbird (the only bird specimen taken) that I
was unable to identify: it appeared to be the male of an undescribed form
of Chlorostilbon, notable for its very large size, but its rather rough prepa-
ration made precise comparisons with other material difficult.

I obtained additional material of this hummingbird during the third
expedition (18 November-2 December 1992). With J. L. Telleria and M.
Dfaz of the Universidad Complutense of Madrid, I observed and collected
birds in the vicinity of the expedition’s base camp in the northern part of
the Sierra at 0°56'N, 72°42'W at the site called “el Valle de los Menhires”
(Valley of the Monoliths) by Estrada and Euertes (1993). Seven specimens
(3 males, 4 females) of the Chlorostilbon were collected and eight others
measured and released; all specimens are housed in the collection of the
Instituto de Ciencias Naturales. On the basis of this sample, I here de-
scribe this hummingbird as

Chlorostilbon olivaresi, sp. nov.

CHIRIBIQUETE EMERALD

HOLOTYPE. — Adult male, no. 31266 of the ornithological collection of the In.stitiito de
Ciencias Naturales (original number FGS 2941), collected on 24 November 1992 in the
Valle de los Menhires, elev. 570m, Sierra de Chiribiquete, Depto. del Caqueta, Colombia
(0°56'N, 72°42'W) by E G. Stiles, J. L. Tellen'a, and M. Diaz.

PARATYPES. — One adult male (ICN 31252, orig. no. FGS 2927) and one sub-adult male
(ICN 31253, orig. no. FGS 2928) taken on 21 Nov. 1992 and three adult females (ICN

4

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

31244, 31245, and 31254, orig. nos. FGS 2915, 2916, 2929) and one subadult female (ICN
31243, orig. no. FGS 2914), taken on 19 and 24 Nov. 1992, all at the type locality.

DIAGNOSIS. — Clearly a member of the genus Chlorostilbon because of the entirely
steel-blue, forked tail of the males, and the distinct dusky face and cheeks and white post-
ocular stripe and uniform gray underparts of the females. Size significantly larger than any
other South American form of Chlorostilbon (exposed culmen always >18 mm, vs 17 mm
or less in all other forms); total culmen >21 mm vs <19.5 mm; wing chord >48 mm in
males, >47 mm in females, vs <47.5 mm and <46.5 mm respectively, except for the
isolated C. melUsugus duidae, which may reach 49.5 mm in both sexes; and weight usually
>3.5 g, vs <3 g). Differs from all forms of C. melUsugus in having the basal half of the
mandible red in males, or with a trace of red in females; in the dull, non-iridescent crown
(males) and extensively grey bases of the lateral rectrices (females). In these characters
resembles C. gibsoni from north and west of the Andes, but differs from all forms of gibsoni
in having the tail much more shallowly forked, with the lateral rectrices more truncate
(males) to rounded (females) at the tip, and the breast and throat much more strongly bluish
in color (males). In the latter feature is most like C. aureoventris of southeastern South
America, but always lacks red on the upper mandible (which is extensively red in the latter);
females of aureoventris also lack gray on the lateral rectrices.

ETYMOLOGY. — I take pleasure in naming this species for Fr. Antonio Olivares in honor
of his many pioneering contributions to Colombian ornithology and his indefatigable labor
in building the bird collection of the Institute de Ciencias Naturales. The English name
refers to the isolated mountain range which includes the type locality and evidently encom-
passes the entire distribution of the species.

DESCRIPTION OF HOLOTYPE.— (Color nomenclature follows Smithe 1975, 1981).
Crown, nape, back, and rump metallic green, near 162, Shamrock Green; a few feathers of
the nasal area of the anterior forehead more brilliant golden green (near 158, Chartreuse);
a small grayish-white postocular spot. Upper tail-coverts more bluish green (163, Emerald
green); tail dark steely blue, nearest 90, Blue-black. Facial area and sides of neck brilliant
Emerald Green, with Chartreuse reflections, passing abruptly to brilliant blue-green (between
164, Cyan, and 65, Turquoise Blue, depending on viewing angle) over the entire throat and
upper breast, this passing to brilliant green (near 62, Spectrum Green) on the lower breast
and belly; lower tail coverts more bluish, near Emerald Green; a small tuft of downy white
feathers on the thigh. Remiges blackish with faint bluish gloss (near 73, Indigo). Basal 3/4
or more of lower mandible red (between 13, Geranium Pink, and 10, Ruby); rest of bill,
legs, and feet black. Exposed culmen 19.5 mm, total culmen 21.8 mm, wing chord 49.6
mm, tail length 25.6 mm, tarsus 4.4 mm, weight 3.4 g. Adult male, left testis 2.4 X 2.3
mm, no fat; tiny diptera in stomach.

DESCRIPTION OF ADULT FEMALE (based on ICN 31244). — Upperparts somewhat
more bronzy-green than in the adult male, nearer 60 (Parrot Green), somewhat duller on the
crown; the more worn leathers of the back with more bluish tips (near Emerald Green),
producing a slightly scaly effect; longest upper tail coverts and most of central rectrices
more bluish (between Emerald Green and Cyan but darker); the second and third pairs of
rectrices similar, but shading to dark blue (near 173, Indigo Blue) at the tips. The outer two
rectrices are extensively pale gray basally (near 86, Light Neutral Gray), with the medial
portion Indigo Blue; the outermost rectrix is tipped broadly, the fourth and third progres-
sively more nairowly, with Light Neutral Gray. The malar and auricular areas are dark sooty
gray, tinged with dull bionze; a white stripe extends trom the eye back over the auriculars.
The underparts are pale gray, slightly tinged brownish (near 79, Glaucous, but paler), av-
eraging palest on the thioat and darkest on the upper breast. The basal 1/4 of the lower
mandible is tinged with dark red, the rest of the bill and feet are black.

Sliles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

5

PLUMAGE VARIATION IN THE TYPE SERIES.— Variation among the adults of the
type series is slight in both sexes, reflecting chiefly feather wear (the degree of scaliness
in the dorsal plumage, the degree to which the crown is dull and soiled in females); one
female (ICN 31245) has the tip of the first rectrix tinged with blue. The subadult male (ICN
31253) is approaching adult plumage, with the brilliant feathers below colored as in the
adults but more scattered over a dull, dusky green Juvenal plumage; traces of a dusky mask
and whitish postocular and malar stripes remain; the tail is less forked than in the adult
males, and the outer rectrix is very narrowly tipped with dull gray. The subadult female
(ICN 31243) is a more uniform, bronzy green dorsally with no bluish tinge to the upper tail
coverts or central rectrices, and a darker, duller gray below.

ADDITIONAL SPECIMENS EXAMINED.— Between May and August 1993, Diego Silva
and Tomas Walschburger of the Eundacion Puerto Rastrojo, conducted studies of birds in the
area of the Rio Mesay, just south of the border of the National Park (0°4'N, 72°26'W), some
85 km SSE of our study area. They observed and collected birds along a transect from the
river (ca 230 m elevation) to the top of a low mesa (ca 360 m), an isolated southern outlier
of the main Sierra. Four specimens of C. olivaresi were taken in the scrub atop the mesa in
July 1993 and were available for examination. These specimens agree perfectly in measure-
ments and coloration with the type series, allowing for differences in plumage wear: in par-
ticular, the golden-green reflections of the facial area and sides of the neck of the males are
stronger, increasing the contrast with the bluish green of the throat. The adult male specimen
from the 1990 expedition does not differ from the males of the type series. Thus, the characters
of C. olivaresi appear to be uniform over most or all of the Sierra de Chiribiquete.

THE SIERRA DE CHIRIBIQUETE AND THE
ECOLOGY OF C. OLIVARESI

The Sierra de Chiribiquete consists of a series of sandstone mesas and
buttes some 125 km long and 30 km wide, extending in an arc that curves
from NW-SE in the north to NE-SW in the south, between the latitudes of
1°20'N and 0°20'N, centered along the line of 73°W longitude (Eig. 1). The
mesas rise abruptly from the surrounding flat lowlands to heights of 800-
900 m in the north and 600 m in the south, often presenting several levels
or terraces separated by vertical cliffs. Many of the larger mesas are riven
by spectacular chasms or cracks; the flat upper surfaces are drained by
streams flowing through vertical cracks that emerge as waterfalls at the edges
of the mesas. The thin, sandy soil of the table-tops supports a scrubby veg-
etation interspersed with areas of naked rock: the stature of the vegetation
reflects the depth of the soil (or the absence thereof) at any given point.
Eurther details of the topography, vegetation, and geology of the Sieira de
Chiribiquete are given by Estrada and Euertes (1993). Extending south of
the main part of the Sierra are a series of progressively lower mesas (ca
300-350 m) which reach the Rfo Caqueta at Araracuara (ca 0°30'S).

The base camp of our expedition was situated on the flat middle level
of a large (ca 3 km long) mesa at an elevation of 570 m. To the north
and east, a line of cliffs rises abruptly to the top level of the mesa (ca
700 m); to the west, the mesa is bounded by a steep-sided canyon some

6

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Fig. 1. Left: location of the Sierra de Chiribiquete (inset) in relation to the major moun-
tain ranges (stippled) of Colombia, and the major rivers of SE Colombia and adjacent areas
(Roman numerals). The rivers are I. Orinoco; II. Guaviare; III. Vaupes; IV. Apaporis; V.
Caqueta; VI. Putumayo; and VII. Amazon. Other localities indicated are a. Magdalena Valley
and b. Sierra de la Macarena. Right: The Sierra de Chiribiquete and adjacent areas showing
major rivers and the collecting localities for Chloro.stilbon olivare.si. The 300, 400, 500, and
600 m contour lines are indicated. Localities: 1. Type locality (Valle de los Menhires); 2.
Collecting site of December 1990 specimen (Valle de las Abejas); 3. Rfo Mesay site. Re-
drawn from Estrada and Fuertes 1993, in part.

Fig. 2. Landforms and vegetation at the type locality. A. General topography, showing
the butte and our campsite, with adjacent stunted forest and surrounding Bonnetia scrub of
the middle level of the mesa; taller forest at the base of the fringing cliffs in foreground
and other me.sas just visible in the background. B. Aspect of Bonnetia scrub, the most
important habitat of C. olivare.si. Note areas of naked rock and patches of low vegetation
(mainly Navia garcia-harrigae) in foreground, the stiff, coriaceou.s-succulent leaves of many
shrubs (Grajfenriedia sp. in foreground, Cln.sia chirihiqueten.si.s at center. The low shrub L
of center is Decagonocarpti.s cornutu.s, the most important nectar source of C. oHvaresv,
taller shrubbery in background (with large white flowers) is Bonnetia niartiana.

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

7

8

THE WILSON BULLETIN • Vol. 108, No. 1. March 1996

100-150 m deep that slopes steeply down towards the surrounding low-
lands. The base camp was located some 100 m SE of the base of an
isolated butte that rises precipitously to a height of ca 75 m.

The vegetation around the base camp consisted of open scrub domi-
nated by the shrub Bonnetia martiana (Theaceae); patches of dense scrub,
in which the taller shrubs attained a height of 2-3 m, alternated with
patches of low, savannalike vegetation and areas of naked rock (Fig. 2).
The principal shrubs of the area had thick, coriaceous leaves and twigs
suggesting adaptation to resist drought, doubtless reflecting the shallow,
sandy soil with low nutrients and minimal water retention capacity. The
most common shrub species included, besides Bonnetia, Clusia chiribi-
qiietensis (Guttiferae), Tepuianthus savannensis (Tepuianthaceae), Graf-
fenriedia sp. (Melastomataceae), and Decagonocarpus cornutus (Ruta-
ceae). Between the shrubs, and in slight depressions where a thin layer
of sandy soil accumulated, occurred a low herbaceous vegetation domi-
nated by Xyridaceae, Eriocaulaceae, Burmanniaceae, Cyperaceae, and
Vellozia phantasmagorica (Velloziaceae) with only occasional grasses
(Gramineae); where drainage was impeded, pools formed after every rain
in which Utricnlaria spp. (Lentibulariaceae) were abundant. Areas of bare
rock were colonized by a terrestrial bromeliad, Navia garcia-barrigae. A
low, dense, tangled forest (canopy height 3-5 m) dominated by Clusia
spp. and Licania sp. grew around the base of the butte.

The level of the mesa sloped very gently towards the fringing cliffs to
the north and east, at the bases of which grew a much taller forest (canopy
height 20-25 m) with numerous palms and a relatively well-developed
understory; in the canyon bottom grew a forest of similar canopy height
but with few palms and notably low tree species diversity and a few large-
leaved monocots (Heliconia, Calathea, Costus, Phenakospennum) in the
understory. The northern part of the mesa was drained by a stream along
which grew an extremely dense, tall (canopy ca 5 m) stand of Bonnetia
whose tangled aspect suggested that of a young mangrove swamp. A
general inventory of the avifauna and its biogeographical affinities (Stiles
et al. 1995), and a detailed analysis of mist-net captures in relation to
vegetation characteristics (Diaz et al. 1996) will appear elsewhere.

In the study area, Chlorostilbon olivaresi was fairly common in the
open Bonnetia scrub of the mesa, where 12 of the 15 individuals were
captured; in fact, it was the bird most frequently captured in mist-nets in
the open areas of the mesa. In this habitat, both sexes were observed
visiting only the red-orange flowers of Decagonocarpus cornutus (see
Frontispiece) and were frequently noted gleaning small arthropods from
the foliage and flowers of Bonnetia and other shrubs and flycatching at
breaks in the vegetation, especially in late afternoon. Pollen samples taken

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

9

from the beaks of eight mist-netted birds with transparent Scotch tape
contained only pollen of Decagonocarpus, usually <10 grains (5 cases),
three grains of pollen of Decagonocarpus and one of Bonnetia (one biid),
or no pollen at all (2 birds). The visit to Bonnetia was almost certainly
to obtain insects such as thrips which were often present on the large,
open, and fragrant camellialike flowers of this species (which do not
produce nectar). Stomach contents of collected individuals contained tiny
insects, mainly flies, but a few thrips and/or microhymenopterans were
present in at least two stomachs. The only other hummingbirds to occur
regularly in the Bonnetia scrub were the Versicolored Emeiald {Amazilia
versicolor), which also visited Decagonocarpus but was much scarcer,
and the Black-throated Mango (Anthracothorax nigricollis), which ap-
peared to be more insectivorous and was seen to visit only the flowers
of a small tree of the Bombacaceae in the low forest at the base of the
butte. We never saw any interactions between any of these species.

The only other habitat in which we regularly recorded C. olivaresi was
the forest at the base of the fringing cliffs at the north end of the mesa.
Here, small numbers occurred in the understory and at gaps; three birds
were captured, and on several occasions a female was observed visiting
the flowers of a small understory tree of the Violaceae. In several days
of observation and netting, we never encountered C. olivaresi in the forest
of the canyon bottom. In visits to other areas at the base of the Sierra,
including mist-netting, other members of the expedition (A. Repizzo, B.
Ortiz) never encountered C. olivaresi in the forests at lower elevations.
During their work at the Rfo Mesay site, Silva and Walschburger only
recorded C. olivaresi in the Bonnetia scrub on the top of the mesa or in
the adjacent low forest (340-360 m), never lower or closer to the river.
The 4—6 months difference between the dates of our observations and
those of Silva and Walschburger would appear to preclude the possibility
of extensive seasonal movements (e.g., into the surrounding lowland for-
ests) by this hummingbird. Thus, C. olivaresi may be restricted to the
scrubby vegetation and adjacent forests on the mesas of the Siena, and
it is quite probably absent from the surrounding forested lowlands. It
evidently occurs widely in the Sierra de Chiribiquete, given the distance
between the Rfo Mesay site and the type locality; the 1990 specimen was
taken some 20 km south of the type locality, also in Bonnetia scrub at a
slightly lower elevation (ca 420 m). The southern limit of olivaresi re-
mains to be determined but is evidently somewhere between the Rfo Me-
say and the Rfo Caqueta, since it was not taken during intensive collecting
by H. Romero at Araracuara in similar Bonnetia scrub.

Of the specimens of C. olivaresi we collected, one adult male and two
females (as well as both of the subadults) had the gonads small and

10

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

undeveloped; one male and one female had moderately enlarged gonads.
Among the birds measured and released, one female had a developing
brood patch. The two subadult specimens were at least several months
old, as they had lost their bill corrugations. Taken together, these data
indicate that C. olivaresi was at the start of its breeding season during
our observations, although courtship displays were not observed, and we
found no active nests. None of the individuals captured were molting,
and most were in slightly to moderately worn plumage, consistent with
the inference that the breeding season was beginning (cf Stiles 1985b).
Although meteorological data for the Sierra de Chiribiquete do not exist,
extrapolation from the data for other sites in Colombian Amazonia sug-
gest that our expedition took place at the very beginning of the dry season
(Estrada and Fuertes 1993). Flowering of at least Decagonocarpus, evi-
dently the most important nectar source at this time for C. olivaresi, was
definitely increasing during our observations (many flower buds, few de-
veloping fruits). Fruiting of several species was also increasing or about
to begin in late November 1992, and a number of other bird species were
evidently just starting gonadal maturation, suggesting that the dry season
might include the main breeding period for a considerable segment of the
Chiribiquete avifauna (Stiles et al. 1995).

The four Rfo Mesay specimens, taken in July 1993, were for the most
part in very fresh plumage, and one female was in heavy molt; neither
of the two with gonad data were in breeding condition. If the seasonality
of the two sites is similar, this suggests that the breeding season of C.
olivaresi falls between late November or December and perhaps May.
Molt in the population extends from perhaps May through at least July,
as both males and one female ot the Rfo Mesay specimens had completed
molt by July, in hummingbirds it is not unusual for males to molt a month
or so ahead of females (cf Stiles 1985b).

Like other members of its genus, C. olivaresi is a rather quiet humming-
bird under most circumstances. TJie only vocalization heard was a shaip,
dry, scratchy cht , similar to the calls of other Chlorostilbon but somewhat
louder, and given by birds foraging at flowers. We never heard it sing.
Compared to others of the genus 1 have observed, C. olivaresi is less nervous
and flighty at flowers and shows much less of the incessant rapid flicking
or pumping of the tail while foraging (cf Stiles and Skutch 1989).

PATTERNS OF VARIATION IN THE C. MELLISUGUS COMPLEX IN
NORTHWESTERN SOUTH AMERICA

No fewer than 10 forms in this complex are recognized by most authors
frorn Colombia and adjacent areas of northwestern South America (see
able , Fig. 3). Because C. olivaresi appears to be a member of this

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

Table 1

Currently accepted Subspecies of the Chlorostilbon mellisugus Complex in
Northwestern south America, with their Distributions and Diagnostic characters'*

Subspecies

Distribution

Characters'

puniHus

W Colombia from Pacific slope
of W Andes E to W edge of
Magdalena Valley

Bill black. M: tail moderately
forked; glittering crown;
breast with little or no blue.

F: trace of grey in outer rec-
trices, extensive green in cen-
tral rectrices.

inelanorhynchus^

Extreme SW Colombia (Narino)
and W Ecuador

Like pumilus but averaging larg-
er.

gibsoni

Upper and middle Magdalena
Valley Colombia

Lower mandible largely red. M;
tail very deeply forked, outer
rectrices attenuate; breast
green; crown dull. F: outer
rectrices with extensive grey
bases, broad whitish tips.

chrysogaster

N lowlands of Colombia E to

Like gibsoni but M: tail even

Santa Marta; W of Lago de
Maracaibo, Venezuela-Colom-
bia

more deeply forked, fore-
crown glittering, throat and
breast tinged blue; F: grey of
outer rectrices darker, less ex-
tensive.

nitens^'

Arid N coast of extreme NE Co-
lombia and NW Venezuela

Like chrysogaster but tail of M
slightly less deeply forked.

caribaeus^

Most N coastal region of Vene-
zuela, S to Orinoco region
and Llanos of Colombia

Bill black. M; glittering crown;
breast with trace of blue; tail
shallowly forked. F; at most a
trace of dusky on lateral rec-
trices, otherwise blue.

napensis^

SE Colombia S of Llanos; E Ec-
uador; adjacent NE Peru

Like caribaeus but averaging
larger. M: tail very shallowly
forked, nearly truncate, breast
strongly tinged blue. F: exten-
sive green flecking on sides.

phaeopygos

E Peru to NE Bolivia, incl. adja-
cent Brazil

Like napensis but M; belly dull-
er, darker, less contrast with
blue of breast; F: much less
green flecking below. Also
averages larger, especially tail.

subfurcatus

S + SE Venezuela E to Guayana,
and adjacent NW Brazil

Like caribeus but M: tail less
forked, breast more bluish.
Averages slightly larger.

duidae

Mt. Duida, SE Venezuela

Like siibfuraliis but decidedly
larger; M: tail longer, more
forked; breast less bluish.

“ Characters and distributions from Meyer de Schauensec (1964, 1966); Meyer dc Schauensee and Phelps (1978); Zimmer
(1950). and Zimmer and Phelps (1952).

May not be distinct from pumihts.

‘ May not be distinct from clir\sof;asler.

Includes narttts from the Orinoco region.

'Often lumped with phaeopyf-os.

' M = male. F = female.

12

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

complex, I decided that to evaluate properly its status and affinities a
broader analysis of the patterns of variation in the complex over north-
western South America would be required.

Previous studies. — The first comprehensive study of the mellisugus
complex was that of Zimmer (1950) who concluded that the mosaic nature
of the patterns of geographic variation, plus the allopatric distributions of
most forms, justified considering all forms of the complex as subspecies
of a single, variable species mellisugus. However, several points of con-
fusion exist in his analysis. Evidently, Zimmer started by considering two
forms from northern South America, nitens and chrysogaster, as members
of different subspecies groups and later concluded that they might not be
separable at all, thus favoring lumping the “eastern” and “western”
groups.

In his initial study of the birds of Colombia, Meyer de Schauensee
(1948-1952) had considered all Colombian forms of the complex as sub-
species of gibsoni, but he later (1960, 1966) noted that two groups of
races could be distinguished on the basis of bill color: those with the
lower mandible largely red {gibsoni, chrysogaster, and nitens) and those
with all-black bills (all other forms). He noted that the red-billed forms
were also those with the most deeply forked tails and attenuated outer
rectrices in the males, and with the most gray in the rectrices of the
females. He noted apparent sympatry of a red-billed (gibsoni) and black-
billed ipumilus) form without evident intergradation at two localities on
the western edge of the Magdalena Valley in Colombia and concluded
that two species should be recognized in Colombia: the red-billed gibsoni
(including chrysogaster and nitens) and the black-billed mellisugus (in-
cluding pumilus, melanorhynchus, caribaeus, and phaeopygus {=napen-
sis).

Wetmore (1968) adopted Meyer de Schauensee’s character of bill color
as a criterion for distinguishing species in the complex when he separated
the black-billed assimilis of southwestern Costa Rica and W Panama from
the red-billed canivetii of farther north. The arrangement of Meyer de
Schauensee (1960, 1966) has been followed by recent authors for the
South American forms of the complex (Meyer de Schauensee and Phelps

Eig. 3. Tails of members of the C. melli.sugu.s complex from Colombia, drawn to the
same scale. Left: tails of adult males, showing form of rectrices; in all, rectrices are uniform
steely blue-black. Right: tails of adult females, showing form, pattern, and colors of the
rectrices. Solid; steely blue-black; heavy stipple: green; light stipple: grey; open: white. The
forms illustrated are: a. chrysogcLster, b. gibsoni', c. olivaresi', d. napensis', e. caribaeus-, f.
pumilus.

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

13

14

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

1978, Hilly and Brown 1986, Sibley and Monroe 1990). However, be-
cause the characters of olivaresi, as detailed above, appear to be in part
a mosaic of features of gibsoni and mellisiigus as cuirently recognized,
this arrangement requires a detailed reevaluation. For this analysis, I con-
sulted specimens of the mellisiigus complex available in the following
museums (numbers of specimens in parentheses): the Institute de Ciencias
Naturales, Universidad Nacional de Colombia (69); the Universidad de
La Salle, Bogota (12); the Unidad de Investigacion “Federico Medem”
(UNIFEM), INDERENA (6), Bogota; and the Colegio de San Jose, Me-
dellin (8). Eor each specimen, I measured the exposed culmen, total cul-
men, wing chord, and tail length to the nearest 0.1 mm with dial calipers.
Eor males only, I measured the depth of the tail fork as the difference
between the lengths of the first and fifth rectrices. In addition, I took
detailed notes on color characters of each form and made detailed draw-
ings of the rectrices of representative individuals of both sexes of all
forms (Eig. 4). Eield work in various parts of Colombia has also given
me the opportunity to capture and measure an additional 18 individuals
of various forms of the mellisiigus complex. Eor these birds, I made the
preceding measurements except depth of tail fork, also with dial calipers,
and to the same level of accuracy; I also weighed all birds to the nearest
0.1 g with a 10 g Pesola spring balance. Because in previous studies I
had found that my field measurements did not differ from those taken in
the museum and that my measurements of birds in the field agree closely
with remeasurements of the same birds prepared as museum specimens
(Stiles 1985a, 1995), I have included both types of measurements in the
quantitative analyses below. In addition, M. Marin kindly measured 10
specimens of napensis and pheaopygus in the collection of the Museum
of Natural Science, Louisiana State Univ. Having worked with and mea-
sured birds previously with Marin, I am confident that his measurements
are comparable with mine and have incorporated them into the analyses.

In all, measurements from 1 1 1 specimens of the following forms were
used in the analyses: olivaresi, pumilus, gibsoni, chrysogaster, caribaeus,
and napensis (Table 2). Of the remaining forms in Table 1, none occurs
in areas adjacent to olivaresi and is likely to bear upon its status; I was
able to examine only two specimens of nitens, and none of melanorhyn-
clnis, subfurcatus, duidae, or phaeopygus. Eor the characters of these
forms (Table 1), 1 have relied upon the descriptions of Zimmer (1950)
and Zimmer and Phelps ( 1952). For the six forms mentioned above, mea-
surements were analyzed by one-way analysis of variance (ANOVA) for
each sex separately. Where a significant result was obtained in the ANO-
VA, Tukey a posteriori tests were performed to determine which forms
dilfered significantly with respect to the measurement in question (Zar

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

15

s purmliLS
Wh mtlanorhy'nchus
^ gibsoni

chryso^cLster

nitens
caribae^us
subfurcatws
A dmdoL
• oUvatzsi
napcnsis
'////, phaeopggos

Eig. 4. Distributions of the members of the C. mellisugus complex in northwestern South
America. Note that limits of some forms are imperfectly known, as are possible zones of
contact; especially east of the Andes, distributions are probably more continuous than shown,
but collecting localities are often widely scattered.

1988). For the measurement of depth of tail fork, I derived a measure of
“relative tail fork” by dividing the difference r5-rl by the tail length for
each specimen, to control for differences in absolute size between the
various forms.

The most striking result of the analyses of variance of the measure-
ments (Table 3) is the clear-cut separation of olivaresi (at a significance
level of P < 0.001) from all other forms of the Chlorostilhon mellisugus
complex, with respect to bill length (both exposed and total culmen), wing

16

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

•7

d

NO

CO

S

00

o

00

CO

in

CN

NO

CN

(N

(N

CN

(N

CO

CN

CN

«3

B

d

d

d

d

d

d

d

d

d

d

<

-g

+ !

+1

+ 1

+1

+1

+1

+1

+1

+1

+1

S

0

On

00

CO

r-

00

o

<

CQ

'O

iT)

r-;

q

q

in

q

q

q

CN

(N

CN

(N

CN

(N

CN

(N

d

d

z

in

u

On

<

ON

NO

ro

Tt

(N

O

CO

»— «

o

c

O

o

o

O

<N

o

O

V ^

> t

d

d

d

d

d

d

d

D

Z

-- ,o
— ^

+1

+1

+1

+1

+1

+1

+1

<

“v ^

2 5

(N

CN

On

CO

'tt

<

NO

o

On

00

5

5

<N

CN

O

d

d

d

d

d

d

d

o

j

o

U

(N

(N

NO

in

r-

O

On

CO

00

O

ON

r-

q

q

(N

q

NO

NO

00

ON

q

r-

in

00

S

o

a:

Uh

JZ

■« ^

+1

+ 1

+1

+ 1

+1

+1

d

+1

d

+1

d

+1

d

+1

+ 1

d

+1

d
+ 1

d

+1

X

o

00

00

o

CN

m

r-

NO

in

CO

u

r--

q

ro

00

00

q

CN

q

q

q

q

-J

cu

Tt

rn

d

d

d

d

d

d

—

— 1

d

d

CN

(N

(N

(N

CN

<N

<N

CN

CN

CN

CN

CN

CN

CN

o

u

NO

r— .

On

On

NO

CN

CO

00

CO

o

•o

o

::d

o

o^

r-

O

NO

CO

On

q

1—

CO

NO

NO

d

d

d

d

d

d

d

d

d

d

d

<N

acn
■S o

+1

+1

+1

+1

+ 1

+1

+1

+1

+1

+ i

+1

+ 1

+1

+1

W

d

»-g

(N

00

o

in

NO

00

ON

00

r-

r-

o

-J

O

O)

O)

q

q

00

q

q

00

q

q

oa

>

d

Tt

d

d

CN

d

d

d

d

CN

d

CN

On

00

<

o

'tt

H

oa

to

rs

(N

00

NO

00

On

NO

CO

00

00

SJ

ON

00

NO

NO

»r)

NO

m

(n

On

ON

00

CN

NO

NO

O

d

d

d

d

d

d

d

d

d

d

d

d

d

•-J

c

3 c

+1

+1

+ 1

+1

+1

+ !

+1

+1

+ 1

+1

+1

+ 1

+1

+1

U

c2 -

H a

ON

00

NO

00

in

m

00

CO

00

CO

in

o

Tt

O)

m

q

q

q

ON

q

O)

q

O)

X

d

d

00

d

00

d

00

NO

00

d

d

CN

CN

1-^

•-H

—

.M

1— M

CN

CN

a.

O

CO

00

(N

VO

in

ON

CN

On

m

in

a:

\c

r-

vn

IT)

to

(N

in

r-

in

NO

Tj-

00

o

lu

d

d

d

d

d

d

d

d

d

d

d

d

d

d

CO

O

8.^

+1

+1

+1

+ !

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

Q

X 3

ON

00

O

o

o

CO

00

o

r-

ON

o

CO

2

<

W

q

m

00

Ov

CN

CO

q

00

q

CN

On

d

d

d

loi

d

NO

d

d

d

d

d

d

>

■ — '

CN

CN

u.

O

n

"7

<N

o

(N

On

On

in

in

O

NO

00

NO

CN

CO

— H

•—

z

in

z

UJ

al

Sex

U-.

S

P-i

S

P-

2

[u

U-

s

LU

Ph

D

CO

<

LU

•-“v

c

to

CO

cd

8

z

Co

C3

Form

to

-5

1

5

o

to

-i:)

oc

0

to

C

•S!

00

§ ti

3

3

C

s:

Ci

3

a

q

s

Oo

tj

C

Mean and standard deviation. Masses are in g; culmen, wing chord, and tail length are in

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

17

Table 3

Result of Analyses of Variance Comparing Measurements of Members of

ChLOROSTILBON MELUSUGUS COMPLEX FROM NORTHWESTERN SOUTH AMERICA

Measurement

Sex F

Result of Tukey tests”

-/ ///-

ExpKjsed culmen M 109.19 <0.001 PU CA CC CN GI NA OL

/-

-/

-///-

F 94.75 <0.001 PU CA GI NA CC CN OL

Total culmen

/-

-/

-///-

M 53.90 <0.001 PU CA CC CN NA GI OL

Wing chord

-///-

F 71.83 <0.001 PU NA CA GI CC CN OL

-II-

-h

-/

-///-

M 48.81 <0.001 CC CA NA CN PU GI OL

-/-

-///-

F 65.53 <0.001 NA CA CC CN PU GI OL

Tail length

-/// //

M 53.08 <0.001 CA NA PU OL GI CC CN

-II-

-/

F 17.57 <0.005 NA CA PU GI CC OL CN

Body mass

-///-

M 29.64 <0.001 PU CA NA GI OL

-///-

F 42.47 <0.001 NA CA PU GI OL

-///-

Relative tail fork M 103.89 <0.001 NA OL CA PU GI CN CC

“ Abbreviations: CA = carihaeus', CC = chrysogaster — N coast of Colombia; CN = chrysogaster — Norte de Santander,
NA = napensis\ GI = gihsonv, OL = olivaresr, PU = pumilus.

Forms not significantly different (P < .05) in a given measurement are connected by solid lines. Breaks in the lines
indicate significant differences, and the number of slashes indicates the degree of significance: one slash = P < 0.05; two
slashes = P < 0.01; three slashes = P < 0.001.

chord, and mass. The larger size of olivaresi is most dramatically shown
by the data on body mass, perhaps the best indicator of overall size; this
form is fully 35-40% heavier than all the other forms, which do not differ
among themselves. In fact, olivaresi is apparently the largest form in the

18

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

entire genus Chlorostilbon, exceeding in size even the Cuban Emerald C.
ricordii (cf Ridgway 1911, Howell 1993). Regarding measurements of
the tail, the separation of olivaresi is more equivocal, chiefly because this
form, because of its overall large size, has a tail length more similar to
those of the gibsoni group with their relatively long, deeply forked tails,
but in its shallowly forked tail, olivaresi resembles closely various mem-
bers of the mellisugus group.

Although some of the remaining forms differed significantly from oth-
ers in one or another dimension, similar clean-cut separations between
groups were the exception. Males of the gibsoni group differed from those
of the mellisugus group significantly in tail length and highly significantly
in relative depth of tail fork; however, in most other dimensions the pat-
tern was less clear: often differences significant in one sex were not so
in the other, and the rank order of the forms often varied from one mea-
surement to another and/or between the sexes for a given measurement
(Tables 2, 3). This is undoubtedly at least in part a reflection of the mostly
small sample sizes (Table 2). One interesting result was the significant
difference in wing length between the northern and cis-Andean popula-
tions of chrysogaster, however, these populations were similar in other
dimensions. Aside from olivaresi, the most distinctive form, in terms of
measurements, was pumilus, with its unique combination of a very short
bill and a relatively long wing (Table 3; see below).

Patterns of plumage variation. — In eastern Colombia, specimens of
caribaeus showed no obvious variation between Arauca and Meta (in
particular, those from Arauca showed no tendency to approach chryso-
gaster). To the south, napensis differed in having a shorter, less forked
tail and more strongly bluish throat in the males, more lateral green fleck-
ing below and less pale tipping on the outer rectrices in the females, and
a somewhat longer bill in both sexes. I could detect no consistent differ-
ences in size or color between specimens of napensis from extreme south-
ern Meta, northwestern Caqueta, eastern Vaupes, or southern and eastern
Amazonas in Colombia, or (in size at least) eastern Ecuador or extreme
northeastern Peru. The largest male measured was from eastern Peru,
where the range of this form approaches that of the possibly indistin-
guishable phaeopygos. I have not seen specimens of phaeopygos from
Peru, so cannot comment on the validity of separating napensis', however,
female specimens of the latter I have seen do show the extensive green
flecking below used by Zimmer (1950) to distinguish this race from
phaeopygos and caribaeus. Although napensis and caribaeus differed sig-
nificantly in relative tail fork in this analysis (Table 3), it is worth noting
that this difference is apparently bridged by the adjacent subfurcatus and
duidae, to judge from the measurements of these forms given by Zimmer

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

19

and Phelps (1952). I should note in passing that my measurements are in
reasonably close agreement with those of Zimmer for those forms which
we both measured.

The range of olivaresi appears to be nearly completely nested within
that of napensis. Although I have seen no specimens of the latter from
localities adjacent to the Sierra de Chiribiquete, those from northwest,
south, and east of this range show no approach to the characters of oli-
varesi; in fact, napensis differs from olivaresi in most characters at least
as much as does any other cis-Andean form of the mellisugus complex.
This indicates that gene flow between olivaresi and napensis does not
occur; given the great difference in size, it is highly questionable whether
interbreeding could occur at all.

Within the gibsoni group, chrysogaster males have a glittering fore-
crown and usually, at least, a faint bluish tinge to the throat (more than
either pumilus or gibsoni). The tail is even more deeply forked than in
gibsoni; in females, the pattern of the tail is similar, although the gray
area at the bases of the outer rectrices tends to be darker and less exten-
sive. The difference in wing length between the populations of this form
east and west of the Serrania de Perija is interesting: at the very least, it
appears that the former shows no approach to the geographically adjacent
form of mellisugus, caribaeus. In color, pattern, and tail shape, the cis-
Andean population of chrysogaster also appears as different from cari-
baeus as does the trans-Andean population, suggesting that gene flow
between these populations is not occurring. The little information I have
on nitens suggests that this form is very similar to the trans-Andean chry-
sogaster. Although data are required on possible zones of contact between
nitens and caribaeus, the information currently available again favors
maintaining separate species status for the gibsoni group (including chry-
sogaster and nitens) and mellisugus.

The case of C. pumilus. — This form of western Colombia agrees with
members of the mellisugus group in having an all-black bill, in the glit-
tering crown of the males, and in the virtual absence of grey in the outer
rectrices of the females. However, the males have significantly longer,
more deeply forked tails and less bluish breasts than do any of the eastern
forms of this group. Females of pumilus differ from those of the eastern
races in their slightly forked to double-rounded tails, with extensive green
in most rectrices, and in the more extensive pale tips to the outer rectrices,
as well as in showing some dark gray at the bases of these rectrices. In
overall proportions, pumilus has a shorter bill and longer wing and tail
than do the eastern races, sex for sex.

In several of these characters, pumilus rather resembles the adjacent
gibsoni of the Magdalena Valley; however, it differs strongly from gibsoni

20

THE WILSON BULLETIN • Vol. 108. No. I, March 1996

in bill color, tail shape, bill length, crown color (males), and the color of
the outer rectrices (females). Moreover, these differences appear to be as
great in specimens from areas where the two forms approach each other
(western Tolima, western Antioquia), as in areas far from the possible
contact areas. Hence, the evidence presently available favors continuing
to separate pumilus and gibsoni at the species level.

Considering the next form to the north, assimilis, pumilus also shows
a number of clear-cut differences based upon the measurements, weights,
and descriptions of the latter in Wetmore (1968). In males, the crown of
pumilus is highly iridescent, while in assimilis it is plain; the gray of the
outer rectrices of the females is much more extensive in pumilus. In mea-
surements, pumilus has a much shorter bill, longer wing, and shorter tail
and also weighs considerably less in both sexes. The ranges of the two
forms are separated by a wide gap in eastern Panama (cf Wetmore 1968)
but from my observations of both forms in the field, I suspect that they
differ in ecology, with pumilus preferring significantly wetter areas than
does assimilis. I conclude from this analysis that pumilus is best consid-
ered an allospecies in the mellisugus superspecies rather than as a sub-
species of mellisugus (sensu stricto); this also eliminates the disjunct dis-
tribution of the latter (with pumilus separated by gibsoni from the other
forms of mellisugus).

Although I was unable to examine specimens of melanorhynchus, there
seems little reason to doubt that this form is conspecific with pumilus',
Zimmer (1950) noted that the two differ little, if at all, in coloration and
overlap in measurements and, in fact, might not be separable. The two
forms were supposed to differ in their preferred elevations, with melan-
orhynchus occurring mostly above 2000 m, and pumilus at lower eleva-
tions; however, this difference also tended to break down in the series of
specimens available to Zimmer (1950). If pumilus and melanorhynchus
are considered to comprise a separate allospecies in the mellisugus com-
plex, as I feel the evidence indicates, this species will have to be called
C. melanorhynchus (since melanorhynchus Gould 1860 has priority over
pumilus Gould 1872). Without having seen specimens of melanorhynchus
itself, I cannot comment on the advisability of synonymizing pumilus and
recommend further study.

On the basis of a phylogenetic species concept (Cracraft 1983, Mc-
Kitrick and Zink 1988), melanorhynchus (with pumilus) would also clear-
ly be entitled to species status; it possibly differentiated from other forms
of the mellisugus complex in the Choco humid forest refugium during
one of the dry epochs of the Pleistocene (cf Haffer 1974). I suggest the
English name West Andean Emerald for this form in view of its basically

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

21

Andean distribution, which lies to the west of all other members of the
mellisugus complex in South America.

Origin of the Chiribiquete Emerald. — C. olivaresi represents an anom-
aly among the cis-Andean populations of the mellisugus complex: its
characters are sharply discordant with the general trends of variation over
this broad area. Although large size is a character shared by another iso-
lated mountain endemic, duidae of Venezuela, the differences in color
and morphology of the latter relative to the forms occurring elsewhere in
southern and southeastern Venezuela, caribaeus and particularly subfur-
catus, are differences of degree only: there seems little reason to doubt
that duidae is but a derivative of one or the other (apparently neither has
been recorded from the adjacent lowlands), and there is little reason to
suspect that it is reproductively isolated from them (Zimmer and Phelps
1952). In fact, duidae was not considered by Mayr and Phelps (1967),
although it was recognized as distinct by Meyer de Schauensee and Phelps
(1978). The case of olivaresi is clearly not comparable to that of duidae:
it is certainly not an isolated derivative of napensis (or any other cis-
Andean form). The many points of resemblance to gibsoni (bill color, the
dull crown of the males, and the rectrix color of the females) suggest that
the origins of olivaresi might lie with that form, at least in part. At the
present time, the ranges of olivaresi and gibsoni are separated by at least
300 km of forested lowlands and foothills, which constitute unsuitable
habitat for both forms which are evidently adapted to dry or edaphically
scrubby habitats (and which are inhabited, at least at present, by napen-
sis). It is likely that the eastern foothills of the Andes in southeastern
Colombia represented a humid enclave or “forest refugium” through
Pliocene-Pleistocene times (Haffer 1974, 1985), such that the range of
gibsoni has probably never included or approached the Sierra de Chiri-
biquete. A more likely scenario is for one or a few individuals of the
ancestral population of gibsoni to have dispersed across the forested low-
lands to Chiribiquete, perhaps during a dry period in the Pliocene or early
Pleistocene. During such a period, the low passes south and west of the
Sierra de la Macarena might well have supported a more xeric, scrubby,
vegetation suitable for gibsoni, facilitating its arrival on the eastern slope
of the Andes. Consisting of sandstones of Permian age or earlier, the
Sierra de Chiribiquete antedates at least the final uplift of the Andes (cf
Gal vis, in Estrada and Fuertes 1993), and would have been available for
colonization by forms adapted to xeric or open habitats throughout the
latter part of the Cenozoic. The final population of Chlorostilbon to col-
onize the Sierra de Chiribiquete might even have been the result of hy-
bridization of the newly arrived gibsoni-iype birds with the form of mel-
lisugus already present in the adjoining areas. At least, this would provide

22

THE WILSON BULLETIN • VoL 108, No. I, March 1996

a tentative explanation for the meUisugus-\\](..& characters of olivaresi (very
shallowly forked tail and blue breast of the males). Once established, this
population of possibly hybrid origin might well have developed large size
in the process of adapting to its peculiar insular habitat. Large size is a
feature of some island populations (cf Grant 1965) including those of
Chlorostilbon (cf Ridgway 1911, Howell 1993), and the large size of
duidae may represent a result of the same processes, albeit at an earlier
and more incomplete stage. A possibly analogous case of presumptive
hybrid origin of a currently stable and well-differentiated form among the
white-eyes of Reunion Island has been discussed by Gill (1973). Genetic
studies in the mellisugus complex would assuredly shed light on the or-
igins of olivaresi, and might help to resolve several other questions re-
garding geographical variation and species limits in the complex. In any
case, by both biological and phylogenetic criteria, olivaresi appears as
entitled to separate (allo)species status as does any other form within the
entire mellisugus complex.

Among the avifauna of the Sierra de Chiribiquete, C. olivaresi is also
an anomaly. Virtually all of the species of the upper levels of the Sierra
whose subspecific allocations could be determined, belong to forms
whose distributions include the Llanos, the Orinoco region, and/or the
savannas of eastern Vaupes and adjacent Brazil, rather than the Amazo-
nian lowlands and Andean foothills to the west and south (Stiles et al.
1995). The lack of differentiation of these forms suggests a vicariance
pattern, probably a relict of more a continuous distribution of savanna or
scrub-adapted birds during the dry periods of the Pleistocene. The present
isolation of these birds from their relatives to the north and east probably
dates back no more than 15,000—20,000 years, much too short a time to
have permitted significant differentiation among birds of the Amazon
drainage (Capparella 1988).

Interestingly, one other species of Chiribiquete appears to have dis-
persed from the xeric upper Magdalena Valley rather than having its af-
finities with forms to the north and east. This species, Hemitriccus mar-
garitaceiventer, also appears to have differentiated in Chiribiquete from
its putative ancestral population in the Magdalena Valley, although to a
much lesser extent. This population clearly represents a distinct subspe-
cies of margaritaceiventer (Stiles, unpubl. data), probably of more recent
origin than C. olivaresi. In general terms, the Chiribiquete avifauna is
basically a relict of a formerly more continuous Orinoquian avifauna upon
which is superimposed a small number of forms that have dispersed from
similar habitats to the west, at different periods of the geological past (cf
Mayr and Phelps 1967). The most distinctive, and probably the oldest, of

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

23

these forms is undoubtedly the Chiribiquete Emerald, Chlorostilbon oli-
varesi.

AN OVERVIEW OF THE C. MELLISUGUS COMPLEX

The most recent treatment of this complex as a whole is that by Sibley
and Monroe (1990). These authors consider as allospecies of the melli-
sugus superspecies canivetii of Mexico and northern Central America,
assimilis of southern Costa Rica and western Panama, and mellisugus
(including pumilus and melanorhynchus) of northern South America,
while excluding gibsoni (including chrysogaster and nitens). This repre-
sents the opposite extreme from the broad treatment of Zimmer (1950)
who considered all these forms as subspecies of mellisugus. Especially in
view of the characters of olivaresi, which combines certain features of
both gibsoni and the mellisugus group, I would agree with Zimmer that
all of these forms are representatives of a common stock. In particular, I
see no justification for including the canivetii group while excluding gib-
soni from this complex: they share characters, such as the red lower man-
dible and deeply forked tail of the males, and extensive gray bases and
broad white tips to the lateral rectrices of females (Zimmer 1950). These
features could well represent parallel adaptations to more open, seasonal
habitats than are occupied by neighboring forms.

In general, I feel that too little attention has been paid to the ecology
of these hummingbirds — in particular, the range of elevations and humid-
ity conditions occupied by each form. Although all members of the group
prefer open, brushy habitats, so far as known, there are a number of
differences that could have bearing on their status. Eor instance, pumilus
appears to be primarily a bird of humid foothills, entering well into the
subtropical zone where it has been recorded to elevations of 2000 m or
more (Hilty and Brown 1986; pers. obs.); the same evidently is true of
melanorhynchus (Zimmer 1950). The preferred habitat of gibsoni appears
to be in hotter, drier areas; it occupies the floor of the middle and upper
Magdalena Valley, up to ca 2000 m or more in dry valleys, mainly on
the western side; chrysogaster and nitens are mainly or exclusively low-
land forms, with the latter occupying the driest areas along the northern
coast (Meyer de Schauensee 1948-1952). Also occurring primarily in the
lowlands east of the Andes are caribaeus (probably throughout the Lla-
nos) and napensis in more humid areas farther south, apparently largely
in areas of riverine scrub. Subfurcatus, by contrast, occurs mostly above
1000 m in the Gran Sabana and tepuis of eastern and southeastern Ven-
ezuela and adjacent Brazil and Guyana (Meyer de Schauensee and Phelps
1978); duidae is evidently isolated at similar elevations on Cenro Duida
(Zimmer and Phelps 1952). Different habitat preferences may help to

24

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

promote or reinforce reproductive isolation of different forms where they
meet, as perhaps is the case between pumilus and gihsoni on the western
side of the Magdalena Valley. This would likely be the case for napensis
and olivaresi, which appear restricted to the upper elevation Bonnetia
scrub of the Sierra de Chiribiquete. A detailed study of the displays of
these hummingbirds (cf Stiles 1983) might also help to clarify species
limits.

On present evidence, I recommend that the relationships of these forms
are best expressed by considering the mellisugus complex to comprise a
single {mellisugus) superspecies. In South America, the forms melano-
rhynchus (including pumilus), gibsoni (including chrysogaster and ni-
tens), olivaresi, and mellisugus (including caribaeus, napensis, phaeo-
pygos, subfurcatus, duidae, and one or more races in eastern South Amer-
ica beyond the scope of this study; see Zimmer 1950) should be consid-
ered allospecies within this superspecies.

In this connection, an interesting analysis of the northern forms of the
mellisugus complex has recently appeared (Howell 1993). Using argu-
ments similar to those employed here, Howell has presented strong evi-
dence for recognizing auriceps of western Mexico and forficatus of Coz-
umel Island, as distinct species. He also advocates recognizing salvini of
northern Central America as distinct from canivetii of southeastern Mex-
ico and extreme western Guatemala (considering the two to comprise a
superspecies). In the latter case, however, depth of tail fork (in both sexes)
appears to be the only reasonably clear-cut difference. This difference is
on the same order of magnitude as that between caribaeus and napensis
(in fact, the mean ratio of the relative tail forks of males of the latter two
forms is ca 2.19, whereas the corresponding mean ratio for canivetii and
salvini is only 1.30). The differences in patterns of the rectrices cited by
Howell between the two forms (cf his hg. 1) are of a magnitude that, in
my experience, could be subsumed by individual variation. In the absence
of a more detailed analysis of geographic variation within each form, 1
am hesitant to follow Howell (1993) in separating these forms at the
species level. For the present, I prefer to adopt a more conservative spe-
cies criterion and consider canivetii and osberti (the form of salvini in
question) as only subspecifically distinct, particularly as there does not
appear to be any significant ecological difference between them; this is
also more consistent with species criteria in the mellisugus complex as a
whole (contra Howell 1993). This problem clearly merits further study.
Also, in keeping with the mosaic pattern of variation in this complex as
discussed here, I consider auriceps and forficatus as allospecies of the
mellisugus superspecies, rather than as species apart.

I should note here that I had previously (Stiles and Skutch 1989) ques-

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

25

tioned the distinctness of assimilis from canivetii, based upon apparently
intermediate birds that I had observed and trapped (but in most cases not
collected) along the Pacific coast of southcentral Costa Rica. While wide-
spread hybridization and introgression does occur between Hoffmann’s
Woodpeckers Melanerpes hojfmanni of northwestern Costa Rica and Red-
crowned Woodpeckers (M. mbrucapillus) of the southern Pacific area
following deforestation (cf Stiles and Skutch 1989; various specimens), I
now feel that the evidence for the same in Chlorostilbon is still too weak
to justify lumping assimilis and canivetii and recommend further study
of this problem. For the present, it seems best to follow current practice
(A.O.U. 1983, Sibley and Monroe 1990) in considering these forms to
be allospecies; indeed, a limited zone of hybridization would not be in-
consistent with this interpretation.

In conclusion, I recommend that the ^‘'mellisugus complex be recog-
nized as a single superspecies rather than being divided into a mixture of
separate species and superspecies, some with disjunct distributions. From
Mexico to northern South America, I recommend recognition of the fol-
lowing allospecies: auriceps, forficatus, canivetii, assimilis, melanorhyn-
chus, gibsoni, olivaresi, and mellisugus. Together, the members of the
mellisugus superspecies show a fascinating pattern of allopatric distribu-
tions and mosaic geographical variation that bespeaks a complex evolu-
tionary history fully in keeping with the complicated geological history
of Central America and northern South America during the latter part of
the Cenozoic. These hummingbirds would seem to comprise a most fruit-
ful group for further genetic, phylogenetic, and biogeographical studies.

ACKNOWLEDGMENTS

For logistic and financial support during the expedition to the Sierra de Chiribiquete, I
am grateful to the Institute de Cooperacion Tecnica of Spain, the Institute de Ciencias
Naturales, and INDERENA. For help and companionship in the field, I thank Jose Luis
Tellen'a, Mario Dfaz, Bernardo Ortiz, Augusto Repizzo, and all the other members of our
expedition; the surgical expertise of Carlos Castano was particularly appreciated. Diego
Silva and Tomas Walschburger of the Fundacion Puerto Rastrojo kindly shared with me
their observations from the Rio Mesay area. The following kindly provided access to spec-
imens of Chlorostilbon in their care: Hno. Roque Casallas of the Museo de La Salle, Bogota;
Hernando Chirivi and Jorge Morales of INDERENA-UNIFEM; and Hno. Marco A. Serna
of the Museo del Colegio de San Jose, Medellin. I am grateful to M. Man'n for providing
measurements of specimens in the Louisiana State Museum of Natural Science. Arturo
Rodriguez contributed valuable curatorial assistance in the Institute de Ciencias Naturales.
S. N. G. Howell and M. Robbins provided useful comments on the manuscript.

LITERATURE CITED

American Ornithologists’ Union. 1983. Check-list of North American birds, 6th edition.

A. O. U., Washington, D.C.

26

THE WILSON BULLETIN • Vol. JOS, No. 1, March 1996

Capparella, a. 1988. Genetic variation in Neotropical birds: implications for the speciation
process. Acta XIX Intnl. Ornithol. Congr.: 1658-1664.

Chapman, E M. 1917. The distribution of bird-life in Colombia: a contribution to a bio-
logical survey of South America. Bull. Amer. Mus. Nat. Hist. 36:1—728.

Cracraft, J. 1983. Species concepts and speciation analysis. Current Ornithol. 1:159-187.

DIaz, M., E G. Stiles, and J. L. Telleri'a. 1996. Las comunidades de aves de las partes
superiores de la Sierra de Chiribiquete, Depto. del Caqueta, Colombia. Ardeola (in
press).

Estrada, J. and J. Euertes. 1993. Estudios botanicos en la Guyana colombiana V. Notas
sobre la vegetacion y la flora de la Sierra de Chiribiquete. Revta. Acad. Colomb. Cienc.
18:483-497.

Gill, E B. 1973. Intra-island variation in the Mascarene White-eye, Zosterops borbonica.
Ornithol. Monogr. 12:1—58.

Grant, P. R. 1965. The adaptive significance of some size trends in island birds. Evolution
19:355-367.

Haffer, j. 1974. Avian speciation in tropical South America. Publ. Nuttall Ornithol. Club
14:1-390.

. 1985. Avian zoogeography of the Neotropical lowlands. Ornithol. Monogr. 36:

113-146.

Hilty, S. L. and W. L. Brown. 1986. A guide to the birds of Colombia. Princeton Univ.
Press, Princeton, New Jersey.

Howell, S. N. G. 1993. Taxonomy and distribution of the hummingbird genus Chlorostil-
bon in Mexico and northern Central America. Euphonia 2:25-37.

Mayr, E. and W. H. Phelps, Jr. 1967. The origins of the bird fauna of the South Venezuela
highlands. Bull. Amer. Mus. Nat. Hist. 136:269-328.

McKitrick, M. C. and R. M. Zink. 1988. Species concepts in ornithology. Condor 90:1-14.

Meyer de Schauensee, R. 1948-1952. The birds of the Republic of Colombia. Caldasia
5(22-26):25 1-1212.

. 1960. The birds of Colombia and adjacent areas of South America. Livingston
Publ. Co., Narberth, Pennsylvania.

. 1966. The species of birds of South America with their distributions. Livingston
Publ. Co., Narberth, Pennsylvania.

AND W. H. Phelps, Jr. 1978. A guide to the birds of Venezuela. Princeton Univ.
Press, Princeton, New Jersey.

Ridgway, R. 1911. The birds of North and Middle America, part V. Bull. U.S. Natnl. Mus.,
vol. 50.

Sibley, C. G. and B. L. Monroe, Jr. 1990. Distribution and taxonomy of birds of the
world. Yale Univ. Press, New Haven, Connecticut.

Smithe, E 1975, 1981. Naturalists’ color guide. Amer. Mus. Natl. Hist. Press, New York,
New York.

Stiles, E G. 1983. Systematics of the southern forms of Selasphorus (Trochilidae). Auk
100:31 1-325.

. 1985a. Geographic variation in the Fiery-throated Hummingbird, Panterpe insig-
nis. Pp. 23-30 in Neotropical ornithology (P. A. Buckley, M. S. Foster, E. S. Morton,
R. R. Ridgely, and E G. Buckley, eds.). Ornithol. Monogr. 36.

. 1985b. Seasonal patterns and coevolution in the hummingbird-flower community
of a Costa Rican subtropical forest. Pp. 757—787 in Neotropical ornithology (P. A.
Buckley, M. S. Foster, E. S. Morton, R. R. Ridgely, and F. G. Buckley, eds.). Ornithol.
Monogr. 36.

Stiles • A NEW COLOMBIAN EMERALD HUMMINGBIRD

27

. 1995. Distribucion y variacion en el Ermitafio Carinegro {Phaethornis anthophilus)

en Colombia. Caldasia 18:119—130.

AND A. E Skutch. 1989. A guide to the birds of Costa Rica. Cornell Univ. Press,

Ithaca, New York.

, J. L. TellerIa, and M. Diaz. 1995. Observaciones sobre la ecologia, composicion

taxonomica, y zoogeografia de la avifauna de la Sierra de Chiribiquete, Depto. del
Caqueta, Colombia. Caldasia 17:481-500.

Wetmore, a. 1968. The birds of the Republic of Panama, part 2. Smithsonian Misc. Coll,
vol. 150, pt. 2.

Zar, J. H. 1988. Biostatistical analysis. Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

Zimmer, J. T. 1950. Studies of Peruvian birds #58. The genera Chlorostilbon, Thalurania,
Hylocharis, and Chrysuronia. Amer. Mus. Nat. Hist. Novitates, no. 1474:1—12. •

, AND W. H. Phelps, Jr. 1952. Two new birds from Venezuela. Amer. Mus. Nat.

Hist. Novitates, no. 1544:1-5.

COLOR PLATE

Publication of the frontispiece painting has been made possible by an endowment estab-
lished by George Miksch Sutton.

Wilson Bull, 108(1), 1996, pp. 28-35

REPRODUCTION AND MOVEMENTS OF MOUNTAIN
PLOVERS BREEDING IN COLORADO

Fritz L. Knopf and Jeffery R. Rupert

Abstract.— North American populations of Mountain Plovers {Charadrius montanus)
have declined 63% since 1966. Using radiotelemetry, we monitored plover nesting and
brood rearing during the 1993 and 1994 breeding seasons in Weld County, Colorado. Our
objectives were to extend preliminary breeding studies of 1992 (Miller and Knopf 1993)
and to determine the minimum area required for successful reproduction by this species.
Plovers began arriving on the breeding grounds in mid-March. Due to a high incidence of
predation, eggs hatched in only nine of 34 (26%) nests in 1993 and in 20 of 54 (37%) nests
in 1994. Daily survival rates of chicks were 0.957 (442 telemetry days, 44 chicks) and 0.951
(610 telemetry days, 42 chicks) each year, respectively. Plover broods moved an average
337 ± 46.5 m/day (N = 30) and 298 ±41.9 m/day (N = 14) in 1993 and 1994 (r = 1.10,
P = 0.27), respectively. Plovers that raised chicks to fledging used between 28 and 91 ha,
averaging 56.6 ±21.5 ha. Daily movement rates (r = 0.7, P = 0.48) and total area used (/
= 1.4, P = 0.17) were similar between broods where ^ one chick fledged and broods from
which no chicks fledged. Success of plovers in raising chicks appeared related either to
overall fox activity in the area or to how effectively the adult detected and distracted foxes.
Most plovers left the breeding grounds in mid-to-late July, four months after arrival. Pop-
ulation declines of Mountain Plovers appear independent of any recent landscape fragmen-
tation within this breeding stronghold of the species. Received 10 May 1995, accepted 1
Oct. 1995.

North American Mountain Plover {Charadrius montanus) populations
declined 63% between 1966 and 1991 (Knopf 1994) and the decline has
continued through 1993 (Knopf, in press). This species nests across the
western Great Plains and eastern Colorado Plateau region, with the breed-
ing stronghold being in Weld County, Colorado (Graul and Webster
1976). Mountain Plover nests are located in microsites of native short-
grass prairie dominated by blue grama (Bouteloua gracilis) with 30% or
more of the area being bare ground (Knopf and Miller 1994; Knopf and
Rupert 1996). A female plover lays a clutch for the male to incubate then
a second clutch for herself (Graul 1973). The precocial chicks receive
uniparental care and move up to 800 m from the nest shortly after hatch-
ing (Graul 1975). Brood-rearing habitat is described as areas with forbs
or objects such as fence posts where chicks can find shade to avoid mid-
day heat (Graul 1975).

In this study, we used radiotelemetry to monitor movements and de-
termine the area requirements of Mountain Plovers during the breeding
season. We also provide information on nest success and chick survival
to supplement preliminary findings of Miller and Knopf (1993) in 1992.

U.S. National Biological Service, 4512 McMurry Ave., Fort Collins, Colorado 80525-3400.

28

Knopf and Rupert • MOUNTAIN PLOVERS

29

We interpret our findings relative to conservation of this rapidly declining
species.

METHODS

We studied Mountain Plovers during the 1993 and 1994 breeding seasons on the Pawnee
National Grassland (PNG), a 781-km^ shortgrass prairie. Graul (1973) summarized the phys-
iography, vegetation, and climate of this region. We concentrated our study on three grazing
allotments (Keota, Keota Steer, and Owens) in the vicinity of Keota, Colorado, and on three
other allotments (Reno, Sand, and Wildhorse) northwest of Briggsdale, Colorado.

Once a nest was located, we monitored the progression of incubation (29 days) by egg
flotation every 3-5 days. During nest monitoring, we drove a vehicle near the nest and,
remaining in the vehicle, reached down to remove an egg to avoid spreading human scent
in the vicinity. Within 2-A days before hatching we captured adult birds on nests with a
fishing line snare or swing-door box trap made of wire mesh. In 1993, some juveniles from
unknown nests were captured by hand and aged per Miller and Knopf (1993) to supplement
survival rate data on older chicks.

All birds were banded with a U.S. Fish and Wildlife Service numbered metal band on
one leg and colored plastic bands on the other. Plovers (N = 43) were fitted with 1.5— 3.0
g radio transmitters with 15 cm antennae (Holohil Ltd., Woodlawn, Ontario, Canada, and
Advanced Telemetry Systems, Isanti, Minnesota. Mention of commercial products does not
constitute endorsement by the U.S. Government). The transmitter was affixed by applying
a light coating of waterproof epoxy adhesive (Titan Corp., Lynnwood, Washington) and
sliding it under the upper back feathers. Care was taken not to expose the skin to the epoxy.
The transmitter was not visible on the birds and the whip antenna was difficult to see even
with binoculars at close (<20 m) distances. Transmitter life was about 90 days, and trans-
mitters remained attached to adult birds until the prebasic molt. Birds were not handled after
initial transmitter attachment.

All chicks in monitored nests were color banded on the day of hatching, then never
rehandled. We subsequently located broods almost daily at distances up to 1000 m with a
TRX-1000 Wildlife Materials Inc. (Carbondale, Illinois) receiver and a hand-held, three-
element Yagi antenna. All relocations were from a vehicle, again to avoid spreading human
scent. Specific locations of plovers were recorded using the Magellan NAV 5000 Global
Positioning System [Magellan Systems Corporation, San Dimas, California]. Position read-
ings were recorded to the nearest hundredth of a minute and only if a satellite geometric
quotient registered > 7 on a scale of 1-9. Based upon readings at a known (surveyed)
benchmark, accuracy of our specific instrument was calculated at 7.2 ± 1 .4 (x ± SD) m
latitude and 8.4 ± 1 .6 m longitude for an area error of ± 60.6 m^.

The area used by broods was estimated by superimposing a grid over a map of the study
area and counting the number of new cells (10 X 10 m) visited over time. This is an
adaptation of the grid-cell home range estimation method first proposed by Siniff and Tester
(1965). Whereas di.scontinuous telemetry data often result in successive locations being
several cells apart, we conservatively assumed that birds traveled in a straight line between
locations. Thus, area-of-use descriptions are considered to be the minimum (vs actual) area
requirements for raising a brood.

RESULTS

Egg/chick survival. — Plovers were first observed on the PNG on 17
March 1993 and 21 March 1994, although the exact date of the first bird

30

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Table 1

Productivity and Chick Survival Rates of Mountain Plovers on the Pawnee
National Grassland, Weld County, Colorado, 1992-1994

X No./Successful nest

Year

No.

nests

Nest

success

(%)

Eggs

hatched

Chicks

fledged

Chicks

migrated®

Daily

survival

rate

1992'=

14

50

2.6

1.21

0.74

0.977

1993

34

26

2.4

0.26

0.22

0.957*=

1994

54

37

2.6

0.35

0.17

0.95 U

“Calculated from average of 18 days that chicks remained on breeding grounds after fledging (Miller and Knopf 1993).
•’ Data from Miller and Knopf (1993).

' Ba.sed on 442 telemetry days, 44 chicks
Based on 610 telemetry days, 42 chicks.

arriving may have been a day or two earlier. We monitored 34 nests in
1993 and 54 nests in 1994. Three clutches were abandoned each year. In
1993, two clutches were abandoned after being partially predated. The
other clutch was abandoned after it had been flooded from 4 to 6 June
and then incubated again from 7 to 20 June. In 1994, two clutches were
abandoned for unknown reasons. One other was abandoned after being
incubated two weeks beyond normal hatching time. Adult plovers from
three of the six nesting efforts stayed in the nest area until migrating,
whereas the other three left the area within 48 hours and could not be
relocated.

Predation rates on both eggs and chicks were high (see Miller and
Knopf [1993] for a list of potential predators), and egg and chick survival
were low (Table 1). Survival probabilities of the 1994 chicks partitioned
into 10-day intervals (day 1-10 = 0.935, 1 1-20 = 0.957, 21-30 = 0.970,
and 31^0 = 0.971) indicate that survival increased with age of the chick.
Generally, daily survival rates were only slightly lower than reported for
1992 (Miller and Knopf 1993). However, the lower daily rates resulted
in a drastic decline in the number of chicks produced per nesting effort
if projected to the time when fledged chicks left the breeding area.

Movement patterns. — We captured and placed transmitters on 17 and
26 adult plovers with broods in 1993 and 1994, respectively. Each brood
was tracked until either all members of the brood were killed by a pred-
ator or fledged chicks left the nest vicinity. Of adults fitted with trans-
mitters, we obtained prolonged movement and area-of-use data on seven
and 14 broods in 1993 and 1994, respectively. An additional 26 chicks
were aged and fitted with transmitters in 1993, of which 23 provided
usable information. Thus, data were obtained on 30 broods in 1993 and

Knopf and Rupert • MOUNTAIN PLOVERS

31

14 broods in 1994. The average duration of radio-tracking was 10.3 ±
5.2 and 22.9 ± 13.4 days for each brood in 1993 and 1994, respectively.

Based upon 290 telemetry days, plover broods moved an average of
337 ± 46.5 m/day in 1993 (range 61 to 600 m). The 320 telemetry days
in 1994 indicated an average move of 298 ± 41.9 m/day (range 85 to
651 m). Distances moved were similar (r = 1.10, P = 0.27, 608 df)
between years. Of the 42 broods, 38 moved 100—500 m/day. Of the re-
maining four broods, two were monitored for only five days and moved
<100 m/day. Two others, monitored four days and 13 days, moved 1000
and 1085 m/day, respectively.

Most losses of chicks were to swift foxes (Vulpes velox). Average daily
movements were similar {t = 0.7, P = 0.48, df = 40) among broods in
which all chicks were killed by predators (400 ± 313 m, N = 15) and
broods where chicks fledged (340 ± 226 m, N = 27). The total distances
moved during the first 10 days after hatching of eggs (when predation
rates were highest) were also similar {t - 0.50, P - 0.62, df = 14)
between broods in which all chicks were lost and broods where chicks
fledged.

Brood-rearing area. — Due to the high rates of predation and some
transmitters on older chicks becoming dislodged, we were only able to
track six broods from hatching to fledging. The total minimum area used
by those six broods ranged from 28 ha to 91 ha and averaged 56.6 ±
21.5 ha. As suspected from the movement data, the average area used on
a daily basis by broods where chicks fledged (2.5 ± 1.6 ha, N = 27) vs
broods that lost all chicks to predators (3.4 ± 2.4 ha, N = 15) were
similar {t = 1.4, P = 0.17, df = 40).

Timing of departure. — At least two adults renested after losing a clutch
or brood early in the breeding season (May, early June). One adult whose
chicks hatched on 21 May 1993 was located on a second nest 140 m
away on 21 June. The transmitter had failed, so knowledge of the fate of
the first brood of three chicks is not certain. This bird abandoned its
second clutch after it was partially depredated; it remained in the vicinity
with other plovers until 14 July when it left the area, returned 26-29 July,
then left the area for the season.

This example illustrates the general pattern of flocking and departure
from the breeding grounds. A few adult plovers began congregating in
flocks in mid-June each year. Fledglings started to appear in these flocks
in July. Many adults also were undergoing a prebasic molt at this time,
precipitating an increased rate of transmitter loss. Plovers with transmit-
ters began leaving the study areas gradually after mid-June, but a major
exodus of plovers occurred mid-late July. Seventeen plovers in 1993 and
eight plovers in 1994 which were being located daily were known to have

32

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

left the study area between 14 July and 3 August. In 1993, an aerial
survey of Weld County from 3 to 9 August confirmed that not one of 17
transmittered birds remained in the vicinity. The few adults (transmittered
and untransmittered) that we could find in early August of 1993 all had
large chicks incapable of flight.

DISCUSSION

Productivity. — Miller and Knopf (1993) reported that adult survival
and productivity on the PNG were similar to historical studies. We re-
cently documented that survival rates of birds wintering in California are
also high (Knopf and Rupert 1995). In this study, we recorded daily
survival rates of chicks in 1993 and 1994 similar to those reported for
1992. The more intensive efforts described here detected an increasing
probability of survival with age of the chick, as reported by Graul (1975).
Thus, the 1992 daily survival rate (0.977) was inflated slightly by a high
incidence of older chicks being transmittered that year.

An unknown percentage of female plovers lay two clutches, the male
incubates the first as the female lays a second clutch for herself (Graul
1973). The incidence of male-incubated clutches may increase with food
abundance (Graul 1976). Thus, whereas 0.26 and 0.35 chicks fledged per
nest compares poorly with fledging rates of some other North American
plovers (Page et al. 1983, Haig and Oring 1988, Prindiville Gaines and
Ryan 1988), the actual productivity per pair may be up to twice that value
in a given year.

The number of eggs hatched per clutch has remained stable when com-
pared to studies conducted 10 (McCaffery et al. 1984) and 25 (Graul
1975) years earlier. Most reproductive losses were due to fox predation.
We believe that the greater losses of nests and chicks to predators in 1993
and 1994 compared to 1992 were attributable indirectly to reduced food
resources. Mountain Plover reproductive efforts (Graul 1976), as those of
other grassland birds (George et al. 1992), are less successful in years of
drought. The two seasons of this study were drought years on the PNG.
Grasshoppers are a major food item of both plovers and foxes, and their
populations were very low both years. Low grasshopper populations
would increase the time spent foraging for both species, thus increasing
the probability of eggs and chicks being detected by foxes.

Movement patterns. — Reproductive success of many birds appears to
be a tradeoff between acquiring food and risking predation (Martin 1992).
From an evolutionary perspective, numerous traits of Mountain Plovers
may reduce detection by predators, including the cryptic coloration of
chicks and crypsis to avoid detection (Sordahl 1991), two clutches in-
cubated separately by the two adults (Graul 1973), shell removal at hatch-

Knopf and Rupert • MOUNTAIN PLOVERS

33

ing (Graul 1975), rapid movement of chicks away from the nest (Graul
1975), predator distraction displays by adults (McCaffery et al. 1984),
and the ability of chicks to fly at only 70% of adult body weight (Miller
and Knopf 1993).

Mountain Plovers led hatchlings away from the nest as soon as they
were dry. We regularly tracked the directional movement of a brood up
to 2 km within two or three days of hatching. Many plovers moved broods
to areas of disturbed prairie (Knopf and Rupert 1996) and then remained
in those general areas. In contemporary prairie landscapes, such distur-
bances are either areas frequented for watering and loafing by cattle or
fallow agricultural fields. After the initial move, many broods remained
in the vicinity of these areas where chicks foraged on small insects.

Area requirements. — This study was precipitated by the conservation
need for information on the minimum area necessary for plovers to raise
chicks. The minimum area within which a brood was raised was 28 ha.
Plovers raised chicks in broadly overlapping areas, with two or three
broods sometimes occurring in a general vicinity (such as around a cattle
watering tank). Thus, the potential exists for a suitable area to meet the
needs for more than one bird to raise chicks successfully.

The movement rates and area-of-use by plovers that successfully raised
their chicks to fledge did not differ from those of plovers that lost the
entire brood. Swift foxes were often seen hunting during daylight hours
(especially after mid-June as pups began eating prey) in addition to night
foraging. The eventual success of a plover in raising its chicks appeared
to be the result of either the overall fox activity in the immediate area
being used by the brood or the effectiveness of the adult plover in de-
tecting and distracting a fox.

Departure from breeding grounds. — Most Mountain Plovers that bred
in Weld County left by 1 August each year. Occasional flocks of plovers
were seen after early August, with the latest being on 8 October 1993.
We assumed the later flocks were of birds moving south from more north-
erly nesting areas. The first plovers arrive on California wintering grounds
in mid-October. The prolonged period of migration to the wintering areas
is in sharp contrast to the apparently direct, nonstop flight from California
back to Colorado in March (Knopf and Rupert 1995). Plovers spend ap-
proximately four months on their Colorado breeding grounds, five months
on the California wintering grounds, and three months moving from the
former to the latter.

CONCLUSIONS

Miller and Knopf (1993) concluded that recent population declines of
Mountain Plovers can be attributed either to long-term declines in repro-

34

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

ductive success or to phenomena occurring at nonbreeding areas. The
comparatively lower survival rates of chicks observed in 1993 and 1994
support that preliminary conclusion. Although swift fox densities seem
high on the PNG, the fox is only locally distributed within the breeding
range of the plover. We would not expect, for example, such a high rate
of predation in Phillips County, Montana (the second major breeding
locale) due to the absence of foxes in that state.

The shortgrass prairie, like many native vegetative associations, has
become highly fragmented in the last century. The PNG is no exception,
with 781 km^ occurring in 130 parcels ranging from 16 to 23,895 ha (B.
Ladd, pers. commun.). Although some of the smaller PNG parcels are
<28 ha, the minimum area required for brood rearing, all are contiguous
to private lands that are likewise managed as rangelands. We feel that
these private lands also provide suitable Mountain Plover habitat. Thus,
we conclude that virtually all relatively flat, grazed shortgrass prairie par-
cels provide potential habitats for plovers on the PNG. The current decline
of the North American Mountain Plover population appears independent
of recent fragmentation of landscapes within the native shortgrass prairie
of northeastern Colorado, the breeding stronghold (Graul and Webster
1976) of the species.

ACKNOWLEDGMENTS

We thank the U.S. Lorest Service, for financial assistance, and personnel of the Rocky
Mountain Region, Arapaho-Roosevelt National Lorest, and the Pawnee National Grassland
District for technical assistance. We are especially grateful to Larry Mullen for administra-
tive support. The Colorado Division of Wildlife conducted aerial reconnaissance, and we
specifically thank Jim Dennis for his efforts. Susan Skagen and Tex A. Sordahl provided
helpful comments on the manuscript. Clif Knopf provided assistance in the field.

LITERATURE CITED

George, T. L., A. C. Fowler, R. L. Knight, and L. C. McEwen. 1992. Impacts of a severe
drought on grassland birds in western North Dakota. Ecological Applications 2:275—
284.

Graul, W. D. 1973. Adaptive aspects of the Mountain Plover social system. Living Bird
12:69-94.

. 1975. Breeding biology of the Mountain Plover. Wilson Bull. 87:6—31.

. 1976. Food fluctuations and multiple clutches in the Mountain Plover. Auk 93:

166-167.

AND L. E. Webster. 1976. Breeding status of the Mountain Plover. Condor 78:

265-267.

Haig, S. and L. W. Oring. 1988. Mate, site, and territory fidelity in Piping Plovers. Auk
105:268-277.

Knopf, F. L. 1994. Avian assemblages on altered grasslands. Stud. Avian. Biol. 15:247—
257.

. 1996. Prairie legacies — birds in Prairie conservation: preserving North America’s

Knopf and Rupert • MOUNTAIN PLOVERS

35

most endangered ecosystem (E B. Samson and E L. Knopf, eds.)- Island Press, Covelo,
California (in press).

AND J. R. Rupert. 1995. Habits and habitats of Mountain Plovers in California.

Condor 97:743-751.

AND . 1996. Declining species on private lands: Mountain Plovers on plowed

ground. Wildl. Soc. Bull, (in press).

Martin, T. E. 1992. Interaction of nest predation and food limitation in reproductive suc-
cess. Current Ornithol. 9:163-197.

McCaffery, B. J., T. a. Sordahl, and P. Zahler. 1984. Behavioral ecology of the Moun-
tain Plover in northeastern Colorado. Wader Study Group Bull. 40:18—21.

Miller, B. J. and E L. Knopf. 1993. Growth and survival of Mountain Plovers. J. Field
Ornithol. 64:500—506; 65:193.

Page, G. W., L. E. Stenzel, D. W. Winkler, and C. W. Swarth. 1983. Spacing out at
Mono Lake: breeding success, nest density, and predation in the Snowy Plover. Auk
100:13-24.

Prindiville Gaines, E. and M. R. Ryan. 1988. Piping Plover habitat use and reproductive
success North Dakota. J. Wildl. Manage. 52:266—273.

SiNiFF, D. B. and j. R. Tester. 1965. Computer analysis of animal movement data obtained
by telemetry. BioScience 15:104—108.

Sordahl, T. A. 1991. Antipredator behavior of Mountain Plover chicks. Prairie Nat. 23:
109-115.

Wilson Bull., 108(1), 1996, pp. 36-52

TRIGEMINAL REPELLENTS DO NOT PROMOTE
CONDITIONED ODOR AVOIDANCE IN
EUROPEAN STARLINGS

Larry Clark

Abstract. — Birds, and in particular European Starlings (Sturnus vulgaris), avoid con-
sumption of fluid and food treated with the natural plant products, methyl anthranilate and
o-aminoacetophenone. Avoidance is an unlearned response most likely mediated via chem-
ically sensitive fibers of the trigeminal nerve. The trigeminal nerve codes for chemical
irritation and pain. Starlings are not repelled by the odor of the compounds, nor is olfaction
important in the avoidance response. Moreover, starlings fail to learn to avoid the odor of
the repellents, even after direct oral contact with liquid repellent. While trigeminal irritants
can be powerful repellents, the absence of associative learning for these repellents will
influence the application strategy for formulation and use. More broadly, the difference in
learning abilities associated with trigeminal repellents and those commonly responsible for
conditioned avoidance learning have implications for the structure of chemical defenses of
fruits and the prevention of untimely frugivory. Received 24 Feb. 1995, accepted I Sept.
1995.

Nonlethal bird repellents are important components of an integrated
wildlife management strategy. Repellents can be used to protect birds
from human activities (Clark and Shah 1993) or to minimize damage
caused by birds (Mason and Clark 1992). The social emphasis on safe,
nonlethal methods to resolve conflicts between humans and birds has
resulted in numerous attempts to identify new repellents (Dolbeer 1986,
Crocker and Perry 1990, Clark and Shah 1991). However, reported effi-
cacy of nonlethal repellents is highly variable (Mason and Clark 1992).
In part, this is due to a misunderstanding about how repellents work. To
minimize failure rates in the field, several fundamental questions about
mode of action and formation of avoidance response remain to be re-
solved.

Nonlethal chemical repellents operate via two distinct mechanisms (Za-
horik 1976), conditioned avoidance and nonlearned avoidance. In con-
ditioned avoidance learning, birds learn to avoid sensory cues paired with
a stimulus that causes illness (Garcia et al. 1966). The magnitude and
persistence of the avoidance response depends on the toxic potential of
the sickness producing agent and the localization of the illness. Pelchat
et al. (1983) found conditioned avoidance was strongest in the rat when

United States Dept, of Agriculture, Animal and Plant Health Inspection Service, Animal Damage Control,
Denver Wildlife Research Center and Monell Chemical Senses Center, 3500 Market Street, Philadelphia,
Pennsylvania 19104. (Present Address; United States Dept, of Agriculture, Animal Plant Health Inspection
Service, National Wildlife Research Center, 1716 Heath Parkway, Fort Collins, Colorado 80524).

36

Clark • BIRD REPELLENTS

37

sickness was localized in the upper region of the small intestine. In birds,
ingestion of carbamate insecticides (e.g., methiocarb) and fungicides (e.g.,
thiram, ziram) causes gastrointestinal sickness. Substances causing sick-
ness have been used to condition birds to avoid tastes (Schuler 1983),
odors (Clark and Mason 1987), and visual cues (Mason and Reidinger
1983). In a nonlearned avoidance response, substances possess taste,
smell, or irritating qualities that are perceived as unpalatable by birds.
Generally, in the quantities ingested, these substances do not cause sick-
ness (Clark and Mason 1993).

Previous studies indicated that acetophenone and anthranilate bird re-
pellents must be present in high concentrations to be effective (Clark et
al. 1991). High concentrations of repellents (hundreds to thousands ppm)
can be delivered orally, in food or fluid, or to the eye via aerosols. Re-
sponsiveness to only high concentrations suggests mediation by the tri-
geminal system as opposed to olfaction or taste (Walker et al. 1986).
Chemically sensitive fibers of the trigeminal nerve mediate response to
irritating and painful stimuli (Green et al. 1990). Yet coding for pain or
irritation does not necessarily imply tissue damage (Clark 1995).

Common questions are whether the odors of acetophenones and an-
thranilates are repellent or whether birds can learn to avoid the odor of
these repellents. These questions imply avoidance behavior mediated via
olfaction and the formation of a conditioned avoidance response. Deter-
mining whether avoidance is influenced by olfaction or trigeminal cues
is critical to the conceptualization and implementation of delivery strat-
egies of repellents and to the understanding of how birds may respond to
natural plant or insect chemical defenses.

The experiments in this paper address these question and are used to
argue the point that acetophenone and anthranilate compounds are trigem-
inal irritants (repellents). As a test of the appropriateness of the experi-
mental paradigm, the effect of short-term water deprivation on subsequent
drinking assays was evaluated in Experiment 1. Experiment 2 tested
whether naive starlings avoided the odor of a repellent and whether oral
contact with a repellent was a sufficiently adverse experience to train
starlings to subsequently avoid the odor of a repellent. This experiment
also tested the effect of stimulus sequence on the outcome of the drinking
assays. Experiment 3 tested the effect of prolonged exposure to orally
delivered repellents on the subsequent response to the odor of the repel-
lent. In Experiment 4, the role of olfaction in the mediation of the avoid-
ance response was assessed.

METHODS

Study subjects. — European Starlings (Sturnus vulgari.<;) were decoy-trapped at Sandusky,
Ohio, and transported to the laboratory in Philadelphia where they were kept in group

38

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

housing until selected for experimentation. Starlings were maintained on chick starter mash
{ad libitum) supplemented with a vitamin mixture and fresh apples (weekly). Tap water was
available continuously, except during testing. Because starlings exhibit a seasonality in their
olfactory ability (Clark and Smeraski 1990), all tests were conducted during the spring when
starlings have good olfactory acuity and discrimination ability. Starlings were maintained
on a constant temperature (23°C), 14:10 h light: dark cycle during their residence in the
laboratory.

Test stimuli. — Fluid intake for o-aminoacetophenone (OAP) and methyl anthranilate (MA)
was evaluated using standard 6-h one-bottle (no-choice) drinking tests (Clark and Shah
1994). These compounds were selected as representative of nonlethal acetophenone and
anthranilate bird repellents. Both compounds are natural plant metabolites that have organ-
oleptic characteristics that humans perceive as a musky/foxy odor for OAP and a grapey
odor for MA (Acree et al. 1990). Concentrations selected for testing were based upon water
solubility limits of the least soluble compound (MA) and the least practical concentration
yielding reliable rejection of treated fluid by most starlings.

Experiment 1 : Ejfects of water deprivation on fluid consumption. — The objectives of the
first experiment were to (1) determine the effect of diurnal water deprivation on overnight
water consumption and (2) determine the effect of a three-day diurnal water deprivation
schedule on post-test water intake. Determining the potential magnitude of the carry-over
effect attributable to water deprivation was important for the interpretation of other exper-
iments relating to the effects of chemical repellents. While previous studies showed that the
6-h drinking test used here is not sufficient to cause a severe water deficit in starlings (Clark
and Shah 1991), it was decided that a revisitation of these questions was in order.

The experimental design consisted of a standard one-bottle, 6-h assay. On the first day
(the pre-test period), 18 birds were randomly assigned to three groups (N = 6/group) and
presented with tap water in graduated Richter tubes. The recording period began at 10:00
and ended at 16:00 (hereafter defined as the diurnal test period). The Richter tubes were
replaced with a second set of tubes and water was available ad libitum through the period
16:00 to lights out at 19:00, throughout the night, and for the period from lights on at 7:00
to 10:00, the start of the next test sequence (hereafter defined as the overnight period). As
a precondition for further testing, similarity for average diurnal water consumption among
groups was verified using a one-way fixed effects analysis of variance (anova).

On the second day, birds within the groups were presented with one of three randomly
assigned treatments. Birds within the control group were presented with tap water. Birds
within the second group were presented with a 28 mM solution of o-aminoacetophenone
(OAP). Because OAP is a potent bird repellent, it was anticipated that this group would
experience voluntary water deprivation by avoiding the treated water (Clark and Shah 1991).
Birds within the third group were water deprived by physically excluding them from the
Richter tubes. The protocol for recording water consumption for the diurnal and overnight
periods followed that described above. Birds retained their water presentation treatment
assignments for days three and four, and water intake for the diurnal and overnight periods
followed the format described above.

On the fifth day (the post-test period), birds within all three treatment groups were pre-
sented with untreated tap water and intake was monitored according to the prescribed pro-
tocol.

The first question, “. . . does diurnal fluid deprivation affect overnight water consump-
tion?”, was addressed using a 3 X 3 repeated measures fixed-effect analysis of covariance
(ancova). The dependent variable was overnight water consumption. The previous day’s
diurnal fluid intake was used as a covariate. Because the covariate was measured each day
along with the dependent variable, it was treated as a changing covariate, i.e. separate

Clark • BIRD REPELLENTS

39

residuals were calculated for each day. Days were treated as a repeated measure, and group
was treated as a between measures effect. Multivariate criteria (Rao R, Wilk’s lambda) were
used to simultaneously test repeated measures contrasts on adjusted means when there were
more than two levels on the repeated measure. This approach is to be preferred because it
does not rely on the assumption that the repeated measures are independent. In circumstances
where only univariate statistics were estimable, i.e., one degree of freedom, the F statistic
was used to assess significance. The latter case was most common in planned comparison
of treatment levels by contrasts. All post-hoc tests were analyzed using a Scheffe’s test. This
and subsequent analyses were carried out using the STATISTICAL (1994) software pack-
ages (MANOVA procedures). This software is particularly suited to repeated measures and
other longitudinal designs.

The second question, “. . . did the intervening treatments affect post-test water intake
relative to the pretreatment period?” was addressed using a 2 X 3 repeated measures fixed-
effect ANOVA. The dependent variable was diurnal fluid intake, period (pre vs post test)
was a within measures (repeated) effect, and group (i.e., water deprivation schedule) was a
between measures effect.

Experiment 2: Effects of odor and treatment sequence on water consumption. — Starlings
readily form conditioned odor avoidance responses when the unconditional stimulus pro-
duces a strong gastrointestinal illness (Clark and Mason 1987). Starlings also avoid some
substances upon initial contact, indicating that the avoidance in not learned (Clark and
Mason 1993). This latter response indicates a different mechanism for avoidance, one that
suggests hedonic attributes (e.g., taste, smell, irritation) form the basis of acceptance or
rejection. Clark and Mason termed these compounds sensory repellents. The objectives of
this experiment were to (1) determine whether oral and/or nasal exposure to GAP influenced
fluid intake, (2) determine whether starlings could learn to avoid the odor of GAP once they
had an aversive oral exposure to the substance, and (3) assess whether the order of presen-
tation of GAP in solution or odor influenced the behavioral response.

Thirty-six starlings were randomly assigned to six groups of six birds each. Gn the first
day, birds were tested for baseline water consumption in a standard 6-h drinking test fol-
lowing procedures outlined in Experiment 1. Attached to both sides of the Richter tube
sipping port was a 40 mm diameter, opaquely screened polypropylene disk containing a
wick saturated with tap water.

During the three-day test period, birds were presented with one of three treatment con-
ditions (Richter tube/wick disk combinations): [w/w] a tap water filled Richter tube and a
tap water soaked wick, [o/w] a Richter tube filled with 28 mM GAP solution and tap water
soaked wick, and [w/o] a tap water filled Richter tube and an GAP .soaked wick. Presen-
tations of the Richter tube/wick combinations to experimental cohorts were established ac-
cording to two possible Latin square designs (Table 1). Gn the fifth day (post-treatment
period), all birds were presented with the control tube/wick pairing (w/w). Diurnal and
overnight fluid intake were monitored following procedures outlined in Experiment 1.

Data were analyzed using a fixed-effects 3X3X6 nested ANGVA design with repeated
measures. Diurnal fluid intake was the dependent variable and day was a repeated measure
with three levels. Sequence (two levels) and group (six levels) were between measures
effects, with group nested within sequence (Table 1 ). To te.st whether the context of stimulus
presentation is important for repellency, contrasts between w/w and o/w, and w/w and w/o
were made for the first day of testing. These comparisons addressed the question of whether
GAP in solution (where olfactory, gustatory, and trigeminal systems may influence respond-
ing) or whether GAP as an odor stimulus (where olfactory and trigeminal systems may
influence responding) were important in the formation of an avoidance response. By con-
sidering only the first day of testing, these compari.sons controlled for possible learning

40

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Latinized Designs for the Presentation Order of Treatments®

Design 1

Design 2

Day

i

2

3

Day

i

2

3

Group 1

w/w

o/w

w/o

Group 4

w/o

o/w

w/w

2

o/w

w/o

w/w

5

o/w

w/w

w/o

3

w/o

w/w

o/w

6

w/w

w/o

o/w

“ Experimental codes are: (w/w) tap water presented in the Richter tube with blank odor disks on either side of the
drinking port; (o/w) 28 mM OAP solution in the Richter tube with blank odor disks on either side of the drinking port;
(w/o) water presented in the Richter tube with saturated OAP wicks in the odor disks on either side of the drinking port.

effects while considering whether repellency is mediated via nasal/ocular (i.e., volatile) or
oral/nasal/ocular (i.e., contact and volatile) mechanisms. Similar contrast analyses comparing
w/w to w/o were performed for days two and three. These tests allowed for the possibility
of learning while addressing the question whether consumption for birds exposed to odor
differed from that of the control condition.

OAP is a good sensory repellent when delivered orally. However, can oral presentation
of OAP be used to condition birds to avoid the odor of OAP? This question was addressed
by inspection of contrasts of the day X treatment interaction term. Intake for birds presented
with treatment levels w/w and w/o on day 2 were compared for those birds presented with
o/w on day 1, i.e., the putative training day. The overall sequence effect throughout the test
was determined by consideration of the main effect and by inspection of post hoc differences
among means grouped by treatment category using the Scheffe’s test.

Experiment 3: The conditioned odor avoidance response. — In the absence of water de-
privation effects (Experiment 1), sequence effects (Experiment 2), and an apparent inability
for short exposure to orally delivered OAP to act as an unconditional stimulus (Experiment
2), the length of time starlings were orally exposed to a chemical repellent was increased
to determine whether a conditioned avoidance response towards odor could be attained.

Separate experiments on two bird repellents were carried out, and the experimental design
of these tests was as follows. Starlings were tested serially for fluid intake in a standard no-
choice (one-bottle) 6-h drinking assay. On the first day, eighteen starlings were drawn at
random from the group housing pool, randomly assigned to one of two groups (N = 9/
group) and pre-test water consumption was monitored according to methods described in
Experiment 1 . Attached to both sides of the Richter tube sipping port was a 40 mm diameter,
opaquely screened polypropylene disk containing a wick saturated with tap water (w/w,
following the nomenclature convention in Experiment 2).

Eor each of three successive days (2-4), starlings within groups were presented with
Richter tubes containing a 28 mM solution of a sensory repellent (OAP or MA) and diurnal
fluid intake was monitored as described above (Experiment 1). Attached to both sides of
the Richter tube sipping port was a wick saturated with tap water (o/w). The strong odor
derived from the repellent emanated from the drinking port of the Richter tube, i.e., repellent
solution.

On the fifth and sixth day, starlings were presented with Richter tubes containing tap
water and fluid intake was monitored as above. During this period the wick inside the disks
was saturated with repellent (w/o). Thus, the strong odor of the repellent was present in the
apparatus but did not originate from the .solution.

Data were analyzed using a fixed effects one-way ANOVA with repeated measures on

Clark • BIRD REPELLENTS

41

days/treatment effect. The confounding of day and treatment on water intake was considered
unimportant. Thus, in the absence of carryover effects due to diurnal water deprivation
(Clark and Shah 1991, Experiment 1), any day/treatment effect was to be interpreted pri-
marily as a treatment effect. In addition to the day/treatment effect, a single a prion planned
comparison was made using contrasts. The w/w condition on day 1 was compared to the
w/o condition (days 5-6) as an assessment for the formation of a conditioned odor avoidance
response. The results for the GAP and MA tests were analyzed separately.

Experiment 4: The role of olfaction in the avoidance response. — Odors can stimulate the
olfactory system, but they can also gain access to chemically sensitive receptors of the
trigeminal system via the mouth, nose, or eye. Thus, avoidance of an odor might be based
upon its unpleasant smell or on its irritating/painful properties.

The object of Experiment 4 was to determine the influence of olfaction on fluid intake
when odor was present in the fluid to be consumed. The experimental design was as follows.
Starlings were tested for fluid intake in a standard no-choice (one-bottle) six hour drinking
assay. The experiment comprised two surgical conditions. (Surgery effect: sham or bilateral
olfactory nerve cut [BONG]), two repellent concentration conditions (Concentration/Group
effect: 14 mM and 28 mM), and two test periods (day/treatment effect: pretreatment = water
presentation and treatment = chemical repellent presentation). The confounding of time and
treatment was considered to be unimportant (Experiment 1), and any differences were as-
sumed to be attributable to the presence of repellent. Data were analyzed using a fixed
effects, 3-way analysis of variance with repeated measures on the day/treatment effect.

Sixteen starlings were selected at random from the group housing pool and randomly
assigned so that eight received a BONG and the other eight received a sham surgery.
Starlings were anesthetized with choral hydrate/penabarbitol at a dose of 2 ml/kg, intra-
peritoneally and placed in a head-holder. Surgery for BONG and sham treatments followed
procedures outlined in Glark and Mason (1987). Following surgery, starlings were housed
individually in cages (61 X 36 X 41 cm). Two weeks following surgery, birds were adapted
to experimental conditions. The two-week latency period was estimated to provide sufficient
time for degeneration of olfactory afferents into the olfactory bulb, yet not sufficiently long
to allow olfactory nerve regeneration (Wenzel and Salzman 1968). Following adaptation
one half of the birds within the BONG and sham surgery treatments was randomly assigned
to concentration groups, resulting in four birds per experimental cell. On the first day of
testing (pretreatment period), beginning at 09:30 h, consumption of tap water was recorded
every 2 h for the next 6 h. On the second day, starlings were presented with their preassigned
concentration of OAP and fluid intake was monitored every 2 h for the next 6 h. Fluid was
presented in 120 ml graduated Richter tubes. Starlings were visually isolated from one
another as well as from the contents of the Richter tubes. After five days rest, the same
birds used in the above experiment were re-randomized with respect to concentration group
assignment and tested with MA using the above protocol.

At the end of the experiments, birds were killed with pentobarbital and sequentially
perfused with physiological saline, 10% formalin and a 10% formalin and 30% sucrose
mixture. Brains were removed and the effect of the nerve section on olfactory bulb structure
was evaluated in a double blind sequence. Briefly, the anterior tips of the hemispheres
(containing the olfactory bulbs) were cut into 50 p-m slices. Every fifth slice was stained
with Nissl stain to highlight the outline of the glomeruli. The observer (familiar with normal
histological structure of avian olfactory bulbs) categorized the coded histological series as
being either in a degenerative state or normal. Goncordance of the observer s scoring and
the surgical status were compared using a chi-square analysis. Degeneration of the glomer-
ular structure was taken as evidence for lack of olfactory nerve input into the olfactory bulb
(Wenzel and Salzman 1968).

42

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

45

1 2 3

1 2 3

1 2 3

COE COE COE

10 -

5 -

0

2

Night

3

Eig. 1 . Average overnight water consumption as a function of time and water deprivation
schedule (Experiment 1). Open circles = control group (C); solid circles = OAP group (O);
Shaded circles = water excluded group (E). Horizontal bars depict statistically similar water
consumption (Scheffe s test, P > 0.05). Letters compare average intake across treatment
categories within a single night. Numerals compare average intake across nights within a
single treatment category. Vertical capped bars depict ± one standard error.

Zinc sulfate (ZnS04) also has been used to render birds anosmic. However, because ZnSOj
strips away the olfactory epithelium (Burd 1993), it may also affect epithelial layers con-
taining trigeminal free nerve endings. There are no data on this latter point. Because of this
uncertainty, surgical manipulation of the olfactory nerve was deemed to be the best way to
determine the role of olfaction in the avoidance response.

Directly eliminating trigeminal input is not feasible. In birds, only the ophthalmic branch
of the trigeminal nerve (OBTN) is easily accessible for denervation studies (Getty 1975).
Manipulation of the maxillary and mandibular branches would require too much destruction
of bone and muscle tissue. All three branches could be eliminated at the Gasserian ganglion
located in the eye orbit, but this would require permanent blinding of the bird.

RESULTS

Experiment I : Effects of water deprivation on fluid consumption. — The
overnight fluid intake for the treatment groups, adjusted for the covariate
of diurnal fluid intake, differed across days (Rao r = 4.55, P < 0.007).
The post-hoc analysis indicated that the overnight water consumption
within treatment groups was similar across nights (Fig. 1). However, water
consumption differed among treatments within nights. On the first night,
all birds consumed similar amounts of water. On the second and third

Clark • BIRD REPELLENTS

43

Fig. 2. Average diurnal water consumption as a function of treatment group and time
(Experiment 1). Open circles = control group; solid circles = OAP group; shaded circles
= water excluded group). Vertical capped bars depict ± one standard error.

nights the control birds drank less water than the birds that were physi-
cally excluded from water during the day. Overnight consumption for
birds presented with OAP did not differ from the other two experimental
groups. There was an inverse, albeit nonsignificant (P < 0.099), relation-
ship between daytime fluid intake and subsequent overnight water con-
sumption (Fig. 1). Birds experiencing fluid deprivation by exclusion tend-
ed to have the highest overnight water intake (36.0 ml ± 3.7 SE), fol-
lowed by birds experiencing deprivation by repellency (30.4 ml ± 3.6
SE). Control birds that had ad libitum access to water showed the lowest
overnight water consumption (24.3 ml ± 2.7 SE). Treatment category did
not appear to affect baseline diurnal water consumption immediately fol-
lowing the tests (Fig. 2). There were no group {P < 0.453), test period
{P < 0.889), or interaction (P < 0.362) effects.

Experiment 2. — The context of the stimulus is important for repellency.
With no prior experience, starlings presented with OAP in solution con-
sumed less fluid than the controls (P = 94.86, df = 1,30, P < 0.001).
This observation stands in contrast to the similarity of diurnal fluid intake
for the controls (w/w) and birds presented with water and OAP odor (w/
o) (P < 0.45). Thus, odor alone is not a sufficiently strong stimulus to

44

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Test Day 113232 131223 121323

Test Group 614325 265143 463521

Eig. 3. Average, diurnal fluid intake as a function of time and treatment (Experiment
2). Birds in the control group (w/w, open bars) were presented with tap water in the drinking
tubes and disks contained water soaked wicks, birds in the OAP group (o/w, solid bars)
were presented with a 28 mM solution of o-aminoacetophenone in the drinking tubes and
disks contained water soaked wicks, birds in the odor group (w/o, hatched bars) were pre-
sented with water filled drinking tubes and wick soaked in OAP. Average intake is ranked
within treatment type and nonsignificant {P > 0.05) differences among means are depicted
by the horizontal bar. Test Day and Test Group are the column and row label designations
of the two Latinized presentation designs described in Table 1 . Vertical capped bars depict
one standard error.

result in avoidance. This pattern persists even for experienced birds. On
days two and three the comparison w/w to w/o indicates similar levels of
fluid intake {P < 0.966 and 0.21, respectively), irrespective of the pre-
vious treatment exposure. The lack of an overall test sequence {P <
0.305) or day {P < 0.121) effect on diurnal fluid intake is apparent in
Fig. 3.

Starlings do not readily acquire a conditioned avoidance response to
the odor of OAP when orally delivered OAP is used as the unconditional
stimulus. Comparison by contrasts of w/w to w/o on the second day of
testing when both groups received o/w on the first day failed to uncover
differences in fluid consumption {P < 0.591).

As was the case for Experiment 1, the intervening three day treatment
schedule had no effect on baseline water consumption. There were no
group, day or interaction effects for the pre-test vs post-test comparison

Clark • BIRD REPELLENTS

45

E.

0)

ro

■4—'

_c

■TD

LL

Treatment Day

Fig. 4. Average fluid intake for OAP as a function of time and solution treatment (Ex-
periment 3). The left most open bar depicts intake of tap water paired with a water soaked
wick in the odor disc (w/w). The middle solid bars depict fluid intake of a 28 mM o-
aminoacetophenone (OAP) solution paired with a water soaked wick in the odor disc (o/
w). The right most hatched bars depict fluid intake of tap water paired with an OAP soaked
wick in the odor disc (w/o). Vertical capped bars depict one standard error.

of diurnal water consumption (P < 0.351, < 0.434, < 0.674, respec-
tively).

Experiment 3: The conditioned odor avoidance response. — There were
significant day/treatment effects for both the OAP and MA tests (Rao’s r
= 9.56, P < 0.024 and r = 12.48, P < 0.15, respectively). The planned
comparisons indicated that fluid consumption was similar for the w/w and
w/o treatment conditions for both OAP (P < 0.72, Fig. 4) and MA (P <
0.87, Fig. 5). Thus, three days exposure to orally administered repellent
was not sufficient for the formation of a conditioned avoidance response
to that odor.

Experiment 4: The role of olfaction. There was a strong day/treatment
effect for OAP (P = 74.57, df = 1,12, P < 0.001), showing that overall
fluid intake was suppressed when the starlings were presented with either
28 or 14 mM OAP solutions (Fig. 6). None of the other effects achieved
probability levels less than P = 0.15. The planned comparisons between
sham-OAP and BONC-OAP within each concentration showed that sur-
gery had no effect on avoidance of OAP (P < 0.619, 0.62, respectively).

46

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

35

30

^ 25

£

^ 20
ro

1

^ 10
5
0

0 1 2 3 4 5 6 7

Treatment Day

Fig. 5. Average fluid intake for MA as a function of time and solution treatment (Ex-
periment 3). The left most open bar depicts intake of tap water paired with a water soaked
wick in the odor disc (w/w). The middle solid bars depict fluid intake of a 28 mM methyl
anthranilate (MA) solution paired with a water soaked wick in the odor disc (o/w). The right
most hatched bars depicted fluid intake of tap water paired with a MA soaked wick in the
odor disc (w/o). Vertical capped bars depict one standard error.

Similar consumption patterns were observed for the MA trials. There
was a strong period effect for MA (F 83.14, df = 1,12, P < 0.001),
showing that fluid intake was suppressed across all surgery and concen-
tration levels when starlings were presented with MA solutions (Fig. 7).
No other effect achieved probability levels less than P = 0.22. The
planned comparisons between sham-MA and BONC-MA within each
concentration showed that surgery had no effect on avoidance of MA (P
< 0.215, 0.399, respectively).

Post-mortem visual inspection of the glomerular layer of sham surgery
birds showed a diffuse pattern seen in normal starlings, typical for the
spring breeding season. In contrast, BONC birds showed a complete break
down of the glomerular structure. The birds were euthanized 20—25 days
after surgery. The observer scoring the slides, but blind to the experi-
mental identity, categorized all BONC tissue as having degenerative glo-
meruli (N = 8) and all SHAM tissue as being normal (N = 8) =

16.0, df = 1, P < 0.001).

Clark • BIRD REPELLENTS

47

o-Aminoacetophenone

Sham BONG Sham BONG

28 mM 14 mM

Treatment

Eig. 6. Average fluid intake for OAP as a function of a solution’s molar concentration,
surgery and treatment period. Starlings were presented with tap water during the pretreat-
ment period (hatched bars) and an OAP solution during the treatment period (solid bars).
Capped vertical bars depict one standard error.

DISCUSSION

When deprived of water during the day, starlings drank more water
than normal during the overnight recovery period. However, there was
no difference between pre- and post-test diurnal water consumption, ir-
respective of the experimental water deprivation schedule (Experiment 1 ).
Together these data suggest that the effects of short-term water depriva-
tion are ameliorated during the 18-h recovery period. This lack of car-
ryover, i.e., day effect, is consistent with previously reported results for
similar experiments (Clark and Shah 1991).

The use of the two Latinized square treatment presentation sequences
allowed a detailed analysis of the influence of delivery route on the avoid-
ance response (Experiment 2). Naive starlings that were exposed to only
the odor of OAP did not avoid the drinking apparatus, suggesting that
the odor of OAP was not repellent. Only when OAP was allowed to come
in contact with the mouth did starlings show an avoidance response.

Starlings failed to avoid the odor of OAP even after they had come in

48

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

E

CD

B

■g

LL

Methyl Anthranilate

Sham BONG
28 mM

Sham BONG
14 mM

Treatment

Lig. 7. Average fluid intake for MA as a function of a solution’s molar concentration,
surgery, and treatment period. Starlings were presented with tap water during the pretreat-
ment period (hatched bars) and a MA solution during the treatment period (solid bars).
Capped vertical bars depict one standard error.

oral contact with the substance (Experiment 2). Moreover, the prolonged
exposure to repellents (OAP and MA) over several days failed to result
in a learned odor avoidance (Experiment 3). These findings suggest that
starlings do not readily learn to avoid solutions associated with the odor
of repellents. The failure to form a conditioned avoidance is independent
of the history of exposure to odors or repellent solutions (Experiment 2).
Experiment 4 showed the lack of importance of olfactory cues in avoiding
repellents. Surgical elimination of the olfactory nerves did not cause birds
to increase consumption of repellent-bearing solution as was expected if
olfaction was the mediating mechanism for avoidance.

The lack of odor avoidance and contribution of the olfactory system
in the avoidance response is consistent with previous speculation that
sensory repellents are mediated by the trigeminal system, although the
role of taste cannot be ruled out on the basis of these experiments alone.
However, there is other evidence consistent with involvement of the tri-
geminal nerve for perception of OAP and MA. Mason et al. (1989)

Clark • BIRD REPELLENTS

49

showed that bilateral section of the OBTN virtually eliminated the avoid-
ance response of starlings to anthranilates. There is also electrophysio-
logical evidence that the OBTN is highly responsive to OAP and MA
(Clark, unpubl. data). Molecular modelling evidence suggests birds have
a receptor mechanism that is responsive to anthranilates and acetophen-
ones and is analogous to the vanalloid pain receptors in mammals (Clark
and Shah 1994). Finally, even after central taste nuclei are ablated, chick-
ens continue to avoid MA (Benowitz 1964), suggesting that taste is not
critical in mediating the avoidance response.

The unsuitability of oral exposure to repellents as the unconditional
stimulus may be attributable to its mode of action and/or localization. The
trigeminal repellents tested appear to be unpalatable but do not cause
illness (at least for the quantities voluntarily consumed by birds). In ad-
dition, the aversive effect is localized in the mouth (in this experiment)
and not in the gastrointestinal system. Thus, repellents that are unpalatable
and cause oral stimulation, in combination or alone, may be ineffective
as unconditional stimuli relative to repellents that cause illness and gas-
trointestinal stimulation. This does not suggest that trigeminal repellents
are poorer repellents than associative repellents. Both can be potent and
result in strong avoidance behavior. The difference between the two re-
pellent types is that birds fail to learn from their aversive experience with
trigeminal repellents.

From a practical viewpoint, failure to learn about the sensory attributes
of a repellent is not crippling to the objective of bird repellency. Labo-
ratory and field studies show birds continuously sample small quantities
of substances treated with sensory repellents (e.g.. Mason et al. 1985,
Clark and Shah 1991, 1994; Avery and Decker 1994). While total
amounts consumed are small, how does the trigeminal repellent work to
repel birds away from an area? The effectiveness of the repellent is two-
fold. In the first case, the repellent has a direct effect on individuals. When
movement is not restricted, a bird encountering a sensory repellent quick-
ly moves on to a more palatable resource patch (Mason et al. 1985). This
observation is consistent with optimal foraging theory (Charnov 1976)
and the proposition that low rates of energy return in a patch will favor
the probability of abandoning that patch (Lima 1985). The repellent can
be viewed as signalling low energy return or physically achieving the
fact. Regardless, the outcome is the same. Second, as a consequence of
reduced residency time in a patch, the recruitment opportunities to that
patch are reduced. Thus, the number of birds that might visit the patch
because they observe other individuals in the patch is reduced (Krebs
1974).

The inability to learn to stay away from sensory cues associated with

50

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

trigeminal repellents has implications for the evolution of plant-bird in-
teractions as well. The behavior for free-ranging birds is to sample fruit
throughout its development, apparently giving more attention to the sa-
lient chemical cues associated with palatability rather than ancillary sen-
sory cues (Willson and Comet 1993). To prevent untimely frugivory, it
is to the plant’s advantage to employ transient protection to fruits and
have birds return to the fruits at a later time when the seeds are ready for
dispersal, rather than have birds form a strong conditioned avoidance
response (e.g., Brower 1969, Guarino et al. 1974, Mason 1989). These
observations provide intriguing interpretive possibilities about how chem-
ical defenses of fruit should be structured to exploit the sensory systems
and learning capabilities of birds. Chemical defenses in unripe bird dis-
persed fruit should consist of trigeminal repellents and not compounds
that cause gastro-intestinal illness. Finally, a better understanding of how
plants prevent untimely frugivory will prove invaluable in the design of
nonlethal repellents for safe wildlife management strategies.

ACKNOWLEDGMENTS

I thank D. Coleman for assistance in the laboratory. C. A. Smeraski kindly performed
the histological preparations. This study was supported by the United States Dept, of Ag-
riculture cooperative agreement 12-34-41-0040 between the Monell Chemical Senses Center
and the Denver Wildlife Research Center. All experimental procedures meet guidelines set
forth by Monell’s institutional animal care committee.

LITERATURE CITED

Agree, T. E., E. H. Lavin, R. Nishida, and S. Watanabe. 1990. The serendipitous dis-
covery of ortho-aminoacetophenone as the foxy smelling component of labruscana
grapes. Chem. Eng. News 9:80.

Avery, M. L. and D. G. Decker. 1994. Responses of captive Fish Crows to eggs treated
with chemical repellents. J. Wildl. Manage. 58:261-266.

Benowitz, L. 1972. Effects of frebrain ablations on avoidance learning in chickens. Physiol.
Behav. 9:601-608.

Brower, L. P. 1969. Ecological chemistry. Sci. Am. 220:22-29.

Burd, G. D. 1993. Morphological study of the effects of intranasal zinc sulfate irrigation
on the mouse olfactory epithelium and olfactory bulb. Micros. Res. Tech. 24:195-213.
Charnov, E. L. 1976. Optimal foraging: the marginal value theorem. Theor. Popul. Biol.
9:129-136.

Clark, L. 1995. Modulation of avian responsiveness to chemical irritants: effects of pros-
taglandin El and analgesics. J. Exp. Zool. 271:432-4^40.

AND J. R. Mason. 1987. Olfactory discrimination of plant volatiles by the European

starling. Anim. Behav. 35:227-235.

AND . 1993. Interactions between sensory and postingestional repellents in

starlings: methyl anthranilate and sucrose. Ecol. Appl. 3:262-270.

AND P. S. Shah. 1991. Nonlethal bird repellents: in search of a general model

relating repellency and chemical structure. J. Wildl. Manage 55:538-545.

Clark • BIRD REPELLENTS

51

AND . 1993. Chemical bird repellents: possible use in cyanide ponds. J.

Wildl. Manage. 57:657-664.

AND . 1994. Tests and refinements of a general structure-activity model for

avian repellents. J. Chem. Ecol. 20:321—339.

, , and j. R. Mason. 1991. Chemical repellency in birds: relationship be-
tween chemical structure and avoidance response. J. Exp. Zool. 260:310-322.

AND C. A. Smeraski. 1990. Seasonal shifts in odor acuity by starlings. J. Exp. Zool.

177:673-680.

Crocker, D. R. and S. M. Perry. 1990. Plant chemistry and bird repellents. Ibis 132:300-
308.

Dolbeer, R. a. 1986. Current status and potential of nonlethal means of reducing bird
damage in agriculture. Pp. 474-483 in ACTA XIX Congresses Internationalis Orni-
thologici. Vol. 1 (H. Ouellet, ed.). Univ. of Ottawa Press, Ottawa, Ontario, Canada.

Garcia, .1. R., R. Kovner, and K. E Green. 1966. Cue properties vs. palatability of flavors
in avoidance learning. Psychonomic Sci. 20:313-314.

Getty, R. 1975. Sisson and Grossman’s The anatomy of the domestic animals. Vol. 2. Pp.
2022-2037. W. B. Saunders Company, Philadelphia, Pennsylvania.

Green, B. G., J. R. Mason, and M. R. Kare. 1990. Chemical senses, Vol. 2, Irritation.
Marcel Dekker, New York.

Guarino, j. L., W. E Shake, and E. W. Schafer. 1974. Reducing bird damage to ripening
cherries with methiocarb. J. Wildl. Manage. 38:338—342.

Krebs, J. R. 1974. Colonial nesting and social feeding as strategies for exploiting food
resources in the Great Blue Heron (Ardea herodias). Behaviour 51:99-134.

Lima, S. L. 1985. Sampling behavior of starlings foraging in simple patchy environments.
Behav. Ecol. Sociobiol. 16:135—142.

Mason, J. R. 1989. Avoidance of methiocarb-poisoned apples by red-winged blackbirds.
J. Wildl. Manage. 53:836-840.

AND L. Clark. 1992. Nonlethal repellents: the development of cost-effective, prac-
tical solutions to agricultural and industrial problems. Proc. Vert. Pest Conf. 15:115-
129.

AND R. E Reidinger. 1983. Importance of color for methiocarb-induced food aver-
sions in red-winged blackbirds (Agelaius phoencieus). J. Wildl. Manage. 47:383-393.

, M. A. Adams, and L. Clark. 1989. Anthranilate repellency to starlings: chemical

correlates and sensory perception. J. Wildl. Manage. 53:55-64.

, L. Clark, and P. S. Shah. 1991. Ortho-ami noacetophenone repellency to birds:

similarities to methyl anthranilate. J. Wildl. Manage. 55:334—340.

, J. F. Glahn, R. a. Dolbeer, and R. F. Reidinger. 1985. Field evaluation of

dimethyl anthranilate as a bird repellent livestock feed additive. J. Wildl. Manage. 49:
636-642.

Pelchat, M. L., H. j. Grill, P. Rozin, and J. Jacobs. 1983. Quality of acquired respon.ses
to tastes by Rattus norvegicus depends on type of associated discomfort. J. Comp.
Psychol. 97:140-153.

Schuler, W. 1983. Respon.ses to sugars and their behavioural mechanisms in the starling
{Stunms vulgaris L.). Behav. Ecol. Sociobiol. 13:243-251.

Statistica®. 1994. StatSoft, Inc., Tulsa, Oklahoma.

Walker, J. C., D. B. Walker, C. R. Tambiah, and K. S. Gilmore. 1986. Olfactory and
nonolfactory odor detection in pigeons: elucidation by a cardiac acceleration paradigm.
Physiol. Behav. 38:575—580.

Wenzel, B. M. and A. Salzman. 1968. Olfactory bulb ablation or nerve section and
behavior of pigeons in nonolfactory learning. Exper. Neurol. 22:472-^79.

52

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Willson, M. E and T A. Comet. 1993. Eood choices by Northwestern Crows: experiments
with captive, free-ranging and hand-raised birds. Condor 95:596—615.

Zahorik, D. M. 1976. Associative and non-associative factors in learned food preferences.
Pp. 181-200 in Learning mechanisms in food selection (L. M. Barker, M. R. Best, and
M. Domjan, eds.). Baylor Univ. Press, Waco, Texas.

Wilson Bull., 108(1), 1996, pp. 53-60

NEST-SITE SELECTION BY HOODED WARBLERS IN
BOTTOMLAND HARDWOODS OF SOUTH CAROLINA

John C. Kilgo, Robert A. Sargent,

Brian R. Chapman, and Karl V. Miller

Abstract. — We measured habitat features at 45 nests of Hooded Warblers (Wilsonia
citrina) and 45 non-use sites in bottomland hardwood habitats in the coastal plain of South
Carolina during the breeding seasons 1993—1994 to determine features that affect nesting
success. Hooded Warblers nested in switchcane {Arundinaria gigantea) and hardwood sap-
lings or shrubs that averaged 1.76 ± 0.10 m {SE) in height. Nests were more concealed
from above (P = 0.001) and from the side (P = 0.002) than surrogate nests placed at non-
use sites but were less concealed from below (P = 0.002). Nest sites also had a greater
number of potential substrates (P = 0.014) in the nest patch (5-m radius) and greater mea-
sures of vegetation density (P < 0.05) in the nest patch than non-use sites. Successful nests
differed from unsuccessful nests only in the amount of fern cover in the nest patch (greater
for successful nests, P = 0.012). Fern cover may influence nesting success through an effect
on behavioral defense strategies. Nesting success of Hooded Warblers may largely be un-
related to fine-scale differences in vegetative characteristics of the nest site. Received 28
Mar. 1995, accepted 20 Sept. 1995.

Because availability of suitable nest sites may be the most critical de-
terminant of habitat selection (and thus perceived habitat quality) by some
birds (Steele 1993), knowledge of what constitutes a suitable nest site, or
more importantly a successful nest site, is necessary (Martin 1993a). For
example, Martin and Roper (1988) found that successful Hermit Thrush
(Catharus guttatus) nests were characterized by a greater density of white
fir (Abies concolor) saplings in the 5-m radius circle surrounding the nest.
Such specific habitat features that affect nest fate should be identified for
other species.

Hooded Warblers (Wilsonia citrina) have been classified by the Part-
ners In Flight prioritization scheme as a species of “very high concern”
in the Southeast (Hunter et al. 1993a, b). We examined nest-site selection
patterns of Hooded Warblers to determine habitat differences between
successful and unsuccessful nests. We measured variables at two scales,
the nest site and the nest patch. Hooded Warblers inhabit moist mature
deciduous forests of eastern North America (Bent 1953, Powell and Rap-
pole 1986, Evans Ogden and Stutchbury 1994). In the coastal plain of
the southeastern United States, Hooded Warblers occur almost exclusively
in forested wetlands (Bent 1953) and reach their greatest abundance in
bottomland hardwood forests (Oak-Gum-Cypress [Quercus-Nyssa-Taxo-
dium] association).

Daniel B. Warnell School of Forest Resources, The Univ. of Georgia, Athens, Georgia 30602-2152.

53

54

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

STUDY AREA AND METHODS

We conducted this study at the U.S. Dept, of Energy’s Savannah River Site (SRS), a
78,000-ha tract in Aiken, Barnwell, and Allendale counties. South Carolina. These counties
lie in the Upper Coastal Plain physiographic province. Elevation ranges from <25 m at the
Savannah River to 80 m at headwater streams (Workman and McLeod 1990). Bottomland
hardwood forests are found along stream courses and may be seasonally flooded, usually
during late winter-early spring. Dominant canopy species include sweetgum {Liquidambar
styraciflua), swamp tupelo {Nyssa sylvatica var. biflora), red maple {Acer rubrum), water
oak {Quercus nigra), and diamond-leaf oak {Q. laurifolia). Dominants in the mid-story
include American holly, (Ilex opaca), sweet bay {Magnolia virginiana), red bay {Persea
borbonia), and ironwood {Carpinus Carolina). Switchcane {Arundinaria gigantea) and dog
hobble {Leucothoe axillaris) dominate the shrub layer, and ferns, primarily netted chain fern
{Woodwardia areolata) and Christmas fern {Polystichum acrostichoides), are the dominant
ground cover (Workman and McLeod 1990). Bottomland study sites ranged in width from
<50->1000 m and were adjacent to closed-canopy pine {Pinas elliottii and P. palustris)
forest.

We located Hooded Warbler nests in 1 1 bottomland hardwood strips during May-June
1993 and 1994 by observing adult behavior and by searching potential nesting habitat. We
found most nests during the incubation stage. We were unable to determine whether nests
were first or second attempts because individual birds were not marked and territories were
not mapped. We monitored nests at 3^ day intervals (Ralph et al. 1993) to determine nest
fate. Nests containing nestlings on the last visit before the expected fledging date were
assumed to have fledged. We defined successful nests as those that fledged at least one
nestling. Vegetation measurements were made following termination of the nesting attempt.
We made measurements at the nest plant and in the nest patch, defined as the 5-m radius
circle centered on the nest plant (Martin and Roper 1988). Vegetation measurements then
were repeated at an unused site. We located non-use sites by pacing 35 m (Ralph et al.
1993) upstream or downstream (determined by coin toss) in a direction parallel to the general
bearing of the bottomland strip. This procedure located non-use sites outside of the nest
patch but within the bottomland habitat. Non-use sites were centered on the plant stem
nearest to the 35-m point that was of the same species and approximate size as the substrate
plant (Ralph et al. 1993). Thus, equal numbers of nest sites and non-use sites were sampled.
Success data were obtained from 36 nests, 15 nests in 1993 (8 successful, 7 unsuccessful)
and from 21 nests in 1994 (10 successful, 1 1 unsuccessful). Eight additional nests in 1993
and one nest in 1994 that were empty when found were sampled and included in the
comparison of nest sites versus non-use sites but not in the analyses relating to nest success
(Martin and Roper 1988).

Measurements taken at the nest site included plant species used as the nesting substrate,
nest height, plant height, and percentage of nest concealment. Concealment indices (0-4: 0
= 0% concealed, 1 = 1-25% concealed, 2 = 26-50% concealed, etc.) were estimated by
viewing the nest from above and below and at nest level from a distance of 1 m in each of
the four cardinal directions (Martin and Roper 1988, Hoi way 1991 ). Lor concealment esti-
mates at non-use sites, an empty Hooded Warbler nest was placed at nest-height (i.e., the
height of the nest corresponding to the non-use site) in the suiTogate substrate plant (Holway
1991).

Measurements taken in the nest patch included overstory canopy cover, stem density of
potential nest substrates and trees, fern cover, other herbaceous ground cover, and vegetation
profile. Canopy cover above the patch was estimated by five hit-miss readings through an
ocular tube (James and Shugart 1970), one at the nest plant and one in each of the cardinal

Kilgo et al. • HOODED WARBLER NEST SITES

55

directions from the perimeter of the patch. Potential substrate and tree (woody stems > 3
m tall) densities were measured in five 1-m^ quadrats located randomly along the four
cardinal directions (transect and position on transect were randomized). Potential substrates
were defined as switchcane >1 m tall and other woody species 1—3 m tall. Percent foliage
cover of ferns and of other herbaceous ground cover also was estimated (0-4) within the
quadrats. Vegetation profile of the patch, which may be viewed as an index of concealment
at the scale of the patch, was determined using a 3-m tall vegetation profile board (Nudds
1977, Noon 1981) against which percentage cover was estimated (0^) for each 0.5-m
interval. The profile board was located at the nest plant and was read from a distance of 5
m in each of the cardinal directions.

For comparisons involving potential substrate density, nests were classified as either
“switchcane” or “other”, depending on the species of their actual substrate. Stem density
of switchcane then was determined for switchcane nests and of other for other nests. Thus,
the potential substrate variable was a nest-specific measurement which circumvented the
problem, e.g., of comparing average switchcane density across all nests when only a portion
of nests were in switchcane. Similarly, one vegetation profile measurement (nest level) was
nest-specific. The profile readings for the 0.5-m interval corresponding to the height of each
nest were compiled for the variable vegetation profile at nest level.

Univariate comparisons were made between Hooded Warbler nest sites and non-use sites
and between successful and unsuccessful nest sites for each habitat variable. Variables es-
timated with the 0^ index were compared with a Wilcoxon rank-sum test. All other com-
parisons were made with a two-sample f-test. Variances were assumed to be equal for
comparisons in which the sample sizes were the same. When sample sizes differed, the F-
test for equality of variance was used to test the equal variance assumption. Equal variance
tests always were appropriate. Because no differences (P > 0.05) were found between years
for any variables, data from both years were pooled.

RESULTS

Hooded Warblers selected saplings of nine different species as nest
substrates: switchcane, 20 (44%); red bay, 7 (16%); common gallberry,
5 (11%); American holly, 5 (11%); water oak, 2 (4%); diamond leaf oak,
2 (4%); blueberry (Vaccinium spp.), 2 (4%); wax myrtle {Myrica ceri-
fera), 1 (2%); and black oak (Quercus velutina), 1 (2%). Mean height of
the nest plant was 1.76 ± 0.10 m (SE). With one exception, in which the
nest was located in an upright branch of an American holly, nests were
placed in crotches of the main stem and primary branches of the substrate
plant. Nest height averaged 0.98 ± 0.36 m.

Hooded Warbler nest sites differed from non-use sites in several ways.
Concealment of nests from above and from the side was greater {P <
0.005) at nest sites than at non-use sites, but from below was lower (f
= 0.002) at nest sites than at non-use sites (Table 1). Potential substrate
density was greater {P = 0.014) at nest sites (3.79 ± 3.37 stems/m^) than
at non-use sites (2.08 ± 1.88 stems/m^). Vegetation profile measures for
all heights and at nest level were greater {P < 0.05) at nests sites than at
non-use sites (Table 1). Conversely, only one difference was determined

56

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Comparison of Microhabitat Variables (x ± SE) at Hooded Warbler Nest sites (N
= 45) WITH Those at Random Sites (N = 45) within Bottomland Hardwoods,
Savannah River Site, South Carolina 1994-1994

Parameter

Nest site

Random site

Nest site

Nest height (m)

0.98

-h

0.05

1,01

-1-

0.05

Plant height (m)

1.76

-h

0.10

1.78

-h

0.14

Concealment

Side

1.93

0.10

1.50

4-

0.08

0.002

Above

3.36

-i-

0.13

2.40

4-

0.21

0.001

Below

1.58

-+-

0.16

2.09

4-

0.14

0.002

Nest patch

Canopy coveC

4.36

-+-

0.10

4.49

4-

0.10

0.360

Potential substrate density^

3.79

-+-

0.60

2.08

4-

0.33

0.014

Pern cover“

1.19

-+-

0.12

1.31

4-

0.16

0.842

Ground cover“

1.48

-+-

0.10

1.69

4-

0.12

0.268

Vegetation profile'"

0.0— 0.5 m

3.07

-+-

0.10

2.68

4-

0.13

0.025

0.5— 1.0 m

3.39

-+-

0.10

2.83

4“

0.14

0.000

1.0-1. 5 m

3.12

-H

0.09

2.36

4-

0.14

0.000

1. 5-2.0 m

2.94

-+■

0.13

2.14

4-

0.14

0.000

2.0-2.5 m

2.53

4-

0.14

2.02

4-

0.22

0.002

2. 5-3.0 m

2.30

-i-

0.14

1.79

4-

0.13

0.007

Mean

2.88

-h

0.08

2.31

4-

0.10

0.000

Nest level

3.47

-h

0.08

2.86

4-

0.11

0.000

“ Index of percent coverage: 0 = 0%, 1

= 1-25%, 2 =

26-50%. 3

= 51-75%, 4

=

76-100%.

Index values were

compared with the Wilcoxon rank-sum test; all other comparisons were made with a two-sample r-test.

^ Estimated as the sum of five hit-miss readings taken within the patch (5 = total canopy closure).

Stem density (# stems/m^) of the plant species used as substrate. Substrates were categorized as switchcane or other
(woody stems 1. 0-3.0 m). Sample includes switchcane nests from 1993 and 1994 (N = 20) and other nests from 1994 (N
— 12; total sample = 32).

for the comparison between successful and unsuccessful nests: fern cover
was greater (P — 0.012) around successful nests.

DISCUSSION

Nest-site characteristics. — Switchcane best provides the structural fea-
tures of a nest substrate sought by Hooded Warblers of all plant species
occurring in bottomland hardwood habitats at SRS. It apparently is the
preferred substrate species throughout much of the southeastern United
States (Sprunt and Chamberlain 1949, Burleigh 1958). Switchcane is a
woody grass (Poaceae) that may grow to 10 m (Radford et al. 1964) but
normally ranges from 1-3 m. It commonly forms extensive thickets, or
canebrakes, in southeastern swamps. In addition to the nests that were in

Kilgo et al. • HOODED WARBLER NEST SITES

57

switchcane, many of the other nests were in saplings growing in cane-
brakes (i.e., switchcane provided most of the cover for most nests).

A variety of other plant species also were used as nest substrates. Hood-
ed Warblers reportedly prefer mountain laurel (Kalmia latifolia), Ameri-
can holly, and fetterbush {Lyonia spp.) in other parts of their range (Bent
1953). All of these species are thicket-forming shrubs (during the sapling
stage for American holly). Thus, Hooded Warblers may select shrubs,
regardless of species, not only for their microsite characteristics but also
for their thicket-forming properties, as evidenced by the greater density
of potential substrates within the nest patch than at non-use sites. Holway
(1991) found that species preference by another shrub-nesting warbler,
the Black-throated Blue Warbler (Dendroica caerulescens), also was site-
specific; they selected the understory shrub that offered the best protection
from weather and predators.

Hooded Warblers selected nest sites that were less concealed from be-
low than nests at non-use sites. Bent (1953, p. 613) quotes one author
who said that “the easiest way to locate a [Hooded Warbler] nest was to
place [his] head close to the ground, scan the low open spaces and look
for a clump of leaves, which sooner or later proved to be a nest.” The
adaptive advantage of an opening immediately below the nest is unclear,
though it may be related to escape strategies. Although Hooded Warblers
normally do not approach or leave the nest near the ground (Odum 1931),
when flushed, the female often drops from the nest straight to the ground
before flying away just above the ground for a short distance (J. C. Kilgo,
pers. obs.; Evans Ogden and Stutchbury 1994). Alternatively, such open-
ings may result simply from the greater shading provided by the under-
story.

Murphy (1983) and Martin (1992, 1993a) have suggested that preda-
tion, because it is the primary cause of nest failure, should be the key
factor influencing nest-site selection. Selection of nest sites with dense
vegetation theoretically can inhibit predator efficiency by visually screen-
ing the nest and parent activity, by providing too many potential nest sites
for the predator to search, and by physically impeding predators (Holway
1991). Our results indicate that Hooded Warblers may utilize each of these
strategies in their selection of nest sites. Hooded Warblers selected nest
sites that were better concealed from the side and from above than non-
use sites. Furthermore, nest patches contained a greater density of poten-
tial substrates and denser vegetation profiles at all heights than non-use
patches.

Effect of nest-site characteristics on success. — We detected no differ-
ence in concealment from any angle between successful and unsuccessful
nests. Similarly, Howlett and Stutchberry (in press) detected no effect of

58

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

nest concealment on predation of Hooded Warbler nests in Pennsylvania,
and Holway (1991) was unable to detect a relationship between conceal-
ment and the nest success of Black-throated Blue Warblers. We also failed
to detect a difference in the number of potential substrates between suc-
cessful and failed nests, as predicted by Martin and Roper (1988) and
Martin (1993a). Several factors may make statistical distinction of these
subtle habitat features difficult. First, Holway (1991) suggested that pred-
ators using olfactory cues would be less inhibited by visual concealment.
Furthermore, nest predation sometimes can be random, with some nests
being found by chance alone. Second, human visitation of nest sites dur-
ing monitoring may have increased the likelihood of predation, and thus
masked any effects of habitat on predation (Westmoreland and Best 1985,
Martin 1992). Finally, all nests of shrub-nesting woodland birds should
be concealed because selection of poorly concealed sites should be elim-
inated by natural selection (Wray and Whitmore 1979). The latter predic-
tion is contradicted by several authors who have detected differences in
concealment between successful and failed nests (e.g., Nolan 1978, Wray
and Whitmore 1979, Martin and Roper 1988). However, Wray and Whit-
more (1979), suggest that the apparently nonadaptive trait to select poorly
concealed nest sites may be maintained in Vesper Sparrows (Pooecetes
gramineus) because annual variation in their environment may permit the
occupancy of a variety of nest sites to be adaptive over time. Although
such temporal variation is probably great in the early successional habitats
of Vesper Sparrows, the environments of mature forests are relatively
stable. In addition, nest predation generally is higher in shrub and grass-
land habitats than in mature forests (Martin 1993b). Thus, a relationship
between concealment and success should not be as evident in forested
habitat. Studies of woodland shrub-nesting passerines support this con-
tention (Best and Stauffer 1980, Conner et al. 1986, Holway 1991, How-
lett and Stutchbury, in press; but see Martin and Roper 1988), whereas
results of studies of birds in earlier successional habitats are more variable
(Caccamise 1977, Best 1978, Nolan 1978, Wray and Whitmore 1979).
Much of the predation on shrub-nesting woodland birds may largely be
unrelated to fine-scale differences in concealment (Holway 1991).

The difference in fern cover between successful and failed nests is
intriguing. This finding may be related to nest-defense strategies. Female
Hooded Warblers almost invariably drop to the ground when flushed from
the nest, and rather than flying away, they often engage in a distraction
display, which consists of running through the underbrush with wings
drooped and tail spread (J. C. Kilgo, pers. obs.; Evans Ogden and Stutch-
bury 1994). This behavior likely is their primary (if not only) means of
nest defense. If insufficient ground cover exists in the patch to make this

Kilgo et al. • HOODED WARBLER NEST SITES

59

technique effective (i.e., if the bird must itself escape and is not able to
risk distracting the predator) the nest may be rendered more susceptible
to predation. Ferns may provide structure that conceals the displaying
female yet is sufficiently open to allow the predator to detect her. Thus,
degree of fern cover may be one of the subtle habitat features that deter-
mines nest fate of Hooded Warblers. This may also explain why the more
obvious measures of concealment and vegetation density did not differ
between successful and failed nests.

ACKNOWLEDGMENTS

This study was funded by the United States Dept, of Energy, Savannah River Site, the
United States Forest Service, Savannah River Forest Station Biodiversity Program, the Univ.
of Georgia, and Mclntire-Stennis Project No. GEO-0074-MS. J. Blake provided logistical
support. We thank the many field assistants who helped in locating and monitoring nests
and measuring vegetation. We thank G. W. Ware for providing statistical advice and A. S.
Johnson, K. C. Parker, R. J. Warren, and D. H. White for reviewing the manuscript. L. B.
Best and B. J. Stutchbury also provided valuable editorial comments.

LITERATURE CITED

Bent, A. C. 1953. Life histories of North American wood warblers. Bull. 203, U. S.
National Museum, Washington, D.C.

Best, L. B. 1978. Field Sparrow reproductive success and nesting ecology. Auk 95:9-22.
AND D. E Stauffer. 1980. Factors affecting nesting success in riparian bird com-
munities. Condor 82:149-158.

Burleigh, T. D. 1958. Georgia birds. Univ. of Oklahoma Press, Norman, Oklahoma.
Caccamise, D. E 1977. Nesting success and nest site characteristics in the Red-winged
Blackbird. Wilson Bull. 89:396—403.

Conner, R. N., M. E. Anderson, and J. G. Dickson. 1986. Relationships among temtory
size, habitat, song, and nesting success of Northern Cardinals. Auk 103:23-31.

Evans Ogden, L. J. and B. J. Stutchbury. 1994. Hooded Warbler {Wilsonia citrina). In
The birds of North America, No. 1 10 (A. Poole and F. Gill, eds.). The Acad. Nat. Sci.
of Philadelphia, Philadelphia, Pennsylvania; The American Ornithologists’ Union,
Washington, D.C.

Holway, D. a. 1991. Nest-site selection and the importance of nest concealment in the
Black-throated Blue Warbler. Condor 93:575-581.

Howlett, j. S. and B. j. Stutchbury. Nest concealment and predation in Hooded Warblers:

experimental removal of nest cover. Auk (in press).

Hunter, W. C., M. E Carter, D. N. Pashley, and K. Barker. 1993a. The Partners in
Flight prioritization scheme. Pp. 109-1 19 in Status and management of Neotropical
migratory birds (D. M. Finch and P. W. Stangel, eds.). U.S.D.A. For. Ser. Gen. Tech.
Rep. RM-229.

Hunter, W. C., D. N. Pashley, and R. E. F. Escano. 1993b. Neotropical migratory landbird
species and their habitats of special concern within the southeast region. Pp. 159-171
in Status and management of Neotropical migratory birds (D. M. Finch and P. W.
Stangel, eds.). U.S.D.A. For. Ser. Gen. Tech. Rep. RM-229.

James, F. C. and H. H. Shugart. 1970. A quantitative method of habitat description.
Audubon Field Notes 24:727—736.

60

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Martin, T. E. 1992. Breeding productivity considerations: what are the appropriate habitat
features for management? Pp. 455^73 in Ecology and conservation of Neotropical
migrant landbirds (J. M. Hagan and D. W. Johnston, eds.). Smithsonian Institution Press,
Washington, D.C.

. 1993a. Nest predation, nest sites, and birds: new perspectives on old patterns.

BioScience 43:523-532.

. 1993b. Nest predation among vegetation layers and habitat types: revising the

dogmas. Am. Nat. 141:897-913.

AND J. J. Roper. 1988. Nest predation and nest-site selection of a western population

of Hermit Thrush. Condor 90:51—57.

Murphy, M. T. 1983. Nest success and nesting habits of Eastern Kingbirds and other
flycatchers. Condor 85:208-219.

Nolan, V. 1978. The ecology and behavior of the Prairie Warbler Dendroica discolor.
Ornithol. Monogr. 26.

Noon, B. R. 1981. Techniques for sampling avian habitats. Pp. 42-51 in The use of mul-
tivariate statistics in studies of wildlife habitat (D. E. Capen, ed.). USDA Eor. Serv.
Gen. Tech. Rep. RM-87.

Nudds, T. D. 1977. Quantifying the vegetative structure of wildlife cover. Wildl. Soc. Bull.
5:113-117.

Odum, E. P. 1931. Notes on the nesting habits of the Hooded Warbler. Wilson Bull. 43:
316-317.

Powell, G. V. N. and J. H. Rappole. 1986. The Hooded Warbler. Pp. 827—853 in Audubon
wildlife report 1986 (R. L. Di Silvestro, ed.). National Audubon Society, New York,
New York.

Radford, A. E., H. E. Ahles, and C. R. Bell. 1964. Manual of the vascular flora of the
Carolinas. Univ. of North Carolina Press, Chapel Hill, North Carolina.

Ralph, C. J., G. R. Guepel, P. Pyle, T. E. Martin, and D. F. DeSante. 1993. Handbook
of field methods for monitoring landbirds. USDA For. Serv. Gen. Tech. Rep. PSW-144.

Sprunt, a. and E. B. Chamberlain. 1949. South Carolina bird life. Univ. of South Carolina
Press, Columbia, South Carolina.

Steele, B. B. 1993. Selection of foraging and nesting sites by Black-throated Blue War-
blers: their relative influence on habitat choice. Condor 95:568-579.

Westmoreland, D. and L. B. Best. 1985. The effect of disturbance on Mourning Dove
nesting success. Auk 102:774—780.

Workman, S. W. and K. W. McLeod. 1990. Vegetation of the Savannah River Site: major
community types. Publication SRO-NERP-19, National Environmental Research Park
Program. Savannah River Ecology Laboratory, Aiken, South Carolina.

Wray, T. and R. C. Whitmore. 1979. Effects of vegetation on nesting success of Vesper
Sparrows. Auk 96:802-805.

Wilson Bull., 108(1), 1996, pp. 61-71

CHANGE IN BODY MASS OF FEMALE COMMON
GOLDENEYES DURING NESTING AND
BROOD REARING

Michael C. Zicus' and Michael R. Riggs^

Abstract. — We measured body mass of female Common Goldeneyes {Bucephala clan-
giila) during nesting on fish bearing lakes in northcentral Minnesota, in 1982-1985. Median
body mass during egg laying was 775 g. Female mass during incubation varied among lakes
and possibly years. Mass at the start of incubation (698—715 g) was 10.7—11.0% greater
than that at hatching. Females regained most of the mass lost during incubation by the time
they abandoned their class IIC or class III ducklings. Goldeneyes in Minnesota weighed
less at the start of nesting than those studied on predominately fishless Ontario lakes; pro-
portional mass loss during incubation was also substantially less than that reported in Ontario
(approximately 20%). Differences in body mass dynamics may be related to the relative
ease of food acquisition during nesting; foods might be acquired more easily in more pro-
ductive wetlands despite the presence of fish. Received 24 Feb. 1995, accepted 1 Sept. 1995.

Relationships among incubation behavior, female body mass, and types
of nutrients and energy sources used by temperate nesting waterfowl are
understood reasonably well. In general, species that begin nesting earlier
have greater body mass, forage relatively less while nesting, rely more
on endogenous resources, and lose proportionately more mass during in-
cubation than do later nesting species (see review in Afton and Paulus
1992). Common Goldeneyes {Bucephala clangula) deviate somewhat
from this pattern. Although they are relatively small-bodied, females be-
gin nesting soon after arrival when many wetlands are still ice-covered.
Foraging territories also are defended vigorously during laying and early
incubation (Savard 1984, Zicus and Hennes 1993). In addition, laying
rates are low compared to other similar-sized waterfowl (cf Palmer 1976),
and clutch mass can exceed female mass (Zicus, unpubl. data). These
traits suggest that although females arrive with some stored reserves, ex-
ogenous nutrient sources may be important for clutch completion and
female maintenance during incubation.

Mallory and Weatherhead (1993) recently predicted that female Com-
mon Goldeneyes lose approximately 18.5% of their body mass during
incubation. Their prediction was based on relationships proposed by Af-
ton and Paulus ( 1 992) and appeared to be supported by data from an
Ontario study where wetlands had been influenced extensively by acid

‘ Minnesota Dept, of Natural Resources, Wetland Wildlife Populations and Research Group, 102 2.3rd
St., Bemidji, Minnesota 56601.

^ Minnesota Dept, of Natural Resources, Wildlife Populations and Research Unit, 500 Lafayette Road,
St. Paul, Minnesota 55155.

61

62

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

deposition (Mallory et al. 1994). Harvey et al. (1989) concluded that
Wood Duck {Aix sponsa) body mass dynamics during incubation varied
substantially among and within individuals. They speculated that vari-
ability in Wood Duck incubation mass was related in part to fluctuating
environmental conditions. Unfortunately, few published data relate local
or regional environmental factors to body mass changes in other species
of incubating waterfowl.

We describe body mass changes during nesting and brood rearing for
female Common Goldeneyes in Minnesota. We examined the effects of
different lakes and years and of changing reproductive stage on female
body mass. Goldeneyes have been studied previously in areas where fish
were absent or where goldeneyes appeared to favor fishless wetlands
thereby avoiding dietary competition with fish during the reproductive
period (Eriksson 1979, Eadie and Keast 1982, Blancher et al. 1992, and
others). Whereas many lakes in Mallory and Weatherhead’s 1993 study
area were fishless (Mallory et al. 1994), each of our study lakes supported
fish communities. Thus, our data should improve the understanding of
goldeneye nesting biology where they commonly occupy fish-bearing
lakes.

STUDY AREA AND METHODS

Lemale Common Goldeneyes were weighed during nesting on three lakes in Beltrami and
Itasca counties of northcentral Minnesota. Refuge Pond, North Twin Lake, and Island Lake
differed in size (46, 117, and 1250 ha, respectively), amount and type of public use, and
use by goldeneyes (Zicus et al. 1995). The two larger lakes supported fish populations
(dominated by Centrarchidae, Percidae and Esocidae) and were characterized by morpho-
edaphic indices (MEI) (Ryder 1965) of 6.75 and 18.01, respectively (Minn. Dept. Nat.
Resour., Section of Lisheries, unpubl. data). These values are near optimum for highly
productive fish communities (Ryder et al. 1974). The smallest lake supported only minnows.
Lemales with hatched young were captured and weighed on Refuge Pond and Island and
North Twin Lakes as well as on 10 additional lakes, each of which supported productive
fish populations with species composition similar to those in the lakes where nesting ducks
were weighed. These additional lakes had MEIs ranging from 4.52 to 21.78 (Minn. Dept.
Nat. Resour., Section of Lisheries, unpubl. data).

Lemale Common Goldeneyes were captured before incubation began with nest traps (Zi-
cus 1989) and again when we inspected nest boxes for use. Lemales were leg banded with
U.S. Lish and Wildlife Service bands. Incubating female mass was measured when possible
during weekly nest checks, and females accompanying broods were weighed when they
were caught nest prospecting (Zicus and Hennes 1989) and during annual leg banding
(Johnson 1972). Mass was determined to the nearest 5 g using spring scales and was un-
adjusted for female structural size.

Reproductive stage for egg-laying females was defined relative to the start of incubation,
and that of incubating and brood-rearing females was referenced to the departure of young
from nests. This differs from the convention often used for incubating females. However,
we believe it is preferable because it allows corresponding days to be compared more
appropriately. Reproductive stage of females with young from unmonitored nests was esti-

Zicus and Riggs • GOLDENEYE BODY MASS

63

mated from the age of the majority of the ducklings in the brood. Duckling age was deter-
mined by comparison with known-age ducklings in various stages of plumage development.
Plumage stages were assigned the following ages: IB — 10 days, IC — 18 days, II A 27 days,
IIB — 35 days, and IIC — 44 days.

We examined mass change of females before incubation using linear regression (PROC
GEM; SAS Institute Inc. 1988). Year and location effects could not be examined because
too few females were captured. Most females were weighed more than once during incu-
bation, so we investigated their mass change using a generalized linear mixed model
(GLMM) with maximum likelihood estimators (PROC MIXED; SAS Institute Inc. 1992).
This approach allows measurements on subjects to be repeated within and across years.
Dependencies among repeated measures are modelled explicitly and ensuing tests are ad-
justed for this dependence based on the underlying covariance structure (Laird and Ware
1982, Ware 1985). We determined (Jennrich and Schluchter 1986) that a compound sym-
metry covariance structure was optimal for our models. We modelled the effect of lake,
year, linear, quadratic, and cubic effects of incubation day, and their interactions on female
mass. When interactions were not significant (a = 0.05), we used a reduced model. Log-
likelihood ratio statistics were used to evaluate model goodness of fit, and simultaneous
paired comparisons were made using a Bonferroni adjustment to pairwise differences in the
time-adjusted means (Dobson 1990). Brood-rearing females were measured only once and
their mass change was examined using linear regression (PROC GLM; SAS Institute Inc.
1988). We ignored possible lake and year effects because too few brood-rearing females
were measured.

We further examined nonsignificant statistical results using post hoc power analyses
(Anonymous 1995). Regression results were evaluated using a SAS MACRO (Latour 1992).
Power calculations for the generalized linear mixed model were based on adjusted least
squares effects estimated by PROC MIXED (SAS Institute Inc. 1992).

RESULTS

From 1982 to 1985, 45 females were weighed prior to or during egg-
laying. In addition, 82 females were weighed repeatedly (1-5 times each
year) for a total of 213 times at known points during incubation or with
hatched young in the nest. One female was weighed in four years, five
in three years, 25 in two years, and 51 in only one year for a total of 120
within-year time-series. During brood rearing, 63 females were weighed.

Prelaying and laying. — Reproductive status of females weighed before
the start of incubation varied (Table 1). Those considered known nesters
successfully incubated nests that we observed. Known nesters were
weighed from one to 30 days before incubation (median = 15.5 days)
and most likely represented females that were beginning to lay the clutch
that they eventually incubated. In contrast, the sample of unknown status
likely included females nesting elsewhere as well as those laying eggs
parasitically when captured. We could not detect any linear trend in mass
of known nesters during the laying period (mass change = 1.5 g/day,
95% confidence interval = -0.4 to 3.5).

Incubation. — We fit the GLMM to measurements of mass for females
that successfully incubated a clutch. We detected no significant interac-

64

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Mass of Eemale Common Goldeneyes Captured During the Laying Period in
Northcentral Minnesota, 1982-85

Mass (g)

Stage (daysp

Status

X

SD

Median

Range

N

Median

Known nesters

768

34.0

775

685-810

16

15.5

Unknown

733

43.8

720

660-830

29

7

* Days prior to the start of incubation.

tions among the main effects (lake and year) and linear, quadratic, or
cubic measures of incubation day (all Ps > 0.46) that affected mass loss.
Two models, one linear and one cubic with incubation day, indicated mass
was influenced by stage of incubation. The cubic model fit the data best
(G“ = 16.48, df = 2, P < 0.001) and indicated a curvilinear effect of
stage on female mass (Table 2 and Fig. 1).

Body mass also was influenced by conditions related to specific lakes
(P = 2.72; df = 2, 79; P = 0.072). Female body mass was greater
throughout incubation on Island Lake than on North Twin Lake (P =
4.29; df = 1, 79; P = 0.041), but mass did not differ between Refuge
Pond and either North Twin (P = 0.00; df = 1, 79; P = 0.980) or Island
Lake (P - 1.94; df = 1, 79; P = 0.167) (Fig. 1).

We detected no effect due to different years with the GLMM (P =
0.75; df = 3, 125; P = 0.523), but graphical comparison of body mass
of females for whom we had measurements in 1982 and 1983 suggested
year might have an effect. Each of the four females was weighed at
similar points during incubation both years, and each was 20-^0 g lighter
in 1983 than 1982 (Fig. 2).

Four of eight females that abandoned their nests during incubation had

Table 2

Estimated Coefficients for Polynomial Measures of Incubation Day on Mass
Change in Eemale Common Goldeneyes from Three Northcental Minnesota Lakes,

1982-1985

Measure

P

SE (P)

pa

Day

6.3027

1.1799

<0.001

Day^

-0.4127

0.0995

<0.001

Day^

0.0083

0.0023

<0.001

“Two-sided test 3 = 0.

Zicus and Riggs • GOLDENEYE BODY MASS

65

Fig. 1. Maximum likelihood estimates of cubic temporal trend in female Common Gold-
eneye body mass during incubation on three northcentral Minnesota lakes, 1982-1985.

a body mass greater than that predicted for successful females at the
comparable point in incubation, and four were less than predicted.

Brood rearing. — Females in the composite sample regained body mass
at approximately 2 grams/day (F = 51.7; df = 1, 62; P < 0.001) from
the low they had reached when the young departed the nest (Fig. 3).
Females apparently regained most of the body mass lost during incubation
by the time they left their broods of class IIC or class III ducklings to
molt.

DISCUSSION

Our results differed markedly from previous measurements of Common
Goldeneye mass. Minnesota goldeneyes appear to weigh less at the onset
of nesting than do Ontario birds. Mallory (1991:17) reported that prelay-
ing females averaged 875 g at his Wanapitei study site and 842 g at a
site farther east. These values are 67-100 g more than our median at the
start of laying and 12^5 g more than our heaviest females. Whether
females truly differ to this extent is unclear. Our sample of prelaying
females was small, and we weighed females at various times during pre-
laying and laying. Furthermore, the Wanapitei values are estimates ob-
tained by adjusting female mass determined during incubation to account

66

THE WILSON BULLETIN • VoL 108, No. 1, March 1996

Days Before Departure From Nest

Fig. 2. Body mass of individual female Common Goldeneyes measured during incu-
bation while nesting on three northcentral Minnesota lakes, 1982 and 1983. Individual fe-
males are identified by unique symbols.

for incubation mass loss and that assumed lost in the course of laying a
clutch (Mallory 1991:16-17). Goldeneye females have been reported to
lose approximately 22 g/egg (i.e., approximately 1 1 g/day) during egg
laying (H. G. Lumsden, unpubl. data cited in Mallory 1991). We did not
detect this sort of change in Minnesota, and there was sufficient power
(>0.99) to detect a change of as little as 6 g/day. Projected mass of
Wanapitei females may be biased high because mass loss per egg in
Ontario is now believed to be <22 g (M. L. Mallory, Environ. Can., pers.
comm.). Nonetheless, model predictions of mass for Minnesota females
at the start of incubation ranged from 698-715 g depending on factors
associated with the nesting lake. In contrast, Ontario goldeneyes appear

ZicLis and Riggs • GOLDENEYE BODY MASS

67

Fig. 3. Linear regression predictions (±95% confidence limits) and observed mean body
mass (sample sizes) for female Common Goldeneyes with hatched young on 12 northcentral
Minnesota lakes, 1982—1985.

to begin incubation at 752-829 g (calculated from Mallory and Weath-
erhead 1993).

Minnesota goldeneye females lost proportionately less mass during in-
cubation than Ontario birds. Rate of mass change did not differ among
nesting lakes or years, although our analysis indicated location and pos-
sibly year affected overall incubation mass. The proportionate mass loss
by goldeneyes from the start of incubation until broods departed the nest
(31 days) was 10.7-11.0% depending on the nesting lake. In comparison,
Mallory and Weatherhead (1993) estimated mean mass loss variously and
reported changes for Ontario goldeneyes of 16.7% (page 853), 17.8%
(page 856), and 24.5% (calculated from equation page 853 using 31 days
of incubation).

68

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Incubating Common Goldeneyes apparently lose body mass in a non-
linear fashion (Mallory and Weatherhead 1993, this study). However, our
analysis suggested a different pattern than that described for birds in On-
tario. Minnesota goldeneyes experienced a rapid initial decline during
early incubation, reduced rate of loss in mid-incubation, and an increased
rate of loss again prior to hatching and duckling departure. Mass loss
during incubation for Ontario females has been modelled as decreasing
monotonically with the lowest mass occurring on the last day of incu-
bation (Mallory and Weatherhead 1993:equation p. 853). However, Mal-
lory and Weatherhead (1993:853) also indicated that some females
reached their lowest mass as early as day 18 of incubation, after which
some mass was regained.

Female mass when broods departed the nest was similar in Ontario and
Minnesota, but again measurements are not directly comparable. Mallory
and Weatherhead (1993:856) reported that mean body mass of 10 females
at the end of incubation was 626 g. However, two of the females included
in their sample had deserted their nests after 18 and 24 days of incubation,
respectively. By comparison, Minnesota females were predicted to weigh
615-633 g on the departure day (model intercepts) depending on the lake.
Unadjusted arithmetic mean mass on the day broods departed the nest
was 616 g (N = 20).

Mallory and Weatherhead (1993) speculated that female goldeneyes
that lost too large a proportion of their initial incubation weight or
dropped below approximately 600 g might be more prone to nest aban-
donment than others. We observed no indication that mass of females
deserting their nests differed from that of successful females, but more
data are needed. Furthermore, seven of 20 successful females weighed
the day ducklings departed the nest were less than 600 g and five were
<580 g. Some females remained below 600 g well into brood rearing.

Methodology alone does not explain the marked differences in incu-
bation mass and proportional mass change during incubation in our study
versus those reported for Ontario goldeneyes. At least two explanations
are tenable. In addition to the well known difference existing in geese,
body size has been shown to vary geographically in some passerines
(Aldrich and James 1991, Twedt et al. 1994). Ontario females may be
structurally larger than those nesting in Minnesota and thus able to return
to breeding areas with more stored reserves. Alternatively, Ontario fe-
males may be similar in size but may return with proportionately more
stored resources (i.e., better condition). Several studies (e.g., Gatti 1983,
Harvey et al. 1989, Aldrich and Raveling 1983) have reported that heavier
individuals lost a greater proportion of their body mass in incubation than
lighter conspecifics. Gatti (1983) reasoned that heavier Mallards {Anas

Zicus and Riggs • GOLDENEYE BODY MASS

69

platyrhynchos) could afford to lose more mass than those in poorer con-
dition. In contrast, Kennamer and Hepp (1987) reported that double-
brooded Wood Duck females lost a smaller proportion of their body mass
after their first nesting than single-brooded females and may have been
in better condition as a result.

Whether Ontario females are structurally larger than those in Minnesota
or just begin incubation in better condition, incubation constancy was
similar in the two locations (Mallory and Weatherhead 1993, Zicus et al.
1995). Together with the disparate mass loss and comparable weights at
the end of incubation, these results indicate that either differences in. for-
aging time during incubation recesses exist or else Ontario females con-
sume less food or food of lower quality during incubation than do Min-
nesota females. Most resident nesting goldeneyes from Island Lake and
Refuge Pond appeared to forage exclusively on their respective lakes,
whereas most females nesting on North Twin Lake foraged elsewhere.
Zicus and Hennes (1993) observed nesting female goldeneyes feeding at
least as much as most small-bodied waterfowl which rely extensively on
exogenous resources. They also reported that time devoted to foraging
during nesting varied among years and concluded that females foraged
less when food was most available. Harvey et al. (1989) likewise believed
that reduced food availability in some years contributed to a greater rel-
ative mass loss during incubation in Wood Ducks. Furthermore, incuba-
tion mass was lowest on Refuge Pond, the location among our three study
sites where Zicus et al. (1995) reported low incubation constancy in a
concurrent study. They speculated that low constancy was a consequence
of increased foraging time because of more difficult food acquisition.

Goldeneye mass during incubation varied among the lakes that we stud-
ied. Although we detected no differences among years with the GLMM,
measurement of a small sample of the same females in two consecutive
years suggested that yearly differences of at least 20 g might exist in
some years. The GLMM analysis had low power (0.46) at a = 0.05 to
detect such a difference. Nonetheless, the among lake and year mass
differences we measured were less than differences between Minnesota
and Ontario. Mann and Sedinger (1993) suggested that Northern Pintail
{Anas acuta) females nesting in Alaska relied more on endogenous re-
sources than temperate nesting congeners because of less productive high
latitude wetlands. Goldeneyes might use different nesting strategies de-
pending on average environmental conditions and food availability in the
regions they occupy. In some regions, wetland productivity may be suf-
ficient to provide adequate goldeneye food availability even in the pres-
ence of fish (DesGranges and Gagnon 1994:220). Resource acquisition
would then be less constrained by female foraging time because food

70

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

could be acquired easily. Patterns of body mass change before and during
incubation indicate this situation is likely in Minnesota. Alternatively,
females might rely more on resources acquired before arrival to nesting
areas where wetland productivity is low. If nesting strategy within the
species is flexible, relationships among habitat quality, female mass, and
reproductive effort and success (Mallory et al. 1994) may need to be
reexamined.

ACKNOWLEDGMENTS

We thank N. L. Weiland, biologist on the Chippewa National Lorest Blackduck District,
and private landowners on North Twin and Island Lakes for permission to work with their
waterfowl nest boxes. Numerous summer technicians with the Minnesota Department of
Natural Resources helped capture and weigh brood females during routine leg banding.
Discussions with and comments by M. A. Hanson, S. J. Maxson, R. T. Eberhardt, and D.
R Rave improved the manuscript. M. L. Mallory also reviewed an earlier draft of the
manuscript.

LITERATURE CITED

Afton, A. D. AND S. L. Paulus. 1992. Incubation and brood care, Pp. 62-108 in Ecology
and management of breeding waterfowl (B. D. J. Batt, A. D. Afton, M. G. Anderson,
C. D. Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, eds.). Univ. of Minnesota
Press, Minneapolis, Minnesota.

Aldrich, J. W. and E C. James. 1991. Ecogeographic variation in the American Robin
(Turdus migratorius). Auk 108:230—249.

Aldrich, T. W. and D. G. Raveling. 1983. Effects of experience and body weight on
incubation behavior of Canada Geese. Auk 100:670—679.

Anonymous. 1995. Journal news. J. Wildl. Manage. 59:196-198.

Blancher, P. j., D. K. McNicol, R. K. Ross, C. H. R. Wedeles, and P. Morrison. 1992.
Towards a model of acidification effects on waterfowl in Eastern Canada. Environ.
Pollut. 78:57-63.

Desgranges, J-L. and C. Gagnon. 1994. Duckling response to changes in the trophic web
of acidified lakes. Hydrobiologia 279/280:207—220.

Dobson, A. J. 1990. An introduction to generalized linear models. Chapman and Hall.
London, England.

Eadie, j. M. and a. Keast. 1982. Do goldeneyes and perch compete for food? Oecologia
55:225-230.

Eriksson, M. O. G. 1979. Competition between freshwater fish and goldeneyes Bucephala
clangula (L.) for common prey. Oecologia 41:99-107.

Gatti, R. C. 1983. Incubation weight loss in the Mallard. Can. J. Zool. 61:565-569.

Harvey, W. E, IV, Hepp, G. R., and R. A. Kennamer. 1989. Body mass dynamics of wood
ducks during incubation: individual variation. Can. J. Zool. 67:570—574.

Jennrich, R. I. AND M. D. SCHLUCHTER. 1986. Unbalanced repeated-measures models with
structured covariance matrices. Biometrics 42:805—820.

Johnson, L. L. 1972. An improved capture technique for flightless young goldeneyes. J.
Wildl. Manage. 36:1277-1279.

Kennamer, R. A. and G. R. Hepp. 1987. Frequency and timing of second broods in Wood
Ducks. Wilson Bull. 99:655-662.

Zicus and Riggs • GOLDENEYE BODY MASS

71

Laird, N. M. and J. H. Ware. 1982. Random-effects models for longitudinal studies.
Biometrics 38:963-974.

Latour, K. R. 1992. %Power: a simple macro for power and sample size calculations.
Proc. SAS Lfser’s Group Int. 17:1173—1177.

Mallory, M. L. 1991. Acid precipitation, female quality, and parental investment of Com-
mon Goldeneyes. M.S. thesis, Carleton Univ., Ottawa, Ontario.

AND P. J. Weatherhead. 1993. Incubation rhythms and mass loss of Common

Goldeneyes. Condor 95:849-859.

, D. K. McNicol, and P. j. Weatherhead. 1994. Habitat quality and reproductive

effort of Common Goldeneyes nesting near Sudbury, Canada. J. Wildl. Manage. 58:
552-560.

Mann, E E. and J. S. Sedinger. 1993. Nutrient-reserve dynamics and control of clutch
size in Northern Pintails breeding in Alaska. Auk 110:264—278.

Palmer, R. S. 1976. Handbook of North American birds, Vol. 3. Yale Univ. Press, New
Haven, Connecticut.

Ryder, R. A. 1965. A method for estimating the potential fish production of north-tem-
perate lakes. Trans. Amer. Fish. Soc. 94:214—218.

, S. R. Kerr, K. H. Loftus, and H. A. Regier. 1974. The morphoedaphic index, a

fish yield estimator — review and evaluation. J. Fish. Res. Board Can. 31:663—688.

SAS Institute Inc. 1988. SAS/STAT user’s guide, release 6.03 edition. SAS Institute Inc.,
Cary, North Carolina.

. 1992. SAS technical report p-229: SAS/STAT software changes and enhancements,

release 6.07. SAS Institute Inc., Cary, North Carolina.

Savard, J.-P. L. 1984. Territorial behavior of Common Goldeneye, Barrow’s Goldeneye
and Bufflehead in areas of sympatry. Ornis Scand. 15:211-216.

Twedt, D. j., W. j. Bleier, and G. M. Linz. 1994. Geographic variation in Yellow-headed
Blackbirds from the northern Great Plains. Condor 96:1030—1036.

Ware, J. H. 1985. Linear models for the analysis of longitudinal studies. Am. Stat. 39:95-

101.

Zicus, M. C. 1989. Automatic trap for waterfowl using nest boxes. J. Field Ornithol. 60:
109-111.

AND S. K. Hennes. 1989. Nest prospecting by Common Goldeneyes. Condor 91:

807-812.

AND . 1993. Diurnal time budgets of breeding Common Goldeneyes. Wilson

Bull. 105:680-685.

, , AND M. R. Riggs. 1995. Common Goldeneye nest attendance patterns.

Condor 97:461^72.

Wilson Bull., 108(1), 1996, pp. 72-79

INTERSPECIFIC VARIATION IN THE CALLS OF
SPHENISCUS PENGUINS

Nina N. Thumser,' - Jeffrey D. Karron,' and
Ml LUCENT S. FiCKEN'

Abstract. — We compared the vocal repertoires of Jackass (Spheniscus demersus), Hum-
boldt (5. humboldti), and Magellanic (5. rnagellanicus) penguins. Discriminant and cluster
analyses of the bray call indicate that Jackass and Magellanic penguins are more similar to
each other than either is to the Humboldt penguin, and all three are distinct from the Rock-
hopper penguin (Eudyptes chrysocome). The congruence of the vocal analyses with phylog-
enies based on allozyme data suggests that differences in vocalizations reflect gradual di-
vergence over time, not character displacement. Received 1 Jan. 1995, accepted 20 Sept.
1995.

Vocalizations frequently are used to assess taxonomic relationships in
birds and the use of song in avian systematics has been thoroughly re-
viewed by Payne (1986). Vocalizations also have been used to determine
phytogenies of non-passerine species such as the caledrine sandpipers
(Miller et al. 1988). Jouventin (1982) found calls to be the best behavioral
criterion for classifying penguins. The degree of variation in calls was
used to infer subspecies and species status in island populations of nine
penguin taxa. Vocalizations have been shown to be of primary importance
in the communication of many penguin species (Pettingill 1960, Stone-
house 1960, Boersma 1974, Spurr 1975, Jouventin 1982). Although be-
havior has been studied in all of the Spheniscus penguins, only prelimi-
nary information exists concerning their vocalizations (Boersma 1974,
1976, Eggleton and Siegfried 1979, Jouventin 1982, Scolaro 1987).

Species status and phylogenetic relationships in the genus Spheniscus
are not clearly defined. There is insufficient detail in the fossil record to
distinguish among species (Simpson 1976). Morphological studies led
Clancey (1966) to classify Jackass penguins (5. demersus) as a subspecies
of Magellanic penguins {S. rnagellanicus). O’Hare (1989) used 22 mor-
phological characters to clearly differentiate Spheniscus from other pen-
guin genera. However, he was unable to determine the taxonomic rela-
tionships among species within the genus. Utilizing data from DNA-DNA
hybridization, Sibley and Monroe (1990) proposed that S. demersus be
viewed as a superspecies containing demersus, rnagellanicus, and the
Humboldt penguin (S. humboldti). Recent allozyme analyses suggest that

' Dept, of Biological Sciences, P.O. Box 413, Univ. of Wisconsin-Milwaukee, Milwaukee, Wisconsin
53201.

^ Present Address: Dept, of Biological Sciences. 150 W. University Blvd. Florida Tech. Melbourne, Flor-
ida 32901-6988.

72

Thumser et al. • SPHENISCUS VOCAL COMPARISON

73

Humboldt penguins form a distinct species, but that Jackass and Magel-
lanic penguins are closely related (Grant et al. 1994, Thumser and Karron
1994).

We quantitatively analyzed the vocalizations of three Spheniscus spe-
cies (Jackass, Humboldt, and Magellanic penguins) and one outgroup
(Rockhopper penguins, Eudyptes chrysocome). Their vocal repertoires
were compared to determine if species consistently differ in the acoustical
structure of their calls. A resulting phylogeny was compared to an inde-
pendent phylogeny based on protein polymorphisms (Thumser and Kar-
ron 1994).

METHODS

Vocalizations of 21 Humboldt penguins were recorded at the Milwaukee County Zoo in
Wisconsin (February 1986— May 1987, February— March 1988), the Brookfield Zoo in Chi-
cago, Illinois (November 1987- April 1988), and the St. Louis Zoo in Missouri (October
1988). Recordings of 12 Jackass penguins were made at the Henry Villas Zoo in Madison,
Wisconsin (January-April 1988), the Knoxville Zoo in Tennessee (May 1989), and the
Racine Zoo in Wisconsin (February-March 1990). Seven Magellanic penguins were record-
ed at the Cincinnati Zoo in Ohio (April 1988) and by Jim Klinesteker at the John Ball
Zoological Gardens in Grand Rapids, Michigan (Spring 1989). Eleven Rockhopper penguins
were recorded at the Cincinnati Zoo (April 1988) and the St. Louis Zoo (October 1988).

This study was performed exclusively on captive penguins. Although the majority of
recorded Jackass penguins were born in captivity, most of the Magellanic, Rockhopper, and
Humboldt penguins were born in the wild. The birds comprising these captive populations
may have been drawn from a limited number of wild populations. However, the results from
this study should be representative since seabirds usually have limited variation in their
vocalizations, particularly at or below the species level (Pierotti 1987).

Observations were made during breeding periods, mainly prior to and just after egg laying,
since most of the calls occurred at those times. A microphone was placed inside the exhibit,
but observations were made from outside tbe exhibit to minimize disturbance of the birds.
Individuals were identified by tag color. General behavior was observed throughout the day
(from 08:00 to 17:00 h CST). Notes were recorded on videotape (Hitachi HJ 5000) and by
hand. Vocalizations were recorded throughout the period on a cassette recorder (Aiwa HSJ
500) using a Nakamichi (CM 100) microphone. Whenever possible the identity of the caller
was noted. Peak periods of vocalization were simultaneously videotaped and tape-recorded.

The recorded calls were analyzed at 150 Hz bandwidth using a Kay 7800 Digital Sona-
Graph and digitized using a Sigma Scan (1988) Program. For the bray call, the number of
syllables per call, total duration of the call, sum of the inter-syllable intervals, duration of
the longest syllable, and minimum, main, and maximum frequency of the longest syllable
were recorded. The main frequency represented the darkest band in the sonagram of the
call. These seven variables were selected to assess the acoustic structure of the bray call
based on both frequency and temporal components. In addition, these parameters were
selected because they could be measured precisely.

The vocalizations were analyzed using discriminant and cluster analysis in SYSTAT (Wil-
kinson 1990). The bray call was selected for analysis since nested ANOVA of individuals
within populations within species indicated significant differences at the species level for
more than one parameter (Thumser 1993). The data set included all recorded bray calls of
the three Spheniscus penguins and the Rockhopper penguins. There were 109 calls from

74

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Discriminant Eunction 1 Showing the Correlations between Conditional Dependent
Variables and Dependent Canonical Eactors in Eour Penguin Species

Factor

Variable

1

2

3

Number of syllables

-0.289

-0.710

-0.024

Total duration

0.337

-0.652

0.401

Inter-syllable interval

0.270

0.018

-0.345

Duration of longest syllable

0.622

-0.229

0.559

Maximum frequency

0.008

-0.635

-0.075

Minimum frequency

-0.145

0.464

0.083

Main frequency

-0.292

0.257

0.544

Chi-square

496.75

213.17

39.61

df

21

12

5

P

<0.001

<0.001

<0.001

Correlation

0.764

0.644

0.339

Humboldt, 77 calls from Jackass, 38 calls from Magellanic, and 106 calls from Rockhopper
penguins. Each of the bray call variables was standardized by converting its values to z-
scores prior to analysis. In discriminate analysis known groups were used to generate linear
models which gave the best fit for that grouping. The data were also analyzed to determine
how well the model predicts the actual groupings. Another multivariate technique, cluster
analysis, was used to detect natural groupings in data with no prior expectations. In this
case, Pearson’s distance measures and the single-linkage method were performed by cal-
culating the mean of each of the standardized variables for each species.

RESULTS

For the first discriminant function. Factor 1 arranged the four species
primarily on the basis of the duration of longest syllable and the total
duration of the call, while Factor 2 was primarily based on the number
of syllables, the total duration of the call, and the maximum frequency
of the longest syllable (Table 1). Overall, the analysis correctly catego-
rized 86% of Humboldt, 82% of Jackass, 52% of Magellanic, and 79%
of Rockhopper penguin calls. These vocal parameters clearly separated
the Spheniscus penguins from the outgroup, Rockhopper penguins (Fig.
1 A). Therefore, the outgroup was removed from the analysis and a second
discriminant analysis was run to increase the spread among the Spheniscus
penguins. In this discriminant function. Factor A arranged the three spe-
cies primarily on the basis of syllable number and maximum frequency
of the longest syllable, and Factor B was based primarily on the duration
and main frequency of the longest syllable (Table 2). The analysis cor-
rectly predicted 91% of Humboldt, 71% of Jackass, and 61% of Magel-
lanic penguin calls. Within the Spheniscus penguins, there was consid-

Thumser et al. • SPHENISCUS VOCAL COMPARISON

75

FACTOR(I)

Fig. 1. (A) A scatterpiot of the similarity between the bray calls of three Speniscus

species (Jacka.ss, Magellanic, and Humboldt) and an outgroup (Rockhopper penguins). (B)
A scatterpiot of the similarity between the bray calls of the three Spheniscus species. Both
scatterplots have ellipses around 50% of the data points for Humboldt (dark star), Jacka.ss
(open .square), Magellanic (dark circle), and Rockhopper (cross) penguins.

76

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 2

Discriminant Function 2 Showing the Correlations between Conditional Dependent
Variables and Dependent Canonical Factors in Three Spheniscus Species

Factor

Variable

A

B

Number of syllables

0.711

-0.020

Total duration

0.311

0.319

Inter-syllable interval

-0.136

-0.274

Duration of longest syllable

-0.225

0.735

Maximum frequency

0.441

-0.065

Minimum frequency

-0.257

0.074

Main frequency

-0.041

0.423

Chi-square

287.22

60.23

df

14

6

P

<0.001

<0.001

Correlation

0.804

0.491

erable overlap in the vocalizations of Magellanic and Jackass penguins
(Fig. IB). In fact, 37% of the Magellanic penguin calls were incorrectly
classified as Jackass penguin calls, and 12% of Jackass penguin calls were
incorrectly classified as Magellanic penguin calls. By contrast, there was
less similarity in the bray call parameters of Humboldt and Jackass pen-
guins (Fig. IB). Only 6% of the Humboldt penguin calls were incorrectly
classified as Jackass penguin calls and 17% of the Jackass penguin calls
were incorrectly classified as Humboldt penguin calls. Humboldt and
Magellanic penguins were the least similar in their calls (Fig. IB). In both
species 3% of their calls were incorrectly classified as the other species.
These results were supported by the cluster analysis shown in Table 3
and Figure 2. The distance matrix and tree show the Magellanic and
Jackass penguins clustering closely together, the Humboldt penguins more
distant, and the Rockhopper penguins the most distant.

Table 3

Pearson’s Distance Matrix Showing the Distance between Four Penguin Species
Based on Seven Variables of the Bray Call

Species

1

2

3

1 Humboldt Penguin

2 Jackass Penguin

1.304

3 Magellanic Penguin

1.436

0.260

4 Rockhopper Penguin

1.462

1.664

1.509

Thiimser et al. • SPHENISCUS VOCAL COMPARISON

77

A)

Magellanic

Jackass

Humboldt

Rockhopper

I 1 1 1

0.00 0.50 1.00 1.50

DISTANCE

B)

Magellanic

Jackass

Humboldt

Rockhopper

King

-+-

0.00

0.20 0.40

DISTANCE

0.60

Fig. 2. (A) A tree using the single-linkage method based on Pearson’s distances of

parameters of the bray call. (B) UPGMA tree based on modified Rogers distance of allozyme
data (Thumser and Karron 1994).

DISCUSSION

Overall, the Spheniscus penguins have retained a complement of calls
that are similar in structure and function (Thumser 1993). The bray call
is used to establish a territory and to advertise availability for pairing.
The bird stands with its head pointing up and calls while slowly flapping
its wings. This was the only call which showed sufficient species-level
variation for phenetic analysis.

The analyses of selected vocal parameters of the bray call clearly dis-
tinguish Humboldt from both Jackass and Magellanic penguins. However,
discriminant and cluster analyses often could not distinguish between the
Magellanic and Jackass penguin calls. This may reflect the evolutionary
relationships among the species or may have resulted from other factors.
Since Humboldt and Magellanic penguins occur sympatrically in South
America, another possible explanation for the differences in their breeding
calls is character displacement. By contrast, the similarity of South Amer-
ican Magellanic and African Jackass penguin calls is unlikely to result
from convergence.

In order to determine whether character displacement has occurred it
is necessary to know which vocal characters are ancestral. It is difficult
to root a tree based on vocalizations and determine the most ancestral
species because vocalizations can be subject to strong selection. However,
a comparison of allozyme variation enhances these results because protein
markers are subject to weaker and different selective forces than those

78

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

influencing behavioral traits. Trees based on allelic characters are also
more easily rooted using outgroup taxa.

The allozyme data were consistent in two different studies (Grant et
al. 1994, Thumser and Karron 1994). Grant et al. (1994) based their
analysis on 15 polymorphic loci in 75 captive (Humboldt, Magellanic)
and wild (Jackass, Rockhopper, and Macaroni [Eudyptes chrysolophus])
penguins. Thumser and Karron (1994) studied nine polymorphic loci in
165 captive (Jackass, Humboldt, Rockhopper, King [Aptenodytes pata-
gonicus]) and wild (Humboldt, Magellanic) penguins. In both studies.
Jackass and Magellanic penguins were very closely related and Humboldt
penguins clearly formed a distinct species. There is a striking similarity
of phenetic trees based on the allozyme data and the vocal analysis pre-
sented here (Fig. 2). Cladistic analysis of the allozyme data confirmed
that Spheniscus penguins form a monophyletic group (Grant et al. 1994,
Thumser and Karron 1994). These findings suggest that differences in
vocalizations between the Humboldt and Magellanic penguins are not due
to character displacement, but rather reflect gradual genetic divergence of
separate evolutionary lineages. Although the Humboldt and Magellanic
penguins occur sympatrically, they have lower genetic identities and
greater vocal differences than the more closely related Jackass and Mag-
ellanic penguins.

ACKNOWLEDGMENTS

This research was partially funded by grants from the Ruth Walker Scholarship Lund, the
Univ. of Wisconsin— Milwaukee Graduate School, and the Institute of Museum Services (IC-
10197-91). We thank the following zoos for allowing access and providing assistance in
recording penguin vocalizations; Brookfield, Cincinnati, Henry Villas Park, Knoxville, Mil-
waukee County, Racine, and St. Louis. We are also grateful to Jim Klinesteker at the John
Ball Zoo for providing us with recordings of a Magellanic penguin population. P. D. Boers-
ma, G. Miller, and an anonymous reviewer provided several helpful suggestions on an earlier
version of the manuscript.

LITERATURE CITED

Boersma, P. E. 1974. The Galapagos penguin; a study of adaptations for life in an unpre-
dictable environment. Ph.D. diss. Ohio State Univ., Columbus, Ohio.

. 1976. An ecological and behavioral study of the Galapagos penguin. Living Bird

15;43-93.

Clancey, P. a. 1966. The penguins Spheniscus clemersus Linnaeus and Spheniscus niagel-
lanicus Forster. Ostrich 37;237.

Eggleton, P. and W. R. Siegfried. 1979. Displays of the Jackass penguin. Ostrich 50; 139-
167.

Grant, W. S., D. C. Duffy, and R. W. Leslie. 1994. Allozyme phylogeny of Spheniscus
penguins. Auk 1 I 1;7 16-720.

JouvENTiN, P. 1982. Visual and vocal signals in penguins, their evolution and adaptive
characters. Adv. Ethology 24.

Thumser et al. • SPHENISCUS VOCAL COMPARISON

79

Miller, E. H., W. W. H. Gunn, and B. N. Veprintser. 1988. Breeding vocalizations of
Baird’s Sandpiper Calidris bairdii and related species, with remarks on phylogeny and
adaptation. Ornis Scand. 19:257-267.

O’Hare, R. J. 1989. Systematics and the study of natural history, with an estimate of the
phylogeny of living penguins (Aves: Spheniscidae). Ph.D. diss. Harvard Univ., Boston,
Massachusetts.

Payne, R. B. 1986. Bird songs and avian systematics. Pp. 87—116 in Current ornithology.
Vol. 3 (R. E Johnston, ed.). Plenum Press, New York, New York.

Pettingill, O. S. Jr. 1960. Creche behavior and individual recognition in a colony of
Rockhopper penguins. Wilson Bull. 72:209—221.

PiEROTTi, R. 1987. Isolating mechanisms in seabirds. Evolution 41:559-570.

SCOLARO, J. A. 1987. A model life table for Magellanic penguins {Spheniscus magellanicus)
at Punta Tombo, Argentina. J. Eield Ornithol. 58:432-441.

Sibley, C. G. and B. L. Monroe, Jr. 1990. Distribution and taxonomy of birds of the
world. Yale Univ. Press, New Haven, Connecticut.

Sigma Scan. 1988. Scientific Measurement Program, Version 3.90. Jandel Scientific, Corte
Madera, California.

Simpson, G. G. 1976. Penguins past and present, here and there. Yale Univ. Press, New
Haven, Connecticut.

Spurr, E. B. 1975. Communication in Adelie Penguins. Pp. 449-501 in The biology of
penguins (B. Stonehouse, ed.). Macmillan, New York, New York.

Stonehouse, B. 1960. The King Penguin (Aptenodytes patagonicus) of South Georgia.
Talk. Isl. Dep. Surv. Sci. Rep. 23.

Thumser, N. N. 1993. Phylogenetic relationships among Spheniscus penguins based on the
analysis of vocal and allozyme data. Ph.D. diss. Univ. of Wisconsin-Milwaukee, Mil-
waukee, Wisconsin.

AND J. D. Karron. 1994. Patterns of genetic polymorphism in five species of

penguins. Auk 111:1018-1022.

Wilkinson, L. 1990. SYSTAT: the system for statistics. SYSTAT, Inc., Evanston, Illinois.

Wilson Bull., 108(1), 1996, pp. 80-93

THE BREEDING BIOLOGY OF THE WILLOW TIT IN
NORTHEASTERN SIBERIA

Vladimir V. Pravosudov and Elena V. Pravosudova

Abstract. We studied the breeding biology of the Willow Tit {Parus montanus) during

19g7_1990 in the Magadan region of northeastern Siberia. Clutch size and number of fledg-
lings averaged 7.5 and 6.5, respectively, and both were correlated negatively with the date
of the first egg. Nestling growth rate was correlated positively with the date of the first egg,
but was not related to brood size. Body mass at fledging was related negatively to brood
size. Males fed nestlings significantly more often than did females, while females spent
more time attending nestlings. The number of parental visits per young did not change
significantly with brood size. During the first 13 days of the nestling period, female feeding
rate per young was positively related to brood size while for males this relationship was
negative. The nestling diet consisted mostly of Lepidoptera larvae, Arachnoidea, and Dip-
tera. Received 20 April, 1995, accepted 21 Sept. 1995.

Most studies of the Willow Tit {Parus montanus), a small hole-nesting
Palearctic passerine, have focused on the species’ non-breeding biology
in northern Europe (see Ekman 1989, Matthysen 1990). Eewer studies
have been done on its breeding biology (Orell 1983, Orell and Ojanen
1983, Pravosudov 1987, Orell and Koivula 1988, see review by Cramp
and Perrins 1993). Here, we report details of breeding biology of the
Willow tit in northeastern Siberia and compare it with breeding biology
of other parids. Our study area is close to the easternmost part of this
species’ range.

METHODS

We collected data between 1987 and 1990 in the .southern part of the Magadan legion of
northeastern Siberia (60°N, 150°E). The habitat was comprised mainly of larch {Lati.x ca-
jandery), poplar (Populus suoveolens), and chosenia {Chosenia arbutifolia). A detailed de-
scription of the study area has been published previously (Pravosudov 1993a, b). Most of
the birds were fitted with unique combinations of color bands. Nests were opened with a
knife, after which a patch of bark was used to cover the opening. Nestlings in 13 nests were
weighed to the neare.st 0.1 g and the fifth primary was measured to the nearest millimeter.
Using nonlinear regression, we found that the logistic equation was the most suitable for
describing rate of body mass increase, and we used the growth constant K to estimate
nestling growth rate. As another index of nestling growth, we used the growth of the fifth
primary. The rate of increase in length of this feather is very close to linear, so we used the
slope of the linear regression as an estimate of the growth rate. The rates at which nestlings
were fed were studied by repeatedly observing eight nests for 1-h intervals over the whole
nestling period. All observations were made from 35^0 m, and the parents did not appear

Institute of the Biological Problems of the North, Academy ol Sciences of Russia, Far East Division, K.
Marx pr. 24, Magadan, Russia. (Present address; Behavioral Ecology Group, Department ot Zoology,
The Ohio State University, 1735 Neil Avenue, Columbus, Ohio, 43210-1293.)

80

Pravosudov and Pravosudova * WILLOW TITS

81

to be disturbed by our presence. To collect food brought to each nestling by its parents, we
used a thread tied loosely around each nestling’s neck. After the parents made a number of
feeding visits corresponding to the number of young in a nest, we collected the food loads
from each nestling’s throat with forceps and stored each load in a separate vial. Parents
resumed feeding as soon as collared nestlings were put back in the nest. This method did
not appear to harm the young tits. In 1987, we determined only the frequency of different
prey items in the diet. In 1989 and 1990, we also weighed to the nearest mg each food load
taken from a nestling and each individual item such loads contained. Over all years, we
collected 296 food items contained in 108 individual loads from 11 broods, 185 of which
were weighed. The number of food loads collected per nest was 5—23 loads. The food loads
were collected from nestlings 6—12 days old, with most collected during 7—11 days of
nestling age.

Multi-way general linear models (GLM) and multiple and simple linear regression anal-
yses were used for the majority of tests. General linear model (GLM) is a multivariate
analysis that is used to perform analysis of variances (ANOVA) with balanced and unbal-
anced designs, analysis of covariance (ANCOVA), and regression (Neter et al. 1990, Anon-
ymous 1991). All analyses were performed using MINITAB routines (Anonymous 1991).
All tests of the slopes in regression analyses and for covariates in the GLMs were two-
tailed. For statistical analyses of parental feeding rates, we used two methods (1) we ana-
lyzed averages per nest for the whole nestling period (16 days), and (2) we used all obser-
vations (147 1-h periods) made repetitively at the eight nests in a GLM where each nest
served as a factor. All models dealing with parental feeding rate consisted of a nest as a
factor and nestling age and brood size as covariates. Analyses of the length of time that
parents spent in the nest during one visit while feeding nestlings was done on three nests
only, and all 579 such observations were used in a GLM where each nest served as a factor
and brood size and nestling age served as covariates. For statistical analyses of variance in
food loads and main prey type in the diet among different years and different broods, we
used average frequency of every prey type in a load per nest in ANOVA. Thus, the brood
was the primary sampling unit. To analyze variation in the diet among different broods, we
used average frequency of every prey type in a load and the load was the primary sampling
unit. The variation among broods was tested separately for each year.

RESULTS

Willow Tits almost always excavated their own cavities. Only two of
22 nests were in pre-existing natural holes, and in both cases, the birds
altered the original hole by shaping the cavity. The average start of egg-
laying differed significantly among years, occurring a week later during
1990 than during the two preceding years (ANOVA, ^2,8 = 5.7, P =
0.012; Table 1). Mean clutch size was 7.5 eggs and varied significantly
among years. Birds that started laying eggs later had smaller clutches
(ANCOVA, effect of year, F, ,7 = 10.1, P = 0.006; date of the first egg,
slope = -0.15, P = 0.006; Table 1). The number of young that fledged
was 6.5 and did not vary significantly among years, but pairs that started
laying eggs later had significantly fewer fledglings (ANCOVA, effect of
year, F, ,6 = 2.4, P = 0.124; date of the first egg, slope = -0.19, P =
0.03; Table 1). Reproductive success as fledglings per egg averaged 0.85
and was nearly always higher than 0.7 (Table 1). The nestling body mass

82

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table

Breeding Characteristics of Willow Tits

1

IN THE Magadan Region of Siberia*"

Parameter

1987

1989

1990

Total

Range

No. of nests

9

6

7

22

Mean date of first egg

30 May

30 May

6 June

1 June

25 May-7 June

1.6 days

1.9 days

1.7 days

5.6 days

Clutch size

7.1

8.0

7.7

7.5

6-10

0.4

0.5

0.4

1.1

No. of fledglings

6.4

6.2

6.8

6.5

3-9

0.6

0.7

0.7

1.6

Reproductive success**

0.89

0.75

0.89

0.85

0.42-1.00

5.2

6.1

6.1

15.5

K constant

0.376

0.372

0.411

0.385

0.331-0.436

0.017

0.015

0.016

0.035

Primary growth rate

3.28

3.44

3.45

3.39

3.02-3.60

(mm/day)

0.07

0.06

0.07

0.16

Body mass of young

11.91

12.49

12.54

12.33

11.34-13.16

on day 14 posthatch

0.43

0.69

0.27

0.56

•‘Within each cell of the four rows on the left, the upper number is the mean and the lower number is the standard
deviation.

Reproductive success = fledglings/eggs.

growth constant (K) did not vary significantly among years or with brood
size (ANOVA, P > 0.2). The nestlings in the later nests, however, grew
significantly faster (regression on the date of the first egg, b = +0.005,
P - 0.011, constant K = 0.213 + 0.005 date; Fig. 1). Body mass of
nestlings on day 14 posthatch did not vary significantly among years
(Table 1) or with date of first egg (ANOVA, P > 0.2). However, day- 14
body mass appeared to be related to brood size; nestlings in smaller
broods were heavier (regression on the brood size, b = —0.163, P =
0.064, N = 13, mass = 13.4 — 0.163 brood size). Body mass of fledglings
was not significantly related to their growth rate (regression, P = 0.80).
The growth rate of the fifth primary varied similarly with the growth
constant, K. Feather growth was not significantly affected by either year
or brood size (ANOVA, P > 0.2), but the nestlings in later broods tend
to grow their fifth primary faster (regression on the date of the first egg,
b = 0.017, F = 0.082, N = 13, rate = 2.83 + 0.017 date).

Males made significantly more feeding trips to nest than did females
(paired t-test, t — 3.56, N = 8 nests, P = 0.009, Fig. 2). Females increased
the number of feeding trips per brood as brood size increased (Regression,
b = 0.82, P = 0.046, N = 8 broods, number of trips = 1.18 + 0.82
brood size), while total number of visits and number of visits by males
per nest were not significantly affected by brood size (F > 0.2, calculated

Pravosudov and Pravosudova • WILLOW TITS

83

H

<

H

CZJ

Z

O

U

u

H

<

a

H

o

a

o

cn

<

>-

Q

O

a

0.45

0.44

0.43

0.42

0.41

0.40

0.39

0.38

0.37

0.36

0.35

0.34

0.33

0.32

0.31

0.30

I I I I L

25 30 4 9 14

MAY JUNE

DATE OF THE FIRST EGG

Fig. 1 . Relationship between body mass growth rate constant, K, of Willow Tits and
the date of the first egg.

on averages per nest for the entire nestling period). GLM analysis using
all feeding observations and brood as a factor during the entire nestling
period showed no significant relationships between the number of visits
per nest (total, male and female separately) and brood size {P > 0.2).
Both total number of visits per nest (slope = 3.12, P < 0.01) and number
of visits by females (slope = 3.95, P < 0.01) significantly increased as
nestlings grew older, although male feeding rate was affected by nestling
age only suggestively (slope = 1.56, P = 0.095). The number of feeding
trips per young (total, male and female separately) was also not signifi-
cantly affected by brood size {P > 0.4, calculated both for nest averages
for all nesting period and with a nest as a factor for all feeding obser-
vations). Both total number of feeding trips (GLM with a brood as a

84

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Eig. 2. Number of feeding visits of Willow Tits per nestling per hour as related to
nestling age. Open circles and squares represent feeding rates of males and females, re-
spectively. Triangles represent rates for males and females combined. Vertical lines represent
± one SD of the means.

factor, slope = 2.64, P < 0.01) and number of visits by female per young
(slope = 3.30, P < 0.01) increased significantly as nestlings grew older,
while male feeding rate was not significantly affected by nestling age
(slope = 1.29, P = 0.22). Since some young left their nest on the 14th
day posthatch and nestling body mass (Fig. 3) and parental feeding rate
(Fig. 3) started slowing down by the 13th day posthatch, we analyzed the
frequency of parental feeding visits separately for the first 13 days of
nestling age (N = 116 1-h periods for 8 nests). This separate analysis of
the first 13 days may be very important since the young reach their fledg-
ing body mass and are able to fledge by that time, so the heaviest pressure
on parents should fall in this period. The results were strikingly different

Prcivosudov and Pravosndova • WILLOW TITS

85

Fig. 3. Nestling body mass of Willow Tits as related to nestling age.

from those of the whole nesting period (Table 2). Both total number of
feeding trips and number of feeding trips by female per brood and per
young increased significantly as nestlings grew older. Male feeding rate,
though, did not change significantly with nestling age (Table 2). Both
males and females increased number of feedings per nest as brood size
increased. Feeding frequency per young showed a dramatic difference
between sexes (Fig. 4). Females made more feeding trips per young in
larger broods, while males made significantly fewer of them as brood size
increased (Table 2, Fig. 4).

Females, but not males, spent more time attending nests containing
fewer young during the entire nesting period (GLM with nest and brood
size as factors, ^2,579 = 8.27, P < 0.001). Both male and female decreased
their attendance time as nestlings grew older (GLM with a nest as a factor,
P < 0.01, Fig. 5).

Arachnoidea, Lepidoptera, and Diptera comprised the majority of the

86

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Table 2

Relationships between Parental Leeding Rate and Nestling Age (Day 1-13) and

Brood Size in the Willow Tit

No. of visils/h

Age

Brood

size

Slope

p

Slope

p

Male per brood

0.16

0.13

1.10

0.00

Lemale per brood

0.44

0.00

1.37

0.00

Total per brood

0.59

0.00

2.49

0.00

Male per young

0.03

0.11

-0.12

0.01

Lemale per young

0.06

0.00

0.08

0.01

Total per young

0.09

0.00

-0.03

0.58

nestlings’ diet (79%, Tables 3 and 4). The number of prey items contained
in each individual food load brought to nestlings averaged 3.0 and did
not vary significantly among years (Fj g = 1.76, P = 0.23, Table 5) or
different broods (no two pairs were different at F = 0.05, Tukey’s test).
Average mass of a load taken during 1989-1990 also was not statistically
different (r-test, t = 1.74, P = 0.22, N = 7) between these years (Table
3), although there were significant differences among broods during both
years (in 1989, one nest was different from one out of 3 broods and in
1990, one nest was different from the rest, all differences at F = 0.05,
Tukey’s test). The mass of a food load was positively and significantly
related to the number of prey items in a load when differences among
years and broods were accounted for (ANCOVA with year and brood as
factors and the number of items in a load as a covariate, individual load
is the primary sampling unit; slope = 15.64, t = 5.49, F < 0.001). Even
though all major prey items showed a great deal of variation there were
no significant difference in average frequency of any major prey type per
load either among years or among broods (no pairs compared were dif-
ferent at F = 0.05, Tukey’s test).

DISCUSSION

The breeding biology of the Willow Tit in northeastern Siberia and
western Europe appears to be very similar (Orell and Ojanen 1983). Av-
erage clutch size (7.62) and number of fledglings (6.19) in Einland (Orell
and Ojanen 1983) are nearly identical to those in Siberia (7.5 eggs and
6.5 fledglings). The trend toward reduced clutch size with later breeding
seems to be general not only for Willow Tits but for many other bird
species as well (Orell and Ojanen 1983). Nestling growth rate (K) of
Siberian Willow Tits (0.331-0.436) is very similar to that of Willow Tits

Pravosudov and Pravosudova • WILLOW TITS

87

B

X

X

u

CL.

2.2

2.0

Fig. 4. Feeding rates by male and female Willow tits per brood (A) and per nestling
(B) as related to brood size. Feeding rates are taken as averages for the first 13 days of
nestling age, and lines represent standard deviations of the means. Males are represented by
circles and females by squares.

88

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Eig. 5. Eength of time spent in the nest cavity during one visit by male and female
Willow Tits as related to nestling age. Vertical lines represent standard deviations of the
means. Males are shown by circles and females by squares.

in other studies (0.380-0.406; Foster and Godfrey 1950, Inozemtsev
1962, Song 1980, Orell and Ojanen 1983). The negative coirelation be-
tween fledgling body mass and brood size has also been demonstrated
before (Nur 1984, Orell and Koivula 1990).

It seems that birds that start breeding earlier typically produce more
fledglings. However, we found that nestling growth rate was positively
correlated with the start of breeding. Nestlings that hatched later grew
significantly faster although there was no effect of the timing of breeding
on fledgling body mass. The fledgling body mass was also unrelated to
their growth rate. This result may suggest that although late-breeding pairs
produce fewer fledglings, their young could be better off nutritionally.
Timing of breeding has long been thought of as a life-history trade-off.

Pravosudov and Pravosudova • WILLOW TITS

89

Table 3

Frequency of Different Food Types in the Diet of Nestling Willow Tits in Siberia

Prey taxon

1987

1989

1990

Total

Percent

Arachnoidea

9

20

20

49

16.5

Total Lepidoptera

25

63

45

135

45.6

Larvae

18

59

45

122

41.2

Imago

7

4

0

13

4.4

Total Hymenoptera

15

3

2

20

6.8

Larval Hymenoptera

0

2

0

2

0.7

Tentridinidae

1

0

0

1

0.3

Pamphilidae

0

1

0

1

0.3

Formicidae

1

0

2

3

1.0

Total Coleoptera

16

4

2

22

7.4

Larval Coleoptera

15

2

2

19

6.4

Diptera

38

10

2

50

16.9

Limoniidae

25

0

0

25

8.4

Tipulidae

10

9

2

21

7.1

Unidentified

3

1

0

4

1.3

Plecoptera

2

4

0

6

2.0

Chloroperlidae

0

3

0

3

1.0

Unidentified

2

1

0

3

1.0

Homoptera

1

2

0

3

1.0

Gastropoda

0

1

1

2

0.7

Fish bones

2

0

0

2

0.7

Total number

111

1 14

71

296

Birds may benefit by breeding earlier, possibly because food abundance
is higher. However, for the Willow Tits breeding in Siberia, this is prob-
ably not true since when they start breeding there is still snow cover and
food becomes abundant only later (pers. obs.). Other studies have shown
that in some resident passerine birds, including the Willow Tit, there is
another important benefit from breeding earlier; young fledged earlier
stand a better chance of recruitment into both winter flocks and into the
next season’s breeding population (Nilsson 1988, Koivula et al. 1993,
Pravosudov 1993b). Our results suggest that there may be a trade-off
between early and late breeding in the Willow Tit. If nutritional condition
of fledglings were independent of the start of breeding, birds breeding
earlier would produce young that would have a higher chance of estab-
lishing themselves into a subsequent breeding population. However, birds
breeding later may produce better nourished young, although fewer of
them. The optimal time of breeding may thus reflect a trade-off between
breeding early enough to produce early dispersers (and successful re-
cruits) and breeding late enough to produce well nourished young.

90

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 4

Mass (mg) of Different Lood Types in the Diet of Nestling Willow Tits

Prey taxon

1989

1990

Total

Percent

Arachnoidea

833

702

1535

20.5

Lepidoptera

2392

2531

4923

65.8

Larvae

2175

2531

4706

62.9

Imago

217

0

217

2.9

Hymenoptera

160

85

245

3.3

Pamphilidae

60

0

60

0.8

Lormicidae

0

85

85

1.1

Larvae

100

0

100

1.3

Coleoptera

57

104

161

2.1

Larvae

36

104

140

1.9

Diptera

310

119

429

5.7

Tipulidae

290

119

409

5.5

Plecoptera

84

0

84

1.1

Chloroperlidae

68

0

68

0.9

Homoptera

76

0

76

1.0

Gastropoda

16

14

30

0.4

Total

3928

3555

7483

Parental behavior in which male Willow Tits fed young more often
than females but females spent more time attending nestlings during the
whole nesting period was quite similar to that of the Mountain Chickadee
{Pams gambeli) (Grundel 1987). However, during the first 13 days of
nestling life. Willow Tits differed significantly from Mountain Chickadees
in the way the sexes responded to changes in brood size. In our study,
both males and females increased the number of feeding visits per brood
as the brood size increased. However, the number of feeding trips per
young was negatively correlated with brood size in males but positively
correlated in females. This result, and also the fact that male feeding rate

Table 5

Measurements of Nestling Leeding in the Willow Tit

Parameter

1987

1989

1990

Total

Number of loads

24

49

35

108

Number of broods

4

4

3

1 1

Number of food items/load

4.2 (2.6)“

2.2 (0.7)

2.5 (0.8)

3.0 (1.8)

Mass of a food load

—

81.1 (22.2)

127.9 (42.4)

101.1 (38.4)

" so.

Pravosudov and P ravosudova • WILLOW TITS

91

did not change much with nestling age, may indicate that males worked
as hard as possible and that is why their feeding rate was similar for
broods of different size and age throughout the entire nesting period. On
the other hand, females appeared to adjust their feeding rate to variation
in demand caused by different ages of young and different brood sizes.
Females, but not males, also adjusted their attendance time to brood size
while feeding nestlings.

It is known that parental feeding rate by itself is not necessarily a good
indicator of parental investment since the biomass of food brought by
parents to their young per visit may differ between males and fernales
(Grundel 1987). However, Grundel (1987) found in the Mountain Chick-
adee that differences in total volume of food per nestling in broods of
different size were due to changes in feeding frequency rather than in
prey size or load size. Therefore, the patterns of feeding frequencies by
male and female Willow Tit parents of large and small broods appear to
be a valid representation of their investment. The pattern of feeding rate
with an increase in the beginning of the nestling stage, a plateau in the
middle, and a decrease before fledging seems to be very common and
has been shown for many parids (Gibb 1950, Royama 1966, Grundel
1987). Because of the sexes’ opposite trends in feeding rate per young
with an increase in brood size, the total number of feedings per young in
Siberian Willow Tits was not significantly different in broods of different
size. This pattern has not been found in many studies (Gibb 1950, 1955;
Royama 1966, Walsh 1978, Grundel 1987), although it has been dem-
onstrated before (Pinkowski 1978). Assuming that the changes in feeding
rate associated with changes in brood size truly represent a change in
parental investment, our results are consistent with the individual adjust-
ment hypothesis (Hogstedt 1980, Pettifor et al. 1988) which assumes that
birds adjust their clutch size to their own capabilities of raising young.
The absence of any relationship between brood size and nestling growth
rate also appears to support this hypothesis. The negative relationship
between fledgling body mass and brood size seems to go against the
individual adjustment hypothesis since post-fledging survival is known to
correlate positively with fledging body mass (Nur 1984, Orell and Koivula
1990). Since young in large broods tend to be lighter compared with
young from smaller broods, one can assume that those from larger broods
should have lower survival. However, experimental results from studies
of Willow Tits in Finland did not fully support this assumption, suggest-
ing that any relationship among brood size, nestling weight, and juvenile
survivorship can be complicated by environmental variability (Orell and
Koivula 1990). Also, from all the relationships described above, we can
assume that lighter young fledge earlier (earlier breeding start results in

92

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

larger brood size which results in lighter body mass of fledglings) and,
if they survive, would have a greater chance of breeding.

ACKNOWLEDGMENTS

We thank Elena Zimireva for her assistance in the field. T. C. Grubb, Jr., R. A. Mauck,
and T. A. Waite, O. Hogstad, and C. R. Blem provided valuable comments on an earlier
draft of the manuscript.

LITERATURE CITED

Anonymous. 1991. Minitab reference manual: release 8. Minitab, State College, Pennsyl-
vania.

Cramp, S. and C. M. Perrins. 1993. The birds of the western Palearctic. Volume VII.
Flycatchers to Shrikes. Oxford Univ. Press, Oxford, England.

Ekman, J. 1989. Ecology of non-breeding social systems of Parus. Wilson Bull. 101:263—
288.

Foster, J. and C. Godfrey. 1950. A study of the British Willow Tit. Brit. Birds 43:351-
361.

Gibb, J. A. 1950. The breeding biology of the Great and Blue titmice. Ibis 92:507—539.

. 1955. Feeding rates of Great Tits. Br. Birds 48:49-58.

Grundel, R. 1987. Determinants of nestling feeding rates and parental investment in the
Mountain Chickadee. Condor 89:319-328.

Hogstedt, G. 1980. Evolution of clutch size in birds: adaptive variation in relation to
territory quality. Science 210:1148-1150.

Inozemtsev, A. A. 1962. Notes on the ecology of tits in the Moscow region. Pp. 169-199
in Materialy po faune i ekologii zhivotnykh. (S. P. Naumova, ed.). (in Russian).

Koivula, K., K. Lahti, M. Orell, and S. Rytkonen. 1993. Prior residency as a key
determinant of social dominance in the Willow Tit {Parus montaniis). Behav. Ecol.
Sociobiol. 33:283-287.

Matthysen, E. 1990. Nonbreeding social organization in Parus. Curr. Ornithol. 7:209-249.

Neter, j., W. Wasserman, and M. H. Kutner. 1990. Applied linear, statistical models.
IRWIN Press, Homewood, Illinois.

Nilsson, J.-A. 1988. Causes and consequences of dispersal in Marsh Tits: time as a fitness
factor in establishment. Ph.D. diss.. University of Lund, Lund, Sweden.

Nur, N. 1984. The consequences of brood size for breeding Blue Tits. II. Nestling weight,
offspring survival and optimal brood size. J. Anim. Ecol. 53:497-517.

Orell, M. 1983. Nestling growth in the Great Tit Parus major and the Willow Tit P.
montanus. Ornis Fenn. 60:65-82.

AND M. Ojanen. 1983. Breeding biology and population dynamics of the Willow
Tit Parus montanus. Ann. Zool. Fenn. 20:99-1 14.

AND K. Koivula. 1988. Cost of reproduction: parental survival and production of
recruits in the Willow Tit Parus montanus. Oecologia 77:423^32.

AND . 1 990. Effects of brood size manipulations on adult and juvenile sur-
vival and future fecundity in the Willow Tit, Parus montanus. Pp. 297—306 in Popu-
lation biology of passerine birds, an integrated approach (J. Blondel, A. Gosler, J.-D.
Lcbreton, and R. McCleery, cds.), NATO ASI Series, vol. G-24. Springer- Verlag, Hei-
delberg, Germany.

Peitifor, R. a., C. M. Perrins, and R. H. McCleery. 1988. Individual optimization of
clutch size in Great Tits. Nature 336:160-162.

Pruvosudov and Pravosudova • WILLOW TITS

93

PiNKOWSKi, B. C. 1978. Feeding of nestling and fledgling Eastern Bluebirds. Wilson Bull.
90:84-98.

Pravosudov, V. V. 1987. Ecology of two closely related species of tits in the northwestern
part of the USSR. Ornitologia (Moscow) 22:68-75 (in Russian).

. 1993a. Breeding biology of the Eurasian Nuthatch in northeastern Siberia. Wilson

Bull. 105:475-482.

. 1993b. Social organization of the Nuthatch Sitta europaea asiatica. Ornis Scand.

24:290-296.

Royama, T. 1966. Factors governing feeding rate, food requirement and brood size of
nestling Great Tits Parus major. Ibis 108:313—347.

Song, Y. 1980. Studies on the breeding ecology and feeding habits of Willow tits. Acta
Zool. Sinica 26:370-377 (in Chinese with English summary).

Walsh, FI. 1978. Food of nestling Purple Martins. Wilson Bull. 90:248-260.

Wilson Bull., 108(1), 1996, pp. 94-103

CENSUSING WINTERING POPULATIONS OF
SWAINSON’S WARBLERS: SURVEYS IN THE
BLUE MOUNTAINS OF JAMAICA

Gary R. Graves

Abstract. — Census methods developed for breeding populations of Nearctic-Neotropic
migrant passerines are largely ineffective for determining the distribution and abundance of
Swainson’s Warbler {Limnothlypis swainsonii) on its wintering grounds in the Caribbean
basin. Using playback of tape-recorded call notes interspersed with advertising songs, I
found the warbler to be widespread and relatively common in montane forests of the Blue
Mountains of Jamaica. Detection rates with playback varied from 17.8 to 29.2 warblers/10
h along five census transects. Census efficiency was increased by an estimated factor of five
to 10 times with the use of tape playback. Received 23 March 1995, accepted 20 Aug. 1995.

For conservationists, the apparent decline of some Nearctic-Neotropic
migratory birds is a serious concern (Terborgh 1989, Robbins et al. 1989)
that demands both rigorous management and quantitative monitoring of
breeding and wintering populations. Effective census techniques have
been developed for many songbird species (Passeriformes) during the
breeding season (Ralph and Scott 1981, Robbins et al. 1986). However,
songbirds rarely sing on their wintering grounds and during fall migration,
are feathered in duller basic plumages, and hence, are more difficult to
detect and identify than during the breeding season. Yet most studies of
migrants in the Neotropics employ censusing techniques designed for
breeding birds of higher latitudes, with little or no attempt to compensate
for the decreased detectability of wintering populations. As this report
will communicate, standard point count and transect censusing methods
can be relatively ineffective in determining the true abundance of secre-
tive migratory species that winter in dense tropical habitats.

The breeding and wintering biologies of Swainson’s Warbler {Lim-
nothlypis swainsonii) are among the most poorly known of the migratory
wood warblers (cf Meanley 1971, Eddleman 1978, Graves 1992, Brown
and Dixon 1994, Thomas 1994). Wintering birds have been reported in
the Caribbean basin from the Bahamas, Cuba and the Isle of Pines, Ja-
maica, Cayman Islands, and the Yucatan Peninsula, casually east to Puerto
Rico and the Virgin Islands, and from southern Veracruz south to Hon-
duras (Meanley 1971, AOU 1983, Pashley 1988, Winker et al. 1992).
Most wintering records are anecdotal (e.g., Eaton 1953), and, at present,
wintering sites where more than one or two individuals can be consis-

Dept. of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washington,
D.C. 20560.

94

Graves • CENSUSING SWAINSON’S WARBLERS

95

tently encountered per day are known only from certain regions of Cuba
(G. Wallace, pers. comm.; Kirkconnell et al., unpubl. ms.). This paper
documents the utility of playback techniques for censusing wintering pop-
ulations of Swainson’s Warbler and reports the discovery of what appears
to be a major wintering area for the species in the Blue Mountains of
Jamaica.

METHODS

Study area. — I surveyed Swainson’s Warbler populations along the same trails and roads
at Hardwar Gap that were studied by the Lacks, from 2 to 10 February 1995. Hardwar Gap,
at the divide forming the boundary between St. Andrew and Portland Parishes in the Port
Royal Mountains (an outlier of the Blue Mountains), is accessible along a hard-surfaced
road (B 1 ) from Kingston to Buff Bay on the northern coast. Hardwar Gap is one of the
most frequent destinations in Jamaica of resident and visiting ornithologists and bird watch-
ers. Both Hardwar Gap and the adjacent Hollywell Park now lie within the borders of the
Blue and John Crow Mountains National Park (Muchoney et al. 1994).

Three census transects were established along trails (“forest” transects in Table 1) that
pass through a peninsula of upper montane rain forest connected to a much larger block of
forest eastward in the national park (see Asprey and Robbins 1953 and Muchoney et al.
1994 for descriptions of habitat). The Hardwar Gap region was extensively affected by
Hurricane Gilbert in 1988 (Wunderle et al. 1992), and patches of early successional vege-
tation were present along all three forest transects, but there was little evidence of cutting
or other human disturbance away from roadsides. Transect terrain was extremely steep.
Slopes varied from 15° to 70° and were estimated to average over 40°. Significant portions
of forest trail No. 3 (Table 1) were at the summit of a knife-like ridge.

The fourth and fifth census routes followed the main road (Bl) from a point just below
Hardwar Gap to the village of Section, Portland Parish. This road forms the boundary
between relatively undisturbed montane forest (upslope) and a patchwork of lightly modified
montane forest, coffee plantations, residential gardens, and agricultural scrub (downslope)
in the 1 km buffer zone of the national park. The upper and lower sections of the road,
although contiguous, were treated as separate census routes (Table 1).

The length of “forest” transects was estimated by averaging the number of steps counted
on three downslope trips (at 0.76 m/step). The road transects were measured with a vehicle
odometer. Elevations were estimated by multiple readings of a calibrated altimeter.

Census methods. — I used techniques developed for monitoring breeding populations of
Swainson’s Warblers in the southeastern United States (Graves, unpubl.). Loop cassette tapes
(20 s and 60 s) were prepared from recordings (made a with Marantz PMD430 cassette
recorder and a Sennheiser ME 80 directional microphone in Virginia and Louisiana) of
Swainson’s Warbler call notes interspersed with a primary song every 15 s (one song on 20
s loop). Advertising songs were included because they are more easily heard above the low
frequency noise caused by ru.stling leaves and dripping water in the forest understory. The
opposite proportion of songs and calls are used to census breeding populations.

Recordings were broadcast from a dual-speaker “boom box” (Sanyo). Power output was
variable due to battery use and atmospheric conditions, but the audio output was adjusted
before every census run so that call notes were audible to me at a distance of 60 m and
songs at a somewhat greater distance (unobstructed by vegetation). The direction of the
speakers was rotated every 5 to 10 s until a probable response was detected. During clear
weather at Hardwar Gap, Swainson’s Warblers responded to tape broadcasts from distances

96

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Censuses of Swainson’s Warblers at Hardwar Gap, Blue Mountains, Jamaica

Census transect

Census

run

Starting

time

Census

duration

(m)

Weather'

Number

recorded

Cumu-

lative

number

Lorest trail

1

15:30

45

PC, O, L

1

1

no. U

2

06:55

120

PC, L

2

2

3

14:27

68

C

4

4

U8.0 warblers/ 10 h

Lorest trail

1

07:14

40

C

2

2

no. 2"^

2

07:55

69

C, PC

3

4

3

16:21

33

C

0

4

4

16:55

54

C, PC

2

4

5

13:21

50

O, L

1

4

6

14:15

45

O, L

1

4

18.6 warblers/10 h

Lorest trail

1

12:02

98

O, M, R

1

1

no. 3* *=

2

13:40

84

M, R

1

2

3

13:18

92

O, L

6

6

4

14:52

73

L R

2

6

5

07:16

178

C

6

10

6'

10:30

70

C, PC

7

12

23.2 warblers/ 10 h

Road (upper)®

1

07:05

190

PC

8

8

2

10:22

98

PC

7

11

3’’

10:00

60

O, PC

2

13

29.2 warblers/10 h

Road (lower)''

1

07:38

54

L M, R

1

1

2

08:35

57

L M, R

1

2

3

1 1:31

69

L M

1

3

4

12:42

70

L M

1

3

5

06:45

72

O

6

8

6

07:58

84

O

2

9

17.8 warblers/10 h

“ F = heavy fog; M = mist; R = rain; O = overcast; PC = partly cloudy; C = clear.

^ Length = 1604 m; elevation = 1 140-1200 m.

Total number of warblers recorded over all census runs standardized by duration ol censuses.

Length = 530 m; elevation = 1 140-1200 m.

'Length = 1808 m; elevation = 1140-1325 m.

•^Playbacks only at significant distributional gaps (at distances >150 m from nearest flagging marker).

* Length = 2980 m; elevation = 990-1 140 m.

^Length = 1409 m; elevation - 945-990 m.

Graves • CENSUSING SWAINSON’S WARBLERS

97

as great as 80 to 90 m (usually upslope or downslope). Response was significantly curtailed
during periods of heavy mist and rain, probably because the effective range of the tape
broadcast was attenuated by noise.

After a warbler responded and its location was marked with plastic flagging, I walked
quietly from the area until I could no longer hear the calling bird (80-90 m during clear
weather), at which point I resumed playback. Otherwise, tape playback was continuous and
only intermittently stopped to listen for suspected responses.

Territory size in wintering Swainson’s Warblers is unknown. “Playback-and-follow” trials
were performed on more than a dozen individuals. Eive warblers followed the playback
more than 50 m (66, 80, 113, 119, 122 m, respectively) from the initial point of response
(measured from a point perpendicular to the trail or road). Judging from this evidence and
the relatively large size of breeding territories (Graves 1992, unpubl. data), I conservatively
considered any subsequent response within 150 m of the original discovery point to refer
to the same bird unless two or more individuals were heard calling simultaneously in the
zone. When in doubt, I attempted to determine the approximate territorial boundary with
“playback-and-follow” trials. When a warbler’s territorial boundaries were reasonably un-
derstood, playback on subsequent census runs in that territory was halted after a response
was elicited.

RESULTS

Behavioral responses of Swainson’s Warblers to tape playback were
variable, ranging from close approaches (4-5 m) to the observer and
frequent back-and-forth flights accompanied by vigorous calling, to a sub-
dued approach (8-14 m) accompanied by a few faint chips and ventri-
loquial “seep” notes, following by rapid disappearance (within 30 s) of
the warbler.

On several occasions, vigorous responses often elicited counter-calling
from other individuals in adjacent territories (N = 5 different “counter-
calling” pairs). In one case, a highly agitated Swainson’s Warbler flew
across the road only to be promptly attacked by a counter-calling bird.
Only once did a warbler respond to playback by giving a primary song
(three somewhat incomplete renditions). Although my observations in
each territory were necessarily brief, no evidence of territorial overlap or
pairing was observed. The intense border displays and skirmishes ob-
served are the first evidence of intraspecific territoriality of Swainson’s
Warbler on its wintering grounds (cf Eaton 1953). With the exception of
the single warbler that sang (presumably male), the sex of individuals
was unknown. However, the timid vocal and behavioral responses of some
warblers were similar to the behavior of females on breeding territories
(Graves, unpubl. data).

Interspecific responses of other wintering wood warblers at Hardwar
Gap to Swainson’s Warbler call notes and songs were variable. Vigorous
counter-calling was elicited from the Ovenbird (Seiurus aiirocapillus).
Common Yellowthroats (Geothlypis trichas) frequently approached the
speakers and counter-called, and the single Louisiana Waterthrush (Seiu-

98

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

ms motacilla) encountered during the censuses approached from a dis-
tance of >50 m along a stream channel and called. More subdued re-
sponses were recorded for Black- throated Blue Warbler (Dendroica ca-
€ml€sc€ns). Black-and-white Warbler (^Mniotilto vofici), American Red-
start {Setophaga ruticilla), and Worm-eating Warbler (Helmitheros
vermivorus). The few Prairie Warblers (D. discolor) encountered along
the road transect and in bracken-dominated swards appeared to exhibit
little interest in Swainson’s Warbler playback.

Forty-two Swainson’s Warblers (68 census records) were recorded on
the five census transects. The number of detections ranged from four to
13 warblers/transect (Table 1), and relative abundance varied from 2.13
to 7.77 warblers/km. Playback of taped calls and songs was instrumental
in detecting and mapping the locations of individuals. Call notes were
heard on only four occasions when playback was not being used. Detec-
tion rates were consistently high under clear or overcast skies, but were
markedly lower during periods of heavy mist and light rain or immedi-
ately afterwards (e.g., forest trail no. 3: census runs 1, 2, and 4 vs 3, 5,
and 6). In general, a minimum of three or four census runs under clear
or overcast skies was necessary to locate and map Swainson’s Warbler
territories along each census transect.

Rates of Swainson’s Warbler detection along the five transect segments
(Table 1) varied from 17.8 to 29.2 warblers/10 h. Because of territorial
spacing, the number of individuals recorded along a transect is clearly
influenced by its length and the spatial distribution of high-quality habitat.
A few tentative comparisons can be made between my data (Table 1 ) and
Lack’s (1976: appendices 19 and 20), assuming our rates of movement
(0.5-2. 1 km/h vs. slow walk) were similar. Lack’s “slow walk’’ censuses
through montane forests yielded three Swainson’s/10 h (N = 32 h of
censuses) but none (N = 12 h) along roadsides at Hardwar Gap. Com-
parable figures with playback ranged from 18.0 to 23.2 Swainson s War-
blers/10 h (N = 18.7 h of censuses) along forest trails and from 17.8 to
29.2 warblers/10 h (N = 12.6 h of censuses) along roadsides. The contrast
between the Lacks’ and my roadside data are particularly striking.

Wunderle et al. (1992) failed to detect Swainson’s Warbler with fixed-
radius point count censuses in montane forest at Hardwar Gap either
before (December 1987) or after (January 1989) Hurricane Gilbert. Based
on my observations and previous work at Hardwar Gap (especially that
of the Lack’s 1972, 1976), I estimate that the use of playback is five to
ten times more efficient than standard point count or transect methods for
censusing wintering Swainson’s Warblers.

1 estimated that tape playback of Swainson’s Warbler calls and songs
sampled a path approximately 150 m wide along census transects in the

Graves • CENSUSING SWAINSON’S WARBLERS

99

Hardwar Gap area. Detectability is a function of territory placement. The
farther a territorial boundary is from the transect, the less likely the oc-
cupant will be detected, with or without tape playback. Thus, the cumu-
lative totals in Table 1 represent estimates of the minimum number of
warblers along each transect.

Mist-netting studies are likely to be less efficient in determining the
distribution and abundance of wintering Swainson’s Warblers. Because
Swainson’s Warblers appeared to be territorial and birds were dispersed
in the Blue Mountains (2.13 to 7.55 warblers/km of transect), mist nets
would have to be scattered widely to capture more than a few resident
birds, regardless of the number of net hours. Winter foraging behavior
appeared to be remarkably similar to that observed during breeding season
(Graves, unpubl. data; see Lack 1976). All foraging motions of Swain-
son’s Warblers observed during February 1995, were directed toward dead
leaves, most often in the leaf litter, but also toward dead leaves on fallen
logs or inclined tree trunks within 1 .5 m of the ground. Undisturbed birds
rarely flew more than a few meters and usually close to the leaf litter. If
observations from breeding territories might serve as a guide, foraging
Swainson’s on their wintering territories may walk up to mist nets, appear
to inspect the linear net lane and the net itself, and then proceed to walk
under or around them. The consequence of large and dispersed territories
combined with a terrestrial mode of foraging produce low rates of mist-
net capture, even when the habitat is “saturated” by wintering birds. In
any event, a large area (ca 100 ha) that would require hundreds of man
hours to sample adequately with mist nets may be efficiently censused
with playback in one or two days by a single observer.

As a note of caution, density estimates cannot be calculated from data
in Table 1 . Sections of the census transects were bordered by habitat that
appeared to be unoccupied by Swainson’s Warblers in the Hardwar Gap
area (e.g., bracken-dominated swards along forest transects; coffee, resi-
dential gardens, landslides, and agricultural scrub along the road). Tran-
sect data also need to be corrected for “switch-back” or “hair-pin” sam-
pling. This situation occurs when census transects switch directions
abruptly and the same area is sampled one or more times by the playback
broadcast. Wintering density at Hardwar Gap can be estimated only after
habitat patchiness and “hair-pin” sampling are accounted for (Graves,
unpubl. data).

In summary, Swainson’s Warblers appeared to be a widespread and
relatively common wintering resident in undisturbed and slightly modified
montane forest in the Hardwar Gap region of the Blue Mountains. At
present, more than 50,000 ha of apparently suitable habitat occurs within
the borders (including buffer zone) of the Blue and John Crow Mountains

100

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

National Park (Muchoney et al. 1994). If Swainson’s Warbler densities
are relatively uniform throughout this montane region, the park may sup-
port 17,000-25,000 wintering individuals (Graves, unpubl. data). Regard-
less, the 42 individuals found at Hardwar Gap constitute the largest
known wintering population of Swainson’s Warbler.

DISCUSSION

History and status of Swainson’s Warbler in Jamaica. — The distribu-
tion and abundance of Swainson’s Warbler in Jamaica is inadequately
known, perhaps owing to its secretive behavior and cryptic appearance.
The first specimen from Jamaica was collected by Edward Newton at
Hope, St. Andrew Parish, on 8 February 1879 (Newton 1879). This spec-
imen was the first of a series of eight collected by Newton from 1 October
to 8 April (1879-1882) at Hope, Hermitage, and Mt. Elizabeth, St. An-
drews Parish, and at Port Royal, St. Thomas Parish (Merriam 1885). Bond
(1940) overlooked Merriam’s paper and referred implicitly to Newton’s
first specimen as the only record for the island. Five years later. Bond
(1945:118) noted that Swainson’s Warbler “winters in Jamaica (at least
nine records from October 1 to April 8).’’

Foreshadowing the findings of Fack some twenty years in the future,
Tordoff (1952:321) reported, “I collected three female Swainson’s War-
blers in the winter of 1946-47 . . . within 15 miles of Kingston. In ad-
dition, I saw at least nine others between December 31 and February 7.
On two occasions I saw three in one day. My observations indicate that
this warbler winters in Jamaica in fair numbers. Six of the individuals
that I saw were in dry lowland woods; the rest were in damp forests in
the hills north of Kingston (at Hermitage, St. Andrew Parish).’’

In the early 1970s, Fack (1976:174-175) found Swainson’s Warbler to
be “regular in lowland woods and montane forest with thick undergrowth,
and also in the tall dark forest with a dense canopy but no undergrowth.”
Fack recorded 1-3 birds/ 10 h of observation in dry limestone forest near
sea-level at Morant Point and Negril, in arid ruinate forest on Fong Moun-
tain, and in rich secondary forest in the FeiTy River valley and Mona
Wood. In montane forest at Hardwar Gap in the Blue Mountains, he
observed three Swainson’s Warblers/ 10 h on slow walks, and six warblers/
10 h on fast walks along forest trails, but none along the roadside edges
of the forest. Fack also noted that five Swainson’s were banded but none
was observed in the overgrown gardens at Greenhills, about 1.5 km ENE
of Hardwar Gap in Portland Parish.

Compilations of the Gosse Bird Club (1963—1993) indicate that Swain-
son’s Warbler was banded (cumulative total of 25 individuals) or observed
nearly annually from 1963 through 1977, most frequently at Mona Woods

Graves • CENSUSING SWAINSON’S WARBLERS

101

near Kingston (see Diamond and Smith 1973). The last published record
for Jamaica in the 1970s was reported from the Blue Mountains in Oc-
tober 1977. Inexplicably, there were no additional published reports of
Swainson’s Warbler in Jamaica during the next 15 years until December
1992.

Three recent unpublished observations have come to my attention. Pe-
ter Marra (pers. comm.) banded 3^ Swainson’s Warblers on a 5-ha study
site in dry limestone forest at Kew Park, Westmoreland Parish. Russell
Greenberg (pers. comm.) observed one or two per day in the Hardwar
Gap area during six days of field work in January 1984. Finally, Robert
Sutton (pers. comm.) banded Swainson’s Warblers at Greenhills in 1978,
1982, 1985, and 1986 (N = 10 individuals). In summary, the field studies
of Tordoff (1952), and especially those of Lack and Lack (1972) and
Lack (1976), combined with scattered sight and banding records compiled
by the Gosse Bird Club (1963-1993), as well as the results of the present
study suggest that Swainson’s Warbler is a widespread wintering species
occurring at low densities in Jamaica.

Conclusions. — The most significant recent advance in population stud-
ies of wintering songbirds or resident Neotropical species has been the
use of tape-recorded playback of call notes and songs of focal species
(e.g., Parker 1991). Among the wood warblers (Parulini; taxonomy of
Sibley and Monroe 1990), tape playback has been used successfully to
census Black-throated Blue Warblers in Jamaica (Holmes et al. 1989,
Sliwa and Sherry 1992) and Puerto Rico (Wunderle 1992), American
Redstarts in Jamaica (Holmes et al. 1989, Sliwa and Sherry 1992), Ken-
tucky Warblers (Oporornis formosus) in Panama (Mabey and Morton
1992), and Hooded Warblers (Wilsonia citrina) in the Yucatan Peninsula
(Lynch et al. 1985).

Swainson’s Warbler may be added to the list of Nearctic-Neotropic
migrants that are most effectively monitored on their wintering grounds
with tape playback. Although playback census techniques are perhaps less
crucial for wintering species that are behaviorally conspicuous and occupy
open habitats (e.g.. Palm Warbler [Dendroica palmarum]), results from
the cited field studies suggest that censusing efficiency can be significantly
enhanced for all wintering wood warblers with species-specific tape
broadcast.

At Hardwar Gap, the entire wintering warbler fauna could be censused
along the same census transects by including call notes of the nine most
common species on a single loop cassette or audio compact disc (CD)
Black-throated Blue Warbler {[Dendroica caerulescens]. Prairie Warbler
[D. discolor]. Black-and-white Warbler [Mniotilta varia], American Red-
start [Setophaga ruticilla]'. Worm-eating Warbler [Helniitheros vermivo-

102

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

rus}, Swainson’s Warbler [Limnothlypis swainsonii], Ovenbird [Seiurus
aurocapillus], Louisiana Waterthrush [S. motacilla], and Common Yel-
low-throat [Geothlypis trichas]). Extended cuts of particular species could
be played as needed (cut selection on CD or separate loop cassettes).
Tape broadcast of “spishing” noises, call notes of a single species, or
owl calls, while better than nothing, are less effective in multi-species
censuses of wintering warblers (Graves, unpubl.).

ACKNOWLEDGMENTS

I thank Russell Greenberg, Catherine Levy, Brooke Meanley, John Rappole, George Wal-
lace, David Wiedenfeld, and Kevin Winker for insightful comments on the manuscript.
Robert Sutton provided banding data. Logistics in Jamaica were greatly facilitated by David
Smith and Susan Anderson (Jamaica Conservation and Development Trust, JCDT) and
members of the Gosse Bird Club. Harold Thomas, Hyacinth Pascoe, and Dwight Pryce
assisted my work at Hollywell. Permission to work in the Blue and John Crow Mountains
National Park was granted by Lranklin McDonald of the Natural Resources Conservation
Authority. Lield work was supported by the JCDT in cooperation with Smithsonian Insti-
tution.

LITERATURE CITED

American Ornithologists’ Union. 1983. Check-list of North American birds, 6th ed.
A.O.U., Washington, D.C.

Asprey, G. E and R. G. Robbins. 1953. The vegetation of Jamaica. Ecol. Monogr. 23:
359^12.

Bond, J. 1940. Check-list of birds of the West Indies. Acad. Nat. Sci., Philadelphia, Penn-
sylvania.

. 1945. Check-list of birds of the West Indies. Acad. Nat. Sci., Philadelphia, Penn-
sylvania.

Brown, R. E. and J. G. Dixon. 1994. Swainson’s Warbler {Limnothlypis swainsonii). In
The Birds of North America, No. 126 (A. Poole and L Gill, eds.). Academy of Natural
Sciences, Philadelphia; American Ornithologists’ Union, Washington, D.C.

Diamond, A. W. and R. W. Smith. 1973. Returns and survival of banded warblers wintering
in Jamaica. Bird Banding 44:221—224.

Eaton, S. W. 1953. Wood warblers wintering in Cuba. Wilson Bull. 65:169-174.

Eddleman, W. R. 1978. Selection and management of Swainson’s Warbler habitat. MSc.
thesis, Univ. of Missouri, Columbia, Missouri.

Gosse Bird Club. 1963-1993. Broadsheets Nos. 1-60.

Graves, G. R. 1992. A case of aggregated nest placement and probable polygyny in the
Swainson’s Warbler. Wilson Bull. 104:370—373.

Holmes, R. T, T. W. Sherry, and L. Reitsma. 1989. Population structure, territoriality and
overwinter survival of two migrant warbler species in Jamaica. Condor 91:545-561.

Kirkconnell, A, G. E. Wallace, and O. H. Garrido. 1995. Notes on the status and
behavior of the Swainson’s Warbler in Cuba. Wilson Bull 108:175-178.

Lack. D. 1976. Island biology illustrated by the land birds of Jamaica. Univ. California
Press, Berkeley, California.

AND P. Lack. 1972. Wintering warblers in Jamaica. Living Bird 11:129-153.

Lynch, J. E, E. S. Morton, and M. E. van der Voort. 1985. Habitat segregation between
the sexes of wintering Hooded Warblers (Wilsonia citrina). Auk 102:714-721.

Graves • CENSUSING SWAINSON’S WARBLERS

103

Mabey, S. E. and E. S. Morton. 1992. Demography and territorial behavior of wintering
Kentucky Warblers in Panama. Pp. 329-336 in Ecology and conservation of Neotropical
migrant landbirds (J. M. Hagen and D. W. Johnston, eds.). Smithsonian Institution Press,
Washington, D.C.

Meanley, B. 1971. Natural history of the Swainson’s Warbler. North American Fauna No.
69. U.S. Dept. Interior, Washington, D.C.

Merriam, C. H. 1885. Swainson’s Warbler in Jamaica. Auk 2:377.

Muchoney, D. M, S. Iremonger, and R. Wright. 1994. A rapid ecological assessment of
the Blue and John Crow Mountains National Park, Jamaica. Nature Conservancy, Ar-
lington, Virginia.

Newton, A. 1879. Exhibited bird-skins obtained in Jamaica. Proc. Zool. Soc. London 1879:
552-553.

Parker, T. A., III. 1991. On the use of tape recorders in avifaunal surveys. Auk 108:443-
444.

Pashley, D. N. 1988. Warblers of the West Indies 11. The western Caribbean. Carib. J. Sci.
24:112-126.

Ralph, C. J. and J. M. Scott. 1981. Estimating numbers of terrestrial birds. Studies in
Avian Biology no. 6.

Robbins, C. S., D. Bystrak, and P. H. Geissler. 1986. The breeding bird survey: Its first
fifteen years, 1965-1979. U.S. Fish and Wildlife Serv. Resource Pub. 157.

, J. R. Sauer, R. S. Greeenberg, and S. Droege. 1989. Population declines in

North American birds that migrate to the neotropics. Proc. Natl. Acad. Sci. 86:7658-
7662.

Sibley, C. G. and B. L. Monroe, Jr. 1990. Distribution and taxonomy of birds of the
world. Yale Univ. Press, New Haven, Connecticut.

Sliwa, a. and T. W. Sherry. 1992. Surveying wintering warbler populations in Jamaica:
point counts with and without broadcast vocalizations. Condor 94:924—936.

Terborgh, j. 1989. Where have all the birds gone? Princeton Univ. Press, Princeton, New
Jersey.

Thomas, B. G. 1994. Habitat selection and breeding status of Swainson’s Warbler. Thesis,
University of Missouri, Columbia.

Tordoff, H. B. 1952. Notes on birds of Jamaica. Auk 69:320-322.

Winker, K., R. J. Oehlenschlager, M. A. Ramos, R. M. Zink, J. H. Rappole, and D. W.
Warner. 1992. Avian distribution and abundance records for the Sierra de los Tuxtlas,
Veracruz, Mexico. Wilson Bull. 104:699-718.

Wunderle, j. M., Jr. 1992. Sexual habitat segregation in wintering Black-throated Blue
Warblers in Puerto Rico. Pp. 299-307 in Ecology and conservation of Neotropical
migrant landbirds (J. M. Hagen and D. W. Johnston, eds.). Smithsonian Institution Press,
Washington, D.C.

, D. J. Lodge, and R. B. Waide. 1992. Short-term effects of Hurricane Gilbert on

terrestrial bird populations on Jamaica. Auk 109:148-166.

Wilson Bull., 108(1), 1996, pp. 104-114

COLONY-SITE AND NEST-SITE USE BY COMMON
CRACKLES IN NORTH DAKOTA

H. Jeffrey Homan,' George M. Linz,^

William J. Bleier,' and Robert B. Carlson^

Abstract. — We searched 638 quarter sections (0.8 X 0.8 km) for Common Crackle
{Quiscalus quiscula) nesting sites in Benson County, North Dakota, in 1989 and 1990. We
found 3596 active nests in 202 colonies on 177 quarter sections. Colonies in shelterbelts
next to inhabited farmsteads were found at greater than expected frequencies {P ^ 0.05),
whereas colonies in vegetation associated with potholes and miscellaneous habitats (woods,
ravines, railroad easements, and lakesides) occurred below expected frequencies. Nest sites
in stands of vegetation >100 m from farmstead residences occuiTed less frequently than
expected (P < 0.05). Within colonies, nest sites in blue spruce {Picea pungens), Siberian
elm (Ulmus pumila) and black poplar (Populus nigra) were found at greater than expected
frequencies (P ^ 0.05) according to these species’ availabilities, while green ash {Fraxinus
pennsylvanica), American elm {Ulmus americana), and Russian olive (Elaeagnus angusti-
folia) were used below expected frequencies. The Common Crackle’s preference for shel-
terbelts near inhabited farmsteads affected the physical and vegetative characteristics of
colony sites and nest sites; with the exception of hawthorn {Crataegus rotundifolia), colo-
nized stands had species compositions typically found in multi-rowed farmstead shelterbelts
in North Dakota. Received 7 Feb. 1995, accepted 25 Aug. 1995.

The North Dakota population of breeding Common Crackles {Quis-
calus quiscula) has more than doubled to 768,000 pairs (Nelms et al.
1994) from initial estimates made in 1967 (Stewart and Kantrud 1972).
In the northern Great Plains, a region of intensive agricultural production,
rows of shrubs and trees (shelterbelts) may be important nesting habitats
for Common Crackles (Yahner 1982). In South Dakota and Minnesota,
more than 50% of the birds nesting in multi-rowed shelterbelts were Com-
mon Crackles (Field 1971, Yahner 1982). North Dakota was historically
dominated by prairie grasslands, and Common Crackles were restricted
to nesting in the vegetation of riparian habitats, wetlands, and towns
(Coues 1878). Recent plantings of numerous shelterbelts for agricultural
and other purposes (e.g., insulation and beautification of farmsteads) may
have enhanced the Common Crackle’s access to prime nesting sites in
North Dakota.

Records from the Cornell Laboratory of Ornithology indicate that co-
niferous and deciduous trees are most frequently used by grackles nest
sites with 24% and 14% of the nests, respectively (Maxwell et al. 1976).

' Dept, of Zoology, North Dakota State Univ., Fargo, North Dakota 58105.

^ U.S. Dept, of Agriculture, Denver Wildlife Research Center, North Dakota Field Station, Stevens Hall,
North Dakota State Univ., Fargo, North Dakota 58105.

^ Dept, of Entomology, North Dakota State Univ., Fargo, North Dakota 58105.

104

Homan et al. • COMMON CRACKLE NESTING

105

However, quantitative comparisons of nest-site use by Common Crackles
have been made only by Field (1971) and Yahner (1982). These studies
involved nest-site use at the substrate level, and broader perspectives of
colony-site use and habitat use were not investigated.

Our objectives were to determine habitat, colony-site, and nest-site use
by Common Crackles in northcentral North Dakota. Our data on the pre-
ferred nesting sites of Common Crackles may benefit participants in shel-
terbelt-planting efforts (e.g., the North Dakota Centennial Tree Planting
Project) who wish to avoid creating favorable nesting habitat for this
species because it can damage crops.

STUDY AREA AND METHODS

Benson County is located in northcentral North Dakota in the north-
eastern Drift Plain Physiographic Region (Stewart 1975). The topography
is flat to gently rolling, consisting of croplands interspersed with numer-
ous potholes, temporary wetlands, and shelterbelts. The county is pri-
marily cropland (74%). The remaining land area is dedicated to rangeland
and pasture (17%) and woodlands, federal non-croplands, and other lands
(9%). About 97% (5666 ha) of Benson County’s native woodlands are in
the east in the Devils Lake and Wood Lake regions. Water bodies >16
ha represent 2% of the county. Siberian elm (Ulmus pumila) is the main
species found in single-row shelterbelts. Multi-row shelterbelts consist of
various combinations of species including Siberian elm, caragana (Car-
agana aborescens), green ash (Fraxinus pennsylvanicd), boxelder {Acer
negundo), plains cottonwood (Populus deltoides), and blue spruce (Picea
pungens). In low-lying areas, willow (Salix spp.), quaking aspen {Populus
tremuloides), and plains cottonwood grow naturally. Hawthorn {Cratae-
gus rotundifolia), chokecherry {Prunus virginiana), and wild plum {Pru-
nus americana) occur frequently in pastures and uncultivated areas. Large
stands of bur oak {Quercus macrocarpa) are found in the hill region
surrounding Devils Lake Basin.

Long-term average precipitation is 44 cm, with 72% of it falling in
April-September (North Dakota Agricultural Statistics Serv. 1990). In
May, the peak breeding period for Common Crackles in Benson County,
the average temperature is 12°C. Average dates of first and last frosts
(0°C) are 13 September and 23 May, respectively.

From 18 May through 10 June 1989-1990, we located active nests by
systematic walk-through surveys on 638 randomly selected quarter sec-
tions (0.8 km X 0.8 km). All vegetation capable of supporting a Common
Crackle nest was searched. Surveys were made daily from 09:00 to 18:
00 h. A nest was considered active if it contained eggs, nestlings, or was
defended by adults. An extendable pole with a mirror was used to check

106

THE WILSON BULLETIN • Vo/. 108, No. I, March 1996

for eggs and young. All nest sites were marked with colored mylar tape
attached at the base of the nesting substrate.

We defined a nest site as the substrate on which a nest was built. A
colony site was the stand of vegetation in which a nest occurred, with a
stand being any continuous body of vegetation separated from all other
such bodies by at least 50 m. We selected 50 m because colony sites were
usually not defended at distances >50 m and thus could be considered
distinct from other stands (Gutzwiller and Anderson 1987).

At each colony site, five nest sites and five potential nest sites (controls)
were randomly selected. Controls included vegetation > 1 m in height and
capable of supporting a nest. The controls were selected by randomly
drawing numbers and converting these numbers to meters on an x-y axis
defined by the length and width of the colony site. If the coordinates did
not fall on a suitable control, coordinates were redrawn until five controls
were chosen. When <5 nest sites were present at a colony site, data were
collected for all nest sites. We recorded plant species and trunk diameter
at breast height (DBH), vegetation height, nest height, and distances to
nearest permanent water (DPW) and residence (DRS) in 0—100, 101—300,
301-500, and >500 m categories. In 1990, we measured distance from
nest sites to edge (DEG). Edge was defined as the border of any opening
>5 m across where vegetation was :^1 m. Heights of nests, nesting sub-
strates, and controls were estimated with a telescoping pole or clinometer.
Distance measurements >500 m were estimated with an optical range
finder. A measuring wheel was used for distances <500 m.

If uncolonized stands were in a colonized quarter section (quarter), we
gathered data from five controls allocated randomly among the stands.
Typically, uncolonized stands were single-row shelterbelts of Siberian elm
or low-lying areas dominated by willow, quaking aspen or plains cotton-
wood. If only one uncolonized stand was present, data for all five controls
were drawn from this stand. Data were pooled across colonized quarters,
and controls from uncolonized stands were compared against controls
from the colonized stands.

Additionally, we made comparisons between colonized and uncolo-
nized quarters. Controls from 70 uncolonized quarters that were surveyed
for Common Crackles during 1989-1990 were compared against the com-
bined controls from the colony sites and unused stands in the colonized
quarters. Eive controls were randomly selected from each of the colonized
and uncolonized quarters.

Colony sites and uncolonized stands were classified according to the
following habitats: inhabited farmstead shelterbelts, abandoned farmstead
shelterbelts, windbreaks (agricultural shelterbelts), towns, potholes, pas-
tures, and miscellaneous. When delineating habitats for uncolonized

Homan et al. • COMMON CRACKLE NESTING

107

stands, potholes were considered as habitat only if surrounded by shrubs
or trees; no Common Crackle nest sites in cattail (Typha spp.) were ob-
served during our two years of surveys in Benson County. Because of
the continuous nature of vegetation distributions in towns, all nests in
habitat classified as town were attributed to a single colony site.

County-wide habitat availabilities were estimated with the non-map-
ping technique (Marcum and Loftsgaarden 1980). We selected 200 quar-
ters from our 1989-1990 surveys. Only quarters with vegetation capable
of supporting Common Crackle nests were used. Five controls from each
quarter were selected by placing an 80-grid, transparent sheet on an aerial
photograph of the quarter and randomly selecting grids. Only colonizable
habitats were selected, and non-nesting areas (e.g., croplands, roads, and
water bodies) were not used.

We tested four null hypotheses: (1) DRS and DPW categories and
vegetation were used as nest sites according to their availabilities in the
colonies, (2) use of stands in colonized quarters was independent of both
vegetation composition and DRS and DPW categories, (3) use of quarters
was independent of vegetation and DRS and DPW categories, and (4)
colony sites were distributed among the seven habitat categories in pro-
portion to county-wide habitat availabilities. We used G-tests for goodness
of fit to determine if actual use differed from expected (null) use for
habitats, nest-site vegetation, and DRS and DPW categories in the colony
sites (Sokal and Rolf 1981). If the G-tests were significant (P < 0.05),
preference and avoidance were estimated using the Bonferroni method
with an a = 0.05 (Neu et al. 1974, Byers et al. 1984, Thomas and Taylor
1990). We used G-tests of independence to compare colony sites with
uncolonized stands and to compare colonized and uncolonized quarters.
Nest-site vegetation used <2% of the time was combined into a miscel-
laneous category. Vegetation height and DBH and DEG variables could
not be transformed to approximate normality; therefore, Wilcoxon two-
sample tests were used (Sokal and Rolf 1981). Pairwise comparisons of
vegetation heights and DBH were made only for preferred and avoided
species as determined by Bonferroni tests.

RESULTS

During our two-year study, we found 202 colonies with 3596 active
nests on 177 of the 638 quarters surveyed. Thus, Common Crackles had
a mean colony and nest density of 0.49 colonies and 8.81 nests per km^
in Benson County, with 28% of the quarters occupied. In decreasing order,
the most frequently used nesting substrates were blue spruce (N = 924),
Siberian elm (N = 819), boxelder (N = 427), caragana (N = 238), and
hawthorn (N = 230). Habitat classified as inhabited farmstead shelterbelt

108

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 1

Comparisons" of Use and Availability of Seven Habitats Occupied by 202 Common
Crackle Colonies in Benson County, North Dakota, during 1989-90

Habitat

Availability
(N = 1000)

%

Use

(N = 202)

%

Preference'’

Windbreak

25.6

24.3

0

Miscellaneous'^

24.1

11.4

—

Pothole

22.7

6.9

—

Inhabited farmstead

10.9

32.2

+

Pasture

7.7

9.9

0

Abandoned farmstead

7.5

12.9

0

Town

1.5

2.5

0

^ G-test for goodness of fit: G = 90.2, df = 6, < 0.0001.

” indicates preference, “ — ” avoidance, and “0” use according to availability. Preference was determined with
Bonferroni confidence intervals (a = 0.05) placed on use.

Composition of the miscellaneous habitat category: lakesides (35%), woods (32%). roadsides (14%), ravines (14%),
ditches (3%), and railroad easements (2%).

accounted for 57.8% of the nests. Towns, which occurred on five quarters,
had the highest mean number of nests (N = 5, jc = 142.5 nests/km^, SE
= 42.8).

Habitats were not colonized in proportion to their availabilities (G =
90.2, 6 df, P < 0.0001). Shelterbelts of inhabited farmsteads were used
more frequently than expected, while potholes and miscellaneous habitats
(e.g., woods, ravines, railroad easements, and lakesides) were used below
expected frequencies (Table 1). Abandoned farmstead shelterbelts, pas-
tures, towns, and windbreaks were used according to availabilities. Col-
onies were larger (Wilcoxon Two-sample Test: Z = 4.9, P < 0.0001) on
quarters with inhabited farmsteads {x = 50.0 nests/km^, SE = 5.5, N =
65) than on quarters with abandoned farmsteads (x = 14.4 nests/km^, SE
= 4.1, N = 26).

The use of quarters depended on plant species composition (G = 251.1,
18 df, P < 0.0001). Green ash, blue spruce, wild plum, and hawthorn
occurred more frequently on colonized quarters (Table 2). Uncolonized
quarters were typified by quaking aspen, plains cottonwood, and willow.
We failed to detect differences between controls of colonized and unco-
lonized quarters for either vegetation heights or DBHs (all Ps > 0.05).
Distance categories of controls differed between colonized and uncolon-
ized quarters for farmsteads (G = 379.3, 3 df, P < 0.0001) and permanent
water (G = 49.8, 3 df, P < 0.0001), with more controls on uncolonized
quarters >500 m from both of these features.

Use of stands within colonized quarters was dependent on vegetation

Honum et al. • COMMON CRACKLE NESTING

109

Table 2

Comparisons^ between Randomly Selected Control Vegetation (N = 5) from
Quarter Sections Colonized by Common Crackles and Unused Quarter Sections

Vegetation

Uncolonized
quarter sections
(N = 350)

%

Colonized
quarter section^

(N = 880)

%

Preference^

Siberian elm (Ulmus pumila)

19.1

16.2

0

Blue spruce (Picea pungens)

0.0

3.8

+

Boxelder (Acer negundo)

8.6

11.9

0

Hawthorn (Crataegus rotundifolia)

1.4

4.2

4 .

Caragana (Caragana arborescens)

4.9

5.8

0

Green ash (Fraxinus pennsylvanica)

7.7

11.9

+

Chokecherry (Prunus virginiana)

6.9

8.4

0

Willow (Salix spp.)

21.4

12.5

—

Lilac (Syringa vulgaris)

1.1

1.9

0

Wild plum (Prunus americana)

0.3

3.0

+

Black poplar (Populus nigra)

0.0

1.0

0

Miscellaneous'*

1.4

1.6

0

Bur oak (Quercus macrocarpa)

2.0

0.0

0

Honeysuckle (Lonicera tatarica)

0.6

1.5

0

Juneberry (Amelanchier canadensis)

0.6

0.8

0

American elm (Ulmus americana)

0.9

2.3

0

Aspen (P. tremuloides)

11.7

5.1

—

Russian olive (Elaeagnus angustifolia)

2.0

2.2

0

Cottonwood (P. deltoides)

9.4

5.9

—

“ G-lesl of independence: G = 251.1, df = 18, P < 0.0001.

•’One colonized quarter section consisted only of an abandoned shed surrounded by wheat and was not used in the
analysis of vegetation.

+ indicates preference, “ — ” indicates avoidance, and “O'* indicates use according to availability. Selection was
determined with Bonferroni confidence intervals (a = 0.05) placed on the vegetation from colonized quarter sections.

^ Vegetation comprising <2% of the combined categories.

composition (G = 252.4, 17 df, P < 0.0001) (Table 3). Stands with
Siberian elm, green ash, boxelder, caragana, hawthorn, blue spruce, lilac
(Syringa vulgaris), and American elm (Ulmus americana) were more like-
ly to be colonized than stands consisting of willow, plains cottonwood,
and quaking aspen. Plains cottonwood had a larger DBH (P = 0.04)
within colony sites (x = 39.0 cm, SE = 4.2) than in uncolonized stands
(x = 26.2 cm, SE = 1.9). No differences in heights of control vegetation
were detected between colony sites and uncolonized stands (all Ps >
0.05). The DRS categories were not independent between used and un-
used stands (G = 221.4, 3 df, P < 0.0001), with stands of vegetation in
the 0-100 m and 101-300 m DRS categories colonized more frequently
than stands in 301-500 and >500 m categories. The DPW categories
between colony sites and unused stands were independent (G = 4.7, 3
df, P = 0.192).

110

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 3

COMPARISON‘‘ WITHIN QUARTER SECTIONS USED BY NESTING COMMON GRACKLES BETWEEN

Randomly Selected Controls (N = 5) from Unused Stands of Vegetation and

Colonized Stands

Species

Unused stands
(N = 405)

%

Colonized stands
(N = 979)

%

Preference*’

Willow

33.6

8.0

—

Cottonwood

13.8

4.3

—

Siberian elm

11.8

17.7

+

Chokecherry

8.6

7.2

0

Aspen

6.2

4.0

—

Green ash

4.0

13.0

+

Boxelder

3.7

14.0

+

Wild plum

3.5

3.8

0

Caragana

3.0

7.0

+

Miscellaneous‘S

2.7

2.4

0

Russian olive

2.5

1.8

0

Hawthorn

2.2

4.8

+

Blue spruce

1.5

3.9

+

Honeysuckle

1.0

1.7

0

Black poplar

0.7

0.9

0

Juneberry

0.7

1.1

0

Lilac

0.2

2.1

+

American elm

0.2

2.2

+

“ G-lest of independence: G = 252.4, df = ]1, P < 0.0001.

^“ + " indicates preference, “ — ” indicates avoidance and “0’' indicates use according to availability. Selection was
determined with Bonferroni confidence intervals (a = 0.05) placed on the vegetation from colonized stands.
Miscellaneous category consisted of species that formed <2% of the combined categories.

Within colony sites, nesting vegetation was not used according to avail-
ability (G = 203.6, 17 df, P < 0.0001). Blue spruce, Siberian elm, and
black poplar were used at greater than expected frequencies, whereas
green ash, willow, American elm, quaking aspen, Russian olive, and
plains cottonwood were used below expected frequencies (Table 4). The
DRS categories were not used in proportion to their availabilities in col-
ony sites (G == 30.6, 3 df, P < 0.0001). Nest sites <100 m from farm-
steads were used more frequently than expected; all other DRS categories
were used below expected frequencies. The DRW categories were not
distributed randomly between controls and nest sites (G = 8.8, 3 df, P =
0.032); nest sites in the 301-500 m category were used more frequently,
while nest sites >501 m from permanent water were used less frequently.
Nest-site heights were greater than controls (all Ps < 0.05) for Siberian
elm, blue spruce, green ash, and American elm. The DBHs of nest sites
were larger than controls for green ash and Siberian elm, while nest-site

Homan et al. • COMMON CRACKLE NESTING

Table 4

Comparisons" within Colonies between Randomly Selected Vegetation (N = 5) Used
BY Nesting Common Crackles and Randomly Selected Unused Vegetation

Species

Use

(N = 799)

%

Availability
(N = 979)

%

Preference^

Siberian elm

23.6

17.7

+

Blue spruce

15.5

3.9

+

Boxelder

11.9

14.0

0

Hawthorn

7.5

4.8

0

Caragana

7.5

7.0

0

Green ash

6.3

13.0

—

Chokecherry

5.4

7.2

0

Willow

4.8

8.0

—

Lilac

4.3

2.2

0

Wild plum

3.5

3.8

0

Black Poplar

2.8

0.9

+

Miscellaneous*'

2.5

2.4

0

Honeysuckle

1.1

1.7

0

Juneberry

1.1

1.1

0

American elm

0.8

2.2

—

Aspen

0.8

4.0

—

Russian olive

0.6

1.8

—

Cottonwood

0.1

4.3

—

“ O-test for goodness of fit: G = 203.6, df = 17, P < 0.0001.

^ “+’* indicates preference. “ — ” indicates avoidance and “O’" indicates use according to availability. Selection was
determined with Bonferroni confidence intervals (a = 0.05) placed on use.

Miscellaneous category consisted of substrates that formed <2% of the combined use and availability categories.

DBH was smaller for caragana. In 1990, nest sites were placed randomly
with respect to DEG (Z = 1.6, P = 0.12).

DISCUSSION

Common Crackles prefer shelterbelts of inhabited farmsteads over six
other habitat categories. Windbreaks, which offer structurally similar nest-
ing substrates and are often adjacent to farmstead habitats, are used only
according to availability. Windbreaks in North Dakota are usually single-
rowed structures of Siberian elm and lack the areal extent and species
heterogeneity of multi-rowed farmstead shelterbelts. Areal extent and spe-
cies heterogeneity, however, can not account for the preference shown by
Common Crackles for shelterbelts next to inhabited rather than abandoned
farmsteads. Both classes of farmstead shelterbelts are of comparable size
and species composition. Common Crackles may prefer shelterbelts of
inhabited farmsteads because of the increased access to invertebrates.
During the 1989 nesting season in Benson County, Common Crackles

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

1 12

used invertebrates as their major food (Homan et al. 1994). By establish-
ing colonies near the maintained landscapes of active farmsteads, the birds
may improve their foraging success for invertebrates (Yahner 1982). An-
thropogenic supplementation of food and water resources (e.g., spilled
grains, food, and water for livestock and pets) may further encourage
Common Crackle colonization by inhabited farmsteads (Martin 1978.

141).

Habitat use was the broadest measurement of nesting distribution m

our study. The vegetation and DRS data show that Common Crackles
primarily nest by farmsteads. Excepting hawthorn, the vegetation char-
acteristics associated with Common Crackle colonies were typical of plant
species compositions found in multi-rowed farmstead shelterbelts. The
use of pastures for colony sites probably cause hawthorn (and to a lesser
extent wild plum) to be included with the shelterbelt vegetation. The
apparent discrepancies among our vegetation analyses of Common Crack-
le colony and nest-site use were due to the scalar design of our study.
For example green ash, which is avoided by Common Crackles within
colony sites, is often planted in farmstead shelterbelts with more preferred
nesting substrates (e.g., blue spruce); thus, green ash becomes an indicator
of colonized quarters by association with preferred species.

In colony sites. Common Crackles apparently prefer blue spruce, Si-
berian elm, and black poplar over other plant species. These species are
profusely branched, which aids in nest attachment; moreover, their dense
foliage probably offers concealment and protection from excessive heat
loss or gain. A warmer microclimate may allow for earlier initiation of
egg laying (Erskine 1971). The foliage of Siberian elm and black poplar
does not appear until May in northcentral North Dakota, and blue spruce
(or other dense conifers) is the only nesting substrate favorable for ini-
tiating nests in April, the beginning of the breeding season in the state
(Stewart 1975).

The rarity of blue spruce in habitats not classihed as farmsteads made
it impossible to directly separate the influence of human activity from the
influence exerted by the structural characteristics of blue spruce. However,
the combination of a preferred macrohabitat (inhabited farmstead) with a
preferred microhabitat (blue spruce) may present the most favorable en-
vironment for nesting Common Crackles. The infrequent use of pothole
habitat may have confounded the avoidance shown by nesting Common
Crackles for structurally open vegetation, such as quaking aspen and
plains cottonwood, but open-structured shelterbelt vegetation (e.g., green
ash, American elm, and Russian olive) within colony sites was also used
below expected frequencies. Our data support those of Yahner (1982) and
Field ( 1971 ) who observed that green ash and other open shelterbelt plant

Homan et al. • COMMON CRACKLE NESTING

113

species were used infrequently or avoided by nesting Common Crackles.
Avoidance is probably a result of the lack of secure nest attachments
inherent in open-structured vegetation.

The presence of permanent water probably affects nesting behavior of
Common Crackles. Similar observations concerning the association be-
tween Common Crackle colonies and water have been made (Erskine
1971, Bent 1958:398-399, Martin 1978:141). However, the birds may be
responding to vegetation supported by water rather than to water itself
(Erskine 1971). We located 14 colonies by potholes; although this habitat
is avoided compared to availability, potholes may influence the associa-
tion of colonies with permanent water.

Nesting substrate height may also be involved in nest-site selection,
with Common Crackles displaying a preference for taller vegetation. Us-
ing taller vegetation allows for building nests at greater heights, which
may provide for earlier detection of predators (Cutzwiller and Anderson
1987, Bekoff et al. 1987). Additionally, males often use taller vegetation
for displaying (Petersen and Young 1950, Wiens 1965, Wiley 1976); these
displaying sites may later become nest sites.

Planting more green ash and less blue spruce in farmstead shelterbelts
may help reduce nesting densities of Common Crackles in this type of
habitat. Structurally open vegetation, including open coniferous species
such as ponderosa pine {Pinus ponderosa), probably deters nesting be-
cause of the lack of suitable attachments for nests. The planting of blue
spruce windbreaks at distances >500 m from inhabited residences for
reducing soil erosion should not encourage colonization.

ACKNOWLEDGMENTS

We thank D. Bergman, K. Fitzner, D. Foulk, K. Maier, C. Nelms, B. Osborne, K. Schei-
decker and M. Soehren for their many hours of help in the field and lab. R. Dolbeer and
G. Nuechterlein contributed valuable comments at the beginning of this study. D. Twedt
assisted with the study design. We also thank landowners of Benson County for access to
their properties. A. Barras, D. Bergman, D. Caccamise, R. Dolbeer, M. Kenyon, and an
anonymous reviewer commented on earlier drafts of this manuscript. This research was
supported by the Denver Wildlife Research Center, United States Dept, of Agriculture,
Animal Plant Health Inspection Service, [Contract No. 1 2-34-4 1-0020(CA)] and the Dept,
of Zoology, North Dakota State Univ., Fargo.

LITERATURE CITED

Bekoff, M., A. C. Scott, and D. A. Conner. 1987. Nonrandom nest-site selection in
Evening Grosbeaks. Condor 89:819-829.

Bent, A. C. 1958. Life histories of North American blackbirds, orioles, tanagers and allies.
U.S. Natl. Mus. Bull. 211.

Byers, C. R., R. K. Steinhorst, and P. R. Krausman. 1984. Clarification of a technique
for analysis of utilization-availability data. J. Wildl. Manage. 48:1050—1053.

114

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

COUES, E. C. 1878. Field notes on birds observed in North Dakota and Montana along the
forty-ninth parallel during the seasons of 1873 and 1874. U.S. Dept. Int. Geolog. and

Geograph. Surv. Bull. IV No. 3. i i

Erskine, a. J. 1971. Some new perspectives on the breeding ecology of Common Crackles.

Wilson Bull. 83:352—370.

Field, N. H. 1971. Use of eastern South Dakota shelterbelt by nesting birds. South Dakota
Bird Notes 23:43^5.

Gutzwiller, K. J. and S. H. Anderson. 1987. Multiscale associations between cavity-
nesting birds and features of Wyoming streamside woodlands. Condor 89:534-548.
Homan, H. J., G. M. Linz, and W. J. Bleier. 1994. Effect of crop phenology and habitat
on the diet of Common Crackles. Am. Midi. Nat. 131:381—385.

Marcum, C. L. and D. O. Loftsgaarden. 1980. A nonmapping technique for studying
habitat preferences. J. Wildl. Manage. 44:963—968.

Martin, T. E. 1978. Diversity and density of shelterbelt bird communities. M.S. thesis.
South Dakota State Univ., Brookings, South Dakota.

Maxwell, G. R., J. M. Nocilly, and R. I. Shearer. 1976. Observations at a cavity nest
of the Common Crackle and an analysis of grackle nest sites. Wilson Bull. 88:505-
507.

Nelms C. O., W. J. Bleier, D. L. Otis, and G. M. Linz. 1994. Population estimates of
breeding blackbirds in North Dakota, 1967, 1981-1982 and 1990. Am. Midi. Nat. 132:
256-263.

Neu, C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-

availability data. J. Wildl. Manage. 38:541-545.

North Dakota Agricultural Statistical Service. 1990. North Dakota agricultural sta-
tistics 1990. Agricultural Statistics No. 58. North Dakota State Univ., Fargo.

Petersen, A. and J. T. Young. 1950. A nesting study of the Bronzed Grackle. Auk 67:
466-467.

SOKAL, R. R. AND E J. Rohlf. 1981. Biometry, second ed. W. H. Freeman, San Francisco,
California.

Stewart, R. E. 1975. Breeding birds of North Dakota. Tri-college Center for Environmental
Studies, Fargo, North Dakota.

AND Kantrud, H. a. 1972. Population estimates of breeding birds in North Dakota.

Auk 89:766-789.

Thomas, D. L. and E. J. Taylor. 1990. Study designs and tests for comparing resource
use and availability. J. Wildl. Manage. 54:322—330.

Wiens, J. A. 1965. Behavioral interactions of Red-winged Blackbirds and Common Crack-
les on a common breeding ground. Auk 82:356—374.

Wiley, R. H. 1976. Communication and spatial relationships in a colony of Common
Crackles. Anim. Behav. 24:570—584.

Yahner, R. H. 1982. Avian nest densities and nest-site selection in farmstead shelterbelts.
Wilson Bull. 94:156-175.

Wilson Bull., 108(1), 1996, pp. 115-122

EVIDENCE OF DUAL BREEDING RANGES FOR THE
SEDGE WREN IN THE CENTRAL GREAT PLAINS

Paul A. Bedell

Abstract. — Sedge Wrens {Cistothorus platensis) are very rare breeders in the central
plains states but show a pattern of mid-summer arrival dates. I examined their status in
central Nebraska in August 1994 by conducting six Breeding Bird Survey routes and by
searching suitable habitat. I recorded Sedge Wrens on three of the six survey routes. Most
wrens occurred on sub-irrigated native meadows, but a variety of grassland types were used.
Most clutches were initiated by the second week of August. Observers should be aware of
the potential for late-summer breeding in other portions of their range. Received 22 Mar.
1995, accepted 15 Aug. 1995.

Although the breeding range of the Sedge Wren {Cistothorus platensis)
includes much of the midwestern and northeastern United States (AOU
1983), it is apparently common and widespread only in the upper mid-
west, e.g., Minnesota (Janssen 1987) and Wisconsin (Robbins 1991). In
much of the remainder of their breeding range in the United States, Sedge
Wrens are rather rare, local, and erratic in occurrence. The breeding status
of the Sedge Wren is an enigma as there are few or no nest records in
many areas, combined with peculiar mid- to late-summer arrival dates.
Such is the case in Alabama (Imhof 1962), Arkansas (James and Neal
1986), Kansas (Thompson and Ely 1992), Kentucky (Mengel 1965), Mis-
souri (Robbins and Easterla 1992), Nebraska (Lingle and Bedell 1989),
and Tennessee (Robinson 1990). McNair (1983) discussed summer oc-
currences of Sedge Wrens in the southeastern states and questioned
whether these records indicate possible breeding activity. Late-summer
breeding may occur in areas where wrens are absent in early summer,
although nest records are very few. This phenomenon has been observed
in Arkansas (Meanly 1952), Kansas (Schwilling 1982), and Nebraska
(Lingle and Bedell 1989). Coincident with mid-summer arrival dates in
more southerly areas is a “shifting about” in and out of nesting territories
during mid-July in Minnesota (Burns 1982) and Illinois (Kroodsma and
Verner 1978).

Sedge Wrens are highly opportunistic breeders that show little or no
site fidelity (Walkinshaw 1935, Burns 1982), probably due to habitat in-
stability (Kroodsma and Verner 1978). The ephemeral nature of wet grass-
land and marsh edge habitat dictates that species breeding in these habitats
should be good dispersers (Remsen and Parker 1990). Some populations
of Sedge Wrens are double-brooded (Burns 1982), and they are polygy-

10120 Silverleaf Ter., Richmond, VA 23236.

115

116

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

nous (Crawford 1977). Males build multiple nests which may be used for
courtship activities, dormitories, or predator decoys (Burns 1982) and
which are referred to in this study as dummy nests. One of these nests is
lined and used by the female as a brood nest. However, nest building by
males does not in itself prove breeding, as they may not attract a mate
(Crawford 1977, McNair 1983).

In order to determine if territorial Sedge Wrens occurred in suitable
habitat in late summer throughout central Nebraska, I censused selected
Breeding Bird Survey (BBS) routes in August for the presence or absence
of wrens. I also checked suitable habitat on a random basis in August
and compared this to atlas data, nest records, and local ornithological
literature. Clutch data came from this study and from field work which I
conducted during the summers of 1988-1992 in Hall County, Nebraska.

METHODS

Previous field work in the Platte River valley indicated that in August Sedge Wrens often
occur in the same type of wet meadows favored by Bobolinks {Dolichonyx oryzivorus).
Therefore, I chose BBS routes based on relatively high numbers of that species. The chosen
BBS routes had no previous records of Sedge Wrens. The mean numbers of Bobolinks per
BBS route conducted from late May through mid-June from 1967—1992 were provided by
the Breeding Bird Survey (B. Peterjohn, unpubl.). I also chose routes ba.sed on geographic
coverage of central Nebraska and the Sandhills region. I conducted Nebraska BBS route
numbers 007, 018, 026, 029, 041, and 116, between 8-14 Aug. 1994 and recorded the
presence or absence of Sedge Wrens at half-mile intervals. I compared August Sedge Wren
occurrence to June Bobolink occurrence at identical BBS survey stops with a Chi-square
test (Ambrose and Ambrose 1987).

I also checked suitable habitat for the presence of Sedge Wrens throughout the central
Nebraska area from 2—23 Aug. 1994 by stopping and listening from the roadside as I
travelled. Several locations were reported by other observers. This area was bounded ap-
proximately by the Platte River valley, 102°W. Longitude in Cherry and Keith Counties,
98°W. Longitude in Hamilton County, and the South Dakota border. BBS routes were in
Nebraska counties Buffalo, Cherry, Holt, Loup, and Wheeler. When access permitted, I
searched for nests by first observing the birds, then by carefully pushing aside the vegetation
as I walked through their territories.

I categorized wrens in a ranking of probable breeding evidence; (A) singing in suitable
habitat, (B) nest building or dummy nests observed, (C) nest with eggs or adult feeding
young. Because Sedge Wrens can build nests yet remain mateless, I considered breeding to
be confirmed only when nests with eggs or offspring were observed or adults were observed
carrying food. Clutch initiation was defined as the date the first egg was laid. Backdating a
clutch to initiation was determined as one egg laid per day until clutch completion, thirteen
days for incubation, and another fourteen days until fledging (Walkinshaw 1935, Burns 1982,
pers. obs.).

Locations were marked on county maps and compared to Nebraska Breeding Bird Atlas
data (Molhoff, unpubl.) and to June breeding records from state literature (Bruner et al.
1904, Cink 1973).

Bedell • SEDGE WREN BREEDING RANGE

1 17

Table 1

Numbers of Sedge Wrens Recorded on Breeding Bird Survey (BBS) Routes in

Nebraska

BBS route

# Years run

Total June

August

number'

1967-1992

occurrences

1994

001

24

2

—

022

17

2

—

007

23

0

1

018

1 1

0

0

026

18

0

2 •

029

20

0

0

041

3

0

0

116

3

0

10

* Rows 001 and 022 are in northeastern Nebraska; routes 007. 018, 026, 029, 041. and 1 16 are in central Nebraska.

RESULTS

Three of the six BBS routes had Sedge Wrens singing on territories in
August despite no previous June records. Only two BBS routes in Ne-
braska have ever reported Sedge Wrens, and then only rarely from the
eastern border of the state. These records are from routes #001 in Otoe
County, with individual reports in two out of 24 years, and #022 in Thurs-
ton County, also with individual reports in two out of 17 years (Table 1).

I found no wrens in the central or western Sandhills region even though
there appears to be much suitable habitat. Sedge Wrens occurred in this
study only east of 100°W. longitude.

Even though the highest incidence of Bobolinks and Sedge Wrens oc-
curred on the same route, there was no significant association between
survey stops that recorded Bobolinks in June and identical stops that
recorded wrens in August (x^ = 2.25, df = 1, E > 0.05).

Seventeen additional sites with territorial Sedge Wrens were located in
August 1994, mostly in sub-irrigated native meadows of the Platte River
valley (Table 2). But one site was on Conservation Reserve Program
(CRP) land idled for one season. Another site was on a three-year old
dry 1.2 ha. native prairie planting. This prairie restoration site in Hamilton
County, Nebraska, contained an active nest initiated on 9 August 1994.

Nine additional nests were discovered at Mormon Island Crane Mead-
ows (MICM) in Hall County, Nebraska, in August 1989 and 1992. Two
more nests in central Nebraska are described by Lingle and Bedell (1989)
from August 1988. Most of these twelve clutches were initiated by the
second week of August (Fig. 1). In addition, an adult wren with four

118

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 2

Habitat Use by Sedge Wrens in Nebraska during August 1994

Occurrences

Breeding evidence category

Habitat

A

B

c

Subirrigated native meadow

6

3

1

Other wetlands

3

0

0

Prairie restoration and CRP land

2

0

1

Upland prairie

0

0

1

recently fledged young were reported from Buffalo County, Nebraska, in
an upland prairie on 26 September 1994 (G. Lingle, pers. comm.).

DISCUSSION

Since Sedge Wrens can easily be overlooked, there is a possibility that
they are present in these areas in early summer but delay the onset of
breeding. This is unlikely due to the paucity of atlas, BBS, and other
published records over many years. Stronger evidence of their early-sum-
mer absence comes from the Mormon Island Crane Meadows where ten

Lig. 1. Sedge Wren clutch initiation date.s by week in Nebraska.

Bedell • SEDGE WREN BREEDING RANGE 1 1 9

of the twelve active nests were located. Thirteen years of Breeding Bird
Census studies conducted between 1980 and 1994 at MICM between 23
May and 20 June have recorded only single males in 1990 and 1994, and
there are no records prior to June.

Although most wrens occurred on native wet grasslands, they showed
a range of adaptability from partly flooded sites to dry prairie. The active
nest in Hamilton County was on a dry level site devoid of any wetland
vegetation. The adult with fledglings in Buffalo County was seen on a
dry upland prairie. The Conservation Reserve Program (CRP) site was
characterized by a rank growth of annuals up to 3 m in height. Sedge
Wrens have also been recorded as using CRP lands in the Dakotas and
Minnesota (Johnson and Schwartz 1993). Vegetative growth of at least
0.5 m was common to all sites.

The standard breeding phenology of most avian migrants includes a
spring migration to a breeding region, establishment of a breeding site
and raising offspring, and a return migration. The only North American
migrant bird species known to possibly deviate from this pattern by uti-
lizing dual breeding ranges is the Phainopepla (Phainopepla nitens)
(Walsberg 1977). Phainopeplas breed in the Colorado Desert of California
in March and April, then apparently migrate and renest in coastal oak
woodlands from May through July. However, there is no direct evidence
of this. The evidence for dual breeding ranges for Sedge Wrens is also
conjectural in that there is no proof that they have first nested elsewhere
before appearing in the central plains states in July and August. But the
circumstantial evidence is compelling. The period of “shifting about”
described by Burns (1982) and Kroodsma and Verner (1978) coincides
with the arrival of Sedge Wrens in the central plains and in other states.
The ephemeral habitat and low philopatry indicate a need and the ability
for undertaking this highly unusual breeding strategy. The high mobility
of Sedge Wrens may not be unique. There is an intriguing report of
Yellow Rails (Coturnicops noveboracensis), which nest in similar habitat,
nesting in late August in North Dakota (Lambeth 1994).

The breeding range of the Sedge Wren includes eastern Nebraska
(Johnsgard 1979) and eastern Kansas (Thomson and Ely 1992), but this
is based on very little actual evidence. A survey of historical early-sum-
mer breeding records from local ornithological literature, Nebraska Breed-
ing Bird Atlas data (Molhoff, unpubl.), and the N.A. Nest Record program
revealed only two June records of active nests in Nebraska (Bruner et al.
1904, Cink 1973). There are no June breeding records from Kansas and
only two Kansas BBS routes, 020 in Barton County and 026 in Jefferson
County, have reported Sedge Wrens, with a combined historical total of
six birds. Sch willing (1982) described two active nests and recently

120

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

fledged young from Atchison County, Kansas, from August 1980. There
is one North American nest record card of a probable dummy nest for
Nebraska and none for Kansas. There are a handful of additional records
of singing birds in appropriate habitat that possibly represent breeding
activity (Cink 1973, Labedz 1984). Recently, July-August records in the
Platte River valley of Hall County in central Nebraska indicate a regular
mid-summer migration and nesting season in an area where they are large-
ly absent until mid-July (Bedell 1987, Lingle and Bedell 1989). Observers
have become aware of the influx of Sedge Wrens into the area m mid-
summer where they have been recorded in numerous locations (Bedell
1987, Clausen 1989). A similar phenology has been noted for the Konza
Prairie near Manhattan, Kansas (Zimmerman 1993). The breeding status
of Sedge Wrens on the Konza Prairie has yet to be determined, where
they occur annually in drainages that have ample stands of Spartina.
When I visited the area on 25 Aug. 1994, I failed to find any activity,
although wrens were present through at least 17 Aug. (J. Zimmerman,
pers. comm.).

If Sedge Wrens have indeed first bred elsewhere in their range, then
where are they coming from and why risk the hazards of migration? Theie
seems to be little benefit for individuals that may have been successful
in raising a first brood on a good territory to undertake the risks inherent
in migration and of establishing another territory, so this migration should
consist of individuals that accrue benefits in fitness that offset the costs.
A possible explanation may be that because of the incidence of simulta-
neous polygyny (19% in Crawford 1977), bachelor males migrate to sec-
ondary areas such as the central plains anticipating a possible influx of
females. If changing habitat conditions force Sedge Wrens to search for
new territories in mid-season, these males would have an advantage, and
the females would have an established territory to quickly move into. But
if habitat conditions in the primary range remain favorable, few if any
females may undertake this mid-summer migration. This may explain the
lack of breeding evidence in some summers in Nebraska and elsewhere.
Another good possibility is that they originate from the northern portion
of their breeding range where a shorter season prevents them from raising
a second brood. Or, perhaps these are individuals that occupied marginal
territories which became unsuitable by July. This interesting problem cer-
tainly deserves further study.

If Sedge Wrens expand their breeding range into the central plains
states in late summer, the same phenomenon should occur in othei areas
considered beyond their normal range. To understand the extent of this
breeding range shift will require other observers to be aware of this pos-
sibility, and to take a closer look at the occunence of late-summer Sedge

Bedell • SEDGE WREN BREEDING RANGE

121

Wrens. This aspect of their breeding strategy needs to be determined for
any understanding of the status of this possibly declining species.

ACKNOWLEDGMENT

This research was supported by the Margaret Morse Nice award of the Wilson Ornitho-
logical Society and the E. A. Bergstrom fund of the Association of Eield Ornithologists.
Bruce Peterjohn provided data and maps from the Breeding Bird Survey program. Gary
Lingle of the Platte River Whooping Crane Habitat Maintenance Trust and Bill Whitney of
the Prairie/Plains Resource Institute provided helpful information on field locations and
permitted property access. Wayne Molhoff provided data gathered during the Nebraska
Breeding Bird Atlas project. C. Blem, G. Lingle, and J. Zimmerman provided helpful com-
ments on earlier versions of the manuscript.

LITERATURE CITED

Ambrose, H. W. and K. P. Ambrose. 1987. A handbook of biological investigation. Hunter
Textbooks, Inc., Winston-Salem, North Carolina.

American Ornithologists’ Union. 1983. Check-list of North American Birds, 6th ed. Am.
Ornithol. Union, Washington, D.C.

Bedell, P. 1987. Early fall migration of Sedge Wrens. Nebr. Bird Rev. 55:86-88.

Bruner, L., R. H. Wolcott, and M. H. Swenk. 1904. A preliminary review of the birds
of Nebraska. Klopp and Bartlett, Omaha, Nebraska.

Burns, J. T. 1982. Nests, territories, and reproduction of Sedge Wrens (Cistothorus platen-
sis). Wilson Bull. 94:338-349.

CiNK, C. 1973. Summer records of the Short-billed Marsh Wren in Nebraska. Nebr. Bird
Rev. 41:17-19.

Clausen, M. K. 1989. Recent Sedge Wren observations in Nebraska. Nebr. Bird Rev. 57:
92-93.

Crawford, R. D. 1977. Polygynous breeding of Short-billed Marsh Wrens. Auk 94 359-
362.

Imhof, T. a. 1962. Alabama birds. Univ. of Alabama Press, University, Alabama.

James, D. A. and J. C. Neal. 1986. Arkansas birds: their distribution and abundance. The
Univ. of Arkansas Press, Fayetteville, Arkansas.

Janssen, R. B. 1987. Birds in Minnesota. Univ. of Minn. Press, Minneapolis, Minnesota.
JoHNSGARD, P. A. 1979. Birds of the Great Plains: breeding species and their distribution.
Univ. of Nebr. Press, Lincoln, Nebraska.

Johnson, D. H. and M. D. Schwartz. 1993. The Conservation Reserve Program and
grassland birds. Cons. Biol. 7(4):934— 937.

Kroodsma, D. and j. Verner. 1978. Complex singing behaviors among Cistothorus wrens.
Auk 95:703-716.

Labedz, T. 1984. Sedge Wren, a new bird species for Mormon Island Crane Meadows.
Nebr. Bird Rev. 52:65-66.

Lambeth, D. O. 1994. Territorial Yellow Rails in late August in Grand Forks County. North
Dak. Nat. Science Soc. Newsletter 1 1(I):8-I0.

Lingle, G. R. and P. Bedell. 1989. Nesting ecology of Sedge Wrens in Hall County,
Nebraska. Nebr. Bird Rev. 57(2):47-49.

McNair, D. B. 1983. The significance of breeding season records of Sedge Wrens in the
southeast states. The Oriole 48(4):49-52.

Meanly, B. 1952. Notes on the ecology of the Short-billed Marsh Wren in the lower
Arkansas rice fields. Wilson Bull. 64:22-25.

122

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Mengel, R. M. 1965. The birds of Kentucky. Ornithol. Monographs No. 3. Amer. Ornithol.

Rem^n,°J. V. AND T. A. Parker III. 1990. Seasonal distribution of the Azure Gallinule
(Porphyrula flavirostris), with comments on vagrancy in rails and gallmules. Wilson

Bull. 102(3):380-399. .

Robbins, M. B. and D. A. Easterla. 1992. Birds of Missouri: their distribution and abun-
dance. Univ. of Missouri Press, Columbia, Missouri.

Robbins, S. D. Jr. 1991. Wisconsin birdlife: population & distribution: past & present. The

Univ. of Wise. Press, Madison, Wisconsin.

Robinson, J. C. 1990. An annotated checklist of the birds of Tennessee. The Univ. of Tenn.

Press, Knoxville, Tenn. • . , c on

SCHWILLING, M. D. 1982. Sedge Wrens nesting into September. Kansas Ornithol. Soc. Bull.

Thompson, M. C. and C. Ely. 1992. Birds in Kansas, Vol. II. Univ. Press of Kansas,
Lawrence, Kansas.

Walkinshaw, L. 1935. Studies of the Short-billed Marsh Wren (Cistothorus stellans) m
Michigan. Wilson Bull. 52:362-369.

Walsberg, G. E. 1977. Ecology and energetics of contrasting social systems in Phaino-
pepla nitens (Aves: Ptilogonatidae). Univ. Calif. Publ. Zool. 108:1-63.

Zimmerman, J. 1993. The Birds of Konza: the avian ecology of the tallgrass prairie. Univ.
Press of Kansas, Lawrence, Kansas.

Wilson Bull., 108(1), 1996, pp. 123-128

DIETS OF NORTHERN PYGMY-OWLS AND
NORTHERN SAW- WHET OWLS IN
WEST-CENTRAL MONTANA

Denver W. Holt and Leslie A. Leroux

Abstract. — One hundred ninety-four prey from 31 Northern Pygmy-Owls {Glaucidium
gnoma) and 388 prey from 23 Northern Saw-whet Owls {Aegolius acadicus) were compared.
Thirty-six percent of the pygmy-owl’s prey was birds, whereas, 98.0% of the saw-whet
owl’s prey was small mammals, particularly voles. Food niche breadth and dietary evenness
was 10.6 and 0.69 for pygmy-owls vs 3.3 and 0.89 for saw-whet Owls. Body mass of prey
killed by both species was about 38 g. Dietary overlap between these two owl species was
37.0%, indicating that they fed on different prey assemblages. Received 4 April 1995, ac-
cepted 28 Aug. 1995.

Northern Pygmy-Owls {Glaucidium gnoma) and Northern Saw-whet
Owls {Aegolius acadicus) overlap throughout much of their range in the
western United States (AOU 1983). The natural history of Northern Pyg-
my-Owls is poorly known (Holt and Norton 1986, Holt et al. 1990), while
that of Northern Saw-whet Owls is more certain (Cannings 1993).

In west-central Montana, Northern Pygmy-Owls and Northern Saw-
whet Owls occur sympatrically from mixed deciduous and coniferous
forested valley bottoms (975 m) to higher elevation (1584 m) coniferous
forests (Holt and Hillis 1987). Both species are obligate cavity nesters,
dependent upon woodpeckers or natural sites for nests. Both species for-
age similarly, using a perch and pounce technique. Northern Pygmy-Owls
are crepuscular or diurnal, and Northern Saw-whet Owls are nocturnal.
The diet of Northern Pygmy-Owls has been reported at the class level,
while that of Northern Saw-whet Owls has been specihc and thoroughly
reviewed (Marks and Doremus 1988, Holt et al. 1991, Swengel and S wen-
gel 1992). Several authors have compared the diets of sympatric owl’s
(Maser et al. 1970, Knight and Jackman 1984, Marks and Marti 1984,
Nilsson 1984, Bosakowski and Smith 1992), but Herrera and Hiraldo
(1976) in Europe, and Hayward and Carton (1988) in North America,
have compared the diet of small cavity nesting forest owls. Herein, we
compare their diet, prey biomass, food niche breadth (FNB), dietary even-
ness (DE), and dietary overlap (DO).

Pellets and pellet fragments from Northern Pygmy-Owls and Northern
Saw-whet Owls were collected below roost trees near Missoula, Montana,
during the non-breeding season — October through February 1987 to
1992. Pellets were dissected by hand, and prey species were identified

Owl Research Institute, RO. Box 8335, Missoula, Montana 59807.

123

124

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

and quantified using skulls and mandibles. Diurnal field observations of
hunting Northern Pygmy-Owls with prey were also included. To evaluate
these owls’ trophic niches, we first compared prey species frequencies
and percentages. We then computed the Shannon-Weaver diversity index
to compare their FNB: where H' = - S p.logp, and p, represents the
proportion of each species in the prey sample (see Marti 1987). These
values range from one to N, with larger values suggesting a broader food
niche breadth. Dietary evenness was calculated using the equation; F -
_ i)/(N| — 1), where N, is the antilog of the Shannon-Weaver index
(H'), and N, is the reciprocal of Simpson’s index (1/D) (Marti 1987). The
dietary evenness values range from zero to one. As prey proportions
become more equal, evenness values approach unity. To compare dietary
overlap, we used the equation; O = S PyP.k/^S P.j^ ^ where p,j and
Pii, are proportions of prey species in the diets of owls j and k, respectively
(Marti 1987). The dietary overlap value ranges from zero to one, with
zero meaning no dietary overlap and one meaning complete dietary over-
lap. We multiplied the values by 100 and report them as percentages for
easier interpretation. Body mass of prey was set as the midpoint of the
range. We did this because of inconsistencies with using the mean body
mass from the literature, and age differences among prey species are not
always delineated (Marti 1987, Holt et al. 1991, Holt 1993). Prey body
mass data were taken from Dunning (1984) for birds and from Burt and
Grossenheider (1976) for mammals. Prey was identified to species foi the
FNB, DE, and DO equations.

One hundred ninety-four prey items were recorded from 31 Northern
Pygmy-Owls. Thirteen bird and four mammal species were eaten (see
Table 1 for list and scientific names of prey items). Mammals represented
60.8% of the prey and birds at least 36.6%. Microtus voles represented
53.6% of the total prey eaten and 88.1% of the mammals eaten (Table
1). House Sparrows {Passer domesticus) represented 13.9% of the total
prey eaten and 35.5% of the total birds eaten (Table 1). Food niche
breadth was 10.6 (N = 99), and this value suggests a wide trophic niche.
Dietary evenness was 0.69, which suggests that few prey species were
evenly distributed in the diet. Prey body mass ranged from 3—167 g, x —
38.4 g.

Three hundred eighty-eight prey items were recorded from 23 Northern
Saw-whet Owls. Six mammals and one bird species were eaten (see Table
2 for list and scientific names of prey items). Mammals represented at
least 98.5% of the total prey, with deer mice, montane voles, and meadow
voles, representing 92.0% (Table 2). When combined, Microtus species
were more frequently eaten then Peromyscus, 57.9% vs 34.8%. Birds
were numerically insignificant. Food niche breadth was 3.3 (N = 366)

Half and Lerou.x • PYGMY-OWL AND SAW- WHET OWL DIETS

125

Table 1

Prey Species from 31 Northern Pygmy-Owls

Species

No.

%

MP

Range

Biomass

(g)

BIRDS

House Sparrow {Passer domesticus)

27

13.9

27

20-34

729

Pine Siskin (Carduelis pinus)

1 1

5.7

15

10-20

165

Evening Grosbeak (Coccothraustes vespertinus)

9

4.6

62

38-86

558

House Finch (Carpodacus mexicanus)

7

3.6

22

19-25

154

Dark-eyed Junco {Jiinco hyemalis)

5

2.6

20

14-26

. 100

Bohemian Waxwing (Bomhycilla garndus)

4

2.1

58

46-69

232

Northern Flicker {Colaptes auratus)

2

1.0

144

121-167

288

Black-capped Chickadee {Pams atricapillus)

2

1.0

11

8-13

22

Song Sparrow {Melospiza melodia)

2

1.0

20

11-29

40

Mountain Chickadee {P. gambeli)

I

tr.2

1 1

8-14

1 1

American Robin {Tardus migratorius)

1

tr.

83

63-103

83

American Goldfinch {Carduelis tristis)

1

tr.

14

8-20

14

Long-billed Marsh Wren {Cistothorus palustris)

1

tr.

1 1

9-13

1 1

Waxwing spp. {Bombycilla spp.)

1

tr.

—

—

—

Bird spp.

2

1.0

—

—

—

subtotal

76

36.6

—

8-167

2407

MAMMALS

Vole spp. {Microtus spp.)

88

45.4

57

8-85

5016

Meadow Vole {Microtus pennsylvanicus)

13

6.7

49

28-70

637

Montane Vole {M. montanus)

3

1.5

57

28-85

171

Deer Mouse {Peromyscus maniculatus)

8

4.1

27

18-35

216

Vagrant Sbrew {Sorex vagrans)

2

1.0

4

3-6

8

Mammal spp.

4

2.1

—

—

subtotal

118

60.8

—

3-85

6048

Total

194

100.0

—

3-167

8455

“ tr. = (race amounts <1%.

suggesting a narrow trophic niche. Dietary evenness was 0.89, suggesting
few species were eaten in similar proportions. Prey body mass ranged
from 3 to 130 g, x = 37.7 g. Mean mammalian prey was 38.4 g.

Food niche breadth of the two species was strikingly different, with
Northern Pygmy-Owls feeding on greater than three times as many spe-
cies as Northern Saw-whet Owls (10.6 vs 3.3). Evenness values were also
strikingly different (0.69 vs 0.89), and suggested that Northern Pygmy-
Owls were not as restricted in their diet as Northern Saw-whet Owls.
Thus, Northern Pygmy-Owls in our study area fed on a wider assemblage
of prey than did Northern Saw-whet Owls. Dietary overlap was 37.0%,
again indicating that these two species used different prey assemblages.
At the generic level for mammals however, Microtus voles represented

126

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 2

Prey of 23 Northern Saw-Whet Owls

Species

MAMMALS

Deer Mouse (Peromyscus maniculatus)
Montane Vole (Microtus montanus)

Meadow Vole (M. pennsylvanicus)

Vole spp. (Microtus spp.)

Shrew Spp. (Sorex spp.)

Vagrant Shrew (S. vagrans)

Masked Shrew (5. cinereum)

Northern Pocket Gopher (Thymomas talpoides)
subtotal

BIRDS

Cedar Waxwing (Bombycilla cedrorum)

Bird Spp.
subtotal
Total

No.

%

MP

Range

Biomass

(g)

134

34.5

27

18-35

3618

122

31.4

57

28-85

6954

101

26.0

49

28-70

4949

10

2.6

57

28-85

570

10

2.6

5

3-7

50

5

1.3

4

3-6

20

1

tr.

4

3-7

4

2

tr.

104

78-130

208

385

98.5

—

3-130

16,373

1

tr.

33

25^0

33

2

tr.

—

—

—

3

tr.

—

—

—

388

100.0

—

3-130

16,406

53.6% of the Northern Pygmy-Owl’s diet and 60. 1 % of the Northern Saw-
whet Owl’s diet. This comparison suggests that Microtus voles were al-
most equally important to both species of owls.

Northern Pygmy Owls ate prey that averaged 38.4 g, with the smallest
being a shrew spp. (4 g) and the largest a Northern Flicker {Colaptes
auratus-, 142 g). Northern Saw-whet Owls ate prey that averaged 37.7 g,
with the smallest being a shrew (4 g) and the largest a northern pocket
gopher (104 g). Yet the Northern Pygmy-Owl is the smaller of these two
species. Indeed, average body mass for museum specimens of both spe-
cies are Northern Pygmy-Owls; males 61.9 g, range 54-74 (N = 42) and
females 73.0 g, range 64-87 (N = 10) and Northern Saw-whet Owls;
males 74.9 g, range 54-96 (N = 27) and females 90.8 g, range 65-124
(N = 18) (Earhart and Johnson 1970), but also see Cannings (1993) for
live weights.

This is the hrst quantitative review of the Northern Pygmy-Owls diet
in North America and the first to compare its diet with another small
sympatric forest owl. Previous authors (Holman 1926, Norton and Holt
1982, Holt and Norton 1986, Bull et al. 1987) have reported dietary data
for Northern Pygmy-Owls. In these studies however, sample sizes weie
small (<35), and prey species were not always identified. An interesting
similarity arises from these studies however. The percentages of birds in
the Northern Pygmy-Owls diet were about 25 to 50% of the total prey.

Holt and Lerou.x • PYGMY-OWL AND SAW-WHET OWL DIETS

127

26.0%, 32.0%, 47.0%, 36.0%, respectively. These data are similar to our
results, and we know of no other North American owl species that shows
such a preponderance of birds in its diet. During the breeding season
however, Holman (1926) reported 42.0% lizards, and Norton and Holt
(1982) and Bull et al. (1987) also reported 3.2% and 30.0% insects, re-
spectively.

Diet of Northern Saw-whet Owls was consistent with other studies
reporting their feeding ecology (Marks and Doremus 1988, Holt et al.
1991, Swengel and Swengel 1992). Prey body mass reported here (37.7
g) was in the upper limits of those reported by Cannings (1993). We
believe this reflects the high proportions of Microtus voles in the owls
diet from our study area.

Marks and Marti (1984) compared the trophic niche of breeding Long-
eared Owls {Asia otus) and Barn Owls {Tyto alba). They concluded that
competition could not be stated as shaping the owl’s FNB. Hayward and
Carton (1988) compared the diets of Boreal Owls {Aegolius funereus).
Northern Saw-whet Owls, and Western Screech-Owls (Otus kennicottii).
These owls ate similar sized prey, but sample sizes were too small for
meaningful conclusions to be drawn. Bosakowski and Smith (1992) com-
pared the trophic niche of Eastern Screech-Owls (D. asio). Barred Owls
iStrix varia), and Great Horned Owls {Bubo virginianus) and concluded
that the low dietary overlap was a result of size differences and habitat
use between these owl species.

There is little conclusive proof about which mechanisms structure com-
munities. Wiens (1989) listed two conditions that must be met for inter-
specific competition to exist (1) species must share resources, and (2)
joint exploitation of those resources must negatively effect one or all
species involved. We cannot conclude that diet is shaping the sympatric
distribution of Northern Pygmy-Owls and Northern Saw-whet Owls in
western Montana. Perhaps diel activity rhythms contribute to these owls’
sympatry and reduced dietary overlap — Northern Pygmy-Owls are diur-
nal or crepuscular and Northern Saw-whet Owls are nocturnal. Prey ac-
tivity rhythms may also influence spatial overlap between these species,
and these type of data need to be incorporated into future studies of owl
feeding ecology.

ACKNOWLEDGMENTS

We thank Charles Blem, Mike Maples, and Scott Swengel for comments on the manu-
script.

LITERATURE CITED

American Ornithologists Union. 1983. Check-list of North American birds, 6th ed.

A.O.U., Washington, D.C.

128

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Bosakowski, T and D. G. Smith. 1992. Comparative diets of sympatric nesting raptors in
the eastern deciduous forest biome. Can. J. Zool. 70:984— 991.

Bull, E. L., J. E. Hohmann, and M. G. Henjum. 1987. Northern Pygmy-Owl nests in
northeastern Oregon. J. Rap. Res. 21:77—78.

Burt, W. H. and R. G. Grossenheider. 1976. A field guide to the mammals. Houghton
Mifflin Co., Boston, Massachusetts.

Cannings, R. J. 1993. Northern Saw-whet Owl (Aegolius acadicus) in The birds of North
America, No. 42 (A. Poole and E Gill, eds.). The Academy of Natural Sciences of
Philadelphia, The American Ornithologist’s Union, Washington, D.C.

Dunning, J. B., Jr. 1984. Body weights of 686 species of North American birds. West.
Bird-Banding Assoc. Monog. No. 1.

Earhart, C. M. and N. K. Johnson. 1970. Size dimorphism and food habits of North
American owls. Condor 72:251-264.

Hayward, G. D. and E. O. Garton. 1988. Resource partitioning among forest owls in the
River of No Return Wilderness. Oecologia 75:253-265.

Herrera, C. M. and F. Hiraldo. 1976. Food-niche and trophic relationships among Eu-
ropean owls. Ornis Scand. 7:29-41.

Holman, E C. 1926. Nesting of the California Pygmy Owl in Yosemite. Condor 28:92-93.

Holt, D. W. 1993. Trophic niche of nearctic Short-eared Owls. Wilson Bull. 105:497-503.

, E. Andrews, and N. Claflin. 1991. Non-breeding season diet of Northern Saw-

whet Owls {Aegolius acadicus) on Nantucket Island, Massachusetts. Can. Field-Nat.
105:382-385.

AND J. M. Hillis. 1987. Current status and habitat associations of forest owls in

western Montana. Pp. 281-288 in Biology and conservation of northern forest owls (R.
Nero, R. J. Clark, R. J. Knapton, R. H. Hamre, eds.). USDA For. Serv. Gen. Tech. Rep.
RM-142, Rocky Mt. For. and Range Exper. Stat., Ft. Collins, Colorado.

, R. Kline, and L. S. Holt. 1990. A description of “tufts” and concealing posture

in Northern Pygmy-Owls. J. Rap. Res. 24:59-63.

and W. D. Norton. 1986. Observations of nesting Northern Pygmy-Owls. J. Rap.

Res. 20:39-Jl.

Knight, R. L. and R. E. Jackman. 1984. Food-niche relationships between Great Horned
Owls and Common Barn Owls in eastern Washington. Auk 101:175-179.

Marks, J. S. and C. D. Marti. 1984. Feeding ecology of sympatric Barn Owls and Long-
eared Owls in Idaho. Ornis Scand. 15:135-143.

AND J. H. Doremus. 1988. Breeding season diet of Northern Saw-whet Owls in

southwestern Idaho. Wikson Bull. 100:690-694.

Marti, C. D. 1987. Raptor food habits studies. Pp. 67—80 in Raptor management techniques
manual (B. G. Pendleton, B. A. Milsap, K. W. Kline, and D. A. Bird, eds.). Nat. Wildl.
Fed. Sci. and Tech. Sen No. 10, Washington, D.C.

Maser, C., E. W. Hammer, and S. H. Anderson. 1970. Comparative food habits of three
owl species in central Oregon. Murrelet 51:29-33.

Nilsson, I. N. 1984. Prey weight, food overlap and reproductive output of potentially
competing Long-eared and Tawny Owls. Ornis Scand. 16:176-182.

Norton, W. D. and D. W. Holt. 1982. Simultaneous nesting of Northern Pygmy-Owls
and Northern Saw-whet Owls in the same snag. MuiTelet 63:94.

SwENGEL, S. R. AND A. B. SwENGEL. 1992. Diet of Northern Saw-whet Owls in southern
Wisconsin. Condor 94:707-71 1.

Wiens, J. A. 1989. The ecology of bird communities. Vol. 1, Cambridge Univ. Press,
London England.

Wilson Bull., 108(1), 1996, pp. 129-136

EFFECTS OF EGG TYPE ON DEPREDATION OF
ARTIFICIAL GROUND NESTS

Richard H. Yahner and Carolyn G. Mahan

Abstract. — We examined depredation of artificial ground nests containing three egg
types (brown chicken, white chicken, or Northern Bobwhite [Colinus virginianus]) in rela-
tion to plot age (clearcut vs uncut) and time period (trials 1—5) at the Barrens Grouse Habitat
Management Area, Centre County, Pennsylvania, from May-July 1993. One hundred thir-
teen (38%) of the total nests were disturbed. Fewer nests were disturbed in clearcut (32%)
than in uncut plots (43%) (P < 0.05). Clearcut plots had higher densities of brushy vege-
tation near ground level which better concealed nests and reduced foraging efficiency of
predators. Rates of nest disturbance varied with time period (P s 0.005); in general, rates
were greater in trials 1—3 than in trials 4—5, partially because of gypsy moth {Lymantria
dispar) defoliation during trials 1-3. Nest fate also differed significantly (P < 0.001) with
egg type. Rates of disturbance were lower with nests containing brown chicken eggs (24%)
compared to nests containing white chicken eggs (46%) or Northern Bobwhite (43%) eggs.
Nests with brown chicken eggs were better camoflaged and, hence, less likely to be dis-
turbed. Based on our findings, we recommend that brown chicken eggs be used as an
alternative to Japanese Quail {Coturnix coturnix japonica) eggs when simulating nests of
Ruffed Grouse {Bonasa umbellus) or Wild Turkey (Meleagris gallopavo) in artificial ground
nest studies. Received 28 Feb. 1995, accepted 1 June 1995.

The effects of egg size (e.g., Reistma et al. 1990) and egg color (e.g.,
Westmoreland and Best 1986, Yahner and DeLong 1992) have been ex-
amined in experimental studies designed to infer predation rates on nests
of bird species with relatively small eggs. In addition, the effects of egg
size on predation rates by American Crows {Corvus brachyrhnchos) have
been investigated in meadows using large chicken eggs, small chicken
eggs, and white painted Japanese Quail {Coturnix coturnix japonica) eggs
(Montevecchi 1976). Relatively large eggs, including brown chicken, Jap-
anese Quail, and Northern Bobwhite {Colinu.s virginianus) eggs also have
been used in a variety of artificial nest studies as a means of determining
rates of predation on nests simulating those of larger birds such as gal-
linaceous birds (e.g., Boag et al. 1984, Yahner and Wright 1985). How-
ever, no studies to our knowledge have examined differences in rates of
nest disturbance on artificial ground nests in forested habitats using large
eggs that differ in both size and color. This information is important in
the experimental design of artificial nest studies intended to obtain esti-
mates of predation on natural ground nests in various landscapes (e.g.,
Storaas 1988, Willebrand and Marcstrbm 1988). Our objective was to
compare rates of depredation among artificial ground nests containing

School of Forest Resources, The Pennsylvania State Univ., University Park, Pennsylvania 16802.

129

130

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

three egg types that vary in both size and color and placed in forested
plots of two age classes.

Study area and methods. — We conducted this study at the 1166-ha
Barrens Grouse Habitat Management Area (HMA), State Game Lands
176, Centre County, Pennsylvania, the site of four previous studies of
depredation of artificial ground nests (Yahner and Wright 1985, Yahner
et al. 1989, 1993; Yahner and Mahan 1996). The Barrens Grouse HMA
has been managed via forest clearcutting since 1976 by the Pennsylvania
Game Commission to create habitat for Ruffed Grouse {Bonasa umbellus)
(Yahner 1991, 1992). It contains an uncut (reference) and a cut (treated)
sector of similar size. The treated sector is subdivided into 50% and 75%
cut areas, corresponding to the amount of forest clearcutting and con-
tained 136 contiguous, 4-ha blocks (e.g., see Yahner 1993, Yahner et al.
1993); 76 and 60 blocks are in the 50% and 75% areas, respectively.
Each block is subdivided into four 1-ha (100 X 100 m) plots arranged in
a checkerboard pattern (plots A-D). Our study was focused in the 75%
area; in this area, plot A (western plot) in each block was clearcut during
winter 1975-1976, and plot B (northern plot) was cut during winter 1980-
1981, plot C (eastern plot) was cut in winters 1985-1986 or 1986-1987,
and plot D (southern plot) was uncut.

Overstory trees (>7.5 cm dbh and >1.5 m tall) in plot D of the 75%
area were about 70 years old and consisted primarily of quaking aspen
{Pop ulus tremuloides), bigtooth aspen (P. grandidentata), oak {Quercus
spp.), and pitch pine (Pinus rigida). Common understory trees (2. 5-7. 7
cm dbh) and shrubs (<2.5 cm dbh) in all plots of the 75% area were
aspen, dwarf chinkapin oak (Q. prinoides), scrub oak {Q. ilicifolia), and
blueberry (Vaccinium spp.) (Yahner 1993).

Gallinaceous birds nesting at ground level at the Barrens Grouse HMA
were Ruffed Grouse and Wild Turkey {Meleagris gallopavo) (Yahner et
al. 1989, Yahner 1993). Potential predators on ground nests were Amer-
ican Crow, Blue Jay {Cyanocitta cristata), Virginia opossum {Didelphis
virginianus), eastern chipmunk {Tamias striatus), gray squirrel (Sciurus
carolinensis), red squirrel (Tamiasciurus hudsonicus), red fox (Vulpes
vulpes), gray fox {Urocyon cinereoargenteus), black bear (Ursus ameri-
canus), raccoon {Procyon lotor), striped skunk {Mephitis mephitis), and
weasel {Mustela spp.) (Therres 1982, Yahner et al. 1993).

We placed artificial ground nests during five time periods from late
May to late July 1993 in the 75% area (Table 1). Each nest consisted of
three fresh eggs put in a slight depression in leaf litter adjacent to a log,
overstory tree, or stump (Yahner and Wright 1985, Yahner et al. 1993).
Each nest contained one egg type: brown chicken, white chicken, or
Northern Bobwhite. Based on a sample of 10 eggs/type, mean length and

Yahner and Mahan • ARTIFICIAL GROUND NESTS

131

Table 1

Fate of 299 Artificial Ground Nests in Relation to Age of Plot, Time Period, and
Egg Type at the Barrens Grouse Habitat Management Study Area, Centre County,

Pennsylvania, 1993

Nesl fate

Undisturbed Disturbed

Variable

Level

n

%

N

%

Age of plot

Clearcut

101

68

48

32

Uncut

85

57

65

43

Time period

Trial 1

37

62

23

38

Trial 2

35

58

25

42

Trial 3

26

43

34

57

Trial 4

46

77

14

23

Trial 5

42

71

17

29

Egg type

Brown chicken

76

76

24

24

White chicken

53

54

46

46

Northern Bobwhite

57

57

43

43

Total

186

62

113

38

width of brown chicken eggs were 52 X 40 mm, white chicken eggs were
56 X 42 mm, and bobwhite eggs were 30 X 24 mm. Brown chicken eggs
are light brown to buffy in color; white chicken and bobwhite eggs were
dull or creamy white (Harrison 1975). Ruffed Grouse eggs were 39 X 30
mm (buffy), and Eastern Wild Turkey eggs are 63 X 45 mm (pale buff
or buffy white) (Harrison 1975).

A trial was six days in length, with eight days between trials (meth-
odology follows that of Yahner and Scott 1988). During each trial, 15
clearcut plots (plot C) and 15 uncut plots (plot D) were selected randomly.
Two nests were placed in each plot; nests were separated by 30-35 m
and placed 5 m from the edge of the plot. When placing nests, we wore
rubber gloves and boots to minimize human scent at nests (Nol and
Brooks 1982). This experimental design gave 60 nests/trial equally divid-
ed between the two plot ages and among the three egg types (total = 300
nests; one of the 300 nests was omitted from analysis due to incorrect
placement).

We checked nests six days after placement between sunrise and 12:00
h (DST) to determine the fate (undisturbed, disturbed by avian predator,
disturbed by unknown predator) of each nest (Yahner and Wright 1985).
A disturbed nest was characterized by > one broken or missing egg on
day 6 of a given trial. Appearance and mode of disturbance of the eggs
were used to identify predators as avian (e.g., peck hole in egg) or un-

132

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

known (Rearden 1951, Boag et al. 1984, Yahner and Wright 1985). In
addition, eggs and eggshells were removed on day 6 of each trial.

We examined the dependency of nest fate (undisturbed vs disturbed)
on plot age (clearcut vs uncut), time period (trials 1-5), egg type (brown
chicken, white chicken, or Northern Bobwhite), using a four-way test-of-
independence (Dixon 1990). Likelihood ratios (G^) were used to test for
interactions of nest fate with the three other variables, using log-linear
models (Sokal and Rohlf 1981, Dixon 1990). Likelihood ratios are ap-
propriate when analyzing attribute variables in multi-way contingency
tables. If nest fate were significantly dependent on a given variable, we
used a posteriori G-tests for goodness-of-fit about the cell (level) of in-
terest (Sokal and Rohlf 1981). Because corvids are major predators on
artificial nests at the Barrens Grouse HMA (e.g., Yahner and Wright 1985,
but see Yahner et al. 1993), the frequency of nests disturbed by avian
predators was compared among the three egg types using a G-test for
goodness-of-fit.

RESULTS

One-hundred thirteen (38%) of 299 artificial ground nests were dis-
turbed during the five trials (Table 1). Regardless of egg type, nest fate
was associated with age of plot (G = 4.7, df = \, P < 0.05). Fewer nests
(all egg types combined) were disturbed in clearcut plots (N = 48, 32%)
than in uncut plots (N = 65, 43%). Nest fate also was associated with
time period (G = 18.0, df = 4, P < 0.005). The frequency of total
disturbed nests in trial 3 (N = 34, 57%) was significantly higher than
expected (G = 6.3, df = 1, P < 0.025), whereas frequency of total dis-
turbed nests in trial 4 (N = 14, 23%) was significantly lower than ex-
pected (G = 4.7, df = 1, P < 0.05). In general, the percentage of disturbed
nests/trial was greater in trials 1—3 (38—57%) compared to that in trials
4-5 (23-29%).

Nest fate varied with the three egg types (G = 14.0, df = 1, ^ < 0.001)
(Table 1). The frequency of disturbed nests with brown chicken eggs (N
= 24, 24%) was considerably lower than expected (G = 8.2, df = 1, F*
< 0.005), but the frequencies of disturbed nests with white chicken (N
= 46, 46%) or Northern Bobwhite eggs (N = 43, 43%) were not different
from expected (Gs < 2.7, df = 1, F > 0.10). Moreover, there was a
significant interaction among nest fate, egg type, and time period (G =
17.9, df = 8, P < 0.05). In particular, fewer nests with brown chicken
eggs were disturbed in trial 1 (N = 2, 2%) than expected (G = 6.0, df
— \, P < 0.05), and more nests with Northern Bobwhite eggs were dis-
turbed in trial 3 (N = 16, 14%) than expected (G = 7.9, df = 1 P <
0.005).

Yahner and Malum • ARTIFICIAL GROUND NESTS

133

Thirty-six (32%) of the 113 disturbed nests were preyed upon by avian
predators, principally Blue Jays and American Crows. The frequency of
nests lost to avian predators differed among the three egg types (G =
11.9, df = 2, P < 0.001). Of the total nests disturbed by birds, eight
(22%) were those containing brown chicken eggs, 22 (61%) had white
chicken eggs, and six (17%) had Northern Bobwhite eggs. Avian preda-
tors destroyed more nests with white chicken eggs than expected (G —
11.6, df = 1, P < 0.001) and less with Northern Bobwhite eggs than

expected (G = 5.1, df = 1, P < 0.05).

Discussion. — Our finding that rates of disturbance of artificial ground
nests were lower in clearcut plots than in uncut plots concurs with results
obtained in other studies of artificial nests at the Barrens Grouse HMA
(e.g., Yahner and Wright 1985, Yahner and Cypher 1987, Yahner and
Scott 1988). Clearcut plots were characterized by higher densities of
brushy vegetation near ground level, which presumably better concealed
artificial nests and reduced foraging of nest predators such as crows (Pi-
cozzi 1975) and raccoons (Bowman and Hanis 1980). Moreover, uncut
plots contained overstory trees that served as perch sites for avian nest
predators (Yahner et al. 1989).

Most studies of artificial nests at the Barrens Grouse HMA have not
documented significant differences in rates of nest disturbance over time
(e.g., Yahner et al. 1989, 1993). However, as in the present study, Yahner
and Wright (1985) found reduced rates of nest disturbance later in the
breeding season, possibly because family groups of crows move to com-
munal roosting sites and agricultural feeding sites (Cross 1946). Another
possible explanation for greater nest disturbance in earlier trials of our
study may be related to gypsy moth (Lymantria dispar) defoliation. Al-
though we did not quantify the extent of defoliation caused by gypsy
moth larvae, it was greatest during trial 3, which preceded the pupal stage
of the life cycle and corresponded to the period of most extensive defo-
liation on the study area in spring and summer 1993 (Yahner and Mahan
1996). Extensive defoliation by gypsy moths has been shown to increase
rates of artificial nest predation (Thurber et al. 1994).

Nests with brown chicken eggs in our study were better camouflaged
and, hence, less likely to be disturbed by predators than other egg types,
particularly by avian predators that rely on vision when foraging. A 24%
disturbance of nests with brown chicken eggs was comparable to the rate
found in a previous study at the Barrens Grouse HMA (Yahner et al.
1993). Our rate of disturbance of nests with Northern Bobwhite eggs
(38%) was higher than that reported with an artificial nest study in Vir-
ginia using Northern Bobwhite eggs (20%) (Leimgruber et al. 1994).

In contrast to nests with brown chicken eggs, nests with more visually

134

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

conspicuous egg types, i.e., white chicken and Northern Bobwhite, were
lost at a comparable rate despite appreciable differences in size between
the two types. These results concur with those of Montevecchi (1976),
who found similar rates of predation by American Crows on white eggs
of three sizes placed in meadows. Perhaps because corvids are common
nest predators at the Barrens Grouse HMA (Yahner and Wright 1985,
Yahner and Scott 1988), color rather than egg size was the major factor
influencing nest disturbance in our study. Although nests with Northern
Bobwhite eggs were preyed upon as expected, conceivably many of the
nests with missing Northern Bobwhite eggs at the end of trials could have
been those in which eggs easily were carried away by large avian nest
predators such as the American Crow (see Montevecchi 1976). For in-
stance, of the 39 nests with no eggs present at the end of a given trial,
the majority (N = 30, 77%) were those with eggs of Northern Bobwhite.

Some concern has been raised about size of eggs used in artificial nest
studies (e.g., Boag et al. 1984, Reitsma et al. 1990, Roper 1992). Eggs
of Japanese Quail used in artificial nest studies, for example, are consid-
erably larger than those of songbirds, e.g., warblers, thereby potentially
reducing rates of nest disturbance by smaller-sized mammalian predators
that are less efficient at handling a larger egg (e.g., red squirrels and
eastern chipmunks (Boag et al. 1984, Reistma et al. 1990). Thus, eggs
used in our study, which were intended to simulate egg size of larger
birds (e.g.. Ruffed Grouse, Wild Turkey), probably were too large for
handling by smaller predators.

Various investigators have often used either brown chicken eggs (e.g.,
Andren and Angelstam 1988, DeGraaf and Anglestam 1993, Yahner et
al. 1993) or Japanese Quail eggs (Boag et al. 1984, Ratti and Reese 1988)
as part of the experimental design of artificial nest studies in forested
habitats. Based on our findings, we recommend brown chicken eggs as a
suitable alternative to Japanese Quail eggs, both in terms of size and color,
when simulating nests of Ruffed Grouse and Wild Turkeys in artificial
ground nest studies.

ACKNOWLEDGM ENTS

We thank J. R. Gillis, S. M. Partridge, and B. D. Ross for field assistance. This study
was funded by the Pennsylvania Agricultural Experiment Station and the Max McGraw
Wildlife Loundation.

LITERATURE CITED

Andrew, P. and P. Angelstam. 1988. Elevated predation rates as an edge effect in habitat
islands: experimental evidence. Ecology 69:544—547.

Boag, D. A., S. G. Reebs, and M. A. Schroeder. 1984. Egg loss among spruce grouse
inhabiting lodgepole pine forests. Can. J. Zool. 62:1034-1037.

Yahner and Mahan • ARTIFICIAL GROUND NESTS

135

Bowman, G. B. and L. D. Harris. 1980. Effect of spatial heterogeneity by ground-nest
depredation. J. Wildl. Manage. 44:806-813.

Cross, A. O. 1946. Eastern Crow. Pp. 226—259 in Life histories of North American jays,
crows, and titmice (A. C. Bent, ed.). U.S. Natl. Mus. Bull. 191.

DeGraaf, R. M. and P. Angelstam. 1993. Effects of timber size-class on predation of
artificial nests in extensive forests. For. Ecol. Manage. 61:127-136.

Dixon, W. J. (chief ed.). 1990. BMDP statistical software manual. Univ. California, Berke-
ley, California.

Harrison, H. H. 1975. A field guide to birds’ nests of the eastern United States. Houghton
Mifflin Co., Boston, Massachusetts.

Leimgruber, P, W. j. McShea, and J. H. Rappole. 1994. Predation on artificial nests in
large forest blocks. J. Wildl. Manage. 58:254—260.

Montevecchi, W. a. 1976. Egg size and the egg predatory behavior of crows. Behaviour
57:307-320.

Nol, E. and R. j. Brooks. 1982. Effects of predator exclosures on nesting outcome of
killdeer. J. Field Ornithol. 53:263-268.

Picozzi, N. 1975. Crow predation on marked nests. J. Wildl. Manage. 39:151-155.

Ratti, j. T. and K. P. Reese. 1988. Preliminary tests of the ecological trap hypothesis. J.
Wildl. Manage. 52:484^91.

Rearden, j. D. 1951. Identification of waterfowl nest predators. J. Wildl. Manage. 15:386—
395.

Reitsma, L. R., R. T. Holmes, and T. W. Sherry. 1990. Effects of red squirrels, Tamias-
ciurus hudsonicus, and eastern chipmunks, Tamias striatus, on nest predation in a north-
ern hardwood forest: an artificial nest experiment. Oikos 57:375-380.

Roper, J. J. 1992. Nest predation experiments with quail eggs: too much to swallow? Oikos
65:528-530.

SoKAL, R. R. AND F. J. Rohlf. 1981. Biometry, second ed. W. H. Freeman and Co., San
Francisco, California.

Storaas, T. 1988. A comparison of losses in artificial and naturally occurring Capercaillie
nests. J. Wildl. Manage. 52:123-126.

Therres, G. D. 1982. Ruffed Grouse populations and relative predator abundance on a
grouse management area in central Pennsylvania. M.S. thesis. The Pennsylvania State
Univ., University Park, Pennsylvania.

Thurber, D. K., W. R. McClain, and R. C. Whitmore. 1994. Indirect effects of gypsy
moth defoliation on nest predation. J. Wildl. Manage. 58:493-500.

Westmoreland, D. and L. B. Best. 1986. Incubation continuity and the advantage of
cryptic egg coloration to Mourning Doves. Wilson Bull. 98:297-300.

WiLLEBRAND, T. AND V. Marcstrom. 1988. On the danger of using dummy nests to study
predation. Auk 105:378-379.

Yahner, R. H. 1991. Avian nesting ecology in small even-aged aspen stands. J. Wildl.
Manage. 55:155-159.

. 1992. Dynamics of a small mammal community in a fragmented forest. Amer.

Midi. Nat. 127:381-391.

. 1993. Effects of long-term clear-cutting on wintering and breeding birds. Wilson

Bull. 105:239-255.

AND B. L. Cypher. 1987. Effects of nest location on depredation of artificial ar-
boreal nests. J. Wildl. Manage. 51:178-181.

AND C. A. DeLong. 1992. Avian predation and parasitism on artificial nests and

eggs in two fragmented landscapes. Wilson Bull. 104:162-168.

136

THE WILSON BULLETIN • Vol. JOS, No. I. March 1996

■ AND C. G. Mahan.

1996. Depredation of artificial ground nests in a managed

forested landscapes. Conserv. Biol. 10:1^.

^ and C. a. DeLong. 1993. Dynamics of depredation on artificial ground

nes’ts in habitat managed for Ruffed Grouse. Wilson Bull. 105:172-179.

^ T. E. Morrell, and J. S. Rachael. 1989. Effects of edge contrast on depredation

of artificial avian nests. J. Wildl. Manage. 53:1 135-1 138. ■ i

and D. P. Scott. 1988. Effects of forest fragmentation on depredation of artificial

nests. J. Wildl. Manage. 52:158 161. tr- t j

and a. L. Wright. 1985. Depredation on artificial ground nests: effects of edge

and plot age. J. Wildl. Manage. 49:508-513.

Wilson Bull., 108(1), 1996, pp. 137-150

FOOD AVAILABILITY AND FEEDING
PREFERENCES OF BREEDING FULVOUS
WHISTLING-DUCKS IN LOUISIANA RICEFIELDS

William L. Hohman, Timothy M. Stark, and
Joseph L. Moore

Abstract. — Expansion of the breeding distribution of the Fulvous whistling-duck (Den-
drocygna bicolor) into the southeastern United States after the mid- 1800s coincided with
the establishment of rice (Oryza sativa) cultures in Texas, Louisiana, and Florida. In southern
Louisiana, where approximately 80% of rice is aerially seeded in water, Fulvous whistling-
ducks are suspected of feeding extensively on planted rice and are considered a nuisance.
To determine the extent of rice utilization by ducks nesting in southwestern Louisiana, we
estimated food availability in ricefields and assessed feeding preferences. We also examined
effects of sex and stage of reproduction on food selection. Feeding sites in Louisiana rice-
fields that were tilled and flooded in preparation for spring planting, contained abundant
foods (mean ± SE = 109.0 ± 18.0 g/m^), especially seeds of moist soil plants such as
signalgrass (Brachiaria exten.sa), beakrush (Rhynchospora sp.), and flatsedge (Cyperus iria).
Diets of males and females were similar {P = 0.080), but varied through the reproductive
cycle (P = 0.008). Consumption of plant material was slightly reduced during the period
of rapid ovarian follicle growth in females; however, ingestion of animal foods never ex-
ceeded 4%. Fulvous whistling-ducks exhibited feeding preferences {P < 0.001) with aquatic
earthworms (Oligochaeta) and wild millet seeds (Echinochloa sp.) being preferred over other
food taxa. Rice made up <4% of the diet and was selected in proportion to its availability
before and during period of rapid follicle development. Almost 25% of the diet of incubating
females consisted of rice; however, we concluded that crop depredation by Fulvous whis-
tling-ducks (<0.1%) was of minor importance relative to other potential sources of crop
loss. Indeed, use of ricefields by whistling-ducks may actually benefit farmers if ingestion
of seeds of undesirable plants reduces the need for costly herbicide treatments. Received 18
April 1995, accepted 22 Sept. 1995.

Private lands provide critical habitat for many wildlife species, but
wildlife use of these areas sometimes results in significant economic
losses (e.g., crop depredation) or conflicts with intended land uses (e.g.,
designation as critical habitat for threatened or endangered species). Since
1987, >1 million ha of rice {Oryza sativa) have been planted annually in
the United States, mostly in the Mississippi Alluvial Valley, Gulf Coastal
Plain, and Central Valley of California. In these regions, ricefields similar
to other seasonally flooded habitats receive high use by shorebirds, wad-
ing birds, and waterfowl (hereafter waterbirds). Rice prairies in eastern
Texas, for example, provide wintering habitat for >2 million waterfowl
(Hobaugh et al. 1989). In Louisiana and California, harvested ricefields

National Biological Service, Southern Science Center, 700 Cajundome Blvd., Lafayette Louisiana 70506-
3152.

137

138

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

are used extensively by feeding and resting waterbirds in winter and dur-
ing fall migration (Miller 1987, Heitmeyer et al. 1989, Rave and Cordes
1993). Indeed, flooding of ricefields after harvest to provide wintering
and migrational habitat for waterbirds has been actively promoted by
some farmers’ groups, agricultural extension services, state and federal
wildlife agencies, representatives of the rice industry, and private conser-
vation organizations. Advantages to rice farmers participating in winter
flooding programs include enhanced waterfowl hunting (leasing) and
viewing opportunities, as well as potential for positive public image, re-
tention of nutrients and topsoil, weed control, stubble removal, and low-
ered tillage costs.

Ricefields may also receive high use by spring-migrating and nesting
waterbirds (Helm et al. 1987, Hohman et al. 1994), but avian use of fields
after they have been prepared for planting until harvest is actively dis-
couraged. Waterbird use of ricefields in spring and summer may be es-
pecially great in areas such as southern Louisiana where most rice is
planted in water (“water seeding’’); that is, pregerminated seed is aerially
dispersed over fields following discing, flooding, leveling or dragging
with a blade, and settling of particulate matter. Water-seeded fields gen-
erally are drained within 24 hours of planting, but are reflooded from 7—
14 days after rice has sprouted until 2—3 weeks before harvest. Elsewhere
rice is mostly broadcast or drilled in dry fields ( dry seeding ). Both
dry- and water-seeded fields may be flooded in winter and are managed
similarly after rice has sprouted, so the principal difference between plant-
ing methods is the presence of water in fields immediately before spring
planting. In spite of increased risks of seed depredation by waterbirds,
water seeding is preferred by Louisiana rice farmers to control weeds.

Fulvous whistling-ducks {Dendrocygna bicolor, hereafter whistling-
ducks) occur worldwide in tropical and semitropical regions (Johnsgard
1978). Their expansion into the southeastern United States in the late
1800s coincided with the establishment of rice cultures in Texas, Loui-
siana, and Florida (Lynch 1943, Bolen and Rylander 1983, Turnbull et
al. 1989). The first breeding records for whistling-ducks in Louisiana were
obtained in 1939 (Lynch 1943). Their numbers in Louisiana increased
rapidly in the 1940s to perhaps 10,000 ducks but soon decreased because
of hazing practices adopted by rice farmers to reduce crop depredation
(McCartney 1963). Introduction of aldrin (a pesticide used to protect seed
against larvae of rice water weevil {Lissorhoptrus oryzophilus]) in 1960,
further depressed whistling-duck populations in Texas and Louisiana. Al-
though the Louisiana population has recovered somewhat since 1970
when use of aldrin-treated seed was discontinued, numbers of whistling-
ducks remain below peak counts observed in the 1940s in spite of in-

Hohman et al. • WHISTLING DUCK DIET

139

creased acreages of planted rice (Flickinger et al. 1977, Zwank et al.
1988).

Whistling-ducks nesting along the western Gulf Coastal Plain are mi-
gratory (Flickinger et al. 1973, Hohman and Richard 1994), arriving in
southern Louisiana in February or March (McCartney 1963) when rice-
fields are being flooded in preparation for planting. Because of their pres-
ence in ricefields around the time of planting, occasionally in flocks of
>2000 birds (Davis et al. 1944 in McCartney 1963), whistling-ducks are
suspected of feeding on planted rice; consequently, they and, secondarily,
other waterbirds (e.g., shorebirds and wading birds) are actively hazed
from fields by rice farmers. We conducted this study to determine the
extent of rice utilization by whistling-ducks nesting in southwestern Lou-
isiana. Specifically, we estimated food availability in ricefields, assessed
feeding preferences, and examined effects of sex and stage of reproduc-
tion on food selection by whistling-ducks.

METHODS

Whistling-ducks were collected on private agricultural lands in southwestern Louisiana,
9-15 May 1992 and 18 March-8 May 1993. Ducks were collected throughout the diurnal
period and most were observed feeding for a minimum of 15 min before collection. Alcohol
was injected into the gullets of specimens immediately after collection to minimize post-
mortem digestion of foods (Bailey and Titman 1984). We assigned pair status to birds on
the basis of observations made before collection. Paired individuals were those showing
active association, i.e., copulation, mutual display, female tolerance of the male or nonran-
dom spacing. Sex was assigned on the basis of cloacal characteristics (Hochbaum 1942).
Specimens then were wrapped in paper towels and frozen in sealed plastic bags.

In the laboratory, thawed specimens were dissected and esophageal contents were re-
moved, weighed (±0.01 g), and frozen. Carcasses were retained for contaminant analyses
and proximate analyses of fat and protein composition. Ovaries removed from females were
weighed (±0.01 g) and inspected for evidence of postovulatory follicles. We assigned fe-
males and their mates to the following reproductive categories, based in part on Krapu
(1974): Prenesting — females with ovary mass <3 g and no post-ovulatory follicles; Rapid
follicle growth (REG) — preovulatory females with ovary mass >3 g and ovulating females;
Incubation — birds collected at nest sites with embryo development si day (Weller 1956,
Caldwell and Snart 1974).

Food availability was sampled at feeding sites by using a 6.1 -cm diameter corer inserted
to a substrate depth of at least 10 cm. Three or five core samples were taken at each feeding
site. Corer contents (water column and substrate) were emptied into individual plastic bags
and frozen. Thawed esophageal and core samples were hand-sorted to remove all macro-
scopic plant and animal material. Core .samples were initially washed through a .series of
screens with 0.0625-4.0 mm^ openings. Plant and animal taxa were .separated, identified,
and dried to constant ma.ss (±0.001 g) at 50°C. Common names of invertebrates and plants
followed Pennak (1989) and Scott and Wasser (1980), respectively. Food habits and avail-
ability were summarized on an aggregate percentage of dry mass basis (Swanson et al.
1974). Only food samples from birds containing >five items were included in the analysis
(Reinecke and Owen 1980).

The proportion of plant material in the diet was compared by reproductive status and sex

140

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

using two-way analysis of variance with Type III sum of squares on arsine square _root
transformed dL (PROC GLM, SAS Institute, Inc. 1987). To determine if whistling duck
diets differed between day and night, we compared occurrences (presence or absence) of
rice in incubating birds collected before and after 08:00 h with a Chi-squared test (Conover
1980) Incubators collected before 08:00 h and found to have food in their esophagi were
assumed to have fed at night. Food preferences were assessed on a dry mass basis by using
PREFER, a computer program that assesses preferences using nonparametnc proc^ures
(Johnson 1980). Only foods having an aggregate percentage of dry mass > an -
frequency of occurrence in use or availability samples were included m the analysis These
melded foxtail {Alopecurus caroUnianus), rice, junglerice barnyardgrass iEchtnochloa co-
lonum), broadleaf signalgrass {Brachiaria extensa), rice flatsedge (Cyperus ina), other flat-
sedges {Cyperus spp.), spikerush {Eleocharis sp.), beakrush (Rhynchospora sp.), razorsedge
(Scleria sp.), mudplantain (Heteranthera limosa), buttercup {Ranunculus spp.), lesser ^me-
cress {Coronopus didvmus), morningglory {Ipomoea sp.), and aquatic earthworms (O igo-
chaeta) Aggregate percentages of dry mass of taxa collected at feeding sites were assumed
to represent food available to whistling ducks at those sites. Significance level was set a

priori at a = 0.05.

RESULTS

Food availability.— Forty-nine cores were taken at 13 feeding sites in
five southwestern Louisiana ricefields. Estimates of food density in in-
dividual ricefields ranged from 53.1-171.5 g/m^ and averaged (± SE)
109.0 ± 18.0 g/nrF- Plant material consisted almost exclusively of seeds
and made up >98 aggregate percentage of dry mass of available foods
(Table 1). A minimum of 28 plant taxa were identified in availability
samples of which only four taxa contributed >5%. Although animal foods
made up <2% of available foods, they were present at all feeding sites.
Only one animal taxon (aquatic earthworms) contributed appreciably to

available foods. • . •

Food use. — Eighty-five of 121 whistling-ducks collected in this study

had >5 food items in their esophagi. Four males collected without mates
were of unknown reproductive status and excluded from subsequent anal-
yses. The proportion of plant material in the diet of breeding whistling
ducks was similar in males and females (E[, 75] = 3.15, P - 0.080)
throughout the reproductive cycle (^[2,75] = 1.31, P = 0.276), but vaned
among reproductive categories (F[2,75] = 5.21, P = 0.008). Plant food
consumption was somewhat reduced during REG relative to other repro-
ductive categories, but never decreased below 96% even in females. We
found no difference in the prevalence of rice in esophagi of incubating
whistling-ducks collected before and after 08:00 h (x^ = 0.024, 1 df, P
= 0.84). Plant foods eaten by whistling-ducks consisted almost exclu-
sively of seeds from >29 taxa, 14 of which contributed >1% dry mass
or occurred with >50% frequency (Table 2). Aquatic earthworms were
the only animal food contributing appreciably to the diet.

Feeding preferences. — Whistling-ducks exhibited feeding preferences

Hohnicm et al. • WHISTLING DUCK DIET

141

Table 1

Food Availability at Fulvous Whistling Duck Feeding Sites in Five Southwestern

Louisiana Ricefields

Food taxa'^

Aggregate %
dry mass

Dry mass

(g/m-)

occurrence

Mean ± SE

Range

Plant

98.2

100.0

106.8 ± 17.2

52.6-165.4

Seeds

96.0

100.0

101.4 ± 16.5

52.6-165.4

Alopecurus carolinianus

2.3

100.0

1.5 ± 1.0

0.0-5.6

Oryzci saliva

4.2

60.0

4.1 ±2.1

0.0-12.6

Echinochloa colonum

1.2

100.0

1.0 ± 0.4

0.2-2. 1 ■

Brachiaria extensa

30.6

100.0

34.4 ± 9.2

2.4-60.7

Cyperus iria

10.7

80.0

8.9 ± 4.6

0.0-23.7

Cyperus spp.

2.3

100.0

2.3 ± 0.8

0.3-4.5

Eleocharis sp.

3.9

80.0

3.1 ± 2.2

0.0-12.9

Rhynchospora sp.

19.0

100.0

29.5 ± 16.3

0.0-96.6

Heteranthera limosa

6.9

100.0

3.1 ± 2.6

0.0-14.5

Ranunculus spp.

4.1

100.0

3.6 ± 1.5

0.6-9.7

Coronopus didynius

1.2

80.0

0.9 ± 0.6

0.0-3.3

Ipomoea sp.

1.2

80.0

1.1 ± 0.5

0.0-3.2

Miscellaneous'’

8.4

100.0

8.0 ± 2.5

1.0-17.0

OtheU

2.1

40.0

5.3 ± 4.7

0.0-26.4

Animal

1.8

100.0

2.2 ± 0.9

0.5-6. 1

Oligochaeta

1.0

80.0

1.3 ± 0.9

0.0-5. 5

Miscellaneous*'

0.9

100.0

0.8 ± 0.2

0.5-L6

’ Includes only laxa with aggregate percentage of dry mass a I and frequency of occurrence £50%.

Miscellaneous seeds were from Motiugo verticillata, Cerastium viscosum, Commelina sp., EcUpta spp., Serinea oppos-
itifolia. Scleria sp., Fimhrisrylis miliacea, Sisyrinchium sp., Digilaria scmguinalis, Lolium temulenlum, Panicum spp., Pluil-
aris sp.. Polygonum hyciropiperoides, Polygonum porloricen.se, Solanum americanum, and Verbena sp.

'Other plant material included unidentified roots and tubers.

Miscellaneous animals included unidentified vertebrate and invertebrate eggs, Copepoda, Coleoptera (larvae and adults),
Chironomidae (larvae and pupae), Corixidae (adults), Formicidae (adults), and Gastropoda.

during both the prenesting (F[,3 22j = 36.60, P < 0.001) and RFG periods
(^[12.16] ~ 10.68, P < 0.001). Aquatic earthworms and junglerice barn-
yardgrass were preferred over other food items during both reproductive
periods (Table 3). Spikerush, flatsedge, and beakrush seeds were under-
represented in the diets, whereas rice was eaten in proportion to its abun-
dance (Table 3).

DISCUSSION

Food availability. — Density of potential foods, especially seeds of
moist soil plants, was high in Louisiana ricehelds used by feeding whis-
tling-ducks in spring. Our estimate of seed density at whistling-duck feed-
ing sites (101.2 ± 16.5 g/m^) was comparable to that (range, 90-134.4
g/m2) in impoundments in the Mississippi Alluvial Valley managed spe-

142

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Table 2

Foods of Male (M) and Female (F) Fulvous

Whistling Ducks Collected in

Agricultural Areas in Southwestern Louisiana

Aggregate % dry

mass

<RFG"

RFG

M + F

M

F

M + F

Food taxa*”

N = 35

N = 15

N = 16

N = 31

Plant

98.1

99.0

96.1

97.5

Seeds

97.6

98.9

96.0

97.4

Lolium temulentum

0.0

0.6

3.4

2.0

Triticum aestivum

0.0

0.0

0.0

u.u

Phalaris sp.

tc

0.2

0.3

0.2

Oryza sativa

3.6

1.4

6.1

3.8

Echinochloa colonum

8.9

9.3

13.3

1 1.3

Brachiaria extensa

27.8

49.3

45.2

47.2

Panicum spp.

0.3

tr

tr

tr

Cvperus iria

30.3

0.2

1.7

1.0

Cvperus spp.

0.2

1.4

0.8

1.1

Eleocharis sp.

0.6

0.0

0.3

0.2

Rhynchospora sp.

12.1

32.4

21.6

26.8

Scleria sp.

3.0

3.4

0.8

2.1

Heteranthera limosa

8.6

tr

tr

tr

Ranunculus spp.

1.1

0.2

0.5

0.4

Miscellaneous'^

0.5

0.5

2.0

1.3

OtheF

1.1

0.1

0.1

0.1

Animal

1.9

1.0

3.9

2.5

Oligochaeta

1.7

0.6

2.6

1.6

Miscellaneous'^

0.2

0.4

1.3

l).v

« Renmductive catefiories' <RFG = prenestmg females (and ineir mates; wiiii uvai^ ..v. ^

folliclL; RFC = preo^vulatory females with ovary mass >3 g and ovulating females; >RFG - birds collected at nest sites

with embryo development a 1 day. -,cnw

” Includes only taxa with aggregate percentage of dry mass al or frequency of occurrence _50%.

^ Trace (tr), aeeregate % dry mass <0.1. .

<■ Miscellaneous seeds were from Coronopus didymus. Commelina sp., Echpta alba, Ecltpta sp., opposi if .

Ipomoea sp., Cyperus compressu.i. Fimbristylis miUacea. Sisyrincbium sp., Alopecurus caroUmamis. D, guana sangumahs.
Phalaris sp., Triticum aestivum. Polygonum hydropiperoides, and Verbena sp.

'Other nlani material included unidentified roots, tubers, and other parts.

tMisclll^aneLs animals included unidentified invertebrate eggs, Coleoptera (larvae and adults), Chironomidae (larvae
and pupae), Tabanidae (larvae), Formicidae (adults), and Mollusca (Gastropoda and Pelecypoda).

cifically for production of moist soil plants (Reid et al. 1989) and 2^
times greater than densities of seeds and all other plant foods in nearby
coastal marshes (Jemison and Chabreck 1962, Hohman et al. 1990, Bie-
lefeld and Afton 1992, Manley et al. 1992). Observed seed densities were
substantially greater than previous estimates obtained in Louisiana rice-
fields in late winter (4.3—38.0 g/m^; Harmon et al. 1960, Davis et al.
1961); however, if selection of feeding sites by whistling-ducks was in-

Hohman et al. • WHISTLING DUCK DIET

143

Table 2
Extended

% Occurrence

>RFG

<RFG

RFG

>RFG

M + F

M+F

M

F

M + F

M + F

N = 15

N = 35

N = 15

N = 16

N = 31

N = 15

99.8

100.0

100.0

100.0

100.0

100.0

99.8

100.0

100.0

100.0

100.0

100.0

2.7

0.0

20.0

18.8

19.4

53.0

5.1

0.0

0.0

0.0

0.0

13.3

0.1

54.3

33.3

62.5

48.4

20.0

24.4

60.0

13.3

31.3

22.6

46.7

16.6

91.4

80.0

87.5

83.9

80.0

28.9

94.3

86.7

100.0

93.6

80.0

0.0

57.1

6.7

6.3

6.5

0.0

0.1

68.6

26.7

31.3

29.0

46.7

tr

45.7

46.7

18.8

32.3

13.3

tr

62.9

0.0

6.3

3.2

20.0

20.7

20.0

86.7

68.8

77.4

46.7

0.5

54.3

26.7

12.5

19.4

46.7

0.1

68.6

6.7

12.5

9.7

33.3

0.1

74.3

46.7

50.0

48.4

33.3

0.4

82.9

53.3

93.8

74.2

33.3

tr

28.6

26.7

18.8

22.6

6.7

0.2

68.6

60.0

100.0

80.7

26.7

0.1

51.4

46.7

68.8

58.1

13.3

0.1

40.0

60.0

93.8

77.4

26.7

fluenced by food availability (i.e., bird avoidance of sites with reduced
food availability), then we likely overestimated food density in ricefields.

Abundance of potential foods in ricefields and their availability to feed-
ing waterbirds vary temporally and geographically in relation to farming
practices. Seed density in ricefield sediments is probably maximal im-
mediately after autumn harvest (Harmon et al. 1960, Miller et al. 1989)
and declines thereafter as a result of granivory, germination, physical
degradation or destruction (e.g., tilling or burning), burial, and dispersal
of seeds (McGinn and Glasgow 1963). To control noxious weeds such as
red rice {O. sativa van), most rice farmers in southern Louisiana practice
a two-year planting cycle with rice cultivated in rotation with fallow,
crayfish (Decapoda) aquaculture, pasture, or row crops (e.g., soybeans,
milo, or wheat). Fields sampled in this study had been flooded and me-
chanically treated (disced and bladed) in preparation for water seeding of

144

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

UJ

-J

<

H

<

z

<

o

-!

z

OS

tu

H

on

W

O

oo

c

J

l£

tu

W

u

C

tu

H

U

tu

_)

o

U

c/?

u

o

z

X

t/)

3

o

>

J

o

z

E

tu

tu

u

pa

t/i

tu

tj

tu

oi

tu

tu

tu

oi

CL

O

z

5

tu

tu

Uh

u

a,

GO

U

u

J

w

o

CL

<

Pi

CQ

H

U

z

<

Pi

o

Pi

o

X

Pi

Pi

o

u

u

GO

X

u

UJ

J

o

Oi)

c

to

(U

c

1)

Z

<

Pi

O

CL

Pi

O

u

X

X

u

p,

v>

u

<

Pi

oa

Pi

u

o

X

<

Pi

o

H

UJ

X

X

Pi

-J

o

00

r^\

<u II

1 Z

£ :S

o

Q, »-(

OC

^ T3

£ §

rt Li-

0

s' QC

1 R

3 &

o

>■ C

QC

O II

r 0-

CL uo

U

3

Pt

X W
tu

■3 2

2 I
s-^
§ "
S P

o ^

tKC ?

r 53
Q.'c:

CO 5

^

11 " s

J < d

u ai A
^ CCi

S -<

s :5

II

2 i3

. 0.

01)

c

c

P3

^ *5 b
s: J

~ 'U

;§ a: ;

tu
II H

-d X y ■=
(U r 1 X 2

,<u ^

c/J

-2

15

E

01)

1)

0> u

a. >

CO

c^ s:

2
^ 5

CO
GO

o y

c O U
C 00 S

J- 5 D

II or e <-

a c

Hohman el ul. • WHISTLING DUCK DIET

145

rice (i.e., rice had not been planted in these helds in the preceding growing
season). Seeds found in our samples presumably were produced during
the previous growing season. Thus, it is apparent that the farming prac-
tices implemented between rice plantings may have a large influence on
seed abundance in Louisiana ricefields in spring. Outside of the Gulf
Coastal Plain, rice is mostly dry-seeded with or without crop rotation.
Flooding of ricefields, as is practiced by farmers that water-seed rice, is
necessary for waterbirds to gain access to potential foods. The effect on
food availability of mechanical treatments performed in flooded ricefields
is unclear, but the appearance of large numbers of birds (shorebirds as
well as waterfowl) in fields following such treatments, especially blading
(W. L. Hohman, pers. obs.), suggests that food availability may be en-
hanced.

Feeding preferences. — Greater than 96% of the diet of male and female
whistling-ducks nesting in southwestern Louisiana was composed of plant
material. Animal foods were actively selected by whistling ducks before
and during RFG (i.e., period of high protein demand in females), and
animal food consumption increased slightly (females only) during RFG.
Nonetheless, the amount of animal food eaten by whistling ducks during
RFG was less than that reported for any other small-bodied waterfowl
species (Krapu and Reinecke 1992: tables 1-5). Other female ducks, even
those that are primarily herbivorous (e.g., Gadwall {Anas strepera], An-
kney and Alisauskas 1991), substantially increase their consumption of
animal foods to offset high protein costs of reproduction (Krapu and Re-
inecke 1992), but that apparently is not necessary for female whistling
ducks. Black-bellied whistling-Ducks (Dendrocygna autumnalis) also eat
only small amounts (<10%) of animal foods during the nesting period
(Bourne 1981). Although the amount of animal material at whistling-duck
feeding sites was low relative to plant material, our estimate of animal
food density (2.2 ± 0.9 g/m^) was comparable to that (2.65-2.87 g/m^)
found in freshwater coastal marshes where spring-migrating blue-winged
teal (A. discors) consumed >56% animal material (Manley et al. 1992).
This result suggests that whistling-ducks fed inefficiently on animal foods
or that not all foods found in core samples were available to birds. It
further suggests that proteins required for production of eggs must come
from exogenous or endogenous sources in addition to those contained in
animal foods eaten by birds during the daytime.

Whereas previous studies reported that whistling ducks using ricefields
eat mostly rice (Imler 1944 in Meanley and Meanley 1959; Bruzual and
Bruzual 1983), we found only limited consumption of rice by ducks nest-
ing in southwestern Louisiana. Rice made up <4% of the diet and was
selected in proportion to its abundance before and during RFG. Almost

146

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

25% of the diet of incubating whistling-ducks consisted of rice, but we
were unable to assess feeding preferences of incubating whistling ducks
because feeding sites were unknown and food availability therefore could
not be determined. Whistling-ducks are known to feed in flooded ricefields
at night (Meanley and Meanley 1959). The potential for crop depredation
presumably is greatest at night when whistling-ducks can feed undisturbed.

It is possible that our sampling of birds only during the daytime underes-
timated rice utilization by whistling-ducks; however, prevalence of rice in
esophagi of incubating birds collected before 08:00 h (assumed to have fed
at night) was similar to that of incubators collected after 08.00 h. Sample
size used for this comparison was limited, but we interpret this result as
evidence that bias associated with time of collection was minimal. We
therefore concur with Meanley and Meanley (1959) that, relative to other
seeds, rice was of minor importance in overall diet of whistling-ducks
nesting in southwestern Louisiana.

Our conclusion that whistling-ducks ate small amounts of rice relative
to seeds of other plants should not be interpreted as evidence that they
caused no damage in ricefields. Based on energy requirements calculated
for nesting whistling-ducks, we estimate maximum daily consumption of
rice to be 44.5 g/bird or 44.5 kg/day for the entire population in southern
Louisiana (Table 4). Thus, we determined the potential for crop depre-
dation in southern Louisiana during the 60-day planting period to be
^0.1% of seeded rice (Table 4). Previous estimates of crop loss caused
by feeding whistling-ducks ranged from 0.25-2.0% (Imler 1944 in Mean-
ley and Meanley 1959; McCartney 1963, Bourne and Osbourne 1978),
but estimates made before 1965 probably do not accurately represent
losses under current farming practices. Use of pregerminated seed, for
example, greatly reduces the duration of flooding after planting and there-
by limits availability of rice to feeding ducks. (We observed no diurnal
feeding by whistling-ducks in dewatered fields.) Removal of water within
48 hours of planting also minimizes puddling or trampling of seeded rice
(i.e., reduced sprouting caused by birds stepping on and burying rice seed,
McCartney 1963). Concentrated feeding by large flocks of whistling-
ducks may result in localized crop losses greater than those projected in
this study, but we believe that such instances are uncommon. Under cur-
rent farming practices, depredation is restricted to fields planted early in
the growing season (before 1 April) when whistling ducks occui in flocks
and temperatures are cool, requiring farmers to hold water on seeded
fields >48 h. We further suggest that crop losses caused by whistling-
ducks are of minor importance relative to other potential sources of crop
loss such as other seed predators, variable seed germination rates, weather,
and disease. Indeed, we suggest that use of ricefields by whistling-ducks

Hohrnan el at. • WHISTLING DUCK DIET

147

Table 4

Energy Requirements of Breeding Fulvous Whistling Ducks and Potential Crop
Losses in Southern Louisiana Ricefields

Calculation assumptions

Sources

Total seeded rice in southern Louisiana = 22,662,080 kg
Rice acreage = 202,340 ha

Anonymous (1995)

80% of acreage was water-seeded (i.e., available)

R. Levy (pers. comm.)

Seeding rate =140 kg/ha

Anonymous (1995)

Maximum daily rice consumption/bird = 45.9 g

Whistling Duck Diet = 100% rice'"

True metabolizable energy of rice = 3.34 kcal/g

Reinecke et al. (1989)

Body mass = 756 ± 4 g

W. L. Hohrnan (unpubl. data)

Basal Metabolic Rate (BMR) = 75 * (body mass
[kg])“’^

Owen and Reinecke (1979)

= 61.3 kcal/day

Daily energy expenditures = 2.5 BMR

Owen and Reinecke (1979)

= 153.3 kcal/day

Maximum seasonal rice consumption/population
= 27,539 kg

Planting season = 60 days

Anonymous (1995)

Population = 10,000 ducks

Flickinger et al. (1977)

" Actual range = 3.6-24.3% (This study).

may actually benefit farmers if ingestion of seeds of undesirable plants
reduces the need for costly herbicide treatments.

ACKNOWLEDGMENTS

We thank the county agents in Acadia (R. Levy), Evangeline (A. Mire), and Vermillion
(C. McCrory [deceased], H. Cormier, and M. Shirley) Parishes, and also E Bowers (U.S.
Fish and Wildlife Service) and C. Cordes (Southern Science Center) for their support of
this study. We are sincerely grateful to the many rice farmers and other cooperators who
shared information with us and/or granted us permission to conduct this study on their land.
R. E. Olson and G. A. Weisbrich assisted in the field or laboratory. Helpful reviews of our
manuscript were provided by S. Linscombe, Z. A. Malaeb, S. W. Manley, K. J. Reinecke,
B. A. Vairin, and M. W. Weller. Collections were authorized under scientific permits issued
to the senior author by the U.S. Fish and Wildlife Service (PRT-747267) and Louisiana
Dept, of Wildlife and Fisheries (LNHP-93-02 and LNHP-94-21). Before implementation of
this study, procedures involving use and care of birds were reviewed and approved by the
Anim.al Use and Care Committee, Southern Science Center.

LITERATURE CITED

Anonymous. 1994. Rice industry guide. Rice J. 97:33-39.

. 1995. Rice varieties and management tips. Louisiana State Univ. Agricultural

Center Pub. 2270.

148

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Ankney, C. D. and R. T. Alisauskas. 1991. Nutrient-reserve dynamics and diet of breeding
female Gadwalls. Condor 93:799-810.

Bailey, R. O. and R. D. Titman. 1984. Habitat use and feeding ecology of postbreeding
Redheads. J. Wildl. Manage. 48:1 144-1 155.

Bielefeld, R. R. and A. D. Afton. 1992. Canvasback food density in the Mississippi River
Delta, Louisiana: habitat and temporal differences. Proc. Annu. Conf. Southeast. Assoc.
Game Pish Agencies 46:97-103.

Bolen, E. G. and M. K. Rylander. 1983. Whistling ducks: zoogeography, ecology, anat-
omy. Mus. Spec. Publ. 20, Texas Tech. Univ., Lubbock, Texas.

Bourne, G. R. 1981. Pood habits of Black-bellied Whistling Ducks occupying rice culture
habitats. Wilson Bull. 93:551—554.

and D. R. Osborne. 1978. Black-bellied whistling duck utilization of a rice culture

habitat. Interciencia 3:152-159.

Bruzual, J. and I. Bruzual. 1983. Peeding habits of whistling ducks in the Calabozo
ricefields, Venezuela, during the non-reproductive period. Wildfowl 34:20-26.

Caldwell, P. J. and A. E. Snart. 1974. A photographic index for aging Mallard embryos.
J. Wildl. Manage. 38:298-301.

Conover, W. J. 1980. Practical nonparametric statistics. John Wiley and Sons, New York,
New York.

Davis, J. P, C. H. Thomas, and L. L. Glasgow. 1961. Poods available to waterfowl in
fallow rice fields of southwestern Louisiana, 1960—1961. Proc. Annu. Conf. Southeast.
A.SSOC. Game Pish Comm. 15:60-66.

Plickinger, E. L., K. a. King, and O. Heyland. 1973. Pen-reared Fulvous Tree Ducks in
movement studies of wild populations. J. Wildl. Manage. 37:171-175.

, D. S. Lobpries, and H. A. Bateman. 1977. Fulvous Whistling Duck populations

in Texas and Louisiana. Wilson Bull. 89:329—331.

Harmon, B. G., C. H. Thomas, and L. Glasgow. 1960. Waterfowl foods in Louisiana
ricefields. Trans. N. Am. Wildl. Nat. Resour. Conf. 25:153—161.

Heitmeyer, M. E., D. P. Connelly, and R. L. Pederson. 1989. The Central, Imperial, and
Coachella Valleys of California. Pp. 475-506 in Habitat management for migrating and
wintering waterfowl in North America (L. M. Smith, R. L. Pederson, and R. M. Ka-
minski, eds.). Texas Tech Univ. Press, Lubbock, Texas.

Helm, R. N., D. N. Pashley, and P. J. Zwank. 1987. Notes on the nesting of the Common
Moorhen and Purple Gallinule in southwe.stern Louisiana. J. Field Ornithol. 58:55-61.

Hobaugh, W. C., C. D. Stutzenbaker, and E. L. Plickinger. 1989. The rice prairies. Pp.
367-383 in Habitat management for migrating and wintering waterfowl in North Amer-
ica (L. M. Smith, R. L. Pederson, and R. M. Kaminski, eds.). Texas Tech Univ. Press,
Lubbock, Texas.

Hochbaum, H. a. 1942. Sex and age determination of waterfowl by cloacal examination.
Trans. N. Am. Wildl. Nat. Resour. Conf. 7:299-307.

Hohman, W. L., D. W. Woolington, and J. H. Devries. 1990. Food habits of wintering
Canvasbacks in Louisiana. Can. J. Zool. 68:2605-2609.

, J. L. Moore, T. M. Stark, G. A. Weisbrich, and R. A. Coon. 1994. Breeding

waterbird use of Louisiana rice fields in relation to planting practices. Proc. Annu. Conf.
Southeast. Assoc. Fish Wildl. Agencies 48:31-37.

AND D. M. Richard. 1994. Remigial molt in Fulvous whistling ducks nesting in

Louisiana. Southwest. Nat. 39:190-192.

Jemison, E. S. and R. H. Chabreck. 1962. The availability of waterfowl foods in coastal
marsh impoundments in Louisiana. Trans. N. Am. Wildl. Nat. Resour. Cont. 27.288—
300.

Hohnum et at. • WHISTLING DUCK DIET

149

JoHNSGARD, P. A. 1978. Waterfowl of the world. Univ. Nebraska Press, Lincoln, Nebraska.

Johnson, D. H. 1980. The comparison of usage and availability measurements for evalu-
ating resource preference. Ecology 61:65-71.

Kraph, G. L. 1974. Feeding ecology of Pintail hens during reproduction. Auk 91 :278-290.

AND K. J. Reinecke. 1992. Foraging ecology and nutrition. Pp. 1-29 in Ecology

and management of breeding waterfowl (B. D. J. Batt, A. D. Alton, M. G. Anderson,
C. D. Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, eds.). Univ. Minnesota
Press, Minneapolis, Minnesota.

Lynch, J. J. 1943. Fulvous tree-duck in Louisiana. Auk 60:100—102.

Manley, S. W., W. L. Hohman, J. L. Moore, and D. Richard. 1992. Food preferences in
migrating Blue-winged Teal in southwestern Louisiana. Proc. Annu. Conf. Southeast.
Assoc. Game Fish Agencies. 46:46—56.

McCartney, R. B. 1963. The Fulvous Tree Duck in Louisiana. M.S. thesis, Louisiana State
Univ., Baton Rouge, Louisiana.

McGinn, L. R. and L. L. Glasgow. 1963. Loss of waterfowl foods in ricelields in southwest
Louisiana. Proc. Annu. Conf. Southeast. Assoc. Game Fish Comm. 17:69-79.

Meanley, B. and a. G. Meanley. 1959. Observations on the Fulvous Tree Duck in Lou-
isiana. Wilson Bull. 71:33-45.

Miller, M. R. 1987. Fall and winter foods of Northern Pintails in ihe Sacramento Valley,
California. J. Wildl. Manage. 51:405-414.

, D. E. Sharp, D. S. Gilmer, and W. R. Mulvaney. 1989. Rice available to wa-
terfowl in harvested fields in the Sacramento Valley, California. Calif. Fish Game 75:
113-123.

Owen, R. B., Jr. and K. J. Reinecke. 1979. Bioenergetics of breeding dabbling ducks. Pp.
71-93 in Waterfowl and wetlands — an integrated review (T. A. Bookhut, ed.). LaCrosse
Printing Co., Inc., LaCrosse, Wisconsin.

Pennak, R. W. 1989. Fresh-water invertebrates of the United States. John Wiley & Sons,
New York, New York.

Rave, D. P. and C. L. Cordes. 1993. Time-activity budgets of Northern Pintails using
nonhunted rice fields in southwest Louisiana. J. Field Ornithol. 64:21 1-218.

Reid, F. A., J. R. Kelley, Jr., T. S. Taylor, and L. H. Fredrickson. 1989. Upper Mi.ssis-
sippi Valley wetlands — refuges and moist-soil impoundments. Pp. 181-202 in Habitat
management for migrating and wintering waterfowl in North America (L. M. Smith.
R. L. Pederson, and R. M. Kaminski, eds.). Texas Tech Univ. Press, Lubbock, Texas.

Reinecke, K. J. and R. B. Owen, Jr. 1980. Food use and nutrition of Black Ducks nesting
in Maine. J. Wildl. Manage. 44:549-558.

, R. M. Kaminski, D. J. Moorhead. J. D. Hodges, and J. R. Nassar. 1989. Mi.s-

sissippi Alluvial Valley. Pp. 203—247 in Habitat management for migrating and win-
tering waterfowl in North America (L. M. Smith, R. L. Pederson, and R. M. Kaminski,
eds.). Texas Tech Univ. Press, Lubbock, Texas.

SAS Institute, Inc. 1987. SAS/STAT guide for personal computers. Version 6 ed. SAS
Inst., Inc., Cary, North Carolina.

Scott, T. G. and C. H. Wasser. 1980. Checkli.st of North American plants for wildlife
biologists. The Wildlife Society, Washington, D.C.

Swanson, G. W, G. L. Krapu, J. D. Bartonek, J. R. Serie, and D. H. Johnson. 1974.
Advantages in mathematically weighing waterfowl food habits data. J. Wildl. Manage.
32:302-307.

Turnbull, R. E., F. A. Johnson, and D. H. Brakhage. 1989. Status, distribution, and foods
of Fulvous whistling ducks in south Florida. J. Wildl. Manage. 53:1053-1057.

150 THE WILSON BULLETIN • Vol. JOS, No. I, March 1996

Weller, M. W. 1956. A simple field candler for waterfowl eggs. J. Wildl. Manage. 20:

111-113. . ,

ZwANK, P. J., R M. McKenzie, and E. B. Moser. 1988. Lulvous whistlmg-duck abundance

and habitat use in southwestern Louisiana. Wilson Bull. 100:488^94.

Wilson Bull., 108(1), 1996, pp. 151-153

SHORT COMMUNICATIONS

Do standardized brood counts accurately measure productivity? — A standardized
method for estimating raptor reproductive success and productivity is crucial for making
valid comparisons among years and populations (Steenhof 1987). Steenhof and Kochert
(1982) recommended that reproductive surveys be conducted when nestlings reach 80% of
the average fledging age because nestlings are easily counted (Steenhof 1987), and little
mortality occurs after this age and prior to fledging (Millsap 1981, Steenhof 1987). The
utility of this recommendation would be enhanced if the number of nestlings reaching 80%
of average fledging age was also indicative of a pair’s productivity later in the nesting cycle.
We tested this relationship in Prairie Falcons {Falco mexicanus) by measuring productivity
at 80% of fledging age and at the end of parental care (approximately 35 days later or 27
days after fledging; McFadzen and Marzluff, unpubl. data). This latter measure of produc-
tivity was selected because attributes of habitat and land-use around nesting areas may affect
productivity throughout the nesting period until parental care ceases and fledglings disperse.

In conjunction with a multi-year study of Prairie Falcon productivity (Lehman et al. 1993,
Marzluff et al. 1993) in the Snake River Birds of Prey National Conservation Area in
southwestern Idaho (area described in U.S.D.I. 1979), we knew the number of nestlings
reaching 80% of fledging age and the number subsequently dispersing from the natal ter-
ritory for 58 broods during the 1992 and 1993 breeding seasons. This sample included all
sites we knew failed to produce fledglings (N = 24) and 34 sites where we radio-tagged
nestlings (N = 141) and monitored their survival to dispersal (McFadzen and Marzluff,
unpubl. data).

The number of nestlings attaining 80% of fledging age was significantly correlated with
the number of young dispersing from a territory (Fig. 1; r = 0.83, P < 0.001, N = 58
broods), but the number of young per brood that survived to disperse differed substantially
from the number of nestlings attaining 80% of fledging age. This difference was significant
for broods of four (Paired t = -3.2, 11 df, P = 0.008) and broods of five to six (Paired t
= —3.9, 16 df, P = 0.001) but not for broods of one to three (Paired t = -1.6, 4 df, P =
0.18). The relative inaccuracy of predicting the number of dispersers from the number of
nestlings attaining 80% of fledging age was indicated by the large 95% confidence intervals
associated with the mean difference between these two measures. The mean number of
nestlings attaining 80% of fledging age was expected to be larger than the number of
dispersers by up to 1.47, 3.91 and 4.37 nestlings, respectively for broods of three size classes
(1-3, 4, and 5-6 nestlings; upper 95% confidence limits for difference between two means
using one-sided confidence intervals; Hahn and Meeker 1991).

Variation in the number of young dispersing from broods of a given size required brood
size to differ by at least two nestlings before significant differences in the numbers of
dispersers were observed. The number of young dispersing from broods of one, two, or
three (pooled x ± SE = 1.2 ± 0.51, N = 5) did not differ significantly from the number
dispersing from broods of four (x ± SE = 2.7 ± 0.33, N = 12; Tukey’s pairwise comparison,
P = 0.08). The number dispersing from broods of four was not different from the number
dispersing from broods of five or six (pooled x ± SE = 3.5 ± 0.4, N = 17; Tukey’s pairwise
comparison P = 0.20). Broods of five or six produced more dispersers than broods of one,
two, or three (Tukey’s pairwise comparison, P = 0.001).

The significant correlation between the number of nestlings attaining 80% of fledging age
and the number dispersing verifies that standardized nestling counts are indicative indices
of a pair’s productivity. However, we suggest that caution be applied in interpretations of

151

Number of young dispersing from natal territory

1

THE WILSON BULLETIN • Vol. I OH, No. I, March 1996

Number of young attaining 80% of fledging age

Eig. 1. Relationship between the number of young attaining 80% of fledging age (brood
size) and the number surviving to disperse from the natal temtory 35 days later. Sample
sizes for each brood size are totalled at the top of the figure and listed next to each point
with multiple observations. The least-squares regression line (solid line) and associated 95%
confidence (dotted lines) and prediction intervals (dashed lines) are plotted.

brood counts as a measure of productivity for two reasons. First, brood counts may not
accurately reflect the number of young fledging from the site. From 16% to 50% of young
attaining 80% of fledging age died before they fledged (McFadzen and Marzlutt, unpubl.
data), and many more died before they dispersed (Fig. 1). Second, there was considerable
variation in the number of dispersers produced from larger broods so that prediction mtei vals
were large (Fig. 1) and brood counts needed to differ by at least two nestlings for differences
to remain significant (P < 0.05) until the end of parental care. If a study compared a large
number of broods between treatments (e.g., years, study areas, etc.), the resulting statistically
powerful test could provide misleading conclusions about dilferences in productivity by
showing that brood counts differing by less than two nestlings were significantly different
(P < 0.05).

SHORT COMMUNICATIONS

153

We support the use of standardized measures of productivity but urge researchers to
measure productivity as late in the nesting cycle as possible. Measurements at earlier stages,
such as Steenhof (1987) proposes, can be used to rank pairs in order of their productivity
and to discriminate most successful pairs from failed ones. This may be adequate for studies
designed to survey avian use of an area and to contrast the probability of successful repro-
duction among treatments. However, studies designed to compare brood counts among treat-
ments, understand the demography of a population, or understand the factors that influence
the reproductive success of individuals should not rely on measures of productivity made
early in the nesting cycle because these are unlikely to correlate precisely with the number
of young surviving to later stages.

Acknowledgments. — We thank the Raptor Research and Technical Assistance Center, Na-
tional Biological Service, Boise, Idaho, for providing us with nesting data and field support.
This note is a result of a cooperative research project between the U.S. Bureau of Land
Management and the Idaho Army National Guard. Funding for the work was provided by
the Idaho Army National Guard through the U.S. Army Chemical Research, Development,
and Engineering Center to Greenfalk Consultants, contract # DAA 05-90-C-0135. M. Ko-
chert, K. Steenhof, M. Vekasy, L. Schueck, D. Conner, and K. Beal provided helpful com-
ments on earlier drafts of the manuscript.

LITERATURE CITED

Hahn, G. and W. Meeker. 1991. Statistical intervals. John Wiley and Sons, New York,
New York.

Lehman, R. N., K. Steenhof, M. N. Kochert, and L. B. Carpenter. 1993. Raptor abun-
dance and reproductive success in the Snake River Birds of Prey Area. Pp. 12-39 in
Snake River birds of prey 1993 annual report (K. Steenhof, ed.). U.S. Dept. Interior B.

L. M., Boise, Idaho.

Marzluff, J. M., L. S. Schueck, M. Vekasy, B. A. Kimsey, M. McFadzen, R. R. Town-
send, AND J. O. McKinley. 1993. Influence of military training on the behavior of
raptors in the Snake River birds of Prey Area. Pp. 40-125 in Snake River birds of prey
1993 annual report (K. Steenhof, ed.). U.S. Dept, of Interior B. L. M., Boise, Idaho.
Milsap, B. a. 1981. Distributional status of Falconi formes in west-central Arizona with
notes on ecology, reproductive success, and management. U.S. Dept, of Interior B. L.

M. , Tech. Note 355.

Steenhof, K. 1987. Assessing raptor reproductive success and productivity. Pp. 157-170
in Raptor management techniques manual (B. A. Giron Pendleton, B. A. Milsap, K. W.
Cline, and D. M. Bird, eds.). National Wildlife Federation, Washington, D.C.

and M. N. Kochert. 1982. An evaluation of methods used to estimate raptor

nesting success. J. Wildl. Manage. 46:885-893.

U.S. Department of the Interior. 1979. Snake River birds of prey special research Rep.
Bur. Land Manage., Boise, Idaho.

John M. Marzluff and Mary McFadzen, Greenfalk Consultants. H210 Gantz Ave.. lioi.se.
Idaho 83709. Received 4 April 1995. accepted 3 June 1995.

154

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Wilson Bull., 108(1), 1996, pp. 154-159

Comparative foraging behavior of sympatric Snow Geese, Greater White-fronted
Geese, and Canada Geese during the non-hreeding season.— Interspecific comparisons
of behavior provide a way to organize information for several species that can lead to
hypotheses regarding the functional significance of observed interspecific differences (Clut-
ton-Brock and Harvey 1984). Previous studies of goose time-activity budgets (e.g., Frederick
and Klaas 1982, Giroux and Bedard 1990, Black et al. 1991, Ely 1992) have focused on
single species and collectively were conducted under widely differing environmental con-
ditions. Certain environmental factors are known to affect goose behavioral patterns and
may confound direct interpretation of interspecific comparisons (Table 1). These environ-
mental factors include geographic region, weather, presence of heterospecifics, group size,
habitat and vegetation type, year, season, age, social status, and gender. We are aware of no
studies that have controlled for environmental variation and examined differences in time-
budgets solely as a function of species membership. The objective of this study was to
identify interspecies differences (and similarities) in foraging behavior of geese during the
non-breeding season, while accounting for sources of environmental variation.

Study area and methods.— study was conducted from November 1991 to February
1992 and October 1992 to February 1993 southwest of Houston, Texas. The area, known
as the rice-prairie region of Texas, lies inland from the coastal marshes and extends from
Port Lavaca, Texas, eastward to the Louisiana border (Hobaugh et al. 1989, Gawlik 1994).
We studied Snow Geese {Chen caerulescens). Greater White-fronted Geese {Anser albi-
frons), and Canada Geese {Branta canadensis', small races), the three most abundant species
of geese wintering in the mid-continental United States (Haskins 1993). A fourth species,
the Ross’ Goose (C. rossii), also occurred in the study area but was much less common

than the other three species (Harpole et al., in press).

We selected four groups of agricultural fields as sample sites, two each in Colorado and
Wharton counties. Each site was approximately 2000 ha in size and contained the three
most common types of ground cover (i.e., plowed soil, rice stubble, and annual plants)
(Gawlik 1994). The sequence of visitation to sites was chosen randomly to reduce biases.
Each site was observed from a vehicle driven along a pre-established route starting within
one hour of sunrise, except when postponed by heavy rain, and ending by early to mid
afternoon. Because we were interested in the foraging behavior of geese, we selected for
behavior quantification only those flocks in which >50% of individuals were feeding. To
reduce the chance of missing rare behavior or losing sight of individuals altogether, we
selected only those flocks that provided reasonable visibility (<300 m). Flocks were char-
acterized with regard to their species composition and relative abundances. After a 10-min
settling period, flocks were filmed with a high-resolution 8-mm video camera and telephoto
lens for about 15 min. Video tapes were later analyzed to construct 5-12 mm continuous
time budgets for one focal bird (Altmann 1974) of each species visible in mixed-species
flocks and two focal birds of the same species for single-species flocks. If a focal bird
became obscured during an observation period, the next closest individual of the same
species exhibiting the same behavior was chosen to complete the focal sample. Behavioral
categories included feeding stationary, feeding locomotion, non-feeding locomotion, resting
(head pulled close to body or tucked under wing), comfort (preening and wing stretches),
alert (head up), and aggression.

We quantified interspecies differences in the percentage of time spent on each behavior
for each of the species-pairs (i.e.. Snow and Greater White-fronted, Snow and Canada, and
Greater White-fronted and Canada) with paired /-tests. We assessed individual variation m

Table 1

Factors Affecting Time-Activity Budgets of Geese During the Non-breeding Season

SHORT COMMUNICATIONS

155

o 5 ^

Vi ®

O J 'C
CL ^

I O O
v: il ^

^ H 2'o

3 <0
o N

S-a

X o ^

“eI

1) Q. 00
Q 2 W

C/5 (Va c/5

c/5 c/5 c/5 C/5

O

O

O

(U

c/a

O

o

o

o

c/5

0

0

V

o

<u

O

CO

C

o

CO

0

0

o

«U

CO

0

0

I

V

<u

t/D

o

o

o

CO

0

0

O

c/a

0

o

c/5

o

c/5

0

c/a

0

O

CT

a

CO

0

o

*,E

>

O

a

<D

<U

ci^Q-^OOO

Wcd>cd>>>—
CQC>C>>>>^^C>C
^•‘“CuOOOc30«-0'-

<u

w

cd

-o

B

c

o

C/D 02

00

a\

r<“)

^ 00

c ^ c c c
CO C/D 00 C/D C/D C/D O

00
00
On 00

On

00

o

a x:

5

On — 00 —

T3

C

cc3

O

'u

<U

<u

V-.

O.

Oi)

u.

0)

JO

c

<u

TJ

c

Oi

5=

O

•o

C u

03 <U

'-s

O 3

iH S

CQ O

00

c

ON
05

05

00 O
05 05
— 05

TO

!= TO
^ u<

TO ^

'O TD

CQ S

^ CQ —

o3 c

PQ c/i ca O

CN

05

05

5c

s:

C3

Cj

^ ii
a U
E ^
w

05

05

a

CQ

05

05

C

TO

—4

c

cd

TO

«5

cd

c

cd

Cd

cd

c

s

a

X

S

S

u

1)

W)

"5

n

05

05

CO

o

D

4— •

c

u

:>

o

0

cd

u

jd

o

cd

u

3

u

e

-2

CQ

Q

H

o

<

CQ

CQ

s

E3

Tj-

05

05

o

E

x:

o

C/D

s: statistically significant or used as a classification category; ns: statistically non-significant.
Cloud cover was statistically significant, other weather variables were not.

Some age and gender classes were statistically significant, others were not.

156

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

time budgets within a species by assigning members of smgle-species pairs randomly to
one of two groups for subsequent analyses with paired /-tests as outlined above.

Because our sampling framework provided paired data on individuals foraging in the
same location at the same time, we were able to control for sources of environmental
variation, as identified in other studies (Table 1), such as geographic region, weather, pres-
ence of heterospecifics, group size, habitat and vegetation type, year, and season. We ac-
counted for age differences by selecting only adults as focal birds for Snow Geese and
Greater White-fronted Geese based on plumage differences. We could not, however, distin-
guish between adult and juvenile Canada Geese because both age classes have similar
plumages. Gender was a potential source of variation we could not control because differ-
ences in body size or plumage were not discernable at the distances we viewed the geese.
However, gender does not appear to influence significantly time-activity budgets of Snow
Geese (Lrederick and Klaas 1982, Belanger and Bedard 1992) and we do not believe it had
an effect on our analysis. Linally, our sampling scheme precluded our identifying social
status of individuals. We assume that this potential source of variation was distributed ran-
domly among the individuals we examined.

Results and discussion.— i:\\c Greater White-fronted Goose was the only species we ob-
served in single-species flocks often enough to analyze the degree to which individual
differences in behavior within a species affect time-activity budgets. Time spent on any
behavior by Greater White-fronted Geese differed less than 2% among individuals (all tests,
df = 44, p > 0.05). These data provide quantitative support for the notion that individuals
of the same species within a flock behave similarly and thus are not completely independent
(Gauthier et al. 1988, Giroux and Bedard 1990, Ely 1992). We suggest that individual
differences within age and species classes are not a significant source of variation in time-
activity budgets of wintering geese.

All three species spent most of their time feeding or in alert behavior, with substantially
less time in other behavior (Lig. 1 ). The large amount of time spent feeding was not un-
expected because we restricted our analysis to observations ot flocks engaged primarily in
feeding, and feeding has been reported as the primary activity for non-breeding geese in
other areas (Gauthier et al. 1988, Belanger and Bedard 1992, Ely 1992). Alert behavior is
also a common activity in social birds, and indeed, most explanations of why birds forage
in flocks are based on benefits from group feeding or antipredator behavior (Barnard and
Thompson 1985). Eor many species of geese, the greatest cause of direct mortality is hunting
(Boyd 1957, Owen 1980, Francis et al. 1992). In our study sites, hunting pressure was heavy
and geese were frequently disturbed by nearby shooting; thus we viewed hunters as the
main predator on geese. Another potential predator that frequently disturbed feeding geese
was the Bald Eagle {Haliaeetus leucocephalus). This species is known to prey on geese in
other areas (McWilliams et al. 1994) and it occurred regularly at our sites. However, during
two years of study in which we recorded 122 eagle sightings, we observed only two attempts
by eagles to capture living geese, and neither was successful. We observed coyotes (Canus
latrans) in the same field with feeding geese only three times, out of 22 total coyote sight-
ings, and they did not attempt to capture the geese.

Interspecific comparisons showed that time spent feeding differed among species by less
that 1 1%, time spent resting differed by less than 8%, time spent alert differed by less than
4%, and mher behavior collectively differed by less than 1%. None of these differences was
statistically significant for Snow Geese and Greater White-fronted Geese (all tests, df ^ 56,
P > 0.05) and Canada Geese differed only in the manner in which they fed (Fig. 1 ). Canada
Geese spent more time in feeding locomotion than did Snow Geese (df = 15, F = 0.007)
and Greater White-fronted Geese (df = 28, P = 0.0001) and less time feeding stationary
than did Snow Geese (df = 15, P = 0.059) and Greater White-fronted Geese (df = 28, P

SHORT COMMUNICATIONS

157

P“

<c'

.<5^

S' <c'

/

.<5^

<z»

e»

.<5^

<b

4^

.<r

s®

O

/

#

£

Fig. I. Time spent on various activities by Snow Geese (N = 65), Greater White-fronted
Geese (N = 122), and Canada Geese (N = 38) during the non-breeding season. Bars indicate
the mean ± I SE. Means represent percentages pooled within a species using one focal bird/
species/fiock.

= 0.004). These differences in feeding behavior corresponded to differences in diet and
morphology. Diet analysis of 16 geese collected by DEG in fields of plowed soil and annual
plants within the same study area, showed that subterranean portions of plants made up
27% and 17% of esophageal and proventriculi contents for Snow Geese and Greater White-
fronted Geese respectively, whereas Canada Gee.se contained no subterranean material

158

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

(Gawlik 1994). White-fronted Geese and Snow Geese are larger than Canada Geese and
more similar to each other in bill morphology. Overall, the smaller-bodied Canada Geese
spent a greater proportion of their time walking and feeding on exposed portions of plants,
whereas Snow Geese and Greater White-fronted Geese spent more time feeding m one
location and consuming underground plant parts. Thus, although the type of feeding behav-
ior exhibited by geese was related to morphology and diet, our results suggest that the
overall time devoted to basic daily requirements such as consuming food and avoiding
predation was similar for all members of a flock regardless of species.

Acknowledgments.— Vundmg for this study was provided by the Southeast Texas Wildlife
Eoundation Texas Agricultural Experiment Station, and the U.S. Fish and Wildlife Service
at Attwater’s Prairie Chicken National Wildlife Refuge. DEG also received support m the
form of a Tom Slick Graduate Fellowship. We acknowledge the many landowners who
allowed us access to their property and provided additional insight into the wintering ecology
of geese We thank W. Hobaugh for his comments and suggestions during the study design.
We also thank K. Bildstein, P. DuBowy, M. Clark, E. Klaas, and an anonymous reviewer
for their comments on earlier drafts of this manuscript.

LITERATURE CITED

Altmann, J. 1974. Observational study of behavior; sampling methods. Behaviour 49:227-

267. . . .

Amat, J. a., B. Garcia-Criado, and A. Garcia-Ciudad. 1991. Food, feeding behaviour

and nutritional ecology of wintering Greylag Geese (Anser anser). Ardea 79:271-282.
Barnard, C. J. and D. B. A. Thompson. 1985. Gulls and plovers: the ecology and behav-
iour of mixed-species feeding groups. Columbia Univ. Press, New York, New York.
BfiLANGER, L. AND J. Bedard. 1992. Flock composition and foraging behavior of greater
snow geese {Chen caerulescens atlantica). Can. J. Zool. 70:2410—2415.

Black, J. M. and M. Owen. 1988. Variations in pair bond and agonistic behaviors in
Barnacle Geese on the wintering grounds. Pp. 39-57 in Waterfowl m winter (M. W.
Weller, ed.). Univ. of Minnesota Press, Minneapolis, Minnesota.

and . 1989. Parent-offspring relationships in wintering barnacle geese.

Anim. Behav. 37:187—198.

, C. Deerenberg, and M. Owen. 1991. Foraging behavior and site selection of

barnacle geese {Branta leucopsis) in a traditional and newly colonised spring staging
habitat. Ardea 79:349-358.

^ Black, J. M., C. Carbone, R. L. Wells, and M. Owen. 1992. Foraging dynamics

in goose flocks: the cost of living on the edge. Anim. Behav. 44:41-50.

Boyd, H. 1957. Mortality and fertility of the White-fronted Goose. Bird Study 4:80-93.
Clutton-Brock, T H. and P. H. Harvey. 1984. Comparative approaches to investigating
adaptation. Pp. 7-29 in Behavioural ecology (J. R. Krebs and N. B. Davies, eds.).

Blackwell Scientihc Publications, Oxford, U.K.

Davis, S. E., E. E. Klaas, and K. J. Koehler. 1989. Diurnal time-activity budgets and
habitat use of Lesser Snow Geese Anser caerulescens in the middle Missouri River
valley during winter and spring. Wildfowl 40:45-54.

Ely, C. 1992. Time allocation by greater white-fronted geese: influence of diet, energy

reserves and predation. Condor 94:857-870.

Francis, C. M., M. H. Richards, R. Cooke, and R. F. Rockwell. 1992. Long-term changes
in survival rates of Lesser Snow Geese. Ecology 73:1346-1362.

Frederick, R. B. and E. E. Klaas. 1982. Resource use and behavior of migrating snow

geese. J. Wildl. Manage. 46:601-614.

SHORT COMMUNICATIONS

159

Gauthier, G., Y. Bedard, and J. Bedard. 1988. Habitat use and activity budgets of greater
snow geese in spring. J. Wildl. Manage. 52:191—201.

Gawlik, D. E. 1994. Competition and predation as processes affecting community patterns
of geese. Ph.D. diss., Texas A&M Univ., College Station, Texas.

Giroux, J. F. and J. Bedard. 1990. Activity budgets of greater snow geese in fall. Can. J.
Zool. 68:2700-2702.

Harpole, D. N., D. E. Gawlik, and R. D. Slack. Distribution of Ross’ and snow geese in
Texas. Proceedings of the 48th annual conference of the Association of Fish and Wild-
life Agencies (in press).

Haskins, J. 1993. Analyses of selected mid-winter waterfowl survey data (1955-1993):
region 2 (central flyway portion). U.S. Fish and Wildlife Service, Albuquerque, New
Mexico.

Hobaugh, W. C., C. D. Stutzenbaker, and E. L. Flickenger. 1989. The rice prairies. Pp.
367-383 in Habitat management for migrating and wintering waterfowl in North Amer-
ica (L. M. Smith, R. L. Pederson, and R. M. Kaminski, eds.). Texas Tech Univ. Press,
Lubbock, Texas.

McWilliams, S. R., J. P. Dunn, and D. G. Raveling. 1994. Predator-prey interactions
between eagles and Cackling Canada and Ross’ Geese during winter in California.
Wilson Bull 106:272-288.

Owen, M. 1980. Wild geese of the world. B. T. Batsford, London, U.K.

SCHMUTZ, J. A. 1994. Age, habitat and tide effects on feeding activity of Emperor Geese
during autumn migration. Condor 96:46—51.

Turcotte, Y. and j. Bedard. 1989. Prolonged parental care and foraging of Greater Snow
Goose juveniles. Wilson Bull. 101:500-503.

Ydenberg, H., H. H. Th. Prins, and J. Van Dijk. 1983. The post-roost gatherings of
wintering Barnacle Geese: information centres? Ardea 71:125-131.

Dale E. Gawlik and R. Douglas Slack, Dept, of Wildlife and Fisheries Sciences, Texas
A&M Univ., College Station, Texas 77843-2258. (Present address DEG: Everglades Sys-
tems Research Division, South Florida Water Management District, P.O. Box 24680, West
Palm Beach, Florida 33416-4680.) Received 7 March 1995, accepted 1 Sept. 1995.

Wilson Bull., 108(1), 1996, pp. 159-163

Survival of radio-collared nestling Puerto Rican Parrots. — A remnant population of
the critically endangered Puerto Rican Parrot (Amazona vittata) survives in the Luquillo
Mountains of northeastern Puerto Rico (Snyder et al. 1987). During the last three decades,
intensive research and management have reversed a precipitous population decline to one
of continual population growth (Wiley 1980, Snyder et al. 1987, Lindsey et al. 1989, Lindsey
1992). Prior to Hurricane Hugo in 1989, the wild population had grown from 14 birds
during the mid-1970s to 47 (Meyers 1995). The hurricane reduced the population to about
22-24 birds; however, by early 1994, the population was estimated at 38-39 individuals
(Meyers 1995). Population surveys for late 1994 (post-breeding) and early 1995 were 42
and 33, respectively (F. J. Vilella and F. Nunez, pers. comm.). These data invite optimism
for the full recovery of the population because the species has shown the ability to recuperate
about 83% of its pre-disturbance numbers within five years following a major disturbance.

160

THE WILSON BULLETIN • Vol. 108, No. /, March 1996

Recovery of the species, however, will probably require relocation of wild parrots or releases
of captive-reared parrots in the Luquillo Mountains and other suitable areas of Puerto Rico.

Low survival of free-flying, juvenile Puerto Rican Parrots hinders recovery efforts (Lind-
sey et al. 1994), and telemetry is the only reliable means for determining causes of mortality
during this stage of the life cycle. Radio telemetry was first proposed in the mid-1980s as
a field technique to monitor wild Puerto Rican Parrots. Using Hispaniolan Parrots (/I. ven-
tralis) as surrogates, S. A. Temple and E. Santana C. (pers. comm.) designed one of the
first collar-mounted transmitters intended for future use on Puerto Rican Parrots. Early field
research conducted by J. W. Wiley (Wiley 1983, Wiley et al. 1992) showed that radio
telemetry was not only possible, but imperative, to determine survival rates of wild parrots.
Until recently, telemetry was the only satisfactory method of individually identifying Puerto
Rican Parrots in the field.

Herein, we report the resightings of three radio-collared Puerto Rican Parrots that were
marked during 1985-1987 and present the expected survival for all radio-collared parrots
released during those years. Our objectives are not only to report this information but to
demonstrate that radio telemetry is crucial to our understanding of the Puerto Rican Parrot’s
ecology and indispensable for successful management and conservation.

Studv area and methods. — The extant Puerto Rican Parrot population is confined to the
1 1,200 ha Luquillo Experimental Lorest within the Luquillo Mountains and nearby forests
(max. elev. ca 1000 m) of northeastern Puerto Rico ( 18°19'N, 65°45'W). Parrots nest among
valleys and ridges at elevations of 370-670 m. Rainfall varies from 2000 mm in the foothills
to more than 5000 mm on the highest peaks (Snyder et al. 1987). Lollowing Holdridge’s
(1947) life zone classification scheme, the Luquillo Mountains are classified as subtropical
wet and subtropical rain forest (Ewel and Whitmore 1973). More than 1200 plant species
have been described, of which 240 are trees. Two forest types, tabonuco, characterized by
its predominant species, tabonuco (Dacryodes excelsa), and palo Colorado forest with the
predominant palo Colorado {Cyrilla racemiflora), constitute the dominant vegetative cover
in the parrot nesting areas (Little and Wadsworth 1964, Snyder et al. 1987).

After relatively little disturbance in parrot areas for 50 years, major habitat disturbance
was caused by Hurricane Hugo on 1 8 September 1 989 which damaged much ot the forest
(Walker 1991, Walker et al. 1992). Although the parrot population declined 49%, presumably
caused by the storm, the nesting population recovered within one to two years and produced
six nesting pairs in 1991 and 1992 (Meyers et al. 1993).

Prom 1985 to 1987, 15 nestling parrots were radio-collared and released (Lindsey et al.
1994). In September 1985, three captive-reared parrots were released with radio transmitters
(Snyder et al. 1987). Only one parrot survived the first week after the release. After 17 days,
however, the sole survivor lost its radio 1 .3 km from its release site. This parrot was iden-
tified by its leg band location in the spring of 1986. The second radio-marked panot (No.
553) was hatched from a West Pork nest in 1986 and moved to an East Pork nest as a
hatchling, a distance of about 1.3 km to the SSL. PaiTot 553 successfully fledged on 25
June 1986 with a 6.3-g radio-collar attached (2.3% of its body weight). It was radio-tracked
until 19 December 1986. The third parrot was one of the remaining 13 radio-collared and
released from nests (wild-reared and foster captive-reared) on the eastern and western range
of the Luquillo Mountains (Lindsey et al. 1994). All parrots were also marked with num-
bered stainless steel leg bands before release.

Results. — The first radio-collared parrot, released in 1985, was observed nesting at East
Pork during 1989 and successfully produced one fertile egg that year (M. H. Wilson, pers.
comm.). It was presumed dead after Hurricane Hugo when it failed to return to the nesting
area with its mate.

During the 1991 breeding season, five years post-fledging, panot 553 was observed with

SHORT COMMUNICATIONS

161

Year

Fig. 1. Expected survival of marked Puerto Rican Parrots in the Luquillo Mountains,
Puerto Rico, 1985-1987. Number of parrots radio-marked and released by year; 1985 = 3,

1986 = 4, 1987 = 7. Mortality estimated at 32.5% from fledging to one year of age, 15.2%
for years 2-4, 8.7% per year after year 4 (Snyder et al. 1987, Lacy et al. 1989), and 50%
(determined from surveys) for all parrots during the year of Hurricane Hugo. Survival for

1987 released parrots was calculated from raw data for minimum number alive at six months
after fledging.

its transmitter still intact al an East Mountain nest, 1.1 km NE of its natal site (B. Roberts,
pers. comm.). It was the male of a breeding pair that successfully nested in a natural cavity
of a tabonuco tree (Meyers et al. 1993). The parrot's identity was confirmed when its radio
was shed inside the cavity during the breeding season and later di.scovered during a nest
check (H. Abreu, pers. comm.). During the 1992 breeding season, this pair nested again in

162

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

the same cavity and produced fledglings (Meyers et al. 1993). By early 1993, parrot 553
had disappeared and was presumed dead when the female was seen with a new mate (B.
Roberts, pers. comm.).

The third radio-collared parrot was sighted in the South Fork nesting area on 18 February
1993 by B. Roberts (pers. comm.). Roberts noted a large bulge in the parrot’s neck feathers,
attributing it to an unseen transmitter. Her observations were later verified by other observers
when the parrot (with its radio still attached), replaced the male of the South Fork lA nest-
pair in early June of the same year.

Discussion. — Based on survival estimates for Puerto Rican Parrots (Snyder et al. 1987,
Lacy et al. 1989) and assuming from previous experience no effect on survival due to the
transmitter, we believe that two to three of the 15 parrots marked in 1985-1987 would be
alive in 1993 (Fig. 1). Verifying that at least two of the radio-collared parrots were alive in
that year demonstrates that this type of radio attachment may have little influence on juvenile
survival. Another study of 15 parrots of iour Amazona species in Puerto Rico during 1991-
1993 resulted in no mortalities as a consequence of similarly designed radio-collars that
remained attached for up to 1.8 years (Meyers, unpubl. data, Meyers 1995). One the 15
radio-collared parrots (female) and its unmarked mate successfully fledged two nestlings.
We believe that important information gained by radio telemetry more than compensates
for minor effects that this technique may have on the parrots. Because most parrots spend
considerable time in groups, telemetry provides invaluable life-history and survival infor-
mation.

Most of the radio-collared parrots probably lost their transmitters within 1—2 years. Radio
attachments should be sufficiently secure to retain the transmitter for the predicted battery
life. For two Puerto Rican Parrots in this study, however, the attachment material (heavy
gauge cotton thread) had not deteriorated sufficiently to allow the successful shedding of
the transmitter for almost four years after the battery had expired. Research is underway
using a new radio-collar design and attachment mechanism to test the practicality of low
carbon steel connectors that rust and allow the parrot to shed the collar after the battery
expires (Meyers 1995).

Acknowledgments. — O. H. Pattee, J. H. Rappole, D. H. White, and J. W. Wiley provided
helpful comments for improving the manuscript. The Puerto Rican Parrot Project is a co-
operative program of the Puerto Rico Dept, of Natural and Environmental Resources; U.S.
Dept, of Agriculture, Forest Service; and U.S. Dept, of Interior — Fish and Wildlife Service
and National Biological Service.

LITERATURE CITED

Ewel, j. j. and j. L. Whitmore. 1973. The ecological life zones of Puerto Rico and the
U.S. Virgin Islands. U.S.D.A. For. Serv. Res. Paper ITF-18, Inst. Trop. For., Rio Piedras,
Puerto Rico.

Holdridge, L. R. 1947. Determination of world plant formations from simple climatic data.
Science 105:367—368.

Lacy, R. C., N. R. Flesness, and U. S. Seal. 1989. Puerto Rican Parrot population viability
analysis and recommendations. Captive Breeding Specialist Group, Apple Valley, Min-
nesota.

Lindsey, G. D. 1992. Nest guarding from observation blinds: strategy for improving Puerto
Rican Parrot nest success. J. Field Ornithol. 63:466-472.

, W. J. Arendt, and j. Kalina. 1994. Survival and causes of mortality in juvenile

Puerto Rican Parrots. J. Field Ornithol. 65:76-82.

, M. K. Brock, and M. H. Wilson. 1989. Current status of the Puerto Rican Parrot

conservation program. Pp. 89—99 in Wildlife management in the Caribbean islands.

SHORT COMMUNICATIONS

163

Proc. Fourth Meet. Caribbean Foresters. U.S.D.A. Forest Service, Inst. Trop. For. and
Caribbean Natl. For., Rio Piedras, Puerto Rico.

Little, E. L., Jr. and F. H. Wadsworth. 1964. Common trees of Puerto Rico and the
Virgin Islands (Second printing, 1989). U.S.D.A. For. Serv. Agric. Handbook No. 249,
Washington, D.C.

Meyers, J. M. 1995. Puerto Rican parrots in Our living resources 1994: a report to the
nation on the distribution, abundance and health of U.S. plants, animals, and ecosystems
(E. T. LaRoe III, G. S. Farris, C. E. Puckett, and P. D. Doran, eds.). U.S.D.I. National
Biological Service, Washington, D.C. (in press).

. 1995. Evaluation of three radio transmitters and collar designs for Amazona. Wildl.

Soc. Bull. 23 (in press).

, F. J. ViLELLA, AND W. C. Barrow, Jr. 1993. Positive effects of Hurricane Hugo:

record years for Puerto Rican Parrots nesting in the wild. Endangered Species Tech.
Bull. 28:1,10.

Snyder, N. F. R., J. W. Wiley, and C. B. Kepler. 1987. The parrots of Luquillo: natural
history and conservation of the Puerto Rican Parrot. Western Foundation Vertebrate
Zoology, Los Angeles, California.

Walker, L. R. 1991. Tree damage and recovery from Hurricane Hugo in Luquillo Exper-
imental Forest, Puerto Rico. Biotropica 23:379-385.

, J. VOLTZOW, J. D. Ackerman, D. S. Fernandez, and N. Fetcher. 1992. Immediate

impact of Hurricane Hugo on a Puerto Rican rain forest. Ecology 73:691—694.

Wiley, J. W. 1980. The Puerto Rican Parrot {Amazona vittata): its decline and the program
for its conservation. Pp. 133-159 in Conservation of new world parrots. (R. F. Pasquier,
ed.). Int. Counc. for Bird Preserv. Tech. Publ. No. 1, Smithsonian Inst. Press, Wash-
ington, D.C.

. 1983. The role of captive propagation in Puerto Rican Parrot conservation. Pp.

441-^53 in Jean Delacour/IFCB Symposium on breeding birds in captivity (A. C. Ris-
ser, Jr. and F. S. Todd, eds.). International Foundation for Conservation of Birds, North
Hollywood, California.

, N. F. R. Snyder, and R. S. Gnam. 1992. Reintroduction as a conservation strategy

in parrot conservation. Pp. 165— 2(X) in New World parrots in crisis: solutions from
conservation biology (S. R. Beissinger and N. F. R. Snyder, eds.). Smithsonian Inst.
Press, Washington, D.C.

J. Michael Meyers, National Biological Service, Patuxent Environmental Science Center,
P.O. Box N, Highway 191 Km 4.4, Palmer, Puerto Rico 00721-0501 (Present address:
National Biological Service, Southeastern Biological Science Center, Southeast Research
Station, D. B. Wamell School of Forest Resources, The Univ. of Georgia, Athens, Georgia
30602-2 1 52)-, Wayne J. Arendt, U.S. Forest Service, Internationa! Institute of Tropica!
Forestry, Call Box 25000, Rio Piedras, Puerto Rico 00928-2500', and Gerald D. Lindsey,
National Biological Service, Patuxent Environmental Science Center, Laurel, Maryland
20708. (Present address: National Biological Service. Hawaii Research Station, P.O. Box
44, Hawaii National Park, Hawaii 96718). Received 28 March 1995, accepted 1 Sept. 1995.

164

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Wilson Bull., 108(1), 1996, pp. 164-166

New nesting area of Puerto Rican Parrots. — The Puerto Rican Parrot (Amazona vittata)
population has been isolated in a 1 1,300-ha tract of the Luquillo Mountains of northeastern
Puerto Rico for ca 80 years and has been using only a small area of the forest for nesting
since the 1950s. During the last 50 years, it was known to have nested only in cavities of
old growth palo Colorado (Cyrilla racemiflora) with one exception (Snyder et al. 1987). In
1974, an occupied cavity was found in a laurel sabino (Magnolia splendens). Three to six
nests have been monitored annually for the last 20 years, often in the same cavity trees for
up to 11 years (Snyder et al. 1987, Meyers et al. 1993). The almost exclusive use of palo
Colorado trees for nesting has been caused by the scarcity of cavities in other tree species
(Snyder et al. 1987). A recent study (Meyers and Barrow, unpubl. report) indicates, however,
that potential palo Colorado nesting trees have declined up to 72% since 1976 because of
lack of natural recruitment. Only 7% of palo Colorado stems were <30 cm diameter breast
height (dbh), which indicates that further declines are expected in the future.

Snyder et al. (1987) conducted interviews that revealed that earlier in this century, the
Puerto Rican Parrot occasionally nested in tabonuco (Dacryodes excelsa) and caimitillo
(Micropholis chrysophylloides) cavities in the Luquillo Mountains. In northwestern Puerto
Rico, Rio Abajo residents reported that parrots commonly nested in corcho (Pisonia suh-
cordata), aguacate (Persea americana), jacana (Poiiteria multiflora), Puerto Rico royalpalm
(Roystonea borinquena), and pot holes in limestone cliffs. Gundlach (1878, cited in Snyder
et al. 1987) also reported that parrots nested in palms in the coastal plain near Quebradillas.
Clearly, the Puerto Rican Parrot used a diversity of nesting sites in the past and might do
so again in the future if nesting sites were available. I report here the habitat of a new
nesting area of the Puerto Rican Parrot and suggest procedures that may encourage further
expansion and changes in nesting habits of this critically endangered species.

Methods. — I located parrot activity sites during evening and morning surveys from a tree
platform (20 m high) and by plotting the location of breeding pairs on maps with a compass
and rangefinder. Habitat data were collected from 0.05-ha and 0.20-ha concentric, circular
plots centered at the nesting tree (nesting site) and at activity sites of the pair in the nesting
area. Data collected in the 0.05-ha plot were (1) elevation (m), (2) slope (%), (3) aspect
(compass direction), (4) canopy height (m) measured by a rangefinder or Haga tree altimeter,
and (5) vegetative strata. Midstory cover from 5-12 m high and canopy cover >12 m high
were estimated by three persons using a scale of 1 to 4 (1 ^ 25%, 2 = 26—50%, 3 = 51 —
75%, 4 > 76%). Small trees, >2 m high and <30 cm dbh, were identified to species and
counted. Large trees, >30 cm dbh, were identified to speeies and measured for dbh and
height (m). Cavity openings were categorized by diameter classes: ^5.0 cm, 5.1—20.0 cm,
20.1-35.0 cm, 35.1-50.0 cm, >50 cm. Potential parrot ne.st trees (>49 cm dbh, see Snyder
et al. 1987) were tallied, identified as to species, and measured for dbh and height in 0.2-
ha plots.

Results. — W. Abreu (O. Carrasquillo, pers. comm.) found the new nesting cavity in 1991.
These parrots nested successfully and fledged young in 1991 and 1992 at East Mountain
(Meyers et al. 1993), but the male was presumed dead in 1993 when its mate was seen at
the cavity with another male (B. Roberts, pers. comm.). East Mountain is 1.1 km northeast
of East Pork, a traditional parrot nesting area at higher elevations (500 m) in the palo
Colorado forest of the Luquillo Mountains. The newly discovered nesting cavity was in a
21-m high tabonuco tree (73 cm dbh) at an elevation of 370 m in tabonuco forest. Steep
northwest slopes (55%) were prominent at the nesting site. Seven of eight trees >49 cm
dbh and surrounding the cavity tree (().2-ha plot) were potential nest trees. The nest cavity
opening was 35-50 cm. Two smaller cavities openings (5—20 cm) were found in palo col-

SHORT COMMUNICATIONS

165

orado trees nearby. Density of trees >30 cm dbh was 220 stems/ha, of which 35 stems were
potential nest trees. Forest canopy and midstory cover in the area surrounding the nest tree
was sparse to moderate (25-50% cover). At the nesting site, many colonizing trees (1640
stems/ha), such as trumpet-tree {Cecropia sheberiana), had recently sprouted (<6 cm dbh)
after passage of Hurricane Hugo.

Habitat used by the breeding pair in the nesting area (activity sites, N = 3) was within
150 m of the nest and was oriented towards the northwest with slopes of 8^4%. Fewer
potential nesting trees were found at these sites (12 fewer trees/ha). Density of trees >30
cm dbh was 100 stems/ha, of which 21 stems were potential nest trees. Forest canopy was
broken and sparse (25-50% cover) at heights of 17.4-26.4 m. Palicourea riparia (626 stems/
ha), Psychotria berteriana (373 stems/ha), and Tetrazygia urbanii (387 stems/ha), all small
trees or shrubs (<6 cm dbh), were the predominant cover in the understory where few
trumpet-trees were found (347 stems/ha).

Discussion. — These were the first Puerto Rican Parrots reported nesting in a natural cavity
of a tabonuco tree since intensive research began in the 1950s (Rodriguez- Vidal 1959,
Snyder et al. 1987). Old-growth tabonuco forest at East Mountain was selected for nesting
by this pair in contrast with old growth palo Colorado forest used in the past (Snyder et al.
1987, Meyers 1994). Few large trees (25 stems/ha >60 cm dbh) surrounded the nesting tree,
which was considerably less than reported for the nearby East Fork nesting area (73 stems/
ha >60 cm dbh) for palo Colorado forest by Snyder et al. (1987) before Huiricane Hugo.
The hurricane, however, reduced the number of potential nesting trees (>49 cm dbh) at East
Fork to 23 stems/ha (Meyers and W. C. Barrow, Jr, unpubl. report). Preserving old-growth
tabonuco forest in the Luquillo and other mountains of Puerto Rico may be important for
the recovery of the species.

The high density of colonizers (e.g., trumpet-tree) at the nest site used by parrots at East
Mountain was the consequence of extensive canopy openings caused by Hurricane Hugo.
Parrot nesting areas at East Mountain and East Fork received significantly more hurricane
damage than nesting areas on western slopes, where young trumpet-trees were almost non-
existent (0—20 stems/ha) in nesting plots (Meyers and Barrow, unpubl. report). This change
in habitat, without the loss of large nest-trees, may have little effect on the parrot. Habitat
damage, even from a severe storm, may not be detrimental to parrots and may actually
stimulate the population growth rate (Meyers et al. 1993). To understand this effect, however,
forest habitat simulation models may prove valuable in predicting potential effects of hur-
ricanes and availability of potential nest-trees and food.

Although 47 cavities in palo Colorado trees were enhanced for Puerto Rican Panots in
1990, only one has been used in five years (E. Garcia, pers. comm.). The site selection for
enhancing cavities was based on parrot activity in the area which was determined by qual-
itative surveys. A recent study, however, revealed that parrot activity sites, based on quan-
titative surveys, are different from nesting sites for an important habitat characteristic. Nest-
ing sites have 16—20 more potential nesting tree.s/ha than activity sites of panots (Meyers
and Barrow, unpubl. report). It may be beneficial to create clusters of cavities in potential
nesting trees (palo Colorado, tabonuco, caimitillo, and laurel sabino) in habitat similar to
that used by the parrot for nesting, i.e. with a high density of potential nesting trees.

Acknowledgments. — Steven C. Latta, Eugene P. Odum, Paul W. Sykes, Jr., and two anon-
ymous reviewers provided helpful comments for revising the manuscript. Wylie C. Banow,
Jr., Rafael de Leon, Jennifer D. Horn, and Keith L. Pardieck collected parrot activity and
habitat data. The Puerto Rican Parrot Project is a cooperative program of the Puerto Rico
Dept, of Natural and Environmental Resources; U.S. Dept, of Agriculture, Forest Service;
and U.S. Dept, of Interior — Fish and Wildlife Service and National Biological Service.

166

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

LITERATURE CITED

Meyers, J. M. 1994. Old growth forests and the Puerto Rican Parrot. Endangered Species
Tech. Bull. 29:12.

, E J. ViLELLA, AND W. C. Barrow, Jr. 1993. Positive effects of Hurricane Hugo:

record years for Puerto Rican Parrots nesting in the wild. Endangered Species Tech.
Bull. 28:1-10.

Rodriguez- Vidal, J. A. 1959. Puerto Rican Parrot {Amazona vittata vittata) study. Com-
monwealth of Puerto Rico, Dept, of Agric. and Commerce, San Juan, Puerto Rico.

Snyder, N. F. R., J. W. Wiley, and C. B. Kepler. 1987. The parrots of Luquillo: natural
history and conservation of the Puerto Rican Parrot. West. Found. Vertebr. Zool., Los
Angeles, California.

J. Michael Meyers, National Biological Service, Patuxent Environmental Science Center,
P.O. Box N, Highway 191 Km 4.4, Palmer, Puerto Rico 00721-0501 . (Present address:
National Biological Service, Southeastern Biological Science Center, D. B. Warnell School
of Forest Resources, The Univ. of Georgia, Athens, Georgia 30602-2152.) Received 17 May
1995, accepted 19 Sept. 1995.

Wilson Bull, 108(1), 1996, pp. 166-170

Neotropical migrants in marginal habitats on a Guatemalan cattle ranch. — Recent
studies of migratory birds overwintering in Central America and the Caribbean have focused
on bird communities in particular types of disturbed habitats, such as citrus orchards (Rogers
et al. 1982, Mills and Rogers 1992) or agricultural fields in varying stages of succession
after human abandonment (Waide 1980, Kricher and Davis 1989), while others have at-
tempted to discern broader patterns of species occurrence across a wide variety of habitat
types (Waide et al. 1980, Leek 1985, Lynch 1989, Robbins et al. 1992, Wunderle and Waide
1995). Although a few in the latter category have included a small amount of data from
cattle ranches. Central American cattle ranches have received little attention in the ornitho-
logical literature. This is unfortunate because conversion to cattle ranching is the single
largest threat to the remaining undisturbed lands in Central America (Myers 1980, Busch-
bacher 1986, Lynch 1989). Given the amount of land already used for cattle ranching in
Central America and the amount likely to be converted in the near future, knowledge of
patterns of species occurrence on land modified for cattle ranching is critical for formulating
future conservation strategies.

We mist netted birds on an active cattle ranch in the Pacific lowlands of Guatemala to
investigate the extent to which narrow riparian corridors and other marginal habitat set in
a matrix of open cattle pasture serve as usable habitat for overwintering migratory birds.

Study area and methods. — We conducted the study from 2 February to 2 March 1995 on
Finca Caobanal, a working cattle ranch in the Pacific lowlands of Guatemala. The region is
characterized by relatively flat topography with elevation ranging from sea level to approx-
imately 200 m. The average annual temperature is 25°C, and annual rainfall averages 200
cm, with a pronounced dry season between November and April (Universidad Rafael Lan-
dfvar 1987). Although the native vegetation type is subtropical humid forest, the vast ma-
jority of the region has been converted to agricultural land, particularly cattle pasture, and
more recently, sugar cane.

SHORT COMMUNICATIONS

167

Finca Caobanal comprises approximately 1000 ha situated along the Maria Linda River,
about 30 km southeast of the town of Escuintla. The ranch is a mosaic of open pastures
separated by hedgerows and artificial canals lined with narrow corridors of secondary growth
vegetation and one 250-ha parcel of secondary growth forest. The riparian corridors range
in width from approximately 10-80 m; lower and mid-story vegetation is generally dense
and is dominated by tall grasses, Heliconia sp.. Acacia hindsii, Ricinus communis and Piper
sp. Isolated overstory trees include Pithecolobium saman, Bombax ellipticum. Acacia
hindsii, Enterolobium cyclocarpum, Salix sp. and Ficus sp. In the secondary growth forest,
dominant tree species include Terminalia oblonga, Cedrela odorata, Ficus sp., Triplaris
malaenodedron and Cecropia sp., and the average canopy height is 15-20 m. The understory
is relatively open, but in some places is dominated by large patches of Heliconia sp.

Although previous studies that surveyed avian communities on cattle ranches in Central
America (Saab and Petit 1994, Robbins et al. 1992, Lynch 1989) focused on open pasture,
deliberately avoiding edges and hedgerows, we focused our mist netting efforts in the veg-
etation on the margins of actively grazed cattle pastures, in the riparian corridors along the
irrigation canals, and in the 250 hectare forest. Two to seven black nylon 30 mm mesh nets
(depending on habitat shape and size) were operated daily, generally in a line with approx-
imately 1 2m between each net, and one or two nets situated perpendicular to the others. In
order to avoid heat stress on the birds, nets were checked every 15 min and were closed
between approximately 10:30 and 15:00. Nets were moved every three days to lessen the
problem of net shyness and were open for a total of 446 net-h. Birds were marked with
indelible ink underneath the wing in order to identify recaptures.

Mist netting is strongly biased against species that spend most or all of their time in the
canopy, and this investigation should therefore be viewed as a study of primarily understory
birds. Additionally, using mist net data to infer absolute species abundance directly is prob-
lematic because some species have a higher probability of capture than others; nonetheless,
when interpreted cautiously such data can still serve as a meaningful index of relative
species abundance (Karr 1981).

Results. — We caught 258 individual birds of 45 different species; 49% of individual birds
were long-distance migrants, and 51% were residents (Table 1). This approximately 1:1 ratio
of overwintering migrants to residents is consistent with other published netting studies
conducted in secondary growth habitats in Central America and the Caribbean (Lynch 1989,
Waide 1980).

The five migratory songbird species most frequently captured were Yellow Warbler, Den-
droica petechia (N = 22), Common Yellowthroat, Geothlypis trichas (N = 16), Northern
Waterthrush, Seiurus noveboracensis (N = 10), Yellow-breasted Chat, Jcteria virens (N =
8), and Painted Bunting, Passerina ciris (N = 15). The first four species are widely reported
to be abundant on disturbed and/or agricultural lands in Central America (Rogers et al.
1982, Kricher and Davis 1992, Petit et al. 1992, Robbins et al. 1992, Lynch 1992, Mills
and Rogers 1992), but the Painted Bunting is less well-documented as an inhabitant of
severely human-modified environments in Central America.

Capture rates of birds in the riparian corridors were strikingly different from those in the
secondary growth forest; working in the corridors we caught an average of 76 birds per 100
net-h, compared to only 20 birds per 100 net-h in the forest (x^ = 30.5, P < 0.01). The
low capture rate in the forest may have resulted partly from the fact that the average forest
canopy height was considerably greater than in the corridors, and consequently canopy-
foraging birds may never have flown low enough in the forest to be captured. We do not
believe this entirely explains the disparity, however, as casual observations also indicated a
greater density of birds in the riparian corridors. Although casual observations may have
been biased by a lower probability of detection of birds in the forest canopy, the corrobo-

168

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Table 1

Bird Species Netted at Linca Caobanal

Species

Number

capiured

Migratory

status

Ruddy Ground-Dove {Columbina talpacoti)

3

R

Common Ground-Dove (C passerina)

1

R

Groove-billed Ani (Crotophaga sulcirostris)

1

R

Cinnamon Hummingbird {Amazila rutila)

28

R

Rubv-throated Hummingbird {Archilochus colubris)

1

M

Amazon Kingfisher {Cholorceryle amazona)

1

R

Smoky-brown Woodpecker {Venilornis fumigatus)

1

R

Ivory-billed Woodcreeper {Xiphorynchus flavigaster)

2

R

Barred Antshrike (Thamnophilus doliatus)

1

R

Rufous-breasted Spinetail {Synallaxis erythrothorax)

8

R

Rose-throated Becard {Pachyramphus aglaiae)

2

M

Dusky-capped Llycatcher {Myiarchus tuberculifer)

2

M

Unknown Llycatcher (Empidonax sp.)

1

R

Least Llycatcher (E. minimus)

1 1

7

Yellow-olive Llycatcher (Tolmomyias sulphurescens)

1

R

White-throated Spadebill (Flatyrinchus mystaceus)

1

R

Unknown Pewee (Contopus sp.)

7

Common Tody-flycatcher (Todirostrum cinereum)

4

R

Mangrove Swallow {Tachycineta albilinea)

1

R

Barn Swallow (Hirundo rustica)

17

M

Northern Rough-winged Swallow {Stelgidopteryx serripennis)

3

M

Rufous-naped Wren {Camplylorhynchus rufinucha)

4

R

Swainson’s Thrush (Catharus uslulatus)

4

M

Clay-colored Robin {Turdus greyi)

25

R

Bell’s Vireo {Vireo bellii)

]

M

Tennessee Warbler {Vermivora peregrina)

1

M

Black-and-White Warbler (Mniotilta varia)

1

M

Magnolia Warbler {Dendroica magnolia)

M

Yellow Warbler {D. petechia)

M

MacGillivray’s Warbler (Oporornis tolmiei)

2

M

Kentucky Warbler {O. formosus)

1

M

Hooded Warbler {Wilsonia citrina)

1

M

Worm-eating Warbler (Helmithers vermivorus)

1

M

Ovenbird (Seiurus aurocapillus)

3

M

Northern Waterthrush (S. novehoracensis)

10

M

Common Yellowthroat (Geothlypis trichas)

16

M

Yellow-breasted Chat {Icteria virens)

8

M

American Redstart (Setuphaga ruticilla)

1

M

Northern Oriole (Icterus galbula)

1

M

Blue-gray Tanager (Thraupis episcopus)

4

R

Painted Bunting (Passerina c/m)

15

M

Indigo Bunting (P. cyanea)

4

M

White-collared Seedeater (Sporophila torqueola)

28

R

Blue-black Grassquit (Volatina Jacarina)

9

R

Grayish Saltator (Saltator coerulescens)

1

R

SHORT COMMUNICATIONS

169

ration of the mist-net data is highly suggestive that there is indeed a greater density of birds
in the riparian corridors than in the secondary forest patch.

Discussion. — Our results suggest that lands already under use for cattle ranching should
not simply be written off as lost by conservation advocates; riparian vegetation corridors
and other marginal habitat patches on cattle ranches can be valuable habitat refuges for a
large variety of migrant and resident species. Although large-scale clearing of forest for
cattle ranching is surely disastrous to many birds, as well as other animal and plant species,
in an area like the Pacific lowlands of Guatemala, where virtually all of the forest has
already been cleared (Universidad Rafael Landivar 1987), there may still exist valuable
conservation opportunities.

In the region where we conducted our study, there exists a wide variety of land manage-
ment strategies on different cattle ranches and farms; some allow relatively lush corridors
of vegetation and overstory trees to grow along irrigation canals and fence lines, while
others bulldoze virtually all vegetation. These differing practices surely have an enormous
impact on bird communities. Even at Finca Caobanal, riparian corridors are massively cut
back once or twice a year; the seasonal timing of this event may profoundly affect the
suitability of the area for overwintering migrants or the nesting success of residents.

Conservation advocates in North America as well as Central America must concern them-
selves not only with securing the last, isolated parcels of undisturbed habitat for preservation,
but just as importantly, with influencing land-management practices on lands that have
already been modified by humans. Future research should provide a basis for prescribing
beneficial management practices, so that the harm done in clearing forests for cattle ranching
or agriculture is mitigated to the greatest extent possible.

Acknowledgments. — We thank Francois and Nini Berger and FUNDAVES for hosting us
at Finca Caobanal. Ann T. Brice of the Psittacine Research Project at U.C. Davis provided
financial and logistical support, and W. W. Weathers provided helpful criticism on the manu-
script. This project was supported in part by U.S.A.I.D. Science and Technology Grant HRN-
5600-G-00-2I6-00 to ATB.

LITERATURE CITED

Buschbacher, R. J. 1986. Tropical deforestation and pasture development. Bioscience 36:
22-28.

Karr, J. R. 1981. Surveying birds with mist nets. Pp. 62-77 in Estimating numbers of
terrestrial birds (C. J. Ralph and J. M. Scott, eds.). Allen Press, Inc., Lawrence, Kansas.
Kricher, j. C. and W. E. Davis, Jr. 1992. Patterns of avian species richness in disturbed
and undisturbed habitats in Belize. Pp. 240-246 in Ecology and conservation of Neo-
tropical migrant landbirds (J. Hagen and D. Johnston, eds.). Smithsonian Institution
Press, Washington, D.C.

Leck, C. F. 1985. The use of disturbed habitats by North American birds wintering in
Mexico. Biotropica 17:263-264.

Lynch, J. F. 1989. Distribution of overwintering nearctic migrants in the Yucatan Peninsula,
I: General patterns of occurrence. Condor 91:515-544.

. 1992. Distribution of overwintering nearctic migrants in the Yucatan Peninsula,

II: use of native and human-modified vegetation. Pp. 178-195 in Ecology and conser-
vation of Neotropical migrant landbirds (J. Hagen and D. Johnston, eds.). Smithsonian
Institution Press, Washington, D.C.

Mills, E. D. and D. T. Rogers, Jr. 1992. Ratios of neotropical migrant and neotropical
resident birds in winter in a citrus plantation in central Belize. J. Field Ornith. 63:109-
116.

Myers, N. 1980. Conversion of tropical moist forests. Natl. Acad. Sci., Washington, D.C.

170

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

Petit, D. R., L. J. Petit, and K. G. Smith. 1992. Habitat associations of migratory birds
overwintering in Belize, Central America. Pp. 247-255 in Ecology and conservation
of Neotropical migrant landbirds (J. Hagen and D. Johnston, eds.). Smithsonian Insti-
tution Press, Washington, D.C.

Robbins, C. S., B. A. Dowell, D. K. Dawson, J. A. Col6n, R. Estrada, A. Sutton, R.
Sutton, and D. Weyer. 1992. Comparison of neotropical migrant landbird populations
wintering in tropical forest, isolated forest fragments, and agricultural habitats. Pp. 207-
220 in Ecology and conservation of Neotropical migrant landbirds (J. Hagen and D.
Johnston, eds.). Smithsonian Institution Press, Washington. D.C.

Rogers, D. T, Jr., D. L. Hicks, E. W. Wischusen, and J. R. Parrish. 1982. Repeats,
returns, and estimated flight ranges of some North American migrants in Guatemala. J.
Field Ornithol. 53:133—138.

Saab, V. A. and D. R. Petit. 1994. Impact of pasture development on winter bird com-
munities in Belize, Central America. Condor 94:66-71.

Universidad Rafael LandIvar. 1987. Perfil Ambiental de la Republica de Guatemala.
Guatemala City: Universidad Rafael Landivar.

Waide, R. B. 1980. Resource partitioning between migrant and resident birds: the use of
irregular resources. Pp. 337—352 in Migrant birds in the Neotropics: ecology, behavior,
distribution and conservation (A. Keast and E. S. Morton, eds.). Smithsonian Institution
Press, Washington, D.C.

, J. T. Emlen, and E. j. Tramer. 1980. Distribution of migrant birds in the Yucatan

Peninsula: A survey. Pp. 165-171 in Migrant birds in the Neotropics: ecology, behavior,
distribution and conservation (A. Keast and E. S. Morton, eds.). Smithsonian Institution
Press, Washington, D.C.

WUNDERLE, J. M., Jr. and R. B. Waide. 1995. Distribution of overwintering nearctic mi-
grants in the Bahamas and Greater Antilles. Condor 95:904—933.

Rodney B. Siegel, Dept, of Avian Sciences, Univ. of California, Davis, California 95616;
and Marco V. Centeno, 2da Calle 21-31, Zona 15, Vista Hermosa 1, Guatemala Ciudad,
Guatemala. Received 18 May, 1995, accepted 1 Oct. 1996.

Wilson Bull., 108(1), 1996, pp. 170-175

Ungulate ectoparasite removal by Black Caracaras and Pale-winged Trumpeters in
Amazonian forests. — Interspecific interactions in which an organism eats the ectoparasites
of another, usually larger, organism (sometimes referred to as “cleaning symbioses”; Wit-
tenberger 1981) comprise a relatively common form of mutualism. In terrestrial vertebrates,
these associations are chiefly represented by a few bird species which routinely remove ticks
and hematophagous diptera from large mammals. Such interactions, however, are by no
means regularly distributed across different macrohabitats and appear to be more common
in tropical savannas where recent radiations of large herbivores and their parasites are most
impressive. Cleaning mutuali.sms are thus perhaps best illustrated by certain savanna bird
species of sub-Saharan Africa such as Yellow-billed (Buphagus africanu.s) and Red-billed
oxpeckers (B. erythrorhvnchus) which are highly specialized in plucking ticks from a wide
range of wild and domestic ungulate hosts for the mainstay of their diet (Attwell 1966,
Bezuidenhout and Stutterheim 1980, Hart et al. 1990). In other open habitats, similar inter-
actions also occur less frequently, for example, between Fan-tailed Ravens {Corvus rhipi-

SHORT COMMUNICATIONS

171

durus) and camels {Camelus dromedarius) (Lewis 1989), Pale-winged Starlings (Onychog-
nathus nabouroup) and mountain zebras {Equus zebra) (Penzorn and Horak 1989), Black-
billed Magpies (Pica pica) and moose (Alces alces) (Samuel and Welsh 1991), and Yellow-
bellied Bulbuls {Alophoixus phaeocephalus), and klipspringers (Oreotragus oreotragus)
(Roberts 1993). Cattle Egrets {Bubulcus ibis) are perhaps the best known case of open-
habitat cleaning mutualists in the New World (Burger and Gochfeld 1982), but here domestic
bovid herds have largely replaced the aborigine megaherbivores with which this species was
formerly associated.

In comparison, documented cases of avian species conducting similar lifestyles under
close-canopy tropical forests are apparently rare. It remains to be seen, however, whether
this should be attributed to the fact that (1) these habitats support far fewer species and
lower densities (or smaller surface area) of large-bodied terrestrial herbivores, or (2) far less
is known about the behavior and interspecific associations of forest vertebrates. Here I report
two distinct cases of previously undocumented bird-ungulate mutualistic interactions ob-
served at different Amazonian forest sites. The bird species involved in these interactions
are the Black Caracara {Daptrius ater) and Pale-winged Trumpeter {Psophia leucoptera)
which were observed providing “cleaning” services to Brazilian tapirs (Tapirus terrestris)
and gray brocket deer {Mazama gouazoubira), respectively. Moreover, despite the disparate
ecological and phylogenetic differences between these bird species, it is suggested that these
interactions may be consistent throughout much of the Amazon basin.

Black Caracaras are small (330^45 g) raptors more closely related to polyborine falcons
rather than to true falcons and hawks (Griffiths 1994). The Black Caracara is widely dis-
tributed throughout the Amazon basin from the eastern slopes of the Andes east to Ma-
ranhao, north to the Guianas, and south to the woodland fringes of northern Mato Grosso,
Brazil. In Brazilian Amazonia, this species is more commonly found along rivers, forest
edges, and associated habitats than in vast undisturbed areas of unflooded (terra firme) forest
interior far removed from large rivers (Peres and Whittaker 1991, unpubl. data). Small family
flocks of 3—5 individuals are often seen along rivers where they scavange primarily on small
carrion, prey on arthropods and nestling birds, and occasionally feed on ripe fruit pulp
(Brown and Amadon 1968, pers. obs.). Daptrius ater, therefore, diverges from its widely
sympatric congener, the Red-throated Caracara (D. americanus), which is more social, far
more frugivorous, but also specializes on raiding collonial wasp and bee nests (Thiollay
1991, pers. obs.).

Black Caracaras were observed cleaning a tapir on 12 September 1993 (late dry season)
along a small stream draining a “cacaia” (i.e., water-logged forest patch where trees undergo
a sudden die-off following a change in stream channel) on the right bank of the upper
Tarauaca river, western Acre, Brazil (9°23'S, 71°54'W). A field assistant and I saw a group
of eight Black Caracaras, ca 25 m from us, immediately next to an adult male tapir. An
additional 10 or more caracaras were perched in the understory nearby, apparently awaiting
their turn to descend upon the tapir. The tapir partly exposed its ventral parts and one of its
inner hind-thighs, while lying on its side in a rather relaxed posture. The single juvenile
and several adult caracaras (differentiated by the lemon-yellow and bright-orange facial skin,
respectively) were actively searching for and plucking ticks attached to the large tapir which
was later determined to be a bull. Four of these caracaras were searching for and grooming
ticks while perched directly on top of the tapir. Another bird meticulously inspected the
ventral side of the tapir between its front- and hind-quarters, whereas a third caracara partly
circled around the back and dorsal flanks of the tapir, also examining its body surface. We
were unable to observe whether the other caracaras were also foraging in this manner, for
they were obstructed from view on the opposite side of the tapir. Upon detecting our pres-

172

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

ence 4-5 min later, all caracaras suddenly retreated to nearby perches, thus startling the tapir
which immediately got up and fled from us through Maranthaceae-dominated undergrowth.

Interviews with subsistence hunters of three local communities of different parts of the
Brazilian Amazon — Kaxinawa Indians of the Rio Jordao Reserve in western Acre, Kayapo
Indians of A’Ukre, southeastern Para, and caboclo (non-tribal) hunters of Vila Moura, upper
Tefe river, central Amazonas — revealed that the associations between Black Caracaras and
tapirs witnessed in the upper Tarauaca also occurred at these forest sites. Cleaning sessions
in the upper Xingu basin (Rio Riozinho), as reported by Kayapo warriors, usually involve
one adult tapir and 3-4 D. ater, and may last several hours while tapirs “roll over on both
sides to facilitate tick removal from all body parts”. Most intriguingly, experienced Kayapo
hunters reported that tapirs and Black Caracaras typically are able to approach one another
through a series of vocal exchanges (whistles in the case of tapirs; rasping calls in the case
of D. ater), which apparently helps to coordinate their encounters within a closed-canopy
habitat where visibility is inherently poor. Both Black Caracaras and tapirs, however, are
reported to discontinue counter-calling and to remain very quiet once they eventually get to
within sight of one another and during actual cleaning sessions. Indeed, hunters well aware
of such rallying calls have reported cueing onto vocalizations of Black Caracaras suspected
of “searching for tapirs” in order to home in on a potentially easy kill, for tapirs are the
largest-bodied and often most preferred game species of native Amazonians. Should the
occurrence of such mutual “approach calls” be confirmed, they would suggest that such
cleaning mutualism between Black Caracaras and tapir is relatively stable and has a long
history. Interestingly, both Black Caracaras and tapirs are more commonly found in river-
edge, backwater palm swamps, and river- or stream-disturbed habitats, rather than in high
terra firme forest on well drained soils (e.g., Bodmer 1990, Peres in press). Lurthermore,
Black Caracaras have not been reported to pick ticks from any other Amazonian ungulate,
which perhaps suggests a high degree of specificity in these associations.

The second set of observations involves a group of trumpeters (Psophiidae), which are
highly social, large-bodied (ca 1200 g), terrestrial frugivore-insectivore birds foraging almost
entirely on the forest leaf-litter. The three recognized species of trumpeters White-winged
Trumpeter (P. leiicoptera). Green-winged Trumpeter {P. viridis), and Gray-winged Trum-
peter (P. crepitans) are of widespread occurrence in Amazonian and the Guianan Shield
forests, and comprise the only representatives of the entire family Psophiidae. Closed mem-
bership monospecihc groups of 4—10 individuals maintain relatively large, stable territories
which are actively defended against neighboring groups (Sherman 1991). I made these
observations at a remote terra firme forest 4 km inland from the headwaters of the Urucu
river, Amazonas, Brazil (4°50'S, 65°16'W). Trumpeter group size at this site averaged 6.2
individuals (N = 12). On the morning of 26 September 1988 (late dry season), as I sat
under a stationary group of woolly monkeys {Lagothrix lagotricha cana), I observed an
approaching group of seven White-winged Trumpeters walking slowly alongside an adult
male Gray Brocket Deer (Mazama gouazouhira). Pour of those trumpeters immediately next
to the deer were clearly gleaning over specific parts of its pelage surface from ground level
while striving to maintain hxed positions relative to the front- or hind-quarters of this cervid.
Horse Hies (Diptera: Tabanidae) of 15-25 mm swarming around the brocket deer were being
rapidly snatched by the trumpeters through precise pecking and neck-stretching maneuvers,
some of which involved a quick upward leap off the ground. I observed three tabanids
beeing successfully, captured by two different trumpeters during an observation period of
45 sec. Moreover, I suspect that ticks which may have been attached to the deer (which
unfortunately could not be resolved through a pair of 10 X 40 Zeiss binoculars) were also
being removed and eaten: several pecks were directed at specific leg and ventral parts of
the animal for no other obvious reason. Once the birds and the deer had already cleared the

SHORT COMMUNICATIONS

173

nearest distance between their path and the point where I had been sitting, one of the birds
sounded an alarm call, presumably directed at me. This clearly triggered the deer to sprint
away in leaps and bounds, and the typical collective response of trumpeters under contexts
of predation threat, which involves increased group vigilance and alarm-calling, followed
by a rapid withdrawal from the source of threat once it has been identified (pers. obs.).

As suggested by several interviews, both sets of observations reported here may have a
more widespread occurrence across the geographic distribution of Daptrius ater and Psophia
spp. in Amazonia. Of particular interest. Sick (1984) named Daptrius ater as “Gaviao-de-
Anta” (“tapir-hawk”), but although he reports that this species “sometimes removes ticks
and maggots from wild animals” (my translation), he does not specify which species of
“animals” those might actually be. The larger, more terrestrial and better known relatives
of forest caracaras (Milvago), which typically inhabit cattle-raising districts and savannas
farther south, also engage in similar interactions with other species of large marrimals.
Yellow-headed Caracaras (Milvago chimachima), for instance, are well reknowed to be
professional “tick-eaters”, frequently picking ectoparasites off livestock and capybaras (Hy-
drochaeris hydrochaeris) in the Brazilian pantanal and cerrado (Sick 1984, F. Olmos, pers.
comm.). M. chimachima and a number of other avian species which remove ectoparasites
from large mammals (e.g., magpies: Goodwin 1986) can, however, take advantage of these
associations to nibble at flesh and, therefore, maintain or enlarge branding wounds on their
mammalian hosts (Brown and Amadon 1968). Although this clearly brings nutritional ad-
vantages to these birds, it remains to be determined whether or not Daptrius ater and
Psophia leucoptera “cheat” in these otherwise mutualistic interactions by also eating live
tissue from open sores.

At oxbow lake- and river-edge habitats along white-water rivers of western Amazonia,
Giant Cowbirds (Scaphidura oryzivora) are commonly observed catching tabanid flies, and
perhaps ticks, parasitizing capybaras (J. Terborgh, pers. comm.), and Robinson (1988) once
observed a Giant Cowbird foraging on the back of a tapir. One of the few other cases of
ectoparasite removal from Tapirus was reported for white-nosed coatis (Nasua narica)
grooming a few individuals of Baird’s tapirs (T. bairdii) near the Barro Colorado Island
Field Station in Panama (McClean 1992). This interaction, however, was apparently learned
locally and represented a human artefact in that both species had been routinely fed in the
laboratory clearing, and thus spent a disproportionately large amount of time together.

The three-way interactions reported here, although easily overlooked in forest habitats,
may be of great importance to hematophagus arthropods, large mammalian hosts, and bird
species, such as caracaras and trumpeters, which successfully take advantage of such for-
aging opportunities. Ticks and tabanid flies, particularly if engorged with a full bloodmeal,
may represent a key nutritional supplement to the birds. For the ungulate ho.sts, too, this
may represent the most efficient way of eliminating unwelcome ectoparasites, which could
serve as vectors of major debilitating diseases. Of special interest, Amazonian forest un-
gulates, such as tapirs, both brocket deer species {Maz.ama americana and M. gouazouhira)
and peccaries (Tayassu tajacu and T. pecari) appear to be particularly susceptible to infes-
tations of large ticks. Hunters throughout the region often report examining carcasses of
these species with conspicuous ectoparasite loads, suggesting that infestation rates may be
faster than removal rates by mutualists or spontaneous withdrawal of the ectoparasite. In
the case of tapirs, kills are reported either to be heavily infested with ticks or conspicuously
“clean” (carrying no ticks), which suggest that tick-removal bouts by avian mutualists may
be relatively sporadic. Indeed, several species of ungulate may lose physical condition if
subject to chronic ectoparasite infestations, but whether or not such associations depress
ectoparasite populations to any significant extent is yet to be determined (but see Samuel
and Welsch 1991). Indeed, greater population densities of open-habitat mutualists (such as

174

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

that of Yellow-headed Caracaras in central Brazil) appear to be available to provide ecto-
parasite removal services to savanna and forest-edge grazers than are their equivalent forest
species to browser-frugivore ungulates. Perhaps the evolutionary opportunity presented by
the inherently low densities (and smaller body surface area) of forest megaherbivores
which could otherwise subsidize larger populations of large-bodied hematophagous arthro-
pods—was never significant enough to be claimed by cleaning mutualists more specialized
than caracaras and trumpeters.

Acknowledgments. — The faunal inventories which made these observations possible were
funded by the Wildlife Conservation Society and the Brazilian Science Council (CNPq). I
thank Josimar Pinheiro (Kaxinawa), Karuro and Batidn (Kayapo), and Edvar Dias for their
valuable assistance during fieldwork at sites surveyed in the upper Tarauaca, upper Xingu,
and upper Tefe rivers, respectively.

LITERATURE CITED

Attwell, R. I. G. 1966. Oxpeckers, and their associations with mammals in Zambia. Puku
4:17^8.

Bezuidenhout, J. D. and C. J. Stutterheim. 1980. A critical evaluation of the role played
by the red-billed oxpecker Buphagus erythrorhrynchus in the biological control of ticks.
Onderstepoort. J. Vet. Res. 47:51—75.

Bodmer, R. E. 1990. Responses of ungulates to seasonal inundations in the Amazon flood-
plain. J. Trop. Ecol. 6:191-201.

Brown, L. and D. Amadon. 1968. Eagles, hawks and falcons of the world. McGraw-Hill,
New York, New York.

Burger, J. and M. Gochfeld. 1982. Host selection as an adaptation to host-dependent
foraging success in the cattle egret (Bubulcus ibis). Behaviour 79:212-229.

Goodwin, D. 1986. Crows of the world, 2nd ed. Univ. of Washington Press, Seattle, Wash-
ington.

Griffiths, C. S. 1994. Syringeal morphology and the phylogeny of the Falconidae. Condor
96:127-140.

Hart, B. L., L. A. Hart, and M. S. Mooring. 1990. Differential foraging of oxpeckers
on impala in comparison with sympatric antelope species. Afr. J. Ecol. 28:240-249.

Lewis, A. D. 1989. Notes on two ravens Corvus spp. in Kenya. Scopus 13:129-131.

McClean, D. 1992. The rise and fall of a mutualism? Coatis, tapirs, and ticks on Barro
Colorado Island, Panama. Biotropica 24:220-222.

Penzhorn, B. L. and I. G. Horak. 1989. Starlings, mountain zebras and ticks. Koedoe 32:
133-134.

Peres, C. A. In press. Nonvolant mammal community structure in different Amazonian
forest types in Mammals of the Neotropics, Vol. 3 (J. E Eisenberg, ed.). Univ. of
Chicago Press, Chicago, Illinois.

and a. Whittaker. 1991. Annotated checklist of bird species of the upper Rio

Urucu, Amazonas, Brazil. Bull. Brit. Ornith. Cl. 111:156—171.

Roberts, S. C. 1993. Yellowbellied bulbul gleaning on a klipspringer. Ostrich 64:136.

Robinson, S. K. 1988. Foraging ecology and host relationships of Giant Cowbirds in south-
eastern Peru. Wilson Bull. 100:224—235.

Samuel, W. M. and D. A. Welsh. 1991. Winter ticks on moose and other ungulates: factors
influencing their population size. Alces 27:169-182.

Sherman, FT 1991. The ecology and social behavior of the white-winged trumpeter
(Psophia leucoptera). Ph.D. diss., Univ. of California, Davis, California.

Sick, H. 1984. Ornitologia brasileira: uma introdu9ao. Vol. 1. Editora Univ. de Brasilia,
Brasilia, Brazil.

SHORT COMMUNICATIONS

175

Thiollay, J. M. 1991. Foraging, home range use and social behaviour of a group-living
rainforest raptor, the red-throated caracara Daptrius americanus. Ibis 133:382—393.
WiTTENBERGER, J. F. 1981. Animal social behavior. Duxbury Press, Boston, Massachusetts.

Carlos A. Peres, Dept, of Ecology, Universidade de Sdo Paulo, Caixa Postal 1 1.46I, Sdo
Paulo-S.P. 05422-970, BRAZIL. (Present address: CSERGE, School of Environmental Sci-
ences. University of East Anglia, Norwich NR4 7TJ, ENGLAND). Received 25 April 1995,
accepted I Sept. 1995.

Wilson Bull., 108(1), 1996, pp. 175-178

Notes on the status and behavior of the Swainson’s Warbler in Cuba. — The Swain-
son’s Warbler (Limnothlypis swainsonii) is one of the less common North American warblers
(Morse 1989). Although data from the Breeding Bird Survey suggest that the species has
undergone a significant range-wide population increase during the period 1966-1988 (Sauer
and Droege 1992), regional Neotropical migrant prioritization schemes for the midwestern
(Thompson et al. 1993) and southeastern (Hunter et al. 1993) United States consider the
Swainson’s Warbler among the more vulnerable Neotropical migrants based on its low
population, threats on the breeding and wintering grounds, and its restricted range. Consid-
ering its vulnerability, the status of the Swainson’s Warbler is poorly known in its breeding
range (Hunter et al. 1993), and even more so in winter. Here, we summarize recent and
historical records for the Swainson’s Warbler in Cuba, re-assess its status there, and describe
aspects of its foraging and flocking behavior based on casual observations, previously pub-
lished information, and anecdotal reports.

The Swainson’s Warbler winters in the northern Bahama Islands, Cuba, the Cayman
Islands, Jamaica, the Yucatan Peninsula, and Belize (AOU 1983). There are also sight and
banding records from Puerto Rico (AOU 1983; J. Faaborg, pers. comm.) and sight records
from St. John (Raffaele 1989). It is reported as casual on the Swan Islands (AOU 1957). In
Cuba, the Swainson’s Warbler has been considered a rare winter resident (Garrido and Garcia
Montana 1975). The first report for Cuba was provided by Gundlach (1876) who knew of
a single sight record from La Habana (Fig. 1). In the 150 years prior to 1991, it is unknown
exactly how many Swainson’s records exist for Cuba, but we are aware of only 21 (Fig. 1).

Banding activities carried out during the winters of 1991-1994 by Cuban researchers of
the Institute of Ecology and Systematics (lES) of the Ministry of Science, Technology, and
Environment, the Cuban National Museum of Natural History; and by a cooperative forest
bird survey project of the lES, the Canadian Wildlife Service (CWS), and the Long Point
Bird Observatory (LPBO) have provided many new records of Swainson’s Warbler. Recent
bird-watching tours have contributed additional sight records of the species. In total, 58
individuals were observed, netted, or collected at 17 sites during the winters of 1991-1994
(Fig 1). Highest numbers were at El Cenote, Cienaga de Zapata, Matanzas Province, where

13 (1.80/100 net-h) were captured 1 1-14 February 1991, and at Camino al Sitio Viejo, Cayo
Coco, Ciego de Avila Province, where 12 (1.67/100 net-h) were captured 27-30 January
1994. Seasonally, Swainson’s Warblers have been observed in Cuba from 15 September to

14 April (Garrido and Garcia Montana 1975; Garrido and Kirkconnell, unpubl. data).

Historical and recent Cuban records indicate that Swainson’s Warblers occur in the low-
lands, montane regions, and in swampy areas. They apparently prefer semideciduous forest
with high shrub and tree stem density, complete, or nearly complete, canopy cover, abundant

176

THE WILSON BULLETIN • Vol. 108. No. 1, March 1996

The first numeral, followed by a colon, indicates number of sites at that location trom which
records are derived and is followed by the number of individuals in each of the following
categories: C = Collected, N = Netted and usually banded, R = Recapture of bird banded
previous year, S = Sight record, ? = exact number of individuals unknown.

dry leaf litter, and humid, shady areas, sometimes near streams, or near waterholes in lime-
stone bedrock near the coast. While they seem to inhabit mostly larger forest tracts, they
also occasionally reside in smaller forest fragments in the vicinity of larger forests, or, in
the keys, in patches of low coastal scrub.

Inter-winter site fidelity has been documented for the Swain.son’s Warbler in Jamaica (Di-
amond and Smith 1973). In the recent Cuban banding surveys, a Swain.son’s Warbler was
recaptured one year after banding and 120 m from the original capture site (McNicholl 1992).

Swainson's Warblers spend most of their time on the ground or less than 1-2 m above it.
Lack and Lack (1972) reported that Swainson’s Warblers in Jamaica always feed on the
forest floor where they rummage and probe in leaf litter, sometimes tossing leaves aside.
Our observations in Cuba support this, but include sightings of the species insect gleaning
from surfaces of leaf litter or bare ground, and gleaning prey in slow moving water. Whereas
insects seem to be the principal food items, Eaton (1953) discovered the bones of small
lizards, perhaps anoles (Anoli.s sp.) or geckos {Sphaerodactyhis sp.), in stomach samples of
Swainson’s Warblers wintering in Cuba.

Although Eaton (1953) classified Swainson’s Warblers as solitary foragers, we frequently
observed them foraging in close association with other warblers, particularly Ovenbirds
{Seinni.s aurocapilla.s) and Worm-eating Warblers {Helniitheros vermivorus). When the three
species were together, Swainson’s Warblers foraged in the wettest areas with Worm-eating
Warblers, while Ovenbirds foraged in drier areas. At La Giiira, Pinar del Rfo Province in
Feb-Mar 1986, Garrido twice observed Kentucky Warblers (Opororis formosus) and Hooded
Warblers (Wilsonia cilrina), joining such mixed-species Hocks, typically feeding within a
few inches of the ground, and sallying or hovering in a manner similar to the American
Redstart (Seiophaf^a ruticiUa) (Bennett 1980).

In many ol our sightings, we observed an apparent association between Swainson s War-
blers and Ovenbirds. Like Swainson’s Warblers, Ovenbirds pick insects oft leal litter or the

SHORT COMMUNICATIONS

177

ground, but, in general, they use their bills to rummage in dry leaves more frequently than
Swainson’s Warblers. During four of our sightings, we observed a Swainson’s Warbler feed-
ing in the "wake” of an Ovenbird, sometimes following the Ovenbird within a few centi-
meters. During many other observations Ovenbirds and Swainson’s Warblers foraged in
close proximity without any apparent aggression. Ovenbirds and Swainson’s Warblers were
frequently captured side-by-side in mist nets suggesting that they were moving together
through the forest. We suggest that the body movements of walking Ovenbirds, which often
include repeated cocking of the tail, and the disruption of the leaf litter with the bill and
feet, make otherwise cryptic insects move or fly, facilitating their capture by Swainson’s
Warblers (the “beater effect” of Powell 1985). Other potential explanations of this trailing
behavior include the possibility that active foraging by Ovenbirds signals the presence of
food resources to Swainson’s Warblers, as proposed for a variety of mixed-species assem-
blages (e.g., Gannon 1934, Rand 1954, Sealy 1973, Turner 1965); or that the association
provides enhanced predator detection for both species, as proposed for other mixed-species
flocks (e.g., Cody 1971).

In summary, the recent evidence provided by bird banding and regular, intensive searching
during bird-watching tours suggests that the Swainson’s Warbler is a more common winter
resident in Cuba than previously believed. The total of 58 individuals recorded during the
winters of 1991-1994 is nearly three times the total recorded in the previous 150 years. We
believe that the paucity of early records is attributable to the cryptic plumage of the species,
its elusive behavior in dense vegetation, and, particularly, the lack of intensive surveys
utilizing mist nets.

Acknowledgments. — We thank the Cuban Ministry of Science, Technology, and Environ-
ment, the Canadian Wildlife Service, Great Auk Nature Tours, Long Point Bird Observatory,
and the Canadian Nature Federation for financial and logistical support. Graeme Gibson of
Great Auk Nature Tours and CubaTour made the bird-watching trips to Zapata and La Guira
possible. The following people assisted in the field: Martin Acosta, Vincente Berovides, Pedro
Blanco, Jane Bowles, Beverly Collier, Ronel Concepcion, Donald Fillman, Esteban Godinez,
Raul Gomez, Hiram Gonzalez, Colleen Hyslop, Judith Kennedy, Alejandro Llanes, Frank
Loyola, Miriam Martinez, Jonathan McCracken, Martin McNicholl, Lourdes Mugica, Ramona
Oviedo, Alina Perez, Ronald Ridout, Days! Rodriguez, Barbara Sanchez, Eliser Socarras,
Steven Wendt, and Daily Zuniga. Joseph Walter of the Geographic Resource Center, Univ. of
Missouri produced the map. Michael Baltz, John Faaborg, Herbert Raffaele, Elizabeth Wallace,
and an anonymous reviewer contributed valuable comments on the manu.script.

LITERATURE CITED

American Ornithologists’ Union. 1957. Check-list of North American birds, 5th ed.
A.O.U., Washington, D.C.

. 1983. Check-list of North American birds, 6th ed. A.O.U., Washington, D.C.

Bennett, S. E. 1980. Interspecific competition and the niche of the American Redstart
(Setophaga ruticilla) in winter and breeding communities. Pp. 919-335 in Migrant birds
in the Neotropics: ecology, behavior, distribution, and conservation (A. Keast and E.
S. Morton, eds.). Smithsonian Inst. Press, Washington, D.C.

Cody, M. L. 1971. Finch flocks in the Mojave Desert. Theor. Popul. Biol. 2:142-158.
Diamond, A. W. and R. W. Smith. 1973. Returns and survival of banded warblers wintering
in Jamaica. Bird-Banding 44:221-224.

Eaton, S. W. 1953. Wood warblers wintering in Cuba. Wilson Bull. 65:169-174.

Gannon, G. R. 1934. Associations of small insectivorous birds. Emu 34:122-129.

AND F. Garcia Montana. 1975. Catalogo de las aves de Cuba. Acad. Cien., La

Habana.

178

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Gundlach, J. 1876. Contribucion a la Ornitologia Cubana. Imprenta La Antilla.

Hunter, W. C., D. N. Pashley, and R. E. E Escano. 1993. Neotropical migratory landbird
species and their habitats of special concern within the southeast region. Pp. 159—171
in Status and management of Neotropical migratory birds (D. M. Finch and P. W.
Stangel, eds.). USDA Forest Serv., Gen. Tech. Rep. RM-229.

Lack, D. and P. Lack. 1972. Wintering warblers in Jamaica. Living Bird 11:179-183.

McNicholl, M. K. 1992. Surveys of Nearctic migrant and Neotropical resident birds win-
tering in Cuban forest ecosystems: report of the 1992 field season. Unpubl. rep.. Long
Point Bird Obs., Port Rowan, Ontario.

Morse, D. H. 1989. American warblers, an ecological and behavioral perspective. Harvard
Univ. Press, Cambridge, Massachusetts.

Powell, G. V. N. 1985. Sociobiology and adaptive significance of interspecific foraging
flocks in the neotropics in P. A. Buckley et al. (eds.). Neotropical Ornithology. Ornithol.
Monographs 36.

Raffaele, H. a. 1989. A guide to the birds of Puerto Rico and the Virgin Islands, 2nd ed.
Princeton Univ. Press, Princeton, New Jersey.

Rand, A. L. 1954. Social feeding behavior of birds. Fieldianna, Zool. 36:1-71.

Sanchez, B. and D. Rodriguez. 1992. Relacion de aves y nuevo reporte de Bijirita de
Swainson (Limnothlvpis swainsonii) en el Rincon del Guanal, sur Isla de Juventud,
Cuba. Comunicaciones Breves de Zoologia: 15— 1 8. Editorial Academia, La Habana.

Sauer, J. R. and S. Droege. 1992. Geographic patterns in population trends of Neotropical
migrants in North America. Pp. 26^2 in Ecology and conservation of Neotropical
migrant landbirds (J. M. Hagan III and D. W. Johnston, eds.). Smithsonian Inst. Press,
Washington, D.C.

Sealy, S. G. 1973. Interspecific feeding assemblages of marine birds of British Columbia.
Auk 90:796-802.

Thompson, E R., S. J. Lewis, J. Green, and D. Ewert. 1993. Status of Neotropical migrant
landbirds in the Midwest: identifying species of management concern. Pp. 145-158 in
Status and management of Neotropical migratory birds (D. M. Finch and P. W. Stangel.
eds.). USDA Forest Serv., Gen. Tech. Rep. RM-229.

Turner, E. R. A. 1965. Social feeding in birds. Behavior 24:1-46.

Arturo Kirkconnell, Museo Nacional de Historia Natural. CapitoHo Nacional, Lm Ha-
hana, Cuba', George E. Wallace, Long Point Bird Obserx’atory, Box 160. Port Rowan,
Ontario, NOE I MO Canada', and Orlando H. Garrido, Museo Nacional de Historia Nat-
ural. CapitoHo Nacional, La Habana, Cuba. Received 5 Dec. 1994, accepted I Oct. 1995.

Wilson Bull., 108(1), 1996, pp. 178-180

Comments on a probable gynandromorphic Black-throated Blue Warbler. — Sexual
plumage differences in passerine birds are believed to be controlled genetically and only
minimally influenced by hormones (Murton and Westwood 1977). Bilateral gynandromorphs
are among the most striking manifestations ot chromosomal regulation of plumage (Crew
and Munro 1938, Cock I960, Witschi 1961). In these rare individuals, plumages of the left
and right sides of the body are demarcated along the midline and presumably reflect gonadal
placement. In most cases, an ovary and female plumage are found on the lelt side, a testis
and male plumage on the right. Several hypotheses have been advanced to explain the

SHORT COMMUNICATIONS

179

genesis of gynandromorphs and mosaics, but it is possible that several genetic mechanisms
or ontogenetic tracks may produce a variety of phenotypic outcomes collectively lumped as
bilateral gynandromorphs. This idea is suggested by variation in the degree of plumage
asymmetry in gynandromorphs. For example, traces of definitive male plumage may appear
on the "female” side and vice versa (e.g., Laybourne 1967). One problem with categorizing
gynandromorphic birds, is that there are no standard conventions of analysis, such as those
developed for diagnosing hybrids (Graves 1990). The following example illustrates some
of the difficulties.

Patten (1993) reported an unusually plumaged Black-throated Blue Warbler (Dendroica
caerulescens) that was photographed but could not be collected at Stovepipe Wells, Death
Valley National Monument, Inyo County, California, in October 1987. Patten described (p.
696) this individual as “sexually dimorphic with respect to each lateral half of the bird,
with the left side appearing to be male and the right looking like a female.” While we
believe the bird was correctly identified as a bilateral gynandromorph, the first such record
for the subfamily Parulinae, a reappraisal of photographs suggests that its phenotype was
qualitatively different from bilateral gynandromorphs previously reported among passerines
(Crew and Munro 1938, Kumerloeve 1987).

We examined two color transparencies taken by Dunn (Visual Resources for Ornithology
[VIREO] archives catalog numbers v06/23/00]-v06/23/002) as well as those examined by
Patten (taken by Paul E. Lehman, VIREO v06/ 1 2/00 l-v06/l 2/004). The left side of the bird
appears to be in male first-basic plumage (hatching year), as indicated by brownish second-
aries and primaries and an unusual white spot on the lower eyelid. The left side is clearly
demarcated from the right side along the midline of the mantle. The right side has an
enigmatic appearance not matched by any of the 1370+ study specimens of Black-throated
Blue Warbler in the National Museum of Natural History (USNM), Smithsonian Institution.
The principal characters of the right side include (1) pale supercilium that extends from the
base of the bill posteriorly to the rear of the auriculars where it becomes broader, (2)
extensively white lower eyelid, (3) grizzled black and white throat, whiter at the center, (4)
pale malar mark, (5) yellowish wash on the belly and lower sides, (6) a relatively wide
black stripe on the side that extends from the lower edge of the throat posteriorly past the
bend of the wing to the base of the primary coverts, (7) large white triangle at the base of
the primaries, and (8) mantle and crown, olive-gray, distinctly less “blue” than the left side.

A comparative description of the right side of this gynandromorph and basic plumages
of male and female Black-throated Blue Warblers follows. White superciliary and lower
eyelid markings occur rarely in fall male Black-throated Blue Warblers in first basic plumage
but are unknown in males in definitive basic plumage. In contrast, nearly all females have
pale superciliaries and lower eyelid spots (.see Parkes 1979). The face pattern of Patten’s
bird was more strongly pronounced than that of any male or female we have examined.

Throat feathers of males in first basic plumage are often tipped with white, imparting a
grizzled appearance. With very rare exceptions, females have unmarked throats. Patten’s
bird had a grizzled throat, whiter near the center, and a pale malar mark, thus appearing
more male-like than female in this character.

A black stripe extends from the side of the throat posteriorly along the sides to the lower
flanks in basic-plumaged males, whereas females in all plumages have unmarked sides. Pat-
ten’s bird exhibited a distinctive black stripe tbal began at the right side of the throat and
continued to the upper flanks, becoming more diffuse posteriorly. The extent of the black
stripe was well within the range exhibited by males in first basic plumage. The remainder of
the underparts of Patten’s bird more closely resembled those of females in basic plumage.

In sum, the right side of Patten’s bird appeared to be a mosaic of distinctive elements
from both male and female plumages, but weighed more heavily toward male characters.

180

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

while the left side was typically male. However, the true nature and cause of this plumage
asymmetry will never be known without a specimen.

Crew and Munro (1938) concluded that bilateral gynandromorphism in birds is of three
types: (1) the finch or sparrow type where plumage is genetically determined and reflective
of lateral chromosome distribution; (2) the chicken or fowl type in which sexual differences
in plumage are subject to hormonal regulation so that perfect bilaterahty of plumage is
impossible- and (3) the pheasant type where plumage on one half of the bird is normal and
the other half is an intersexual mosaic (Danforth 1937a,b). Nearly forty years later, Hollan-
der’s (1975) review of sectorial mosaics in pigeons suggested that mosaics of sex-linked
plumage may be caused by bipaternity and the subsequent incorporation of tissue derived
from supernumerary sperm into an embryo. This process may result m an asymmetrical
patchwork of male and female plumage such that one side appears to be normal and the
other side an intersexual mosaic. In any case, little has been learned about gynandromor-
phism in passerines other than the fact that departure from bilateral asymmetry is variable
(see Laybourne 1967, Kumerloeve 1987). The appearance of Patten’s bird, normal male on
the left side and intersexual mosaic on the right, marks an observed extreme within the
order Passeriformes.

LITERATURE CITED

Cock, A. G. 1960. Pour half-and-half mosaic fowls. Genet. Res. 1 :275-287.

Crew, P. A. E. and S. S. Munro. 1938. Gynandromorphism and lateral asymmetry m birds.
Proc. Royal Soc. Edinburgh 58:114—135.

Danforth, C. H. 1937a. Artificial gynandromorphism and plumage m Phasianus. J. Genet.
34:497-506.

. 1937b. An experimental study of plumage in Reeves Pheasants. J. Exp. Zool. 77:

Graves, G. R. 1990. Systematics of the “green-throated sunangels” (Aves: Trochilidae):

valid taxa or hybrids? Proc. Biol. Soc. Wash. 103:6-25.

Hollander, W. E 1975. Sectorial mosaics in the domestic pigeon: 25 more years. J. He-
redity 66:177— 202.

Kumerloeve, H. 1987. Le gynandromorphisme chez les oiseaux — recapitulation des don-
nees connues. Alauda 55:1-9.

Laybourne, R. C. 1967. Bilateral gynandrism in an Evening Grosbeak. Auk 84:267-272.
Murton, R. K. and N. j. Westwood. 1977. Avian breeding cycles. Clarendon Press, Ox-
ford, England.

Parkes, K. C. 1979. Plumage variation in female Black-throated Blue Warblers. Cont.
Birdlife 1:133-135.

Patten, M. A. 1993. A probable bilateral gynandromorphic Black-throated Blue Warbler.
Wilson Bull. 105:695—698.

WIT.SCHI E. 1961. Sex and secondary sexual characters. Pp. 1 15-168 in Biology and com-
parative physiology of birds, Vol. II (A. J. Marshall, ed.). Academic Press, New York.

Gary R Graves, Dept, of Vertebrate Zoology, National Mii.seuni of Natural History, Smith-
sonian hutitution, Washington, D.C. 20560- Michael A. Patten, Department of Biology,
Univer.sitv of California, Riverside, California 9252 C, and Jon L. Dunn, 153 Grange Hall
Road. Dayton. Ohio 45430. Received 19 Jan. 1995, accepted 5 Sept. 1995.

SHORT COMMUNICATIONS

181

Wilson Bull., 108(1), 1996, pp. 181-182

Rufous crown feathers on adult male Tennessee Warblers. — The presence of rufous
crown feathers in Tennessee Warblers [Vennivora peregrinci) is an undescribed feature that
supports taxonomic affinities with other Vennivora warblers, although the possibility of
hybrid origins exists. In 1980, among Tennessee Warblers collected from Aquatuk Lake in
northeastern Ontario, J. A. Dick noted two adult males with crown feathers with some rufous
coloration among the normally grayish crown feathers. This phenomenon seems to have
been overlooked in descriptions of this species, except for a very brief passing comment by
Chapman (1917). S. V. Nash and R. D. James located another example in northeastern
Ontario in 1982. The Royal Ontario Museum has six additional birds displaying some rufous
in the crown. This note describes in greater detail the extent of rufous crown feathers in
Tennessee Warblers.

J. A. D. asked curators of several other North American museums, if they had Tennessee
Warblers with rufous crown feathers. Nine of 850 (about I %) were reported to have them.
The proportion may be higher, as the coloration can be overlooked easily (see below) and
in Royal Ontario Museum collections eight examples among 136 specimens (nearly 6%)
show some rufous on crown feathers.

The rufous feathers, as is typical of other Vennivora, are found on adult males that
normally have completely gray crowns. On any individual feather, the rufous is in the center
of the feather, including the rachis, although one vane may be more highly colored. It may
be fairly distinct and roughly tear-drop shaped, and the gray distal border of the feather is
as wide as the rufous drop. Thus, the rufous might be scarcely, if at all, visible when the
feathers are in place, as only the gray terminal edges of the feathers would show. On some
other birds, the rufous is much less distinct, being a very light suffusion among the gray,
and extends closer to the distal ends of the crown feathers. This type is usually visible with
the feathers in place but can be so pale as to be scarcely noticeable.

The rufous color is not as orange as on an Orange-crowned Warbler (U. celata), nor as
chestnut as on a Nashville Warbler (V. ruficapilla). It is somewhat intermediate, but more
orange than chestnut. It approximates the orange-rufous of Smithe (1975), color 132C, but
is very much paler than that illustrated, making it easy to overlook. The number of colored
feathers also varies considerably from bird to bird. Some have only one or a very few,
usually clo.se to the front center of the crown and usually, but not always, the more strongly
colored; one bird had only a single very pale colored feather. The paleness and the small
numbers and size of the feathers makes them difficult to detect. On several others there
were numerous colored feathers, usually less intensely colored and lying in a “patch” across
the center to hind crown. The rufous coloration becomes more obvious if the feathers are
ruffled to show the length of individual feathers. But it is akso possible that the rufous color
has faded on the older specimens, making them even more difficult to detect. One bird
showed some stronger coloring toward the front center as well as more lightly colored
feathers toward the back of the crown. Another showed a suffusion of color across only the
fore crown, strongest toward the front center.

There is only one known hybrid specimen involving the Tenne.s.see Warbler, Carnegie
Museum of Natural History 152341, collected at the museum’s Powdermill Nature Reserve,
Rector, Pennsylvania, on 26 Augu.st 1979. The other putative parent species was identified
by K. C. Parkes and R. C. Leberman as the Nashville Warbler, a species in which adult
males have an obvious rufous crown patch. The hybrid specimen, however, is a male in
first basic plumage. Although 70% of males of V. r. ruficapilla of this plumage stage in the
Carnegie collection have crown patches, the patches are unknown in young Tennes.see War-
blers, and there is no sign of one in the hybrid (K. C. Parkes, in litt.). Intrageneric hybrids

182

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

are rare in the Parulinae (Parkes 1978; Bledsoe, 1988), and it would seem unlikely that the
crown patch in Tennessee Warblers is a consequence of hybridization. None of those we
have examined with rufous crown feathers would appear to be anything but typical Ten-
nessee Warblers in other respects. With more than one percent of the population exhibiting
rufous crown feathers, it seems much more likely that the rufous is a vestigial plumage
pattern of the type typically found in most other members of the genus Vermivora.

Among other Vermivora warblers in which the males exhibit obvious rufous crown patch-
es, it might be presumed that these patches serve a display function. It seems very unlikely,
however, that the coloration now serves any display function in Tennessee Warblers, since
it is either very restricted or very pale.

Acknowledgments — We thank S. V. Nash for assistance in the field. S. L Bailey, Univ. of
California, Berkeley, J. W. Litzpatrick, Lield Museum of Natural History, J. Hinshaw, Univ.
of Michigan, Ann Arbor, H. Ouellet, Canadian Museum of Nature, K. C. Parkes, Carnegie
Museum of Natural History, and T. Webber, Llorida State Museum kindly examined spec-
imens in their respective institutions for us. We would also like to thank K. C. Parkes and
J. D. Rising for comments on an earlier draft of this paper.

LITERATURE CITED

Bledsoe, A. H. 1988. A hybrid Oporornis Philadelphia X Geothlypis trichas, with com-
ments on the taxonomic interpretation and evolutionary significance of intrageneric
hybridization. Wilson Bull. 100:1—8.

Chapman, P. M. 1917. The warblers of North America. D. Appleton & Co., New York,
New York.

Parkes, K. C. 1978. Still another parulid intergeneric hybrid (Mniotilta X Dendroica) and
its taxonomic and evolutionary implications. Auk 95:682-690.

Smithe, L B. 1975. Naturalist’s color guide. New York, American Museum of Natural
History.

James A. Dick and Ross D. James, Dept, of Ornithology, Royal Ontario Museum, 100
Queen’s Park Crescent, Toronto, Ontario, Canada, MSS 2C6. Received 19 Mar. 1995,
accepted 20 Aug. 1995.

Wilson Bull., 108(1), 1996, pp. 182-186

American Goldfinch nests in purple loosestrife.— Bird foraging, nesting, and other ac-
tivities are often closely related to vegetation characteristics. Introduced plants may alter the
architecture and chemistry of the plant community, potentially affecting the food base and
nest substrate available to birds. One non-indigenous plant, purple loosestrife {Lythrum sal-
icaria), is said to have little value to North American wildlife, and a biological control
program is predicted to dramatically reduce American populations of loosestrife (Malecki
et al. 1993). Here I report American Goldfinch {Carduelis tristis) use of loosestrife as nest
substrate.

The American Goldfinch is a widespread breeding bird in North America and nests in a
variety of habitats that include parks and yards with ornamental vegetation, weedy waste
grounds, forest edges, fence rows, old fields, abandoned orchards, shrub swamps, and marsh-
es. Nickell (1951) and Smith (1988) suggested that pre-Columbian habitats were beaver

SHORT COMMUNICATIONS

183

Table 1

American Goldfinch Nests in Purple Loosestrife, Southeastern New York

Dimension

Mean

Median

IQR"

Range

N

Nest rim to soil (cm)

140.3

134

121-182

90-190

7

Plant height from soil (cm)

229.7

224

200-246

192-315

7

Nest height fraction of plant height (%)

60.9

60.3

54-66.7

45-74

7

Maximum plant diameter at level of nest (cm)

94.2

94

67-1 14

58-152

12

Nearest woody plant > 1 m tall (m)“

14.6

11.5

6.5-21

2-34

12

Nearest open water >1 m wide (m)"'

9.2

4

1-16

1-37

1 1

Nearest shore (upland) (m)^

19.4

17

0-29

0-71

11

• Paced or estimated from maps.
Interquartile range.

meadows, wetlands, lake and river banks, and bum areas. Purple loosestrife, a robust, shmb-
like forb introduced to North America ca 1800 (Thompson et al. 1987), is common in the
moister types of habitats used by nesting goldfinches in New York.

From 1971 to 1994, in the course of other field work, I found 15 American Goldfinch
nests in purple loosestrife in Dutchess and Ulster counties in the Hudson Valley (Table 1 ).
I have deposited voucher photographs at Visual Resources for Ornithology (VIREO catalog
numbers v06/22/001 through v06/22/004; Academy of Natural Sciences of Philadelphia,
Philadelphia, Pennsylvania). Three nests were active and 12 abandoned, the abandoned nests
identified by a persistent accumulation of nestling feces on the nest rim (Walkinshaw 1939,
Berger 1971:258).

The goldfinch nests were above water or intermittently saturated soil as follows: five in
flooded nontidal marshes, two in the upper intertidal zone of a freshwater-tidal marsh, two
at pond margins, two in patches of unmowed wet meadow within mowed fields, one in a
patch of wet meadow in a dry old field, two in extensive wet old fields, and one in a small
roadside wetland. The nests were attached-statant and resembled types 1, 2, and 6 illustrated
by Nickell (1958) for the goldfinch. The nests were typically woven around several primary
vertical stems of loosestrife and their ascending secondary branches, with the branches
preventing the nests from sliding down the primary stems. Nests were attached to stalks of
the current year (8), previous year (1), or both years (2), and remained identifiable through
one or even two winters (Smith 1988). The nests were in tall, wide, many-stemmed loose-
strife clumps near or touching other robust loosestrife (Table 1). The nests were not wider
than high (contra Allen 1934), probably the result of attachment to the vertical stalks of
loosestrife rather than being saddled on horizontal branches of woody plants.

Goldfinch nesting habitat, nest sites (plant species), and nest height vary greatly (Berger
1971), but nests are not placed on wholly artificial substrates or on the ground (Nickell
1951). I compiled literature reports of 5991 nest substrate records, predominantly (94%)
from the Great Lakes region, representing 87 plant species or genera. Of these records,
87.5% are in broad-leaved trees or shrubs and only 1 1 .6% in herbs (almost all thistles
[Cirsium]). The same records comprise 55% nests in plant taxa native (Gleason and Cron-
quist 1991, Peattie 1991) to the American Goldfinch breeding range, 1% in introduced taxa,
and 44% not determinable as native or introduced (mostly hawthorns [Crataegus] and wil-
lows [^a/ix]). The records are 32% wetland plant taxa following the classification of Reed
(1988), 22% upland taxa, and 46% ambiguous (mostly hawthorns, elms [Ulmus], and maples
[Acer]). The data compilation and a list of sources have been deposited at the National

184

THE WILSON BULLETIN • Vol. I OH, No. I, March 1996

Technical Information Service (NTIS Accession Number PB95-226965; 5285 Port Royal
Road, Springfield, Virginia 22161).

Nickell (1958) remarked that American Goldhnch and Red-winged Blackbird (Agelaius
phoeniceus) are exceptions to the rule that bird species nesting above the ground do not
attach nests to herbs because they are weak, insufficiently branched, and not fully grown at
the peak of the nesting season. Goldfinch nests with shorter distances from woody plants
(Table 1) suggest loosestrife was occasionally selected in preference to apparently suitable
woody species close by, and nests with longer distances indicate loo.sestrife allowed gold-
finches to nest in the interiors of herbaceous wetlands not otherwise usable. Purple loose-
strife is intermediate in growth form between the shrubs and the thistles used by nesting
goldfinches. Like the thistles, loosestrife shoots develop late; they do not reach full height
until August (Rawinski 1982:27). Most goldfinches begin nest construction mid-June to mid-
August. The late development of loosestrife may make it unsuitable for some marsh birds
that build elevated nests in May and June.

McCabe (1991:50, 55) believed concealment from predators, weather, sunlight, and con-
specifics the key factor governing taxonomic choice of nest site in the Willow Flycatcher
Empidona.x traillii Goldfinch eggs, nestlings, and sitting females are vulnerable to over-
heating in the sun and to chilling in rain and wind (Mayer 1981, Kleinhenz 1984). Eggs
and nestlings can also drown when tightly constructed nests fill with rain (Allen 1934).
Kleinhenz (1984) found that successful nests had more overhead vegetation cover, and that
goldfinches selected broad-leaved rather than nanow-leaved hawthorns. Mature loosestrife
has a dense leafy crown that presumably shelters goldfinch nests from sun, wind, and rain;
tall, dense loosestrife growth probably also conceals nests from predators and the Brown-
headed Cowbird (Molothrus ater).

Wetlands can be refuges from predation and brood parasitism (Kiviat 1989). Expanses of
soft wet soil, stream channels, or frequent flooding may deter some mammals and snakes
from reaching nests. Many potential avian predators and the Brown-headed Cowbird do not
venture far into marshes that lack tall shrubs or trees. Low rates of nest predation and brood
parasitism of blackbirds. Swamp Sparrow {Melospiza georgiana), and Song Sparrow {M.
melocJia) have been recorded in extensive marshes (Johnston 1956, Friedmann 1963, Mean-
ley and Webb 1963, Ortega and Cruz 1991).

Native birds commonly nest in certain introduced plants, e.g., shrubby honeysuckles (Lo-
nicera spp.), common buckthorn (Rhamnus cathartica), and multiflora rose (Rosa multiflora)
DeGraaf et al. 1975, Whelan and Dilger 1992), whereas other aliens such as tamarisk (Ta-
marix spp.) (Brush 1983, Ohmart et al. 1988:156-157) are rarely used, and not all native
species are used equally (Berger 1971:217). Although the American Goldhnch more often
nests in native than introduced plants, a few abundant alien taxa are used. The breeding
range of the American Goldhnch (AOU 1983) and the American range of purple loosestrife
(Thompson et al. 1987:19) are nearly conterminous, making loosestrife a potential ne.st
substrate for the goldhnch over much of its range. In extensive marshes lacking woody
vegetation, loosestrife could have facilitated an ecological extension of goldhnch nesting
habitat. Native forbs that form tall clumps and dense patches in marshes and wet meadows
rarely provide a structure as dense, leafy, and sturdy as that of mature clumps of purple
loosestrife.

Acknowledgments. — Charles Leek, Brooke Meanley, Thomas J. Rawinski, Gretchen Ste-
vens, Bryan L. Swift, and Julie Zickefoose reviewed a dralt, and 1 am also grateful to
referees Alex L. A. Middleton, Charles R. Smith, and Doris Watt. This is Bard College
Field Station - Hudsonia Contribution 31.

SHORT COMMUNICATIONS

185

LITERATURE CITED

Allen, A. A. 1934. American bird biographies. Comstock Publishing Co., Ithaca, New
York.

American Ornithologists’ Union. 1983. Check-list of North American birds, 6th ed.
A.O.U., Washington, D.C.

Berger, A. J. 1971. Bird study. Dover Publications, New York, New York.

Brush, T. 1983. First nesting of a New World woodpecker in tamarisk (Tarnarix chinensis).
Southwestern Naturalist 28(1 ):1 13.

DeGraaf, R. M., H. R. Pywell, and J. W. Thomas. 1975. Relationships between nest
height, vegetation, and housing density in New England suburbs. Transactions of the
Northeast Section, Wildlife Society 32:130-150.

Friedmann. H. 1963. Host relations of the parasitic cowbirds. U.S. National Museum Bul-
letin 233.

Gleason, H. A. and A. Cronquist. 1991. Manual of vascular plants of northeastern United
States and adjacent Canada, 2nd ed. New York Botanical Garden, Bronx, New York.

Johnston, R. F. 1956. Population structure in salt marsh Song Sparrows. Part II. Density,
age structure, and maintenance. Condor 58:254—272.

Kiviat, E. 1989. The role of wildlife in estuarine ecosystems. Pp. 437-475 in Estuarine
ecology (J. W. Day et al., eds.). John Wiley and Sons, New York.

Kleinhenz, P. C. 1984. Nest site microclimates and their energetic significance to nesting
American Goldfinches (Carduelis tristis). M.S. thesis, Ohio State University, Columbus,
Ohio.

Malecki, R. a., B. Blossey, S. D. Hight, D. Schroeder, L. T. Kok, and J. R. Coulson.
1993. Biological control of purple loosestrife. BioScience 43:680-686.

Mayer, L. P. 1981. The importance of seasonal microclimate utilization of two small birds,
Carolina Chickadee (Pams caroUnensis) and American Goldfinch (Carduelis Irislis).
Ph.D. diss. Ohio State Univ. Columbus, Ohio.

McCabe, R. A. 1991. The little green bird; ecology of the Willow Flycatcher. Corrected
printing. Rusty Rock Press, Madison, Wisconsin.

Meanley, B. and j. S. Webb. 1963. Nesting ecology and reproductive rate of the Red-
winged Blackbird in tidal marshes of the upper Chesapeake Bay region. Chesapeake
Science 4:90—100.

Nickell, W. P. 1951. Studies of habitats, territory, and nests of the Eastern Goldfinch. Auk
68:447^70.

. 1958. Variations in engineering features of the nest of several species of birds in

relation to nest sites and nesting materials. Botanical Studies, Butler University (Indi-
anapolis, Indiana) 13:121-139.

Ohmart, R. D., B. W. Anderson, and W. C. Hunter. 1988. The ecology of the lower
Colorado River from Davis Dam to the Mexico — United States international boundary:
a community profile. U.S. Fish and Wildlife Service Biological Report 85(7.19).

Ortega, C. P. and A. Cruz. 1991. A comparative study of cowbird parasitism in Yellow-
headed Blackbirds and Red-winged Blackbirds. Auk 108:16-24.

Peattie, D. C. 1991. A natural history of western trees. Houghton Mifflin, Boston.

Rawinski, T. j. 1982. The ecology and management of purple loosestrife (Lyihrum salicaria
L.) in central New York. M.S. thesis, Cornell University, Ithaca, New York.

Reed, P. B., Jr. 1988. National list of plant species that occur in wetlands: national sum-
mary. U.S. Department of the Interior, Fish and Wildlife Service, Biological Report
88(24).

Smith, C. R. 1988. American Goldfinch, Carduelis tristis. P. 496 in The atlas of breeding

186

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

birds in New York State (R. L Andrle and J. R. Carroll, eds.). Led. New York State
Bird Clubs, New York State Dept, of Environmental Conservation, and Cornell Univ.
Laboratory of Ornithology, Ithaca, New York.

Thompson, D. Q., R. L. Stuckey, and E. B. Thompson. 1987. Spread, impact, and control
of purple loosestrife {Lythrum salicaria) in North American wetlands. U.S. Dept, of the
Interior, Fish and Wildlife Service, Fish and Wildlife Research 2.

Walkinshaw, L. H. 1939. Life history studies of the Eastern Goldfinch. Part II. Jack-Pine
Warbler 17:12-21.

Whelan, C. J. and M. L. Dilger. 1992. Invasive, exotic shrubs: a paradox for natural area
managers? Natural Areas Journal 12:109-110.

Erik Kiviat, Hudsonia Ltd., Bard College Field Station, Annandale, New York 12504.
Received 22 April 1994, accepted 22 Sept. 1995.

Wilson Bull., 108(1), 1996, pp. 186-187

Opportunistic winter water acquisition by Pine Grosbeaks. — The sparse documenta-
tion of water acquisition by birds in cold regions is limited primarily to observations of
consumption of water in a frozen form. Methods reported include Pine Siskins (Carduelis
pinus) eating snow. Cedar Waxwings {Bombcilla cedrorum) catching snowflakes during a
storm (both by Dr. Glover Allen as cited in Gordon (1934) and in Allard (1934)), and “Song
Thrush” chipping ice (Harding 1986). Allard (1934) also reported “starlings” eating snow
and catching snowflakes. Other species observed eating snow include the “Redwing (T.
iliacus) and Blackbird (T. merula)” (editors note following Harding [1986]), and Bohemian
Waxwing (B. garrulus) (pers. obs.).

This note documents the opportunistic exploitation of free water droplets in a cold region
in winter by Pine Grosbeaks (Pinicola enucleator). It also notes that this species has the
ability to hover, somewhat like a hummingbird, for short periods.

Lone Pine Grosbeaks were observed on 7 November 1992 and again 20 December 1994
at a site about 24 km northeast of Anchorage, Alaska, in Eagle River Valley (61°19'N/
149°28'W), flying from a cottonwood tree (Populus sp.) perch to hover briefly below an
icicle as a droplet of water formed. The droplet was sipped off the end of the icicle and
then the bird returned to its perch in the tree about 1.5 m away. This process was repeated
5—10 times over a 5-min period. The icicles were forming off the roof of a cabin located
on a south-facing 27° slope. The area is under what Viereck et al. (1992) classifies as an
“open poplar” (I.B.2.C) or “open spruce-poplar” (I.C.2.d) forest and receives its first mea-
surable snowfall in September and is snowfree by mid April (pers. obs). The area is visited
intermittently throughout the winter months by Pine Grosbeaks. Temperatures prior to both
observations varied somewhat (±3°C) but were consistently subfreezing (x = -10°C) at
night (National Weather Service, pers. comm.; NO A A 1992) and at or above freezing (0-
5°C range) during the day at the 610 m elevation observation site. Both observations were
preceded by snowfall of 25 cm (6—7 November 1992) (NOAA 1992) to 46 cm (15—18
December 1994) (pers. obs.). These conditions led to droplet-producing icicles along the
south-facing roof pitch.

Acknowledgments. — I thank A. Carter, Alaska Bird TLC; D. Irons, USFWS; and anony-
mous reviewers for their helpful suggestions on the manuscript. I also appreciate the research
assistance received from University of Alaska— Fairbanks and USFWS library staffs.

SHORT COMMUNICATIONS

187

LITERATURE CITED

Allard, H. A. 1934. How some birds satisfy thirst. Science 80:116-117.

Gordon, S. 1934. The drinking habits of birds. Nature 133:436^37.

Harding, B. D. 1986. Song Thrush chipping ice. Br. Birds. 79:405.

NOAA. 1992. Climatological data, Alaska, November 1992. v. 78, #11. Department of
Commerce, National Oceanic Atmospheric Administration, National Climate Data Cen-
ter, North Carolina.

ViERECK, L. A., C. T. Dyrness, a. R. Batten, and K. J. Wenzlick. 1992. The Alaska
vegetation classification. Gen Tech. Rep. PNW-GTR-286. Portland, Oregon. USDA For-
est Service, Pac. NW Res. Sta.

David F. G. Wolfe, Wildlife Ecologist, P.O. Box 101572, Anchorage, Alaska 99510-1 572.
Received 16 May 1995, accepted 30 Sept. 1995.

Wilson Bull., 108(1), 1996, pp. 187-189

Evidence of nest parasitism in Mottled Ducks. — Intraspecific nest parasitism, which is
the fostering of one or more eggs into the nest of a conspecific, is widespread in waterfowl
(Yom-Tov 1980, Eadie 1991). Intraspecific nest parasitism is most common among cavity
nesting waterfowl and waterfowl that nest in colonies (Rohwer and Freeman 1989). In
contrast, it is rare in solitary, upland-nesting waterfowl, including most Anatini (Eadie et
al. 1988, Rohwer and Freeman 1989), except when they nest in high densities (e.g., Drewien
and Fredrickson 1970, Titman and Lowther 1975, Hines and Mitchell 1984). Dense-nesting
situations may facilitate parasitism by reducing the time, energy, and risk associated with
finding host nests (Rohwer and Freeman 1989). Some authors (e.g., Jones and Leopold
1967, Erskine 1990) have suggested nest parasitism may also occur as a consequence of
nest site competition when waterfowl nest in high densities. However, parasitism persists in
cavity nesting ducks when nest sites are abundant (Semel and Sherman 1986) and evidence
for nest site competition in non-cavity nesting waterfowl is equivocal (Rohwer and Freeman
1989). Intraspecific nest parasitism has been documented for only six species of Anatini
from North America: Northern Shoveler (Anas clypeata). Green-winged Teal (A. crecca).
Cinnamon Teal (A. cyanoptera). Mallard (A. platyrhynchos), American Black Duck (A.
rubripes), and Gadwall (A. strepera) (reviewed in: Eadie et al. 1988, Rohwer and Freeman
1989, Sayler 1992). Here we report the first evidence of intraspecific nest parasitism in the
Mottled Duck (A. fulvigula).

We found 132 Mottled Duck nests during searches of six islands in the Atchafalaya Delta
Wildlife Management Area (29°26'N, 9I°20'W), Saint Mary Parish, Louisiana, during
March through August 1994. When we found a nest, we estimated incubiUion stage (Weller
1956) and individually marked all eggs. Newly laid eggs were marked on subsequent nest
checks and incubation stage was estimated again. Incubation period for Mottled Ducks was
assumed to be 26 days (Stutzenbaker 1988).

We found four cases of apparent nest parasitism. (1) On 8 April, we found a nest con-
taining 12 eggs, which we estimated at 19 days incubation. On 16 April, the nest contained
several recently hatched eggs and one unhatched egg. We opened the unhatched egg, which
contained a 15 day-old embryo (Caldwell and Snail 1974). We believe that this was a non-
term egg (an egg laid after the onset of incubation, Morse and Wight 1969) and not an

188

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

embryo that died during development, because it had not deteriorated. (2) On 16 April, we
found a laying stage nest containing 10 eggs. The nest contained 15 eggs on 29 April. Four
of these eggs clearly differed in color from the rest of the clutch and were estimated to have
been incubated <four days. Egg color difference can accurately distinguish parasitic eggs
in some birds (Lyon 1993) and has been used to identify parasitic eggs in waterfowl (Jones
and Leopold 1967). Egg color difference, together with an approximate eight day difference
in incubation stage, suggests that the four eggs were laid by a hen other than the host. On
12 May, the nest contained 10 ducklings and eight unpipped eggs. Seven of the unhatched
eggs differed in color from shells of the hatched eggs. The large clutch size (18 vs typical
range of 8-13 eggs, Stutzenbaker 1988) was further evidence of nest parasitism. (3) On 7
May. we found a nest containing 12 eggs, most of which had 12-day-old embryos. Three
eggs, however, differed in color from the others and were 5-10 days less developed. On 18
May, there were hatching movements (>23 days of incubation) in all eggs except the three
differently colored eggs. On 25 May, there was evidence of hatched eggs, but no intact eggs
remained in the nest. (4) On 6 July, we found a laying stage nest containing six eggs. It
contained nine eggs on 20 July, one of which was a different color and was unincubated
(<3 days, Weller 1956). On 30 July, the nest still contained nine eggs, several of which had
hatching movements, but the off-color egg’s incubation stage was 1 1 days. An 8 August
nest check revealed evidence of hatched eggs, but two unhatched eggs remained, one of
which was the egg of different color.

All four cases of suspected parasitism occurred on one 22-ha island, where most (N =
82) nests were found. The estimated parasitism rate (minimally 5%) on this island was
similar to other studies of island nesting Anatini (Rohwer and Freeman 1989). We were
unable to estimate nest densities because we did not systematically search islands. However,
during March and April, areas searched on islands were approximately equal and twice as
many nests were found on the 22-ha island as on all other islands combined. Three of the
other islands were >40 ha. The other two were <20 ha and were often flooded. Failure to
detect parasitism on other islands may reflect smaller samples of nests or lower nest den-
sities.

Acknowledgments. — Financial support was provided by the Fur and Refuge Division,
Louisiana Dept, of Wildlife and Fisheries. We are grateful to the Atchafalaya Delta Wildlife
Management Area staff for logistical support provided during field work. R R. Garrettson
and W. L. Hohman provided helpful comments on the manuscript.

LITERATURE CITED

Caldwell, P. J. and A. E. Snart. 1974. A photographic index for aging Mallard embryos.
J. Wildl. Manage. 38:298-301.

Drewien, R. C. and L. F. Fredrickson. 1970. High density Mallard ne.sting on a South
Dakota island. Wilson Bull. 82:95—96.

Eadie, j. McA. 1991. Constraint and opportunity in the evolution of brood parasitism in
waterfowl. Proc. International Ornithological Congress 20:1031 — 1040.

, F P. Kehoe, and T. D. Nudds. 1988. Pre-hatch and post-hatch brood amalgamation

in North American Anatidae: a review ol hypotheses. Can. J. Zool. 66:1709—1721.
Erskine, a. j. 1990. Joint laying in Bucephala ducks — “parasitism” or nest-site competi-
tion? Ornis Scand. 21:52—56.

Hines, J. E. and G. J. Mitchell. 1984. Parasitic laying in nests of Gadwalls. Can. J. Zool.
62:627-630.

Jones, R. E. and A. S. Leopold. 1967. Nesting interference in a dense population of Wood
Ducks. J. Wildl. Manage. 31:221—228.

SHORT COMMUNICATIONS

189

Lyon, B. E. 1993. Conspecific brood para.sitisni as a flexible female reproductive lactic in
American Coots. Anim. Behav. 46:91 1-928.

Rohwer, F. C. and S. Freeman. 1989. The distribution of conspecific nest parasitism in
birds. Can. J. Zool. 67:239-253.

Sayler, R. D. 1992. Ecology and evolution of brood parasitism in waterfowl. Pp. 290-322
in Ecology and management of breeding waterfowl (B. D. J. Batt, A. D. Alton, M. G.
Anderson, C. D. Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, eds.). Univ.
Minnesota Press, Minneapolis, Minnesota.

Semel, B. and P. W. Sherman. 1986. Dynamics of nest parasitism in Wood Ducks. Auk
103:813-816.

Stutzenbaker, C. D. 1988. The Mottled Duck: its life history, ecology, and management.
Texas Parks and Wildl. Dept., Austin, Texas.

Titman, R. D. and j. K. Lowther. 1975. The breeding behavior of a crowded population
of Mallards. Can. J. Zool. 53:1270-1283.

Weller, M. W. 1956. A simple field candler for waterfowl eggs. J. Wildl. Manage. 20:
1 1 1-1 13.

Yom-Tov, Y. 1980. Intraspecific nest parasitism in birds. Biol. Rev. Cambridge Philos. Soc.
55:93-108.

William P. Johnson, School of Forestry, Wildlife and Fisheries, Louisiana Agricultural
Experiment Station, Louisiana State Univ. Agricultural Center, Baton Rouge, Louisiana
70803\ Frank C. Rohwer, School of Forestry, Wildlife and Fisheries, Louisiana Agricul-
tural Experiment Station, Louisiana State Univ. Agricultural Center, Baton Rouge, Louisi-
ana 70803'. AND Michael Carloss, Louisiana Dept, of Wildlife and Eisheries, 2415 Darnall
Road, New Iheria, Louisiana 70560. Received 13 April 1995, accepted / Sept. 1995.

WiLson Bull., 108(1), 1996, pp. 189-190

Eight new host species for the parasitic blow fly genus Protocalliphora (Diptera:
Calliphoridae). — Larvae of Protocalliphora blow flies (Diptera: Calliphoridae) are obligate
hematophagous parasites that reside in nests of birds with nidicolous young where they
intermittently attach to the nestlings to feed. Only one species of Protocalliphora, P. hraueri,
is known to be an obligate subcutaneous parasite (Sabrosky et al. 1989). Protocalliphora
blow flies appear to have little host specificity (Bennett and Whitworth 1992), and, with the
exception of birds whose nest structure is not conducive to blowfly retention and develop-
ment (e.g., loosely arranged stick nests, very wet nests), eventually all nidicolous bird spe-
cies within the range of these blow flies are likely to be recorded as hosts (Sabrosky el al.
1989).

During a three-year study of interactions between Protocalliphora blow flies and Neo-
tropical migratory bird species, bird nests were collected from study plots in Arkansas in
the Ozark National Forest, in 1991, 1992, and 1993, and from the Ouachita National Forest
in 1993. In Idaho, ne.sts were collected during 1992 and 1993 from Targhee National Fore.st.
Nests were located and monitored following the protocols detailed in Marlin and Geupel
(1993). When the nests were no longer active (i.e. after fledging, death, or depredation),
they were collected in plastic bags, taken to the laboratory, and searched for Protocalliphora
larvae and pupae. Larvae were collected from nestlings and fledglings whenever noted. The
larvae and pupae were reared to maturity and identified using the taxonomic key provided

190

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

in Sabrosky et. al. (1989). Voucher specimens were deposited in the Univ. of Arkansas
Museum of Entomology.

I recorded eight new host species for this parasitic genus of blow flies. In Arkansas, P.
deceptor larvae were collected from Acadian Flycatcher {Empidonax virescen.s). Hooded
Warbler (Wilsonia citrina), and Bachman’s Sparrow (AimophUa aestivalis) nests. Larvae ol
P. braueri were collected from the nests of Black-and-white Warbler (Mniotilta varia) and
Kentucky Warbler (Oporornis formosus). Subcutaneous larvae of P. braueri were collected
from Kentucky Warbler nestlings and from a single Yellow-throated Vireo (Vireo flavifron.s)
fledgling. In Idaho, an unknown species of Protocalliphora larvae was collected from Veery
{Catharus fuscescens) nests and P. metallica and an unknown Protocalliphora species from
MacGillivray’s Warbler (Oporornis tolmiei) nests. Unknown species could not be identified
due to damage incurred during transport. These appear to be the first records of Protocal-
liphora parasitism in these bird species.

Acknowledgments. — I thank C. Sabrosky for identifying blow fly specimens. Numerous
graduate students and field assistants collected nests in Arkansas and Idaho. J. Johnson and
the University of Arkansas Cooperative Fish and Wildlife Research Unit provided trans-
portation and laboratory space. T. Martin, L. Garner, S. Garner, and their field assistants
collected nests in Idaho. A special thanks to N. Ball, S. Foster, R. King, and C. Sagers for
thoughtful comments on earlier versions of this manuscript.

LITERATURE CITED

Bennett, G. F. and T. L. Whitworth. 1992. Host, nest, and ecological relationships of
species of Protocalliphora (Diptera: Calliphoridae). Can. J. Zool. 70:51—61.

Martin, T. E. and G. R. Geupel. 1993. Nest-monitoring plots: Methods for locating nests
and monitoring success. J. Field Ornith. 64:507—519.

Sabrosky, C. W, G. F. Bennett, and T. L. Whitworth. 1989. Bird blow flies {Protocal-
liphora) in North America (Diptera: Calliphoridae), with notes on the Palearctic species.
Smithsonian Institution Press, Washington, D.C.

Mia Revels, Dept, of Biological Sciences, Univ. of Arkan.sas, Fayetteville, Arkansas 72701.
Received 20 April 1995, accepted I Sept. 1995.

Wilson Bull., 108(1), 1996, pp. 190-192

Observations of shorebird predation by snapping turtles in eastern Lake Ontario. —

Accounts of snapping turtle (Chelydra serpentina) predation on birds other than waterfowl
are rare. These include Laughing Gull (Larus atricilla) (Alexander 1921), Semipalmated
Sandpiper (Calidris pusilla) and Lesser Yellowlegs (Tringa flavipes) (Street 1989), and the
possible predation of a Forster’s Tern (Sterna forsteri) chick (Fraser 1994).

The present observations were made at a freshwater dune ecosystem at the Nature Con-
servancy’s El Dorado Beach Preserve in Jefferson County, New York. Thick, partially-
submerged algal mats accumulate there annually in the shallow embayments of Lake On-
tario. Large quantities of a green, filamentous algae (Cladophora glomerata) break oft from
underwater rocky substrates when the lake temperature exceeds 25°C, as in late summer
(Vetterle 1976). The resulting offshore algal mats entrap invertebrates, including freshwater
crustaceans, gastropods, and insects and some vertebrates such as small fish. This concen-

SHORT COMMUNICATIONS

191

tration of food organisms is subject to predation by northern water snakes {Nerodia sipedon),
painted turtles {Chrysemys picta), snapping turtles, and many species of birds, especially
migrating shorebirds.

On 7 August 1994 at 16:15 EST, a Semipalmated Sandpiper was observed being pulled
underwater, through an algal mat. The captured sandpiper was peeping loudly with only its
head visible above water. The nearby shorebirds (Semipalmated Sandpipers and Lesser Yel-
lowlegs) also vocalized excitedly. A Lesser Yellowlegs fluttered briefly above the site while
vocalizing frenziedly. A large turtle’s carapace was felt at the immediate location where the
captured sandpiper was last observed; the turtle had presumably consumed the bird imme-
diately. At 12:30 on 8 August 1994 another Semipalmated Sandpiper was pulled under the
same algal mat by a snapping turtle. The bird was repeatedly pulled underwater, but resur-
faced each time with its wings fully extended outward. This reaction made it difficult for
the turtle to pull the bird through the algae. The event was photographed at the site. After
I nudged the turtle’s carapace with my foot, the turtle released the sandpiper. The- bird
quickly flew away, landed nearby, and began preening. The sandpiper shivered and preened
for approximately 10 minutes, at which time it joined a close flock of feeding shorebirds.
There was no visible damage to the bird.

At 12:25 on 11 August 1994 a large snapping turtle was observed crossing the sandy
beach from an inland pond in order to enter the algal mat. The turtle was estimated to weigh
between 10 and 15 kilograms, and had a carapace approximately 35 cm long. The shorebirds
exhibited a distinct pattern of aggression in response to the fully exposed turtle. Three Black-
bellied Plovers (Pluvialis squatarola) “escorted” the turtle closely. A Killdeer (Charadrius
vociferus) stood further back and vocalized. Six Lesser Yellowlegs called and periodically
fluttered in the air above the turtle. The remaining shorebirds (130-135 Semipalmated Sand-
pipers, four Sanderlings [Calidris alba], four Semipalmated Plovers [Charadrius semipal-
matus], and one Spotted Sandpiper [Actitus macularia]) remained in a tight, distant group
until the turtle was submerged under the algae. Within 15 min, the birds resumed normal
feeding activity in the vicinity of the submerged turtle. It was later observed that there were
two snapping turtles under the algal mat.

At 10:00 on 13 August 1994 a Lesser Yellowlegs was observed being pulled under the
algae, with only its head exposed. Three other Lesser Yellowlegs were frantically peeping
and fluttering above the victim. I again waded into the algae and nudged a large snapping
turtle, which released the bird. The yellowlegs flew nearby, preened, and apparently had no
damage done to its body or legs. It soon joined a feeding flock of shorebirds and could be
identified by the algae and duckweed (Lemna sp.) remaining on its undertail coverts and
legs.

It appears that individual snapping turtles can become efficient seasonal predators of
shorebirds. Contrary to popular assumptions, even large snapping turtles did not rely entirely
on strong jaw musculature to capture prey, but instead displayed a furtive hunting tactic.
The delicate legs of those birds that escaped were not visibly damaged by the turtles’ sharp,
heavy mandibles, as might otherwi.se be expected. Populations of shorebirds, which may
seem to be unlikely prey for sluggish snapping turtles, can be reduced during migration
when feeding in areas where the turtles occur. Three confirmed events of snapping turtle
predation on shorebirds, at the same site and within six days, represent a potential impact
on migratory shorebird numbers.

LITERATURE CITED

Alexander, E. G. 1921. Laughing Gull (Larus atricilla) captured by snapping turtles. Auk
38:596.

192

THE WILSON BULLETIN • Vol. JOS, No. J, March 1996

Eraser, G. 1994. Possible predation of a Forster’s Tern chick by a snapping turtle. Prairie
Nat. 26:33-35.

Street, H. M. 1989. Semipalmated Sandpiper captured by turtle. Ont. Birds 7:70.
Vetterle, P. 1976. The importance of the macroinvertebrate benthos in Lake Ontario Cla-
dophora mats. Masters’ thesis, Professional Studies Division, College of Arts and Sci-
ences, SUNY College at Oswego, Oswego, New York.

Gregory S. Pryor, RD# 8 Box 194, Oswego, New Yoric 13126. Received 28 April 1995,
accepted 8 Sept. 1995.

Erratum

The paper entitled “Gray Flycatcher predation on a hummingbird,” in Wilson Bulletin
107:565-567, actually refers to observations of the Gray Kingbird. Although the correct
scientific name is given, the substitution of “flycatcher” for “kingbird” escaped two ref-
erees, a proofreader, this editor, and (apparently) the author.

Wilson Bull., 108(1), 1996, pp. 193-204

ORNITHOLOGICAL LITERATURE

Edited by William E. Davis, Jr.

Arena Birds: Sexual Selection and Behavior. By Paul A. Johnsgard, illus. by the
author. Smithsonian Institution Press, Washington, D.C. 1994: 330 pp., 38 color plates, 70
pen-and-ink illustrations. $39.95 (cloth). — Arena birds are those species that exhibit elabo-
rate courtship behavior in well-defined, often communal areas called arenas. Familiar ex-
amples from North America include many of the grouse species, some duck species, as well
as certain shorebirds such as the Pectoral (Calidris melanotos) and Buff-breasted (Tringites
subruficoUis) sandpipers. Familiar examples from outside of North America include the New
Guinea and Australia birds-of-paradise and bowerbirds, the contingas and manakins of the
Neotropics, and the European and African bustards. Other examples include species in such
diverse groups as the parrots (the Kakapo [Strigops hahroptilus] of New Zealand), the
hummingbirds (the hermit species as well as others), African widowbirds and whydahs, and
the Australian lyrebirds. Obviously, arena courtship has evolved independently in numerous
taxonomically distant bird groups. It is one form of sexual selection, a pervasive evolution-
ary force in animals ranging from arthropods to apes. In this volume, Paul Johnsgard, known
for his prolific popularized treatments of various of the world’s bird families, has again
attempted a broad review, this time not taxonomically oriented but evolutionarily oriented:
arena courtship in birds.

Johnsgard’s book is comprehensive and well referenced. It will prove important to stu-
dents and researchers interested in sexual selection as manifested in arena courtship. A total
of 460 references are included in the literature cited, making this volume clearly the best
available review of this subject as it applies to birds. In addition, students will find the
thorough 26 page glossary most helpful in that many definitions (Bateman’s principle, hand-
icap hypothesis, male dominance polygyny, sexy son hypothesis) are treated as short para-
graphs, providing a very sound and quickly referenced summary of the complex terminology
and conceptual hypotheses that currently pepper the field of sexual selection theory.

The book is divided into 12 chapters, most of which deal with specific taxonomic groups
in which arena behavior is particularly dominant. The first two chapters are introductory,
one to introduce sexual selection as an evolutionary process, the other to focus on arenas,
courts, and leks. Both of these chapters try to provide a sound overview of relevant evo-
lutionary theory, introducing (perhaps too briefly) the various models of sexual selection
and the kinds of data sets that support sexual selection as an explanation for a particular
courtship pattern. These chapters provide an adequate introduction to the field although they
are best read after one has a relatively sound understanding of the principles of natural
selection. Johnsgard gives perhaps an overly concise introduction to Darwin’s thinking in
conceiving of sexual selection (there is only a reference to “Descent of Man and Sexual
Selection,” none to “The Origin of Species,” where the idea was first introduced) as well
as to Alfred Russel Wallace’s views on Darwin’s theory. But one must remember that this
is not a book on all of sexual selection but on only one form of it as shown in some birds.

There are seven tables in chapter 1 that will prove useful in gaining an overview of the
numerous hypotheses that envelop this complex subject: models of sexual selection among
birds, hypotheses that relate to male showiness among normally monogamous birds, hy-
potheses that relate to sexual monomorphism among nonmonogamous birds, costs and ben-
efits influencing female mate choice strategies in lekking, costs and benefits influencing
male clustering strategies, relative individual mating success among males of selected arena
species, and examples of male age-related dominance/fitness ratios in lekking birds. Johns-
gard's discussion is meant to summarize, although he does go into much greater depth when

193

194

THE WILSON BULLETIN • Vol. JOS, No. 1, Inarch 1996

discussing various species treated in the main body of the text. Chapter 2 is brief and also
summary in nature. Johnsgard provides definitions of arenas, courts, and leks and attempts,
as in the preceding chapter, to outline all of the relevant terminology in tabular form.
Included in this chapter is a useful table listing all of the 60 arena bird species, describing
for each its dispersion pattern (lek, exploded lek, mobile lek), color pattern (dimorphic,
nondimorphic, reverse dimorphic), mass pattern (never clearly defined in the table but un-
doubtedly referring to males having greater, equal, or less mass than females), display types
(aerial, ground, joint male, stage or arena, tree or shrub), and a principal reference source
from the ornithological literature.

The remaining 10 chapters deal with specific examples. For example, chapter 9, “Man-
akins: spectacular soloists and dazzling duets,” spotlights the 24 species of lekking Neo-
tropical manakins, some of which have been intensively studied and for which much is
therefore known. Even readers quite familiar with this group will find Johnsgard s treatment
praiseworthy. He discusses well-chosen examples, provides excellent pen-and-ink diagrams
of the behaviors he describes (a very useful feature throughout the book), and summarizes
the essential hypotheses that have been advanced to account for the elaborate behaviors.
Other chapters are similar in approach.

What is missing from the volume is a final chapter that insightfully brings together the
numerous threads contained within the main body of the text. There is no concluding chapter
of any kind. Missing, therefore, are comparisons among taxonomically distant groups, sub-
stantive discussions of convergent evolution, and suggestions about how such diverse taxa
evolved independently into arena birds. There is, for instance, no real comparison of the
ecology of Neotropical species such as the Guianan Cock-of-the Rock (Rupicola rupicola)
with that of the various species of New Guinea birds-of-paradise, two cases in which a
frugivorous diet has been suggested as influential in structuring the birds’ reproductive
ecologies. This quibble aside, anyone interested in this subject cannot but help to find
Johnsgard’s book of great use. It is a most thorough popular review of the literature, greatly
enhanced by numerous well-crafted illustrations. — John C. Kricher.

My Double Life; Memoirs of a Naturalist. By Frances Hamerstrom. University of
Wisconsin Press, Madison. 1994: 316 pp., 72 black-and-white photographs, 42 line draw-
ings. $16.95 (paper); $35 (cloth). — This autobiographical .sketch consists of 90 snippets —
recollections and reminiscences — mostly two-to-four pages in length, which provide a win-
dow into the life of an accomplished naturalist and conservationist. The first 85 pages or
so deal with Frances’s childhood — glimpses into the secret life of a bright, alert, questioning,
secretive, naughty, awful, and sometimes manipulative child — the kind that drives parents
and governesses to distraction but often produces accomplished adults. The descriptions, for
example, of a child of six smoking cigarettes, cutting her new Christmas doll’s hair, con-
ducting a formal funeral for a Blue Jay, or cutting up her mother’s white kid opera gloves
to make jesses for her pet kestrel, are sometimes sad, sometimes humorous, sometimes
poignant. There are undertones of strife against grownups who "forbade wild pets and tried
to squelch my companionship with creepy crawly creatures . . .” She found grownups "ei-
ther weak or not to be trusted.” Another 20 pages deal with the metamorphosis of a tomboy
into a young woman.

The remainder of the book deals largely with the prairie chicken work she shared with
her husband of 59 years, Frederick, with occasional glimpses and hints of raptor research.
We follow their lives in a series of deserted farmhouses in central Wisconsin during the
Great Depression and war yeans — behavioral studies from blinds, harsh winters with frozen

ORNITHOLOGICAL LITERATURE

195

pumps and prairie chicken censuses on snowshoes, graduate school with Aldo Leopold, a
succession of 7000 “boomers,” or helpers, with the chicken work, parenting under primitive
conditions (two children), and local politics and conservation initiatives.

This book is well written, well illustrated, and provides an interesting perspective on
some facets of ornithological research, particularly during the first half of the twentieth
century. What were the two lives of this most interesting ornithologist/naturalist? Read this
delightful book and find out for yourself — I highly recommend it. — William E. Davis, Jr.

Annotated Bibliography of the Loons, Gaviidae. By Judith W. McIntyre and Norma
G. Cutler. North American Loon Fund, Gilford, New Hampshire. 1995: 170 pp. Available
in paperback for $12 plus $2 S&H within U. S. A. from: NALF, 6 Lilypond Road, Gilford,
NH 03246. — This bibliography contains 1650 citations with the most recent 1994. The
introduction lists 44 sub-categories (e.g., acid rain, food, predation) for which separate sub-
bibliographies may be obtained. Anyone purchasing the entire bibliography is entitled to
one free sub-bibliography, and additional ones at cost, ordered from the above address. Not
available on disk. About half of the entries are annotated, and annotations run up to nine
lines of type. This bibliography will be useful for anyone interested in loon biology or
conservation. — William E. Davis, Jr.

Atlas of the Breeding Birds of New Hampshire. By Carol R. Foss (ed.). Audubon
Society of New Hampshire, Dover N. H. 1994. 459 pp., many b/w sketches, 241 maps.
$39.95 (cloth). — This publication from a New England state continues the expansion of the
breeding bird atlas shelf in libraries. The birders of the Granite State have produced a fine
work which matches the high standards of the earlier Atlases.

The book follows the well-known format, with a page of text and a sketch of the species
facing a map occupying a full page. The resulting large maps for a small state are perhaps
the most easily interpreted of any in the atlases I have seen. The atlas blocks were set up
on the standard 6 blocks per 7.5 minute topographic sheet, but as with several other states
having limited manpower only one block per sheet was selected as a “priority block.”

New Hampshire has varied habitat and the second greatest altitudinal range of any eastern
state (from sea level to over 6200 feet.) This variety has been organized under 16 physical
divisions. The Atlas workers found breeding evidence for 204 species with 176 of these
“Confirmed”. Twenty-one other species are listed as “Historically or potentially breeding”
species which were not found during the atlas project. No attempt is made to consider
populations using Breeding Bird Survey data or something similar as other state atlases have
done.

The species accounts, written by a large panel of experts, provide a variety of natural
history information, some it from New Hampshire publications but much of it from standard
sources such as the Bent series. Besides the customary table summarizing the number of
blocks for which “Confirmed”, “Probable”, and “Possible” status was obtained for the
species, another table detailing the number of records meeting each of the criteria for con-
firmation (i.e.. Nest Building, Nest with eggs, etc.) together with the range of dates for each.
This useful information is not included in other atlases.

Besides the species accounts and the usual summary of the geography of the state, there
are two essays that are unique for this atlas. A chapter entitled “Major Changes in Breeding
Avifauna of New Hampshire Since Its First Settlements by Europeans in 1623” gives a

196

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

very lucid historical discussion of the effect of settlement and habitat change, as well as
more recent range changes. Accompanying this is an appendix showing range maps for the
period 1963-1980 of several species. Some of these have greatly disappeared in this short
interval. Another fascinating chapter by Tudor Richards is a discussion of the occurrence
of species along the great altitudinal gradient in the state. Richards attempted to count the
species present at 500 foot intervals at stations throughout the state. He classifys his results
results into seven avifaunal regions. The highest count was 134 species in the interval
between 1000 and 1500 feet, and the lowest count was one species (Dark-eyed Junco) in
the 5500-6000 feet interval.

I have only one minor negative comment. The maps of land elevations and forest types
are done in a 6-interval “gray” scale. It is next to impossible to distinguish between the
three darkest categories.

The field work for this atlas was completed in 1986, and the long delay in publication
seems to be characteristic of the atlas business. Other states have experienced similar delays.
In any event, this is a well-done compilation and welcome addition. — George A. Hall.

A Birder’s Guide to Coastal North Carolina. By John O. Eussell III. Univ. of North
Carolina Press, Chapel Hill. 1994: 540 pp., 10 black-and-white photographs, 44 maps.
$16.95 (paper); $29.95 (cloth). — The purpose of this book is to describe and facilitate vis-
iting the better birding sites along the North Carolina coast. It accomplishes this goal in
fine fashion. The book is divided into three parts. Part 1 consists of three chapters, the first
of which describes the climate, physiography, and habitat of the various sections of the
coast. The second chapter provides general planning and travel advice and introduces birders
to the pleasantries of ticks, fire ants, and poisonous snakes. Chapter 3 is a list of more than
350 species which may be found in the region, with annotations on status, season, and
habitat and sometimes specific site information of from one to nine lines of text; a list of
41 accidentals is included. Part 2 is a site guide consisting of six chapters, five of which
describe (from north to south) sections of the coast and associated tidewater or outer
coastal plain. The sixth deals with pelagic birding. Each chapter has a map of the entire
section of coast described, and detailed maps for particularly interesting sites. There is
usually a section on logistics, and the text guides the reader on a “tour of each site with
distances described to the nearest tenth of a mile, notable birds enumerated, and habitats
described. Telephone numbers, e.g., for pertinent wildlife refuge headquarters, are included,
as are helpful hints about avoiding problems with hunting seasons. The text is highlighted
by ample, common sense advice. For the site 1 had most recently visited (fall 1994), I found
the directions and descriptions excellent; I assume that this is the rule rather than the ex-
ception. Certainly, the text reads smoothly and the directions are clear and detailed. Part 3
consists of more detailed descriptions of status, season, distribution, habitat, and special
features of 141 bird species “of special interest.’ The accounts average a third to half a
page, but some run to two pages. An appendix contains detailed bar graphs describing
seasonal occurrence and status (common, rare, etc.) of all but accidental species. The graphs
are large, detailed, and user friendly, with a status key on each pair ot facing pages.

The book is generally well done, but is not without problems. It is a big book — it won t
lit in your pocket or in most automobile glove boxes. This large size at least partially results
from redundancy (which the author acknowledges) among the annotated bird list (chapter
3), the section on birds of special interest (part 3), and the bar graphs — triple coverage for
141 species. Integrating the annotated lists would probably have been a good idea. Nowhere
in the book is there a map of the entire region to help a reader untamiliar with the region

ORNITHOLOGICAL LITERATURE

197

put into a visual context the physiography and habitat descriptions in the opening chapter.
But despite these few detracting features, 1 would recommend the book and consider it
indispensable for anyone planning to visit the North Carolina coast. — Wili.iam E. Davis, Jr.

The Birds of Kentucky. By Burt L. Monroe, Jr. Indiana Univ. Press, Bloomington,
Indiana. 1994: 145 pp., numerous color plates. $49.95. — “The birds of Kentucky” is an
attractive “coffee table” book describing all birds known to have occuned in that state. The
late author was the dean of Kentucky ornithologists and one of the grand men of our science.
However, anyone looking for new or comprehensive information about the birds of Kentucky
will be disappointed by this book. Why such beautiful tomes are done with such thin detail
continues to bewilder me. On the other hand, the illustrations in the book are outstanding.
William Zimmerman is one of the best illustrators of living birds in the world and the
figures in this book demonstrate his abilities. It is unfortunate that not all species were
illustrated. The reference section is comprehensive and well done. I wish the rest of the
book was as thorough in its coverage. — C. R. Blem.

The Chronicles of the Rowleys. By Peter Rowley. Huntingdonshire Local History So-
ciety, Huntingdon, Cambridgeshire, England. 1995: 158 pp. Available in U.S.A. from Peter
Rowley, 815 Park Avenue, New York, New York 10021. $25 (cloth). — This book provides
glimpses of English life for about a century (ca 1780-1880) from the perspective of one
family, the Rowleys. People with such phonic names as Owsley Rowley stride across the
stage of English history. All of the chapters should be of interest to those with a general
fascination for history, but one chapter, “The Rigid Squire and the Eclectic Ornithologist”
(25 pages), is a biographical account of ornithologist Dawson Rowley (1822-1878). Daw-
son, under the influence of famed ornithologist Alfred Newton, compiled two massive vol-
umes on the Great Auk (Pinguinus impennis) which, although never published, are appar-
ently still extant. After Rowley abandoned the Great Auk project, he turned his attention to
the publication of three volumes of Ornithological Miscellany, published in parts as a mag-
azine beginning in 1875 and before his death in 1878. These issues apparently included
hand-colored lithograph plates by John Gerard Keulemans.

The book is a scholarly work based on primary sources (mostly correspondence) and
should appeal to those interested in the history of ornithology. — William E. Davis, Jr.

Before the Echo: Essays on Nature. By Pete Dunne. Univ. Of Texas Press, Austin.
1995: 152 pp., 20 line drawings. $19.95 (cloth). — In contrast to his previous books of essays,
which dealt mostly with birds, this collection of 30 essays deals largely with other facets
of the natural world. Most of the essays were first published in the New Jersey edition of
the Sunday New York Times. The essays treat a variety of rather mundane subjects, such
as the first snowfall of the year, confrontations with mice in an old farmhouse, or burning
leaves in the fall. My favorite, “Before the Echo,” deals with hunting, and is not a polemical
anti-hunting statement but rather a consideration of hunting as a natural phenomenon. Many
of the essays center on poignant reminisces of youth (e.g., catching fireflies), and most have
a mildly polemical, but not offensive tone (after all, Pete Dunne is an environmentalist who
has worked most of his professional life with the New Jersey Audubon Society). He does
go after an occasional Sacred Cow, such as the grass lawns of suburbia.

198

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

Pete Dunne has become one of the premier natural history writers of North America, and
his essays are beautifully written. You may not always agree with his biases, but the stories
he tells and the yarns he spins are delightful. Even though there isn’t much about birds, I
would recommend reading this book to anyone interested in the interaction of people with
the natural world. — William E. Davis, Jr.

Birds of the Cayman Islands. Revised edition. By Patricia Bradley, photography by
Yves-Jacques Rey-Millet. Caerulea Press, Italy. 1995: 261 pp., 77 color photos, 4 maps. No
price given (cloth). — This excellent little guide has been substantially revised, and the status
of individual bird species updated, with several new species added and the status of many
changed. The photographs are much the same as in the first edition (1985), but there have
been a few replacements and the number of habitat photos increased. In the copies I ex-
amined, the new edition photos were not as sharp as those of the first edition, but the color
seemed to be more realistic, at least in some photographs. The maps have been redrafted
and updated, and the bibliography has been considerably expanded. A new appendix pro-
vides a check-list of breeding birds with status and distribution for each of the three islands.
Most of the minor errors mentioned by Jon Barlow in his thorough review of the first edition
in The Wilson Bulletin (1987, 99:512-514) seem to have been corrected, although curiously,
the subspecific designation bairdi for the local Jamaican Oriole {Icterus leucopteryx) is still
misspelled.

This book should be indispensable to anyone visiting the Cayman Islands and remains
an important component of Caribbean ornithological literature. — William E. Davis, Jr.

Swifts. By Phil Chantler and Gerald Driessens. Pica Press, East Sussex, U.K. 1995:237
pp. £ 26.— This field guide is similar to others on seabirds, waterfowl, sparrows, warblers,
and shorebirds published by Houghton-Miffiin and Princeton Univ. Press. As such, this
volume becomes an important addition to the library of those interested in knowing and
identifying the birds of the world. The present book carries on the tradition of the others in
that it is considerably more than just a field guide. The book contains a rich variety of
infoimation about swifts. It provides maps of the distribution of the swifts of the world,
gives detailed descriptions of status, relative abundance, migration, breeding, adverse factors,
habits and habitat, and describes problems of identification. I greatly enjoyed the sections
on swift biology, the possibility of new (undescribed) species, conservation, and “how to
watch swifts.” The reference section is comprehensive and modern. The binding and format
are excellent and the book is attractive. — C. R. Blem.

Contributions to the History of North American Ornithology. William E. Davis, Jr.
and Jerome A. Jackson (eds.). Memoirs Nuttall Omithol. Club. No. 12. Cambridge MA.
1995: vii +501 pp., many b/w photos. $40 (cloth) — This book is an outgrowth of a sym-
posium on the history of North American ornithology given at the 1991 meeting of the
Association of Field Ornithologists. The editors gathered the material of that symposium
together with a number of invited chapters into an engaging and valuable preamble to a
definitive history. The history of American ornithology is a neglected subject, and many
current workers show a lack of interest in it, possibly because they lack any knowledge of

ORNITHOLOGICAL LITERATURE

199

it. A history was contemplated as a part of the Centennial celebration of the AOU but this
did not eventuate.

The bulk of the book consists of 1 1 chapters detailing the development of ornithology at
as many noted centers of ornithological research; 7 museums and 4 universities (also with
museums.) Since the authors of these accounts are usually noted and important participants
in their subject institutions, the result is a set of fascinating stories. Thus we have the
Academy of Natural Sciences of Philadelphia (by Gill); U.S. National Museum of Natural
History (Banks); Harvard’s Museum of Comparative Zoology (Barrow); Univ. of Kansas
(R. Johnston); American Museum of Natural History (W. Lanyon); Field Museum
(Lowther); Carnegie Museum of Natural History (Parkes); Univ. of California’s Museum of
Vertebrate Zoology (Johnson); Cornell Univ. (Butcher and McGowan); National Museum
of Canada (Ouellet); and the Royal Ontario Museum (Barlow).

The format and content of these chapters varies from institution to institution. Some are
straightforward historical summaries, but some include delightful anecdotes about people
whom we all know by reputation, but who have lived before our time. Most of the accounts
detail the major contributions of the institution, and some itemize the strength of their
collections. The MVZ account contains a series of short biographies of many of the people
associated with that museum, and also includes an “academic genealogy” of Joseph Grinnell
involving a list 50 distinguished present-day ornithologists who are “academic grandchil-
dren”. The Cornell and Kansas Universities each has a list of the doctoral degrees granted
in ornithology. The problems of the present day come into view in Henri Ouellet’s low
keyed description of the regrettable situation at the Canadian National Museum.

Most of the accounts are illustrated by excellent photographs. These range from formal
studio shots, through group photos of departmental staffs, to informal shots in the field,
highlighted by one of a group from the MVZ taking a bath in a Mexican stock tank at the
end of an expedition. One impressive group photo taken in 1984 shows 40 people associated
with the University of Kansas. I was most interested to see photos of Annie Alexander, C.
E. Hellmayr, J. T. Zimmer, and Witmer Stone. (If the reader cannot identify these people,
he will learn much from reading this book.) It is also interesting to see youthful photos of
some of the “Elders of Our Clan”, whom we perhaps knew only in their declining years.

In addition to these chapters there is a chapter on ornithological research within the U.S.
Forest Service (by R. Conner) and one on the history of Canadian ornithology (by M. G.
Ainley).

The three final chapters are more general. Edward Burtt and Alan Peterson discuss “Al-
exander Wilson and the Founding of North American Ornithology.” Besides a brief outline
of Wilson’s life there is a detailed discussion of Wilson’s contribution to the taxonomy of
North American birds. Wilson claimed to have first described 51 species, but the AOU
Check-list today gives him credit for only 20. Burtt and Peterson provide a table correlating
Wilson’s taxonomy with that of the present day. The authors then discuss in turn their ideas
of Wilson’s contributions to aspects of ornithology unrelated to taxonomy, his contributions
to nature writing, and his influence on bird illustration.

In a chapter derived from a paper given at the 1990 WOS meeting Fran9ois Vuilleumier
and Allison V. Andors describe the “Origin and Development of North American Avian
Biogeography.” They argue convincingly that this history is a good example of Kuhn’s
“paradigm” model of scientific progress.

The final chapter by Davis and Jackson is an annotated listing of “The Literature of the
History of North American Ornithology.” About 300 references are discussed in a classified
fashion, in which they hope to guide to researchers interested in the history.

In summary, this is an excellent beginning. I recommend it to all with no reservations.
It may be a selfish thought to want more of a good thing, but I for one do. The reader can

200

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

think of several institutions (perhaps his own) which are not included. Other subfields of
ornithology could proht by having historical reviews done. The present editors hint that
they might prepare a second volume. Indeed, Dr. Jackson has hinted to me that they might
eventually attempt a full scale history. Let us hope that they do. — George A. Hall.

A Naturalist in Indian Territory: The Journals of S. W. Woodhouse, 1 849-5Q Edited
and annotated by John S. Tomer and Michael J. Brodhead. University of Oklahoma Press,
Norman, Oklahoma. 1992: 304 pp., 26 figs., 4 maps. $29.95 (cloth). — In 1849 and 1850 —
over two decades after the Creeks had been forcibly resettled into Indian Territory (now the
state of Oklahoma) — the Corps of Topographical Engineers finally surveyed the northern
and western boundary of their new homeland. The individual selected to serve as surgeon-
naturalist for the expedition was Samuel Washington Woodhouse (1821 — 1904), an assistant
resident physician at the Philadelphia Hospital. Woodhouse had first become interested in
natural history as a teenager. Soon he discovered the Academy of Natural Sciences of
Philadelphia, where he spent long hours pouring through the specimen collections, inter-
acting with some of the leading naturalists of his day, and learning the rudiments of orni-
thological theory and practice. A lack of employment prospects prevented the young man
from seriously considering the idea of ornithology as a vocation. Instead, Woodhouse de-
cided to pursue the more traditional careers of farming and medicine. However, he did not
completely lose his taste for natural history, and when the opportunity to join the Creek
boundary survey presented itself, the single, twenty-seven-year-old doctor quickly seized it.

By the mid-nineteenth century, the Creek Nation contained numerous scattered settle-
ments, and Woodhouse was the first naturalist to undertake a systematic scientific inventory
of the area. The expedition’s slow pace and relatively light medical caseload allowed him
ample time to examine the flora and fauna of the region. After two seasons in the field,
Woodhouse brought back over two thousand specimens, including 15 animal forms which
proved new to science. Among these were a new kind of Mourning Dove (Ectopistes niar-
ginella Woodhouse, now Zenaida macroura morginella). In 1 849, Woodhouse also collected
the skin of the previously undescribed Prairie Ealcon (Falco mexicanus), but he failed to
gain credit for the discovery because he carelessly mistook the specimen for a Peregrine
Ealcon (F. peregrinus). One year later, before anyone caught the mistake, the German or-
nithologist Hermann Schlegel named and described the Prairie Ealcon based on another
specimen taken in Nuevo Leon, Mexico. The editors of the volume at hand were able to
track down this and many other interesting episodes associated with the expedition by
examining surviving specimens at the Academy of Natural Sciences of Philadelphia, the
National Museum of Natural History, the Museum of Comparative Zoology, and other
natural history repositories.

In addition to many valuable specimens, Woodhouse also brought back three private
journals which Tomer and Brodhead have carefully edited for publication. These documents
contain fascinating descriptions of the survey party's experiences in the field, detailed ac-
counts of medical practice on the frontier, as well as invaluable observation.s about the
environment and people of the Creek Nation in the mid-nineteenth century. A lengthy
introduction and careful annotations help to illuminate and place into broader context Wood-
house’s sometimes cryptic entries. Included in the introduction are a thorough (though oc-
casionally tedious) summary of early western exploring expeditions, a brief history of the
Creek boundary expedition, an examination of the significance of Woodhouse’s natural
history work in Indian Territory, and a sketch of his life. The annotations run the gamut
from discussion of current scientific names of plants and animals mentioned in the text to

ORNITHOLOGICAL LITERATURE

201

identification of the people, places, and objects Woodhouse encountered on his journey.
Tomer and Brodhead have done an excellent job of making these important journals acces-
sible to a broad audience, although this reader would like to have seen a few more of
Woodhouse’s fine field sketches reproduced.

Shortly after his return from Indian Territory, Woodhouse returned to the field as a mem-
ber of the Zuni and Colorado rivers expedition (1851 — 1852) and a private expedition to
Central America (1853). For reasons that are still unclear, he devoted the remainder of his
life to professional medical practice. Although his collections and publications were signif-
icant, they were soon overshadowed by the work of the well-funded and highly publicized
Pacific Railroad Surveys of the 1850s. Woodhouse and his scientific contribution faded into
obscurity. Tomer and Brodhead have done a great service in resurrecting the legacy of this
important but forgotten naturalist. — Mark V. Barrow, Jr.

The Eastern Screech-Owl; Life History, Ecology and Behavior in the Suburbs and
Countryside. By Frederick R. Gehlbach, Texas A & M Univ. Press, College Station, Texas.
1994: 302 pp., a color frontispiece with no caption, ten chapters, 34 black-and-white pho-
tographs, 27 tables, 25 figs., $45.00 (hardcover). — This book is a personal narrative of 25
years of Gehlbach’s studies of Eastern Screech-Owls {Otus asio) in central Texas. For eleven
of those years (1976 to 1987), Gehlbach compared the life history of suburban and rural
populations. Gehlbach’s passion for these owls and for natural history in general surfaces
throughout the book.

In chapter one — “On Studying Screech Owls,” Gehlbach outlines the beginnings of
what he terms the exploratory (1967 to 1975), or trial and eiTor period, and the confirm-
atory period (1976 to 1991). There is also a brief overview of statistical procedures.
Chapter two — “Landscapes,” a straight forward description of the suburban and rural
habitats — vegetation, nesting and roosting environments. Chapter three — “Food Supplies
and Predation.” Food niche, prey species, seasonality, and prey body mass are all cal-
culated from cached prey, but no pellet analysis was conducted. Gehlbach provides data
on hunting tactics and periodicity from direct observations. The brief discussion on mob-
bing is interesting, as Gehlbach convincingly correlates resident mobbers with their po-
tential as prey. Chapter four — “Adult Weight, Coloration and Molt.” Comparative weights
of 80 males and 166 females provide ample samples for some of Gehlbach’s analysis. In
particular, the weight dynamics seem to be tied to weather and food fluctuations. There
is a lengthy discussion on red and grey color morphs and speculation as to why. The molt
data, although brief, seems adequately covered. Chapter five — “Eggs and Incubation.”
This chapter chronicles the start of the nesting season from egg laying and replacement
clutches through clutch and egg sizes, laying intervals, incubation duration, and hatching.
Of particular note, is the re-mating and re-laying by one female whose mate was killed
by a car while she incubated one egg. She abandoned the egg after six days, re-mated
three days later and re-nested about 12 days after that. Chapter six — “Chicks and Fledg-
lings.” In his longest chapter, Gehlbach uses male roosting proximity to nests as a barom-
eter of egg hatching, brooding and fledgling. Hatch dates, sequences, brooding, nest mi-
croclimate, parental responsibility, growth rates, mortality, nest-cavity symbiosis, fledging
and dispersal are all well covered. Two bits of information I found particularly interesting
were; (1) “coincidental hatching” in which the typical asynchronous hatching of most
owls was complicated by the fact that 13 (23.6%) and eight (14.5%) of 55 clutches,
respectively, had two and three eggs respectively hatch within a 24 hour period (p. 105);
and (2) “nest-cavity symbiosis”, in which Eastern Screech-Owls bring live Texas blind

202

THE WILSON BULLETIN • Vol. 108, No. 1, March 1996

snakes {Leptotyphlops dulcis) back to the nest cavity and these snakes apparently eat
insects that might compete with nestlings for stored food items. In fact, young in nests
with these snakes grew faster and had better survival than nests without them. Gehlbach
concludes this is probably coincidental and not selective. In any case it is very intriguing.
Chapter seven — “Vocalizations.” An overview of song types, hoots, barks, non-vocal
sounds, and juvenile development is given. Gehlbach describes in which context and
chronology the various sounds are produced. This perhaps is the weakest chapter. Chapter
eight — “Lifetime Reproduction.” Discussions on age and size, nest sites, mates, recruit-
ment, and inheritance are provided. Most of the data, however emphasizes females because
sample sizes were to small for males. Gehlbach summarizes this chapter by stating that
about half the females each season are yearlings, of which about half disappear after their
first attempt at breeding. Those yearlings that reproduce successfully are usually larger
and tend to show site fidelity in following years. Life long monogamous pair bonds are
the general rule, with few polygynous relationships. Interestingly, females began to have
smaller clutches and fledged fewer young after about age five. Chapter nine — “Population
Structure and Elux.” Topics include age classes and survival, productivity, use of space,
densities and cycles. This is a good overview of Gehlbach’s marked individuals and com-
parative data between the suburban and rural populations are given. Survival is better in
the suburbs, but surprisingly these owls have short life spans, except in a few cases.
Breeding densities seemed to fluctuate in relation to environmental factors and predators,
and Gehlbach concludes that the suburban owls exhibit a nine year cycle, similar to the
9.3 year lunar cycle. Chapter ten — “The Suburban Advantage. Gehlbach sums up the
advantages that his suburban screech owls have over rural screech owls. The sections are
pre-adaptations of Eastern Screech-Owls, connections, and prescriptions for the future.

Throughout the entire book, Gehlbach has carefully compared many aspects of these owl’s
life history. In short, the suburban birds seem to have an easier life than the rural birds.
Whether this is really important for the screech-owl population of central Texas is for the
reader to decide. In any case, this small owl seems well-adapted to a wide range of habitats
and habitat modification due to people.

Although it appears that the book was intended to be readable for all audiences interested
in natural history, it is not. It is a technical book. The book is beyond the interest of average
bird watchers or naturalists. Even the serious researcher must review each page carefully to
fully understand the message. Gehlbach’s continuous emphasis throughout the text concern-
ing the exploratory period gets a bit old. I agree that it often takes a few years to become
familiar with the organism one is studying, but I believe that an eight-year exploratory
period is not necessary. I also would have liked to see more reference to Van Camp and
Henny (1975, The screech owl; its life history and population ecology in northern Ohio.
North American Fauna 71 ). This was also a long-term study that could have provided a lot
of comparative data to Gehlbach’s work. Gary Ritchison and his graduate students have also
published a number of good papers on Eastern Screech-Owl vocalizations and dispersal in
Kentucky, but deserved more recognition. Perhaps less comparisons with Boreal Owls {Ae-
golius fiinereus) and more with other species of Otus also would have been preferable.

Should you buy this book? Absolutely. There is an incredible amount of information
throughout the book. I recommend this book for anyone .studying birds. Gehlbach’s 25 years
of screech-owl watching have given him much time to think ol various ways to analyze
data. There are many good ideas for other species. Although one may not agree with his
methods or conclusions, these types of long-term data are rare. Although the book at $45.00
is expensive, it has a hard cover, includes a great deal of inlormation and will last a long
time — buy it! — Denver W. Holt.

ORNITHOLOGICAL LITERATURE

203

The Birds of Nigeria. By J. H. Elgood et al. British Ornithologists’ Union Check-list
No. 4 (second edition). 1994. 306 pp., hardback, 7 figs., 10 tables, gazetteer, 16 color plates
of habitat and 48 of Nigerian birds. £21.00 (UK), £23.00 (overseas) including postage from
British Ornithologists’ Union, % The Natural History Museum, Akeman Street, Tring, Herts.
HP23 6AP, UK. — Nigeria is among the ten most populous nations on earth, with a size and
human density comparable to that of Pakistan. It is hardly a premier destination either for
ecotourists or ornithologists. In 1976, the British Ornithologists’ Union began to publish
check-lists of avifaunas for little-known countries and regions of the world. “The Birds of
Nigeria” first appeared in 1981, compiled by J. H. Elgood, a former professor of zoology
( 1949 to 1965) at Ibadan University in that West African nation, who described the country’s
only endemic bird species, the Ibadan Malimbe {Malimhus ibadanensis). When the first
edition of “The Birds of Nigeria” sold out ten years after publication, the B.O.U. prevailed
upon its author to spearhead its revision and updating, which happily appeared in his 85th
year.

As a national check-list, this work is exemplary. Not only does it contain an annotated
list of about 900 species reported in the country to date (884 being admitted to tbe official
list), but it also has excellent summaries of Nigeria’s environment — its topography, geology,
climate, weather, and vegetative zones. Migration and breeding both are insightfully cate-
gorized and summarized. It has a list of names of people who have recorded birds in the
country, several maps plus an excellent gazetteer, a compilation of banding recoveries (none
intra-African), and a comprehensive bibliography. In short, I found everything one might
hope for in a work of this nature.

The color photos, including an attractive cover photo of a Red-throated Bee-eater (Merops
hullocki), plus a lovely color painting of the endemic Ibadan Malimbe by Martin Woodcock,
all are welcome enhancements to this second edition. Other changes since the first edition,
in addition to adding and updating species accounts and various summaries, include the
renaming and reordering of the avifauna to conform largely to “The Birds of Africa” series
(“BoA", Academic Press 1982 ff). Appendices list the numerous changes to both scientific
and English names between editions. BoA was a sensible choice for a taxonomic and no-
menclatural model, but unfortunately it is not yet complete; the four volumes published to
date cover only slightly more than 60% of Nigeria’s species. Elgood and his team had access
to some BoA work in progress, but for species towards the end of the systematic order, parts
of the original nomenclature, including such quirky English names as Exclamatory Paradise
Whydah {Vidua interjecta), can be found. It might have been better to have followed the
well-researched taxonomy and nomenclature of Dowsett and Forbes- Watson’s “Cbecklist of
Birds of the Afrotropical and Malagasy Regions,” Vol. I (Tauraco Press 1993), cited in the
bibliography, for the remaining species.

If there is any overall problem with this work, it involves the authors’ difficulty in re-
stating the present status of Nigeria’s birds after considering the massive changes in the
Nigerian environment since Elgood left the country some thirty years ago. Much of the
book is based on data gathered by him and several essentially contemporaneous expatriates,
updated whenever possible by more recent material. Sadly, the latter is at best spotty. Details
of evidence (existence and location of cataloged specimens, photographs, sight report doc-
umentation, etc.) to support many species’ stated status also would have enhanced the check-
list’s authority. A more conservative assessment might have admitted somewhat fewer spe-
cies to the official list, and more cunent information if available might have decreased the
stated distribution and abundance of many species. In relation to this work’s overall u.se-
fulness, however, these comments are minor.

When asked to update his original section on Nigerian vegetation, Ronald Keay revisited
the country and reported wryly, “There is no vegetation left in Nigeria!” With most local

204

THE WILSON BULLETIN • Vol. 108, No. I, March 1996

biological work to date having been undertaken by foreigners, the importance of a work
such as “The Birds of Nigeria” cannot be understated. It perhaps was not coincidental that
the Nigerian Conservation Eoundation (NCE) was formed locally shortly after the first edi-
tion was published. The NCE has been instrumental in establishing the framework for sound
national conservation policy. This second edition gives a concise statement of the present
conservation situation. One can hope that now expanded, improved, and more attractive,
this new edition will inspire more direct local interest in the country’s birds and their
protection.

B.O.U.’s support of basic efforts like compiling check-lists worldwide may have more
ultimate impact on protecting the earth’s birdlife than any other modest single measure I
can think of. I salute both the B.O.U. and the authors of “The Birds of Nigeria,” as should
we all. At less than $40 in hardback, this book is a worthwhile bargain for any student of
West African avifaunas. Its purchase also further encourages the B.O.U. to continue its fine
program. — P. William Smith

ADDENDUM TO “BIRDS OF CONIFEROUS FOREST ON
MOUNT GRAHAM, ARIZONA,” Wilson Bulletin 107(4):7I9-722

Audio evidence recorded by the author, Joe T. Marshall, and engineered by Michael A.
Wascher, is available gratis as a 66-min stereo recording to accompany this article. The 17
tracks are arranged from lower to higher elevations along Swift Trail, Mount Graham. Be-
sides nearby birds, other species heard in the background at each altitudinal station are listed
in an accompanying printout. Send written or electronic mail requests stating your scientific
or other interests in Mt. Graham birds to Joe T. Marshall, National Museum of Natural
History, Room 378, Washington, D.C. 20560-0111 (email: mnhvzll3@sivm.si.edu). Indi-
cate whether your preference is to borrow a digital audiocassette, borrow a CD, or receive
one of a limited supply of analog audiocassettes.

This i,ssue of The Wilson Bulletin was published on 1 March 1996.

The Wilson Bulletin

Editor CiiAKl.i'.s R. Bi.km

Department of Biology
Virginia Commonwealth University
816 Park Avenue
Richmotul, Virginia 23284-2012

Assuitonl Editors Leann Bi.EM

Ai.beut E. Conway

Editorial Board Kath> C. Bkai,

Bichahi) N. Connek
Thomas M. Ha(;(;ehtv
John A. Smai.ewood

Review Editor WiEEIAM E. Davis, Jh.

127 East Street

Eoxboro, Massachusetts 02035

Index Editor Kathy G. Beai,

616 Xenia Avenue
Yellow Springs, Ohio 45387

SUCGESTIONS TO AUTHOUS

See Wilson Bulletin. 107:574—575, 1995 for more detailed “Information for Authors.”
Manuscripts intended for publication in The Wilson Bulletin should be submitted in trij)licate,
neatly typewritten, double-spaced, with at least 3 cm margins, and on one side only of good
quality white paper. Do not submit xerographic copies that are made on slick, heavy paper. Tables
should be typed on separate sheets, and should be narrow and deep rather than wide and shallow.
Eollow the AOU Check-list (Sixth Edition, 1983) insofar as scientific names of U.S., Canadian.
Mexican, Central .American, and West Indian birds are concerned. Abstracts of major papers
should be brief but quotable. In both Major Papers and Short Communications, where fewer than
5 papers are cited, the citations may be included in tbe text. Eollow carefidly the style used in
this issue in listing the literature cited; othenvise, follow the “CBE Style Manual” (AIBS, 1983).
Photographs for illustrations should have good contrast and be on glossy paper. Submit prints
unmounted and attach to each a brief but adequate legend. Do not write heavily on the backs of
jihotograpbs. Diagrams and line drawings shouki be in black ink aiul ibeir lettering large enougb
to permit reduction. Original figures or photographs submitted must be smaller than 22 X 28 cm.
Alterations in copy after the type has been set must be charged to the author.

NoTtCE OE CtlANGE OE Al)l)tiE.S.S

If your address changes, notify the Society immediately. Send your complete new address to
Ornithological Societies of North America. P.O. Box 1897. Lawrence. KS 66044-8897.

The permanent mailing address of the Wilson Ornithological Society is: c/o The Museum of
Zoology. The University of Michigan, Ann Arbor. Micbigan 48109. Persons having business with
any of the officers may address them at their various addresses given on the back of tbe front
cover, and all matters pertaining to the Bulletin should be sent directly to the Editor.

MEMItEti.SIlIt’ lN(,)littilE.S

Membership inquiries should be sent to Dr. John .Smallwood. Dept, of Biology, Montclair State
Univ.. Upper Montclair. New jersey 07043.

CONTENTS

MAJOR PAPERS

A NEW SPECIES OF EMERALD HUMMINGBIRD (TROCHILIDAE, CHLOROSTILBON) FROM THE SIERRA DE
CHIRIBIQUETE, SOUTHEASTERN COLOMBIA, WITH A REVIEW OF THE C. MELLISUGUS COMPLEX

F. Gary Stiles

REPRODUCTION AND MOVEMENTS OF MOUNTAIN PLOVERS BREEDING IN COLORADO

Fritz L. Knopf and Jeffery R. Rupert

TRIGEMINAL REPELLENTS DO NOT PROMOTE CONDITIONED ODOR AVOIDANCE IN EUROPEAN STAR-
LINGS. Clark

NEST-SITE SELECTION BY HOODED WARBLERS IN BOTTOMLAND HARDWOODS OF SOUTH CAROLINA

John C. Kilgo, Robert A. Sargent, Brian R. Chapman, and Karl V. Miller

CHANGE IN BODY MASS OF FEMALE COMMON GOLDENEYES DLIRING NESTING AND BROOD REARING

Michael C. Zicus and Michael R. Riggs

INTERSPECIFIC VARIATION IN THE CALLS OF SPHENISCUS PENGUINS

Nina N. Thumser, Jeffrey D. Karron, and Millicent S. Ficken

THE BREEDING BIOLOGY OF THE WILLOW TIT IN NORTHEASTERN SIBERIA

Vladimir V. Pravosudov and Elena V. Pravosudova

CENSUSING WINTERING POPULATIONS OF SWAINSON’S WARBLERS: SURVEYS IN THE BLUE MOUNTAINS
OF JAMAICA Gary R. Graves

COLONY-SITE AND NEST-SITE USE BY COMMON CRACKLES IN NORTH DAKOTA

H. Jeffrey Homan, George M. Linz, William J. Bleier, and Robert B. Carlson

EVIDENCE OF DUAL BREEDING RANGES FOR THE SEDGE WREN IN THE CENTRAL GREAT PLAINS

Paul A. Bedell

DIETS OF NORTHERN PYGMY-OWLS AND NORTHERN SAW-WHET OWLS IN WEST-CENTRAL MONTANA

Denver W. Holt and Leslie A. Leroux

EFFECTS OF EGG TYPE ON DEPREDATION OF ARTIFICIAL GROUND NESTS

Richard H. Yahner and Carolyn G. Mahan

FOOD AVAILABILITY AND FEEDING PREFERENCES OF BREEDING FULVOUS WHISTLING-DUCKS IN LOUIS-
IANA RICEFIELDS - William L. Hohman, Timothy M. Stark, and Joseph L. Moore

SHORT COMMUNICATIONS

DO STANDARDIZED BROOD COUNTS ACCURATELY MEASURE PRODUCTIVITY? -

John M. Marzluff and Mary McFadzen

COMPARATIVE FORAGING BEHAVIOR OF SYMPATRIC SNOW GEESE, GREATER WHITE-FRONTED

GEESE, AND CANADA GEESE DURING THE NON-BREEDING SEASON

Dale E. Gawlik and R. Douglas Slack

SURVIVAL OF RADIO-COLLARED NESTLING PUERTO RICAN PARROTS --

J. Michael Meyers, Wayne J. Arendt. and Gerald D. Lindsey

NEW NESTING AREA OF PUERTO RICAN PARROTS — J- Michael Meyers

NEOTROPICAL MIGRANTS IN MARGINAL HABITATS ON A GUATEMALAN CATTLE RANCH —

Rodney B. Siegel and Marco V. Centeno

UNGULATE ECTOPARASITE REMOVAL BY BLACK CARACARAS AND PALE-WINGED TRUMPETERS
IN AMAZONIAN FORESTS - Carlos A. Peres

NOTES ON THE STATUS AND BEHAVIOR OF THE SWAINSON’S WARBLER IN CUBA -

Arturo Kirkconnell, George E. Wallace, and Orlando H. Garrido

COMMENTS ON A PROBABLE GYNANDROMORPHIC BLACK-THROATED BLUE WARBLER

Gary R. Graves, Michael A. Patten, and Jon L. Dunn

RUFOUS CROWN FEATHERS ON ADULT MALE TENNESSEE WARBLERS

James A. Dick and Ross D. James

AMERICAN GOLDFINCH NESTS IN PURPLE LOOSESTRIFE Erik Kiviat

OPPORTUNISTIC WINTER WATER ACQUISITION BY PINE GROSBEAKS David J . G. Wolfe

EVIDENCE OF NEST PARASITISM IN MOTTLED DUCKS

William P. Johnson, Erank C. Rohwer, and Michael Carloss

EIGHT NEW HOST SPECIES FOR THE PARASITIC BLOW FLY GENUS PROTOCALLIPHORA (DIPTERA:
CALLIPHORIDAE)-. Kevels

OBSERVATIONS OF SHOREBIRD PREDATION BY SNAPPING TURTLES IN EASTERN LAKE ONTARIO
Gregory S. Pryor

1

28

36

53

61

72

80

94

104

115

123

129

137

151

154

159

164

166

170

175

178

181

182

186

187

189

190
193

ORNITIIOIXMSICAL LITERATURE

ThcWsonBulktin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. 108, NO. 2 JUNE 1996 PAGES 205-396

(ISSN (X)43-5643)

The Wilson Ornithological Society
Founded December 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President— Keith L. Bildstein, Hawk Mountain Sanctuary, RR 2, Box 191, Kempton, Pennsylvania
19529-9449.

First Vice-President— Edward H. Burtt, Jr., Department of Biology, Ohio Wesleyan University,
Delaware, Ohio 43015.

Second Vice-President-John C. Kricher, Biology Department, Wheaton College, Norton, Mas-
sachusetts 02766.

Editor— Charles R. Blem, Department of Biology, Virginia Commonwealth University, P.O. Box
842012, Richmond, Virginia 23284-2012.

Secretary-John A. Smallwood, Department of Biology, Montclair State University, Upper Mont-
clair, New Jersey 07043.

Treasurer— Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana
46556.

Elected Council Members— Donald F. Caccamise and Laurie J. Goodrich (terms expire 1996),
Carol A. Corbat and William E. Davis (terms expire 1997), and Margaret C. Bnttingham an
Herbert T. Hendrickson (terms expire 1998).

Membership dues per calendar year are: Active, S21.00; Student, $15.00; Family, $25.00; Sus-
taining, $30.00; Life memberships $500 (payable in four installments).

The Wilson Bulletin is sent to all members not in arrears for dues.

The Josselyn Van Tyne Memorial Library

The Josselyn Van Tyne Memorial Library of the Wilson Ornithological Society, housed in the
University of Michigan Museum of Zoology, was established in concurrence with the University
of Michigan in 1930. Until 1947 the Library was maintained entirely by gifts and bequests ol
books, reprints, and ornithological magazines from members and friends of the Society. Two mem-
bers have generously established a fund for the purchase of new books; members and friends are
invited to maintain the fund by regular contribution, thus making available to all Society members
the more important new books on ornithology and related subjects. The fund will be administered
by the Library Committee, which will be happy to receive suggestions on the choice of new books
to be added to the Library. William A. Lunk, University Museums, University of Michigan, is
Chairman of the Committee. The Library currently receives 195 periodicals as gifts and in ex-
change for The Wilson Bulletin. With the usual exception of rare books, any item in the Library
may be borrowed by members of the Society and will be sent prepaid (by the University of
Michigan) to any address in the United States, its possessions, or Canada. Return postage is paid
by the borrower. Inquiries and requests by borrowers, as well as gifts of books, pamphlets, reprints,
and magazines, should be addressed to: The Josselyn Van Tyne Memorial Library, University of
Michigan Museum of Zoology, Ann Arbor, Michigan 48109. Contributions to the New Book Fund
should be sent to the Treasurer (small sums in stamps are acceptable).

The Wilson Bulletin
(ISSN 0043-5643)

THE WIUSON BULLETIN (ISSN 004:1-5643) is published quarterly in March, June. September, and December by
0:ilo,ica. Society. 8,0 East lO.h Street. I.wrence

elsewhere, is $40.00 per year. Second-class postage paid at Lawrence. KS. PUS 1 MAS I hrt. Sena a g

HII! I KTIN PO Box 1897, I.awrence, KS 66044-8897. , j j

Back issues or single copies are available for $12.00 each. Most back issues of the Bulletin are available and may be ordered

from the Treasurer. Special prices will be quoted for quantity orders. , , i i .1 i , iho Pdiinr F,-

All articles and communications for publications, books and publications for reviews should be ^

changes should 1. addressed to The Josselyn Van Tyne Memorial Library Museum Ti^tnce.

Subscriptions, changes of address and claims for undelivered copies should be sent to the OSNA, .

KS 66044-8897. Phone: (913) 84.3-1221; FAX: (913) 84,3-1274.

© Copyright 1996 by the Wilson Ornithological Society

Printed by Allen Press, Inc., Lawrence, Kansas 66044, U.S.A.

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

THE WILSON BULLETIN

A QUARTERLY MAGAZINE OF ORNITHOLOGY

Published by the Wilson Ornithological Society

VoL. 108, No. 2 June 1996 Pages 205-396

Wilson Bull., 108(2), 1996, pp. 2Q5-221

GEOGRAPHIC VARIATION AND SPECIES LIMITS IN
CINNYCERTHIA WRENS OF THE ANDES

Robb T. Brumfield' - and J. V. Remsen, Jr.'

Abstract. — Few studies have quantified geographic variation in widely distributed An-
dean birds despite the fact that the linearity of their distributions provides unique opportunity
to assess latitudinal geographic variation. We examined geographic variation of morpho-
metric and plumage characters in populations currently treated as a single species, the Sepia-
brown Wren (Cinnycerthia peruana), that inhabits humid montane forests from northern
Colombia to central Bolivia. Our analysis supports the recognition of three biological species
(olivascens, peruana, and fulva) based on discrete morphometric differences as well as
marked plumage differences. Size variation within populations is inconsistent with the pre-
dictions of Bergmann’s Rule, whereas variation across species runs counter to the predic-
tions, with the smallest species occurring farthest from the Equator. Received 20 June 1995,
accepted 10 Dec. 1995.

Before hypotheses concerning the origin and maintenance of geograph-
ic variation in bird species can be formulated, the patterns of variation
must be well-described (Zink and Remsen 1986). Although the humid
slopes of the Andes mountains are potentially one of the world’s most
productive areas for the study of geographic variation (Remsen 1984a,
Graves 1985, 1988), few workers (e.g.. Graves 1982, 1985, 1991, Remsen
et al. 1991) have quantified geographic variation in widely distributed
Andean birds. We present here an evaluation of moiphometric and plum-
age variation in populations currently treated as a single species, the Se-
pia-brown Wren (Cinnycerthia peruana), that inhabits humid montane
forests in the temperate and subtropical zones of the Andes from northern
Colombia to central Bolivia (Fig. 1). We also re-evaluate species limits
among these populations.

' Museum of Natural Science, Louisiana State Univ., Baton Rouge. Louisiana 70803.

^ Present address: Laboratory of Molecular Systematics, National Mu.seum of Natural History, Smithson-
ian Imstitution MRC 534, Washington. D.C. 20560, and Dept, of Zoology, Univ. of Maryland, College
Park, Maryland 20742.

205

206

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Brumfield and Remsen • VARIATION IN C/NNYCERTHIA WRENS 207

The four cun'ently recognized subspecies of Cinnycerthia peruana
{bogotensis, olivascens, peruana, and/w/va; Paynter 1960, Ridgely and
Tudor 1989, Fjeldsa and Krabbe 1990) were all originally described as
distinct species based on size and plumage differences. Known only
from the eastern Andes of Colombia, bogotensis differs from its geo-
graphically closest relative from the rest of the northern Andes, olivas-
cens, by its darker coloration (Hellmayr 1934, Fjeldsa and Krabbe
1990). Nominate peruana is smaller in size and less olivaceous than
olivascens. Hellmayr (1934) noted that two of nine specimens from Ec-
uador had a faint greyish tinge in the postocular region; he interpreted
this as revealing the close relationship between olivascens and peruana.
It is not clear, however, whether Hellmayr realized that juveniles of
olivascens typically have a grey postocular area. Hellmayr (1934) in-
cluded the preceding taxa as subspecies within Cinnycerthia peruana
but considered the most southern subspecies, fulva, to be a separate
species (C. fulva) based on its distinctly smaller size and well-defined
buffy-white superciliary. Until Hellmayr’s (1934) revision, /w/va was not
considered to be congeneric with the other taxa (all in Cinnycerthia),
but was placed in the genus Thryophilus. Paynter (1960) placed /w/va
as a subspecies within C. peruana and retained the other three subspe-
cies; no reasons were published for the merger of fulva into peruana.
Subsequent works on South American birds have followed Paynter’s
treatment (e.g., Meyer de Schauensee 1966, 1970; Ridgely and Tudor
1989; Fjeldsa and Krabbe 1990).

METHODS

We examined six mensural characters with dial calipers (measured to the nearest 0.05
mm) on 235 (118 males and 1 17 females, according to gonad information on label) study
skins: (1) wing-length (chord of unflattened wing from bend of wing to longest primary);
(2) bill-length (of exposed culmen); (3) bill-width (at its base); (4) bill-depth (at its base);
(5) tail-length (measured from point of insertion of central rectrices to tip of longest rectrix);
and (6) tarsus length (from the joint of tarsometatarsus and tibiotarsus to the lateral edge of
last undivided scute). In addition to the specimens at the Museum of Natural Science,
Louisiana State University (hereafter LSUMZ), specimens were examined from five other
museums with major collections of Andean birds (see Acknowledgments).

Specimens in juvenal or downy plumage or with damaged, extensively worn or molting
wing and tail feathers were excluded. Because many specimens in “adult” plumage lacked
data on age (e.g., skull pneumatization), we also perfonned separate analyses using only those
specimens with skull pneumatization S:90% to determine the effect of inclusion of young birds

<—

Fig. 1. Distribution of Cinnycerthia taxa based on samples included in this study. Triangles
are bogotensis. dark circles are olivascens, hollow circles are peruana, and squares are fulva.

208

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

in adult plumage. Similarly, most specimens did not possess data on body mass, a potentially
useful size character. Using only specimens of known mass, an ANOVA was used to detect
sex and subspecies differences. Mass was not used in the multivariate analyses.

Geographic coordinates for each locality were taken from Paynter et al. (1975), Paynter
and Traylor (1977, 1981), and Stephens and Traylor (1983). SAS software (SAS Institute,
Inc. 1982) was used to calculate all univariate statistics and perform regressions, analyses
of variance (ANOVA, MANOVA), principal component analyses (PCA), and discriminant
function analyses (DEA). Morphometric data were log,o-transformed for all analyses to cor-
rect for a non-normal distribution. All ANOVAs were two-way a posteriori comparisons of
least squares means with the rejection level set according to the number of comparisons by
using the Bonferroni method.

Plumage colors of the 235 specimens were compared to a published color standard (Ridg-
way 1912) under a combination of fluorescent and natural overhead lighting. For bogotensis,
olivascens and peruana, the postocular area of each specimen was scored as (0) no post-
ocular stripe evident, (1) postocular stripe inconspicuous, poorly demarcated, and probably
not visible in the field, or (2) postocular stripe conspicuous, reasonably well-defined, and
probably visible in the field. For all specimens, the amount of white feathering on the head
{in fulva group, excluding the superciliary) was scored as (0) none, (1) partial or complete,
small, white eye-ring, (2) a portion of the forecrown with white feathers, (3) both 1 and 2,
(4) extensive white feathering on face and forecrown, or (5) much of the head white, es-
pecially most of the face and crown. Because the degree of development of the superciliary
and the extent of white feathering on the head are continuously distributed characters, our
scoring system reflects an arbitrary typology.

Specimens collected from the same region more than 70 years apart showed no sign of
post-mortem color change. There is some seasonal variation in coloration caused by wear.
For example, in nominate peruana, and especially in the fulva group, specimens from Oc-
tober to June tended to be more richly colored and ochraceous than those taken from July
to September. This is the only Andean forest bird with which we are familiar that shows
such seasonal plumage wear.

RESULTS

Sexual dimorphism. — ANOVAs of the six skin measurements indicate
a significant {P < 0.05) sex effect on wing-length in all subspecies except
bogotensis (Table 1). This was also found in the analysis of specimens
with skull pneumatization >90%. Differentiation in other characters de-
pend on the subspecies examined. The larger sample sizes of olivascens
and peruana may account for their having significant sexual dimorphism
in more characters than /w/va or bogotensis. Because of the apparent dif-
ferences between sexes, all subsequent analyses were performed sepa-
rately on males and females.

Univariate geographic variation. — Except for wing length, bill length,
and mass, males and females show similar patterns of mensural character
differentiation among subspecies. ANOVAs indicate that wing, tail, and
tarsus lengths are significantly different among all subspecies except be-
tween bogotensis and olivascens. Other differences depend on the sub-
species compared (Table 1). The elimination of specimens lacking data
on skull pneumatization caused a loss of significance in those compari-

Brumfield and Remsen • VARIATION IN CINNYCERTHIA WRENS

209

Table 1

Morphometric Character Means with Standard Deviations'*

Character

Males

C

bogoiensis

C o.

olivascens

C. peruana

C. fulva

Mass

—

25.9 ± 4.1

*20.6 ± 1.7

18.4 ± 2.7

(15)#

(36)#

(6)

Wing length

67.9 ± 0.6

69.5 ± 3.0

*62.0 ± 1.6

*57.9 ± 1.9

(4)

(42)#

(57)#

(14)#

Bill length

14.4 ± 0.7

14.1 ± 0.7

*13.3 ± 0.6

*12.8 ± 0.8

(4)

(42)

(56)#

((4)

Bill width

4.9 ± 0.1

5.2 ± 0.3

*4.8 ± 0.3

4.8 ± 0.2

(4)

(43)

(56)

(14)

Bill depth

5.1 ±0.1

5.1 ± 0.2

*4.7 ± 0.3

4.5 ± 0.2

(4)

(40)#

(56)

(13)

Tail length

64.9 ± 1.2

65.4 ± 2.3

*58.8 ± 1.9

*54.9 ± 1.3

(4)

(41)#

(56)#

(14)

Tarsus length

25.3 ± 0.6

24.6 ± 1.3

*23.1 ± 0.9

*21.7 ± 1.1

(4)

(42)#

(55)#

(13)#

Females

C, o.

C. o.

Character

bogoiensis

olivascens

C. peruana

C. fulva

Mass

—

23.0 ± 2.7

*18.4 ± 1.3

*14.8 ± 1.6

(22)#

(32)#

(10)

Wing length

68.3 ± 1.7

*65.9 ± 2.2

*60.2 ± 2.2

*55.3 ± 1.7

(5)

(52)#

(42)#

(18)#

Bill length

13.8 ± 0.1

14.0 ± 0.6

*12.8 ± 0.5

12.6 ± 0.7

(4)

(51)

(42)#

(17)

Bill width

5.1 ± 0.2

5.1 ± 0.3

*4.7 ± 0.2

4.7 ± 0.3

(5)

(52)

(41)

(18)

Bill depth

4.8 ± 0.5

4.9 ± 0.2

*4.6 ± 0.2

4.5 ± 0.3

(5)

(49)#

(42)

(16)

Tail length

62.5 ± 1.7

62.6 ± 2.5

57.1 ± 2.1

*53.3 ± 1.9

(5)

(51)#

(41)#

(18)

Tarsus length

24.4 ± 2.6

23.5 ± 1.3

*22.4 ± 0.9

*20.7 ± 0.8

(5)

(49)#

(42)#

(17)#

“ Sample sizes are in parentheses.

* Significant difference between subspecies, P < 0,05.

# Significant difference between sexes. P < 0.05.

sons that were only weakly significant in the full analysis. Differences in
wing length, tail length, and tarsus length remained strongly significant
except in the comparison of peruana ‘dx\d fulva, for which tail length was
no longer significantly different.

In all six size characters, olivascens differs significantly from its south-
ern neighbor peruana. To determine whether the samples of olivascens

210

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

c

o

c

o

Q.

E

o

U

"E

a

fi

5.0

2.5

0.0*

-2.5-

-5.0

□

bogotensis

° olivascens

□

□ B

° peruana

□

□

□ t g

Q S

□

* fulva

□

0

Q

□

□

S

o

a

□

o

o

o

§

e

o

•

O

° 1

o

o

8o

0

° f

8 ®

OA

o 1

o

o

A

A

r ^

A

A

A

A

o

o

o

o

d

Latitude

Fig. 2. Plot of first principal component axis against latitude in males. Larger circle
denotes olivascens specimen that clusters with peruana (LSUMZ 88567). Negative numbers
represent northern latitudes.

from north of the Equator contributed to a larger size difference between
olivascens and peruana as a function of geographic distance, we used an
ANOVA to compare populations of olivascens north and south of the
equator. None of the morphometric characters showed significant differ-
ences in either sex.

A discrete size difference exists between olivascens and peruana where
the two meet. Two adult olivascens (males; LSUMZ 1 17378, 1 1739) were
taken on the east slope of the Cordillera Colan at “30 km by road E
Florida on road to Rioja,” depto. Amazonas. Only 16 km SW on the west
slope of Colan four adult peruana from “33 road km NE Ingenio on road
to Laguna Pomacochas,” depto. Amazonas, were collected (males:
LSUMZ 82126, 153098-9; female: LSUMZ 82125). A clue as to where
the two taxa meet comes from a series taken east of La Peca Nueva. From
site two, one male specimen (ca. 2680 m elev.; LSUMZ 88567; skull
completely pneumatized) is peruana based on the morphometric charac-
ters (in multivariate space this individual sorts out with peruana; see large
circle in Fig. 2). Although Graves (1980) considered this specimen as
evidence for introgression because of a “few white feathers in the eye-
ring,” it is clearly typical of peruana in size and plumage as well as in
presence of white feathers. The other nine specimens taken from the La

Brumfield and Remsen • VARIATION IN CINNYCERTHIA WRENS 2 1 1

Peca Nueva sites are all olivascens based on size, and all lack white facial
feathering. This suggests sympatry between the two forms without intro-
gression. When individual skin measurements are compared for all west
slope sites, the discrete difference in size between the two subspecies is
apparent (Table 2).

In both sexes, peruana and /m/vc/ differ signihcantly in wing, tail, and
tarsus lengths. Our most northern sample of fulva is from a northern spur
of the Cordillera Vilcabamba (depto. Cuzco, Peru), a ridge bounded on
the west by the “deeply incised” Apurfmac Valley (Haffer 1974). Our
southernmost sample of peruana is only 150 km NW on a slope of the
Cordillera Occidental, west of the Apurfmac (Yuraccyacu, depto. Aya-
cucho, Peru). All specimens except one of unknown age (female: Amer-
ican Museum of Natural History 820507; hereafter AMNH) are clear
representatives of their respective subspecies (Table 2). The aberrant spec-
imen was taken from the Cordillera Vilcabamba and is peruana in tail
length.

Specimens from the Cordillera Vilcabamba represent a distinctive new
subspecies based on plumage (Remsen and Brumfield, unpubl. data). Be-
cause these populations are allopatric from populations of peruana and
fulva from the main Andes, they are less likely to show signs of intro-
gression between the two forms. To determine if there is any evidence of
introgression between peruana and fulva, we compared the southernmost
samples of peruana (from Yuraccyacu) with our northern most sample of
fulva from the main Andes (males: Field Museum of Natural History
311813-4, from Pillahuata, depto. Cuzco; hereafter FMNH). Although
bo\h fulva specimens were typical of fulva in wing-length, one individual
(31 1814) was typical of peruana in both tail-length and tarsus-length, two
characters that are reliable for distinguishing all subspecies except bo-
gotensis (Table 2). This may indicate some introgression between popu-
lations of fulva from the main Andes and peruana.

North of the equator, the elevational distribution of specimens is not
correlated with latitude (Spearman r = 0.25, ns). South of the equator,
the elevations of specimens increase slightly but significantly with dis-
tance from the equator (Spearman r = 0.29, P < 0.0005). A MANOVA
revealed that there is no significant elevation effect on the morphometric
characters in either sex (Wilk’s Lambda = 0.07, ns).

Multivariate geographic variation. — Because individual characters do
not vary independently, we performed a principal component analysis to
determine which variables best account for the variation expressed across
the range of the species. In specimens missing one or two characters (N
= 17), the subspecies mean was substituted for the missing value. Eigen-
values and eigenvectors were extracted from the correlation matrix. Load-

Table 2

Individual Measurements of Specimens Near Potential Contact Zones (Refer to Figures 1 and 2; Subspecies Means are for All

Other Specimens)

212

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

~ a.
CD w
T3

OB

CD =

ob-£

C OB

5 5

+1

CN

O r' 00 m
cn iri enco-
re CN (N (N O'!

(N

+ 1

yn

so

(N

d

+1

d

+1

CN

in

r-

d

+1

o
d

r- so so

00 (N 00
d — CN

so ■

so

00

in

(N o Os
^ in in

o so (N (N
in in in in in

m (N r- m

cn sd in d

q

cn

+ 1
in
On
sO

+ 1

q

in

(N

C

C3

<U

E

r- so m r- r-

m so r-
so r- so

so so

os

d

+1

(N

q

f!

00

00

in

CN

d

+1

d
+ 1
00

so

d

+1

cn

d

q

+ 1
q

(N

so

q

q

q

q

q

q

q

q

00

+1

rn

CN

<n

CN

CN

—

CN

CN

CN

CN

(N

CN

(N

CN

CN

CN

21.7

r^inONr^osOOcNoo —

OS

ooor-osr-"^inin

insomsominininmin

so 00 On 00 00
^

^ P 9 P P +1

o Os m 00

(N CN — — \0

d

CN

r- 00 cn SO

'Sf Tj- -rf

CNOOOsOsOOOCNOsqq
in in ^ in

— inocNr-'qqqq

^cncndcNcndcNcn

q

S

ON ON 00 so o in o

o

so

(N

SO

(N

SO

in CN

d d

CN

£

C^I

>5

o

B

OJ

B

3

3

3

3

3

3

O

a

O

cd

O

03

O

03

O

3

O

3

’ly)

’c/5

’oo

>.

>>

0

f.

o

o

o

O

U

U

C^l

CN

CN

CN

CN

o

o

o

o

O

CJ

ex

03

3

3

03

3

3

u

ob

ob

ob

_ob

ob

tj

3

3

3

3

3

3

£

12

12

12

12

>-

>-

>-

>-

+ 1

q

•d

in

CN

d

+1

»n

CN

d

+1

00
d
+ 1
00
CN

q

+ 1

q

d

m

q

CN

+ 1

q

00

C3 a

-O X)

S E

C3 cd

X) X
C3 CC

JJ JJ

> >

S S 2 S

S "S (u (u

3 3 X =

C

03

O

E

a

>

u

Brumfield and Reinsen • VARIATION IN CINNYCERTHIA WRENS 2 1 3

ON

00

= 5

— H

d

d

+i

CN

o

00

(N

On

m

+1

00

00

o

+1

(d

d

—

d

P ^

lO

(N

(N

CN

CN

CN

CN

CN

CN

CN

CN

C^J

CN

r-

C<)

CN

d

<N

CN

CN

lO

CO

CN

CN

0

+1

o

d

00

vd

id

(N

id

m

id

00

vd

+1

r-

vd

O)

cd

CO

vd

+1

SO

VD

VO

NO

VO

m

in

in

in

in

in

cn

(N

cd

VO

»n

in

m

(N

cn

d

d

d

+1

00

ON

m

o

in

+1

in

in

1^

ON

in

+1

CQ U

lO

iri

id

id

id

ON

VD

in

•d

m

cn

m

C/D

a>

d

d

d

E

+1

vO

CN

(N

in

00

+1

ON

in

X

CN

+1

CQ--

>o

iri

»d

id

id

id

id

id

OJ

Uh

lo

'd

(N

Q

o

in

r-

UJ

-1

oa

<;

z

p

z

Bill

nglh

d

+1

(N

lo

m

id

in

q

q

rd

d

+1

(N

(N

00

CN

00

CN

cd

d

+1

H

0

u

q

q

(N

— 1

X

CN

(N

CN

r-

(N

CN

—

ap-S

+1

On

o

o

in

+1

ON

00

ON

00

CN

+1

^ J

00

ON

d

vd

ON

ON

CN

vd

id

On

VO

VO

vO

VO

'O

r-

vO

in

CN

in

m

in

in

in

m

tri

d

id

vO

VO

in

m

X

(N

75

75

+1

o

IT)

uo

in

o

o

o

o

+1

ON

+1

vS

IT)

id

id

id

r-'

»d

o

(N

CN

n

CN

C^l

IN

C^l

—

00

d

00

c

cd

cd

cd

cd

M

X

X

XJ

(U

E

S

Cd

E

Cd

E

Cd

E

Cd

c

X)

X

X

X

Cd

cd

cd

Cd

v

o

o

()

c>

p

c

C^l

<n

m

m

c

>

>

>

>

cd

V

O

(U

V

o

V

V

V

u

cd

>5

’c/D

’c^

'55

s:

2

cd

u

<U

cd

u

V

Cd

u

0)

cd

u

(1)

P

C3

0

— *

CN

CN

C^J

C^J

CN

CN

CN

C^l

!:x

u

o

"P

EEE

q

o

U

-£P

_W)

_ob

_ob

_ob

-S?

-£P

_oo

cd

u

3

■o

u

O

XJ

Ui

0

TD

u

O

X)

u

0

£

IE

iE

iE

IE

iE

lE

u

u

u

u

<o

214

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 3

Character Loadings from Principal Component Analysis of Morphometric

Characters

PCA I

PCA II

PCA III

Males

Females

Males

Females

Males

Females

Wing length

0.458

0.451

-0.279

-0.294

-0.133

-0.173

Bill length

0.394

0.398

0.213

-0.341

-0.781

0.628

Bill width

0.322

0.361

0.854

0.630

0.327

0.293

Bill depth

0.408

0.372

-0.040

0.544

0.272

-0.056

Tail length

0.458

0.456

-0.183

-0.321

-0.038

0.051

Tarsus length

0.393

0.402

-0.335

-0.037

0.435

-0.695

Percentage of variance

62.7

63.8

12.2

1 1.7

8.3

8.8

ings on the first principal component (PCA 1) are large and positive
(wing-length and tail-length had the highest loadings) for all characters
(this was also found in an analysis of known adults) indicating that gen-
eral size accounts for most (males 63%, females 64%) of the variation
(Table 3).

A regression of PCA 1 against latitude illustrates a significant negative
correlation between size and latitude in males both north (r = —0.47, P

< 0.05, N = 26) and south (r =-0.79, P < 0.0001, N = 91) of the
equator, and in females both north (r = —0.47 P < 0.05, N = 32) and
south (r = -0.79, P < 0.0001, N = 85) of the equator (Fig. 2; Table 1).
It should be noted that the two taxa that occur north of the equator,
bogotensis and olivascens, were represented by small sample sizes from
all collecting localities (Fig. 1). Latitudinal variation within populations
south of the equator is interesting in that measurements of olivascens are
positively correlated with latitude in both males (r = 0.50, P < 0.05, N
= 21) and females (r = 0.50, P < 0.05, N = 24). Populations of peruana,
with the greatest range in latitudinal sampling, however, show no signif-
icant correlation with latitude in either males (/• = 0.26, ns, N = 56) or
females {r = 0.05, ns, N = 42). Finally, populations oi fulva are signif-
icantly negatively correlated with latitude in both males (r = —0.62, P

< 0.05, N = 15) and females (r = —0.51, P < 0.05, N = 18). If the
subspecies examined were the result of clinal variation (i.e., not distinct
evolutionary units), then we would expect size to be negatively correlated
with latitude both within as well as among subspecies. The subspecies
with the best latitudinal sampling, peruana, showed no correlation of size
with latitude. More samples of bogotensis, olivascens, <xnd fulva are need-

Brumfield and Remsen • VARIATION IN CINNYCERTH/A WRENS

215

ed to allow sufficient examination of clinal variation within each subspe-
cies.

A discriminant function analysis was used to determine the percentage
of specimens that could be classihed correctly solely on size measure-
ments. This analysis is similar to PCA, but assumes a priori the identi-
fication of the groups based on the museum label. By making this as-
sumption, DFA can determine which characters cause maximal separation
of the groups. We plotted discriminant functions for each specimen along
an axis. Specimens were considered unidentifiable if their discriminant
function value overlapped with any values from their neighbor taxon.
Specimens of bogotensis could not be distinguished from olivascens.
When compared only to each other, 79% of olivascens and 92% of per-
uana were identified unambiguously. When compared only to each other,
87% of peruana and 73% of fulva were identified unambiguously. All
unidentifiable specimens were examined individually for geographic prox-
imity to their neighbor taxon and size relative to the rest of their taxon.
Most unidentifiable specimens occurred far from their neighbor taxon and,
therefore, probably represent extreme within-subspecies variation and not
introgression from subspecific neighbors. Three of the five individuals
from our most northern specimens of fulva (Cordillera Vilcabamba), how-
ever, could not be differentiated from peruana based on size measure-
ments. These three are, however, unambiguously diagnosable based on
plumage (presence of buffy-white superciliary stripe in fulva).

Plumage Coloration

1. Eastern Andes of Colombia and bogotensis. — Beginning near the
northeastern extreme of latitudinal distribution of our sample, specimens
examined (N = 14) from the eastern Andes of Colombia (deptos. Huila,
Putumayo, and Narino) can be distinguished from other populations in
the Andes of Colombia and Ecuador by their darker, redder brown un-
derparts. In being darker than olivascens, these specimens match Hell-
mayr’s (1934) brief notes on bogotensis, which were described from “Bo-
gota” specimens that lacked precise locality data. However, we were un-
able to locate specimens from the possibly disjunct population (see map
in Hilty and Brown 1986) farther north in deptos. Santander, Boyaca, and
Cundinamarca, a region more likely to contain the true type locality for
“Bogota” specimens, as noted for this taxon in particular by Olivares
(1969). Hellmayr (1934), who examined the type specimen of bogotensis,
assigned a specimen from Andalucia, depto. Huila (presumably the same
specimen in our sample, AMNH I 16874) to this taxon. Therefore, we
treated our sample as bogotensis but with some hesitancy. In any case,
these specimens differ strongly from our nearest specimens from adjacent

216

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Ecuador in depto. Napo in being a darker, richer, redder (particularly in
depto. Narino) brown on the underparts, closer to Front’s Brown or Van
Dyke Brown than to the Cinnamon Brown or Saccardo’s Umber of the
Ecuadorian specimens, and in having much darker throats. Such levels of
differentiation in color are typically accorded subspecies status.

2. Central and western Andes of Colombia to northern Peru. — Popu-
lations from the central Andes and the western Andes of Colombia, the
eastern Andes of Ecuador, the eastern Andes of Peru north of the Maranon
Valley, and some localities south of the Maranon in the eastern Andes of
extreme northern depto. Amazonas, Peru, are all similar in general col-
oration. Their underparts are fairly dark brown, closest to Cinnamon
Brown or Saccardo’s Umber, with the throats slightly paler; their upper-
parts are an even darker brown, closest to Chestnut Brown or Van Dyke
Brown, with a paler, often ochraceous-tinged forecrown. We are unable
to distinguish by using any color character the specimens from the two
extremes in latitudinal distribution (central Colombia vs. northern Peru),
except that populations from Ecuador south to Peru show an increasing
tendency towards having a trace of a postocular stripe (Table 4). Also,
several specimens from Colombia have white feathering on the forehead,
a feature not yet known from populations from Ecuador and Peru. These
populations represent the taxon olivascens, the type locality of which is
Santa Elena on the eastern slope of the Central Andes near the northern
limit of the taxon. We found only one specimen each from the western
Andes of Colombia and Ecuador. Both are within the range of color of
olivascens.

In northern Peru, where olivascens meets nominate peruana, geograph-
ic variation in color is complex. All but one specimen from three localities
in extreme northern Peru on both sides of the Maranon River are virtually
indistinguishable from one another. These specimens are from (1) the
Cerro Chinguela area in depto. Cajamarca, north of the Maranon (see
Parker et al. 1985), (2) the Cordillera Colan area in central depto. Ama-
zonas, south of the Maranon, and (3) the Abra Patricia area in extreme
northern depto. San Martin. The only difference between the populations
on opposite sides of the Maranon was a tendency towards faint postocular
stripes in populations on the south side (Table 4). These represent the
southernmost populations of what is currently considered C. p. olivascens
(Ejeldsa and Krabbe 1990).

However, one specimen from Cordillera Colan (ca 2680 m elev.;
LSUMZ 88567; skull completely pneumatized) differs dramatically from
the other 1 1 from the region in being paler and less reddish throughout
and in having a distinct postocular streak. As noted above, it is also
distinctly smaller than males from the same area (wing-length 63.7 mm

Brumfield and Remsen • VARIATION IN CINNYCERTHIA WRENS

217

Table 4

Prominence of Postocular Area in Populations of
(Numbers in Rows Refer to the Number of Spec

CiNNYCERTHiA Wrens in Peru
imens with Each Score)

Postocular score

Region^

0>'

u

2d

C. o. bogotensis

E. Andes

7

0

0

C. o. new subspecies?

depto. Narino

8

0

0

C. o. olivascens

W. Andes (e. slope), Colombia

19

0

0

Central Andes, Colombia

7

0

0

Ecuador

14

2

0

depto. Cajamarca

15

0

0

Cordillera Colan

9

2

0

Abra Patricia

2

4

0

Total (C. o. olivascens)

66 (89%)

8 (11%)

0

C. peruana

Cordillera Colan

0

0

1

Ne. of Ingenio; La Lejia

0

0

4

Puerta del Monte

0

0

4

depto. La Libertad

0

6

6

depto. Huanuco

1

31

2

depto. Pasco

2

23

2

depto. Jum'n

0

14

4

depto. Ayacucho

0

7

0

Total (C. peruana)

3 (3%)

81 (76%)

23 (21%)

" Arranged from north to south.

0 = No postocular stripe evident.

' I = Postocular stripe inconspicuous, poorly demarcated, and probably not visible in the field.
■■ 2 = Postocular stripe conspicuous, reasonably well-defined, and probably visible in the field.

VS. 72,0 mm average for other two males), and it also has an indistinct
white eye-ring, a condition found in many specimens of nominate peru-
ana but unknown in olivascens. In these respects, therefore, it matches
specimens of nominate peruana from localities to the south. It differs,
however, from nominate peruana from deptos. San Martin and La Lib-
ertad in being less ochraceous. As noted above. Graves (1980) considered
this specimen to be an olivascens with signs of introgression from nom-
inate peruana. However, other than being collected in the same area as
true olivascens, it has no characters of that taxon, and its paler, less red-
dish coloration is not really intermediate between that of olivascens and

218

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

typically more ochraceous peruana. For those reasons, Graves (pers.
comm.), with a much larger series of specimens available to him than in
1980, now considers the specimen to represent peruana and not an in-
tergrade. Because Cinnycerthia wrens are sedentary, secretive, under-
growth-dwelling species with short, rounded wings, it seems unlikely that
this individual represented a long-distance wanderer, but rather a sym-
patric resident population.

Three adults from “33 road km NE Ingenio on road to Laguna Po-
macochas,” depto. Amazonas (LSUMZ 82125-26, 153098; elevs. ca.
2315 and 2135 m), are like typical nominate peruana in having prominent
postocular streaks (Table 4), in being a paler, less reddish brown overall,
and in being small (see Morphometries section above). One (female:
LSUMZ 82125) is distinctly ochraceous on the underparts and has a white
eye-ring and whitish feathers on the forehead; it is indistinguishable from
nominate peruana. The other two (both males), however, lack the ochra-
ceous tones and are virtually identical to LSUMZ 88567 from Cordillera
Colan; one has a partial white eye-ring. The locality from which these
three specimens come is just south of the main part of the Cordillera
Colan and only about 10 km from localities that have typical olivascens.
This locality, however, is also between the Cordillera Colan and a locality
farther southeast, Abra Patricia, where typical olivascens is found. An-
other specimen (male: AMNH 234998) from slightly farther south at La
Lejia, N. of Chachapoyas, depto. Amazonas, is slightly less reddish
throughout and less ochraceous on the throat than typical nominate per-
uana from farther south, but it is more reddish throughout than the three
above-mentioned specimens. We suspect that a thorough elevational tran-
sect in this region would show that, where the two taxa come together,
olivascens will be found at high elevations and peruana at low elevations.

Whether the above specimens represent intermediates between olivas-
cens and nominate peruana or whether they represent geographic varia-
tion at the northern extreme of nominate peruana cannot be determined
without larger series of specimens, preferably accompanied by genetic
samples. We currently favor the latter treatment because the specimens in
question are like nominate peruana in size, presence of white facial feath-
ering, and prominence of the postocular streak. The only tendency to-
wards nominate peruana in the series of olivascens in the region is the
presence of class 1 postocular streaks in some specimens (Table 4). There-
fore, if there is gene flow between the populations, it is only faintly
expressed in the phenotypes of the populations there. The differences
between nominate peruana and olivascens are of the same general mag-
nitude as those between olivascens and C. unirufa, which are clearly two
separate species (Hellmayr 1934, Parker et al. 1985). We therefore believe

Brumfield ami Remsen • VARIATION IN CINNYCERTHIA WRENS

219

that the burden of proof falls on those who would treat olivascens and
nominate peruana as a single, freely interbreeding biological species, and
we regard C. olivascens Sharpe as a separate species. We propose “Sharpe’s
Wren’’ as an English name for C. olivascens and “Peruvian Wren,” the
name used by Hellmayr (1934), for C. peruana, which is endemic to Peru.

3. Eastern Andes of central Peru. — A series of 106 adult specimens
from Puerta del Monte, depto. San Martin, south through deptos. La Lib-
ertad, Huanuco, Pasco, Junm, and Ayacucho are all relatively uniform in
coloration but with some individual variation that shows little geographic
pattern. Their upperparts are a dark, rich brown closest to Brussels Brown
or Russet Brown but darker; most specimens have an ochraceous tinge
to the forecrown. Their underparts are generally closest to Cinnamon
Brown. Their throats range from buffy white to ochraceous. All but three
specimens have either a class 1 or class 2 postocular streak (Table 4),
and 38% (N = 41) have some white feathering on the head (Table 5).
We concur with Graves (1980) that this white feathering is found only
on birds with nearly or completely pneumatized skulls and that it is found
in both males and females (cf. Gochfeld 1979).

4. Eastern Andes of southern Peru and northern Bolivia. — Reference
works (Paynter 1960, Ridgely and Tudor 1989, Fjeldsa and Krabbe 1990)
currently treat all populations from depto. Cuzco, Peru, south to depto.
Cochabamba, Bolivia, as one taxon, C. peruana fulva. However, this re-
gion includes at least three discrete taxa, as we outline below.

At the northern extreme in latitude of this region, specimens (N = 7)
from the isolated Cordillera Vilcabamba, depto. Cuzco, differ dramatically
from any other Cinnycerthia wren in having a well-defined dark crown.
These are clearly a distinct, undescribed taxon most closely related to C.
{p}. fulva (Remsen and Brumfield, unpubl. data). The Cordillera Vilca-
bamba specimens do not, however, differ significantly in any size char-
acter from fulva populations in Cuzco or depto. Puno, Peru, south to
depto. Cochabamba, Bolivia.

The type locality fulva is in the main Andes of depto. Cuzco. Un-
fortunately, few specimens o^ fulva have been collected in that region,
and these wrens are notably rare there (T. A. Parker, pers. comm.); in
fact, we can find no specimens or published localities from west of the
Rio Apurimac Valley through the Urubamba drainage east to at least the
Rfo Paucartambo valley. Because humid forest in this region has been
thoroughly sampled (e.g.. Chapman 1921, Parker and O’Neill 1980), we
suspect that the absence of Cinnycerthia wrens in this region may be a
true gap in their distribution. In addition to the type specimen, taken at
Huasampilla, we found only two other specimens in depto. Cuzco; all
three are in the Rfo Madre de Dios drainage in the humid eastern portion

220

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

X

O

H

a:

LJ

u.

UJ

(A

o

CxC

00

O'

UJ

CQ

z

<

>

Ij

o

CQ

Q

z

<

D

o'

UJ

UJ

Qc:

CX

O

u

Z

C/O

00

X

Z

u

<

UJ

X

in

H

m

3:

S

00

Z

UJ

§

j

03

<

g

H

1

u

UJ

0

a.

00

u.

O

u.

O

c/5

a^

X

CQ

z

o

H

<

J

X

X

Z

a.

O

D-

Q

<

LU

d:

z

o

o

z

E2

UJ

X

h-

<

UJ

X

u.

o

H

z

X

o

<

C W3
O (U

o o o o o o o

o o o o o o o

o o o o o

(N

(N

in (N O O O O

o o o o o o o

Os

00

m 'O »r)

vO

Pi

o

00

o

U

o

(U

CL

oo

Xi

3

X

E

^ o

O U

T3

C

<

ui

u

o

iC

'C

cj

Z

CL

(U

TD

'X)

;j

»3

u

• <0
O "O

^ c

CO <<
(D ^
—
c
<

CO

o

u

CO

E

cd

O cd

O o

'c5’ 2 S

U U C3

. =: Cu

° "H S

^ D b S' -9

^ U UJ "O U <!

o

-a

C3

3

oooomooo

o o

00 C^l

X

r-

(N — (N CN ON

o o o o

o o

(N

CN

— — O CN (N

ON

o

o —

(N

X

OcNCNoooNOmc^

— (N — ’ X

to

U

t2

c

"o

U

(U

cd -o
J OJ C3
.. c t:

11^

C

Q

s

L.

Ci.

u

0)

« 3

S -a J K

o

X

o

3

O

cd

<

cd

-X ^ t.

u . (D
0^3

U Z CL

c

s

Cl

U

o o o o ^

Cl cx Cl x< Cl

0) 1) !U 1) D

•o -O “O -O T) o

H

Brumfield ami Remsen • VARIATION IN CINNYCERTHIA WRENS

IT)

uj ^

j S

CQ H

H U

o o o o o o

o o o o o o

o o

00

CN O m

O (U

(N

O ^ O 00

o o o o o o

^ (N Tf o so

>

o

c

X)

a

a

3

o

u

0/j

^3

u

(N

r-

00

(N

03

X

e

a.

3

0

N

3

X

bO

o

o

N

O

c

a

0.

03

X

o

C3

3

3

ci

O

u

cu

U

o

o

O

o

a

a

D-

a

(U

<u

dJ

03

TD

X)

T3

X3

,0

H

“The amount of white feathering on the head (\n fulva, excluding the superciliary) was scored as: (0) none; (1) partial or complete, small, white eye-ring; (2) a portion of the forecrown
with white feathers; (3) both I and 2; (4) extensive white feathering on face and forecrown; or (5) much of the head white, especially most of the face and crown.

222

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

of the departamento. The two specimens that we examined (FMNH
311813-814, from Pillahuata) are extremely similar to each other and
differ dramatically from specimens from depto. Puno, Peru, to depto.
Cochabamba, Bolivia, in having a superciliary that is less conspicuous
because it is ochraceous rather than buff-white, and because it is not as
broad. Also, the broad, dark eye-line is less conspicuous because the
auriculars are much darker, and the back and underparts are slightly
darker. In these features, the Cuzco specimens could be regarded as show-
ing variation in the direction of peruana. We suspect that this is why
Paynter (1960) merged fulva into peruana. Even if the gap in distribution
in western depto. Cuzco is a sampling artifact, peruana would be sepa-
rated from fulva by the canyon of the Rio Urubamba, and so no test of
sympatry is possible for direct assessment of species limits. Therefore,
whether peruana and fulva should be treated as different biological spe-
cies depends on the importance one assigns to the apparent variation
towards peruana shown in Cuzco specimens of fulva.

We propose that fulva be elevated to species status because fulva, par-
ticularly those populations from the Cordillera Vilcabamba and those
from depto. Puno south, differs in plumage pattern more than peruana
does from C. olivascens or than even C. olivascens does from C. unirufa.
All populations currently included in fulva differ from C. peruana in
having a distinctive, well-defined superciliary that extends from the fore-
crown at the nares posteriorly to the nape. The auriculars are pale like
the superciliary, thereby sharply demarcating a broad, dark eye-stripe, and
the auriculars are only slightly darker than the pale throat. Thus, fulva
has a patterned face, with a conspicuous pale superciliary and dark eye-
stripe, whereas peruana, which only occasionally has a hint of a post-
ocular superciliary on an otherwise uniform reddish-brown face, has no
face pattern. The head pattern of fulva gives the bird a different “look”
from C. peruana, a difference that doubtless led to these populations’
original placement in a different genus, Thryophilus (now merged in Thry-
othoriis). As noted by Ridgely and Tudor (1989), the face pattern of fulva
recalls that of the sympatric Mountain Wren {Troglodytes solstitialis). The
rest of the plumage, however, is similar to other Cinnycerthia species.
The breast and belly, closest to a pale dull Sayal Brown, are paler and
less reddish than those of C. peruana. The upperparts are like those of
C. peruana but are also paler and slightly less reddish. Finally, the ex-
tensive white feathering found in some peruana is unknown in fulva
(Table 5). We propose that Hellmayr’s (1934) English name for fulva,
“Superci hated Wren,” be used for this species. We suspect that compar-
isons of vocalizations will be especially important in resolving species

Brumfield and Remsen • VARIATION IN CINNYCERTH/A WRENS

223

limits; T. A. Parker (pers. comm.) has observed that fulva is rather quiet
compared to the often noisy peruana.

The distinctively small, pale specimens from depto. Puno, Peru, and
deptos. La Paz and Cochabamba, Bolivia, seem to be indistinguishable
from one another in plumage or size; the paleness of a recent LSUMZ
series of 13 July- August specimens from the Chuspipata region of La Paz
seems attributable to seasonal wear. The Puno-to-Cochabamba population
represents an undescribed subspecies (Remsen and Brumfield, unpubl.
data) that differs from the depto. Cuzco population in the ways described
above.

Although white facial feathering has been reported in C. peruana sensu
latu from Ecuador south (Fjeldsa and Krabbe 1990), we cannot find any
explicit reference to such white feathering mfulva. In a series of 39 adults
of fulva from depto. Puno to depto. Cochabamba, eight individuals show
“extra” white feathers on the forecrown above the superciliary and three
of these have “extra” white feathers around the eyes (Table 5); none,
however, shows the class 4 or class 5 white feathering found regularly in
peruana.

DISCUSSION

Zoogeography. — Somewhere in the Eastern Andes of extreme northern
Ecuador or southern Colombia, bogotensis must meet olivascens. We-
know of no geographic barriers in the region that might separate the two.
The southern limit of olivascens also does not correspond to any prom-
inent geographic boundary. The deep, arid Marafion valley has long been
recognized as a major barrier to dispersal for birds occurring in humid
montane forest. Parker et al. (1985) noted 18 pairs of allospecies separated
by the Rfo Marafion. However, C. olivascens is found on both sides of
the barrier, and our limited samples suggest that it interdigitates with C.
peruana in the Andes Just south of the Marafion. The southern limit of
peruana corresponds to one of the most important biogeographic bound-
aries in the humid Andes, the Rio Apurfmac valley (Haffer 1974).

The distinctive new subspecies oi fulva in the Cordillera Vilcabamba
adds another taxon endemic to this major outlying range. The gap in
distribution between the distinctive taxon in the Cordillera Vilcabamba
and the range of nominate /m/vcz in southeastern depto. Cuzco is a pattern
also found in the Rufous-naped Brush-Finch (Atiapetes rufinucha) (Rem-
sen 1993).

Nominate fulva and the undescribed southern subspecies must meet
somewhere in southern depto. Cuzco or northern depto. Puno. Several
other pairs of Andean bird taxa share this pattern (e.g., Marcapata Spi-
netail [Cranioleuca marcapatae] and Light-crowned Spinetail [C. alhi-

224

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

ceps\, Remsen 1984b; White-throated Spadebill [Platyrinchus mystaceus
zamorae and P. m. partridgei], Remsen et al. 1991). Precisely where this
turnover takes place is not known, but the most likely geographic barrier
in the region is the Rio Marcapata valley in extreme southeastern depto.
Cuzco.

Species limits. — Our analysis of geographic variation in plumage and
morphology in populations currently treated as Cinnycerthia peruana sup-
ports recognition of three biological species. The northernmost, olivas-
cens, shows large, discrete morphometric differences from the next pop-
ulation to the south, peruana, as well as marked plumage differences.
Some specimens in the region of contact, however, have intermediate
plumage characters that could be interpreted as evidence of past or present
gene flow. The presence of these two different forms in such geographic
proximity suggests that either an abrupt step dine serves to differentiate
the groups or, more likely, past geological events (e.g., Rio Maranon)
isolated the forms long enough for differentiation to occur. The absence
of unambiguously intermediate forms and the presence of an individual
of peruana within a population of olivascens suggest that these two taxa
are best treated as separate but closely related species.

Although contact between peruana and/w/va is prevented by the Apu-
rimac Valley, the distinctive plumage pattern of fulva relative to other
Cinnycerthia species suggests that fulva should be treated as a separate
species; in fact, no reason was ever published for the merger (by Paynter
1960) of fulva into peruana.

In summary, we recommend the following taxonomic treatment;

Cinnycerthia olivascens Sharpe’s Wren

C. o. hogotensis (eastern Andes of Colombia)

C. o. olivascens (central Andes and western Andes of Colombia,
eastern Andes of Ecuador, eastern Andes of Peru south extreme
northern depto. Amazonas, Peru)

Cinnycerthia peruana Peruvian Wren (eastern Andes of Peru from dep-
to. Amazonas to depto. Ayacucho)

Cinnycerthia fulva Superciliated Wren

C. f (undescribed subspecies) (Cordillera Vilcabamba, depto. Cuzco,
Peru)

C. f fulva (eastern Andes of depto. Cuzco, Peru)

C. f (undescribed subspecies) (depto. Puno, Peru, to depto. Cocha-
bamba, Bolivia)

Bergmann’s Rule. — The linearity of the distributions of Andean forest
birds (Remsen 1984a, Graves 1988) creates a unique opportunity to assess

Brumfield and Remsen • VARIATION IN CINNYCERTHIA WRENS

225

Bergmann’s Rule, which predicts that the body size of organisms increas-
es in colder and drier areas as a response to changes in climatic conditions
(James 1970). Although climatological data are unavailable for the humid
Andes, we assume that temperatures in the Andes are colder with increas-
ing distance from the Equator; rainfall data from LSUMZ collecting lo-
calities in Peru and Bolivia also suggest that localities farthest from the
equator are drier. The elevational distribution of our specimen localities
remains constant over latitude north of the Equator and increases with
latitude south of the equator. Therefore, we do not believe that elevation
per se is a variable that would influence body size, or if so, the influence
would be in producing colder temperatures with increasing elevations
south of the Equator.

Although Bergmann’s Rule, including both the pattern and the expla-
nation for the pattern (James 1970), seems deeply entrenched in vertebrate
biology, 68% of the studies on intraspecific variation compiled by Peters
(1992) do not support the rule (see also McNab 1971, Zink and Remsen
1986, Geist 1987). Only four studies have investigated Bergmann’s rule
in Andean birds. Graves (1991) found a positive correlation between body
size and latitude in the Carbonated Elowerpiercer (Diglossa carbonaria)
in both directions away from the equator, as did Kratter (1993) for the
Yellow-billed Cacique (Amblycercus holosericeus). Remsen et al. (1991)
found that body size in the White-throated Spadebill {Platyrinchus mys-
taceus) decreased south of the equator, counter to the prediction of Berg-
mann’s rule, but the analysis included only populations south of the equa-
tor. Additionally, Remsen (1993) found that body size decreased away
from the equator in Atlapetes rufinucha, but only those populations from
depto. Cuzco (13°S. Eat.) to depto. Santa Cruz, Bolivia (18°S. Eat.) were
analyzed. Although Bergmann’s Rule is intended to apply primarily to
intraspecific variation, we believe that its application to allopatric sister-
species (allospecies), as in the Cinnycerthia populations analyzed herein,
is also appropriate. Our results indicate that body size is negatively cor-
related with latitude in Cinnycerthia wrens both north and south of the
equator, counter to Bergmann’s Rule. The only exception appears within
populations of olivascens south of the equator, where size may be posi-
tively correlated with latitude. Obviously, congruence between large num-
bers of a wide variety of taxa is necessary to determine the applicability
of Bergmann’s “Rule” to Andean birds.

ACKNOWLEDGMENTS

We thank the curatorial staffs of the Academy of Natural Sciences of Philadelphia, Amer-
ican Museum of Natural History (New York), Carnegie Museum of Natural History (Pitts-
burgh), Field Museum of Natural History (Chicago), and Delaware Museum of Natural

226

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

History (Dover) for generous loans of specimens. We thank A. P. Capparella, R. T. Chesser,
J. C. Coulson, G. R. Graves, A. W. Kratter, T. S. Schulenberg, D. A. Wiedenfeld, and two
anonymous reviewers for providing helpful comments on the manuscript. G. R. Graves re-
examined material from northern Peru. J. M. Bates, S. J. Hackett, T. A. Parker III, K. V.
Rosenberg, and T. S. Schulenberg provided to RTB immeasurable help and encouragement
in the project’s development. S. A. Juliano assisted with statistical analyses. The morpho-
metric analyses of this paper initially were prepared by RTB as a research project for JVR’s
ornithology course.

LITERATURE CITED

Chapman, L M. 1921. Distribution of bird life in the Urubamba Valley of Peru. Bull. U.S.
Nat. Mus. No. 1 17.

LjeldsA, j. and N. Krabbe. 1990. Birds of the high Andes. Zool. Mus., Univ. of Copen-
hagen, Denmark.

Geist, V. 1987. Bergmann’s Rule is invalid. Can. J. Zool. 65:1035-1038.

Graves, G. R. 1980. Relationship of white facial feathering to age and locality in Peruvian
Cinnycerthia peruana. Bull. Brit. Orn. Club 100:149—150.

. 1982. Speciation in the Carbonated Llower-piercer {Diglossa carbonaria) complex

of the Andes. Condor 84:1-14.

. 1985. Elevational correlates of speciation and intraspecific geographic variation in

plumage of Andean forest birds. Auk 102:556—579.

. 1988. Linearity of geographic range and its possible effect on the population

structure of Andean birds. Auk 105:47-52.

. 1991. Bergmann’s rule near the equator: Latitudinal dines in body size of an

Andean passerine bird. Proc. Nat. Acad. Sci. 88:2322—2325.

Haffer, j. 1974. Avian speciation in tropical South America. Publ. Nuttall Orn. Club 14:
1-390.

Hellmayr, C. E. 1934. Catalogue of birds of the Americas. Lield Mus. Nat. Hist. Publ.,
Zool. Ser. 13, pt. 6.

Hilty, S. L. and W. L. Brown. 1986. A guide to the birds of Colombia. Princeton Univ.
Press, Princeton, New Jersey.

James, P. C. 1970. Geographic size variation in birds and its relationship to climate. Ecology
51:365-390.

Kratter, A. W. 1993. Geographic variation in the Yellow-billed Cacique, Amblycercus
holosericeus, a partial bamboo specialist. Condor 95:641—651.

McNab, B. K. 1971. On the ecological significance of Bergmann’s Rule. Ecology 52:845-
854.

Meyer De Schauensee, R. 1966. The species of birds of South America. Livingston Publ.
Co., Narberth, Pennsylvania.

. 1970. A guide to the birds of South America. Livingston Publ. Co., Narberth,

Pennsylvania.

Olivares, A. 1969. Aves de Cundinamarca. Publicaciones, Direccion de Divulgacion Cul-
tural, Bogota, Colombia.

Parker, T. A., Ill and J. P. O’Neill. 1980. Notes on little known birds of the upper
Urubamba Valley, southern Peru. Auk 97:167-176.

, T. S. Schulenberg, G. R. Graves, and M. J. Braun. 1985. The avifauna of the

Huancabamba region, northern Peru. Pp. 169—197 in Orn. Monogr, Vol. 36 (P. A Buck-
ley, M. S. Foster, E. S. Morton, R. S. Ridgely, and E G. Buckley, eds.).

Paynter, R. a., Jr. 1960. Family Troglodytidae. Pp. 379-440 in Check-list of the birds ot
the world. Vol. 9. Mus. Comp. Zool., Cambridge, Massachusetts.

Brumfield and Remsen • VARIATION IN CINNYCERTHIA WRENS

227

AND M. A. Traylor, Jr. 1977. Ornithological gazetteer of Ecuador. Mus. Comp.

Zool., Cambridge, Massachusetts.

AND M. A. Traylor, Jr. 1981. Ornithological gazetteer of Colombia. Mus. Comp.

Zool., Cambridge, Massachusetts.

, M. A. Traylor, Jr., and B. Winter. 1975. Ornithological gazetteer of Bolivia.

Mus. Comp. Zool., Cambridge, Massachusetts.

Peters, R. H. 1992. A critique for ecology. Cambridge Univ. Press, New York, New York.
Remsen, J. V., Jr. 1984a. High incidence of “leapfrog” pattern of geographic variation in
Andean birds: implications for the speciation process. Science 224:171-173.

. 1984b. Geographic variation, zoogeography, and possible rapid evolution in some

Cranioleuca spinetails (Furnariidae) of the Andes. Wilson Bull. 96:515-523.

. 1993. Zoogeography and geographic variation in Atlapetes rufinucha (Aves: Em-

berizinae), including a distinctive new subspecies, in southern Peru and Bolivia. Proc.
Biol. Soc. Wash. 106:429-435.

, O. Rocha, C. G. Schmitt, and D. C. Schmitt. 1991. Zoogeography and geograph-
ic variation of Platyrinchus mystaceus in Bolivia and Pena, and the Circum-Amazonian
distribution pattern. Ornitologia Neotropical 2:77-83.

Ridgely, R. S. and G. Tudor. 1989. The birds of South America: the oscine passerines.
Univ. Texas Press, Austin, Texas.

Ridgway, R. 1912. Color standards and color nomenclature. Published by the author. Wash-
ington, D.C.

Sas Institute, Inc. 1982. SAS user’s guide: statistics. 1982 edition. SAS Institute, Cary,
North Carolina.

Stephens, L. and M. A. Traylor, Jr. 1983. Ornithological gazetteer of Peru. Mus. Comp.
Zool., Cambridge, Massachusetts.

Zink, R. M. and Remsen, J. V., Jr. 1986. Evolutionary processes and patterns of geographic
variation in birds. Cum Ornithol., Vol. 4 (R. F. Johnston, ed.). Plenum Press, New York,
New York.

Wilson Bull., 108(2), 1996, pp. 228-245

NEST ATTENTIVENESS IN HUMMINGBIRDS

William H. Baltosser

Abstract. — With few exceptions, nest building, incubation, and care of young are the
responsibilities of the female in hummingbirds (Trochilidae). The time females spent on
these tasks (collectively defined as attentiveness) was determined for three species of hum-
mingbirds nesting in southeastern Arizona and southwestern New Mexico, i.e., the Broad-
billed (Cynanihus latirostris), Violet-crowned (Amazilia violiceps), and Black-chinned {Ar-
chilochus alexandri) hummingbirds. In addition to elucidating attentiveness, the data from
my study also provide insights into the allocation of time for maintenance and other activ-
ities. This information, combined with that from other studies, shows a remarkable unifor-
mity in attentiveness in the Trochilidae regardless of the taxa or geographic areas involved.
Received 1 Mar. 1994, accepted 15 Aug. 1995.

With few exceptions, nest building, incubation, and care of young are
the responsibilities of the female in hummingbirds, with males rarely par-
ticipating in reproduction beyond copulation. In addition to carrying out
these responsibilities, female hummingbirds must also tend to their own
needs, such as food procurement, plumage maintenance, and predator
avoidance. The allocation of time to this array of tasks is also influenced
by a number of other factors, including the availability of resources,
weather conditions, and interactions among individuals and taxa. With
such factors and the demands of non-reproductive activities in mind, I
examined nest attentiveness in nesting Broad-billed {Cynanthus latiros-
tris), Violet-crowned {Amazilia violiceps), and Black-chinned {Archilo-
chus alexandri) hummingbirds in southwestern New Mexico and south-
eastern Arizona. My focus was on determining the time allotted by in-
dividual females to the tasks of nest building, incubation, and care of
young. These findings are presented along with comparisons to hum-
mingbird attentiveness in other studies. In general, hummingbirds show
similar attentiveness patterns, particularly as regards incubation and to
lesser degrees with other stages of the reproductive cycle.

STUDY AREAS AND METHODS

Three areas were selected for this study; all lacked artificial food sources such as hum-
mingbird feeders or extensive areas of cultivated plants. The first and largest of these was
Guadalupe Canyon, situated along the United States-Mexico border in southwestern New
Mexico (Hidalgo Co.) and southeastern Arizona (Cochise Co.). Hummingbirds regularly
nesting in this area include the Broad-billed, Violet-crowned, and Black-chinned, while
Costa’s Hummingbird {Calypte costae) nests occasionally (Baltosser 1986b, 1989a, 1989b)
and the Lucifer Hummingbird (Calothorax lucifer) has nested at least once (Scott 1994).

Dept, of Biology, Univ. of Arkansas at Little Rock, Little Rock, Arkansas 72204.

228

Baltosser • NEST ATTENTIVENESS

229

The second area was a segment of Rucker Canyon which lies on the west side of the
Chiricahua Mountains in Cochise Co. in southeastern Arizona. Black-chinned Humming-
birds were the primary species at this site, with a few Magnificent Hummingbirds {Eugenes
fulgens) also nesting (Baltosser 1986b). The third and most northern of the areas was along
the Gila River near the town of Cliff, Grant County, New Mexico (Baltosser 1986a). The
Black-chinned Hummingbird was the only hummingbird to nest at this site (for map and
extensive description of each area see Baltosser 1986b).

I recorded periods of attentiveness in nesting female hummingbirds during 209 h of
observation in the summer of 1976. My data sets are based on six nests and 17.5 h of
observation for Broad-billed Hummingbirds, five nests and 26.1 h for Violet-crowned Hum-
mingbirds, and 46 nests and 165.6 h for Black-chinned Hummingbirds (25 nests and 67.4
h for Guadalupe Canyon, 1 1 nests and 50.5 h for Rucker Canyon, and 10 nests and 47.7 h
for Cliff). Only sessions (attentive periods) and recesses (periods away from nest) thought
to have begun and ended spontaneously have been included in my study in order to eliminate
unnaturally short or long intervals (see Skutch 1962). Nests were randomly selected for
intensive study (stratified random sampling) and they were monitored from vantage points
far enough away to prevent my interfering with natural events. Events were timed using a
stopwatch and chronicled with the aid of a tape recorder.

My periods of observation were 1-4 h, and I stratified these into morning (06:00-10:00),
mid-day (1 1:00-16:00), and evening (17:00-20:30) segments to gather data when ambient
temperatures were lowest (morning hours), highest (mid-day), and intermediate (evening).
Observations scheduled in this manner also allowed me to gather data for the period fol-
lowing the nocturnal fast (morning) and prior to the onset of fasting (evening). A tally of
the hours that I devoted to each period by species and area was maintained, so that the
observer effort would be similar for each.

Observations of nest attentiveness were further broken down by stage of the nesting cycle,
which I categorized as (1) nest construction, (2) incubation, (3) care of smaller young, and
(4) care of larger young. My reason for having two categories for the care of young was to
allow for the detection of any differences in attentiveness due to differences that might exist
in the energy needs of the chicks. Small young in my study are defined as nestlings of
approximately 1-7 days in age, whereas large young are defined as nestlings in excess of
seven days of age.

In this study, I depict nest attentiveness in terms of the average duration (data pooled) of
sessions and recesses for each stage, species, and area. These data have in turn been stan-
dardized to compare the frequency of the females’ arrivals and departures (number on/off
bouts) on an hourly basis (collectively referred to as their activity budget). Diurnal fluctu-
ations in attentiveness (e.g., morning vs evening) are based on the duration of sessions and
recesses during morning, mid-day, and evening periods. Data within these three periods
were averaged and then ranked for each species, stage of nesting, and area. Rankings show
the relative duration of sessions and recesses throughout the day (e.g., morning sessions the
shortest of the day, afternoon sessions the longest, and evening sessions of intermediate
duration). The frequency of “unscheduled” disruptions (e.g., sudden presence of a conspe-
cific, another species, predator, or a passing vehicle near the nest) and their influence upon
attentive patterns was enumerated via direct observation.

Statistical comparisons were made using nonparametric analy.ses of variance (Kruskal-
Wallis), coupled with multiple comparison tests (.see Day and Quinn 1989). Modified non-
parametric Student-Newman-Keuls’ multiple range tests (SNK) and Tukey’s honestly sig-
nificant difference tests (HSD) were used because they represent the extremes in perfor-
mance of a posteriori multiple comparison methods (Pimentel and Smith 1990). SNK tends
to form too many nonsignificant subsets (high error rate) and HSD too few (lower error

230

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 1

Duration (min) of Sessions (on) and Recesses (off) of Hummingbirds in Guadalupe

Canyon for All Stages of Nesting^

Stage

Broad-billed

(A)

Violet-crowned

(B)

Black-chinned

(C)

Significance
among species (*)

On

Off

On

Off

On Off

Species

On Off

Construction

N =

1

N

= 1

N = 8

Mean

1.5

1.1

0.6

0.5

1.2 3.0

A-B

SD

2.1

0.7

0.2

0.04

0.7 2.7

A-C

Effort

6.1

3.2

42.2

29.6

194.6 480.8

B-C

* *

Incubation

N =

3

N

= 2

N = 16

Mean

17.3

4.7

15.1

5.6

9.1 3.8

A-B

SD

7.6

1.9

8.2

1.5

5.8 2.4

A-C

*

Effort

435.0

125.4

247.9

101.3

1274.8 520.1

B-C

Small young

N

= 1

N = 4

Mean

—

—

10.2

7.1

5.5 9.9

A-B

SD

—

—

4.4

0.5

2.8 6.8

A-C

Effort

—

—

165.5

126.2

314.0 405.1

B-C

Large young

N =

2

N

= 3

N = 5

Mean

2.2

17.9

0.8

22.8

1.8 26.7

A-B

SD

2.0

5.5

0.3

10.4

1.7 9.8

A-C

Effort

49.9

209.7

24.7

658.2

45.2 475.2

B-C

• N = Number of nests sampled. Effort = total observation time (minutes).

rate). I used SNK to gauge the results of HSD and, of 72 comparisons, the two methods
differed in only 14 instances (19%). The 14 differences were distributed among the four
stages as follows: construction (12%), incubation (4%), small-young (3%), and large-young
(0%). I have been conservative in that my findings are based on Tukey’s HSD, with the
level of significance set at P < 0.05.

RESULTS

Duration of Alternating Bouts

Hummingbirds in Guadalupe Canyon. — During nest construction a rap-
id pace of sessions and recesses occurs (Table 1), punctuated by extended
recesses. Black-chins completed an average of 10 on/off bouts averaging
1.2 and 3.0 minutes, respectively, before taking recesses that averaged
15.8 min. Violet-crowns completed an average of 14 on/off bouts before
recesses averaging 9.2 min were taken. Data for Broad-bills are limited,
but additional observations (untimed) suggest that attentiveness in nest
construction is similar to the above. Attentiveness among species for the
remaining stages of the nesting cycle showed generally similar patterns
(Table 1 ). The only significant differences in incubation were in the length

Baliosser • NEST ATTENTIVENESS

231

Table 2

Duration (min) of Sessions (on) and Recesses (off) of Black-Chinned Hummingbirds
Among Three Areas for All Stages of Nesting"*

Stage

Guadalupe

(G)

Rucker

(R)

Cliff

(C)

Significance
among areas (*)

On

Off

On

Off

On

Off

Area

On Off

Construction

N =

8

N

= 1

N

= 2

Mean

1.2

3.0

0.9

2.4

0.5

0.5

G-R

SD

0.7

2.5

0.4

1.6

0.3

0.01

G-C

* *

Effort"*

194.6

480.8

11.4

21.5

16.4

24.5

R-C

*

Incubation

N =

16

N

= 10

N

= 7

Mean

9.1

3.8

8.9

3.9

17.1

7.0

G-R

SD

5.8

2.4

4.9

2.2

10.4

3.0

G-C

* *

Effort"*

1274.8

520.1

863.8

339.0

545.9

224.9

R-C

* *

Small young

N =

4

N

= 3

N

= 5

Mean

5.5

9.9

13.1

17.0

1 1.7

12.8

G-R

* *

SD

2.8

6.8

6.8

9.0

4.1

5.3

G-C

Effort"*

314.0

405.1

268.8

421.4

356.2

392.4

R-C

Large young

N =

5

N

= 2

N

= 3

Mean

1.8

26.7

0.8

13.5

0.7

25.9

G-R

SD

1.7

9.8

0.1

1.2

0.3

6.0

G-C

Effort"*

45.2

475.2

8.2

149.1

15.0

605.5

R-C

■ N = Number of nests sampled. Effort = total observation time (min).

of sessions between Broad-bills and Black-chins. For both the small-
young and large-young stages, there were no significant differences
among species.

Black-chinned Hummingbirds. — I found no significant differences
among stages between Guadalupe and Rucker canyons, except for the
small-young stage (Table 2). In this case, sessions and recesses were
shorter in Guadalupe Canyon. Comparisons between Guadalupe Canyon
and Cliff showed significant differences in the construction and incubation
stages. Sessions/recesses were longer in Guadalupe Canyon during con-
struction, whereas they were longer during incubation at Cliff. Between
Rucker Canyon and Cliff, significant differences existed between con-
struction and incubation. Construction sessions were longer in Rucker
Canyon (recesses did not differ between the two areas), while on/off bouts
during incubation were significantly shorter in Rucker Canyon.

Activity Budgets

Nest construction. — Female Broad-billed Hummingbirds spent 59.8%
of each hour on the nest, while the remaining 40.2% was spent away

232

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

Table 3

Comparisons among Stages of Nesting Showing the Percentage of Time Nesting
Lemale Hummingbirds Spent on and off Nests and the Associated Number of Bouts

(Scaled for 1-h Intervals)

Stage

Broad-billed

Guadalupe

Violet-

crowned

Guadalupe

Black-chinned

Guadalupe

Black-chinned

Rucker

Black-chini

Cliff

Construction

On

59.8

56.4

29.2

28.0

47.6

Bouts

23.4

56.4

14.8

19.0

59.0

Off

40.2

43.6

70.8

72.0

52.4

Bouts

23.0

56.0

14.0

18.4

59.0

Incubation

On

84.3

76.6

75.0

74.0

76.7

Bouts

2.9

3.0

4.9

4.9

2.7

Off

15.7

24.4

25.0

26.0

23.3

Bouts

2.0

2.6

4.0

4.0

2.0

Small young

On

—

64.7

36.6

43.7

57.3

Bouts

—

3.8

4.0

2.0

2.9

Off

—

35.3

63.4

56.3

42.7

Bouts

—

3.0

3.9

2.0

2.0

Large young

On

10.8

3.8

9.0

6.9

3.5

Bouts

3.0

3.0

3.0

5.0

3.0

Off

89.2

96.2

91.0

93.1

96.5

Bouts

3.0

2.5

2.1

4.1

2.2

(Table 3). They averaged 23.4 attentive bouts (mean duration 1.5 min.
Table 1) and 23.0 inattentive bouts (mean 1.1 min, Table 1) per hour
(Table 3). Violet-crowned Hummingbirds were similar to Broad-bills but
they differed from syntopic Black-chins in that the latter spent less time
at the nest and had fewer bouts. Black-chins in Rucker Canyon were
similar to those in Guadalupe Canyon, whereas those at Cliff were inter-
mediate between these two populations and Broad-bills and Violet-crowns
in Guadalupe Canyon (Table 3).

Incubation. — Attentiveness for this stage of the nesting cycle was sim-
ilar for all species and areas in my study. Black-chins in Guadalupe and
Rucker canyons and Violet-crowns spent virtually the same percentage
of time on and away from nests, but this was achieved through different
strategies (Table 3). Black-chins exhibited a greater frequency of on/off
bouts (Table 3), and these were of shorter duration than those of Violet-

Baltosser • NEST ATTENTIVENESS

233

crowns (Table 1). Broad-bills spent more time on their nests than either
of the above species, were away less, and had fewer bouts (Table 3).
Black-chins at Cliff were similar to all other species, although the fre-
quency of on/off bouts was most like Broad-bills (Table 3).

Small young. — In one segment of Guadalupe Canyon, I was able to
observe a Violet-crowned nest and a Black-chinned nest simultaneously.
The Black-chinned under surveillance was generally present when the
Violet-crowned was away and absent when it was present. Additionally,
the Violet-crowned was on its nest about the same average time that
Black-chins in the canyon were away and vice versa (Table 3). Black-
chins at Rucker and Cliff spent more time on their nests and were away
for shorter periods than Black-chins in Guadalupe Canyon. Black-chins
at all three sites had shorter sessions and longer recesses than Violet-
crowns (Table 3).

Large young. — All species in Guadalupe Canyon showed similar at-
tentiveness patterns, although Violet-crowns averaged less time on the
nest than Black-chins and Broad-bills (Table 3). Attentiveness in Black-
chins in Rucker Canyon was similar to that in this species in Guadalupe
Canyon, except that the former had shorter recesses (Table 1 and Table
2) and thus more bouts (Table 3). Black-chins at Cliff differed from those
in Guadalupe and Rucker canyons in terms of the number of bouts and
in total time spent on and away from nests. In these respects, this pop-
ulation was more like Violet-crowns than other Black-chinned popula-
tions.

Feeding young. — The time females spent feeding young did not differ
significantly among species or areas. Feeding sessions in Broad-bills av-
eraged 44 sec (N = 13, SD = 19.3), which is the same as in Violet-
crowns (N = 46, SD = 18.1). Black-chins in Guadalupe and Rucker
canyons had the same average time, which amounted to 5 1 sec (N = 30,
SD = 20.4 and N = 20, SD = 23.2, respectively). The feeding sessions
in Black-chins at Cliff averaged 43 sec (N = 29, SD = 16.3).

Diurnal Patterns

Hummingbirds in Guadalupe Canyon. — Attentiveness during incuba-
tion was shortest in the morning, longest during mid-day, and intermediate
during the evening (Fig. 1). Recesses for all species were longest in the
evening, which is the period preceding the nighttime fast. Violet-crowns
differed from Broad-bills and Black-chins (morning recesses shortest and
intermediate during mid-day) in that their recesses were of intermediate
duration during mornings and shortest during mid-day.

Data for Broad-bills during the small-young stage are limited to casual
observations, and thus comparisons are restricted to Black-chins and Vi-

234

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Broad-billed

Violet-crowned

Guadalupe

Guadalupe

On Off

On Off

Incubation

Morning

■ ■

■1

Mid-day

Evening

■1

Black-chinned

Black-chinned

Black-chinned

Guadalupe

Rucker

Cliff

On

Off

On Off

On Off

■

■

■i

■ ■

■■ *

Small Young

Morning

Mid-day

Evening

Large Young

Morning

Mid-day

Evening

Fig. 1. Diurnal patterns in attentive behavior among species and areas for different
stages of nesting; duration of bouts shortest (smallest polygon), duration intermediate m
length (medium polygon), and duration longest (largest polygon; bout duration based on
morning, mid-day, and evening averages).

olet-crowns (Fig. 1). Sessions were shortest during evening hours, longest
during mid-day, and of intermediate length during mornings, which is the
first opportunity to renew energy levels following nocturnal fasting. Re-
cesses for both species were of intermediate length during morning hours.
They were longest for Violet-crowns during mid-day and for Black-chins

during the evening.

When large young were present, each of the three species exhibited a
different pattern with regard to the duration of sessions (Fig. 1). However,
all species exhibited the same pattern for recesses in that they were short-
est in the morning, longest during mid-day, and of intei mediate duration
during evening periods. Comparisons among species show no other
shared patterns within stages of nesting, and relatively few patterns were
the same when comparisons were made among stages (Fig. 1).

Black-chinned Hummingbirds. — Comparisons among aieas reveal that
a number of patterns were similar in this species (Fig. 1 ). For example,
incubation sessions in Guadalupe and Rucker canyons were shortest dur-
ing mornings, which was in marked contrast to Cliff where they were
longest (response to lower ambient temperature?). Recesses were shortest
during mid-day in Rucker Canyon and Cliff, whereas they were longest
during evening hours in Guadalupe and Rucker canyons.

The single feature in common among areas was during the small-young

Baltosser • NEST ATTENTIVENESS

235

Stage (Fig. 1), with sessions shortest during evening hours (the last op-
portunity to renew energy reserves prior to nocturnal fasting). At Rucker
Canyon and Cliff, sessions and recesses were longest during morning
hours (first opportunity after the nighttime fast to renew energy levels).
Mid-day recesses were of intermediate duration in Rucker Canyon and
shortest in Guadalupe Canyon and Cliff.

When large young were present, sessions in Rucker Canyon and Cliff
were shortest during mid-day, longest in the morning, and of intermediate
duration during evening hours (Fig. 1). In Guadalupe Canyon, sessions
were shortest in the morning and longest in the evening. Recesses were
shortest during mornings in Guadalupe Canyon and Cliff, but in Rucker
Canyon they were longest. Recesses were of intermediate duration in the
evening for Guadalupe and Rucker canyons (they were most lengthy at
Cliff).

Nest Defense

Females often reacted to intrusions by other animals (even machinery)
when these transgressions were into areas near their nests (Table 4). In
some instances, this involved scolding or even attack by the female. Dis-
ruptions at other times resulted in the female disappearing for a brief
period, whereas in other cases they evidenced “curiosity” at the intrusion.
Conspecifics were the major cause of nest defense by Black-chins in
Guadalupe and Rucker canyons. Black-chins rarely responded in this
manner at Cliff, even though in several instances incubating females were
within sight of each other. Interspecific intrusions were also the major
cause of nest defense in Violet-crowns. In Broad-bills, nest defense was
less frequent but it was elicited equally by incursions by conspecifics and
other hummingbirds.

DISCUSSION
Diurnal Patterns

The incubation and small-young stages of the nesting cycle were most
similar to one another (Table 1 ). By contrast, the construction and large-
young stages had little in common and were each very different from the
other two stages. These relationships (or lack thereof) can be interpreted
in a number of ways. However, 1 assume that the similarity between the
incubation and small-young stages was due, at least in part, to thermal
constraints, small young being somewhat ectothermic after hatching. On
the other hand, large young are farther along in their development and
thus would not be expected to make the same demands upon the female
as developing embryos or small young, which is also true of nest con-

236

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 4

Lactors that Induced Lemale Broad-billed (BB), Violet-crowned (VC), and Black-
chinned (BC) Hummingbirds to Leave Their Nests“

Distractions/h by area
and hummingbird species

Guadalupe Rucker

Canyon Canyon Cliff

Source of distraction

BB

VC

BC

BC

BC

Cooper’s Hawk {Accipiter cooperii)

—

—

—

0.02

—

American Kestrel {Falco span’erius)

—

0.04

0.02

—

—

Broad-billed Hummingbird (Cynanthus latirostris)

0.11

0.08

—

—

—

Violet-crowned Hummingbird (AmaziHa violiceps)

0.17

—

0.02

—

—

Magnificent Hummingbird {Eugenes fut gens)

—

—

—

0.08

—

Black-chinned Hummingbird {Archilochus alexandri)

0.06

0.19

0.28

0.18

0.02

Hummingbird spp. (Trochilidae)

—

—

0.06

0.02

—

Acorn Woodpecker {Melanerpes formicivorus)

—

—

—

0.02

—

Gila Woodpecker {Melanerpes uropygialis)

—

—

0.02

—

—

Ash-throated Llycatcher {Myiarchus cinerascens)

—

—

0.06

—

—

Brown-crested Llyeatcher {Myiarchus tyrannulus)

—

—

0.02

—

—

Cassin’s Kingbird {Tyrannus vociferam)

—

—

0.05

—

—

Thick-billed Kingbird {Tyrannus crassirostris)

—

0.08

0.02

—

—

Gray-breasted Jay {Aphelocoma ultramarina)

—

—

0.05

—

—

Northern Mockingbird (Mitnus polyglottos)

—

—

0.02

—

—

Summer Tanager {Piranga rubra)

—

0.04

0.05

—

0.02

Black-headed Grosbeak {Pheucticus melanocephalus)

—

—

—

0.02

—

Canyon Towhee {Pipilo fuscus)

—

—

0.03

—

—

Hooded Oriole {Icterus cucullatus)

—

0.04

0.12

—

—

Evening Grosbeak {Coccothraustes vespertinus)

—

—

—

0.02

—

Unknown bird speeies

—

—

0.02

0.04

—

Cattle {Bos taunts)

—

0.04

—

—

—

People {Homo sapiens)

0.06

—

0.03

0.02

0.06

Bumblebee {Bombus sp.)

—

—

—

—

0.02

Pipevine Swallowtail Butterfly {Battus philenor)

—

—

0.02

—

—

Unknown

0.06

0.04

—

—

—

Total number of distraction.s/hour

0.46

0.55

0.89

0.42

0.12

• Broad-billed observations based on six nests and 17,5 h of direct observation; Violet-crowned observations based on
live nests and 26 I h of direct observation; Black-chinned observations for Guadalupe Canyon based on 25 nests and 67.4
h. for Rucker Canyon I I nests and 50.5 h. and for Cliff 10 nests and 47.7 h of direct observation.

struction. I have confined my discussion of diurnal patterns in attentive
behavior (Fig. 1) to the incubation and small-young stages and, in turn,
to hummingbirds in Guadalupe Canyon. This was done to avoid the con-
founding effects of differing selective pressures during other stages in the
nesting cycle and among geographic areas.

Incubation. — A consistent pattern of nest attentiveness relative to time
of day was evident among the hummingbirds in Guadalupe Canyon (Fig.

Ballosser • NEST ATTENTIVENESS

237

1). Morning sessions were shortest, which may well have resulted from
the needs of females to feed themselves after the nocturnal fast, balanced
against meeting the requirements of the developing embryos. Mid-day
sessions were longest during incubation when temperatures were highest
(average ambient temperatures [shade] of 32.6°C [SD = 0.8], with ex-
tremes of 31 to 38°C). This is contrary to many published expectations
(e.g., Kendeigh 1952, 1963; von Haartman 1956; White and Kinney 1974)
which suggest the existence of an inverse correlation between temperature
and attentiveness. The intermediacy of evening sessions (Fig. 1) and the
duration of evening recesses (longest of day for each species) are believed
to represent periods of crop-filling anticipatory to evening fasting. Such
an explanation is consistent with the feeding behavior of both free-living
and captive hummingbirds (Beuchat et al. 1979, Wheeler 1980, Powers
and Nagy 1988, Tiebout 1989, Powers 1991).

My findings regarding mid-day sessions might be interpreted as females
shading their eggs (see Wolf 1964, Vleck 1981). However, female posture
on the nest, and the fact that nests were not in direct sunlight, does not
support this interpretation. Instead, the results might indicate that females
remained on the nest as a way of minimizing energy expenditures or be
a strategy for heat dissipation without excessive water loss (see Ricklefs
1971, 1974; Calder 1974). Another possibility is that females remained
at the nest to foil ectothermic predators (e.g., snakes), many of which are
most active during the warmer portion of the day.

Small young. — Attentive patterns of Violet-crowned and Black-chinned
hummingbirds generally were similar (Fig. 1 ; data for Broad-bills lack-
ing). Afternoon sessions were longest, which strengthens the argument
that temperature and attentiveness were not inversely related. The inter-
mediacy of sessions in both species during morning sessions contrasts
with that for incubation when sessions were shortest. This may indicate
that newly hatched young either ( 1 ) required more parental care than
developing embryos or (2) that females expended less nighttime energy
and could “afford” to remain at the nest longer. That evenings were used
to “tank up” prior to fasting is supported in Black-chins by the fact that
sessions were shortest and recesses longest during this stage (same pattern
as in incubation).

Violet-crowned Hummingbird attentiveness during the evening period
differed from that of Black-chins, being marked by the shortest sessions
and recesses of the day. When the female was not at her nest, she was
often perched in sight of it for relatively long periods. This might suggest
that this period was not important for replenishing nectar supplies. How-
ever, plentiful and nearby Parry agaves {Agave parryi) were beginning to
produce large quantities of nectar at this time (Baltosser 1989b), thus

238

THE WILSON BULLETIN • Vol. JOS. No. 2, June 1996

providing an abundance of food. This fact, plus the dominance of the
Violet-crowned over Black-chinned and Broad-billed hummingbirds (Bal-
tosser 1989b), may well have allowed the former to “refuel” more quick-
ly (see Howell and Dawson 1954 for pertinent discussion).

Attentiveness Strategies

In most birds in which only one sex tends the eggs (excluding certain
nidifugous species), the eggs are incubated 60-80% of the daylight hours
(Skutch 1962). These levels of attentiveness curtail foraging, and under
some circumstances, incubating birds may have difficulty finding enough
food to maintain themselves (Skutch 1962, Walsberg 1983, Williams
1993). When levels fall below 60%, hatching may be retarded. However,
an increase in levels above 70-80% does not necessarily shorten the in-
cubation period (Skutch 1962).

In my study, the percentage of time hummingbirds covered their eggs
during incubation was similar among the three species (Table 3). Black-
chins and Violet-crowns in Guadalupe Canyon were very similar, as was
the case for Black-chins among areas (x = 75.2%, SD = 1.4). However,
despite the uniformity in overall time spent at the nest, the way in which
it was partitioned varied among species. For example. Broad-billed and
Violet-crowned hummingbirds had relatively few on/off bouts, which
were relatively longer (Table 1). By contrast, Black-chinned Humming-
birds had shorter sessions and recesses (at least for Guadalupe and Rucker
canyons) which resulted in more bouts.

Disruptions

Nest attentiveness during incubation and care of small young is clearly
affected by what might be termed “intrinsic” considerations, e.g., ther-
moregulation, self-maintenance, and feeding of young. In addition, “ex-
trinsic” considerations also affect attentiveness, the most important of
which is intrusions into the nest area by other animals. Whether the in-
truders are potential competitors, predators, or neutral in intent, female
hummingbirds can respond in one of two ways: passively or actively.
Two opposing tendencies are thus exhibited by female hummingbirds
while sitting on their nests: (1) to passively reduce interactions centered
around the nest and (2) to actively defend the nest against intruders (Wolf
and Wolf 1971, present study).

Comparative material is virtually lacking, but in the Purple-throated
Carib (Eiilampis jugularis). Wolf and Wolf (1971) found that nest defense
was strongest against birds that posed the greatest source of danger; it
declined as this potential diminished. Female Caribs often, though not
inevitably, responded to intrusions by leaving the nest in pursuit. Large

Baltosser • NEST ATTENTIVENESS

239

passerines were responsible for 51% of all disruptions, followed in de-
scending order by small passerines (29%), conspecifics (12%), and other
hummingbirds (9%).

Responses similar to those of the Purple-throated Carib were seen in
my study when females were faced with intrusions by animals near their
nests. However, compared to the Caribs, intrusions by conspecifics and
other hummingbirds played a more prominent role (Table 4). Conspecifics
were the major single source of disruption for Black-chins in both Gua-
dalupe and Rucker canyons, amounting to 34% and 43%, respectively.
The impact of other hummingbirds on this species was less in both areas,
being only 9% and 24%, respectively. In Guadalupe Canyon, the impact
of all other species on Black-chins was 57%, of which most were attrib-
utable to large passerines (Table 4). For Black-chins in Rucker Canyon,
34% of all interactions resulted from other non-hummingbird species,
primarily large passerines. Black-chins were the only nesting humming-
bird at Cliff, and despite the presence of an occasional migrant Broad-
tailed {Selasphorus platycercus). Rufous (Selasphorus rufus), or Calliope
{Stellula calliope), were not disrupted by these hummingbirds. Overall,
disruptions at this site were minimal, with 50% being generally attribut-
able to human interference.

Conspecific intrusions were not a factor for Violet-crowns, as the spe-
cies was relatively rare and nests were widely dispersed. However, intru-
sions by Black-chinned and Broad-billed hummingbirds accounted for
50% of all cases in which Violet-crowns left their nest. Larger passerines
were the other major source of disruption, accounting for 29% of the
intrusions to which Violet-crowns responded. In Broad-billed Humming-
birds, conspecific intrusions accounted for 25% of the aggressive re-
sponses by this species. In addition, defense of the nest against other
hummingbirds accounted for 50% of the responses of nesting Broad-bills.

The potential impact of intrusions around the nest can be examined by
noting how frequently females left their nests and their fledging success.
Comparisons of species nesting in Guadalupe Canyon show a significant
inverse correlation between these parameters (r = —0.996, df = 1, P <
0.05), based on 6663 minutes (111 h.) of intensive observation at 36 nests
(Baltosser 1986b). On average, fledging success for Black-chins was 44%,
with nearly one intrusion/hour (Table 4), 60% for Violet-crowns with just
over 0.5 intrusions/hour (Table 4), and 66.7% for Broad-bills with just
under 0.5 intrusions/hour (Table 4). Black-chinned nesting success vs in-
trusion rate was not significantly correlated in Rucker Canyon or at Cliff.

Comparing attentiveness at successful nests vs those that were unsuc-
cessful is another way of assessing the impact of intrusions around the
nest. Because of limited sample size, this is possible only for Black-

240

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

chinned Hummingbirds, and then only during incubation (stage with
greatest mortality, see Baltosser 1986b). Attentiveness for Black-chins in
Guadalupe Canyon that fledged averaged 75% (SD = 7.4, Range = 66-
84%) and unsuccessful nests averaged 67% (SD = 10.7, Range = 50-
81%). In Rucker Canyon, attentiveness was reversed, i.e., percentages for
nests that fledged were lower (T = 71%, SD = 10.4, Range = 61-84%)
than those which failed {x — 76%, SD = 3.9, Range = 71—81%). Black-
chins at Cliff that were successful had greater attentiveness {x = 77%,
SD = 7.2, Range = 68-86%) than those that were unsuccessful (x =
71%, SD = 5.8, Range = 67-75%).

Attentiveness in the Trochilidae

The data presented in this paper on nest attentiveness during incubation
are compared with those on other hummingbirds in Table 5. In addition
to studies already cited, this table includes data from Orr (1939), Skutch
(1951, 1958, 1964, 1967), Calder (1971, 1975), Smith et al. (1974), and
Montgomerie and Redsell (1980). Data are arranged by species and by
the primary area in which each nests (i.e., temperate, subtropical, and
tropical latitudes). Collectively, these data show that hummingbirds are
similar in incubation attentiveness regardless of species and latitude. This
suggests the existence of a fixed requirement for this stage of reproduc-
tion, which hummingbirds have been able to meet effectively throughout
a wide variety of nesting habitats (Vleck 1981).

Aside from the present study, comparative data for other stages of the
nesting cycle are generally unavailable for most hummingbirds. However,
comparisons are possible based on Wolf and Wolf’s study (1971) of the
Purple-throated Carib and that of Bene (1940) on the Black-chinned Hum-
mingbird. Female Caribs with small young spent 59.5% of each hour at
the nest, compared to 57-59% for Black-chins in the study by Bene.
These percentages are similar to my findings on Violet-crowned Hum-
mingbirds in Guadalupe Canyon and nearly identical to Black-chins at
Cliff (Table 2). Attentiveness in Black-chins in Guadalupe Canyon at this
stage was only 36.6%, while that in Rucker Canyon was 43.7%.

For nests having large young, data from Wolf and Wolf (1971) shows
attentiveness to average 8.5% (range 5.7—13.4%), based on three days of
observation. This average is nearly the same as the 9% obtained for
Black-chins by Bene (1940), and these are very similar to the 10.8%- in
Broad-bills and 9.0% in Black-chins 1 found in Guadalupe Canyon. In
contrast, my figures were 3.8% for Violet-crowns, 3.5% for Black-chins
at Cliff, and 6.9% for Black-chins at Rucker Canyon.

Wolf and Wolf (1971) found that feeding bouts for young in Purple-
throated Caribs averaged 38 sec (SD =11). This is similar to the 44 sec

Boltosser • NEST ATTENTIVENESS

241

Table 5

Comparisons among Temperate, Subtropical, and Tropical Nesting Hummingbirds
Showing Uniformity in Attentiveness During Incubation

Species of
hummingbird

Average length (min.)

Sessions Recesses ;

Percent

attentive

Source

Temperate

Black-chinned Hummingbird

6.3

2.3

70

Vleck 1981

(Archilochus alexandri)

8.9

3.9

74

Baltosser, present study

17.1

7.0

77

Baltosser, present study

9.1

3.8

75

Baltosser, present study

Anna’s Hummingbird

14.5

6.4

69

Calder 1975

(Calypte anna)

15.5

2.8

84

Howell and Dawson 1954

8.5

2.2

79

Smith et al. 1974

8.1

3.0

75

Vleck 1981

Costa’s Hummingbird

(Calypte costae)

15.8

2.4

83

Vleck 1981

Broad-tailed Hummingbird

7.7

2.1

78

Calder 1975

(Selasphorus platycercus)

—

—

72

Montgomerie and Redsell

1980

Allen’s Hummingbird

(Selasphorus sasin)

4.6

1.4

77

Orr 1939

Calliope Hummingbird

(Stellula calliope)

7.3

2.0 77

Subtropical

Calder 1971

Violet-crowned Hummingbird

(Amazilia violiceps)

15.1

5.6

77

Baltosser, present study

Broad-billed Hummingbird

(Cynanthus latirostris)

17.3

4.7

84

Baltosser, present study

White-eared Hummingbird

(Hylocharis leucotis)

9.1

4.0

Tropical

70

Skutch 1962

Violet Sabrewing (Campylop-

terns hernileucurus)

42.8

24.0

64

Skutch 1967

Purple-throated Carib

(Eu lamp is jugularis)

—

—

69

Wolf and Wolf 1971

Violet-headed Hummingbird

(Klais guimeti)

40.6

15.2

73

Skutch 1958

White-crested Coquette

(Lophornis adorabilis)

1 1.9

6.4

64

Skutch 1962

Scaly-breasted Hummingbird

(Phaeochroa cuvierii)

20.9

6.9

76

Skutch 1962

Little Hermit

24.7

1 1.7

67

Skutch 1951

(Phaethornis longuemareus)

40.7

16.4

71

Skutch 1962

36.9

15.3

70

Skutch 1964

242

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

that I found in Broad-billed and Violet-crowned hummingbirds and to the
43 sec in Black-chins at Cliff. Black-chins in Guadalupe and Rucker
canyons had bouts that averaged 51 sec, which is similar to what can be
deduced for this species from Bene (1940).

SUMMARY AND CONCLUSIONS

Hudson (1920) and Woods (1927) were among the first to comment
on the uniformity of general life-history traits among hummingbirds. Pi-
telka (1942) noted that this degree of uniformity is perhaps as extreme
as that in any group of similar taxonomic rank. The data from the present
study, coupled with those cited within, demonstrate considerable unifor-
mity in nest attentiveness within the Trochilidae. This is best exemplified
during incubation, but evidence of behavioral consistency is shown in all
stages. This is remarkable, given the disparity among the species and
latitudes from which comparative data were obtained.

The three hummingbird species that I studied in Guadalupe Canyon
exhibited similar attentiveness during the incubation and small-young
stages, i.e., stages of development when embryos and young were perhaps
most vulnerable to thermal fluctuations. Anticipatory feeding was com-
mon and often characterized the behavior of female hummingbirds during
hours immediately preceding the nocturnal fast. During morning hours
when ambient temperatures were lowest, the need to replenish energy
reserves depleted during the night was apparent in the behavior of nesting
females. However, presumably because of thermoregulatory needs of em-
bryos and newly-hatched young, attentive patterns seem to reflect a com-
promise between the female’s needs and those of her brood.

Lengths of sessions and recesses can provide insight into resource
availability and the competitive environment among hummingbirds. Fac-
tors that increase the comings and goings from a nest beyond those es-
sential for self-maintenance, incubation, or the feeding of young may play
a critical role. Success vs failure may hinge on extrinsic items that in-
crease the number of unscheduled departures from the nest. For example,
escalating the number of departures from a nest may increase predation
through nest betrayal (Skutch 1949, 1962). Low constancy during incu-
bation may also result in the death of embryos and young (e.g.. Grant
1982) or prolong incubation (Pienkowski 1984) and, thereby, the risk of
nest predation (Byrkjedal 1985).

Correlation is not the strongest form of inference (e.g., see Eberhardt
1970, Romesburg 1981), but results such as mine can provide valuable
insights into biological processes. Given the potential data to be gained
by focusing on interactions at the nest, it is surprising that more studies
have not pursued this line of investigation. My research and that con-

Bultosser • NEST ATTENTIVENESS

243

ducted by Wolf and Wolf (1971) demonstrate the need to investigate the
extent to which female hummingbirds are induced by outside forces to
leave their nests. The consequences of such disruptions (including fre-
quency and duration) are potentially great and are a much overlooked
aspect of the breeding biology of hummingbirds.

ACKNOWLEDGMENTS

I thank R. J. Raitt and J. R Hubbard for their help throughout all aspects of this study.
Appreciation is extended to S. M. Russell, E N. White, W. M. Shepherd, C. R. Blem, and
three anonymous reviewers for their guidance and encouragement. I am indebted to the
Magoffin and Hadley families of Guadalupe Canyon, the personnel of the Coronado National
Forest in Rucker Canyon, and the Hunt family at Cliff. I also am extremely grateful to Mr.
and Mrs. W. W. Baltosser, my wife Ginger, and my daughter Dianna. This work was sup-
ported in part by the New Mexico State Univ. Dept, of Biology, the New Mexico Dept, of
Game and Fish (contract No. 519-68-06), and Chapman Memorial grants (61462, 84472,
and 110115) from the American Museum of Natural History.

LITERATURE CITED

Baltosser, W. H. 1986a. Seasonal analysis of a southwestern New Mexico riparian bird
community. Western Birds 17:115-131.

. 1986b. Nesting success and productivity of hummingbirds in southwestern New

Mexico and southeastern Arizona. Wilson Bull. 98:353-367.

. 1989a. Costa’s Hummingbird: its distribution and status. Western Birds 20:41-62.

. 1989b. Nectar availability and habitat selection by hummingbirds in Guadalupe

Canyon. Wilson Bull. 101:559-578.

Bene, F. 1940. Rhythm in the brooding and feeding routine of the Black-chinned Hum-
mingbird. Condor 42:207-212.

Beuchat, C. a., S. B. Chaplin, and M. L. Morton. 1979. Ambient temperature and the
daily energetics of two species of hummingbirds, Calypte anna and Selasphorus rufous.
Physiol. Zool. 52:280-295.

Byrkjedal, I. 1985. Time-activity budget for breeding Greater Golden-Plovers in Norwe-
gian mountains. Wilson Bull. 97:486-501.

Calder, W. a. 1971. Temperature relationships and nesting of the Calliope Hummingbird.
Condor 73:314-321.

. 1974. Consequences of body size for avian energetics. Pp. 86-151 in Avian en-
ergetics (R. A. Paynter, ed.). Publ. Nuttall Ornithol. Club No. 15.

. 1975. Daylength and the hummingbirds’ use of time. Auk 92:81-97.

Day, R. W. and G. P. Quinn. 1989. Comparisons of treatments after an analysis of variance
in ecology. Ecol. Monogr. 59:433-463.

Eberhardt, L. L. 1970. Correlation, regression, and density dependence. Ecology 51:306-
310.

Grant, G. S. 1982. Avian incubation: egg temperature, nest humidity, and behavioral
thermoregulation in a hot environment. Ornithol. Monogr. No. 30.

Howell, T. R. and W. R. Dawson. 1954. Nest temperatures and attentiveness in the Anna
Hummingbird. Condor 56:93—97.

Hudson, W. H. 1920. Birds of La Plata. J. M. Dent and Sons Ltd., London and Toronto.
Kendeigh, S. C. 1952. Parental care and its evolution in birds. 111. Biol. Monogr. 22:1—357.

244

THE WILSON BULLETIN • Vol. 108, No. 2. June 1996

. 1963. Thermodynamics of incubation in the House Wren, Troglodytes aedon. Proc.

Int. Ornithol. Congo 13:884-904.

Montgomerie, R. D. and C. A. Redsell. 1980. A nesting hummingbird feeding solely on
arthropods. Condor 82:463—464.

Orr, R. T. 1939. Observations on the nesting of the Allen Hummingbird. Condor 41:17-24.
PiENKOWSKi, M. W. 1984. Breeding biology and population dynamics of Ringed Plovers
{Charadrius hiaticula) in Britain and Greenland: nest-predation as a possible factor
limiting distribution and timing of breeding. J. Zool. (London) 202:83—114.

Pimentel, R. A. and J. D. Smith. 1990. Biostat I: a univariate statistical toolbox (Version
2.0). Sigma Soft, Placentia, California.

PiTELKA, E A. 1942. Territoriality and related problems in North American hummingbirds.
Condor 44: 1 89-204.

Powers, D. R. 1991. Diurnal variation in mass, metabolic rate, and respiratory quotient in
Anna’s and Costa’s hummingbirds. Physiol. Zool. 64:850—870.

AND K. A. Nagy. 1988. Eield metabolic rate and food consumption by free-living

Anna’s Hummingbirds (Calypte anna). Physiol. Zool. 61:500-506.

Ricklefs, R. E. 1971. Eoraging behavior of Mangrove Swallows at Barro Colorado Island.
Auk 88:635-651.

. 1974. Energetics of reproduction in birds. Pp. 152—297 in Avian energetics (R. A.

Paynter, ed.). Publ. Nuttall Ornithol. Club No. 15.

Romesburg, H. C. 1981. Wildlife science: gaining reliable knowledge. J. Wildl. Manage.
45:293-313.

Scott, P. E. 1994. Lucifer Hummingbird {Calothorax lucifer). In The birds of North Amer-
ica, No. 134 (A. Poole and E Gill, eds.). The Academy of Natural Sciences of Phila-
delphia; The American Ornithologists’ Union, Washington, D. C.

Skutch, a. E 1949. Do tropical birds rear as many young as they can nourish? Ibis 91:
430-455.

. 1951. Life history of the Longuemares’ Hermit Hummingbird. Ibis 93:180-195.

. 1958. Life history of the Violet-headed Humminghird. Wikson Bull. 70:5-19.

. 1962. The constancy of incubation. Wilson Bull. 74:1 15-152.

. 1964. Life histories of hermit hummingbirds. Auk 81:5-25.

. 1967. Life histories of Central American highland birds. Publ. Nuttall Ornithol.

Club No. 7.

Smith, W. K., S. W. Roberts, and P. C. Miller. 1974. Calculating the nocturnal energy
expenditure of an incubating Anna’s Hummingbird. Condor 76:176-183.

Tiebout, H. M. 1989. Tests of a model of food passage rates in hummingbirds. Auk 106:
203-208.

Vleck, C. M. 1981. Hummingbird incubation: female attentiveness and egg temperature.
Oecologia 51:199—205.

VON Haartman, L. 1956. Der einfluss der temperatur auf den brutrhythmus experimentell
nachgewiesen. Ornis. Fenn. 33:100-107.

Walsberg, G. E. 1983. Avian ecological energetics. Pp. 161-220 In Avian biology (D. S.

Earner and J. R. King, eds.). Vol. 8. Academic Press, New York, New York.

Wheeler, T G. 1980. Experiments in feeding behavior of the Anna Hummingbird. Wilson
Bull. 92:53-62.

White, F. N. and J. L. Kinney. 1974. Avian incubation. Science 189:107-115.

Williams, J. B. 1993. Energetics of incubation in free-living Orange-breasted Sunbirds in
South Africa. Condor 95:1 15-126.

Baltosser • NEST ATTENTIVENESS

245

Wolf, L. L. 1964. Nesting of the Eork-tailed Emerald in Oaxaca, Mexico. Condor 66:51—55.

AND J. S. Wolf. 1971. Nesting of the Purple-throated Carib Hummingbird. Ibis

1 13:306-315.

Woods, R. S. 1927. The hummingbirds of California. Auk 44:297—318.

Wilson Bull, 108(2), 1996, pp. 246-267

SEASONAL POPULATION SURVEYS AND NATURAL
HISTORY OL A MICRONESIAN BIRD COMMUNITY

Robert J. Craig

Abstract. — I replicated quarterly population surveys of landbirds on Saipan, Mariana
Islands at two environmental scales; habitat specific and island-wide. I determined popula-
tion densities and the degree of seasonal fluctuation in counts to compare densities in native
vs disturbed habitat and to observe whether populations exhibited characteristics of those
at either saturation or below saturation densities. I also gathered new data on the natural
history of largely unknown species. For seven of the nine forest birds examined, inter-
seasonal census variation was greater than intra-seasonal variation, suggesting that most of
the species undergo seasonal shifts in population or breeding status (the latter case was
indicated for four forest species). The principal difference uncovered between the two census
scales was that the Micronesian Honeyeater {Myzomela ruhrata) was relatively uncommon
in native forest but regular on island-wide counts. Otherwise, forest species showed nu-
merous similarities in count trends at both scales. However, habitat-specific data showed
that for many species, counts and computed densities were greater in native forest than in
disturbed habitat. Independent density assessment (based on a new procedure) for the Bridled
White-eye (Zosterops conspicillatus) was of the same order ot magnitude as that obtained
through censusing. The densities reported here, particularly for the Rufous Fantail (Rhipi-
diira rufifrons). Bridled White-eye, and Golden White-eye (Cleptornis marchei), are among
the highest ever reported for birds (>1900/km^) and are almost certainly at habitat saturation.
Interspecific competition is expected in such a case, and interspecific aggression was prev-
alent, particularly among ecologically similar species. Received 27 April 1995, accepted 1
Dec. 1995.

Land birds of the Mariana Islands, Micronesia have received limited,
mostly qualitative study (e.g., Marshall 1949, Baker 1951, Pratt et al.
1987, Reichel and Glass 1991), and the quantitative ecology of most
species remains unknown. Jenkins (1983) reviewed aspects of the natural
history of the now mostly extinct (Savidge 1987) avifauna of the south-
ernmost island of Guam, and Engbring et al. (1986) reported population
estimates, based on one survey, for Rota, Aguiguan, Tinian, and Saipan.
Quantitative scrutiny has been given only to the Nightingale Reed-War-
bler (Acrocephalus luscinia) (Craig 1992a) and Bridled White-eye {Zos-
terops conspicillatus) and Golden White-eye {Cleptornis marchei) (Craig
1989, 1990).

The island of Saipan presently has the most diverse, albeit meager,
assemblage of land birds in the Marianas. It consists of three medium-
sized predators, the Yellow Bittern {Ixobrychus sinensis). Collared King-
fisher {Halcyon chlori.s) and Nightingale Reed-warbler; two herbivores.

Northern Mariana.s College, P.O. Box I2.‘i(), Commonwealth No. Mariana fslands, Saipan. MP 969.‘)0.
Present address: 90 Liberty Highway. Putnam. Connecticut 06260.

246

Craig • MICRONESIAN BIRD COMMUNITY

247

the Mariana Fruit-Dove (Ptilinopus roseicapilla) and White-throated
Ground Dove {Gallicolumba xanthonura)', four omnivores, the Microne-
sian Megapode {Megapodius laperoiise), Micronesian Starling {Ap/onis
opoco). Golden White-eye, and Bridled White-eye, a nectarivore, the Mi-
cronesian Honeyeater {Myzoniela rubrata)\ and two small insectivores,
the Island Swiftlet {Aerodramus vanikorensis) and Rufous Fantail (Rhip-
idura rufifrons). Most of these species, or at least their local subspecies,
are endemic to the Marianas or Micronesia. Prehistorically, perhaps twice
as many species were present (Steadman 1992). Two other species, the
Javanese Turtle Dove (Streptopelia bitorquata) and Eurasian Tree Spar-
row {Passer montanus), are present but not native.

This study reports replicated, quarterly population surveys I made of
these species on Saipan. They were made at two environmental scales,
habitat specific and island-wide, to determine population densities and the
degree of fluctuation in populations or breeding activity. Because all spe-
cies are nonmigratory, I hypothesized that populations might build to the
maximum density sustainable by available resources and that little fluc-
tuation in densities generally occurs. I also gathered new natural history
data on many species.

STUDY AREAS AND METHODS

Habitats. — The island of Saipan is predominantly a raised coral island 22 km long and
3-10 km wide. It has a climate with uniform temperatures but seasonal rainfall. Typically,
and during this study, the dry season is December-June and the wet season is July-Novem-
ber. Reduced rainfall, establishment of easterly trade winds (Young 1989), and decline in
flowering, fruiting, and growth by certain native tree and vine species characterize the dry
season. During the wet .sea.son rain increases, particularly August-September, trade winds
break down (Young 1989), and beginning in the late dry season, many native trees and
vines flower and fruit. Typhoons are frequent during the latter half of the year and exert a
strong influence on the structure of natural habitats (Fosberg I960).

Much of Saipan likely was once forested, particularly on limestone soils (Fosberg 1960),
Such limestone forest is relatively xerophytic except at the highest elevations (ca 300-466
m), where near cloud forest conditions prevail. This forest is typically den.se, with a canopy
dominated by two widespread Indo-Pacific trees, Pisonia grandis and Cynometra ramijiora,
and understory of mostly C. ramijiora and the Mariana endemic Guamia mariannae (Craig
1992b). Other natural habitats, including ravine forest, swordgrass (Mi.scanthus fioridus)
savannah (both occurring on exposures of volcanic soil), mangrove swamp, freshwater-
swamp, reed (Phragmites karka) marsh, strand forest, and coastal scrub are also present.
Combined, native habitats presently cover roughly 30% of the island.

The remainder of Saipan’s natural habitats have developed on disturbed sites. Level areas
are largely abandoned agricultural lands (Fosberg 1960) vegetated by elephant grass {Pen-
nisetum ptirpureum) meadows, and tangantangan (Leucaena leucocephala) thickets. Sec-
ondary forests of introduced species, particularly sosuge (Acacia confusa), white monkeypod
(Albizia lebheck), and flametree {Delonix regia) are also common, as are areas of “agrifo-
rest” (Fngbring et al. 1986) where trees such as coconut {Cocos nucifera) and r-nango
(Mangifera indica) are frequent.

248

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Bird censuses— \ performed two types of bird censuses; variable circular plot (Reynolds
et al. 1980) surveys and U.S. Eish and Wildlife Service breeding bird surveys. The former
were conducted in limestone forests of the Marpi region of northern Saipan (Eig. 1). I
censused two separate locations. The first was an old Japanese hiking trail, the Banadero
Trail, located along the west slope of a steep escarpment known as Suicide Chtf. The second
was a modern hiking trail along the Laderan Tangke cliffline. Marpi is characterized by
steep limestone escarpments with the most extensive native forest remaining on the island.
The breeding bird survey traversed the island from north to south and covered a variety of

habitats. . , .

1 used the variable circular plot (VCP) technique, chosen because of its utility in rough

tropical terrain (Scott et al. 1986), to survey 15 points each at the two census routes (30
total points). Based on the frequency with which birds provided cues, I established count
periods of 8 min/station. Points were 100 m apart except at Laderan Tangke, where one set
of stations was placed 200 m apart and another 300 m apart to avoid disturbed habitat. At
each point, 1 recorded all birds seen or heard and estimated the distance of each bird Irom

Craig • MICRONESIAN BIRD COMMUNITY

249

Table 1

Detection Distances (m) Used for Computing Population Estimates of Birds in

Disturbed Habitats, Limestone Forest, and

1986)

IN THE

1982 Survey (Engbring et al.

Disturbed

Limestone

habitats

forest

1982

Species

Distance

N

Distance N

Distance

Micronesian Megapode

0

0

80

32

105

White-throated Ground Dove

a

—

20

51

80

Mariana Fruit Dove

70

20

50

155

159

Collared Kingfisher

70

56

50

283

193

Rufous Fantail

20

127

10

706

58

Nightingale Reed-warbler

50

34

0

0

87

Micronesian Starling

40

30

15

220

66

Micronesian Honeyeater

25

54

15

52

58

Bridled White-eye

15

590

10

2291

33

Golden White-eye

20

70

10

615

42

* Because an insufficient sample was available, the distance estimate for limestone forest was used in computations.

the point. Censuses began at sunrise and were conducted under conditions of minimal rain
and low wind (although wind averaged higher in the dry season). Practice censuses were
conducted in October 1990, and censuses were made at quarterly intervals thereafter in
January, April, July, and October, 1991-1992. Replicate data were, therefore, available for
each year and for the wet and dry seasons. I also made five replicate censuses each at the
two routes from late April to mid-May 1993 in order to assess within-season variation in
counts.

Although 1 attempted to calibrate distance estimates by placing plastic flagging at 10, 15,
and 20 m intervals (the maximum distance easily visible in limestone fore.st) from selected
census points and by walking from the point to distantly vocalizing birds during practice
censuses, distances were difficult to estimate (Table 1). Indeed, correctly estimating the
distance to the roughly 15 birds/census point, under varying conditions of topography, veg-
etation density, and orientation of the bird to the point, even for an ob.server with 20 years
of censusing experience, seemed an unrealistic expectation. Hence, population densities
derived from such estimates are of limited accuracy. I report computed densities and make
independent assessments of their utility but use direct counts for many analyses and rec-
ommend that future studies compare counts rather than densities when possible.

In addition to these regularly performed surveys, I employed the VCP procedure at three
disturbed sites to provide data comparable with those for native forest. Using the same
procedures outlined above (except that points were 150 m apart to improve sampling in-
dependence in habitats in which birds could be detected farther), 1 censused 25 points at
Laderan Hakmang (Kagman), 17 points at Sabanan Fiiang, and 17 points at Mt. Takpochao
(Tapotchau) in March 1993. Laderan Hakmang, the site of a former World War II fighter
field complex, is presently a xeric mosaic of meadows, tangantangan thickets, and .scattered
introduced and native tree species. Sabanan Fiiang, formerly a World War II hospital site,
is similarly xeric and largely overgrown by tangantangan and scattered introduced and native
trees, particularly ironwood (Casuarina equisetifolia). The Mt. Takpochao area, at least in

250

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

part a pre-war coffee (Coffea arahica) plantation (ca 300 m elevation), is a mesic mosaic
of meadows, swordgrass savannah, thickets, and copses of native forest.

The breeding bird survey involved censusing for 3 min at each of 50 roadside stops
placed 0.8 km (0.5 mi) apart. Counts began 15 min before sunrise on days with low wind
and little rain, and all birds seen or heard to 0.4 km away were recorded. Because the quality
of back roads on Saipan is poor, the entire survey took ca. 4.5-5 h to complete. Moreover,
the limited availability of back roads necessitated breaking the route into two segments
(after station 23) in order to traverse the entire island. Surveys also were conducted quarterly
in 1991-1992.

Additional observations. — to investigate additional aspects of avian populations and social
systems, I mist netted and color banded small passerines in the Marpi native forest and at
Capitol Hill. This second site facilitated study of the Micronesian Honeyeater which was
uncommon in limestone forest but common in suburban settings.

Intensive banding of Bridled W^hite-eyes at Capitol Hill provided an assessment of pop-
ulation densities independent of those obtained through bird censuses. I banded white-eyes
intermittantly from Lebruary 1992 to June 1993, and in Lebruary and May 1993, I recorded
the proportion of banded vs unbanded birds present within a 50 m radius of the banding
site. To determine population distribution, in May 1993 I also assessed the proportions of
banded vs unbanded birds at 50 m intervals to 300 m from the banding site.

I made incidental observations on all Saipan land birds throughout my investigations. I paid
particular attention to occurrences of interspecific aggression, and I assessed intraspecific ag-
gression by playing back recorded songs to selected species. Data on breeding, foraging, and
microhabitat use also were gathered. Lrom 1988 to 1993, I made additional observations on
the nearby Mariana Islands of Tinian (4 d). Rota (69 d), and Aguiguan (6 d).

Analysis. — I used the procedure described by Scott et al. ( 1988) and followed by Engbring
et al. (1986) to compute population densities.

Lor two loudly vocal and wide ranging species, the Mariana Eruit Dove and Collared
Kingfisher, VCP census points 100 m apart were inadequate to ensure that observations
from each point were independent. For these species, I computed population densities based
on 16 alternate census points (at least 200 m apart). Micronesian Megapodes also were
detectable at long distances, but because they were .sedentary and rare, I was able to distin-
guish the locations of all individuals encountered.

To obtain independent population estimates for the Bridled White-eye, I employed the
Jolly-Seber procedure (Tanner 1978) to analyze capture-recapture data from banding oper-
ations. In addition, I used the Lincoln-Peterson index (Tanner 1978) to evaluate populations
based on the relative proportions of banded and unbanded birds observed directly around
(to 50 m) the banding site (the region assumed to include intersections ot home ranges of
all birds banded). An assumption of the Jolly-Seber procedure, random sampling of banded
and unbanded members of the population, may not have been met because previously cap-
tured birds might become net shy. Moreover, the Lincoln-Peterson index requires that no
recruitment occur during the study period, an assumption not met during the extended study
period. Hence, population estimates based on both methods, particularly the latter, are likely
inflated.

To compute population densities, P, from the above indices, 1 employed data gathered on
the dispersion of marked birds from the banding site. I developed a relation using the number
of birds with home ranges intersecting the banding site (A/) as generated lrom the two
methods above, the area (A) of each of / zones radiating from the banding site at 50 m
intervals, the multiple (aj of the basal zone (0-50 m from the banding site) area (A;) of
each A, and the proportion of birds banded in each of these areas {Pi).P = N/X|(r//>,). P was
converted to birds/ha by dividing it by the area ol the basal zone, 7853 m .

Craig • MICRONESIAN BIRD COMMUNITY

251

Table 2

Comparative Counts of Birds (Birds/10 Stations) for I99I-I992 (Limestone Forest),

THE 1993 Survey of Disturbed Sites, and the 1982 Survey (Engbring et al.

1986)

Species

Limeslone

forest

Disturbed

sites

1982

survey

1991

1992

Micronesian Megapode

1.0

1.3

0.2

Yellow Bittern

—

—

0.2

0.6

White-throated Ground Dove

1.7

2.1 (2.8)“

0.7

0.6

Mariana Fruit Dove

5.9 (5.6)'’

5.9 (6.8)“ (7.0)'’

3.6

20.0

Collared Kingfisher

10.6 (12.5)'’

10.3 (12.3)'’

10.0

13.6

Rufous Fantail

25.3

28.4

23.0

45.0

Nightingale Reed-warbler

—

—

6.1

1 1.8

Micronesian Starling

8.5

7.5

5.4

4.7

Micronesian Honeyeater

2.0

2.3

9.8

22.6

Bridled White-eye

87.3

88.7

107.1

77.0

Golden White-eye

22.1

23.9

12.5

30.4

• April counts of species with seasonal shifts in calling frequency.
Based on 1 6 stations spaced 200 m apart.

SPECIES ACCOUNTS

Micronesian Megapode. — Believed to have become extinct on Saipan
after the early 1930s, it was rediscovered in 1978 by Pratt et al. (1987).
This present population, estimated at 25-50 by Glass and Aldan (1987)
is suspected to be descended from birds brought to Saipan from more
northern Mariana islands by island inhabitants (Engbring et al. 1986).
During this study, I estimated 14 birds to be present in native forests (and
occasionally in adjacent tangantangan thickets) of the Marpi region. In
1989 I also heard a bird farther south at Laderan Papao, although I found
none there in later years. Despite intensive surveys, I located none at
Naftan Point, the southernmost point on Saipan, where Glass and Aldan
(1987) previously reported individuals. Hence, populations are likely de-
clining. Both direct counts (Table 2) and density estimates (Table 3) for
limestone forest were greater than those recorded in 1982 by Engbring et
al. (1986), but the present VCP transects overlapped the only remaining
range of the species on Saipan, whereas Engbring et al. (1986) surveyed
all habitats throughout the island.

No firm evidence of breeding by this endangered species is known from
Saipan. However, in 1991 I located the possible remains (soil and rotting
vegetation) of an old nest mound in the Marpi forest, similar in dimen-
sions to those which I have observed in interior forests of the Palau
Islands (where a different subspecies occurs). Glass and Aldan (1987)

252

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 3

Comparative 1991-1992 Density Estimates (Birds/Km^) for Birds of Limestone
Lorest, the 1993 Survey of Disturbed Sites, and the 1982 Survey (Engbring et al.

1986)

Species

Limestone

forest

Disturbed

sites

1982

survey

1991

1992

Micronesian Megapode

2

3

—

1

White-throated Ground Dove

58

72

24

2

Mariana Fruit Dove

27 (26)='

27 (32)

1 1

25

Collared Kingfisher

43 (51)

42 (50)

26

1 1

Rufous Eantail

2160

2423

647

456

Micronesian Starling

403

356

48

32

Micronesian Honeyeater

123

138

205

203

Bridled White-eye

5904

5994

3992

2221

Golden White-eye

1935

2095

390

532

“Numbers in parentheses are population estimates based on 16 stations 200 m apart.

suspected a peak in calling (and breeding activity) in January, but in 1991
both limestone forest and island-wide surveys showed a calling peak in
July (Figs. 2A, 3A), the month in which I saw two birds engaging in
apparent courtship chases (the birds otherwise foraged together and
showed no evidence of aggression). This pattern was not repeated in 1992,
although replicate counts performed in April 1993 (Table 4) suggested
that census variation between seasons was greater than that within a sea-
son. Individuals or pairs were sedentary, responded vigorously to play-
back, and appeared to defend all-purpose territories. Birds could be found
in the same areas even between years, although during the study period
they invaded new locations on two occasions, thus providing evidence
for either territory relocation or reproduction.

Baker (1951) reported that the Micronesian Megapode is omnivorous,
although field observations on foraging are virtually nonexistent. I re-
corded feeding only once, when I observed an individual capture an in-
sect. Foraging birds generally scratched leaf litter with the feet and, at
least occasionally, scratched alternately with one foot and then the other.

Eig. 2. Mean 1991-1992 population trends of land birds for island-wide counts. (A)
MIME — Micronesian Megapode, WTGD — White-throated Ground Dove, MEDO — Mariana
Emit Dove, COKI — Collared Kingfisher, MIHO — Micronesian Honeyeater, YEBI — Yellow
Bittern; (B) RUEA — Rufous Eantail, MIST — Micronesian Starling, BWEY — Bridled White-
eye, GWEY — Golden White-eye, NRWA — Nightingale Reed-warbler.

Craig • MICRONESIAN BIRD COMMUNITY

253

c

D

O

o

Month

MIME -B- WTGD MFDO

COKI MIHO YEBI

RUFA -e- MIST BWEY
GWEY NRWA

254

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

-m- MIME WTGD -A- MFDO

COKI MIHO

-m- RUFA -e- MIST BWEY GWEY

Lig. 3. Mean 1991-1992 population trends of land birds for limestone forest counts.
See Lig. 2 for legend abbreviations.

Craig • MICRONESIAN BIRD COMMUNITY

255

Table 4

Coefficients of Variation for 1 99 1- 1 992 Limestone Forest and Island- wide Surveys,
AND FOR 1993 Replicated (5 times) VCP Surveys

Species

Limestone forest

Island-wide

survey

1991-1992

1993

1991-1992

Micronesian Megapode

43.2

27.2

82.8

Yellow Bittern

—

—

33.9

White-throated Ground Dove

57.0

64.4

52.9

Mariana Fruit Dove

50.3

35.2

62.7

Collared Kingfisher

15.1

22.9

15.0 .

Rufous Fantail

13.4

8.5

14.4

Nightingale Reed-warbler

—

—

29.9

Micronesian Starling

19.0

9.8

22.6

Micronesian Honeyeater

32.4

17.6

11.6

Bridled White-eye

13.4

4.3

18.5

Golden White-eye

18.5

5.0

12.9

Gut contents from two individuals collected on islands north of Saipan
contained spiders, insects, seeds, and plant fragments (Stinson 1993a).

Yellow Bittern. — This species is typically categorized as a water bird
(e.g. Engbring et al. 1986), and it indeed foraged in ponds, marshes, tidal
flats, and shorelines. However, the Yellow Bittern also inhabited upland
habitat mosaics in which grasslands were an important part. It was absent
from the forests of the VCP transects (Table 2) but occurred uncommonly
on the island-wide survey and showed weak October peaks each year
(Fig. 2A). Direct counts were low at disturbed sites compared to those
reported by Engbring et al. (1986) (Table 2), but my sample size was too
limited for a valid comparison to be made or for population denities to
be computed.

I recorded nesting in February (eggs) in a patch of elephant grass sur-
rounded by tangantangan. Birds were seen in pairs and were observed
flying hundreds of meters, thus suggesting that no all purpose territory
existed. Observations of foraging were limited to two captures of lizards
in upland habitat.

White-throated Ground Dove. — Although fairly common (but reported
as rare by Jenkins and Aguon 1981), based on the frequency with which
flying birds were seen, the species was otherwise visually inconspicuous
and called infrequently. Such characteristics resulted in its being poorly
censused. However, birds were usually encountered at close range (Table
1) and, therefore, densities computed (Table 3) were high relative to fruit
doves.

256

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Despite under-representation on censuses, three of four annual counts
made at the two environmental scales peaked in April-July (Figs. 2A,
3A). Such a trend likely indicated an increase in breeding activity during
those months (most census detections were of vocalizing birds). Indeed,
Stinson (1993b) reported that all 14 nests in the Division of Fish and
Wildlife (DFW) files were found between April and September. However,
Jenkins (1983) provided evidence that the extinct Guam population could
breed year round.

The cyclic nature of counts led to high coefficients of variation for
census data compared to Saipan passerines, although even variation with-
in a season was high (Table 4). Based on direct counts, the species ap-
peared more frequent in native forest than in disturbed habitats and more
frequent than found by Engbring et al. (1986) (Table 2). Computed den-
sities followed similar trends (Table 3).

The White-throated Ground Dove on Saipan, Rota, and Aguiguan used
a range of forest strata (Table 5), including the ground (N = 60). These
observations contrast with those of Marshall (1949), Baker (1951), Jen-
kins (1983), and Engbring et al. (1986) who considered the species to be
largely or entirely arboreal. On Saipan, it occurred in native forest, sec-
ondary forest, agriforest, tangantangan thickets, and habitat mosaics that
included fields. In such habitats, it flew for at least 300 m above the
canopy, suggesting that it did not defend all purpose territories.

Foraging observations included feeding on the ground on seeds and
probing leaf litter (4), feeding on fruits of the native trees Melanolepsis
multiglandulosa (1) and Premna obtusifolia (2), and inspecting papaya
(Carica papaya) fruits (1). Marshall (1949), Jenkins (1983), and Villa-
gomez (1987) list additional fruits eaten. Many members of the genus are
forest understory herbivores (Beehler et al. 1986), but the White-throated
Ground Dove appears to be more of a microhabitat generalist.

Mariana Fruit Dove. — All counts showed evidence of a population
peak in April-July, although the trend was most pronounced in island-
wide data (Figs. 2A, 3 A). Data from 1983—84 and 1987 roadside call
counts on Saipan (Villagomez 1987) showed a similar trend. As with the
White-throated Ground Dove, these peaks appeared to represent increases
in breeding activity rather than population cycles. Most census detections
were of calling birds, and fruiting peaks by certain common native tree
and vine species (e.g.. Ficus spp., Premna obtusifolia, Jasminum marian-
um\ unpubl. data) corresponded with these high counts. Wet season in-
creases in breeding are known for New Guinea Fruit Doves (Frith et al.
1974). 1 recorded breeding in February (carrying nesting material). May
(egg), and July (nestling), and Stinson (1993b) reported that 35 of 38
nests in DFW files were found between April and September.

Craig • MICRONESIAN BIRD COMMUNITY

257

Table 5

Percent Use of Forest Zones by Native Mariana Island Doves

Forest

zone

Species

Top

Middle

Lower

Ground

White-throated Ground Dove

45.0 (27)“

20.0 (12)“

5.0 (3)“

30.0 (18)“

Mariana Fruit Dove

76.6 (49)"

21.9 (14)“

1.6 (1)“

0

= N.

Direct counts showed that the Mariana Fruit Dove was more common
in limestone forest than in disturbed habitat, but uncommon compared to
that reported by Engbring et al. (1986) (Table 2). However, because I
encountered fruit doves at closer range than Engbring et al. (1986) (Table
1), density estimates for limestone forest were similar to those from 1982
(Table 3). Like the White-throated Ground Dove, the cyclic nature of
counts produced high variation in seasonal census results. Variation was
comparatively low within a season, although still higher than for passer-
ines (Table 4).

The Mariana Fruit Dove appeared to focus activities (N = 64) in upper
and mid-forest strata (Table 5). Otherwise, it occupied a range of habitats
similar to that of the White-throated Ground Dove. Also, like the preced-
ing species, it flew >100 m (often in pairs) above the canopy, suggesting
that it did not defend all-purpose territories. I saw individuals feeding on
fruits of the native Ficus spp. (9) and Premna obtusifolia (2) trees, the
vine Jasminum marianum (1), and the introduced Muntingia calabura (1).
Jenkins (1983) and Villagomez (1987) list additional fruits eaten by the
Mariana Fruit Dove. Like many members of the genus (Beehler et al.
1986), this species is a canopy frugivore.

Collared Kingfisher. — Both limestone forest and island-wide surveys
showed that three of four annual counts peaked in October (Figs. 2A,
3A). Seasonal variation in censuses was similar to that obtained for pas-
serines, but lower than that for doves. Variation within was greater than
that between seasons (Table 4), which illustrated the difficulty in census-
ing a species that regularly flew >300 m above the forest canopy. Den-
sities computed (Table 3) are likely exaggerated because of the liklihood
of overcount from flight calls made during these long flights. The Collared
Kingfisher was encountered with similar frequency on limestone forest
and disturbed habitat counts (Table 2), although because it was observed
at greater distances in disturbed habitat (Table 1 ), its computed density
was comparatively low. Engbring et al. (1986) found birds with similar
frequency to that of this study (Table 2) but with lower computed density

258

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

(Table 3) because detection distances were lower in the present study
(Table 1).

I detected breeding in January (recently fledged nest). May (nest), and
June (nestlings, incubating). The Collared Kingfisher was present in every
habitat, including shorelines, wetlands, savannah, disturbed sites, and
limestone forest. In forest, I usually observed it flying above the canopy,
at the forest edge, or perched in the top of canopy trees. However, it also
entered the forest interior, where I mist netted individuals twice. It often
occurred in pairs or groups of three to four birds, which probably were
family groups.

I observed 15 feeding attempts by the Collared Kingfisher. Prey items
included a millipede, grasshopper, unidentified insect, lizards (four con-
firmed, three apparent), Micronesian Honeyeater, Golden White-eye (at-
tempt), and Bridled White-eye (capture, attempt, apparent capture). Mar-
shall (1949) had previously listed insects, spiders, crabs, lizards, and mice
as prey, and he also described frequent but unsuccessful attacks on Bri-
dled White-eyes. Engbring et al. (1986) reported an instance of predation
on a Rufous Fantail. The Collared Kingfisher is the only extant native
predator on birds in the Marianas.

The importance of this species as a bird predator is reflected in obser-
vations of mobbing, scolding, and alarm calls directed against it by other
species, including the Bridled White-eye (6), Golden White-eye (1), Ru-
fous Fantail (1), and Micronesian Honeyeater (1). I also saw a Collared
Kingfisher chase a Yellow Bittern (1) and fight with a Black Drongo
{Dicrurus macrocercus) on Rota.

Island Swiftlet. — Because of its crepuscular nature, this species was
poorly censused by the techniques employed, and census data are not
reported. However, incidental dawn/dusk observations and data from the
island-wide census demonstrated that, unlike most species, it was found
unevenly about the island. Its distribution appeared correlated with the
occurrence of suitable nest caves. Hence, it was common in mountainous
areas around Takpochao, Capitol Hill, Navy Hill, As Teo, eastern Marpi,
Laderan Hakmang, and Sabanan Nanasu but largely absent from western
Marpi and flat lowlands throughout the island. I gathered no breeding or
behavioral data on the Island Swiftlet.

Rufous Fantail. — Annual counts consistently peaked in October (Figs.
2B, 3B). Seasonal variation in counts was lower than that for doves, but
variation within a season was still lower (Table 4), suggesting that pop-
ulations or breeding status changed seasonally. The species was found
with similar frequency in limestone forest and disturbed habitats (Table
2), although because birds were detected at greater distances in disturbed
habitat (Table I ), computed densities there (Table 3) were relatively low.

Craig • MICRONESIAN BIRD COMMUNITY

259

Engbring et al. (1986) recorded the Rufous Fantail more frequently than
did this study (Table 2), but their computed densities were far lower
(Table 3) than in limestone forest, because my detection distances in forest
were lower (Table 1).

Breeding was recorded for January (nest construction, eggs, fledglings,
juveniles), February (eggs, juveniles), March (nest), April (nestlings), Oc-
tober (nest construction, eggs), and November (nestlings). Jenkins (1983)
reported breeding in January— April, June, and November on Guam, and
Marshall (1949) believed, based on gonad condition of specimens, that
breeding occurred year round. Birds occurred commonly in a variety of
wooded and thicket habitats, including beach strand and suburban habi-
tats, but they were largely absent from swordgrass savannah.

Frequently observed food begging in small flocks of three to four birds
indicated that these were family groups. Color banding further showed
that groups remained at a single location. At such locations, males en-
gaged in song duels with neighbors and responded aggressively to taped
playback of songs. Hence, individuals appeared to defend all purpose
territories.

Observations of interspecific aggression were restricted to one instance
each of supplanting a Bridled White-eye at a perch and chasing a foraging
Golden White-eye from near a fantail nest. More frequently, I saw birds
following Golden (10) and Bridled white-eyes (10) to capture insects
flushed by the foraging activities of these species.

Nightingale Reed-warbler. — I regularly recorded birds only on island-
wide and disturbed site counts (Table 2). Birds detected on limestone
forest surveys were almost all calling from outside the forest. Island-wide
counts showed no clear seasonal trend (Fig. 2B). Previous studies at Mar-
pi demonstrated a drop in territorial activity in the wet season by up to
24% (Craig 1992a). Indeed, my only breeding record was for February
(fledgling). My inability to detect a similar island-wide trend by this loud-
ly vocal species likely meant that the census data were inherently variable,
although the local trend I found might not have been general.

In five years of observing on Saipan, I located Nightingale Reed-war-
blers in interior forest on only three occasions. These birds did not appear
to be territory holders, because on subsequent visits to the same site they
were absent. At the disturbed census sites (surveyed in March), I found
individuals with lower frequency (Table 2) and density (Table 3) than did
the more comprehensive Engbring et al. (1986) survey (made in May),
although the species was generally widespread and common on the island.
It occurred in all thicket-meadow mosaics, forest edge, reed marshes, and
forest openings, but it was absent from beach strand and swordgrass sa-
vannah.

260

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Because Nightingale Reed-warblers usually are concealed in thick veg-
etation, I recorded foraging rarely despite intensive study. Observations
included eating insects (3), gleaning invertebrate from leaves (3) and a
dead leaf (1), and probing a dead stub (1). Marshall (1949) reported in-
sects, spiders, snails, and lizards as prey. Although the species was in-
traspecifically aggressive and defended all-purpose territories (Craig
1992a), I saw no interaction between it and other species.

Micronesian Starling. — This species showed little clear seasonal trend
in censuses (Figs. 2B, 3B). Although seasonal variation in counts was
low compared to doves, within-season variation was even lower (Table
4), suggesting a seasonal shift in populations or breeding status. Com-
pared to limestone forest, it was encountered less frequently (Table 2)
and observed to greater distances (Table 1) in disturbed habitats. Hence,
computed densities (Table 3) were lower in disturbed habitats. It also was
found more commonly in limestone forest than by Engbring et al. (1986)
(Table 2).

Micronesian Starlings were usually seen in pairs, family groups (based
on observations of adults attending food begging juveniles, mist netted
juveniles with an aggressive adult nearby) or juvenile flocks (all birds in
juvenal plumage). Larger flocks (5-11, not 50 as reported by Marshall
1949), made up mostly of birds in juvenal plumage were likely the prod-
uct of several nestings by a single adult pair. As Jenkins (1983) reported
for Guam, I observed single pairs nesting at the same location nearly year
round. Banding showed that birds maintain a home range. The species
used virtually all habitats from beach strand to interior forest and sword-
grass savannah.

Jenkins (1983) listed a variety of fruits and seeds taken by the Micro-
nesian Starling, and Marshall (1949) described it as omnivorous. I ob-
served birds feeding on fruits of Ficus spp. (4), papaya (Carica papaya)
(3), camachile (Pithecellobium diilce) (1) and an insect (1). Reichel and
Glass (1990) reported that it preys on seabird eggs.

Micronesian Honeyeater. — No clear seasonal trend emerged in census
data at either environmental scale (Figs. 2A, 3A). In limestone forest,
between season variation was high compared to other passerines and to
within-season variation (Table 4), suggesting that seasonal shifts occurred
in populations or breeding status. Jenkins (1983) reported breeding on
Guam year round, although he was uncertain if breeding peaks occuired.
I recorded breeding in February (nest building) and May (courtship). It
was uncommon in limestone forest compared to disturbed habitats, as
well as to other passerines (Tables 2, 3). Engbring et al. (1986) found
that the frequency (Table 2) and computed density of the Micronesian

Craig • MICRONESIAN BIRD COMMUNITY

261

Honeyeater was higher than I found for birds in limestone forest (Table
3).

The Micronesian Honeyeater was aggressively territorial against con-
specifics, chased individuals and dispersed flocks of Golden White-eyes
(4), and chased Rufous Fantails (2). I also saw a Micronesian Starling
supplant a Micronesian Honeyeater at a perch.

At Capitol Hill, a color banded male had a territory of ca 0.7 ha. Two
additional banded territorial males were observed to within 150 m from
the banding site. However, repeated mist netting at one site yielded regular
capture of unhanded birds (mostly females or juveniles, based on plumage
and measurements) which indicated the existence of a population of non-
territorial birds. These floaters or nomadic birds may account for the
seasonal variation in census data, because they may opportunistically fol-
low ephemeral nectar sources as do certain of the Hawaiian Honeycreep-
ers (Scott et al. 1986).

The species occupied a variety of habitats, including beach strand, man-
groves, upland forest, suburban areas, and disturbed habitats. It was large-
ly absent from swordgrass savannah, but particularly common in the vi-
cinity of coconut {Cocos nucifera) groves where it fed on nectar. Foraging
is discussed in detail by Craig and Beal (ms), and Table 6 lists 1 1 nectar
sources that I recorded.

Bridled White-eye. — No clear pattern emerged in counts at either en-
vironmental scale, although January counts averaged lowest, probably
because higher winds at this season reduced the detectability of this can-
opy species (Figs. 2B, 3B). Like most passerines, variation in counts was
relatively low, and variation between seasons was greater than within a
season (Table 4). Although even more individuals were encountered in
disturbed habitats than in limestone forest (Table 2), birds detected were
at greater distances so that population densities (Table 3) were lower in
disturbed habitat. 1 recorded more Bridled White-eyes (Table 2), and den-
sities computed were far greater than reported by Engbring et al. (1986)
(Table 3), because I detected birds at closer range (Table 1).

Banded birds declined in frequency of occurrence, p, from the banding
site to 300 m in an empirically fitted quadratic relationship (r- = 0.99):

p = 1.47x2 - 1.21x + 53.82,

where x has values from one for the basal zone (0-50 m from the banding
site) to six for the outermost zone (251-300 m). Based on this relation-
ship, I solved Equation (3) for my Lincoln-Peterson (8754 bird/km^) and
Jolly-Seber (7770 birds/km^) population estimates, which yielded densi-
ties of the same order of magnitude as those obtained through VCP cen-
susing (Table 3).

262

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 6

Plant Species Led upon by Three Small Forest Passerines

Plant species

Bridled

White-eye

Golden

White-eye

Micronesian

Honeyeater

se“ fr

fl

ne

se fr

fl

ne

se fr fl

ne

Vines:

Momordica charantia

X X

X X

Mikania scandens

7

X

Passiflora foetida

X

Opercidina ventricosa

X

Jasminum marianum

X

X

Trees:

Pisonia grandis

X

X

X

Cynometrci ramiflora

X

X

X

Premna obtusifolia

X

7

X

X

Ficus spp.

X

X

Melanolepsis multiglandulosa

X

7

X

Erythrina variegata

X

X

X

X

Psychotria mariana

X

X

X

X

Morinda citrifolia

X

X

Artocarpus spp.

X

X

Aidia cochinchinensis

X

X

X

Pipturus argenteus

X

Bikkia mariannensis

X

Hibiscus tiliaceus

X

X

Delonix regia

X

Lantana camara

X

X

X

X

Albizia lebbeck

X

X

Carica papaya

X

X

Leucaena leucocephala

X

Cocos nucifera

X

Muntingia calabura

X

X

Herbs:

Biclens pilosa x

■’ se = seed, fr = fruit, fl = flower, ne = nectar.

The Bridled White-eye was found in all habitats from beach strand to
disturbed habitats, suburban areas, and forest. It was less common in
swordgrass savannah. I recorded breeding in January (carrying nesting
material), February (nestlings, carrying nesting material), August (eggs,
carrying nesting material), and October (carrying food). Moreover, I ob-
served food begging by juveniles year-round. Jenkins (1983) also reported
that the Guam Bridled White-eye bred year-round. Although it is not
territorial, banding demonstrated that birds remain in a home range, and

Craig • MICRONESIAN BIRD COMMUNITY

263

individuals could be attracted to playback of various flocking calls. Other
than scolding Collared Kingfishers, interspecific social interactions in-
volved only an observation of a Bridled White-eye following a foraging
Rufous Fantail. No interspecific aggression initiated by Bridled White-
eyes was noted. I recorded feeding on seeds, nectar, flowers, and fruit of
22 plant species (Table 6) in addition to invertebrates.

Golden White-eye. — No clear pattern emerged in counts at either en-
vironmental scale (Figs. 2B, 3B). Although as with other passerines, sea-
sonal variation in censuses was relatively low, intraseasonal variation was
still lower (Table 4). The Golden White-eye was decidedly more common
in limestone forest than in disturbed habitats (Table 2, 3), although slight-
ly less frequently encountered than by Engbring et al. (1986). Neverthe-
less, computed densities were greater in this study (Table 3) because I
encountered birds at closer range (Table 1).

I recorded breeding in January (gathering nesting material, eggs, fledg-
lings), February (eggs), March (eggs). May (recently fledged nest), June
(eggs), July (copulation, carrying nesting material, eggs, nestlings), Au-
gust (nest construction), and October (eggs). Moreover, I heard song and
observed food begging year-round, except during the protracted dry sea-
son of 1993, when I heard no singing during June despite my almost
daily presence in the field. This latter observation may help to explain
Marshall’s (1949) failure to detect any song in this species. Other than
limited observations reported by Stinson and Stinson (1994), little other
data on breeding exist.

Aggression involved chases and dispersing flocks of Bridled White-
eyes (20) and Rufous Fantails (2). Golden White-eyes were territorial.
Banded males defended territory boundaries against other males and re-
sponded, although not vigorously, to playback of recorded songs. Family
groups of 3-4 (as demonstrated by food begging of Juvenal plumaged
birds) were typical. The Golden White-eye occurred in all wooded hab-
itats, including strand forest and suburban areas, although it was generally
absent from swordgrass savannah. Foraging is discussed by Craig and
Beal (ms); I observed feeding upon invertebrates and the nectar, flowers,
and fruit of 13 plant species (Table 6).

DISCUSSION

For seven of the nine forest bird species examined, census variation in
limestone forest between seasons was greater than that within a season.
The remaining two species possessed behavioral characteristics which
made them particularly difficult to census. Hence, most or all forest spe-
cies likely undergo seasonal shifts in either populations or conspicuous-
ness (i.e., increased vocalizations related to breeding). In the case of

264

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

doves. Strong shifts in vocalizing related to breeding activity are indicated.
Many tropical forest passerines increase breeding activity during the wet
season (Beehler et al. 1986), but only two species showed consistent peak
counts at this time. These, the Collared Kingfisher and Rufous Fantail,
are also the only Mariana Island forest species with widespread Indo-
Pacific distributions. Other species showed divergence in seasonal counts
between years, suggesting that no regular pattern in counts existed. Com-
bined with evidence for year-round breeding for such species as the Mi-
cronesian Starling, Micronesian Honeyeater, Bridled White-eye, and
Golden White-eye, seasonal variation in counts may, therefore, be caused
by actual population shifts or differing peak breeding times related to
resource availability. Storms, the vagaries of seasonal patterns, and atten-
dant alteration in food supplies may drive such population or breeding
shifts.

For many forest species, peak counts (7 of 9) and computed densities
(8 of 9) were greater in native forest than in disturbed habitat. Therefore,
native limestone forest, with its comparatively high diversity of tree spe-
cies, its cooler, wetter microclimate, and variety of food sources, is likely
to be superior habitat for most forest species. Only the Micronesian Hon-
eyeater was noticeably more common outside native forest. Presumably,
nectar is insufficiently abundant or consistently available to support high
densities of this species in limestone forest. However, on nearby Aguig-
uan, which has forests similar to those on Saipan (Chandran et al. 1992),
the Micronesian Honeyeater was common in native forest (Craig et al.
1992). Extensive stands of the introduced Lantana camara, are found
adjacent to forest on Aguiguan but not Saipan. This shrub flowers year-
round and is frequently visited by Micronesian Honeyeaters (Craig et al.

1992).

Censusing at two environmental scales uncovered few clear differences
in seasonal trends. The principal difference uncovered was that the Mi-
cronesian Honeyeater was relatively uncommon in limestone forest but
regular on island-wide counts. Otherwise, forest species showed numer-
ous similarities in counts at both scales, thereby suggesting that a wide

range of habitat was suitable for most.

Results of the Engbring et al. (1986) population survey of Saipan biids
showed direct counts of roughly the same order of magnitude as those
reported in this study. Major differences in counts probably result from
this study’s survey of primarily native forest, as opposed to all habitats
in the 1982 survey. However, the much higher numbers ot Mariana Fruit
Doves and Rufous Fantails found by Engbring et al. (1986) are not easily
accounted for and may indicate population declines in these species. Re-

Craig • MICRONESIAN BIRD COMMUNITY

265

cent surveys on Aguiguan also detected declines in counts of Mariana
Fruit Doves compared to 1982 (Craig et al. 1992).

Although direct counts exhibited similarities, densities reported here
are generally well above those computed by Engbring et al. (1986). Most
of this difference may be attributed to the shorter distances at which I
detected species. Recomputation of densities for my counts using the
Engbring et al. (1986) distance estimates indeed yielded similar densities
to those they obtained. Shorter detection distances were a consequence
of my surveying only in dense, interior forest, whereas Engbring et al.
(1986) censused in all habitats. Despite the large difference in results, I
believe my density estimates are realistic. Independent density computa-
tions for the Bridled White-eye were of the same order of magnitude as
those obtained through censusing. Moreover, Craig et al. (1992) pointed
out that densities determined for such small passerines as the Golden
White-eye translated to encountering one family group of four directly
on the transect line once roughly every 100 m. Such a frequency is con-
sistent with field experience for these species.

The densities reported here, particularly for the Rufous Eantail, Bridled
White-eye, and Golden White-eye are among the highest ever reported
for birds and are similar to those obtained for the most abundant of the
Hawaiian Honeycreepers (Drepanidinae) (Scott et al. 1986). Indeed, per-
sonal observations of the Common Amakihi (Hemignathus virens) and
Apapane (Himatione sanguinea) in the heart of their present range on
Hawaii indicated that densities of Marianas small passerines were similar
to those of these Hawaiian species. In temperate forests, in contrast, the
density of the Ovenbird (Seiurus aurocapillus), the most abundant breed-
ing species in two typical northeastern forest tracts, averaged 149.5-1 16.8
birds/km^ (Magee 1989— 1993a, 1989— 1993b). These densities are about
1/15 that of the similarly territorial Golden White-eye and 1/44 that of
the flocking Bridled White-eye. That such immense densities occur sug-
gests that at least some forest birds in the Marianas exist at the maximum
densities allowed by resources available in the habitat.

I cannot definitively attribute census variation to population fluctua-
tions, because differences in breeding activity can also produce census
variation. Eurther study is required to demonstrate that populations are at
carrying capacity. However, existing at saturation densities is a charac-
teristic of avian communities that is predicted to elicit interspecific com-
petition, particularly between ecologically similar species (MacArthur
1972, Wiens 1977). Interspecific competition is most obviously mani-
fested through aggression, and indeed the most ecologically similar of the
small passerines, the two species of white-eyes (Craig 1989, Craig and
Beal, unpubl. data) are those for which I observed the most aggression.

266 THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

The larger Golden White-eye was clearly socially dominant over the Bri-
dled White-eye, and much of the aggression observed involved chases of
foraging Bridled White-eyes. Such behavior suggested that the contested
resource was food. Based on observations on the four small passerines,
the order of social dominance appeared to be Micronesian Honeyeater,
Golden White-eye, Rufous Fantail, and Bridled White-eye. In contrast
with temperate systems in which bird species can overlap widely in eco-
logical space with little aggression because populations are rarely at sat-
uration densities (Wiens 1977, Craig 1987), this study found aggression
prevalent between species that were only generally similar in their ecol-
ogy (Craig 1989, Craig and Beal, unpubl. data). The existence of popu-
lations at the carrying capacity of the habitat most likely accounts foi this
difference.

ACKNOWLEDGMENTS

My research benefitted from discussions on censusing procedures with J. Engbring. It
was funded in part by Pittman-Robertson Eederal aid to wildlife.

LITERATURE CITED

Baker, R. H. 1951. The avifauna of Micronesia, its origin, evolution, and distribution.
Univ. Kansas Publ. Mus. Nat. Hist. 3:1—359.

Beehler, B. M., T. K. Pratt, and D. A. Zimmerman. 1986. Birds of New Guinea. Princeton
Univ. Press, Princeton, New Jersey.

Chandran, R., R. J. Craig, Z. Keys, C. Sheu, and J. Dubrall. 1992. Sti-ucture and com-
position of native forest on Aguiguan. Proc. Marianas Res. Sympos. 1.51—56.

Craig, R. J. 1987. Divergent prey selection in two species of waterthrushes (Seiurus). Auk

104:180-187.

. 1989. Observations on the foraging ecology and social behavior ol the Bridled

White-eye. Condor 91:187—192.

, 1990. Foraging behavior and microhabitat use of two species of white-eyes (Zos-

teropidae) on Saipan, Micronesia. Auk 107:500—505.

1992a. Territoriality, habitat use, and ecological distinctness ot an endangered

Pacific island reed-warbler. J. Field Ornithol. 63:436-444.

1992b. Ecological characteristics of a native limestone fore.st on Saipan, Mariana

Islands. Micronesica 25:85-97.

, R. Chandran, and A. Ellis. 1992. Bird populations on Aguiguan: a ten year

update. Proc. Marianas Res. Sympos. 1:8-15.

Engbring, J., E L. Ramsey, and V. J. Wildman. 1986. Micronesian forest bird survey,
1982: Saipan Tinian, Aguijan, and Rota. U.S. Fish Wildl. Serv. Report, Honolulu,
Hawaii.

Fosberg, E R. 1960. The vegetation of Micronesia. Amer. Mus. Nat. Hist. Bull. 199:1-76.
Frith, H. J., L. W. Braithwaite, and T. O. Wolfe. 1974. Sexual cycles of pigeons in a
tropical environment. Ausl. Wildl. Res. 1:117-128.

Glass, P. O. and D. T. Aldan. 1987. Micronesian Megapode surveys and research. Pp.

131 — 153 in Commonwealth No. Mariana Is., Div. Fish Wildl. Prog. Rep. 1982—1987.
Jenkins, J. M. 1983. The native forest birds ol Guam. Orn. Monogr. 31.

Craig • MICRONESIAN BIRD COMMUNITY

267

AND C. F. Aguon. 1981. Status of candidate endangered species on Saipan, Tinian,
and Rota of the Mariana Islands. Micronesica 17:184-186.

MacArthlir, R. H. 1972. Geographical ecology. Harper and Row, New York.

M.agee, a. 1989-1993a. Climax hemlock-white pine forest with transition hardwoods. J.
Field Orn. 60:38, 61:49, 62:57-58, 63:73-74, 64:72-73.

. 1989-19936. Second growth hardwood forest. J. Field. Orn 60 26-27 61-30 62-
35, 63:37, 64:37-38.

Marshall, J. T. Jr. 1949. The endemic avifauna of Saipan, Tinian, Guam, and Palau.
Condor 51:200-221.

Pratt, D. H., P. L. Bruner, and D. G. Berrett. 1987. A field guide to the birds of Hawaii
and the tropical Pacific. Princeton University Press, Princeton, New Jersey.

Reichel, j. D. and P. O. Glass. 1990. Micronesian Starling predation on seabird eggs
Fmu 90:135-136.

AND P. O. Glass. 1991. Checklist of the birds of the Mariana Islands. Flepaio 51-
3-11.

Reynolds, R. T, J. M. Scott, and R. A. Nussbaum. 1980. A variable circular-plot method
for estimating bird numbers. Condor 82:309-313.

Savidge, j. a. 1987. Extinction of an island forest avifauna by an introduced snake. Ecol-
ogy 68:660-668.

Scott, J. M., S. Mountainspring, F. L. Ramsey, and C. B. Kepler. 1986. Forest bird
communities of the Hawaiian Islands: their dynamics, ecology, and conservation. Stud-
ies Avian Biol. 9.

, E. Garton, and j. Jeffrey. 1988. Guideline for field surveys of birds. U.S. Fish
Wildl. Serv. Report, Honolulu, Hawaii.

Steadman, D. W. 1992. Extinct and extirpated birds of Rota, Mariana Islands. Micronesica
25:71-84.

Stinson, D. W. 1993a. Micronesian megapode research. Pp. 217-233 in Commonwealth
No. Mariana Is., Div. Fish Wildl. Prog. Report 1987-1992.

., 1993b. Dove research. Pp. 81-84 in Commonwealth No. Mariana Is., Div. Fish
Wildl. Prog. Report 1987-1992.

Stinson, C. W. and D. W. Stinson. 1994. Nest size, clutch size, and incubation behavior
in the Golden White-eye. J. Field Orn. 65:65-69.

Tanner, J. T. 1978. Guide to the study of animal populations. Univ. of Tennessee Press,
Knoxville, Tennessee.

Villagomez, E. C. 1987. Game bird (dove) survey and research. Pp. 37-46 in Common-
wealth No. Mariana Is., Div. Fish Wildl. Prog. Ann. Report.

Wiens, J. A. 1977. On competition and variable environments. Am. Sci. 65:590-597.

Young, F J. 1989. Soil survey of the islands of Aguijan, Rota, Saipan, and Tinian, Com-
monwealth of the Northern Mariana Islands. U.S. Soil Cons. Serv.

Wilson Bull., 108(2), 1996, pp. 26S-219

NATURAL HISTORY AND CONSERVATION STATUS
OF THE TAMARUGO CONEBILL IN NORTHERN CHILE

Cristian F. Estades'

Abstract. I studied Tamarugo Conebill {Conirostrum tamarugense) populations at the

Pampa del Tamarugal National Reserve in northern Chile between October 1993 and July
1994. The estimated population of conebills was 35,107 individuals in 10,787 ha of tama-
rugo forests. A strong relationship was found between forest foliage volume per hectaie and
breeding conebill density. The Tamarugo Conebill breeds in the tamarugal September-De-
cember and then probably migrates to the highlands of southern Peru. Although the species’
breeding locality presently is protected, it faces some important threats including the pump-
ing of the subterranean aquifers upon which the tamarugal vegetation depends, attempts to
control the butterfly upon whose larvae the species forages, and human disturbance on the
wintering grounds. Received 15 May 1995, accepted 30 Nov. 1995.

The western slope of the Andes, comprising the arid regions of northern
Chile and southern Peru, is one of 57 areas of high avian endemism in
South America (Bibby et al. 1992b). Among the birds of this region, the
Tamarugo Conebill {Conirostrum tamarugense) is one of the rarest spe-
cies, and has been known to science only since 1969 (Mayr and Vuilleu-
mier 1983). The species was formally described by Johnson and Millie
(1972) from six specimens collected at the Pampa del Tamarugal, northern
Chile. Afterward, the few documented records of the species (see Mc-
Farlane and Loo 1974, McFarlane 1975, Tallman et al. 1978, and Schu-
lenberg 1987) have provided little information about its natural history
and general status. All records are of solitary individuals or small gioups,
generally in mixed flocks with the Cinereous Conebill (C. cinereum)
(McFarlane 1975). Thus far, there have been no estimates of its population
size, habitat requirements, or seasonal movements. Recently Estades and
Lopez-Calleja (in press) have reported the nesting of the species at the
Pampa del Tamarugal. The conservation status of the Tamarugo Conebill
is uncertain, and the Chilean Eorest Service (CONAE) considers it “in-
sufficiently known” (Glade 1988). More recently Rottmann and Lopez-
Calleja (1992) considered the Tamarugo Conebill as “vulnerable,” while
Collar et al. (1992) also categorized the species as “insufficiently known”
in the list of threatened birds of the Americas.

The present paper reports on an assessment of the conservation status
of the Tamarugo Conebill (C. tamarugense) in northern Chile. The ob-
jectives of the study were to obtain estimates of the species’ population

' Depl. dc Mancjo dc Recur.so.s Foreslale.s, Univ. de Chile, Ca.silla 9206, Santiago, Chile. (Present address.
Dept, of Wildlife Ecology, Univ. of Wisconsin, I6.J9 Linden Dr.. Madison, Wisconsin 53706.)

268

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

269

70

Fig. 1. Geographical location of the study area.

size, describe its habitat requirements, assess its seasonal movements, and
determine principal threats to its survival.

STUDY AREA AND METHODS

I studied conebills at the Pampa del Tamarugal National Reserve (20°24'S, 69°44'W) in
the Tarapaca Region, northern Chile (Fig. I). This area, at an elevation of 1000 m, has an
extreme dry climate with a mean annual rainfall of 0.3 mm (di Castri and Hajek 1976).
Annual temperatures range between an absolute minimum of -12°C and a maximum of
35°C. Relative humidity undergoes wide daily fluctuations (3-8% during the day to 80-
100% during the night in October; Sudzuki 1985). The Pampa del Tamarugal is a vast salt
flat over subterranean aquifers. It is dominated by savannas of tamarugo (Prosopis tama-
rugo), a highly drought-resistant tree. Shrub and herbaceous layers are almost nonexistent,
with only scattered individuals of Atriplex utacamensis. Tessaria ahsinthioides, Caescdpinia
aphyllci, and the extreme halophyte DistichUs spicata (Gajardo 1994). During the last four
centuries, the Pampa del Tamarugal was intensively exploited for the production of lirewood
and charcoal, mainly to support the mining industry. This cutting drastically reduced the
extent ot the tamarugo forest, leaving few remnants of forest, mainly low density stands of
little economic interest. In the 1930s, a reforestation program was started, which by the
1970s had generated 14,600 ha of tamarugo plantations (Aguime and Wrann 1985). In
addition, approximately 1900 ha of algarrobo (Prosopis cdha) plantations were established
at the Reserve. In 1983 the administration of the plantations passed to the Chilean Forest
Service which created the 109,842-ha Pampa del Tamarugal National Reserve (Fig. 1). Since
Its creation, the reserve has been used for the production of forage for sheep which feed on

270

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

tamarugo pods, for the production of firewood and charcoal, and, occasionally, for the
production of lumber.

I visited the Pampa del Tamarugal three times between October 1993 and July 1994. The
hrst visit (25 Oct-2 Nov) was during the austral spring, the second (14-21 March) during
late summer and the third (19-26 July) during mid-winter. The reserve is divided into four
major areas: Pintados, Bellavista, La Tirana, and Zapiga (Fig. 1). Due to time constraints,
only the first three plots were assessed. These areas were selected for the following reasons:
Pintados is the largest stand and holds the oldest tamarugo plantations, Bellavista is at the
southern limit of the Reserve and might correspond to the southern limit of the Tamarugo
ConebilFs range, and the La Tirana plot holds the last natural tamarugo stands.

In order to estimate population densities 18 line transects were established at Pintados
and six at Bellavista in the spring of 1993. Because no birds were found at La Tirana during
the surveys, no transects were established and density was considered to be zero. Three of
the transects at Pintados were located in a 60-ha managed stand. The trees of the latter stand
had been pruned in 1991 for the production of timber and charcoal. As a result of this
management, these trees lacked almost the entire lower half of their foliage. Additionally,
four transects were surveyed at an algarrobo plantation at Pintados.

The transects used were of the “fixed belt” type (Bibby et al. 1992a). In each transect,
all the birds seen or heard between two rows of trees were counted. Thus the width of the
belt was two times the distance between tree rows. Because this distance was different
between years of planting (10, 13.5 or 15 m) the length of the transects varied from 300 to
200 m, in order to equalize the surveyed areas. Preliminary observations carried out in the
afternoon (i.e., after 13:00 h) consistently gave lower densities than observations m the
morning. Therefore, all surveys were made between 7:00 and 12:00 h. Transects were sur-
veyed 3.1 times, on average, during the first visit.

During March and July 1994, four of the transects at Pintados were surveyed a total of
eight times per visit in order to compare densities between seasons. The other transects were
not formally assessed.

Habitat along transects was characterized in terms of foliage volume per ha. The esti-
mation of each tree's volume (m^) was based on the formula volume — (0.5 diameter) X
pi X % height. The estimated dimensions of the trees were total height, crown diameter,
and presence or absence of the lower third of the foliage. To convert the results to volume
per ha, the number of trees in the transect was recorded.

To describe the patterns of microhabitat use by foraging conebills, I divided trees visually
into six sections. The parts of the tree that contained the leaves and flowers were separated
from those that consisted of leafless branches, and the tree was divided into three horizontal
layers. All observed birds were assigned to the section in which they were first observed.
Individuals were captured in mist nets and banded to study the species’ migratory move-
ments. Finally, non-systematic observations were made to expand knowledge of the species
general natural history, both in the reserve and in adjacent localities. Mean densities were
compared among seasons using one-way ANOVA. The el feet of tamarugo pruning on biid
density was assessed using a /-test.

RESULTS

Population size. — Densities derived from transects were calculated se-
paratedly for the different types of tamarugo forests surveyed during
spring 1993. During this period, many conebills were observed through-
out the tamarugo plantations but at varying local densities (Table 1). No
individuals were found at La Tirana. The total estimated conebill popu-

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

271

Table 1

Densities (ind/ha) and Population Sizes of Tamarugo Conebill in Pampa del
Tamarugal National Reserve, Northern Chile, during October 1993

Stand

Stand

age

Stand
area (ha)

Conebill density
mean (SD)

Total population

Pintados

46

964

9.27 (2.85)

8940

Pintados

27

2881

4.84 (3.23)

13,947

Pintados

21

4833

2.38 (3.39)

1 1,517

Bellavista

21

2109

0.33 (1.49)

703

All stands

—

10,787

3.25 (0.68“)

35,107'’

“Weighted mean's standard deviation (Cochran 1980)

" 95% confidence interval: 18,970-51,244; 11.3 degrees of freedom.

lation was 35,107 individuals. At Pintados, significant differences were
found between conebill densities in different seasons (Fig. 2). Densities
declined (F = 74.37, df = 2, F = 0.000) over time from a high of 8.3
ind/ha m October 1993 (breeding season) to almost no birds at all in July
1994 (nonbreeding season).

Fig, 2. Mean conebill densities (±SD) on four transects in a 46-year-old forest stand at
Lintados for three periods. The empty circle with the inteiTogative symbol repre.sents the
possibility that the population had increa.sed in November-December due to the addition of
rst-year birds following the breeding .sea.son. The asterisk indicates individuals were ob-
served in the area, but none was recorded on the transects.

Conebill Density [ind/ha]

272

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Foliage Volume (Thousands of mVha)

Lig. 3. Relationship between tamarugo foliage volume per ha and conebill density dur-
ing October (breeding season) 1993 at Pintados. Points represent the mean conebill density
on each transect.

Habitat use. — A significant relationship was observed between forest
maturity and density of breeding conebills (October 1993). At Pintados,
a high percentage of the variance (r- = 0.744; N = 18; P < 0.0001) of
conebill density is explained by the total foliage volume (Fig. 3). Surveys
of algarrobo plantations show that the Tamarugo Conebill does not use
this forest type; on the four transects, no individuals were seen (October
1993). During the breeding season, the managed stand at Pintados held a
signihcatively lower density of conebills than the unmanaged stands {t-
test, P = 0.0000).

The species was found wintering (July 1994) at the localities of Pica
and Mamina (see Fig. 1 ). At the first site, a few individuals were observed
foraging with C. cinereutn in Citrus plantations. The same situation was
found at Mamina, where conebills were foraging in mixed flocks in ri-
parian scrubs of Baccharis petiolata and Tessaria ahsinthioides.

Microhahitat use. — During the breeding season, the Tamarugo Conebill
showed a marked tendency to use the outer and upper parts of tamarugo

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

273

Fig. 4. Percent of sightings of conebills in each tamarugo microhabitat zone

trees (Fig. 4). The species was observed foraging mainly on Lepidoptera
larvae, specifically on the species Leptotes trigemmatus Butler (Lycaen-
idae) which feeds on tamarugo leaves, buds, and flowers (Cogollor et al.
1982). Conebills preferentially selected the descending terminal branches
of the trees for nesting (Estades and Lopez-Calleja, in press). In the oldest
forests, breeding pairs held territories that included only a few trees
around the nest tree. During the nonbreeding season, the species did not
show a clear pattern of microhabitat use at Pampa del Tamarugal. At the
other sites surveyed, the species foraged mainly inside the scrub.

Because conebills principally used the upper half of tamarugo trees for
foraging, I thought that the removal of the lower half of the crown (as a
result of the pruning of the trees in the managed area) might have little
impact on the birds, but their density was lower in the managed stand.
The hypothesis that pruning created microclimatic changes that may have
affected the habitat suitability of the trees was tested. In July 1994 air
temperatures were measured at ground level and at 1 .5 m and at 4.5 m
at the center of 15 pruned and 15 non-pruned trees and then averaged.
Data were collected at 7:00 and 12:00 and repeated over two days. No
significant differences were found between pruned and non-pruned trees
at 7:00 h (r-test, P = 0.635). At 12:00 h, mean temperature was signifi-
cantly lower at the pruned plot (/-test, P = 0.009), probably due to the
convective cooling of wind.

Breeding. The nesting of C. tamarugense was first recorded at Pin-
tados m October 1993, austral spring (Estades and Lopez-Calleja, in

274

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

press). These authors found several active nests and presented evidence
that the species has been breeding at the Pampa del Tamarugal for many
years.

Breeding success was not evaluated. However, during March 1994 (late
summer), conebills at the reserve were observed flying in small flocks
composed of approximately 60% adults (individuals with the rufous su-
perciliary line, throat, upper breast, and undertail coverts) and 40% im-
matures. The only juvenile conebill captured was slightly smaller than
the adults and showed a dirty grayish color instead of the characteristic
rufous pattern.

During July 1994 (winter), very few individuals were observed at the
reserve (0-6/day). It was not possible to determine their age since all
were seen from a distance. However, many of the conebills observed at
the Mamina oasis lacked the rufous adult color.

DISCUSSION

Habitat and distribution. — Even though C. tamanigense is not restrict-
ed to tamarugo forests (McFarlane 1975, Tallman et al. 1978, this report
and several personal communications), the species depends on this type
of vegetation for breeding. Moreover, given the high concentration of
individuals and reproductive activity (Estades and Lopez-Calleja, in press)
and the absence of significant tamarugo forests outside the reserve, it
seems possible that most, if not all, of the species’ population breeds in
this area. Maturity of the forest seems to be an important factor in deter-
mining the habitat suitability for conebills. The strong relationship be-
tween foliage volume per hectare and the density of breeding conebills
parallels patterns described by Mills et al. (1991). Considering that the
number of flowers and the density of larvae in these flowers at Pintados
are correlated positively with tamarugo foliage volume (Lopez-Calleja
and Estades, unpubl. data), the relationship between conebill density and
foliage volume (Fig. 3) suggests that, above a certain foliage volume,
conebill density could be limited by territoriality instead of availability
of food. Below approximately 2000 m^ of foliage per ha, the tamarugo
forest is not a suitable habitat for the Tamarugo Conebill (Fig. 3). This
hypothesis could explain the absence of the species at La Tirana. Even
though foliage density was not formally assessed at this site, it was clearly
below 2000 m\ as trees are separated by 30 m or more in this area.

Low density of conebills at Bellavista could be explained by the scar-
city of Leptotes larvae feeding on tamarugo flowers there (Lopez-Calleja
and Estades, unpubl. data). This situation could be due to the high degree
of isolation of the plot which may reduce immigration of these butterflies.

The high proportion of time spent by the species at the top of trees

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

275

(see Fig. 4), might be mainly due to the concentration of flowers and
Leptotes larvae there, rather than a territorial behavior. In fact, breeding
conebills, instead of singing from high branches, call constantly while
looking for food, as stated by Fjeldsa and Krabbe (1990).

Johnson and Millie (1972) and Tallman et al. (1978) suggested that the
Tamarugo Conebill could be a high-altitude species that breeds during the
summer (highlands breeding season), then in winter moves into lowlands
of Tarapaca until the spring (late lowlands breeding season). Our present
study does not support this hypothesis, indicating instead that breeding
occurs at Pampa del Tamarugal between September and December during
the tamarugo flowering and the outbreak of Leptotes trigenimatus larvae.
Individuals may then move upwards (to 2500-3500 m) and to the north,
following a chain of small, mid-elevation, vegetated valleys. Possibly a
small group could also migrate to the north from or through Zapiga to
the lowlands of Arica, where the species has been observed (McFarlane
1975) (Fig. 5). Given the results of this study, isolated individuals or small
flocks of the species seen in lowlands near Arica during the breeding
season (Sept. -Nov.; McFarlane 1975, Sallaberry pers. comm.) are prob-
ably wandering, nonbreeding birds. This uncommon pattern of migration
would enable the species to use an important food resource almost ex-
clusively during its breeding period (no other bird species was observed
foraging on Leptotes larvae at the Pampa del Tamarugal) and thereafter
to share seasonally rich food resources with the species of the highlands
of northern Chile and southern Peru during their breeding season (Jan.-
March).

It IS possible that some conebills could remain near the Pampa del
Tamarugal during the entire year, since in May-June the tamarugos have
a second flowering period (“devareo”). However, the extent and intensity
of this phenomenon seems to be insufficient to support a large conebill
population.

Conservation status. The relationship observed between the vegeta-
tion and Tamarugo Conebill density suggests that, in the year of the spe-
cies’ discovery (1969) the oldest of the reforested forests (which then had
a mean age of 23 years) were just beginning to provide a suitable habitat
for the species, allowing its populations to reach a level that permitted
detection by ornithologists. This hypothesis implies that the present pop-
ulation of the species is derived from a small number of individuals that
survived a major population decrease during the period of deforestation
at Pampa del Tamarugal.

My estimates for the total population of C. tamarugense at the study
area are larger than any of the general estimates for minimum viable
populations (see Shaffer 1987). In addition, assuming that the relationship

276

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

70

Lig. 5. Known localities (symbols) and hypothetical migratory movements (arrows) of
C. tamcirugense.

between vegetation and conebill density will not change during future
years, the total population of the species should increase with the forest
volume as younger plantations mature. However, even if Tamarugo Cone-
bill populations are increasing and viable, there are at least three major
threats that could reduce the species’ short-term survival probability. (1)
Presently the tamarugo is still managed mainly for production of sheep
forage (pods) and for production of timber and charcoal, and pruning of
trees reduces habitat suitability for conebills. However, management tech-
niques could be modified in order to reduce the negative impact on the

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

277

birds. The threat of forage production is quite different, because manage-
ment attempts to control populations of Leptotes trigemmatus, which feed
on flowers and drastically reduce the yield of pods. Initially it was in-
tended to control the species by using chemicals (Cogollor et al. 1982).
At present, the Chilean Forest Service is studying the use of parasitoid
microhymenoptera as an alternative way of control. The success of such
a control program could, without doubt, be extremely harmful for conebill
populations. (2) Since the mean annual rainfall in the area is 0.3 mm, the
only water available for the tamarugos are the subterranean aquifers that
are found beneath the Pampa del Tamarugal. At present, these waters are
pumped at a rate of 700 Its/sec (CONAF rangers, pers. comm.) in order
to supply the requirements of the city of Iquique (see Fig. 1). The eco-
logical consequences of this water removal are not known. It is not clear
whether these aquifers are recharged by rainfall in the mountains, or if
they are fossil waters. In the last few years, there has been an increas-
ing number of dead trees observed at several sites in the reserve. How-
ever, this phenomenon has been related to an excessive salt accumulation
within the xylem vessels rather than to water deficit (Donoso et al. 1989).
(3) Almost all localities where C. tamarugense has been observed, out of
the Pampa del Tamarugal National Reserve, seem to lack of any type of
legal protection. During the non-breeding season, the Tamarugo Conebill
has been reported foraging in sites dominated by tree species such as
Polylepis, shrubs such as Gynoxys in Peru (Tallman et al. 1978), and scrub
{Tessaria, Baccharis), trees {Schmus) and cultivated species {Citrus) in
Chile (McFarlane and Loo 1974, McFarlane 1975, this study). Even
though there are no direct threats to conebill populations at the species'
wintering localities, all these areas are characterized by extreme aridity,
and there is increasing human disturbance (e.g., Polylepis cutting). The
security of the species’ wintering grounds is, therefore, uncertain (Collar
et al. 1992).

I suggest the following: (1) To give the species the lUCN conservation
status of “Vulnerable,” as suggested by Rottmann and Lopez-Calleja
(1992). Further research and population monitoring should continue to
clarify its status. (2) To stop or limit the control of Leptotes trigemmatus.
An economic assessment would be necessary in order to estimate the cost
of not controlling this insect. (3) To study the short- and long-term effects
of tamarugo management strategies on conebill populations in order to
design management techniques that best benefit all tamarugo users. (4)
To study the effects of water pumping on the survival of Tamarugo Cone-
bills.

278

THE WILSON BULLETIN • Vol. 108, No. 2, June J996

ACKNOWLEDGMENTS

I thank M. V. Lopez-Calleja, H. J. Hernandez, and P. B. Aguirre for their valuable field
assistance. The Chilean Eorest Service (ConaO gave permission to work at the Pampa del
Tamarugal National Reserve and provided some assistance. G. H. Rosenberg kindly shared
with me his field observations. I thank J. Jimenez for his review of an earlier draft of the
manuscript. I also thank J. M. Bates, J. P. O’Neill, C. R. Blem, and S. A. Temple for their
helpful comments on this paper. This study was supported by the Pan American section of
BirdLife International and the U.S. Pish and Wildlife Service.

LITERATURE CITED

Aguirre, J. J. and J. Wrann. 1985. Especies del genero Prosopis y su manejo en la Pampa
del Tamarugal. Pp. 3-33 in Estado actual del conocimiento sobre Prosopis tamarugo.
(M. A. Habit, ed.). United Nations, Pood and Agriculture Organization, Santiago, Chile.
Bibby, C. j., N. D. Burgess, and D. A. Hill. 1992a. Bird census techniques. British Trust
for Ornithology/Royal Society for the Protection of Birds. Academic Press, London,
England.

, N. J. Collar, M. J. Crosby, M. E Heath, Ch. Imboden, T. H. Johnson, A. J.

Long, A. J. Stattersfield, and S. J. Thirgood. 1992b. Putting biodiversity on the
map: priority areas for global conservation. International Council for Bird Preservation,
Cambridge, England. *

Cochran, W. G. 1980. Tecnicas de muestreo. Compania Editorial Continental S.A., Ciudad
de Mexico, Mexico.

CoGOLLOR, G., M. Cheul, and M. Poblete. 1982. Evaluacion del dano producido por
insectos en Tamarugo Prosopis tamarugo Phil. Universidad de Chile and Corporacion
Nacional Porestal, Santiago, Chile.

Collar, N. J., L. P. Gonzaga, N. Krabbe, A. Madrono Nieto, L. G. Naranjo, T. A. Parker
III, AND D. C. Wege. 1992. Threatened birds of the Americas. The ICBP/IUCN Red
Data Book. 3d ed., part 2. International Council for Bird Preservation, Cambridge,
England.

Di Castri, P. and E. R. Hajek. 1976. Bioclimatologi'a de Chile. Universidad Catolica de
Chile, Santiago, Chile.

Donoso, j., R. Rosende, I. Ulloa, and E. Cuevas. 1989. Estudio de mortalidad de arboles
en la Pampa del tamarugal. Universidad de Chile and Corporacion Nacional Forestal,
Santiago, Chile.

Estades, C. E and M. V. Lopez-Calleja. First nesting record of the Tamarugo Conebill.
Auk (In press).

FjeldsA, j., and N. Krabbe. 1990. Birds of the high Andes. Zoological Museum, Univ.

Copenhagen and Apollo Books, Svendborg, Denmark.

Gajardo, R. 1994. La vegetacion natural de Chile. Clasificacion y distribucion geografica.
Editorial Universitaria, Santiago, Chile.

Glade, A. A. (Ed.). 1988. Libro rojo de los vertebrados terrestres de Chile. Corporacion
Nacional Forestal, Santiago, Chile.

Johnson, A. W. and W. R. Millie. 1972. A new species of conebill (Coni rostrum) from
northern Chile. Pp. 3-8 in Supplement to the birds of Chile and adjacent regions of
Argentina, Bolivia and Peru (A. W. Johnson, ed.). Platt Establecimientos Graficos, Bue-
nos Aires, Argentina.

Mayr, E. and F. Vuilleumier. 1983. New species of birds described from 1966 to 1975.
J. Ornithol. 124:217-232.

Estades • TAMARUGO CONEBILL IN NORTHERN CHILE

279

McFarlane, R. W. 1975. The status of certain birds in northern Chile. Bull. Intern Council
Bird Preserv. 12:300-309.

AND E. Loo. 1974. Food habits of some birds in Tarapaca. Idesia 3:163-166.

Mills, G. S., J. B. Dunning, Jr., and J. M. Bates. 1991. The relationship between breeding
bird density and vegetation volume. Wilson Bull. 103:468-479.

Rottmann, j. and M. V. Lopez-Calleja. 1992. Estrategia nacional de conservacidn de
aves. Servicio Agricola y Ganadero, Division de Proteccion de los Recursos Naturales
Renovables, Serie Tecnica 1, Santiago, Chile.

SCHULENBERG, T. S. 1987. Observations on two rare birds, Upucerthia albigula and Coni-
rostrum tamarugense, from the Andes of southwestern Peru. Condor 89:654-658.

Shaffer, M. 1987. Minimum viable populations: coping with uncertainty. Pp. 69-86 in
Viable populations for conservation. (M. E. Soule, ed.). Cambridge Univ. Press, Cam-
bridge, England. ’ •

SuDZUKi, F. 1985. Utilizacion de la humedad ambiental por Prosopis tamarugo Phil. Pp.
35-50 in Estado actual del conocimiento sobre Prosopis tamarugo. (M. A. Habit, ed.).
United Nations, Food and Agriculture Organization, Santiago, Chile.

Tallman, D. a., T. a. Parker III, G. D. Lester, and R. A. Hughes. 1978. Notes on two
species of birds previously unreported from Peru. Wilson Bull. 90:445-446.

Wilson Bull., 108(2), 1996, pp. 280-291

AVIAN ABUNDANCE IN RIPARIAN ZONES OF
THREE FOREST TYPES IN THE
CASCADE MOUNTAINS, OREGON

Robert G. Anthony,' Gregory A. Green, ^ Eric D. Forsman,'^ and

S. Kim Nelson'

Abstract. We surveyed bird populations along headwater streams of old-growth, ma-

ture, and young coniferous forests of the Oregon Cascade Mountains during summer and
winter. Brown Creepers {Certhia americana). Chestnut-backed Chickadees (Parus rufes-
cens). Golden-crowned Kinglets (Regulus satrapa). Evening Grosbeaks {Hesperiphona ves-
pertinus), and Winter Wrens {Troglodytes troglodytes) were common in all stand types.
During the summer, abundances of Brown Creepers, Hammond s Flycatchers {Empidonax
hammondii), Hermit/Townsend’s warblers {Dendroica occidentalis), and Chestnut-backed
Chickadees were significantly higher in old-growth and mature forests compared to young
forests. Species richness and densities generally were not signihcantly different among the
stand types during winter. However, numbers of Chestnut-backed Chickadees, Evening Gros-
beaks, Golden-crowned Kinglets, Hairy Woodpeckers {Picoides villosus), and Winter Wrens
were much higher in the winter than in summer. Swainson s Thrushes {Cothorus ustulotus)
and Rufous Hummingbirds {Selasphorus rufus) were more abundant in riparian areas in this
study compared to other studies in upland forests and may be riparian associates along these
headwater streams. Received 4 Oct 1994, accepted 20 Oct. 1995.

Complexity of streamside vegetation associated with the aquatic and
terrestrial interface creates unique habitats which are generally high in
wildlife diversity and abundance (Thomas et al. 1979, Bull and Skovlin
1982). This is especially true in arid land environments or along large
streams and rivers where riparian vegetation is markedly different from
upland vegetation (Johnson and Jones 1977). However, vegetative com-
munities along small mountain streams within western coniferous forests
are less distinct from upland areas because conifers dominate and suppress
the deciduous components of the riparian community (Swanson et al.
1982). The importance of these mountain streams as habitat for birds has
been largely overlooked to date, especially in the dense forest lands of
the western Cascade Mountains.

The rapid liquidation of old-growth forests and their recognition as
unique wildlife habitat has made old-growth management a major for-
estry-wildlife issue in the Pacific Northwest (Meslow et al. 1981). Ini-
tially, the old-growth issue focused on the Northern Spotted Owl {Strix

■ National Biological Service, Oregon Cooperative Wildlife Research Unit, Dept, of Fisheries and Wild-
life, Oregon State Univ., Corvallis, Oregon 97331.

^ Parametrix, Inc. 5808 Lake Washington Blvd. NE, Kirkland, Washington 98033.

■’ USDA Forest Service, Pacific Northwest Research Station, 3200 Jefferson Way, Corvallis, Oregon
9733 I .

280

Anthony et al. • AVIAN ABUNDANCE IN RIPARIAN ZONES

281

occidentalis), but other studies have indicated that many more species of
wddlife may find optimum habitat in old-growth forests (Meslow 1978;
Thomas 1979; Verner and Boss 1980; Anthony et al. 1982; Mannan 1980,
1982; Manuwal 1991).

In this study, we compared abundance of birds in small riparian zones
among old-growth, mature, and young-aged forests. Specifically, species
richness and density of the avian communities were compared among
stand types and between summer and winter seasons.

STUDY AREAS AND METHODS

The study was conducted in the western hemlock {Tsuga heterophylla) zone of the Oregon
Cascade Range on the Blue River and McKenzie River Ranger Districts of Willamette
National Forest, Lane County, Oregon. Study areas were located along riparian zones of
second and third-order streams within old-growth (400-450 yrs), mature (130-200 yrs), and
young (25-35 yrs) forest stands. Five study areas were selected in both old-growth and
mature stands and two in young stands. Old-growth stands were natural, unharvested forests,
and mature stands were relatively even-aged and originated from an extensive wildfire dur-
ing the 1 850s. Young stands originated from clear-cut forest practices. Study areas ranged
in mean elevation from 490-975 m and varied in slope and aspect. Annual rainfall is
approximately 180 cm.

Douglas-fir (Pseudotsuga menziesii) dominated the forest community in all successional
stages. Western hemlock and western red cedar {Thuja plicata) contributed to the old-growth
canopy. Western red alder (Alnus rubra) was a conspicuous component of the young stands
and was co-dominant with big-leaf maple {Acer macrophyllum) and willow {Salix spp.).
The shrub layer included salmonberry {Ruhus spectabilis), red huckleberry {Vaccinium par-
vifohum), vine maple {Acer circinatum), salal {Gaultheria shallon). Pacific rhododendron
{Rhododendron macrophyllum), dwarf Oregon grape {Berberis nervosa), western swordfern
{Polystichum munitum), and common bracken fern {Pteridium aquilinum). The young-aged
stands had the lowest shrub cover because of the dense overstory of young Douglas-fir and
western red cedar; however, early serai plants such as willow and snowbrush ceanothus
{Ceanothus velutinus) were common.

We established five plots at 200-m intervals along the riparian zone in each of the 12
study areas. Inclusion of more than five plots was not possible because of difficulty of
locating homogeneous habitats that were more than 1 km long. Stream noise was minimal,
because the streams were small (1—3 m wide). Birds were surveyed using the variable
circular plot method (Reynolds et al. 1980). Each plot was sampled once per week for 10
min during the period of dawn to 10:00 h. Each bird seen or heard during the sample period
was identified, and the distance from their location to the plot center was estimated and
verified with range finders. Surveys were repeated five to seven times during summer (May-
June) and six times during winter (Jan-Feb). The third and fourth authors conducted the
surveys and alternated visits to a stand to minimize observer bias.

Estimates of bird density were determined by the method first described by Emien (1971),
and later modified by Ram.sey and Scott (1981), to distance sampling of birds. At least 15
detections per species over all stands were required for the algorithm to estimate detection
distances and generate density estimates. Densities were not estimated for species with <15
detections.

Differences in densities between serai stages were tested using analysis of variance with
Duncan’s multiple range test for mean separation (Steel and Torrie 1980:187). All statistical

282

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 1

Common Names, Scientific Names, and Abbreviations for Bird Species Observed

DURING the Study

Common name

Scientific name

Abbreviation

Black-capped Chickadee

Parus atricapillus

BCCH

Brown Creeper

Certhia americana

BRCR

Chestnut-backed Chickadee

P. rufescens

CBCH

Common Raven

Corvus corax

CORA

Dark-eyed Junco

Junco hyemalis

DEJU

Evening Grosbeak

Hesperiphona vespertinus

EVGR

Golden-crowned Kinglet

Regulus satrapa

GCKI

Gray Jay

Perisoreus canadensis

GRJA

Hammond’s Elycatcher

Empidonax hammondii

HAFL

Hairy Woodpecker

Picoides villosus

HAWO

Hermit Warbler

Dendroica occidentalis

HEWA

MacGillivray’s Warbler

Oporornis tolnuei

MGWA

Olive-sided Flycatcher

Contopus borealis

OS EL

Pine Siskin

Corduelis pinus

PISI

Pileated Woodpecker

Dryocopus pileatus

PIWO

Red-breasted Nuthatch

Sitta canadensis

RBNU

Rufous Hummingbird

Selasphorus rufus

RUHU

Steller’s Jay

Cyanocitta sfelleri

STJA

Swainson’s Thrush

Catharus ustulatus

SWTH

Vaux’s Swift

Chaetura vauxi

VASW

Varied Thrush

Ixoreus naevius

VATH

Warbling Vireo

Vireo gilvus

WAVI

Pacific Slope Flycatcher

Empidonax difficilis

WEFL

Western Tanager

Piranga ludovicianus

WETA

Wilson’s Warbler

Wilsonia pusilla

WIWA

Winter Wren

Troglodytes troglodytes

WIWR

tests were performed at the 0.05 level of significance. A key to bird species codes, common
names, and scientific names is provided in Table 1 .

RESULTS

Species composition. — Forty-six species were detected during the
study. The Brown Creeper, Chestnut-backed Chickadee, Golden-crowned
Kinglet, Evening Grosbeak, and Winter Wren were common in all stands
during both seasons (Table 2). In addition, Hammond’s Flycatcher, Her-
mit/Townsend’s warbler, Swainson’s Thrush, and Western Flycatcher were
found in all stands during the summer. Because the study was conducted
in the zone of hybridization between Hermit and Townsend’s warblers,
we could not distinguish the two species by song. Vaux’s Swift (4 ob-
servations) and Black-capped Chickadee (8 observations) were observed

Anthony et al. • AVIAN ABUNDANCE IN RIPARIAN ZONES

283

Table 2

Avian Population Densities (#/40 ha) in Old-growth, Mature, and Young Seral
Stages in the Western Cascade Mountains, Oregon, during Summer and Winter 1984

Old-growth (N = 5)

Mature (N

= 5)

Young (N = 2)

Species-’

x

± 2 SE

X

± 2 SE

X

± 2 SE

BRCR

24.10

6.50

Summer

9.50

4.50

1.00

1 8**

CBCH

42.00

17.20

33.40

19.20

22.40

1 1 3*

CORA

0.70

1.20

0.00

0.00

0.70

0.80

DEJU

9.00

9.70

1.60

1.50

26.30

36.90

EVGR

0.10

0.10

0.60

0.30

0.20

0.3*

GCKI

14.30

6.70

12.10

8.00

1 1.90

14.10

HAFL

30.90

10.60

30.70

10.90

7.40

0.0*

HAWO

0.50

0.50

0.50

0.30

0.50

1.00

HETH

0.90

1.00

0.30

0.20

1.80

2.40

HEWA

5.90

4.70

4.70

2.30

1.70

2.00

MGWA

1.60

3.20

0.00

0.00

8.80

17.70

ROBI

0.10

0.10

0.80

0.80

0.50

0.60

RUHU

66.70

29.80

29.50

31.40

40.00

80.00

STJA

0.80

0.70

1.20

0.90

3.50

2 1**

SWTH

6.50

3.90

14.30

9.70

25.50

3 0*

VATH

3.60

2.00

1.20

1.30

0.00

0.00

WAVI

0.00

0.00

0.00

0.00

1 1.70

23.30

WEFL

1 1.70

4.80

12.50

11.30

14.60

0 00

WETA

1.40

1.20

2.20

1.40

5.40

10.70

WIWA

1.90

2.00

3.90

3.50

3.00

0.30

WIWR

39.30

5.00

47.00

13.20

18.80

27.40

Total

261.70

58.50

205.70

43.80

205.30

169.30

BRCR

6.20

2.90

Winter

7.50

2.20

3.60

7.10

CBCH

103.20

30.00

91.70

32.20

97.70

77.60

DEJU

3.90

4.50

1.30

2.50

6.00

4.90

EVGR

5.00

4.80

10.90

6.70

92.90

144.0*

GCKI

55.80

1 1.00

57.40

31.90

72.10

13.60

HAWO

5.40

4.50

2.60

3.30

1.80

3.70

WIWR

171.10

24.00

218.90

88.70

148.30

70.00

Total

350.50

36.90

390.20

8 1 .00

422.30

139.70

“ Table 1 .

* P < 0.05, analysis of variance.
** p < 0.0 1, analysis of variance.

only in old-growth stands, and Olive-sided Flycatcher (10 observations)
were observed in two old-growth and one mature stand during the sum-
mer. Varied Thrush was recorded only in old-growth and mature stands
and during summer. Pileated Woodpecker was detected predominately in
old-growth and mature forests. Similarly, the Hairy Woodpecker was re-

284

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

corded mostly in old-growth and mature forests during winter, but some
were observed in one young-aged stand. Twenty-one observations of War-
bling Vireos occurred in a single young-aged stand during summer.

Summer densities and species richness. — No major differences in spe-
cies richness were observed between the three forest types with 26, 29,
and 32 species recorded in mature, old-growth, and young stands, re-
spectively. Thirty-nine species of birds were detected in the 12 stands
during the summer, of which 21 were observed >15 times to estimate
density. Brown Creepers, Chestnut-backed Chickadees, Hermit/Town-
send’s warblers. Rufous Hummingbirds, and Varied Thrushes had highest
densities in old-growth stands (Table 2). Abundances of Brown Creepers
and Chestnut-backed Chickadees were significantly {P < 0.05) higher in
the old-growth and mature forests than in young forests. Abundances of
Hammond’s Flycatchers (P < 0.05), Hermit/Townsend’s warblers, and
Winter Wrens were similar in old-growth and mature forests and greater
than densities in young stands. Common Ravens were equally abundant
in old-growth and young stages, but densities were very low, and there
was no significant difference among all three serai stages. Numbers of
Evening Grosbeaks, Wilson’s warblers, and Winter Wrens were highest
in the mature stands; but only densities of the Evening Grosbeak were
significantly {P < 0.05) higher. Species with highest abundance in the
young successional stages were Dark-eyed Junco, MacGillivray’s warbler.
Stellar’s Jay, Swainson’s Thrush, Warbling Vireo, and Western Tanager.
The Warbling Vireo was found only in one young stand and at low num-
bers. Abundances of Hairy Woodpecker, Golden-crowned Kinglet, and
Pacific Slope Elycatcher were similar across all forest types. Abundances
of Brown Creeper, Varied Thrush, Chestnut-backed Chickadee, and Her-
mit warbler were greater with increasing age of forest stands, while num-
bers of Steller’s Jays, Swainson’s Thrushes, and Western Tanagers were
less with increasing age of forest stands.

Winter densities. — Species richness during winter differed little among
the three forest types, with 12, 15 and 15 species recorded in old-growth,
mature, and young stands, respectively. Twenty species were recorded in
all 12 stands during the winter survey period, and densities could be
estimated for seven of these. Except for Evening Grosbeaks, only minor
differences in abundance were noted among the three forest types (Table
2). Evening Grosbeak densities were 10-20 times greater {P < 0.05) in
the young stands as in mature and old-growth forests. Highest numbers
were found in the old-growth forests for Chestnut-backed Chickadees and
Hairy Woodpeckers, although neither were significantly different among
forest types. Abundances of Chestnut-backed Chickadees were high and
similar throughout all forest types. Brown Creepers and Winter Wrens

Anthony et al. • AVIAN ABUNDANCE IN RIPARIAN ZONES

285

had their highest numbers in the mature forests. Abundances of the Dark-
eyed Junco, Evening Grosbeak, and Golden-crowned Kinglet were high-
est in young forests. Hairy Woodpecker numbers were greater with in-
creased stand age; Evening Grosbeak numbers were less with stand age.

Seasonal changes. — Abundances of Chestnut-backed Chickadees, Eve-
ning Grosbeaks, Golden-crowned Kinglets, Hairy Woodpeckers, and Win-
ter Wrens were much higher in winter as compared to summer (Figs. 1,
2)- These species are the most common winter residents, and some form
large winter flocks. In contrast, numbers of Dark-eyed Junco were gen-
erally lower in the winter than in summer (Fig. 2c). All the warblers,
flycatchers, and thrushes were not present on the study area during winter
because they migrated out of the area.

DISCUSSION

Overall, population densities were highest in old-growth stands during
summer and young stands during winter. These results are different from
those of Manuwal and Huff (1987) for the Washington Cascades where
species richness and abundance were greater in old-growth versus young
forests during winter. Birds present in the summer period were, for the
most part, breeders, and the greater structural diversity of vegetation in
the old-growth forests likely provided more nesting and foraging habitat,
resulting in greater abundance. Total overall bird numbers for the three
successional stages were approximately twice as high during the winter
as compared to the summer. This was likely a result of “flocking” of
migrants from higher elevations and latitudes, and there could be more
seed- and fruit-bearing vegetation in young stands.

Brown Creepers were relatively more abundant in old-growth stands
than in young and mature stands during summer; they were more abun-
dant with increasing age of stands. This is consistent with other studies
(e.g., Thomas 1979, Verner 1980, Mannan 1982). Fidelity to old-growth
was not as great during winter, as Brown Creeper numbers were similar
in both old-growth and mature stands. We found highest numbers of Var-
ied Thrushes in old-growth stands during summer, and they were totally
absent from the young stands. Varied Thrushes appear to reach their high-
est breeding densities in older coniferous forests (Ramsden et al. 1979,
Mannan 1982). We observed greater abundances of Chestnut-backed
Chickadees and Golden-crowned Kinglets in old-growth forests. These
results compare well with findings by Hagar (1960), Buckner et al.
(1975), and Manuwal (1991) who found both species to prefer the late-
successional forests. We found Hammond’s Flycatchers equally abundant
in the old-growth and mature forests and four times greater than in the
young stands during the summer, which is consistent with the reports of

Density per 40 ha Density per 40 ha Density per 40 ha

286

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Anthony et al. • AVIAN ABUNDANCE IN RIPARIAN ZONES

287

Hagar (1960), Verner (1980), McMillan and Walter ( 1981), and Manuwal
(1991) for Douglas-fir forests. However, Mannan (1982) found Ham-
mond’s Flycatcher densities to be equal for both 85 and >200-year-old
stands in mixed-conifer forests in northeastern Oregon. Evening Gros-
beaks were found at consistently low numbers (<1/40 ha) during summer
but were significantly more abundant in mature stands. They were about
100-fold more abundant during winter with highest densities found in
young stands. Our results are consistent with those of Thomas (1979) and
McMillan and Walter (1981) who found that Evening Grosbeaks breed
in dense stands of older conifer forests and move to younger forests dur-
ing winter. Winter Wrens were most abundant in three mature stands
during both seasons, and all of these sites were at low elevations and
contained much shrub cover. Hagar (1960) also found Winter Wren den-
sities to be highest in mature stands, while Peterson and Peterson (1983)
stated that slash and brush cover positively influenced Winter Wren den-
sities, not age class.

In summary, the results of this study provide information on species
that attain their highest densities in small riparian areas of late-succes-
sional forests (>120 yrs old) or are found predominately in old-growth
forests. During summer. Brown Creepers, Varied Thrushes, Hermit/Town-
send’s warblers, and Chestnut-backed Chickadees were more abundant
with increasing age of stands. Hammond’s Flycatchers were more abun-
dant m old-growth and mature forests and much less common in young
forests. Evening Grosbeaks were most abundant in mature forest stands
Vaux’s Swifts, Black-capped Chickadees, Olive-sided Flycatchers, and Pi-
leated Woodpeckers were not abundant enough to estimate densities, but
most observations of these species were in old-growth stands. Likewise,

numbers of Hairy Woodpeckers were greater with age of forest stands
during winter.

Of the above list of species, the Brown Creeper, Chestnut-backed
Chickadee, Black-capped Chickadee, Evening Grosbeak, Varied Thrush,
and Hairy Woodpecker are common and widely distributed species, so
any dependencies on late-successional forests are unlikely. However,'the
group contains a number of cavity-nesting species (Chestnut backed
Chickadee, Black-capped Chickadee, Brown Creeper, Vaux’s Swift and
Plicated Woodpecker) that may be responding to higher densities of snags

Fig. 1 . Densities of Chestnut-backed Chickadee ( 1 a), Evening Grosbeak ( I b) and Gold-
en-crowned Kinglet (Ic) in three forest types during summer (S) and winter (W) western
Oregon. Circles and bars represent mean ± 2 SE; where only circles are present, error bars

Density per 40 ha Density per 40 ha Density per 40 ha

288

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Anthony et uL • AVIAN ABUNDANCE IN RIPARIAN ZONES

289

Table 3

Ranked Abundance of Bird Species in Young, Mature, and Old-growth Forests of
THE Oregon and Washington Cascades

Young

Mature

Gilbert

and

This Allwine Manuwal This
Species" study 1991 1991 study

Gilbert

and

Allwine Manuwal
1991 1991

RUHU

CBCH

WIWR

HAFL

BRCR

GCKI

WEFL

DEJU

SWTH

HEWA

RBNU

VATH

PIS I

GRJA

VASW

1

4

5
9

16

7

6
2
3

3 6

2 2

4

9

3

5 4
5

1 1

6 8
7

10

4
2
I
3
8
7
6

10

5

4 5

2 1

3

6

2

6 4

8

1 3

5 9
7

10

“ See Table 1 for list of common and scientific names of birds.

Old-growth

Gilbert

and

This Allwine Manuwal

study 1991 1991

1

2 3 3

3 1 1

4 4

5 . 7

6 2

7 5 4

8 8
9

2 9

6

6 5

10

or logs in late-successional forests. The importance of snags to these
species in relation to their association with late-successional forests needs
further clarification. Hairy Woodpeckers were the only species that dis-
played an association with old-growth forests during winter. Based on
these findings and the common and widely distributed nature of Hairy
Woodpeckers, winter bird surveys in riparian areas are probably not high
priority to assess associations with late-successional forests.

Riparian associates — Similar studies on avian communities in upslope
habitats have been conducted in young, mature, and old-growth forests
in the Oregon Cascades (Gilbert and Allwine 1991 ), Oregon Coast Range
(Carey et al. 1991), and Washington Cascades (Manuwal 1991). These
studies provide estimates of relative bird abundance but only qualitative
comparisons of ranked abundances can be made (Table 3), because the

f—

Fig. 2. Densities of Hairy Woodpecker (2a), Winter Wren (2b), and Dark-eyed Junco
(2c) in three forest types during summer (S) and winter (W), western Oregon. Circles and
bars represent mean ± 2 SE; where only circles are present error, bars are too small to
graph.

290

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

previous studies did not estimate densities. All the species we detected in
riparian areas were also found in upland areas in the above studies so we
did not identify any riparian obligate species. However, Swainson’s
Thrushes and Rufous Hummingbirds were abundant along headwater
streams in this study, but were detected infrequently in upland areas in
the above studies. Further study may reveal some association with riparian
areas for these two species. In contrast, Hermit/Townsend’s warblers. Var-
ied Thrushes, and Red-breasted Nuthatches were abundant in upland areas
of the above three studies but were not among the top 10 most abundant
species in riparian areas in this study. These species may be associated
with upland habitats.

ACKNOWLEDGMENTS

This study was funded by the U.S.D.A. Lorest Service, Pacific Northwest Research Sta-
tion, Olympia, Washington. The research was contracted through the Oregon Cooperative
Wildlife Research Unit; Oregon State Univ., Oregon Department of Lish and Wildlife, Na-
tional Biological Service, and Wildlife Management Institute cooperating.

LITERATURE CITED

Anthony, R. G., R. L. Knight, G. T Allen, B. R. McClelland, and J. I. Hodges. 1982.
Habitat use by nesting and roosting bald eagles in the Pacific Northwest. Trans. N.
Amer. Wildl. Natur. Resour. Conf. 47:332-342.

Buckner, C. H., A. J. Erskine, R. Lidstone, B. B. McLeod, and M. Ward. 1975. The
breeding bird community of coast forest stands on northern Vancouver Island. Murrelet
56:6-11.

Bull, E. L. and J. M. Skovlin. 1982. Relationships between avifauna and streamside
vegetation. Trans. N. Amer. Wildl. Natur. Resour. Conf. 47:496—506.

Carey, A. B., M. M. Hardt, S. P. Horton, and B. L. Biswell. 1991. Spring bird com-
munities in the Oregon Coast Range. Pp. 123—142. in Wildlife and vegetation of un-
managed Douglas-fir forests. (L. P. Ruggerio et al., eds.) USDA Lorest Seivice. Gen.
Tech. Rep. PNW-GTR-285, Portland, Oregon.

Emlen, j. T. 1971. Population densities of birds derived from transect counts. Auk 88:323-
342.

Gilbert, P. L and R. Allwine. 1991. Spring bird communities in the Oregon Cascade
range. Pp. 145-158. in Wildlife and vegetation of unmanaged Douglas fir forests. (L.L
Ruggerio et al., eds.) USDA Lorest Service. Gen. Tech. Rep. PNW-GTR-285, Portland,
Oregon.

Hagar, D. C. 1960. The interrelationships of logging, birds, and timber regeneration in the
Douglas-fir region of northwestern California. Ecology 41:1 16-125.

Horvath, D. 1963. Contributions to nesting ecology of forest birds. M.S. thesis. Univ. of
British Columbia, Vancouver, British Columbia.

Johnson, R. R. and D. A. Jones (eds.) 1977. Importance, pre.servation, and management
of riparian habitat: A symposium. U.S. Lor. Serv., Gen. Tech. Rept. Rm-43. Port Collins,
Colorado.

M ANNAN, R. W. 1980. Assemblages of bird species in western coniferous old-growth for-
ests. Pp. 357-368 in R. M. Degraff (ed.). Management of western forests and grasslands
for nongame birds. Gen. Tech. Rept. INT-86. U.S. Lor. Serv., Odgen, Utah.

Anthony et al. • AVIAN ABUNDANCE IN RIPARIAN ZONES

291

Mannan, R. W. 1982. Bird populations and vegetation characteristics in managed and old-
growth forests, northeastern Oregon. Ph.D. diss. Oregon State Univ. Corvallis, Oregon.

Manuwal, D. a. 1991. Spring bird communities in the southern Washington Cascade
Range. Pp. 161-174 in Wildlife and vegetation of unmanaged Douglas-fir forests (L. E
Ruggerio et al., eds.). USDA Forest Service. Gen. Tech. Rep. PNW-GTR-285, Portland,
Oregon.

AND M. H. Huff. 1987. Spring and winter bird populations in a Douglas-fir forest
sere. J. Wildl. Manage. 51:586-595.

McMillan, A. and M. Walter. 1981. Wildlife habitats and species management relation-
ships program. Vol. III. Birds. Siuslaw Nat. For., U.S. For. Serv., Corvallis, Oregon.

Meslow, E. C. 1978. The relationship of birds to habitat structure-plant communities and
successional stages. Pp. 12-18 in Proceedings of workshop on nongame bird habitat
management in the coniferous forests of the western United States. U.S. For. Serv.,
Gen. Tech. Rept. WO-1. U.S. For. Serv., Washington, D.C.

, C. Maser, and J. Verner. 1981. Old-growth forests as wildlife habitat. Trans. N.
Amer. Wildl. Natur. Resour. Conf. 46:329-335.

Peterson, E. B. and N. M. Peterson. 1983. Summer bird densities in relation to forest
types in western North America: Annotated bibliography and analysis of literature. Can.
Wildl. Serv., Delta, British Columbia.

Ramsden, D. j., L. j. Fyon, and G. L. Halvorsen. 1979. Small bird populations and
feeding habitats — western Montana in July. Amer. Birds 33:11-16.

Ramsey, F. L. and J. M. Scott. 1981. Analysis of bird survey data using a modification
of Fmlen’s method. Pp. 483-387 in Estimating numbers of terre.strial birds. Studies in
avian biology No. 6 (C. J. Ralph and J. M. Scott, eds.). Cooper Ornith. Soc. Publ. Allen
Press, Fawrence, Kansas.

Reynolds, R. T, J. M. Scott, and R. A. Nussbaum. 1980. A variable circular-plot method
for estimating bird densities. Condor 82:309-313.

Swanson, F. J., S. V. Gregory, J. R. Sedell, and A. B. Campbell. 1982. Land-water
interactions: the riparian zone. Pp. 267-291 in Analysis of coniferous forest ecosystems
in the western United States. US/IBP Synthesis Series 14 (R. L. Edmonds, ed.). Hutch-
inson Ross Publ. Co., Stroudsburg, Pennsylvania.

Thomas, J. W. (ed.) 1979. Wildlife habitats in managed forests: the Blue Mountains of
Oregon and Washington. Agric. Handb. 553. U.S. For. Serv., Washington, D.C.

, C. Maser, and J. F. Rodiek. 1979. Riparian zones. Pp. 40-47 in Wildlife habitats
in managed forests: the Blue Mountains of Oregon and Washington. Agric. Handb. 553
(J. W. Thomas, ed.). U.S. For. Serv., Washington, D.C.

Steel, R. G. D. and J. H. Torrie. 1980. Principles and procedures of statistics. McGraw-
Hill, New York, New York.

Verner, J. 1980. Bird communities of mixed-conifer forests of the Sierra Nevada. Pp. 198-
223 in Management of western forests and grasslands for nongame birds. (R. M. De-
graff, ed.). U.S. For. Serv., Gen. Tech. Rept. INT-86. Ogden, Utah.

• AND A. S. Boss (eds.) 1980. California wildlife and their habitats: western Sierra

Nevada. U.S. For. Serv., Gen. Tech. Rept. PSW-37. Berkeley, California.

Wilson Bull., 108(2), 1996, pp. 292-301

HABITAT CHANGES AND SUCCESS OE ARTIEICIAL
NESTS ON AN ALKALINE ELAT

Marcus T. Koenen,' David M. Leslie, Jr.,“ and Mark Gregory^

Abstract. We studied habitat changes and success of artificial ground nests on an ex-

pansive alkaline flat at Salt Plains National Wildlife Refuge (NWR), Oklahoma, in 1993
and 1994. Aerial photographs of the refuge taken during 1941-1942, 1966, and 1989 were
digitized to evaluate changes of the alkaline flat, herbaceous rangeland, and shrub rangeland
that was dominated by saltcedar (Taniarix spp.). Vegetation cover increased by about 600
ha between 1941 and 1989, and alkaline flat decreased by >240 ha. Field experiments were
conducted to determine predator and flooding impacts on artificial nests that simulated Least
Tern {Sterna antillarum) and Snowy Plover (Charadrius alexandrinus) nests. Experimental
nest plots were placed on the alkaline flat adjacent to, 500 m from, and 1000 m from
herbaceous rangeland, shrub rangeland, and stream habitat that was not associated with
vegetation. Plot comparisons were made by calculating the probability of nest success with
a modified Mayfield Method. Nests near vegetation had significantly higher losses to mam-
malian predators (P < 0.05) but significantly lower losses to flooding (P < 0.05) than nests
at 500 or 1000 m from vegetation. Encroaching vegetation will likely continue to reduce
habitat for ground-nesting birds and simultaneously increase predation lates on nests. Re-
ceived 20 June 1995, accepted 10 Jan. 1996.

The interior population of the Least Tern {Sterna antillarum) has been
listed as endangered since 27 June 1985 (U.S. Fish and Wildl. Seiv.
1985). The inland population of the Snowy Plover {Charadrius alexan-
drinus) is currently listed as a Category 2 species (U.S. Fish and Wildl.
Serv. 1991), and the coastal population of the Snowy Plover was federally
listed as threatened on 5 March 1993 (U.S. Fish and Wildl. Serv. 1993).
The population decline of the interior Least Tern has been attributed large-
ly to loss of breeding habitat due to river channelization and construction
of impoundments (U.S. Fish and Wildl. Serv. 1990). Snowy Plovers use
similar habitat and likely are affected by the same habitat changes that
caused the population decline of the interior Least Tern.

An alkaline flat at Salt Plains NWR contains the largest concentration
of breeding Least Terns in Oklahoma (U.S. Fish and Wildl. Serv. 1990).
Predation and flooding have been identified as the major causes of Least
Tern and Snowy Plover egg losses on the flat (Grover and Knopf 1982,
Hill 1985, Utych 1993). Coyotes {Canis latran.s) are the main egg predator,
and rain causes sheet flooding on the flat which can wash eggs out of nests.

' Oklahoma Cooperative Fish and Wildlife Research Unit, Dept, of Zoology, Oklahoma State Univ.,

Stillwater, Oklahoma 74078. ■ ■

2 U.S. National Biological Service, Oklahoma Cooperative Fish and Wildlife Research Unit, Dept, of
Zoology, Oklahoma State Univ., Stillwater, Oklahoma 74078.

^ Dept, of Agronomy, Oklahoma State Univ., Stillwater, Oklahoma 74078.

292

Koenen et al. • HABITAT CHANGES AND NEST SUCCESS

293

Predation and flooding are likely consequences of habitat changes that
have occurred at Salt Plains NWR since its creation in the 1930s. Res-
ervoir consti action and the spread of saltcedar {Tamarix spp.) have altered
riparian habitats in the southwestern United States, including parts of
Oklahoma (Block 1994, Stinnet et al. 1987). Several studies have reported
higher predation lates on artificial bird nests close to forest-prairie or
forest-farmland edge habitats than on nests farther from the edge (Andren
and Angelstram 1988, Burger 1988, Baton 1993, Wilcove et al. 1986),
but no studies have examined predation rates on Least Terns and Snowy
Plover eggs relative to proximity to vegetated habitat. We assessed habitat
changes at Salt Plains NWR and evaluated their impact on predation and
flooding of Least Tern and Snowy Plover nests. As our alternate hypoth-
eses, we predicted that predation rates would be higher on nests situated
close to vegetation that provides cover for predators than nests farther
from vegetation and that nests closer to streams would be more suscep-
tible to flooding losses than nests farther away from streams.

STUDY AREA AND METHODS

Salt Plains NWR is located in Alfalfa County in northcentral Oklahoma and currently
contains an alkaline flat of 5095 ha where Least Terns and Snowy Plovers nest (Grover and
Knopf 1982, Hill 1985, Utych 1993, Schweitzer 1994). The alkaline flat is closed to the
general public except for a small public use area on the southwestern corner of the flat that
is open between 1 April and 15 October for collecting selenite crystals (Koenen 1995).

The alkaline flat is nearly level and poorly drained. The water table is at the surface in
some areas and up to 1-m deep in others (Williams and Grover 1975). Salt Plains NWR
receives an average annual rainfall of 68 cm, of which about 60% occurs in spring and
summer. Rain can cause 1-3 cm sheets of moving water on the alkaline flat, which can
remain for several hours to several days. Sheet flooding can wash eggs out of nests and
submerge entire colonies.

The alkaline flat at Salt Plains NWR is nearly bare; sparse vegetation includes sea purslane
{Sesuvium verrucosiim) and inland salt grass (Distichlis stricta). Vegetation forms well-
defined borders at the edge of the alkaline flats. The east and south sides of the alkaline flat
are dominated by exotic saltcedar; the north and west sides of the alkaline flat are primarily
bordered by grazed mixed-grass prairie.

Creeks flow across the alkaline flat into the Great Salt Lake of Oklahoma and are ephem-
eral and multibranched. The Great Salt Lake was created by a dam across the Salt Fork of
the Arkansas River in 1941. The resulting re.servoir flooded about 30% of the original 1 1,
137-ha alkaline flat (Purdue 1976).

Least Terns nest in loo.sely defined colonies along the west branch of the Salt Fork of the
Arkansas River, Clay Creek, Cottonwood Creek, and Spring Creek and in scattered patches
throughout the alkaline flat (Hill 1985, Schweitzer 1994). The average distance between
Least Tern nests is 70 m (Schweitzer 1994). Snowy Plover nests are widely scattered and
can be found in Least Tern colonies and areas not used by terns. Nests of both species are
shallow scrapes (ca 0.5-4 cm deep X 5-10 cm wide) and typically contain 1-3 eggs (Har-
rison 1979).

We identified and manually digitized habitat at Salt Plains NWR from 1:16,000 and 1:

294

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

20,000 aerial photographs obtained from the Natural Resource Conservation Service, U.S.
Dept, of Agriculture, using an Altek graphic digitizer board and GRASS#4.0 (USACERL
1991) Geographical Information System software. Aerial photographs were taken on (1) 24
November 1941 and 22 January 1942, (2) 10 June 1966, and (3) 2 December 1989. The
refuge boundary was superimposed over the 1966 photograph, digitized as a separate file,
and laid over each digitized habitat map. Digitized habitat (vector) maps were labeled ac-
cording to major cover types and converted to 30-m resolution raster maps for habitat
analyses (Johnson 1993). Cover (ha) was calculated for herbaceous rangeland, shrub range-
land, alkaline flat, and the Great Salt Lake, including an island within Salt Plains NWR’s
boundary south of Highway 11.

To evaluate the impact of habitat changes and predators on survival of nests of ground
nesting birds, we conducted controlled experiments with artificial nests and Japanese Quail
(Coturnix coturnix) eggs that simulated Least Tern and Snowy Plover nests. Artificial nests
were used because we could control nest placement in desired treatment areas. The goal was
to identify potential predators, areas sensitive to predation, and areas sensitive to flooding.

In 1993, 60- X 90-m plots were placed adjacent to, 500 m from, and 1000 m from shrub
rangeland and the west branch of the Salt Fork of the Arkansas River in the northeastern
corner of the alkaline flat (N = 3 plots). In 1994, the experiment was expanded with 15
new plots. A new plot was established about 1000 m north of each of the 1993 plots (N =

3 plots) to replicate assessment of proximity to shrub rangeland. Paired plots were placed
about 1000 m from one another and adjacent to, 500 m from, and 1000 m from grazed
herbaceous rangeland on the west side of the alkaline flat (N = 6 plots). Similarly, paired
plots were placed adjacent to, 500 m from, and 1000 m from Cottonwood Creek, not
associated with vegetation cover (N = 6 plots).

Each plot contained 12 artificial nest scrapes placed 30 m apart, and each nest contained
2—3 quail eggs. Eggs were placed in the nests for three 21 -day trials (17 May— 6 June; 16
June-6 July; and 16 July-6 August) in 1993 and 1994 to imitate the incubation period of
Least Terns and Snowy Plovers (Hill 1985). The remaining eggs were removed after each
21 -day trial, and scrapes were left empty for 10 days before the beginning of the next trial.
Nests were monitored every 3—4 days to determine predation rates and other factors causing
nest losses.

Mayfield’s method (1975) was used to compare the probability of nest success in each
plot. We used a slight modification of the Mayfield method to compare effects of flooding
and predation. The Mayfield method determined daily mortality rates and the probability of
ne.st success based on nest failure over the number of days that nests were observed (ex-
posure). Because nest failure may have been the result of flooding, predation, abandonment,
or other factors, we modified the Mayfield method to calculate separate rates of daily mor-
tality due to predators (predator mortality) and flooding (flooding mortality). Only clutches
lost to predators were considered to have failed when calculating predator mortality. Simi-
larly, only clutches lost to flooding were considered to have failed when calculating flooding
mortality. The number of days that >one egg remained in nest scrapes was used to determine
exposure. Predator and flooding mortality rates were compared among treatments with 95%
confidence intervals; significant differences (P < 0.05) were demonstrated by non-overlap-
ping confidence intervals (Johnson 1979). Nest success comparisons were made by pooling
data for similar treatments (i.e., similar distance from a vegetation type) and trial periods
when there were no significant differences.

A nest was considered successful if sone egg remained in the nest scrape at the end ol
the 21 -day trial. Nests were considered predated if crushed eggs, large shell fragments, and/
or predator footprints were located at a nest. Nests were considered flooded if eggs were

Koenen et al. • HABITAT CHANGES AND NEST SUCCESS

295

Table 1

Aerial Coverage of Dominant Habitat (ha) at Salt Plains National Wildlife

Refuge

1941/1942

1966

1989

Herbaceous rangeland

311

197

336

Shrub rangeland (north)

1546

1853

2142

Shrub rangeland (south)

1044

937

1014

Salt flat

5342

5688

5095

Great Salt Lake

3811

3290

3349

Total area

12,175

12,175

12,175

washed out of nests and relocated in the area. Nests without clear signs of outcome were
categorized as unknown and were not included in the final analysis.

RESULTS

Habitat changes. — The refuge boundary south of Highway 1 1 encom-
passed 12,175 ha (Table 1). Herbaceous rangeland decreased by 114 ha
between 1941 and 1966 and increased by 139 ha between 1966 and 1989.
The most dramatic changes occurred on the northeastern side of the refuge
where shrub rangeland increased from 1546 to 1853 ha from 1941 to
1966 and increased to 2142 ha between 1966 to 1989. The shrub vege-
tation spread 12.3 ha/yr between 1941 and 1966 and 12.6 ha/yr between
1966 and 1989. Total herbaceous and shrub rangeland cover increased
from 2901 ha in 1941 to 3492 ha in 1989, which represented a 3.4 ha/yr
spread from 1941 to 1966 and a 22.0 ha/yr spread from 1966 to 1989.
The alkaline flat decreased from 5342 ha in 1941 to 5095 ha in 1989,
and the Great Salt Lake decreased 462 ha over the same period. This
represented a net loss of 709 ha of alkaline flat and lake cover.

Predator and flooding impact. — In 1993, there were no significant dif-
ferences iP > 0.05) in overall nesting success of artificial nests at various
distances from shrub rangeland (Table 2). In 1994, overall nest success
adjacent to shrub rangeland (0.42) and adjacent to Cottonwood Creek
(0.47) was significantly higher {P > 0.05) than nest success in plots 500
m from Cottonwood Creek (0.16) (Table 3). Nest success was not sig-
nificantly different among plots placed adjacent to, 500 m, and 1000 m
from shrub rangeland or herbaceous rangeland. In contrast, nest success
was higher adjacent to Cottonwood Creek than 500 m from the Creek.

Coyotes were the only mammalian nest predator positively identified
on the salt flats by sight or tracks; however, some tracks may have been
from feral dogs {Canis familiaris). Canids left distinctive eggshell remains
at nests where eggs were eaten; egg remains were similar to those shown

296

THE WILSON BULLETIN • Vol. 108, No. 2. June 1996

Table 2

Probability of Nest Success (Mayfield Method) and 95% Confidence Interval of
ALL Nest Plots Placed at Three Intervals (0 m, 500 m, and 1000 m) from Shrub
Rangeland at Salt Plains National Wildlift; Refuge, Oklahoma, in 1993 (N — 36)

Adjacent to

Shrub rangeland

500 m

1000 m

Overall nest success

0.49“

0.37“

0.28“

Confidence interval (95%)

0.35-0.68

0.25-0.55

0.21-0.50

Nest success based on canid mortality

0.24

0.31

0.15

Confidence interval (95%)

0.08-0.69

0.12-0.79

0.02-1.21

Nest success based on flood mortality

0.52

0.43

0.49

Confidence interval (95%)

0.24-1.09

0.18-0.98

0.23-0.99

“ Probabilities followed by the same letter are not significantly different among columns; 95% confidence interval com
parisons (Johnson 1979).

and discussed by Scoter (1946). Ring-billed Gulls {Lams delawarensis)
increased substantially on the alkaline flat in late July and August 1993
and 1994. They predated up to 83% of artificial nests in treatment plots,
but this was not included in our analysis of predation because they did
not predate the relatively few Least Tern and Snowy Plover nests that
remained that late in the nesting season (Hill 1985, Schweitzer 1994,
Koenen 1995). Therefore, our estimates of predator mortality on artificial
nests represented predation by only canids, primarily coyotes.

In 1993, there were no significant {P < 0.05) differences in nest success
among plots associated with shrub rangeland based on mortality due to
canids (Table 2). However, comparison of canid predated plots in 1994
indicated significant lower nest success for plots adjacent to shrub and
herbaceous rangelands and within 500 m of herbaceous rangeland than
500 m and 1 000 m from shrub rangeland, 1 000 m from herbaceous range-
land, and all Cottonwood Creek plots (Table 3).

Losses of artificial nests associated with shrub rangeland due to flood-
ing were not significantly different among plots in 1993 (Table 2). Com-
parison of flooded nests in 1994 indicated significantly {P > 0.05) higher
nest success adjacent to shrub rangeland (0.89) than on plots located 500
m and 1000 m from shrub rangeland, 1000 m from herbaceous rangeland,
and all Cottonwood Creek plots (Table 3). Plots adjacent to herbaceous
rangeland also had higher nest success than plots located 500 m and 1000
m from Cottonwood Creek.

DISCUSSION

The absence of high scouring floods due to flood control by reservoir
construction has resulted in dense saltcedar stands in sandy floodplains

Table 3

Koenen el al. • HABITAT CHANGES AND NEST SUCCESS

297

u

(N

-§

d «
d

00

CN

'O

00

o

d

lo

VO

d

I

d

o\

^ O

3:2

CN

r- 9

3^

r-

00

ro

C^

d

fo

o

00

o

O

00

sD

00

d

(N

C^

?

lo

00

On

?

lO

r-

d

d (N
d

00

lo

d

I

r--

r-J

VD

lO

?■

IT)

(N

On

;3

ON

d

NO

NO

?

NO

d

r-

d

ON

(N

NO

On

NO

(N

00

d

I

o

VO

d

NO

00

d

I

VO

VO

o

o

o

— H

o

X

lo

On

d

X

d

X)

On

d

1

r-*

00

nA

VO

1

VO

(N

d

r-

d

d

d

d

00

r-

X

ON

r-

d

d

ON

d

1

00

1

VO

1

CN

d

(N

r-

d

VO

d

d

d

ON

VO

On

VO

VO

ON

?

ON

(O

?

On

00

d

1

o

(O

d

ON

CN

d

o

00

d

d

d

•n

“O

*c

0

VO

C3

VO

0

VO

On

o

ON

c

ON

c

c

o

o

>

u

“O

>

u

-o

>

u<

O

o

<U

o

(U

c/5

c/5

C

c

3

c

.—

X)

X

(U

00

o

c/5

O

u

c

c/5

O

o

c

c/5

<u

CJ

c

o

o

. M

o

o

<u

T3

u

3

"O

o

13

T3

vn

C

3

c/5

c

o

X

c

3

c/5

c

o

X

c

o

U

(y5

OJ

g

0

u

C4

<U

E

0

U

298

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

in the southwestern United States, including Oklahoma (Kerpez and
Smith 1987, Stinnet et al. 1987). Sandbars along the Canadian River in
western Oklahoma decreased from 68% to 15% of the total floodplain
between 1954 and 1983; shrub dominated wetlands increased from 13%
to 46%. Schulenberg and Ptacek (1984) noted that encroaching salt cedar
on sandbanks in Kansas reduced nesting habitat available to Least Terns.

Vegetation in Least Tern colonies generally does not cover >20% of
the ground surface (Thompson and Slack 1982, Gochfeld 1983). En-
croaching vegetation and related habitat changes may cause terns to aban-
don a site (Gochfeld 1983, Burger 1984, U.S. Fish and Wildl. Serv. 1990,
Boyd and Rupert 1991, Ziewitz 1992). Saltcedar tolerates the salt levels
found on the alkaline flats at Salt Plains NWR (Ungar 1966) and domi-
nated the shrub rangeland that encroached about 709 ha onto the alkaline
flat and waterways between 1941 and 1989. The rate of spread on the
northeastern alkaline flats was similar from 1941 to 1966 and from 1966
to 1989. Because of the high water table and often saturated soils on the
salt flat, the saltcedar-dominated shrub rangeland will likely continue to
spread away from the west branch of the Salt Fork of the Arkansas River
and onto the alkaline flat. Herbaceous rangeland cover fluctuated slightly
between periods analyzed; however, it does not appear to be encroaching
onto the salt flat habitat.

Our artificial nest experiment indicated that overall nest success was
relatively similar near vegetated or away from vegetated areas. However,
causes of nest failure differed among experimental areas. In support of
our alternate hyotheses, nests adjacent to shrub rangeland had greatest
nest failure from predators in 1994 and areas near streams had highest
nest losses due to flooding. However, areas adjacent to a stream with
shrub rangeland had lower nest losses due to flooding than areas near a
stream with no vegetation. The saltcedar-dominated rangeland may have
encroached into these areas because flooding did not regularly occur there.
In contrast, dense saltcedar stands also can stabilize substrate and alter
fluvial processes (Stinnet et al. 1987). After established, vegetation may
have channeled water away from the salt flats or acted as a barrier to
sheet flooding. Saltcedar has not become fully established on the west
bank of the west branch of the Salt Fork of the Arkansas River; continued
saltcedar encroachment may further alter fluvial processes and have pos-
itive or negative consequences for ground nesting birds. Accelerated salt-
cedar growth along the Rio Grande, Pecos, and Gila rivers, for example,
stabilized channel sediments, reduced stream velocity, accelerated sedi-
mentation, and increased flood risks (Blackburn et al. 1982).

The increase in shrub rangeland over the last 50 years at Salt Plains
NWR likely increased habitat favorable for canids and their predation of

Koenen el cil. • HABITAT CHANGES AND NEST SUCCESS

299

tern and plover nests. Coyotes have been implicated as major nest pred-
ators of Least Tern and Snowy Plover nests at Salt Plains NWR; about 5
to 60% of monitored nests have been lost to predators annually between
1977 and 1994 (Grover and Knopf 1982, Hill 1985, Utych 1993, Koenen
et al. 1996). Artificial nests may not be as vulnerable to predation as real
nests (Angelstram 1986, Andren and Angelstram 1988, Martin 1987, Wil-
lebrand and Marctstrom 1988, Paton 1993); however, there was no sig-
nificant difference in the rate of canid predation of artificial nests and real
nests in our study (Koenen 1995). Because of similarity between artificial
nests and real nests, we contend that management must account for the
different factors that cause nest losses on different areas of the alkaline
flat. Nesting areas near vegetated areas should receive greater protection
from predators, while areas farther from vegetation should be managed
to minimize effects of flooding (Koenen et al. 1996). Long-term man-
agement plans also should consider vegetation control to maintain nesting
areas, reduce impact of predators, and monitor changes in flooding po-
tential.

ACKNOWLEDGMENT

We thank S. L. Gale Koenen, K. A. Shannon, and S. Smith, for checking nest plots and
the G. M. Sutton Avian Research Center and hundreds of Coturnix Quail for donating eggs.
We greatly appreciate comments on an earlier draft of this paper from E Knopf, P. Hendricks,
and two anonymous reviewers. Appreciation also is extended to the refuge personnel, par-
ticularly manager R. Krey, for their immense and gracious support. Eunding for this study
was provided by Region 2 of the U.S. Fish and Wildlife Service, Salt Plains NWR, and the
Oklahoma Cooperative Fish and Wildlife Research Unit (U.S. Nat. Biol. Serv., Okla. Dept.
Wildl. Conserv., Okla. State Univ., and Wildl. Manage. Inst., cooperating).

LITERATURE CITED

Angelstram, P. 1986. Predation on ground-nesting birds’ ne.sts in relation to predator den-
sities and habitat edge. Oikos 47:365-373.

Andren, H. and P. Angelstram. 1988. Elevated predation rates as an edge effect in habitat
islands: experimental evidence. Ecology 69:544-547.

Blackburn, W. H., R. W. Knight, and J. L. Schuster. 1982. Saltcedar influence on sed-
imentation in the Brazos River. J. Soil Water Conserv. 37:286—301.

Boyd, R. L. and J. Rupert. 1991. Habitat management and population ecology studies of
the Least Tern in Kansas. Kansas Dept. Wildl. and Parks Nongame Wildl. Proj. Rep.,
Baldwin City, Kansas.

Brock, J. H. 1994. Tamari.x spp. (Salt Cedar), an invasive exotic woody plant in arid and
semi-arid riparian habitats of western USA. Pp. 27-44 in Ecology and management of
invasive riverside plants (L. C. de Waal, L. E. Childs, P. M. Wade, and J. H. Brock,
eds.). International Centre of Landscape Ecology, John Wiley and Sons, New York,
New York.

Burger, J. 1984. Colony stability in Least Terns. Condor 86:61-67.

Burger, L. D. 1988. Relation between forest and prairie fragmentation and depredation of
artificial ne.sts in Missouri. M.S. thesis. Univ. Missouri, Columbia, Missouri.

300

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Gochfeld, M. 1983. Colony site selection by Least Terns; physical attributes of sites.
Colon. Waterbirds 6:205-213.

Grover, P. B. and L. L. Knopf. 1982. Habitat requirements and breeding success of char-
adriiform birds nesting at Salt Plains National Wildlife Refuge, Oklahoma. J. Lield
Ornithol. 53:139-148.

Harrison, H. H. 1979. A field guide to western bird’s nests. Houghton Mifflin Co. Boston,
Massachusetts.

Hill, L. A. 1985. Breeding ecology of interior Least Terns, Snowy Plovers, and American
Avocets at Salt Plains National Wildlife Refuge, Oklahoma. M.S. thesis, Oklahoma
State Univ., Stillwater, Oklahoma.

Johnson, D. H. 1979. Estimating nest success: the Mayfield Method and an alternative.
Auk 96:65 1-661 .

Johnson, L. B. 1993. Ecological analysis using geographic information systems. Pp. 27-
38 in GIS applications in mammalogy (S. B. McLaren and J. K. Braun, eds.). Oklahoma
Museum Nat. Hist., Norman, Oklahoma.

Kerpez, T a. and N. S. Smith. 1987. Saltcedar control for wildlife habitat improvement
in the southwestern United States. U.S. Lish and Wildl. Serv. Res. Publ. 169.

Koenen, M. T. 1995. Breeding ecology and management of Least Terns, Snowy Plovers,
and American Avocets. M.S. thesis, Oklahoma State Univ., Stillwater, Oklahoma.

^ R. B. Utych, and D. M. Leslie Jr. 1996. Methods used to improve Least Tern

and Snowy Plover nesting success on alkaline flats. J. Lield Ornithol. 67:281-291.

Martin, T. E. 1987. Artificial nest experiments: effects of nest appearance and type of
predator. Condor 89:925—928.

Mayfield, H. L 1975. Suggestions for calculating nest success. Wilson Bull. 87:456-466.

Baton, P. W. 1993. The effect of edge on avian nest success: how strong is the evidence?
Conserv. Biol. 8:17-26.

Purdue, J. R. 1976. Adaptations of the Snowy Plover on the Great Salt Plains, Oklahoma.
Southwest. Nat. 21:347-357.

ScHULENBERG, J. H. AND M. B. Ptacek. 1984. Status of the interior Least Tern in Kansas.
Am. Birds 38:975-981.

Schweitzer, S. H. 1994. Abundance and conservation of endangered interior Least Terns
nesting on salt flat habitat. Ph.D. diss., Oklahoma State Univ., Stillwater, Oklahoma.

SooTER, C. A. 1946. Habits of coyotes in destroying nests and eggs of waterfowl. J. Wildl.
Manage. 10:33-38.

Stinnet, D. P, R. W. Smith, and S. W. Conrady. 1987. Riparian areas of western Okla-
homa. U.S. Pish and Wildlife Service, Ecological Services, Tulsa, Oklahoma.

Thompson, B. C. and R. D. Slack. 1982. Physical aspects of colony selection by Least
Terns on the Texas coast. Colon. Waterbirds 5:161-168.

Ungar, I. A. 1966. Salt tolerance of plants growing in saline areas of Kansas and Okla-
homa. Ecology 47:154-155.

USACERL. 1991. Grass4.0 user’s manunal. U.S. Army Construction Engineering Research
Laboratory, Champaign, Illinois.

U.S. Pish and Wildl. Serv. 1985. Interior population of the Least Tern determined to be
endangered. Led. Reg. 50:21784-21792.

. 1990. Interior population of the Least Tern {Sterna antiHanun): recovery plan.

Twin Cities, Minnesota.

1991. Endangered and threatened wildlife and plants; animal candidate review for

listing as endangered or threatened species. Led. Reg. 56(225):58804-588 10.

1993. Endangered and threatened wildlife and plants; determination of threatened

Koenen el al. • HABITAT CHANGES AND NEST SUCCESS

301

status tor the Delta Smelt and the Pacihc Coast population of the Western Snowy Plover.
Fed. Reg. 58(42): 12864-12874.

Utych, R. B. 1993. Compatibility of selenite crystal digging with breeding ecology of
Least Terns and Snowy Plovers at Salt Plains National Wildlife Refuge. M.S. thesis,
Oklahoma State Univ., Stillwater, Oklahoma.

WiLLEBRAND. T. AND V. Marctstrom. 1988. On the danger of using dummy nests to study
predation. Auk 105:378-379.

Williams, G. E. and E. S. Grover. 1975. Soil survey of Alfalfa County, Oklahoma. U.S.
Soil Conserv. Serv., Stillwater, Oklahoma.

WiLCOVE, D. S., C. H. McLellan, and A. P. Dobson. 1986. Habitat fragmentation in the
temperate zone. Pp. 237-256 in Conservation biology (M. E. Soule, ed.). Sinauer As-
sociates, Inc., Sunderland, Massachusetts.

ZiEWiTZ, J. W., J. G. Sidle, and J. J. Dinan. 1992. Habitat conservation for nesting Least
Terns and Piping Plovers on the Platte River, Nebraska. Prairie Nat. 24:1-20.

Wilson Bull., 108(2), 1996, pp. 302-316

NESTING ECOLOGY OF SCISSOR-TAILED
FLYCATCHERS IN SOUTH TEXAS

Kenneth R. Nolte and Timothy E. Fulbright

Abstract. — We examined nest-site selection and nesting success of the Scissor-tailed
Flycatcher {Tyrannus forficatiis) on the Rob and Bessie Welder Wildlife Foundation Refuge,
San Patricio County, Texas in 1992-1993. Mesquite {Prosopis glandulosa) comprised 22%
of available shrubs; however, Scissor-tailed Flycatchers used shrubs out of proportion to
their availability, placing 91% of their nests in mesquite. Scissor-tailed Flycatcher nests were
placed in taller shrubs with less vertical cover and patchiness, with less total horizontal
cover, and with greater heterogeneity than in random sites. The majority of nests were
oriented to the northwest (18%), north (17%), and northeast (23%). Successful nests were
in shrubs with less vertical patchiness and horizontal cover and with greater vertical cover
(<1 m) and horizontal heterogeneity than unsuccessful nests. Nest-site selection appeared
to be a tradeoff between orienting nests to provide protection from abiotic factors while
minimizing horizontal cover to allow sufficient visibility for nest defense. Received 7 April
1995, accepted 1 Dec. 1995.

Nest-site selection is closely tied to fitness (Martin and Roper 1988)
by influencing losses caused by predators and weather. Tyrant flycatchers
{Tyrannus spp.) breed later (Robins 1970) and have longer nesting cycles
than most other open-nesting passerines. Except for three species of phoe-
bes (Sayornis spp.), the Vermilion Flycatcher {Pyrocephalus rubinus), and
the Acadian Flycatcher {Empidonax virescens), most flycatchers raise
only one brood per year (Bent 1942, Robins 1970, Murphy 1989). Scis-
sor-tailed Flycatchers {T. forficatus) tend to have the largest clutch size
among the tyrannids (Murphy 1989), slower growth rates for nestlings,
and more time spent in the nest (Murphy 1988). Scissor-tails breed
throughout the south-central United States, with the core nesting range
being located in north-central Texas (Fitch 1950). Like other flycatchers,
Scissor-tailed Flycatchers typically place nests in relatively conspicuous
locations, often near the canopy edge (Bent 1942) and at heights ranging
from 1.5 to 12.2 m (Fitch 1950). Scissor-tailed Flycatchers tend to forage
and nest along roadways in open prairies dotted with few trees (Bent
1942). Use of roadways and ditches by mammalian predators may render
Scissor-tailed Flycatcher nests more susceptible to predation as well as
increasing the possibility of predation on adults while foraging. The in-
herently greater diversity of snakes and avian predators in southern lati-

' Caesar Kleberg Wildlife Research Institute. Campus Box 218, Texas A&M Univ.-Kingsville. Kingsville,
Texas 78363. Present address: Buenos Aires National Wildlife Refuge, RO. Box 109, Sasabe, Arizona
8563^

2 Dept' of Agronomy and Resource Sciences, Campus Box I.S6. Texas A&M Univ.-Kingsville, Kingsville,
Texas 78363.

302

Nohe and Fidhhght • SCISSOR-TAILED FLYCATCHER NESTING

303

tudes may impose an additional cause of potential nest failure. These
factors, combined with the intense heat, high winds, heavy rainfall, and
high humidity typical of the summer months in south Texas, are predom-
inant factors influencing the environment and, therefore, nesting success.
This leads to the prediction that Scissor-tailed Flycatchers have evolved
specialized nest placement to mitigate these influences. Scissor-tailed Fly-
catchers should select the largest available shrubs within an area as nest
sites because tall shrubs with greater volume facilitate placement of nests
at locations inaccessible to terrestrial predators, provide protection from
abiotic factors, and allow for nest defense from reptilian and avian pred-
ators and avian nest parasites.

Our objectives were to quantify nesting ecology and to test the hy-
pothesis that nest-site selection and nesting success of the Scissor-tailed
Flycatcher is a positive function of vertical cover and a negative function
of horizontal cover of the nest shrub. Predictions based on this hypothesis
were that (1) successful Scissor-tailed Flycatcher nests are placed in
shrubs with greater vertical cover than shrubs containing unsuccessful
nests; (2) successful nests are placed in shrubs with less horizontal cover
than at unsuccessful nests; and (3) nests are placed within shrubs at lo-
cations inaccessible to mammalian and reptilian predators, i.e., a negative
relationship should exist between relative nest height and relative hori-
zontal distance of nest from main stem to shrub canopy.

STUDY AREA AND METHODS

We conducted this study on the Rob and Bessie Welder Wildlife Foundation Refuge which
encompasses 3156 ha and is 80 km northeast of Corpus Christi in northern San Patricio
County, Texas. The primary habitat associated with the study area was a mesquite-mixed
grass community (Drawe et al. 1978) and was composed of moderately dense stands of
honey mesquite interspersed with den.se clusters of chapairal and interstitial areas of grass.
Other common brush species include huisache {Acacia sniallii), spiny hackbeny {Celtis
pallida), agarito (Berberis trifoliata), lotebush {Ziziphus ohtusifolia), and lime pricklyash
(Zanihoxylum fagara). The soil associated with the mesquite-mixed grass community is
Victoria clay (0-1% slope). Prevailing winds are from the southeast and may reach average
speeds of 56 km/h (Guckian and Garcia 1979). Peak periods of rainfall occur during April,
May, and June.

We found Scissor-tailed Flycatcher nests from May through August 1992 and 1993 by
traversing pastures and by using an extensive network of unimproved roads. Nests were
located by observing Scissor-tailed Flycatchers and by visually inspecting shrubs. We
marked nests, using florescent flagging placed on a shrub or structure adjacent to the nest
shrub, and we revisited at three-day intervals, recording the number of eggs and/or young
at each nest to determine nest fates. An extendible minor-and-pole device was used to view
the contents of nests. Evidence of nest success included observations of young fledging
from a nest or the presence of young near a nest. Nests were considered successful if >1
nestling fledged. Failure was assumed when nest contents disappeared before the anticipated
fledging date or when the nest was damtigcd or blown out of the shrub. Not all nests were

304

THE WILSON BULLETIN • Vo/. 108, No. 2, June 1996

found before the onset of egg-laying or incubation; therefore, success was quantified by
using Mayfield’s method (Mayfield 1961, 1975) to compensate for exposure. We tested for
a difference in nest success between years with an F-test (Johnson 1979).

Nest sites were revisited to conduct vegetation measurements following fledging of young
or upon nest failure. Since some nests were lost because of abiotic factors or predation or
because some nests were inaccessible, only 60 nest sites were measured. Vegetation mea-
surements were also taken at randomly selected sites (N = 30 each year) to represent
available nest sites. Available shrubs were selected by pacing 100 m in a random direction
from each Scissor-tailed Llycatcher nest site and then choosing the shrub nearest the end of
the 100-m distance. We recorded shrub species and determined proportions of each species
at used and at available sites for preference/avoidance analysis. Lrequency of nest placement
among available shrubs was compared using chi-square analysis. If a chi-square test resulted
in rejecting the null hypothesis that a species was used in proportion to availability, a
Bonferroni z-statistic was used (Neu et al. 1974) to estimate whether a Scissor-tailed Lly-
catcher selected or avoided that shrub species.

Variables were grouped into two levels of resolution: nest placement within the shrub
and vertical and horizontal structure of the nest shrub to six m from the nest (Table 1). A
6-m-radius from the nest was selected to describe horizontal structure, since most nests were
in uniform habitats composed of mesquite trees with canopies <10 m in diameter. We
quantified horizontal and vertical structure (cover) and patterning (patchiness) of the vege-
tation using a method similar to the “bird centered view” described by (Weins and Roten-
berry 1985). We quantified vertical and horizontal cover using a 2-cm diameter rod marked
at 0.1 m increments. We recorded the number of 0.1 m increments touching vegetation out
of the total possible number of increments within each of three height (vertical) or distance
(horizontal) classes (0-1 m, 1-3 m, and 3-6 m). Lor example, within the 1-3 m class (a
distance of 2 m) there were 20 possible increments. If vegetation touched 10 of the 20
increments, cover was 50%. Vertical cover was measured by extending the rod from the
ground to the canopy projecting through the nest. Horizontal structure was quantified by
extending the rod parallel to the ground, at nest height, in each of the four cardinal directions
from the nest. Cover was calculated from the mean of the four cover estimates within each
of the three distance classes. If the nest was too high to be reached from the ground or it
could not be reached by climbing the tree, a ladder was used to measure horizontal cover.
At the nest placement level, we determined average three-dimensional cover surrounding
the nest. The structure rod was oriented vertically to nest height and structure measurements
were recorded at 0.5-m increments moving away from the nest in each of the four cardinal
directions. Mean percent cover was determined within each of three distance classes (0-1
m, 1-3 m, and 3-6 m) horizontally from the nest and extending from the ground to the
outer canopy of the shrub. The coefficient of variation (CV) for structure variables repre-
sented an index of the patchiness of the measured variable. We also calculated a horizontal
heterogeneity index (HHI) (Rotenberry and Weins 1980) using horizontal structure data.

We compared means and variances for statistical differences between nest sites and ran-
dom sites and between successful and unsuccessful nests to explore the relationship between
nest success and nest-site selection, (Ratti et al. 1984). Homogeneity of variance tests can
indicate aspects of nest-site .selection not readily detectable by comparing sample means
alone. We considered nest-site .selection to have occuned when ( 1) Scissor-tailed Llycatchers
selected nest-site characteristics with different means but similar variance as random sites,
(2) nest sites had similar means but less variance than at random sites, and (3) nest sites
had different means and less variance than at random sites (Lig. 1). In the second situation,
traditional compari.sons of sample means would not have detected habitat selection, whereas.

Nolte ami Fulhhght • SCISSOR-TAILED FLYCATCHER NESTING

305

Table 1

Measured and Calculated Vegetation Variables Used to Quantify Nest Sites and

Available Random Sites

Variable

Description

Shrub characteristics^

TOTHT

VI

NSDIAM

NSVOL

VCOVOI

VCOV13

VCOV36

TVCOV

CVVCOV

HCOVOl

HCOV13

HCOV36

THCOV

CVHCOV

HHI

AVEDNW

CVDNW

Placement characteristics

Total height of the shrub (m)

Shrub vigor (1 = <25%, 2 = 26 < 50%, 3 = 51 < 75%, 4 =
76 100%) based on the percent of living material

Average diameter of the shrub at nest height (m)

Shrub volume (pil3) (NSDIAM*NSDIAM*TOTHT/2) (mO
Vertical cover projected through nest from 0-1 m (%)

Vertical cover projected through nest from 1-3 m (%)

Vertical cover projected through nest from 3-6 m (%)

Total vertical cover 0-6 m (%)

Coefficient of variation (CV) for the three vertical cover vari-
ables

Average horizontal cover 0-1 m from nest (%)

Average horizontal cover 1-3 m from nest (%)

Average horizontal cover 3-6 m from nest (%)

Total horizontal cover (%)

Coefficient of variation (CV) for the three horizontal cover
variables

Horizontal heterogeneity

Average distance to the nearest shrub (m)

Coefficient of variation (CV) of distance to the nearest shrub
in each of the four compass directions

NESTHT

RELHT

TOPDIST

ORIENT

DTRK

TOTTRK

RELDIST

NSTANGL

COVOl

COV13

COV36

TCOV

CVCOV

Height of nest (m)

Ratio between nest height and total height of shrub

Vertical distance from nest to canopy of shrub (m)

Orientation of the nest within the shrub (degrees)

Horizontal distance from the main stem to the nest (m)

Distance from main stem to the canopy through nest (m)
Relative horizontal distance of nest between main stem and
shrub canopy

Angle of main branch supporting nest (degrees)

Three dimensional cover from 0-1 m around the nest (%)

Three dimensional cover from 1-3 m around the nest (%)

Three dimensional cover from 3-6 m around the nest (%)

Total cover in a 6-m radius cube around the nest (%)

CV for the three cover measurements

“All shrub variables were used to compare flycatcher nests and random sites.

306

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

Mean value

Variance not different different

not different

1.

no selection

2.

selection

less at nest site

3.

selection

4.

selection

less at random site

5.

no selection

6.

no selection

Fig. 1 . Possible combinations of means and variances at nest sites and random sites that
indicate habitat selection or random choice.

an instance when sample means were different but variance was less at random sites would
not indicate habitat selection.

Treatment means (nest vs random, successful vs unsuccessful) for nest-site variables were
compared with a completely random design and a two-way factorial treatment structure with
the general linear model (GLM) procedure (SAS 1988). When an interaction occuiTcd be-
tween years, contrasts were used to compare treatments by year. Percentage or proportional
data were arc-sine transformed before statistical analyses (Sokol and Rohlf 1973). Analyses
were performed with SAS (Statistical Analysis Institute 1988) and conclusions were based
on a = 0.05 unless otherwise indicated.

RESULTS

Reproductive success. — Forty-eight nests were monitored during 1992-
1993, resulting in 789 nest-days of observations (Table 2). The first nest-

Nolle ami Fulbri^ht • SCISSOR-TAILED FLYCATCHER NESTING

307

Table 2

Estimates of Nest Success, Confidence Intervals, and Sources of Nest Failure for

SCISSOR-TAILED FLYCATCHERS BREEDING

1992 1993

Nest days

186

603

Number of eggs (.f ± SD) (N)

4.4 ± 0.5 (I2)A

4.5 ± 0.5 (19)A

Number of young (x ± SD) (N)

3.2 ± 1.6 (7)A

3.0 ± 1.0(10)A

Daily mortality rate

5.9%

2.2%

Mayfield estimate (N)

94.1% (17)

97.8% (31)

95% confidence interval

90.6-97.6%

96.6-99.0%

Probability of survival to fledging

15. 6% A

50.7%B

95% confidence interval

4.9-26.3%

34.8-66.6%

Sources of nest failure

Weather (N)

45.5% (5)

7.7% (1)

Predation (N)

36.4% (4)

15.4% (2) ■

Abandonment (N)

18.1% (2)

7.7% (1)

Unknown (N)

0.0% (0)

69.2% (9)

^ Means followed by the same letter are not significantly different {P > 0.05).

ing activity was recorded on 5 May 1992 and 3 May 1993. Egg laying
began on 28 May 1992 and 25 May 1993. Mean fledging dates respec-
tively were 1 July 1992 (N = 7) and 3 July 1993 (N = 10). Based on
complete nests, the number of eggs/nest and number of fledglings/nest
did not differ {P > 0.05) between years (Table 2). Nineteen nests were
destroyed during storms or were removed (used as nesting material) by
other birds before vegetation could be quantified. These and all other nests
that could not be visually inspected because they were inaccessible were
excluded. Of the remaining 48 nests (17 in 1992, 31 in 1993) used to
calculate success, 31 were found before initiation of egg-laying. Proba-
bility of nesting success was greater {P = 0.03) during 1993 than 1992.
Nest success was 39% when years were pooled. There was a year X nest
success interaction for the number of eggs/nest {P — 0.03) and for the
number of young/nest (P = 0.0001) between successful and unsuccessful
nests. Both the number of eggs/nest {P = 0.09) and the number of young/
nest were similar (P = 0.06) at successful and unsuccessful nests during

1992. Number of eggs/nest and young/nest were greater {P — 0.01, P =
0.0001, respectively) at successful nests than at unsuccessful nests during

1993.

Rainfall during 1992 (128.4 cm) and 1993 (102.8 cm) was above the
annual average of 88.9 cm for the Welder Refuge and therefore may not
reflect average conditions. Abiotic factors accounted for the largest per-
centage (46%) of nest failures in 1992. All five nests lost to abiotic factors

308

THE WILSON BULLETIN • VoL 108, No. 2, June 1996

were found on the ground near the shrub following storms. Predation
accounted for 36% of nest failures, and the remaining 18% of failures
were because of abandonment. Predation was assumed when nest contents
disappeared under suspicious circumstances, i.e., eggs or young disap-
peared between consecutive visits or when contents disappeared following
observations of predators near the nest site. In 1993, the majority (69%)
of nest failures were because of unknown causes.

Frequency of shrub selection. — Scissor-tailed Flycatcher nests (N = 60)
were placed nonrandomly among the available shrub species (x^ = 170.46,
df = 4, P < 0.0001). Frequency of available shrubs in the habitat was
mesquite (22%), huisache (20%), spiny hackberry (18%), lime pricklyash
(20%), brazil (Ziziphus obtusifolia) (8%), agarito (5%), Texas persimmon
(Diospyros texana) (3%), sugar hackberry {Celtis laevigata) (2%), and
wolfberry (Lycium berlandieri) (2%). Scissor-tailed Flycatchers selected
mesquite and avoided all other shrubs during 1992 and 1993. Ninety-one
percent of the nests (N = 55) were placed in mesquite. Nests were also
placed in huisache (N = 1), lime pricklyash (N = 2), sugar hackberry (N
= 1), and under a transformer on a telephone pole (N = 1).

Nest-site characteristics. — Nests were placed 2.8 ± 0.8 m high and 1.9
± 1.0 m (x ± 1 SD; N = 60) from the main stem of the shrub. Relative
height of the nest within the shrub and relative horizontal distance from
main stem to the shrub canopy were 0.60 ±0.11 and 0.49 ± 0.18, re-
spectively. There was no correlation (r = -0.15, P = 0.26) between
relative nest height and relative horizontal distance. Average height, di-
ameter, and volume of nest shrubs were 4.7 ± 0.9 m, 7.6 ± 2.7 m, and
172.0 ± 133.0 m-\ respectively. Mean nest orientation was to the south-
east; however, a majority (58%) of nests were oriented northwest (18%),
north (17%), and northeast (23%) (Fig. 2).

Flycatcher v.v random comparisons. — Thirteen of the 17 vegetation
characteristic means differed {P < 0.05) between flycatcher nests and
random sites (Table 3). The year X treatment (used or random) interaction
was significant {P = 0.0001) for total horizontal cover. Total horizontal
cover was greater (P < 0.05) at random sites than at nest-sites during
1992 but was not different {P > 0.05) during 1993. Scissor-tailed Fly-
catchers chose shrubs that were taller {P < 0.001), greater in diameter (P
< 0.001) and volume (P < 0.001), and had less (P < 0.001) variation in
vertical cover than random shrubs. They also chose shrubs with less ver-
tical cover from 0-1 m (P < 0.001), from 1-3 m (P = 0.001), and from
0-6 m (total vertical cover) (P < 0.001). However, there was more (P <

0. 001) vertical cover from 3-6 m at their nests than at random shrubs.
Scissor-tailed Flycatchers selected sites that were more (P = 0.002) open,

1. e., a greater distance to the nearest shrub and shrubs that were patchier

Nolte and Fidhright • SCISSOR-TAILED FLYCATCHER NESTING

309

s

Fig. 2. Percent of Scissor-tailed Flycatcher nests (N = 60) oriented within each of the
eight cardinal compass directions on the Rob and Bessie Welder Wildlife Refuge, 1992-
1993.

{P < 0.001), i.e., greater variation in the amount of horizontal cover, than
random sites. Nests were also placed in shrubs with greater {P < 0.001)
horizontal heterogeneity than random sites.

Tests for homogeneity of variance indicated differences {P < 0.05) for
seven characteristics (Table 3). Scissor-tailed Flycatchers selected shrubs
with greater variation in shrub diameter {P < 0.001) and volume {P <
0.001) than random shrubs. Random shrubs had greater {P < 0.001) vari-
ance for vertical cover from 0-1 m, from 1-3 m and for total vertical
cover, with less iP < 0.001) variance for vertical cover from 3-6 m than
shrubs selected by flycatchers. The variance of average distance to the
nearest woody vegetation was greater {P < 0.001) at nests than at random
sites.

Successful unsuccessful nest comparisons. — There was no difference
(P > 0.05) between relative height or relative horizontal distance at sue-

310

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 3

Comparison of Vegetation Characteristics at Scissor-tailed Elycatcher Nests (N =

60) vs Random Sites (N = 60)

Ho: Ho:

Flycatcher Random equal equal

means variance

Variable T SD x SD P-value P-value

Shrub characteristics

TOTHT*

4.7

0.9

2.0

0.9

0.000

0.841

VI

3.3

0.9

3.6

0.7

0.051

0.060

NSDIAM

7.6

2.7

2.0

1.5

0.000

0.000

NSVOL

172.0

133.0

10.3

27.0

0.000

0.000

VCOVOl*

0.23

0.15

0.70

0.26

0.000

0.000

VCOV13*

0.19

0.12

0.39

0.30

0.001

0.000

VCOV36

0.33

0.17

0.04

0.10

0.000

0.000

TVCOV*

0.25

0.08

0.39

0.15

0.000

0.000

CVVCOV*

64.4

33.4

108.2

40.0

0.000

0.170

HCOVOl

0.53

0.27

0.59

0.21

0.199

0.083

HCOV13

0.37

0.21

0.34

0.27

0.345

0.075

HCOV36*

0.20

0.25

0.26

0.23

0.037

0.466

THCOV*‘>

0.32

0.20

0.41

0.19

0.000

0.674

CVHCOV*

73.5

43.0

65.4

38.0

0.000

0.351

HHI*

1.7

0.6

1.0

0.5

0.000

0.239

AVEDNW

13.8

24.7

3.8

2.4

0.002

0.000

CVDNW

40.0

25.2

47.1

26.9

0.129

0.624

“Significant (P < 0.05) interaction between year and treatment (used and random). THCOV at used and random sites
differed (P < 0.0001) for 1992 and was similar (P = 0.6545) for 1993.

* Habitat selection.

cessful and unsuccessful nests. Analysis of cover measurements at Scis-
sor-tailed Flycatcher nests indicated vertical cover from 0-1 m was great-
er (P = 0.036) at successful nests (Table 4). Successful nests were also
placed in shrubs with less (P = 0.013) patchiness of vertical cover than
unsuccessful nests. The year X treatment (nest success) interaction was
significant (P < 0.05) for horizontal cover from 1-3 m and from 3-6 m
and for total horizontal cover. Contrasts indicated that three cover attri-
butes at successful and unsuccessful nests differed for 1993 (P < 0.05)
but not for 1992 (P > 0.05).

Tests for homogeneity of variance indicated differences for four char-
acteristics (Table 4). Successful nests were placed in shrubs with less
variation in vertical cover from 0-1 m (P = 0.027), CV of mean distance
to the nearest shrub in each of the four cardinal compass directions (P =
0.005), and horizontal heterogeneity (P = 0.005). Variance for average
distance to the nearest shrub (P = 0.009) was greater at successful nests.

Nolle and Fidhright • SCISSOR-TAILED FLYCATCHER NESTING

311

Table 4

Comparison of Vegetation Characteristics at Successful (N = 17) and
Unsuccessful (N = 39) Scissor-tailed Flycatcher Nests on the Rob and Bessie
Welder Wildlife Refuge, San Patricio County, Texas, 1992-1993

Ho: Ho:

Successful Unsuccessful equal equal

means variance

Variable T SD T SD P-value P-value

Shrub characteristics

TOTHT

VI

NSDIAM

NSVOL

VCOVOl* *

VCOV13

VCOV36

TVCOV

CVVCOV*

HCOVOl

Hcovn*^'

HCOV36*“

THCOV^^"

CVHCOV

HHI*

AVEDNW

CVDNW*

Placement characteristics
NESTHT
RELHT
TOPDIST
ORIENT
DTRK
TOTTRK
RELTRK
NSTANGL
COVOl
COV13
COV36
TCOV
CVCOV

4.6

0.8

4.7

3.4

1.0

3.2

7.7

2.7

7.7

173.0

128.0

175.0

0.29

0.12

0.21

0.20

0.12

0.18

0.31

0.15

0.33

0.27

0.06

0.24

48.8

25.7

71.6

0.45

0.25

0.56

0.33

0.22

0.39

0.15

0.19

0.24

0.26

0.16

0.36

73.6

49.0

72.5

1.6

0.3

1.8

17.9

34.6

12.7

37.4

14.8

40.6

2.8

0.9

2.8

0.60

0.13

0.60

1.8

0.6

1.9

171.0

68.0

1 13.0

1.7

1.1

2.1

3.8

1.4

3.9

0.46

0.21

0.50

37

27

31

0.27

0.07

0.23

0.23

0.06

0.23

0.20

0.13

0.19

0.24

0.06

0.23

32.7

35.7

31.8

0.9

0.220

0.668

0.9

0.801

0.648

2.7

0.426

0.939

137.0

0.368

0.797

0.16

0.036

0.027

0.11

0.342

0.953

0.18

0.172

0.131

0.09

0.218

0.231

32.6

0.047

0.304

0.27

0.200

0.811

0.21

0.003

0.686

0.28

0.000

0.153

0.22

0.000

0.232

42.5

0.812

0.453

0.6

0.285

0.005

20.8

0.401

0.009

29.3

0.776

0.005

0.8

0.176

0.643

0.11

0.409

0.326

0.5

0.912

0.373

68.0

0.066

0.867

1.0

0.466

0.900

1.4

0.426

0.939

0.18

0.775

0.243

22

0.483

0.324

0.08

0.066

0.326

0.07

0.140

0.672

0.1 1

0.208

0.394

0.06

0.279

0.685

25.1

0.463

0.077

•Significant (P < 0.05) interaction between year and treatment (successful and unsuccessful). HCOV13, HCOV36. and
THCOV at successful and unsuccessful nests differed (P < 0.05) during 1993 but were similar (P > 0.05) during 1992.

* Habitat selection.

312

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

DISCUSSION

The nesting success rate (39%) for Scissor-tailed Flycatchers on the
Welder Refuge was less than that reported for other flycatchers except
Eastern Kingbirds (25.6%) in Kansas (Murphy 1986). Scissor-tailed Fly-
catchers had the highest success rate 81% (N = 16) (Murphy 1983) and
clutch size 4.69 (N = 16) (Murphy 1988) of all tyrannids reported. Mean
clutch size on the Welder Refuge was similar to that reported by Murphy
(1988), suggesting Fitch’s (1950) estimate of nest success was low, pos-
sibly because of the inclusion of incomplete nests. Similarities between
clutch sizes indicate nests on the Welder Refuge suffered a greater mor-
tality rate during the post-laying period. Above-average rainfall during
the 1992 and 1993 breeding seasons was partially responsible for lower
nesting success because high winds and heavy rains dislodged nests from
shrubs. Murphy (1986) reported losses caused by weather were mostly
from wind blowing nests from trees. Abiotic factors were also believed
to have accounted for many of the nest failures from unknown causes
during 1993. However, since most of the these nests could not be found,
or the nest contents had disappeared, the exact cause of failure remains
uncertain.

Nest success may be affected at two spatial scales: habitat in the im-
mediate vicinity of the nest (shrub characteristics) and habitat surrounding
the nest (characteristics of the nest patch) (MacKenzie and Sealy 1981,
Martin and Roper 1988). This study focused on nesting success and nest-
site selection at the nest shrub scale. We believe this degree of resolution
was sufficient to describe nest-site selection by Scissor-tailed Flycatchers.
Nests were placed, on average, 2.8 m high in shrubs 4.7 m tall. Mean
height of available shrubs was only 2 m and therefore would not provide
much horizontal obstruction for their nests placed in adjacent shrubs.
Because of the nest height, conspicuous nest placement, and orientation
away from prevailing winds, it appears that Scissor-tailed Flycatchers se-
lect attributes related to nest shrubs rather than the surrounding habitat.

Site selection for Scissor-tailed Flycatchers may have been a function
of selecting characteristics that allowed adults to monitor and defend the
nest site since horizontal cover was less at successful nests. Ricklefs
(1977) found that a strong correlation existed between nest conspicuous-
ness and intensity of nest defense in tropical passerines. Scissor-tailed
Flycatcher nests were generally found >100 m apart and were located in
open stands of mesquite on the Welder Refuge. Fitch (1950) noted that
nests were never found within 76 y of each other. Spacing of Scissor-
tailed Flycatcher nests is partially a function of the open habitat selected
and partially because of the size and aggressive defense of individual

Nolte and Fidbhght • SCISSOR-TAILED FLYCATCHER NESTING

313

territories. Placing nests in open shrubs (less vertical and horizontal cover)
would allow for sufhcient air space in which the birds can maneuver to
attack intruders. However, less total cover may also increase the risk of
nest failure from abiotic factors.

Nest orientation relative to the center of the shrub should influence
losses of nests because of abiotic factors, including prevailing southeast
winds and numerous thunderstorms originating in the Gulf of Mexico
during the nesting season. Since only 25% of the nests were oriented to
the east, southeast, or south (toward prevailing winds), the birds appeared
to place nests so as to minimize the effects of abiotic factors. Placement
of Scissor-tailed Flycatcher nests within shrubs appeared to minimize hor-
izontal cover while favorable nest orientation may have provided some
respite from the wind, rain, and sun, thus partially mitigating the effects
of mortality from overexposure to the sun. Murphy (1985) described nest-
ling deaths from overexposure to sun in Eastern Kingbirds, as a source
of mortality.

Murphy (1983) noted that predation was the driving force behind nest-
site selection in Eastern Kingbirds, as nests placed extremely low or ex-
tremely high within trees had the lowest probability of fledging young.
He added that maximum success occurred at relative nest heights and
relative horizontal distances from the tree center to the shrub canopy edge
of about 0.5. Our results differ somewhat from the above. Although nests
were placed at relative heights and horizontal distances of 0.6 and 0.5,
respectively, there was no difference between successful and unsuccessful
nests for either variable.

Nest concealment was greater at low predation nests than at high pre-
dation nests for woodland birds including the Hermit Thrush {Hylocichla
guttata). Prairie Warbler (Dendroica discolor). Mourning Dove (Zenaida
macroura), and Eastern Kingbird (Murphy 1983, Westmoreland and Best
1985, Martin and Roper 1988). Great-tailed Grackles (Cassidix mexican-
us) appeared to be the primary avian predator of Scissor-tailed Flycatcher
nests. Large groups of grackles were observed harassing Scissor-tailed
Flycatchers at their nest sites on numerous occasions. Raccoons (Procyon
lotor), and opossums (Didelphis virginianus) were the only common
mammalian predators present on the study site capable of depredating the
flycatcher nests. We documented no incidence of mammalian predation
on nests during the two years of this study; however, these mammalian
predators are nocturnal and direct observations would be unlikely. Vertical
cover <1 m was greater at successful nests than at unsuccessful nests.
Greater ground cover may inhibit some terrestrial predators from locating
nests, although reptilian predators may actually benefit.

Nineteen nests were abandoned before egg-laying for unknown reasons.

314

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Frequent visitation to nests has been documented to affect nesting success.
However, daily intrusions, including removal of young for measurements,
did not adversely affect nesting success of Scissor-tailed Flycatchers
(Fitch 1950). Because nests of unknown fate were often found intact
although empty, snake predation likely contributed to nest failure. Had
mammalian and avian predators been responsible, it is likely that shell
fragments or other signs would have been left at the nest site. Several
species of snakes known to prey on eggs and nestlings were present
throughout the study area. Kingsnakes (Lampropeltis spp.), yellow-bellied
racers {Coluber constrictor flaviventris), and western coachwhips {Mas-
ticophis flagellum testaceus) were observed at nest sites on several oc-
casions and were observed or suspected to be the source of nest failure
on numerous other occasions for Dickcissels {Spiza americana). Mourn-
ing Doves, Northern Mockingbirds {Mimus polyglottos), and Northern
Cardinals (Cardinalis cardinalis) on the Welder Refuge (Nolte, pers. ob-
serv.).

Strength of attachment of the nest to the shrub may be an important
component of nest success. Many nests failed before hatching or fledging
because they were dislodged from the nest shrub following storms. Con-
spicuous placement of nests may render them more vulnerable to unpre-
dictable, heavy rainfall and wind events than those of most other passer-
ines which nest in short, dense shrubs in south Texas. Additional research
should be conducted to determine if the firmness of nest attachment is
related to nesting success.

On the Welder Refuge, randomly available shrubs appeared to be of
insufficient size to accommodate placement of flycatcher nests. Mesquite
seemed to afford the best compromise by providing optimal cover and
by allowing nests to be placed at locations inaccessible to terrestrial pred-
ators. Previous investigators have reported that Scissor-tailed Flycatchers
nest in any species of tree that is isolated and is open-foliaged (Bent 1942,
Fitch 1950). The structural attributes provided by mesquite may be only
partially responsible for nest-site selection by Scissor-tailed Flycatchers.
They selected for total height of nest shrubs; therefore, size relative to
other available shrubs may also have a role in shrub selection on the
Welder Refuge.

Based on the results of this study, we accept the hypothesis that Scissor-
tailed Flycatchers select nest sites based on horizontal structure of the
shrub. Our results indicate that nest-site selection in Scissor-tailed Fly-
catchers appears to be a trade-off between providing air space around the
nest (less horizontal cover) for defense from predators and at the expense
of increasing exposure to abiotic influences such as wind, rain, and solar
radiation. Results did not, however, support the predictions that nest-site

Nolle and Fidhright • SCISSOR-TAILED FLYCATCHER NESTING

315

selection was a function of vertical cover or that a negative relationship
existed between relative nest height and relative horizontal distance within
the nest shrub.

About 400,000 ha of rangelands in Texas are annually treated with
herbicides, often with the goal of decreasing the density of mesquite.
Brush management practices, to one degree or another, result in setting
back succession. We found evidence that Scissor-tailed Flycatchers show
shrub-specific site tenacity. In 1993, six nests were placed in shrubs that
contained a Scissor-tailed Flycatcher nest in 1992. Subsequent observa-
tions in 1994 indicated that 25 nests were in shrubs containing Scissor-
tailed Flycatcher nests in at least one of the two previous years. If areas
used as nest-sites are subsequently altered via some brush management
practice, returning pairs of Scissor-tailed Flycatchers may attempt to re-
nest in dead shrubs. We documented eight occasions when nests were
placed in shrubs that were dead before initiation of nesting activity, and
in all eight cases the nests failed. The widespread use of such practices
could decrease the available nesting habitat for Scissor-tailed Flycatchers.
Our results indicate this will undoubtedly result in a greater rate of nest
mortality. Other passerine species, including cavity nesters or those that
require larger shrubs for nest placement and support, could be equally
affected. Managers should consider leaving strips or patches of untreated
brush when large acreages of rangeland are managed. Another manage-
ment strategy could be to leave dispersed mature mesquite in an area
following treatment. Brush control on sites without mesquite should allow
for the preservation of individuals or loose clumps of the largest trees
available.

ACKNOWLEDGMENTS

We are grateful to James Teer and the Rob and Bessie Welder Wildlife Foundation for
support during this study. We thank P. A. Moody for assisting with field work and data
management. We also thank Fred Guthery, Eric Hellgren, Ralph Bingham, Michael T. Mur-
phy, Richard Conner, and several anonymous reviewers for valuable comments on this
manuscript. This is Rob and Bessie Welder Wildlife Foundation contribution 447.

LITERATURE CITED

Bent, A. C. 1942. Life histories of North American flycatchers, larks, swallows, and their
allies. U.S. Natl. Mus. Bull. 179.

Drawe, D. L., a. D. Chamrad, and T. W. Box. 1978. Plant communities of the Welder
Wildlife Refuge, 2nd ed. Cont. No. 5, revised. Welder Wildlife Foundation, Sinton,
Texas.

Fitch, R. W, Jr. 1950. Life history and ecology of the Scissor-tailed Flycatcher, Muscivoro
forficata. Auk 67:144-168.

Guckian, W. J. and R. N. Garcia. 1979. Soil survey of San Patricio and Aransas Counties,
Texas. U.S.D.A., Soil Conserv. Serv.

316

THE WILSON BULLETIN • Vol. 108, No. 2. June 1996

Johnson, D. H. 1979. Estimating nest success: the Mayfield method and an alternative.
Auk 96:651-661.

Mackenzie, D. I. and S. G. Sealy. 1981. Nest site selection in Eastern and Western
Kingbirds: a multivariate approach. Condor 83:310-321.

Martin, T. E. and J. J. Roper. 1988. Nest predation and nest-site selection of a western
population of the Hermit Thrush. Condor 90:51—57.

Mayfield, H. 1961. Nesting success calculated from exposure. Wilson Bull. 73:255—261.

. 1975. Suggestions for calculating nest success. Wilson Bull. 87:456-466.

Murphy, M. T. 1983. Nest success and nesting habits of Eastern Kingbirds and other
flycatchers. Condor 85:208—219.

. 1985. Nestling Eastern Kingbird growth: effects of initial size and ambient tem-
perature. Ecology 66:162—170.

. 1986. Temporal components of reproductive variability in Eastern Kingbirds {Tyr-

annus tyrannus). Ecology 67:1483-1492.

. 1988. Comparative reproductive biology of kingbirds (Tyrannus spp.) in eastern

Kansas. Wilson Bull. 100:357-376.

. 1989. Life history variability in North American breeding tyrant flycatchers: phy-
togeny, size or ecology? Oikos 54:3-14.

Neu, C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-
availability data. J. Wildl. Manage. 38:541-545.

Ratti, j. T, D. L. Mackey, and J. R. Alldredge. 1984. Analysis of spruce grouse habitat
in north-central Washington. J. Wildl. Manage. 48:1188-1196.

Ricklefs, R. E. 1977. Reactions of some Panamanian birds to human intrusion at the nest.
Condor 79:376-379.

Robins, J. D. 1970. The relationship of food supply to the timing of breeding in aerial
foragers. Kansas Ornithol. Soc. Bull. 21:9-15.

Rotenberry, j. T. and J. A. Wiens. 1980. Habitat structure, patchiness, and avian com-
munities in North American steppe vegetation: a multivariate analysis. Ecology 6:1228-
1250.

Statistical Analysis Institute. 1988. SAS/STAT user’s guide. SAS Inst. Cary, North
Carolina.

Sokol, R. R. and R. J. Rohlf. 1973. Introduction to biostatistics. W.H. Ereeman, San
Francisco.

Weins, j. a. and j. T. Rotenberry. 1985. Response of breeding pas.serine birds to range-
land alteration in a North American shrubsteppe locality. J. Appl. Ecol. 22:655—668.

Westmoreland, D. and L. B. Best. 1985. The effect of disturbance on Mourning Dove
nesting success. Auk 102:774—780.

Wilson Bull., 108(2), 1996, pp. 317-334

BREEDING BIOLOGY OF THE BROWN NODDY ON
TERN ISLAND, HAWAII

Jennifer L. Megyesi' and Curtice R. Griffin^

Abstract. — We observed Brown Noddy (Anoiis stolidiis pileatus) breeding phenology
and population trends on Tern Island, French Frigate Shoals, Hawaii, from 1982 to 1992.
Peaks of laying ranged from the first week in January to the first week in November;
however, most laying occurred between March and September each year. Incubation length
was 34.8 days (N = 19, SD = 0.6, range = 29-37 days). There were no differences in
breeding pairs between the measurements of the first egg laid and successive eggs laid within
a season. The proportion of light- and dark-colored chicks was 26% and 74%, respectively
(N = 221) and differed from other Brown Noddy colonies studied in Atlantic and Pacific
oceans. The length of time between clutches depended on whether the previous outcome
was a failed clutch or a successfully fledged chick. Hatching, fledging, and reproductive
success were significantly different between years. The subspecies (A. s. pileatus) differs in
many aspects of its breeding biology from other colonies in the Atlantic and Pacific oceans,
in regard to year-round occurrence at the colony, frequent renesting attempts, large egg size,
proportion of light and dark colored chicks, and low reproductive success caused by in-
clement weather and predation by Great Frigatebirds (Fregata minor). Received 31 Mar.,
1995, accepted 5 Dec. 1995.

The Brown Noddy (Anous stolidus) is the largest and most widely
distributed of the tropical and subtropical tern species (Cramp 1985). The
breeding biology of the nominate subspecies A. s. stolidus has been stud-
ied extensively in its Atlantic range (Dorward and Ashmole 1963, Rob-
ertson 1964, Morris and Chardine 1992); however, few studies exist for
A. s. pileatus which ranges over most of the Pacific and Indian oceans
(Cramp 1985). Brown (1973, 1977) published most of the information on
Brown Noddies breeding in the Hawaiian Islands; however, his obser-
vations spanned only two breeding seasons. Here, we present observations
of the breeding biology of this species in the Northwestern Hawaiian
Islands, the northern-most part of its range, and compare it to previous
reports from other Hawaiian Islands and with observations of the biology
of the subspecies A. s. stolidus in the Atlantic Ocean.

STUDY AREA AND METHODS

French Frigate Shoals is a crescent-shaped atoll situated approximately midway in the
Hawaiian Archipelago (23°45'N, 166°17'W) and is part of the Northwe.stern Hawaiian Is-
lands National Wildlife Refuge administered by the U.S. Fish and Wildlife Service
(USFWS). The atoll contains 10 well-established islands and as many as nine sandy islets
that are seasonally awash (Fig. 1). Tern Island, located near the northwestern tip of the atoll,

' RO. Box 741, Truro, Massachusetts 02666.

^ Dept. Forestry and Wildlife Management, Univ. of Massachusetts, Amherst, Massachusetts OUKH.

317

318

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

HAWAIIAN ISLANDS

Gf^'

ME'

v\^'

166 20' W

FRENCH FRIGATE SHOALS

166'l0' W

23' 50' N

TRIG

. ■ - TERN •

-

..SHARK

•

WHALESKATE

23'50'N

ROUND
• • MULLET

LA PEROUSE

EAST /

\

•

23 40'N

. kciN

'■ %LITTLE GIN

r

DISAPPEARING

•-

23 40'N

/

166 20' W 166 10' W

Fig. 1 . French Frigate Shoals, with insert of the Hawaiian Islands.

is the largest island and the only human-made island at French Frigate Shoals. Originally
4.4 ha in size, the island is now approximately 15.0 ha, measuring 945 m in length and
99.1 m wide (Amerson 1971, USFWS unpubl. data). The island was expanded to accom-
modate military operations during the 1940s. An active, crushed-coral runway, approxi-
mately 76.2 m wide, extends the length of the island. Sixteen species of seabirds nest on
Tern Island (Amerson 1971, USFWS). No mammalian land predators occur within the atoll.
At French Frigate Shoals, Brown Noddies nest on Tern, La Perouse, Whaleskate, East and
Little Gin islands. Tern Island has the largest nesting colony, with approximately 1700-

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

319

Fig. 2. Location of Brown Noddy study plots A, B, C, and D on Tern Island, French
Frigate Shoals, Hawaii.

2500 nesting pairs. East Island, approximately 13 km southeast of Tern Island, is the next
largest colony with approximately 500 breeding pairs.

Four study plots were chosen on Tern Island because of high numbers of accessible Brown
Noddy nests (Fig. 2). Study plot A oceurred on the north side of Tern Island. The plot
measured 78 m in length and ranged from 10.3 to 7.8 m wide. The plot consisted of non-
vegetated areas of coralline rubble, dense stands of the grass Lepturus repens on its northern
edge and sparse patches of small Chenopodium oahuense shrubs on its southern edge; two
small Tournefortia argentea bushes were in the middle of the plot. Study plot B was on
the south side of the island, extending 120 m east to west and 3.5 m north to south from
the edge of the runway. Small (<2 m in height) Tournefortia argentea bushes, Lepturus
repens, Boerhavia repens, Portulaca spp., and Eleusine indica were the major vegetative
cover in study plot B. A coralline berm extended along the plot’s length, approximately 1.5
m from the north edge of the plot. Habitat and total area of study plot C were similar to
that of study plot A. Study plot D was used to determine breeding success of Brown Noddy
nests > 1 m above the ground in Tournefortia argentea and Pluchea spp. bushes. Study plot
A was used in all years of the study, while study plot B was used in 1989-1992; plots C
and D were used in 1980, 1981, and 1982.

We made observations of Brown Noddy breeding phenology and population trends on
Tern Island from 1982 to 1992. Data on other aspects of breeding biology, including incu-
bation behaviors, egg measurements, chick polymorphism and growth, and parental care
and breeding success were collected in 1980-1982 and 1989-1992 (Table 1). Observations
in 1980 through 1987 were made by several U.S. Fish and Wildlife Service personnel;
observations in 1988-1992 were made primarily by the senior author.

Beginning each breeding season, from 1982 to 1992, we conducted island-wide searches
and recorded dates of first arrival for Brown Noddy adults, first eggs laid, and first chicks
hatched and fledged. Because of protracted laying and numerous ne.sting attempts, only the
minimum number of breeding pairs on the island was determined each breeding season by
recording the highest number of nests containing either an egg or a chick on island-wide
monthly counts of ne.sts during 1982-1985 and by island-wide counts made every 36 days
(the mean incubation length of Brown Noddies reported by Brown |1977|), during 1985-

320

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 1

Data Collection on Tern Island, Lrench Lrigate Shoals, Hawaii

Data collected

Year collected

Study ploP

Breeding phenology

1982-1992

Island-wide

Number of breeding pairs

1982-1992

Island-wide

Incubation behaviors

1980-1981, 1989-1992

A, B, C, D

Egg morphometries

1980-1982, 1989-1992

A, B, C, D

Chick polymorphism

1980-1981, 1988

A, C, D

Chick growth

1980

A, C, D

Relay intervals

1989-1992

B

Parental care of nestling

1989-1992

B

Breeding success

1982, 1989-1992

A, B, C, D

“Study plots C and D were primarily used only in 1980 and 1981 and in 1982 for breeding success.

1992. Study plots were checked every other day while nests were active, and every five
days when there were no active nests, except during inclement weather (winds exceeding
8.6 km/h in combination with rain). Because laying occurs throughout the year on Tern
Island but is extremely synchronous for noddies, we defined breeding season as the period
between the laying of the first egg to the laying of the last egg within the study plots,
regardless of calendar year. In 1980 and 1981, breeding occurred from March to October.
However, from November 20, 1988 to October 7, 1990, Brown Noddies nested continuously.
We considered the beginning of the 1990 breeding season to be December 21, 1989, when
there were no chicks less than four weeks of age left on the island and within the plots and
when a new peak of laying occurred. Lrom this peak, study plot A was monitored contin-
uously for three successive breeding seasons, December 21, 1989 to October 7, 1990, March
15 to September 10, 1991, and Lebruary 2 to November 12, 1992. In 1980 and 1981, all
nests within study plots A, C, and D were marked with individually numbered metal tags,
and the fate of each nest was recorded. In 1989-1992, nesting success in study plot A was
observed by marking each nest with a blue-painted piece of coral placed 5—12 cm north of
the nest. Lor all of study plot A, the total number of new eggs, eggs lost, and newly hatched
chicks were recorded at each check. All chicks that reached three weeks of age were banded
with a size 3 incaloy USLWS band. All dead chicks found within the plot were recorded.

Beginning in October 1989, breeding adults were color-banded within study plot B to
monitor individual breeding success and effort. Twenty-one pairs were banded during 1989.
All breeding attempts for these color-banded pairs were recorded. Thereafter, pairs were
captured at the beginning of each season so that the number of breeding attempts for each
pair could be monitored for the entire breeding season. Thirteen pairs were banded in 1990,
and 29 pairs were banded in 1991. All birds were captured by hand, or by using a long-
handled minnow net during daytime hours. They were then banded with a unique color-
band combination consisting of three plastic bands and a size 3 incaloy, USLWS band. A
total of 1 14 birds, representing 62 pairs, were color-banded.

All nests within study plot B belonging to color-banded pairs were marked with a uniquely
numbered rock. Dates of laying, hatching, fledging, and chick departure from the island for
each nesting attempt were recorded. Eggs in marked nests were weighed with a 100-g Pesola
scale to the nearest 1 g; eggs three days of age or older were not weighed. Measurements
of egg length and breadth were taken with digital or manual vernier calipers to the nearest

Megyesi and Griffin • BROWN NODDY BREEDING BIOEOGY

321

0.1 mm. The late of each marked nest was recorded (i.e., failed egg or dead chick). In
addition to daily searches, observations of color-banded birds were conducted from May I
to May 27, 1992 at 06:00—07:45, 12:00—13:45, and 20:00-21:45 h. All sightings of color-
banded birds on any part of the island were recorded during 1989-1992. Incubation lengths
were gathered periodically throughout 1989—1992 for nests with known dates for laying and
hatching. Incubation shifts were followed for a total of 10 pairs in 1980 and 57 pairs in
1981; one bird from each pair was marked on the crown with picric acid. Incubation shifts
were recorded every 2 h for a total of 72 h during each of five observation periods. The
sex was not known for any of the 67 pairs of birds, and pairs were in various stages of
incubation at the time of monitoring.

Proportions of chicks with light down versus dark down were determined in 1980, 1981,
and 1989. Although there were intermediate plumages, chicks were considered light-colored
if they did not contrast markedly with the coralline rubble and dark if they resembled the
color ol the adult plumage. In 1980, chick growth for an initial sample of 28 chicks in study
plot A was measured every 3—4 days until chicks fledged. Chicks were weighed with a
Pesola scale to the nearest 0.5 g; culmen, wing chord, and tarsus lengths were measured
with vernier calipers to the nearest 0.1 mm. In 1989-1992, fledglings were weighed (Sone
week post-fledging) using a 500-g Pesola scale to the nearest 2.0 g, and culmen and wing
chord were measured with vernier calipers to the nearest 0.1 mm. In all years of study,
chicks were considered fledged at the first sign of sustained flight when nests were being
checked. Island-wide searches for fledglings were conducted, and their nests were checked
periodically after dusk between 19:00 and 22:00 h. After 14 days of no sightings, a fledgling
was considered absent from the island.

RESULTS

Breeding phenology and population estimates. — The minimum nesting
population on Tern Island appeared to increase during the ten years of
the study from at least 375 pairs in 1982 to at least 2410 pairs in 1992.
Breeding phenology varied widely between years = 15824.28; df =
88; P < 0.001); however, the largest peaks of laying occurred between
March and September each year (Fig. 3). Peaks of laying varied widely
during these months from 1982 through 1985 and in 1991; while in 1986
and 1987, peak laying occurred in late April, and in late June in 1992.
From 1988 through 1989, Brown Noddies nested continuously.

Noddies usually nested on the ground in open, coralline rubble areas.
Nests varied, ranging from inornate to elaborately lined with pieces of
colored plastic, shells, crab carapaces, and vegetation. Brown Noddies
nested as high as 1.0 m above the ground in Pluchea spp., Chenopodium
ohauense, and Lepturus repens, in which nests consisted of dense plat-
forms of vegetation as large as 0.5 m in diameter and 20 cm thick.

Egg measurements, incubation behaviors, and renesting. — Of 2,889
nests observed on Tern Island, all contained one egg that ranged in color
from all white to heavily speckled brown. There was little variation in
mean egg length and breadth between years (N = 304 and 303, respec-
tively); however, egg mass differed in some years (N = 253;, one-way
ANOVA, Table 2). Mean length, breadth, and mass of color-banded pairs’

Number of Nests

322

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

J MM J S N J MM J S N J MM J S N J MM J S N J MM J S N J MM J S N

1982 1 1983 I 1984 | 1985 | 1986 | 1987

J MM J S N J MM J S N J MM J S N J MM J S N J MM J S N
1988 I 1989 1 1990 | 1991 1 1992

Year

Fig. 3. Nesting phenology and numbers of nests containing eggs of Brown Noddies on
Tern Island, French Frigate Shoals, Hawaii, 1982—1992.

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

323

Mean Egg Measurements (±

Table 2

SD; N IN Parentheses) for Tern Island Brown Noddies

Year

Length

(mm)

Breadth

(mm)

Mass

(g)

1981

53.7 ± 2.0

36.9 ± 1.0

36.0 ± 2.4

(30)

(30)

(30)

1982

53.4 ± 2.2

36.7 ± 1.2

37.2 ± 3.0

(150)

(149)

(124)

1989

53.3 ± 1.5

36.6 ± 1.6

38.2 ± 2.5

(20)

(20)

(2)

1990

53.2 ± 1.8

36.9 ± 0.9

39.0 ± 2.7

(94)

(49)

(48)

1991

53.2 ± 2.1

36.8 ± 1.1

37.9 ± 3.0

(55)

(55)

(49)

0.28

0.53

7.89

P

0.89

0.71

<0.00D

* ANOVA tests for differences between years; df = 4, 299, for length; 4, 298, for breadth; and 3, 247, for mass.

"Egg mass in 1990 differed significantly from 1981 and 1982; in 1991 egg mass differed significantly from 1981; 1989
was not compared due to small sample size.

first eggs were not different from eggs laid later in the same season (F2 /02
= 0.29, P — 0.02 and F2gj = 2.63, respectively; P > 0.05 for all three
measures).

Incubating noddies shifted more frequently in the morning (02:00-
08:00 h) and night (20:00-02:00 h) than in the midday (08:00-14:00 h)
and evening (14:00-20:00 h) (N = 67; == 31.5; df = 3; F < 0.001),

with adults averaging as much as 2.0 shifts during the night period (Table
3). There was no difference in mean shift length between years, so data
were pooled (F.^g/ = 0.02; P > 0.05). Average shift length for 67 pairs
during 360 hrs of observation was 11.3 h (SE = 0.9; range = 5.3-25.2
h). Average duration from laying to hatching was 34.8 d (N = 19, SD =
0.6, range = 29-37 d).

Of the total number of pairs, 47.0% (N = 34) and 34.3% (N = 35)
renested after the first nest failed in 1990 and 1991, respectively. Thirteen
of the renests occurred after the loss of an egg, and 24 occurred after
losing a chick; one pair relaid twice after losing a chick on each of the
two previous attempts (Table 4). There were no differences in the pro-
portion of pairs that relaid in 1990 compared to 1991 (x^ = 1.17; df =
\ \ P > 0.05), nor were pairs more likely to relay after losing an egg
versus losing a chick (x^ = 1.56; df = 1; P > 0.05). Average age of
chicks lost among pairs that relaid was 7.1 d (N = 26; SD = 7.1; range
= 1-20 d). Further, relay interval lengths (the time from laying the first

324

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 3

Incubation Behaviors for Brown Noddies on Tern Island During Each of Live 72-h

Periods of Observation

Mean
incubation
shift length
(hours)

Total number
of shifts
per pair

Number of shifts per lime period'

Year

Morning

Midday

Evening

Night

1980

10.4 ± 6.7'’

6.0 ± 2.3'=

1.6 ± 1.1

1.5 ± 1.3

1.6 ± 1.1

1.3 ± 0.9

(50)

(10)

(10)

(10)

(10)

(10)

1981

March

1 1.1 ± 9.5

7.7 ± 3.4

1.9 ± 1.6

0.4 ± 0.9

1.5 ± 1.0

2.3 ± 1.4

(73)

(14)

(14)

(14)

(14)

(14)

May

12.6 ± 7.9

6.4 ± 1.9

1.5 ± 1.1

1.1 ± 1.2

1.5 ± 0.9

1.2 ± 1.1

(57)

(15)

(15)

(15)

(15)

(15)

June

1 1.7 ± 7.7

6.9 ± 1.96

1.3 ± 1.0

1.0 ± 0.9

0.9 ± 1.0

1.7 ± 1.2

(68)

(15)

(15)

(15)

(15)

(15)

August

10.5 ± 7.8

7.2 ± 2.6

1.3 ± 1.6

1.2 ± 1.1

1.4 ± 0.8

1.6 ± 1.3

(67)

(13)

(13)

(13)

(13)

(13)

All years

1 1.3 ± 0.9“

6.8 ± 0.7

1.5 ± 1.3

0.96 ±1.1

1.1 ± 1.0

2.0 ± 1.1

“ Morning: 02:00

-08:00 h; midday

; 08:00-14:00 h;

evening: 14:00-

-20:00 b; night: 20:00-02:00 h.

*’ Mean ± SD; number of shifts per observation period in parentheses.
'Mean ± SD; number of pairs in parentheses.

Mean ± SE.

egg to laying the replacement egg) were not different whether the pair
lost an egg, a chick < 1 week of age, or a chick > 1 week of age (Kruskal-
Wallis, H = 4.11; df = 2, N = 7, 11 and 7, respectively; P > 0.05).
However, pairs that successfully fledged a chick waited longer to nest
again than did pairs that lost an egg or a chick (Kruskal-Wallis, H =
16.08; df = 2; N = 6; 7* < 0.001). These data should be inteipreted with
caution, however, as the sample sizes were small.

Chick polymorphism, growth and parental care. — There were no dif-
ferences between years in the proportion of light- and dark-colored chicks
in 1980, 1981, or 1989 (N = 221; = 0.253; df = 2; P > 0.05). Down

color for chicks during these three years was 26% light- and 74% dark-
colored. For 84 of 160 chicks, average age prior to disappearance from
the nest did not differ with respect to color (N = 19 and 65, mean = 9.9
d and 7.5 d, SD = 6.9 and 4.6 for dark and light-colored chicks, respec-
tively).

Average chick mass at one day of age was 28.8 g (N = 29, SD =
3.83); chicks attained average adult mass (200.4 g, N = 122) at 34 d of
age. Chick mass increased until 38 d post-hatching and began decreasing
on day 43 (Fig. 4). At the end of the measurement period (range = 47-
51 d), the average chick mass was 195.8 g (N = 13, SD = 18.39). There

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

325

as

c

m t3

*00

+1 'Z'

Tl VO

<U

r<^ ^ IN

U.

oi o

"Z x>

<N _:

w

j

OQ

<

ON

On

Q

2

<

O

Ov

On

U

D

Q

c/5

u

s

Q

O

:z:

2

O

a:

CQ

a:

O

u,

c/5

-J

<

>

q:

u

H

2

X

U

H

D

_)

u

Q

2

<

C/)

>■

<

J

UJ

OC

u.

O

oc

UJ

03

D

z

00 a.
^ P
- 1

> ^
S--

.5 a

>^2

—

dj

QC

5 O

Is

IT)

c/5

Os

03

"O

+1

NO

i

rn

r->

>*
C3
"O

ON

(N ^

(N ^

+ 1

NO
(N

+ 1

>,

CQ

r-

00

I

(N

a

u

o

X «

c

=

.w

^ d>

?r 00
o

_0 c/5
■P +1
> =

ra

ii aj

Weighi (g) Culmen (mm)

326

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

40

35 -
30

25 -
20 -
15 -
10

0/f r-

hatch 5

300 ■

250
200
150 -
100 -
50 -
0

. .M-

I

10 15 20 25 30 35 40 45 fledge

Age of chick in days

hatch 5

— 1 1 i 1 1 i 1 I

10 15 20 25 30 35 40 45 fledge

300

250 -

200

'b

•S. 150

QO

c

^ 100
50
0

• I • .

i ; 1 1 1 1 1 I I t

hatch 5 10 15 20 25 30 35 40 45 fledge

40

35

30

25

20

15

o'4

Age of chick in days

r'

hatch 5 10 15 20 25 30 35 40 45 fledge

Age of chick in days

Age of chick in days

Fig. 4. Growth of Brown Noddy chicks on Tern Island.

was no difference in weight loss between newly fledged chicks and chicks
that were weighed ^ one week post-fledging (t = 1.01, df = 21, P >
0.05). There was relatively little growth in wing length until 10 d of age,
followed by steady growth up to 50 d post-hatching, when the measure-
ment period ended (Fig. 4). Culmen length increased throughout the 50
d of measurement, while tarsus length increased until 20 d of age (Fig.
4). Mean duration from hatching to fledging was 47.8 d (N = 31, SD =
4.3, range = 39-61 d); however, chicks remained in the nest and were
fed by their parents an average of 65.1 d after fledging (N = 14, SD =
13.4, range = 38-88 d). Three birds returned to within 10 m of their
natal site 11, 21, and 24 months after leaving the island. The youngest
bird observed breeding was two years and 1 1 months old.

Breeding success. — In 1990, a combination of severe storms during
peak laying period and depredation of chicks by Great Frigatebirds {Fre-
gata minor) resulted in poor reproductive success in comparison with
1991 and 1992 (Table 5). Similarly, breeding success was poor from
February 1 to June 23, 1992; only three of 79 eggs laid hatched, and no
chicks survived to fledging. However, after this date, nearly all eggs laid
hatched and resulted in fledged chicks (Table 5). There was a difference
in hatching and fledging success between study plots; both plots B and
D had greater hatching success than A and C, while fledging success was
lower in plot B and greater in plot D when compared to plots A and C

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

327

Table 5

Brown Noddy Hatching, Eledging, and Reproductive Success

1982

1990

1991

1992

Total # eggs laid

222

1239

702

726

Total # eggs lost

51

644

128

161

Total # eggs hatched

171

595

574

565

Total # chicks found

dead in nests

0

42

12

3

Total # fledged

62

15

263

324

Hatching success (%)“

77

48*’

82'--

78”

Fledging success (%)“

36

3*’

46^

57”

Reproductive success (%)“

28

1.2

38

45

^ Hatching success is the number of chicks hatched from the total number of eggs laid; fledging success is the number
of chicks fledged from the total number of chicks hatched, and reproductive success is the number of chicks fledged from
the total number of eggs laid.

Lower than in other years; X‘ = 313.2 for hatching success and 430.4 for fledging success; df - 3; P < 0.001.

Higher than in other years; P < 0.001 .

(X" - 22.3 and 60.4 for hatching and fledging, respectively; df = 3; P <

0.001).

DISCUSSION

Collectively, the Hawaiian Islands support between 89,500 and 150,000
breeding pairs of Brown Noddies. Nearly 32% of these occur in the north-
western Hawaiian Islands, with the largest population found on Nihoa
Island (Harrison et al. 1983). Breeding Brown Noddies were documented
by the Rothschild expedition at French Frigate Shoals as early as 1891,
when thousands were observed nesting on Tern Island (Munro 1941).
Wetmore observed 500 pairs nesting on Tern Island in 1923; however,
they were not observed nesting on the island from 1953 to 1969 (in
Amerson 1971). This period of non-nesting corresponds with its use by
Coast Guard personnel as a LORAN station and the presence of cats and
dogs on the island (Richardson 1954, Amerson 1971). In June 1967, an
estimated 10,182 Brown Noddies were observed at French Frigate Shoals.
However, the maximum number of nests recorded was 1675, occurring
mainly on East and Whaleskate islands (Amerson 1971). Breeding nod-
dies were observed again on Tern Island in 1977, shortly before the island
was abandoned by the Coast Guard (M. Rauzon, pers. comm.). It is dif-
ficult to interpret whether an increase in the nesting population at French
Frigate Shoals has occurred since 1891, but the population at Tern Island
has expanded greatly since 1977 (Fig. 3; USFWS unpubl. data).

Time of arrival for breeding adult Brown Noddies and their occurrence

328

THE WILSON BULLETIN • Vol. 108. No. 2, June 1996

at a breeding colony vary among Pacific, Indian, and Atlantic ocean col-
onies (Table 6). In the Hawaiian Islands, Brown (1973) observed adults
year-round on Manana Island but noted their absence during the day from
December to March in 1971 and 1972. Similarly, noddies are year-round
residents on Nihoa, Necker, and Laysan islands, but fewer adults are pres-
ent during the day from December to March (Ely and Clapp 1973, Clapp
et al. 1977, Clapp and Kridler 1977). Woodward (1972) noted that Brown
Noddies are entirely absent from January to March on Kure Atoll. We
observed adults on Tern Island year-round except during an El Nino event
which occurred from October 1990 to March 1991, when adults aban-
doned the island and left fledglings still dependent on parental care to
starve (Eig. 3). In contrast, Murphy (1936) stated that noddies from all
of the sub-tropical South Atlantic Islands migrate from their nesting
grounds between May and December. Similarly, Morris and Chardine
(1992) reported that noddies were absent annually from Cayo Noroeste,
Culebra, Puerto Rico from September to March. In Elorida, Robertson
(1964) noted that Brown Noddies departed the Dry Tortugas as early as
May, and Watson (1908) reported that all birds were absent annually from
these colonies by the end of September. We suggest that the variation in
occurrence of Brown Noddies at Pacific and Indian Ocean breeding col-
onies versus those in the Atlantic may be due to a variety of factors,
including food availability, ocean currents, and water temperature. How-
ever, there are no data to confirm these relationships.

In the Hawaiian Islands, the Brown Noddy’s breeding season is erratic.
Laying occurs from May to August on Manana Island (Brown 1973),
while on Nihoa, Necker, and Laysan islands egg laying has been docu-
mented throughout the year (Ely and Clapp 1973, Clapp et al. 1977, Clapp
and Kridler 1977). Woodward (1972) documented egg laying at Kure
Atoll annually from April through August (Table 6). We observed that
laying on Tern Island could occur throughout the year, although most
eggs were laid between March and September during 1982 to 1992 (Eig.
3). Brown Noddy breeding phenology and synchrony has been explained
as a response to food availability (Ashmole 1963). Morris and Chardine
(1992) attributed variation in Brown Noddy breeding phenology to geo-
graphic differences in feeding regimes, ffrench (1990) stated that inclem-
ent weather on Soldado Rock, Trinidad, prolonged the breeding season
so that two peaks of laying occurred in 1966, but during other years, only
one peak of laying was recorded. High loss of eggs and chicks to high
tides, inclement weather, and predation by Pied Crows {Cot-vus alhus) on
Aldabra Atoll in the Seychelles also caused variability in laying patterns
(Diamond and Prys-Jones 1986). Likewise, Dorward and Ashmole (1963)
attributed the double peaks of laying on Ascension Island to relaying or

Table 6

Breeding Phenology of Brown Noddies in Pacific, Atlantic, and Indian Ocean Colonies

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

329

u

o

c/5

Os

O

O

U

’O

c

UJ

•o

D

-C

H

X

r-

ON

r-*

r-

ON

3

0

r-*

u

13

r-

On

?3.

X

C/D

5

*0

00

A

c

’3-

*3

r-

3

On

x

On

w

3

3

(U

3

0

C

a

a

3

!ZI

u

3

a

a

(U

u

3

3

O

3

D.

u

3

C/D

U

u

CQ

<

CN

ON

On

C

cd

j::

U

■o

c

C3

fc

O

cn

X

On

X

00

ON

a

(U

tU

3

*0

O

E

c/5

X

Ol

<

a

■o

•o

c

3

3

3

X5

T3

u

3

3

0

£

u

O

3

Q Q

r-

ON

s

c

o

c/)

Xi

O

D. ^

a o

.g-§

t

C3

D

C

c3

00

3

<

D,

<

<U

X)

6

o

a.

V

on

(U 0) — O (U

U u
CO CO
> >

*o -o
c c

3 3

O O

Lh U

CO 3

'cO

a a

- *0

■o c

C 3

i2

at

^ C
C 3
1> a
(U >Y

O J

3

O

X

LTi

B

3

00

'u

Un

<u

X

E

<u

3 ^

WJ ^

- V 3 3

I S

D.

(U

a

<

Xi

2

'C

:>

Ofl

I ^

^ 9

S £

99

D.

< s

a

<

I

>

o

Z

T3 -O “O T3
C C C C
3 3 3 3

O O O O

3 3 3 3

U. n; dj (U O

< >. >H >, >.

> (U c

§ E S

•c 2 ^ i

5* k2

< u. Z >

3

.*E

'u

"O

C .-

iH 3

E p:

3 ^

S —

3 -E a

u ^

O »-

H CO

>N X ^
u .- <D

Q Z Z

u

X

'o

3

O,

c

o

O)

c

X

o

-o

c

iS c

c/5 3

dJ

3 O

3

(A

o

o

5

(U

3

CU

3*^

C

3

a>

u

o

^ ^ I

JJ XJ J3

•E <

3

u

<u

i -o
^ c

^ ^ 3

O) O X
dJ 00 nX
0 3^

*- X 3

Z t2 -2

O u S

W u
3 X czi

U J <

(U

X

o

>.

dJ

C/D

O-

dJ

C/D

i

D.

<

■o

c

3

o

3

V

T3

b

>

>

>

00

b

>

w

w

3

’5b

o

00

•ct

o

o

X

fN

in

o

o

3*

in

o

3-

0

r-

o

ON

(N

o

b

C3

o

CN

CN

b

(N

C N

3-

0

c

o

r-

X

CN

X

X

in

X

in

o

o

o

X

in

Q

""

00

—

—

*— '

X

X

3"

flj

..

"O

3

Z

Z

Z

z

Z

Z

Z

Z

Z

Z

00

3

un

b

rj

b

in

X

b

in

Z-.

00

r-

9t

in

J

(N

»n

(T)

p

(N

(N

in

p

00

un

ro

bi

X

00

9-(

r^

On

b

fNl

(^1

(N

(N

(N

—

o

O

•o O
<

C

3

(U

O

O

c

3

*■5

C

*0

c

3

3

E

X

U

330

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

to birds laying for the first time that had been prevented from nesting due
to high seas. On Tern Island, Brown Noddy pairs attempted to renest as
often as four times during a breeding season after losing an egg or chick.
We know of no other study that has documented Brown Noddies relaying
after losing their second and third clutches within the same breeding sea-
son. Ashmole (1963) observed that Sooty Terns will renest more fre-
quently if loss of an egg occurs early rather than late in the incubation
period. He also suggested that predation of Sooty Tern chicks by Ascen-
sion Frigatebirds {Fregata aquila) may have contributed to the breeding
cycles observed on Ascension Island, where birds laying during the peak
season are more likely to fledge young. The renesting efforts we observed
on Tern Island could explain the multiple peaks in laying and, conse-
quently, the variation in Brown Noddy breeding phenology. Predation on
chicks by Great Frigatebirds, on eggs by Ruddy Turnstones {Arenaria
interpres), and inclement weather resulting in nesting failure may all con-
tribute to the multiple laying peaks we observed.

Brown Noddy eggs on Tern Island were significantly larger than those
laid on Manana Island and at other Pacific and Atlantic ocean colonies
(One-way ANOVA, P < 0.05; Table 7). Morris and Chardine (1992)
stated that differences in egg sizes between Pacific and Atlantic ocean
colonies were probably not due to a difference in body size. However,
the body mass of 122 adults studied on Tern Island averaged larger than
those on Cayo Noroeste, Culebra; Ascension Island (mistakenly reported
as a Pacific Ocean colony by Morris and Chardine 1992); and Manana
Island (SD = 16.7, range = 165-242 g, F = 65.5, df = 3, F < 0.001).
Egg weight as a function of body weight has also been discussed by Rahn
et al. (1975), Verbeek and Richardson (1982), and Pierotti and Bellrose
(1986). Our data support the results of Monis and Chardine (1992) who
found no differences in the measurements of a pair’s first egg in com-
parison with subsequent eggs laid during the breeding season.

Mean incubation period did not differ for noddies nesting at Tern Island
or Cayo Noroeste, Culebra, although both of these colonies had shorter
incubation periods than did noddies on Manana Island (One-way ANO-
VA, F = 1 18.6, df = 2, F < 0.001; Brown 1973, Morris and Chardine
1992). Mean incubation shift lengths for noddies on Tern Island were also
shorter than those observed on Manana Island (Brown 1973) and the Dry
Tortugas, where Watson (1908) reported shifts occumng every two hours
during the day. Finally, birds at both Tern and Manana islands switched
most often during the morning and night.

Although the proportion of light- and dark-colored chicks was not dif-
ferent between years on Tern Island, it did differ among five colonies in
the Pacific and Atlantic oceans (Table 7; = 72.92, df = 4; F < 0.001).

1

Table 7

Aspects of Brown Noddy Breeding Biology Compared to Some Pacific, Atlantic, and Indian Ocean Colonies

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

oo

q

q

q

q

q

O

o

o

cn

irj

f-l,

NO

NO

d

f<N

r--

NO

CN

in

in

On

o

o

o

q

o

q

q

q

q

q

O

O

NO

iri

00

Tt

d

u

X

nO

(N

cn

in

N-

rt

00

O

un

(N

m

in

Z

NO

(N

j

On

in

j

CN

(N

cn

in

(N

q

in

CN

q

(N

CN

CN ^

d

r-

CO

lo

eigl

Ig)

+ 1

(>•

+1

00

NO

1

+i S

.1 CN

Tl ^

O'

q

CN

q

q O

q ^

NO

cn

m

cn

cn

m

ON

00

00

SI — S

— *

?-•

d

d ^

d ^

O'

■5 £

L- C

1

+1

O

+ 1 c^.

-f-l

ON

CN

+1

+ 1 2

+ 1

00

fO

q

o

T c

(n C-

O

NO

lO

NO

in

NO

NO

cn

m

m

m

(n

(N

(N

00

o

s

CN

O'

CN

CN ^

O'

oe £

1

+ 1

o

-t-l

O'

+1

ON

CN

1

+ 1 ^

+1 Z3

+ 1

V £

52.0

00

q C.

w

(T)

rn

CN

CN

d

in

in

in

in

in

in

Q

00

O'

in

O'

NO

O'

CN

d

d

O

\0 I
ro

O

r--

CXD r-)

^ A

m On
(N

o

o

NO

I

q

lO

ON

^ r-
o

NO 1
ro

rj

r-

ON

-
- "O

1)

o

5

cd

"O

0

1

“O

3

(>0

Um

u-

“O

c

r3

I

cd

q

u

P--

c/f

id

00 ^
3 §

^ c

c

o

u

CQ

q

ir!

ON

o

<

c

o

I

o

-C

"O

n ^

C

<U .«

u j
u

>Y

u

Q

3

JZ

c

3

2

c/5

*0

c

3

o

d

3

3

0

c/5

(U

0

u

3

c/5

3*

3

3

O'

3

C/D

U

O

E

NO

r-

O

2

■fc

o

C

3

On

“O

3

<

On

o

s

S

3

cn

3

U

(U

c

o

on

3
O
J=

* c

"c 2

cd on

2 -u

c

cd

(N

On

C 2
2 —
^ —

£ X

nO

On

^ 5

2 5

C4l

« c
E 3

W (D
c/5 o

o

u

331

■■ Mean ± SD, N in parentheses.

Brown (1973) placed I 10 chicks in a gray category; these chicks were omitted from the analysis. N = number of chicks analyzed.
* Egg dimensions from these colonies compared to Tern Island, one-way ANOVA, P < 0.05 for all three measures.

332

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

While there were more light-colored chicks at Kure Atoll and Cayo No-
roeste, Culebra, there were more dark-colored chicks on Ascension, Man-
ana, and Tern islands. Because of the subjective methods used at all five
colonies to classify chicks as either light or dark, some observer bias may
explain the differences between colonies.

Brown Noddy chick growth rates on Tern Island were similar to those
reported by Dorward and Ashmole (1963), Brown (1973), and Morris
and Chardine (1992) at other Pacific and Atlantic ocean colonies. How-
ever, chicks on Tern Island fledged at a later age than those observed on
Cayo Noroeste, Culebra (mean = 45.7 d, cf Morris and Chardine 1992)
and Manana Island, although this difference may be the result of observer
bias in determining when a chick has fledged. Gibson-Hill (1951) re-
marked that chicks are fed by their parents some weeks after learning to
fly, and Brown (1973) observed a chick being fed more than 100 days
after fledging. It seems likely that Brown Noddy chicks are attended at
the colony and learn to forage during this time rather than leaving the
colony with adults as is the case for Sooty Terns and Red-tailed Tropic-
birds (Phaethon rubricauda) (Ainley et al. 1986). This is further sup-
ported by our observation of both members of a pair back at the breeding
colony within 12 d after the departure of their fledglings from the island.
Finally, although our data support Burger’s (1980) conclusion that terns
defer breeding until their third year, our study documents that birds can
return to their natal site as early as 1 1 months after independence.

Ashmole (1963) attributed poor breeding success on Ascension Island
to chick starvation. Likewise, Morris and Chardine (1992) concluded that
higher breeding success at Cayo Noroeste, Puerto Rico, was the result of
a reliable food source for adults. In contrast. Brown (1973) reported that
on Manana Island all eggs hatched after June 1 1 were preyed upon by
Black-Crowned Night-Herons {Nycticorax nycticorax), and although
breeding seasonality was governed by food supply, the ultimate factor
affecting fledging and success of the breeding season was predation on
Brown Noddy chicks. Robertson (1964) also reported predation by Mag-
nificent Frigatebirds {Fregata magnificens) on Brown Noddy chicks on
the Dry Tortugas, although neither he nor Watson (1908) quantified the
effect on reproductive success. In our study. Great Frigatebird predation
was responsible for nearly all chick losses in 1990-1991 (Megyesi 1995).
Anecdotal observations collected by USFWS personnel in previous years
since 1979 confirm that Great Frigatebird predation occurs regularly on
Tern Island. Although most Brown Noddy chick loss was attributed to
frigatebird predation in this study, in 1992, Brown Noddies, Black Nod-
dies {Anoiis tenuirostris). Sooty Terns, and White Terns {Gygis alba) ex-
perienced complete nesting failure, where incubating adults abandoned

Megyesi and Griffin • BROWN NODDY BREEDING BIOLOGY

333

pipped and hatching eggs and newly hatched chicks up to June 23. Fol-
lowing this date, Brown Noddies renested, and reproductive success was
higher than had been observed in previous years of study. We suggest
that this series of events in 1992 may be related to El Nino Southern
Oscillation (Schreiber and Schreiber 1984; unpubl. data, USFWS).

The differences in fledging success between plots B and D are probably
a result of Great Frigatebird predation. Plot B lies parallel to the runway
and is more readily accessible to hunting frigatebirds. In contrast, plot D
contained nests built in low growing vegetation that might have provided
more cover from hunting frigatebirds. We suggest that lower reproductive
success for Brown Noddies on Tern Island in comparison with other col-
onies is the result of inclement weather and constant predation pressure
by Great Frigatebirds.

ACKNOWLEDGMENTS

We thank the numerous U.S. Fish and Wildlife Service staff and volunteers on Tern Island
who helped to collect a decade of biological observations. We especially thank Ken Mc-
Dermond and Ken Niethammer for their encouragement and support. We also thank Kyle
Jones, Sheila Conant, and an anonymous reviewer who read an earlier draft of this manu-
script.

LITERATURE CITED

Ainley, D. G., L. B. Spear, and R. J. Boekelheide. 1986. Extended post-fledging parental
care in the Red-tailed Tropicbird and Sooty Tern. Condor 88:101-102.

Amerson, a. B., Jr. 1971. The natural history of French Frigate Shoals, Northwestern
Hawaiian Islands. Smithsonian Institution, Atoll Res. Bull. 150. Washington, D.C.

AND R C. Shelton. 1976. The natural history of Johnston Atoll, central Pacific

Ocean. Smithsonian Institution, Atoll Res. Bull. 162. Washington, D.C.

Ashmole, N. P. 1963. The biology of the wideawake or Sooty Tern (Sterna fuscata) on
Ascension Island. Ibis 103b:294-364.

Brown, W. Y. 1973. The breeding biology of Sooty Terns and Brown Noddies on Manana
or Rabbit Island, Oahu, Hawaii. Ph.D. diss., Univ. of Hawaii, Honolulu, Hawaii.

. 1977. Temporal patterns in laying, hatching and incubation of Sooty Terns and

Brown Noddies. Condor 79:133-136.

Burger, J. 1980. The transition to independence and postHedging ptirental care in seabirds.
Pp. 367-440 in Behavior of marine animals, current perspectives in research. Vol. 4 (J.
Burger, B. L. Olla, and H. E. Winn, eds.). Plenum Press. New York, New York.

Clapp, R. B. and E. Kridler. 1977. The natural history of Necker Island, northwestern
Hawaiian Islands. Smithsonian Institution, Atoll Res. Bull. No. 206. Washington. D.C.

’ > and R. R. Fleet. 1977. The natural history of Nihoa Island, Northwestern

Hawaiian Islands. Smithsonian Institution, Atoll Res. Bull. No. 207. Washington, D.C.

Cramp, S. 1985. Birds of Europe, the Middle East and North Africa: the birds of the
Western Palearctic. Oxford Univ. Press. New York. New York.

Diamond, A. W. and R. P. Prys-Jones. 1986. The biology of terns nesting at Aldabra Atoll,
Indian Ocean, with particular reference to breeding seasonality. J. Zool. Lond 2I0'527-
549.

334

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Dorward, D. E and N. P. Ashmole. 1963. Notes on the biology of the Brown Noddy
Anous stoliclus on Ascension Island. Ibis 103b:447-457.

Ely, C. a. and R. B. Clapp. 1973. The natural history of Laysan Island, northwestern
Hawaiian Islands. Smithsonian Institution, Atoll Res. Bull. No. 171. Washington, D.C.

FFRENCH, R. 1990. The birds and other vertebrates of Soldado Rock, Trinidad. Liv. Wrld.
Jour. Trinidad and Tobago Field Nat. Club; 16-19.

Gibson-Hill, C. a. 1951. Notes on the nesting habits of seven representative tropical
seabirds. J. Bombay Nat. Hist. Soc. 48:214—235.

Harrison, C. S., T. S. Hida, and M. P. Seki. 1983. Hawaiian seabird feeding ecology.
Wildlife Mongr., No. 85. Bethesda, Maryland.

Megyesi, J. L. 1995. Breeding biology of the Brown Noddy Anous stolidus pileatus on
Tern Island, Hawaii. Master’s thesis. Univ. of Massachusetts Amherst, Amherst, Mas-
sachusetts.

Morris, R. D. 1984. Breeding chronology and reproductive success of seabirds on Little
Tobago, Trinidad, 1975—1976. Colon. Waterbirds 7:1—9.

and j. W. Chardine 1992. The breeding biology and aspects of the feeding ecology

of Brown Noddies Anous stolidus nesting near Culebra, Puerto Rico, 1985-1989. J.
Zool (Lond). 226:65-79.

Munro, G. C. 1941. Birds of Hawaii and adventures in bird study; an ocean cruise. ’Elepaio
2:63-64.

Murphy, R. C. 1936. Oceanic birds of South America, vol. II. American Museum of Natural
History. New York, New York.

PiEROTTi, R. AND C. A. Bellrose. 1986. Proximate and ultimate causation of egg size and
the “third-chick disadvantage” in the Western Gull. Auk 103:401-407.

Rahn, H., C. V. Paganelli, and A. Ar. 1975. Relation of avian weight to body weight.
Auk 92:750-765.

Richardson, E 1954. Notes on the birds of French Frigate Shoal. Part II. General accounts
of visits of December 1953 and March 1954. ’Elepaio 14:73—75.

Robertson, W. B., Jr. 1964. The terns of the Dry Toitugas. Bull. Flor. State. Mus. 8:1-94.

SCHREIBER, R. W. AND E. A. SCHREIBER. 1984. Central Pacific seabirds and the El Nino
southern oscillation: 1982 to 1983 perspectives. Science 225:713-716.

Sprunt, a., Jr. 1948. The tern colonies of the Dry Tortugas Keys. Auk 65:1-19.

Stonehouse, B., and S. Stonehouse. 1963. The frigate bird Fregata aquila of Ascension
Island. Ibis 103b;409-422.

Verbeek, N. a. M. and H. Richardson. 1982. Limits to egg size in gulls: another point
of view. J. Field Ornithol. 53:168—170.

Watson, J. B. 1908. The behavior of Noddy and Sooty Terns. Carnegie Inst., Washington,
Papers from the Tortugas Laboratory 2:187—255.

Woodward, P. W. 1972. The natural history of Kure Atoll, Northwestern Hawaiian Islands.
Smithsonian Institution, Atoll Res. Bull. No. 164. Washington, D.C.

Wilson Bull., 108(2), 1996, pp. 335-341

DISCRIMINATION BETWEEN REGIONAL SONG
FORMS IN THE NORTHERN PARULA

Daniel J. Regelski' - and Ralph R. Moldenhauer'

Abstract. — Distinctly different territorial (Type A) song forms characterize western and
eastern populations within the breeding range of the Northern Parula (Parula americana).
We conducted playback experiments to determine if territorial males respond differentially
to the two song forms. Male response is stronger to Type A songs of their own population
than to the songs of the other population (two-tailed Wilcoxon test, P < 0.001 ). The possible
basis for this discrimination is discussed. Received 3 Dec. 1993, accepted 15 Oct. 1995.

Many oscine species exhibit geographic song variation (Thielcke 1969),
and some of these species’ songs vary microgeographically, with two or
more small scale local dialects (Kroodsma 1981, Tomback et al. 1983,
Kroodsma et al. 1984). In species such as the Mourning Warbler {Opo-
rornis Philadelphia), there is macrogeographical song variation, with two
or more distinct and widespread regional song forms or song populations
(Pitocchelli 1990).

The Northern Parula {Parula americana) has two primary song types.
Types A and B (Moldenhauer 1992). Spectrographic analysis of Type A
songs from throughout the breeding range by Moldenhauer (1992) re-
vealed an eastern and a western song population (Fig. 1) whose songs
are characterized by distinctly different terminal notes (Fig. 2). Playback
experiments have shown that in species whose song varies geographically,
territorial males can often distinguish between different dialects or song
forms, as inferred from the intensity of response to playback (Kroodsma
et al. 1984, Ritchison 1985). Usually, the response is stronger to local or
familiar song forms. In the present study we conducted playback exper-
iments with eastern and western Northern Parulas to determine if terri-
torial males would respond differentially to eastern and western forms of
the Type A song.

METHODS

Our experimental design and data analysis follow Kroodsma et. al. (1984) and Kroodsma
(1989), with certain modifications. We obtained over 100 Northern Parula Type A songs
from the Texas Bird Sound Library (TBSL) at Sam Houston State Univ. in Huntsville, Texas,
the Cornell Library of Natural Sounds (CLNS) at Cornell Univ., and The BoiTor Laboratory
of Bioacoustics (BLB) at The Ohio State Univ.

We used REAL TIME SPECTROGRAM .software (by Engineering Design, Belmont,
MA) to measure the trill portion of each .song for two characteristics: trill duration (TD)

' Dept, of Biological Sciences, Sam Houston State Univ., Huntsville, Texas 77341.

^ Pre.sent Addre.ss: Dept, of Zoology, The Ohio State Univ., 1735 Neil Ave., Columbus. Ohio 43210.

335

336

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

IH EASTERN FORM “ RECORDING LOCATIONS

Eig. 1. Map of the distributions of the western and eastern song populations within the
breeding range of the Northern Parula, as well as the recording locations for the songs used
to make the playback tapes and the locations of the playback experiments.

and trill rate (TR), the number of trill syllables per second. The trill is the portion of the
.song preceding the last syllable or terminal note. The eastern and western populations differ
significantly in the mean values for these two variables (Moldenhauer 1992). Two values
were calculated for each song: TD/mean TD and TR/mean TR, using the mean values for
the appropriate song population as reported by Moldenhauer (1992). The songs were ar-
ranged into pairs of one eastern and one western song that were matched. as closely as
possible for the values of TD/mean TD and TR/mean TR, to pair .songs that were similar
to one another, while taking into account the characteristics of each of the two song popu-
lations. Using this method, 12 playback tapes were made, using songs recorded at various

FREQUENCY IN KHz oo o FREQUENCY IN KHz

Regelski and Moldenhauer • NORTHERN PARULA SONG

337

0

0.5 1.0

TIME IN SECONDS

1.5

EASTERN FORM

0

0.5 1.0

TIME IN SECONDS

1.5

Fig. 2. Representative western and eastern forms of the Northern Parula Type A song. Notice
the difference in the terminal note. These songs were used to make playback tape number 10.

338

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

dates from throughout the ranges of the two song populations (Lig. 1 ). Each tape consisted
of 25 repetitions of one eastern song spaced 1 2 sec apart on the right channel, and the same
treatment of one western song on the left channel. The two channels were staggered by 6
sec so that a song was broadcast from alternating speakers every 6 sec. On half the tapes
an eastern song was broadcast first, on the other half, a western song was broadcast first.
Erom 9 April to 14 May 1993, playback experiments were conducted with 12 western males
near Huntsville in Walker County, Texas. Erom 19 to 24 May 1993, playback experiments
were conducted with 12 eastern males in Liberty County, Elorida, as well as Bulloch, Cam-
den, Jenkins and Screven Counties, Georgia (Fig. 1). The same 12 tapes were used for the
playbacks to both western and eastern males. This reciprocal design is discussed in Kroods-
ma (1989).

Songs were played on a Uher 4200 tape recorder through a pair of Radio Shack Minimus
0.8 self-amplified speakers placed 18 m apart. The playback level was set by ear to ap-
proximate a singing bird. Ribbons were placed every 2 m between the speakers, and the
location of the bird during the playback period with respect to the measured 18 m was
recorded every 6 s. After the first 5 min. playback period with the western song broadcast
from position 0 m and the eastern song from position 18 m, the speaker cables were switched
and the tape played a second time.

For the western trials, the median position of the bird during the first playback period
was subtracted from the median position during the second playback period. For the eastern
trials, the median position of the bird during the second playback period was subtracted
from the median position during the first playback period. Positive values for difference in
median positions (measured in meters) indicate a closer approach to and/or more time spent
in proximity to the speaker broadcasting the bird’s own song form, while negative values
indicate a closer approach and/or more time spent in proximity to the speaker broadcasting
the other song form. Comparing two positive values, the larger reflects a closer approach
and/or more time spent in proximity to the speaker broadcasting the bird’s own song form
than the smaller value. Kroodsma (1989) suggests that such differences in median positions
can be used as an “index of response’’ to the two stimuli. The greater the positive difference
in median position, the stronger the response to the bird’s own song type. The rationale for
this method is presented by Kroodsma et. al. (1984). Therefore, we interpret closer approach
to, and more time spent in proximity to a speaker broadcasting a song form, as evidenced
by the difference in the bird’s median position during two playback periods, as a stronger
aggressive response to that song form. Hereafter, we will use the terms “stronger response”
and “responded more strongly” to reflect this interpretation. Trials where the male did not
approach and spend time between the speakers were not included in the analysis.

RESULTS

Ten of twelve western birds responded more strongly to their own song
form, and two responded more strongly to the other song form. The bias
in response is significant (two-tailed Wilcoxon test, = 8, P < 0.05).
Nine of twelve eastern birds responded more strongly to their own song
form, and three responded more strongly to the other song form. The bias
in response is significant (two-tailed Wilcoxon test, Ts = 12, P < 0.025).

To justify pooling both data sets, we compared the means of the dif-
ferences in median position, a measure of the average strength of response
for all 12 subjects to their own song type. These values, 4.48 m for
western birds and 3.78 m for eastern birds, are not statistically different

Regelski ami Moldenhauer • NORTHERN PARULA SONG

339

C/D

u

<

PQ

<

CU

uu

O

Qi

W

03

S

D

2

DIFFERENCE IN MEDIAN POSITIONS (m)

Fig. 3. Summary of all the playbacks to Northern Parulas. Each cell represents a play-
back session to a different male. The letter in each cell indicates whether the playback was
to a (W)estern or (E)astern bird, and the number is the playback tape used. The playbacks
were conducted in Texas, Florida, and Georgia. The abscissa is the difference between the
median positions during the first and second playback periods. Positive values reflect a
stronger response to the subject’s own song form, a negative value reflects a stronger re-
sponse to the other song form, and zero (0) indicates no preference. Note: for this figure,
the median differences were rounded to the nearest whole number.

(two-tailed Mann-Whitney U-test P > 0.20). Applying Bonferroni’s In-
equality (Lehman 1991), the a for the test of the pooled data cannot
exceed 0.049. Pooling the data gives the following results: 19 out of 24
birds responded more strongly to their own song form, and five out of
24 responded more strongly to the alternate song form (Fig. 3). The bias
in response is significant (two-tailed Wilcoxon test, = 35, P < 0.001).
These results suggest that the birds respond more strongly to their own
song form.

DISCUSSION

Territorial male Northern Parulas respond to both regional forms of the
Type A song but seem to be able to discriminate between the two, as

340

THE WILSON BULLETIN • VoL 108, No. 2, June 1996

evidenced by stronger response (closer approach, more time spent in prox-
imity) to playback of their own song form. This resembles the pattern of
response that has been observed in the Chiffchaff {Phylloscopus colly-
bita), a species with macrogeographic song variation comparable to that
of the Northern Parula (Thielcke and Linsenmair 1963).

In Song Sparrows {Melospiza melodia), the strength of response to
playback of various songs increases with similarity to the subject’s own
song (McArthur 1986). There is also evidence that the learned association
of song with aggressive behavior or territorial disputes (Payne 1986, Rich-
ards 1979) or visual stimuli (Crook 1984, Murray and Gill 1976) is in-
volved in song form discrimination.

The two Type A song forms of the Northern Parula are relatively sim-
ilar (Fig. 2), and the song probably is partially learned (Kroodsma and
Baylis 1982). It is unlikely that any of the subjects had been exposed to
the other song form on the breeding or wintering grounds, as the play-
backs were conducted in areas of allopatry, and preliminary evidence
suggests that the two song populations may have separate wintering
grounds (Moldenhauer, unpubl. data). Thus, the song discrimination re-
ported in this paper may be a result of (1) stronger response to the more
structurally similar song form, (2) association of the bird’s own song form
with territorial disputes or aggressive interactions, (3) association of the
bird’s own song form with visual stimuli, such as conspecific plumage,
or (4) some combination of the above. Determining how the song dis-
crimination reported here affects gene flow in this species will require
further study, especially in the area of female choice between song forms.

ACKNOWLEDGMENTS

We thank Bill and Martha Lovejoy for their hospitality; Andrew Dewees and Cecil Hallum
for statistical assistance; Monte Thies for help with the figures; Donald Kroodsma, Lrank
B. Gill, Robert B. Payne, J. B. Dunning, Richard Bradley, and Tom C. Grubb, Jr.’s reading
group and an anonymous reviewer for their comments, Lrederick Weinzierl for making his
property available for playback experiments, and the staff of the A. J. Brown (Parker Creek)
Wastewater Treatment Plant for allowing access to Northern Parula habitat.

LITERATURE CITED

Crook, J. R. 1984. Song variation and species discrimination in Blue-winged Warblers.
Wilson Bull. 96:91—99.

Kroodsma, D. E. 1981. Geographical variation and functions of song types in warblers
(Parulidae). Auk 98:743—751.

. 1989. Suggested experimental designs for song playbacks. Animal Behavior 37:

600-609.

and j. R. Baylis. 1982. Appendix: a world survey of evidence for vocal learning

in birds. Pp. 311-337 in Acoustic communication in birds, Vol. 2. (D. E. Kroodsma
and E. H. Miller, eds.). Academic Press, New York, New York.

Regelski and Moldenhauer • NORTHERN PARULA SONG

341

, W. R. Meservey, a. L. Whitlock, and W. M. VanderHaegen. 1984. Blue-winged
Warblers (Vennivora pinus) “recognize” dialects in type II but not type I songs. Behav.
Ecol. Sociobiol. 15:127-131.

Lehman, R. S. 1991. Statistics and research design in the behavioral sciences. Wadsworth
Publishing, Belmont California.

McArthur, P. D. 1986. Similarity of playback songs to .self song as a determinant of
response strength in song sparrows (Melospiza melodia). Anim. Behav. 34:199-207.

Moldenhauer, R. R. 1992. Two song populations of the Northern Parula. Auk 109:215-

222.

Murray, B. G. and E B. Gill. 1976. Behavioral interactions of Blue-winged and Golden-
winged warblers. Wilson Bull. 88:231-254.

Payne, R. B. 1986. Bird songs and avian systematics. Current ornithology 3:87-126.

PiTOCCHELLi, J. 1990. Plumage, morphometric, and song variation in Mourning {Oporoniis
Philadelphia) and MacGillivray’s (O. tolniei) warblers. Auk 107:161-171.

Richards, D. G. 1979. Recognition of neighbors by as.sociative learning in Rufous-sided
Towhees. Auk 96:688-693.

Ritchison, G. 1985. Responses of neighboring conspecifics to typical and atypical songs
of a Rufous-sided Towhee. J. Eield Ornith. 56:280-282.

Thielcke, G. a. 1969. Geographic variation in bird vocalizations. Pp. 311-339 in Bird
vocalizations (R. A. Hinde, ed.). Cambridge University Press, New York, New York.

AND K. Linsenmair. 1963. Zur geographischen Variation des Gesanges des Zilp-
zalps, Phylloscopus collybita, in Mittel- und Siidwesteuropa mit einem Vergleich des
Gesanges des Fitis, Phylloscopus trochilus. J. Ornith. 104:372-402.

Tomback, D. E, D. B. Thompson, and M. C. Baker. 1983. Dialect discrimination by White-
crowned Sparrows: reactions to near and distant dialects. Auk 100:452-460.

Wilson Bull., 108(2), 1996, pp. 342-356

DISPERSAL AND HABITAT USE BY POST-FLEDGING
JUVENILE SNOWY EGRETS AND
BLACK-CROWNED NIGHT-HERONS

R. Michael Erwin,' John G. Haig,'-^ Daniel B. Stotts,' and

Jeff S. Hatfield'

Abstract. — We studied the post-fledging dispersal movements and habitat use of juvenile
Snowy Egrets (Egretta thida) (SNEG) and Black-crowned Night-Herons (Nycticorax nyc-
ticorax) (BCNH) in coastal Virginia using a dye (picric acid) and radiotelemetry. Results
from monitoring radiomarked birds revealed significant differences both years between spe-
cies, with SNEGs dispersing more widely than BCNHs. BCNH juveniles usually remained
south of Delaware, but SNEGs often moved into Delaware and southern New Jersey. The
maximum dispersal distance found for a SNEG was ca 340 km north of the natal colony.
Temporal patterns of movement followed logistic relationships, with rapid initial move-
ments, but relatively few movements after about 2—3 weeks for most birds. Cumulative
distances moved by juvenile SNEGs during August-September differed from 1992 to 1993.
No such year difference was found for BCNHs. Compared to SNEGs, BCNHs used man-
made impoundments relatively more often than natural wetlands; however no quantitative
assessment of habitat preferences could be made. Received 25 May 1995, accepted 9 Dec.
1995.

Little is known about survival, movements, or habitat use during the
post-breeding period for most North American migratory birds (Finch and
Stangel 1993). Even for large, conspicuous species such as colonial wa-
terbirds, few quantitative dispersal data have been published. Numerous
anecdotal reports indicate that many terns, gulls, and wading birds move
northward along the coasts and large interior rivers of the United States
after the nesting season. In some regions, the movement pattern is con-
strained by geography, with waterbirds often following major drainages
(Gill and Mewaldt 1979) regardless of the cardinal direction or following
a peninsula or island archipelago (e.g., in Florida, Powell and Bjork 1990,
Strong and Bancroft 1994 on Great White Herons, Ardea herodias oc-
cidentalis). Studies of individual species revealed some extensive north-
ward dispersal after breeding; e.g.. Black-crowned Night-Herons (Nyctic-
orax nycticorax) in the eastern United States (Bartsch 1952, Byrd 1978),
Little Blue Herons (Egretta caerulea) in Mississippi (Coffey 1943) and
elsewhere in the southeast (Townsend 1931), Great White Herons in Flor-
ida (Powell and Bjork 1990), and Cattle Egrets (Buhulcus ibis) in Africa
(Siegfried 1970).

Studies of dispersal can yield information that has both basic and ap-

‘ National Biological Service, Patuxent Environmental Science Center. Laurel, Maryland 20708.

2 Present addre.ss; 59 Ramona, San Franci.sco California 94103.

342

Envin et al. • DISPERSAL BY YOUNG HERONS

343

plied value. Better estimates of movement rates, habitat use, and mortality
tor most migrant species would assist in developing population viability
analyses that are currently in demand for threatened or endangered spe-
cies. Further, determining levels of site fidelity and identifying habitat
types used during dispersal could have immediate management implica-
tions.

In this study, we followed movements during dispersal of juvenile
Snowy Egrets (hereafter SNEGs) and Black-crowned Night-Herons (here-
after BCNHs) to address the following questions: (1) How do young birds
move during the post-fledging period? Are there differences between spe-
cies and years? (2) How much site fidelity do birds show during repeated
observations? (3) What types of habitats do the species use during this
period, and are there species differences?

STUDY AREA AND METHODS

We studied nesting ecology of two mixed-species colonies near the town of Chincoteague,
Accomack County, Virginia (Fig. 1). The colonies are located in Iva frutescens shrubs along
the margins of a saltmarsh island complex. The Causeway Colony, the largest in Virginia
(Erwin et al., in press), had 230-640 SNEGs and 32-37 BCNHs in 1992-1993 in addition
to large numbers of five other species. The Willis Colony had 300-390 SNEGs and 47-90
BCNHs in the two years and also included other wader species.

The habitat surrounding the colonies consisted of natural Spartina saltmarshes, creeks,
and pannes, as well freshwater ponds and large impounded brackish marshes at the Chin-
coteague National Wildlife Refuge to the east.

Radio telemetry. — When nestlings were approximately two weeks old, we fixed 10 g
radio transmitters to aluminum U.S. Fish & Wildlife Service bands and attached the package
above the tarsometatarsal joint (see Erwin et al., in press). Radios were equipped with a
mortality sensor that resulted in a 50% increase in pulse rate when the radio remained
stationary for 12 h. The transmitters had a range of about 1-2 km on the ground and 5-18
km from aircraft depending on altitude and habitats. The battery life was rated at three
months. We attached the radiotransmitters only to the largest chicks (presumably the A
chick) in the brood to insure that we would have a reasonable sample size for monitoring
post-fledging movements. We used radios only at the Causeway colony. A total of 10 and
19 radiomarked BCNHs fledged in 1992 and 1993, respectively (of totals of 10 and 20
initially marked); for SNEGs, the comparable figures were 19 and 20 (of totals of 20 initially
marked both years).

In late July, when most young were fully feathered and capable of short flights (age 40-
50 d), we conducted radio checks of the colonies 2-5 times weekly to estimate fledging age
and to determine when to begin broader surveillance for dispersing individuals. During the
initial period (1-2 weeks) after fledging, we concentrated our searches by vehicle and boat
in the Chincoteague vicinity, e.specially on the Chincoteague National Wildlife Refuge about
6 km to the east. We began airplane surveys after all the birds had fledged. We made four
flights between 12 August and 24 September in 1992 and six flights between 27 July and
2 October in 1993. We followed a regular search pattern that included the coastal areas of
the DelMarVa peninsula and Delaware Bay. The basic route followed the Chesapeake shore
of the peninsula from Cambridge, Maryland south to Cape Charles, Virginia, continued
north along the Atlantic shore of the peninsula into Delaware Bay up to the Che.sapeake

344

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Lig. 1. Study area in coastal Virginia showing the south end of Assateague Island Na-
tional Seashore, including Chincoteague National Wildlife Refuge, and two colony sites.

and Delaware (C and D) Canal, and finished along the New Jersey shoreline from Salem
to Cape May. We made additional searches 1-2 times each year up the Delaware River to
an area 10 km N of Philadelphia, and along the Atlantic shoreline of New Jersey from Cape
May north to Manasquan. We made two flights in each year on the Chesapeake Bay side
of the peninsula north of Cambridge to the C and D canal. On 1 October 1993, we made
one search along the western shore of Chesapeake Bay from the lower Patuxent River south
to the mouth of the James River in Virginia. Our search route and the limited range of the
transmitters could have introduced some bias because we did not survey (except once) the
western shore of Chesapeake Bay, nor did we effectively survey the interior of the peninsula.
Bunding and time precluded more exhaustive surveys. Nonetheless, species comparisons are
still meaningful.

During aerial surveys, we usually maintained an altitude of 500-1000 m, scanning for
birds with strut-mounted Yagi antennae. Upon detecting a signal, we circled to determine a
location for the individual and identified a landmark (e.g., a named creek or impoundment)
whenever possible. We recorded locations on maps, and later converted them to UTM co-
ordinates. Because the precision of location was probably ±0.5 km, we used the center of

Erwin et cil. • DISPERSAL BY YOUNG HERONS

345

a large impoundment for the UTM coordinate when appropriate. All wetland locations were
identified as either natural or manmade (ponds or impounded marshes).

Colonncirking. — Just before most young SNEGs in the colonies fledged, we conducted a

roundup with 7—12 participants. We held captured birds in wooden duck crates, dyed
them with picric acid on the wings and back, and banded each prior to release. We marked
123 SNEGs in 1992 and 182 in 1993. After marking the birds, we continued to monitor the
colonies for at least two weeks to record survival of radiomarked young and to look for
any moribund or dead dyed birds. We dyed egrets in both the Causeway and Willis colonies.

To stimulate the reporting of colormarked egrets by the public and colleagues in other
natural resource agencies, we sent information packages to national wildlife refuges, parks,
state wildlife management areas, and state wildlife administrations from New Jersey to North
Carolina. We also sent releases to a number of newspapers and magazines in the DelMarVa
region. We solicited the following information: date, time, name of wetland used by the
bird, wetland type, nearest town/village, band colors (if applicable), and size of associated
group. Upon receiving a report, we converted map locations to UTM coordinates and, when-
ever possible, recorded the type of wetland used (impoundment, pond, creek).

Statistical tests. — We used the Multi-Response Permutation Procedure (MRPP) to test for
large-scale differences in spatial distributions between species and years (Biondini et al.
1988). The MRPP compares the distribution of one group of points to another group using
a permutation procedure (Manly 1991) and tests whether the two distributions are identical.
We used this test only on the radio locations determined for birds that had left the Chin-
coteague vicinity. Locations reported for colormarked birds were biased because of the over-
representation of sightings at public-access wetlands (e.g., refuges). These sightings were
valuable, however, because they revealed locations that were not included in our aerial
survey route.

To further investigate movements of the two species, we computed the number of relo-
cations, total number of different wetlands used, and the cumulative distance (km) moved
during the entire tracking period for each individual. Because of some locational imprecision
when conducting aerial searches, we did not record a “new location” from a radiomarked
bird unless it had moved ca 1—2 km from its previous location.

Because sampling intensity and duration differed between years (four aerial surveys over
6 weeks in 1992, six surveys over nine weeks in 1993), we selected a subset of radiomarked
birds for standardized periods. To permit a reasonable time period for dispersal, we used
data for only those individuals (N = 33 out of 48) resighted at least two times and followed
at least 30 days from the time they left the colony site. After examining the raw data plots
for each bird, we applied the logistic model for describing cumulative distances. The ma-
jority of individuals for which we had more than five observations revealed a sigmoidal
(logistic) pattern in distances moved. We fit a separate 3-parameter logistic curve (Draper
and Smith 1981) to each individual (using PROC NLIN procedure in SAS). Erom each of
these logistic curves, we estimated the cumulative distances travelled by Day 5, 15, 30, 45,
and 60. Next, to test for species and year effects, we conducted two-way ANOVAs on the
cumulative distance variables for Days 5, 15, 30, 45, and 60 with Tukey tests to discern
pairwise differences.

We wanted to determine gross habitat preferences for each species by comparing use of
natural wetlands and manmade impoundments or ponds. However, because of our limited
(coastal edge) search route and the difficulty in obtaining wetland area on the local (county)
scales, we were unable to assess habitat preferences. Instead, we compared the number of
individuals of each species found in the two types of habitats, and also the total number of
the two wetland types used (regardless of how many individuals used them).

346

THE WILSON BULLETIN • Vo/. 108, No. 2, June 1996

Table 1

Summary of Numbers of Juvenile Snowy Egrets (SNEG) and Black-crowned Night-
Herons (BCNH) Radiomarked in 1992 and 1993 at Chincoteague, Virginia

Species

Year

No. marked
that fledged

No.

censored'*

Local

relocation’’-'

SNEG

1992

19

2

10

1993

20

6

6

Total uncensored

31

BCNH

1992

10

0

9

1993

20

1

15

Total uncensored

29

“ Not relocated.

Relocated at least once in the Chincoteague-S. Assateague Island block (Chincoteague National Wildlife Refuge) after
fledging from colony.

'Chi-square test for local vs nonlocal relocation by species (using uncensored totals): x’ = 6.54, P = 0.01 1.

RESULTS

Dispersal. — From an earlier analysis, we found that juveniles fledged
and left the colony when they were between 53 and 58 days old (Erwin
et ah, in press). Of the 29 and 39 radiomarked juveniles in 1992 and
1993, respectively, 19 and 21 individuals were recorded at least once in
the wetlands within about 10 km of the nesting colony. Birds usually
returned to the vicinity of the colony site at least once, or moved to the
Chincoteague National Wildlife Refuge, about 6-10 km east of the col-
onies (Table 1). We found more juvenile BCNHs at least once in the local
area (24/29 or 83%) than juvenile SNEGs (16/31 or 52%; = 6.54, P

= 0.011). The remainder (as well as later dispersers from the Chinco-
teague vicinity) dispersed primarily to the north, but also west and south
(Figs. 2, 3, and 4). We never located a few of the fledglings (=censored
observations. Table 1); in 1992, we never located two SNEGs after fledg-
ing, whereas in 1993, we failed to locate one BCNH and six SNEGs. In
1992, SNEGs moved north into Delaware and New Jersey, whereas
BCNHs were confined to Maryland and Virginia. Delaware Bay seemed
to act as a partial barrier, with few birds crossing the bay into New Jersey.
We received 99 reports of colormarked individual SNEGs, the majority
from the Chincoteague National Wildlife Refuge. These sightings added
a few new locations in 1992, with the northernmost bird reported from

Fig. 2. Locations of radiomarked Snowy Egrets and Black-crowned Night-HerOns in
1992 during August-September.

Erwin et al. • DISPERSAL BY YOUNG HERONS

347

348

THE WILSON BULLETIN • VoL 108, No. 2, June 1996

Fig. 3. Locations of radiomarked Snowy Egrets and Black-crowned Night-Herons in
1993 during late July-early October. All locations outside the Chincoteague block (nesting
colony locale) are shown.

Erwin et al. • DISPERSAL BY YOUNG HERONS

349

Fig. 4. Locations of dye-marked Snowy Egrets in 1992 and 1993. If two different sym-
bols abut each other, the same location is indicated.

350

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 2

Summary of Multi-response Permutation Procedure"’ Comparing Spatial Patterns of
Dispersing Snowy Egrets (SNEG) and Black-crowned Night-Herons (BCNH), 1992

AND 1993

Comparison

P Value

Radiomarked:

SNEG 92 vs 93

<0.001

BCNH 92 vs 93

0.02

SNEG vs BCNH (1992)

0.01

SNEG vs BCNH (1993)

<0.001

Dye-marked:

SNEG 92 vs 93

0. 18 ns

Dye vs radiomarked (SNEG):

1992

0.002

1993

0.09 ns

“See Biondini et al. 1988.

the Tinicum National Wildlife Refuge south of Philadelphia about 210
km from the nesting colony (Fig. 3).

In 1993, both species generally dispersed slightly greater distances (cf
Figs. 2, 3, and 4) and, the number of “lost” (censored) birds also in-
creased. The radiomarked BCNHs moved north into Delaware, and more
SNEGs were found farther north in New Jersey (Figs. 2 and 3). In 1992,
no dyed SNEGs were reported west of the Chesapeake Bay, but three
were seen in 1993 (Eig. 4). A total of 141 colormarked SNEGs were
reported in 1993, mostly from the Chincoteague area. Also, we received
a report of a dyed SNEG from far northern New Jersey in Lyndhurst,
near the extensive tidal Hudson River wetlands known as the Meadow-
lands (Fig. 4). This location is 342 km north of the nesting colonies. In
contrast, the longest distances recorded for radiomarked BCNHs were
only 85 km in 1992 and 102 km in 1993.

Spatial pattern. — MRPP test results revealed significant differences {P
< 0.05) for comparisons between years within species and between spe-
cies within years for the radiomarked birds (Table 2). Dye-marked SNEGs
showed no yearly difference, but their distribution was significantly dif-
ferent from radiomarked SNEGs in 1992 (Table 2). Although the MRPP
test reveals that both species and yearly “geographic centroids” differed,
the test does not indicate how the patterns differed.

Temporal pattern of dispersal. — We found a high degree of variability
among individual SNEGs and BCNHs during both years of the study; A

Ei-n’in et al. • DISPERSAL BY YOUNG HERONS

351

Table 3

Summary of Dispersal Movement of Juvenile Snowy Egrets (SNEG) and Black-

CROWNED Night-Herons (BCNH) during Late Summer, 1992 and 1993

Group Year N“

Day"

Mean cum.
distance‘s
(±SD)

Mean Mean no.

total different

relocations‘^ locations'^

SNEG 1992 11

5

37.2 (30.2)

15

50.6 (35.6)

30

52.5 (39.4)

45

52.5 (39.4)

60

52.5 (39.4)

Total

4.3 3.9

SNEG 1993 8

5

97.0 (66.7)

15

121.3 (61.5)

30

121.3 (61.5)

45

121.3 (61.5)

60

124.9 (53.9)

Total

6.1 4.4

BCNH 1992 6

5

17.5 (14.2)

15

18.4 (13.3)

30

21.3 (1 1.4)

45

28.1 (14.4)

60

44.2 (37.3)

Total

16.2 5.0

BCNH 1993 8

5

40.6 (47.2)

15

44.1 (45.1)

30

44.3 (44.9)

45

44.8 (44.4)

60

44.8 (44.4)

Total

4.3 3.0

^ Only birds followed for >30 days, with two or more locations, were included in the analysis.

•’To adjust for sampling intensity differences between years, values for specific numbers of days since colony departure
(i.e., 5. 15, 30, 45. and 60) were predicted from the logistic curve for each bird.

' Means of cumulative distances moved by individuals from the colony site (origin), based on predictions from individual
logistic curves.

Mean number of relocations per individual during ground and air surveys in 1992 and 1993.

' Mean number of different wetland locations per individual during surveys. Because of locational uncertainty from aerial
surveys, a “move” to a new location required an estimated 1-2 km actual relocation.

few individual herons and egrets made

return trips

to the Chincoteague

vicinity after short dispersals.

, while most moved

increasing distances

from the colony vicinity. The general pattern was for an initial move to

local wetlands near the colony site and at the Chincoteague NWR. From
here, a rapid movement occurred during the next 1—2 weeks, with less
movement afterward (Table 3, Fig. 5). In both years, individuals of both
species moved relatively little after about 30 days, with many individuals

352

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Fig. 5. Examples of cumulative distances moved over time by individual Snowy Egrets
since fledging (1992 solid line, 1993 dotted line). Predicted logistic curves were generated
using a three-parameter logistic equation. Symbols represent actual location distances for
two individuals.

remaining in the same wetland complex (see last column in Table 3).
Extreme cases in 1992 included a BCNH that spent 73 days in an im-
poundment at Chincoteague NWR from August 3 to October 14 (24 total
relocations) and a SNEG that spent 24 days in a marsh also at Chinco-
teague NWR (9 relocations). In 1993, a BCNH spent 57 days (July 21
to September 15) at the same location as the previous year (13 relocations)
and a SNEG spent 53 days from August 10 to October 1 at Eortesque
Wildlife Management Area on the Delaware Bay shore of New Jersey.

We found several significant differences in the cumulative distances
moved by dispersing birds (Table 4). Eor SNEGs, individuals averaged
more than twice the cumulative distance in 1993 than in 1992, whereas
BCNHs showed no year effect. Also, in 1993, we found a significant
species effect (Tables 3 and 4).

Habitat use. — More birds of both species used natural wetlands than
manmade ones (Table 5); however, habitat availability could not be as-
sessed for the study area (see Methods). BCNHs were associated with
manmade wetlands relatively more often than SNEGs in both years (Table
5). Because manmade impoundments only comprise a small fraction (15%
in Delaware) of the total area of coastal estuarine emergent wetland in

Erwin et al. • DISPERSAL BY YOUNG HERONS

353

Table 4

Differences in Cumulative Dispersal Distance between Snowy Egrets (SNEG) and
Black-crowned Night-Herons (BCNH) in 1992 and 1993“

Comparison

Day

Difference
between means
(km)

Significance

level

(P < 0.05)

SNEG, 1992-1993

5

59.8

*

15

70.7

*

30

68.9

*

45

68.9

*

60

72.4

*

BCNH, 1992-1993

5

23.1

ns

15

25.7

ns

30

23.1

ns

45

16.7

ns

60

0.6

ns

1992, SNEG-BCNH

5

19.7

ns

15

32.2

ns

30

31.2

ns

45

24.4

ns

60

8.3

ns

1993, SNEG-BCNH

5

56.4

ns

15

77.2

*

30

77.0

*

45

76.6

*

60

80.1

*

"Two-way ANOVAs performed on cumulative distances at each predicted day, with Tukey multiple comparison test
results. The corresponding one-way reparameterized ANOVAs with Tukey tests were used for interaction terms.

Table 5

Habitat Use of Young Radiomarked BCNHs and SNEGs in the Delmarva Region
DURING THE DISPERSAL PERIOD (AUGUST-SePTEMBER, 1992 AND 1993)

Number using habitat:"

Species

Natural

Impounded

Total

BCNH

1992

14(11)

11 (9)

25 (20)

1993

18 (15)

7 (10)

25 (25)

SNEG

1992

30 (22)

4(7)

34 (29)

1993

27 (12)

4(6)

31(18)

“ First number listed
number of individual

is the number of locations in
birds using these habitats. An

that habitat type with one or more individuals; in parentheses is the
individual may be counted more than once if it was found in more

than one wetland.

354

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

our region, our results suggest that BCNHs may show some preference
for manmade over natural wetlands.

DISCUSSION

Dispersal pattern. — The MRPP test results suggest that juvenile
SNEGs and BCNHs disperse in somewhat different ways, with SNEGs
more free ranging (see Eigs. 4-5). The greater tendency for BCNHs to
remain in the local region of their natal colony is reflected by the high
frequency with which we located individual BCNHs within the Chinco-
teague region in both years (Table 1).

Reports of the dye-marked SNEGs indicated that our radiomarking data
were probably underestimating the overall movements of birds. At least
one dye-marked SNEG was found 342 km north of the natal colony
(about 130 km farther than the farthest radiomarked individual and well
beyond the 240 km northward extent of our aerial surveys). Reports of
dye-marked SNEGs also were reported from the southwestern shore of
the Chesapeake Bay in 1993, an area beyond our normal aerial search
route. Perhaps the censored birds rapidly left our DelMarVa search area.
However, radiotelemetry studies are often plagued by a failure to distin-
guish radio failures from movements beyond the survey area.

The results from our dyed egrets and those from studies in southern
New Jersey (W. Crans, Rutgers Univ., unpubl. data), from which SNEGs
moved into southern New England after dispersal, suggest that SNEG
dispersal is greater northward than our telemetry results suggests.

Both species showed a rapid initial movement away from the natal
colony area, but made relatively restricted movements thereafter, resulting
in a logistic relationship for cumulative distance traveled. Such a pattern
may result from local depletion of prey resources during the long nesting
period in the colonies. Others have noted a similar movement pattern.
Powell and BJork (1990) found that most radiomarked juvenile Great
White Herons in Elorida Bay rapidly moved north from the Bay to central
Florida in summer. Strong and Bancroft (1994) described a rapid (>20
km the first 10 days) northward dispersal of young White-crowned Pi-
geons {Columba leucocephala) from the Florida Keys. Immature Spanish
Imperial Eagles {Aquila adalberti) also make long solo flights soon after
leaving the natal area (Gonzalez et al. 1989).

Variable water conditions may have contributed to yearly differences
in dispersal by juvenile SNEGs. In 1993, the eastern shore region of
Virginia suffered from a severe drought (B. Truitt, pers. comm.). Unlike
the wetter conditions in 1992, by late July and early August 1993 during
the wading bird dispersal period, most of the impoundments at the Chin-
coteague NWR were dry. This may explain why the cumulative distances

£m-//i et al. • DISPERSAL BY YOUNG HERONS

355

of radiomarked SNEGs averaged more than twice as far in 1993. Snowy
Egrets seem to search widely for other wading birds and often concentrate
in “drawdowns” in impoundments and tidal pools (Kushlan 1978, Erwin
1983). We received a report of seven dyed SNEGs in one salthay im-
poundment in New Jersey in August 1993 (W. Grans, unpubl. data).
BCNHs, in contrast, are more generalized in feeding habitat use (Davis
1993). Because they may use freshwater ponds and creeks as well as tidal
areas, they probably do not have to move as far to find adequate food.

Habitat use. — Habitat selection during the post-breeding dispersal
phase for juveniles could not be fully evaluated since we lacked quanti-
tative estimates of wetland types for our limited survey region. The gen-
eral pattern seemed to show a tendency for BCNHs to use impounded
areas, including both farm ponds and large wildlife impoundments, to a
greater relative extent than did SNEGs.

Data on movements of juvenile Great White Herons in central-south
Florida (Powell and Bjork 1990) suggested a seasonal component to hab-
itat selection. Early and late in the season, birds seemed to settle into
human-modified habitats towards the central part of the state, whereas
during the middle period, they traveled shorter distances and settled into
natural marshes in the southern Everglades. Gill and Mewaldt (1979)
described the dispersal patterns of SNEGs and BCNHs leaving a south
San Francisco Bay heronry. Both species followed nonrandom patterns,
with BCNHs following the Bay edge and tributaries, while SNEGs mostly
moved <50 km to the “nearest appropriate habitats.” The farthest dis-
tance travelled by a SNEG in their study was 209 km from the natal
colony. Neither of these studies addressed the issue of habitat use versus
availability.

For young inexperienced juveniles, a critical ecological factor may be
locating a concentrated source of prey, whether it be a small sewage
treatment lagoon, a large impoundment, or a natural drying freshwater
slough (Kushlan 1978, Powell and Bjork 1990). Site tenacity seemed to
be high for many radiomarked individuals, with an average of only 3-5
different locations per species-year group (Table 3). Variation in the
amount of time specific sites were used was enormous, ranging from one
to 73 days at Chincoteague NWR.

ACKNOWLEDGMENTS

We thank John Schroer and his staff at Chincoteague National Wildlife Refuge, for the
field logistical support, and the many volunteers (including Jason and Lindsay Erwin and
David Jachowski) who assisted. W. Link helped with statistical advice, K. Fontaine assisted
in preparing the manuscript, and M. Banker and K. Boone provided graphics help. C. Blem,
T. Custer, C. Henny, J. Smith, and B. Watts provided useful comments on earlier drafts of
the manuscript.

356

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

LITERATURE CITED

Bartsch, P. 1952. A note on the first bird banding in America. Bird-Banding 33:59-60.

Biondini, M. E., P. W. Mielke, Jr., and K. J. Berry. 1988. Data-dependent permutation
techniques for the analysis of ecological data. Vegetatio 75:161-168.

Byrd, M. A. 1978. Dispersal and movements of six North American ciconiiforms. Pp. 161-
185 in Wading birds (A. Sprunt, IV, J. C. Ogden, and S. Winckler, eds.) National
Audubon Society Rept. No. 7, New York, New York.

Coffey, B. B., Jr. 1943. Post-juvenal migration of herons. Bird-Banding 14:34-39.

Davis, W. E. 1993. Black-crowned Night-Heron. Pp. 1-20 in The birds of North America,
No. 74 (A. Poole and P. Gill, eds.) The Academy of Natural Sciences, Philadelphia and
The American Ornithologists Union, Washington, D.C.

Draper, N. R. and H. Smith. 1981. Applied regression analysis. 2nd edition. J. Wiley and
Sons, Inc., New York, New York.

Erwin, R. M. 1983. Peeding habitats of nesting wading birds: spatial use and social influ-
ences. Auk 100: 960—970.

, J. G. Haig, D. B. Stotts, and J. S. Hatfield. 1996. Growth, nest success, and

survival of Black-crowned Night-Heron {Nycticorax nycticorax) and Snowy Egret
(Egretta thula) chicks in coastal Virginia. Auk (in press).

Pinch, D. M. and P. W. Stangel. 1993. Status and management of neotropical migratory
birds. Gen. Tech. Rep. RM-229, Rocky Mountain Exper. Sta., U.S. Forest Service, Ft.
Collins, Colorado.

Gill, R., Jr. and L. R. Mewaldt. 1979. Dispersal and migratory patterns of San Francisco
Bay produced herons, egrets, and terns. North Amer. Bird Bander 4:4-13.

Gonzalez, L. M., B. Heredia, J. L. Gonzalez, and J. C. Alonso. 1989. Juvenile dispersal
of Spanish Imperial Eagles. J. Field Ornithol. 60:369—379.

Kushlan, j. a. 1978. Feeding ecology of wading birds. Pp. 249-298 in Wading birds (A.
Sprunt, IV, J. C. Ogden, and S. Winckler, eds.) National Audubon Society Rept. No. 7,
New York, New York.

Manly, B. F. T. 1991. Randomization and Monte Carlo methods in biology. Chapman and
Hall, New York, New York.

Powell, G. V. N. and R. Bjork. 1990. Studies of wading birds in Florida Bay: a biological
assessment of the ecosystem. National Audubon Society report to the E. Ordway Dunn
Foundation, National Audubon Soc., Tavernier, Florida.

Siegfried, W. R. 1970. Mortality and dispersal of ringed Cattle Egrets. Ostrich 41:122-
135.

Strong, A. and G. T. Bancroft. 1994. Postfledging dispersal of White-crowned Pigeons:
implications for conservation of deciduous seasonal forests in the Florida Keys. Con-
serv. Biol. 8:770—779.

Townsend, C. W. 1931. The post-breeding northern migration of North American herons.
Proc. Intern. Ornithol. Congr. 7:366-369.

Wilson Bull.. 108(2), 1996, pp. 357-368

NEST-SITE SELECTION OF RED-SHOULDERED AND
RED-TAILED HAWKS IN A MANAGED FOREST

Christopher E. Moorman''^ and Brian R. Chapman'

Abstract. — We compared nest-site macro- and microhabitat selection of Red-shouldered
(Biiteo lineotus) and Red-tailed hawks (B. jamaicensis) and examined potential relationships
between habitat selection and nest success in a managed forest in central Georgia. We
located 12 Red-shouldered and 10 Red-tailed hawk nests during the 1994 breeding season.
Circular plots (1 km-) were mapped around each hawk nest and 100 random points, and
selected macrohabitat characteristics within the plots were measured and compared. Red-
shouldered Hawk nest-site macrohabitat was characterized by significantly more bottomland
hardwood habitat, less older age (>50 yr) pine habitat, and larger nest-site stands than
random plots. Red-tailed Hawk nest plots contained significantly more agriculture habitat,
more young (6-20 yr) pine habitat, less upland hardwood habitat, less total amount of edge,
fewer number of stands, and larger average stand size than random plots. Red-shouldered
Hawk nest sites (0.04 ha) had more large (>69 cm DBH) trees and lower percent total
canopy cover than random points. Red-tailed Hawk nests were placed close to habitat edges
and openings in the canopy, and nest sites had taller trees, larger (>69 cm) trees, and greater
percent understory cover than random points. Successful Red-tailed Hawk nests were placed
in shorter trees than unsuccessful nests. On the study site, large floodplain forests offering
mature trees were important to breeding Red-shouldered Hawks, and mature pine forest
edges near openings created by silvicultural and agricultural practices were important to
breeding Red-tailed Hawks. Received 7 April 1995, accepted 15 Nov. 1995.

Forest management practices in Georgia often are directed towards in-
creasing the production of pine timber. Silvicultural treatments result in
forest modifications that include alterations in horizontal and vertical
structural diversity, stand diversity, size class distribution, and vegetative
species composition (Nelson and Titus 1988). Hardwood species are usu-
ally removed from the overstory in managed pine stands. The remaining
stands of pine probably provide little habitat that is suitable for raptors
(Edwards 1978). Declines in Red-shouldered Hawk (Buteo Uneatus) pop-
ulations elsewhere in its range have been attributed to alterations of nest-
ing habitat, especially riparian habitat, and replacement by the Red-tailed
Hawk (B. jamaicen.si.s) which is more xeric-adapted (Stewart 1949 Henny
etal. 1973).

Nest-site selection of the Red-shouldered and the Red-tailed hawk sel-
dom has been studied at the landscape level. Few studies have described
nest-site habitat selection of sympatric populations of the Red-shouldered
and the Red-tailed hawk (Titus and Mosher 1981, Bednarz and Dinsmore
1982), and none have described nest-site selection for either of the two

^ Daniel B. Warnell School of Forest Resources, The Univ. of Georgia. Athens. Georgia 30602-2152
'Present address: Dept. Forestry. 261 Lehotsky Hall, Clemson Univ., Clemson, South Carolina 29634.

357

358

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

species in the southeastern United States. We conducted a study in a
managed forest (1) to determine whether Red-shouldered and Red-tailed
hawks establish nest sites in proximity to specific micro- and macrohabitat
types; (2) to determine how silvicultural practices could create or modify
these specific habitat types; (3) to describe potential relationships between
reproductive success and nesting habitat; and (4) to determine whether
habitat partitioning exists between the two species.

STUDY AREA AND METHODS

Field investigations took place at Bishop E Grant Memorial Forest, a 5718 ha wildlife
management area (WMA) owned by the Univ. of Georgia School of Forest Resources and
operated in cooperation with the Georgia Dept, of Natural Resources. The WMA is located
in Putnam County approximately 14.5 km north of Eatonton, GA. The property lies within
the southern Piedmont physiographic province, a region of broad, gently sloping topography
with occasional steep or strongly sloping terrain around the major drainage basins. A ma-
jority of the existing upland habitat types are dominated by loblolly pine {Pinus taeda).
Present silvicultural treatments in the pine forests range from thinning and prescribed burn-
ing to clear-cutting and replanting. Bottomland hardwood habitats dominate along the pro-
perty’s three largest creeks (Glady Creek, Big Indian Creek, and Little River), and upland
hardwood habitats exist in some areas along the drainage basins associated with these creeks.
The Univ. of Georgia Agricultural Experiment Station grazes cattle on several large pastures
that lie within the WMA.

We used a number of techniques to locate active Red-shouldered and Red-tailed hawk
nests during the 1994 breeding season. We searched for old nests during the preceding
winter months and later returned to check for signs of activity. Because taped calls of
conspecific vocalizations have proven effective in locating nesting raptors (Rosenfield et al.
1988, Kimmel and Yahner 1990), Red-shouldered Hawk alarm calls were broadcast from a
cassette recorder near potential nest sites. Vocalizations were played for 15-s periods dis-
tributed evenly over 5 min. Broadcasting was repeated every 10-20 min while an investi-
gator moved through the wooded area (Mosher et al. 1990). When Red-shouldered Hawks
were present, they normally responded to the vocalizations either by calling or flying towards
the broadcasts. We made extensive nest searches in areas where birds responded. Because
Red-tailed Hawks are relatively conspicuous visually, we located nests by searching areas
where birds were seen perched or soaring. To prevent bias, nest searches also were conducted
in all forested stands (trees >20 yrs) in areas where birds were not seen or heard. Stands
were searched on foot by walking transects that were spaced to permit observation of most
tree crowns. These searches began in late April and were continued through mid-June.
Occupied nests of both species were monitored every 7-10 days and outcomes were re-
corded. Nests that fledged at least one young were considered successful.

Macrohabitat analy.sis. — We classihed habitat types on a digital database developed with
the geographical information system (GIS) software package ARC/INFO (Environmental
Systems Research Institute 1987). Pine habitats were separated into hve types based on age
(numbers indicate age of forest in years): <6PINE; 6-20PINE; 21-30PINE; 31-50PINE;
>50PINE. Three additional habitats types included in the analysis were bottomland hard-
woods (BOTTOM), upland hardwoods (UPLAND) and agricultural land such as pastures
and Helds (AGR). Once nests were located and verified as occupied, they were recorded on
the GIS database within the habitat types in which they occurred. To characterize available
habitat, we selected 100 random points from a UTM coordinate grid using a random number

Moorman and Chapman • HAWK NEST-SITE SELECTION

359

geneiator. Since hawks require large trees as nest substrates, only random points that fell
within forested habitat greater than 20 years old were selected for analysis.

When analyzing macrohabitat preference, it is important to know the scale at which
selection occurs. Sedgwick and Knopf (1992) analyzed nesting habitat within three concen-
tric circles of increasing size using the nest site as the center. Lehmkuhl and Raphael (1993)
also assessed owl habitat pattern within three concentric circles centered on foraging loca-
tions. Using a similar technique and a GIS, we mapped concentric circles of increasing size
around each nest site and random point. The central circle, or mesoplot, was 1 km^ (radius
= 564 m) which was approximately equal to the smallest Red-shouldered Hawk home range
(D. L. Howell, pers. comm.). We selected the smallest home range size to minimize sampling
outside of territories (Sedgwick and Knopf 1992). The innermost circle (radius = 399 m),
or microplot, had an area approximately half that of the home range. An outer circle (radius
— 798 m), or macroplot, encompassed twice the area of the mesoplot. Eor comparison and
because home range size of Red-tailed Hawks in the Southeast is unknown, circles of the
same size were generated around Red-tailed Hawk nests.

We measured macrohabitat characteristics of nest plots and random plots with the GIS.
Area of each habitat type, amount of edge (TOTEDGE), average patch size (AVGSIZE),
and number of patches within circles (#STAND) were compared for each circle size. The
patch size containing the plot center (STSIZE) for each nest site and random point also was
tested for differences.

Microhabitat analysis. To quantify available habitat, we used the same random points
as for the macrohabitat analysis. Field locations of the UTM coordinates selected were found
using a global positioning system. At the end of the nesting season (July-Aug.), we mea-
sured nest-site vegetation using a modification (Noon 1981) of the James and Shugart (1970)
technique. In this study, we defined the nest site as a 0.04 ha circular plot with the nest tree
as its center. Distance to water (DISTWAT), distance to a road (DISTROAD), distance to
a break in the overstory canopy (DISTOP), and distance to a change in habitat type (DIS-
TEDGE) were determined with the GIS. Except for the nest-tree-specific variables (Table
1), the sampling was the same at random points as at nest sites. A spherical densiometer (4
samples per site) was used to measure percent canopy cover and an ocular tube (20 samples
per site) was used to determine percent ground cover and percent understory cover. Heights
of four dominant trees in the 0.04 ha plot were determined using a clinometer and their
average was used as the site canopy height. For each species, only random points that fell
within the same habitat as nests were used in comparative analysis. Limiting the random
sites prevented comparing nest sites to habitats where hawks were known not to nest.

Statistics. — We used Wilcoxon rank-sum tests to compare nest sites and random points.
Red-shouldered and Red-tailed hawk nest sites, and successful and unsuccessful nest sites
of each species. Nonparametric analyses were used because some sample sizes were small
and most data were non-normally distributed. Since nonparametric statistics were used, non-
normal, percentage, and count data did not require transformation (Zar 1974). All statistical
analyses were performed using the Statistical Analysis System (SAS Institute Inc. 1982).

RESULTS

Macrohahitat analysis. — Because plot scale had little effect for the two
species (Moorman 1995), we used mesoplot (1 km^) values for all anal-
yses. Ten Red-shouldered Hawk nests were located in bottomland hard-
wood habitat and two were found in upland hardwood ridges bordering
bottomland forest. Two of these nests were located in areas where hawks
previously were not seen or heard. Red-shouldered Hawk nest sites were

360

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 1

List of Additional Nest Site and Random Point Variables and Explanation of Their

Mnemonics

Variable

Description

3-8cm(#)

Number of stems within the 0.04 ha plot with DBH between

3 and 8 cm

9-15cm(#)

Number of stems within the 0.04 ha plot with DBH between

9 and 15 cm

16— 23cm(#)

Number of stems within the 0.04 ha plot with DBH between

16 and 23 cm

24-38cm(#)

Number of stems within the 0.04 ha plot with DBH between
24 and 38 cm

39-53cm(#)

Number of stems within the 0.04 ha plot with DBH between
39 and 53 cm

54-69cm(#)

Number of stems within the 0.04 ha plot with DBH between
54 and 69 cm

>69cm(#)

Number of stems within the 0.04 ha plot with DBH greater
than 69 cm

BASALAREA

Total basal area per hectare

SHRUBDEN

Estimate of the number of shrubs per hectare

CANHT(m)

Average height of four dominant trees in the 0.04 ha plot

GRCOVER(%)

Percent ground cover determined with an ocular tube

CANCOVER(%)

Percent total canopy cover determined with a spherical den-
siometer

UNCOVER(%)

Percent understory cover determined with an ocular tube

NESTHT(m)

Height of the nest determined with a clinometer

NETREEHT(m)

Height of the nest tree determined with a clinometer

NETREEDBH(cm)

DBH of the nest tree

PER.NESTHT

Percent of the nest height of the nest tree height

NEARNGHBR(m)

Distance to nearest nest of the same species

located in larger stands (x = 194.15 ha) than random points (x = 63.8
ha) (Table 2). Nest plots had significantly more BOTTOM (x = 28.43
ha) and less >50PINE (x = 16.61 ha) than random plots (x = 9.08 ha
and 33.94 ha, respectively).

Eight Red-tailed Hawk nests were in >50PINE habitat, one was in 30-
50PINE habitat, and one was located in a loblolly pine within UPLAND
habitat. One of the ten nests was located in an area where Red-tailed
Hawks previously had not been observed. Nest plots had more AGR (x
= 31.71 ha) and less UPLAND (x = 16.92 ha) and 6-20PINE (x = 7.87
ha) than random plots (x = 7.49 ha, 23.79 ha, 13.79 ha, respectively). In
addition. Red-tailed Hawk nest plots had less TOTEDGE, less #STAND,
and a greater AVGSIZE (Table 2).

Red-shouldered and Red-tailed hawk nest-site macrohabitats were sep-

Moorman am! Chapman • HAWK NEST-SITE SELECTION

361

Table 2

Mean ± one Standard Error of Mesoplot Variables Measured at Red-shouldered
Hawk Nests, Red-tailed Hawk Nests, and 100 Random Points

Red-shouldered Red-tailed Random

Variable Hawk (1.0 km-) Hawk ( 1 .0 km-) (1.0 km^)

STSIZE(ha)

194.15

-F

42.32-

39.66

4-

11.13

63.80

-F

7.81

BOTTOM! ha)

28.43

-+■

3.21-

4.07

-F

2.06

9.08

-F

1.35

UPLAND! ha)

26.67

-h

3.24*=

16.92

-F

2.50”

23.79

-F

0.89

AGR!ha)

4.89

H-

L53‘^

31.71

-F

7.62*’

7.49

-F

1.10

<6PINE!ha)

4.18

-F

2.15

8.29

-F

3.84

5.26

-F

0.87

6-20PINE!ha)

13.78

-F

3.27

7.87

-F

4.63*’

13.79

-F

1.10

21-30PINE!ha)

2.03

-F

1.15

0.37

-F

0.36

1.59

-F

0.44

31-50PINE!ha)

3.43

-F

2.22

5.04

-F

2.91

5.07

-F

0.83

>50PINEtha)

16.61

-F

3.18“

25.74

-F

5.06

33.94

-F

1.66-

TOTEDGE! km)

25.00

-F

1.39^

20.10

-F

1.77*’

25.40

-F

0.39

#STAND

25.58

-F

3.01

17.30

-F

2.53*’

22.96

-F

0.57

AVGSIZE!ha)

4.46

-F

0.45

7.45

-F

1.42*’

4.60

-F

0.14

N

12

10

100

® Significant differences {P ^ 0.05) between Red-shouldered Hawk nest plots and random plots according to Wilcoxon
rank-sum tests.

^Significant differences {P < 0.05) between Red-tailed Hawk nest plots and random plots according to Wilcoxon rank-
sum tests.

•= Significant differences {P ^ 0.05) between Red-shouldered Hawk and Red-tailed Hawk nest plots according to Wilcoxon
rank-sum tests.

arated by significant differences in several variables (Table 2). Red-tailed
Hawk macrohabitat was characteristic of upland habitat; Red-shouldered
Hawk macrohabitat represented bottomland habitat. Red-shouldered
Hawk nests were located in larger stands (T = 194.15 ha) and nest plots
had more BOTTOM {x = 28.43 ha), more UPLAND (T = 26.67 ha), less
AGR {x = 4.89 ha), and more TOTEDGE (x = 25.0 km) than Red-tailed
Hawk nest plots (L = 39.66 ha, 4.07 ha, 16.92 ha, 31.71 ha, 20.1 km,
respectively).

We found no significant differences between successful and unsuc-
cessful nesting macrohabitats for Red-shouldered or Red-tailed hawks.
Four of 12 Red-shouldered Hawk nests failed to fledge at least one young.
One nest was abandoned during incubation and one was damaged during
a severe storm. The causes of nest failure were unknown for the other
two. Five of the 10 Red-tailed Hawk nests were successful in fledging
young. Causes of nest failures were not known.

Microhahitat analysis. — One Red-shouldered Hawk pair nested in a
loblolly pine, but the remaining 1 1 pairs placed nests in deciduous trees.
Four were in American sycamores (Platanus occidentalis), two in sweet-
gums (Liquidambar styraciflua), two in southern red oaks {Quercus fal-

362

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

cata), one in a water oak {Q. nigra), one in a green ash {Fraxinus
pennsylvanica), and one in an eastern cottonwood (Populus deltoides).
Red-shouldered Hawks only nested in hardwood habitat, so only random
points located in hardwood habitat were used in comparative analyses.
Red-shouldered Hawk nest sites had significantly more large trees (>69
cm) and lower CANCOVER than other sites within hardwood habitat
(Table 3).

All Red-tailed Hawk pairs nested in loblolly pines. Nest sites were
located in either pine or upland hardwood-pine habitat, so only random
points located in these habitat types were used in statistical comparisons.
Red-tailed Hawks built their nests significantly closer to edges and closer
to openings in the canopy than random points (Table 3). Nest sites had
greater UNCOVER and more understory trees (9-15 cm) than random
points. Nest sites had more large trees (>69cm) and more tall trees
(CANHT) than other points within potential nesting habitat.

Red-shouldered Hawk nests were placed closer to water {x = 68 m)
and farther from edges (x = 57.3 m) and openings (x = 139.1 m) than
Red-tailed Hawk nests (x = 355.3 m, 14.9 m, 19.6 m, respectively). Red-
shouldered Hawk nest sites had greater CANCOVER and lower SHRUB-
DEN than Red-tailed Hawk nests (Table 3). We also compared nest-tree
variables and nearest neighbor distances between the two species (Table
4). Red-tailed Hawks nested higher (NESTHT) and higher in the tree
(PERNESTHT) than Red-shouldered Hawks. Nearest neighbor distances
ranged from 448 m to 4195 m for Red-shouldered Hawks and from 1389
m to 2971 m for Red-tailed Hawks. Red-shouldered Hawk nearest neigh-
bor distances generally were smaller (x = 1322 m) than those of Red-
tailed Hawks (x = 1827 m), but one pair nested in a solitary location
(4195 m from its nearest neighbor). When nest-site habitat variables were
compared, successful Red-tailed Hawk nests were determined to be in
significantly (P = 0.02) shorter trees (x = 32.0 m, N = 5) than unsuc-
cessful (x = 37.9 m, N = 5) nests.

DISCUSSION

The effects of plot scale were minimal for both Red-tailed and Red-
shouldered hawks. Degree of differences gradually decreased with in-
crease in circle scale, but some differences existed at the largest scale.
Because the size of the mesoplot circle was based on actual Red-shoul-
dered Hawk home ranges, final results and discussions of management
implications may be most appropriately based at this scale; Red-tailed
Hawk home range analysis is needed for the heavily forested Southeast.
Once home range size is determined, applicability of our plot sizes could

Moorman and Chapman • HAWK NEST-SITE SELECTION

a:

E

X

H

Q

Z

<

CO

W

H

(y5

00

LU

Z

<

X

Q

LU

U

<

H

I

Q

LU

o:

Q

z

<

w

Q

J

D

O

X

Q

uj

Qi

H

Z

o

CL,

r<i

UJ

-j

DC

<

H

J

D O
00 Z

< 5
s z

^ g

CO 00

UJ w

Qi

O

2 U
<

<

t

5

<

I

o

cc

O

u.

O

0^

O

a:

ai

UJ

Q

cc

<

Q

Z

<

(—

C/3

UJ

z

O

+ 1

z

<

UJ

s

Eos

o'«

QSZ

s

X3

X5

<N

X

X

X

NO

<N

X

NO

T3

II

CN

cn

ON

ON

ID

CN

q

q

o

O

o

NO

<N

(N

(N

d

d

d

d

d

(D

d

d

d

2

■o

Z

+ 1

+ 1

+ 1

+ 1

+ 1

+1

+1

+1

+1

+1

+1

+1

+ 1

+1

+1

+1

+1

U

NO

O'

ro

ON

(N

m

NO

On

(N

o

q

o

o

?3

X

iri

ON

Os

d

iri

ro

(N

d

d

ON

(N

(N

*— i

IT)

NO

— H

CN

NO

m

00

LTl

oc

o ^

liz

2 II
oZ

^ S

LJ r3

CliX

■o

On

ON

•o

q

T3

q

•o

q

<D

q

in

CD

(N

o

q

oo

On

ID

m

o

tN

o

tn

o

CN

d

d

d

d

d

d

NO

d

d

d

d

+ 1

+1

+ 1

+ f

+ i

+1

+1

+1

+1

+1

+1

+ 1

+ 1

+1

+1

+1

+1

o

(N

q

o

q

in

o

q

q

q

q

o

d

On

00

r-'

ON

d

CD

(N

d

d

CD

d

ro

On

NO

<N

CD

ON

CN

00

00

NO

CN

in

(N

o\

•q

00

•o

17. 2>

q

r--

CN

00

CN

D

NO

m

o

o

o

(N

O

CN

NO

(N

d

d

d

d

d

d

d

d

d

d

+ 1

+ 1

+ 1

+ 1

+ f

+1

+1

+1

+1

+ 1

+1

4-1

-fl

-rl

+ 1

+1

+1

00

O)

00

■q

CD

— -

CN

00

(N

q

q

■q

q

q

q

CD

CD

•*=t

CD

ON

ID

(N

d

d

CD

CD

d

ON

NO

o

00

CN

CD

CN

CD

m

ON

in

ro

Dj

CN

NO

00

On

C

NO

■q

q

D

N-

CN

«

CN

q

00

CD

o

o

m

o

CN

CD

D

CD

— H

d

d

d

d

d

CD

CD

CN

d

d

d

-H

4-1

■M

-HI

4-1

-Hi

+1

4-1

-HI

+1

+1

4-1

-HI

-HI

+1

+i

+1

q

Di

-H

— ■

On

CD

q

q

q

00

q

ON

D

q

o

q

00

d

ON

On

d

CD

CN

d

d

d

CN

00

d

00

NO

ID

CD

r-

CN

CD

CN

00

CD

m

oo

in

UJ

Q

□

E

<

<

Q

IX

O

W

O

oi

H

H

H

H

cn

cn

00

cn

5

S

5

Q

*

c

o

=tt=

.-EE

C O o O ^

O fn 00 m On

n (N rn i/N “

T I si ^ Os ,

m ON — cs| m iri

*

c

UJ

cc:

Z

u

p

a;

aJ

UJ

'c

<

Q

UJ

>

c

o

r"

j

m

H

>

O

ON

E

<

D

X

o

u

NO

1

o

ON

00

aj

Z

u

z

1

'^4'

NO

<

X

<

aJ

<

D

A

m

cn

u

a

u

UJ

>

O

u

z

D

363

Significant difference {P ^ 0.05) between Red-tailed Hawk nest sites and their corresponding random sites.
Significant difference {P ^ 0.05) between Red-shouldered and Red-tailed hawk nest sites.

Sample size = 50.

364

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 4

Mean ± One Standard Error of Nest-tree Variables and Nearest Neighbor
Distances Measured at Red-shouldered and Red-tailed Hawk Nest Sites

Red-shouldered

Red-tailed

Variable

Hawk (N = 12)

Hawk (N = 10)

NESTHT(m)

21.32 ± 1.49“

28.22 ± 1.18

NETREEHT(m)

37.17 ± 2.72

34.97 ± 1.33

NETREEDBH(cm)

67.17 ± 7.02

53.70 ± 2.63

PERNESTHT

56.0 ± 0.02“

81.0 ± 0.02

NEARNGHBR(m)

1322.99 ± 317.12

1827.47 ± 193.00

“Significant difference (P < 0.05) between Red-shouldered and Red-tailed hawk nests according to Wilcoxon rank-sum
tests.

be addressed. However, for the purpose of this study. Red-tailed Hawk
nest eircle mesoplot values probably sufficed for comparative analyses.

Habitat studies are often based on nests located without a random or
complete search of the study site because of the time required to locate
nests. Instead, raptor nests are usually located in an area where the birds
previously were seen or heard. We were able to locate an additional one
Red-tailed Hawk nest and two Red-shouldered Hawk nests because we
searched in areas where there were no previous hawk observations. How-
ever, these nests were not in habitat types different than the other nests
in our study and probably did not alter the results of our tests.

Red-shouldered Hawk nesting habitat was characterized by greater area
of bottomland habitat with nests located in large stands. Both Bednarz
and Dinsmore (1981, 1982), in Iowa, and Bosakowski et al. (1992), in
New Jersey, determined quantitatively that bottomland and other wetlands
are important habitats for breeding Red-shouldered Hawks. Stewart
(1949), Henny et al. (1973), Portnoy and Dodge (1979), and Woodrey
(1986) also reported riparian forests as the predominant nesting habitat.
Bednarz and Dinsmore (1981) suggested a critical floodplain forest size
of 250 ha, which was much larger than the 100 ha proposed by Robbins
(1979). Bednarz and Dinsmore (1982) also suggested that upland habitat
surrounding smaller floodplain forests may provide sufficient habitat for
Red-shouldered Hawks and act as a buffer against Red-tailed Hawk en-
croachment. Red-tailed Hawks historically have been described as open
country raptors often found in association with agriculture and forest
clearings (Bent 1937). All of the Red-tailed Hawks nested at or near the
edge between forested habitat and either pasture or recently clearcut hab-
itat. We often observed pairs foraging at the edge of expansive cow pas-
tures or from snags in relatively large (40 and 264 ha) clearcuts. Red-

Moorman and Chapman • HAWK NEST-SITE SELECTION

365

tailed Hawk nests were located near these foraging sites. Therefore, nest
plots had a lower amount of edge and fewer and larger stands than random
plots. Bednarz and Dinsmore (1982) also reported that Red-tailed Hawks
seemed to prefer larger hunting areas with less interspersion.

Because Red-shouldered Hawks were associated with bottomland hab-
itats and Red-tailed Hawks with upland sites, differences in nest-site mac-
rohabitat characteristics were not surprising. Red-shouldered Hawks se-
lected large areas of hardwood habitat, and Red-tailed Hawk nesting mac-
rohabitat had more agricultural area. Red-shouldered Hawk nest plots also
contained more edge. In their comparison of Red-shouldered and Red-
tailed hawk macrohabitats, Bednarz and Dinsmore (1982) also determined
that edge and number of feeding areas were important to Red-shouldered
Hawks, which used numerous small marshes interspersed with forest
when foraging.

If the primary step in choosing a nest site is habitat type selection, then
it is important to determine what cues within that habitat type are involved
in final nest-site selection. Within nesting habitat, larger trees (>69 cm)
and lower percent canopy cover were the structural differences between
Red-shouldered Hawk nest sites and random sites. Outside of the South-
east, nesting Red-shouldered Hawks also are associated with mature forest
in or near wetland habitat. Pairs nested closer to water (Titus and Mosher
1981, Bosakowski et al. 1992) and in microhabitats characterized by larg-
er, more mature trees (Titus and Mosher 1981, Morris and Lemon 1983,
Woodrey 1986) than random sites. In Ohio, Woodrey (1986) described
Red-shouldered Hawk nests as having greater percent canopy cover in
association with more large trees. Because our nest sites also were char-
acterized by more large overstory trees, a lower total canopy cover may
be the result of a reduced number of understory and midstory trees in
nest sites.

All Red-tailed Hawk nests were in loblolly pines in either pine or
upland hardwood-pine habitat. No other study has shown an exclusive
use of conifers as nest trees (Bent 1937, Fitch et al. 1946, Seidensticker
and Reynolds 1971, Titus and Mosher 1981). Because deciduous trees
were readily available, loblolly pines may have some important structural
characteristics preferred by Red-tailed Hawks. Perhaps the loblolly pine’s
straight growth form or open canopy provides easier access to the nest
(Bednarz and Dinsmore 1982).

The large diameter trees and well-developed understory at Red-tailed
Hawk nest sites are characteristic of mature pine microhabitats. Nest sites
were closer to openings and edges, had a greater canopy height, and had
taller trees than other sites within pine and hardwood-pine habitat. Each
of these characteristics favors easy nest access. Other studies also deter-

366

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

mined that nest access was important in Red-tailed Hawk nest-site selec-
tion. Speiser and Bosakowski (1988) determined that Red-tailed Hawks
nested closer to forest openings and on steeper slopes than random sites.
Titus and Mosher (1981) and Bednarz and Dinsmore (1982) also found
that pairs nested on steeper slopes.

Red-shouldered Hawks nested in sites with greater percent canopy cov-
er and lower shrub density, which are both probably correlates of the
habitat in which the birds nested. Floodplain forests tended to have a
sparser shrub layer and a more dense canopy than upland pine habitat.
By placing nests high in the nest tree near forest canopy openings. Red-
tailed Hawks may have improved access from above. However, Red-
shouldered Hawks placed nests low in the canopy, maybe improving ac-
cess from below, where these agile flyers typically approach the nest.
Nesting low in the canopy may protect Red-shouldered nestlings from
insolation and adverse weather (Bednarz and Dinsmore 1982), and pre-
dation by large avian species (Morris et al. 1982, Woodrey 1986).

Each species selected mature forests offering more nest sites with larger
trees when compared to available areas. Therefore, it may be important
to leave some stands of older, larger trees in both pine and hardwood
habitats to maintain these species. Encroachment by Red-tailed Hawks on
Red-shouldered Hawk breeding territories was probably of minimal im-
portance. In the study site, silvicultural activities were limited to upland
pine habitat, and bottomland corridors were left undisturbed. The number
of nesting Red-shouldered Hawks was relatively high and nest density
was only slightly smaller than the highest recorded density (0.22/100 ha;
Bosakowski et al. 1992). Although many intraspecific confrontations were
observed for the species during the study, no interspecific competition for
territory was noted. The minimum distance between a Red-shouldered
and a Red-tailed hawk nest was 650 m, and the Red-shouldered Hawk
nest successfully fledged young. Bednarz and Dinsmore (1982) suggested
that forest clearing and development of pastures along drainage areas
might shift the competitive advantage from Red-shouldered to Red-tailed
hawks. Bryant (1986) also reported that selective cutting in woodlots and
failure to maintain uncut buffer zones around traditional Red-shouldered
Hawk nest sites may result in local extirpation of the species. We also
agree that contiguous floodplain forests must be left relatively undisturbed
to conserve this species. Large bottomland corridors should exclude Red-
tailed Hawks because they provide poor canopy access from above. How-
ever, pine timber management on upland sites probably does not adversely
affect nesting Red-shouldered Hawks and silvicultural and agricultural
practices provide the edges and openings important to nesting Red-tailed
Hawks.

Moorman and Chapman • HAWK NEST-SITE SELECTION

367

ACKNOWLEDGMENTS

We thank D. L. Howell, J. H. Brunjes, and C. W. Eberly for continued help with field
research. We also thank A. Harris and J. Gallagher for providing housing and on-site support
and K. V. Miller and B. D. Shiver for their reviews of earlier manuscripts. This manuscript
was improved by comments provided by J. C. Bednarz and two anonymous referees. Fund-
ing for this project was provided by the Daniel B. Warnell School of Forest Resources,
Univ. of Georgia and Mclntire-Stennis ProJ. No. GEO-0074-MS.

LITERATURE CITED

Bednarz, J. C. and J. J. Dinsmore. 1981. Status, habitat use, and management of Red-
shouldered Hawks in Iowa. J. Wild. Manage. 45:236-41.

AND . 1982. Nest sites and habitat of Red-shouldered and Red-tailed hawks

in Iowa. Wilson Bull. 94:31-45.

Bent, A. C. 1937. Life histories of North American birds of prey. Pt. 1. U.S. Nat. Mus.
Bull. 167, Washington, D.C.

Bosakowski, T, D. G. Smith, and R. Speiser. 1992. Status, nesting density, and macro-
habitat selection of Red-shouldered Hawks in northern New Jersey. Wilson Bull. 104:
434-446.

Bryant, A. A. 1986. Influence of selective logging on Red-shouldered Hawks, Buteo li-
neatiis, in Waterloo Region, Ontario, 1953-1978. Can. Field-Nat. 100:520-525.
Edwards, M. G. 1978. Raptor management. Pp. 129-134 in Proc. of the workshop man-
agement of southern forests for nongame birds (R. M. DeGraaf, technical coordinator).
U.S. Dep. Agric. For. Serv., Gen. Tech. Rep. SE-14, Southeast. For. Exp. Stn., Asheville,
North Carolina.

Environmental Systems Research Institute. 1987. ARC/INFO. Redlands, California:
ESRI.

Fitch, H. S., F. Swenson, and D. F. Tillotson. 1946. Behavior and food habits of the Red-
tailed Hawk. Condor 48:205-237.

Henny, C. j., F. C. Schmid, E. M. Martin, and L. L. Hood. 1973. Territorial behavior,
pesticides, and the population ecology of Red-shouldered Hawks in central Maryland,
1943-1971. Ecology 54:545-554.

Howell, J., B. Smith, J. B. Holt, Jr., and D. R. Osborne. 1978. Habitat structure and
productivity in Red-tailed Hawks. Bird-Banding 49:162-171.

James, F. C. and H. H. Shugart, Jr. 1970. A quantitative method of habitat description.
Aud. Field Notes 24:727-736.

Kimmel, j. T. and R. H. Yahner. 1990. Response of Northern Goshawks to taped conspe-
cific and Great Horned Owl calls. J. Raptor Res. 24:107-112.

Lehmkuhl, j. F. and M. G. Raphael. 1993. Habitat pattern around Northern Spotted Owl
locations on the Olympic Peninsula, Washington. J. Wildl. Manage. 57:302-315.
Moorman, C. E. 1995. Nest-site selection of Red-shouldered and Red-tailed hawks in a
managed forest in central Georgia. M.S. thesis. University of Georgia, Athens, Georgia.
Morris, M. M. J. and R. E. Lemon. 1983. Characteristics of vegetation and topography
near Red-shouldered Hawk nests in southwestern Quebec. J. Wildl. Manage. 47:138-
145.

, B. L. Penak, R. E. Lemon, and D. M. Bird. 1982. Characteristics of Red-shoul-
dered Hawk, Buteo lineatus, nest sites in .southwestern Quebec. Can. Field-Nat. 96:
139-142.

Mosher, J. A., M. R. Fuller, and M. Kopeny. 1990. Surveying woodland raptors by
broadcast of conspecific vocalizations. J. Field Ornithol. 61:453-461.

368

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Nelson, B. B. and K. Titus. 1988. Silviculture practices and raptor habitat associations in
the Northeast. Pp. 171-179 in Proc. northeast raptor management symposium and work-
shop (B. G. Pendleton, M. N. Lefranc, Jr., M. B. Moss, C. E. Ruibal, M. A. Knighton,
D. L. Krahe, eds.). Nat. Wildl. Led. Sci. Tech. Sen No. 13.

Noon, B. R. 1981. Techniques for sampling avian habitats. Pp. 42-49 in The use of mul-
tivariate statistics in the study of wildlife habitat (D.E. Capen, ed.). USDA Lorest Ser-
vice General Technical Report RM-87.

Portnoy, J. W. and W. E. Dodge. 1979. Red-shouldered Hawk nesting ecology and be-
havior. Wilson Bull. 91:104-1 17.

Robbins, C. S. 1979. Effects of forest fragmentation on bird populations. Pp. 198-212 in
Management of north-central and northeastern forests for nongame birds (R. M. De-
Graaf and K. E. Evans, eds.). Gen. Tech. Rep. NC-51, St. Paul, Minnesota: U.S. Dept.
Ag., Forest Service, North Central Forest Expt. Station.

Rosenfield, R. N., J. Bielefeldt, R. K. Anderson, and W. A. Smith. 1988. Effectiveness
of broadcast calls for detecting breeding Cooper’s Hawks. Wildl. Soc. Bull. 16:210-
212.

SAS Institute Inc. 1982. SAS user’s guide: statistics. 1982 edition. Carey, North Carolina.

Sedgwick, J. A. and F. L. Knopf. 1992. Describing Willow Flycatcher habitats: scale per-
spectives and gender differences. Condor 94:720-733.

Seidensticker, j. C. and H. ’V. Reynolds. 1971. The nesting, reproductive performance,
and chlorinated hydrocarbon residues in the Red-tailed Hawk and Great Horned Owl
in south-central Montana. Wilson Bull. 83:408-418.

Speiser, R. T. and T. Bosakowski. 1988. Nest site preferences of Red-tailed Hawks in the
Highlands of southeastern New York and northern New Jersey. J. Field Ornithol. 59:
361-368.

Stewart, R. E. 1949. Ecology of a nesting Red-shouldered Hawk population. Wilson Bull.
61:26-35.

Titus, K. and J. A. Mosher. 1981. Nest-site habitat selected by woodland hawks in the
central Appalachians. Auk 98:270—281.

WOODREY, M. S. 1986. Characteristics of Red-shouldered Hawk nests in southeast Ohio.
Wilson Bull. 98:466-469.

Zar, j. H. 1974. Biostatistical analysis. Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

Wilson Bull.. 108(2), 1996, pp. 369-371

SHORT COMMUNICATIONS

Avoidance of cabbage fields by Snow Geese. — Available evidence suggests that herbi-
vores generally avoid sulfuroiis odors and volatile fatty acids (Mason et al. 1994) because
such odors are associated with carnivore urine and feces (Nolte et al. 1994). Alternatively,
or in addition, herbivores may avoid sulfur volatiles because they signal the bioaccumulation
of toxicants such as selenium (Morris 1970) or because they indicate the presence of mi-
crobial degradation products that are toxic to vertebrates (Guildford et al. 1987).

Greylag Geese {Anser anser) avoid skatol (Neuhaus 1963), a (albeit nitrogenous) volatile
present in the feces of predatory civet cats (Civetriciis civetta) and in the fruits of some
plants. Contrary to prevailing belief, the olfactory performance of birds in general is on a
par with that of mammals (Clark and Mason 1989). Anecdotes provided by farmers suggest
that geese rarely forage on winter cover crops in fields where cabbage had been planted tbe
previous summer. A variety of sulfurous volatiles result from the decomposition of cabbage
(Brassica oleracea capitata). including hydrogen sulfide, methyl disulfide, dimethyl disul-
fide, and various methyl mercaptans (Dateo et al. 1957, Self et al. 1963). We patterned the
present series of observations to test the hypothesis that the odors of decaying cabbage repel
Snow Geese {Chen caerulescens). We chose Snow Geese as our model species for three
reasons. First, this bird is a strict herbivore. Second, large numbers of Snow Geese over-
winter in southern New Jersey, our study area. They routinely forage on the winter cover
crops planted in fields (Mason and Clark 1994). Thus it was possible to obtain a large
number of fields in which cabbage had been planted the previous summer, and within which
geese could possibly feed. Third, Domestic Geese {Anser anser) respond to plant odors
(Neuhaus 1963, Wurdinger 1979), and captive Snow Geese will avoid high concentrations
of Deer Away Big Game Repellent (IntAgra, Minneapolis, Minn.) (Mason, unpubl. obs.).
The repellency of this commercially available product depends upon the production of sulfur
odors and volatile fatty acids (Bullard et al. 1978).

Study area and methods. — We selected 16 fields near Cedarville, New Jersey, for study.
All were physically similar (30—40 ha in size, adjacent to other agricultural fields on at least
3 sides) and within 5 km of Delaware Bay marsh habitat used by >30,000 overwintering
Snow Geese (L. Widjeskog, N. J. Div. Fish and Game, pers. commun.).

Cabbage had been planted in eight of the fields during the 1994 growing season. Peppers
{Capsicum frutescens, N = 6) or soybeans {Glycine max. N = 2) had been planted in the
other eight fields. Conversations with farmers indicated that cabbage, peppers, and soybeans
were rotated among all of the fields in multiyear cycles. During the observation period, all
of the fields were planted with rye {Secale cereale). The maturity of the rye in all 16 fields
was similar.

We paired cabbage and control fields on the basis of proximity; no member of any pair
was more than 200 m apart. At the middle of each field, we established a 100-m transect
parallel to the longest axis of the field and marked the ends of each transect with 0.4 m
long wooden survey stakes.

Between 30 October 1994 and 20 March 1995, we visited all fields at seven-day intervals.
During each visit, we walked each transect and collected all goose droppings with 0.5 m of
the transect midline. We took the droppings to the laboratory and dried them in an oven at
37 C to a constant mass or for 72 h. We used these masses as an indication of goose activity
(Mason et al. 1993, Mason and Clark 1994). We did not attempt to analyze the cover crop
or the soil for the presence of sulfurous compounds, although sulfurous odors were readily
apparent to us during our visits to cabbage fields.

369

370

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

1/2 1/9 1/16 1/23 1/30 2/6 2/13 2/20 3/6 3/13 3/20

Dates

Lig. 1. Mean snow goose feces/transeet meter in cabbage fields and control fields be-
tween January 2 1995 (week 1) and March 20, 1995 (week 1 1). Whiskers represent standard
errors of the means.

We used a two-factor (sampling date, field type) repeated measures analysis of variance
to evaluate the data. Subsequently, we used Tukey tests (Winer 1962:198) to isolate signif-
icant differences among means {P < 0.05).

Results. — No droppings were found in any field (cabbage or control) until the beginning
of January. Lor that reason, only those data collected between 2 January 1995 and 20 March
1995 were evaluated. Overall, masses of droppings increased over time {F = 18.3; 10,40
df; P < 0.0001) and were greater in control fields than in cabbage fields {F = 65.4; 10,40
df; P < 0.0002). The significant interaction between time (dates) and field type {F = 5.2;
10,40 df; P < 0.0002) showed that differences of masses of droppings between cabbage
and control fields decreased as the season progressed (Lig. 1). There were no significant
differences by late March.

Di.scussion. — Snow Goose activity levels were significantly less in cabbage fields than in
control fields. Although the data do not unambiguously address the issue of sulfur repellency,
we believe that the activity difference is consistent with avoidance of the former and not
preferenee for the latter. Sulfurous volatiles were readily apparent to us during our visits to
cabbage fields throughout the study period. Similar odors were not detected in control fields.
If sulfurous volatiles were important, then avoidance could reflect some characteristic of the
cover crop (e.g., unpalatability acquired through the absorption and translocation of degra-
dation products) or it could reflect an aversion to ambient (and readily detectable) volatiles
in the field (Guildford et al. 1987). Regardless, our data are consistent with the notion that
sulfurous volatiles may repel Snow Geese, at least within a feeding context. Perhaps sulfur

SHORT COMMUNICATIONS

371

containing substances could be developed as a method for goose grazing control. Many
avian species have an acute sense of smell (Davis 1973, Clark and Mason 1989, Clark et
al. 1993), with variability among species similar to that observed for mammals (Fazzalari
1978). While threshold data are unavailable, there is evidence that geese are highly respon-
sive to odorous cues (Neuhaus 1963, Wurdinger 1979, 1982).

LITERATURE CITED

Bullard, R. W., T. J. Leiker, J. E. Peterson, and S. R. Kjlburn. 1978. Volatile components
of fermented egg, an animal attractant and repellent. J. Agr. Food Chem. 26:155-159.

Clark, L., K. V. Avilova, and N. J. Bean. 1993. Chemical repellency in birds: relationship
between chemical structure and avoidance response. J. Exp. Zool. 260:310-322.

AND J. R. Mason. 1989. Sensitivity of Brown-headed Cowbirds to volatiles. Condor
91:922-932.

Dateo, G. R, R. C. Clapp, D. A. M. Mackay, E. J. Hewitt, and T. Hasselstrom. 1957.
Identification of the volatile sulfur components of cooked cabbage and tbe nature of
the precursors in the fresh vegetable. Food Res. 22:440-445.

Davis, R. G. 1973. Olfactory psychophysical parameters in man, rat, dog, and pigeon. J.
Comp. Physiol. Psychol. 85:221-232.

Fazzalari, F. A. 1978. Compilation of odor and taste threshold values data. American
Society for Testing and Materials, Philadelphia, Pennsylvania.

Guildford, T, C. Nicol, M. Rothschild, and B. P. Moore. 1987. The biological roles of
pyrazine: evidence for a warning odour function. Biol. J. Linn. 31:1 13-128.

Mason, J. R., L. Clark, and N. J. Bean. 1993. White plastic flags repel snow geese {Chen
caerulescens). Crop Prot. 12:497-500.

AND L. Clark. 1994. Evaluation of plastic and mylar flagging as repellents for

snow geese (Chen caerulescens). Crop Prot. 13:531-534.

, G. Epple, and D. L. Nolle. 1994. Semiochemicals and improvements in rodent
control. Pp. 327-346. in Behavioral aspects of feeding. (B. G. Galef, M. Mainardi, and
P. Valsecchi, eds.), Harwood Academic Publishers, Switzerland.

Morris, V. C. 1970. Selenium content of foods. J. Nutri. 100:1385-1386.

Neuhaus, W. 1963. On the olfactory sense of birds. Pp. 111-115. in Olfaction and taste
(Y. Zotterman, ed.), Macmillian, New York.

Nolle D. L., J. R. Mason, G. Epple, E. Aronov, and D. L. Campbell. 1994. Why are
predator urines aversive to prey? J. Chem. Ecol. 20:1505-1516.

Self, R., J. C. Casey and T. Swain. 1963. The low boiling point volatiles of cooked foods.
Chem. Indus, p. 863.

Winer, B. J. 1962. Statistical principles in experimental design. McGraw-Hill, New York,
New York.

Wurdinger, I. 1979. Olfaction and feeding behavior in juvenile geese (Anser a. Anser and
Anser domesticus). Zeit. Tierpsychol. 49:132-137.

. 1982. Olfaction and home learning in juvenile geese (Anser and Branta species).

Biol. Behav. 5:347-351.

J. Russell Mason and Larry Clark, USDA/Animal and Plant Health In.spection Service.
Animal Damage Control, Denver Wildlife Re.search Center. % Monell Chemical Senses
Center, 3500 Market Street. Philadelphia. Penn.sylvania I9J04-330H. Received 25 Max
1995, accepted 7 Nov. 1995.

372

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Wilson Bull., 108(2), 1996, pp. 372-374

Taxonomic status of the Cuban form of the Red-winged Blackbird. — The Cuban Red-
winged Blackbird was described as a species, Agelaius assimilis, by Gundlach in Lembeye
in 1850 (Ridgway 1902, Blake 1968). It was treated as a species by Ridgway (1902) who
noted that it differed from the widespread Red-winged Blackbird {A. phoeniceus) because

( 1 ) the female plumage was uniformly black, unlike any subspecies of A. phoeniceus and

(2) the male was smaller than almost any form of A. phoeniceus. Hellmayr ( 1937), however,
treated it as a subspecies of the Red-winged Blackbird with only the following explanation:
“A. p. assimilis is nothing but a small race of the American Red-wing with a very dark
female.” Subsequent authors (e.g., Barbour 1943, Bond 1956, Blake 1968, Orians 1985,
Sibley and Monroe 1990) have followed Hellmayr’s taxonomy, although Mayr and Short
(1970) considered assimilis a sibling species. Recent fieldwork by Whittingham et al. (1992)
has shown that the form assimilis differs from other populations of A. phoeniceus in voice
and social behavior. This new evidence, combined with the similar plumage of male and
female assimilis, leads us to conclude that this taxon is best treated at the species level.
Below we summarize the evidence.

Plumage dichromatism. — Although the Red-winged Blackbird shows much geographic
variation in size over its large range (e.g.. Power 1969, 1970; Dickerman 1974), the basic
plumage pattern of the female, brown and heavily streaked, is consistent throughout its
range, except in guhernator. This includes populations closest to Cuba, A. p. bryanti of the
Bahamas and A. p. richmondi of the tropical lowlands of Middle America. In the subspecies
of the Mexican plateau, A. p. gubernator, female streaking is greatly reduced and limited
to the throat, the remaining plumage is very dark brown (but not as black as assimilis). In
the Californian subspecies, A. p. californicus and A. p. mailliardorum, streaking is also
reduced in females and the plumage is dark brown, although not to the degree that it is in
gubernator. Although not stated explicitly, the tendency of these populations to vary in
female plumage color in the direction of assimilis almost certainly influenced Hellmayr’s
and others’ decisions to regard the latter as only an end-point of the variation in female
plumage of A. phoeniceus.

In our opinion, however, the female plumage of assimilis differs qualitatively from being
merely an unstreaked, dark extreme in plumage variation because the plumage is unifomily
coal-black, like the males and not brown, as in even the darkest forms currently treated as
subspecies of A. phoeniceus. Lurthermore, the evidence for maintaining guhernator as a
subspecies of A. phoeniceus is weak (see Hardy and Dickerman 1965). Linally, in the
Tricolored Blackbird (A. tricolor), the female has a relatively less-streaked plumage that
differs from that of the male less than do male and female plumages of sympatrically
breeding A. phoeniceus. The sexual dichromatism of assimilis is even less than that of A.
tricolor. Therefore, differences in female plumage in Agelaius are associated with differ-
ences in taxa designated as .separate species.

The plumages of nestlings and second year males also differ between A. phoeniceus and
A. assimilis. In A. assimilis, nestling plumage is entirely dull black and some nestlings have
reddish-brown lesser wing coverts (presumably males; Kirkconnell pers. obs.). In contrast,
the nestling plumage of A. phoeniceus is entirely streaked brown (Pyle et al. 1987). Second
year (SY) male A. assimilis are entirely black except for the orange epaulets which are
mottled with black (Kirkconnell pers. obs.). In contrast, the plumage of SY males of A.
phoeniceus is blackish with heavy white or buff streaking (Pyle et al. 1987).

Vocalizations. — Whittingham et al. (1992) compared the vocalizations of Red-winged
Blackbirds in North America and Cuba. Sonographic analyses showed that male phoeniceus
and assimilis songs were similar in structure; however, male assimilis songs were shorter

SHORT COMMUNICATIONS

373

and had a greater frequency range (see Fig. 1 in Whittingham et al. 1992). In contrast, songs
ot female phoeniceits differed dramatically from those of female cissimilis. The latter were
nearly identical to male assimilis songs (see Fig. 1 in Whittingham et al. 1992). In contrast,
phoeniceits females sing two song types (Beletsky 1983), each a series of individual notes
that differ distinctly from songs of male phoeniceits or of either sex in assimilis. The vocal
behavior of phoeniceits and assimilis also differs dramatically; assimilis males and females
often sing their songs in a duet (Whittingham et al. 1992), whereas phoeniceits males and
females sing only solo songs.

Mating system. — Duetting is generally associated with prolonged monogamous pair bonds
(Farabaugh 1982), which suggests that the mating systems of phoeniceits and assimilis also
differ. Further, studies of color-marked birds show that male and female assimilis are ob-
served only in pairs whether on their breeding territories or while foraging away from their
territories (Kirkconnell, pers. obs.). These observations further support the idea of a mo-
nogamous mating system in assimilis. In contrast, phoeniceus is polygynous throughout its
range (reviewed in Whittingham and Robertson 1994). In some cases, males may have as
many as 15 females breeding on their territory at one time (Beletsky and Orians 1990).

Validity of A. assimilis subniger. — Bangs and Zappey (1905) recognized the population
on the Isle of Pines (now Isle of Youth) as A. assimilis. Bangs (1913) later described the
population as A. subniger based on its coloration being very dark brown and “. . . the bill
has a tendency to be rather longer and with a slightly rounded, less flattened culmen.”
However, the validity of these characters was questioned because the specimens Bangs
examined were mostly immature (Todd 1916). Todd (1916) stated “. . . all but one of the
male specimens are clearly in the immature stage. . ..the culmen is slightly flatter, it is true,
in the Cuban specimens, but I believe that even this difference would disappear in a large
series; at any rate, it is certainly too trifling a difference upon which to base the recognition
of even a subspecies. Garrido (1970), in his revision, agreed with Todd’s comments and
considered the taxon subniger a synonym of assimilis.

In summary, the sexes are similar in phenotype and vocalizations in assimilis, whereas
these characteristics differ dramatically between the sexes in phoeniceus. Furthermore, the
plumage of nestlings and SY males as well as the mating system differs between assimilis
and phoeniceus. We believe that the evidence strongly favors treatment of the taxon endemic
to Cuba as a species, Agelaius assimilis.

Acknowledgments. — We thank J. V. Remsen Jr. and L. A. Whittingham for comments on
the manuscript. (Ed. note: L. A. Whittingham helped immensely to expedite publication of
this paper.)

LITERATURE CITED

Bangs, O. 1913. New Birds from Cuba and the Isle of Pines. Proc. New England Zool.
Club 4:92.

AND W. R. Zappey. 1905. Birds of the Isle of Pines. Am. Nat. 39:179-215.

Barbour, T. 1943. Cuban Ornithology. Memoirs of the Nutthall Ornithological Club, Cam-
bridge, Massachusetts. No. 9, p. 123.

Beletsky, L. D. 1983. Aggressive and pair-bond maintenance songs of female Red-winged
Blackbirds {Agelaius phoeniceus). Z. Tierpsychol. 62:47-54.

AND G. H. Orians. 1990. Male parental care in a population of Red-winged Black-
birds, 1983-1988. Can. J. Zool. 68:606-609.

Blake, E. R. 1968. Family Icteridae. Pp. 138-202 in Checklist of birds of the world, vol.
XIV (R. A. Paynter Jr., ed.) Museum of Comparative Zoology, Cambridge, Massachu-
setts.

374

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

Bond, J. 1956. Checklist of birds of the West Indies. Acad. Nat. Sci. Phil., Philadelphia,
Pennsylvania.

Dickerman, R. W. 1974. Review of Red-winged Blackbirds (Agelaius phoeniceus) of east-
ern, west-central, and southern Mexico and Central America. Amerc. Mus. Novitates
2538:1-8.

Larabaugh, S. 1982. The ecological and social significance of duetting. Pp. 85-124. in
Acoustic communication in birds (D. E. Kroodsma. and E. H. Miller, eds.). Academic
Press, New York, New York.

Garrido, O. H. 1970. Variacion del genero Agelaius (Aves: Icteridae) en Cuba. Poeyana
68:1-18.

Hardy, J. W. and R. W. Dickerman. 1965. Relationships between the two forms of the
Red-winged Blackbird in Mexico. Living Bird 4:107-130.

Hellmayr, C. E. 1937. Catalogue of birds of the Americas and adjacent islands. Eield Mus.
Nat. Hist. Zool. Series Vol. XIII, Chicago, Illinois.

Mayr, E. and L. L. Short. 1970. Species taxa of North American birds. Cambridge,
Massachusetts.

Orians, G. H. 1985. Blackbirds of the Americas. Univ. of Washington Press, Seattle, Wash-
ington.

Power, D. M. 1969. Evolutionary implications of wing and size variation in the Red-
winged Blackbird in relation to certain geographic and climatic factors: a multiple
regression analysis. Syst. Zool. 18:363—373.

. 1970. Geographic variation of Red-winged Blackbirds in central North America.

Univ. Kansas Publ. Mus. Natur. Hist. 90:1-83.

Pyle, P, S. N. G. Howell, R. P. Yunick, and D. E Desante. 1987. Identification guide to
North American passerines. Slate Creek Press, Bolinas, Calif.

Ridgway, R. 1902. The birds of North and Middle America. Part II. Washington, D.C.

Sibley, C. G. and B. L. Monroe, Jr. 1990. A world checklist of birds. Yale Univ. Press,
New Haven, Connecticut.

Todd, W. E. C. 1916. The Birds of the Isle of Pines. Ann. Carnegie Museum, Vol. 10.

Whittingham, L. a., a. Kirkconnell, and L. M. Ratcliffe. 1992. Differences in song
and sexual dimorphism between Cuban and North America Red-winged Blackbirds
(Agelaius phoeniceus). Auk 109:928—933.

AND R. J. Robertson. 1994. Pood availability, parental care and male mating suc-
cess in Red-winged Blackbirds (Agelaius phoeniceus). J. Anim. Ecol. 63:139-150.

Orlando Garrido and Arturo Kirkconnell, Museo Nacional de Historia Natural, Cap-
itol io National, La Hahana, Cuba. Received 12 Sept. 1995, accepted 6 Dec. 1995.

Wilson Bull., 108(2), 1996, pp. 374-377

Nest adoption by Monk Parakeets. — Monk Parakeets (Mylopsitta monachus) are un-
usual, being the only non-cavity nesting psittacines. Rather than using tree holes, burrows,
or crevices as other parrots typically do, they build large domed nests of twigs (Porshaw
1989). Their nests often include several compartments, each with a separate entranee, and
several nests may be built in the same tree or in neighboring trees. Monk Parakeets are non-
migratory and use their nests year-round for roosting as well as for breeding. Nests typically
are built in trees, as well as on a variety of man-made structures (windmill towers, utility

SHORT COMMUNICATIONS

375

poles, sign posts, etc.). One sub-species, M. monachus luchsi, builds stick nests on cliffs
(Tanning 1991).

During a visit to southern Buenos Aires Province, Argentina, Humphrey and Peterson
(1978) noted an association between nests ot the Firewood Gatherer (Anumbius anniimbi)
and those of Monk Parakeets. They found that parakeets frequently used Anumbius nests as
a foundation on which to build their own nests. From Humphrey and Peterson’s observa-
tions, it is not clear whether or not the nests added to by parakeets had been abandoned by
their original builders. In the resulting duplex nests, twigs added by the parakeets frequently
engulfed the original nest, but the nest cavities were separated by a double wall, and the
parakeets built and used a separate entrance tunnel.

Here, I report a similar association found during a study of Monk Parakeets in Entre Rios
Province, Argentina. A large proportion of Monk Parakeet nests were found to be remodeled
nests ot the Brown Cacholote (PseucJoseisura lophotes). The occasional adoption of Brown
Cacholote nests by Monk Parakeets has been previously noted (Nores and Nores 1994); I
show that adopted nests are widely used by Monk Parakeets and suggest that the association
may provide a clue to understanding the evolution of the Monk Parakeet’s domed nest.

Study area and methods. — During the austral spring/summers of 1993-94 and 1994-95,

I studied the breeding behavior of Monk Parakeets on a portion of Estancia Santa Ana de
Carpinchorf, a cattle ranch in northern Entre Ri'os Province, Argentina. Parts of the ranch
have been cleared, but much of it retains its native savannah woodland vegetation, which
is dominated by three xerophytic trees: Acacia caven, Prosopis affinis and Prosopis nigra.

I monitored the occupancy and breeding activity of all nests that were found on 1000 ha of
uncleared land and that were accessible with a 7-m ladder. The locations of all nests in the
study area were mapped, and for each one I measured (or estimated, in the case of very
isolated nests) the distance to the nearest neighboring nest. The height to the center of each
nest was measured as well. Only nests that were occupied (used either for roosting or
breeding) by parakeets for at least a portion of the study are discussed here. All nests were
scored as being either original Monk Parakeet nests or adopted Brown Cacholote nests that
had been remodeled by parakeets. This determination was made by visual inspection of the
twigs used in a nest’s construction. Adopted nests are recognizable because a portion (gen-
erally the back and/or underside) of the nest comprises twigs of more variable and greater
thickness than those used by parakeets (see Results).

Results and discussion. — Monk Parakeets were observed to construct nests using thorn
twigs clipped from nearby A. caven, P. affinis, and P. nigra trees (usually <100 m from
nest site). Twigs from other species of trees, or picked up from the ground, were used <1%
of the time (JRE, unpubl. data). Parakeets consistently used the terminal ends of twigs for
nest-building and occasionally used twigs stolen from nearby parakeet nests. Brown Cach-
olotes, on the other hand, use a variety of types and sizes of twigs (Nores and Nores 1994),
resulting in a nest that, though similar in shape and size to that of the Monk Parakeet, is
readily distinguishable because of the nesting materials employed in its construction. To
document this difference in sizes of twigs used by the two species, I measured the mid-twig
diameters of 100 randomly chosen twigs/nest from three cacholote nests and three parakeet
nests. The variance and median did not vary significantly among nests within each species,
so data were pooled within each species for the analyses presented here. Variance in twig
diameter is much higher in cacholote nests than in nests built by Monk Parakeets (F-test;
F = 5.67, df = 299, P 0.0001), and twigs used by the parakeets are significantly thinner
(Mann-Whitney U test: Z = -14.55, P = 0.0001, N = 600).

In the course of the two field seasons, I monitored a total of 39 accessible and occupied
Monk Parakeet nests, some of which were occupied during both years. Of these 39 nests,
20 (51%) were originally cacholote nests that had been adopted and remodeled to varying

376

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

Table 1

Nest and Nest-site Characteristics of Monk Parakeets at Estancia Santa Ana de
CarpinchorI, Entre Rios Province, Argentina

Mean nest

Mean nearest

Original

Number

height

neighbor dist.

Number used

builder

(% total)

(X ± SE)

(X ± SE)

for breeding

M. monachus

19 (48.7)

5.32 (± 0.3)

39.2 (± 16.2)

1 1

P. lophotes

20 (51.3)

4.5 (± 0.3)

150.8 (± 46.6)

1

extents by Monk Parakeets. This is likely to be a conservative estimate, since extensive
remodeling of a cacholote nest by the parakeets could eventually engulf its foundation,
leading me to score some adopted nests as originally being parakeet nests. Parakeets ap-
peared to adopt nests that had been abandoned by their original owners and had begun to
fall apart, creating an opening in the nest chamber. In their study of Brown Cacholote nesting
behavior, Nores and Nores (pers. comm.) also found that monk parakeets usually moved
into abandoned Brown Cacholote nests (8 of the 9 cases they observed). When remodeling
cacholote nests, parakeets add twigs to the roof and around the entrance. For the six nests
that I found in early stages of remodeling, parakeets had built a new entrance tunnel; how-
ever, Nores and Nores (pers. comm.) found that in eight of the nine cases of nest adoption
they observed, the parakeets used the cacholote nest’s original entrance. Unlike the nest
association found by Humphrey and Peterson (1978) in which Monk Parakeets used An-
umbius nests as foundations upon which to build their own nesting compartment, parakeets
adopting cacholote nests always re-used the original nest’s chamber.

The main site characteristics, nest height and nearest-neighbor distance and whether or
not the nest was used for breeding, are summarized in Table 1. The heights of nests built
by Monk Parakeets and cacholote nests adopted by parakeets did not differ significantly
(Mann-Whitney U test: Z = -1.548, N = 39, P = 0.12). Adopted nests were more isolated,
as reflected by their significantly longer nearest-neighbor distances (Mann-Whitney U test:
Z = -2.757, N = 39, P < 0.01). Breeding attempts occurred in 12 nests, and most (92%)
of these were in nests originally built by parakeets themselves. This significant preference
(X^ = 12.82, df = 1, P < 0.005) for breeding in non-adopted nests is probably due to the
fact that these nests were more likely to be in colonies (a colony was defined as groups of
nests with nearest-neighbor distances of less than 100 m). In the single case in which a
breeding attempt took place in an adopted nest, the nest had been enlarged and contained
two compartments, both of which were occupied.

The Monk Parakeet’s willingness to adopt and remodel the nests of another species is
particularly interesting in light of the fact that the Monk Parakeet is the only parrot species
that builds a nest that is completely dis.sociated from a cavity. Nest adoption may originally
have arisen in this species’ ancestor as an alternative nesting strategy used by pairs unable
to find or .successfully compete for nesting cavities. The adoption behavior may have pre-
ceded the evolution of more complex nest-building behavior. The ability to construct a nest
would then have emancipated them from a dependence on cavities or other species’ nests
for breeding. By giving pairs flexibility in choosing nest sites, nest-construction may in turn
have facilitated the strong tendency of Monk Parakeets to breed colonially.

Acknowledgment.'!. — During the preparation of this paper, I was supported by Princeton
University and an NSF pre-doctoral fellowship. I thank the Ortiz Basualdo family and the
employees of Estancia Santa Ana de Carpinchori for their hospitality and logistical support

SHORT COMMUNICATIONS

377

in the field. K. Harms assisted in the collection of field data, and P. Grant, K. Harms, and
H. Horn made helpful comments on early drafts of the manuscript. I also thank D. Panning
for comments on the manuscript, as well as A. Nores and M. Nores for allowing me to
include unpublished data.

LITERATURE CITED

Forshaw, J. M. 1989. Parrots of the world. 3rd edition. Lansdowne Editions, Sydney,
Australia.

Humphrey, P. S. and R. T. Peterson. 1978. Nesting behavior and affinities of Monk Par-
akeets of southern Buenos Aires Province, Argentina. Wilson Bull. 90:544-552.
Panning, D. V. 1991. Distribution and nest sites of the monk parakeet in Bolivia. Wilson
Bull. 103:366-372.

Nores, A. I. and M. Nores. 1994. Nest building and nesting behavior of the Brown
Cacholote. Wilson Bull. 106:106-120.

Jessica R. Eberhard, Ecology and Evolutionary Biology, Princeton Univ., Princeton, New
Jersey 08544-1003. Received 20 April 1995, accepted 15 Oct. 1995.

Wilson Bull., 108(2), 1996, pp. 377-378

Vermilion Flycatcher and Black Phoebe feeding on fish. — We describe our observa-
tions of two species of flycatchers, the Vermilion Flycatcher (Pyrocephalus rubinus) and
the Black Phoebe (Sayornis nigricans) feeding on fish. Observations of Black Phoebes
capturing fish have been noted as unusual (Bent 1942, Lawson 1975), and this is the first
account of a Vermilion Flycatcher feeding on fish (Bent 1942, Terres 1980).

We made these observations at the Hassayampa River Rest Area approximately 6 km
southeast of Wickenburg, Maricopa County, Arizona. On 2 Dec. 1993, we observed an adult
male Vermilion Flycatcher eating a small fish. The flycatcher was first observed perched in
a mesquite tree (Prosopsis velutina) approximately 12 m from the Hassayampa River. With
binoculars we could clearly see the distal half of a fish protruding from the flycatcher’s bill.
It was unknown if the flycatcher captured or scavenged the fish. Vermilion Flycatchers most
commonly feed by hawking for insects, but occasionally they land on the ground to feed
on terrestrial invertebrates (Bent 1942, Terres 1980, Ehrlich et al. 1988 Rosenberg et al
1991).

Andrews returned to the area on 4 Dec. 1993 and observed an adult male vermilion
Flycatcher on a small mesquite tree branch 2.5 m directly over the water. After several min
of observation, the flycatcher flew down, breaking the surface of the water. It then hovered
just above the water for several seconds before again darting into the water. The bird hovered
then darted into the water two more times. All four attempts were unsuccessful. It then
returned to the same me.squite branch above the water. The depth of the water at this location
was approximately 12 cm. Suspecting that the flycatcher may have been diving into the
water after insects, we looked for insects or other aquatic invertebrates. No insects were
observed in or over the water in the area where the flycatcher was hunting. Several large
schools of longfin dace (Agosia chrysoga.ster), an abundant native fish of the family Cy-
prinidae, were observed at the site where the flycatcher had been hovering and diving.
Attempts to photograph the flycatcher’s feeding behavior were unsuccessful.

378

THE WILSON BULLETIN • Vol. JOS. No. 2, June 1996

Approximately one hour after the Vermilion Llycatcher observation, Andrews saw a Black
Phoebe capture a small fish, probably a longfin dace, in the same area. Although Black
Phoebes feeding on small fish have been noted in the literature and reported as an unusual
diet item (Bent 1942, Lawson 1975), it is noteworthy to describe here the capture and kill
method used. Using binoculars, Andrews observed a Black Phoebe perched on the edge of
the river looking into the water. It quickly jumped into the shallows and emerged with a
small fish in its bill. The phoebe returned to the bank with the wiggling fish and forcibly
threw the fish on the ground three times. When the fish ceased to move it was swallowed
headfirst by the phoebe. This method of immobilizing the fish was similar to that described
by Lawson (1975) who reported a Black Phoebe repeatedly striking a captured fish against
a tree branch until it ceased to struggle then swallowed it, apparently headfirst. We hypoth-
esize that the two species of flycatchers’ feeding behavior was an opportunistic response to
the abundance and visibility of small fish in shallow water.

Acknowledgments. — We are grateful to K. A. King for his encouragement and constructive
comments on early drafts of this manuscript. We thank T. A. Gatz, for his literature search
and review of the manuscript, S. Robertson for his review of the final draft, and J. Hanson
for her grammatical expertise.

LITERATURE CITED

Bent, A. C. 1942. Life histories of North American flycatchers, larks, swallows, and their
allies. U.S. Natl. Mus. Bull. 179:159-163, 306-307.

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The birders handbook; a field guide
to the natural history of North American birds. Simon and Schuster, New York, New
York.

Lawson, C. S. 1975. Pish catching by a Black Phoebe. West. Birds 6:107-109.

Rosenberg, K. V., R. D. Ohmart, W. C. Hunter, and B. W. Anderson. 1991. Birds of
the Lower Colorado River Valley, Univ. of Ariz. Press, Tucson, Arizona.

Terres, J. K. 1980. The Audubon Society encyclopedia of North American birds. Alfred
A. Knopf, New York, New York.

Brenda J. Andrews and Marie Sullivan, U. S. Fish and Wildlife Service, 2321 W. Royal
Palm Rd., Suite 103, Phoenix, Arizona 85021, and J. David Hoerath, Bureau of Land
Management, 2015 W. Deer Valley Rd., Phoenix, Arizona 85027. Received 29 Aug. 1995,
accepted 20 Oct. 1995.

Wilson Bull., 108(2), 1996, pp. 378-380

Nest-site reuse in the Western Wood-Pewee. — Reuse of the same nest site within a
territory from one year to the next is well documented for birds such as colonial breeders
(Shields 1984), cavity nesters (Harvey et al. 1979, Newton 1994), and species nesting on
natural ledges and artificial structures (Bent 1942). In these groups, nest site reuse is pro-
moted by the scarcity of suitable nest sites. Pew non-colonial, open-nesting passerines have
been documented reusing nest sites between years. Breeding studies that compare nest lo-
cations between years for this nesting guild generally report that nest sites are not reu.sed
(Hendricks 1991, Martin and Roper 1988) or are rarely reused (Nolan 1978). Howeyer,
.some open nesting tyrannid flycatchers, i.e.. Eastern Kingbird (Tyrannus tyrannus) (Blancher

SHORT COMMUNICATIONS

379

and Robertson 1985), Western Flycatcher {Empidonax difficilis), and Eastern Wood-Pewee
{Contopus Virens) (Bent 1942), regularly reuse nest sites between years. We report several
instances of nest site reuse in another flycatcher, the Western Wood-Pewee (C. sordidulus).

During 1992-1994, we monitored the nesting dynamics of birds breeding in pinyon pine
(Finns ednlis) - one-seed juniper (Jimiperus monospennci) habitat in Colfax County, north-
eastern New Mexico. During this period, we located 46 Western Wood-Pewee nests that
reached the egg-laying stage. All of these nests were in the dominant tree species, pinyon
pine. In 1993, two of seven nest sites u.sed during 1992 were reused, and in 1994, three of
15 nest sites used during 1993 were reused. One nest site was used in all three years, with
the new nests constructed by adding material to the remaining portion of the previous year’s
nest. In other instances where the previous year’s nest had fallen off the branch over the
winter, the new nest was built in the same location. Because we did not band pewees on
our site, individual recognition was not possible. However, we suspect that reuse involved
the return of at least one individual of a pair from the previous year.

In 1993, a pewee nest which had been depredated during incubation was reused in the
same season. The second clutch, initiated less than a week after the depredation event, was
raised successfully, and this nest site was also reused successfully the following year. It
seems unusual for birds to reuse a depredated nest or nest site (Harvey et al. 1979, Dow
and Fredga 1983). A previously depredated nest might be more vulnerable to future pre-
dation than would be a new nest site, since some predators (e.g., corvids) appear to search
the locations of nests that they have previously depredated, even a year later (Sonerud and
FJeld 1987).

Several explanations for the reuse of nest sites by the Western Wood-Pewee on our study
area are possible, including (1) high quality nest sites may be in limited supply, despite the
abundance of pinyons in the breeding habitat, (2) nest site reuse may be an extreme behav-
ioral expression of site fidelity, and (3) pewees may benefit from time and energy savings
by not searching for new nest sites or nest materials.

Acknowledgments.~We thank Scott Adair, Thomas Adams, Earl Bell, Sharon Childers,
Jeff Kozma, Peter Skylstad, Boyd Trolinger, Jerrad Van’t Hul, and Van Wu for their help
in observing nests. The NRA Whittington Center and V-7 Ranch provided access to their
lands and logistical support for this study. Steven R. Beissinger provided helpful comments
on this manuscript. This research is part of the BBIRD (Breeding Biology Research and
Monitoring Database) program, which is funded by the Global Climate Change Research
Program of the National Biological Service. This is BBIRD Publication # 5.

LITERATURE CITED

Bent, A. C. 1942. Life hi.stories of North American flycatchers, larks, swallows and their
allies. U.S. Natl. Mus. Bull. 179.

Blancher, P. J. and R. J. Robertson. 1985. Site consistency in kingbird breeding perfor-
mance: implications for site fidelity. J. Anim. Ecol. 54:1017-1027.

Dow, H. AND S. Fredga. 1983. Breeding and natal dispersal of the goldeneye, (Bucephala
clangula). J. Anim. Ecol. 52:681-695.

Harvey, P. H., P. J. Greenwood, and M. Perrins. 1979. Breeding area fidelity of great tits
(Pams major). J. Anim. Ecol. 48:305-313.

Hendricks, P. 1991. Site fidelity and renesting of female American pipits. J. Field Ornithol
62:338-342.

Martin, T. E. and J. J. Roper. 1988. Nest predation and nest-site selection of a western
population of the Hermit Thrush. Condor 90:50-57.

Newton, I. 1994. The role of nest sites in limiting the number of hole-nesting birds: a
review. Biol. Conserv. 70:265-276.

380

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

Nolan, V., Jr. 1978. Ecology and behavior of the Prairie Warbler (Dendroica discolor).
Ornithol. Monograph No. 26.

Shields, W. M. 1984. Lactors affecting nest and site fidelity in Adirondack barn swallows.
Auk 101:780-789.

SONERUD, G. A. AND P. E. Fjeld. 1987. Long-term memory in egg predators: an experiment
with a Hooded Crow. Ornis Scand. 18:323-324.

David R. Curson, Christopher B. Goguen, and Nancy E. Mathews, Univ. of Wisconsin-
Madison, Dept, of Wildlife Ecology, 226 Russell Lxibs, J630 Linden Drive, Madison, Wis-
consin 53706-1598. Received 21 June 1995, accepted 20 Dec. 1995.

Wilson Bull., 108(2), 1996, pp. 380-381

Nest sharing by a Lesser Scaup and a Greater Scaup. — Nest sharing has been loosely
defined as two females sharing a nest, incubating their eggs together, and (perhaps) sharing
in the care of the young (Torres 1982). It is a relatively uncommon phenomenon, reported
infrequently in ornithological literature (see Torres 1982, for a brief review). While con-
ducting field studies of nesting waterfowl on the islands of the North Arm of Great Slave
Lake (approximately 62°30'N 1 15°10'W) in June 1993, we discovered a clutch of 26 scaup
eggs which was being incubated by two females, one a Lesser Scaup (Aythya afftnis) and
one a Greater Scaup (Aythya niarila). Both females flushed at close range (although not
simultaneously) and were identified visually via wing stripe characteristics and size. Incu-
bation status was determined by female behavior, egg warmth, and amount of down present
at the nest.

The clutch of 26 eggs consisted of 17 “large” and nine “small” eggs, and may have
been the product of more than two females. Two eggs were cracked, possibly indicating
some aggressive interaction between the females. We measured a sample of eggs using
vernier calipers. Three large eggs averaged 63.9 X 43.5 mm, whereas four small eggs
averaged 56.8 X 42.5 mm. These measurements lie within the ranges reported for Greater
and Lesser scaup, respectively (Bent 1923, Bellrose 1976, Palmer 1976). The eggs were
laid in an oval-shaped depression lined with grass and were marginally concealed by a
clump of grass. This aiTangement provided ample room for two females to sit side by side,
probably in direct contact with each other, and thereby incubate virtually the entire clutch
simultaneously.

Subsequent inspection of this nest in late July revealed that it had been partially suc-
cessful. Seven membranes from hatched eggs were observed. In addition, six eggs were
found intact in the nest, four dead ducklings were still in their partially opened egg shells,
two dead ducklings were outside their egg shells but still in the nest, and one dead duckling
was found outside the nest. One egg which had been destroyed by a predator and was
assumed to belong to the same nest was found nearby. The fate of the remaining five eggs
could not be determined.

Skutch (1961) stated that unless the young of the two nest sharing species hatch at about
the same time, and are of similar size and feeding habits, it is unlikely that the young of
both species will survive. Given the ecological similarities between the two species of scaup,
it is unlikely that any resulting combinations of females and ducklings that survived through
departure from the nesting island would experience anything beyond the normal threats to
their survival. For example, mixed age (and thus mixed size) broods and broods attended

SHORT COMMUNICATIONS

381

by two or more Females have been reported for both species of scaup (Bellrose 1976, Palmer
1976) and are relatively commonly observed among breeding scaup in the Great Slave Lake
area (Fournier and Hines, unpubl. data).

LITERATURE CITED

Bellrose, F. C. 1976. Ducks, geese, and swans of North America. Stackpole Books. Har-
risburg, Pennsylvania.

Bent, A. C. 1923. Life histories of North American wild fowl. Part one. U.S. National
Museum. Bull. 126.

Palmer, R. S. 1976. Handbook of North American birds. Vol. 3, Waterfowl (Part 2). Yale
Univ. Press. New London, Connecticut.

Skutch, a. F. 1961. Helpers among birds. Condor 63:198-199.

Terres, J. K. 1982. The Audubon Society encyclopedia of North American birds. Alfred
A. Knopf Inc. New York, New York.

Michael A. Fournier and James E. Hines, Canadian Wildlife Sendee, P.O. Box 637, Yel-
lowknife, NWT, XI A 2N5. Received 20 Aug. 1995, accepted 9 Nov. 1995.

Wilson Bull., 108(2), 1996, pp. 381-382

Garni vory observed in the Cedar Waxwing.— On 9 Aug. 1992, I observed an adult
Cedar Waxwing (Bombycilla cedrorum) fly to a tree branch approximately 12 m above the
ground with an unidentified nestling bird in its bill. Through 7 X 35 binoculars it appeared
the nestling was naked, approximately 3 cm in length and being held by the tail. It had
been eviscerated, with the stomach hanging down to one side. The Cedar Waxwing paused
briefly after landing, then swallowed the nestling whole in approximately three successive
swallowing motions. Cedar Waxwings are frugivorous, with the exception of a relatively
small proportion of invertebrate prey (Tyler 1950). To my knowledge, carnivory has not
been reported for this species.

The species of the nestling was not known. It is unlikely it was a Brown-headed Cowbird
(Molothrus ater), since cowbirds are not abundant in the area (pers. obs.), and Cedar Wax-
wings do not readily accept cowbird eggs (Friedmann 1963). It seems most probable the
nestling was a Cedar Waxwing and was eaten either as a form of infanticide or was a dead
nestling removed from the nest for hygienic reasons.

Infanticide has been reported for a variety of avian taxa (Stanback and Koenig 1992).
Non-nutritional motivations for infanticide include the removal of unrelated offspring by a
replacement mate and the lowering of the reproductive output of competitors (Stanback and
Koenig 1992). A number of other individuals of this species were observed in the area,
providing opportunity, and perhaps the motive, for infanticidal behavior. Although Cedar
Waxwings may be less predisposed to cannibalism because of their largely frugivorous diet
(Stanback and Koenig 1992), they are colonial nesters (Tyler 1950) which, combined with
the tendency to swallow foods whole (Tyler 1950), could potentially increase their predis-
position towards cannibali.sm (Mock 1984, Stanback and Koenig 1992). Whether or not this
observation represented cannibalism or predation of another species, it certainly represents
a bizarre deviation from cu.stomary Cedar Waxwing diet.

382

THE WILSON BULLETIN • Vo/. 108, No. 2, June 1996

LITERATURE CITED

Eriedmann, H. 1963. Host relations of the parasitic cowbirds. Bull. U.S. Natl. Mus. Bull.
233.

Mock, D. W. 1984. Infanticide, siblingicide, and avian nesting mortality. Pp. 3—33 in In-
fanticide: comparative and evolutionary perspectives (G. Hausfater and S. B. Hrdy,
eds.). Aldine, New York.

Stanback, M. T. and W. D. Koenig. 1992. Cannibalism in birds. Pp. 277-298 in Canni-
balism: ecology and evolution among diverse taxa (M. A. Elgar and B. J. Crespi, eds.).
Oxford Univ. Press, New York, New York.

Tyler, W. M. 1950. Cedar Waxwing, in Life histories of the North American wagtails,
shrikes, vireos and their allies (A. C. Bent, ed.). Bull. U.S. Natl. Mus. 197.

David I. King, Dept, of Forestry and Wildlife Conservation, Univ. of Massachusetts, Am-
herst, Massachusetts 01003. Received 3 Oct. 1995, accepted 1 Dec. 1995.

Wilson Bull., 108(2), 1996, pp. 382-384

A case of cooperative breeding in the Hooded Warbler. — Cooperative breeding in-
volves one or more individuals, in addition to the genetic parents, giving parental care to
offspring (Stacey and Koenig 1990, Emlen 1991). In birds, this parental care may take
several forms, such as feeding nestlings, nest construction, incubation, defense against pred-
ators, and territory defense (Stacey and Koenig 1990). The social organization of cooperative
breeders occurs in a variety of forms including (1) nonreproductive adults helping their
parents raise young, (2) “plural breeders” where more than one monogamous pair within a
social group breeds simultaneously, (3) “highly gregarious” monogamous cooperative
breeding groups, and (4) polyandrous or polygynandrous cooperative breeding groups (Sta-
cey and Koenig 1990, Krebs and Davies 1991). Cooperative breeding in birds is relatively
rare, existing in only 2.4% (220 of 9000) of avian species (Stacey and Koenig 1990, but
see Emlen and Vehrencamp 1983). Herein, we report the first documented case of cooper-
ative breeding in a warbler, the Hooded Warbler (Wilsonia citrina).

Methods. — Hooded Warblers are small, migratory songbirds that breed in selectively-
logged mixed hardwood deciduous forests. We conducted this research in Crawford County,
Pennsylvania (41°N, 79°W) as part of a two-year mating system study from May- August
1994 and 1995. The mating system is socially monogamous, with one male and one female
occupying a single breeding territory (Stutchbury et al. 1994, Evans Ogden and Stutchbury
1994). We discovered the nest where cooperative breeding occurred on June 16, 1995. When
the nestlings were five days old, a banded female, a banded male (B) and an unbanded male
(U) were caught with mist nets near the nest. Unbanded adults were banded with U.S. Fish
and Wildlife aluminum bands and unique color band combinations to identify individuals.
Upon returning nestlings to the nest after banding them, the female and U male began
chipping rapidly near the nest while the B male chipped rapidly approximately 10-15 m
away. This peripheral male then flew to the nest and fed one of the nestlings.

To determine if both males were feeding nestlings, the nest was video-taped from 08:00-
09:00 EDT each day for four days. Playback experiments were also conducted to determine
the role of the males in territory defense. A recording of male “repeat” and “mixed” mode
song patterns (Wiley et al. 1995) was used. After feeding rate observations were complete

SHORT COMMUNICATIONS

383

on the fourth day, three playback experiments of a lO-min. duration were conducted in mid-
moining in the center of the teixitory. A model of a male Hooded Warbler was placed beside
the playback speaker. Playbacks were conducted at approximately 30-min. intervals. Re-
peating the playback experiment confirmed that the same male was responding each time.

Results. — Mean feeding rates (±SE) (N = 4 h) for the female was 7.4 ± 3.0 trip.s/h.
Mean feeding rates for the B and U males were 9.6 ± 1.6 trips/h and I.I ± 0.2 trips/h
respectively. The feeding rate for the B male was significantly higher (unpaired /-test, / =
5.29, df — 6, P = 0.002) than that of the U male. Feeding rates did not differ significantly
between either the female and the B male (/-test, / = -0.65, df = 6, P = 0.54) or the U
male (/-test, / = 2.12, df = 6, P = 0.08).

The degree of aggressive behavior during playbacks varied between the two males. Both
males chipped rapidly and countersang within one minute of the start of the playback.
However, only the B male circled repeatedly around the model (at a radius of 10-15 m),
changed its perch frequently and flew within 5 m of the model twice. The U male stayed
about 20-25 m away. Neither male physically attacked the model. Each time I approached
the nest, all three adults flew to the nest in defense.

Discussion. Previously in this study area, feeding behavior to nestlings has been ob-
served for 5-6 h/nest for about 60 nests, with no prior instances of cooperative breeding
(Evans Ogden and Stutchbury 1994, Neudorf, unpubl. data). Therefore, cooperative breeding
is rare (1/60 or 1.5%) in Hooded Warblers. The “auxiliary” male (U) could have shared
paternity with the “dominant” male (B) because about 40% of female Hooded Warblers
produce extra-pair young from fertilizations with neighboring males (Stutchbury et al. 1994).
High levels of extra-pair matings could increase the likelihood of cooperative breeding, just
as It may favor adoption of fledglings in this species (Stutchbury and Evans Ogden, unpubl.
data). In 1994, the auxiliary male was pair-bonded with this same female on a nearby
territory, but was genetically unrelated to his offspring that year (Tarof, Stutchbury and
Piper, unpubl. data). This cuckoldry, along with subsequent mate-switching the following
year by the female, may be explained by the relative low quality of this auxiliary male.
Why the dominant male permits an ASY (after-second year) auxiliary male to help raise
offspring is a question that has perplexed researchers since cooperative breeding was first
reported (Skutch 1935), particularly with a territorial species such as Hooded Warblers.
Although males were never seen at the nest together, no overt aggression was observed. We
expect that cooperative breeding may be widespread in warblers, although it likely occurs
at very low frequencies within a species.

Acknowledgments. — Diane Neudorf and Joan Howlett provided advice and pertinent com-
ments on the manuscript. We appreciate the assistance of Margaret Tarof, Diane Neudorf,
Lesley Evans Ogden, and Trevor Pitcher with fieldwork. A Natural Sciences and Engineering
Research Council of Canada scholarship to S.A.T. supported this research.

LITERATURE CITED

Emlen, S. T. 1991. Evolution of cooperative breeding in birds and mammals. Pp. 301-337
in Behavioural ecology: an evolutionary approach. Third ed. (J. R. Krebs and N. B.
Davies, eds.). Blackwell, Oxford, England.

AND S. L. Vehrencamp. 1983. Cooperative breeding strategies among birds. Pp.
93-133 in Perspectives in ornithology (A. H. Brush and G. A. Clark Jr., eds.). Cam-
bridge Univ. Press, Cambridge, England.

Evans Ogden, L. and B. J. Stutchbury. 1994. Hooded Warbler. In The Birds of North
America, No. 1 10 (A. Poole, P. Stettenheim, and E Gill, eds.). Philadelphia: The Acad-
emy of Nat. Sci., Washington, D.C.

384

THE WILSON BULLETIN • Vol. JOS, No. 2, June 1996

Krebs, J. R. and N. B. Davies. 1991. Behavioural ecology: an evolutionary approach.
Blackwell, Oxford, England.

Skutch, a. L 1935. Helpers at the nest. Auk 52:257-273.

Stacey, R B. and W. D. Koenig. 1990. Cooperative breeding in birds: long-term studies
of ecology and behaviour. Cambridge Univ. Press, Cambridge, England.

Stutchbury, B. J., J. M. Rhymer, and E. S. Morton. 1994. Extra-pair paternity in hooded
warblers. Behav. Eeol. Sociobiol. 5:384—392.

Wiley, R. H., R. Godard, and A. D. Thompson Jr. 1995. Use of two singing modes by
hooded warblers as adaptations for signalling. Behaviour 129:243-278.

Scott A. Tarof and Bridget J. Stutchbury, Dept, of Biology, York Univ. 4700 Keele St.,
North York, Ontario, M3J IPS, Canada. Received 3 Oct. 1995, accepted 1 Jan. 1996.

Wilson Bull., 108(2), 1996, pp. 385-394

ORNITHOLOGICAL LITERATURE

Edited by William E. Davis, Jr.

A BIRDER s GUIDE TO ARKANSAS. By Mel White. American Birding Association, Inc.,
Colorado Springs, Colorado. 1995:259 pp., 53 maps, 20 black-and-white photographs, 16
line drawings. $16.95 (soft wrap-around cover). — This is the seventh in the series of the
American Birding Association’s birdfinding guide series, which is a revision and extension
of the Lane series which served birders well for many years. Like the other six guides
published so far, this book is well-written, well-edited, and well-designed and a must for
anyone interested in birds visiting an area covered by one of tbe series. The book is sturdy,
printed on heavy paper with a wire-o binding and a wrap-around cover wbicb serves to
protect the book and/or serve as a book mark. Its roughly 15 by 21.5 cm size is small
enough to fit into a car’s glove box and some jacket pockets. It is designed to be used.

An introductory chapter has sections on the physiography of the state, the birding cal-
endar, and historical and ecological notes. The book is written with humor, and the section
entitled Weather and other pests,” which provides useful information on ticks, chiggers,
mosquitos, poisonous snakes, etc., concludes with “Try not to walk off any cliffs.” The
major portion of the text describes 59 birding sites clustered into five geographic regions
bounded by county lines. Each regional cluster of sites has an introduction and map of site
locations, followed by detailed descriptions, including a map, of each site. The site maps
are clear and readable, and descriptions of directions are given to a tenth of a mile — an
excellent feature. The usual list of useful telephone numbers for the local birding “hotline,”
travel information, etc., are supplemented by the numbers to call for information for any of
the sites in state or federal parks or preserves. Following the site guide chapters are sections
on specialty birds and bar graphs for each species giving status (e.g., “hard to miss,”
lucky to find”) throughout the year. Additional appendices list “seldom-seen” and “ac-
cidental” birds, and lists of Arkansas mammals, amphibians, reptiles, and butterflies.

I cannot imagine anyone interested in birds visiting Arkansas without taking a copy of
this book along. — William E. Davis, Jr.

Stories I like to tell: an autobiography. By H. Elliott McClure. Privately printed,
Camarillo, California. 1995:373 pp., 110 black-and-white photograph figures. Available in
paperback for $14 (S&H inch) from H. Elliott McClure, 69 E. Loop, Camarillo, California
93010-2327. — This is the story of an entomologist turned ornithologist and naturalist who
began his professional career as a wildlife biologist studying doves in the mid-west and
ended it nearly 35 years later with involvement in an international bird-banding .scheme in
Asia. This very personal narrative includes recollections of youthful adventures from dodg-
ing streetcars with his bicycle to an awe-inspiring viewing of a local doctor’s butterfly
collection. The pages are full of interesting anecdotes, including many which are bird related.
Twenty-five years were spend in Asia, beginning with an encephalitis study in Japan, and
many adventures describe exotic birds and places. There are chapters on hornbills and
bulbuls and lots of birdwatching adventures in, for example, Australia, the Philippines,
Thailand, and Malaya. A lot of historical information on M.A.P.S. (Migratory Animals
Pathological Survey) Asian project is woven into the latter part of the book. Several hundred
photographs (usually several per figure) present a photographic collage of everything from
family photos to scorpions and hornbills.

Despite the minor annoyance of the many typos, occasional grammatical collapses, lack

385

386

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

of an index, and a table of contents with no page numbers, I found this a delightful book
full of the fun and adventures of an interesting ornithological career. The narrative is en-
gaging, and there was enough about birds, conservation, and natural history to satisfy my
ornithological cravings. — William E. Davis, Jr.

The hornbills. By Alan Kemp, illustrated by Martin Woodcock. Oxford University Press,
New York. 1995: 302 pp., 14 color plates, black-and-white line drawings, 53 range maps,
7 tables, glossary. $60. — The hornbills, (Order Buceroti formes sensu Sibley and Monroe.
1990. “Distribution and Taxonomy of Birds of the World.” Yale Univ. Press, New Haven,
Connecticut) are separated by Kemp into two families, including one genus and two species
of very large (to 4000 g) ground hornbills (Bucorvidae) and 52 species and eight genera of
medium to large (83-3300 g) true hornbills (Bucerotidae). The taxonomy of hornbills seems
to be in a state of flux, and there are some taxonomic changes included in this book —
primarily the elevation of subspecies to species level and the inclusion of some new common
names. Lor example, Sibley and Monroe (1990) refer to Tockus albicristatus as the White-
crested Hornbill, a translation of the specific name, whereas Kemp refers to it as the Long-
tailed Hornbill. Kemp apparently used Long-tailed Hornbill to reduce confusion with the
White-crowned Hornbill (Aceros comatus) which has also at times been called the White-
crested Hornbill.

The ground hornbills are found only on the savannas of Africa; the true hornbills are
forest birds of Africa, south and east Asia, and many of the islands of the Indian Ocean.
Most of the hornbills have a large bill with a casque which has suggested functions ranging
from sex and age recognition to sound reception, to use in foraging, nest maintenance, or
territorial defense. In a dramatic introduction, territorial male Great Helmeted Hornbills
(Buceros vigil) rivals are described as meeting in a manner analogous to the behavior of
rams — by flying at one another and butting heads in mid-air! Unfortunately, it turns out that
the behavior has yet to be substantiated (Cranbrook and Kemp, 1995. Ibis 137:588-589).

Hornbills are cavity nesters, most taking advantage of natural tree cavities. In the true
hornbills, the female seals herself in and undergoes a complete molt of flight feathers during
the nesting effort. As a result of their large size and associated need for large trees for their
cavity nests and because of the limited distribution of many species, hornbills may be
particularly susceptible to becoming endangered as a result of elimination of old growth
forest. Range maps are shown for all but the Visayan Wrinkled Hornbill (Aceros waldeni),
but these are of historic ranges and populations of many species are already fragmented and
declining. At least some species will use nest boxes, and an appendix provides guidelines
for captive management and breeding. These management tools are, however, just that —
tools. They are not answers to long-term survival. Only protection of natural habitats and
ecosystems can provide a future for these birds. Kemp acknowledges the use of hornbills
as human food and for medicinal and religious purposes in various cultures and suggests
that any efforts to manage the species must consider these human needs. Since most of the
larger species produce two eggs, but raise only a single chick, he suggests removal of the
.second chick as one means of meeting the.se needs.

Tables include a wealth of morphological and ecological data begging for further analysis
and comparison with other taxa. Plates include not only superb field guide style composites,
showing each species with sex and age variants, but also ones showing all Asian and all
African hornbills in flight. Colored photographs illustrate some habitats, nests, and a skull
carved by a Chinese craftsman. The bibliography of nearly 600 references includes some
unpublished reports and citations from an incredible array of journals and often obscure

ORNITHOLOGICAL LITERATURE

387

series. Certainly, this is a resource that in itself will facilitate further advancement of our
knowledge ot this group. The three-page glossary dehnes some hornbill-specific terms and
also a strange assortment of other ornithological and non-ornithological (e.g., “selective
logging,” “swidden agriculture”) terms. “Rectrix” is misspelled in the glossary.

The Hornbills is obviously a labor of love, dedication, and deep understanding, and I
thoroughly enjoyed reading it. It is a model tor monographic coverage that would be difficult
to surpass. — Jerome A. Jackson.

Woodpeckers. By Hans Winkler, David A. Christie, and David Nurney. Houghton Mifflin
Company, Boston. 1995: 406 pp., 64 color plates, some black-and-white figures, 214 range
maps. $40. (hardcover). — This guide to the woodpeckers of the world follows on the heels
of somewhat similarly comprehensive monographs on woodpeckers by Short (1982. “Wood-
peckers of the World,” Delaware Museum of Natural History Monogr. Ser. No. 4, Green-
ville, Delaware) and Frugis, Malaguzzi, Vicini and Cristina (1988, “Guida ai Picchi del
Mondo. Museo Regionale di Scienze Natural! Monogr VIE Torino, Italy.” In Italian.). How
does Winkler et al. compare to these other monographs?

Winkler et al. recognize 214 species, following the systematic arrangement of Sibley and
Monroe (1990. “Distribution and Taxonomy of Birds of the World.” Yale Univ. Press, New
Haven), which was based largely on Short (1982). Short and Frugis et al. recognize “about”
200 species, with Frugis et al. following Short closely. The taxonomic differences in Winkler
et al. result in general from recognition of Short’s subspecies of two piculets and several
woodpeckers as distinct species.

Introductory material in Winkler et al. includes 35 pages that describe the layout of species
accounts, the nature of taxonomic decisions made, descriptions of tribes and genera of
woodpeckers, basic zoogeography, morphology, plumages and molt, foods and foraging,
habitats, ecology, behavior, and interactions with man. Short’s introductory coverage is sim-
ilar in scope and depth; Frugis et al. is a skeleton by comparison, covering a similar scope
of material in much less depth but including useful drawings illustrating foraging patterns
and a bit more discussion of conservation problems than found in introductory material of
the others.

Color plates by Nurney in Winkler et al. include several species/subspecies per page in
somewhat field guide fashion. These focus primarily on adult males but also show heads of
adult females. The plates in general are excellent, although the bills of Campephilus species
seem too slender and “weak.” Detailed captions face each plate; captions in the other books
are essentially limited to species identification. The plates by Sandstrom in Short (1982) are
larger but with generally fewer species per page, sometimes showing species-specific habitat
characteristics. The plates by Cristina and Vicini in Frugis et al. often include only one or
two species per page, are generally adequate but less refined, and not as well reproduced
as in the other books. The plates in Short and Frugis et al. focus only on adult males.

Species accounts in Winkler et al. include a range map, statement of identifying charac-
teristics and distribution, description of habitat, sex and age differences, geographic varia-
tion, sample measurements, and descriptions of voice, habits, food, and timing of breeding.
Short’s species accounts provide a similar range and depth of information. Species accounts
in Frugis et al. include both English and Italian common names, listing of subspecies, and
very brief descriptions of distinctive characteristics, habitat, distribution, and for some spe-
cies, notes on behavior and diet. There are no range maps in either Short or Frugis et al.

In format and in being more recent, Winkler et al. have the most useful and comprehensive
of these monographs. Their “Woodpeckers” is much more than indicated by the subtitle on

388

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

the dust jacket (one of three subtitles used!): “An Identification Guide to the Woodpeckers
of the World.” All of these books are useful introductions to the diversity, behavioral ecol-
ogy, and adaptive strategies of the Picidae. Frugis et al. was intended only as an illustrated
checklist of the woodpeckers of the world, and it succeeds quite well in that. In comparing
the two more comprehensive books, however, one must also consider the backgrounds of
the authors, for these certainly have influenced the nature of the finished products. Winkler
has studied woodpeckers in many parts of the world and has emphasized vocalizations in
much of his work. As might be expected, text material for European species is much stron-
ger. Short has spent a lifetime studying woodpeckers throughout the world with an emphasis
on taxonomic studies. His studies in the New World and in Africa have been extensive and
are strongly reflected in his monograph.

Having established the similarities and differences in these works, I will now focus on
the details included in Winkler et al. With the incredible amount of information included in
Winkler et al., some problems could be expected and some are pre.sent. On p. 12, the authors
refer to Dryocopus magellanicus as the Magellanic Woodpecker, stating that it is “often
seen as a good intermediate between Dryocopus and Celeus. . .” Surely this is in error, since
elsewhere throughout the book (and among other authors) the Magellanic Woodpecker is
treated as Campephilus magellanicus. Furthermore, the species generally considered inter-
mediate to Dryocopus and Celeus is Dryocopus galeatus, the Helmeted Woodpecker, a
species with very limited range in Brazil, Paraguay, and Argentina.

As a measure of accuracy, I most carefully examined accounts of species with which I had
the greatest familiarity. These accounts were generally very good with very up-to-date infor-
mation, but I found several minor problems with the Red-cockaded Woodpecker (Picoides
borealis) account. The range map for this species appropriately shows a very fragmented
distribution but inappropriately seems to show the species still occupying Missouri, Tennessee,
and Maryland and a much wider range in Kentucky than has ever been known. It also shows
the species absent from some areas of the mid-South where it is known. In the caption tor
Plate 34, in reference to the adult female Red-cockaded Woodpecker, the authors indicate that
this is the “only ‘ladder-backed’ species in its range.” Since this is a guide intended for
identification, and since this species is considered endangered, it is important to note the
“ladder back” of the very common Red-bellied Woodpecker (Melanerpes carolinus) which
occurs throughout the range of the Red-cockaded. Finally, the list of pine species used as nest
sites by the Red-cockaded is incomplete (e.g., Finns virginiana isn’t included) and seems to
inappropriately downplay the importance of loblolly pine {Pinus taeda) relative to others.

Another problem is the perpetuation of the notion that juvenile female Hairy Woodpeckers
(Picoides villosus) normally have an “orange-red patch on crown” (pp. 35, 291). The fe-
males very rarely have even a reduced orange-red patch. Eye color mentioned for the Ivory-
billed Woodpecker (Campephilus principalis) follows Short (1982) and is given as “white
to creamy-white” (p. 354), whereas numerous specimen labels and descriptions of early
naturalists who closely observed the species almost always refer to the yellow color of the
iris. The range map for the Pileated Woodpecker (Dryocopus pileatus) shows the species
much farther north than it is known to occur in Canada. These are all minor problems that
are sometimes ambiguous or incorrect elsewhere in the literature, illustrating that there is
still much to be learned about woodpeckers.

In general, “Woodpeckers” is an up-to-date and thorough compendium of information
about the world’s woodpeckers. It is put together in a user friendly format with good doc-
umentation and outstanding illustrations. Winkler et al. have brought us to a new plateau in
our understanding of woodpeckers and have made it easier for us to surge on ahead. —
Jerome A. Jackson.

ORNITHOLOGICAL LITERATURE

389

Endangered ecosystems of the united states: a preliminary assessment of loss and
DEGRADATION. By Reed E Noss, Edward T. LaRoe III, and J. Michael Scott. U.S. Department
of the Interior, National Biological Service, Biological Report 28, Washington, D.C. 1995:
58 pp., three numbered text figs., eight black-and-white photographs with captions, five
appendices, no charge (paper). — The status of natural ecosystems throughout the United
States is cuiTently a topic of considerable concern. Unfortunately, minimal information is
available to determine the geographical extent and condition of the ecosystems of the United
States, and the quality of information varies greatly from state to state. In spite of these
difficulties, Noss et al. have done a superb job of defining ecosystems to evaluate and
compiling and analyzing large databases in order to estimate the loss and degradation of
United States ecosystems. The authors’ arguments for managing and protecting ecosystems
to prevent the necessity to have to deal with threatened and endangered species focuses on
the primary problem of endangered species management.

The booklet is well written and treatments of topics are presented clearly in a logical
fashion. The authors list ecosystems as critically endangered (30), endangered (58), or threat-
ened (38), with the real “meat” of the publication located in the appendices where the
extent of loss for each ecosystem is presented. The authors indicate that degradation and
losses of ecosystems have been most pronounced in the South, Northeast, Midwest, and
California. Akso evaluated is the potential risk for further loss within each ecosystem. Be-
cause pristine sites of many ecosystems are already nearly nonexistent, the authors suggest
that restoration be an integral part of ecosystem management. The authors show where
significant information gaps occur. The illumination of these gaps should help state and
federal agencies adjust methodologies and priorities for data collection. This publication is
essential for researchers and managers who work with threatened, endangered, or sensitive
species and any aspect of ecosystem management. The authors should be commended for
this excellent publication. — Richard N. Conner.

This fragile land. A natural history of the Nebraska sandhills. By Paul A. Johns-
gard. Univ. Nebraska Press, Lincoln, Nebraska. 1995:xv -F 256 pp. 48 numbered figs., 5
tables. $35.00 (cloth).— Warning! Read this book and you will never again travel across the
Great Plains on 180 single-mindedly driven by purple mountains majesty over the western
horizon. You will be obliged to make a significant diversionary journey northward into the
largest area of sand dunes in the Western Hemisphere. But once on the blue highways, of
which there are very few, you will discover, as Johnsgard writes, “Most roads in the sand-
hills lead nowhere.” But that is their charm, “a land of no straight lines . . . patiently shaped
by water, wind, and time.” It’s a landscape laid down in the Miocene, formed during the
post-glacial, stabilized by prairies, richly endowed with rivers, brooks, fens, and marshes
and with a human population of about one person per square mile. This is an ecological
natural history that illustrates the diverse biotic communities in a region that comprises
almost a quarter of the area of the state. While grasshoppers and grasshopper mice (complete
with sonogram), midges, and sticklebacks (yes, they do it in Nebraska just like they do it
for European ethologists) are described, birds are well represented, including some bird
transect data gathered by H. Elliot McClure in the mid-l940’s. Part One takes you around
the periphery — the valley of the Niohrara, the (ponderosa) pine ridge and High Plains, the
Platte, tallgrass praiiie, and cornfields. Part Two takes you into the interior with chapter
subtitles such as Sandreed, Sicklehills, and ’Roo Rats;” “Boots, Burrowing Owls and Box
Turtles; Whispers, Bells, and Trumpets.” Part Three de.scribes the advent of the ranching
economy in the nineteenth century and the development of center-pivot irrigation during

390

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

the last third of the twentieth century. Irrigation has not depressed the water table very much
because of the rapid recharge rates; but as a consequence of the ease at which water returns
to the aquifers deep in the sand, irrigation has led to widespread ground-water pollution by
nitrates, herbicides, and pesticides. But there is hope. Johnsgard tells us that the people who
are making their living in the sandhills have banded together to thwart external economic
and political pressures to increase short-term profit at the expense of long-term productivity.
The seven appendices in the book comprise 30% of its length and provide, for example, a
checklist of the vertebrates and vascular plants indicating the biotic communities in which
they occur, a distributional list of sandhills birds, an annotated list of natural areas in the
sandhills, and a very helpful glossary. Most of you have read other books by the author.
This is Johnsgard at his poetic best, not only in the prose, but also in his delightful line
drawings that complement the text. — John L. Zimmerman.

Where to watch birds in South America. By Nigel Wheatley. Princeton Univ. Press,
41 William St., Princeton, New Jersey 08540. 1995: 431 pp., 52 black-and-white bird figs.,
108 maps, $35.00 US (hardback). — This is a highly useful work that discusses an entire
continent and related offshore archipelagos. As some recent books on South American birds
have ignored the Galapagos and Falklands, it was good to see them included. Each country
(and archipelago) has an introduction, with notes on size, getting around, lodging and food,
health and safety concerns, climate and best months to visit, habitats, and conservation;
along with country species totals, highlights, and endemic species. The author is a numbers
gatherer with many accountings and tables of birds and endemic families per country;
species numbers per size of country; statistics on trip lists, day lists; top ten sites for species
lists; and endemic family lists per country. The most useful lists are each country’s endemics
with brief notes on where such endemics are found (with 179 in Brazil, for instance).

The sites within each country are clustered, rather than alphabetized, and well-labeled
and numbered in each country’s locator map. The nearly 100 site maps are well executed.
One would wish that there were maps for all 206 sites. A typical site account has a good
introduction, .separate lists of endemics, specialties, others (birds), other wildlife (usually a
few mammals, good feature), and access notes. Many sites are published for the first time
in a book with broad availability.

The highly competitive business of bird li.sting companies competing with each other for
the same clients has led to many great spots being kept semi-secret. Independent travelers
have often remarked on how rarely they weaseled information out of such companies or
their trip leaders. Some tips on where to go would end up on competitor’s itineraries. The
outdated code of ethics that birders must share all finding information with no charge for
cost of gathering or preparing such information for strangers was in place before the tour
businesses. Grossly underpaid bird tour leaders pay top dollar for legal, medical, financial
and other consulting advice, but when those same professionals are birders they demand
our consulting advice for free. This book will save tour leaders countless hours of pro hono
work answering the questions so well answered in this work.

The actual species lists given for each locale are confusingly broken up into three cate-
gories. What a pain to have to search through three incomplete lists running like prose text.
1 would have much preferred two columns side-by-side with annotations as to endemics,
specialties, and others (via bolds and italics), along with crude abundance symbols (where
such information is available). As “no one reads introductions,” the reader will have a hard
time figuring out what the asterisks mean (tip: it’s buried on page 18). Space does preclude
complete lists in a book treating an entire continent, but I felt that North American migrants
and winterers were rarely covered.

ORNITHOLOGICAL LITERATURE

391

How good are the locale lists? To answer this, I compared my notes from three recent
annual visits to the excellent Costanera Sur Faunal Reserve adjacent to downtown (city
centre) Buenos Aires, Argentina. Taking civilized non-birders en route to Antarctica at mid-
morning and mid-afternoon hours, with no tapes or playbacks, and without wading in the
marshes, I found 82 species. We saw four of his 12 specialties listed, 36 of his 51 “Others,”
and 42 species not mentioned. The author does state that common and widespread species
are purposely left off such lists. Not listed in the book were such species as White-faced
Ibis (Plegadis chihi) (much commoner than two other ibises listed), the “southern” endemic
Coscoroba Swan {Coscoroba coscoroba) (large numbers in “dry” years). Red Shoveler
(A>ias clypeata) (always there), rarities such as Masked Duck {Oxyura dominica) and Pam-
pas Paintedsnipe {Nycticryphes semicollciris) and interesting common species for first-time
visitors such as Rufous Hornero (Funmrius rufus) and Chalk-browed Mockingbird (Mimus
satuniinus).

Ornithologists should note that not a single scientific name appears in the book; other
checklists will have to be handy. Fortunately, South America was nearly devoid of British
colonies and a confusing plethora of “common” names in English (as has existed in Africa
and Asia) did not evolve. Common sense common names with unique modifiers that tied
species to correct group-names were agreed upon in an early (recent) period by several
American ornithologists. It is interesting to note that it was a British birder who put together
this work despite the much greater number of Americans traveling there. As expected, there
was no mention of the Malvinas name in the FalkJands chapter. There was also no mention
of the 22 pioneering chapters on South America in “Finding Birds Around the World”
(Peter Alden and John Gooders, Houghton Mifflin, Boston, 1981).

Mistakes of commission appeared few. Like so many writers, he says the train climbs UP
to the ruins of Machu Picchu (7374 feet) from Cusco (10,200 feet) in Peru. Listing of
significant inbound tour operators and lodge addresses is most useful, although I doubt the
Venezuelan Audubon Society is the main avenue for booking Hato Pinero in the llanos. The
listing of Antarctic ships in this 1995 guide has no mention of the new inexpensive Russian
ships on the scene for several years, while mentioning the Ocean Princess which has been
out of service since April 1993.

Despite the high price tag for a book with no color plates, this is a work that will repay
Its cost easily m directing casual and serious visitors to the most representative accessible
parks and wild areas on the bird continent. — Peter Alden.

A GUIDE TO the birds OF MEXICO AND NORTHERN CENTRAL AMERICA. By Steve N.G.
Howell and Sophie Webb. Oxford Univ. Press, New York. 1995. 1,010 pp., 1087 maps 71
color plates, 44 linecuts. $39.95 paper, $75.00 cloth.— This long awaited handbook covers
all birds known from Mexico, Belize, Guatemala, Honduras, and western Nicaragua. The
fine color plates by Sophie Webb cover many plumages and subspecies for the first time.
The plates in my copy appear to be overexposed, as the colors seem a bit washed out
compared to .some originals I’ve seen. Most North American winterers are not illustrated,
so that visitors will again have to take along one of the North American field guides. The
range maps (all at the species level) are excellent and show where birds are present in
migration, as well as breeding and wintering areas. These maps are done on an outline of
provinces which is far better than doing so with a few rivers or no provinces as some books
present them. The species accounts are exhaustive, with good attention to vocalizations,
similar species, ranges, relative abundances, habitats, taxonomic questions, and major sub-
species. With such excellent plates, it appears that the descriptions are too long. Every shade.

392

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

streak, and pattern is included in minute detail, often running on for many hundreds of
words.

Steve N. G. Howell, from Britain, embarked on this project in November 1981 and
proceeded to spend the better part of the next dozen years gathering identification, ecolog-
ical, distribution, and vocalization information in thousands of sites in the field. Based at
the Point Reyes Bird Observatory, he queried hundreds of (chiefly American) colleagues
for details (note the five pages of acknowledgments) and supervised the production of plates
and range maps. Most of the field guide work in Latin America has focused on areas from
Costa Rica southward since the early 1970s when no less than four were published on the
Mexico/Guatemala region. It is so nice to see such a compendium of knowledge come out
on this area after a gap of roughly 20 years, especially since this area is an important one
for wintering Nearctic birds, contains many endemics and has a rapidly diminishing stock
of pristine habitat.

The book suffers from an overuse of abbreviations, and the various keys to this plethora
of them are scattered in too many places. Some abbreviations are listed on unnumbered
page 87, some on unnumbered pages xv and xvi, others in acknowledgments (by extrapo-
lation), and others such as INIREB, ICACH, RSMHN, and SEOUL, on page 798. They
should all have been listed from A-Z on both endpapers for reference. Why can’t Chiapas
be Chia instead of Chis? Couldn’t LCUNAM be shortened somehow? Under each species
account the sections are split up into ID, SS, SO, RA, and NB, which could have been
spelled out or at least given as Ident., Sim. sp.. Status, Distr., Range and Note. In a short
sample species account such as Worm-eating Warbler Helmitheros vermivorus the reader
will have to translate ID, SE, SD, L, C, S, Ver, U, R, S, Tamps, Nay, PA, SNGH, PP, L, U,
NE, R, W, DL, RA, E, and SE. This makes for unpleasant reading. It took a long time to
figure out what the asterisk indicated. On page 87 you will find that it means different things
depending on where it is placed.

The maps on pages 2—6 appear to have enough space to have most names of islands,
provinces, and selected cities and towns written out at the correct site rather than resorting
to a endless letters and numbers. The biogeographic maps on pages 8-9 are yet another
nightmare of letters and numbers. Despite this, the geography and bird distribution section
is well organized. The various tables on migration, visitors and species that withdraw from
parts of ranges are of great interest. Note that all the microscopic asterisks are actually tiny
letter a, b, and c symbols explained on page 35. I was pleased to see the country-by-country
chapters on conservation. Maps showing sites of present day reserves would have been
helpful.

Rare indeed these days is the author of regional guides who fails to “lump and split”
any number of species and attempt to improve English names. This book is no exception.
The sequence of orders and families follows the A.O.U. Checklist of North American Birds
(1983) rather than the more controversial DNA-DNA sequence. The “lumping/splitting”
pendulum swings over to splits here with four species created out of the Green Parakeet
(Aratinga holochroa) and four species created out of the Fork-tailed Emerald (Chlorostilhon
canivelii). There are several dozen other accepted splits involving such genera as Amazona
parrots, Glaiicidium pygmy-owls, Caprinmlgus nightjars, Campylopterus sabrewings, Cy-
nanthiis and Amazilia hummingbirds, Trogon, Formicarius ant-thrushes, Platyrinchus spa-
debills, Progne martins, Stelgiclopleryx swallows, Hylorchilus and Troglodytes wrens, Vireo,
Icterus orioles, and Junto. The Yellow-namped Warbler (Dendroica coronata) is not divided.
Many birders will be surprised to see the flowerpiercer genus Diglossa placed in the Eni-
herizinae.

As for English names he fails to use Louisiana Heron as a good alternate to Tricolored
Heron (Egretta tricolor), coins Eared Quetzal (better than Trogon) for Euplilos neoxemus.

ORNITHOLOGICAL LITERATURE

393

and fails to adopt Whitestart lor Myiohoriis. Members of Myiohorus genus (all of which
have white outer tail feathers) have no red in the tail and should not retain the grossly
incorrect name applied to Old World thrushes of the genus Phoenicurus. I note that the
silver-like color is spelled grey. An informal group of ornithologists involved with English
names in the 1980’s proposed that the Americans might use grey (not gray) in return for
the British dropping the use of the u in colour and harbour.

Useful appendices include lists of species known from islands and cays of both the Pacific
and Atlantic coasts, descriptions of fifty species of eastern Honduras that are omitted from
the main text, and a 26 page bibliography that somehow omits my “Finding Birds in Western

Mexico” (University of Arizona Press, Tucson, 1969) that would have been listed first!

Peter Alden.

Current ornithology. Volume 12. Dennis M. Power (ed.). Plenum Press, New York,
New York. 1995:278 pp.$79.50. — Plenum Press continues to provide outstanding service to
the community of scientific ornithologists by publishing this series. The present volume
includes chapters on (1) testosterone and polygyny in birds by Les D. Beletsky, David E
Gori, Scott Freeman, and John C. Wingfield, (2) use of migration counts to monitor landbird
populations by Erica H. Dunn and David J. T. Hussell, (3) ptilochronology by Thomas C.
Grubb, Jr., (4) individual voice discrimination by Marcel M. Lambrechts and Andre A.
Dhondt, (5) evolution of bird coloration and plumage elaboration by Udo M. Savalli, and
(6) hatching asynchrony and the onset of incubation by Scott J. Stoleson and Steven R.
Beissinger. As has been true in all previous volumes, the editing and writing of these chap-
ters is excellent and the content remarkably good, particularly given the length of the series.

I will not describe the contents of the chapters here as the titles pretty much tell the story.
The accounts on migration and hatching asynchrony were particularly interesting to me and
promise to have wide application in ornithology. The chapters on voice discrimination,
evolution of color, and ptilochronology forced me to revise and expand my course notes for
my university class in ornithology. The chapter on polygyny illustrates how physiology and
behavior can be integrated to better understand birds. As usual, these sources are a gold
mine for researchers, graduate students, and teachers. — C. R. Blem.

Last of the curlews. By Fred Bodsworth. Counterpoint, Washington D.C. 1995. $15
(cloth). This little book of less than a hundred pages has become a conservation classic
since it was published in 1955 by Dodd, Mead & Company. This edition has been expanded
by including a Forward by Pulitzer Prize-winning poet and conservationist W.S. Mervin, a
brief Epilogue by Bodsworth, and a 45 page Afterword by Nobel Prize-winning particle
physicist Murray Gell-Mann. Gell-Mann’s Afterward begins with the question: “Can the
human race learn, while there is still time, how to coexist with the great diversity of bird
life on this planet? In the pages which follow, he examines the question of extinction in
the context of geologic time and past periods of catastrophic extinction of species and then
focuses on the changes to the biosphere which humans have produced in the past few
centuries and the enormous stress we continue to put on the ecological systems of the planet.
He eventually enters an arena which is perhaps best described as philo.sophical or political
with a discussion of quality of human life versus quantity, at one point asking: “Why
squander quality of life for the sake of mere numbers of humans?” In a world already
overcrowded with humans, this is perhaps the most fundamental question one could ask.

394

THE WILSON BULLETIN • Vol. 108. No. 2, June 1996

The book itself contains 1 1 chapters which comprise a biographical sketch of the hypo-
thetical “last of the curlews,” an Eskimo Curlew (Numenius borealis). It is a touching story
told by a biologist with a deep understanding of shorebird biology. The skillful avoidance
of anthropomorphism is quite remarkable, as is the author’s use of language to evoke an
emotional response from the reader. The chapters are separated by short sections entitled
“The Gantlet” [gauntlet] which consist of vignettes quoted from published historical doc-
uments which delineate the slaughter of Eskimo Curlews by man and the virtual disap-
pearance of the species.

This classic story of human exploitation and extinction, together with the penetrating
analysis of global problems by Gell-Mann, are well worth reading and discussing, particu-
larly in times of political and economic change. — William E. Davis, Jr.

Erratum

In the paper “Detectability and population density of Scaly-naped Pigeons before and
after Hurricane Hugo in Puerto Rico and Vieques Island,” (Wilson Bulletin, 107[4];727—
733), the first sentence should be: “Hurricane Hugo hit northeastern Puerto Rico with sus-
tained winds of 30-40 m/s (gusting to 50-60 m/s) on 18 September 1989 (see Boose et
al., 1994).”

INFORMATION FOR AUTHORS

The Wilson Bulletin publishes significant research and review articles in the field of
ornithology. Mss are accepted for review with the understanding that the same or similar
work has not been and will not be published nor is presently submitted elsewhere, that all
persons listed as authors have given their approval for submission of the ms, and that any
person cited as a personal communication has approved such citation. All mss should be
submitted directly to the Editor.

Text. Manuscripts should be prepared carefully in the format of this issue of The Wilson
Bulletin. Mss will be returned without review if they are not properly prepared. They
should be neatly typed, double-spaced throughout (including tables, figure legends, and

Literature cited ), with at least 3 cm margins all around, and on one side of good quality
8.5" X 1 1" paper. Do not use erasable bond. Mss typed on low-quality dot-matrix printers
are not acceptable. The ms should include a cover sheet (unnumbered) with the following:

( 1 ) Title, (2) Authors, their institutions, and addresses, (3) Name, address, and phone number
of author to receive proof, (4) A brief title for use as a running head. All pages of the text
through the Literature cited should be numbered, and the name of the author should
appear in the upper right-hand corner of each. The text should begin in the middle of the
first numbered page. Three copies should be submitted. Xerographic copies are acceptable
if they are clearly readable and on good quality paper.

Tables. Tables are expensive to print and should be prepared only if they are necessary.
Do not repeat material in the text in tables. Tables should be narrow and deep rather than
wide and shallow. Double space all entries in tables, including titles. Do not use vertical
rules. Lise tables in a recent issue of the Bulletin as examples of style and format. Tables
should be typed on separate unnumbered pages and placed at the end of the ms.

F/gMrci.— Illustrations must be readable (particularly lettering) when reduced in size.
Final size will usually be 1 1.4 cm wide. Illustrations larger than 22 X 28 cm will not be
accepted, and should be reduced photographically before submission. Legends for all figures
should be typed on a separate page. Photographs should be clear, of good contrast, and on
glo,ssy paper. Drawings should be in India ink on good drawing board, drafting paper, or
blue-lmed graph paper. Figures produced by high-quality laser printers may be acceptable.
All lettering should be done with a lettering instrument or printer. Do not use typewriter
lettering and do not submit figures having adhesive transfer letters affixed to them. Designate
the top of each illustration and label (on the back in soft pencil) with author’s name, ms
title, and figure number. Submit two duplicates or readable xerographic copies of each figure
as well as the original or high-contrast glossy photo of the original.

Authors of accepted papers are urged to submit voucher photographs of their work to
Visual Re.sources for Ornithology (VIREO) at the Academy of Natural Sciences of Phila-
delphia. Accession numbers from VIREO will then be published within appropriate sections
of the paper to facilitate access to the photographs in subsequent years.

Style and format.— The current issue of The Wilson Bulletin should be used as a guide
for preparing your ms; all mss must be submitted in that format. For general matters of
style authors should consult the “CBE Style Manual,” 5th ed.. Council of Biology Editors,
Inc., Bethe.sda, MD, 1983. Do not use footnotes or more than two levels of subject sub-
headings. Except in rare circumstances, major papers should be preceded by an abstract, not
to exceed 5% of the length of the ms. Abstracts should be informative rather than indicative,
and should be capable of standing by themselves. Most units should be metric, and com-
pound units should be in one-line form (i.e., cm-.sec ^-). The continental system of dating
(19 Jan. 1950) and the 24 hour clock (09:00, 22:00) should be u.sed, and the Standard Time
specified (e.g., EST for Eastern Standard Time) at first reference.

References.— In both major papers and general notes, if more than four references are
cited, they should be included in a terminal “Literature cited” section. Include only refer-

395

396

THE WILSON BULLETIN • Vol. 108, No. 2, June 1996

ences cited in the ms, and only material available in the open literature. (“In-house” reports
and the like should not be cited.) Use recent issues of the Bulletin for style, and the most
recent issue of “BIOSIS,” BioScience Information Service, Philadelphia, PA, for abbrevi-
ations of periodical names. If in doubt, do not abbreviate serial names. Manuscripts with
fewer than hve references should be cited internally, e.g., (Sprenkle and Blem, Wilson Bull.
96:184—195) or Sprenkle and Blem (Wilson Bull. 96:184-195).

Nomenclature. — Common names and technical names of birds should be those given in
the 1983 A.O.U. Check-list (and supplements as may appear) unless justification is given.
Lor other species the Bulletin uses the common names in Sibley and Monroe, “Distribution
and Taxonomy of Birds of the World.” Common names of birds should be capitalized. The
scientific name should be given at first mention of a species both in the abstract and in the
text.

The editor welcomes queries concerning style and format during your preparation of mss
for submission to the Bulletin. — Charles R. Blem, Editor.

This issue of The Wilson Bulletin was published on 1 June 1996.

The Wilson Bulletin

Editor Charles R. Blem

Department of Biology
Virginia Commonwealth University
816 Park Avenue
Richmond, Virginia 23284-2012

Assistant Editors Leann Blem

Albert E. Conway

Editorial Board Kathy G. Be At.

Richard N. Conner
Thomas M. Haooerty
John A. Smali.wood

Review Editor William E. Davis, Jr.

127 East Street

Foxboro, Massachusetts 02035

Index Editor Kathy G. Beal

616 Xenia Avenue
Yellow Springs, Ohio 45387

Suggestions to Authors

See Wilson Bulletin, 108:395—396, 1995 for more detailed “Information for Authors.”
Manuscripts intended for publication in The Wilson Bulletin should be submitted in triplicate,
neatly typewritten, double-spaced, with at least 3 cm margins, and on one side only of good
quality white paper. Do not submit xerographic copies that are made on slick, heavy paper. Tables
should be typed on separate sheets, and should be narrow and deep rather than wide and shallow.
Follow the AOU Check-list (Sixth Edition, 1983) insofar as scientific names of U.S., Canadian,
Mexican, Central American, and West Indian birds are concerned. Abstracts of major papers
should be brief but quotable. In both Major Papers and Short Communications, where fewer than
5 papers are cited, the citations m.ay be included in the text. Follow carefully the style used in
this issue in listing the literature cited; otherwise, follow the “CBE Style Manual” (AIBS, 1983).
Photographs for illustrations should have good contrast and be on glossy paper. Submit prints
unmounted and attach to each a brief but adequate legend. Do not write heavily on the backs of
photographs. Diagrams and line drawings should be in black ink and their lettering large enough
to permit reduction. Original figures or photographs submitted must be smaller than 22 X 28 cm.
Alterations in copy after the type has been set must be charged to the author.

Notice of Change of Address

If your address changes, notify the Society immediately. Send your complete new address to
Ornithological Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

The permanent mailing address of the Wilson Ornithological Society is: c/o The Museum of
Zoology, The University of Michigan, Ann Arbor, Michigan 48109. Persons having business with
any of the officers may address them at their various addresses given on the back of the front
cover, and all matters pertaining to the Bulletin should be sent directly to the Editor.

Member.ship Inquiries

Membership inquiries should be sent to Dr. John Smallwood, Dept, of Biology, Montclair State
Univ., Upper Montclair, New Jersey 07043.

CONTENTS

MAJOR PAPERS

GEOGRAPHIC VARIATION AND SPECIES LIMITS IN CINNYCERTHIA WRENS OF THE ANDES

Robb T. Brumfield and J. V. Remsen, Jr.

NEST ATTENTIVENESS IN HUMMINGBIRDS William H. Baltosser

SEASONAL POPULATION SURVEYS AND NATURAL HISTORY OF A MICRONESIAN BIRD COMMUNITY ...
Robert J. Craig

NATURAL HISTORY AND CONSERVATION STATUS OF THE TAMARUGO CONEBILL IN NORTHERN CHILE
Christian F. Estades

AVIAN ABUNDANCE IN RIPARIAN ZONES OF THREE FOREST TYPES IN THE CASCADE MOUNTAINS,
OREGON Robert G. Anthony, Gregory A. Green, Eric D. Eorsman, and S. Kim Nelson

HABITAT CHANGES AND SUCCESS OF ARTIFICIAL NESTS ON AN ALKALINE FLAT

Marcus T. Koenen, David M. Leslie, Jr., and Mark Gregory

NESTING ECOLOGY OF SCISSOR-TAILED FLYCATCHERS IN SOUTH TEXAS

- Kenneth R. Nolte and Timothy E. Eulbright

BREEDING BIOLOGY OF THE BROWN NODDY ON TERN ISLAND, HAWAII

Jennifer L. Megyesi and Curtice R. Griffin

DISCRIMINATION BETWEEN REGIONAL SONG FORMS IN THE NORTHERN PARULA

Daniel J. Regelski and Ralph R. Moldenhauer

DISPERSAL AND HABITAT USE BY POST-FLEDGING JUVENILE SNOWY EGRETS AND BLACK-CROWNED
NIGHT-HERONS R. Michael Erwin, John G. Haig, Daniel B. Stotts, and Jeff S. Hatfield

NEST-SITE SELECTION OF RED-SHOULDERED AND RED-TAILED HAWKS IN A MANAGED FOREST

Christopher E. Moorman and Brian R. Chapman

SHORT COMMUNICATIONS

AVOIDANCE OF CABBAGE RELDS BY SNOW GEESE J. Russell Mason and Larry Clark

TAXONOMIC STATUS OF THE CUBAN FORM OF THE RED-WINGED BLACKBIRD

— - Orlando Garrido and Arturo Kirkconnell

NEST ADOPTION BY MONK PARAKEETS Jessica R. Eberhard

VERMILION FLYCATCHER AND BLACK PHOEBE FEEDING ON FISH

Brenda J. Andrews, Marie Sullivan, and J. David Hoerath

NEST-SITE REUSE IN THE WESTERN WOOD-PEWEE

David R. Curson, Christopher B. Goguen, and Nancy E. Mathews

NEST SHARING BY A LESSER SCAUP AND A GREATER SCAUP

- Michael A. Fournier and James E. Hines

CARNIVORY OBSERVED IN THE CEDAR WAXWING David /. King

A CASE OF COOPERATIVE BREEDING IN THE HOODED WARBLER

- Scott A. Tarof and Bridget J. Stutchbury

ORNITHOLOGICAL LITERATURE.....

20!

12\

24(

26f

28C

292

302

317

335

342

357

369

372

374

377

378

380

381

382

385

V

Tlie Wilson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. 108, NO. 3 SEPTEMBER 1996 PAGES 397-606

(ISSN (X)43-5643)

The Wilson Oknitholikhcal Society
Founded December 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President — Keith L. Bildstein, Hawk Mountain Sanctuary, RR 2, Box 191, Kempton, Pennsylvania
19529-9449.

First Vice-President — Edward H. Burtt, Jr., Department of Biology, Ohio W'esleyan University,
Delaware, Ohio 43015.

Second Vice-President — John C. Kricher, Biology Department, Wheaton College, Norton, Mas-
sachusetts 02766.

Editor — Charles R. Blem, Department of Biology, Virginia Commonwealth University, P.O. Box
842012, Richmond, Virginia 23284-2012.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper Mont-
clair, New Jersey 07043.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana
46556.

Elected Council Members — Carol A. Corbat and William E. Davis (terms expire 1997), and
Margaret C. Brittingham and Herbert T. Hendrickson (terms expire 1998), Peter C. Frederick
and Danny J. Ingold (terms expire 1999).

Membership dues per calendar year are: Active, $21.00; Student, $15.00; Family, $25.00; Sus-
taining, .$30.00; Life memberships $500 (payable in four installments).

The WiI-SON Bulletin is sent to all members not in arrears for dues.

The Jo-sselyn Van Tyne Memorial Library

The Josselyn Van Tyne Memorial Library of the Wilson Ornithological Society, housed in the
University of Michigan Museum of Zoology, was established in concurrence with the University
of Michigan in 1930. Until 1947 the Library was maintained entirely by gifts and bequests of
books, reprints, and ornithological magazines from members and friends of the Society. Two mem-
bers have generously established a fund for the purchase of new books; members and friends are
invited to maintain the fund by regular contribution, thus making available to all Society members
the more important new books on ornithology and related subjects. The fund will be administered
by the Library Committee, which will be happy to receive suggestions on the choice of new books
to be added to the Library. William A. Lunk, University Museums, University of Michigan, is
Chairman of the Committee. The Library currently receives 195 periodicals as gifts and in ex-
change for The. Wilson Bulletin. With the usual exception of rare books, any item in the Library
may be borrowed by members of the Society and will be sent prepaid (by the University of
Michigan) to any address in the United States, its possessions, or Canada. Return postage is paid
by the borrower. Inquiries and requests by borrowers, as well as gifts of books, pamphlets, reprints,
and magazines, should be addressed to: The Josselyn Van Tyne Memorial Library, University of
Michigan Museum of Zoology, Ann Arbor, Michigan 48109. Contributions to the New Book Fund
should be sent to the Treasurer (small sums in stamps are acceptable).

The Wilson Bulletin
(ISSN 0043-5643)

THE WILSON BULLETIN (ISSN 004.5-.364.3) is published quarterly in March. June. September, and December by the W ilson
Omitholonical .Society. 810 East lOtli Street. Ijiwrence. KS 66044-8897. The subscription price, both in the United Slates and
elsewhere, is $40.(M) per year. Berimlicals postage paiti at I,.awrence. KS. POSTMASTER; Send address changes to THE WILSON
BULLETIN. P.O. Box 1897. Uwrence. KS 66044-8897.

Back issues or single copies are available for $I2.(K) each. Most back issues of the Bulletin are available and may be onlered
from the Treasurer. Special prir-es will be quoted for (piantily ortlers.

,411 articles and communications for publications. b(H)ks and publications for reviews should be addressed to the Editor. Ex-
changes should be addressed to The Josselyn Van Tyne Memorial Library. Museum of Zoology. Ann .Arbor. Michigan 48109.

Subscriptions, changes of address and claims for undelivered copies should be sent to the O.SNA. P.O. Box 1897. Liwrence.
KS 66(H4-8897. Phone: (91.-}) 848-1221; FAX: (918) 848-1274.

© Copyright 1996 by the Wilson Ornithological Society
Printed by Allen Press. Inc.. Lawrence, Kansas 66044, U.S.A.

@ This paper meets the requirements of ANSi/NISO Z39.48-1992 (Permanence of Paper).

Frontispiece. Adult (right) and juvenile (left) Acrobatornis fonsecai. Pink-legged Grav-
eteiro, a new genus and species in the Furnariidae from southeastern Bahia, Brazil. Painting
by Paul Donahue.

THE WILSON BULLETIN

A QUARTERLY MAGAZINE OF ORNITHOLOGY

Published by the Wilson Ornithological Society

VoL. 108, No. 3 September 1996 Pages 397-606

Wilson Bull., 108(3), 1996, pp. 397-433

A NEW GENUS AND SPECIES OF FURNARIID
(AVES: FURNARIID AE) FROM THE COCOA-GROWING
REGION OF SOUTHEASTERN BAHIA, BRAZIL

Jose Fernando Pacheco,' Bret M. Whitney,'- and Luiz P Gonzaga'

Abstract. — We here describe Acrobatornis fonsecai, a new genus and species in the
Furnariidae, from the Atlantic Forest of southeastern Bahia, Brazil. Among the outstanding
features of this small, arboreal form are: black-and-gray definitive plumage lacking any
rufous; juvenal plumage markedly different from adult; stout, bright-pink legs and feet; and
Its acrobatic foraging behavior involving almost constant inverted hangs on foliage and
scansorial creeping along the undersides of canopy limbs. Analysis of morphology, vocal-
izations, and behavior suggest to us a phylogenetic position close to Asthenes and Crani-
oleuca, in some respects, it appears close to the equally obscure Xenerpestes and Meto-
pothrix. New data on the morphology, vocalizations, and behavior of several funiariids
possibly related to Acrobatornis are presented in the context of intrafamilial relationships.
We theorize that Acrobatornis could have colonized its current range during an ancient
period of continental semiaridity that promoted the expansion of stick-nesting prototypes
from a southern, Chaco-Patagonian/Pantanal center, and today represents a relict that sur-
vived by adapting to build its stick-nest in the relatively dry, open, canopy of leguminaceous
trees of the contemporary humid forest in southeastern Bahia. Another theory of origin
places emphasis on the fact that the closest relatives of practically all (if not all) other birds
syntopic with Acrobatornis are of primarily Amazonian distribution. Acrobatornis fonsecai
has a most unusual distribution in a restricted region in which lowland Atlantic Forest has
been converted virtually entirely to cocoa plantations. Until very recently a lucrative and
vitally important source of income for Bahia, the economic base for cocoa production has
suffered catastrophic, apparently irrecoverable, decline owing to “witch's broom” disease,
which has proven resistant to all forms of control. The predictable wave to cut and sell the
tall trees shading failing cocoa plantations has already begun in earnest with the consequence
that the remnant forest canopies in this region, upon which Acrobatornis fonsecai is totally
dependent, are being rapidly destroyed. This remarkable new furnariid and the secrets it
holds for elucidation of phylogeny, evolutionary history, speciation patterns, and zoogeog-
raphy, if not safeguarded immediately, when its habitat is still for sale, could disappear in
the coming decade. Received 23 April 1996, accepted 21 May 1996.

' Instituto de Biologia, Depto. de Zoologia. Cidade Universitaria, Univer.sidade Federal do Rio de Janeiro
21941-000, Rio de Janeiro. RJ, Brasil.

= Museum of Natural Science. 1 19 Foster Hall, Louisiana State Univ., Baton Rouge, Louisiana, 70803.

397

398

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Resumo. — Acrobatomis fonsecai, um novo genero e nova especie de Lurnariidae, e des-
crito da Mata Atlantica do sudeste da Bahia, Brasil. Centre os aspectos notaveis desta
pequena ave arbon'cola estao: a sua plumagem definitiva cinza e preta sem qualquer aver-
melhado; plumagem juvenil bem diferenciada da adulta; patas e pes fortes de colora9ao rosa
vivo; e seu comportamento acrobatico de forrageamento de envolve uma quase permanente
posigao invertida dependurada na folhagem e uma “escalada negativa” ao longo das su-
perficies inferiores dos galhos da copa. Atraves de analises da morfologia, vocalizagoes, e
comportamento, e sugerido uma posi^ao filogenetica de Acrobatomis proximo aos Asthenes
e Cranioleuca\ em alguns aspectos o novo genero mostra-se proximo aos igualmente ob-
scuros Xenerpestes e Metopothrix. Novos dados sobre a morfologia, vocaliza96es, e com-
portamento de varios Lurnariidae relacionados a Acrobatomis sao apresentados no contexto
das afinidades intrafamiliares. E teorizado que Acrobatomis colonizou sua atual area de
distribui^ao durante um periodo remoto de semi-aridez continental, que promoveu a expan-
sao oriunda do sul, Chaco e Patagonia, dos prototipos construtores de ninhos de graveto.
Lie hoje representaria uma “relfquia” que conseguiu sobreviver, no sudeste da Bahia, ao
periodo umido contemporaneo, por adaptar construir seu ninho de gravetos nas copas re-
lativamente secas e abertas das arvores leguminosas. Uma outra teoria sobre a origem en-
fatiza o fato de que os parentes mais prdximos de quase a totalidade (se nao todos) das
aves sintopicas com Acrobatomis possuem uma distribui^ao principalmente amazonica.
Num padrao incomum de distribui9ao, Acrobatomis fonsecai ocorre numa regiao restrita da
Mata Atlantica de tabuleiro do sudeste da Bahia, que tern sido quase completamente con-
vertida em planta9oes de cacau. A cacauicultura, ate muito recentemente lucrativa e sub-
stancial fonte de receita para a Bahia, tern sofrido um catastrofico, aparentemente irrecu-
peravel, declmio na produ9ao devido a dissemina9ao da “vassoura-de-bruxa,” doen9a cau-
sada por fungos, que tern resistido a diversas formas de controle. De maneira grave, as
grandes arvores copadas, remanescentes da floresta Integra, das quais Acrobatomis fonsecai
6 totalmente dependente, ja estao sendo cortadas e vendidas. Este notavel novo furnarldeo,
e os segredos que ele guarda para elucida9ao da filogenia, historia evolutiva, padroes de
especia9ao e zoogeografia, se nao imediatamente salvaguardados, quando o seu habitat ainda
esta a venda, poderao desaparecer na proxima decada.

From a continental perspective, the Neotropical family Furnariidae has
undergone a geographical and ecological radiation, paralleled by a diver-
sity of form and function, without equal in the world of birds. The Fur-
nariidae contains more than 230 species in 53 genera (as currently de-
fined; Sibley and Monroe 1990, 1993). To this magnificent assemblage
we must now add one more genus and species, a member of such singular
appearance, ecology, and distribution as to mark it truly outstanding even
in a family characterized by adaptive extremes. Perhaps most remarkable,
however, is our discovery of this new form in the remnant Atlantic Forest
of southeastern Bahia, a heavily populated region in which we never
would have predicted the contemporary existence of a furnariid of such
affinities.

On 26 January 1988, in search of habitat that might harbor the little-
known Stresemann’s Bristlefront {Merulaxis stresemanni), Whitney lo-
cated a slope cloaked in undisturbed Atlantic Forest in the serra das Lon-

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 399

tras above the village of Itatingui in the cocoa-growing region of southern
Bahia. Although observations were precluded by rain, the area was tar-
geted for a detailed avifaunal survey. In November 1994, following an
ornithological investigation of selected points in interior Minas Gerais
and Bahia ending in the city of Salvador, Whitney suggested to Pacheco
and Paulo Sergio M. da Fonseca that they make an introductory visit to
Itatingui along their return drive to Rio de Janeiro. Documentation of a
largely unknown avifauna in Atlantic Forest fragments in the highest parts
of the serra de Ouricana approximately 125 km WNW of Itatingui (Gon-
zaga et al. 1995), and recent discoveries of undescribed species there
(Gonzaga and Pacheco 1995, Pacheco and Gonzaga 1995), indicated an
urgent need to explore the Atlantic Forest of southern Bahia north of the
Rio Jequitinhonha and interior from the relatively well known coastal
forests. Thus, on the morning of 17 November 1994, while observing
birds in a mixed-species flock foraging in the canopy of trees shading a
cocoa plantation at the edge of undisturbed forest above Itatingui, Fonseca
called Pacheco s attention to a strange pair of birds, one gray-and-black
and the other largely brown, creeping along the undersides of limbs, and
hanging acrobatically on clusters of foliage and flowers of a tall Croton
tree. They were able to observe the birds for several minutes, noting
details of the plumage and foraging behavior. It was clearly a species
unknown from Brazil, and quite possibly unknown altogether. After dis-
cussion of this exciting news with Gonzaga and Whitney, Pacheco, Fon-
seca, and Claudia Bauer returned to Itatingui in late January 1995 and
obtained four specimens, including two of each “morph” they had iden-
tified previously.

It was clear even on cursory inspection that the specimens represented
a species unknown to science, and further examination indicated a strong
probability that a new genus would have to be erected as well. Rather
than describe the new form immediately, it was decided to return to Ita-
tingui in October, when the birds would likely be breeding and vocal, so
that more ecological and distributional information could be gathered.
Thus, from 2 to 12 October 1995, Pacheco, Fonseca, Whitney, and Barth
explored the serra das Lontras above Itatingui, and a number of other
localities in the general region. This expedition was highly successful,
resulting in: the discovery and collection of the nest and the location of
52 additional nest sites which defined, we think to a large extent, the
distributional limits of the new bird; the tape-recording of several indi-
viduals, which documented the vocal repertoire of the species almost
completely; the tape-recording and collection of two additional voucher
specimens, which permitted preparation of a skeleton and preservation of
additional biochemical material; and the production of photographs and

400

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Fig. 1. Adult Acrobatomis fonsecai, gen. nov. sp. nov. Video image captured from Hi-8
format original (11 Oct. 1995 near Camacan, Bahia; video by Whitney).

video recordings of habitat, nests, and foraging maneuvers. Furthermore,
we determined that the known distribution of the new species was entirely
coincident with the cocoa-growing province of southern Bahia — the con-
sequences of which appear to be devastating for the continued survival
of the new bird. Finally, in March 1996, Whitney located the new species
at additional localities in the cocoa-growing region slightly north and west
of previously documented ones.

Consideration of all these data, informed through extensive personal
field experience with the Furnariidae, including all but one genus and 1 1
currently recognized species, and comparison with museum anatomical
and skin specimens from a cross-section of potentially related genera,
convinces us that the unknown form indeed represents a new genus and
species in the Furnariidae. The descriptions follow.

Acrobatomis gen. nov.

TYPE-SPECIES: Acrobatomis fonsecai Pacheco, Whitney, and Gonzaga.

DIAGNOSIS. — A small, arboreal furnariid (weight 14 g) completely lacking rufous in
definitive plumage, and largely orange-tawny in Juvenal plumage. Adults basically gray with
black wings, tail, and cap (Frontispiece, Fig. 1). Tail shorter than the wing (with tail/wing
ratio varying from 0.86 to 0.92, with an average of 0.89); strongly graduated and slightly

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 401

stiffened, composed of 12 rectrices. Two inner pairs of rectrices strongly acuminate at the
tips and deeply excised on the inner web; the acumination at the tips of the rectrices de-
creases gradually from innermost to outermost, so that outermost are almost blunt, cuneiform
in shape and only slightly excised on the inner web. Outer pair of rectrices about 60% of
the length of the innermost pair, and projecting far beyond the under tail coverts. Wing
relatively long in comparison to most furnariids of similar mass, and extending well beyond
base of the tail. Seventh to ninth primaries longest and approximately equal in length; sixth
primary only slightly shorter; tenth and fifth primaries of nearly equal length and about 5
mm shorter than longest ones.

Bill compressed, short and pointed; culmen slightly decurved, skull pseudo-schizorhinal
(Fig. 2A). Tarsi short and strong (Fig. 2A), conspicuously pink-colored. Hind toe thicker
than front toes; claw of hind toe slightly shorter than the length of the toe itself. Sternum
two-notched. Syrinx typically furnarioid, with well-defined Membranae tracheales. Process!
vocales without “horns,” and two pairs of intrinsic syringeal muscles (Fig. 3). Membranae
tracheales limited posteriorly by A-2 (A-3 vestigial), and anteriorly by a drum formed by
the partial fusion of two elements.

SPECIMENS EXAMINED: SKINS. — Only specimens actually measured are listed; many
others were compared superficially. Acrobatornis fonsecai, gen. nov, sp. nov.: Brazil: Bahia,
2 males (Museu de Zoologia da Univ. de Sao Paulo [hereafter MZUSP] No. 74154 [holo-
type], Museu Paraense Emilio Goeldi [hereafter MPEG] No. 52345), 1 female (MZUSP
74155), and 3 sex unk. (MZUSP 74156 [juv.], MPEG 52346 [juv.], and Louisiana State
Univ. Museum of Natural Science [hereafter LSUMZ] 160000 [ad.]). Cranioleuca pyrrho-
phia: Bolivia: Santa Cruz, 3 males (LSUMZ 124033, 124036, 124040). C. curtata: Peru:
San Martin, 1 male (LSUMZ 86368); Pasco, 1 male (LSUMZ 130232); Ayacucho, 1 male
(LSUMZ 69428); Bolivia: Cochabamba, 1 male (LSUMZ 37666). C. pallida: Brazil: Sao
Paulo, 1 male (LSUMZ 63352). C. albiceps: Bolivia: La Paz, 3 males (LSUMZ 95950,
101981, 101983). Asthenes dorbignyi: Peru: Arequipa, 2 males {arequipae: LSUMZ 1 14142,

1 19197); Bolivia: La Paz, 2 males {consobrina: LSUMZ 101995, 101996). A. baeri: Bolivia:
Santa Cruz, 3 males (LSUMZ 153692, 153693, 153698); Argentina: Corrientes, 1 male
(LSUMZ 54632); Uruguay: 1 male (Academy of Natural Sciences [hereafter ANSP] No.
169843). A. patagonica: Argentina: Chubut, 2 (LSUMZ 73269, ANSP 186350). Thripopha-
ga fusciceps: Bolivia: Beni, 2 males, 1 female (LSUMZ 124062, 124063, 124065). Pha-
cellodomus sibilatrix: Bolivia: Santa Cruz, 2 males, 1 female (LSUMZ 153700, 153701,
153702); Argentina, Chaco, 1 male (LSUMZ 83932). P. rufifrons: Brazil: Minas Gerais, 2
males (LSUMZ 65165, 65166), Mato Grosso, 1 male (LSUMZ 80295); Bolivia: Santa Cimz,

2 males (LSUMZ 124070, 124074). Siptomis striaticollis: Peru: Cajamarca, 3 males
(LSUMZ 87015, 87016, 87017); Colombia: Huila, 1 male (ANSP 155470). Xenerpestes
minlosi: Panama: Darien, 1 sex unk., (ANSP 150153); Colombia: Bolivar, 1 female (ANSP
160747). X. singulars: Ecuador: Morona-Santiago, 1 female (ANSP 176812), Zamora Chin-
chipe, 1 female (ANSP 185397); Peru: San Martin, 2 females, 1 sex unk. (LSUMZ 84690,
84691, 84692). Metopothrix aurantiacus: Ecuador: Napo, 2 males (LSUMZ 70898, 82963);
Peru: Loreto, 1 male (LSUMZ 1 19657). Margaromis squamiger: Bolivia: La Paz 3 males
(LSUMZ 95985, 95989, 95991).

SPECIMENS EXAMINED: ANATOMICALS.— Skulls illustrated: Acrobatornis fonse-
cai, gen. nov., sp. nov.: Brazil: Bahia (MPEG 3762 from skin specimen MPEG 52345).
Cranioleuca pyrrhophia: Bolivia: Santa Cruz (LSUMZ 125825). C. albiceps: Bolivia: La
Paz (LSUMZ 101317). Asthenes baeri: Bolivia: Santa Cruz (LSUMZ 153909). Asthenes
dorbignyi: Bolivia: La Paz (LSUMZ 101323). Thripophaga fusciceps: Bolivia: Beni (bill;
LSUMZ 124064). Phacellodornus sibilatrix: Bolivia: Santa Cruz (LSUMZ 153910). Xener-
pestes singularis: Peru: San Martin (bill; LSUMZ 84692). Metopothrix aurantiacus: Peru:

402 THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

A

Eig. 2. Cranial and tarsal profiles of: A. Acrohatornis fonsecai gen. nov. sp. nov. (in-
terorbital septum and frontal destroyed by shot); B. Asthenes baeri\ C. A. dorbignyr, D.
Cranioleuca albiceps\ E. C. pyrrhophia to show the pseudo-schizorhinal skull of Acroba-
tonii.s and to permit general comparisons, especially bill shapes and relative thicknesses and
lengths of tarsi, with some relevant taxa (continued).

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 403

Fig. 2. (continued) Cranial and tarsal profiles of: F. Thripophaga fusciceps (no skull
available); G. Phacellodomus sibilatrix\ H. Xenerpestes singularis (no skull available); I.
Metopothrix aurantiacus; J. Margaromis squamiger for comparison with Acrohatoniis fon-
secai gen. nov. sp. nov. in 2A. Drawings by Dan Lane.

404

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Eig. 3. Syrinx of Acrobatornis fon.secai gen. nov. sp. nov. showing well-defined Mem-
branae tracheales and two pairs of intrinsic syringeal muscles, features that place it in the
Furnarioidea. The lack of “horns” on the Process! vocales (PV) unequivocally places the
new genus in the Furnariidae (as opposed to the Dendrocolaptidae). MT = M. tracheola-
teralis; MVD = M. vocalis dorsalis; MW = M. vocalis ventralis; MS = M. sternotrachealis;
PV = Processus vocalis; A-1 and B-1 = cartilaginous elements. Nomenclature follows Ames
(1971). Drawing by Gonzaga.

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 405

Loreto (LSUMZ 51912). Margarornis sqiiamiger. Peru: Pasco (LSUMZ 129867). Syrinx:
Acrobatoniis fonsecai: Brazil: Bahia (from holotype).

SPECIMENS EXAMINED: TAPE-RECORDINGS. — The voices of all but about 20 spe-
cies of furnariids were included in initial comparisons with the vocalizations of Acrobatoniis
gen. nov. sp. nov. Selected recordings from which sound spectrograms were made are listed,
with pertinent data, in the figure legends. All recordings will be archived at the Library of
Natural Sounds (hereafter LNS), Cornell Laboratory of Ornithology, Ithaca, New York, and
those made in Brazil, also at the Arquivo Sonoro Elias P. Coelho (ASEC), Universidade
Federal do Rio de Janeiro, Rio de Janeiro.

BIOCHEMICAL SPECIMENS. — All biochemical specimens stored at LSUMZ, with the
following tissue collection catalog numbers: B-26329 (voucher MZUSP 74154, holotype;
male; blood); B-26330 (voucher MZUSP 74155; female; blood and liver in separate tubes);
B-26331 (voucher MPEG 52345; male; blood and liver in separate tubes).

ETYMOLOGY. — From the Greek words akrobatos, and ornis, referring to the acrobatic
climbing and hanging foraging behavior of this bird; the root akrobatos is immediately
understood in a remarkably diverse set of languages. The name is masculine in gender.

Acrobatornis fonsecai sp. nov.

Pink-legged Graveteiro
Acrobata (Portuguese)

HOLOTYPE. — MZUSP No. 74154; adult male from 15°1 1'S, 39°23'W, at approximately
550m elevation in the serra das Lontras above Itatingui, Municipality of Arataca, Bahia,
Brazil; 25 January 1995; collected by J. F. Pacheco, prepared by L. P. Gonzaga. Blood
sample housed at LSUMZ, No. B-26329. Not tape-recorded.

DISTRIBUTION. — Apparently restricted to the region of southeastern Bahia between the
drainage of the Rio de Contas in the north (known to just north of Ibirataia, 14°02'S,
39°40'W) and the Rio Jequitinhonha in the south (southernmost record near Teixeira do
Progresso, approximately 15°45'S, 39°28'W), occurring west at least as far as Ipiaii ( 14°06'S,
39°42'W) and east as far as the vicinity of Itabuna (14°48'S, 39°17'W) (Fig. 4). Altitudinal
distribution from near sea-level to approximately 550 m.

DESCRIPTION OF HOLOTYPE. — The two descriptive colors, “medium-gray” and
“dark-gray,” correspond to Munsell® Soil Color Chart (1994 ed.) 7.5YR 5/1 and 7.5YR
4/1, respectively. Crown black. Frontal and loral feathers (which are normal, not elongated
or stiffened), posteriorly to approximately the anterior edges of the orbits, medium-gray with
black apical margins imparting a grizzled appearance to these regions. Tiny feathers of malar
and suborbital regions margined blackish more narrowly than the frontal feathers. Supercil-
iary stripes subtly whiter, about 2.5 mm wide and 10 mm long posterior to orbit, blending
into medium-gray of sides of neck. Post-ocular stripes (about 3.5 mm wide and 10 mm
long) blackish and surrounded by medium-gray of headside. Nuchal and mantle feathers
dark-gray with conspicuous blackish margins, imparting a weakly scalloped effect. Small,
anteriormost scapular feathers largely blackish; larger, posterior ones wholly dark-gray. Mid-
dle back posteriorly through rump and upper tail coverts pure-medium-gray. Underparts
from chin to belly same medium-gray as headsides, but appearing narrowly streaked (or
mottled, on throat) with white, widest on lower breast, owing to white feather shafts and
margins of these feathers. Medium-gray of flanks and undertail coverts very weakly tinged
olivaceous, and streaking in these regions obsolete. Tail and wing in molt. Rectrices dark-
gray, blackish at tips, lighter overall on undersurface; shafts blackish dorsally and whitish
ventrally. Outermost rectrices with narrow but conspicuous whitish fringes on inner webs.
Feathers at wrist whitish flecked with medium-gray. Alula and upper primary coverts black.

406

THE WILSON BULLETIN • Vol. JOS, No. 3, September 1996

Eig. 4. Distribution of Acrobatornis fonsecai in southeastern Bahia, Brazil, showing
principal rivers, cities, and the major highway BR-lOl. Triangle is the type locality in the
serra das Lontras. Stars indicate trees with nests; a few sites very close together were mapped
as a single point. Circled stars mark nests confirmed active in October 1995. The roughly
linear distribution of stars reflects the fact that we conducted searches mostly from roadsides,
“n” marks areas we searched for nests but found none. The single “?” is in a region we
suspect holds Acrobatornis but which we were unable to check. Shaded areas are above
500 m elevation. Dotted line is the Bahia/Minas Gerais state boundary. We expect Acro-
batornis occurs locally as far west as about 40°W, which is approximately the western limit
of cocoa cultivation in this region.

forming a conspicuous, diagonally oriented slash along bend of folded wing; the other upper
wing coverts black with light-gray or whitish borders producing, in effect, a gray-edged
panel in the wing-covert region. Remiges blackish, narrowly margined (except two outer-
most pairs) light-gray on the proximal portion of outer webs (these margins widest where
remiges meet the wing-coverts), and whitish on proximal one-half to two-thirds of inner

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 407

webs. Secondaries and tertials slightly paler than primaries, and similarly pale-fringed on
outer webs; some worn feathers in this region, including at least one probable juvenal
scapular feather with a brownish tinge. Soft parts in life: irides pale-gray; maxilla dark-
brown; mandible pink; tarsi and feet clear, bright-pink; claws brownish-pink.

MEASUREMENTS OF HOLOTYPE (mm). — Wing (chord) 65.1; tail 56.0; culmen from
base (at skull) 11.2; culmen from anterior edge nares 7.5; bill depth at anterior edge nares
3.6; bill width at anterior edge naes 3.0; tarsus 17.3; mass 15 gr.

DESCRIPTION OF FEMALE.— The single known female (MZUSP No. 74155) is like
the holotype except mantle with almost no black, instead being concolor with the medium-
gray back and rump. Wing with tips of primaries broken; tail 52.3; culmen from base (at
skull) 11.6; culmen from anterior edge nares 7.5; bill depth not measurable; bill width at
anterior edge nares 2.9; tarsus 17.2; mass unknown.

DESCRIPTION OF JUVENILE. — There are two specimens, both unsexed, quite similar
in plumage, and patterned basically like the adults, but with gray regions instead largely
reddish-yellow; MZUSP No. 74156 is described here. Frontal and supraloral feathers red-
dish-yellow (nearest 7.5YR 6/8; slightly oranger than Tawny, Color No. 38, of Smithe
[1975]), a few, irregularly scattered ones, posteriorly to about the posterior edge of orbits,
with conspicuous black tips or with thin black margins (possibly not juvenal feathers?).
Crown feathers in orbital region slightly paler reddish-yellow, the pale shafts of which
overlay darker, rather worn, brownish (7.5YR 4/2) feathers, producing a sublty streaked or
mottled effect. Posterior half of crown more completely brownish, carrying a weak tinge of
forecrown color. Superciliary stripes blending posteriorly into subtly darker sides of neck
and nearly complete nuchal collar. Post-ocular stripes (perhaps extending anteriorly through
loral region) slightly paler than crown, weakly contrasting with rather grizzled headsides.
Mantle and scapulars pale-brownish (7.5YR 5/2), most washed with reddish-yellow, dis-
tinctly paler than crown, and with two or three medium-gray (7.5YR 5/1) feathers (probably
not juvenal) in the anterior region. Rump and upper tail coverts rather bright yellowish-red
(5YR 5/8). Underparts generally same reddish-yellow as sides of neck, but with a blotchy
appearance owing to differences in color saturation of individual feathers (effects of wear?),
grayish feather bases showing on some feathers, and small, pale, subapical spots on most
throat feathers. There are also a few wholly gray feathers (probably not juvenal) on the side
of the breast. Most rectrices are abraded at tips; central pairs slightly narrower than adult
and not excised. The outermost three pairs largely orange-rufous (nearest 2.5YR 5/8), with
this color concentrated on outer web. Dark-gray (7.5 YR 4/1) bases present on all rectrices,
increasing in extent from outermost to innermost such that three innermost pairs have largely
dark proximal webs. Wing patterned as in adult (black regions same), but all gray feather
margins instead yellowish-red (5YR 5/8). Secondaries and tertials with wider, more con-
spicuous margins, and primaries with thin, yellowish-red apical fringes instead of .solidly
blackish. Soft parts as in adults, with pale-grayish irides. Wing (chord) 61.1; tail 56.1;
culmen from base (at skull) 11.3; culmen from anterior edge nares 7.2; bill depth not
measurable; bill width at anterior edge nares 3.0; tarsus 17.3; mass 14 gr.

ETYMOLOGY. — We are pleased to name this distinctive new furnariid for Paulo Sergio
Moreira da Fonseca (“P.S.”) of Rio de Janeiro, our multi-talented friend of many years, not
only because he was the first to gasp in wonder at the living bird, but also in recognition
of his unending encouragement and deep generosity. Through his excellence in the identi-
fication and observation of birds, PS. has contributed much valuable data to our continuing
studies of the Brazilian avifauna.

We designate the English name Pink-legged Graveteiro to call attention to conspicuous
morphological and ecological features of the bird: the stout, bright-pink legs and feet, and
the fact that it gathers twigs and sticks (“gravetos” in Portuguese) to construct its ne.st, as

408

THE WILSON BULLETIN • Vol. JOS, No. 3, September 1996

do several other groups of furnariids, such as the canasteros (“basket-makers” in Spanish).
Indeed, to call Acrobatornis a “canastero” or “thornbird” or “spinetail” or any other ex-
isting English name seems inappropriate, regardless of its phylogenetic affinities.

The Portuguese name Acrobata refers to the highly acrobatic foraging behavior of the
new species, and this is the name we became accustomed to use while studying it in the
field.

REMARKS

Variation in the type series. — The type series comprises four adults
and two specimens in largely juvenal plumage. The adult specimens, aside
from the holotype and the female described above, are an adult male
(MPEG No. 52345), and an unsexed bird (LSUMZ No. 160000), both of
which are virtually like the holotype except that they have considerably
less black feather-edging in the mantle region, more closely approaching
the adult female than the holotype. The second juvenile specimen, MPEG
No. 52346, is very much like MZUSP No. 74156, but with all orangish
feathers slightly paler, and with the lower back and rump largely medium-
gray. This specimen weighed only 12 g.

Habitat. — Mori (1989) summarized specific climatic data from the low-
land “moist forest” of southeastern Bahia, and characterized the region
as generally hot and humid, without a distinct dry season (but with short,
unpredictable dry periods of one to three months), and with rainfall great-
er than 1300 mm/year. Acrobatornis fonsecai inhabits the canopy and
subcanopy in a restricted section (Eig. 4) of the moist Atlantic Eorest
domain of southeastern Bahia, hereafter refened to as the Itabuna-Ca-
macan region. Native forest within its known range, where not removed
altogether, has been converted to cocoa (Theobroma cacao) plantations
virtually completely; we found no intact forest habitat below about 400
m elevation and almost none below 600 m. Traditionally, cocoa is culti-
vated by thorough removal of the native forest understory, and thinning
of the canopy to about 25 trees/hectare to provide the necessary shade
for the growing cocoa, a system known as “cabruca” (Mori et al. 1983).
Within the known range of Acrobatornis, cocoa is cultivated to elevations
of about 600 m (pers. observ.); indeed, there is little land above this
elevation. We found Acrobatornis fonsecai only in canopy trees left to
shade cocoa plantations (Eig. 5).

When first discovered at 550 m in the serra das Lontras (which turned
out to be the highest elevation at which we ever found it), we expected
that Acrobatornis was a montane forest species, like other undescribed
forms recently discovered in southern Bahia. It soon became apparent,
however, that Acrobatornis was absent from undisturbed montane forest
and, at these higher elevations, was to be found only in the tall trees
shading cocoa, beyond the forest edge. Eollowing our October field ex-

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 409

Fig. 5. Habitat of Acrobatornis fonsecai gen. nov. sp. nov. Leguminosae and other trees
(thinned, “cabruca” canopy) shading cocoa monoculture at the type locality in the serra das
Lontras of southeastern Bahia, Brazil. Three nests of Acrobatornis are visible in the canopy
of the thin tree right of center (two nests above the horizon, one below). Photo by Whitney.

pedition, all evidence indicates that Acrobatornis has spread from a nar-
row distributional center in the contiguous lowlands into the higher ele-
vations of the serra das Lontras and serra Bonita, exactly following the
opening of these denser, more humid, montane canopies for the propa-
gation of cocoa.

The broken, cabruca canopies in the Itabuna-Camacan region, and the
continuous canopies of undisturbed forests above about 500 m, have
many trees heavily laden with bromeliads, lianas, orchids, and mosses.
We also noted many species and individuals of Leguminosae (e.g., Ery-

410

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

thrina verna, Senna multijuga, Schizolobium parahyba, Inga spp.), the
small leaves and relatively open canopies of which allow much greater
penetration of sunlight and wind and, consequently, support almost no
epiphytic growth. Acrobatornis favors Leguminosae for nesting (Whitney
et al. 1996). Other species of Leguminosae listed by Lewis (1987) as
common in the cocoa-growing region of southern Bahia are Diplotropis
incexis, Platycyamus regnellii, Sweetia friiticosa, Parkia pendula, and
Plathymenia foliolosa. In late January, Acrobatornis also foraged in large,
flowering, Croton sp. trees (Euphorbiaceae; see frontispiece).

The native habitats of the Itabuna-Camacan region have suffered ex-
tensive alteration for so long that it is now difficult, perhaps impossible,
to reconstruct the natural habitats there beyond a basic structural descrip-
tion. What is clear is that Acrobatornis has been able to tolerate radical
changes in the habitat in which it evolved, in both a paleoclimatic time-
frame, over probably millions of years, and a recent-historical context,
over the past century or so. Today, Acrobatornis persists as a fairly com-
mon species along roads, including even the major Brazilian highway,
BR-101 (we have even observed it flying across this thoroughfare), and
forages and nests in close proximity to human presence (see Fig. 10 in
“Conservation,” below).

We suspect that Acrobatornis is absent from coastal forests east and
south of Itabuna, where recent, pluviomarine soils (RADAMBRASIL
1981; see discussion under “Origins” below) and forest structure (pers.
observ., unquantified) differ from those in the Itabuna-Camacan region.
We failed to find Acrobatornis along the principal roads from just east of
Arataca east to Una (practically on the coast), then north along the coast
to near Ilheus. Indeed, we noted few cocoa plantations east of the vicinity
of Arataca (which might reflect the different soil type?). We did not con-
duct tape-playback presentations to attempt to find birds, but base the
above suggestion on the fact that we noted no nests of Acrobatornis (see
below). Acrobatornis might occur somewhat farther east, to the north of
Ilheus, where soils and cocoa plantations quite near the coast are appar-
ently much like those in the Itabuna-Camacan region; we did not have
an opportunity to explore this area (see “?” in Fig. 4). We expect it occurs
west to about 40°W, which is near the western limit of cocoa cultivation
in this region.

Breeding. — As appears to be true of most species of passerines in the
Atlantic Forest (pers. observ.), Acrobatornis breeds in the Septem-
ber/October, spring period, at least. Two adults collected in early October
appeared to be in breeding condition and, at this season, the birds were
singing conspicuously and feeding both nestlings and fledged, food-beg-
ging juveniles. On 4 October 1995, when we were actively looking for

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 4 1 I

potential nests of the new species, it was the food-begging calls of nest-
ling Acrobatornis that first called our attention to an unusually sized and
shaped stick-nest in the canopy of a leafy Erythrina tree. Nests were
characteristic, often conspicuous masses of sticks and twigs in the cano-
pies of tall trees, easily seen from roadsides. We confirmed that the pres-
ence of nests was a highly reliable indicator of the presence of Acroba-
tomis, as previously undetected birds responded within 1 min to playback
of tape-recordings presented below nests. The nest and nesting ecology
of Acrobatornis fonsecai, and implications for intrafamilial relationships,
were described by Whitney et al. (1996).

We observed adult Acrobatornis feeding young in four nests at widely
scattered localities (Fig. 4). At one active nest, we determined that both
adults feed the young, although they usually were not at the nest simul-
taneously. Feeding intervals between 08;30 and 11:00 averaged about
once every 10 min., with more regular feedings earlier in the period.
Feeding intervals probably vary considerably with the age of the young;
noisiness of the young in this nest suggested that they were fairly well-
developed. We also noted that adults (perhaps only the male?) usually
sang once from near the nest entrance immediately after feeding the
young. At another active nest in which no young birds were audible, we
suspect that the adult (presumed female, as the presumed male was sing-
ing some 50 m away at the time) was incubating or brooding small young,
because it once stayed inside the nest for 13 mins. We suspect that the
clutch size is two or three, because we saw pairs of adults accompanied
by two young birds several times and, on two occasions, by three pre-
sumed offspring (one of which could have been from a previous nesting).
Gonads of the two adult specimens collected in late January were largely
destroyed by shot (one could not be sexed), such that reproductive con-
dition was not possible to determine, but Pacheco, Fonseca, and Bauer
noted that Acrobatornis was not vocal and was generally inconspicuous
then. Similarly, Whitney noted relatively poor response to tape playback
m early March, and presumed young birds in the company of adults were
not food-begging.

Molt. — Of the two (adult) specimens of Acrobatornis collected in early
October, during the breeding season, one (female) showed no sign of molt,
and one (male) was molting the inner primaries; tail molt had not com-
menced. The two adults collected in late January had little evidence of
molt in the head and body. One (male; holotype) was molting primary
No. 7 and molt of the rectrices was well underway. It had retained at
least one brownish-tinged scapular feather on the left side. The other
specimen (unsexed) was molting some primaries but no rectrices. Juvenal
plumage is apparently replaced in the postjuvenal molt. Both juveniles

412

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

collected in late January showed at least one ingrowing, adult-colored
rectrix and several black or gray head and body feathers. If these birds
were fledged in the previous breeding season, definitive feathering is
probably attained by about six months age. Although we have no firm
evidence that Acrobalornis has a distinct subadult plumage (one ingrow-
ing rectrix with a rufous margin on the right side of MZUSP No. 74156
seems different from adult), as do Scytalopus tapaculos, for instance, this
merits further attention.

Behavior and ecology. — Acrobatornis fonsecai foraged in the canopy
and subcanopy of tall trees, virtually always in the company of mixed-
species flocks of small insectivores and frugivores. Consistent flock as-
sociates in cocoa plantations were Gray Elaenia (Myiopagis caniceps).
Chestnut-crowned Becard {Pachyramphus castaneus). Black-capped Be-
card {P. marginatus). Streaked Xenops (Xenops rutilans), Rufous-browed
Peppershrike {Cyclarhis gujanensis). Red-eyed Vireo (Vireo olivaceus)'.
Tropical Parula {Parula pitiayumi), Bananaquit (Coereba flaveola).
Flame-crested Tanager (Tachyphonus cristatus), Sayaca Tanager (Thrau-
pis sayaca). Palm Tanager (T. palmarum). Violaceous Euphonia {Eu-
phonia violacea). Chestnut-bellied Euphonia {E. pectoralis). Green-head-
ed Tanager {Tangara seledon). Red-necked Tanager (T. cyanocephala),
and Blue Dacnis (Dacnis cayana). At higher elevations. Rufous-headed
Tanager (Hemithraupis ruficapilla) was a near-constant member of mixed-
species flocks, and a variety of other species occasionally joined near
undisturbed forest borders. At about 550 m in the serra das Lontras, Cran-
ioleuca pallida (Pallid Spinetail) foraged in some of the same mixed-
species flocks as Acrobatornis, generally keeping to lower heights and in
more tangled vegetation, performing gleans and short reaches in vines
and dead leaves, but also gleaning from bark.

Acrobatornis traveled through the treetops with a variety of acrobatic
maneuvers, seldom spending more than about 10 sec. at a foraging site.
The birds also flew strongly, adults sometimes traversing more than 300
m in a single flight to join mixed-species foraging flocks after feeding
young at the nest. The most characteristic foraging maneuvers were in-
verted hangs and inverted creeping or hitching along limbs (ranging in
diameter from about 3 mm to 8 cm), with the tail parallel to the substrate;
we estimated these behaviors constituted at least 80% of search maneu-
vers (terminology follows Remsen and Robinson 1990). Most other
searches were scansorial along the uppersides of limbs. Individuals
crawled with agility over and through the terminal leaf- and flower-clus-
ters of tall trees, hanging and swinging as they poked their heads into the
foliage. Their powerful legs and feet (Fig. 2A) allowed them to perform
these maneuvers without fluttering the wings for balance. Unfortunately,

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 413

we were not able to determine the precise orientation of the legs and feet
dining scansorial locomotion, but some of these behaviors were video-
taped, and might show sufficient detail to be informative in future study.
Almost all attack maneuvers were near-perch, mostly gleans in live fo-
liage and flowers, and probes in moss coating trunks and limbs and the
bark of dead limbs. They also reached into new, unopened (still curled)
leaves, probing deeply with the bill. On only one occasion did we note
dead-leaf searching. An adult Acrobatornis hitched along terminal bran-
chlets to reach a small cluster ot dead leaves, then poked and probed in
these for several seconds. Most of the trees in which Acrobatornis foraged
(Leguminosae) did not hold dead leaves or leaf-clusters, and their tall
canopies trapped few dead leaves fallen from other trees.

On two occasions (one of which was partially video-taped) we ob-
served Acrobatornis hitching upwards for distances of about 1 m on the
principal trunks of trees at least 25 cm in diameter at foraging height
(about 20 m above ground), using the tail as a brace or prop (a behavior
rarely reported in Furnariidae), and probing intently in moss and at the
bases of small epiphytic ferns and a bromeliad (spending nearly 3 min at
this latter site). In general, trees in which Acrobatornis foraged held (apart
from thin patches of moss) almost no epiphytic growth.

On the afternoon of 1 1 October, following a brief but hard rain, we
observed a family group of Acrobatornis foraging in the mostly leafless
canopy of a tree at least 35 m tall. One adult performed two aerial, fly-
catching maneuvers to capture large, winged termites. It flew about 3 m
upwards from the crown of the tree, stalled as it took the insect in the
bill, then fluttered back to land near the other birds. Each time, a food-
begging juvenile followed the adult after it landed, but was not fed. In-
stead, the adult bounded away quickly through the branches, and once
we were able to see it hold the termite with its foot as it pecked it against
the limb and swallowed it.

Apparently, most prey were very small; the only items we were able
to see clearly in the field were the winged termites, several small cater-
pillars (fed to nestlings and food-begging juveniles), and a fairly large
moth (gleaned from a dead limb) that fluttered in the bird’s grasp for
several seconds before being subdued and swallowed. Stomachs of three
specimens collected in January contained remains of tiny arthropods.
These were principally Coleoptera (including several Curculionidae and
one Staphylinidae), which were present in all stomachs and represented
58% of all (N = 90) identifiable food items. Ants and insect larvae (in-
cluding catterpilars) were also present in all stomachs, but the first totaled
only 11% and the latter only 7% of the items. Other insects included
winged Hymenoptera (including a minute apoid), Hemiptera, and Ho-

414

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

moptera (including two nymphs). Insect eggs, oothecae of an orthopter-
oid, and spiders were also found, each in only one of the stomachs ex-
amined.

Singing posture was nearly vertical, with the bill parallel to the ground
and opened fairly widely. Sometimes, as the birds sang, they leaned for-
ward and swung the head and neck from side to side, broadcasting the
song more widely. In response to tape playback of songs in October,
individuals (often just the presumed male) usually responded almost im-
mediately by flying in to perch in the tree nearest the tape recorder. They
perched in an upright posture and remained silent for up to several min-
utes. They then either sang one song before departing or, most frequently,
flew off to rejoin the female and sang once from there. With repeated
tape playbacks, males twice descended to near the ground, and females
and immatures sometimes came to trees overhead.

Vocalizations. — The vocal repertoire of Acrobatornis fonsecai is typi-
cal of that of most furnariids (pers. observ.). Whitney recorded 29 songs
(at least 19 of which were in response to playback) from 12-14 individual
adult Acrobatornis, and other vocalizations from several (number unde-
termined) individuals, four of which were immatures or juveniles. The
natural (unsolicited) song may be generally characterized as a simple
series of very short, piercing syllables at about 5.5 kHz that begins with
syllables delivered slowly enough to be counted (5-8/sec), then gradually
accelerates in pace while decreasing slightly in amplitude, finishing with
syllables spaced tightly. It is almost always introduced by 2-4 more ir-
regularly spaced, sharp syllables, and lasts from about 4-8 sec (Fig. 6A,
B, C). Some parameters of songs vary slightly, mostly with respect to
overall duration. The shortest songs always begin as described above, but
then do not achieve a delivery rate greater than about 15 syllables/sec.
Longer songs, and those given in response to playback, finish with syl-
lables delivered at a rate of 20-26/sec, and the series tends to drop ap-
proximately 0.5 kHz in frequency as it loses amplitude (Fig. 6B, C).
Songs given in response to playback may be nearly 12 sec in duration
and usually have a few stutters in the fastest section, after the halfway
point (Fig. 6C). Songs are audible to at least 200 m. Acrobatornis pro-
duces a duet in which the two members of the pair sing different parts.
The presumed male sings a normal version of the song and the female
joins in with irregularly paced bursts of sharp chips much like the intro-
ductory syllables of songs (Fig. 6D). Few duets were heard, and these
were in response to tape playback.

Calls given by foraging Acrobatornis are short, sharp, single syllables
delivered at irregular intervals; flight calls are similar. Food-begging ju-
veniles utter a slightly higher-sounding version of this call and sometimes

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 4 1 5

give doublets with the two elements about 0.17 sec apart and the first
slightly louder (adult and juv. calls Fig. 6E). Also given by adults, al-
though rather rarely (heard only twice), is a chattery vocalization that
may be a pair greeting. It is delivered when one member of a pair that
has been foraging apart flies in to land near its mate. It may be described
as a jumble of 10-12 syllables, quietest in the middle, then loudest
through the final three or four syllables, which are more distinctly sepa-
rated from each other (Fig. 6F). Complete songs are sometimes delivered
immediately following this vocalization. We did not hear the scolding or
mobbing vocalization of Acrobatornis.

SYSTEMATIC RELATIONSHIPS, ORIGIN, AND DESTINY

Phylogeny and classification of the Dendrocolaptidae/Furnariidae com-
plex, or of its many subgroups, have been the subjects of some important
recent studies (Vaurie 1971, 1980; Feduccia 1973; Sibley and Ahlquist
1985; Rudge and Raikow 1992a,b; Clench 1995), all of which, however,
have focused on morphological comparison with limited or inconsistent
discussion of other characters. Vaurie (1980) focused especially on nest
location and architecture. Owing to the variety of taxa judged “interme-
diate in one or another kind of analysis, and the fact that many species
and several genera have not been included in most of these analyses, even
familial limits remain the subject of considerable debate. Detailed docu-
mentation of a new genus, then, especially one well-differentiated from
all others and perhaps basal to some contemporary groups, seems partic-
ularly desirable. Placement of the new genus in the Furnariidae requires
comparison to several other, possibly related genera. Much further anal-
ysis, incorporating morphological, vocal, ecological, and biochemical data
(a great deal of which are now available) is needed to construct a well-
corroborated phylogeny of this large and complex assemblage of birds.

In Table 1 (comparative mensural data from selected furnariids). Fig.

2 (skulls and tarsi), and Figs. 6-9 (spectrograms of songs and other vo-
calizations, many examples from near type localities), we present com-
parative data for Acrobatornis and a variety of selected genera (and se-
lected species of each) that we judge, from field and museum experience,
to be relevant, regardless of degree of actual relatedness. These are: Cran-
ioleuca, Asthenes, Thripophaga, Phacellodomus, Xenerpestes, Metopoth-
rix, and Margarornis\ aspects of each will be described and discussed
relative to Acrobatornis. We consider each of them, except Asthenes, to
be monophyletic as currently defined, or close enough to monophyletic
(e.g., Cranioleuca) for the purposes of the present comparisons. Perhaps
the monotypic Siptornis should be considered as well, but we have in-
sufficient data (no nest, no recordings of song). Tails of these various

Frequency ( kHz)

416

THE WILSON BULLETIN • Vol. JOS, No. 3, September 1996

Time (sec)

Fig. 6. Sound spectrograms of vocalizations of Acrobatoniis fonsecai gen. nov. sp. nov.
for comparison with tho.se of other taxa shown in subsequent figures. All songs in this and
subsequent figures (except A in this figure) shown on same time scale; all other vocalizations
are on a scale 4X that of songs to show greater detail. A. Natural song, short version,
presented on scale 2X that of other songs. Serra das Lontras, 5 Oct. 1995, 475 m. B. Natural
song, long version, with more rapidly paced ending. Serra das Lontras, 8 Oct. 1995, 480 m.

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 4 1 7

genera are at least moderately graduated and composed of 12 rectrices
that are somewhat stiffened, except Thripophaga and Xenerpestes, in
which rectrices are soft. Most of the following discussion of these genera
stems from Whitney’s unpublished observations.

Cranioleuca. Bill length varies substantially in Cranioleuca (e.g., Fig.
2B, C), with even the smallest-billed species (such as C. curtata Ash-
browed Spinetail) considerably longer-billed than Acrobatornis (Table 1).
Body mass is roughly similar (Table 1 ). Tail/wing ratios for the three
representatives of Cranioleucci in Table 1 average 1.1 (and all species
have the tail longer than the wing, although the albiceps complex has
low tail/wing ratio), whereas Acrobatornis has the tail shorter than the
wing with an average ratio of 0.89. Rectrices of almost all Cranioleuca
species are acuminate and, in many species, the innermost are excised on
the inner web near the tip, forming a point without an exposed spine. The
excised tips of the rectrices of some Cranioleuca taxa closely approach
the shape of those of Acrobatornis. W^ings and tails of all members of
Cranioleuca are almost entirely rufous, but many species show darker
feathers in the alula region, and some have blackish proximal webs on
the primary coverts. The obvious difference in color aside, none shows
as marked a pattern, or the same pattern, on the wing as Acrobatornis,
but we do see a vague similarity. Almost all species of Cranioleuca have
pale superciliary stripes contrasting with a dark (brown or rufous, streaked
in two) crown or cap; a few have dark superciliaries contrasting with
buffy or white crowns. The superciliaried/capped pattern of Cranioleuca
seems much like that of Acrobatornis, but this pattern is pervasive in
Furnariidae. All members of Cranioleuca, except the marsh-inhabiting C.
sulphurifera (Sulphur-bearded Spinetail), forage arboreally and almost ex-
clusively with mixed-species flocks, moving through the middle strata and
subcanopies (C. albiceps complex in understory) of forest and woodland
with short hops and hitches along horizontal and vertical substrates, per-
forming reaches and gleans from vines, bark, dead leaves, and tangles.

C. Song after tape playback of conspecific song, which is essentially identical to the long
natural song shown in B. 7 km W Camacan, 12 Oct. 1995, 105 m. D. Duet in which one
bird (adult male?) delivers a long song and is joined by the other bird giving irregularly
paced bursts of similar syllables. Serra das Lontras, 7 Oct. 1995, 475 m. E. Calls: forag-
ing/contact of adult (three on left) and food-solicitation of one accompanying Juvenile (three
on right). 4 km W Arataca, 1 1 Oct. 1995, 1 10 m. E Chatter that seems to serve as a pair
greeting call, and is sometimes followed immediately by the song. Same loc. and date as
E. All recordings by Whitney. Sound spectrograms produced with “Canary” 1 .2 of the
Bioacoustics Research Program of the Cornell Laboratory of Ornithology, Ithaca, New York,
and Canvas” of Deneba Software, Miami, Florida.

418

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Standard measurements of acrobatornis fonsecai and Selected other Genera and Species of Furnarjidae

Acrobatornis

fonsecai

Asthenes

Cranioleuca

Thripophaga

fusciceps

Phacellodo-
mus sibllaira

Xenerpesles

singularis

Metopoihra

auranliacus

Margaromis

squamiger

baeri

dorbignyi

albiceps

pyrrhophia

curlauj

Bill width

S± SD
(N) range

3.0 ±0,05
(4) 2.9-3.0

3.0 ±028
(5) 2.6-3.6

3.0 ± 0.13
(4) 2.9-32

3.2 ± 0.06
(3) 32-32

2.8 ± 021
(3) 2.6-3.0

3.3 ± 0.29
(4) 3.0-3.7

4.1 ±0.10
(3) 4,0-42

32 ± 023
(4) 3.0-3 8

32 ±0.18
(5) 3.0-3.4

32 ± 0.06
(3)3.1-32

3.1 ±0.15
(3) 3.0-32

Bill depth
s 1 SD
(N) range

3,6 ± 0.07
(2) 3.5-3.6

3.5 ± 0.24
(4) 3 4-3,9

3.7 ± 026
(4) 3.5-4.0

3 .9 ±0.10
(3) 3.8-4.0

3,4 ± 0.06
(3) 3.4-3 5

3.7 ± 0.26
(4)3.54.1

4.7 ± 0.15
(3) 4.5-4 8

3.9 ±0.14
(2) 3.8; 4.0

32 ±0.15
(5)31-3.5

32 ± 0.07
(2) 3.3; 3.4

32 ±0.0
(3) 32-32

Bill length
fixtm nares
kiSD
(N) range

7.4 ± 0.16
(4) 72-7.5

7.6 ± 0.48
(5) 6.9-8.1

9.9 ± 0.25
(4) 9.6-102

102 ± 0.15
(3) 10.0-10.3

10.5 ±0.55
(3) 9.9-10.9

9.5 ± 0.88
(4) 8,9-10.8

10.0 ±0.17
(3) 9.9-102

8.5 ±0.61
(4) 8,1-9 4

7.7 ± 0.57
(5) 6.8-82

8.1 ±020
(3) 7.9-82

7.9 ±0-10
(3) 7 8-8-0

Bill length

Grom skull
«±SD
(N) range

11.4 ± 021
(4) 112-11,6

122 ± 0,52
(5) 11.8-13 1

15.4 ± 0.32
(4) 15.1-15.8

13.9 ± 029
(3) 14.9-15.1

16.2 ± 0.42
(3) 15.7-16.5

15-1 ± 1.09
(4) 14.4-16.7

15.6 ±022
(3) 15,4-16.0

13.5 ±0,84
(4) 12.8-14.7

11.5±0.5l
(5) 10.9-12.1

11,7 ±0.06
(3) 1I.6-II.7

13.9 ± 029
(3) 13.6-14.1

Tarsus
s±SD
(N) range

17.5 ± 0.42
(5) 17.2-182

19.7 ± 1.09
(4) 18.5-21.0

24.4 ±0.14
(4) 242-24.5

202 ±0.21
(3) 20.1-20.5

19-7 ± 1.08
(3) 18.9-20.9

18.8 ±0.55
(3) 18.3-19.4

21.8 ±0.15
(3)21.6-21,9

18.5 ± 0.17
(3) 18,4-18.7

15.5 ±0.12
(3) 15.4-15.6

15.0 ± 021
(3) 14.8-152

20.5 ± 0.66
(3) 20.1-212

Tail

S ± SD
(N) range

55.4 ±2,10
(5) 52J-58.0

62.9 ± 5.58
(5) 56.5-70.4

74.6 ± 1-91
(4) 722-76.2

65.2 ± 0.87
(3) 64 5-66.5

75.6 ± 0.42
(3) 752-76.1

73.4 ± 423
(3) 702-78.3

87.4 ±3.75
(2) 84 8; 90.1

62.4 ± 1.82
(4) 59.7-63.6

49.1 ± 125
(4) 48 4-51.0

51.5 ± 121
(3) 50.1-522

722 ± 1.42
(3) 70.8-73.6

Wing

USD
(N) range

63.6 ± 3.54
(4) 60.2-67.8

58.1 ±3.59
(5) 55.3-62.1

67.1 ±4.22
(4) 62.6-71.4

64.9 ± 0.80
(3) 642-65.8

64.6 ± 2.40
(3)619-66 4

65.7 ± 1.57
(4) 64.0-672

74.4 ±3.70
(3) 70.1-76.7

55.8 ± 125
(4) 54,7-57,5

55.9 ± 1-51
(5) 54-0-57.8

57,9 ± 0.58
(3) 57.5-58.6

762 ± 1.15
(3) 752-77.5

Mass
k±SD
(N) range

13.7 ± 1.26
(4) 12.0-15,0

14.4 ± 0.55
(3) 14.0-15.0

18.4 ± 0.48
(4) 18.0-19.0

15.7 ± 1.05
(3) 14.6-16.7

12,9 ± 0.72
(3) 12.1-13.5

16,7 ± 1.06
(2) 16.0: 17 5

25.2 ± 1.04
(3) 24.0-25.9

15.6 ±0,61
(3) 14.9-16.0

11.6 ±048
(4) 11.0-12,0

112 ± 021
(2) 11.2; 11.5

162 ±2.07
(3) 14-5-18-5

and probes in mosses and epiphytic growth. Cranioleuca spinetails rarely
hang inverted for more than 1—2 sec. There are some published reports
of scansorial foraging in Cranioleuca (e.g., Skutch 1969, Vuilleumier note
81 in Vaurie 1980), and we have observed this behavior ourselves many
times. We do not believe that any member of Cranioleuca is as highly
scansorial, or hangs as much, as Acrobatornis, which we suspect is re-
flected in the relatively short and stout tarsus of Acrobatornis (compare
Fig. 2A with 2D, E).

Cranioleuca shows remarkable homogeneity in vocalizations across its
18 or so members, especially the montane and far-southern species, which
is most of them. These songs are short (generally less than 2 sec) series
of thin, spritely, syllables on a level or slightly descending frequency (the
last syllable or two almost always at lower frequency), introduced by two
or three relatively loud and widely spaced syllables with subsequent syl-
lables spaced more closely and quieter (Fig. 7 A, B; two species selected
from recordings of all species). Cranioleuca songs of this type, except
for their relatively rapid beginning and truncated length, are structured
much like and sound similar to the song of Acrobatornis. Songs of C.
gutturata (Speckled Spinetail) and the two or three members of the C.
vulpina (Rusty-backed Spinetail) complex, all principally of lowland Am-
azonian distribution in river-created habitats, differ distinctly from those
of the other members of the genus (and from each other) and do not
approach the song of Acrobatornis. Cranioleuca calls are short, sharp.

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 4 1 9

1 0

8 -
6 -

^ 4

N

^ 2-1

C. pallida

n

C. pyrrhophia

T. macroiira

B

'"111

li

liiljfllii

.11
ilii.

^ ' 1 1 1 1 I I 1 1 1 1 “

3120120123456

C. albiceps C. pyrrhophia

D

E

1 1 1 1, I

; / 1

'' w mi

iLliJ

! o 1 M M

* j ■ i j ;

O

c

0)

3

o*

<D

0.5 1.0

Time (sec)

1.5

Fig. 7. Sound spectrograms of vocalizations of selected Cranioleuca species and Thri-
pophaga macroura. A. Cranioleuca pallida, natural song. Brazil; Bahia, about 10 km E.
Boa Nova, 8 Aug. 1993, 900 m. Recording by Gonzaga. B. C. pyrrhophia, natural song.
Argentina. Salta, about 75 km E. J. V. Gonzalez, 30 Dec. 1987, 330 m. Close resemblance
m species shown in A and B is characteristic of the montane and far-southern members of
the genus (thus, most species). C. Thripophaga macroura, natural song. Bahia, near Al-
madina, 9 Oct. 1995, 425 m. D. Cranioleuca albiceps, scold. Bolivia: La Paz, 31 Mar. 1993,
3180 m. E. C. pyrrhophia, calls (two on left) and scold (two series on right). Argentina:
Salta, about 60 km E. J. V. Gonzalez, 29 Dec. 1987, 330 m. LNS 43819. Scolds of most
Cranioleuca species share this two- to six-syllable, rapidly paced structure. Foraging and
pair-contact calls of most are similar as well. All recordings except A by Wbitney.

single syllables (e.g., Fig. 7E [left side]) that sound much like the calls
of Acrobatornis. Scold or mobbing vocalizations of Cranioleuca (we
know most species) consist of 2-6 emphatic, sharp syllables (vertical
orientation in spectrograms) delivered in rapid succession (e.g.. Fig. 7D
and E [two examples on right side]). In Cranioleuca, pairs rarely duet,
although both members may sing simultaneously in an unsynchronized
manner. As we observed for Acrobatornis, duets are most frequently
heard in response to tape playback.

Asthenes. — To place our discussion of relationships of Asthenes to Ac-
robatornis in perspective, it is necessary to express our view ihtxi Asthenes
includes two, possibly three, distinct lineages, the evolutionary histories
of which have probably been independent since pre-Andean times. It
serves to separate the two most obvious ones, generally, by habitat and
nest architecture. The “stick-nesting” group inhabits woody brush and
rocks and is not dependent on grassland (e.g., Chaco woodland and scrub;

420

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

arid and semi-arid interandean valleys; barren altiplanos; one in Brazilian
serra), and builds sturdy nests of sticks (depending on their availability)
lined mostly with wool and grass. The second, “grass-nesting,” group is
completely dependent on grassland (with or without scattered shrubs and
rocks; e.g. grassy paramos; Festuca-dovaxndXcd hillsides and valleys; one
in grassy coastal marshes from extreme southern Rio Grande do Sul to
Buenos Aires, Argentina) and builds nests of grasses and other herbaceous
material; most species use no or few sticks (far-southern A. anthoides
[Austral Canastero] may be an exception). A thorough definition of this
division, which finds parallel in distinct morphotypes, vocalizations, and
ecologies, is beyond the scope of this paper. Species limits within Asthe-
nes have been the subject of much debate over the years and, unfortu-
nately, recent modifications that seem to have gained general acceptance
(e.g., the A. dorbignyi [Creamy-breasted Canastero] complex) have been
extraordinarily poorly documented. For comparison with Acrobatornis,
we selected some typical members of our “stick-nesting” group of As-
thenes (see “Specimens Examined,” above. Table 1, and Fig. 9).

Bill lengths and overall shapes vary considerably within stick-nesting
Asthenes, but small-billed species like A. baeri (Short-billed Canastero;
Fig. 2D, Table 1) and A. patagonica (Patagonian Canastero) closely ap-
proach Acrobatornis. These are also the lightest members of the group,
with body mass essentially like Acrobatornis (data for baeri in Table 1).
Like Cranioleuca, Asthenes have tail/wing ratios greater than 1. Stick-
nesting Asthenes have blunt-tipped or weakly acuminate rectrices that are
not excised. They forage on the ground or low in brush and rocks with
short gleans, reaches, and probes, often moving quickly between places
of concealment, seldom ascending to heights of more than about 1 m
except to sing or for nesting.

Two types of songs or song-like vocalizations are shared by stick-
nesting Asthenes, each of generally uniform pattern and cadence (but a
few differ appreciably in auditory quality) across the group. One begins
with well-separated syllables then rapidly accelerates through several (at
least 5) seconds, often descending slightly through the terminal third.
Examples are shown in Fig. 8A and C (selected from recordings of all
species of Asthenes). Figure 8E shows a variant of this song-type in which
pace is much slower through about the first two-thirds, and syllables are
finely modulated (and sound harsher), probably communicating a different
message. In all respects, including auditory quality (which is especially
difficult to judge from a spectrogram), this song-type of Athenes is re-
markably similar to the song of Acrobatornis (compare with Fig. 6A, B,
C). The second song-type is a rattling series of rapidly paced syllables
that begins quietly, quickly crescendos, and then trails off as it slows

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 42 1

down (Fig. 8B, D, F). We have not heard a song of this type from Ac-
robatornis. Among all stick-nesting Asthenes, the songs of A. baeri (of
semiarid brush and woodland from extreme southern Brazil to central
Argentina) and A. clorbignyi (of arid and semiarid, brushy valleys and
slopes of the Andes of southern Peru to central Argentina) are most sim-
ilar to songs of Acrobatornis\ indeed A. dorbignyi and A. baeri appear to
be sister species elevationally allopatric in the north and perhaps para-
patric farther south, in Mendoza, Argentina. Like Acrobatornis, calls of
Asthenes species (Fig. 8A, E, G) are sharp, single syllables similar to the
introductory syllables of the song. Some canasteros occasionally sing du-
ets, especially at territorial encounters or in response to tape-recording
playback; an example is shown in Fig. 8H (A. patagonica).

Thripophaga. — The bill of T. fusciceps (Plain Softtail; Fig. 2F, Table
1 ) is proportioned much like that of Acrobatornis, but is larger; body
mass is correspondingly greater as well. The rectrices of the four species
currently included in Thripophaga are wide, soft, and blunt-tipped (the
poorly known T. berlepschi [Russet-mantled Softtail], which probably
does not belong in this genus, somewhat more pointed). We have field
experience only with T. macroura (Striated Softtail) and T. fusciceps
(Plain Softtail). Both forage with mixed-species flocks, mostly in the mid-
dle strata of dense, vine-rich, humid, lowland forest, by hitching and
crawling upwards and laterally through tangles with the tail partially
spread, reaching and gleaning (rarely hanging) from bark, vines, and es-
pecially dead foliage trapped in such places. They are not scansorial.
Songs and calls of T. macroura and T. fusciceps are very similar in struc-
ture and auditory quality. A long song of T. macroura, selected to show
maximum similarity to Acrobatornis, is shown in Fig. 7D; most songs
are introduced by a single or two well-separated, loud syllables followed
immediately by a short burst of tightly spaced syllables, the whole lasting
less than about 1.5 sec. Although cadence and pace of the long song
resemble the song of Acrobatornis, structure of individual syllables, and
thus auditory quality, are quite different.

Phacellodomus. — The smallest member of Phacellodomus, P. sibilatrix
(Little Thornbird), is slightly larger than Acrobatornis (Table 1, tail/wing
ratio 1.1). Like other members of the genus, its bill is also differently
shaped, with a hump above the nares and steeply sloping culmen (Fig.
2G). Rectrices of Phacellodomus thornbirds are blunt-tipped. They forage
primarily with reaches and gleans on the ground and in low brush and
thickety growth, often hopping on the ground. Across the genus (we do
not know P. dorsalis [Chestnut-backed Thornbird], of the middle Mara-
non valley of n. Peru), male thornbird songs are delivered from prominent
perches (bush- or treetops, for instance, always near the nest) and con-

Frequency { kHz)

422

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

A. baeri

A. baeri

Time (sec)

Fig. 8. Sound spectrograms of vocalizations of selected stick-nesting Asthenes species.
A. A. baeri, song after conspecific tape playback (same as natural). Introductory syllables
are like one common type of pair-contact call. Argentina: Salta, 05 km S. Rivadavia, 27
Oct. 1989, 260 m. LNS 46142. B. A. baeri, a second .song-type. Argentina: Salta, 50 km
N. J. V. Gonzalez, 28 Oct. 1989, 390 m. LNS 46166. C. A. dorbignyi, natural song. Argen-
tina: Jujuy, 13 km N. Humahuaca, 4 Jan. 1988, 3300 m. LNS 43944. D. A. dorbignyi, a

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 423

sistently comprise series of evenly spaced syllables on a steady or slightly
ascending and descending frequency, the whole lasting about 2—4 sec and
repeated at intervals ot several seconds. Examples are shown in Fig. 9
(A, C, D; selected trom recordings of all species except dorsalis). Females
and immatures sing and sometimes duet with males (Fig. 9A). Individuals
are capable of singing several variants of the song; some (especially those
in duets) are considerably longer and more complex than the “normal”
songs shown in the figures. Calls of Phacellodomus delivered while for-
aging are sharp, penetrating syllables (e.g.. Fig. 9B). Somewhat different,
single-syllable calls are given while scolding or mobbing. Although songs
of Phacellodomus are not similar to those of Acrobatornis in overall pat-
tern, some variants (e.g., female part of duet in Fig. 9A) sound like the
beginning of songs of Acrobatornis.

Xenerpestes. — The two species of Xenerpestes are little-known, and
their relationship within the Furnariidae (in fact, whether they are fur-
nariids at all), is poorly corroborated. One anatomical specimen exists
(National Museum of Natural History), but has not been analyzed. Men-
sural data for X. singularis presented in Table 1 (see Fig. 2H for bill and
tarsus shapes) are remarkably similar to those of Acrobatornis, although
Acrobatornis is about 15% heavier and has a thicker, more powerfully
clawed tarsus. The average tail/wing ratio of four Xenerpestes singularis
is 0.88, essentially identical to Acrobatornis (0.89). Within Furnariidae,
the black-and-gray definitive plumage of Acrobatornis is approached, al-
beit superficially, only by the two members of Xenerpestes, both of which
have restricted distributions in nw South America: X. minlosi (Double-
banded Graytail) and X. singularis (Equatorial Graytail). Both are basi-
cally grayish above and have pale superciliaries; X. minlosi is black-cap-
ped, like Acrobatornis-, X. singularis lacks a contrasting cap and has a
rufous-streaked forecrown. Xenerpestes minlosi is mostly whitish below,
and has two fairly conspicuous, white wingbars; singularis has yellowish.

■second song-type. Argentina: Jujuy, 30 km W. La Quiaca, 05 Jan. 1988, 3640 m. LNS
43952. E. A. dorbignyi, harsh version of natural song-type shown in C (this example also
natural). Same individual (and part of same recording) as in D. LNS 43952. E A. potagonica,
natural song. Argentina: Rio Negro, 8 km E. Villa Regina, 7 Jan. 1988, 485 m. G. A.
patagonica, calls of a single individual. Argentina: Chubut, 08 km NNW Trelew, 12 Jan.
1988. LNS 46108. Context for these calls unknown; possibly pair-contact calls. H. A. pci-
tagonica, pair duet involving same individual shown in F, after playback of song shown in
E LNS 43960. Note almost identical songs, both types, of A. baeri and dorbignvi-, also
similarity of the three Asthenes species shown in B, D, and H, and of these to song of
Xenerpestes minlosi in Fig. 9E. Note close similarity of songs of A. baeri and dorbignyi A
and C to songs of Acrobatornis in Fig. 6B and C. All recordings by Whitney.

Frequency (kHz)

424

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

M. aurantiacus M. SQuamiger

G

H ,, i u'- u . Jiu. 1,

/l /' A A

1 1 i I ' ^ '' f» '

- f r V f

1

\ 1

1 1 1 1

1 1 1 —I 1 r-

0 1 2 0 0.5 1.0 1.5 2.0

Time (sec)

Fig. 9. Sound spectrograms of vocalizations of some other stick-nesting furnariids, and
the highly scansorial Margarornis .squaniiger. A. Phacellodomus sibilatri.x, natural duet at
nest. Presumed male begins with more widely spaced syllables centered at about 4 kHz,
joined by presumed female singing rapid series at higher frequency. Argentina; Salta, 50
km N. J. V. Gonzalez, 28 Oct. 1989, 390 m. LNS 46163. One common type of male song
is similar to that shown in this duet. The presumed female song shown here is similar to
the beginning of the song of Acrobatornis (Fig. 6B and C). B. P. sibilatrix, natural calls,
apparently pair-contact. Argentina; Salta, 75 km E. J. V. Gonzalez, 30 Dec. 1987, 330 m.
LNS 4383 \.C. P. rufifron.s, natural song. Song-type repeated at fairly regular intervals from
a stationary perch, often in active nest tree. Brazil; Mato Grosso do Sul, NW. Miranda at
“Pousada Caiman,” 19 July 1995. D. P. rufifrons, natural song. Different song-type from
that shown in C. We believe individuals give both song types, and some others; thornbirds

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 425

black-streaked underparts and plain, grayish wings. Rectrices of both Xe-
nerpestes are grayish and blunt-tipped. They forage actively in the sub-
canopy of humid forest, almost always with mixed-species flocks, hop-
ping and hitching along the uppersides of limbs, occasionally hanging on
leaf-clusters, rarely inverted from limbs, performing gleans and short
reaches. They are not scansorial, but Ridgely and Tudor ( 1994; 129) noted
for X. singidaris that “occasionally one briefly creeps along a branch.”
Xenerpestes minlosi often forages in dead leaves and leaf-clusters, and
often hangs briefly from foliage. Similar behaviors of X. singularis were
described by Parker and Parker (1980). The voice of X. minlosi is a long,
extraordinarily rapidly paced, chattering series of syllables on a steady
frequency, beginning quietly and quickly attaining an increased amplitude
with slight pulses at irregular intervals; the whole lasts from 3-12 sec
(Fig. 9E). We believe pairs sometimes duet. Calls are sharp, thin, almost
bisyllabic notes (Fig. 9F). The song differs from that of Acrobatornis in
its rapid beginning and overall rapid pace, but is approached by the tightly
spaced ending of long songs of Acrobatornis. One individual singing in
the Acrobatornis duet shown in Fig. 6D sounded much like Xenerpestes
minlosi shown in Fig. 9E. In our opinion, the song of Xenerpestes minlosi,
together with other evidence gathered to date, clearly places this genus
in the Furnariidae.

Metopothrix. — Metopothrix aurantiacus (Orange-fronted Plushcrown;
sole member of the genus), like Xenerpestes, is a small, phylogenetically
obscure furnariid bearing marked structural resemblance to Acrobatornis
(Fig. 21, Table 1). Its tarsus and foot appear to be somewhat stronger than
those of Xenerpestes, thus closer to Acrobatornis, and the average
tail/wing ratio of three specimens exactly matches Acrobatornis (0.89).
Rectrices of Metopothrix are grayish and blunt-tipped or slightly pointed.
Metopothrix usually forages with mixed-species flocks in the subcanopy

<—

have complex vocal repertoire.s. Brazil: Bahia, 06 km NW Boa Nova, 9 Mar. 1996. E.
Xenerpestes minlosi, song after playback (same as natural but longer). Dark band at about
7 kHz is insect noise. Panama: Darien, Cana airstrip in Parque Nacional Darien. Note
similarity with songs of stick-nesting Asthenes species in Fig. 8B, D, and F, and with .songs
of Acrobatornis in Fig. 6B and C, especially the rapidly paced ending of those songs. It
also seems quite similar to one of the individuals singing in the duet of Acrobatornis in
Fig. 6D. F. X. minlosi, natural pair-contact calls. Same individual as E. G. Metopothrix
aurantiacus, natural song. Peru: Madre de Dios, near Atalaya above “Hacienda Amazonia”
lodge, 8 July 1989, about 600 m. Songs vary from 2 to 5 syllables (3 most common). H.
Margarornis squamiger, natural calls of foraging birds, context unknown. Songs of all
Margaromis species are sharp, high-frequency syllables like these, but are delivered in a
rapid, steady series lasting about I .sec. All recordings by Whitney.

426

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

of forest and well-developed second-growth, mostly in river-created hab-
itats (medium/old whitewater islands, river edge, and foothill forest near
river floodplains) in upper Amazonia. It performs gleans and reaches in
live foliage and vine tangles, often at or near the periphery of trees,
paying particular attention to the undersides of leaves, including large
leaves like those of Cecropia species, and very small leaves like those of
Leguminosae. Individuals may hang inverted for several seconds at a
time, and sometimes crawl in an inverted position. The song of Meto-
pothrix is an inconspicuous series of 3-5 evenly spaced, high, thin syl-
lables lasting 1.5-2. 5 sec (Fig. 9G). The birds nearly always travel in
pairs or family groups, and we have heard a variety of other, quiet, in-
trafamilial vocalizations. The song and these other vocalizations show
little resemblance to those of Acrobatornis.

Margarornis. — The three or four (including M. bellulus [Beautiful
Treerunner]) currently recognized species of Margarornis treerunners are
small furnariids (Rudge and Raikow 1992a) that inhabit humid montane
forest of southern Middle America and the Andes. The bill of Marga-
rornis squamiger is shaped much like that of Acrobatornis, but is slightly
longer, as is the tail/wing ratio (average of 3 specimens 0.95); body mass
is also greater (Fig. 2J; Table 1). Rectrices are acuminate with the rachis
extending slightly beyond the feather vanes. Margarornis species forage
with mixed-species flocks in the forest interior. They are highly scansorial,
almost constantly creeping along limbs, both upper- and undersides, and
sometimes up tree trunks, performing gleans and probes in moss, epi-
phytic growth, and dead leaf clusters. They sometimes use the tail as a
prop when stationary (Slud 1964, Wetmore 1972, pers. observ., all spe-
cies). Individuals may hang inverted on balls of moss or leaf clusters for
several seconds. The scansorial behavior of Margarornis species appears
to be much like that of Acrobatornis, although we have not been able to
determine whether the locomotory and grasping action of the legs and
feet of the two are the same. Dissection of hindlimb musculature of Ac-
robatornis, in comparison with that reported by Rudge and Raikow
(1992b) for their Margarornis assemblage, would perhaps be enlightening
in this regard. Acrobatornis spends much time hanging from and crawling
through foliage, a search-method rarely practiced by Magarornis species.
Songs of all Margarornis treerunners consist of short, sharp, high-fre-
quency syllables delivered in a steady series lasting about 1 sec. Calls are
like individual syllables of the song, but are delivered at irregular intervals
(i.e., not in repetitive series; Fig. 9H, selected from recordings of all
species). Their vocalization seem quite different from those of Acroba-
tornis.

So what is Acrobatornis most likel — Overall morphological compari-

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 427

sons trom both Table 1 and Fig. 2 show particularly close parallels in bill
shape, tail/wing ratio, and all standard absolute measurements between
Acrobatornis, Metopothrix, and Xenerpestes, and lesser similarity of these
with Margarornis. The acuminate and strongly excised central rectrices
of Acrobatornis are different from the blunt rectrices of these others, but
are like several species of Cranioleuca. Among stick-nesting Asthenes,
A. baeri and A. patagonica seem close to Acrobatornis. We judge plum-
age parallels (overall pattern and regions of contrast) strongest with Cran-
ioleuca (notwithstanding that no species are colored like Acrobatornis)
and Xenerpestes. Comparisons of vocalizations, especially songs and
adult contact (foraging) calls, show a strikingly close similarity to one
song-type of some stick-nesting Asthenes, and notable parallels with Xe-
nerpestes. Single-syllable foraging or pair-contact calls of Acrobatornis
appear most like those of montane Cranioleuca species, but this is more
complicated by the difficulty of judging homology of these vocalizations.
The hanging and inverted-scansorial foraging maneuvers oi Acrobatornis
are approached most notably by Margarornis and Cranioleuca and, to a
lesser extent, Metopothrix and Xenerpestes, and some other arboreal fur-
nariids. Asthenes are highly terrestrial foragers, and bear no resemblance
to Acrobatornis in this regard.

Our intrafamilial comparisons of morphology, vocalizations, and for-
aging behavior, together with nest architecture and placement discussed
by Whitney et al. (1996), suggest that Acrobatornis is related to stick-
nesting Asthenes canasteros and Cranioleuca spinetails, and may be the
closest relative of the Xenerpestes graytails and the monotypic Metopoth-
rix. It is, of course, impossible at the moment to know how many of the
similarities among these genera might be the result of convergence. Our
hypotheses are, at least, supported by multiple, corroborative parallels,
and may be helpful in orienting future, especially biochemical, investi-
gation. Pertinent specimens are at hand.

On the origin of Acrobatornis. — As pointed out by Remsen (1984),
“Although reconstructions of historical zoogeographic events often con-
tain more speculation than warranted by available data, such historical
hypotheses are a necessary part of a zoogeographic and taxonomic anal-
ysis.” We postulate that Acrobatornis fonsecai represents a relict from an
expansion of stick-nesting furnariids from a (probably arid or semi-arid)
Chaco-Patagonian/Pantanal distributional center that survived in arboreal
habitats on ancient and stable soils in southeastern Bahia. Soils in the
restricted region from Feira de Santana south to the vicinity of Camacan
and Pau-Brasil are formed primarily of decomposed crystalline rock of
lower Precambrian age known as the Jequie Complex, and are among the
oldest soils on the planet (Caldeira 1954, RADAMBRASIL 1981). The

428

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

entire known distribution of Acrobatornis is encompassed in this area.
Acrobatornis appears to be absent from forests on the sandy fluvio-marine
plain (“Planicie fluviomarinha” RADAMBRASIL 1981). It may inhabit
evergreen seasonal forest (“mata de tabuleiro”), however, where the co-
coa-growing region approaches the coast between Itabuna and Ilheus.
This forest is also mostly on soils of the Jequie Complex. The fact that
these soils are extremely old and have apparently been stable for a long
period of time may help explain the peculiar, relictual distribution of
Acrobatornis.

As the arid and semi-arid periods that had originally allowed it to reach
the region eventually gave way to a more mesic environment with taller
trees, Acrobatornis survived by gradually adapting to build its nests in
the relatively xeric crowns of trees, especially Leguminosae. These wide,
leguminaceous canopies, with their small leaves and open, layered strat-
ification of limbs, are indeed hot and dry, supporting virtually no epi-
phytic growth. Acrobatornis may have managed to survive through suc-
cessful ecological adaptation; we suspect the nest and vocalizations have
undergone little structural modification.

A quite different hypothesis of origin of Acrobatornis in the humid
forest of southeastern Bahia gives primary importance to analysis of dis-
tributions of birds syntopic with it today. Almost all (if not all) birds
inhabiting the lowland forest of southeastern Bahia, from terrestrial spe-
cies to canopy species, are unequivocally most closely related to (and
probably derived from) Amazonian stock (Willis 1992, Whitney et al.
1995). Why should Acrobatornis be any different? Fox Acrobatornis, un-
like all other birds in this region of Bahia, it is difficult to identify and
justify a contemporary link to an Amazonian relative. In this context,
however, we recall various similarities to upper Amazonian Metopothrix
and, in the subtropical forests of the east slope of the Andes and in trans-
Andean lowlands, Xenerpestes. Could there be (or could there have been)
an Acrobatornis-\\V.Q bird in the vast and little-known canopies of Ama-
zonia? The answer to this riddle may lie in elucidation of such poorly
understood factors as edaphic habitat maintenance and variable distribu-
tion of vegetation types.

Conservation: the destiny of Acrobatornis. — The area of occurrence of
Acrobatornis fonsecai lies entirely within the nucleus of the cocoa-grow-
ing region of southern Bahia. Up to now, it has maintained apparently
viable numbers in harmony with intense anthropogenic alteration of its
habitat and pervasive human presence (Fig. 10). Although virtually no
undisturbed forest remains in the lowlands of the Rio Jequitinhonha-Rio
de Contas interfluvium, it is clear that the persistence of the “cabruca”
canopies shading the cocoa guarantees the survival of a considerable por-

Pacheco el al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 429

Fig. 10. House and cocoa plantation on the side of the major highway BR-101, about
14 km N. of Itabuna, Bahia. Tall, native trees shading cocoa plantations such as this provide
acceptable breeding habitat for Acrobatornis fonsecai and a wide variety of other birds. A
nest of Acrobatornis is visible as a dark spot in the upper left of the tall tree in the center
of the photograph. Photo by Whitney.

tion of the indigenous canopy fauna and flora, including the avifauna
(Alves 1990, pers. observ.), and maintains corridors for fauna only sea-
sonally or temporally dependent. The cabruca system of growing cocoa
has preserved, to a considerable extent, the humid microclimate upon
which many forms of life depend, and must be responsible on a large
scale for maintenance of the local hydrologic regime. Cabruca canopies
are in some areas almost continuous over large tracts, and hold magnif-
icent adult trees of a wide variety of species. Many individuals of these
mature trees must be near the end of their lifespans, however, and there
are no seedlings to regenerate them because of constant weeding of the
understory in cocoa plantations. Consequences of this were summarized
by Mori et al. (1981): “A system of clear-cut and burn with the subse-
quent replacement of native trees by monotypic stands of [introduced]
Erythrina fusca has been developed. In this system, banana trees are used
to shade the young cocoa plants and to provide income until the cocoa
trees produce fruit. Fertilizers, herbicides, and pesticides are used in both
systems [this and cabruca] but more .so in the cut-and-burn system.” This
system is called “derruba total” (Alger and Caldas 1996).

430

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Cocoa was introduced into southern Bahia from the Amazon in 1746,
and has been the most important crop in that region ever since (Mori et
al. 1983). In their concise and well-balanced discussion of the history of
cocoa cultivation and the current socioeconomic problems confronting the
cocoa growers of southeastern Bahia, Alger and Caldas (1996) reported
that during a brief period in the 1970s, cocoa from southern Bahia was
the second most important export for Brazil, after coffee. During this era,
more than half a million hectares were planted in cocoa, and the industry
flourished. Between 1986 and 1992, however, the international price for
cocoa fell from US $2500 to $1000 per ton, hitting the industry hard.
Then, in 1989, came the accidental introduction of the fungus Crinipellis
perniciosa, which causes the disease known as “witch’s-broom.” Ac-
cording to the Executive Commission for the Economic Recovery of the
Cocoa Industry (CEPLAC), as of March 1995, the disease had spread
through more than 70% of the cultivated region. Control of the fungus
was expensive and ultimately not effective. Cocoa production plummeted,
with dire consequences for the livelihoods of tens of thousands of people
and, concomitantly, for the remnant forests of southeastern Bahia. For
example, the newspaper A Tarde of Salvador, Bahia, for 16 Feb. 1996
(article by Rosa M. Carvalho), reported that Hershey, the largest chocolate
producer in the United States and perhaps the biggest buyer of Bahian
cocoa, was on the verge of terminating contracts with Bahian growers
because of the drastic decline in cocoa production. And Alger and Caldas
(1996) reported how plantation owners, although often with reluctance,
have little choice but to cut their valuable old trees and sell the timber
for immediate cash. They explained, “The shift from cocoa to other crops
or for livestock pasture is not considered remunerative. Deforestation for
timber sale is generating pasture more because of lack of alternatives than
enthusiasm for livestock raising.” They went on to describe how many
of the largest plantation owners, faced with steady losses, are highly ame-
nable to preserving their remnant forests, but receive little support from
the government to do so, and emphasized the importance of establishing
economic incentives for land preservation, such as subsidies for formation
of private reserves.

As it stands now, there is not one officially protected area of forest in
the range of Acrobatornis, although it may eventually be found in the
nearby Una Reserve, the avifauna of which is poorly known. This reserve
is the stronghold of one of the rarest primates in the world, Leontopithecus
chrysomelas (Golden-headed Lion Tamarin), which has a distribution al-
most exactly coincident with that of Acrobatornis: remnant lowland forest
between the Rio de Contas and the Rio Jequitinhonha. Even a tribe of
Indians, the Kamakan, was indigenous to the still more restricted region

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 43 1

between the Rio de Contas and the Rio Pardo; the modern town, Cama-
can, takes their name. The last member of the Kamakan died in 1938
(Viveiros de Castro 1986).

The only practical means that we can imagine of preserving a popu-
lation of Acrobatoniis and the other fauna in this region is the immediate
purchase of two large, separate blocks, one in the serra das Lontras and
one in the serra Bonita, that encompass forest from the highest elevations
in these ranges down to near sea-level. Such blocks must incorporate
extensive cabruca canopies in contact with healthy forest canopies. The
cocoa could then be removed and seedlings of native tree species planted
to replace the aging adults. Local people, especially large land-owners,
must be integrally involved in establishment and maintenance of these
forest reserves. This land is for sale today, and the price and socio-polit-
ical climate are favorable (Alger and Caldas 1996).

We will never cease to be amazed at how this striking little bird that
constructs conspicuous stick-nests in treetops along the congested high-
way BR-101 could have been overlooked so completely. Its discovery
serves to remind us of how much remains to be learned, even as it fades
from existence.

ACKNOWLEDGMENTS

This description benefitted from the generous collaboration of Phyllis Isler (who produced
the sound spectrograms from our recordings), P.S.M. da Fonseca (whose assistance and
enthusiasm in the field were unflagging), Dan Lane (who illustrated the skulls and tarsi),
Claudia Bauer (who helped in the field and in making the map), Robert Barth (who assisted
in and funded part of the October 1995 field work); and Morton Isler (who produced the
morphological table from our data): their contributions are tremendously appreciated. We
are grateful to J. V. Remsen, Jr. and Steven Cardiff of LSUMZ, and Robert Ridgely, David
Agro, and Sally Conyne of ANSP for allowing us to examine specimens in their care. Steven
Cardiff and Donna Dittmann also oversaw preparation of the skeleton of the new genus,
and Steve prepared the study skin that documents its identity. Hannah Gould of the Univ.
of Texas at Austin helped us locate pertinent references, and Johann Becker of the National
Museum of Rio de Janeiro gave us valuable advice on questions of nomenclature. Francisco
Mallet-Rodrigues and Inge M. Schloemp helped us in a variety of ways. Howard Wilson
assisted m making black-and-white images from Hi-8 format video tape. Gary Graves and
J. V. Remsen, Jr. gave us helpful criticism of the manuscript, and Charles Blem assisted in
seeing that it was published promptly. Mark and Allison Duffel of “Mr. Wizard’s” in Austin,
Texas have been especially helpful in keeping our recording equipment in optimal condition,
often on short notice, and Holland Photo of Austin produced the black-and-white photos in
the figures. Field Guides Incorporated, of Austin, Texas, generously financed part of our
expenses for research in southern Bahia, and tour participants Jane Brooks, James Plyler,
Thomas Raque, John and Barbara Ribble, and Polly Rothstein helped Whitney make im-
portant observations in March 1996. Finally, we are grateful to Paul Donahue for his at-
tractive frontispiece painting of the new genus and species.

432

THE WILSON BULLETIN • Vol. 108. No. 3, September 1996

LITERATURE CITED

Alger, K. and M. Caldas. 1996. Cacau na Bahia. Decadencia e amea9a a Mata Atlantica.
Ciencia Hoje 117:28—35.

Alves, M. C. 1990. The role of cocoa plantations in the conservation of the Atlantic Forest
of southern Bahia, Brazil. M. S. thesis Univ. of Florida, Gainesville, Florida.

Ames, P. L. 1971. The morphology of the syrinx in passerine birds. Peabody Mus. Nat.
Hist. Yale Univ. Bull. 37.

Caldeira, C. 1954. Fazendas de cacau na Bahia. Ministerio da Agricultura, Servi^o de
Informa9ao Agricola, Rio de Janeiro, Brazil.

Clench, M. H. 1995. Body pterolysis of woodcreepers and ovenbirds (Dendrocolaptidae
and Furnariidae). Auk 1 12:800-804.

Feduccia, J. a. 1973. Evolutionary trends in the Neotropical ovenbirds and woodhewers.
Ornith. Monogr. No. 13.

Gonzaga, L. P. and J. F. Pacheco. 1995. A new species of Phylloscartes (Tyrannidae)
from the mountains of southern Bahia. Bull. Brit. Orn. Club 115:88-97.

, , C. Bauer and G. D. A. Castiglioni. 1995. An avifaunal survey of the

vanishing montane Atlantic forest of southern Bahia, Brazil. Bird Conserv. International
5:213-224.

Lewis, G. P. 1987. Legumes of Bahia. Royal Botanical Gardens. Kew, United Kingdom.

Mori, S. A. 1989. Eastern, extra- Amazonian Brazil. Pp. 427—454 in Floristic inventory of
tropical countries: the status of plant systematics, collections, and vegetation, plus rec-
ommendations for the future (D. G. Campbell and H. David Hammond, eds.). New
York Botanical Garden, New York, New York.

, B. M. Boom, and G. T. Prance. 1981. Distribution patterns and conservation of

eastern Brazilian coastal forest tree species. Brittonia 33:233-245.

, A. M. DE Carvalho, and T. S. dos Santos. 1983. Southern Bahian moist forests.

Bot. Rev. (Lancaster) 49:155-232.

Munsell®. 1994. Soil Color Charts. Revised Edition. Macbeth Division of Kollmorgan
Instruments Corporation, New Windsor, New York.

Pacheco, J. F. and L. P. Gonzaga. 1995. A new species of Synallaxis of the ruficapil-
lalinfiiscatci complex from eastern Brazil (Passeriformes: Furnariidae). Ararajuba 3:3-
1 1.

Parker, T. A. Ill and S. A. Parker. 1980. Rediscovery of Xenerpestes singularis. Auk 97:
203-205.

RADAMBRASIL. 1981. Projeto Radambrasil, levantamento de recursos naturals, (vol. 24.
Folha SD. 24: Salvador). Ministerio das Minas e Energia, Secretaria Geral, Rio de
Janeiro, Brazil.

Remsen, j. V., Jr. 1984. Geographic variation, zoogeography, and possible rapid evolution
in some Cranioleuca spinetails (Furnariidae) of the Andes. Wilson Bull. 96:515-523.

AND S. K. Robinson. 1990. A classification scheme for foraging behavior of birds

in terrestrial habitats. Pp. 144-160 in Avian foraging: theory, methodology, and appli-
cations (M. L. Morrison, C. J. Ralph, J. Verner, and J. R. Jehl, Jr., eds.). Studies in
Avian Biol. 13.

Ridgely, R. S. and G. Tudor. 1994. The Birds of South America. Vol 11. University of
Texas Press, Austin, Texas.

Rudge, D. W. and R. j. Raikow. 1992a. The phylogenetic relationships of ihe Margarornis
Assemblage (Furnariidae). Condor 94:760—766.

. 1992b. Structure, function, and variation in the hindlimb muscles of the

Pacheco et al. • A NEW FURNARID FROM SOUTHEASTERN BRAZIL 433

Margaroniis assemblage (Aves: Passeriformes: Furnariidae). Ann. Carnegie Mus 61-
207-237.

Sibley, C. G. and J. E. Ahlquist. 1985. Phylogeny and classification of new world sub-
oscme passerine birds (Passeriformes, Oligomyodi: Tyrannides). Pp. 396-428 in Neo-
tropical Ornithology (P. A. Buckley, M. S. Foster, E. S. Morton, R. S. Ridgely, and F.
G. Buckley, eds.). Ornith. Monogr. No. 36.

AND B. L. Monroe, Jr. 1990. Distribution and Taxonomy of Birds of the World.
Yale University Press, New Haven, Connecticut.

’ • 1993. A supplement to Distribution and Taxonomy of Birds of the World.

Yale Univ. Press, New Haven, Connecticut.

Skutch, a. F. 1969. Life histories of Central American birds III. Pacific Coast Avifauna
No 35.

Slud, P. 1964. The birds of Costa Rica. Bull. Amer. Mus. Nat. Hist. 128.

Smithe, F. B. 1975. Naturalist’s color guide. Amer. Mus. Nat. Hist., New York, New York.

Vaurie, C. 1971. Classification of the Ovenbirds (Furnariidae). Whitherby, London, U.K.

. 1980. Taxonomy and geographical distribution of the Furnariidae (Aves, Passeri-
formes). Bull. Amer. Mus. Nat. Hist. 166.

ViVEiROS DE Castro, E. B. 1986. Curt Nimuendaju, 104 mitos indi'genas nunca publicados.
A redescoberta do etnologo teuto-brasileiro. Rev. Patr. Hist. Art. Bras. 21:64-1 1 1.

Wetmore, a. 1972. The birds of the Republic of Panama. Part. 3. Smithsonian Institution
Press, Washington, D. C.

Whitney, B. M., J. F. Pacheco, P. S. M. da Fonseca, and R. H. Barth, Jr.. 1996. The
nest and nesting ecology of Acrobatornis fonsecai (Aves: Furnariidae), with implica-
tions for intrafamilial relationships. Wilson Bull. 108(3): 434-448.

’ ’ Parrini. 1995. Two species of Neopelma in southeastern Brazil

and diversification within the Neopelma/Tyranneutes complex: implications of the sub-
species concept for conservation. Ararajuba 3:43-53.

Willis, E. O. (1992) Zoogeographical origins of eastern Brazilian birds. Omit Neotrop 3-
1-15.

WILSON ORNITHOLOGICAL SOCIETY MEETING

The 78th Meeting of the Wilson Ornithological Society will be held 17-20 Apr., 1997,
Kansas State Univ., Manhattan, KS. The Society will meet jointly with the Kansas Orni-
thological Society. Registration materials and call for papers will be sent to members in
December 1996. Inquiries about the scientific program should be directed to John C.
Kricher, Biology Dept., Wheaton College, Norton, MA 02766 (508-286-3950;
jkricher@wheatonma.edu). Local chair is John L. Zimmerman, Div. Biology-Ackert Hall,
KSU, Manhattan, KS 66506-4901 (913-532-6659 or -6615).

Wilson Bull., 108(3), 1996, pp. 434-448

THE NEST AND NESTING ECOLOGY OE
ACROBATORNIS FONSECAI (EURNARIIDAE), WITH
IMPLICATIONS EOR INTRAFAMILIAL
RELATIONSHIPS

Bret M. Whitney, ‘ ^ Jose Fernando Pacheco,'

Paulo Sergio Moreira da Fonseca,^ and Robert H. Barth, Jr.'*

Abstract. — Descriptions of the nest and nesting ecology of Acrobatornis fonsecai (Pink-
legged Graveteiro), a newly described genus and species in the Furnariidae, are presented.
Nests, constructed of twigs and sticks, are single-chambered, well-lined with mosses and
leaves (one examined in detail), and situated in the canopy of tall trees. In October 1995,
we located 131 nests in 72 trees at 54 sites. The average number of nests/tree was 1.8 with
a maximum of five nests in a single tree; apparently only one nest/tree is active. “Extra”
nests were often smaller than active nests, and at least sometimes had no entrance or cham-
ber. We postulate that these nest-like structures represent dummy or cock nests to confuse
predators or parasites (it certainly worked on us), and may serve as resource stores (i.e.,
construction materials and nest foundations). Brief observations indicated that immatures
(probably offspring) help adults in nest construction, and may help feed food-begging ju-
veniles. Comparison with other, possibly related, furnariids, suggests that nest architecture
of A. fonsecai is most similar to that of the “stick-nesting” group of Asthenes canasteros,
for which nests are relatively well known, but is also similar to some Cranioleuca spinetails
and perhaps to the Xenerpestes graytails and the Metopothri.x plushcrown, which are poorly
known. Our data supplement the discussion of morphological, vocal, and behavioral com-
parisons of the same groups presented by Pacheco et al. (1996). We postulate that stick-
nesting in Furnariidae arose in a pre-Andean, Chaco-Patagonian/Pantanal center, and provide
some theories on the evolution of this behavior. Received 23 April 1996, accepted 21 May
1996.

Resumo. — Descrigao do ninho e dados ecologicos da nidifica9ao dc Acrobatornis fonsecai
(Acrobata), um novo genero e especie de Furnariidae recentemente descrito sao apresenta-
dos. Os ninhos, construidos de gravetos, possuem uma unica camara bem forrada com
musgos e folhas (N = 1 examinado em detalhe), e sao situados na copa de arvores altas.
Em outubro de 1995 foram localizados 131 ninhos em 72 arvores em 54 pontos diferentes.
O niimero medio de ninhos por arvore foi de 1,8, com um maximo de cinco ninhos em
uma unica arvore; aparentemente apenas um ninho por arvore e ativo. Os ninhos “extras”
sao geralmente menores do que os ninhos ativos e, ao menos as vezes, nao apresentam
entrada ou camara. E postulado que estas estruturas representem “ninhos falsos” para con-
fundir predadores ou parasitas (como aconteceu com os autores) ou, talvez, para servir como
reserva de recursos (i.e., material de constru9ao e “alicerce” de ninhos). Breves observa9oes
indicam que imaturos (provavelmente filhotes de uma ninhada anterior) ajudam adultos na
constru9ao do ninho e, talvez, colaborem na alimenta9ao de jovens. Compara96es com
outros, possivelmente aparentados, furnan'deos, sugere que a arquitetura do ninho de A.

' Institute de Biologia, Depto. de Zoologia, Cidade Universitaria, Universidade Federal do Rio de Janeiro
21941-000, Rio de Janeiro, RJ, Brasil.

- Mu.seum of Natural Science, 1 19 Foster Hall, Louisiana State Univ., Baton Rouge, Louisiana, 70803.

^ Rua Diamantina 20, Apto. No. 201, 22461, Rio de Janeiro, RJ, Brasil.

■* Dept, of Zoology, Patterson Laboratory, Univ. of Texas, Austin, Texas, 78754.

434

Whitney et cil. • NESTING ECOLOGY OF ACROBATORNIS FONSECAI 435

fonsecai tenha mais semelhan^a com aqueles do grupo de Asthenes “construtores de ninhos
de graveto, para o qual os ninhos sao relativamente bem conhecidos, mas tambem apresenta
similaridades com aqueles de algumas Crcinioleuca e talvez com aqueles de Xenerpestes e
Metopothrix, os quais sao insuficientemente conhecidos. Os dados sobre o ninho e nidifi-
ca^ao complementam a discussao de mortologia, vocaliza^ao e comportamento em com-
paragao aos mesmos grupos conforme apresentados por Pacheco et al. (1996). E teorizado
que a contru^ao de ninhos de gravetos em Furnariidae surgiu num periodo pre-andino no
centro Chaco-Patagonia-pantanal, e sao fornecidas algumas teorias sobre a evolu^ao deste
comportamento.

Nest architecture and placement have been considered important in
judging systematic relationships in the Neotropical family Furnariidae
(Vaurie 1971, 1980), and we concur that data on nests, at least insofar as
definition of basic nest-type, are desirable in any analysis of intrafamilial
relationships of this complex assemblage of birds. We are also in full
agreement with Narosky et al. (1983) that caution regarding assumptions
of relatedness based on current classifications or on nesting similarities
must be maintained. Data on nests should be overlaid with as many other
potentially informative data as possible. This report on the nest and nest-
ing ecology of the recently discovered and described Acrobatornis fon-
secai (Pink-legged Graveteiro; Pacheco, et al. 1996), a new genus and
species in the Furnariidae, supplements the discussion of its relationships
based on intrafamilial comparisons of morphology, vocalizations, and be-
havior presented by Pacheco et al. (1996). Distribution of A. fonsecai in
the remnant lowland forest of southeastern Bahia, Brazil, was mapped by
Pacheco et al. (1996; Fig. 4).

Description of the nest of Acrobatornis fonsecai. — As seen from the
ground, the nest of Acrobatornis fonsecai is a globular, ovoid, or roughly
rectangular structure of twigs and sticks, usually situated in a fork of
branches, sometimes on top of a limb if angled less than about 30° above
the horizontal, and inside the crown of a tall tree surrounded by or ad-
jacent to other tall trees (Fig. 1). Because nests are generally conspicuous
in treetops, sometimes in leafless trees, we located them easily with visual
searches, mostly from roadsides. These nests may have been overlooked
or ignored by the many observers who have traversed this region in the
past because of their superficial similarity to nests of Phacellodonius ruf-
ifrons (Rufous-fronted Thornbird) which, however, typically builds much
longer nests that hang in a vertical column from the periphery of isolated
trees (Skutch 1969a, Thomas 1983). Like those of A. fonsecai, nests of
thornbirds are also variable in shape and size, with some, probably un-
finished, nests being constructed around upright branches or even thin
tree trunks (Figs. 2, 3). Such nests are often smaller than normal, and
similar in outward appearance to those of A. fonsecai.

436

THE WILSON BULLETIN • VoL 108, No. 3, September 1996

Lig. 1. Nest tree of Acrobatornis fonsecai at the edge of a mixed cocoa and banana
plantation along a road near Arataca, Bahia. Most nest trees were in closer proximity to
other tall trees. This Senna multijuga tree held five nests (see Lig. 2). Photos in this and
other figures by Whitney.

Between 4 and 12 October 1995 we plotted a total of 131 nests of
Acrobatornis fonsecai in 72 nest trees at 53 different sites in the Itabuna-
Camacan region of southeastern Bahia (see Fig. 4 in Pacheco et al.
[1996]). At 36 sites we observed only a single nest tree, at 16 sites, two
nest trees, and 1 site had four trees with nests. At sites with more than
one nest tree, we noted that these were generally less than about 100 m
apart (unless on opposite sides of a road), but that they were rarely im-
mediately adjacent. We do not know whether these separate trees were
occupied by separate pairs of A. fonsecai. As can be deduced from the
numbers above, individual trees usually held more than one nest. Of the

Whitney et al. • NESTING ECOLOGY OF ACROBATORNIS FONSECA! 437

Fig. 2. Nest tree (Senna niultijuga', see Fig. 1) of Acrobatornis fonsecai near Arataca,
Bahia. Five nests are visible; we do not know if any were active in October 1995.

total of 72 nest trees we found in October, 31 (43%) held one nest, 27
(37%) held two nests, 1 1 (15%) held three nests, 2 held four nests, and
1 tree had five nests (Figs. 1, 2). The average number of nests/tree was
1.8. Nest-height varied considerably with tree height, but nests were al-
ways in the upper V3, usually in the upper Va, of trees. Most nests were
in excess of about 20 m above ground, and we estimated some as higher
than 30 m.

Acrobatornis fonsecai builds nests mostly in mature trees of the family
Leguminosae. Of the 72 nest trees located, 37 (51%) were members of
the genera of Leguminosae mentioned in the “habitat” section of Pacheco
et al. (1996; except that no nests were found in Inga species), 25 (35%)
were not identified to family (but we suspect that many of these were
male Erythrina species), and 10 were leafless trees (most of which we
suspect were Leguminosae that had dropped their leaves). Leguminosae
are among the dominant trees along roadsides over cocoa plantations in
the range of A. fonsecai. This notwithstanding, we do not estimate that
Leguminosae were so overwhelmingly more numerous than other tall
trees that this could account for the tact that A. fonsecai nested mostly in
Leguminosae. A more plausible explanation might be that the relatively
open nature of the crowns of Leguminosae, and the fact that many had

438

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Eig. 3. Nest tree of Phacellodomus r. rufifrons isolated in a field near Arataca, Bahia.
Three nests are visible. Vertically oriented, pendant nest on lower left of tree is typical of
an active nest. The two spherical, nest-like structures built around limbs are superficially
similar to nests of Acrobatornis.

dropped at least some of their leaves in October, made it easier to see
nests in them than in many other trees. To test this, we searched carefully
in various kinds of trees shading cocoa and presented tape playback of
A. fonsecai in parts of the serra Bonita where no nests were obvious. This
effort resulted, however, in the location of only one nest tree, which
turned out to be a Schizolohium parahyba (Leguminosae).

We were able to collect two of five nests in a Senna multijuga tree
(Figs. 1, 2), neither of which was active. One of these, shown in Fig. 4,
was damaged on removal from the tree, having lost a horizontal (roughly

Whitney et ul. • NESTING ECOLOGY OF ACROBATORNIS FONSECA! 439

Fig. 4. Ne.st of Acrobatornis fonsecai removed from nest tree shown in Figs. I and 2.
This nest is missing the entrance tunnel, but shows the single, moss-lined chamber with its
covering of twigs and sticks.

in the same line as the main axis of the nest) antechamber or entrance
tunnel of undetermined length. We noted such entrance tunnels on all four
active nests located, and numerous other nests, and concluded that it is a
typical feature of active nests of A. fonsecai. Entrance tunnels were al-
ways toward one end of the nest, usually the lower end if the orientation
of the nest was other than horizontal. Bearing in mind that description of
most aspects of the external, stick-structure of this nest are rendered some-
what inaccurate because an undetermined number of sticks was lost, we
present the following observations.

The outer layer, which was made up entirely of a dense weave (lining
materials not visible) of sticks, none of which had thorns or spines, mea-
sured 18 cm long, about 24 cm in diameter (across the center of the
chamber), and 18 cm tall or deep, not including sticks extending out
irregularly from the main body. We dismantled the nest for more detailed
analysis of its architecture. It contained 374 sticks (total mass 115 g)
ranging 1 to 3 mm in diameter, which we separated into three classes by
length: 10-15 cm (276; 74% of number, 50% of mass); 15-20 cm (75;
20% of number, 32% of mass); and over 20 cm (23; 6% of number, 18%
of mass). The longest and thickest sticks, all of which were around the

440

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Eig. 5. Nest-like structure of Acrobatoniis fonsecai, built around the fork of a rather
heavy branch, with no entrance or chamber, and with an EpiphyUum sp. cactus growing out
of it. This structure contained several small ant nests. We suspect that such nest-like struc-
tures, or “extra nests” of Acrobatoniis might serve as dummy or cock nests, or as stores
of sticks and building foundations.

outside edge, were 27 cm long and 3 mm in diameter (N = 4). These
sticks seemed light in weight relative to their length, varying from 0.8—
1.2 g.; the heaviest equaled about 8.5% of the body weight of the bird
(about 14 g.).

Beneath the external layer of sticks was a dense lining, 18 cm in di-
ameter (thus 75% of the total width of the nest) and about 10 cm deep,
with a mass of about 85 g. It was made up primarily (about 70%) of one
type of moss, most of which appeared to be healthy and green when
collected, and even when the nest was dismantled in February 1996. Also

Whitney el cil. • NESTING ECOLOGY OF ACROBATORNIS FONSECA/ 441

woven into this layer were rachises of decomposed leaves, which were
more numerous around the incubation chamber. The chamber was quite
rounded, and measured about 6X6 cm. Concentrated around it were
pieces of Tillandsia usneoides lichens and Marasmius sp. fungus (see Sick
1957). One convoluted strand of the latter measured 90 cm. The chamber
was surrounded with one type of leaf, which seemed to be a species of
bamboo or bamboo-like grass. These leaves were folded or wrapped
around the walls of the chamber, and averaged 12 cm long and about 2.5
cm wide. Some of the leaves had flecks of bird droppings on them. Be-
cause this nest was not active at the time of collection, it is possible that
other birds or mammals had modified the interiormost lining materials.

Some nests are considerably longer (thus, more rectangular in profile)
than others, which we suspect is owing mostly to variation in length of
the entrance tunnel. This variation was not great relative to variation in
the dimensions of the stick-nests of some other furnariids, as in some
species of Phacellodomus, Pseudoseisura, dtnd Asthenes species (pers. ob-
serv.), and we estimated the largest nest we saw to be about 45 cm long
and of average circumference.

Our limited observations suggest that only one nest per tree is active
at one time. This is reported to be the case with Phacellodomus rufifrons,
which often has multiple nests in a single tree (Thomas 1983; Fig. 3). A
variety of furnariids are known to build substantial stick-nests (Vaurie
1980, Narosky et al. 1983) that may persist with little external damage
for months or even years (e.g., Skutch 1969, Nores and Nores 1994,
various taxa pers. observ.). Thus, it seems likely that some of the “extra”
nests of Acrobatornis fonsecai in multiple-nest trees are old nests.

Does Acrobatornis build dummy or cock nests? — A number of obser-
vations suggest that A. fonsecai may frequently construct one or more
dummy or cock nests in trees with an active nest. Of two nests collected
the one described above appeared to be a true or complete nest, with a
single, well-lined chamber occupying most internal space of the nest. It
was not active when collected, but the fact that it was lined, and had
traces of old bird droppings on some of the lining material, may indicate
that it was at one time an incubation nest rather than a dormitory or
dummy nest (Skutch 1969a). The other nest, although quite similar in
overall size, shape, and external composition to the true nest, had no sign
of an entrance or internal chamber; it was simply an oblong ball of sticks
in a fork of a branch with one part of the branch through the middle of
it, thus, not an old nest or a potential dormitory (Fig. 5). This nest had
an epiphytic cactus {Epiphyllum sp.) growing out of it in several direc-
tions, which we suspect had formed the original foundation for construc-
tion, and there were several small ant nests within it. In the orientation

442

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

in which it was in the tree (similar to that in Fig. 5), it measured 32 cm
high, 22 cm wide, and 16 cm deep. Relative to the other nest, this cham-
berless one was constructed of shorter sticks.

A third nest from this same tree fell apart when detached from its
supporting limb. We could not, therefore, examine it in detail, but we
determined that it contained no lining, and suspect that it had no internal
chamber. We noted on many occasions that “nests” in multi-nest trees
showed appreciable variation in size, with the smallest ones often too
small to have an internal chamber.

We have one more, rather fascinating observation relating to the pos-
sible deployment of dummy nests by A. fonsecai. We observed that one
of four nests in a single tree near Camacan was actually a small, arboreal
ant nest that had been decorated with twigs, resulting in its remarkable
similarity to the other three Acrobatornis nests in the tree. It seemed to
us that the sticks had been applied evenly and loosely in a horizontal
orientation, after the ant nest was in place, and not left over after an
interior occupation and outward construction by the ants that resulted in
a uniform distribution of sticks. It also seems almost unimaginable to us
that the birds had perceived that the ant nest was similar in size and shape
to their own nests, then added some sticks to make it a dummy. It is
perhaps more likely that they started adding twigs to the stable substrate
of the ant nest, and built around it to some extent. Regardless of the
birds’ “intent,” it fooled us for a few moments.

We suggest that the normal-sized but chamberless nest (Fig. 5), the
“customized” ant nest, and small, probably chamberless nest-like struc-
tures in multi-nest trees, serve as dummies that may confuse predators as
to the location of the true nest. Such a function has been attributed to
dummy nests atop true nests constructed by pairs of Barred Waxbills
{Estrilda astrild).

The multi-chambered stick nest of Phacellodomus rufifrons may con-
tain dormitories and an incubation chamber or, if not the active nest, just
dormitories, but they apparently always have internal chambers (Skutch
1969, Thomas 1983). Skutch (1969) believed that the complexity of nests
of P. rufifrons made it difficult for predators (and even him) to locate the
eggs and young within, and he proposed that complex construction was
not only designed to confuse predators but also represented “an outlet
for excess energy or a pastime.”

Although neither of the above authors suggested that the multiplicity
of nests of P. rufifrons in a single tree might confuse predators (or nest
parasites like Tapera naevia, the Striped Cuckoo), looking for eggs,
young, or adults, we suspect that this could contribute an important ad-
vantage for thornbirds’ survivorship. The same reasoning applies equally

Whitney et al. • NESTING ECOLOGY OF ACROBATORNIS FONSECAI 443

well to explain the tact that Acrobatornis fonsecai often has more than
one nest or nest-like structure in a tree. Even if non-active nests are
assumed or eventually proven to be nests previously used by the birds as
incubation chambers or dormitories, the fact remains that the birds re-
peatedly select the same tree for nest construction, one reasonable con-
sequence of which is confusion of predators.

-Another, non-exclusive, explanation for the construction of more than
one nest in a single (presumably especially desirable) tree might be that
“extra nests” represent resource stores (i.e., building materials and con-
struction foundations). Sticks take a great deal of time and energy to
gather and transport, and are a valuable enough resource that they are
pirated by conspecifics in some species (Skutch 1969, Thomas 1983), or
even other, unrelated species (pers. observ.). Furthermore, a wide variety
of birds (conspecifics and others) are known to take advantage of old
stick nests for construction materials or in their entirety, for nesting. We
observed Phacellodomus rufifrons stealing sticks from a nest of Acro-
batornis fonsecai', we noted no interspecific interaction, and we were un-
able to determine whether the nest of the latter was active at the time.
Extra nests of Acrobatornis, many of which are smaller than active nests,
might also secure foundation sites, ensuring the rapid construction of a
nest should the primary one be lost or damaged. Thomas (1983) pointed
out that the most difficult and energy-expensive stage of nest construction
for Phacellodomus rufifrons was, by far, the establishment of a founda-
tion.

Does Acrobatornis have “helper” offspringl — In early October, we
observed many Acrobatornis fonsecai occupied principally with feeding
young (see Pacheco et al. 1996), and we observed nest-building or main-
tenance behavior on only one occasion. Late on the afternoon of 1 1 Oc-
tober we saw three of four A. fonsecai remove sticks, one at a time, from
a single nest in a densely foliated tree and carry them to another, slightly
taller, leafless tree about 40 m away that contained two nests. Two adult
birds and one of two brown, immature birds each carried one stick. The
birds took sticks in the bill near the midpoint and, flying rather laboriously
with the neck craned upwards, landed on the nest under construction.
After clambering around on the top and sides of the nest for a moment
with the stick, they deftly placed it in the upper exterior of the nest. We
do not know how sticks are originally gathered (i.e., from the ground or
by breaking them off trees) but, among furnariids, reuse of sticks from
old nests or nests of other species has been reported by Skutch (1969)
and Thomas (1983) for Phacellodomus rufifrons, and for Pseudoseisura
lophotes by Nores and Nores (1994).

The observation of an immature bird involved in construction of a nest

444

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

where adults were also building seems to be rare within the Furnariidae.
Such “helping” behavior of presumed offspring has been previously re-
ported for Phacellodomus rufifrons by Gilliard (1959) and Skutch (1969),
although Thomas (1983) judged that the contribution of young thombirds
to nest construction and maintenance was “minimal.” Additionally, Nores
and Nores (1994) found that young Pseudoseisura lophotes performed a
low level of helping in nest construction. Because few studies sufficiently
detailed to reveal this kind of behavior have been conducted on furnariids,
it is perhaps not surprising that there have been so few reports of young
helping parents in construction of nests. In the case of Acrobatornis fon-
secai, in which adults and immatures (juveniles, at least) are strikingly
dichromatic, a most unusual condition in the Furnariidae, it would be
relatively easy to conduct further observations to determine to what extent
immatures (and, if color-banded, offspring) assist in nest-building or other
activities.

Brown (1987) included Phacellodomus rufifrons in a list of species (his
table 2.2) having helpers at the nest, citing Skutch (1969) and Thomas
(1983). He defined a helper as “an individual that performs parent-like
behavior toward young that are not genetically its own offspring.” How-
ever, the accounts of helping in Skutch (1969) and Thomas (1983) doc-
ument only that presumed young birds occasionally help parents in nest
construction or maintenance. Phacellodomus rufifrons, as presently
known, then, does not fit Brown’s (1987) definition of a helper and should
be removed from his table 2.2. Consequently, no member of Furnariidae
is known to have a helper. On the morning of 1 1 October, Barth saw an
immature (brown-plumaged) Acrobatornis fonsecai feed an insect to a
food-begging juvenile being fed occasionally by a pair of adults. The four
birds probably formed a family group. This observation of apparent help-
ing (sensu Brown 1987; no pun intended) is intriguing, and merits further
investigation.

The extra, possibly dummy nests of Acrobatornis, without entrances or
chambers, if proven to be typical, might be a simpler, primitive form of
false nest, with more complex, derived, dummy nests of some other birds
(e.g., Phacellodomus rufifrons, some Troglodytidae) having evolved to
serve as dormitories as well. It seems worthwhile to advance the possi-
bility, in other words, that these nest-like structures represent the ances-
tral, least-complex state, the derived state of which is dummy nests that
have false chambers that may or may not serve as dormitories. One ap-
parent problem with this idea is that dummy nests have not been reported
for other furnariids (or other suboscines?). We suggest, however, that extra
thornbird nests within a single tree or, in the case of very large nests such
as those of Pseudoseisura species, perhaps even in nearby trees, could be

Whitney el cil. • NESTING ECOLOGY OE ACROBATORNIS FONSECA! 445

dummy nests (i.e., construction of dummy nests might have been mis-
interpreted to a large extent). Interestingly, neither Skutch (1969), Thomas
( 1983), or Nores and Nores ( 1994) advanced any explanation for multiple
stick-nests in a tree, apparently assuming that these were all old nests or
(in the case of Skutch) the result of excess energy. Another problem is
that chamberless “nests,” whether dummies or not, are apparently unre-
ported in birds. Thus, the theory that Acrobatornis fonsecai might be
using such structures as dummy nests or resource stores (as described
earlier) is novel and, of course, untested.

Intrafamilial comparison of nest architecture. — Genera and species
here compared with Acrobatornis fonsecai are the same discussed by
Pacheco et al. (1996) for comparisons of morphology, vocalizations, and
behavior.

Cranioleuca. — Nests (apparently only one per tree) are generally 25-
30 cm in diameter, and are single-chambered, globular or conical masses
of moss, grass, thin vines, and other flexible vegetation, in some species
pendant from limbs at the periphery of trees, in others placed in a fork
of branches or network of supporting vines and other vegetation. Crani-
oleuca pyrrhophia (Stripe-crowned Spinetail, of semiarid scrub and
woodland in Bolivia, Paraguay, and northern Argentina), however, builds
a nest of dry, thorny twigs bound with wool and vegetable fiber, and well-
lined with lichen, or other soft material (Hoy in Vaurie 1980), or of soft
vegetable material with a covering of sticks, in some cases spiny ones
(Narosky et al. 1983). Reports differ regarding the location of the entrance
(near the top or the bottom), and there is apparently no entrance tunnel.

Asthenes. — In comparing nests of Asthenes species with that of Acro-
batornis fonsecai, we follow Pacheco et al. (1996) in limiting discussion
to the “stick-nesting” group of Asthenes. Typical nests (usually one, oc-
casionally two or three, per tree/shrub) are masses of twigs and sticks
roughly 20-40 cm in diameter with single internal chambers built, for
example, inside the crown of a tree or shrub, around the arms and trunk
of columnar cacti, or within piles of rocks or in vegetation clinging to
cliffsides. Some, such as A. patagonica (Patagonian Canastero), have en-
trance tunnels as long as the main body of the nest (Narosky et al. 1983;
pers. observ.).

Thripophaga. — Both T. macroura (Striated Softtail) and T. fusciceps
(Plain Softtail) construct roughly globular nests about 20 cm in diameter
of small twigs and flexible vegetable material, such as grasses, rootlets,
and thin vines. Nests are situated on thin limbs in the crowns of midstory
and subcanopy trees, near the periphery of the tree, inside or at the edge
of tall forest. We have seen only two nests of each species, however, and
have examined none in detail.

446

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Phacellodomus.—^es,i?, of most species are similar to those of P. ruf-
ifrons shown in Fig. 3 and described in detail by Skutch (1969) and
Thomas (1983). As is true of Acrobatornis fonsecai, there is often more
than one nest/tree; active nests are usually situated at the periphery of
trees, well below the crown and, as described earlier, differ from nests of
A. fonsecai in a number of important respects.

Xenerpestes. — The two distinctive species in this genus are poorly
known. Unfortunately, nests remain undescribed, although Ridgely and
Gwynne (1989) suspected that a large stick-nest in eastern Panama be-
longed to the Double-banded Graytail {X. minlosi). Whitney observed a
pair of X. minlosi hopping on a mass of twigs and sticks about 20 cm in
diameter that he suspected was their nest, in the crown of a tall, foliated
and flowering Erythrina tree at the Cana airstrip, Darien, Panama, in
January 1992.

Metopothrix. — The sole member of Metopothrix, M. aurantiacus (Or-
ange-fronted Plushcrown), like Xenerpestes, is poorly known. Nests were
briefly described by Ridgely and Tudor (1994:128) as masses of sticks
nearly 0.5 m across with the entrance at the side, built on lateral branches
of trees from 4 to 20 m above ground, “not dissimilar in overall form
from those of Phacellodomus thornbirds.” These authors stated that Fraga
(1992) had previously “described just such a nest” of Metopothrix, but
Fraga’s paper indicates that he saw birds carrying sticks only; he did not
observe or describe a nest. Metopothrix nests are apparently considerably
larger than those of Acrobatornis and are sometimes placed much nearer
the ground; details of construction remain unknown.

Margarornis. — The only descriptions of the nest of any species appear
to be those of Hilty and Brown (1986:367) and Fjeldsa and Krabbe (1990:
384) for M. squamiger (Pearled Treerunner): “moss ball nest with side
entrance” and “closed nest of moss, placed under a limb or a rock,”
respectively. In exterior architecture and general size and shape, nests of
“stick-nesting” Asthenes are much like nests of Acrobatornis fonsecai,
and appear to be the most similar in the family. Cranioleuca pyrrhophia's
construction of sticks around a well-lined chamber recalls that of A. fon-
secai. This seems to be the only described stick-nest of that genus, al-
though Whitney has recently discovered that C. meulleri (Scaled Spine-
tail) of the lower Amazon region, also builds an arboreal stick-nest (ms.
in prep.). Cranioleuca pyrrhophia lives in habitats with little or no moss
or flexible, herbaceous growth suitable for structural binding (or at least
no reliable sources of such materials), which may have promoted stick-
nesting (or the maintenance of it) in this species. We suspect, however,
that the single-chambered, mossy globes of some of the other Cranioleuca
species would be quite similar to nests of Acrobatornis if covered with a

Whiitiey et al. • NESTING ECOLOGY OE ACROBATORN/S FONSECA! 447

layer of sticks. Just as sticks are abundant and easily accessible in the
habitat ot C. pyrrhophia, they are relatively rare (i.e., soft, hard to break
off, and decompose quickly on the ground) in the humid montane habitats
of most of the other members of the genus. We await documentation and
detailed descriptions of the nests of the two species of Xenerpestes, and
of Metopothrix aurantiacus.

On the origin of stick-nesting in Furnariidae. — There appears to be no
published discussion of the origin of stick-nesting in Furnariidae. We
suspect that powerful environmental factors in place for prolonged periods
would be required for evolution and establishment of such energetically
expensive, sex-shared (i.e., nests are not sexually selected structures), nest
architecture across the broad group of furnariids in which the behavior is
prevalent today. During arid or semi-arid epochs, for example, there may
have been few other construction materials available. Extended periods
of winds, frequent violent weather, or cold would represent substantial
selective forces. Similarly, fortified nests might have evolved to thwart
large reptilian and avian predators and to withstand the shock of regular,
incidental contact of nests and supporting vegetation by large vertebrates.
Operative evolutionary mechanisms aside, we assume that the ancestral
forms of stick-nesting furnariids arose in a southern, Chaco-Patago-
nian/Pantanal (in contemporary terms) distributional center during a pre-
Andean epoch. In light of the overwhelming concentration and diversity
of stick-nesting species surviving in this region of the continent today,
origin of stick-nesting there is a reasonable assumption. This ancient cen-
ter probably extended to interior northeastern Brazil, which today shares
numerous forms with Chaco-northern Patagonia and the Pantanal (Short
1975 and numerous subsequent authors), including such stick-nesting fur-
nariids as the Chotoy Spinetail (Schoeniophylax phryganophila), Phacel-
lodomus rufifrons. Firewood-gatherer {Anumbius annumbi), and Rufous
Cacholote (Pseudoseisura cristata). The contemporary distribution of
stick-nesting Asthenes reaches its northeastern extreme only slightly far-
ther south, in the serras of Minas Gerais {A. luizae Cipo Canastero). Cer-
tain successful forms apparently radiated widely (e.g., Phcicellodomus ruf
ifrons and Anumbius, which are still spreading, following forest clear-
ance), and some, like stick-nesting Asthenes, speciated rapidly as they
colonized a new, vertical stratum of Andean habitats to spread north to
central Peru, where speciation seems relatively incipient. Forest-inhabit-
ing Cranioleuca may have radiated following evolution of a more recent
ancestral form in the forests that must have flourished with the conden-
sation-precipitation (at least) resulting from Andean uplift. Such a se-
quence of events implies that stick-nesting is the primitive condition in
this group of birds, at least.

448

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

ACKNOWLEDGMENTS

We are grateful to Haroldo C. de Lima of the Botanical Garden of Rio de Janeiro for
determining the species of one of the nest trees from dried samples we provided. Claudia
Bauer and Luiz Gonzaga helped in analysis of nest architecture of A. fonsecai. We thank
Hannah Gould of the Univ. of Texas at Austin for helping us locate pertinent references.
Gary Graves and J. V. Remsen, Jr. commented on the manuscript, and Charles Blem assisted
in seeing that it was published promptly. Holland Photo ot Austin produced the black-and-
white photos in the figures. Eield Guides Incorporated, of Austin, Texas, generously financed
part of our expenses for research in southern Bahia.

LITERATURE CITED

Brown, J. L. 1987. Helping and communal breeding in birds: ecology and evolution.

Princeton Univ. Press, Princeton, New Jersey.

FjeldsA, j. and N. Krabbe. 1990. Birds of the high Andes. Zoological Museum, Univ. of
Copenhagen, and Apollo Books, Svendborg, Denmark.

Fraga, R. M. 1992. Nesting behavior of Metopothrix aurantiacus in Ecuador. Hornero, 13:
236.

Gilliard, E. T. 1959. Notes on some birds of northern Venezuela. Amer. Mus. Novitates
1927:1-33.

Hilty, S. L. and W. L. Brown. 1986. A guide to the birds of Colombia. Princeton Univ.
Press, Princeton, New Jersey.

Narosky, S., R. Fraga, and M. de La Pena. 1983. Nidificacion de las aves argentinas
(Dendrocolaptidae y Furnariidae). Asoc. Orn. Plata, Buenos Aires, Argentina.

Nores, a. I. AND M. Nores. 1994. Nest building and nesting behavior of the Brown
Cacholote. Wilson Bull. 106:106—120.

Pacheco, J. E, B. M. Whitney, and L. P. Gonzaga. 1996. A new genus and species of
furnariid (Aves: Furnariidae) from the cocoa-growing region of southeastern Bahia,
Brazil. Wilson Bull. 108(3):397-433.

Ridgely, R. S. and G. Tudor. 1994. The Birds of South America. Vol 11. University of
Texas Press, Austin, Texas.

Short, L. 1975. A zoogeographic analysis of the South American Chaco avifauna. Bull.
Amer. Mus. Nat. Hist. 154:165—352.

Sick, H. 1957. Rosshaarpilze als Nestbau-Material brasilianischer Voegel. J. Orn. 98:421-
431.

Skutch, a. F. 1969. A study of the Rufous-fronted Thornbird and associate birds. Part 1.

Life history of the Rufous-fronted Thornbird. Wilson Bull. 81:5-43.

Thomas, B. T. 1983. The plain-fronted Thornbird: Nest construction, material choice, and
nest defense behavior. Wilson Bull. 95:106-117.

Vaurie, C. 1971. Classification of the Ovenbirds (Furnariidae). Whitherby, London, U.K.
1980. Taxonomy and geographical distribution of the Furnariidae (Aves, Passeri-
formes). Bull. Amer. Mus. Nat. Hist. 166.

Wilson Bull., 108(3), 1996, pp. 449-456

WOODPECKER EXCAVATION AND USE OE CAVITIES
IN POLYSTYRENE SNAGS

Richard N. Conner and Daniel Saenz

Abstract. — We examined woodpecker excavation and use of artificial polystyrene snags
in four forest types in eastern Texas for five years. Twenty-three of 47 artificial snags were
used by Downy Woodpeckers (Picoides pubescens) for cavity excavation and subsequent
nocturnal roosting; they did not use the artificial snags for nesting. Although six other species
of woodpeckers were present in the area, only Downy Woodpeckers excavated cavities in
the artificial cavity substrate. Entrances to cavities in artificial snags became enlarged within
several months of excavation. Other wildlife species using abandoned cavities in artificial
snags were Carolina Chickadees {Pams carolinensis), Prothonotary Warblers (Protonotaria
citrea), southern flying squirrels (Glaucomys volans), and red wasps (Polistes sp.). In one
instance, Carolina Chickadees excavated their own cavity and nested within a polystyrene
snag. Until an artificial cavity substrate acceptable for both woodpecker excavation and
nesting can be found, the utility of artificial snags as a means to augment woodpecker nesting
substrate remains inadequate. Received 18 October 1995, accepted 16 January 1996.

Many woodpecker species and secondary cavity nesters depend on
snags (standing dead trees) for cavity sites that they use for nesting and
roosting (Conner 1978, Evans and Conner 1979, Thomas et al. 1979,
Raphael and White 1984). Harvesting of mature forests can greatly reduce
the availability of substrate for woodpeckers to excavate nest cavities
(Conner 1978, Dickson et al. 1983). Thus, artificial cavity substrate may
benefit nesting woodpeckers in areas where snag availability is low.

Peterson and Grubb (1983) evaluated woodpecker use of 50 artificial
polystyrene snags (242-cm high X 22-cm diameter) over an 11 -month
period in Ohio. Downy Woodpeckers {Picoides pubescens) excavated 51
cavities in 42 of the snags, used them for nocturnal roosting, but failed
to use the cavities for nesting. House Wrens {Troglodytes aedon) and
Carolina Chickadees {Parus carolinensis) nested in cavities excavated by
Downy Woodpeckers. Peterson and Grubb (1983) speculated that other
larger species of woodpeckers might use polystyrene snags if snags
>22-cm diameter were provided, but this idea has never been tested.
Artificial polystyrene snags have also been used to explore sexual differ-
ences in selection of cavity sites by Downy Woodpeckers and to evaluate
cavity entrance orientation and snag selection relative to vegetation in a
regenerating clear cut (Grubb 1982, Petit et al. 1985).

We evaluated woodpecker use of 26-cm diameter X 242-cm high poly-

' Wildlife Habitat and Silviculture Laboratory (Maintained in cooperation with the College of Forestry
Stephen F. Austin State Univ.), Southern Research Station, USDA Forest Service, Nacogdoches, Texas

449

450

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Vegetational Characteristics (Means ± SD) of Mature Pure Pine, Pine-Hardwood,
Upland Hardwood, and Bottomland Hardwood Forest Stands Where Artificial
Polystyrene Snags Were Studied on the Stephen E Austin Experimental Forest in

Eastern Texas

Vegetation variable

Pure pine
(N = 20)

Pine-hardwood
(N = 20)

Upland
hardwood
(N = 20)

Bottomland
hardwood
(N = 20)

Vegetation height (m)

30.0 (3.7)

27.4 (5.5)

20.6 (2.9)

27.1 (5.3)

Pine basal area (m^/ha)

23.5 (3.9)

22.6 (7.3)

3.8 (3.6)

0.2 (0.5)

Hardwood basal area (m^/ha)

0.2 (0.6)

4.0 (3.2)

15.6 (3.5)

18.5 (4.8)

Tree density (#/0.04 ha)

11.5 (3.6)

18.5 (9.6)

10.1 (3.2)

14.0(3.6)

Canopy closure (%)

73.1 (11.1)

71.2 (14.3)

69.3 (13.8)

72.5 (13.0)

Ground cover (%)

2.9 (2.8)

3.5 (2.4)

3.5 (2.7)

9.6 (6.4)

Natural snags (#/0.04 ha)

0.8 (0.8)

0.7 (0.9)

0.7 (0.8)

1.1 (1.0)

Styrene snags in four forest types over a five-year period. We determined
secondary cavity nester use of woodpecker cavities and evaluated cavity
shape and condition with long-term use.

STUDY AREAS AND METHODS

We constructed 47 artificial snags from solid blocks of polystyrene (26-cm diameter X
242-cm high). The 4-cm increase in diameter of the polystyrene snags above what had been
used previously (Peterson and Grubb 1983), placed the substrate diameter within the range
of sizes used by Hairy {Picoides villosus) and Red-bellied {Melanerpes carolinus) wood-
peckers for cavity sites (Conner 1978). Similar to Peterson and Grubb (1983), we painted
the artificial snags with a thick coating of brown latex paint to enhance the snag-like ap-
pearance of the polystyrene snags. After drilling a centrally located 3-cm diameter hole
(parallel to the length of the snag), 80 cm deep into the base of each artificial snag, we
installed it in the field on 20 October 1986 by sliding it onto a 184-cm long “T-pole” (iron
fence post) that had been driven into the ground approximately 110 cm deep. The hole
drilled into the base of each artificial snag was made solely to mount (impale) the snags on
T-poles. All artificial snags were installed as close to vertical as possible, i.e., no lean could
be visually detected. Artificial snags were installed at 1 12-m intervals on four nest box trails
in four forest types (ten snags per trail and one trail in each forest type: mature pure pine
{Pitms spp.l, pine-hardwood, upland hardwood, and bottomland hardwood forest habitats)
located on the Stephen E Austin Experimental Forest (31°29'N, 94°47'W) in southern Nac-
ogdoches County, Texas. Each nest box trail was circular and approximately 1130 m in
length. Cavities for secondary cavity nesters were readily available on each trail, becau.se
20 sites with three nest boxes per site were established at 56-m intervals on each trail as a
part of a different study. Seven additional artificial snags were installed on the edge of
mature pine-hardwood forest next to dirt roads.

Vegetation characteristics were measured at 56-m intervals (20 points) on each of the
four nest box trails (Table 1 ). We measured vegetation height with a clinometer, and tree
basal areas were measured with a one-factor metric prism. Densities of trees and snags >15
cm diameter at breast height were counted within an 1 L3-m radius circular plot. We esti-

Conner and Saenz • WOODPECKERS AND POLYSTYRENE SNAGS 45 1

Table 2

Species Use of Cavities Excavated by Downy Woodpeckers in Artificial Polystyrene
Snags in Four Forest Types on the Stephen E Austin Experimental Forest in

Eastern Texas

Cavity occupant

Number of polystyi

rene snags used

Pure pine
(N = 10)

Pine-

hardwood“

(N = 17)

Upland
hardwood
(N = 10)

Bottomland
hardwood
(N = 10)

Downy WoodpeckeU

0

13

10

0

Carolina Chickadee

0

4“^

2

0

Prothonotary Warbler

0

0

2

0

Southern flying squirrel

0

1

0

0

Red wasps

0

3

2

0

“Artificial snags in forest (N = 10) and edge (N = 7) pine-hardwood habitat combined.

•’ All cavities except one were initially excavated by Downy Woodpeckers.

' In one instance in pine-hardwood edge habitat Carolina Chickadees excavated their own cavity.

mated percent canopy closure and ground cover, using a 4-cm diameter X 12-cm long hollow
tube. We recorded height and compass aspect of pecking and cavity excavation on all
artificial snags from fall 1986 to summer 1991.

Occupants of cavities were determined by checking roosts with a mirror, watching oc-
cupants use a cavity, or flushing the occupant. Artificial snags were visited during the spring
(March— May), fall (September— October), and winter (December— January) during each year
of the study. The species of woodpeckers excavating cavities in artificial snags were deter-
mined by watching the actual excavation or by measuring the final size of the completed
cavity. We also noted claw marks and their relative size to determine if they had been made
by a squirrel or a possible predator (house cat [Fells domesticus] and raccoon [Procyon
lotor]). We were not able to determine nesting success on all of the avian nests detected
because of time and personnel constraints. Artificial snags in the bottomland hardwood area
were monitored only until spring 1989 because flooding lifted the snags off the T-poles and
washed them down the Angelina River.

RESULTS

Except for one case. Downy Woodpeckers were the only species de-
tected excavating and using cavities in the artificial polystyrene snags
(Table 2). We did not observe Downy Woodpeckers nesting in any of the
cavities, but they regularly used the cavities as nocturnal roosts. Downy
Woodpeckers excavated cavities in artificial snags only in the pine-hard-
wood and upland hardwood forest types. Carolina Chickadees were the
most frequent secondary users of cavities excavated by Downy Wood-
peckers (Table 2). In one instance, Carolina Chickadees excavated a cav-
ity during the early spring and successfully nested in it. Prothonotary
Warblers (Protonotaria citrea) successfully nested in two different cavi-
ties in the upland hardwood forest type. Standing water was present in
parts of this area for much of the spring. Five cavities were used by red

452

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Fig. 1. Seasonal appearance of cavities completed by Downy Woodpeckers in polysty-
rene snags (starting in winter 15 months after snag installation), enlargement of cavities by
subsequent use, and use by secondary cavity users during successive winter (WTR), spring
(SPR), and fall (FALL) seasons on the Stephen E Austin Experimental Forest in eastern
Texas.

wasps {Polistes sp.) and one by southern flying squirrels {Glaucomys
volans).

Artificial snags were in place five months before small holes began to
appear in them in the upland hardwood and pine-hardwood areas during
the early spring 1987. Downy Woodpeckers were the only woodpecker
species observed excavating cavities in the artificial snags, and the first
completed cavities (9) appeared in these two habitat types by early Jan-
uary 1988 (15 months after installation) indicating that they had been
excavated during late fall to early winter 1987. Additional completions
of cavities in other artificial snags occurred during the next two years
(Fig. 1). Avian secondary cavity nesters did not begin to use the com-
pleted cavities until more than a year had passed (Fig. 1). Southern flying
squirrels were first detected after two years.

All completed cavity entrances were excavated between 12 and 16 cm
from the top of the artificial snags. It was difficult to detect visually a
preference for cavity orientation. Cavity entrances appeared to be bimodal
in their distribution (Fig. 2). A Rao’s test indicated a non-random orien-
tation of entrances ((/ = 1,591; P < 0.01).

Small holes that seemed to be similar to cavity starts appeared near the
tops of two artificial snags in the pure pine area within five months of
snag installation. Cavities in those two snags, however, were never com-

Conner and Saenz • WOODPECKERS AND POLYSTYRENE SNAGS 453

s

Fig. 2. Aspects of entrances to cavities excavated into artificial polystyrene snags (N =
22) in eastern Texas.

pleted. By January 1988, two other artificial snags in the pure pine area
had small excavations in them but were also abandoned. Artificial snags
in the bottomland hardwood area had small and some large holes exca-
vated within 30 cm of the base of the snags, most likely excavated by
Pileated Woodpeckers (Dryocopus pileatus). But, apparent start holes in
both the pure pine and bottomland hardwood areas were never excavated
beyond several centimeters deep. Artificial snags in all areas had varying
amounts of their surface paint and polystyrene pecked away, as if wood-
peckers or other bark foragers had attempted to forage on them.

Seven cavity entrances became quite enlarged within 8-10 months fol-
lowing cavity completion and subsequent use (Fig. 1). Although entrances
enlarged in all directions, the bottom of each entrance was affected the
most. Polystyrene would erode away 10-15 cm, most likely during the
passage of the occupant, so that entrances gradually became elongated
vertically. Downy Woodpeckers appeared to abandon enlarged cavities.

454

THE WILSON BULLETIN • Vol. JOS, No. 3, September 1996

Claw marks of sufficient size to suggest attempted predation appeared on
four of the artificial snags with cavities during the fall and winter. In one
instance, the cavity entrance was torn open and about half of the cavity
chamber exposed.

DISCUSSION

Our attempt to use large diameter polystyrene snags to encourage some
of the larger woodpeckers to excavate cavities was unsuccessful. Al-
though both Red-bellied and Hairy woodpeckers were present within the
vicinity, neither species apparently excavated cavities in the artificial
snags. Diameters of the artificial snags were sufficient to house cavities
made by these two species (Conner et al. 1975, Jackson 1976). However,
the 3-m height of the artificial snags, which was the tallest block of
polystyrene commercially available, may have been too low for these two
species. Hairy and Red-bellied woodpeckers typically excavate nest cav-
ities at heights above 3 m (Conner 1978). Downy Woodpeckers often nest
in dead tree stubs that are approximately 3 m in height (Conner et al.
1975). They also are known to excavate cavities in very soft, well-de-
cayed natural snags (Conner et al. 1975, 1976). The consistency of poly-
styrene is very similar to that of well-decayed wood tissue found in some
snags used by Downy Woodpeckers for cavity excavation. Both the poly-
styrene and well-decayed wood tissue can be easily excavated by a human
finger nail. Substrate of such little structural strength may be too soft for
the larger woodpecker species.

Although there were woodpeckers within the pure pine and bottomland
hardwood study areas, none of the polystyrene snags in these study areas
was used for cavity excavation. There was an abundance of natural snags
in the bottomland habitat (Conner et al. 1994, Table 1); thus, the attrac-
tiveness of artificial snags was likely less. Natural snags were as common
in the pure pine stand as they were in the pine-hardwood study area (Table
1). The failure of Downy Woodpeckers to use artificial snags in the pure
pine stand is enigmatic.

The long term value of polystyrene snags as an artificial substrate for
woodpecker cavity excavation appears to be relatively low. Only Downy
Woodpeckers excavated cavities, and they did not nest in the cavities
following excavation. The artificial snags do appear to have some value
as roosting sites for Downy Woodpeckers, and the polystyrene material
is well known for its high insulating ability, which would be particularly
valuable during winter at northern latitudes. Although woodpeckers did
not use the cavities for nesting, secondary cavity nesters such as Pro-
thonotary Warblers and Carolina Chickadees successfully nested in the
artificial substrate. Entrances to cavities, however, soon begin to erode

Conner and Saenz • WOODPECKERS AND POLYSTYRENE SNAGS 455

away with use, rendering the cavity unusable after several years. This
problem could be rectihed by reinforcing cavity entrances with wire mesh
or thin wood following the woodpecker’s completion of the cavity cham-
ber.

Still, artificial substrates for woodpecker cavity excavation may have
value. Substrates with a stronger yet brittle structure may be needed to
entice other woodpecker species to excavate cavities and Downy Wood-
peckers to nest. Also, additional structure strength or hardness is needed
on the surface of the artificial snags. Such strength might help deter pred-
ators and provide sufficient hardness and resonance for mutual tapping
behavior and drumming which occur during cavity site selection (Kilham
1958, 1983). Also, further study using larger diameter and taller artificial
snags in areas where natural snags are limited or absent may provide
additional insight.

ACKNOWLEDGMENTS

We thank T. C. Grubb, Jr., D. R. Petit, and J. R. Walters for constructive comments on
an early draft of the manuscript.

LITERATURE CITED

Conner, R. N. 1978. Snag management for cavity nesting birds. Pp. 120-138 in Proc. of
the workshop; management of southern forests for nongame birds (R. M. DeGraaf, tech,
coord.). U.S. Eor. Serv. Gen. Tech. Rep. SE-14.

, R. G. Hooper, H. S. Crawford, and H. S. Mosby. 1975. Woodpecker nesting

habitat in cut and uncut woodlands in Virginia. J. Wildl. Manage. 39:144-150.

, O. K. Miller, Jr., and C. S. Adkisson. 1976. Woodpecker dependence on trees
infected by fungal heart rots. Wilson Bull. 88:575-581.

, S. D. Jones, and G. D. Jones. 1994. Snag condition and woodpecker foraging
ecology in a bottomland hardwood forest. Wilson Bull. 106:242-257.

Dickson, J. G., R. N. Conner, and J. H. Williamson. 1983. Snag retention increases bird
use of a clear-cut. J. Wildl. Manage. 47:799-804.

Evans, K. E. and R. N. Conner. 1979. Snag management. Pp. 214-224 in Management
of north central and northeastern forests for nongame birds (R. M. DeGraaf, tech,
coord.). U.S. Eor. Serv. Gen. Tech. Rep. NC-5L

Grubb, T. C., Jr. 1982. Downy Woodpecker sexes select different cavity sites: an experi-
ment using artificial snags. Wilson Bull. 94:577-579.

Jackson, J. A. 1976. A comparison of some aspects of the breeding ecology of Red-headed
and Red-bellied woodpeckers in Kansas. Condor 78:67-76.

Kilham, L. 1958. Pair formation, mutual tapping and nest hole selection of Red-bellied
Woodpeckers. Auk 75:318-329.

. 1983. Life history studies of woodpeckers of eastern North America. Nuttall Or-

nithol. Club # 20.

Peterson, A. W. and T. C. Grubb, Jr. 1983. Artificial trees as a cavity substrate for
woodpeckers. J. Wildl. Manage. 47:790-798.

Petit, D. R., K. E. Petit, T. C. Grubb, Jr., and L. J. Reichhardt. 1985. Habitat and snag

456

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

selection by woodpeckers in a clear-cut: an analysis using artificial snags. Wilson Bull.
97:525-533.

Raphael, M. G. and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra
Nevada. Wildl. Monogr. No. 86.

Thomas, J. W., R. G. Anderson, C. Maser, and E. L. Bull. 1979. Snags. Pp. 60-77 in
Wildlife habitats in managed forests, the Blue Mountains of Oregon and Washington
(J. W. Thomas, tech. ed.). U.S. For. Serv. Agric. Handbk. No. 553.

Wilson Bull., 108(3), 1996, pp. 457-466

NESTING SUCCESS OF THE PROTHONOTARY
WARBLER IN THE UPPER MISSISSIPPI
RIVER BOTTOMLANDS

David J. Flaspohler

Abstract. — In 1993 and 1994, I studied the breeding biology and nesting success of
Prothonotary Warblers {Protonotaria citrea) at the margin of the species’ breeding range
on the upper Mississippi and Black rivers in west-central Wisconsin. During the severe
flooding of 1993, nesting success was reduced to a third of the level recorded in 1994, a
more typical year. The rate of Brown-headed Cowbird {Molothrus ater) parasitism was the
highest (26.9%) yet reported. House Wrens (Troglodytes aedon) were observed destroying
only one nest, but they were suspected of having a larger role in nest failure as has been
found in other studies (Walkinshaw 1938). Received 30 Mar. 1995, accepted 21 Sept. 1995.

The Prothonotary Warbler {Protonotaria citrea) is a secondary cavity
nester that breeds in floodplain forests of the eastern U.S. Between 1966
and 1987, it experienced regional population declines in the southern U.S.
(James et al. 1991) and in the northern Midwest (Graber et al. 1983). It
is listed as one of ten area-sensitive warbler species (subfamily Parulinae)
(Robbins 1979). Much of its floodplain forest habitat has been lost or
degraded since presettlement times (Fredrickson 1979), and mangrove and
riparian forests of Latin America used by the Prothonotary during the
non-breeding season (Skutch 1989) are being rapidly destroyed or con-
verted to other uses (Terborgh 1 989). In the center of its breeding range
in the southern U.S., less than 25% of the original bottomland forest
remains (Fredrickson 1979, Harris et al. 1984). In Wisconsin, only 8% of
presettlement floodplain forest remains in moderate to high quality con-
dition (Mossman 1988).

Population monitoring of the Prothonotary Warbler across its breeding
range is hampered by the inaccessibility of bottomland forests. In Wis-
consin, where this study was conducted, there has been only one occur-
rence of a Prothonotary Warbler on all Breeding Bird Survey (BBS)
routes from 1966 to 1991 (USFWS, unpubl. data), even though the spe-
cies breeds commonly in suitable habitat (Mossman 1988). Furthermore,
brood parasitism by the Brown-headed Cowbird {Molothrus ater) may be
contributing to population declines by reducing productivity.

Several studies have examined the nesting ecology of the Prothonotary
Warbler (Walkinshaw 1938, 1939, 1941, 1953; Petit 1986, 1989; Blem
and Blem 1991, 1992). However, the majority of nests in these studies
were built in artificial nest boxes. Use of artificial nest boxes may affect

Dept, of Wildlife Ecology, A229 Rus.sell Labs, Univ. of Wisconsin-Madison, Madison, Wisconsin 53706.

457

458

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

breeding parameters such as clutch size and nesting success (Mertens
1977), although preliminary studies do not support the hypothesis that
nest box size affects clutch size for this species (C. Blem, pers. comm.).
Hole nesting birds using nest boxes may also suffer artificially reduced
rates of predation (Nilsson 1984, 1986, Moller 1989) as compared to nests
in natural cavities. Conversely, the greater conspicuousness of nest boxes
may increase predation rates compared to natural cavities. The diameter
of a nest box entrance may also discourage or prevent cowbird parasitism.
Few data exist on nesting success and brood parasitism rates for naturally
occurring nests. The reproductive ecology of the Prothonotary Warbler
has not been studied in detail in the upper Mississippi River region. I
present here nest site characteristics, reproductive success, and rate of
cowbird parasitism for Prothonotary Warblers nesting in the upper Mis-
sissippi River. This study also provides some insight into the effect of
extreme flooding on the reproductive success of Prothonotary Warblers
in this region.

STUDY AREA AND METHODS

I collected nesting data at three sites along a 1 13-km section of the Mississippi River in
west central Wisconsin during 1993 and 1994 (Eig. 1) (pools 5—9, elevations 664—625' asl,
44°09'N, 91°48'W, 43°31'N, 91°14'W). In this area, the river ranges from L3-4.0 km wide
with numerous forested islands 0.25 to 300 ha in size. The river is bordered by steep bluffs
dominated by oak forest (Qiiercm spp.) with patches of remnant prairie on steep south-
facing slopes. Beyond the bluffs lie broad areas of agricultural land with scattered woodlots
where there was once hardwood forest, savanna, and open prairie (Emlen et al. 1986).
Riparian habitat occurs on islands and in strips 0.1 to 1.5 km wide on either shore of the
river. Eastern cottonwood {Populus deltoides) and black willow {Salix nigra) are found on
new alluvial deposits (Olsen and Meyer 1976). Older alluvial sites and mesic areas are
dominated by silver maple (Acer saccharinum), green ash {Fraxinus pennsylvanica), river
birch (Betula nigra), box elder (Acer negundo), and basswood (Tilia americana). American
elm (Ulmus americana), once a dominant canopy species, is now represented only by sap-
lings and young trees, larger trees having succumbed to Dutch Elm disease. Dominant
understory plants include woodbine (Parthenocissus inserta), wood-nettle (Laportea cana-
densis), jewelweed (Impatiens capensis), violet (Viola spp.), poison ivy (Rhus radicans),
button bush (Cephalanthus occidentalis), and grape (Vitis spp.) (Olsen and Meyer 1976).

Information on nesting Prothonotary Warblers was also collected at two sites along the
lower portion of the Black River, two and seven km above its confluence with the Missis-
sippi River. Floodplain vegetation along the lower Black River is similar to that on the
Mississippi River (Barnes 1991). From mid-June to the end ot July 1993, both the Missis-
sippi River and the Black River experienced record-breaking floods which were directly
responsible for numerous nest failures.

From mid-May to the end of July 1993 and 1994, I found nests by walking or canoeing
through promising habitat and by following singing males. I recorded the location and stage
of nesting along with characteristics of the nest and site. I returned to check each nest
approximately every four days. I calculated nest success according to procedures in Mayfield
(1961, 1975) and Caccamise (1977). Nest height was the distance from the ground to' the

Flaspohler • NESTING OF PROTHONOTARY WARBLERS

459

bottom of the cavity opening, and only data from nests that had solid ground under them
for some part of the nesting period were used (Table 1 ). Ratios of the number of young
hatched to the number of eggs laid (H/E) and the number of chicks fledged to the number
of young hatched (F/H) were used as indices of breeding success (Caccamise 1977). The
Mayfield (1961, 1975) method for calculating nest success adjusts for the stage at which a
nest is first discovered.

460

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Nest Parameters of Prothonotary Warblers in the Upper Mississippi River

Bottomlands

Parameter

1993

1994

No. nests“

22

20

No. eggs

90

73

No. hatched

41

23

Hatched/egg

0.46

0.32

No. fledged

25

22

Pledged/hatched

0.61

0.96

No. successful nests

10

6

Percent successfuP

45

30

Mayfield estimate*^^

0.20

0.66

’ Nests include only those found during egg incubation stage.

Success = hedged at least one young.

'Success calculated using Mayfield's (1961) correction for exposure. Estimate includes all nests (1993: N = 28; 1994:
N = 32) found during incubation and nestling stages and represents probability of nest surviving through both stages.

RESULTS AND DISCUSSION

The three greatest sources of mortality for eggs and nestlings during
this study were flooding, predation, and destruction by House Wrens
{Troglodytes aedon). These were also the principal sources of nest mor-
tality in Michigan (Walkinshaw 1938, 1953). Flooding was devastating
in 1993 since the “hundred year” floods coincided precisely with peak
nesting activity in mid-June. In 1993, 36% of all nests were flooded, while
none was lost to flooding in 1994. No nests were abandoned during this
study. Increased predation rates associated with observer nest visits are
unlikely, since most nests were located over water, nest visits were brief,
and a variety of non-terminal routes were used when visiting a nest.

Of 43 nests for which monitoring began during the incubation period,
only one was observed being destroyed by a House Wren. In this case,
the House Wren punctured all four eggs in an unattended nest and
dropped one into the water below the nest. It is possible, based on their
abundance and aggressive habits, that House Wrens were responsible for
other losses attributed to predation.

Mean clutch size (4.31, Table 2) was smaller than that reported by Petit
(1989, X = 4.75, N = 120), Walkinshaw (1941, x = 5.62, N = 1 18), or
Blem and Blem (1992, x = 4.38, N = 266). Mean tree stub diameter at
nest height in this study was greater (29.9 cm) than in Petit’s (1987) study
(13.6 cm). Prothonotary Warblers use cavities excavated by other birds
as well as naturally occurring cavities. Many nests were found in cavities

Flaspohler • NESTING OF PROTHONOTARY WARBLERS

461

Table 2

Characteristics of Prothonotary Warbler Nests

Characteristic

N

X ± SD

Mean clutch size"' (eggs)

36

4.31 ± 0.79

Nest diameter opening (cm)

Least

2.5

Greatest

9.8

Mean

74

5.0 ± 1.35

Mean nest height above ground'’ (cm)

43

219.4 ± 124.3

Mean stub diameter at nest height (cm)

76

29.9 ± 13.7

“ Nests included are only those found during incubation.

Nests from 1993 were not included because all nests were found over highly fluctuating water levels.

that had been expanded through decomposition, and these accounted for
the larger diameter openings.

Prothonotary Warblers glean arthropods from the ground and shrub
layer of riparian forests. While the Prothonotary Warbler is not an obligate
ground forager, it does use the shrub layer extensively when foraging,
and the absence of this layer during much of the 1993 breeding season
may have influenced foraging efficiency.

The record-breaking floods along the Mississippi and its tributaries in
1993 were largely responsible for the lower nest success in 1993 com-
pared with 1994. Peak flood levels occurred precisely during the height
of breeding activity for Prothonotary Warblers in June. The Mississippi
River near Merrick State Park, Wisconsin (Fig. 1, Sites #1, #2) rose nearly
3 m between June 16 and June 26 (U.S. Army Corps of Engineers 1993).
Although no data were available on water levels at study sites on the
Black River, I noted similarly dramatic rates of rise. The Prothonotary
Warbler typically nests within 2-3 m of the water’s surface (Flarrison
1975).

Pieman et al. (1993) found that nest predation near marshes decreased
with increasing water depth. Unusual flood waters may have made some
nests less accessible to predators, thus decreasing nest losses from pre-
dation in 1993. The percentage of nests depredated in this study (27.6%,
N = 28 from 1993 only) was lower than in Walkinshaw’s (1941) study
(41%, N = 27) but higher than in Petit’s (1989) study (20.9%, N = 191).
If nests lost to flooding are removed from the pool of nests available to
predators, a predation frequency similar to Walkinshaw’s (1941) is gen-
erated (44.4%, N = 18 from 1993 only). No attempt was made to distin-
guish predation losses from other nest losses in 1994. Most of Petit’s
nests were in artificial nest boxes with entrance holes smaller than the

462

THE WILSON BULLETIN • Vol. 108. No. 3, September 1996

mean of the entrance holes in this study, which may have influenced
predation rates in that study. Such nest box-specific effects on nest success
will depend on the material used to construct the box (e.g., cardboard vs
wood). House Wrens were absent from Petit’s (1989) and Walkinshaw s
(1941) Tennessee sites while they were common on Walkinshaw’s (1941)
Michigan sites and in this study. The presence of House Wrens may
explain the similar predation rates for non-flooded nests in Michigan and
Wisconsin, while the absence of House Wrens on the Tennessee sites may
account for the lower predation rates reported by Petit (1989).

Potential nest predators observed in this study included Common
Crackle {Quiscalus quiscula). Blue Jay (Cyanocitta cristata). House
Wren, Common Crow (Corvus brachyrhynchos), gray squirrel {Sciurus
carolinensis), and mink {Mustela vison). Species known to prey on Pro-
thonotary Warbler nests include the gray squirrel (Walkinshaw 1938) and
mice of the genus Peroinyscus (Guillory 1987). Other likely predators in
the study area include the raccoon {Procyon lotor), striped skunk (Me-
phitis mephitis), fox squirrel (Sciurus niger), and opossum (Didelphus
virginianus). Although no snakes were seen in 1993, they were seen seven
times in 1994 and have been reported as predators in other studies of
Prothonotary Warblers (Petit 1989). Excluding the single nest destroyed
by a House Wren, 1 was able to identify the predator of a nest by teeth
marks in only three cases. Squirrels (Sciurus spp.) gnawed through the
side of a total of three cavities in both live and dead trees. One such nest
contained three Prothonotary Warbler eggs and five cowbird eggs and
was located on a small island (<1 ha) isolated by approximately 200 m
of swift and deep (>5m) floodwaters and had no dry land on it. Clearly,
islands do not provide complete safety from tree-climbing terrestrial pred-
ators.

Nests were often placed in severely rotted trees in relatively exposed
areas over water where they are vulnerable to damage from storms and
wave action from boats. Of 76 nests in snags and stubs, none was lost
due to the collapse of the tree, although two nest trees collapsed within

one week after the warbler’s fledging.

The incidence of cowbird parasitism was the highest yet reported (Table
3). A comparison of regional cowbird populations indicates that Midwest
cowbird abundance is 2.5 times greater than in the eastern U.S. and is
increasing (Robbins et al. 1986). The bottomland forests of the upper
Mississippi are in agricultural lands that provide foraging habitat for cow-
birds. Petit’s (1989) Tennessee study was conducted in a riparian zone
within a mostly forested landscape. These different land-use patterns may
partly explain the higher parasitism rates in this study. Since Walkin-
shaw’s (1938, 1941) studies in Michigan, cowbird populations have in-

Flaspohler • NESTING OF PROTHONOTARY WARBLERS

463

Table 3

Rates of Cowbird Parasitism of Prothonotary Warblers in the United States

Location

No. nests

% Parasitism

Reference

Iowa

70

25.7

Norris (1890)

Michigan

28

10.7

Walkinshaw (1938)

Louisiana

57

12.3

Goertz ( 1977)

Illinois

154

15.6

Graber et al. (1983)

Tennessee

128

20.3

Petit (1989)

Wisconsin

67

26.9

This study (1996)

Virginia

998

0.013

Blem, unpubl. data

creased across the eastern U.S. This population increase could be partly
responsible for increased parasitism rates since Walkinshaw’s time.

Belles-Isles and Pieman (1986) noted that House Wrens poke holes in
eggs of other species within their territories, often removing the pecked
eggs and disturbing the nest lining. Although I observed House Wrens
destroying only one nest, Walkinshaw (1941) reported that 25% of 413
Michigan Prothonotary eggs and chicks were destroyed by House Wrens.
Walkinshaw spent more time observing nesting behavior than I did during
this study, giving him more opportunities to identify the cause of egg and
nestling loss.

It has long been assumed that cavity-nesting birds are limited primarily
by the availability of nest sites (Hilden 1965, Scott 1979, Mannan et al.
1980) and that House Wrens benefit from nest-destroying behavior by
freeing up nest sites and perhaps decreasing foraging competition. Several
studies have observed that Prothonotary Warblers often compete unsuc-
cessfully for cavities with House Wrens (Smith and Dumont 1944, Graber
et al. 1983). In Walkinshaw’s (1941) comparative study of Prothonotary
Warblers nesting in Michigan and Tennessee, he attributed comparatively
lower nesting success in Michigan to competition from and nest destruc-
tion by House Wrens, a species not common on his Tennessee sites.

In a trial nest-box study conducted in 1994, we placed 20 wooden nest
boxes within past Prothonotary Warbler breeding habitat. House Wrens
occupied 16 (80%) of the nest boxes, and Tree Swallows (Tachycinefa
bicolor) nested in three (15%). No Prothonotary Warblers nested in the
boxes.

House Wrens were the most abundant bird species in the study sites,
as measured by point counts conducted during the study (Flaspohler
1994), and are common and widespread in floodplain forests throughout
Wisconsin (Mossman 1988). I found numerous nests with missing eggs

464

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

and with both disturbed and undisturbed nest linings. Because of the
House Wren’s habit of removing nesting material from the nests that it
destroys (Belles-Isles and Pieman 1986), and thereby disturbing the nest,
one cannot confidently conclude that a disturbed nest implies a mam-
malian predator as proposed by Best (1978) and Petit (1989). Where no
cowbird parasitism was present, I attributed the disappearance of eggs
and nestlings to predation. Where House Wrens are abundant, this method
may tend to overestimate predation rates and underestimate House Wren
nest destruction rates.

ACKNOWLEDGMENTS

I thank K. Goffin, S. Matteson, D. Hinebaugh, and V. Patton, for field assistance, and S.
Temple, P. Arcese, M. Mossman, and D. Sample for helpful comments. Constructive reviews
were provided by C. Blem, L. Petit, and an anonymous reviewer. I am grateful for a variety
of logistical and material assistance provided by G. Evland of the Wisconsin Dept, of Natural
Resources and E. Nelson and S. Lewis of the U.S. Fish and Wildlife Service. C. VanderVeen
provided essential research support during 1993. This research was supported by the Wis-
consin Endangered Resources Cahill Gift Fund, the U.S. Fish and Wildlife Service, and the
Max McGraw Wildlife Foundation.

LITERATURE CITED

Barnes, B. 1991. Tree populations on the islands of the lower Chippewa River in Wiscon-
sin Bull, of the Torrey Bot. Club. 118:424-431.

Belles-Isles, J. C. and J. Picman. 1986. House Wren nest-destroying behavior. Condor

88' 190—193.

Best, L. B. 1978. Field Sparrow reproductive success and nesting ecology. Auk 95:9-22.
Blem, C. R. and L. B. Blem. 1991. Nest box selection by Prothonotary Warblers. J. Field

Ornithol. 62:299—307. , u • a

1992. Prothonotary Warblers nesting in nest boxes: clutch size and

timing in Virginia. Raven 63:15-20. j • j

Caccamise, D. F. 1977. Breeding success and nest site characteristics of the Red-winged

Blackbird. Wilson Bull. 89:396-403.

Emlen, j. T, M. j. DeJong, M. J. Jaeger, T. C. Moermond, K. A. Rusterholtz, and R. P.
White. 1986. Density trends and range boundary constraints of forest birds along a
latitudinal gradient. Auk 103:791-803.

Flaspohler D j 1994 Report on point-counts of breeding birds in the upper Mississippi

River during the spring and summer of 1993. Wisconsin Dept, of Natural Resources.

Madison, Wisconsin. . ^ r u-

Fredrickson L H 1979. Floral and faunal changes in lowland hardwood forests resulting

from channelization, drainage, and impoundment. U.S. Dept. Interior, FWS/OBS-78/9L
Goertz, j. W. 1977. Additional records of Brown-headed Cowbird nest parasitism in Lou-
isiana. Auk:386-389.

Graber, j. W., R. R. Graber, and E. L. Kirk. 1983. Illinois birds: wood warblers. Illinois

Natural History Survey, Biol. Note 118. _

Guillory, H. D. 1987. Cavity competition and suspected predation on prothonotary war-
blers by Peromyscus spp. J. Field Ornithol. 58:425-427.

Harris, L. D., R. Sullivan, and L. Badger. 1984. Bottomland hardwoods: valuable, van-

Flaspohler • NESTING OF PROTHONOTARY WARBLERS

465

ishing, vulnerable. Univ. of Florida School of Forest Resources and Conservation tech-
nical report. Tallahassee, Florida.

Harrison, H. H. 1975. A Held guide to bird’s nests. Houghton Mifflin, Boston, Massachu-
setts.

Hilden, O. 1965. Habitat selection in birds. Ann. Zool. Fenn. 2:53-75.

James, F. C., D. A. Wiedenfeld, and C. E. McCulloch. 1991. Trends in breeding popu-
lations of warblers: declines in the southern highlands and increases in the lowlands.
Pp. 43-56 in Ecology and conservation of neotropical migrant landbirds, (J. M. Hagan
III and D. W. Johnston, eds.). Smithsonian Inst. Press, Washington D.C.

Mannan, R. W, E. C. Meslow, and H. M. Wright. 1980. Use of snags by birds in
Douglas-fir forest. J. Wildl. Manage. 44:787-797.

Mayfield, H. F. 1961. Nesting success calculated from exposure. Wilson Bull 73-255-
261.

. 1975. Suggestions for calculating nest success. Wilson Bull. 87:456-466.

Mertens, j. a. L. 1977. Thermal conditions for successful breeding in Great Tits. II.
Thermal properties of nests and nest-boxes and their implications for the range of
temperature tolerance of Great Tit broods. Oeologia 28:31-56.

Moller, a. P. 1989. Parasites, predators and nest boxes: facts and artifacts in nest box
studies of birds? Oikos 56:421-423.

Mossman, M. j. 1988. Birds of southern Wisconsin floodplain forests. Passenger Pigeon
50:321-337.

Nilsson, S. G. 1984. The evolution of nest-site selection among hole-nesting birds: the
importance of nest predation and competition. Ornis Scand. 15:167-175.

. 1986. Evolution of hole nesting in birds: on balancing selection pressures Auk
103:432-435.

Norris, J. P. 1890. A series of eggs of the Prothonotary Warbler. Ornithologist and Oologist
Vol. 15, No. 12:177-182.

Olsen, K. N. and M. P. Meyer. 1976. Vegetation, land, and water surface changes in the
upper navigable portion of the Mississippi River basin over the period 1939-1973.
Univ. of Minnesota, Inst, of Ag., Forestry, and Home Econ., Remote Sensing Lab.
Research Rep. 76-4. St. Paul, Minnesota.

Petit, L. J. 1986. Factors affecting the reproductive success of Prothonotary Warblers
(Protonotaria citrea) nesting in riverine habitat. M.S. thesis. Bowling Green State Univ.

. 1987. Nest-box use by Prothonotary Warblers (Protonotaria citrea) in riverine
habitat. Wilson Bull. 99:485-488.

. 1989. Breeding biology of Prothonotary Warblers (Protonotaria citrea) in riverine
habitat in Tennessee. Wilson Bull., 101:51-61.

PiCMAN, J., M. L. Milks, and M. Leptich. 1993. Patterns of predation on passerine nests
in marshes: effects of water depth and distance from edge. Auk 1 10:89-94.

Robbins, C. S. 1979. Effect of forest fragmentation on bird populations. Pp. 198-212 in
Management of North Central and Northeastern forests for nongame birds (R. M.
DeGraaf and K. E. Evans, eds.) Gen. Tech. Rep. NC-51, U.S. Forest Service, St. Paul,
Minnesota.

, D. Bystrak, and P. H. Geissler. 1986. The breeding bird survey; its first fifteen
years, 1965-1979. U.S. Fish and Wildlife Service Resource Publ. 157.

Scott, V. E. 1979. Bird responses to snag removal in ponderosa pine. J Forestry 77-26-
28.

Skutch, a. E 1989. Birds of Co.sta Rica. Cornell Univ. Pre.ss, Ithaca, New York.

Smith, E. T. and P. A. Dumont. 1944. Middle-western region. Audubon Mag. 47:104-106.
(Supplement)

466

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Terborgh, J. 1989. Where have all the birds gone? Princeton Univ. Press, Princeton, New
Jersey.

United States Army Corps of Engineers. 1993. Mississippi River data. St. Paul, Min-

nesota. . . n j- n

Walkinshaw, L. H. 1938. Nesting studies of the Prothonotary Warbler, Bird-Bandmg 9:

1939. Additional information on the Prothonotary Warbler. Jack-Pine Warbler 17.

64-71.

1941. The Prothonotary Warbler, a comparison of nesting conditions in Tennessee

and Michigan. Wilson Bull. 53:3—21.

1953. Life-history of the Prothonotary Warbler. Wilson Bull. 65:152-168.

Wilson Bull., 108(3), 1996, pp. 467-479

FACTORS AFFECTING FOOD PROVISIONING OF
NESTLING BLACK-THROATED BLUE WARBLERS

Catherine O’Neill Goodbred' - and Richard T. Holmes’

Abstract. — Using video cameras at nests, we measured rates, quantities, and types of
food delivered by male and female Black-throated Blue Warblers (Dendroica caerulescens)
to nestlings of different ages and at different times of day and nesting season. Based on 89
1.5—2 h observation periods at 18 nests, all of which contained four young, we found that
larval Lepidoptera comprised 60—87% of the estimated prey biomass brought to nestlings
and that the female and male parents delivered approximately equal amounts of food over
the nesting cycle. Food provisioning rates did not vary with time of day or with parental
age, but did increase significantly with age of nestlings and decrease with time of season
(early vs mid-summer). The lower rate and quantity of provisioning in mid-summer was
reflected in significantly slower growth of nestlings in that part of the season, suggesting
constraints on parental food provisioning, perhaps due to lower food availability. Received
27 Aug. 1995, accepted 2 Feb. 1996.

Successful production of offspring is an essential component of indi-
vidual fitness (Stearns 1992), as well as being crucial to the maintenance
of population levels for most, if not all, bird species (Nolan 1978, Viro-
lainen 1984, Sherry and Holmes 1992, Holmes, et al. 1992. Robinson et
al. 1995). Although predation at the nest is probably the single most
important factor affecting breeding success of most passerine birds (Rick-
lefs 1969, Holmes et al. 1992, Martin 1992a), parental care of nestlings
can also be important (Kuitunen and Suhonen 1991). Parental care, which
includes nest building, incubation, food provisioning, vigilance, and
brooding, is not only energy-demanding but also potentially risky to the
survival of the parents and their lifetime reproductive success (Curio
1988). Quantities of food delivered can influence nestling survival as
evidenced by brood reduction due to starvation in many passerine birds
(Magrath 1990) and in some cases by starvation of whole broods (Ro-
denhouse and Holmes 1992, Sherry and Holmes 1992). Also, nestlings
that are not well fed may beg more, which could result in the attraction
of predators, leading to higher nest losses (Skutch 1949, Martin 1992b).
The patterns of food delivery to nestlings and the factors that affect them
are, therefore, important for understanding population processes. Infor-
mation about which factors influence food delivery patterns is also useful
for designing sampling protocols for future studies of reproductive biol-
ogy of particular species.

In this study, we examined the rates and quantities of food delivered

' Dept, of Biological Sciences. Dartmouth College, Hanover, New Hampshire 03755.
^Present address: Norfolk Academy, 1585 Wesleyan Drive, Norfolk, Virginia 23502.

467

468

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

to nestlings by male and female Black-throated Blue Warblers (Dendroica
caerulescens) breeding in a northern hardwoods forest in north-central
New Hampshire. Specifically, we tested whether or not feeding rates and
quantities of food delivered by males and females varied with nestling
age, time of season, time of day, and/or parental age. We also considered
the effect of parental provisioning by measuring the growth rates of young
in the nests being videotaped.

STUDY SITE AND METHODS

This study was conducted in the Hubbard Brook Experimental Forest, West Thornton,
New Hampshire. This forest consists of northern hardwoods, dominated by sugar maple
{Acer saccharum), American beech {Fagus grandifolia), and yellow birch {Betula alle-
ghaniensis), with a shrub layer of hobblebush {Viburnum alnifolium), saplings of sugar
maple and especially beech, and striped maple {A. pensylvanicum). The ground stratum
consists mostly of herbs, tree seedlings, mosses, and ferns (Bormann and Likens 1979,
Holmes 1990).

At Hubbard Brook, Black-throated Blue Warblers nest at an average height of 0.6 m
above ground, mostly in hobblebush but also in other shrub-level vegetation (Holmes 1994).
We located most nests during the building or incubation stages and then checked them
periodically until hatching. Nestlings were weighed with a 10-g Pesola balance on days 2,
4, and 6 (hatching = day 0), the last day they could be handled without causing premature
fledging (Holmes et al. 1992). Parents at each nest were color banded, and aged as yearlings
(i.e., in their first potential breeding season after hatching) or as older individuals (i.e., m
their second or later breeding season), using plumage criteria (U.S. Fish and Wildlife Service
1977, Pyle et al. 1987). Identification of the sexes was unambiguous, due to strong sexual
dichromatism in this species.

To record parental feeding rates and foods brought to nests we used two Sony camcorders
and one Panasonic S-VHS Recorder (VCR) equipped with 8-lOX telephoto lens. Cameras
were set on tripods (0.5-1. 5 m high) at distances of 3-5 m from the nest, depending on
local topography and on density and arrangement of nearby foliage. The video cameras and
tripods were covered with black plastic for rain protection and then draped with burlap for
camouflage. If necessary, overhanging leaves around the nest were pulled aside and tied so
as not to obscure the nest during taping. Any tapes containing evidence that the adults were
disturbed were eliminated from the analyses. Taped sequences ranged from 1.5-2 h de-
pending on battery life; data were converted to number of visits, or quantities of food

delivered, per hour. ^ ,o di v

Between 12 June and 17 July 1991, we obtained 89 video samples at nests of 18 Black-

throated Blue Warbler pairs. We restricted the nests used for analysis to those containing
four young, the mean and modal brood size for Black-throated Blue Warblers at Hubbard
Brook (Holmes et al. 1992). Each nest was attended by one male and one female, and none
of the males was known to be polygynous (see Holmes et al. 1992, 1996). Visits to nests
were video-taped when nestlings were 2, 4, 6, and 8 days of age (hatch = day 0). In this
species, nestlings typically fledge late on day 8 or on day 9 (Holmes 1994). Samples were
obtained from the same nests on different days, but no nests were sampled more than once
per time-of-day/age-of-nestling category. Data from each recording session were categorized
according to age of nestlings (see above), time of the season (“early nests fledged young
before 1 July, “late” nests after 1 July, see Holmes et al. 1992 for nesting chronology), the
time of day when the nests were videotaped (grouped in three time intervals [EDT]; morn-

Goodbred and Holmes • FOOD PROVISIONING OF NESTLINGS

469

ing, 06:30-09:45 h, midday, 10:30-13:30, and afternoon, 14:00-17:00), and the age of
parents (yearling vs older, see above).

From each taped sample, we recorded the number of visits made by each parent and,
where possible, the number, size, and life-form (larval or adult, the latter including Arach-
nida) delivered to the young. Prey size was determined by comparing the length of prey
item (or the prey load when there was more than one item that could not be distinguished)
with the 7 mm-long exposed portion of the Black-throated Blue Warbler bill. These data
were grouped into four size classes (<7 mm, 7—14 mm, 14—21 mm, >21 mm), regardless
of taxa. They were then converted to biomass, based on length-mass regressions for mid-
points of each size class. Conversion factors for the four size classes were 1, 2, 8, and 20
mg, respectively, following the rationale and protocol of Omland and Sherry (1994: Table
1). Estimates of food biomass delivered to the nest per hour were then obtained by summing
the estimated prey biomass on each trip during the sample period, and expressing these as
mg of food delivered brood^' h '.

Data for both the number of feeding trips per hour and the food biomass delivered per
hour were normalized by square root transformation. The relationships between food pro-
visioning and each variable for both females and males were evaluated by analyses of
variance (ANOVA). Differences between the sexes and parental age classes, where appro-
priate, were examined with r-tests. Potential differences in nestling growth in early versus
late season nests was assessed by comparing ( 1 ) the mean body mass of nestlings on day
6 (mean mass per nestling in nests with 4 young, N = number of nests) and (2) the rate of
gain in body mass between day 2 and day 6 (mean mass per nest on day 6 minus that on
day 2, N = number of nests).

RESULTS AND DISCUSSION

The number of feeding trips and the biomass of food delivered to
broods per hour by adult Black-throated Blue Warblers varied signifi-
cantly with nestling age (except for feeding trips by females) and with
time of season (for both sexes), but not with time of day (Table 1 ) or
with parental age (see below). Two and three-way interactions among
these variables were not statistically significant (P values > 0.2). Removal
of time of day from the ANOVAS did not alter these results. As men-
tioned previously, brood size was not a factor in these analyses, because
only nests containing four young were considered.

Effects of nestling age. — The most important factor affecting both the
number of food delivery trips and amount of food delivered per nestling
was nestling age. Both sexes increased their feeding visitation rates from
day 2 through day 8 of the nestling phase (Fig. lA). This trend was not
statistically significant for females (Table 1), probably because they were
already feeding nestlings relatively frequently on day 2 and did not in-
crease their rate substantially between days 4 and 6. In contrast, males
progressively increased their feeding visits as the nestlings became older
(Fig. lA). This difference between the sexes was also evident in com-
parisons of food visitation rates on particular days; on average, females
made significantly more trips to the nest than did males on days 2 and 4
it = 4.01, df = 38, P = 0.000, and t = 3.26, df = 46, P = 0.002,

470

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

A. Feeding trips

B. Food biomass delivered 26

LiG. 1 Leeding rate (A) and estimated food biomass delivered (B) by female and male
Black-throated Blue Warblers to broods of four young on days 2, 4, 6, and 8 following
hatching (Means ± SE, N = number of 1.5-2 h video-taped samples).

Goodhred and Holmes • FOOD PROVISIONING OF NESTLINGS

471

Table 1

Results of Analyses of Variance (ANOVA) Tests of Food Provisioning by Male and
Female Black-throated Blue Warblers at Nests with Four Young“

Females

Males

SS"

df F

P

SS"

df F

P

Number of food deliveries

ModeF

12.50

6

1.74

0.121

41.24

6

15.90

0.000

Age of nestlings

1.25

3

0.35

0.790

38.76

3

29.89

0.000

Time of day

0.95

2

0.40

0.673

0.01

2

0.01

0.988

Time of season

5.18

1

4.34

0.040

2.13

1

4.92

0.029

Residual

1 10.42

88

76.69

88

Food biomass delivered

Model

120.47

6

11.06

0.000

119.47

6

6.93

0.000

Age of nestlings

50.71

3

9.31

0.000

53.15

3

6.16

0.001

Time of day

4.65

2

1.28

0.284

0.69

2

0.12

0.887

Time of season

9.84

1

5.42

0.023

14.24

1

4.95

0.029

Residual

256.62

81

346.52

85

“ Analyses were performed separately on the number of food delivery trips h'' and on the estimated food biomass (mg)
deliveredh"'. ®

^ Sum of squares.

‘Explained variance for full models: female number. E = 0.11; female biomass, E = 0.47, male number E = 0 54-
male biomass H = 0.34.

respectively), but not on days 6 and 8 (r = -0.61, df = 36, P = 0.546,
and t = 0.08, df = 50, P = 0.938, respectively).

The quantity of food delivered to the brood increased significantly with
nestling age for both sexes (Table 1, Fig. IB). There were no significant
differences between the sexes, however, in the quantity of food delivered
per hour at any of the four nestling ages days sampled (r tests, P values
> 0.49). Because males made fewer trips to nests on days 2 and 4 but
were contributing about equal biomass (Fig. IB), they must have been
bringing larger and/or more prey per trip (see below).

The positive relationship between nestling age and feeding/food bio-
mass delivery rate was related, as expected, to the increasing energy de-
mands of the young, either for growth or thermoregulation. This same
pattern has been reported for most species where it has been studied (e.g.,
Morehouse and Brewer 1968, Nolan 1978, Johnson and Best 1982, Bier-
mann and Sealy 1982, Bedard and Meunier 1983, Breitwisch et al. 1986,
Haggerty 1992). The only apparent exception is the Nashville Warbler
(Vermivora ruficapilla) in which both males and females were reported
to feed nestlings at an essentially constant rate (Knapton 1984). This
study, however, examined feeding rates only on days 4 to 8 of the nestling
period.

472

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Differences between male and female passerines in the provisioning of
their nestlings have been noted frequently, but there appears to be no
consistent pattern. In Prairie Warblers (D. discolor) and Eastern Phoebes
(Sayornis phoebe), both sexes fed about equally throughout the nestling
period (Nolan 1978 and Conrad and Robertson 1993, respectively). In
Yellow Warblers (D. petechia), males brought more food to the nest than
did females when nestlings were two days old, but by day 8, males and
females were delivering food at equal rates (Biermann and Sealy 1982).
Female and male Savannah Sparrows {Passerculus sandwichensis) made
approximately equal number of trips to the nest through the entire nestling
period, although this varied some with brood size (Bedard and Meunier
1983). In the latter study, however, when food biomass was considered,
males and females brought different quantities of food at different parts
of the nestling cycle. Similarly, when food biomass estimates were made
for Black-throated Blue Warblers in this study, we found that although
the sexes fed at different rates at different parts of the nesting cycle, they
delivered approximately equal quantities of food over the nestling period.

Effects of time of season.— Black-throaied Blue Warblers made signif-
icantly fewer feeding visits and delivered less food biomass per nest per
hour in mid-summer (July) than earlier in the season in June (Table 1,
Fig. 2). These lowered provisioning rates in the later part of the breeding
season were evident for both female and male parents. Furthermore, there
were no significant differences between the sexes in their rates of feeding
visits or in the quantities of food biomass delivered on any of the four
sample days in either the early or the late parts of the season {t tests, P
values > 0.13). The lower provisioning at nests in mid summer, therefore,
was not due to one sex being a poorer provider at that time.

Other studies have found either no change in food provisioning rates
during the course of a breeding season (Johnson and Best 1982), or if
changes did occur, they were compensated for by an increase in the quan-
tities of prey brought to the nest (Royama 1966). The fewer trips to the
nest/hour and lower food biomass delivered by Black-throated Blue War-
blers in this study could be due to several factors. First, nestlings in mid-
summer may have lower energy requirements because of lower thermo-
regulatory costs in the warmer temperatures, although this seems unlikely
in these relatively cool northern forests. Second, because the nests in mid-

LiG. 2. Leeding rate (A) and estimated food biomass delivered (B) by female and male
Black-throated Blue Warblers to broods of four young in June and July (Means ± SE, N
= number of 1.5-2 h video-taped samples).

Food biomass (mg) delivered per hour Number per hour

Goodbred and Holmes • FOOD PROVISIONING OF NESTLINGS

473

A. Feeding trips by season

Early summer

Mid summer

B. Food biomass delivered by season

474

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

summer were either replacements for ones lost earlier in the season or
were second broods (Holmes unpubl. data), perhaps they may be more
“expendable” if other energy demands for the parents become more im-
portant, e.g., the onset of the annual molt or fat deposition for migration.
Molt in this species, however, does not begin until early August and
departure on migration doesn’t occur until later in the season (Holmes
1994), so this explanation seems unlikely.

A third possible explanation for lower food provisioning of nestlings
in mid-summer is that parents had more difficulty in finding food at that
time of the season. If so, the reduced level of provisioning should result
in slower nestling growth. To test this, we compared growth rates of
nestlings in early and late season nests, as indicated by changes in body
mass of nestlings between day 2 and day 6 following hatching and body
mass on day 6. The mean (±SE) change in body mass of nestlings m
early season nests (5.08 ± 0.16 g, N = 8 nests) was significantlyjiigher
than that for nestlings in late season nests (4.32 ± 0.21, N = 7; r - 2.94,
df = 13, P = 0.012). Similarly, nestlings on day 6 in the early part of
the summer were heavier (7.71 ± 0.02 g, N = 10 nests) than those in
mid-summer (7.03 ± 0.30, N = 8), although these differences were only
marginally significant (r = 1.93, df = 16, P = 0.07). Thus, Black-throated
Blue Warbler nestlings in late season nests grew more slowly than did
their early season counterparts, which could ultimately have a major im-
pact on their post-fledging survival as has been shown by Perrins (1980)
for Great Tits {Pams major). This finding suggests that food provisioning
by Black-throated Blue Warbler parents is not always sufficient to max-
imize growth of their nestlings and is consistent with the proposition that
food is limiting for Black- throated Blue Warblers, at least in some years
or seasons (Holmes et al. 1991, Rodenhouse and Holmes 1992).

Effects of time of day. — Food provisioning by Black-throated Blue
Warblers did not differ significantly with time of day, either for feeding
visitation rate or food biomass delivered (Table 1). Also, there were no
significant differences between the sexes in either feeding visits or food
delivered at different times of day (t tests, P values > 0.22).

Nolan (1978) found little variation in foraging rate during the day for
Prairie Warblers, except for an increase in the morning shortly after dawn
(06:00-07:00) and a smaller increase late in the day (19:00-20:00). Sim-
ilarly, Haggerty found no significant diurnal variation in provisioning
rates of Bachman’s Sparrow {Aimophila aestivalis,). More pronounced
diurnal patterns have been reported for other species, e.g.. Eastern King-
birds (Tyranniis tyrannus, Morehouse and Brewer 1968) and Nashville
Warblers {Vermivora ruficapilla, Knapton 1984), in which feeding rates
decrease in the middle of the day and increase again later. These are

Goodbred and Holmes • FOOD PROVISIONING OF NESTLINGS

475

mainly species of more open habitats in which mid-day temperatures may
depress insect activity or increase physiological stress on foraging adults.
In the present study. Black-throated Blue Warblers nested and foraged
largely in the cool, well-shaded understory of a closed-canopy forest,
where diurnal variation in summer temperature and insect abundance,
especially of Lepidoptera larvae, is not pronounced (R. T. Holmes, un-
publ.).

Ejfects of parental age. — Because of relatively small sample sizes for
each parental age class within a sex and the lack of independence arising
from multiple samples from individual nests, we were unable to include
parental age in the ANOVAs. However, we could test for parental age
effects within a restricted part of the nestling period. To do this, we
compared the feeding rates of yearling and older parents, both females
and males, at nests containing 4 young on day 6 of the nestling period,
the time when our sample was largest. The results of one-way ANOVAs
indicated no significant difference in feeding visitation rate or food bio-
mass delivered between yearling and older females (F, ,7 = 0.06, P =
0.81, and F, ,7 = 0.03, P = 0.881, respectively) or between yearling and
older males (F, ,7 = 0.09, P = 0.768, and F, = 1.51, P = 0.237). Thus,
at least on day 6 of the nestling period, there was no effect of age on
food provisioning of nestlings for either female or male Black-throated
Blue Warblers.

There is little comparative information on food provisioning rates
among parental age classes for passerines. Goossen and Sealy (1982)
suggested that older Yellow Warblers (D. petechia), because of their great-
er experience, should provide better care for their nestlings than first-time
breeding yearlings, but they did not provide supporting data. Studd and
Robertson (1989) found no difference in provisioning rate between age
classes in Yellow Warblers, nor did Omland and Sherry (1994) for male
American Redstarts Setophaga ruticilla. Current evidence thus suggests
that year-old parents do about as well as older, and presumably more
experienced, adults in providing food for their nestlings.

Foods delivered. — For purposes of analysis, food brought to nestlings
was divided into two broad categories recognizable in video images,
namely larval arthropods (almost entirely lepidopteran caterpillars) and
arthropod adults (e.g., crane flies (Tipulidae) and other Diptera, some
Hymenoptera and Lepidoptera, and occasional Coleoptera and Arachni-
da). On a biomass basis, larval insects comprised 60 to 87% of the food
brought to nestlings by both females and males (Table 2). Females deliv-
ered 43 to 50% of the food biomass to the brood over the course of the
nesting cycle. There were no significant differences between the sexes in
larval and adult biomass delivered per hour (/ tests, P values > 0.3),

476

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

Table 2

Estimates (x ± SE) of Food Biomass Delivered to Black-throated Blue Warbler
Nestlings on Days 2, 4, 6, and 8 Following Hatching

Food biomass (mg dry weight)

delivered brood" ' h" '
for nestling age (in days)

2

4

6

8

Food delivered by females

Larval insect biomass'"

Adult insect biomass^"

% larvae

(NF

4.9 ± 0.9
2.5 ± 0.6
63%

(20)

12.2 ± 1.7
3.2 ± 0.8
80%

(23)

14.9 ± 2.4
5.8 ± 1.6
78%

(19)

21.8 ± 3.9
14.0 ± 5.0
60%

(20)

Food delivered by males

Larval insect biomass'*

Adult insect biomass'’

% larvae
(N)

7.7 ± 1.7
1.1 ± 0.5

87%

(18)

12.5 ± 1.9
2.2 ± 0.7

84%

(24)

15.2 ± 2.3
4.9 ± 1.2
79%

(18)

27.4 ± 5.1

11.4 ± 2.5
73%
(26)

% Total food biomass delivered
by females

46%

50%

49%

43%

“ Mostly Lepidoptera larvae.

Flying insects, mostly Diptera, with occasional Hymenoptera and Coleoptera.

' Sample sizes represent number of video samples per nestling age class during which prey

could be classified.

although variances were high (Table 2). There was a trend for males on
day 2 to bring a greater biomass and percentage of larval prey than did
females (Table 2), which compensates for their fewer trips at that stage
(see below, Fig. 1 A). Other studies have noted that male paruline warblers
sometimes bring larger prey than do females, especially during the early
part of the nestling period (see Nolan 1978, Biermann and Sealy 1982,
Omland and Sherry 1994). In the present study, as a result of changing
food loads (and partly by changing prey types), male and female Black-
throated Blue Warblers contributed about equally in terms of prey biomass
delivered to their broods over the course of their nesting cycle.

In conclusion, food provisioning rates of both male and female Black-
throated Blue Warblers varied mostly with age of nestlings (especially for
males) and with time of season for both parents, with males and females
contributing about equally to the provisioning of their offspring. The low-
er rate of feeding visits and of food biomass delivered in mid-summer
was correlated with a decreased rate of nestling growth in those nests
compared to earlier in the season. Thus, food availability, at least in some
times and places, provides a constraint on nestling growth rates which,
in turn, might influence post-fledging survival. Food availability also af-
fects the frequency of double-brooding and, as a consequence, the annual

Goodbred and Holmes • FOOD PROVISIONING OF NESTLINGS

477

reproductive productivity of this species (Holmes et al. 1992). It is im-
portant to acknowledge that these results derive from studies conducted
in only one year and one place. Replication in other locations, and es-
pecially over more years, will be required to determine the generality of
these patterns. Also, because ot large variances and intercorrelations be-
tween variables, larger sample sizes would be helpful for detecting pat-
terns. Finally, some habitats or even territories within habitats occupied
by Black-throated Blue Warblers may be more productive than others,
affecting prey availability, and thus influencing food capture and delivery
rates of the foraging parents. Thus, food provisioning patterns correlated
with measures of food availability in different habitats or on a territory-
to-territory basis would help to clarify causes underlying the patterns
observed in this study.

ACKNOWLEDGMENTS

We thank the 1991 field crew at Hubbard Brook, especially P. P Marra and T. W. Sherry,
for their help, advice, and encouragement throughout this project. Steve Sloan assisted with
the data analysis. The project was supported in part by the Presidential Scholars Program
of Dartmouth College and by an REU supplement to a research grant from the National
Science Foundation to Dartmouth College. P. Marra, T. Sherry, S. Sillett, and P. Hunt pro-
vided helpful comments on the manuscript.

LITERATURE CITED

Bedard, J. and M. Meunier. 1983. Parental care in the Savannah Sparrow Can J Zool
61:2836-2843.

Biermann, G. C. and S. G. Sealy. 1982. Parental feeding of nestling Yellow Warblers in
relation to brood size and prey availability. Auk 99:332-341.

Bormann, F. H. and G. E. Likens. 1979. Patterns and process in a forested ecosystem.
Springer- Verlag, New York, New York.

Breitavisch, R, P G. Merritt, and G. H. Whitesides. 1986. Parental investment by the
Northern Mockingbird: male and female roles in feeding nestlings. Auk 103:152-159.
Conrad, K. F. and R. J. Robertson. 1993. Patterns of parental provisioning by Eastern
Phoebes. Condor 95:57-62.

Curio, E. 1988. Realized life span and delayed cost of parental care Am Nat 131 825-
836.

Goossen, j. P. and S. G. Sealy. 1982. Production of young in a dense nesting population
of Yellow Warblers, Dendroica petechia, in Manitoba. Can. Field Nat. 96:189-199.
Haggerty, T. M. 1992. Effects of nestling age and brood size on nestling care in the
Bachman’s Sparrow {Aimophila aestivalis). Am. Midi. Nat. 128:1 15-125.

Holmes, R. T. 1990. The structure of a temperate deciduous forest bird community: vari-
ability in time and space. Pp. 121-140 in Biogeography and ecology of forest bird
communities (A. Keast, ed.). SPB Academic Publications, The Hague, Netherlands.

. 1994. Black-throated Blue Warbler (Dendroica caerulescens) in Birds of North

America, no. 87 (A. Poole ad F Gill, eds.). The Academy of Natural Sciences of
Philadelphia, Philadelphia, Penn.sylvania; The American Ornithologists’ Union, Wash-
ington D.C.

478

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

^ P p Marra, and T W. Sherry. 1996. Habitat-specific demography of breeding

Black-throated Blue Warblers (Dendroica caerulescens): implications for population
dynamics. J. Anim. Ecol. 65:183—195.

X W. Sherry, P. P. Marra, and K. E. Petit. 1992. Multiple brooding and pro-
ductivity of a neotropical migrant, the Black-throated Blue Warbler {Dendroica caeru-
lescens), in an unfragmented temperate forest. Auk 109:321-333.

^ ’ and E W. Sturges. 1991. Numerical and demographic responses of tem-
perate forest birds to annual fluctuations in their food resources. Acta XX Intern. Or-
nithol. Congr. 111:1559-1567.

Johnson, E. J. and L. B. Best. 1982. Factors affecting feeding and brooding of Gray
Catbird nestlings. Auk 99:148-156.

Knapton, R. W. 1984. Parental feeding of nestling Nashville Warblers: the effects of food
type, brood size, nestling age, and time of day. Wilson Bull. 96:594-602.

Kuitunen, M. and j. Suhonen. 1991. Feeding time and brood-rearing capacity in the
Common Treecreeper {Certhia familiaris). Auk:108:180-184.

Magrath, R. D. 1990. Hatching asynchrony in altricial birds. Biol. Rev. 65:587-622.

Martin, T. E. 1992a. Breeding productivity considerations: what are the appropriate habitat
features for management? Pp. 455-473 in Ecology and conservation of neotropical
migrant landbirds (J. M. Hagan and D. W. Johnston, eds). Smithsonian Press, Wash-
ington D.C.

1992b. Interaction of nest predation and food limitation in reproductive strategies.

Curr. Ornithol. 9:163— 197.

Morehouse, E. L. and R. Brewer. 1968. Feeding of nestling and fledgling Eastern King-

birds. Auk 85:44—54.

Nolan, V. 1978. The ecology and behavior of the Prairie Warbler Dendroica discolor.
Ornithol. Monogr. 26:1-595.

Omland, K. E. and T. W. Sherry. 1994. Parental care at nests of two age classes of male
American Redstart: implications for female mate choice. Condor 96:606-613.

Perrins, C. M. 1980. Survival of young Great Tits, Parus major. Acta XVII Congr. Or-
nithol. Internat. 1:159-174.

Pyle, P, S. N. G. Howell, R. R Yunick, and D. F. DeSante. 1987. Identification guide
to North American passerines. Slate Creek Press, Bolinas, California.

Ricklefs R E. 1969. An analysis of nesting mortality in birds. Smithson. Contrib. Zool. 9.
Robinson, S. K., E R. Thompson III, T. M. Donovan, D. R. Whitehead, J. Faaborg. 1995.
Regional forest fragmentation and the nesting success of migratory birds. Science 267:

1987-1990. , . .

RODENHOUSE, N. L. AND R. T. HOLMES 1992. Results of experimental and natural food

reductions for breeding Black-throated Blue Warblers. Ecology 73:357-372.

Royama, T. 1966. Factors governing feeding rate, food requirement and brood size of

nestling Great Tits Parus major. Ibis 108:313-347.

Sherry T. W and R. T. Holmes. 1992. Population fluctuations in a long-distance neo-
tropical migrant: demographic evidence for the importance of breeding season events
in the American Redstart. Pp. 431-442 in Ecology and conservation of neotropical
migrant landbirds (J. M. Hagan and D. W. Johnston, eds). Smithsonian Press, Wash-

ington D.C.
Skutch, a. 1949.

Do tropical birds rear as many young as they can nourish? Ibis 91.430

455.

Stearns, S. C.
York.

1992. The evolution of life histories. Oxford Univ. Press, New York, New

Goodbred and Holmes • FOOD PROVISIONING OF NESTLINGS

479

Studd, M. V. AND R. J. Robertson. 1989. Influence of age and territory quality on the
reproductive behavior of male Yellow Warblers. Can. J. Zool. 67:268-273.

U.S. Fish and Wildlife Service. 1977. North American bird banding manual, vol. II.
Washington, D.C.

Virolainen, M. 1984. Breeding biology of the Pied Flycatcher Ficedula hypoleuca in
relation to population density. Ann. Zool. Fenn. 21:187-197.

Wilson Bull., 108(3), 1996, pp. 480-495

BREEDING BIOLOGY AND NATURAL HISTORY OF
THE BAHAMA SWALLOW

Paul E. Allen

Abstract. — The Bahama Swallow (Tachycineta cyaneoviridis) is an obligate secondary
cavity-nester endemic to the pine forests of four islands in the northern Bahamas. The near-
threatened status of this poorly known species stems from the limited extent of pine forest
breeding habitat, a history of logging in that habitat, and potential competition from exotic
secondary cavity-nesters. Natural nest sites of Bahama Swallows on Grand Bahama gen-
erally were abandoned woodpecker cavities and nests in all types of cavities were built from
pine needles, Casuarina spp. twigs, and grass. Mean clutch size was 3.0 and the pure white
eggs were slightly larger than those of Tree Swallows (T. bicolor). Both the mean incubation
and nestling periods, 15.8 days and 22.7 days, respectively, were longer than those of Tree
Swallows. Hatching success and nestling success were 87% and 81%, respectively, giving
an overall success rate of 70%. One case of double-brooding was documented, and two
other likely cases were noted. Weekly surveys of adults in pine forest habitat on Grand
Bahama during breeding gave a linear density of 0.18-0.25 pairs-km '. The result from a
single survey on Andros (0.21 pairs-km-') corresponds to survey results on Grand Bahama
in the same period and very roughly agrees with the outcome of a 1988 survey. Received
13 October 1995, accepted 18 February 1996.

The Bahama Swallow {Tachycineta cyaneoviridis), currently listed as
near-threatened (Collar et al. 1992), is a poorly known endemic of the
islands of Andros, Abaco, New Providence, and Grand Bahama in the
northern Bahamas (American Ornithologists’ Union 1983). Like other
members of the Tachycineta genus, the species is an obligate secondary
cavity-nester (Turner and Rose 1989). Bahama Swallows nest mainly m
cavities in Caribbean pine trees {Pinus caribaea), and their breeding sea-
son distribution corresponds to the distribution of the pine forest (Smith
and Smith 1989). Smith and Smith (1989) summarized most known in-
formation about the species from previously published anecdotes and their
own limited observations, yet much remains unknown. Neither its nest
nor eggs has been reliably described (Smith and Smith 1989), contrary
to reports otherwise (Turner and Rose 1989). The need for more infor-
mation about the Bahama Swallow is obvious if we are to understand the
conservation needs of this near-threatened species.

Conservation concerns for the Bahama Swallow stem from the limited
extent of their pine forest habitat and a history of logging in that habitat.
A recent silviculture inventory gave the total area of pine forest in the
Bahamas as 2042 km^ (Allan 1986) and, though the total extent of forest

Bahamas National Trust Rand Nature Centre, East Settler's Way. P_0. Box F-4.M41 Freeport, Grand
Bahama. Bahamas. (Present address: Montana Cooperative Wildlife Research Unit. University of Mon-
tana. Missoula. Montana 59812.)

480

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

481

apparently has not changed due to logging (Henry 1974), most of it is
second growth (Swenson 1986). No logging in the Bahamas has occurred
since the early 1970s (Henry 1974), but history shows how quickly this
limited habitat can be altered. Over 70% of the forest on Grand Bahama
was harvested in just three years during the peak of logging there in the
1950s (Henry 1974). This comprised nearly 30% of all pine forest in the
Bahamas. Given the limited nature of breeding habitat and the extent of
loss possible through logging, concern about the conservation of the Ba-
hama Swallow is appropriate. Conservation problems caused by loss of
habitat could be exacerbated by competition for nest sites with exotic
secondary cavity-nesters. House Sparrows {Passer domesticus) and Eu-
ropean Starlings {Sturnus vulgaris), which are also present in the Baha-
mas.

As the first step in addressing the conservation concerns associated with
the Bahama Swallow, I report here the findings of recent research on their
natural history and breeding biology. I use the Tree Swallow {T. bicolor),
a temperate congener, as the basis for comparing various aspects of Ba-
hama Swallow breeding biology since none of the tropical congeners of
the Bahama Swallow (e.g.. Mangrove Swallow T. albilinea) is as well-
known. I also describe results of surveys that expand upon a pilot survey
in 1988 (Smith and Smith 1989) and which provide baseline information
for monitoring the population size of the Bahama Swallow.

STUDY AREA AND METHODS

I studied breeding Bahama Swallows on Grand Bahama (26°40'N, 78°30'W) in the Ba-
hamas from mid-March through June 1995. I found nests in natural sites throughout the
forested part of the island between Freeport and McClean’s Town, about 75 km east of
Freeport. Most nests in artificial sites were located at an abandoned U.S. Air Force missile
tracking base (hereafter “Missile Base”) near Freetown (26°37'N, 78°21'W) about 35 km
east of Freeport. All nests were either in or adjacent to tracts of secondary pine forest which
make up most of the interior of Grand Bahama east of Freeport. Most nests in natural sites
were found by observing swallows loitering on dead pine trees (“snags”). Nests at the
Missile Base were found by systematic .searches of artificial cavities and by observing
swallows. The presence of a nest in an inaccessible nest site (i.e., most snags) was inferred
by seeing swallows entering a cavity with nest material or by observing birds entering a
cavity on several different occasions.

Most nests in natural sites were observed from the ground once every two or three days,
but some were observed only once or twice in two weeks during the first few weeks of the
breeding .season. Observations generally lasted only long enough to confirm that a nest was
still active. Activity was determined to have cea.sed at a nest when either two 0.5 h obser-
vations on consecutive days showed no activity or when a single 1 h observation revealed
no activity. However, I often made extra observations to confirm lack of activity. If activity
at a nest ceased without my having observed about three weeks of frequent nest visits (which
I assumed to be feeding visits), the nest then was assumed to have failed unless some other
clue (e.g., previous sightings of nestlings looking out of the hole) indicated probable fledg-

482

THE WILSON BULLETIN • Vol. JOS, No. 3. September 1996

ing. Dates of either fledging or failure of a nest were estimated to be the midpoint between
the last observation of activity at the nest and the first observation with no activity. Dates
of clutch completion and hatching were estimated by subtracting the length of the average
incubation and nestling periods (calculated from detailed observations of nests at the Missile
Base) from the estimated fledging date.

I examined some nests in natural sites using a 1-m fiberscope, reaching the cavities with
a 10-m extension ladder (Rohwer 1988). The limited resolution and depth of field of the
fiberscope did not allow precise counts of chicks or eggs in a nest, so I have accurate counts
of eggs or chicks only for those nests I excavated. Cavities were excavated by carefully
enlarging the existing entrance hole of the cavity with a saw. Cut-away pieces from exca-
vations were replaced and secured with wire, thus maintaining the integrity of the cavity.

Some artificial cavities used as nest sites were nest boxes (both standard and Peterson
box designs) or plastic Purple Martin (Progne subis) gourds (Carroll Industries, Van, Texas)

I erected at the Missile Base. Nests in accessible artificial cavities were observed daily prior
to egg-laying, during egg-laying, and for several days prior to hatching. Eggs were measured
with dial calipers to the nearest 0.1 mm and weighed to the nearest 0.1 g with an electronic
balance (Pocket Pro 150-B, Acculab, Newtown, Pennsylvania), generally on the day of
laying.

At nests in accessible artificial cavities, I took measurements of chicks younger than 18
days old nearly every day. Mass was measured with the balance to the nearest 0. 1 g.
Straightened, flattened wing chord of the right wing was taken with dial calipers to the
nearest 0.1 mm until the chord was about 15 mm long and thereafter with a ruler to the
nearest 0.5 mm. Chicks within the same nest were identified by uniquely marking their
wings or legs with a permanent felt-tip marker. These markings were superseded by color
bands and numbered aluminum bands when the chicks were 7-14 days old. To avoid pre-
mature fledging, most chicks older than 17 days were simply counted without handling.
Thus nestling periods are reported on a nest-wise basis (i.e., the period between the first
chick hatching and the last chick fledging) instead of for individual chicks. Three late-season
nests at the Missile Base received only enough visits to determine the number of chicks
hatching and fledging. Permission to salvage and export several nests, eggs, and chicks was
granted by the Bahamas Dept, of Agriculture.

Censusing or surveying highly mobile birds is difficult and my attempts to apply distance
sampling techniques (Buckland et al. 1993) to survey Bahama Swallows along forest roads
on Grand Bahama were unsuccessful. The distance data required for that method was im-
possible to collect, since swallows were often sighted while flying without any reference
object nearby to which distance could be measured or even roughly estimated. Ultimately,
I simply counted all swallows, whether foraging or perched, along separate survey routes
in three areas of Grand Bahama on different days. Since I rarely observed other swallows
on Grand Bahama during the breeding season, I assumed that all swallows I could not
identify were Bahama Swallows. The three routes generally were covered on consecutive
days. Weather on survey days typically was sunny and warm, with the few exceptions being
slight overcast or cloudy conditions. The Eastern Lucaya route was 19 km of sparsely settled,
relatively undeveloped subdivisions covered mostly with secondary forest and a dense net-
work of roads. The second route covered 45 km in the Lucayan Estates subdivision, an area
of secondary forest without any housing but with several farms and a dense road network.
The East End route was 58 km long and used a logging road which ran down the center of
much of the island. The eastern two-thirds of that route was relatively undisturbed secondary
forest, and the remainder went through parts of Lucayan Estates. The routes were driven at
a speed of 10-14 kph using either a moped (in May) or automobile (in June), beginning
between 06:30 and 07:00 EST. I was both driver and observer for all Grand Bahama surveys.

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

483

In an earlier survey of Andros, Smith and Smith (1989) drove 96.4 km through the pine
forest of that island at speeds under 30 kph on two days, 20-21 May 1988. An assistant
and I performed one survey on Andros covering 76.0 km of the 1988 route (P. Smith, pers.
comm.) following the 1988 protocol for the single day of our survey, 26 May 1995. Eor
purposes of estimating the number of breeding pairs in all surveys, I assumed that groups
of either one or two swallows represented one breeding pair and that groups of either three
or four birds represented two breeding pairs. Juveniles identified as such were not counted.

RESULTS AND DISCUSSION

Nesting activity. On 31 March, I noted the first exhibition of nesting
behavior by Bahama Swallows since I had begun observations in mid-
March. Swallows repeatedly flew up to and hovered in front of louvered
access panels on the upper floors of a 12-story building near Freeport. I
interpreted this activity as prospecting for nest sites and saw similar be-
havior at the same building on 2 April, when swallows approached the
undersides of balconies and eaves. The first nesting behavior I observed
at a natural nest site was on 2 April when two to four swallows flew
around and approached a woodpecker hole in a snag. I observed a swal-
low taking nest material into that hole on 10 April. At this and other nests
in natural sites, swallows were active at their nests only between about
07:00 and 1 1:00 during the nest-building stage, and attempts to find nests
by observing adults at natural sites were fruitless later in the day.

Overall, I found 18 nests in natural nest sites: 10 nests seemed to be
successful, five failed without fledging young, and three nests were still
active when I left the island. To establish that nests in natural and artificial
cavities did not differ in obvious ways, I examined nine of the 18 natural
nests with a fiberscope. I found chicks in five of them, eggs in three of
them, and neither eggs nor chicks in the last. In four nest sites that I
excavated, I found incubated clutches of three eggs in each of two cav-
ities, three chicks in another, and a partial nest in the last. What I saw of
nests in natural cavities convinced me that they were similar to nests in
artificial cavities with respect to clutch size and types of nest material
used. I assume that most other aspects of breeding biology do not differ
greatly between swallows nesting in natural and artificial sites.

Nest-site characteristics. — Although Bahama Swallows do use cavities
in live trees (Smith and Smith 1989), all 18 nests I found in natural nest
sites were in pine snags. AH of those sites were abandoned woodpecker
holes, except one which was in a large, cracked branch. The pine snags
used for nesting had a mean diameter at breast height of 22.3 cm (N =
18, SD = 4.26, range: 17.0-28.3 cm). The mean height of the snags was
9.6 m (N = 18, SD = 2.17, range: 6.7-12.8 m), while the mean height
of the cavities was 8.8 m (N = 18, SD = 1.94, range: 6.0-1 1.4 m). The

484

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

closest active nests that I found in natural cavities were about 150 m from
each other.

All but two of 14 nests I found in artificial cavities were at the Missile
Base. Types of artificial cavities used for nesting included housings of
street lights (2 nests), a horizontal pipe (ca 5-6 cm diameter) (1 nest), a
gap (less than 10 cm) between two sections of wall in the side of a
building (1 nest), an electrical conduit box with the only access being a
42 mm diameter hole in the 15 cm X 13 cm floor of the box (1 nest), a
Peterson-style nest box with a 38 mm diameter hole (1 nest), a standard
nest box with a 38 mm diameter hole and 14 cm square floor (1 nest),
an artificial nest gourd with a 55 mm diameter hole (1 nest), and rooftop
ventilation units with rectangular access holes 95 mm wide by 28 mm
high (6 nests in 5 units). The height of these artificial cavities ranged
from 3.0 m for the Peterson box attached to a utility pole to 13.2 m for
the pipe nest. The nest between two wall sections was an exception since
it was at the top of the 12-story building at which I first saw nesting
behavior. The two closest active nests were in ventilation units 8.8 m
apart on top of the same roof.

Nest construction. — During observations of nest building at natural and
artificial nest sites, I never saw both birds of a pair carrying nest material.

I assumed that just one bird of each pair, which I took to be the female,
did most, if not all, of the nest construction. The male often escorted the
female while she was gathering nest material. I observed birds gathering
nest material from the edges of paved and unpaved roads as well as from
the middle of grassy areas that were recently mown. Caribbean pine nee-
dles grass, and Casuarina spp. “needles” (actually fine, segmented twigs
of this exotic tree) formed the bulk of the nests I examined in both arti-
ficial and natural nest sites. I observed one bird travel over 200 m from
its nest site to collect material, but most trips for nest material by other

birds were less than 100 m. ....

The period from when nest building began until clutches were initiated

was 14-18 days in four nests built from scratch in artificial cavities. This
period might be shorter for nests in natural sites since they contained less
nest material than those in artificial sites. The masses of material from
two nests collected from pine snags were quite small (9.0 g and 17.1 g)
compared to the masses of material from four nests m artificial sites built
completely in the season of the study (18.0 g, 41.6 g, 48.2 g, and 114.9
g) This difference may result from the generally larger volume of the
artificial cavities. I collected eight nests, two from snags and six from
artificial sites, after fledging or abandonment of the nests and deposited
them with the Vertebrate Collection at Cornell Univ., Ithaca, New York.

Nest lining materials. — Nests were lined with a variety of materials.

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

485

both naturally occurring and artificial. Flakes of pine bark were common
in nests, and I saw swallows pulling bark directly off trees on several
occasions. Though not every nest contained bark, most contained 1—5
pieces of about 1—3 cm^. Small downy feathers, seemingly originating
from other passerines, were the other common natural lining material.
Nests usually contained 1—5 such feathers, markedly fewer than the scores
of waterfowl feathers often found in Tree Swallow nests (P. Allen, pers.
obs., Winkler 1993). The largest number of feathers I found in a nest was
15-25 flamingo feathers in a nest in a pine snag on the grounds of the
Rand Nature Centre which maintained a small, captive flock of Greater
Flamingos {Phoenicopterus ruber).

Other natural lining materials I found in nests included small dry
leaves, pieces of skin shed from small lizards, yellow flower petals (found
in only one nest), and a pale yellow butterfly wing (found in only one
nest). In one incomplete nest in a snag less than 500 m from the seashore,
I found several dried strands of turtle grass (Thalassia testudinum), a sea
grass which commonly washes up on beaches. This may explain an ob-
servation of Bahama Swallows gathering mouthfuls of “seaweed” and
flying towards the forest (Todd and Worthington 1911).

Artificial material in nest linings was most abundant at the Missile Base
where litter was plentiful. Bits of shredded plastic wrap, small pieces of
newspaper, facial tissue, and regular paper were common in nests there.
These materials were presumably used because of their gross similarity
to feathers. A few nests at the Missile Base also contained paint chips
from paint peeling off buildings. If the birds actually pulled the paint
chips off the buildings instead of picking them off the ground, then the
actions required to do this would be similar to those used for collecting
pine bark.

Eggs. — The eggs I examined from two nests in snags and 10 nests in
artificial cavities were larger, on average, than Tree Swallow eggs in
length, width, and mass (Table 1 ; masses of eggs from the two snags
excluded because the eggs were not freshly laid). The mean clutch size
was 3.0 eggs in these 12 nests and in an additional nest in an artificial
cavity (Table 1). This is the same clutch size reported for a Caribbean
congener, the Golden Swallow (T. euchry.sea) (Turner and Rose 1989)
but, not surprisingly, it is much smaller than for Tree Swallows (Table
1). Of these 1 1 nests in artificial cavities, two had clutches of two eggs,
seven had three-egg clutches, and two had four-egg clutches. The two
nests in natural cavities both had three-egg clutches. As with Tree Swal-
lows (R Allen, pers. obs.; Robertson et al. 1992), the color of freshly laid
eggs was white, but translucent and slightly pinkish, changing to pure
white after a few days of incubation. In all nests where laying was ob-

486

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Comparison of Bahama Swallows and Tree Swallows for Several Aspects oe

Breeding Biology

Bahama Swallow Tree Swallow

N

X ± SD

Range

N

jf ± SD

Range

pd

Egg length

(mmT

36

19.4 ± 0.72

17.2-20.6

2295

19.0 ± 0.90

16.0-22.8

*

Egg width
(mmT

Egg mass (gT

36

30

13.9 ± 0.40
2.0 ± 0.19

12.9-14.6

1. 6-2.3

2292

835

13.6 ± 0.47
1.8 ± 0.17

11.9-15.4
1. 4-2.5

***

•k ^ ^

Clutch size
(eggs)"

13

3.0 ± 0.58

2-4

847

5.4 ± 0.91

3-8

***

Incubation
period (d)”

5

15.8 ± 1.10

15-17

235

14.5 ± 1.13

12-19

*

Hatching

success'’

30

86.7%

33.3-100

10,107

86.9%

—

—

Nestling

suceess'’

26

80.8%

0-100

21,130

83.1%

—

—

Nestling
period (d)‘'‘’

6

22.8 ± 1.21

22-25

554

20.6 ± 1.63

16-29

**

“Tree Swallows in upstate New York (D. Winkler, unpubl. data).

'■Tree Swallows from several studies (Robertson et al. 1992).

“ Calculated for Tree Swallows on a nest-wise basis, just as with the Bahama Swallows.
Results from r-tests comparing means using equal or unequal vanances as appropriate.
*** = P < 0.001.

= P s 0.05, ** = P s 0.01,

served, eggs were laid one per day in the morning. I never observed

Bahama Swallows copulating during this study.

Incubation. — Observations of five nests in artificial cavities yielded no
consistent indication as to when incubation began. Even after the clutches
were complete and incubation had presumably begun, eggs were often
unattended when I checked the nests. This pattern may have resulted from
often visiting nests in the early afternoon, generally the hottest part of the
day. However, defining the incubation period as starting on the day the
last egg was laid and ending on the day the first egg in the nest hatched,
three nests had incubation periods of 15 days, and two nests had incu-
bation periods of 17 days, giving a mean of 15.8 days (Table 1). This is
over one day longer than the incubation period for Tree Swallows m
upstate New York (Table 1). Though I did not often capture birds on the
nest, there was no indication that males shared incubation responsibilities.

Hatching and sur\’ival.—Of the seven nests I visited daily dunng hatch-
ing, the eggs in three nests, two with two-egg clutches and one with a
three-egg clutch, hatched in the same 24-hour period. The eggs in the
four remaining nests, each containing three or four fertile eggs, hatched

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

487

over a period of 48 h. In those seven nests and in three late-season nests,
26 chicks (87%) hatched from 30 eggs (two eggs in one nest and one
egg in a second nest were infertile; an additional egg was missing). Of
the 26 chicks that hatched, 21 (81%) eventually fledged, giving an overall
6gg-to-fledgling success rate of 70%. The hatching and nestling success
rates of these Bahama Swallows correspond closely with those of Tree
Swallows from several studies (Table 1). The success rates for both these
species are for birds nesting in artificial cavities, and success may be
lower in natural cavities.

One three-day-old chick was found dead of unknown causes in its nest,
while its nest mates remained in good health. One entire brood of three
chicks was lost because they fell out of the ventilation unit that housed
their nest. One other chick apparently was killed when the motor was
activated in the ventilation unit that housed its nest. I salvaged three of
these chicks and one of the infertile eggs, depositing them at Cornell.
These are the only such specimens known for the Bahama Swallow
(Smith and Smith 1989).

Nestling period and fledging. — In six successful nests that I monitored
closely, the fledging period was 22 days for four nests, 23 days for the
fifth nest, and 25 days for the remaining nest, giving a mean of 22.7 days
(Table 1). This nestling period is two full days longer than that of Tree
Swallows in upstate New York (Table 1). The siblings from three nests
each containing two chicks fledged in the same 24-hour period. Siblings
from another nest containing three chicks fledged over a 48-hour period.
In two other nests with broods of three young, the fledging period was
unknown because of imprecise counts of the young.

Chick development. — The rate of mass gain for Bahama Swallow
chicks from seven nests was slower than that of Tree Swallows (Fig. 1),
but the period during which chicks rapidly increased mass (days 1-12)
was similar to that for Tree Swallow chicks. The mass of chicks from
both species plateaus near their adult mass at about day 13 (Fig. 1).
Growth rates, calculated by fitting a logistic curve to daily means of mass,
show that Bahama Swallows (K = 0.363) grow more slowly than Tree
Swallows (K = 0.396) using Tree Swallow data from McCarty (1995).
Adult wing lengths of Bahama Swallows are about 4 mm shorter than
those of Tree Swallows (Turner and Rose 1989), and the average length
of Bahama Swallow wing chords was 2.3 mm shorter than, those of Tree
Swallows for days 10-19 (Fig. 2).

In most respects, newly hatched Bahama Swallow chicks were similar
to newly hatched Tree Swallow chicks (R Allen, pers. obs.). The one
exception was that all Bahama Swallow chicks were hatched with several
down feathers (neossoptiles) already formed. Tree Swallows in New York

488

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Age (days)

Fig. 1. Mass (means ± 1 SE) of Tree Swallow (McCarty 1995) and Bahama Swallow
chicks. Upper, dotted horizontal line shows adult mass of 21.3 g for Tree Swallows from
McCarty (1995) Lower, dotted horizontal line shows mean mass of four breeding Bahama
Swallow females captured at the Missile Base (16.3 g). This value corresponds with low
end of mass range (16.3-19.5 g, mean 17.5 g) from museum specimens (Turner and Rose

1989).

often hatch completed naked and only occasionally hatch with one or
more wispy down feathers (P. Allen, pers. obs.). In Bahama Swallows,
dark feather tracts began showing underneath the skin on the wings by
the second day. On the third day, tracts were visible on the head and back
as well. Hair-like shafts of primaries as well as back and chest feathers
began breaking through the skin on the fourth and fifth days. By the
seventh day, the shafts of body feathers were less than 1 mm long. Pri-
mary and tail feathers began emerging from their shafts on about day
nine or ten. By day ten, body feathers were emerged 1-2 mm from their
shafts. The eyes of Bahama Swallow chicks began to open on their fifth
day. Chicks’ eyes were just small slits on days five or six with the slits
widening until being fully rounded by the tenth day.

Parental care. — At least two adults, which I took to be the parents.

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

489

Age (days)

Fig. 2. Straightened, flattened wing chord (mean.s ± 1 SE) of Tree Swallow (McCarty
1995) and Bahama Swallow chicks.

fed young at most nests. At the Missile Base adults from neighboring
nests assisted in defending nests against me during daily visits. In contrast
to Tree Swallow nests, where it is common to find the entire nest well-
covered in fecal matter after fledging (P. Allen, pers. obs.; Robertson et
al. 1992), most Bahama Swallow nests in artificial cavities were clear of
fecal material after chicks fledged. Either Bahama Swallow parents pro-
vided nest sanitation throughout the nestling phase or chicks were able
to defecate out of the entrances to their cavities. However, I did not note
much fouling of the area immediately below nests which would have
indicated that the nestlings were responsible for sanitation.

Fledglings. — Observations of post-fledging chicks were difficult to ob-
tain. In one case, I observed four fledglings (identified by their yellow
gapes) perched in a tree with an adult feeding them. In another instance,
I found four color-marked sibling fledglings in a group less than 500 m
from their nest six days after the last chick had fledged from their nest.

Double broods. — Of the 12 nests at the Missile Base, at least one, and

490

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

possibly three, represented second broods after the successful fledging of
a previous brood. I verified double brooding by a female banded at her
first nest on 17 May while feeding chicks. She was captured again on 21
June on a second nest in a ventilation unit 10.7 m away from the first
nest on the same rooftop. The first brood had fledged on 29 May and the
second clutch was initiated on 18 June in a nest that I had found in April
and identified as being an unused nest from a previous season. Some new
nest material had been added to that nest, and rusty flakes of metal had
been removed from the nest bowl prior to egg-laying. Four chicks fledged
from the first brood, and two chicks fledged from the second.

Another possible double-brood attempt was a clutch of three eggs I
found on 16 June in the same nest from which three chicks had fledged
on 4 June. The second brood in that nest produced two fledglings. Another
possible case of double-brooding was a clutch initiated on 14 June in a
nest box within 25 m of an inaccessible nest that was active until some-
time during the first week of June. The initiation of this late clutch falls
within days of the initiations for the two other double-brood nests, after
a period of more than four weeks without a known clutch initiation at

the Missile Base.

I made a special effort to look for renewed nesting activity at natural
sites in mid-June but was unable to confirm any other possible second-
brood nests. Finding such nests might be especially difficult if nests used
earlier in the season were simply reused without more nest building. The
last search for new nests at the Missile Base was on 26 June, so I do not
know if there were more late-season nests initiated there after that date.

Overall phenology.— The mean date of clutch completion for nine nests
in snags and eight nests in artificial cavities (which excludes the three
late-season nests) was 5 May (SD 6.96, range; 20 April- 15 May). The
average date of hatching for those clutches was 20 May (SD 6.98, range:
5 May-28 May), and the mean fledging date was 11 June (N = 16, SD
7.30, range: 27 May-22 June). For the two late-season nests in which I
observed egg-laying, the average date of clutch completion was 18 June.
The mean estimated dates of hatching and fledging for those two broods

were 4 July and 26 July, respectively.

Interspecific compefinwT.— Bahama Swallows nesting in natural nest
sites had numerous interspecific agonistic interactions with four other cav-
ity-nesting bird species. Two were native species, the Hairy Woodpecker
iPicoides villosus) and La Sagra’s Flycatcher (Myiarchiis sagrae). The
other two. House Sparrow {Passer domesticus) and European Starling
(Sturnus vulgaris), were exotic. In two interactions with woodpeckers,
swallows harassed woodpeckers which were in possession of nest sites,
but in both cases the woodpeckers remained in control of the cavities.

Alien • BAHAMA SWALLOW BREEDING BIOLOGY

491

However, I did find two cases of swallows nesting quite close to wood-
peckers. In one instance, active swallow and woodpecker nests were 30
m apart, and in another, the nests were 75-100 m apart. I observed a
swallow being displaced from the rim of a nest-hole in a snag by a fly-
catcher bringing either food or nest material into the cavity. In two other
cases, flycatchers perching at former swallow nest sites were displaced
by swallows, even though the swallow chicks had already fledged. I found
one instance of flycatchers and swallows nesting within 100 m of each
other. In a case that is difficult to interpret, I excavated a nest site several
weeks after I had seen swallows entering the cavity, and found it filled
with typical House Sparrow nest material but with four rotten flycatcher
eggs at the bottom. I observed no direct interactions between House Spar-
rows and Bahama Swallows. However, one cavity in which swallows
were nest building was later usurped by sparrows which successfully
raised a brood of young there. In another instance, I found a pair of
sparrows inspecting a cavity in which there had been an active swallow
nest with eggs about two weeks earlier. However, I do not have obser-
vations for the intervening period to give any hints as to whether the
swallows had abandoned because of the sparrows. The Missile Base had
a healthy population of breeding House Sparrows which seemed to ex-
clude swallows from nesting in sites they might typically choose in the
absence of sparrows. The sparrows had a monopoly on nest sites under
the eaves of the roofs, while the swallows nested, for the most part, in
sites that gave no means of clinging to the entrance hole or perching
before entering the cavity. Such sites were probably difficult or impossible
for sparrows to access. I observed very few interactions between starlings
and swallows, but I did find one active starling nest within 75 m of a
swallow nest.

Although the exotic cavity nesters have the potential to impact greatly
the Bahama Swallow through competition for nest sites, I found these
species mainly within about one kilometer of human structures or other
disturbance. I never observed either species in undisturbed secondary for-
est, but House Sparrows were at farms in the middle of the secondary
forest. As human development and disturbance encroach on the forest, it
is inevitable that the local ranges of these exotics will spread.

Previous surveys and density estimates. — Emlen (1977) estimated a
Bahama Swallow density of 11.0 birds-km*^ in pine forest during the
breeding season by surveying a total of 21.5 km of transects using the
coefficient of detectability methodology (Emlen 1971). Using Henry’s
(1974) estimate of the total pineland area extant at that time (1782 km^),
Emlen’s density figure results in a population estimate of just under
20,000 birds in the entire species’ breeding range. Caution should be used

492

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

before considering this quick extrapolation as a true reflection of the Ba-
hama Swallow population during that study, since the survey transects
seemed to have been restricted to prime breeding habitat (Emlen 1977).
Also, Emlen (1977) was mainly concerned with making relative, inter-
specific comparisons among species and did not attempt to make absolute
estimates of population sizes.

Smith and Smith (1989) is the only other source of quantitative data
for the Bahama Swallow. From a simple road survey in 1988, they esti-
mated the density of breeding Bahama Swallows at 2.6 birds-km ^ (Smith
and Smith 1989). Using this density estimate and Henry’s (1974) estimate
of pine forest area, the total breeding population of the Bahama Swallow
would have been about 4800 birds. Smith and Smith (1989) conceded the
imprecision of the estimate but felt the result was of the correct magni-
tude, between 1000 and 10,000 living Bahama Swallows. This estimate
is quite different from one derived from Emlen’s data, but it is debatable
whether the difference in the two results reflects an actual decrease in the
Bahama Swallow population, at least of the magnitude indicated. A direct
comparison between the two results can be misleading, since the methods
used were different.

Grand Bahama surveys. — In the surveys I performed, the results from
each of the three individual routes were somewhat irregular (Fig. 3; East-
ern Lucaya range: 0.10—0.56 pairs-km”'; Lucayan Estates range. 0.11 —
0.30 pairs-km-'; East End range: 0.12-0.27 pairs-km"'). A weighted av-
erage of the sightings from the three routes showed a pattern of increasing
frequency of sightings, from 0.17 pairs-km-' to 0.25 pairs-km"', during
the period between the average dates of hatching and fledging (Fig. 4).
This is consistent with adult swallows spending more time foraging in
response to an increased demand for food as their chicks develop. Esti-
mating breeding density from these results would be misleading because
of the assumptions required to do so (e.g., that only birds breeding within
a certain distance of the road were sighted) and because of bias introduced
by what seemed to be an affinity to the road by the birds. Instead, the
survey results should be considered indices to the population size.

Andros survey. — The 1995 Andros survey served to make a direct com-
parison between contemporary survey results and those of Smith and
Smith (1989) without complications in interpretation arising from differ-
ent protocols or routes. In 1988, Smith and Smith ( 1989) observed 0.28
pairs-km-'. In 1995, we saw eight single swallows and eight two-somes
while covering just 70% of the 1988 route, giving a sighting rate of 0.21
pairs-km '. Though the 1995 result represents a 25% decrease from the
1988 survey, the limited nature of the Andros surveys precludes the con-
clusion that the decrease reflects a population decline. However, since the

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

493

Date

Eig. 3. Bahama Swallow pairs-km ' from each of three survey routes sampled on three
roughly consecutive days. Date is the day that the second route was sampled. The large
variance of sightings on the Eastern Lucaya route is probably due to its shorter length.

result of the 1995 Andros survey roughly corresponds to the results of
surveys on Grand Bahama near the same time (0.20 pairs-km“' for 21
May and 0.25 pairs-km-' for 2 June), it seems likely that the 1995 Andros
results may be a reasonable index of the Andros population.

Conclusion. — Like many other species, the greatest threat the Bahama
Swallow faces probably is habitat destruction. The most likely cause of
major habitat alteration loss in the Bahamas will be logging, especially
since much of the secondary forest is now becoming mature after the last
spate of harvesting. However, another source of major habitat loss will
be from housing development, particularly when the residential retirement
and resort communities planned for Grand Bahama are more fully imple-
mented. According to promotional brochures, these subdivisions are to
house over 500,000 people and cover about 170 km^ (R Allen, unpubl.
data), most of which is currently forested. This area was not included in

Pairs per km

494

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Lig. 4. Survey results combining the three routes of each survey using actual number
of birds sighted (dashed line) or estimated number of pairs sighted (solid line). Date is the
day that the second route was sampled.

the pine forest inventory by Allan (1986), but its development will ef-
fectively eliminate about 8% of the breeding habitat currently available
to Bahama Swallows. Hurricanes pose another threat to Bahama Swallow
habitat since they can demolish large portions of the forest on individual
islands and have done so before in the Bahamas. However, if habitat loss
from all sources can be minimized, and possibly mitigated through con-
servation measures such as nest box and snag management programs, the
Bahama Swallow does not seem likely to become endangered. But, given
the limited area of pine forest and the vulnerability of that habitat to
human alteration, it seems unlikely that the conservation status of the
species could ever be upgraded from its current near-threatened status.

ACKNOWLEDGMENTS

Lori Bu.shway, John Confer, Rick and Kathy Oliver, Bill Smith, and David Winkler pro-
vided support and advice that made this project possible. Nancy Jones furnished housing

Allen • BAHAMA SWALLOW BREEDING BIOLOGY

495

and some transportation for the field work. Lori Bushway gave valuable field assistance and
encouragement. Chuck Cavender provided logistical and field assistance on Andros. Phil-
more Russell and Edward Robinson also aided me in the field. I am grateful to the many
landowners who gave me access to their property, especially to the Ministry of Public Works
and Local Government for access to the Missile Base. I thank David Winkler for use of his
Tree Swallow data. Tom Martin, Bill Smith, Bridget Stutchbury, David Winkler, and an
anonymous reviewer provided comments which improved this manuscript. This study was
funded by the Lincoln Park Zoo Scott Neotropic Fund, a Kathleen S. Anderson Award from
Manomet Observatory, the North American Bluebird Society, the Purple Martin Conser-
vation Association, and the Association of Field Ornithologists.

LITERATURE CITED

Allan, T. G. 1986. Management plan for the pine forests of the Bahamas. UTF/BHA/003/
BHA Consultancy Report No. 2. Food and Agriculture Organization of the United
Nations.

American Ornithologists’ Union. 1983. Check-list of North American birds. 6th ed.
A.O.U., Washington, D.C.

Buckland, S. T, D. R. Anderson, K. P. Burnham, and J. L. Laake. 1993. Distance
sampling; estimating abundance of biological populations. Chapman and Hall, London,
England.

Collar, N. J., L. P. Gonzaga, N. Krabbe, A. Madrono Nieto, L. G. Naranjo, T. A. Parker
III, AND D. C. Wege. 1992. Threatened birds of the Americas. The ICBP/IUCN red
data book. Smithsonian Institution Press, Washington, D.C.

Emlen, j. T. 1971. Population densities of birds derived from transect counts. Auk 88:323-342.

. 1977. Land bird communities of Grand Bahama Island; the structure and dynamics
of an avifauna. Ornithological Monograph No. 24. Allen Press Inc., Lawrence, Kansas.
Henry, P W. T. 1974. The pine forests of the Bahamas. British Foreign and Commonwealth
Office Land Resource Study No. 16.

McCarty, J. P. 1995. Effects of short-term changes in environmental conditions on the
foraging ecology and reproductive success of Tree Swallows, Tachycineta hicolor.
Ph.D. diss. Cornell Univ., Ithaca, New York.

Robertson, R. J., B. J. Stutchbury, and R. R. Cohen. 1992. Tree Swallow. {Tachycineta
bicolor) In The birds of North America, No. 1 1 (A. Poole, P. Stettenheim, and F. Gill,
eds.). The Academy of Natural Sciences of Philadelphia, Philadelphia, Pennsylvania,
and The American Ornithologists’ Union, Washington, D.C.

Rohwer, S. 1988. Guyed extension ladder for access to high nests. J Field Ornithol 59-
262-265.

Smith, P. W. and S. A. Smith. 1989. The Bahama Swallow Tachycineta cyaneoviridis; a
summary. Bulletin of the British Ornithologists’ Club 109:170-180.

Swenson, J. 1986. Bahamas forest inventory. UTF/BHA/003/BHA consultancy report No.

7. Food and Agriculture Organization of the United Nations.

Todd, W. E. and W. W. Worthington. 1911. A contribution to the ornithology of the
Bahama Islands. Ann. Carnegie Mus. 7:388-464.

Turner, A. and C. Rose. 1989. A handbook to the swallows and martins of the world.
Christopher Helm, London, England.

Winkler, D. W. 1993. Use and importance of feathers as nest lining in Tree Swallows
(Tachycineta bicolor). Auk 1 10:29-36.

Wilson Bull., 108(3), 1996, pp. 496-506

NEOTROPICAL MIGRATORY BREEDING BIRD
COMMUNITIES IN RIPARIAN EORESTS OP
DIFFERENT WIDTHS ALONG THE
ALTAMAHA RIVER, GEORGIA

Malcolm F. Hodges, Jr.' and David G. Krementz^

Abstract. We surveyed riparian forest corridors of different widths along the lower

Altamaha River in Georgia in 1993 and 1994 to investigate the relationship between forest
corridor width and Neotropical breeding bird community diversity and abundance. Species
richness and abundance of three of six focal species increased with increasing forest corridor
width We suggest if Neotropical breeding bird communities are a target group, that land
managers should consider leaving a 100 m buffer strip along riparian zones. Received 28
Aug. 1995, accepted 13 Feb. 1996.

Studies of the effects of forest fragmentation (Whitcomb et al. 1981,
Lynch and Whigham 1984, Robbins et al. 1989) on communities of breed-
ing birds in discrete forest blocks of different sizes have suggested that
many Neotropical migratory bird species are sensitive to a reduction m
forest area. More difficult to assess is the effect of reduction in width of
riverine forest corridors on bird populations; although forest area may be
reduced by encroachment from the corridor edge, these forests often re-
main contiguous with wider sections of the corridor. Stauffer and Best
(1980) found that bird species richness increased from about 10-30 spe-
cies with increasing width in wooded riparian habitats ranging from 10—
200 m wide in Iowa. They noted, however, that seven of 17 species
previously thought to be area sensitive bred in buffers <20 m wide. In a
study of breeding birds in wooded riparian zones in Maryland and Del-
aware, Keller et al. (1993) found that Neotropical migrants were more
area sensitive than were either short distance migrants or residents. Neo-
tropical migrants increased in richness as corridor width increased, par-
ticularly in corridors <200 m, while richness of the other bird groups
remained relatively stable. Darveau et al. (1995) examined bird densities
in riparian boreal forest corridors of different widths (20-300 m) which
were bordered by recent (<2 yrs old) clearcuts. They found that forest-
breeding birds were sensitive to corridor width and concluded that 60-m
wide corridors were required to maintain forest breeding birds. In all of
these studies, the riparian zones were usually bordered by agricultural or
clearcut fields. Studies in which the riparian zone is bordered by pine

' The Nature Conservancy of Georgia, 1401 Peachtree St. NW. Suite 236. Atlanta, Georgia 30309.

2 National Biological Service, Patuxent Wildlife Research Center, Warnell School of Forest Resources,
Univ. of Georgia, Athens, Georgia 30602-2152.

496

Hodges and Krementz * BIRDS IN RIPARIAN FOREST STRIPS

497

silviculture are apparently lacking. The adjacent habitat could conceivably
ameliorate the species-area effect because the pine plantation might sup-
port some birds that would otherwise be absent in an agricultural or clear-
cut landscape.

In the southeastern United States, forested wetlands are being lost or
converted to pine silviculture at an alarming rate (Winger 1986, Hefner
et al. 1994). However, at present, land managers have no information
about how wide a buffer strip adjacent to streams is necessary to maintain
functional breeding bird communities. Understanding bird-habitat rela-
tionships is important both locally and over broad areas because forest
health may depend on the presence of breeding birds. Forest-dwelling
birds have been shown to control the numbers of insects feeding on tree
foliage (Marquis and Whelan 1994). Further, recreation dollars associated
with more natural forests are becoming a consideration in forest manage-
ment (Wiedner and Kerlinger 1990, Kerlinger 1993).

We investigated the relationship between width of bottomland hard-
wood forest corridors along the Altamaha River in Georgia and breeding
populations of Neotropical migratory birds. Our objectives were to ex-
amine the relationship between forest corridor width and bird species
richness, density, and the probability of encountering a particular bird
species.

STUDY AREA AND METHODS

We studied birds on the Altamaha River floodplain swamp, which begins at the confluence
of the Ocmulgee and Oconee rivers at river km 212 in Montgomery and Jeff Davis counties,
Georgia. These forests extend downstream to the Buffalo and Clayhole swamps at approx-
imately river km 24 in Glynn and McIntosh counties. The three dominant community types
sampled were; willow oak (Quercus phellos), overcup oak (Q. /vra/fl)-water hickory {Carya
aquatica), and bald cypress {Taxodium distichum)-'water tupelo (Nyssa aquatica) (Allard
1990). The first was a wet-mesic river floodplain forest dominated either by willow oak or
laurel oak (Q. laurifolia), with several other hardwoods and bald cypress occurring in the
canopy. The second was a wet river floodplain forest dominated in the canopy by overcup
oak and water hickory. The third, a forested riverine swamp, was dominated by bald cypress
and either water tupelo or Ogeechee lime (N. ogeche) in the canopy and was flooded for
most of the year. Some small stream corridors were sampled which were swamp forests
with the same floristic components as riverine bottomlands but with narrower zonation and
less frequent flooding.

Variable length transects were placed perpendicular to the river, based on corridor width,
shape, lack of disturbance, and presence of adjacent pine plantations. Riverine forest cor-
ridors were rejected if they had been extensively thinned or were in an early succe.ssional
stage (trees <8 cm dbh), excessively convoluted at the pine plantation edge, within 100 m
of other large forest tracts, or in close proximity to a major highway. Transects were located
^500 m apart. Three corridor-width classes were used: narrow (<350 m), medium (400-
700 m), and wide (>1000 m). Generally the Altamaha River meandered within a fairly
defined (about 1500 m wide) bottomland which was bordered by pine plantations. Thus,

498

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Neotropical Migratory Breeding Birds Known to Breed in South Atlantic Coastal
Plain Forests (Adapted from Gauthreaux 1992)“

Common Name (Scientific Name)

American Swallow-tailed Kite* {Elanoides forficatus)

Mississippi Kite {Ictinia mississippiensis)

Broad-winged Hawk (Buteo platyptenis)

Yellow-billed Cuckoo (Coccyzus americaniis)

Ruby-throated Hummingbird {Archilochus colubris)

Eastern Wood-Pewee* {Contopus virens)

Acadian Flycatcher (Empidonax virescens)

Great Crested Flycatcher (Myiarchus crinitiis)

Blue-gray Gnatcatcher {Polioptila caerulea)

Wood Thrush (Hylocichla miistelina)

White-eyed Vireo (Vireo griseus)

Yellow-throated Vireo {V. flavifrons)

Red-eyed Vireo (V. olivaceus)

Northern Parula {Panda americana)

Yellow-throated Warbler {Dendroica dominica)

Prothonotary Warbler {Protonotaria citrea)

Swainson’s Warbler {Limnothlypis swainsonii)

Louisiana Waterthmsh {Seiurus motacilla)

Kentucky Warbler {Oporornis formosus)

Hooded Warbler {Wilsonia citrina)

Summer Tanager {Piranga rubra)

“Those species marked with an asterisk (*) were not recorded in sampling for the present study, although they were noted

during other field work in the Altamaha River basin in 1993 or 1994.

when a wide corridor transect was chosen, the corridor width on the other side of the river
tended to be narrow and vice versa. The Altamaha River, which usually exceeded 100 m
in width, seemed to act as an effective barrier against cross stream movement of five of our
focal species (MFH, unpubl. data). The only exception was the Prothonotary Warbler (sci-
entific names in Table 1) which was observed frequently flying across the river. We selected
10 transect sites in 1993: four narrow (36-135 m), three medium (480-660 m), and three
wide (1320-2088 m). From one to seven points were located along these transects for a
total of 31 points. Points were located 50 m from the pine plantation edge and at 200 m
intervals thereafter. Most transects were located in forested riverine swamps and wet and
wet-mesic river floodplain forests. Exceptions included three narrow transects m which
relatively steep (>30% slope) mesic bluff forests were part of the forest comdor.

In 1994, we sampled six of the 10 transects used in 1993, and 16 additional transects
were added Two medium and two wide transects sampled in 1993 were deleted because of
marginal habitat characteristics. New tran.sects added in 1994 included 14 narrow transects
located in small stream corridors closely associated with the Altamaha River, one medium
transect, and one wide transect. Totals for 1994 were 18 narrow (36-330 m), two medium
(440—510 m), and two wide (1320—1512 m) transects, with 38 points surveyed.

We used the variable circular plot technique (Reynolds et al. 1980) to sample birds at
points along each transect. In 1993, each point was sampled three times during 16-31 May,

Hodges and Krementz • BIRDS IN RIPARIAN FOREST STRIPS

499

I 15 June and 16-30 June. Sampling was confined to the first two periods in 1994 MFH
was the only observer for all sampling. Order of sampling was stratified so that each point

was sampled at least once within 1.5 h after sunrise. All sampling occurred within 4 h after
sunrise.

All birds heard or seen during a 10-min sampling period were recorded, although only
Neotropical migrants were used in these analyses (Table 1 ). We summed the number of
species recorded at all points by corridor width to estimate species richness.

We recorded bird numbers at three distances: <25 m, 25-50 m, and >50 m. Each species
had a maximum detectable range determined by calculating the ratio of birds per unit area
within 50 m of the counting point to birds per unit area beyond 50 m and solving for the
unknown area. The cut-off points for the species used in density estimates were Acadian
Flycatcher— 70 m. Blue-gray Gnatcatcher— 50 m. White-eyed Vireo— 80 m. Red-eyed Vir-
eo — 80 m. Northern Parula — 80 m, and Prothonotary Warbler 100 m.

The structure of the plant community was sampled along each transect at five points at
10 m intervals on either side of each sampling point (10 points total for each sampling
point). Percent canopy cover (>3 m height) and shrub cover (1-3 m height) were estimated
using an ocular tube. The plant community type at each sampling point was deduced from
plant species lists compiled at each sampling point (Allard 1990). We estimated the timber
basal area at the sampling point with a cruising angle. Landsat TM and SPOT satellite
imagery were used to measure the width of the forest corridor from river to pine plantation
edge for each transect.

We estimated density per ha by corridor width and year for each species with sufficient
data. Data consisted of the single count (3 in 1993, 2 in 1994) which had the highest number
of birds detected. Densities of Neotropical migrants were estimated using the program DIS-
TANCE (Laake et al. 1993). Three models (uniform, half-normal, hazard-rate) were eval-
uated using likelihood ratio tests to determine which model best fit the data Once the final
model was selected, we tested for annual differences in densities of each species within
each corridor width by determining if the 95% confidence limits overlapped. We also tested
for annual effects by fitting a nested logistic regression (PROC LOGISTIC, SAS 1990)
model with corridor width and year effects versus corridor width only. The response variable

was presence/absence of each species. Model selection was based on Akaike’s information
criterion (Akaike 1973).

Sensitivity to corridor width was examined using logistic regression (PROC LOGISTIC,
SAS 1990). To produce a data set comparable to that of Keller et al. (1993) who sampled
at the center of each riparian corridor, we subsampled those data taken from the approximate
middle of each transect. At these points, a presence/absence data set was produced for each
species. This response variable was then modelled, using corridor width as a predictor. Based
on the parameter estimates, we calculated the estimated probability of a species occurring
m a corridor of given width (PROC LOGISTIC, SAS 1990:1076).

Next, we examined whether species density was more closely related to the vegetative
characteristics within the corridors or to corridor width alone. Because correlations were
noted among vegetation measurements (unpubl. data), we reduced the three vegetation vari-
ables (percent canopy cover, percent shrub cover, timber basal area) into two variables (PCI,
PC2) using principal components analysis (PROC PRINCOMP, SAS 1990). We then mod-
elled the response variable, density of each species at a point, with the predictor variables
corridor width, distance from the pine edge, PCI, and PC2 (PROC GLM, SAS 1990). To
insure that birds were not responding to inherent differences in the vegetation among cor-
ridor widths, we examined the relationships among coiridor width class and the predictor
variables PCI and PC2 by year and across years.

Finally, we examined the relationship between species richness and corridor width using

500

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

a generalized linear model (PROC GLM, SAS 1990) with year, eorridor width, and

year*corTidor width as predictors.

RESULTS

We detected 48 bird species on the surveys, of which 19 were Neo-
tropical migratory breeding birds (Table 1) Species richness ranged
from four to 15 across all corridor widths with 4-14 in narrow, 11 15
in medium, and 12-15 in wide corridors. The third sampling period m
1993 did not add any unique species to measurement of species richness
beyond those observed during the first two sampling periods. Species
richness varied by corridor width and year (F,, 3, = 3.48, P - 0.07) with
narrow corridors having lower mean species richness (7.8 ± 2.48 S )
than either medium (12.0 ± 0.71) or wide (12.6 ± 1.52) corridors (P <
0.05). More species were recorded in 1993 than in 1994 (F, 31 ■ ■.

P = 0.002). , ^ ^ ...

Six species occurred in sufficient numbers to estimate their densities.

Acadian Flycatcher, Blue-gray Gnatcatcher, White-eyed Vireo, Red-eyed
Vireo, Northern Parula, and Prothonotary Warbler. The highest counts
always occurred during the second survey, followed by the first survey,
and finally the third survey. The third survey in 1993 added little to the
results There were no differences in densities of any species between
years in either the medium or wide corridors (P > 0.05). Comparing all
narrow corridors, densities of Acadian Flycatcher, Blue-gray Gnatcatch-
er, White-eyed Vireo, and Northern Parula in 1994 were lower than m
1993 {P < 0 05). Examining just those narrow transects surveyed in
both years (versus examining all narrow transects surveyed in both
years) no differences in densities were found between years (P > 0.05).
Apparently, the new transects sampled in 1994 were responsible for the
overall differences in densities between years for the narrow corridors^
Thus, we believe that the differences between densities for the 1993 and
1994’ narrow transects reflected habitat differences rather than annual
differences. Those narrow transects added in 1994 were m small stream
corridors closely associated with the Altamaha River, unlike the exclu-
sively riverine corridors used in 1993. Logistic regression indicated that
only the White-eyed Vireo demonstrated a difference (P < 0.05) m
presence/absence between years, with fewer White-eyed Vireos being
seen in 1994. Based on the combination of these two tests, we believe
that there was no strong evidence for an effect of year on densities for
these six species. Thus we combined data from both years in subsequent

analyses. j

For all six species, densities were higher in medium comdors than 11

either narrow or wide corridors (Table 2). However, the differences m

Hodges and Krementz • BIRDS IN RIPARIAN FOREST STRIPS

501

Table 2

Density Estimates per Hectare and 95% Confidence Intervals of Six Focal Species
OF Neotropical Migratory Breeding Birds by Corridor Width (CW), Altamaha

River Basin, 1993 and 1994^

Species

CW

Est. density

Confidence interval

Acadian Flycatcher

N

0.08

0.05-0.15

M

0.10

0.08-0.12

W

0.05

0.04-0.07

Blue-gray Gnatcatcher

N

0.07

0.04-0.15

M

0.1 1

0.08-0.15

W

0.08

0.04-0.17

White-eyed Vireo

N

0.06

0.04-0.10

M

0.12

0.06-0.22*

W

0.06

0.04-0.09

Red-eyed Vireo

N

0.08

0.04-0.14

M

0.16

0.13-0.21

W

0.06

0.05-0.07

Northern Parula

N

0.09

0.05-0.16

M

0.22

0.16-0.31*

W

0.07

0.06-0.08

Prothonotary Warbler

N

0.03

0.01-0.05

M

0.07

0.04-0.13

W

0.05

0.04-0.07

An asterisk (*) denotes that the density estimate is significantly ^P < 0.05) different than either of the other density
imates ror fhar -snerif^c '

"N = narrow corridors; M = medium corridors; W = wide corridors.

density were significant {P < 0.05) only for White-eyed Vireos and
Northern Parulas.

Northern Parulas, White-eyed Vireos and Red-eyed Vireos exhibited
significant relationships between corridor width and probability of occur-
rence {P < 0.05, Fig. 1). The response to increasing corridor width was
most pronounced between 50 and 100 m. There was a greater than 80%
chance of five of the six species, excluding Prothonotary Warbler, being
detected in a 100 m wide corridor strip. Although not significantly so,
both the Blue-gray Gnatcatcher and the Prothonotary Warbler were more
likely to be found in wider corridors, whereas the Acadian Flycatcher was
less likely to be found in wider corridors (Fig. 1).

PCI and PC2 had eigenvalues >1 which is the usual cut-off point for
inclusion in analyses such as ours (Nichols 1977). PCI accounted for
36% of the variation in the three vegetation variables, while PC2 ac-
counted for 32% for a cumulative accounting of 68% of the variation
in vegetation variables. PCI was positively associated with timber basal
area and canopy cover and negatively associated with shrub cover. We

502

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Forest Corridor Width (m)

Fig 1 Probability of detecting six focal species of migratory breeding birds along the
Altamaha River riparian forest corridor. Solid circle— Acadian Flycatcher, square— Blue-
gray Gnatcatcher, triangle (point up)— Northern Parula, triangle (point down)— Prothonotary
Warbler diamond— Red-eyed Vireo, hexagon— White-eyed Vireo. Points represent esti-
mated probabilities of detection at 50, 100, 200, 500, and 1000 m corridor widths.

interpret this score, when high, as reflecting forest stands with a rela-
tively closed canopy and little undergrowth. PC2 was positively asso-
ciated with canopy cover and shrub cover reflecting forests with closed
canopies and more undergrowth. Species-specific modelling revealed
that Prothonotary Warbler and Red-eyed Vireo densities were positively
related to PCI {P < 0.05), while White-eyed Vireo density was nega-
tively related to PCI {P < 0.05). In addition, the density of Red-eyed
Vireos was related positively to the distance from the pine-plantation
edge {P < 0.05), while White-eyed Vireo abundance was positively

related to corridor width {P < 0.05).

We found that vegetation structure differed by corridor width in 1993
(P^ = 4.02, P = 0.04) but not in 1994 (F2,37 = 2.64, P = 0.09). In 1993,

the difference in vegetation structure was due to PC2, the amount of shrub
cover. Combining years, there was no difference in vegetation structure

(F2,54 = 1-97, P = 0.15).

Hodges and Krementz • BIRDS IN RIPARIAN FOREST STRIPS

503

DISCUSSION

The Neotropical migratory breeding bird community that we sampled
along the Altamaha River was typical of what would be expected in the
southeastern United States (Table 1; Hamel 1992). Only two species of
Neotropical migrants were absent from our surveys, American Swallow-
tailed Kite and the Eastern Wood-Pewee. The former probably has one
of the lowest densities among breeding birds of Altamaha River bottom-
land forests, while the latter is not found in bottomland habitats in our
area, preferring mature pine woodlands (MFH, pers. obs., Hamel 1992).
Although the bird community that we monitored was typical, we point
out that the Altamaha River and its associated riparian zone are not typical
of the southeastern United States. The Altamaha River has the largest
watershed in the Southeast (Anonymous 1986) and is famous for its ex-
tensive associated tracts of unbroken bottomland hardwoods (Wharton et
al. 1982). Clearly these forested tracts are unusual today, and so our
findings may not be entirely applicable to other riverine systems in the
Southeast.

Species richness values by corridor width increased with area as Keller
et al. (1993) found. The year effect we observed was, in part, a result of
incorporating small stream corridors in 1994 which had, across the board,
lower species richness values. Despite these differences in habitat types,
all corridor widths had lower species richness in 1994; we are unsure
why.

Of the six species for which we could estimate the probability of de-
tection, five exhibited a trend towards increasing detection with increasing
forest corridor width, three significantly so. Four of these species, Acadian
Flycatcher, Prothonotary Warbler, White-eyed Vireo, and Red-eyed Vireo
showed the strongest relationship between probability of detection and
increasing forest corridor width (Keller et al. 1993). Keller et al.’s (1993)
evidence for the Acadian Flycatcher was opposite from ours though, as
we found Acadian Flycatchers tended to decrease in abundance with in-
creasing corridor width (although not significantly). Robbins et al. (1989)
and Dawson et al. (1993) both found the Acadian Flycatcher to be area
sensitive (positively) in non-riparian forested situations in Maryland. The
Prothonotary Warbler was area sensitive (positively) for Keller et al.
(1993) but not for Robbins et al. (1989) or us.

The relationship observed between densities of the six species and cor-
ridor width was unexpected. We had anticipated that the highest densities
would have occurred in the widest conidors, assuming that the widest
corridors would have contained not only more total habitat but also habitat
of higher quality. Our finding that vegetation structure was not consis-

504

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

tently related to corridor width indicated that wider corridors were not
structurally different than narrower corridors. Despite the lack of differ-
ences in habitat structure that we measured, the differences in densities
suggest either that differences in habitat quality existed, which we did not
measure, or behavioral factors excluded some birds from the widest cor-
ridors. As Van Horne (1983) and Pulliam (1988) have asserted, high an-
imal densities may result from excessive numbers of immigrants who
have been forced away from the preferred habitat by higher ranking (more
dominant) individuals. If true, then the lower densities in the widest cor-
ridors might reflect a predominance of more successful, more dominant
individuals who benefit by controlling densities.

In the generalized linear model examining sensitivity to corridor width,
vegetation parameters, and distance to pine edge. White-eyed Vireo was
the only species whose numbers clearly showed a positive association
with corridor width. During the breeding season, this species is not re-
stricted to forest interiors; it frequently uses young second growth and
other shrubby habitats. White-eyed Vireo’s negative associations with
canopy density and basal area and positive association with shrub density
are easier to predict (Hamel 1992). Other studies (Robbins et al. 1989)
have noted sensitivity to forest area for species common in our area, such
as Acadian Flycatcher and Northern Parula. The much larger data set of
Robbins et al. (1989) may have contributed to their better rate of detection
of significant positive associations with forest area. Red-eyed Vireo num-
bers showed a positive association with distance from the pine-plantation
edge, indicating a preference for forest interiors, and a preference for
closed canopies, the latter having been noted previously (Robbins et al.

1989).

The rapid increase in probability of occurrence and species richness
with increasing corridor width and the apparent asymptotic shape of the
species abundance-area curves suggests that forest corridors of about 100
m should be sufficient to maintain functional assemblages of the six most
common species of breeding neotropical migratory birds. This guideline
exceeds the 60 m recommended by Darveau et al. (1995), but comparing
boreal forests and southeastern bottomland hardwoods may not be appro-
priate. Our guideline does concur with the findings of Keller et al. (1993)
who also recommended a 100 m wide corridor be maintained. Our rec-
ommendation does not take into account the needs of the least common
species of Neotropical migratory birds encountered, which we were not

able to ascertain from these data. • . • j

The value of corridors goes beyond the maintenance of bleeding bir

communities, as Naiman et al. (1993:209) concluded that “effective ri-
parian management could ameliorate many ecological issues related to

Hodges and Krementz • BIRDS IN RIPARIAN FOREST STRIPS

505

land use and environmental quality.” More specifically, Winger (1986)
demonstrated that functional forested corridors assimilate nutrients and
organic matter, hasten the degradation of persistent pesticides and de-
crease the bioavailability of heavy metals. Justification of streamiside cor-
ridors thus goes beyond breeding bird communities alone (see also Brin-
son et al. 1981).

ACKNOWLEDGMENTS

We thank R. Goodloe, C. S. Robbins, D. K. Dawson, and an anonymous reviewer for
their comments on this manuscript. Funding was provided by The Nature Conservancy, the
U.S. Fish and Wildlife Service, and the National Biological Service.

LITERATURE CITED

Akaike, H. 1973. Information theory and an extension of the maximum likelihood prin-
ciple. Pp. 267-281 in Int. symposium on information theory. Second ed. (B. N. Petran
and F. Csaki, eds.). Akademiai Kiadi, Budapest, Hungary.

Allard, D. 1990. Southeastern United States ecological community classification. Version
1.2. The Nature Conservancy Southeastern Regional Office, Chapel Hill, North Caro-
lina.

Anonymous. 1986. Water availability and use report, Altamaha River basin. Draft. Georgia
Environmental Protection Division. Atlanta, Georgia.

Brinson, M. M., B. L. Swift, R. C. Plantico, and J. S. Barclay. 1981. Riparian ecosys-
tems: their ecology and status. FWS/“OBS-81/17. U.S. Fish Wildl. Serv., Biological
Serv. Program, Washington, D.C.

Darveau, M., P. Beauchesne, L. Belanger, J. Huot, and P. Larue. 1995. Riparian forest
strips as habitat for breeding birds in boreal forest. J. Wildl. Manage. 59:67-78.
Dawson, D. K., L. J. Darr, and C. S. Robbins. 1993. Predicting the distribution of breeding
forest birds in a fragmented landscape. Trans. N. Am. Wildl. Nat. Resour Conf 58 35-
43.

Gauthreaux, S. 1992. Preliminary lists of migrants for Partners In Flight neotropical
migratory bird conservation program. Partners In Flight 2:30.

Hamel, P. B. 1992. Land manager’s guide to the birds of the South. The Nature Conser-
vancy, Chapel Hill, North Carolina.

Hefner, J. M., B. O. Wilen, T. E. Dahl, and W. E. Frayer. 1994. Southeast wetlands;
status and trends, mid-1970’s to mid-1980’s. U.S. Dept. Int., Fish and Wildl. Serv.,
Atlanta, Georgia.

Keller, C. M. E., C. S. Robbins, and J. S. Hatfield. 1993. Avian communities in riparian
forests of different widths in Maryland and Delaware. Wetlands 13:137-144.

Kerlinger, P. 1993. Birding economics and birder demographics studies as conservation
tools. Pp. 32-38 in Status and management of neotropical migratory birds. (D. M. Finch
and P. W. Stangel, eds.). USDA Forest Serv. Gen. Tech. Report RM-229.

Laake, j. L., S. T. Buckland, D. R. Anderson, and K. P. Burnham. 1993. DISTANCE
user’s guide V2.0. Colorado Cooperative Eish Wildlife Research Unit, Colorado State
Univ., Fort Collins, Colorado.

Lynch, J. E and D. J. Whigham. 1984. Effects of forest fragmentation on breeding bird
communities in Maryland, USA. Biol. Conserv. 28:287-324.

Marquis, R. J. and C. J. Whelan. 1994. Insectivorous birds increase growth of white oak
through consumption of leaf-chewing insects. Ecology 75:2007-2014.

506

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Naiman, R. J., H. Decamps, and M. Pollock. 1993. The role of riparian corridors in
maintaining regional biodiversity. Ecol. Appl. 3:209—212.

Nichols, S. 1977. On the interpretation of principal components analysis in ecological

contexts. Vegetatio 34:191-197.

Pulliam, H. R. 1988. Sources, sinks, and population regulation. Am. Nat. 132:652-661.
Reynolds, R. T, J. M. Scott, and R. A. Nussbaum. 1980. A variable circular plot method

for estimating bird numbers. Condor 82:309-313.

Robbins, C. S., D. K. Dawson, and B. A. Dowell. 1989. Habitat area requirements of
breeding fwest birds of the Middle Atlantic States. Wildl. Monogr. 103:1-34.

SAS Institute, Inc. 1990. SAS user’s guide: Version 6.02. SAS Inst., Inc., Cary, North
Carolina.

Stauffer, D. E and L. B. Best. 1980. Habitat selection by birds of riparian communities:

evaluating effects of habitat alterations. J. Wildl. Manage. 44:1-15.

Van Horne, B. 1983. Density as a misleading indicator of habitat quality. J. Wildl. Manage.

47:893-901.

Wharton, C. H., W. M. Kitchens, and T. W. Sipe. 1982. The ecology of bottomland
hardwood swamps of the Southeast: a community profile. U.S. Fish Wildl. Serv.,
FWS/OBS-81/37.

Whitcomb, R. E, C. S. Robbins, J. E Lynch, B. L. Whitcomb, M. K. Klimkiewicz and D.
Bystr'ak. 1981. Effects of forest fragmentation on avifauna of the eastern deciduous
forest. Pp. 125-206 in Forest island dynamics in man-dominated landscapes. (R. L.
Burgess and D. M. Sharpe, eds.). Springer- Verlag, New York, New York.

WiEDNER, D. AND P. Kerlinger. 1990. Economics of birding: a national survey of active
birders. Am. Birds 44:209—213.

Winger P. V. 1986. Forested wetlands of the Southeast: review of major characteristics
and role in maintaining water quality. U.S. Fish Wildl. Serv., Resour. Publ. 163.

Wilson Bull., 108(3), 1996, pp. 507-515

DAWN AND DUSK SINGING OF MALE
AMERICAN ROBINS IN RELATION
TO FEMALE BEHAVIOR

Tore Slagsvold

Abstract. It has been assumed that females are particularly fertile during the first hour
after laying when the next egg to be laid is fertilized. In many passerine birds, egg laying
occurs early in the morning. Hence, it may be particularly important for males to sing at
dawn to attract the mate and other females and to repel potential cuckolders. I studied song
activity of mated male American Robins (Turdus migratorius). Egg laying occurred close
to noon. However, males had a peak song activity at dawn but sang little around noon.
Hence, the idea that mated males sing primarily to deter other males just before the ‘fertil-
ization window’ was not supported. Male song activity tended to increase when the mate
visited the nest during the day but less so when she entered the nest to roost at night.
Synchronous emergence of all females at dawn resulted in synchronous termination of the
dawn chorus, whereas a more asynchronous pattern of nest visits by females during the day
and in the evening resulted in asynchronous and scattered periods of song. Received 21
Nov, 1995, accepted 22 Feb. 1996.

In the breeding season, passerine birds typically have a peak song ac-
tivity at dawn and dusk (Armstrong 1973) and, in addition, egg laying
often occurs early in the morning. Females may be particularly fertile
during the first hour after laying, when the next egg to be laid is fertilized
(Birkhead and Mpller 1991). Hence, males may be singing at dawn to
attract the mate and other females and to repel potential cuckolders (Mace
1986, 1987a; Cuthill and MacDonald 1990; Mpller 1991). Such behavior
has been reported in two hole nesting species, the Great Tit {Pams major,
Mace 1986, 1987a, b) and the Willow Tit {P. montanus'. Welling et al.
1995), although alternative explanations for dawn singing in these species
also exist (see Slagsvold et al. 1994). I examined the possible relationship
between song activity of male American Robins (Turdus migratorius) and
female roosting behavior at dawn and dusk and female nest visiting be-
havior near the time of egg laying. Robins are interesting for several
reasons. Egg laying does not occur at dawn but takes place in the middle
of the day (Weatherhead et al. 1991, Scott 1993). Males guard their mates
when fertile, suggesting that extra-pair copulations (EPCs) occur (Gowaty
and Plissner 1987). I discuss how the results may help understand dif-
ferences between species in daily variation of male song activity.

Dept, of Biology, Univ. of Oslo, RO. Box 1050. Blindern. N-0316 Oslo, Norway.

507

508

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

METHODS

The study was conducted during 13-28 April 1994, in a mixed deciduous-coniferous
woodland in St. Edward State Park near Seattle, Washington. The territories of 12 mated
males were visited 20-70 min before the males started singing at dawn. In 10 cases, ob-
servations continued until noon. Observations were also made at 10 territories from 18:30
h until dark. Male song was recorded from a distance of 20-60 m, and one or two males
were observed during each visit. Song activity was expressed as the percentage of minutes
in which the male was heard singing (one or more times) of the total number of minutes
of observation. Males were not banded but were identified by their proximity to the nests
and to the females, their movements, song posts, and roosting sites.

Female robins may spend the night on the nest during the egg-laying period (Brackbill
1985, Slagsvold unpublished data). Video was used to record female presence on the nest,
except at dawn when it was too dark. At dawn, the camera was not placed near the focal
nest until the female had left. A Sony Handycam (CCD-TR805E) on a tripod was placed
3-8 m from the nest and pointed towards it. Female departure from the nest at dawn was
observed in seven cases using lOX binoculars. In addition, the time of first appearance at
the nest at dawn was recorded for another female that did not roost m the nest. Hence, I
was able to examine the possible relationship between song activity and female emergence
time for eight males. Three females were in the egg-laying period, four laid the final egg
on the day of observation, and one was two days after termination of laying. Relative to
the time of civil twilight, time of female emergence was not related to stage of breeding or
date of observation (data not shown), and so the data were pooled. Civil twilight and sunrise
and sunset times were for Seattle (Federal Office Building, Pacific Standard time, adjusted
by one hour to represent summer time; “Climatology of the United States,” no. 40-45,

Washington, D.C., 1961). .

The approximate time of egg laying was recorded by inspecting nests at variable hours

of the day when the female was absent, and also by analyzing at which times the female
was on and off according to the videofilms. Day 1 is defined as the day when the first egg
was laid. Mean clutch size was 3.2 eggs (range 3-4, N = 9). Observations of dawn and
dusk singing were carried out from Day -4 to Day 5. One nest was lost before egg laying
had begun, so the precise stage of breeding was unknown.

RESULTS

In the morning, females left the nest/roost between 05:35-05:57 h {x
= 05:50 ± 8 min [SD], N = 8), which was 9-19 min (x = 13 ± 3, N
= 8) after civil twilight and 16—25 min (jf — 21 2: 3, N 8) before
sunrise. After emergence, females were seen feeding on the ground. Those
that had finished egg laying, or that laid the final egg later that morning,
entered the nest to incubate earlier after sunrise (jc = 3 ± 5 mm, N^- 3)
than those that were still in the laying period {x = 142 ± 114, N^- 4; z
= -2.31, P = 0.021, U-test). The females spent 10-60 min (jc - 31 ±
20, N = 6) on the nest on their first visit, followed by a period off of
10-105 min (jc = 34 ± 34, N = 7) before a new period on. Hence, the
synchronous behavior of the females at dawn, as measured by their emer-

gence times, soon disappeared.

In the evening, females entered the nest to roost between

19:30-20:20

Slagsvold • DAWN SINGING IN AMERICAN ROBINS

509

h (.V = 20:01 ± 17 min, N = 8), which ranged from 33 before to 13 min
after sunset (.f = 4 ± 17, N = 8) and 20—67 min before civil twilight (jc
— 37 ± 17, N — 8). Variation among females was significantly greater
for evening roosting times than for dawn emergence times, when consid-
ered relative to the times of sunset and sunrise, respectively (F = 32.9,
^ ^ 0.001, Variance ratio test). Before entering the nest to roost
at night, the females spent 10-38 min off (.v = 18 ± 10, N = 7). Before
this last period off, they had spent 10-48 min (x = 26 ± 15, N = 7) on
the nest.

Assuming that laying occurred halfway between two nest inspections
(Weatherhead et al. 1991), and using the median value for each female
in case of observations of more than one egg laid (1 — 3 eggs per female)
the median time of laying was 1 1 :50 h (range 10:43-15; 13, N = 9; .v =
12.32 ± 88 min, if using mean of mean values). Including only cases
when the interval between two nest inspections was less than 5 h, the
median was 11:25 h (range 10:43-12:39, N = 6; x = 11:30 ± 42 min,
if using mean of mean values). Using a similar method, an average laying
time of 11:32 EOT was found in eastern Ontario, Canada (Weatherhead
et al. 1991).

One copulation was observed. It occurred on the ground 8 m from the
nest at 06:05 h on 18 April, 10 min after the female left the nest. Between
09:04 and 12:00 h on the same day, the female laid the third egg of a
clutch of four. At another nest, a male tried to copulate with a female
when she was sitting on the nest rim (at 08:23 h; recorded on video), but
she moved away. Between 1 1:12 and 12:05 h on the same day, she’laid
the second egg of a clutch of three.

A peak in song activity occurred at dawn, with little singing thereafter
(Fig. 1). One male did not sing at all at dawn before the female left the
nest. Excluding this male, males started singing from 42 min before to 7
min after civil twilight (x = 12 ± 17 min before civil twilight, N = 11),
or 5-56 min (x = 24 ± 21, N = 8) before the mate left the nest. Relative
to the time of civil twilight, onset of dawn singing was not related to
stage of breeding (Spearman, r, = -0.16, P = 0.62, N = 11; stage of
breeding measured as the number of days elapsing until the day of the
final egg laid; data ranging from five days before this date to two days
after). Because some males did not start singing until soon before female
emergence, song activity did not peak until just before the female left the
nest (Fig. 2). Song activity was low after female emergence (Fig. 2).
Comparing a 10 min period before and after female emergence, song
activity dropped for seven of eight males (no difference in one case; z =
-2.39, P = 0.017, N = 7, Wilcoxon matched pairs test). When males
started singing, the approximate mean distance between the nest and the

510

THE WILSON BULLETIN • Vol. 108. No. 3, September 1996

Lig. 1. Mean (±SE) song activity of ten male American Robins, calculated for consec
utive 30-min periods from dawn to noon.

perch (distance along the ground) was 33 m (SD - 9, range 20 50, N
= 11). Distance was not related to stage of breeding (r^ = 0.03, P -
0 92, N = 11). No males were seen flying to a nest before female emer-
gence. However, later in the day (between 06:57-19:29 h), at least six
males visited the nest without the females being present, and two males
visited the nest to feed the mate.

Song activity was low between dawn and noon (Fig. 1) despite the fact
that most eggs were laid around noon. I compared the data from all
periods when the female was on the nest (1-6 periods combined for each
male, jc = 119 min of observation) with the data from all periods when
the female was off the nest (1-7 periods combined for each male, x -
136 min of observation). Males (N = 9) were observed singing during

Slagsvold • DAWN SINGING IN AMERICAN ROBINS

511

Time since female left roost (min)

Fig. 2. Mean ( + SE) song activity of eight male American Robins, calculated for con-
secutive 10-min periods before and after the female left her night roost at dawn.

0-17% (x - 5 ± 6) of the minutes when a mate was on the nest, and 0-
4% (x — 1 —2.) when she was off the nest. The difference was nearly
significant (z = —1.86, N = 7, f* = 0.063, Wilcoxon matched pairs test;
two males did not sing at all).

Song activity of mated males was low in the evening (Fig. 3). After
18:30 h, three of the ten males included in Fig. 3 did not sing at all. The
other seven males had only one or two periods of song during that period
(separated by at least 10 min without song). The periods of singing lasted
only 1-9 min (T = 4 ± 3, N = 7). When the female entered the nest to
roost at night, two of seven males immediately started singing for eight
and nine minutes, respectively. These were the longest song periods re-
corded of any mated male in the evening. The seven males that did sing
after 1830 h stopped singing from 84 min before to 20 min after sunset
(x = 21 ±35 min before sunset) and 17 to 117 min before civil twilight
(± = 54 ± 35).

Male (N = 7) song activity was recorded after 18:30 h and before the

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

512

Lig. 3. Mean ( + SE) song activity of ten male American Robins, calculated for consec
utive 10-min periods in the evening.

final roosting time of the mate when the female was on the nest (1 3
periods combined per female, t = 36 min of observation) and when off
the nest (1-3 periods combined per female, t = 51 min of observation).
Song activity (percentage of minutes with male heard singing) was low
in both cases (female on the nest: x — 1.1 — 1-9; female off the nest, x
= 0.6 ± 1.7; z = -0.45, P = 0.65, Wilcoxon matched pairs test).

DISCUSSION

Two important results were found. First, mated male robins exhibited
peak song activity at dawn but sang little around the time of egg laying
at noon. Hence, the hypothesis that mated males sing primarily to deter
other males just before the “fertilization window” (Mace 1987a) was not
supported. Second, song activity appeared to be influenced by female
roosting behavior. Singing rates declined when the female left the night

roost.

The principal function of dawn singing in mated male robins seems to
have something to do with the female’s being on the nest. Why should
males be singing then? Dawn song may have several functions (see re-
view in Slagsvold et al. 1994): (1) serve as pair-bond reinforcement and
stimulate hormonally-mediated aspects of breeding in females; (2) tell the
mate that she can emerge from the roost and nest without risk of predation
and of revealing the location to nest predators; (3) attract the mate for

Slagsvold • DAWN SINGING IN AMERICAN ROBINS

513

copulation; (4) attract a new mate in case the former mate has disappeared
during the night; (5) attract other females for EPCs; (6) deter other males
and so avoid that they will try to copulate with the mate; and (7) defend
the nest site and the territory. The hypotheses are not mutually exclusive
and may work in concert (Mace 1987b, Slagsvold et al. 1994). Further
studies are needed to single out the relative importance of each.

The fact that male robins sang little close to the time of egg laying
does not exclude the possibility that song is used as a mate guarding
tactic. The only copulation seen occurred soon after the female left the
roost at dawn, and so dawn singing may be related to this event. Although
Weatherhead et al. (1991) suggested that robins lay eggs around noon
because this may be a suitable time for copulation, no observations of
copulations were reported. Gowaty and Plissner (1987), studying mate
guarding in robins in the fertile period, observed copulations at all hours
of the morning except for the hour after dawn; no observations were made
after noon. However, few copulations were observed, and the authors did
not report the actual frequencies. Further studies are needed to see if dawn
is the regular time of within-pair copulation in robins as has been ob-
served in some other passerine birds (Mace 1987a, Birkhead and Mpller
1991, Davies 1992). In birds, females may actively seek FPCs with males
of high quality (e.g., Kempenaers et al. 1992). No information exists on
when and where FPCs take place in robins. However, as found in another
passerine, the Great Reed Warbler (Acrocephalus arundinaceus; Hasse-
Iquist et al. 1995), they may prefer extra-pair mates with large song rep-
ertoires. In European Starlings (Sturnus vulgaris), females would only
solicit copulation if the social or extra-pair mate sings (Fens and Pinxten
1990).

My results support those of others (e.g., Mace 1986, 1987a; Cuthill
and MacDonald 1990; Part 1991; Otter and Ratcliffe 1993; Slagsvold et
al. 1994; Welling et al. 1995) that male song activity is influenced by
female behavior. Male robins start dawn singing about an hour or less
before female emergence, as do male Great Tits, even though female
Great Tits leave the roost, on average, 49 min later than female robins
when measured relative to civil twilight (Slagsvold et al. 1994). In both
species, males may increase song activity if the mate enters the nest dur-
ing the day and at dusk. Hence, knowledge of factors that influence fe-
male roosting behavior may help explain differences in daily variation of
male song activity within and between species of passerine birds. For
instance, the fact that most of the males within a particular species, like
the robin, stop singing more or less at a constant time in relation to light
intensity at dawn, may not necessarily be related to a sensitivity of males
to light per se but occur because this is the time when their mates leave

514

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

their respective roosts. When the emergence at dawn of female Great Tits
was experimentally delayed, males continued singing for a longer period
(Mace 1986).

Female robins left the roost, on average, 13 min after civil twilight,
and the variation among females was low (range 9-19 min). At that time
of the day, light intensity changes very fast, apparently synchronizing
emergence. A more synchronous emergence in female robins than in fe-
male Great Tits (see Mace 1987b, Slagsvold et al. 1994) may also be due
to the fact that robins do not lay an egg before leaving the night roost in
contrast to the tits. Later in the morning, female behavior was more asyn-
chronous, as were egg laying times, resulting in asynchronous periods of
singing by males. In some other passerines, egg laying occurs in mid-
morning and may be more synchronous than in robins, resulting in stron-
ger synchronization of male song activity. In support of this idea, song
activity of many passerine birds drops after a peak at dawn but reaches
a new but lower peak later in the morning (Klockars 1941, Gyllin 1967,
Lomholt 1971). One explanation for the initial drop is that males need to
forage after dawn singing (Klockars 1941, Lomholt 1971). However, the
drop may also reflect a more synchronous mate guarding by males early
than late in the morning.

I conclude that to understand further daily variation in male song ac-
tivity in passerine birds, more attention should be paid to the role of the
females and, hence, on the factors that influence their copulation and
roosting behavior. Information is needed concerning female emergence
times at dawn and dusk and on the degree of synchronization of these
events between females.

ACKNOWLEDGMENTS

I did this research during sabbatical year at the Burke Museum, Univ. of Washington. I
thank S. Rohwer for arranging the stay in Seattle, J. Fillers for permission to work in St.
Edward State Park, D. Hartmann for climatic data, P. K. Slagsvold for assistance in the
field, and H. Lampe, G.-P. Stetre, and K. Yasukawa for comments on the manuscript. The
study was supported by a grant from the Norwegian Research Council.

LITERATURE CITED

Armstrong, E. A. 1973. A study of bird .song. Dover, New York, New York.

Birkhead, T. R. and A. P. Moller. 1991. Sperm competition in birds. Evolutionary causes
and consequences. Academic Press, London, England.

Brackbill, H. 1985. Initiation of nest-roosting by passerines with open nests. J. Eield

Ornithol. 56:7 1 .

CUTHILL, I. AND W. A. MacDonald. 1990. Experimental manipulation of the dawn and
dusk chorus in the blackbird Turdus merula. Behav. Ecol. Sociobiol. 26:209-216.
Davies, N. B. 1992. Dunnock behaviour and social evolution. Oxford Univ. Press, Oxford,
England.

Slagsvold • DAWN SINGING IN AMERICAN ROBINS

515

Eens, M. and R. Pinxten. 1990. Extra-pair courtship in the starling Sturnus vulgaris. Ibis
132:618-619.

Gowaty, R a. and J. H. Plissner. 1987. Association of male and female American Robins
(Turchis migratorius) during the breeding season: paternity assurance by sexual access
or mate-guarding. Wilson Bull. 99:56-62.

Gyllin, R. 1967. Dygnsrytm hos kornsparven (Einberiza calandra). Var Eagelvarld 26'
19-28.

Hasselquist, D., S. Bensch and T. V. Schantz. 1995. Low frequency of extrapair paternity
in the polygynous great reed warbler, Acrocephalus arundinaceus. Behav. Ecol. 6:27—38.

Kempenaers, B., G. R. Verheyen, M. Van Den Broeck, T. Burke, C. Van Broeckhoven,
AND A. A. Dhondt. 1992. Extra-pair paternity results from female preference for high-
quality males in the blue tit. Nature (London) 357:494-496.

Keockars, B. 1941. Studier over fagelsangens dagsrytmik. Ornis Eenn. 18:73-110.

Lomholt, J. P. 1971. lagttagelser over gulspurvens {Emberiza citrinella) sangaktivitet.
Dansk Orn. Foren. Tidsskr. 65:179-187.

Mace, R. 1986. Importance of female behaviour in the dawn chorus. Anim. Behav. 34:
621-622.

. 1987a. The dawn chorus in the great tit Parus major is directly related to female
fertility. Nature (London) 330:745-746.

. 1987b. Why do birds sing at dawn? Ardea 75:123-132.

M0LLER, A. P. 1991. Why mated songbirds sing so much: mate guarding and male an-
nouncement of mate fertility status. Am. Nat. 138:994-1014.

Otter, K. and L. Ratcliffe. 1993. Changes in singing behavior of male black-capped
chickadees (Parus atricapillus) following mate removal. Behav. Ecol. Sociobiol. 33:
409-414.

Part, T. 1991. Is dawn singing related to paternity insurance? The case of the collared
flycatcher. Anim. Behav. 41:451-456.

Scott, D. M. 1993. On egg-laying times of American Robins. Auk 110:156.

Slagsvold, T, S. Dale and G.-P. S/etre. 1994. Dawn singing in the great tit (Parus major)-.
mate attraction, mate guarding, or territorial defence? Behaviour 131:115-138.

Weatherhead, P. j., R. D. Montgomerie and S. B. Mcrae. 1991. Egg-laying times of
American robins. Auk 108:965-967.

Welling, R, K. Koivula and L. Lahti. 1995. The dawn chorus is linked with female
fertility in the Willow Tit Parus montanus. J. Avian Biol. 26:241-246.

Wilson Bull., 108(3), 1996, pp. 516-523

BREEDING BIOLOGY OF THE CRESTED CARACARA

IN SOUTH TEXAS

Vanessa M. Dickinson'-^ and Keith A. Arnold'

Abstract. — We studied the breeding biology of six nesting pairs of Crested Caracaras
(Caracara plancus) from January to August 1989 in Austin and Colorado Counties, south
Texas. Four of the pairs nested in Macartney rose {Rosa bracteata). All nests were built
below the nest-support canopy. We found caracaras laying eggs between 17 January and 23
June. Eggs hatched from February to April, and in June for two second nesting efforts.
Young from successful first nesting efforts fledged from April to June. By August we did
not see young or adults in the natal area. Nest building and courtship averaged 21 days (N
= 2 pairs). Incubation periods averaged 30 days (N = 4 pairs), nestling dependency periods
averaged 56 days (N = 5 pairs), and post-fledgling dependency periods averaged 33 days
(N = 4 pairs). We believe that two of the pairs each laid a second clutch in June, but the
newly-hatched chicks were killed by red imported fire ants {Solenopsis invicta). Overall
nesting success was 45.7%. Success for first nesting attempts was 72.6%. Received 7 Sept.
1995, accepted 13 Feb. 1996.

In the United States, the Crested Caracara {Caracara plancus) primarily
is found in Texas, Florida, and Arizona (Palmer 1988). The Florida pop-
ulation is listed as threatened under the Endangered Species Act (U.S.
Fish and Wildlife Service 1987). Natural history of the Crested Caracara
in North America has been described (Bent 1938, Brown and Amadon
1968, Oberholser 1974), but there has been little in-depth study of the
species. Knowledge of the Texas population is limited to Oberholser
(1974) and several brief notes. In general, little is known about the Crest-
ed Caracara’s breeding biology. The lack of data on breeding biology
precludes ability to make proper evaluations of population status and
trends. Our objective was to describe the breeding biology, egg measure-
ments, and nesting structures of the Crested Caracara in south Texas.

METHODS

We studied nesting pairs of Crested Caracaras in Austin and Colorado Counties, south
Texas, from January 1989 to August 1989. The study area was centered on the Attwater
Prairie Chicken National Wildlife Refuge (APCNWR), Colorado County (29°40'N,
96°15'W). The study area was at the western boundary of the Gulf Coast Prairie and the
southern boundary of the Post Oak Savannah. The climax vegetation is tallgrass prairie
characterized by big bluestem (Andropogan gerardi) and Indiangrass {Sorghastrum nutans)
with overstory trees such as post oak {Quercus stellata) and blackjack oak {Q. niarilandica)
(Gould 1975). Invading shrubs include Macartney rose {Rosa bracteata), dewberry {Rubus
trivialis), and yaupon {Ilex vomitoira) (Gould 1975).

' Dept of Wildlife and Fisheries Sciences, Texas A&M Univ., College Station, Texas 77843.

2 Present address: Arizona Game and Fish Dept., 2221 West Greenway Road, Phoenix, Arizona 85023.

516

Dickinson and Arnold • BREEDING CRESTED CARACARAS

517

We located nests through aerial and ground surveys. Once we located a nest, we recorded
the following information: nest-tree species and height, height of nest above ground, location
of nest with respect to nest-tree canopy, and nest dimensions (length, width, internal bowl
depth). When a nest was under construction or refurbishment, we recorded the materials
used (species where possible), participation of each sex in nest-building, and duration of
construction or refurbishment. We defined a nest as refurbished if we observed pairs adding
material to an existing nest.

We divided the breeding season into courtship, incubation, nestling, and post-fledging
dependency periods. The onset of courtship was the first day we observed a pair copulating.
We defined post-hedging dependency as the period when juveniles were fed by the adults
within 0.8 km of the nest. The end of the post-fledging dependency period was the date the
young were last seen under these criteria. When we located a nest, we checked for eggs or
nestlings. If there were no eggs or nestlings, we inspected nests daily with a mirror to
determine onset of egg-laying. We determined the laying sequence by marking eggs with a
soft lead pencil. We inspected each nest once a week to record any losses. At the completion
of egg-laying, we recorded mass, length, and width of each egg.

Once we located a nest, we began observations of daily behavior of nesting pairs from
sunrise to sunset. We observed each pair one day a week during the entire breeding season.
Beginning three days before hatching (25 days after the onset of incubation) and three days
before fledging (8 weeks after the onset of incubation), we inspected nests daily to record
the initiation and sequence of hatching and fledging, respectively.

After the young hatched, we monitored nests at one-week intervals to check nestling
survival and to record when young fledged. Nestling ages were estimated from plumage
characteristics as described by Bent (1938) and Oberholser (1974). To facilitate monitoring
young after fledging, we marked nestlings at approximately six weeks of age. Measurements
and mass were taken for each nestling, and a U.S. Fish and Wildlife Service band was
attached to its right leg. Nestlings were measured with a 24-cm ruler and a Mitutoya Model
550-633 caliper as described by Baldwin et al. (1931). Young were color-marked with a 7
cm-long blue, white, or pink vinyl plastic streamer attached to the left leg. Each streamer
had an alphanumeric code painted on each side using NazDar, an ink formulated to fuse
with vinyl surfaces. We were able to distinguish the sexes by size.

We estimated nesting success, using the method developed by Mayfield (1961, 1975) and
a computer program created by J. L. Morrison. We calculated weekly survival for the in-
cubation, nestling, and post-fledgling dependency periods and the total probability of nest
success. We defined a nest as our sample unit and a successful nest as a nest in which at
least one young survived to end of the respective period.

RESULTS

We found six caracara nests in various stages of development through
10 h of aerial and 400 h of ground surveys from late December to mid-
July. Two nests were on the APCNWR and four were on private property.
Five nests were in shrubs; four in Macartney rose and one in yaupon.
One nest was in eastern red-cedar (Juniperus virginiana). All nests were
constructed in the tallest shrub or tree in the immediate area. Measurement
were taken on five nests, as the sixth (Nest 3) deteriorated before we
could measure it. Nest trees averaged 4.3 ± 1.4 m (SE) in height, nests
averaged 3.7 ± 1.2 m in height above ground, and the distance between
the nest and the canopy averaged 55 ± 25 cm. Nests averaged 59 ± 5

518

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Breeding Chronology of Crested Caracaras, South Texas

Nest

no.

Date

found

Eggs laid

Eggs hatched

Nestlings fledged

Last seen
fledglings

Date

No.

Date

No.

Date

No.

l-U

10 Jan

17 Jan

2

13 Feb

1

9 Apr

1

17 May

\-2^

17 Jun

1

16 July

1“

e

—

—

18 Jun

U

—

—

—

—

T

21 Feb

12 Mar

2

—

—

—

—

—

38

27 Mar

—

—

—

—

—

—

—

4

9 Mar

19 Mar

1

16 Apr

1

10 Jun

1

7 July

21 Mar

1

18 Apr

1

12 Jun

1

7 July

5-U

27 Mar

7

7

7

7

25 Apr

2

4 Jun

5-2'’

24 Jun

2

23 July

2^

—

—

—

6

17 Mar

7

7

31 Mar

1

24 May

1

20 Jun

1 Apr

1

24 May

1

20 Jun

2 Apr

1

26 May

1

20 Jun

• Denotes first nesting attempt.

^ Denotes second nesting attempt.

' Egg did not hatch.

■I One nestling killed by red imported fire ants (Solenopsis invicta).
' No data.

' Adults abandoned nest at 31 days of incubation.

* Adults abandoned nest building after three days.

'' Two nestlings killed by red imported fire ants (S. invicta).

cm in length, 50 ± 2 cm in width, and bowls averaged 11 ± 2 cm in
depth. All nests were buried in the canopy and were impossible to see
from the air or ground. Two of the nests constructed in Macartney rose
could be entered and exited by the caracaras through only one opening
in the foliage. Nests were constructed of Macartney rose, dewberry, yau-
pon, or broomweed {Gutierrezia sarothrae) twigs. Nests built in Macart-
ney rose were almost exclusively built with that shrub s twigs.

Four nests were reused in the 1989 nesting season. Nests 4 and 6 were
built by caracaras in 1988 (O. Benton, pers. comm.; J. Holtkamp, pers.
comm.; respectively). Nests 1 and 5 had eggs laid in them in January and
February 1989, respectively, and again in June 1989.

We observed breeding behavior from January through July (Table 1).
We observed pairs copulating as early as 10 January and as late as 19
September. Pair 1 copulated four times during the nestling period (March
and April), and Pair 4 copulated on 19 September, 100 d after the young
had fledged. We observed courtship behavior during 12.5 ± 2.1 d (N =
6 pairs). The incubation, nestling, and post-fledging dependency periods
averaged 29 — 0.5 d (N = 3), 56.2 ± 0.4 d (N = 5), and 31.8 — 2.5 d
(N = 5) in length, respectively.

Dickinson and Arnold • BREEDING CRESTED CARACARAS

519

Table 2

Measurements from Six Crested Caracara Nestlings at Various Stages of
Development in Austin and Colorado Counties, South Texas, 1989

Ne.st

ID

Age

(days)

Wing

chord

(cm)

Tail

(cm)

Tarsus

(cm)

Hallux

(cm)

Culmen

(cm)

Mass

(g)

1

A1

47

28

17

9

1.4

2.32

1100

4

D3

26

16

7

8.5

1.48

2.24

880

4

D3

45

29

16.5

9

1.57

2.49

960

4

El

28

14

6.7

8

1.33

2.15

880

4

D1

47

29

17

8

1.7

2.67

960

6

A2

39

23

11

9.5

1.21

2.51

1200

6

A2

43

26.5

14

10

2.05

2.49

1220

6

D1

37

20.5

9.5

9

1.53

2.48

1160

6

D1

54

35

19

9

1.74

2.28

1160

6

A4

38

24

12

8

2.1

2.53

1100

Pairs laid eggs from January through June (N = 5) (Table 1). Pair 2
laid two eggs on 12 March but abandoned the eggs after 30 d of incu-
bation. We opened the eggs and found them infertile. Pair 3 did not lay
eggs, as the pair abandoned the nest on 29 March after 3 d of nest build-
ing. Pair 4 laid two eggs from 19 March to 21 March. We did not located
Pair 6’s nest in time to determine when the eggs were laid, but we ob-
served three young hatch from 31 March to 2 April. Pairs 1 and 5 laid
second clutchs. Pair 1 laid two eggs on 17 January and laid two eggs
from 17 June to 18 June. Pair 5 fledged two young 25 April and laid two
eggs on 24 June. One chick hatched in Nest 1 and two chicks hatched in
Nest 2 on 16 July and 23 July, respectively. All three nestlings from these
second clutches were killed by red imported fire ants (Solenopsis invicta).

Clutch size was determined for five nests; four nests had two eggs, and
one nest had three eggs. Eggs were laid at 1-2 day intervals (N = 3
nests). Eggs averaged 53.1 ± 0.1 mm in length, 41.3 ± 0.03 mm in width,
and 65.2 ± 0.6 g in mass (N = 4).

Young caracaras were found in the nest from late February to mid-
June, and fledged from early April to mid-June (N = 8). The older chick
always left the nest first. Time between fledglings leaving the nest was
between 1-2 days (N = 5 fledglings). Measurements and masses were
similar for nestlings of similar age (Table 2).

Total probability of nest success for nests initiated in January through
March was 72.6% (N = 5), compared to 45.7% (N = 7) for all attempts
(Table 3). We did not include Pair 3 in the calculation of nest success.

520

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

Table 3

Nesting Success for Crested Caracaras in Austin and Colorado Counties, South

Texas, 1989

Mayfield estimate

Incubation

Nestling

Post-Hedging

dependency

Overall

Total nests observed

Earlier nests (Jan— Mar)

4 OY' 4 (4)'’

4 (4)‘>

Traditional success esti-
mates

0.750

LOO

1.00

Total observed weekly
exposure

13.0

28.0

32.0

—

Mayfield estimated week-
ly survival probabili-
ties

0.923 ± 0.074"

LOO

1.00

Total probability of suc-
cess (Mayfield)

0.726 ± 0.1 16"

1.00

LOO

0.726

Total nests observed

All nests (Jan-Jun)

6 (5)*’ 6 (4)^

4 (4)^’

Traditional success esti-
mates

0.833

0.666

1.00

—

Total observed weekly
exposure

21.0

28.2

32.0

—

Mayfield estimated week-
ly survival probabili-
ties

0.952 ± 0.046"

0.929 ± 0.048^

‘ 1.00

Probability of success for
entire period

0.823 ± 0.083"

0.555 ± 0.082^

‘ LOO

0.457 ± 0.140"

“ Standard deviation.

Number of successful nests is in parentheses.

We observed Pair 3 copulating and nest building for only three days
before they left the study area.

By mid-August, young and adults were not regularly observed on or
around the nest. On 19 September, the family group from Nest 4 was
observed roosting on a powerline 2 km from the nest. Pair 6 was seen 2
km from the nest on 13 September.

DISCUSSION

We report several new findings from our research. Previously unre-
ported for Crested Caracaras has been double brooding, fire ant predation
on newly-hatched nestlings, body measurements and masses of nestlings,
yaupon and eastern-red cedar as nest trees, and duration of the post-
fledging dependency period. Pairs constructed nests from January to June.
Simmons (1925) found nest materials collected as early as December in

Dickinson and Arnold • BREEDING CRESTED CARACARAS

521

Florida, and Levy (1961) observed nest construction in Arizona on 20
March. In this study, caracaras nested in tall, dense Macartney rose stands.
Caracaras nested in the tallest vegetation in the area. Macartney rose on
the APCNWR, and eastern red cedar on the Underwood Ranch. Previous
accounts from North America reported caracaras nested in ebony (Pithe-
collobium spp.), hackberry (Celtis spp.). Macartney rose, mesquite (Pro-
sopis spp.), oaks, palmettos (Sabal spp.), pines (Pinus spp.), saguaros,
yuccas, and on cliffs (Bent 1938, Dillon 1961, Levy 1961, Oberholser
1974, Layne 1978, Farquhar 1986, Ellis et al. 1988). This is the first time
yaupon and eastern-red cedar have been reported as nest supports.

Nest heights were similar to those reported by Oberholser (1974). In
this study caracaras built their nests below the nest-support canopy. Cara-
caras may prefer their nests below the canopy rather than on top. Only
Bent (1938) and Brown and Amadon (1968) report the nests as hard to
locate. Nests were similar in structure and construction materials to those
of other studies. Nests composed entirely of broomweed are reported from
Texas (Bent 1938), although most authors described the nest as simply a
bulky structure of weeds and twigs (Brown and Amadon 1968, Oberhol-
ser 1974, Layne 1978).

In central Texas, two-thirds of 35 nests observed were reoccupied an-
nually, but whether by former owners was not known (Schultze 1904).
In this study, four of six nests had been previously used by caracaras.
Farquhar (1986) reported caracaras using White-tailed Hawk (Buteo al-
bicaudatus) nests on the APCNWR, and Mader (1981 ) noted one caracara
nest refurbishment in Venezuela.

We suspect eggs laid in June were second broods from Pairs 1 and 5,
based on our confidence at identifying the adults from daily behavior
observations. Howell (in Bent 1938) reported one case of double brood-
ing, but whether by the same parents was unknown. Doubling brooding
has never been confirmed, although it was suspected (Slud 1964, Palmer
1988).

We observed copulations throughout the breeding season, which may
help to maintain the pair bond (Newton 1979). Eggs were smaller than
those reported by Bent (1938) and Oberholser (1974) and weighed less
than those reported by Newton (1979). We found most nests had a clutch
size of two, unlike Bent (1938) who reports a larger percentage of clutch
sizes of three. The lengths of incubation and nestling periods were similar
to those of earlier studies (Bent 1938, Layne 1978, Newton 1979).

Timing of nest initiation at our study was comparable to those of other
studies. Four nests in Arizona had nestlings in May (Levy 1961, Ellis et
al. 1988), and one nest in Texas had young in July (Ellis et al. 1988).
Layne (1978) reported that the young fledge at about eight weeks.

522

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

Rivera-Rodriquez and Rodriquez-Estrella (1992) found 83% of 16 cara-
cara nests in Mexico in 1990 were successful. They did not define nesting
success using the Mayfield method. In this study, lower nest success for
the incubation and nestling periods was the result of failure of Nest 2 and
nestlings preyed upon by red imported fire ants in the second nesting
attempt by Pairs 1 and 5.

Our data provides insight into length of the post-fledging dependency
period, which was previously unknown (Newton 1979). Dillon (1961)
last saw a family group in Central Texas in late June or July and did not
see the nesting pair again until the following January. The duration of
the post-fledging dependency period may be longer than reported here,
as we had a narrow definition of this period and we did not use radio-
transmitters to monitor fledglings.

ACKNOWLEDGMENTS

We gratefully acknowledge the assistance and cooperation of personnel at the APCNWR
and numerous volunteers including J. Strater, R. Hicks, J. Weltzin, D. Carrie, R. Carrie, R.
Benson, R. Duran, R. Engelbrecht, K. Northrup, C. and K. Hendry, J. Johnston, J. Holtkamp,
and J. Kent. We also thank O. and M. Benton, D. and L. Underwood, and R. Kaechele for
providing access to their land. We thank J. Morrison, D. Carrie, and two anonymous re-
viewers for their helpful comments on the manuscript. We thank the APCNWR for providing
housing and transportation. This study was funded by grants from the Leliciana Corporation,
International Research Expeditions, and the Texas Ornithological Society.

LITERATURE CITED

Baldwin, S. P, H. C. Oberholser, and L. G. Worley. 1931. Measurements of birds, vol.

11. Sci. Publ. Cleveland Mus. Nat. Hist., Cleveland, Ohio.

Bent, A. C. 1938. Life histories of North American birds of prey, part 2. U.S. Natl. Mus.
Bull. 170.

Brown, L. and D. Amadon. 1968. Eagles, hawks, and falcons of the world. McGraw-Hill
Book Co., New York, New York.

Dillon, O. W, Jr. 1961. Notes on nesting of the caracara. Wilson Bull. 73:387.

Ellis, D. H., D. G. Smith, W. H. Whaley, and C. H. Ellis. 1988. Crested Caracara. Pp.
119-126 in Proc. of the Southwest Raptor Mgmt. Sym. and Workshop (R. L. Glinski,
B. G. Pendleton, M. B. Moss, M. N. LeFranc, Jr., B. A. Millsap, S. W. Hoffman, eds.)
Natl. Wildl. Fed. Sci. Tech. Ser. No. 11, Natl. Wildl. Fed., Washington, D.C.
Farquhar, C. C. 1986. Ecology and breeding biology of the White-tailed Hawk on the
northern coastal prairies of Texas. Ph.D. diss., Texas A&M Univ., College Station,
Texas.

Gould, E W. 1975. Texas plants: a checklist and ecological summary. Texas Agric. Exp.
Stn., College Station, Texas.

Layne, j. N. 1978. Threatened Audubons’s Caracara. Pp. 34-35 in Rare and endangered
biota of Florida, vol. 2. Birds (H. W. Kale, III, ed.). Univ. Presses, Gainesville, Florida.
Levy, S. H. 1961. The caracara nesting in Arizona. Auk 78:99.

Mader, W. j. 1981. Notes on nesting raptors in the llanos of Venezuela. Condor 83:48-51.
Mayfield, H. 1961. Nesting success calculated from exposure. Wilson Bull. 73:255-261.

Dickinson and Arnold • BREEDING CRESTED CARACARAS

523

. 1975. Suggestions for calculating nest success. Wilson Bull. 87:456-466.
Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, South Dakota.
Oberholser, H. C. 1974. The bird life of Texas, Vol. 1. Univ. Texas Press, Austin, Texas.
Palmer, R. S. 1988. Handbook of North American birds, Vol. 5. Yale Univ., New Haven,
Connecticut.

Rivera-Rodriguez, L. B. and R. Rodriguez-Estrella. 1992. Breeding ecology of the
Crested Caracara (Polyborus plancus) in the Cape Region, B.C.S., Mexico J Raptor
Res. 27:91-92.

ScHULTZE, A. E. 1904. Nesting habits of the caracara. Condor 6:106.

Simmons, G. E 1925. Birds of the Austin region. Texas Press, Austin, Texas.

Slud, P. 1964. Birds of Costa Rica. Bull. Am. Mus. Nat. Hist. 128.

U.S. Fish and Wildlife Service. 1987. Threatened status for the Florida population of
Audubon’s Crested Caracara. Fed. Reg. 52:25229-25232.

Wilson Bull., 108(3), 1996, pp. 524-534

BREEDING BIOLOGY OF THE JABIRU IN THE
SOUTHERN LLANOS OF VENEZUELA

Jose A. Gonzalez

Abstract.— I studied the breeding biology of the Jabiru (Jabiru mycteria) at Hato El
Fn'o (state of Apure, Venezuela) during two breeding seasons. I located 22 nests during
19g9_90 and 28 in 1990-91. Jabiru nests were 8-26 m from the ground in ten different
species of trees, with Sterculia apetala the most commonly used (36.4% of the nests). Storks
laid eggs between August and November. The greatest number of clutches were m Septem-
ber. Average clutch size was 3.4 eggs (range: 2-5; N = 17), with four eggs the most frequent
clutch size. Fledglings left their nests in January or February at the age of 12-13 weeks but
were still dependent of their parents for a period of up to two months. Nest success was
47.0% in 1989-90 and 47.6% in 1990-91; productivity (fledglings/active nest) was 0.94
and 1 00 respectively. In 1990-91, only 20% of eggs produced fledglings. Most nests (75%)
failed during incubation. Main causes of nest failure were abandonment, nests falling, and
predation by Crested Caracaras {Polyborus plancus). Received 20 Aug. 1995. accepted 15

Jan. 1996.

The Jabiru {Jabiru mycteria) breeds east of the Andes from southern
Mexico to northern Argentina (Blake 1977, Hancock et al. 1992). Despite
its large size and wide distribution, the breeding biology of this stork i?>
poorly known (Kahl 1971, Luthin 1987, Hancock et al. 1992). Kahl
(1971, 1973) and Spaans (1975) reported on the status, behavior, and
reproduction of the species in Argentina and Surinam. Thomas (1981)
described nesting of the Jabiru in the central llanos of Venezuela (state
of Guarico) and made a behavioral comparison of the species of storks

that coexist in the region (Thomas 1985).

The Jabiru appears to be widespread but not abundant in the llanos oi
Venezuela (Ramo and Busto 1984, Ogden and Thomas 1985). I found no
literature reports on the recent status of the species in this region. Con-
servation of the Jabiru in the llanos is threatened by continuous loss of
forests and foraging sites, the massive use of pesticides m agncultural
lands and the proliferation of artificial dikes for water management (Ay-
arzaguena et al. 1981, Luthin 1987, Morales 1990, Gonzalez 1993). Lu-
thin (1987) strongly suggested that research on the ecology and status o
the Jabiru should be undertaken on each distinct population in order to
develop a global conservation stratregy for the species.

STUDY AREA AND METHODS

1 studied Jabirus at Hato El Frfo, a 78.000-ha private cattle ranch located in the southern
or flooded plains of Venezuela, between the villages of El Saman and Mantecal (7 35

Asociacion Amigos de Donana. Panama 6, 41012-Sevilla, Spam.

524

Gonzalez • BREEDING OF THE JABIRU

525

7 55 N, 68°50'-69°00'W) in the state of Apure. The study area is a tropical wet savanna
with a highly seasonal distribution of rainfall. Mean annual rainfall is 1653 mm (period:
1969-1988), with more than 80% falling between May and October (rainy season), when
much of the land is flooded up to one meter. Rainfall is very scarce between November and
April (dry season), when much of the land becomes dry and water is restricted to a few
streams, lagoons, and deepest marshes. The climate is tropical and mean monthly temper-
ature is more or less uniform throughout the year, ranging from 28.6°C in March to 25.4°C
in July.

Following Ramia s (1967) classification of llanos landscape types, the study area belongs
to the group named “savannas of banco, baji'o and estero”. More than 80% of the land is
occupied by herbaceous savanna vegetation, while the rest is covered by gallery forests and
small isolated forested islands locally called matas. A detailed description of vegetation
communities present at Hato El Frio can be found in Castroviejo and Lopez (1985).

The study area was surveyed daily along fixed ground routes between July and September
in 1989 and 1990, for pairs of Jabirus and for nesting attempts. In 1990, a fixed-wing aircraft
was also used to search for new nests. Nesting attempts were mapped. The status of every
nesting attempt was then monitored weekly until the nest was abandoned by adults or until
fledglings made their first flight. A mirror on an extendable pole (also used to estimate nest
height) was used to observe the contents of lower nests, but in most cases it was necessary
to climb the tree or to use a ladder to reach nest level. Many of the nests were inaccessible
from the ground, and their status in 1989-1990 was determined by prolonged observation
of the birds behavior. During 1990-1991, a fixed-wing aircraft was used to check the
contents of inaccessible nests. The sexes were distinguished by body size and length and
curvature of bill (Kahl 1971), later confirmed by copulation position.

Repeated visits to wading bird nests can severely bias the reproductive parameters studied
(Tremblay and Ellison 1979, Frederick and Collopy 1989b, Kushlan 1992). On each visit,
the distance at which adults left the nest in response to human approach, the total time spent
near each nest and the length of time it took the birds to return to the nest were recorded.
Careful observations were made for possible predation attempts while in the vicinity of the
nest. To minimize disturbance, nests containing chicks more than two weeks old were not
climbed; after this stage, chicks were easily visible using 12X binoculars or a 20X telescope,
but eight visits to nests with older chicks were made to collect regurgitation samples (Gon-
zalez 1993). No nests were climbed more than three times during incubation and the first
two weeks after hatching. In every case, I avoided stormy or rainy days and direct sunshine
to prevent thermal stress to nest contents (Dusi and Dusi 1978, King 1978).

Nesting attempts were categorized as having nest construction activities and permanent
occupation by a pair of Jabirus. Active nests were those in which at least one egg was laid
or, in inaccessible nests, those in which continuous incubation activity by adults was ob-
served. Successful nests were tho,se in which at least one chick fledged, and unsuccessful
nests those that lost all eggs or chicks.

RESULTS

Breeding dates. — The Jabiru nesting period in the southern llanos be-
gan in the latter part of the rainy season and extended to the middle of
the dry season. Laying was from mid-August to mid-November, with the
greatest number of clutches laid in September (66.6% in 1989 and 50%
in 1990, Fig. 1). The nesting season began later in 1989 than in 1990; in

Percent of clutches

526

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

40

Lig. 1. Jabiru laying dates at Hato el Lno in 1989 and 1990.

the first season, no eggs were laid until September, while in 1990, 25%
of the clutches were initiated in the second half of August (Fig. 1).

Nestlings stayed in the nest for about three months (84-93 days; N -
3 nests). Most of the 1989 successful nestlings made their first flight from
1 to 20 January 1990 (64.3%); the remaining ones fledged m February
(Fig. 2). After their first flight siblings stayed together and were fed by
their parents in nearby wetlands for a period of 6-8 weeks (Gonzalez
1993). Therefore, successful Jabiru pairs were involved in reproductive

tasks for almost seven months a year (Fig. 2).

Nest site. — Most of the nests (78.1%) were in small isolated forested
islands (matas) of 0.2-2 ha, 3.1% were in matas of more than 50 ha,
9.4% were in gallery forests and the remaining 9.4% were in solitary
trees. Almost all the Jabiru nests were solitary and owners defended a
breeding territory of 300-500 m around their nests by chasing other Jab-
irus and other wading bird species that flew or foraged in that area. Mean
distance to nearest conspecific nest was 1.9 km (range: 1.2-2. 8). Three
nests were built in the center of mixed-species colonies containing other
wading birds such as White-necked Herons {Ardea cocoi). Great Egrets
{Casmerodius albus), and Maguari Storks {Ciconia maguari).

Nests generally were built at the top of one of ten different species of
large broad-leaved trees. Stercidia apetala was the most frequently used
(36.4% of the nests), followed by Ficus sp. (18.2%) and Pithecellobiiim
.saman (9.1%). Other trees that supported Jabiru nests were Spondias

Percent of nests

Gonzalez • BREEDING OF THE JABIRU 527

Occupied f

— 1 With eggs

^ With chicks

Failed I

— 1 Successful

10 25 10 25 10 25 10 25 10 25 10 25 10 25

AUG SEP OCT NOV DEC JAN FEB

1989-90

31 15 31 15 30 15 31 15 30 15 31 15 3l" 15

JUL AUG SEP OCT NOV DEC JAN FEI

1990-91

Fig. 2. Phenology of the 1989-90 (N = 20 nests) and 1990-91 (N = 24 nests) breeding
seasons.

mombin, Acrocomia sclerocarpa, Ceiba pentandra, Tabebuia sp., Sapiiim
biglandulosum, Vitex sp., and Coccoloba camcasana.

The average height of Jabiru nests on Hato El Fno was 15.4 m (range:
8-26). At a nearby ranch (Hato El Cedral), a nest was built less than 7

528

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

m from the ground. In nests located in dense forests (matas or gallery
forests), the pair usually chose the highest tree around with an excellent
view of the surrounding country. Nests were placed on forks of large
limbs, between a large horizontal limb and the main trunk, or more usu-
ally in places where 3-4 branches crossed. Of the studied nests, 71.9%
were on the top of the tree (crown branches), while the remaining 28.1%
were on bifurcations of the trunk of a dead tree. All of the nests were
less than 500 m from a large temporary or permanent wetland. Many of
the nests (28.1%) were completely surrounded by deep marshes inundated
during the nesting season, 21.9% were close to large permanent lagoons,
and the remaining 50% were on the edge of a river or stream. These
flooded wetlands were used by fledglings for 6-8 weeks after leaving the

nest.

Construction and structure of nests. — All but two of 22 nests occupied
in the 1989-1990 breeding season were reused in 1990-1991. 80% of
those nests included part of their basic structure from one year to the
next. The remaining 20% completely disappeared during the non-breeding
season due to weather, broken supporting limbs, or the piracy of nest
materials by other wading birds (in nests located in mixed-species colo-

nies).

The time spent by Jabirus on nest construction before egg-laying de-
pended on the previous state of the nest. Two nests were completely built
in less than 20 days from a residual group of sticks remaining from the
previous nesting season. Another nest, that kept its structure almost intact,
was repaired in less than a week. The longest time it took a pair of Jabirus
to build their new nest was seven weeks. Both sexes collaborate in nest
construction or repairing, although sticks are mainly gathered by the male
(Kahl 1971, Shannon 1987), at least in early stages of nest-buildmg. In
four nests monitored during an entire day, 66% of the new sticks for nest
construction were gathered by the male and 34% by the female. In another
nest that was apparently finished, all of the sticks and green materials
were gathered by the male while the female remained on a nearby branch;
only sometimes (30.7%), when the male came back with new materials,
she jumped to the nest to position the sticks and then copulate. After egg-
laying the female assumes a more equal role in nest-maintenance, as is

the case in other storks (Kahl 1971).

Jabiru nests are oval to circular structures composed of sticks up to
160 cm long and 3.5 cm thick, with a central area lined with green plant
material (grass, leaves and aquatic vegetation). Five nests measured dur-
ing incubation averaged 205 X 180 cm (range: 180 X 130-220 X 180),
with a central lined area of 100 X 85 cm (range: 80 X 70-120 X 90).

Gonzalez • BREEDING OF THE JABIRU

529

Nest thickness ranged between 40 and 60 cm, although one exceptional
nest was 1 10 cm thick.

The gathering of sticks and green lining material continued without
interruption throughout the nesting season, even in nests containing nest-
lings more than 80 days old and almost ready to fly. Initial size of nests
changed significantly between incubation and fledgling stages. One nest
was 220 X 170 cm during the incubation period and reached 270 X 190
cm two months later; the lined area also increased from 120 X 90 to 170
X 130 cm.

Clutch size and eggs. — Three complete clutches in 1989-1990 and 14
in 1990—1991 were counted. Grouping nests of both breeding seasons,
average clutch size was 3.4 eggs (range: 2—5; N = 17), with six nests of
four eggs and five nests of two eggs. A possible replacement clutch was
found. The owners of this nest lost their original clutch of two eggs in
the first week of October 1990, and the same pair was seen incubating
during several consecutive days in the middle of November. The nest was
abandoned during the first days of December before I could corroborate
the existence of a replacement clutch and its size.

The eggs of the Jabiru are ovate to subelliptical. Coloration is white
but becomes dirty with the passage of time. Six eggs from three nests,
measured early in incubation, averaged 70.4 X 53.6 mm (length; 67.6-
72.9; width: 49.8-55.9), with a mean weight of 110 g (range: 90-120).

Nesting success. — Of the 22 nesting attempts during the 1989-1990
breeding season, five were abandoned before egg-laying or any incubation
activity by Jabirus. Three of these failed attempts were abandoned in an
early stage of nest construction, while the other two nests were apparently
finished and many copulations occurred in them before they were aban-
doned. Territorial pairs of abandoned nests remained in the area for up
to two months, usually perching on the same tree that supported the nest.
It is unlikely that pairs that failed during early stages of construction try
to nest again in another area distant from the original one. Eight of 17
active nests during 1989-1990 fledged at least one chick (Table 1). The
total number of fledglings produced in the study area was 16, with an
average of 0.94 young per active nest. Nine nests (53%) failed, seven
during incubation and two when they contained young chicks. Three nests
fell from trees. Another had two infertile eggs that finally were aban-
doned. In the remaining five nests the cause of reproductive failure could
not be determined. Partial losses were recorded in the eight successful
nests. One egg from a clutch of three was infertile but remained in the
nest for more than a week after the hatching of the other two. In two
nests with initial brood size of three, the two youngest chicks were con-
siderably smaller than their sibling and died during the first month of life.

530

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Reproductive Success of the Jabiru on Hato El Eri'o during the Breeding Seasons

OF 1989-90 AND 1990-91

Nest variable

Occupied nests
Active nests‘‘

Successful nests
% of nests successful*’
Number of fledglings
Lledglings/active nest
Lledglings/successful nest

“ Producing a clutch of eggs.

'> Producing at least one chick to fledging age.

1989-90

22

17

8

47.0

16

0.94

2.0

1990-91

28

21

10

47.6

21

1.00

2.1

probably due to starvation. At another nest, one of three nestlings 65--70
days old died during a prehiature first flight or a fall during flapping
exercises.

In the breeding season of 1990-1991 21 of 28 nesting attempts resulted
in active nests. Ten were successful (47.6%). The total number of fledg-
lings produced in the area was 21 (Table 1). Using data from 14 closely
monitored nests, I estimated that only 20% of the eggs laid produced
fledglings. Eight nests failed during incubation and three after hatching.
One nest with two nestlings 3-4 weeks old fell when the supporting limb
broke. In three failed nests, I observed Crested Caracaras {Polyborus
plancus) eating eggs (2 nests) and nestlings 9-14 days old (1 nest); chicks
and eggs were completely consumed at the nest. It is unknown whether
this was predation or post-abandonment scavenging (see Frederick and
Collopy 1989c). In one of these nests several days before predation, the
pair of Jabirus left the eggs alone for long periods of time, which suggests
that abandonment might have been the actual cause of failure. Brood
reductions occurred in three successful nests, which lost a total of four
nestlings. In two nests with initial brood size of three, the youngest chick
died apparently of starvation; in another nest, with initial brood size of
four, two chicks died during the first month of life.

DISCUSSION

Jabirus begin breeding in the southern llanos of Venezuela in August
and some breeding occurs until February of the following year. Egg-
laying takes place mainly during September, and most fledglings leave
their nests in January. The dates of egg laying coincide with the period
of major flooding, when adult Jabirus can easily find abundant food in

Gonzalez • BREEDING OE THE JABIRU

531

inundated marshes, mainly freshwater eels (Kushlan et al. 1985, Thomas
1985, Gonzalez 1993). Fledglings make their first flight in the middle of
the llanos dry season, when the absence of rains and rapid drying result
in large concentrations of fishes in the few lagoons and ponds that still
preserve some water (Gonzalez 1993). The begining of the breeding sea-
son, as well as nesting success of the Jabiru in the southern llanos, may
be related to the onset and quantity of rainfall and the effect they have
on water level and food availability. This is similar to other wading bird
species that breed in wetland habitats distinguished by a marked seasonal
fluctuation in water level (Ogden et al. 1980, Ayarzagiiena et al. 1981,
Frederick and Collopy 1989a). Breeding dates in the study area are similar
to those reported in literature for Jabirus in other regions. Kahl (1971)
reported that the egg-laying period extends from July to October in Ar-
gentina and Brasil. Spaans (1975) recorded egg laying during August and
September in Surinam. Laying dates in Guyana, Brazil, and Colombia are
also between July and October (in Kahl 1971). In the central llanos of
Venezuela, most of the clutches are laid during September and October
(Thomas 1985).

The majority of the nests found in the study area were located in for-
ested islands (matas) of less than 2 ha, with nests most frequently placed
in Sterculia apetala. Other Jabiru nests reported from the llanos include
11 nests in the palm Copernicia tectorum (Thomas 1981) and two nests
in Pithecellobium saman (Ogden and Thomas 1985). Spaans (1975) men-
tioned one nest in Ceiba pentandra in Surinam and Kahl (1971) reported
six nests in palms and other medium-size trees in Argentina. In general,
Jabirus build nests far from other wading bird nests. However, three nests
in the study area were in the center of mixed-species colonies. Naumburg
(in Kahl 1971) also reported one Jabiru nest in a large colony of Wood
Storks (Mycteria americana) in Brazil.

Average clutch size in the study area was 3.4 eggs. Most completed
clutches were comprised of four eggs. Three clutches (17.6%) were com-
prised of five eggs, a number that is considered rare for Jabirus (Kahl
1971, Hancock et al. 1992). Hagmann (in Bent 1926) reported clutches
of two and three eggs in Jabiru nests, while Lloyd (in Bent 1926) reported
that four eggs was the most frequent clutch size in this species. Kahl
(1971) reported two clutches of four eggs in Argentina. The average size
of six eggs was slightly lower than average size reported by Bent (1926):
73.4 X 58.2 mm (N = 8).

Jabirus exhibit territoriality throughout the nesting season (Kahl 1973,
Thomas 1985, Shannon 1987). As a consequence of this agressive be-
havior, Thomas (1981) suggests that it is unlikely that more than one
female could lay eggs in the same nest. Although I absolutely agree with

532

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

these statements, however, an exceptional event took place in one of the
studied nests: for more than three hours in the morning of 3 October
1989, a male copulated repeatedly with two different females, each of
which took turns in the nest; finally, one of the females was expelled by
the other and left the area. The nest was abandoned three weeks after this
event and no incubation activity was observed during that period.

Nest success in the study area was 47% in 1989-1990 (N = 17 nests)
and 47.6% in 1990-1991 (N = 21 nests), with only 20% of the eggs
producing fledglings (N = 14 nests). There are no other published data
on nesting success of this species in other geographical areas to compare
with my results.

Most of the unsuccessful nests failed in an early stage of incubation.
The abandonment of the nest by adults and predation upon eggs or nest-
lings seem to be the two most important causes of nesting failure in the
studied population; it is very difficult to determine which of them occurs
first (Frederick and Collopy 1989c). The Crested Caracara, an abundant
raptor in the southern llanos, was the main predator of eggs and nestlings
of Jabirus and, in general, of all wading birds in the study area (Gonzalez

1993).

Of 22 nests recorded during the 1989-1990 breeding season, 19 were
reused in 1990-1991 (86.4%). Thomas (1981) reported that in the central
llanos of Venezuela, new nests were made every breeding season because
palms supporting nests died after the first year. In the study area 1 found
only one nest in a palm in 1989-1990; this nest fell during the incubation

period and was not reused in 1990-1991.

Of the 17 active nests monitored in the first breeding season, only eight
were again active during the second one. Considering the successful nests
of 1989-1990 (N = 8), only three of them were active in 1990-1991,
and only two were again successful. If we assume the hypothesis that
Jabiru pairs remain mated in successive seasons and use the same nesting
site, which is supported by several observations on their nesting behavior
and territoriality (Kahl 1973, Thomas 1985, Gonzalez 1993), my data
would indicate that less than half of active pairs in one season are also
active during the following one and that only 25% of successful pairs are
successful in a second consecutive season. This may be due to a mere
coincidence or could suggest that Jabirus have some trouble breeding
successfully in consecutive years, perhaps due to the great amount of time
that adults spend on breeding-related activities (6-7 months). Long-term
studies in the same area that include monitoring of marked birds are
needed to assess these statements.

ACKNOWLEDGMENTS

I thank J Castroviejo for his encouragement during this work. I also thank all the Mal-
donado family, owners of Hato El Frio, for their hospitality and for giving us all the facilities

Gonzalez • BREEDING OF THE JABIRU

533

to develop our work. I am grateful to E Ibanez, V. Rosales, P. Quinones, and the staff of
Estacion Biologica El Fn'o for their assistance during the field work. I am also very grateful
to P. Frederick, C. Ramo, E. de Juana, B. T. Thomas, and C. R. Blem for reviewing earlier
drafts of the manuscript. Financial support was provided by grants of Asociacion de Amigos
de Donana and Universidad Complutense de Madrid.

LITERATURE CITED

Ayarzaguena, J., J. Perez, and C. Ramo. 1981. Los garceros del Llano. Cuadernos La-
goven, Caracas, Venezuela.

Bent, A. C. 1926. Life histories of North American marsh birds. I. U.S. National Museum
Bull. 135, Washington, D.C.

Blake, E. R. 1977. Manual of Neotropical Birds, Vol. 1. Univ. of Chicago Press, Chicago,
Illinois.

Castroviejo, S. and G. Lopez. 1985. Estudio y descripcion de las comunidades vegetales
del “Hato El Frio”, los llanos de Venezuela. Memoria de la Sociedad de Ciencias
Naturales La Salle 124:79-151.

Dusi, J. L. AND R. T. Dusi. 1978. Survey methods used for wading bird studies in Alabama.
Pp. 207-211 in Wading birds (A. Sprunt, IV, J. C. Ogden, and S. Winckler, eds.).
National Audubon Society Research Report 7, New York, New York.

Frederick, P. C. and M. W. Collopy. 1989a. Nesting success of five ciconiiform species
in relation to water conditions in the Florida Everglades. Auk 106:625-634.

AND . 1989b. Researcher disturbance in colonies of wading birds: effects of

frequency of visits and egg-marking on reproductive parameters. Colon. Waterbirds 12-
152-157.

AND . 1989c. The role of predation in determining reproductive success of

colonially nesting wading birds in the Florida Everglades. Condor 91:860-867.
Gonzalez, J. A. 1993. Contribucion al estudio de la ecologia de las ciguefias (Earn. Ci-
coniidae) en los llanos de Venezuela. Ph.D. diss., Univ. Complutense de Madrid, Ma-
drid, Spain.

Hancock, J. A., J. A. Kushlan, and M. P. Kahl. 1992. Storks, ibises and spoonbills of
the world. Academic Press, London, England.

Kahl, M. P. 1971. Observations on the Jabiru and Maguari storks in Argentina, 1969.
Condor 73:220-229.

. 1973. Comparative ethology of the Ciconiidae. Part 6. The Blacknecked, Saddle-

bill, and Jabiru storks (Genera Xenorhynchus, Ephippiorhynchus, and Jabiru). Condor
75:17-27.

King, K. A. 1978. Colonial wading bird survey and census techniques. Pp. 155-159 in
Wading birds (A. Sprunt, IV, J. C. Ogden, and S. Winckler, eds.). National Audubon
Society Research Report 7, New York, New York.

Kushlan, J. A. 1992. Population biology and conservation of colonial wading birds. Colon.
Waterbirds 15:1-7.

, G. Morales, and P. C. Frohring. 1985. Foraging niche relations of wading birds
in tropical wet savannas. Pp. 663-682 in Neotropical ornithology (P. A. Buckley, M.
S. Foster, E. S. Morton, R. S. Ridgely, and F. G. Buckley, eds.). Ornithol. Monogr. No.
36. American Ornithologists’ Union, Washington, D.C.

Luthin, C. S. 1987. Status and conservation priorities for the world’s stork species. Colon.
Waterbirds 10:181—202.

Morales, G. 1990. Conservacion de las aves zancudas en los llanos de Venezuela. Pp. 77-
84 in The Scarlet Ibis (Eiidocimu.s ruber)-, status, conservation and recent research (P.

534

THE WILSON BULLETIN • Vol 108, No. 3, September 1996

C. Frederick, L. G. Morales, A. L. Spaans, and C. S. Luthin, eds.). International Wa-
terfowl and Wetlands Research Bureau, Slimbridge, England.

Ogden, J. C., H. W. Kale, and S. A. Nesbitt. 1980. The influence of annual variation in
rainfall and water levels on nesting by Florida populations of wading birds. Transactions
of the Linnaean Society of New York 9:1 15—126.

and B. T. Thomas. 1985. A colonial wading bird survey in the central llanos of

Venezuela. Colon. Waterbirds 8:23-31.

Ramia, M. 1967. Tipos de sabanas en los llanos de Venezuela. Boletin de la Sociedad
Venezolana de Ciencias Naturales 27:264—288.

Ramo, C. and B. Busto. 1984. Censo aereo de Corocoros y otras aves acuaticas en Ven-
ezuela. Boletin de la Sociedad Venezolana de Ciencias Naturales 39:65-88.

Shannon, P. W. 1987. The Jabiru Stork {Jabiru mycteria) in zoo collections in the United
States. Colon. Waterbirds 10:242-250.

Spaans, A. L. 1975. The status of the Wood Stork, Jabiru, and Maguari Stork along the
Surinam coast. South America. Ardea 63:116-130.

Thomas, B. T. 1981. Jabiru nest, nest building and quintuplets. Condor 83:84-85.

. 1985. Coexistence and behavior differences among the three western hemisphere

storks. Pp. 921-931 in Neotropical ornithology (P. A. Buckley, M. S. Foster, E. S.
Morton, R. S. Ridgely, and F. G. Buckley, eds.). Ornithol. Monogr. No. 36. American
Ornithologists’ Union, Washington, D.C.

Tremblay, J. and L. N. Ellison. 1979. Effects of human disturbance on breeding of Black-
crowned Night Herons. Auk 96:364-369.

Wilson Bull., 108(3), 1996, pp. 535-539

EFFECT OF EGG SIZE ON PREDATION BY
WHITE-FOOTED MICE

R. M. DeGraaf and T. J. Maier

Abstract. — We compared predation by wild-trapped, caged white-footed mice (Pero-
mysciis leucopus) on eggs of Japanese Quail {Coturnix cotiirnix) and Zebra Finches (Poe-
phila guttata) to test the effect of egg size. Nine male and nine female mice were weighed,
acclimated to cages for 24 h, and presented with two wicker nests, one containing a Japanese
Quail egg (33 X 23 mm) and the other a Zebra Finch egg (16 X 12 mm). Nests were
checked at 2, 4, 6, 8, 12, 16, and 24 h; after 24 h, no quail eggs were depredated, but 16
of 18 finch eggs were destroyed. Given their ability to consume small eggs and their ubiquity
and abundance, white-footed mice are potentially significant nest predators. Received 21
Nov. 1995, accepted 15 Feb. 1996.

While evaluating the effect of forest understory density on predation
of artificial nests, we found that white-footed mice (Peromyscus leucopus)
frequently were recorded by remotely-triggered cameras at ground and
shrub nests containing eggs of Japanese Quail {Coturnix coturnix). Many
of these same nests appeared to be undisturbed at the end of the exposure
period (DeGraaf et al., unpubl. data). Eggs of Coturnix commonly are
used to simulate those of passerines in studies of nest predation because
they are the smallest eggs commercially available in large quantity. Many
studies have recently been conducted using Coturnix eggs to assess effects
of habitat fragmentation or structure on forest birds, especially Neotrop-
ical migrants (e.g.. Small and Hunter 1988, Wilcove 1985, Martin 1987),
but none of these studies tested whether mice can open songbird eggs but
not the larger quail eggs. Maxson and Oring (1978) reported that preda-
tion on Spotted Sandpiper {Actitis macularia) eggs essentially was elim-
inated after Peromyscus were trapped out of the study area, indirectly
implicating mice as nest predators. White-footed mice have been docu-
mented as predators on nests of Prothonotary Warblers (Protonotaria ci-
trea) (Guillory 1987). Do Coturnix eggs allow assessment of the role of
mice as predators on artificial nests?

We compared predation by caged white-footed mice on eggs of Japa-
nese Quail and Zebra Finches {Poephila guttata) to test the effect of egg
size. Coturnix eggs averaged approximately 33 X 23 mm, and finch eggs
were approximately 16 X 12 mm. For comparison, egg sizes of several
forest passerines are: Gray Catbird (Dumetella caroUnensi.s), 26 X 19

USDA Forest Service, Northeastern Forest Experiment Station, Univ. of Massachusetts, Amherst, Mas-
sachusetts 01003.

535

536

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

20

16

O)

C

'c 12

’CC

E

CD

^ n

U) 8
O)

LLI

4

0 5 10 15 20 25

Hours post exposure

Lig. 1. Numbers of large Japanese Quail {Coturnix coturnix) and small Zebra Linch
(Poephila guttata) eggs remaining after one egg of each type was simultaneously exposed
to 18 wild-trapped white-footed mice (Peromyscus leucopus) for 24 h.

• • • •

Coturnix couturnix
Poephil^guttata

mm; Wood Thrush (Hylocichla mustelina): 25 X 19 mm, and Black-and-
White Warbler (Mniotilta varia): 17X13 mm (Harrison 1975).

METHODS

Eighteen white-footed mice, nine males and nine non-lactating females, were selected
randomly from 25 individually caged mice wild-trapped over a two-day period in western
Massachusetts. Mice were weighed, sexed, and acclimated for at least 24 h to cages in a
large, unheated vacant barn. Cages were provided with wood-chip litter, water, and food ad
lib. Weights of mice ranged from 1 1.7 to 27.4 g, representing juvenile, subadult, and adult
stages. Mean weight for both males and females was 20.3 g. Each mouse was then provided
with two aviculturists’ wicker nests, one containing a Zebra Finch egg and the other a
Japanese Quail egg, at 09:00 on 13 October 1995. Nests were checked at 2, 4, 6, 8, 12, 16,
and 24 h post-exposure and predation on eggs was recorded.

RESULTS

After 24 h exposure, no Coturnix eggs were depredated, but 16 of 18 finch eggs were
destroyed (Fig. 1). White-footed mice are primarily nocturnal (Baumgardner et al. 1980),
but predation on finch eggs commenced shortly after exposure, i.e., in mid- to late morning.
Mean times until depredation differed (0.05 > 1/| > 0.02) for female (x — 6.5 h) and male
(jp = 15.5 h) mice. We detected no relationship between mouse weight and time to, egg
depredation. The two finch eggs not eaten within 24 h were given to two other randomly

DeGraafand Metier • EGG PREDATION BY MICE

537

chosen mice to determine whether these eggs were unpalatable; both eggs were eaten within
24 h. All but one finch egg were opened from the side and the contents eaten through the
resulting hole; one egg was opened at both ends. No Coturnix eggs were broken after
exposure for 13 days.

DISCUSSION

White-footed mice probably are significant predators of passerine eggs,
but based on the results of our experiments, their effect cannot be exper-
imentally measured using Coturnix eggs. The inability of mice to con-
sume quail eggs may be due to several factors. First, the jaw-gape of
mice may be too small. We measured the Jaw-gape of 22 previously-
frozen, locally wild-trapped adult white-footed mice. The jaws were
opened, using loops of fine wire until resistance was felt, and the distance
between upper and lower incisor tips measured. Jaw-gape ranged from
7.0 to 10.5 mm, with a mean of 8.9 ± 1.1 mm. Such a gape seems too
small to open Coturnix eggs, which were 23 mm in smaller diameter.
Also, eastern chipmunks (Tamias striatus) did not open Coturnix eggs
even though they had previously consumed the contents when presented
with broken eggs (Haskell, 1995a).

Second, eggs from commercial quail farms may have thick shells due
to diet supplements; this was the case with our quail eggs. Shell thickness
may partly explain why Coturnix eggs were not opened by mice in the
present study, but Spotted Sandpiper eggs, which are about the same
size — 32 X 23 mm (Harrison 1975:69) — apparently were depredated by
mice in Maxson and Oring’s (1978) study. We measured shell thickness
of five fragments of each of two Spotted Sandpiper eggs, one from the
Carnegie Museum and another from the University of Massachusetts Mu-
seum of Zoology; both measured 102 microns thick. Two Japanese Quail
eggs similarly measured were 229 and 216 microns thick. Japanese Quail
egg shells were more than twice as thick as Spotted Sandpiper eggs.

Third, egg shape may be a factor. Spotted Sandpiper eggs are oval to
pyriform (Harrison 1975:69); Coturnix eggs are short-oval. Mice may be
able to open the small end of pyriform eggs that are the same size as
short-oval eggs. The eggs of most forest songbirds in the Northeast are
oval or short-oval, rarely long-oval (Harrison 1975).

Coturnix eggs and those of domestic chickens may be useful to assess
predation on eggs of waterfowl and upland game birds, but such eggs are
too large or too thick-shelled to adequately assess predation on those of
smaller passerines. For example, Ovenbird (Seiurus aurocapillus) eggs
average 20.2 X 15.5 mm; those of Chestnut-sided Warbler (Dendroica
pensylvanica) average 16.7 X 12.4 mm (Bent 1953). Haskell (1995a) has
shown that the jaw-gape of chipmunks was large enough to break the

538

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

eggs of most Neotropical migrant passerines that nest in forest or scrub,
but was too small to break Japanese Quail eggs.

The appearance and positioning of artificial nests can be made to ap-
proximate those of breeding birds (e.g., Martin 1987, Yahner and Voytko
1989) and precautions taken to avoid leaving human scent at artificial
nests (e.g., Whelan et al. 1994). But, if the eggs used cannot be broken
by the full suite of potential predators in the habitats being studied, total
predation rates or effects of habitat differences cannot be estimated. For
example, if the relative abundance of predator species varies across hab-
itat fragment sizes, the bias inherent in using too-large eggs can lead to
spurious correlations between fragment size and predation rate. In an in-
vestigation of whether the rate of nest predation by small-mouthed mam-
malian predators varied by fragment size, Haskell (1995b) found that the
number of eggs preyed upon by such predators increased with fragment
size; in large fragments the predominant nest predators were those which
quail egg experiments failed to sample. In both North America (Yahner
and Scott 1988) and Europe (Andren 1992), nest predators in small frag-
ments tend to be large mammals, e.g., raccoon, {Procyon lotor) or corv-
ids. Small mammals, e.g., white-footed mice, in contrast, are ubiquitous
and abundant, inhabiting many habitat types (Lackey et al. 1985) and a
wide range of forest successional stages within extensive forests (Healy
and Brooks 1988). Reitsma et al. (1990) suggest that Peromyscus may be
more important as nest predators than previously thought.

The presence of mice at artificial nests, their demonstrated ability to
consume small eggs, and their ubiquity and abundance render them po-
tentially significant nest predators. Eggs susceptible to mouse depredation
must be used to estimate their impact. We suggest that such eggs be used
in future field studies that use artificial nests to validate mouse predation
on nests of small forest passerines.

ACKNOWLEDGMENTS

We thank W. R. Danielson for providing Zebra Einch eggs, W. M. Healy for help trapping
mice, R. Panza and K. Doyle for providing samples of Spotted Sandpiper eggs, and R. A.
Askins, D. G. Haskell, R. T Holmes, and W. M. VanderHaegen for their critical reviews.
M. A. Sheremeta typed the manuscript, and R. T Brooks prepared the figure.

LITERATURE CITED

Andren, H. 1992. Corvid density and nest predation in relation to forest fragmentation.
Ecology 73:794—804.

Baumgardner, D. J., S. E. Ward, and D. A. Dewsbury. 1980. Diurnal patterning of eight
activities in 14 species of muroid rodents. Anim. Learn. Behav. 8:322—330.

Bent, A. C. 1953. Life histories of North American wood warblers. U.S. National Mus.
Bull. 203.

DeGraaf and Maier • EGG PREDATION BY MICE

539

Guillory, H. D. 1987. Cavity competition and suspected predation on Prothonotary War-
blers by Peromyscus spp. J. Field Ornith. 58:425-427.

Harrison, H. H. 1975. A field guide to birds’ nests. Houghton Mifflin Co., Boston, Mas-
sachusetts.

Haskell, D. G. 1996a. Forest fragmentation and nest predation: are experiments with
Japanese Quail eggs misleading? Auk. 1 12:767-770.

. 1996b. Nest predation in forest track and the decline of migratory songbirds: a
reevaluation of the “fragmentation effect.” Cons. Biol. 9:1316-1318.

Healy, W. M. and R. T. Brooks. 1988. Small mammal abundance in northern hardwood
stands in West Virginia. J. Wildl. Manage. 52:491-496.

Lackey, J. A., D. G. Huckaby, and B. G. Ormiston. 1985. Peromyscus leucopus. Mam-
malian Species No. 247. Am. Soc. Mammal.

Martin, T. E., 1987. Artificial nest experiments: effects of nest appearance and type of
predator. Condor 89:925-928.

Maxson, S. j. and L. W. Oring. 1978. Mice as a source of egg loss among ground-nesting
birds. Auk 6:582-584.

Reitsma, L. R., R. T. Holmes, and T. W. Sherry. 1990. Effects of removal of red squirrels,
Tamiasciurus hudsonicus, and eastern chipmunks, Tamias striatus, on nest predation in
a northern hardwood forest: an artificial nest experiment. Oikos 57:375-380.

Small, M. F. and Hunter, M. L. 1988. Forest fragmentation and avian nest predation in
forested landscapes. Oecologia 76:62-64.

Whelan, C. J., M. L. Dilger, D. Robson, N. Hallyn, and S. Dilger. 1994. Effects of
olfactory cues on artificial-nest experiments. Auk 1 1 1:945-952.

Wilcove, D. S. 1985. Nest predation in forest tracts and the decline of migratory songbirds.
Ecology 66:1211-1214.

Yahner, R. H. and D. P. Scott. 1988. Effects of forest fragmentation on depredation of
artificial nests. J. Wildl. Manage. 52:158-161.

AND R. A. VoYTKO. 1989. Effects of nest-site selection on depredation of artificial
nests. J. Wildl. Manage. 53:21-25.

Wilson Bull., 108(3), 1996, pp. 540-549

CAN CHECKLIST PROGRAMS BE USED TO MONITOR
POPULATIONS OF BIRDS RECORDED DURING THE

MIGRATION SEASON?

Erica H. Dunn,' Jacques Larivee,^ and Andre Cyr^

Abstract. — Quebec’s EPOQ program compiles birders ‘ checklists, each of which re-
ports numbers of birds seen on one day at one site. We analyzed EPOQ data from the
migration season alone (1971-92), to see if these unstandardized counts might monitor
trends in populations that nest farther north. Two sets of trends were computed for each of
58 species, from annual indices based either on abundance or on frequency of detection.
Both spring EPOQ trends were signihcantly correlated with Breeding Bird Survey trends
for Quebec, while only those based on abundance performed well in fall. There was a
positive bias in magnitude of EPOQ trends, but negative EPOQ trends were reliable indi-
cators of negative BBS trends. Analysis of sub-sets of the data showed that sample size had
little qualitative effect. Checklist data should not be relied on for quantitative population
monitoring, but they do contain useful information for detection or corroboration of negative
trends. Received 27 Aug. 1995, accepted 22 Jan. 1996.

Most songbirds that breed in North America are monitored by the Breed-
ing Bird Survey (BBS), a breeding season roadside survey along randomly
chosen routes across the continent (Peterjohn 1994). Certain species are
poorly covered by BBS, however, either because they nest too sparsely or
locally to be covered by an adequate number of routes (many raptors and
colonial birds, for example) or because they breed in remote areas where
BBS routes are largely lacking (e.g., many northern boreal forest breeders).
Counting of birds during their migratory passage has been suggested as a
means of monitoring some of the species missed by BBS and as a means
of corroborating trends detected by other programs. Relatively standardized
daily counts of birds at bird observatories and hawk look-outs have been
shown to document long-term trends in bird numbers similar to those re-
ported by BBS (reviewed in Dunn and Hussell 1995).

Checklist compilation programs potentially offer another source of data
on population trends of migrants. Checklists are pre-printed lists of spe-
cies on which observers can record their observations for an area of any
size and during a period of any length. Compilations of checklist data
have several strong points: they cover broad areas where other data might
be lacking, and they harness the energy of the myriad birders who already
keep careful records of what they see. On the negative side, there is a

' Canadian Wildlife Service, National Wildlife Research Centre. 100 Gamelin Blvd., Hull, PQ, Canada,
K 1 A 0H3

2 Etude des Populations d'Oiseaux du Quebec, 194 rue Ouellet, Rimouski, PQ, Canada, G5L 4R3.

■’ Departement de biologie, Universite de Sherbrooke, Sherbrooke, PQ. Canada, JIK 2R1.

540

Dunn et al. • POPULATION TRENDS FROM CHECKLISTS

541

great deal of “noise” in the data, because observations are made oppor-
tunistically at any site on any date without any limits on duration of
observation or skill of observers. Birders may concentrate on “produc-
tive” locations, and likely are not distributed evenly in time (favoring
weekends and peak migration periods). There is also potential for con-
sistent bias over time; for example, as bird distribution and abundance
change, birders may move to new locations and/or change their search
strategies to keep their birding interesting. Moreover, steady improvement
in birders’ skills and optical aids may have increased detectability of
certain species over the years.

Despite these features of checklist data that might obscure any changes
in bird populations, it is possible that they still contain useful trend in-
formation. Cyr and Larivee (1993) looked for evidence of this, analyzing
spring and fall data from the Etude des Populations d’Oiseaux du Quebec
(EPOQ). This is North America’s longest-running and largest checklist
compilation program, and data are collected according to guidelines de-
signed to maximize scientific value of such projects (Dunn 1995). The
EPOQ trends in Cyr and Larivee’s (1993) study had the same sign (pos-
itive or negative) as BBS trends for Quebec in 62% of the 74 species
analyzed. These results are not strong, but the analysis was of simple
presence/absence data which are limited in ability to detect trends (Bart
and Klosiewski 1989).

The aim of this paper is to examine more closely whether checklists
might contain useful information on population trends and to determine
whether further analyses would be worthwhile.

METHODS

EPOQ data are semi-standardized in that each record contains the number of birds seen
or heard on a single day’s visit to a single locality (within one minute of latitude and
longitude, or roughly 3.2 km^; Cyr and Larivee 1993, 1995). Most lists are submitted by
experienced birders, and the vast majority come from the whole length of the St. Lawrence
corridor in southern Quebec (map in Cyr and Larivee 1995). Data are quite well distributed
over all possible dates (individual days within a year). Although there are fewer than 30
checklists for most dates (54% of spring dates, 89% of fall dates), there are only 29 dates
in the 22-year analysis period with no checklists at all (0.2% of spring dates and 1.2% of
fall dates), all in the early 1970s. The average number of lists per date increased from 4.5
to 25.4 over the study period. We analyzed all available data within the chosen date limits
(see below) regardless of geographic location, length of daily birding trips, number of ob-
servers or weather conditions, but did take into account the seasonal pattern in numbers of
birds seen, as described below.

We selected data from the spring and fall migration “windows” for each of 58 songbirds
(Table 1) for the period 1971-92. This ensured that observations of breeding birds were not
mixed with observations of migrants, as could occur if we used data from a single period
covering the migration periods of all species. To determine these windows, average daily
abundance was plotted against date for each species. Dates were then chosen that included

542

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Species and Codes for Figures, Quebec BBS Trends, and EPOQ Trends (Based on

Abundance, Full Data Set)“

Species

Code

(for

figures)

EPOQ trend

BBS trend

Spring

Fall

Chimney Swift (Chaetura pelagica)

a

-0.4

-1.8

-3.5

Ruby-throated Hummingbird {Archilochus colubris)

b

1.8*

1.7*

1.0

Northern Flicker {Colaptes auratus)

c

-0.7 +

0.2

-2.1 +

Yellow-bellied Sapsucker (Sphyrapicus varius)

d

0.3

-0.6

-0.2

Great Crested Flycatcher {Myiarchus crinitus)

e

0.3

-0.5

-2.5

Eastern Wood-Pewee {Contopus virens)

f

-0.1

-1.0

-1.4

Eastern Phoebe (Sayornis phoebe)

g

0.1

2.0+

0.9

Least Flycatcher {Empidonax minimus)

h

0.6

-0.9

-2.3

Yellow-bellied Flycatcher (£. flaviventris)

i

1.4

-0.3

4.5

House Wren {Troglodytes aedon)

j

-0.8

0.5

-3.8 +

Winter Wren {T. troglodytes)

k

0.7

2.3 +

3.5

Golden-crowned Kinglet {Reguliis satrapa)

1

1.2+

4.0*

-2.0

Wood Thrush {Hylocichla mustelina)

m

-1.3*

1

*

-3.5

Veery {Catharus fuscescens)

n

-0.0

-0.9

-0.1

Swainson’s Thrush (C. ustulatus)

o

-0.0

-3.2*

-2.3

Hermit Thrush (C. guttatus)

p

0.6

-0.1

-0.7

American Robin {Turdus migratorius)

q

1.0

1.0

1.0

Gray Catbird {Dumatella carolinensis)

r

-2.2*

-1.6+

-5.5*

Brown Thrasher {Toxostoma rufum)

s

-2.4*

-2.6

*

SO

1

Solitary Vireo {Vireo solitarius)

t

0.6

4.5*

8.6+

Red-eyed Vireo {V. olivaceus)

u

1.4*

2.0*

2.3*

Warbling Vireo {V. gilvus)

V

2.2*

1.9

1.7

Philadelphia Vireo (V. philadelphicus)

w

2.0*

-0.1

5.2

Tennessee Warbler {Vermivora peregrina)

X

0.5

-1.3

-4.7

Nashville Warbler {V. ruficapilla)

y

0.4

-1.9 +

-4.2

Northern Parula {Parula americana)

Z

0.4

3.7*

-0.3

Black-and-white Warbler {Mniotilta varia)

A

1.1 +

2.0*

4.7 +

Black-thrt. Blue Warbler {Dendroica caerulescens)

B

0.1

2.6*

1.3

Blackburnian Warbler (D. fusca)

C

0.5

4.5*

3.7

Chestnut-sided Warbler (D. pensylvanica)

D

0.5

1.8 +

-6.5

Cape May Warbler {D. tigrina)

F

-1.3*

0.5

-0.2

Magnolia Warbler {D. magnolia)

F

1.3*

1.4

5.8

Yellow-rumped Warbler (D. coronata)

G

0.5

0.4

2.8 +

Black-throated Green Warbler {D. virens)

H

0.6

2.0*

0.0

Bay-breasted Warbler (D. castanea)

I

-1.3 +

-0.5

-9.0

Yellow Warbler {D. petechia)

J

0.8*

0.2

2.9+

Mourning Warbler {Oporornis Philadelphia)

K

0.4

0.0

0.2

Canada Warbler {Wilsonia canadensis)

L

0.7

-0.4

-0.6

Ovenbird {Seiurus aurocapillus)

M

1.2*

-0.0

-0.4

Northern Waterthrush {S. noveboracensis)

N

1.4*

-0.7 .

-0.5

Common Yellowthroat {Geothlypis trichas)

O

0.4

0.2

-2.2 +

American Redstart {Setophaga ruticilla)

P

1.2*

1.5*

-2.0'

Dunn et al. • POPULATION TRENDS FROM CHECKLISTS

543

Table 1

Continued

Species

Code

(for

figures)

EPOQ trend

Spring

Fall

BBS trend

Rose-breasted Grosbeak (Pheucticus ludovicianus)

Q

-0.2

-2.9*

-4.8*

Vesper Sparrow (Pooecetes gramineus)

R

-0.3

-1.0

-6.8*

Savannah Sparrow (Passercuhis sandwichensis)

S

-0.8*

-0.6

-2.3*

Song Sparrow (Melospiza melodia)

T

-0.6

0.4

-0.2

Chipping Sparrow (Spizella passerina)

U

1.1*

2.0*

1.4

Dark-eyed Junco (Junco hyemalis)

V

1.3

1.9+

-3.7

White-throated Sparrow (Zonotrichia albicollis)

W

-0.6

-0.6

-1.9*

Lincoln’s Sparrow {Melospiza lincolnii)

X

1.7*

1.6*

-4.0*

Swamp Sparrow (M. georgiana)

Y

0.3

1.1

-5.0

Bobolink (Dolichonyx oryzivorus)

Z

-1.8*

-1.2+

-6.2*

Eastern Meadowlark (Sturnella magna)

2

-1.5*

-5.3*

Red-winged Blackbird (Agelaiiis phoeniceus)

3

-3.5*

-3.5*

Brown-headed Cowbird {Molothrus ater)

4

-4.3*

-7.2*

Common Crackle {Quiscalus quiscida)

5

-0.0

0.0

Northern Oriole {Icterus galbula)

6

-0.4

-1.6

-2.1

Scarlet Tanager {Piranga olivacea)

7

-0.6

-1.7

-1.8

"Significance of trends (1971-92) shown by: + = 0.05 < P < 0.10, * = P < 0.05.

the seasonal rise and fall of numbers except for about one week at each end of the season,
thus excluding the transitions between migration and stable numbers of either breeding or
wintering birds. Of the 58 species analyzed, migration windows for 50 had also been cal-
culated for Long Point, Ontario (Hussell et al. 1992). Timing of peaks and early/late dates
differed between the provinces, but the “windows” (which excluded extreme dates) were
very similar in both data sets. For convenience, the Long Point dates were used when
available. Fall migration windows in Quebec were not clearly definable from EPOQ data
for Eastern Meadowlark, Red-winged Blackbird, Brown-headed Cowbird, and Common
Crackle; (scientific names in Table 1) so these species were excluded from fall analyses.

We calculated annual indices of abundance for each season for each species, using a
regression procedure that adjusted the daily total of a species according to date within the
season (adapted from the method described in Hussell et al. 1992). If we had merely cal-
culated mean daily count, results would be heavily influenced by numbers seen in peak
migration periods and especially by records from “fall-outs” (when heavy migration is
halted by a weather front). Instead our approach determines whether the average count for
each date (a single day in a single year) is higher or lower than the long-term average count
for that date. The resulting annual index of abundance, therefore, reflects the average degree
of positive or negative deviation from the expected daily values across the entire season.

We did not attempt to correct the data for weather effects or uneven distribution of observers
throughout the season or the province. Such factors introduce variability to annual indices, but
our assumption was that they did not change systematically through time and, therefore, should
not contribute to spurious trends in bird numbers. Those factors most likely to produce consistent
bias over time — improvement of skills or change in birders’ search behavior, see introduction —
cannot in any case be mitigated by data selection or analysis procedures.

544

THE WILSON BULLETIN • Vo/. 108, No. 3, September 1996

Analysis details were as follows. The dependent variable in the regression (run separately
for each species for each season) was log (mean daily count + 1), where “daily count”
was number of birds per hour in the field for a single checklist, and one was added to the
mean to allow log transformation of zeros. Each case was weighted by the number of
checklists used to calculate daily mean abundance. Use of “birds/hr” helps standardize
values from field trips of different lengths. Log transformation addresses the assumptions
of the regression procedure by changing multiplicative to additive effects and by bringing
the distribution of daily counts closer to normality (raw counts are skewed).

Independent variables included first to sixth order terms for day (day = 0 for a day near
the center of the species-specific migration window) and dummy variables for each year
except for one reference year (e.g., Y89 = 1 if year is 1989, otherwise Y89 = 0). The date
terms allowed modelling of a relatively complicated seasonal pattern without adding so
many terms as to produce overfit. Annual abundance indices were calculated from the
coefficients of the dummy variables for year that were estimated in the regression. The
annual abundance index was the value of the adjusted mean for year plus one-half of the
error variance of the regression (so that corrected estimates in the original scale represent
the mean instead of the median; see references in Hussell et al. 1992) back-transformed to
the original scale by exponentiating and subtracting one.

A second analysis, similar to the above, was used to calculate annual indices based on
frequency (the daily proportion of checklists on which the species was reported present).
The only differences were that the dependent variable in the regression was the square root
of the arcsin-transformed daily proportion, with appropriate adjustment prior to transfor-
mation of proportions equal to 0 or 1 (Snedecor and Cochran 1967:327—328), and we did
not add half the error of the variance prior to back-transformation. We refer to this as the
“date-adjusted frequency” index.

Trends were calculated separately for spring and fall indices. Those based on abundance
were calculated from weighted linear regression of the log of the annual indices on year,
('j'here was no need to add a constant before transformation because annual indices were
never equal to zero.) Trends based on frequency were calculated with weighted linear re-
gression of the square root of arcsin-transformed annual indices. In all trend calculations,
weights were proportional to the number of checklists contributed each season during the
species-specific migration period.

The number of lists compiled by EPOQ has increased steadily over the period analyzed
from about 2,000 to about 10,000 annually (Cyr and Larivee 1995). In an attempt to cir-
cumvent possible bias from this source, as well as to determine what sample size might be
sufficient, we reran all procedures on data sets consisting of 1000, then 500, cases selected
randomly from each season each year.

EPOQ trends were compared to trends from the Breeding Bird Survey (BBS) for Quebec
for the same set of years. BBS is a standardized roadside survey in which volunteers make
50 3-min stops every 0.8 km along prescribed routes, recording all birds seen and heard
(Peterjohn 1994). Geographical coverage of Quebec is roughly equivalent in BBS and
EPOQ. BBS trends were calculated using the Canadian Wildlife Survey version of the route
regression analysis method (Erskine et al. 1992). All species analyzed were present on at
least 22 BBS routes in Quebec during the study period. (The recommended number for
meaningful analysis is 15.)

RESULTS

Full data set. — EPOQ trends based on abundance indices, both in
spring and fall, were significantly correlated with BBS trends (Table 2).

Dunn et al. • POPULATION TRENDS FROM CHECKLISTS

545

Table 2

Spearman Rank Correlation Coefficients between EPOQ and BBS Trends for

Quebec, 1971-1992“

EPOQ indices calculated as:

Season

Abundance

(Birds/hr)

(N)

Date-adjusted

frequency

(N)

Full data set

Spring

0.58***

(45,578)

0.51***

(66,821)

Fall

0.55***

(27,682)

0.48***

(39,842)

1000 cases per season

Spring

0.53***

(19,804)

0.38**

(21,864)

Fall

0.47***

(17,334)

0.32*

(20,253)

500 cases per season

Spring

0.50***

(10,728)

0.35**

(11,000)

Fall

0.43***

(10,536)

0.09

(10,959)

“ See methods for definition of the two EPOQ trend calculations. 58 species in spring, 54 in fall. Significance of correlation
(two-tailed tests): * = P < 0.05, ** = p < 0.01, *** — P < 0.001. Total sample size in parentheses.

However, scatter plots showed that correspondence between the programs
was not entirely one-to-one (Figs. lA and IB); that is, points were not
evenly distributed about the dashed line representing equality of trends.
EPOQ produced markedly more positive trends than BBS in those species
that BBS showed to be declining. Significance of trend in EPOQ did not
reflect significance in BBS (Table 3), although trends that were significant
in both programs agreed in sign in all cases but one.

EPOQ trends based on date-adjusted frequency indices were also sig-
nificantly correlated with BBS trends in both seasons (Table 2). The
magnitude of trends based on frequency cannot be compared directly to
BBS magnitude because the scales differ (BBS trends are expressed as
annual percent change in abundance; EPOQ trends are the annual change
in arcsin transformed annual proportions of checklists with the species
present). However, if the two programs monitor the same phenomena,
then the directions of trends should agree. This was largely the case for
trends based on date-adjusted frequency indices for spring (Fig. 2A,
which has a similar pattern to the spring abundance trends in Fig. lA).
However, fall frequency trends based on EPOQ data were much more
likely to be negative than were BBS trends (Fig. 2B) and were also
more negative than EPOQ trends based on abundance (Fig. IB). Sig-
nificance of EPOQ trends based on frequency did not reflect significance
in BBS (Table 3).

546

THE WILSON BULLETIN • Vol. 108, No. 3. September 1996

BBS TREND

Fig. 1. Trends in EPOQ abundance indices for spring (Part A, top) and fall (Part B,
bottom) plotted against BBS trends. Trends expressed as annual percent change in abun-
dance. Dashed line shows one-to-one correspondence. See Table 1 for species codes.

Reduced data set. — When the data set was reduced, analyses gave qual-
itatively similar results to all those presented above. Correlation coeffi-
cients were reduced, however (Table 2), due to increased scatter in EPOQ
trends.

Dunn et at. • POPULATION TRENDS FROM CHECKLISTS

547

Table 3

Number of Species with Significant or Marginally Significant (P < 0. 1 ) Trends in
Quebec BBS and EPOQ (Full Data Set)

EPOQ abundance trend EPOQ frequency trend

Trend

significant in: Spring Fall Spring Fall

EPOQ only 12 15 15 15

Both EPOQ and BBS 11 7 14 9

BBS only 7 9 5 8

DISCUSSION

Comparison of analyses. — EPOQ indices based on abundance of birds
gave the best correspondence to BBS, producing trends that showed the
highest level of agreement in direction and magnitude in both seasons.

Bart and Klosiewski (1989) found that BBS trends based on frequency
indices generally had the same sign as trends based on abundance (pos-
itive or negative), but the two types of trends did not compare well in
magnitude. We had similarly expected that EPOQ trends based on fre-
quency would not correspond as well to BBS trends as those based on
abundance, but this was borne out only by fall results (compare Fig. IB
with Fig. 2B).

Evaluation of checklists in monitoring populations. — The primary uses
made of checklist data do not include population monitoring but rather a
wealth of other applications such as documentation of range, timing of
occurrence in a given region, unusual appearances, and site-specific spe-
cies composition. These applications do not depend on standard obser-
vation protocol and appropriate sampling framework, whereas population
monitoring does if it is to be statistically defensible. Nonetheless, our
results suggest that checklist data, even when uncorrected for likely
sources of spurious variability, do contain information on population
trends, albeit biased. (We assume for the purpose of this discussion that
BBS is an accurate, unbiased indicator of trends, but of course we cannot
be certain of this.)

The positive bias in EPOQ trends (Fig. 1) is just what we might expect
of checklist data as a result of improving skills and optical aids (see Sauer
et al. 1994) or as a result of shifts by birders to more productive birding
spots as species decline in previously-favored sites. The positive bias of
EPOQ trends means that they are less reliable indicators of magnitude
than are BBS trends. Analysis procedures could be altered to reduce vari-
ation introduced to EPOQ indices by factors such as uneven temporal and
geographic distribution of observers, but this would likely help only to

548

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Eig. 2. Trends in date-adjusted frequency indices from ^POQ for spring (Part A, top)
and fall (Part B, bottom) plotted against BBS trends. Trends expressed as annual percent
change in abundance (BBS) and annual change in arcsin transformed annual percentages
(see text). See Table 1 for species codes.

Dunn et al. • POPULATION TRENDS FROM CHECKLISTS

549

improve precision ot trend estimates without altering long-term bias and,
therefore, may not be worth the effort involved. Despite the bias, however,
EPOQ abundance indices produced very few “false negatives” (Fig. 1).
Thus, while an increasing trend in EPOQ does not necessarily indicate a
true increase, a negative EPOQ trend based on abundance is evidently
quite a reliable indicator that some kind of decline is actually taking place.
(EPOQ frequency indices produced false negatives much more often; Fig.
2.) Declines are of more interest for conservation alerts than are increases,
and checklist programs appear to offer a means of detecting some (though
not all) declines in species that are poorly covered by standard population
monitoring programs. It should therefore be of value to analyze EPOQ
data from the migration season for species that breed primarily in tundra
or northern boreal zones and for which we have no other data on popu-
lation trend.

ACKNOWLEDGMENTS

Thanks are due to the thousands of dedicated birders and record-keepers in Quebec with-
out whom this analysis could not have been attempted. Peter Blancher, Brenda Dale, and

an anonymous reviewer made helpful comments on the manuscript.

LITERATURE CITED

Bart, J. and S. P. Klosiewski. 1989. Use of presence-absence to measure changes in avian
density. J. Wildl. Manage. 53:847-852.

Cyr, a. and j. Larivee. 1993. A checklist approach for monitoring neotropical migrant
birds: twenty-year trends in birds of Quebec using EPOQ. Pp. 229-236 in D. M. Finch
and P. W. Stangel (eds.). Status and management of Neotropical migratory birds. U.S.
Forest Serv. Gen. Tech. Rept. RM-229, Fort Collins, Colorado.

AND . 1995. Atlas saisonnier des oiseaux du Quebec. Presses Univ. Sher-
brooke et Soc. Loisir Ornithol. Estrie, Inc., Sherbrooke, Quebec.

Dunn, E. H. 1995. Recommended methods for regional checklist programs. Unpubl. rept.
of North Amer. Migration Monitoring Council. 12 pp. (Available from first author.)

AND D. J. T. Hussell. 1995. Using migration counts to monitor landbird popula-
tions: review and evaluation of current status. Pp. 43-88 in D. M. Power (ed.). Current
ornithology, Vol. 12. Plenum Press, New York, New York.

Erskine, a. j., B. T. Collins, E. Hayakawa, and C. Downes. 1992. The cooperative
Breeding Bird Survey in Canada, 1989-91. Can. Wildl. Serv. Prog. Notes 199.

Hussell, D. J. T, M. H. Mather, and P. H. Sinclair. 1992. Trends in numbers of tropical-
and temperate-wintering migrant landbirds in migration at Long Point, Ontario, 1961-
1988. Pp. 101-1 14 in J. M. Hagan and D. W. Johnston (eds.). Ecology and conservation
of Neotropical migrant landbirds. Smithsonian Inst. Press, Washington, D.C.

Peterjohn, B. G. 1994. The North American Breeding Bird Survey. Birding 26:386—398.

Sauer, J. R., B. G. Peterjohn, and W. A. Link. 1994. Observer differences in the North
American Breeding Bird Survey. Auk I I 1:50-62.

Snedecor, G. W. and W. G. Cochran. 1967, Statistical methods, sixth ed. Iowa State Univ.
Press, Ames, Iowa.

Wilson Bull., 108(3), 1996, pp. 550-555

EFFECT OF MATE REMOVAL ON SINGING BEHAVIOR
AND MOVEMENT PATTERNS OF FEMALE
NORTHERN CARDINALS

David B. McElroy and Gary Ritchison

Abstract. — We temporarily removed the mates of four female Northern Cardinals (Car-
dinalis cardinalis) during the pre-nesting period in an attempt to clarify the functions of
singing by females. Mate removal had no significant effect on singing rates, number of
movements per hour, or distance moved per hour. The failure to increase singing rates after
mate removal suggests that singing by female cardinals is not used to attract new mates and
the absence of any change in movement patterns suggests that females may not actively
seek new mates. When singing, female cardinals were usually accompanied by a singing
male and, prior to and after mate removal, females often duetted with their mates. Female
cardinals may sing and duet with mates to advertise the presence of a female and the mated
status of a male. Duetting may also permit pairs to learn each other’s songs. Such learning
may represent a form of investment important in maintaining a pair bond and may also,
later in the season, permit more efficient intrapair communication. Received 26 Oct. 1 995,
accepted 10 Feb. 1996.

Although singing by females is now well documented in several species
of passerines (e.g., Arcese et al. 1988, Johnson and Kermott 1990, Bap-
tista et al. 1993), the functions of such singing often remain unclear. Most
suggested functions have been based largely on observations, and few
investigators have used experimentation. When used, experimentation has
been limited to playback experiments (e.g., Ritchison 1983, 1986; Arcese
et al. 1988; Baptista et al. 1993). In contrast, investigators examining the
singing behavior of male passerines have used a variety of experimental
techniques, including mate removal experiments. In these studies, the
singing behavior of males is observed prior to and after removal of mates.
Typically, males increase singing rates after removal of their mates, sug-
gesting that singing plays a role either in maintaining contact with mates
or in attracting new mates (Wasserman 1977, Krebs et al. 1981, Cuthill
and Hindmarsh 1985, Otter and Ratcliffe 1993).

Female Northern Cardinals (Cardinalis cardinalis) sing primarily dur-
ing the period after males establish territories and before nesting begins
(Ritchison 1986). Singing females usually are accompanied by singing
mates (Lemon 1968, Ritchison 1986). Although these observations sug-
gest that singing by female Northern Cardinals may be important in in-
trapair communication, perhaps playing some role in the formation of the
pair bond (Ritchison 1986), additional information is needed. The objec-

Dept. of Biological Sciences, Eastern Kentucky Univ., Richmond. Kentucky 4047.5.

550

McElroy and Riichison • SINGING OF FEMALE CARDINALS

551

live ot our study was to use mate removal experiments in an attempt to
clarify the functions of singing by female Northern Cardinals.

METHODS

We conducted this study during the pre-nesting periods (January through early May) of
1990, 1992, and 1993 at the Central Kentucky Wildlife Management Area located 17 km
southeast ot Richmond, Madison County, Kentucky. Male and female cardinals were cap-
tured in mist nets and individually marked with colored leg bands and plastic tape attached
to the tail (Ritchison 1984). Beginning in March, marked cardinals were observed to deter-
mine the identity of pairs and the location and boundaries of territories. Paired females (N
— 4) were subsequently re-captured and fitted with radio-transmitters. Removal experiments
with these females began on 2 April and 20 April 1990, 19 April 1992, and 29 April 1993,
respectively.

Experiments were divided into three periods: pre-removal, removal, and post-removal.
Each period lasted two days and focal females were observed for three hours each day. All
observations were made during the period from sunrise to 1 1 :00. Pre-removal periods began
no sooner than two days after females had been fitted with transmitters. After the pre-
removal period, the female s mate was captured by luring him into a mist net using the
playback of cardinal songs. During the two-day removal period, males were kept in the lab
in a wire mesh cage and provided food and water. The post-removal period began with the
release of the male.

The movements of focal females were monitored using a receiver (Telonic’s TR-2) with
a two-element antenna. The focal female’s location was marked on a map of the study area
every 15 min. All songs were recorded using a Uher 4000 Report Monitor tape recorder
with a Dan Gibson parabolic microphone or a Marantz recorder with a directional micro-
phone. Observations and recordings of females were made from a distance of 20 m or more.

Recordings were analyzed using a Kay Elemetrics Sonagraph (Model 5500). Eor each
observation period, we noted the number of songs and number of bouts, with a bout defined
as a series of songs separated by intervals less than 30 sec. For each song, we determined
the song type (see figures 1 and 2 in Ritchison 1988), number of syllables, and duration.
Each bout was classified as either an accompanied bout or a non-accompanied bout, with
an accompanied bout defined as a bout during which a conspecific male (either in the
female’s territory or in an adjacent territory) sang at some time during the bout. Bouts of
female song during which their mates also sang were classified as duetting bouts.

Possible differences among periods for all variables were tested using repeated measures
analysis of variance (SAS 1989). Repeated measures analysis provides a test for interactions
(Beal and Khamis 1990), and we also examined possible interactions between individuals
and periods. Tests for non-random use of song types were made using likelihood ratio chi-
square tests (SAS 1989). All values are presented as means ± one standard error.

RESULTS

The singing rates of female cardinals did not vary significantly among
test periods (F = 1.34, P = 0.34), with no significant interaction between
individual and period (F = 1.53, P = 0.21). Females uttered an average
of 20.6 ± 4. 1 songs/h during the pre-removal period, 20. 1 ± 6.5 songs/h
during the removal period, and 22.7 ± 8.8 songs/h during the post-re-
moval period.

Female cardinals exhibited no significant variation among periods in

552

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

the number of syllables per song {F - 1.93, P = 0.24). We found a
significant interaction between individual and period (F = 2.47, P =
0.03), however, no clear trend was apparent. Two females used songs
with more syllables during the removal period, one female uttered songs
with more syllables during the post-removal period (although this female
sang only 1 1 songs during the pre-removal and removal periods), and the
songs of the fourth female showed little variation in number of syllables
per song among periods.

Female cardinals exhibited no significant variation among periods in
song duration (F = 0.6, P — 0.58), with no significant interaction between
individual and period (F = 1.67, P = 0.14). Also, we found no significant
differences among periods in intersong intervals (F = 0.21, F = 0.82).
There was a significant interaction between individual and period (F =
4 7^ P = 0.0003), however, no clear trends were apparent. Three females
exhibited little variation in intersong intervals among periods while in-
tersong intervals for the fourth female were longer during the removal
and post-removal periods than during the pre-removal period.

All four female cardinals exhibited significant variation (likelihood ra-
tio tests, P < 0.0001) in the use of song types during the three periods.
However, no consistent trends were found among the four females in the
use of particular song types during particular periods.

During most bouts of song (N = 172 of 248 or 69.4%), female car-
dinals were accompanied by singing males, either their mate (duetting
bouts) or another conspecific male. Females sang with their mates during
88 bouts and with other conspecific males during 84 bouts. During seven
bouts, females sang with both their mate and an intruding male. Females
were accompanied by males in 69 bouts during the pre-removal period,
46 during the removal period, and 57 during the post-removal period. The
number of non-accompanied bouts declined from 45 during the pre-re-
moval period to 18 during the removal period to 13 during the post-
removal period.

Female cardinals showed no significant variation among periods either
in number of movements per hour (F = 1.57, P = 0.3) or distance moved
per hour (F = 0.14, F = 0.63). No significant interactions were noted for
either movements per hour (F = 0.36, F = 0.87) or distance moved per
hour (F = 0.69, F = 0.63). For all females and periods combined, the
mean number of movements per hour was 4.07 — 2.65 while the mean
distance moved per hour was 109.5 ± 59.4 m. All four females remained
on their respective territories after removal of their mates.

Trespassing by conspecific males occurred significantly more often dur-
ing the removal period (F = 19.6, F = 0.004), with an average of 3.16
± 1.02 intrusions per hour during the removal period, 0.46 ± 0.26 intru-

McElroy and Ritchison • SINGING OF FEMALE CARDINALS

553

sions per hour during the pre-removal period, and 2.49 ± 1.11 intrusions
per hour during the post-removal period. The number of singing bouts
by intruding males in the territories of focal females varied significantly
among periods (F = 28.8, P = 0.0018), with an average of 1.64 ± 0.35
bouts/h during the removal period, 0.09 ± 0.05 bouts/h during pre-re-
moval, and 1.07 ± 0.42 bouts/h during post-removal. Focal females sang
with intruding males an average of 0.6 ± 0.19 times/h during removal
and 0. 19 ± 0.04 times/h during post-removal, a significant difference (F
= 12.6, P = 0.011). No singing with intruders occurred during the pre-
removal period.

Three temporarily-removed males maintained their territories and their
mates after being released. One male lost his territory and his mate to
another male during the removal period and was unable to reclaim either
after release.

DISCUSSION

Female Northern Cardinals in our study did not increase singing rates
when mates were removed. In contrast, passerine males typically increase
singing rates after removal of a mate (Wasserman 1977, Krebs et al. 1981,
Cuthill and Hindmarsh 1985, Otter and Ratcliffe 1993). Increased song
output following mate removal may indicate an attempt to re-establish
contact with the absent mate or to attract a new mate (Krebs et al. 1981).
Johnson and Kermott (1990) found that female House Wrens {Troglodytes
aedon) sang primarily after losing contact with mates and these songs
appeared to redirect a mate’s attention to the singing female. Such be-
havior was not observed in our study, and this failure to increase singing
rates after removal of mates suggests that singing by female cardinals is
not used to establish or maintain contact with mates.

There is little evidence that passerine females sing to attract new mates.
Baptista et al. (1993) reported that one widowed female White-crowned
Sparrow (Zonotrichiz leucophrys) produced long bouts of loud song and
suggested that she may have been advertising for a mate. Female Northern
Cardinals did not increase song output after mate removal, and this sug-
gests that singing by female cardinals is not used to attract new mates.

Female cardinals in our study exhibited no significant changes among
periods in either number of movements or distance moved per hour. In
contrast, Klatt and Ritchison (1994) found that female Eastern Screech-
Owls (Otus asio) made significantly more movements and moved signif-
icantly greater distances after mate removal and suggested that such be-
havior would increase the chances either of re-establishing contact with
a mate or attracting a new mate. The absence of any change in movement
patterns after mate removal suggests that female cardinals may not ac-

554

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

lively seek either to re-establish contact with a mate or attract a new mate.
Rather, females may use a more passive approach. The absence of a ter-
ritorial male, as during the removal periods in our study, may indicate to
other males the availability of a teiTitory and, perhaps, of a female. Fe-
male cardinals may wait for neighboring or floater males to trespass, then
seek to pair with the male that eventually obtains the territory.

Our results indicate that singing by female cardinals does not deter
male intrusion into a female’s territory. Male intrusion rates increased
during the removal period even though females continued to sing. Female
song may, however, deter trespassing by other females. If singing by
females is directed primarily to conspecific females, then rates would not
be expected to increase in the absence of the male. Thus, female cardinals
may sing, and duet with mates, to advertise the presence of a female and,
perhaps, the mated status of a male. In further support of this hypothesis,
songs uttered by cardinals during duets are typically normal or high vol-
ume songs that can be heard in neighboring territories (pers. observ.).
Such advertisement may be important in mate retention because unpaired
females (floaters) are present in our cardinal population (Ritchison et al.
1994). Further, some male cardinals participate in extra-pair copulations
(EPCs) (Ritchison et al. 1994). Female cardinals may reduce the likeli-
hood that their mate will participate in EPCs by duetting with and ad-
vertising the mated status of their mate. Reduced participation in EPCs
by mates may be beneficial to females because males that do not engage
in EPCs (or engage in fewer EPCs) may provide more parental care
(Westneat et al. 1990).

Singing by female Northern Cardinals may also permit pairs to learn
each other’s songs. Paired cardinals may sing from or near nests during
the incubation and nestling periods, and Halkin (1990) proposed that fe-
male cardinals on the nest may match her mate’s song type to inform the
male that he need not visit (bring food to) the nest. Eor such communi-
cation to occur, paired cardinals must know the song type repertoires of
their mates. Duetting prior to nesting may permit paired cardinals to be-
come familiar with, and perhaps learn, song types in the repertoire of
their mate. Lending support to this hypothesis, we found that male and
female cardinals sometimes matched song types during duets (pers. obs.).

Matching song types during duets could also serve other functions.
Learning a mate’s song types may represent a form of investment that
may be important in establishing and maintaining a pair bond (Wickler
1980). Thus, if male and female cardinals must learn the songs of their
mates, then duetting may play a role in establishing and maintaining pair
bonds.

In summary, our results suggest that female Northern Cardinals do not

McElroy and Ritchison • SINGING OF FEMALE CARDINALS

555

sing to re-establish contact with an absent mate or to attract a new mate.
Rather, female cardinals sing (and duet with mates) to advertise the fe-
males presence and the paired status of her mate. Duetting may also
permit paired cardinals to learn other’s songs. Such learning may repre-
sent a form of investment important in maintaining a pair bond (Wickler
1980) and may also, later in the season, permit more efficient commu-
nication in the vicinity of nests (Halkin 1990).

ACKNOWLEDGMENTS

We thank Ken Yasukawa and an anonymous reviewer for helpful comments on the manu-
script. Financial support was provided by Eastern Kentucky University.

LITERATURE CITED

Arcese, R, P. K. Stoddard, and S. M. Hiebert. 1988. The form and function of song in
female Song Sparrows. Condor 90:44-50.

Baptista, L. E, P. W. Trail, B. B. Dewolfe, and M. L. Morton. 1993. Singing and its
functions in female White-crowned Sparrows. Anim. Behav. 46:511-524.

Beal, K. G. and H. J. Khamis. 1990. Statistical analysis of a problem data set: correlated
observations. Condor 92:248-251.

CuTHiLL, I. AND A. Hindmarsh. 1985. Increase in Starling song activity with removal of
mate. Anim. Behav. 33:326-328.

Halkin, S. L. 1990. Singing from the nest: intrapair communication in cardinals. Ph.D.

diss., Univ. of Wisconsin, Madison, Wisconsin.

Johnson, L. S. and L. H. Kermott. 1990. Structure and function of female song in a north-
temperate population of House Wrens. J. Field Ornithol. 61:273-284.

Klatt, P. H. and G. Ritchison. 1994. The effect of mate removal on the vocal behavior
and movement patterns of male and female Eastern Screech-Owls. Condor 96:485-493.
Krebs, J. R., M. Avery, and R. J. Cowie. 1981. Effect of removal of mate on the singing
behavior of Great Tits. Anim. Behav. 29:635-637.

Lemon, R. E. 1968. The relation between the organization and function of song in
Cardinals. Behaviour 32:158-178.

Otter, K. and L. Ratcliffe, 1993. Changes in singing behavior of male Black-capped
Chickadees {Pams atricapillus) following mate removal. Behav. Ecol. Sociobiol. 33:
409-414.

Ritchison, G. 1983. The function of singing in female Black-headed Grosbeaks: family-
group maintenance. Auk 100:105-1 16.

. 1984. A new method of marking birds. N. Amer. Bird Bander 9(3):8.

. 1986. The singing behavior of female Northern Cardinals. Condor 88:156-159.

, P. H. Klatt, and D. F. Westneat. 1994. Mate guarding and extra-pair paternity
in Northern Cardinals. Condor 96:1055-1063.

SAS Institute. 1989. SAS user’s guide: statistics. 1989 ed. SAS Institute Inc., Cary, North
Carolina.

Wasserman, F. E. 1977. Mate attraction function of song in the White-throated Sparrow.
Condor 79:125-127.

Westneat, D. E, P. W. Sherman, and M. L. Morton. 1990. The ecology and evolution of
extra-pair copulations in birds. Curr. Ornithol. 7:331-369.

Wickler, W. 1980. Vocal ducting and the pair bond. I. Coyness and partner commitment.
A hypothesis. Z. Tierpsychol. 52:201-209.

Wilson Bull., 108(3), 1996, pp. 556-566

RADIO TELEMETRY DOCUMENTS 24-HOUR FEEDING
ACTIVITY OF WINTERING LESSER SCAUP

Christine M. Custer,' Thomas W. Custer,' and Daniel W. Sparks^

Abstract. — We used radio telemetry to record 198 h of feeding behavior of five Lesser
Scaup (Aythya affinis) on the Indiana Harbor Canal in northwestern Indiana during January
and February 1994. Lesser Scaup fed for short periods of time intermittently during each
24-h period. Lesser Scaup fed a total of 96 min during the day and 226 min during the
night. They fed more between sunset and midnight (31.9% of the period, P = 0.(X)3) than
between sunrise and noon (1 1.6%) or noon and sunset (19.5%); time spent feeding between
midnight and sunrise (26.3%) did not differ from other times of day. Mean dive duration
(22.9 ± 0.64 sec) did not vary by time of day (P = 0.186-0.744). These results are the first
24-h feeding activity reported for individually marked Lesser Scaup. Received 27 Sept. 1995,
accepted 3 Feb. 1996.

Knowledge of both diurnal and nocturnal activity is needed to under-
stand the use of time and energy by waterfowl (Jorde and Owen 1988).
However, estimates of 24-h activity of waterfowl, especially diving ducks,
generally are difficult to obtain and often are imprecise. Night-vision light
intensifiers (NVLI) have been used to document nocturnal activity based
on scan census or focal animal observations (Tamsier 1976, Jorde et al.
1983, Paulus 1984, Takekawa 1987, Bergan et al. 1989). Night obser-
vations, however, are often limited by access and viewing area (Jorde and
Owen 1988, Bergan et al. 1989). Scan counts underestimate feeding ac-
tivity for diving ducks, because some birds are underwater during the
scan (Siegfried 1974). Also, studies of diving ducks using focal animal
methods are impossible to conduct in many situations because it is dif-
ficult to keep track of individual birds in large flocks (C. Custer, pers.
obs.).

Few studies have recorded 24-h activity budgets of Lesser Scaup {Ay-
thya afftnis). Wintering male Lesser Scaup in South Carolina spent <10%
of their time feeding at night and approximately 40% of the day feeding
(Bergan et al. 1989); data were collected using focal-animal sampling
(5-min duration/bird) and NVLI. In contrast, wintering Lesser Scaup on
the Mississippi River in Wisconsin spent 28% of the night feeding and
16% of the day; data were collected using modified scan sampling and
NVLI (Takekawa 1987). We are aware of only one study that quantified
24-h activity budgets of individual waterfowl. The activities of a single
breeding male European Pochard {Aythya ferina) were recorded through-

' National Biological Service, Upper Mississippi Science Center, P.O. Box 818, La Crosse, Wisconsin
54602.

2 U.S. Fish and Wildlife Service, 620 S. Walker, Bloomington, Indiana 47403.

556

Custer et cil. • LESSER SCAUP EEEDING

557

out one 24-h period with the aid of a field glass of high luminosity on a
bright moonlight night (Klima 1966).

The attenuation ot radio signal strength has been used with penguins
to quantify timing and duration of feeding behavior (Trivelpiece et al.
1986). Radio telemetry has not been used to quantify feeding of diving
ducks, however. Our objective was to quantify 24-h feeding activity of
Lesser Scaup wintering on the Indiana Harbor Canal (IHC) (41°38'N
87°28'W) using radio telemetry.

STUDY AREA AND METHODS

The IHC and Grand Calumet River system (Fig. 1) contains some of the last remaining
wildlife habitat within the urban, industrial corridor that dominates the south shore of Lake
Michigan (Brock 1986). Only 50 of 10,000 acres of inland beach-ridge dune and swale
habitat still remain (Bacone 1979), and these wetlands along with the Grand Calumet River
and IHC provide resting, feeding, and loafing habitat for migrating and wintering birds
(Brock 1986), and breeding habitat for Black-crowned Night-Herons {Nycticorax nvctico-
rax). Barn Swallows (Hirundo rustica). Herring Gulls (Larus argentatus), and Mallards
{Anas platyrhynchos). During winter, the IHC is routinely used by 200-300 Lesser Scaup
(J. Simesko, Lake Dock Co., pers. comm.; Custer et al. 1996). Indiana Harbor Canal was
constructed in the early 1900s for navigation and to carry waste discharges from 30 outlets
to Lake Michigan (Bolts 1993). The physical structure of IHC, <50-m wide, open water,
and unlimited access to some observation sites enabled us to monitor 24-h activity budgets.
From the trap site out to Lake Michigan, IHC has straight-sided concrete/steel walls and is
>3m deep with no rooted, submergent vegetation. South of the trap site, soil banks pre-
dominate. The banks slope gradually into the water which becomes shallower (<1.5 m deep)
and supports some aquatic vegetation.

We implanted radio transmitters (164-167 Mhz) in the abdominal cavity (N = 10) or
subcutaneously (N = 2) in 12 male Lesser Scaup trapped in a baited, swim-in corral trap
(Haramis et al. 1987) in IHC (Fig. 1). We stopped trapping after our scaup were radio
marked. Abdominal implants, procured from Advanced Telemetry Systems Inc., were cy-
lindrical (50-mm long, 20-mm diameter) with an internal, coiled antenna and weighed about
20 g. Subcutaneous implants, Holohil Systems Ltd., were disc-shaped (20-mm diameter,
8-mm thick) and implanted in the upper back with an external flexible antenna and weighed
about 5 g. The heaviest transmitter weighed <3% of the duck’s body weight. We implanted
all transmitters within 5 h starting at 20:00 CST, 6 January 1994, and released the scaup at
the trap site at 10:00, 7 January. The transmitters were implanted under sterile conditions;
scaup were anesthetized with isoflurane, the transmitter implanted, and the duck immediately
revived with 100% oxygen (Olsen et al. 1992, Kor.schgen, pers. comm.). We followed ap-
proved Animal Care and Use protocols of Northern Prairie Science Center, Jamestown,
North Dakota. We assumed that the Lesser Scaup we trapped were representative of the
flock present in IHC during this study.

Before establishing data collection protocols, we observed general feeding patterns of
Lesser Scaup for several hours. We found that Lesser Scaup fed while diving in one area
or while slowly swimming, dived and surfaced in a consistent pattern of underwater and
surface times, and did not interrupt feeding with preening, bathing, resting, or other behav-
iors. Feeding individuals were usually >10 m from roosting and resting flocks of Lesser
Scaup and rejoined these flocks after feeding.

We used changes in radio signal strength to determine when a radio-marked Lesser Scaup

558

THE WILSON BULLETIN • Vol. 108. No. 3, September 1996

■ behavioral observation points
□ observation points

Eig. 1. The Grand Calumet River-Indiana Harbor Canal study area. East Chicago, In-
diana, showing the trap site and behavioral observation sites, January-February, 1994.

Custer et al. • LESSER SCAUP FEEDING

559

was feeding, i.e., no signal or a weak signal was received when the duck was under water
(Trivelpiece et al. 1986). Visual observations of a radio-marked Lesser Scaup confirmed
that dives inferred by signal strength were actually dives.

To select an appropriate observation interval to monitor feeding behavior, we listened to
signals from three feeding Lesser Scaup continually for ^30 min each and another radio-
marked scaup for 5 h. Bouts of feeding (N = 15) lasted 11.1 ± 1.39 min (± 1 SE); therefore,
we selected 10 min intervals as the minimum needed to detect feeding behavior.

We established behavioral observation sites at four of 15 locations (Fig. 1) because we
had 24-h access to these four sites. Additionally, these sites were where many of the radio-
marked scaup spent the winter (Custer et al. 1996). On days that we recorded behavior, we
checked each site until we located one or more radio-marked Lesser Scaup. There were
usually ^3 radio-marked Lesser Scaup at a site. We collected behavioral data on the radio-
marked scaup using two methods: 10-min scans and focal-animal sampling (Altmann 1974).
We collected 10-min scan data to estimate percent of time spent foraging. We listened to
each radio signal for 2 min at 10-min intervals and determined whether the duck was feeding
(diving) or not feeding. The rhythmic pattern of ducks diving to feed allowed us to differ-
entiate feeding activity from random changes in signal strength or temporary loss of signal
(Ken ward 1987:130). Secondly, we used focal animal sampling to quantify the duration of
foraging dives. Between 10-min scans, we selected a feeding duck and recorded for 3-6
min the time it spent above and below the water’s surface while feeding. When more than
one scaup was feeding, we alternated focal animal observations equally among the scaup
present.

Behavior was recorded by human observers or by video taping the radio receiver. During
video recordings, we programmed the radio receiver to scan 4—5 frequencies sequentially
for 2 min each. A camcorder was focused on the radio receiver’s display and thus recorded
both the monitored frequency and the audio speaker sounds. The camcorder also recorded
the time of day. Only observations of radio-marked birds whose behaviors were recorded
continuously for >4 h (N = 5 scaup) were included in the analyses. The other seven radio-
marked scaup were located only infrequently or in areas where we could not record contin-
uous behavioral observations.

We recorded behavior during four periods: morning (sunrise to 12:00 h CST), afternoon
(12:01 h to sunset), evening (sunset to 24:00 h), and night (00:01 h to sunrise). Duration of
time periods ranged from 5 h-5 h 20 min for daylight periods and 6 h 20 min-6 h 40 min
during evenings and nights.

For each scaup, we calculated the frequency of consecutive 10-min scans during which
it was feeding and not feeding during each time period. We analyzed frequency data by
time period and by duck with Fisher’s Exact Tests (Zar 1984). Categories for number of
consecutive 10-min scans during which an individual was feeding were 1, 2, and 3 + . Three
or more consecutive scans were combined into one category for frequency analyses to reduce
the number of cells with zeros. When the overall Fisher’s Exact Test was significant, all
pairwise combinations were tested to determine which frequencies differed. An alpha of
0.005 was used for pairwi.se Fisher’s Exact Test comparisons to give an overall P < 0.05
(Neter et al. 1985). We pooled 1 and 2, 3 and 4, 5 and 6, and 1+ consecutive 10-min scans
without feeding for statistical tests to reduce the number of cells with zeros.

Percent of time spent feeding during each time period by each duck was calculated from
the 10-min scan data (Altmann 1974). A fixed-effect, 2-way analysis of variance (ANOVA)
model was used to compare the average percent of time spent feeding among ducks and
time periods. We used Bartlett’s test to te.st the homogeneity of variance assumption of
ANOVA (Zar 1984:181). When variances were not homogeneous, percents were square-
root arcsine transformed. Untransformed percents ± 1 SE are presented in text and tables.

560

THE WILSON BULLETIN • Vo/. 108, No. 3, September 1996

Repeated measures statistics were not possible with these data; therefore, all data were
analyzed and presented by individual duck to account for individual variation. We used an
alpha of 0.05 for all ANOVAs.

We used one-way ANOVA to test our null hypothesis that average time (sec) spent
underwater per dive searching for and retrieving food did not differ among time periods or
among ducks. Empty cells precluded using 2-way ANOVA. Because there were no time
differences, we combined all time periods and tested for differences among individuals.

RESULTS

We recorded 198 h of behavior on five radio-marked Lesser Scaup
between 27 January and 16 February 1994 in IHC. Weather patterns dur-
ing the study were normal; daily maximum temperatures were between
-10° and 0°C with occasional snowfall. The normal high temperature for
February is — 2°C (Bair 1992).

Lesser Scaup fed intermittently for short periods throughout the 24-h
period (Fig. 2). As an example, Lesser Scaup #4566 fed during nine
10-min scan periods between sunset and midnight on 2 February; the first
feeding bout lasted for two consecutive 10-min scans (Fig. 2). The modal
number of consecutive 10-min scans during which Lesser Scaup were
feeding was one for four of five Lesser Scaup (Table 1). Lesser Scaup
#4666 differed from the other four scaup (pairwise Fisher’s Exact Test,
all Ps < 0.002) with a mode of 3+ consecutive 10-min scans with feeding
(Table 1). The median number of consecutive feeding bouts for all scaup
combined was one. Frequency of consecutive 10-min scans with feeding
did not vary by time period (3X4 Fisher’s Exact Test, P = 0.716, N —
98); feeding bouts were not longer or shorter during any particular time
period.

The number of consecutive 10-min scans without feeding did not differ
among ducks (Table 1) or among time periods (4X4 Fisher’s Exact Test
P = 0.318, N = 98). The median number of consecutive scans without
feeding was four for all scaup and time periods combined.

Lesser Scaup fed for 23.7 ± 2.5% (SE) of each 24-h day. They spent
a greater proportion of their time feeding during the evening period (31.9
± 5.07%, N = 16 evening periods) than during the morning (11.6 ±
2.60%, N = 9) or afternoon periods (19.5 ± 4.62%, N = 13). Proportion
of time spent feeding during the night (26.3 ± 4.16%, N = 13) did not
differ from the other three time periods (F = 5.75; df = 3,33; P = 0.003).
The time spent feeding varied among individuals (Table 2), but there was
no interaction between time period and individual duck (P — 0.89, df —
10,33; P = 0.548).

Time spent underwater per dive to search for and retrieve food did not
differ among the four time periods (22.9 — 0.64 sec, N — 57, Ps 0.186

Custer et al. • LESSER SCAUP FEEDING

561

Lesser scaup
#4566

not feeding
feeding

sunset midnight sunrise noon sunset

3 Feb

I I- I ^ I

4 Feb

5 Feb

Fig. 2. Summary of 40+ consecutive hours of 10-min behavior scans for Lesser Scaup
#4566 in Indiana Harbor Canal, East Chicago, Indiana, 2—5 February 1994.

- 0.744). Bird #4566 spent significantly less time underwater than bird
#4616 or bird #4666 (Table 2), however.

DISCUSSION

Our study further demonstrates the need to collect nocturnal data to
better understand the feeding ecology of Lesser Scaup. On some wintering
areas, feeding is more prevalent at night than during the day (Takekawa
1987 and this study). The proportion of time spent feeding during noc-
turnal hours (29%) and diurnal hours (16%) in IHC was strikingly similar
to that of Lesser Scaup on the Mississippi River in Wisconsin which spent
28% of the night feeding and 16% of the day (Takekawa 1987). Tufted
ducks (Aythya fuligula) in Switzerland, a closely related species, also
spent a higher proportion of the night feeding (30-50% of the night) than
they did during the day (<10% of the day) (Pedroli 1982). In contrast.

562

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Number of Consecutive 10-min Scans during Which Radio-marked Lesser Scaup were
Feeding and Not Feeding in Indiana Harbor Canal, East Chicago, Indiana, January-

February 1994

Consecutive

1 0-min

scans

Feeding

Not feeding

Transmitter

frequency

1

2

3 +

1-2

3-4

5-6

7 +

4566

32

17

4

A“

10

8

10

19

A'’

4616

7

2

2

AB

5

2

1

4

A

7029

8

1

0

AB

2

1

1

2

A

4637

7

0

5

B

4

3

0

3

A

4666

0

5

8

C

4

4

1

2

A

^ Frequency distributions sharing same letter are not different among Lesser Scaup, 3X5 Fisher s Exact Test {P <
0.001). Pairwise comparisons to separate individual Lesser Scaup were considered significant at P < 0.005 for an exper-
iment-wide alpha level of P ^ 0.05.

*’4X5 Fisher’s Exact Test {P = 0.699).

male Lesser Scaup in South Carolina fed <10% of the night and approx-
imately 40% of the day (Bergan et al. 1989). Human disturbance is often
cited as the reason for nocturnal feeding (McNeil et al. 1992). Our study
and the study in South Carolina, however, were conducted in areas with
little human disturbance; therefore, this would not explain the difference
between the two studies. Neither cold temperatures (down to -10°C) nor
precipitation affected percent of time spent feeding (Cronan 1957, Nose-
worthy 1981, Takekawa 1987), so temperature should not be a factor
when comparing these two studies.

Although the percent of time spent feeding during day and night in the

Table 2

Average Time Spent Feeding and Time Spent Underwater per Dive by Radio-marked
Lesser Scaup in Indiana Harbor Canal, East Chicago, Indiana during January and

February 1994

Transmitter •
number

Time feeding (%)

Time underwater (sec)

Mean

SE

N'

Mean

SE

N“

4666

39.2

Ab

6.67

1 1

25.0

A*-'

0.82

24

4637

36.7

AC

7.04

6

22.9

AB

1.41

8

4566

17.5

BC

1.86

20

19.4

B

0.92

20

4616

17.4

B

4.69

10

26.3

A

2.43

5

7029

9.0

BC

2.12

4

- N = number of time-of-day periods for which percent of time spent feeding or time underwater was calculated.

Means sharing same letter are not different (F = 5.60; df = 4,33; P = 0.002).

' Means sharing same letter are not different (P = 7.90; df = 3,53; P < 0.001).

Custer et al. • LESSER SCAUP FEEDING

563

IHC was different from some other studies, the total amount of time spent
feeding (5.7 h/d) in IHC was similar to Lesser Scaup in South Carolina
(4 h/d) (Bergan et al. 1989), Lesser Scaup on the Mississippi River (4.1
h/d) (Takekawa 1987), and Tufted Ducks in Switzerland (4 8—5 2 h/d)
(Pedroli 1982).

Lesser Scaup feed for short periods of time (median number of con-
secutive scans with feeding was one) followed by longer non-feeding
periods (median number of consecutive non-feeding scans was 4). The
average length of a feeding bout was 11.1 min. We had the longest con-
tinuous record on scaup #4566, which demonstrated this intermittent feed-
ing pattern continually for two days. The feeding patterns of the other

four Lesser Scaup, although less extensive, were consistent with the pat-
tern of #4566.

During the pre-breeding season (May) in Manitoba, Lesser Scaup re-
peated a foraging, bathing/preening, resting/sleeping cycle about every 3
h dunng daylight hours (Siegfried 1974), which was longer than the ap-
proximately 1-h cycle we found in IHC during winter. European Pochards
in the Bohemian highlands also had a 3—4 h activity which was repeated
regularly during a 24-h period during spring (Klfma 1966). Klima (1966)
hypothesized that the open water habitat with its lack of microhabitat
variation, minimal human disturbance, lack of phototaxis in prey behavior,
and tactile feeding by European Pochards contributed to the similarity of
diurnal and nocturnal feeding patterns. Several characteristics of IHC are
similar to that of Bohemia; IHC is a relatively undisturbed location with-
out hunting and has little recreational or public use. The scaup seem to
have habituated to the industrialized setting, and the bright lights mimic
moon-lit nights which are conducive to nocturnal feeding (Adair 1990;
73). Most of the IHC is deep (>3 m) open water, and availability of
benthic prey, mainly oligochaete worms (T. W. Custer, Natl. Biol. Serv.,
unpubl. data), does not vary by time of day (R. Whitman, pers. comm.).
Duration of feeding cycles may be a function of the type of prey con-
sumed and the time needed to handle, process, and digest it. Oligochaetes
are easy to capture, are very soft, and should be processed through the
digestive system more quickly than other more traditional Lesser Scaup
food items such as molluscs and arthropods (Swanson and Bartonek 1970,
Afton et al. 1991, Custer and Custer 1996).

Time spent underwater per dive (x = 23 sec) by Lesser Scaup in our
study was similar to that of Lesser Scaup wintering in Chesapeake Bay
(x = 23.6 sec, G. M. Haramis, Natl. Biol. Sen, pers. comm.) and by Lesser
and Greater scaup wintering in Connecticut (x = 20.4 sec) (Cronan 1957).
However, shorter dive times have been reported for Lesser Scaup in Man-
itoba during spring (x = 10 sec) (Siegfried 1974, 1976); Lesser and Great-

564

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

er Scaup in the Detroit River (jc = 16.4 sec) (Noseworthy 1981); and
Lesser Scaup in South Carolina (x = 16.6 sec) (Alexander and Hair 1977).
Time spent underwater is not related to water depth (Siegfried 1974) but
may be a function of type and abundance of prey being exploited (Nose-
worthy 1981). Dive time may also be a function of individual behavior
that affects distance covered per foot stroke, i.e. diving efficiency (Lov-
vorn et al. 1991). We do not believe that Lesser Scaup used visual cues
to find food in IHC because they feed extensively at night and because
of the similarity in time spent underwater searching and capturing prey
during the day and night.

We feel that the behavior of these five individuals was representative
of the approximately 200 Lesser Scaup (Custer et al. 1996) that wintered
in IHC. Food was plentiful (>400,000 oligochaetes/m^) in the area where
we made out behavior observations (U.S. Fish and Wildlife Service-Bat-
telle 1993), and we did not observe overt aggressive behavior (CMC,
pers. obs.) that might indicate an abnormal situation. Scaup did not defend
foraging sites during winter in South Carolina (Alexander and Hair 1977)
and aggressive interactions were uncommon (Alexander and Hair 1977,
Bergan et al. 1989).

Radio telemetry is an effective technique to monitor feeding behavior
in diving ducks. We were able to recognize individual ducks and use a
video recorder to acquire data remotely, an important consideration when
continuous 24-h data are needed and availability of personnel is limited.
Radio telemetry overcomes (1) the limitations of night-vision light inten-
sifiers (Bergan et al. 1989) and collecting data during inclement weather
and other conditions of poor visibility, (2) the problem of locating birds
during scan counts (Siegfried 1974), and (3) of keeping track of individ-
uals in large flocks during focal animal observations.

ACKNOWLEDGMENTS

We thank Atla.s Steel, Inland Steel, Lake Dock Company, LTV Steel, and Shell Oil
Company for access to their properties; C. M. Chaffee, J. T Conomy, R. K. Hines, C. O.
Kochanny and A. D. Spicer for help with field work, K. P. Kenow for staying up most of
one night implanting radio transmitters, C. E. Korschgen for advice on radio telemetry; and
J. E. Austin, D. G. Jorde, K. P. Kenow, S. J. Maxson, T. C. Michot, K. J. Remecke, E. J.
Rockwell, and an anonymous reviewer for comments on earlier drafts of the manuscript.

LITERATURE CITED

Adair, S. E. 1990. Factors influencing wintering diving duck use of coastal ponds in south
Texas. M.S. thesis, Texas A&M Univ., College Station, Texas.

Afton, A. D., R. H. Hier, and S. L. Paulus. 1991. Lesser Scaup diets during migration
and winter in the Mississippi flyway. Can. J. Zool. 69:328-333.

Alexander, W. C. and J. D. Hair. 1977. Winter foraging behavior and aggression of diving
ducks in South Carolina. Proc. Southeast. Assoc. Fish Wildl. Agencies 31.226-232.

Custer et til. • LESSER SCAUP FEEDING

565

Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour 49-227-
267.

Bacone, J. a. 1979. Shell Oil dune and swale. Unpubl. Rept., Ind. Dept. Natur. Resour.,
Indianapolis, Indiana.

Bair, E E. 1992. The weather almanac. Sixth ed. Gale Research Inc., Detroit, Michigan.

Bergan, j. F, L. M. Smith, and J. J. Mayer. 1989. Time-activity budgets of diving ducks
wintering in South Carolina. J. Wildl. Manage. 53:769-776.

Bolts, L. 1993. A region of contrasts and dilemmas. Pp. 1-8 in The environment of
Northwest Indiana Contrasts and Dilemmas. Sheffield Press, Hammond, Indiana.

Brock, K. J. 1986. Birds of the Indiana Dunes. Indiana Univ. Press, Bloomington, Indiana.

Cronan, j. M. 1957. Food and feeding habits of the scaups of Connecticut waters Auk
74:459-468.

Custer, C. M. and T. W. Custer. 1996. Food habits of diving ducks in the Great Lakes
after the zebra mussel invasion. J. Field Ornithol. 67:86-99.

’ ' D. W. Sparks, R. K. Hines, and C. O. Kochanny. 1996. Movement

patterns of wintering Lesser Scaup in Grand Calumet River - Indiana Harbor Canal,
Indiana. J. Great Lakes Res. 22:95-99.

Haramis, G. M., E. L. Derleth, and D. G. McAuley. 1987. A quick-catch corral trap for
wintering Canvasbacks. J. Field Ornithol. 58:198-200.

JoRDE, D. G., G. L. Krapu, and R. D. Crawford. 1983. Feeding ecology of Mallards
wintering in Nebraska. J. Wildl. Manage. 47:1044-1053.

AND R. B. Owen, Jr. 1988. The need for nocturnal activity and energy budgets of
waterfowl. Pp. 169-180 in Waterfowl in winter (M. W. Weller, ed.). Univ. Minnesota
Press, Minneapolis, Minnesota.

Kenward, R. 1987. Wildlife radio tagging. Academic Press, London.

Klima, M. 1966. A study of diurnal activity rhythm in the European Pochard, Aythya ferina
(L.), in nature. Zoologicke Listy 15:317-332.

Lovvorn, j. R., D. R. Jones, and R. W. Blake. 1991. Mechanics of underwater locomotion
in diving ducks: drag, buoyancy and acceleration in a size gradient of species J Exp
Biol. 159:89-108.

McNeil, R., R Drapeau, and J. D. Goss-Custard. 1992. The occurrence and adaptive
significance of nocturnal habits in waterfowl. Biol. Rev. 67:381-419.

Neter, j., W. Wasserman, and M. H. Kutner. 1985. Applied linear statistical models, 2nd
ed. R. D. Erwin, Inc., Homewood, Illinois.

Noseworthy, S. M. 1981. Distribution, sex ratios, and behavior of diving ducks wintering
on the Detroit River. M.S. thesis, Univ. Michigan, Ann Arbor, Michigan.

Olsen, G. H., F. J. Dein, G. M. Haramis, and D. G. Jorde. 1992. Implanting radio trans-
mitters in wintering Canvasbacks. J. Wildl. Manage. 56:325-328.

Paulus, S. L. 1984. Activity budgets of nonbreeding Gadwalls in Louisiana. J. Wildl.
Manage. 48:371-380.

Pedroli, J.-C. 1982. Activity and time budget of Tufted Ducks on Swiss lakes during winter.
Wildfowl 33:105-112.

Siegfried, W. R. 1974. Time budget of behavior among Lesser Scaups on Delta Marsh. J.
Wildl. Manage. 38:708-713.

. 1976. Segregation in feeding behaviour of four diving ducks in southern Manitoba.

Can. J. Zool. 54:730-736.

Swanson, G. A., and J. C. Bartonek. 1970. Bias associated with food analysis in gizzards
of Blue-winged Teal. J. Wildl. Manage. 34:739—746.

Takekawa, j. Y. 1987. Energetics of Canvasbacks staging on an Upper Mississippi River
pool during fall migration. Ph.D. diss., Iowa State Univ., Ames, Iowa.

566

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Tamsier, a. 1976. Diurnal activities of Green-winged Teal and Pintails wintering in Lou-
isiana. Wildfowl 27:19—32.

Trivelpiece, W. Z., J. L. Bengtson, S. G. Trivelpiece, and N. J. Volkman. 1986. Eoraging
behavior of Gentoo and Chinstrap penguins as determined by new radiotelemetry tech-
niques. Auk 103:777—781.

U.S. Eish and Wildlife Service-Battelle. 1993. Einal report for the USEPA GLNPO
ARCS Project: Biological assessment of contaminated Great Lakes sediment. Unpubl.
Rept., U.S. Fish and Wildlife Service, Columbia, Missouri.

Zar, j. H. 1984. Biostatistical analysis, 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, New
Jersey.

Wilson Bull., 108(3), 1996, pp. 567-572

BODY MASS AND CARCASS COMPOSITION OF FALL

MIGRANT OLDSQUAWS

James O. Leafloor,' John E. Thompson,' and C. Davison Ankney-

Abstract. We investigated body and organ mass and carcass composition of twenty-
seven migrant Oldsquaws {Clangula hyemalis) killed when they collided with power trans-
mission lines in northeastern Ontario in October 1986. Comparison of the first principal
component (PCI) from eight structural measurements indicated that adult male Oldsquaws
were structurally larger than females; however, organ weights did not differ between sexes
when PCI was included as a covariate (ANCOVA, P > 0.05 in all cases). Carcass com-
position was similar to that reported for spring migrants. Ash-free lean dry weight (AFLDW)
and ash were positively related to structural size, but did not differ between sexes when
PCI was included as a covariate. Lipids comprised 17.5% of whole body mass of females
and 14.1% of males and were sufficient to fuel migration at least to the next likely staging
area in the Great Lakes. Fall migrant Oldsquaws must have stored substantial lipid and
protein reserves after breeding, suggesting that offshore feeding areas in James and Hudson
Bay are critical. Received 9 Oct. 1995, accepted 1 Feb. 1996.

Oldsquaws (Clangula hyemalis) are small-bodied sea ducks that breed
around Hudson Bay and across the Arctic and winter along the Atlantic
and Pacific coasts and on the Great Lakes (Bellrose 1978). Peterson and
Ellarson (1979) reported changes in carcass mass and composition of
Oldsquaws between December and July. Their data were collected from
wintering birds drowned in gill nets on Lake Michigan and from birds
shot on a breeding area in northwest Hudson Bay. They found two peaks
in carcass mass that were primarily associated with increased lipid de-
posits. Peak body mass occurred just before spring migration in May and
in January, a time when Oldsquaws sometimes endure periods of thermal
stress caused by low temperatures (Peterson and Ellarson 1979). However,
carcass composition and body mass data are not available for Oldsquaws
during the postbreeding period and fall migration. It is important to un-
derstand changes in body mass and carcass composition throughout the
annual cycle to identify critical periods for weight gain. The purposes of
this paper are to report body size and carcass composition for a sample
of Oldsquaws obtained during fall migration in northeastern Ontario and
to compare our data to those from Peterson and Ellarson (1979).

METHODS

We obtained 27 adult Old.squaw carca.s.se.s from the Smoky Falls hydroelectric dam located
on the west bank of the Mattagami River in northeastern Ontario (50°04'N, 82°10'W). The

' Ministry of Natural Resources, P. O. Box 190, Moosonee, Ontario, Canada POL lYO. (Present address:
Ministry of Natural Resources, P. O. Box 730. 2 Third Avenue, Cochrane. Ontario POL ICO.)

^ Dept, of Zoology, Univ. of Western Ontario, London, Ontario. Canada N6A 5B7.

567

568

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

birds were killed during an apparent migration movement at around 22:00 h on 26 October
1986 when they hit power transmission lines spanning the river. All birds were frozen and
shipped to the University of Western Ontario for carcass analysis. Whole Oldsquaw car-
casses were thawed and weighed to the nearest 0. 1 g (hereafter referred to as whole body
mass) on a Mettler digital balance. Lollowing Dzubin and Cooch (1992), we measured bill
depth at the base, maximum bill width, culmen 1, tarsus bone, and exposed keel lengths to
the nearest 0.1 mm, using Vernier calipers, and wing chord and total body length to the
nearest mm using a ruler. We also measured wing length, from the body to the distal end
of the outstretched wing, to the nearest mm using a ruler. Next we obtained carcass mass
(whole body minus head, feet, wings, feathers, gastrointestinal tract, and reproductive or-
gans) for comparison to data presented by Peterson and Ellarson (1979). All internal organs
(except lungs and kidneys) and abdominal fat were dissected from the carcasses, patted dry
with paper towels, and weighed to the nearest 0.01 g; contents were removed from gastro-
intestinal organs before weighing.

Our carcass composition analyses were designed to provide data comparable to those of
Peterson and Ellarson (1979), but we also analyzed breast, leg, and liver tissues separately
and included results from a combined homogenate of the head, feet, wings, feathers, gas-
trointestinal tract minus contents, and reproductive organs (hereafter called dry parts ).
The latter procedures provided data for whole birds for future comparisons. Combined values
from breast, leg, liver, and carcass homogenate (but not dry parts) are equivalent to the
“carcass” values from Peterson and Ellarson (1979). Samples were ground twice in a Hobart
meat grinder and oven dried at 90°C to constant weight (Kerr et al. 1982). Dried carcass
homogenate, dry parts homogenate, breast, leg, and liver samples were then homogenized
separately in an electric coffee grinder. Lipid extractions were performed on subsamples (ca
10 g) of each constituent in a modified Soxhlet apparatus using petroleum ether as a solvent
(Dobush et al. 1985). We determined the proportion of lipid in each subsample and multi-
plied this by the dry mass of each constituent sample to determine lipid mass. Lipid mass
was then subtracted from dry weight of each constituent to estimate lean dry mass (LDM;
Ankney and Afton 1988). Lean dry samples were ashed in a muffle furnace for 6 h at 550 C,
and the proportion of ash in each was used to calculate ash for each constituent. Ash weight
was subtracted from LDM to determine ash-free lean dry mass (AFLDM), a measure of
total body protein.

To account for variation in carcass and component masses (lipid, AFLDM, ash) related
to structural size, we used the correlation matrix from bill height, bill width, culmen, tarsus,
wing chord, wing, body, and keel lengths in a principal components analysis (PCA) of all
adults combined. From this we obtained scores for each bird along the first component axis
(PCI) to use as an index of structural size (Alisauskas and Ankney 1987). PCI accounted
for 55% of variation in the characters measured, with a corresponding eigenvalue of 4.44.
Loadings on the first principal component were all positive and ranged from 0.15 for culmen
to 0.44 for wing chord. PCI scores for individual ducks were used as covariates in analyses
of covariance (ANCOVA) comparing lipid, ash, and protein dry weights between sexes.
This technique accounts for variation between sexes that occurs as a result of differences
in body size (Alisauskas and Ankney 1987) and allows comparison of relative amounts of
lipid, ash, and protein. We also compared our data to those from Peterson and Ellarson

(1979).

RESULTS

Male Oldsquaws were larger than females in most external morpho-
logical measurements (Table 1). Masses of esophagus, heart, liver, and

Leafloor et al. • BODY COMPOSITION OF OLDSQUAWS

569

Table 1

Means of Morphological Variables for Adult Oldsquaws Collected in October

1986 in Northeastern Ontario”*’

Variable

Males

Whole body mass

862.8

Carcass mass

633.0

Body length

342.1

Wing length

327.9

Wing chord

223.2

Tarsus

35.6

Culmen

27.1

Bill height

16.4

Bill width

19.7

Keel length

109.2

PCI

1.68

(N - Females (N = 12)

(17.9)A

733.6

(1L6)B

(16.8)A

573.5

(12.7)B

(2.5)A

320.1

(3.5)B

(1.6)A

313.8

(L3)B

(l.l)A

210.2

(l.l)B

(0.3)A

34.7

(0.3)B

(0.3)A

26.5

(0.4)A

(0.3)A

15.9

(0.3)A

(0.2)A

18.2

(0.2)B

(l.O)A

100.6

(0.6)B’

(0.3)A

-1.99 (0.2)B

^ Linear measurements in mm, mass in g, standard error in parentheses.

^ Means followed by the same letter are not different between sexes, /-lest, P > 0.05.

N — 13 males for variables that include the head; two males were decapitated when found.

gizzard differed between sexes (Table 2) but not when PCI was included
as a covariate (ANCOVA, P > 0.05 in all cases). AFLDM and ASH were
positively related to structural size and did not differ between sexes when
PCI was included as a covariate (ANCOVA, P > 0.30), although both
were absolutely larger in males. There was no relationship between total

Table 2

Mean Lengths and Fresh Mass of Internal Organs of Adult Oldsquaws Collected
IN Northeastern Ontario in October 1986”*’

Organ

Males (N = 15)

Females (N = 12)

Small intestine length

133.9

(4. DA

128.6

(4.2)A

Mass

17.0

(0.9)A

15.5

(0.6)A

Large intestine length

6.6

(0.2)A

6.7

(0.2)A

Mass

1.5

(0. 1 )A

1.3

(O.l)A

Caecum length

12.4

(0.6)A

11.1

(0.6)A

Mass

0.7

(O.l)A

0.8

(O.l)A

Esophagus mass

8.1

(0.3)A

6.8

(0.2)B

Gizzard mass

17.0

(2.2)A

10.1

(L2)B

Heart mass

10.5

(0.2)A

6.9

(0.2)B

Pancreas mass

2.4

(0.2)A

1.8

(0.2)A

Spleen mass

0.34

3

b

>

0.34

o

b

>

Liver mass

32.3

(L6)A

25.7

(1.2)B

“ Masses measured in g, lengths in cm, standard error in parentheses.

Means followed by the same letter are not different between sexes, r-tesl (P > 0.05).

570

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Mean Mass of

Table 3

Carcass Components of Adult Oldsquaws Collected in
Northeastern Ontario in October 1986“

Lipid

AFLDM

Ash

Males*’

Breast

2.26 (0.15)

19.97 (0.35)

—

Leg

1.56 (0.13)

5.06 (0.11)

—

Liver

0.89 (0.11)

9.46 (0.55)

—

Homogenate

117.20 (7.64)

105.84 (3.70)

19.87 (0.93)

Dry parts

15.56 (0.57)

29.42 (1.17)

7.71 (0.17)

Total

141.46(8.39)

172.40 (4.51)

28.27 (0.95)

Lemales

Breast

1.94 (0.10)

17.34 (0.29)

—

Leg

1.33 (0.09)

4.46 (0.10)

—

Liver

0.64 (0.09)

7.67 (0.36)

—

Homogenate

124.60 (6.78)

80.08 (2.82)

15.99 (0.45)

Dry parts

13.19 (0.68)

23.99 (0.51)

6.48 (0.17)

Total

141.70 (7.00)

133.56 (2.92)

22.48 (0.47)

“ Mass in grams (g), standard error in parentheses.

A? = 13 males for “dry parts" and “total” values because 2 males were decapitated when found.

lipid mass and body size (ANCOVA, P > 0.75). Furthermore, total lipid
mass was almost identical between male and female Oldsquaws (Table
3) despite the smaller structural size of females. Lipids comprised 17.5%
of whole body weights of females and 14.1% of males.

DISCUSSION

Oldsquaws nest in tundra habitats of southwestern Hudson Bay, and a
migration corridor between James Bay and the Great Lakes has been
postulated (Bellrose 1978). We found little published information on ex-
ternal morphological measurements and gut morphology of Oldsquaws,
but we assume that our sample was unbiased and therefore representative
of Oldsquaws that migrate through northeastern Ontario. Gut measure-
ments were shorter than those reported by Goudie and Ryan (1991: Table
2) for a combined sample of male and female Oldsquaws collected in
coastal Newfoundland during winter. Oldsquaws in Newfoundland fed
mostly upon amphipods and isopods, but the diet of Oldsquaws staging
in James Bay is unknown.

Mean carcass mass of fall migrant female Oldsquaws was similar to
that of spring migrants, and about 70—200 g greater than that during mid-
summer and winter (Peterson and Ellarson 1979). Most seasonal variation
in carcass mass of adult females was attributable to fluctuations in lipid

Leafloor et al. • BODY COMPOSITION OF OLDSQUAWS

571

levels, protein reserves (as indexed by LDM) were about 25 g heavier
during fall migration than in mid-summer but about 15 g less than in
spring. Carcass mass of males was similar in spring and fall; lipid levels
also were similar in spring and fall. Inexplicably, male Oldsquaws had
about 25 g more protein reserves in fall than in spring (Peterson and
Ellarson 1979).

Peterson and Ellarson (1979: Table 1) reported that adult female Old-
squaws lost about 95% of lipid reserves and 30% of LOW between spring
migration and the start of brood rearing in late July. Our data indicated
that fall migrant female Oldsquaws had lipid levels similar to those of
spring migrants, but protein reserves were 11% smaller. This suggests
that (1) substantial increases in lipid and protein stores occur between
late July and October and that (2) female Oldsquaws increase protein
reserves in spring, perhaps for egg production.

Carcass lipids of adult female Oldsquaws on Lake Michigan in Decem-
ber were about 44 g less than, and protein reserves were about equal to
those of fall migrants, suggesting that lipids provided energy during mi-
gration. Peterson and Ellarson (1979:294) estimated that a migrating
Oldsquaw metabolized about 25.76 kcal/h. Assuming stored fat yields
9.45 kcal/g and conversion efficiency is 100%, 44 g of fat yields enough
energy to fly for about 16 h. If Oldsquaws fly at an average speed of
about 80 km/h (Bellrose 1978: 385), an average female in our sample
could travel about 1280 km on 44 g of fat. Carcass lipids of adult males
followed the same pattern as females, averaging 47 g less in winter than
in fall. In addition, they had about 25 g more lean dry mass in fall than
birds sampled in December. The nearest wintering areas for Oldsquaws
from lower James Bay are the lower Great Lakes, i.e., southern Lake
Michigan, Lake Erie, and Lake Ontario (Bellrose 1978). As none of these
locations is >1280 km from lower James Bay, Oldsquaws had more than
sufficient reserves to reach those destinations. Our data suggest that off-
shore habitats in James Bay and Hudson Bay are important to postbreed-
ing Oldsquaws for replenishment and storage of protein and lipid reserves
before fall migration.

ACKNOWLEDGMENTS

This work was supported by the Ontario Ministry of Natural Resources and the Univ. of
Western Ontario. Oldsquaws were collected under permit from the Canadian Wildlife Ser-
vice.

LITERATURE CITED

Alisauskas, R. T. and C. D. Ankney. 1987. Age-related variation in the nutrient reserves
of breeding American Coots (Fulica americana). Can. J. Zool. 65:2417—2420.

572

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Ankney, C. D. and a. D. Afton. 1988. Bioenergetics of breeding Northern Shovelers:
diet, nutrient reserves, clutch size, and incubation. Condor 90:459-472.

Bellrose, E C. 1978. Ducks, geese and swans of North America. Stackpole Books, Har-
risburg, Pennsylvania.

Dobush, G. R., C. D. Ankney, and D. G. Krementz. 1985. The effect of apparatus, ex-
traction time, and solvent type on lipid extractions of snow geese. Can. J. Zool. 63:
1917-1920.

Dzubin, a. and E. G. Cooch. 1992. Measurements of geese: general field methods. Cali-
fornia Waterfowl Association, Sacramento, California.

Goudie, R. I. AND P. C. Ryan. 1991. Diets and morphology of digestive organs of five
species of sea ducks wintering in Newfoundland. J. Yamashina Inst. Ornithol. 22.1—8.

Kerr, D. C., C. D. Ankney, and J. S. Millar. 1982. The effect of drying temperature on
extraction of petroleum ether soluble fats of small birds and mammals. Can. J. Zool.
60:470-472.

Peterson, S. R. and R. S. Ellarson. 1979. Changes in Oldsquaw carcass weight. Wilson
Bull. 91:288-300.

Wilson Bull., 108(3), 1996, pp. 573-583

AVIAN NEST-SITE SELECTION AND NESTING
SUCCESS IN TWO FLORIDA CITRUS GROVES

Mary Crowe Mitchell,'-^ Louis B. Best,' and James P. Gionfriddo' '^

Abstract. — We studied nesting success and nest-site selection of Common Ground-
Doves (Columbina passerina). Northern Mockingbirds (Mimus polyglottos). Brown Thrash-
ers (To.xostoma rufum), and Northern Cardinals (Cardinalis cardinalis) in two Florida citrus
groves in spring 1989. Predation resulted in the loss of more than half of all nests. Fish
Crows (Corx’us ossifragus) and rat snakes (Elaphe obsoleta) seemed to be the major pred-
ators. Nesting success differed between groves and may have resulted from differences in
human activities, predator populations, or vegetation structure. Nesting success of Northern
Cardinals and Brown Thrashers was lower than that reported in other studies and may have
been below the replacement level. Northern Mockingbirds had the most open nest sites with
the largest inter-canopy distances (spacing between tree canopies), whereas Brown Thrashers
seemed to select areas of the groves with the greatest canopy closure. Northern Cardinals
tended to select nest trees with full canopies, perhaps to increase nest concealment. Common
Ground-Dove nests were supported by limbs with small angles (degrees deviation from
horizontal) and the largest diameters. Received 21 June, accepted 25 Jan. 1996.

Citrus groves represent a substantial proportion of breeding habitat
available to birds in Florida, yet nest-site selection and nesting success
have not been studied in these groves. As Florida habitat is converted to
agricultural and other domestic uses, birds are forced to nest in altered
habitats for which they may be poorly adapted (e.g., Dow 1969a, Best
and Rodenhouse 1984). Our objectives were to document nesting success
and characterize nest-site selection in Florida citrus groves. We attempted
to answer the following questions; What preferences do breeding birds
show in selecting their nest sites? Is nesting success affected by nest-site
selection and, if so, how? Are citrus groves suitable nesting habitat for
songbirds?

STUDY AREAS AND METHODS

We used two citrus groves on Merritt Island in Brevard County, Florida for study sites.
Study grove 1, about 71 ha, was privately owned and managed and was almost entirely
planted with orange trees. Study grove 2 was part of the Merritt Island National Wildlife
Refuge, was 45 ha, and had a mixture of orange and grapefruit trees. The major herbaceous
vegetation in the citrus groves was guinea grass (Panicum ma.ximum), poorman’s pepper
(Lepidium virginicum), day-flower (Commelina dijfusa), Richardia (Richardia spp.), prickly
sida (Sida spinosa), Bermudagrass (Cynodon dactylon), vaseygrass (Paspalliim urvellei),
and amaranth (Amaranthus spp.). We studied nests from mid-March through early June in
1989. Nests were found by systematically examining each tree in the groves four times

' Dept, of Animal Ecology, Iowa State Univ., Ames, Iowa 5001 1.

^ Present address: 540 Lower River Rd., Heron, Montana 59844,

^ Present address: Dept, of Forestry and Natural Resources, Purdue Univ., West Lafayette, Indiana 47907.

573

574

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

during the study and by observing bird behavior such as nest building and food carrying.
The location of each nest tree was marked on a map of the grove, and a tree adjacent to
the nest tree was flagged with colored tape. Nests were monitored on alternate days until
they were no longer active. The number and condition of the eggs or young were recorded
during each nest visit. Inaccessible nests were checked by using a pole-mounted mirror, by
climbing the nest tree, or by using a stepladder in the bed of a pickup truck. As part of a
concurrent study, nestlings were weighed and measured during each visit until there was a
risk of inducing premature fledging. Broods of Northern Cardinals (Cardinalis cardinalis)
and Brown Thrashers {Toxostomci rufum) also were ligatured during the nestling period to
collect food samples (see Johnson et al. 1980). To avoid attracting predators to the nest site,
the young were processed at least 10 m from the nest.

Apparent nest success was determined for species with a combined total for both groves
of five or more nests with known outcomes. A nest was considered successful if at least
one nestling fledged. Nests believed to be deserted due to our monitoring activities were
excluded from analyses. Nesting success also was determined by using the number of days
of nest exposure (Mayfield 1975). Because the nesting cycles of species breeding in the
groves differed in length and, hence, the number of exposure days, nesting success was
calculated separately for each species with an adequate sample size. The computer program
MICROMORT (Heisey and Luller 1985) was used to calculate survival rates for the egg
and nestling stages and for the entire nesting cycle.

We used chi Square contingency analysis (2 X 3) to test for differences in nesting out-
comes between groves. Nests were classified as successful, failed due to predation, or failed
due to other causes. Tests were made for all species combined (Common Ground-Doves
[Columhina passerina]. Brown Thrashers, Northern Mockingbirds [Mimus polyglottos], and
Northern Cardinals) and for each species separately, except for the Northern Mockingbird
where the sample size was too small for individual analysis. Red-winged Blackbirds {Age-
laius phoeniceus) were excluded from both analyses because their nests were found only in
Grove 2 in localized areas associated with drainage canals.

After a nest became inactive, we recorded variables characterizing the nest vicinity, nest
substrate, and nest position within the substrate. Inter-canopy (between canopy perimeters)
and inter-tree (between trunks) distances within and between tree rows were determined.
The number of young trees or open spaces where a tree was missing in an area around the
nest tree, which included the eight nearest trees, was used as a measure of the openness of
the nest tree vicinity. In addition, herbaceous ground cover was sampled within a 1-m^
quadrat placed 5 m from the trunk of the nest tree in each of the four cardinal directions.
Within each quadrat, maximum herbaceous cover height was measured with a tape, and the
percent coverages of herbaceous vegetation, bare ground, and plant litter were visually
estimated. Citrus type (orange, grapefruit, or hybrid root stock), nest tree height, canopy
diameter, and the openness of the nest tree canopy (a visual estimate of the percent closure
of the entire canopy) were used to characterize the nest tree. The position of the nest within
the substrate was characterized by nest height, relative nest height (the height of the nest
divided by the height of the nest tree), the number of limbs supporting the nest, the angles
(degrees deviation from horizontal) and diameters of the six largest supporting limbs, and
the percent foliage density of the nest tree, estimated visually above and below the nest in
the area immediately around the nest. Nest-site measurements also were recorded for nests
abandoned before discovery if the species could be identified.

Means and variances were calculated for the variables characterizing the groves and the
nest sites of Common Ground-Doves, Brown Thrashers, Northern Mockingbirds', and North-
ern Cardinals. A series of Student’s /-tests was u.sed to test for differences in variables
between the two groves, between the nest sites and the groves in general, and among the

Mitchell et al. • NEST-SITE SELECTION

575

nest sites of the four species. Because sections within the groves were managed differently,
tree age and height, canopy diameter, inter-canopy and inter-tree distances, and the amount
of herbaceous growth varied. Groves were thus blocked by management units, and vege-
tation was randomly sampled within each unit. For the analyses, 25 samples were randomly
selected from each plot; the distribution of the samples among the management units was
proportional to their area. Student’s r-tests also were used to test for differences in nest-site
characteristics between successful and failed nests of Common Ground-Doves, Northern
Cardinals, and the combined nests of Common Ground-Doves, Brown Thrashers, Northern
Mockingbirds, and Northern Cardinals in Grove 1. (Sample sizes for Brown Thrashers and
Northern Mockingbirds were too small for separate r-tests.) Similarly, successful and failed
nests of Common Ground-Doves, Brown Thrashers, Northern Cardinals, and the combined
nests of Common Ground-Doves, Brown Thrashers, Northern Mockingbirds, and Northern
Cardinals were compared in Grove 2. All significant relationships are presented in the dis-
cussion of the selection of nest-site variables. We tested for correlations between variables
with Spearman’s rank correlation, and found citrus type, inter-tree distance, and the number
of limbs supporting the nest to be highly correlated with other variables. Thus, we eliminated
them from further consideration. Statistical significance was set at P < 0.05 for all tests
unless otherwise stated.

RESULTS AND DISCUSSION

Fifty-four nests representing five species were found in Grove 1, and
65 nests of seven species were discovered in Grove 2. Of these, the
outcome was determined for 46 nests in Grove 1 and 39 nests in Grove
2 (Table 1). The most abundant nests were those of the Northern Cardinal,
Brown Thrasher, and Common Ground-Dove, three of the most common
species found in the Merritt Island citrus groves (Mitchell et al. 1995).

There are numerous potential predators in and around citrus groves,
but Fish Crows (Corvus ossifragus) were probably responsible for most
of the predation. They were seen near nests that had recently been dep-
redated and were observed carrying nestling birds out of the groves on
several occasions. Although otherwise intact, some depredated nests had
their linings pulled up, which also lead us to suspect that crows lifted
young out of nests. Snakes also were thought to be a significant source
of nest loss. A yellow rat snake was seen at the base of a nest tree before
our discovery that the nest had been recently depredated, and a yellow
rat snake was found in another nest consuming nestlings.

Predation caused the loss of more than two-thirds of all nests in Grove
1 (Table 1). All other causes for nest failure accounted for 11% of the
nests. Of the nests found in Grove 1, only 17% were successful. A greater
percentage of the known nests was successful in Grove 2 (33%, Table 1).
Predation also was responsible for most nest failures in Grove 2, but
desertion and other causes accounted for nearly one-fifth of the nesting
failures. Red- winged Blackbirds and Brown Thrashers suffered the great-
est losses from predation in this grove. Fish Crows were thought to be
responsible for four of five instances of predation on Red-winged Black-

576

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Table 1

Nesting Outcomes (Number of Nests) of the Five Most Common Bird Species Nesting

IN Florida Citrus Groves in Spring 1989

Species

Total nests

Successful

fledging

Predation

Desertion

Other causes
of nest
failure”

Gr I

Gr 2

Gr 1

Gr 2

Gr 1

Gr 2

Gr 1

Gr 2

Gr I

Gr 2

Common Ground-Dove

11

6

4

4

6

1

1

0

0

1

Brown Thrasher

13

15

0

4

11

9

1

2

1

0

Northern Mockingbird

4

2

0

1

3

0

1

0

0

1

Red-winged Blackbird

0

5

0

0

0

5

0

0

0

0

Northern Cardinal

18

11

4

4

13

4

0

0

1

3

All nests combined

46

39

8

13

33

19

3

2

2

5

“ Includes deaths from pesticide exposure, starvation, sickness, injury, egg breakage, physical disturbance of the nest by
heavy equipment, and unknown causes.

bird nests. The crows were seen either at or near the nest sites before we
discovered the nest failures. Two of the thrasher nests were depredated
when the citrus fruit was being picked. Because Fish Crows were sighted
more often during or immediately after picking activity, we suspect that
they caused the thrasher nest losses. We found that thrashers often gave
distress calls when we were in the vicinity of their nests. The presence
of fruit harvesters likely would have elicited distress calls from thrashers,
facilitating the ability of crows to find nests.

The frequency of occurrence of successful and unsuccessful nesting
outcomes in the two groves did not differ significantly for Common
Ground-Doves (x^ = 2.4, df = 2), but it did differ significantly for Brown
Thrashers (x^ ~ 4.1), Northern Cardinals (x^ “ 4.4), and for all species
combined (x^ = 7.6).

Daily nest survival rates were similar for all species in the egg stage,
but varied widely in the nestling stage (Table 2). Brown Thrashers in
Grove 1 had the lowest daily nest survival rate for nestlings. Interval
survival rates were higher during the egg stage than the nestling stage for
Brown Thrashers and Northern Cardinals, but not for Common Ground-
Doves. Ground-dove nestlings had a much smaller chance of being de-
stroyed than did the eggs. Nest survival rates spanning both the egg and
nestling intervals were greatest for Common Ground-Doves in both
groves, followed by Northern Cardinals in Grove 2. Brown Thrashers had
the lowest rate of survival, particularly in Grove 1. Survival spanning the
entire nesting cycle was higher in Grove 2 than in Grove 1.

Our Mayfield estimates of nesting success for Brown Thrashers and
Northern Cardinals in the citrus groves were lower than that reported from

Mitchell et til. • NEST-SITE SELECTION

577

Table 2

Reproductive Success of Common Ground-Doves, Brown Thrashers, and Northern
Cardinals in Florida Citrus Groves in Spring 1989

Species

Grove

Exposure days

Daily nest
survival rate

Interval nest
survival rate

Nest survival
rate across
• egg and
nestling
stages

Egg

Stage

Nestling

stage

Egg

Stage

Nestling

stage

Egg

Stage

Nestling

stage

Common Ground-Dove

1

118

104

0.915

0.981

0.32

0.81

0.25

2

57

44

0.947

0.977

0.49

0.78

0.38

Brown Thrasher

1

210

87

0.919

0.736

0.26

0.03

<0.01

2

281

151

0.947

0.894

0.42

0.29

0.12

Northern Cardinal

1

380

132

0.953

0.864

0.48

0.27

0.13

2

101

97

0.960

0.887

0.54

0.34

- 0.18

Other studies. Mayfield nesting success rates of 44% for Brown Thrashers
(Murphy and Fleischer 1986) and 51% for Northern Cardinals (Booth
1980) have been reported. Information on Common Ground-Dove nesting
success is scant, but all the young in seven nests located and monitored
in a pine plantation survived to fledging (Landers and Buckner 1979).

The high failure rates of Brown Thrasher nests in both groves and of
Northern Cardinal nests in Grove 1 attributed to predation may have re-
sulted from our nest monitoring activities. Both species became vocal
when field technicians were near the nest sites. Corvids have learned to
associate human activity and the response of some nesting passerine spe-
cies with the presence of active nests (Gottfried and Thompson 1978,
Best, pers. obs.) and may have discovered more nests because of our
presence. Prior experience with citrus fruit pickers that disturb nesting
birds also may have conditioned the crows. Salathe (1987) found that
crows that successfully depredated European Coot {Fulica atra) nests
would extend their searching around the depredated nests, sometimes re-
sulting in destruction of all nests in the area. He concluded that distur-
bance created by investigator nest monitoring activities affected crow be-
havior by revealing nests and providing positive reinforcement. When
Common Ground-Doves were flushed from the nest, they did not vocalize
but sometimes gave a broken wing display. Those doves that did not
display were probably inconspicuous to predators. Those that feigned in-
jury sometimes continued the behavior as far as several tree rows away
from the nest, perhaps luring predators from the nest site. Although Com-
mon-Ground Dove eggs are white, the dense citrus tree canopies probably
shielded unattended eggs from view from outside the tree canopy. Be-
cause Common Ground-Dove nests were small and often placed on thick

578

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

branches, they were more cryptic than the larger nests of Northern
Cardinals and Brown Thrashers. These differences may have accounted
for the greater nesting success of Common Ground-Doves.

Differences in nesting success between the groves may have resulted
from differences in predator populations, human activities, or vegetation
structure. Grove 1, where nests suffered higher predation rates, was in a
residential area, whereas Grove 2 was within the Merritt Island National
Wildlife Refuge where human disturbance may have been less. The veg-
etation also differed substantially between the two groves and may have
influenced nest predation.

Citrus culture operations were largely responsible for differences in
grove vegetation. Tree hedging, topping, and skirt pruning influenced the
geometry of the tree canopies and the inter-canopy distance, whereas
mowing and herbicide application controlled the amount of herbaceous
cover. Because the two citrus groves were managed differently, intercan-
opy distance was greater {t = —2.0, P = 0.05) in Grove 2 [249.1 ± 223.1
cm (x ± SD)] than in Grove 1 (143.4 ± 130.0 cm). Coverage of herba-
ceous vegetation also was greater (/ = -2.8, P = 0.008) in Grove 2 (44
± 27%) than in Grove 1 (21 ± 30%), but the opposite was true for bare
ground coverage [(r = 3.6, P = 0.001) Grove 2: 10 ± 18% Grove 1: 40
± 37%.] Less vegetative cover in Grove 1 may have resulted in decreased
nest concealment. Although some investigators have found no correlation
between nesting cover and nesting success (e.g., Reynolds 1981, Conner
et al. 1986), Ehrhart and Conner (1986) reported a correlation between
vegetative cover around the nest and nesting success, and Martin and
Roper (1988) found predation to be greater around less well-concealed
nests.

In addition to altering herbaceous and tree-canopy cover, citrus culture
operations may have affected breeding birds by creating disturbances
which may have increased nest desertion, particularly during nest build-
ing. We suspect this because at least two nests were deserted during con-
struction because of our nest monitoring activities. Also, pesticides, her-
bicides, and fungicides were routinely applied in the groves and had the
potential of poisoning adults and nestlings, resulting in decreased survival
and nesting success.

The low nesting success of the breeding birds in the citrus groves
suggests that their reproductive output could have been below the replace-
ment level. Such “sink” populations have been documented in other ag-
ricultural environments (Rodenhouse and Best 1983, Frawley 1989, Bry-
an 1990). Low reproductive success per breeding attempt may be com-
pensated for by the long breeding season in Florida. Common Ground-
Doves are thought to breed year-round in Florida (Baynard 1909 in

Mitchell et ul. • NEST-SITE SELECTION

579

Howell 1932, Landers and Buckner 1979). Northern Mockingbirds and
Northern Cardinals nest from March through August (Woolfenden and
Rohwer 1969), and Brown Thrashers nest from March through July.

In Grove 1, litter coverage was significantly greater around the nest
trees of all species [dove: 55 ± 26% (x ± SD), thrasher: 57 ± 26%,
mockingbird: 46 ± 24%, cardinal: 60 ± 22%] than in the grove in general
(21 ± 25), whereas the coverage of bare ground was significantly less
(dove: 23 ± 27%, thrasher: 19 ± 25%, mockingbird: 19 ± 25%, cardinal:
16 ± 19%, grove: 40 ± 37%). Litter and bare ground coverages around
nest vicinities in Grove 2 did not differ significantly from the grove over-
all, but the coverage of herbaceous vegetation around Northern Mocking-
bird nests (71 ± 19%) was significantly greater than the representative
samples of the grove (44 ± 27%). When all species were compared.
Northern Mockingbird nest vicinities had significantly more herbaceous
vegetation coverage (Table 3). Because Common Ground-Doves, Brown
Thrashers, Northern Mockingbirds, and Northern Cardinals commonly
forage on the ground (De Graaf et al. 1985), the coverages of herbaceous
vegetation, litter, and bare ground may have been important in their se-
lection of a nest vicinity.

Inter-canopy distance was significantly greater around Northern Mock-
ingbird nest trees in Grove 1 (287 ±135 cm) than in the grove in general
(143 ± 130 cm) and was greater around the nest trees of Northern Mock-
ingbirds than around nest trees of the other three species (Table 3). Like-
wise, the number of young trees or open spaces where a tree was missing
near the nest tree, a measure of the openness of the nest vicinity, was
greater around Northern Mockingbird nest sites than around nest sites of
the other species (Table 3). Woolfenden and Rohwer (1969) described
ideal Northern Mockingbird nesting habitat as areas of “spaced” trees
and found that nests were usually located in the more sparsely wooded
or open sections of their plots. Brown Thrasher nest sites in Grove 2 had
significantly smaller inter-canopy distances (144 ± 104 cm) than did a
representative sample of the grove (250 ± 223 cm), suggesting that
thrashers chose sections of the grove with more closed tree canopies.
Inter-canopy distances for Brown Thrasher nest sites were similar in both
groves, and differed from those of both mockingbirds and cardinals (Table
3). Fischer (1980) found that Long-billed Thrasher {Toxostoma longiros-
tre) nests often were placed in thickets with dense leaf canopies that
provided excellent concealment.

Selection of the nest vicinity also may have been influenced by the
grove edges because edge habitats may have been important foraging
areas. Fichter (1959) concluded that the breeding density of Mourning

Table 3

Habitat Variables (x ± SD) Characterizing Nest Sites'

580

THE WILSON BLFLLETIN • Vol. 108, No. 3, September 1996

c-S

o a
ZU

s-s
£ S
“iS

c o

19

El

O 3

U O

o

u u

CQ m
< <
00 IT)

u

u

u

CJ

OQ

<

CQ

<

<

oa

<

<

00

u

<

CQ

<

ON

oa

<

U

X

so

oa

<

t^

CQ

<

f"-.

U

<

<

o

o

(N

cn

NO

(n

in

00

(N

(N

—

(N

—

(n

—

d

d

00

(N

-1-1

+ 1

+ 1

+1

+ 1

+1

+ 1

-M

+1

+1

■M

-l-l

ON

m

r-

NO

r-

00

m

ro

—

00

CN

(n

o

ON

•Nt

m

m

(N

'

in

<N

TO

u

u

u

CQ

U

QJ

u

ON

o

u

oa

u

£ 'B

<

<

u

S

CN

ro

<

<

U

<L> bO
r— c

m

O

o

00

00

(N

(N

m

(N

(N

(N

ON

CN

Z o

-FI

+ 1

+ 1

-M

+ 1

+ 1

+ 1

+ 1

-1-1

-M

s

00

m

in

o

(On

G

ON

cn

m

so

m

NO

O

G

ON

m

SO

U

oa

u

oa

<

00

U U

pa oa

< <

vO 00

odd
+1 +1 +1 +1 +1 +1

o\ o

■ (N IT)

(N

CN

m oa CQ oa

< < < <

O m CN L' _

CJ ro (N — (N

^ <c

8o

CQ

<

o

o

CQ

< oa

— ' in
(N ON

+ 1 +1 +1 +1 +1 +1

NO O cN m iG
' ro

+ 1 +1 +1

CQ

<

o

+1

CQ CQ

oa < <

00 NO NO

d
+1

o

+1

o

+1

oa oa
00 —
+1

+1

lo so

— VsO
rj-

lo 00 CO in —
ON SO
(N

^ (N -•

<

00

<

so

<

00

<

o

<

m

<

r<5

<

<

00

<

<

o

<

ON

<

so

<

in

<

m

<

CN

(N

CN

CN

00

(N

—

CN

00

cn

cn

d

d

00

-hi

+ 1

-h!

+ 1

-hi

+ 1

+ 1

■hi

-h!

-hi

-hi

-hi

■hi

■hi

■hi

00

O

CN

ON

O

00

00

in

CN

r-

ON

<N

(N

cn

r-

— ■

in

in

CN

CN

rt-

<N

—

o

CN

u

B

e

u

x;

^ OJO

s 1

c c

.2 .2

■5 oJ

W) CJ)
<U (U
> >

c/5 c/5

D 3
O O
<U fU
U 0)

u u
Cd <U (U

d: X

c

3

o

u

00

<U 3
O —

C c/5

c3 (U
c/5 ““

D.

^ 5

o cO
w O

It? ^

(U

c

(U

E I

o 03

a>

c

o

V

X)

u

cd

<u

o d
E

E

O

w 0

c ^

u.

o

o

e

'55

>% C

W o
*C/5

c

S

0

X

O (U

S 60

c

o

o

c/5

c/5

CiX)

cd

"O

■O M

C/5

<u

fU

(U

u

c/5

o

‘S

<U C
y 03

o

dj y

c

c

CL C
O c

X

Kl

'H.

X

OJD

cd

60

Cd ^

o

c

V

OJ

(15

^ e

E

X

'X

X

OD

CL ^

o ^

Z

U O F

c/3

1)

z

■9 o

J Um

t2 Z oC

- For each habitat variable, means not sharing the same letter differed significantly (P s 0.05) when compared with a series of Student s r-tests.
'’Foliage density estimates: 1 = cover <25%, 2 = 25-75%, 3 = >75%.

Mitchell et al. • NEST-SITE SELECTION

581

Doves {Zenaida macroura) in Idaho apple orchards was not affected as
much by nesting cover as it was by the adjacent habitat type.

Canopy diameter, which was negatively correlated (Spearman’s rho =
—0.595, P < 0.001, df = 106) with inter-canopy distance, was smallest
for Northern Mockingbird nest trees (Table 3). Also, Northern Mocking-
bird nest tree canopy diameters in Grove 2 (228 ± 65 cm) were signifi-
cantly smaller than canopy diameters in the grove overall (528 ± 156
cm).

The openness of the nest-tree canopy was smallest for Northern
Cardinals and differed significantly from Brown Thrashers (Table 3).
Brown Thrashers chose the tallest trees for nest placement, whereas
Northern Mockingbirds tended to place their nests in the shortest trees.
When all species were compared. Brown Thrasher nest tree heights dif-
fered significantly from those of Common Ground-Doves and Northern
Mockingbirds (Table 3).

Angles of limbs supporting mockingbird nests were significantly larger
than those of the other species (Table 3). Diameter of limbs supporting
nests was similar for all species, except for ground-doves, which had nests
supported by larger limbs.

Relative nest height was greatest for Brown Thrashers and smallest for
Common Ground-Doves (Table 3). When all species were compared,
these two were significantly different from each other.

When successful and unsuccessful nests were compared for each spe-
cies, only six of the nest site variables seemed to be related to nesting
success. The nest vicinity and the placement of the nest in the tree were
important, but the nest tree variables did not seem to be. In Grove 1,
openness near the nest tree was greater for failed Northern Cardinal nests
[1.7 ± 0.9 (x ± SD)] than for successful ones (0.5 ± 0.6). Nest con-
cealment is believed to have a large influence on Northern Cardinal suc-
cess (Ehrhart and Conner 1986), and an open nest vicinity may have
facilitated Fish Crows in detecting activity around the nest site. In Grove
2, the height of herbaceous vegetation in the vicinity of Northern Cardinal
nests was significantly less for successful nests (33 ± 10 cm) than for
unsuccessful nests (55 ± 10 cm), but we have no explanation for this
finding. Successful Common Ground-Dove nests in Grove 2 were placed
in trees with significantly larger inter-canopy distances (267 ± 28 cm)
than were unsuccessful nests (141 ±51 cm), but we cannot explain this
pattern.

Nest placement seemed to affect only Common Ground-Dove nesting
success. Successful Common Ground-Dove nests in Grove 2 had sup-
porting limbs with significantly smaller angles (10 ± 17°) than did un-
successful nests (50 ± 17°). Because Common Ground-Doves build frail

582

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

nests with shallow depressions (Howell 1932), they may have chosen
smaller- angled limbs for added nest stability. Mourning Doves preferen-
tially place their nests on flat, horizontal limbs (Harris et al. 1963, Knight
et al. 1984). Successful Common Ground-Dove nests in Grove 2 also
were significantly closer to the ground (180 ± 12 cm) than unsuccessful
ones (260 ± 34 cm).

Citrus groves seemed to be suitable breeding habitat for songbirds and
doves, based on the number of active nests. Birds seemed to be making
choices about the openness of the nest vicinity, the diameter and openness
of the tree canopy, tree height, limb angle and diameter, and nest height.
These choices may have been based on nest concealment and nest sup-
port, but did not necessarily influence nesting outcome. For example, the
selection of nest trees with closed canopies did not seem to affect nesting
outcome of Brown Thrashers, whereas the choice of small-angled limbs
may have increased nesting success for Common Ground-Doves. Because
citrus groves are unnatural environments subjected to periodic human
disturbances which may inflate predation levels, the choices of some nest
site characteristics that are adaptive in natural habitats may be neutral or
even maladaptive in citrus groves.

ACKNOWLEDGMENTS

We are indebted to Margaret Dexter, Bret Giesler, and Patrick Carroll for their assistance
in the field work. Brooks Humphreys of NASA provided access to the citrus groves on
Merritt Island and taught us about citrus culture. This manuscript benefitted from reviews
by J. J. Dinsmore, L E. Lohrer, and T. E. O’Meara. Eunding for this work was provided by
Mobay Corporation. This is Journal Paper No. J- 16470 of the Iowa Agriculture and Home
Economics Experiment Station, Ames, Iowa. Project No. 3300.

LITERATURE CITED

Best, L. B. and N. L. Rodenhouse. 1984. Territory preference of Vesper Sparrows in
cropland. Wilson Bull. 96:72-82.

Booth, L. M. 1980. Reproductive success of the cardinal in relation to territory size and
arthropod availability in three eastern Texas forests. M.S. thesis, Stephen E Austin State
Univ., Nacogdoches, Texas.

Bryan, G. G. 1990. Species abundance patterns and productivity of birds using grassed
waterways in Iowa rowcrop fields. M.S. thesis, Iowa State Univ., Ames, Iowa.
Conner, R. N., M. E. Anderson, and J. G. Dickson. 1986. Relationships among territory
size, habitat, .song, and nesting success of Northern Cardinals. Auk 103:23-31.

De Graaf, R. M., N. G. Tilghman, S. A. Anderson. 1985. Foraging guilds of North
American birds. Environ. Manage. 9:493-536.

Dow, D. D. 1969. Home range and habitat of the cardinal in peripheral and central popu-
lations. Can. J. Zool. 47:103—114.

Ehrhart, R. L. and R. N. Conner. 1986. Habitat selection by the Northern Cardinal in
three eastern Texas forest stands. Southwest. Nat. 31:191-199.

Mitchell et al. • NEST-SITE SELECTION

583

Lighter, E. 1959. Mourning Dove production in four Idaho orchards and some possible
implications. J. Wildl. Manage. 23:438-447.

Eischer, D. H. 1980. Breeding biology of Curve-billed Thrashers and Long-billed Thrash-
ers in southern Texas. Condor 82:392-397.

Frawlev, B. F. 1989. The dynamics of nongame bird breeding ecology in Iowa alfalfa
fields. M.S. thesis, Iowa State Univ., Ames, Iowa.

Gottfried, B. M. and C. E Thompson. 1978. Experimental analysis of nest predation in
an old-field habitat. Auk 95:304-312.

Harris, S. W., M. A. Morse, and W. H. Longley. 1963. Nesting and production of the
Mourning Dove in Minnesota. Amer. Midi. Nat. 69:150-172.

Heisey, D. M. and T. K. Fuller. 1985. Evaluation of survival and cause-specific mortality
rates using telemetry data. J. Wildl. Manage. 49:668-674.

Howell, A. H. 1932. Florida bird life. Coward-McCann, N.Y. 579 pp.

Johnson, E. J., L. B. Best, and P. A. Heagy. 1980. Food sampling biases associated with
the “ligature method.” Condor 82:186-192.

Knight, R. L., D. G. Smith, D. M. Gaudet, and A. W. Erickson. 1984. Nesting ecology
of Mourning Doves in fruit orchards in north-central Washington. Northwest Sci 58-
230-236.

Landers, L. J. and J. L. Buckner. 1979. Ground Dove use of young pine plantations.
Wilson Bull. 91:467-468.

Martin, T. E. and J. J. Roper. 1988. Nest predation and nest-site selection of a western
population of the Hermit Thrush. Condor 90:51-57.

Mayfield, H. J. 1975. Suggestions for calculating nest success. Wilson Bull. 87:456-466.

Mitchell, M. C., L. B. Best, and D. L. Fischer. 1995. Bird abundance in Florida citrus
groves. Fla. Field Nat. 23.T-9.

Murphy, M. T. and R. C. Fleischer. 1986. Body size, nest predation, and reproductive
patterns in Brown Thrashers and other mimids. Condor 88:446-455.

Reynolds, T. D. 1981. Nesting of the Sage Thrasher, Sage Sparrow, and Brewer’s Sparrow
in southeastern Idaho. Condor 83:61-64.

Rodenhouse, N. L. and L. B. Best. 1983. Breeding ecology of Vesper Sparrows in corn
and soybean fields. Amer. Midi. Nat. 110:265-275.

Salathe, T. 1987. Crow predation on coot eggs: effects of investigator disturbance, nest
cover, and predator learning. Ardea 75:221-229.

WOOLFENDEN, G. E. AND S. A. Rohwer. 1969. Breeding birds in a Florida suburb. Bull.
Florida State Mus. 13:1-83.

Wilson Bull., 108(3), 1996, pp. 584—588

SHORT COMMUNICATIONS

Exponential population growth of Monk Parakeets in the United States. In the Unit-
ed States, at least nine species of introduced parrots now have established breeding popu-
lations (Lever 1987). The most abundant of these is the Monk Parakeet (Myiopsitta mona-
chus). The exact date at which Monk Parakeets established breeding colonies in the United
States is unclear because of uncertainty over when and where birds were released or escaped.
The first confirmed sighting was in 1967 in New York City (Lever 1987), and the species
was breeding there shortly thereafter (Bull 1973). By the early 1970s, the species was so
widespread that the United States Fish and Wildlife Service (USFWS) initiated a control
and removal program on the basis of the species’ reputation in South America as an agri-
cultural pest (Bump 1971, Bucher et al. 1992). By 1975, the year this program ended, the
population of parakeets had been reduced by approximately one-half (Neidermyer and Hick-
ey 1977). Since then, the numbers of Monk Parakeets have increased and the species has
begun a dramatic population expansion to levels far above the pre-control numbers in the
early 1970s. In this paper we document and analyze population trends of the Monk Parakeet
in the United States from 1972 to the present.

Methods. — We summarized Christmas Bird Count (CBC) records published in American
Birds” (now “Field Notes”), personal communications solicited from bird watchers and
ornithologists, and a continuing study of Monk Parakeets in Hyde Park, Chicago, that was
initiated in 1992 (Hyman and Pruett-Jones 1995, Van Bael and Pruett-Jones pers. obs.).

We summarized CBC records from the 1971-1972 count to the 1994-1995 count. In
examining these data for the 1972-1973, 1981-1982, 1986-1987, 1992-1993, and 1994-
1995 counts, we checked records for every reporting locality in the contiguous United States.
For the intervening years, we checked records for every locality within all states that re-
ported at least one Monk Parakeet during at least one of the five counts listed above. For
each CBC locality, we noted the total number of birds reported as well as the number of
party hours. The regional reports were checked each year. For some years. Monk Parakeets
were recorded during the “count week” at a given locality and mentioned in the regional
summaries, but no birds were actually observed on the formal count day. In tabulating
numbers of individuals recorded, we counted “count week records as one parakeet at that
given locality.

To calculate the rate of population growth, we used the standard equation defining ex-
ponential growth N,+ , = N,e" where N,+ , is the population size at time l+U N, is the
population size at time t, r is the rate of population growth, t is the time interval, and e is
the natural logarithm base. This equation can be rewritten as r = (lnN,+ , - lnN,)/t. We
calculated r for each one-year time interval beginning in 1975 (the year the USFWS control
program ended). A plot of r versus population size indicates whether a population is ex-
panding, declining, or has reached a stable equilibrium size. The equation above defining r
can be rewritten as t = ln2/r to calculate the time interval for a population to double in
size.

Results. The Monk Parakeet was already widely distributed in the United States by the

early 1970s. This appears to have been the result of geographically separated releases and
escapes of captive birds (Neidermyer and Hickey 1977). The USFWS control program
reduced the population size very successfully. This reduction is indicated both by the pub-
lished records of the USFWS (Neidermyer and Hickey 1977) and by CBC records. At the
start of the control program, birds were reported from 21 localities in seven states on the

584

SHORT COMMUNICATIONS

585

1972-1973 CBC. Three years later, this was down to seven localities in hve states (1975-
1976 CBC).

Since 1975, the number of states and localities at which Monk Parakeets have been
reported and the total number of individuals counted have all increased. Over the last five
years (since the 1990-1991 CBC), the species has been reported from 76 localities in 15
states (Fig. 1). This includes 62 localities in 13 states from CBC records and regional
summaries and an additional 14 localities in two states from personal communications to
us from ornithologists. The population increase has been dramatic; on CBCs, 1816 birds
were counted in 1994-1995 compared to 33 birds in 1975-1976. Monk Parakeets are not,
however, evenly distributed across the United States. Two states, Florida and Texas, ac-
counted for 1463 (80.6%) of the birds recorded on the 1994-1995 count (see Fig. 1).

The increase in numbers of Monk Parakeets fits an exponential model of population
growth (Fig. 2). Regression of number of individuals recorded per party hour of effort (In)
by year from 1975 to 1995 is linear and statistically significant (Fig. 2, C = 188.94, =

0.908, df = 19, P = 0.0001). The average annual rate of population growth (r) equals 14.6%
(N = 19, range = -58 - 76%), yielding a population doubling time of 4.8 years. A plot
of population growth rate as a function of population size (Fig. 3) shows considerable
fluctuation, but there is not yet any indication that the population is approaching an equi-
librium level. The geographical range of Monk Parakeets has also increased since 1975. A
plot of the number of CBC localities (In) reporting the species since 1975 is linear and
statistically significant (F = 123.01, = 0.865, df = \9, P = 0.0001).

In Hyde Park, Chicago, the population of Monk Parakeets increased from 64 in 1992 to
95 in 1993 (see Hyman and Pruett-Jones 1995). The population was not censused in 1994,
but in 1995 we counted a minimum of 85 nesting chambers, indicating a population size of
approximately 170 adults. This population has almost tripled in three years.

586 THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Year

Fig. 2. Regression of total number (In) of Monk Parakeets recorded on Christmas Bird
Counts in the contiguous United States each year since 1975.

Discussion. — Our analysis shows that the Monk Parakeet is currently experiencing ex-
ponential growth in both its population size and geographical range in the United States.
There is evidence to suggest that this increase is due to reproduction within existing pop-
ulations rather than an increase in observer effort on CBCs or continued releases. First, the
number of states reporting parakeets has remained relatively stable for the last ten years,
fluctuating from five to nine. The increase in localities reporting parakeets has come from
additional localities in those states already reporting the species. This suggests that the

O ^
<D P

-4— > ^

QL c
u o

£ g.

£= O
~ CL

.80-1

.60-

•

.40-

•

.20-

••

0-

-.20-

•

-.40-

-.60-

-.80-

•

1 r

• •

♦ •

T 1 1 1 1

Fig.

United

— I — I — I — I — I — I 1 I I I I I I

0 .05 .10 .15 .20 .25 .30 .35 .40 .45 .50 .55

Population Size (Birds/Party Hours)

3. Plot of annual rate of population growth of Monk Parakeets in the contiguous
States for the period 1975-1994.

SHORT COMMUNICATIONS

587

populations in those states are increasing and expanding into new areas. Second of the 1816
individuals counted on the 1994-1995 CBC, 1253 (69%) were in localities which have had
populations of Monk Parakeets continuously for the last eight to 10 years. At most of these
localities, the numbers ot parakeets reported has increased steadily over this time period.
Finally, a known population in Hyde Park, Chicago, has experienced a dramatic increase in
recent years believed to be entirely due to local production and recruitment of offspring. If
other populations of parakeets are as productive as the birds in Hyde Park, Chicago, the
nationwide pattern of exponential population growth can easily be explained. Although
accidental or purposeful releases of Monk Parakeets probably continue, we consider it un-
likely that these contribute to or explain their population growth.

The total population of Monk Parakeets is obviously much larger than indicated by the
CBCs. In order to estimate the total population size, we need values of two parameters: the
proportion of breeding populations of parakeets that are covered by the CBC count circles
and the proportion of birds resident in the count circles that are actually recorded. As
indicated in Results, over the last five years. Monk Parakeets have been recorded at 76
localities in 15 states. The CBCs comprised 49 (64.5%) of these localities. We can use this
value (0.645) to estimate the proportion of parakeet populations that are covered by the
CBCs. The second parameter, the proportion of resident birds that are actually counted is
much more difficult to estimate. Unfortunately, there are no CBC localities reporting Monk
Parakeets for which separate censuses of parakeets are also available. For example, in Hyde
Park, Chicago, our census data are from an area not included in any of the Chicago CBC
count circles. Without actual data, we cannot estimate this second parameter. If, hypotheti-
cally, the CBCs counted an average of half of the parakeets actually present in any one
count circle, we can calculate what we consider to be a very conservative estimate of the
total population as 1816/(0.645 X 0.50) = 5631. If, in contrast, the CBCs counted only an
average of 10% of the parakeets in an area, the estimate would be 1816(0.645 X 0.10) =
28,155. The large range in these values illustrates how important census data will be to
accurately estimate total population size.

Unless some decision is made to control the population of Monk Parakeets, it seems
likely that the species will continue its range expansion and population increase in North
America. As indicated by the success of the USFWS control program, the species is rela-
tively easy to control through eradication of birds at their colonial nests. Nevertheless, the
social and ethical issues associated with eradicating parakeets have lately proven much more
difficult than the practical issues (cf. Temple 1992). In Hyde Park, Chicago, for example, a
decision by the United States Dept, of Agriculture in the mid-1980s to remove the birds
resulted in the formation of a citizen’s action group to protect the parakeets and a threatened
lawsuit. At present, it appears that in many areas the novelty of having a resident parrot
species and concerns over animal welfare outweigh potential risks of the birds’ becoming
a threat to agriculture. Because Monk Parakeets may have that potential, continued moni-
toring of their populations and initiation of more detailed studies seems warranted.

Acknowledgments. — We thank the numerous individuals that shared their observations of
Monk Parakeets with us, M. Pruett-Jones, D. Tanning, and two anonymous reviewers for
helpful comments on the manuscript, and Dept, of Ecology and Evolution, Univ. of Chicago,
and National Science Foundation for support.

LITERATURE CITED

Bucher, E. H., L. F. Martin, M. B. Martella, and J. L. Navarro. 1992. Social behaviour

and population dynamics of the Monk Parakeet. Proc. 20th Int. Ornithol. Congress.

Christchurch, New Zealand, pp. 681-689.

Bull, J. 1973. Exotic birds in the New York City area. Wilson Bull. 85:501-505.

588

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Bump, G. 1971. The South American Monk, Quaker, or Gray-headed Parakeet. U.S. Fish
and Wildl. Serv., Special Sci. Rep.-Wild. No. 136.

Hyman, J. and S. Pruett-Jones. 1995. Natural history of the Monk Parakeet in Hyde Park,
Chicago. Wilson Bull. 107:510—517.

Lever, C. 1987. Naturalized birds of the world. Longman Scientific & Technical, London,
England.

Neidermyer, W. J. and J. J. Hickey. 1977. The Monk Parakeet in the United States, 1970—
1975. Am. Birds 31:273—278.

Temple, S. A. 1992. Exotic birds: A growing problem with no easy solution. Auk 109:
395-397.

Sunshine Van Bael and Stephen Pruett-Jones, Dept, of Ecology and Evolution, Univ. of
Chicago, 1101 East 57th St.. Chicago, Illinois 60637 (address correspondence to SP-J).
Received 27 Sept. 1995, accepted 10 Feb. 1996.

Wilson Bull., 108(3), 1996, pp. 588—591

Forest gap use by breeding Black-tbroated Green Warblers. Habitat heterogeneity
results from environmental gradients and disturbances that create spatiotemporal patchiness
(White and Pickett 1985). Fine-grained patchiness resulting from forest gaps is a condition
common in temperate forests (Blake and Hoppes 1986) and is typically caused by one to
several tree-falls or tree death (snag) ranging in area from 0.0025 ha to about 0.1 ha (see
Lorimer 1989). The resulting heterogeneity represents a habitat mosaic important to many
species.

While collecting data on the foraging behavior of Black-throated Green Warblers (Den-
droica virens) along the northern shoreline of Lake Huron in Michigan’s eastern Upper
Peninsula, we quantified breeding bird use of forest gaps. Data were collected from 14 June
through 19 July 1994. Transects were established parallel to the Lake Huron shoreline at
distances of 0.4 km (0.25 mile), 0.8 km (0.5 mile), 1.6 km (1.0 mile) and 3.2 km (2.0 mile).
Observers followed these transects for a distance of 6.4 km (4.0 mile), collecting observa-
tions on males (and females if possible) at each established territory (determined by the
presence of a singing male). Because birds were territorial, only one observation per sex
was made at any location. A minimum distance of 100 meters between observations was
established to ensure the independence of data collected (Heijl and Verner 1990).

We used Brokaw’s definition of a forest gap — a hole (minimum of 5 m in diameter) in
the forest canopy extending through all levels down to an average height of two meters
above ground (Brokaw 1982). A bird was considered to be using a gap if it was observed
foraging or singing within 1 m of the canopy edge. We did not count transients — birds flying
through the gap or otherwise obviously not using the gap to obtain food or as a territorial
boundary.

Forest vegetation in the study area consisted of a mixture of conifers including balsam
fir (Abies balsamea), northern white cedar (Thuja occidentalis), white spruce (Picea glauca),
eastern white pine (Pinus strobus) and deciduous species including paper birch (Betula
papvrifera), quaking aspen (Populus tremuloides), balsam poplar (Populus balsamifera) and
red maple (Acer rubrum). Mature canopy was approximately 13.5 m with an understory
principally of balsam fir and white spruce.

Observations within 50 m of roads, open fields or the Lake Huron shoreline were not

SHORT COMMUNICATIONS

589

Fig. 1. Breeding Black-throated Green Warbler gap usage, 1994.

included in the statistical analysis in an effort to reduce possible bias due to edge effect.
We used log-likelihood ratio G tests to analyze for differences in gap use by breeding birds
(Zar 1984). As we did not measure gap area, we were unable to base our expected values
on relative area of gaps versus contiguous forest. Thus, we were conservative in our cal-
culation of expected values, assuming a 1:1 ratio of gap use to forest use.

During the breeding sea.son, males were observed utilizing forest gaps significantly more
and contiguous forest significantly less than expected (G = 5.19, P < 0.05; Fig. I). Males
were observed in, or immediately adjacent to, tree-fall forest gaps in 52.1% of our obser-
vations and in contiguous forest in 30.9% of our observations (N = 94). Females demon-
strated an even greater difference in use, utilizing gaps 55.0% of the time and contiguous
forest 25.0%, however, this was not significant (C = 2.31, F > 0. 10) (Fig. 1). This lack of
significance may be the result of low statistical power due to small female sample size (N
= 20).

These observations, though potentially limited in that data were from only one breeding
season, indicate the potential importance of gaps to breeding birds. Gaps may be important
to breeding birds for a variety of reasons. For instance, foliage insectivores such as the

590

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Black-throated Green Warbler may select preferentially gaps in response to differential prey
abundance. There may be more insects in gaps due to greater primary productivity associated
with higher light levels (Logden 1972, Blake and Hoppes 1986). Previous work has dem-
onstrated differences in assemblages of birds captured in gaps and the surrounding under-
story (Blake and Hoppes 1986, Martin and Karr 1986, McGowan-Stinski, pers. comm.).
These differences have been correlated to an increased insect, fruit, and total foliage abun-
dance in forest gaps (Blake and Hoppes 1986, Martin and Karr 1986).

The increased light penetration in forest gaps (Logden 1972, Blake and Hoppes 1986),
may result in warmer microclimates in gaps as compared to contiguous forest. Warmer
microhabitats could beneht the thermoregulatory physiology of insects, increasing insect
activity and abundance. Birds may thus be attracted to this activity and/or abundance as
they attempt to increase foraging opportunities. Additionally, warmer temperatures within a
gap may benefit a bird as it attempts to budget energy and deal with exigencies faced during
the early breeding season such as cold mornings. Thus, gap foraging may be energetically
advantageous through providing a bird with a concentrated source of prey, as well as a
reducing an individual’s thermoregulatory costs, allowing birds to shunt more energy into
the breeding effort and less into individual maintenance.

The vegetation structure of forest gaps may help birds maximize their foraging profit-
ability. Because gaps typically have a lower vegetation profile as well as a higher density
of foliage (see Martin and Karr 1986), gaps may present birds with a more compact area
to visually search and move about in. Increases in foliage density may yield more insect
prey per unit of search time and reduce an individual s energetic expense in movement.
High foliage density in gaps may also reduce a foraging bird’s vulnerability to predators.

Linally, anecdotal evidence suggests forest gaps may aid males in establishment and
maintenance of territorial boundaries. Black-throated Green males typically spend much of
their time defending territory through visible and acoustical advertisement (Morse 1993).
Utilization of a forest gap, especially a younger gap, as a territorial boundary may benefit
males through increased visibility and song projection. We observed numerous aggressive
male interactions occurring over or immediately adjacent to forest gaps by territorial males.
Singing males were also observed responding to each other across gaps. This evidence
suggests that at least some males at our study site may have been using forest gaps as
territorial boundaries.

Our data suggest the importance of forest gaps to breeding Black-throated Green Warblers
along the northern Lake Huron shoreline. This use suggests that structural heterogeneity in
mature forest provides Black-throated Green Warblers with higher quality habitat than is
found in less heterogeneous, more even-aged forest stands. A better understanding of the
importance small scale disturbances play is critical if forest managers are to maintain high
quality habitat for breeding resident and neotropical migrant birds.

Acknowledgments. — This project was funded in part by The Nature Conservancy — Mich-
igan Chapter, Central Michigan Univ. Graduate School and Biology Dep., and the National
Lish and Wildlife Loundation. We thank M. J. Hamas of Central Michigan Univ. and D.
Ewert of the Michigan Chapter of the Nature Conservancy for assistance in the field as well
as advice and support. The manuscript benefited from the comments of G. Kelley, T. Con-
treras, D. Cimprich and J. McGowan-Stinski.

LITERATURE CITED

Blake, I. G. and W. G. Hoppes. 1986. Influence of resource abundance on use of tree-fall
gaps by birds in an isolated woodlot. Auk 103:j28— 340.

Brokaw, N. V. 1982. The definition of treefall gap and its effect on measures of forest

dynamics. Biotropica 14:158-160.

SHORT COMMUNICATIONS

591

Fogden, M. R 1972. The seasonality and population dynamics of equatorial forest birds in
Sarawak. Ibis 1 14:307-343.

Heijl, S. J. and J. Verner. 1990. Within-season and yearly variations in avian foraging
locations. Stud. Avian Biol. 13:202-209.

Lorimer, C. G. 1989. Relative effects of small and large disturbance on temperate hard-
wood forest structure. Ecology 70:565-567.

Martin, T. E. and J. R. Karr. 1986. Patch utilization by migrating birds: resource oriented?
Ornis Scand. 17:165-174.

Morse, D. H. 1993. Black-throated Green Warbler (Dendroica virens). Pp. 1—20 in The
birds of North America, No. 55. (A. Poole and F. Gill, eds.). Acad. Nat. Sci. of Phila-
delphia; Am. Ornithol. Union, Washington, D.C.

White, P. S. and S. T. Pickett. 1985. Natural disturbance and patch dynamics: an intro-
duction. Pp. 3-13 in The ecology of natural disturbances and patch dynamics. (S. T.
Pickett and P. S. White, eds.). Academic Press, Orlando, Florida.

Zar, j. H. 1984. Biostatistical analysis. 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, New
Jersey.

Robert Smith and Matthew Dallman, Dept, of Biology, Central Michigan Univ., Mt.
Pleasant, Michigan 48858. (Present address RS: Dept, of Biological Sciences, Univ. of
Southern Mississippi, Hattiesburg, Mississippi 39406. Present address MD: The Nature Con-
serx’ancy, 618 Main Street West, Ashland, Wisconsin 54806). Received 3 Nov. 1995, ac-
cepted 1 March 1996.

Wilson Bull., 108(3), 1996, pp. 591-592

Courtship behavior of Golden-cheeked Warblers. — The Golden-cheeked Warbler
(Dendroica chrysoparia) is an endangered species with a known breeding range mostly
confined to the Edwards Plateau of Texas. These warblers inhabit oak-juniper woodlands
and are dependant on Juniperus ashei bark for nesting material (Sexton, Birding 24:373-
376).

Pulich (1967, The Golden-cheeked Warbler, a Bioecological Study, Texas Parks and Wild-
life) described courtship displays in Golden-cheeked Warblers in which males attentively
followed females and briefly displayed before copulation. There are no other published
accounts of this type of behavior in this species. Courtship behavior in Golden-cheeked
Warblers was observed on two occasions during spring 1995. The following observations
were made at Pedernales Falls State Park, Blanco Co., Texas on 27 March 1995 and at
Colorado Bend State Park, San Saba Co., Texas on 4 April 1995. In both cases, a female
Golden-cheeked Warbler was discovered constructing the base platform of a nest prior to
the observations of courtship behavior. The female warbler would make short forays into
neighboring Juniperus ashei to collect strips of bark. All of the nesting material gathered
during these observations was from trees within 15 m of the nest tree. The nest at Pedernales
Falls was 5.3 m high in J. ashei, and the nest at Colorado Bend was 4.7 m high in Ulnius
crassifolia. While observing the female’s activities, a quiet, warbler-like song was heard
that was unlike either of the primary songs of the Golden-cheeked Warbler (Pulich 1976).
This song was muted, but more rapidly paced, than the typical songs of the species. This
combination gave the song a more twittery quality than the primary songs. However, despite
these differences, the song had tonal qualities similar to the other .songs of the species. The

592

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

male warbler was observed singing this song in trees near the nest. He sang for two to three
minutes during which the female continued nest construction; no change in her behavior
was noticed. The male stopped singing as he approached the nest site. On both occasions
the male was carrying strips of juniper bark. The bark was given to the female. The female
then placed bark strips in the nest. While the female was placing the bark, the male quietly
sang the twittery song and spread his tail while slowly lowering and flicking his wings
closed. When the female finished placing the bark strips, she faced the male, quietly chipped,
and crouched with her wings slightly spread and her head down. Similar behavior was noted
in both cases. During the first observation, copulation occurred on the nest platform. The
nest of the second pair was farther along in construction and the copulation occurred next
to the nest. After copulation the male repeatedly sang the same twittery song while flying
from perch to perch, widely circling the nest tree. During this time, the male constantly
flitted its wings and fanned its tail. Tail and wing fanning are not limited to courtship
displays; I have observed similar behavior in Golden-cheeked Warbler territorial interactions
as well as toward Texas rat snakes (Elaphe obsoleta lindheimeri) and when a female Brown-
headed Cowbird (Mothrus ater) was near the nest.

Nest building in Golden-cheeked Warblers is reported to be done entirely by the female
(Pulich 1976, Oberholser, 1976, The bird life of Texas, Univ. of Texas Press). During the
early spring of 1994 and 1995 I observed males carrying nesting material on five occasions.
Other workers also have observed male Golden-cheeked Warblers visiting nests under con-
struction (Keddy-Hector, pers. commun.). This suggests that male warblers may have some
role in nest construction.

These observations suggest that male Golden-cheeked Warblers carrying nesting material
may have been part of courtship. This does not eliminate the possibility that, at least oc-
casionally, the male may have a minor role in nest construction. However, no direct obser-
vations were made of males adding materials to a nest. In addition, no specific male was
observed carrying nesting material on more than one occasion. During previous observations
of male Golden-cheeked Warblers carrying nesting material, there was no evidence of a
female in the vicinity and the twittery song that was given prior to and following the
courtship behavior was not heard.

Acknowledgments. — I thank Dean Keddy-Hector for his input throughout the preparation
of this manuscript. I also thank David Riskind, Leila Gass, Millicent S. Licken, and C. R.
Blem for their comments on earlier drafts of this manuscript.

Mark W. Lockwood, Natural Resource Program, Texas Parks and Wildlife Dept., 4200
Smith School Road, Austin, Texas 78744. Received 14 Nov 1995, accepted 24 Feb. 1996.

Wilson Bull., 108(3), 1996, pp. 593-601

ORNITHOLOGICAL LITERATURE

Edited by William E Davis, Jr.

The Northern Goshawk: ecology and management. By William M. Block, Michael L.
Morrison, and M. Hildegard Reiser, editors. Studies in Avian Biology 16. Cooper Ornitho-
logical Society, % Western Foundation of Vertebrate Zoology, 439 Calle San Pablo, Cam-
arillo, California. 1995: 34 figs., 58 tables. $16.00 (paper). — Honest-to-goodness updates of
species life-history information — as opposed to simple rehashings of previously published
accounts — are always welcome additions to the ornithological literature. This is especially
true when the species in question is one whose known biology is steeped as much in myth
and traditional lore as ecological reality, and even more so when the information in question
focuses on limiting-factors research at multiple sites. The current offering is one such effort.
The published result of a symposium on the biology and management of the Northern
Goshawk (Accipiter gentilis) that was held in conjunction with a Cooper Ornithological
Society meeting in April, 1993, the work represents “a compendium of current information
on goshawk biology and management” in North America.

The symposium was organized because evidence available at the time suggested that
North American populations of goshawks, especially some of those in the western United
States, were declining, and because resources managers had insufficient information upon
which to base practical and effective management efforts. Although the former is no longer
thought to be the case, the resulting proceedings remains a useful and timely contribution
to the avian literature. The 22, 4- to 9-page papers that make up the work were authored
or co-authored by an astounding 41 individuals, testimony to recent interest in the species.
The work includes six contributions on “research approaches and management concepts,”
nine on “resource ecology,” and seven on “population ecology.” Not surprisingly, not all
of the efforts included therein are as scientifically rigorous or as gracefully presented as one
might hope. Several papers, for example, offer but brief and preliminary glimpses of works
that were still in progress at the time. Nevertheless, the proceedings’ editors are to be
congratulated for producing a work that is both editorially clean and reasonably seamless.

Several of the more generically useful papers include ones by Clint Boal on aging nestling
goshawks, and Suzanne Joy, Richard Reynolds, and Douglas Leslie on the costs of benefits
of broadcast survey techniques for breeding goshawks. Although most “management” pa-
pers in the work are clearly aimed at site-specific situations, many of the conclusions reached
should be of general interest to managers of forest raptors. A paper by Pat Kennedy, Johanna
Ward, George Rinker, and Jim Gessaman on post-fledging areas in goshawk home ranges,
for example, insightfully concludes that management strategies need to account for the
habitat requirements of recently fledged young, as well as for the nest-site and foraging-
area requirements of parental adults. Similar bits of wisdom occur in other papers.

Overall, the symposium offers a wealth of new information, as well as a good read. There
is, however, one troubling aspect of the work on which I need to comment. Many of the
management strategies laid out in the symposium are basically “product” rather than “pro-
cess” driven (i.e., most managers still .seem intent on maximizing local goshawk produc-
tion— principally by creating species-specific, “designer habitats” for the species--rather
than on trying to establish fully functional, naturally forested ecosystems typical of the
regions in question). Given that the goshawk does not appear to be severely threatened in
any major portion of its North American range, should we really be trying to maximize its
production in managed forests. Wouldn’t conservation interests be better served by trying
to reestablish a full suite of natural ecosystem functions in “natural” forests. As I read most

593

594

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

of the management papers in the work, I couldn’t help but wonder whether the use of
silvicultural analogs alone (e.g., substitute “numbers of Northern Goshawk nestlings pro-
duced per year per” for “amount of softwood board feet per year”) will ever really work
in forest conservation and whether we need to be looking at management strategies that
extend beyond the current single-species, production-oriented, paradigm.

This latter, not-so-minor, caveat notwithstanding, the current offering should be read by
all who have an interest in the species, as well as by those with interests in the species’
forested habitats. The Cooper Ornithological Society is to be congratulated for making the
proceedings available so quickly, and at such a reasonable price. — Keith L. Bildstein

How BIRDS MIGRATE. By Paul Kerlinger, illus. by Pat Archer. Stackpole Books, Mechan-
icsburg, Pennsylvania 1995: 228 pp., 16 maps, 43 figures, 2 tables. $14.95 (paper). — The
ornithological literature is full of books on bird migration. My own collection takes up more
than a meter of shelf space, and the current offering represents Kerlinger’s second addition
to the group in less than a decade. Even so, “How birds migrate” provides an especially
welcome and long-overdue complement to the existing literature. Indeed, not since Donald
Griffin’s classic work on the subject in 1964 (Bird migration. Doubleday, Garden City, New
York) has anyone managed to distill and highlight the current ornithological literature on
bird migration in such an appealing and productive fashion. Kerlinger’s text, which contains
little of the byzantine modeling, mathematics, and jargon of the current technical literature,
retains all of its essential findings and excitement.

The work consists of 15 chapters, ranging from one on why birds migrate to one on how
conservationists are attempting to protect them. There is also an especially useful annotated
list of additional references. Most chapters focus on single aspects of migration, introducing
and detailing their essential features and concluding with a series of case studies highlighting
the phenomenon in individual species. The work is accompanied by an effective series of
maps that depict the migratory pathways of species, as well as by numerous illustrations of
specific aspects of migration behavior and ecology.

In his previous offering (1989, Flight strategies of migrating hawks, Univ. of Chicago
Press, Chicago, Illinois), Paul Kerlinger demonstrated an extraordinary talent for summariz-
ing the migration literature for his colleagues. His current effort demonstrates a similar talent
for doing so for the much larger lay audience. In fact, Kerlinger sets an enviable standard
for others who would try to do the same. The book’s level of treatment, together with its
modest price, make it an appropriate companion text for introductory courses in ornithology
and avian ecology and behavior. “How birds migrate” should serve as a useful introduction
to bird migration well into the next century. — Keith L. Bildstein

The summer atlas of North American birds. By Jeff Price, Sam Droege, and Amy Price.
Academic Press, San Diego, California. 1995: 364 pp., 463 maps, 16 line drawings. $45
(cloth). — The stated goals of this book are to help birders find birds, and to provide infor-
mation and guidance to conservation organizations and land managers. The book certainly
reaches these goals. The data base of this book is the National Biological Survey s North
American Breeding Bird Survey (BBS). The first chapter provides an overview of the BBS,
including a discussion of the biases and constraints inherent in a survey which is confined
to roadsides and in which most of the data is collected by amateur volunteers. Chapter 2

ORNITHOLOGICAL LITERATURE

595

explains in detail how the BBS data was used to create the maps, and all the problems,
biases, checking, and trouble shooting involved. The bulk of the book, chapter 3, consists
of relative abundance maps, with a level of resolution of 50 km by 50 km blocks, for 450
species and identifiable forms of North American summer birds. Most of the maps show
averages of data collected from hundreds of 1985—1991 BBS routes (50 three-minute point
counts along a 40 km route) in southern Canada and the contiguous United States. Strictly
speaking, the maps do not show breeding bird distribution because non-breeding birds may
be represented. Each map has four levels of abundance displayed (<5, 5-20, 20-50, and
>50 birds detected per route per year), with a minimum mapped value of 0.5. Unfortunately,
the four levels are shown by different levels of intensity of the same color, which makes
some maps difficult to read. Additional maps show species richness patterns for all species
combined, and for several groups of birds (e.g., herons, waterfowl, flycatchers, warblers).
Chapter 4 presents an annotated list of species of 531 species or forms. This includes the
species for which there was insufficient data to create a map, and BBS data from Alaska
and northern Canada. The annotations for each species include a description of habitat, and
three BBS routes that have had consistently high counts for the species. For each route the
information presented includes the average number of individuals per year detected, fre-
quency (e.g., 7/7 indicates that the species was detected in all seven census runs from 1985-
1991), and location. A fifth chapter addresses population trends and conservation issues. It
includes an extensive table of population trends (percentage change), both long term (1966-
1993) and short term (1984—1993), with levels of statistical significance for population
changes indicated. The chapter includes a discussion and analysis of possible causes for
population trends in several groups of birds (e.g., scrub nesters, open water and wetland
species), keying mostly on habitat alteration. The four appendices include a list of scientific
names, references cited and suggested readings, the American Birding Association’s code
of ethics, and selected birdfinding guides and breeding bird atlases for each state. The index
includes only bird species. The 16 drawings by David D. Beadle add an attractive touch to
what is essentially a book of maps.

I find this an interesting and useful presentation of BBS data which should have heuristic
value, and recommend it to anyone interested in avian biogeography, conservation, or bird-
watching.— William E. Davis, Jr.

The atlas of breeding birds of Connecticut. By Louis R. Bevier (ed.). Bulletin 113,
State Geological and Natural History Survey of Connecticut, Dept, of Environmental Pro-
tection, Hartford. 1994: 461 pp., 184 maps, 190 line drawings. Available from: DEP-Pub-
lications, 79 Elm Street, Store-Level-MO, Hartford, CT 06106-5127 (checks payable to
DEP-PUBLICATIONS). $36.95 plus $3 p&h (cloth). — Another in a series of state and
province breeding bird atlases, this book provides baseline data for the breeding birds of
Connecticut. These data should be useful both for future evaluations of natural and human-
induced bird population changes and for immediate use in making informed decisions on
conservation priorities by conservationists, legislators, and state agencies. Because of its
relatively small size, all of the state was censused (larger states have used a priority block
system), dividing the state into 121 7.5 minute quadrangles (USGS quadrangle system),
each divided into 6 blocks of approximately 25 km^ each. More than 500 individuals were
involved in the data collection from 1982-1986. “Block busting” (a few observers spent a
few hours) ensured at least some coverage for every block. Each species was assigned a
breeding status of “Possible,” “Probable,” or “Confirmed” using suitable criteria and maps

596

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

for each species were compiled. Two hybrids and 173 species were in the confirmed cate-
gory, including two species which had not been previously recorded breeding in Connect-
icut.

The book has an interesting design — 24.5 cm wide, 20.4 cm tall — which makes for easy
reading. The maps are large and particularly easy to read. Several sections deal with the
limitations of the study (e.g., level of effort of data collection was neither uniform nor
measurable) and there is an interesting 1 1 page section by George A. Clark, Jr. on inter-
preting the distribution of breeding birds (including associated problems and constraints).
The bulk of the book is occupied by the species accounts. Each species which was either a
confirmed or probable breeder is described in a two (facing) page section which includes
an introductory statement, and habitat requirements, atlas results, and discussion sections.
The introductory statements usually provide current status information (e.g.. Threatened or
Special Concern), and the discussion section generally concentrate on historical accounts
and population trends. An attractive line drawing by Michael DiGiorgio accompanies each
account. More succinct accounts are presented for possible breeding and miscellaneous
species. One appendix summarizes the breeding status (e.g., number of blocks in which
breeding was confirmed), a second presents Breeding Bird Survey population trends for
Connecticut and southern New England, and a third a list of common and scientific names
of plant and animal species (excluding birds) mentioned in the text. The Literature Cited
section contains over 500 references and is a veritable goldmine for local references.

This is a generally excellent book, but is not without problems. I was sorry that the Monk
Parakeet (Myiopsitta monachus) was not included in the data collection, since it has been a
documented breeding species for years, and population changes in this species should be of
future interest. The species accounts, although well done, are a bit thin, in many cases
occupying less than a page of text. There is a lot of white space that might well have been
usefully filled with descriptions of breeding-related behaviors such as nest building, court-
ship displays, foraging behavior during the breeding season, and care of the young. These
are minor quibbles however — this is an important contribution to the breeding-bird literature
of North America. — William E. Davis, Jr.

The West Virginia breeding bird atlas. By Albert R. Buckelew, Jr. and George A.
Hall. Univ. Pittsburgh Press, Pittsburgh. 1994; 232 pp. plus acetate overlays, 173 distribution
maps, 4 introductory tables and 12 figures. $27.95 (cloth).— With this attractive publication.
West Virginia joins the ranks of at least 15 North American states and provinces that have
published breeding bird atlases. In the present volume, more than 300 volunteers conducted
field work between 1983 and 1989. The southeast corner of 7.5' topographic maps was
targeted as the priority block for most of the state. In addition to priority blocks, fieldwork
was conducted in blocks with special features, volunteers’ “favorite” areas, and locations
covered by Brooks Bird Club field trips. Of the potential 2,700 blocks in the state, 516 were
targeted by the study. A total of 171 species is mapped.

The introduction provides background on West Virginia s atlas effort, methods, and or-
ganization. It also includes a brief overview of biogeography, physical features, and climate
of the state that may influence bird distribution. Maps that display physiographic features
are presented here and as seven acetate overlays. Because the state’s avifaunal regions and
habitat types are nicely described and illustrated by Hall (“West Virginia Birds,” 1983),
they are not duplicated in the present volume. Project results are disappointingly slim, with
little reference to survey effort. No indication of the number of hours of fieldwork, for

ORNITHOLOGICAL LITERATURE

597

example, is presented, and the exact number of blocks surveyed is not clear. Survey effort
seems comparable to most other atlases, although the authors state that coverage was not
even. Highlights include first breeding records in West Virginia for the Yellow-bellied Ely-
catcher {Empidonax flaviventris) and Yellow-rumped Warbler (Dendroica coronata). Other
northern species, such as Veery (Cotharus fuscescens) and Dark-eyed Junco (Junco tiymen-
alis), were documented farther south in the western hills than expected.

Like many other states. West Virginia used a priority block system to promote uniform
coverage. However, the inclusion of nonrandom blocks has the risk of accentuating coverage
inconsistency. Opportunistic coverage of blocks selected by volunteers, notably in the Ca-
naan Valley (Blackwater Ealls quadrangle), creates an awkward cluster of observations for
common species, such as Common Yellowthroat {Geothlypis trichas). This problem was
avoided in “The Atlas of Breeding Bird of Michigan” (R. Brewer, G. A. McPeek, and R.
J. Adams, Jr.; 1991) by mapping results at a coarser scale than data were collected.

Species accounts are concise, incorporating the atlas map, text, and a tabular summary
on a single page. Eour map symbols are used to reflect the levels of breeding evidence.

Observed” records are mapped, even though those records are defined as indicating no
evidence of breeding. The map symbols are not, to my eye, intuitively hierarchial, but have
the advantage of being clearly distinguishable. The base map includes state counties and a
grid of topographic maps, creating a busy map. The topographic grid, and corresponding
key in the appendix, makes this one of the easiest atlases in which to pinpoint the location
of records. The accounts are generally free of typographical errors and are well edited. They
provide a concise summary of habitat, distribution, historical patterns, and population trends.
An appendix discusses the status of eight species not confirmed during the atlas efforts. The
literature cited is surprisingly brief and reflects frequent reference to Hall’s previous volume.
The index includes both English and scientific names.

West Virginia has contributed a useful addition to the growing collection of published
breeding bird atlases, targeted well to local interests. I recommend it to students of bird
distribution in the region. — Daniel W. Brauning.

A BIRD-FINDING GUIDE TO ONTARIO. Revised Edition. By Clive E. Goodwin. University of
Toronto Press, Toronto. 1995; 477 pp., 41 maps. $24.95 (paper). — This revised edition
updates the information in the 1982 edition and provides more information on the status of
individual species. The task of providing a bird-finding guide to an area so large and eco-
logically diverse is enormous — Ontario has an area of over one million km^ with deciduous
and boreal forest, tundra, farmland and prairie, and a vast shoreline along the Great Lakes.
It takes five pages to tell you how to use the book (this includes, however, an excellent
section on birding ethics). Chapter 2 introduces the reader to the diversity of habitats and
the birds found in each, as well as a discussion of seasonal variations. Chapters 3-16 are
site guides for individual or clusters of counties in southern Ontario, while chapters 18 and
19 deal with the vast regions of northern Ontario. The chapters start with an overview which
generally includes topography, flora, birding possibilities, and weather and seasonal changes.
The distances are in kilometers, and direction usually given to the neare.st tenth of a kilo-
meter, which is excellent. The maps need to be supplemented by provincial road maps,
however. The coverage of the areas of Ontario I know best (e.g., Algonquin Park) were
adequate but lacked the detail of the American Birding Association birdfinding guide series.
Much useful information, e.g., the Wolfe Island ferry schedule, is included, however. Chapter
19 deals with useful information for the visitor such as accommodations, hot line telephone

598

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

numbers, and important reference books. Chapter 20 is a systematic list of the birds, with
text and bar graphs for each species. An appendix provides common and scientific names
for the mammals, reptiles, and plants mentioned. The index is extensive and printed in the
same type size as the text which makes it easy to read.

The book is generally well written, although a stronger editorial hand could smoothed
out such statements as, “Eifteen species of ducks have been proved to nest. . . ,” and in my
copy page 344 was missing completely. A few of the maps lack scale bars so distances are
hard to judge. The book won’t fit into your pocket but will fit into your glove compartment.
It is a bit overwhelming because of its scope and complexity, but next time I go to Ontario
I will certainly take my copy along. — William E. Davis, Jr.

Birdfinder: a birder’s guide to planning North American trips. By Jerry A. Cooper.
American Birding Association, Inc., Colorado Springs, Colorado. 1995: 374 pp., 40 maps,
22 line drawings. $17.95 (wire-o binding, stiff paper). — This latest addition to the American
Birding Association’s (ABA) “Birder’s Guide” series is in the same format with a wrap-
around rear cover which can act as a book mark and protects the pages. It differs from
previous guides in its scope and intent — it is primarily a trip planner — and should be used
in conjunction with local bird finding guides. The first 19 chapters outline 19 trips, in
chronological order so that they all could be taken in a single year, which theoretically might
allow a birder to see 650 species of birds in North America. An additional “Baker’s dozen”
trips would allow the birder to add a few species to that list. In fairness, list-building is not
the entire focus of this guide — quality of experiences is also emphasized. The chapters
present general itineraries (including map or maps) for the trips scattered about Alaska, the
contiguous 48 states, and Canada, timed at the best season for birds. They also present a
plethora of details on trip planning, including a list of birding guides, telephone numbers
and addresses for outfits which offer special services (e.g., pelagic trips), rare bird alert
telephone numbers, a list of key species that should be looked for, tips on accommodations,
and budget guidelines. A final chapter entitled “The Birdfinder Chart,” which occupies
more than 50 pages, lists the entire ABA checklist of species and indicates on which trips
each is a key, probable, possible, or remotely possible species. This list also functions as
the index to species for the guide — the actual index does not include bird names.

As is typical of the ABA guides, the book is sturdy, easy to read, and well designed. There
are a few typos (mostly words run together) but it is generally well edited. This guide should
be useful to the birder looking for adventure or to the more serious bird student who might
want to have a look around before or after a professional conference. It is loaded with useful
tidbits of information and is certainly worth the money. — William E. Davis, Jr.

Edward Lear: a biography by Peter Levi, Scribner, New York, 1995, 363 pp, 41 black-
and-white illustrations, 2 appendices, bibliography, $30 (cloth). — The English poet, painter,
and travel -writer, Edward Lear (1812-1888), of Lear’s Macaw fame, is best remembered
today for a single child’s verse in which an improbable pair of characters take to sea “in a
beautiful pea-green boat,” court, marry, and “dance by the light of the moon.” “The Owl
and the Pussycat,” written in 1867, was a relatively late addition to the hundreds of songs,
poems, limericks, and “nonsense” rhymes Lear created for the enjoyment ot his friends’
young children. When he published them for a much larger audience, beginning in 1846,

ORNITHOLOGICAL LITERATURE

599

these flights of fancy proved an unexpected commercial success and established Lear as one
of the most beloved children’s writers of all time.

It IS no coincidence that “The Owl and the Pussycat” and so many of Lear’s other stories
feature birds, for the author/artist began his professional career by painting exotic birds at
the London Zoo. By the age of 20, he had published a spectacular monograph on parrots
(“Illustrations of the Family of Psittacidae or Parrots,” 1832) and was providing ornitho-
logical illustrations for many of the most influential British ornithologists of his day, in-
cluding Pndeaux John Selby, William Jardine, T. C. Eyton, and John Gould. His work was
favorably compared with Audubon’s, and he was, in the mid- 1830s, poised at the start of a
career that might have made him the most successful wildlife artist of the 19th century.
How he reached that point, and why he gave it up to pursue a less illustrious life as a poet,
writer, and landscape painter, are among the many interesting aspects of Edward Lear’s life
inadequately addressed in Peter Levi’s new biography.

In his introduction, Levi, a poet and literary historian with several previous biographies
to his credit (including books on Shakespeare, Tolstoy, Tennyson, and Pasternak), tells us
that he set out to write about Lear’s poetry. His scope then broadened to a full biography.
The resulting publication is a sloppy, rambling story that reads more like a dictaphone
transcript than the writing of an Oxford don. Who were the editors, one wonders, who
allowed Levi’s cliche^aden text (“bright as a button,” and “dry as dust”) to lurch between
impenetrable descriptions (“it is useful to notice at once that Lord Derby’s sister Lucy had
married the Rev. Geoffrey Hornby of Winwick, because his grandson, who was born in
1799, married the same Mr. Hornby’s second daughter, who was his first cousin,”) and
irrelevant anecdotes (“as for Poole, one of my uncles flew a Turkish flag there, in front of

his house by the waterside, but not until the 1880’s; in the 1830’s his parents were still in
Istanbul”).

Some simple fact-checking might also have been expected: William Swainson was an
English, not an American naturalist, John Gould’s wife was Elizabeth, not Edith, and it was
the Goulds, not Edward Lear, who illustrated the zoological findings of Charles Darwin’s
voyage on the Beagle.

The book focuses heavily on Lear’s life as an artist, but few of the “wonderful,” “charm-
ing,” “mysterious,” or “heavenly” paintings about which Levi raves are illustrated. This
leaves the reader with a frustrating sensation of having visited an art museum by telephone
and been guided through its disorganized galleries by a docent brimming with strong opin-
ions but few dependable facts. When the artist is allowed to speak for himself, his quotations
are not sourced, further reducing the usefulness of the biography.

Lear deserves better, for he is an appealing, important, and intriguingly complex figure
whose life at times seems more the stuff of 19th century romance than fact. The 20th of 21
children born to a failed stock-broker and an understandably exhausted mother, Lear was
raised by an older sister who gave him his only formal education and his first lessons in
art. Lear s bird paintings at the London Zoo and his ambitious, self-financed monograph on
parrots so impressed the scientific establishment of his day that the Zoo’s chairman. Lord
Stanley (later the 13th Earl of Derby) invited him to illustrate the birds and mammals in
his private menagerie at Knowsley Hall— a project Lord Stanley had previously discussed
with John James Audubon.

Lear enjoyed the financial security and social status such patronage afforded, but he found
the detail of .scientific illustration extremely taxing. “My eyes are so sadly worse,” he wrote
John Gould in 1836, “that no bird under (the size of) an ostrich shall I .soon be able to see
to do.” And so with the same intensity he had lavished on birds, Lear turned to landscape,
traveling first to Ireland and the English Lake District and then to Rome.

Lear found foreign travel much to his liking and went on to visit Malta, Greece, Turkey,

600

THE WILSON BULLETIN • Vol. 108, No. 3, September 1996

Albania, Egypt, and Sinai. From these experiences, he produced several color plate books
in which he described and illustrated his peregrinations.

One of these books, “Illustrated Excursions in Italy” (1846), brought Lear to the attention
of a young Queen Victoria who appointed him as her drawing instructor. Later, Lear returned
to Europe and wrote three more travel books. Then, in 1873, he visited India where he felt
“nearly mad from sheer beauty and wonder.” It was to be the last and most exotic journey
of his life.

Peter Levi traces Lear’s physical, artistic, and emotional travels with enthusiasm and em-
pathy, but his repeated insertions of unrelated personal anecdote, no matter how well intended,
distract from the coherence of the biography. Fortunately, more than a dozen other biographies
of Lear have already been written. The best of these, by far, is Vivien Noakes Edward Lear,
The Life of a Wanderer” (Collins, 1968). While Ms. Noakes’ biography (unfortunately out of
print) is highly recommended, Mr. Levi’s is not. — Robert McCracken Peck

Note. Robert Peck is currently researching for a book on Edward Lear as a natural

history painter. He invites readers with information on this subject to contact him at The
Academy of Natural Sciences of Philadelphia, 1900 Benjamin Franklin Parkway, Philadel-
phia, Pa. 19103; Phone (215) 299-1138; e-mail; peck@say.acnatsci.org.

Polygyny and sexual selection in Red-winged Blackbirds. By William A. Searcy and
Ken Yasukawa. Monographs in Behavior and Ecology, Eds. J. R. Krebs and T. Clutton-Brock,
Princeton Univ. Press, Princeton, New Jersey. 1995: xviii + 312 pp., 61 numbered text figures
and 32 tables. $55 (cloth); $29.95 (paper).— The authors note that there have been about a
thousand papers published on Red-winged Blackbirds {Agelaius phoeniceus). Less than half
of these are cited in this monograph because it is limited to the ecology and evolution of
polygyny in red-wings and the import of this mating system to the process of sexual selection.
The book is no mere synthesis, although the reader is provided with a firm foundation about
the behavioral ecology of the species. There is a certain satisfaction in exploring the biology
of a species about which a great deal is known, since relevant data are available that are
germane to a wide array of questions. Searcy and Yasukawa have used this wealth of published
as well as unpublished data to develop arguments related to both proximate and ultimate
causes for polygyny and past and present sexual selection. Their analy.ses of hypotheses are
carefully accomplished and fully acknowledge confounding factors, pittfalls, and alternate
explanations. Indeed their meticulous, comprehensive approach sometimes caused me to lose
sight of where they were going, but fortunately separate discussions within the text of chapters
are brought to closure with a summary of their conclusions (or lack of conclusions). This
book deserves a longer and more detailed review. I highly recommend that such a review
might be the focus of a senior topics course or graduate seminar since there is much to learn,
not just about the phenomena discussed, but more broadly, about the support of assumptions
and the testing of hypotheses. — John L. Zimmerman.

Bird song; biological themes and variations. By C. K. Catchpole and P. J. B. Slater,
illus by N. Mann. Cambridge Univ. Press, Cambridge. 1995; 248 pp., pen and ink illustra-
tions, 77 figures and 3 tables. $32.95 (cloth).— Bird song arguably is the most intensively
studied area of animal communication and has been fertile ground for researchers m behav-
ioral ecology, ethology, psychology, and neuroscience. As such, the breadth and depth of

ORNITHOLOGICAL LITERATURE

601

the literature can prove intimidating to those new to it. Until recently, someone approaching
the field laced a vast body ot primary literature scattered in a diversity of journals, a handful
of specialized review articles, and a few technical edited volumes. Now, however, two of
the field s most prominent researchers have written an approachable and comprehensive
introduction to the study of bird song.

This book will appeal to graduate students, professionals in the behavioral sciences, and
amateur ornithologists alike. The book stands alone, requiring a minimum of background
knowledge. However, most of the topics are treated in sufficient depth to serve as a useful
academic reference. Catchpole and Slater have not attempted to describe all the available
research on a given topic; instead, they select current and representative examples to illus-
trate their main points. Thus, each section of the book can provide a solid basis for a more
intense literature survey of that particular topic.

The production quality of the book is excellent. Crisp pen and ink illustrations introduce
each chapter and enhance most figures. The figures are clear and pertinent. The authors also
use a standard (author year) citation format — a feature commonly lacking in books aimed

at a more general audience. The modest price of this book belies the excellent production
Job.

In the Tinbergen tradition, the book is organized around questions of mechanism, ontog-
eny, function, and evolutionary history. Chapter 1 introduces basic nomenclature and con-
cepts in animal communication and bird song in particular. Chapter 2 provides an overview
of the physiological and neural mechanisms of song production. This chapter provides a
broad, basic introduction to the neuroethology of bird song. Chapter 3 describes the phe-
nomenon of song learning. Again, the chapter is a solid basic introduction to the area.
Chapter 4 rounds out the discussion of proximate causation in bird song in reviewing how
sound transmission affects singing behavior. This part of the book provides excellent intro-
ductions, but many readers will want to follow up the primary literature.

In the remaining chapters of the book, the authors’ expertise truly comes to the fore in
examining functional causation of song behavior. It is here that bird song is most explicitly
considered a product of inter- and intra-sexual selection. Chapter 5 answers the question of
which birds sing and when do they do it. Chapter 6 focuses on male-male competition and
territory defense in the evolution of singing behavior. Chapter 7 examines mate choice and
the role of female song preferences in the evolution of song complexity. Chapter 8 is an
examination of the variety of songs and singing behaviors among species and describes
competing hypotheses for the evolution of such complexity. Chapter 9 addresses geographic
variation in bird songs and the evolution of dialects. This second section of the book pro-
vides a balanced view of the debates and controversies that have colored the field. The
section also provides an excellent summary of the data we have supporting the functions
of bird song: mate attraction and territory defense. On the whole, no other publication on
the topic has brought together such a wide body of literature.

In summary, “Bird Song” is the most complete and approachable review of the field to
date. Although many experts may find that their particular area has not been covered as
thoroughly as they would have liked, or their favorite species has not been used as an
example, they will not question the books breadth and utility. I plan to hand a copy to any
student interested in starting research in the area. This handsome and affordable volume

should complement the bookshelf of anyone interested in bird behaviour. Scott Mac-

Dougall-Shackleton.

NORTH AMERICAN BLUEBIRD SOCIETY
RESEARCH GRANTS— 1997

The North American Bluebird Society announces the 13th annual grants in aid for orni-
thological research directed toward cavity-nesting species of North America with emphasis
on the genus Sialia. Presently three grants of single or multiple awards are awarded and
include;

BLUEBIRD RESEARCH GRANT

Available to student, professional or individual researcher for a research project focused
on any of the three species of bluebird in the genus Sialia.

GENERAL RESEARCH GRANT

Available to student, professional or individual researcher for a research project focused
on any North American cavity-nesting species.

STUDENT RESEARCH GRANT

Available to full-time college or university students for a research project focused on any
North American cavity-nesting species.

Further guidelines and application materials are available upon request from:

Kevin L. Berner
Research Committee Chairman
College of Agriculture and Technology
State University of New York
Cobleskill, New York 12043

Completed applications must be received by December 1, 1996; funding decisions will be
announced by January 15, 1997.

1996 NABS RESEARCH AWARDS

The North American Bluebird Society is pleased to announce the results of its 12th annual
research grant’s program. The following individuals are recipients of the 1996 research awards:

BLUEBIRD GRANTS

Kristina M. Hannam, University of Miami. Title: Effects of Blowfly Ectoparasites on Eastern
Bluebird Reproductive Success.

STUDENT GRANTS

Karl E. Miller. University of Florida. Title: Nest-site Selection and Reproductive Success
of Secondary Cavity Nesting Birds in Thinned and Unthinned Slash Pine Forests in

Florida.

Paul Doherty, Ohio State University. Title; Metapopulation Dynamics ot a Permanent Res-
ident Forest-dwelling Bird Species Within a Fragmented Landscape; Empirical Data and
Dynamic Programming Models.

Elena V. Pravosudova, Ohio State University. Title: The Effect of Forest Fragmentation on
Social Structure of the Tufted Titmouse.

GENERAL GRANTS

Archibald McCallum, College of Charleston. Title; Reproductive Performance of Flam-
mulated Owls in the Jemez Mountains, New Mexico.

602

603

GRADUATE AND POST-GRADUATE RESEARCH GRANTS

^ The Biological Research Station of the Edmund Niles Huyck Preserve offers grants (max.
- $2,500) to support biological research which utilizes the resources of the Preserve. Among
the research areas supported are basic and applied ecology, animal behavior, systematics,
evolution, and conservation. The 2000 acre Preserve is located on the Helderberg Plateau,
30 miles southwest of Albany. Habitats include northeast hardwood-hemlock forests, conifer
plantations, old fields, permanent and intermittent streams, 10 and 100 acre lakes and several
waterfalls. Facilities include a wet and dry lab, library, and houses/cabins for researchers.
Deadline - February 1, 1997. Application material may be obtained from Dr. Richard L.
Wyman, Executive Director, EN Huyck Preserve and Biological Research Station, PO. Box
189, Rensselaerville, NY 12147.

THE ATLAS OF SOUTHERN AFRICAL BIRDS

Preparation of the manuscript for The Atlas of Southern African Birds is nearing comple-
tion. The Southern African Bird Atlas Project covers Botswana, Lesotho, Namibia, South
Africa, Swaziland and Zimbabwe. Based on seven million distribution records, this is the
largest biodiversity project m Africa. The 1600-page, two- volume atlas contains distribution
maps and texts for 700 species; for many, the ranges are strikingly different from those
shown in current fieldguides and handbooks. 200 vagrants are also covered. To receive
publication information, write to the Avian Demography Unit, University of Cape Town,
Rondebosch, 7700 South Africa, email adu@maths.uct.ac.za or access the Avian Demog-
raphy Unit’s pages at http;//www.uct.ac.za/depts/stats/adu/

ERRATUM

The color frontispiece of Chlorostilbon olivaresi in the March 1996 issue of The Wilson
Bulletin was painted by Eugenia Brieva, staff artist of the Instituto de Ciencias Naturales
Umversidad Nacional de Colombia.

INFORMATION FOR AUTHORS

The Wilson Bulletin publishes significant research and review articles in the field of
ornithology. Mss are accepted for review with the understanding that the same or similar
work has not been and will not be published nor is presently submitted elsewhere, that all
persons listed as authors have given their approval for submission of the ms, and that any
person cited as a personal communication has approved such citation. All mss should be
submitted directly to the Editor.

Text. Manuscripts should be prepared carefully in the format of this issue of The Wilson

Bulletin. Mss will be returned without review if they are not properly prepared. They
should be neatly typed, double-spaced throughout (including tables, figure legends, and
“Literature cited”), with at least 3 cm margins all around, and on one side of good quality
8.5" X 1 1" paper. Do not use erasable bond. Mss typed on low-quality dot-matrix printers
are not acceptable. The ms should include a cover sheet (unnumbered) with the following;

( 1 ) Title, (2) Authors, their institutions, and addresses, (3) Name, address, and phone number
of author to receive proof, (4) A brief title for use as a running head. All pages of the text
through the “Literature cited” should be numbered, and the name of the author should
appear in the upper right-hand corner of each. The text should begin in the middle of the
first numbered page. Three copies should be submitted. Xerographic copies are acceptable
if they are clearly readable and on good quality paper.

Tables. Tables are expensive to print and should be prepared only if they are necessary.

Do not repeat material in the text in tables. Tables should be narrow and deep rather than
wide and shallow. Double space all entries in tables, including titles. Do not use vertical
rules. Use tables in a recent issue of the Bulletin as examples of style and format. Tables
should be typed on separate unnumbered pages and placed at the end of the ms.

Illustrations must be readable (particularly lettering) when reduced in size.
Final size will usually be 11.4 cm wide. Illustrations larger than 22 X 28 cm will not be
accepted, and should be reduced photographically before submission. Legends for all figures
should be typed on a separate page. Photographs should be clear, of good contrast, and on
glossy paper. Drawings should be in India ink on good drawing board, drafting paper, or
blue-lined graph paper. Figures produced by high-quality laser printers may be acceptable.
All lettering should be done with a lettering instrument or printer. Do not use typewriter
lettering and do not submit figures having adhesive transfer letters affixed to them. Designate
the top of each illustration and label (on the back in soft pencil) with author s name, ms
title, and figure number. Submit two duplicates or readable xerographic copies of each figure
as well as the original or high-contrast glossy photo of the original.

Authors of accepted papers are urged to submit voucher photographs of their work to
Visual Resources for Ornithology (VIREO) at the Academy of Natural Sciences of Phila-
delphia. Accession numbers from VIREO will then be published within appropriate sections
of the paper to facilitate access to the photographs in subsequent years.

Style and format.— The current issue of The Wilson Bulletin should be used as a guide
for preparing your ms; all mss must be submitted in that format. For general matters of
.style authors should consult the “CBE Style Manual,” 5th ed.. Council of Biology Editors,
Inc., Bethesda, MD, 1983. Do not use footnotes or more than two levels of subject sub-
headings. Except in rare circumstances, major papers should be preceded by an abstract, not
to exceed 5% of the length of the ms. Abstracts should be informative rather than indicative,
and should be capable of standing by themselves. Most units should be metric, and com-
pound units should be in one-line form (i.e., cm-sec“‘). The continental system of dating
(19 Jan. 1950) and the 24 hour clock (09:00, 22:00) should be used, and the Standard Time
Specified (e.g., ESX for Eastern Standard Time) at first reterence.

References. — In both major papers and general notes, if more than four references are
cited, they should be included in a terminal “Literature cited” section. Include only refer-

604

INFORMATION FOR AUTHORS

605

ences cited in the ms, and only material available in the open literature. (“In-house” reports
and the like should not be cited.) Use recent issues of the Bulletin for style, and the most
recent issue of “BIOSIS,” BioScience Information Service, Philadelphia, PA, for abbrevi-
ations of periodical names. If in doubt, do not abbreviate serial names. Manuscripts with
fewer than five references should be cited internally, e.g., (Sprenkle and Blem, Wilson Bull.
96:184-195) or Sprenkle and Blem (Wilson Bull. 96:184-195).

Nomenclature.— Common names and technical names of birds should be those given in
the 1983 A.O.U. Check-list (and supplements as may appear) unless justification is given.
For other species the Bulletin uses the common names in Sibley and Monroe, “Distribution
and Taxonomy of Birds of the World.” Common names of birds should be capitalized. The
scientific name should be given at first mention of a species both in the abstract and in the

text.

The editor welcomes queries concerning style and format during your preparation of mss
for submission to the Bulletin. — Charles R. Blem, Editor.

This issue of The Wilson Bulletin was published on 4 October 1996.

The Wilson Bulletin

Editor Charles R. Blem

Editorial Board Kathy G. Beal

Department of Biology
Virginia Commonwealth University
816 Park Avenue
Richmond, Virginia 23284-2012

Richard N. Conner
Thomas M. Haggerty
John A. Smallwood

Review Editor WiLl.lAM E. Davis, Jr.

127 East Street

Eoxboro, Massachusetts 02035

Assistant Editors Leann Blem

Albert E. Conway

Index Editor Kathy G. Beal

616 Xenia Avenue
Yellow Springs, Ohio 45387

Suggestions to Authors

See Wilson Bulletin, 108:395—396, 1995 for more detailed “Information for Authors.”
Manuscripts intended for publication in The Wilson Bulletin, should be submitted in triplicate,
neatly typewritten, double-spaced, with at least 3 cm margins, and on one side only of good
quality white paper. Do not submit xerographic copies that are made on slick, heavy paper. Tables
should be typed on separate sheets, and should be narrow and deep rather than wide and shallow.
Follow' the AOU Check-list (Sixth Edition, 1983) insofar as scientific names of U.S., Canadian,
Mexican, Central American, and West Indian birds are concerned. Abstracts of major papers
should be brief but quotable. In both Major Papers and Short Communications, where fewer than
5 papers are cited, the citations may be included in the text. Follow carefully the style used in
this issue in listing the literature cited; otherwise, follow the “CBE Style Manual” (AIBS, 1983).
Photographs for illustrations should have good contrast and be on glossy paper. Submit prints
unmounted and attach to each a brief but adequate legend. Do not write heavily on the backs of
photographs. Diagrams and line drawings should be in black ink and their lettering large enough
to permit reduction. Original figures or photographs submitted must be smaller than 22 X 28 cm.
Alterations in copy after the type has been set must be charged to the author.

Notice of Change of Address

If your address changes, notify the Society immediately. Send your complete new address to
Ornithological Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

The permanent mailing address of the Wilson Ornithological Society is: c/o The Museum of
Zoology. The University of Michigan, Ann Arbor, Michigan 48109. Persons having business with
any of the officers may address them at their various addresses given on the back of the front
cover, and all matters pertaining to the Bulletin should be sent directly to the Editor.

Member.ship Inquirie-S

Membership inquiries should be sent to Dr. John Smallwood. Dept, of Biology, Montclair State
Univ., Upper Montclair, New Jersey 07043.

CONTENTS

MAJOR PAPERS

A NEW GENUS AND SPECIES OF FURNARIID (AVES: FURNARIIDAE) FROM THE COCOA-GROWING REGION

OF SOUTHEASTERN BAHIA, BRAZIL -

Jose Fernando Pacheco, Bret M. Whitney, and Luiz Gonzaga

THE NEST AND NESTING ECOLOGY OF ACROBATORNIS FONSECAl (FURNARIIDAE), WITH IMPLICATIONS
FOR INTRAFAMILIAL RELATIONSHIPS

Bret M. Whitney, Jose Fernando Pacheco, Paulo Sergio Moreira da Fonseca, and

Robert H. Barth, Jr.

WOODPECKER EXCAVATION AND USE OF CAVITIES IN POLYSTYRENE SNAGS

Richard N. Conner and Daniel Saenz

NESTING SUCCESS OF THE PROTHONOTARY WARBLER IN THE UPPER MISSISSIPPI RIVER BOTTOMLANDS

David J. Flaspohler

FACTORS AFFECTING FOOD PROVISIONING OF NESTLING BLACK-THROATED BLUE WARBLERS —

Catherine O’Neill Goodbred and Richard T. Holmes

BREEDING BIOLOGY AND NATURAL HISTORY OF THE BAHAMA SWALLOW Paul E. Allen

NEOTROPICAL MIGRATORY BREEDING BIRD COMMUNITIES IN RIPARIAN FORESTS OF DIFFERENT WIDTHS

ALONG THE ALTAMAHA RIVER, GEORGIA -

Malcolm F. Hodges, Jr. and David G. Krementz

DAWN AND DUSK SINGING OF MALE AMERICAN ROBINS IN RELATION TO FEMALE BEHAVIOR —

Tore Slagsvold

BREEDING BIOLOGY OF THE CRESTED CARACARA IN SOUTH TEXAS

Vanessa M. Dickinson and Keith A. Arnold

BREEDING BIOLOGY OF THE JABIRU IN THE SOUTHERN LLANOS OF VENEZUELA

Jose A. Gonzalez

EFFECT OF EGG SIZE ON PREDATION BY WHITE-FOOTED MICE .... R. M. DeGraaf and T. J. Maier

CAN CHECKLIST PROGRAMS BE USED TO MONITOR POPULATIONS OF BIRDS RECORDED DURING THE
MIGRATION SEASON? Erica H. Dunn, Jacques Larivee, and Andre Cyr

EFFECT OF MATE REMOVAL ON SINGING BEHAVIOR AND MOVEMENT PATTERNS OF FEMALE NORTHERN

David B. McElrov and Gary Ritchison

CARDINALS

RADIO TELEMETRY DOCUMENTS 24-HOUR FEEDING ACTIVITY OF WINTERING LESSER SCAUP

Christine M. Custer, Thomas W. Custer, and Daniel W. Sparks

BODY MASS AND CARCASS COMPOSITION OF FALL MIGRANT OLDSQUAWS

James O. Leafloor, John E. Thompson, and C. Davison Ankney

AVIAN NEST-SITE SELECTION AND NESTING SUCCESS IN TWO FLORIDA CITRUS GROVES

Mary Crowe Mitchell, Louis B. Best, and James P. Gionfriddo

SHORT COMMUNICATIONS

EXPONENTIAL POPULATION GROWTH OF MONK PARAKEETS IN THE UNITED STATES

Sunshine Van Bael and Stephen Pruett-Jones

FOREST GAP USE BY BREEDING BLACK-THROATED GREEN WARBLERS

Robert Smith and Matthew Dallman

COURTSHIP BEHAVIOR OF GOLDEN-CHEEKED WARBLERS Mark W. LockwOod

ORNITHOLOGICAL LITERATURE

397

434

449

457

467

480

496

507

5)6

524

535

540

550

556

561

512

58^

58S

591

59;

The Wilson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. 108, NO. 4 DECEMBER 1996 PAGES 607-848

(ISSN 0043-5643)

The Wilson Oknithological Society
Founded December 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President— Keith L. Bildstein, Hawk Mountain Sanctuary, RR 2, Box 191, Kempton, Pennsylvania
19529-9449.

First Vice-President — Edward H. Burtt, Jr., Department of Biology, Ohio Wesleyan University,
Delaware, Ohio 43015.

Second Vice-President- — John C. Kricher, Biology Department, Wheaton College, Norton, Mas-
sachusetts 02766.

Eflitor — Charles R. Blem, Department of Biology, Virginia Commonwealth University, P.O. Box
842012, Richmond, Virginia 23284-2012.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper Mont-
clair, New Jersey 07043.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana
46556.

Elected Council Members — Carol A. Corbat and William E. Davis (terms expire 1997), and
Margaret C. Brittingham and Herbert T. Hendrickson (terms expire 1998), Peter C. Frederick
and Danny J. Ingold (terms expire 1999).

Membership dues per calendar year are: Active, $21.00; Student, $15.00; Family, $25.00; Sus-
taining, $30.00; Life memberships $500 (payable in four installments).

The Wll.soN Buli.ETIN is sent to all members not in arrears for dues.

The Jossei.yn Van Tyne Memoriai. Library

The Josselyn Van Tyne Memorial Library of the Wilson Ornithological Society, housed in the
University of Michigan Museum of Zoology, was established in concurrence with the University
of Michigan in 1930. Until 1947 the Library was maintained entirely by gifts and bequests of
books, reprints, and ornithological magazines from members and friends of the Society. Two mem-
bers have generously established a fund for the purchase of new books; members and friends are
invited to maintain the fund by regular contribution, thus making available to all Society members
the more important new books on ornithology and related subjects. The fund will be administered
by the Library Committee, which will be happy to receive suggestions on the choice of new books
to be added to the Library. William A. Limk, University Museums, University of Michigan, is
Chairman of the Committee. The Library currently receives 195 periodicals as gifts and in ex-
change for The Wilson Bulletin. With the usual exception of rare books, any item in the Library
may be borrowed by members of the Society and will be sent prepaid (by the University of
Michigan) to any address in the United States, its possessions, or Canada. Return postage is paid
by the borrower. Iiupiiries and requests by borrowers, as well as gifts of books, pamphlets, reprints,
and magazines, should be addressed to: The Josselyn Van Tyne Memorial Library, University of
Michigan Museum of Zoology, Ann Arbor, Michigan 48109. Contributions to the New Book Fund
should be sent to the Treasurer (small sums in stamps are acceptable).

The Wilson Bulletin
(ISSN 0043-5643)

the WILSON BULLETIN (ISSN OOTI-.SM.'i) is [Iiihlished (|uartcrly in March, June. Septenilx-r, and Deceniher hy the Wilson
Ornilholopieal Society. 810 East lOth Street. Lawrence. KS 600+t-8897. The suhscription price, holh iti the United States and
elsewhi-re. is $-K).(K) per year. Periodicals postage paid at Lawrence. KS. POS'I’M ASTER; Sen<l address ehaiiftes to THE WILSON
BULLETIN. P.O. Box 1897. Liwrenee. KS OOOI-t-8897.

Back issues or sinule copies are available for SI 2.00 each. Most hack issues of the HuUetin are availahle and may he ordered
from ifte Treasurer. Special prices will he ipioteil for ipiantity orders.

All articles and eommunieutions for piihlieations. hooks and puhlications for reviews should he addressed to the Editor. Ex-
chanpes should he addressed to The Josselyn Van Tyne Memorial l.ihrary. Museum of Zoolopy. Ann Arhor. Miehipan -18109.

Suhscriplions. ehanpes of adilress and ejaims for undelivered copies sh(aild he sent to the OSNA, P.O. Box 189,. Isiwrenee.
KS (.(,01-1-8897. Phone: (91.1) 81.1-1221: EAX; (91.1) 8-1.1- 1 27-1.

© Copyright 1996 by the Wikson Ornithological Society
Printed by Allen Press, Inc., Lawrence, Kansas 66044, U.S.A.

@ This paper meets the requirements of ANSi/NISO Z39.48-1992 (Permanence of Paper).

Frontispiece. The frontispiece painting by Patrick Ching has been made possible by an
endowment established by George Miksch Sutton.

THE WILSON BULLETIN

A QUARTERLY MAGAZINE OF ORNITHOLOGY
Published by the Wilson Ornithological Society

VoL. 108, No. 4 December 1996 Pages 607-848

Wilson Bull., 108(4), 1996, pp. 607-619

DESCRIPTION OF ADULTS, EGGSHELLS, NESTLING,
FLEDGLING, AND NEST OF THE POO-ULI

Andrew Engilis, Jr.,‘ Thane K. Pratt,^

Cameron B. Kepler,^ A. Marie Ecton,"^ and
Kimberly M. Fluetsch^

Abstract. — The Poo-uli (Melamprosops phaeosoma), a Hawaiian honeycreeper discov-
ered on the island of Maui in 1973 and now near extinction, is represented in museums by
only two specimens. Based on the first observations of a nesting pair and re-examination
of the two specimens, we describe the adult male and female, eggshells, nestling, and
fledgling Poo-uli. Poo-uli are sexually monochromatic but males are brighter. The male is
brown above, whitish below, and has an extensive black mask bordered with gray on the
crown and a distinct white auricular patch. The female differs in having a similar facial
pattern not as sharply demarked and in having a grayish wash below. The observed fledgling
resembled the adults but was paler brown above and whitish below and had a much smaller
black mask and pale mandible. We tentatively assigned both museum specimens to first
basic plumage because they resembled the adult female but retained some pale Juvenal
coloration in the mandible. We also determined from dissection that the holotype was an
immature male; we could not determine sex of the paratype. The nest was an open cup of
twigs and bryophytes with a thin lining of fern rootlets. The nest contained eggshell frag-
ments with brown-gray speckling against a whitish background. The nests, eggshells, and
nestlings resemble those of other Hawaiian honey creepers. Received 1 Dec. 1995, accepted
27 May 1996.

The Poo-uli {Melamprosops phaeosoma), discovered on the island of
Maui in 1973, is the most recent, and presumably last, described extant

' Bernice P. Bishop Museum, P.O. Box 19000- A, Honolulu, Hawaii, 96817. Present address: Ducks Un-
limited, Inc., 3074 Gold Canal Dr., Rancho Cordova, California 95670.

^National Biological Service, Hawaii Field Station, P.O. Box 44, Hawaii National Park, Hawaii 96718.

^ NBS, P.O. Box 44, Hawaii National Park, Hawaii 96718. Present address: NBS, Southeast Research
Station, Wamell Sch. Forest Res. — Univ. of Georgia, Athens, Georgia 30602-2152.

NBS, P.O. Box 44, Hawaii National Park, Hawaii 96718. Present address: Biology Dept., Univ. of
Miami, P.O. Box 24918, Coral Gables, Florida 33124-0421.

’NBS, P.O. Box 44, Hawaii National Park, Hawaii 96718. Present address: 6018 Royal Creek, San
Antonio, Texas 78239.

607

608

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

genus and species of Hawaiian bird (Casey and Jacobi 1974). Despite
doubts about its systematic affinities (Pratt 1992), initial genetic compar-
isons (C. Tarr and R. Fleischer, pers. comm.) suggest placement of the
Poo-uli within the Hawaiian honeycreepers (Fringillidae: Drepanidini).
Several studies have investigated aspects of the morphology, life history,
and conservation status of this endangered bird (Bock 1978, Baldwin and
Casey 1983, Scott et al. 1986, Engilis 1990, Mountainspring et al. 1990,
Kepler et al. 1996).

The original description of the Poo-uli was based on two specimens of
unknown age, identified as “males (?);” no others have been collected.
Subsequent field research has not investigated age and sex differences in
plumage and soft parts. No nests had been found until recently. In 1986,
Kepler et al. (1996) studied two sequential nesting attempts by a pair of
Poo-uli in the Hanawi Natural Area Reserve, Maui. This provided an
opportunity to describe the adults, nestlings, fledgling, and nests observed.
We compare these descriptions with the two museum specimens, with
other Hawaiian honeycreepers and, when appropriate, with cardueline
finches, the group from which the Hawaiian honeycreepers are thought
to have evolved (James and Olson 1991).

METHODS

Details on the study site and nesting events are given in Kepler et al. (1996), and a map
of the study area is provided in Moutainspring et al. (1990) and of the nest site in Engilis
(1990). Our descriptions of Poo-uli are based on (1) repeated field observations by Engilis
and Kepler with color names quoted from written notes and (2) sketches by Engilis and
artist Patrick Ching. We observed birds at the nests from distances of 40 m (nest #1) and
18 m (nest #2) through binoculars (Leitz lOX), spotting telescope (Bushnell Spacemaster
20-60X, nest #1) or Questar telescope (SOX, nest #2) (Kepler et al. 1996). Though events
at nest #2 were photographed, the pictures taken were only marginally useful for plumage
description. Both nests were collected and measured on 16 June 1986, treated by Berlese
extraction on 16-18 June, and deposited at Bernice P. Bishop Museum, Hawaii, with the
catalog numbers of BPBM 162151 for nest #1 and its eggshell fragments and BPBM 162152
for nest #2. We examined the nests and holotype specimen (BBM-X-1471 12) at Bishop
Museum; M. LeCroy examined the paratype (AMNH 810456) at The American Museum
of Natural History (AMNH), New York. We also examined two enlarged black-and-white
photographs comparing both specimens soon after preparation and held by the AMNH Bird
Dept. Library. The photographs lacked identifying numbers.

RESULTS

Adult Male

Of the two adults tending the nest, we assume that the brighter-colored
bird was the male parent because it sang and courted the drab bird, fed
the drab bird and chicks, but did not incubate or brood. Also, for all other

Engilis et al. • PLUMAGES AND NEST OE THE POO-ULI

609

Hawaiian honeycreepers, male plumage is either brighter than or similar
to female plumage (Freed et al. 1987).

Face. — Face with a distinctive black mask. Mask triangular, crisply
bordered, extending from the forehead and chin, around the eye, to a
point beyond the eye and bounded above by a gray crown and below by
a white auricular patch.

Upperparts. — Crown behind mask gray, merging on the nape to dark
brown. Back dark brown. Scapulars and wing coverts dark rufous brown.
Primaries and secondaries dark brown with blackish shafts, outer margin
of primaries buff. Rump and upper tail coverts rufous-brown. Tail dark
brown edged rufous, so short as to be mostly hidden by the folded wings,
not tapered, notched, the feathers lax and pointed.

Underparts. — Auricular patch distinct, creamy white continuing onto
the throat, bordered by gray of the upper flanks. Chin black. Throat white.
Breast white, washed with light gray. Belly white, merging with deep
cinnamon undertail coverts. Flanks, forward white washed with gray, pos-
teriorly becoming more grayish tinged buff, then cinnamon. Leg feath-
ering cinnamon. Undersides of the primaries silver-gray.

Bill glossy black, appearing bluish at a distance. Iris medium brown.
Legs and feet dark pink-brown; foot pads yellowish.

Adult Female

We assume that the drab bird at the nest was the female, because she
incubated and brooded but did not sing. She was similar to but duller
than the male, differing as follows. Black mask smaller, more grayish.
Pale auricular patch suffused with gray and less sharply bounded. Throat
white suffused with gray, but breast and sides to anterior belly gray.
Flanks more washed with gray, becoming golden cinnamon where they
met the primaries. Leg feathering gray. Undertail coverts buff-gray with
darker tips. Bill, iris and legs same as the male. Gape black, anterior roof
of palate pink. Neither adult showed signs of pox lesions or scars.

Nestlings

Three nestlings were observed, one in nest #1, two in nest #2. We saw
nestlings best on 20-22 May at nest #2 when the oldest chick was ca
10-12 days old. Only their heads could be seen, covered with medium
gray natal down; we could not see a black mask. The head of the smaller
(younger?) chick, when first seen on 21 May, appeared “mottled” black
and gray. For both chicks, the iris was blackish and the bill light gray
with bright yellow rim and red spot at the corner of the gape; inside the
mouth was reddish pink. On 29 May at ca 19 days-old and 2 days prior

610

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Fig. 1. Detail of Poo-uli chick on 29 May, ca 19 days of age and two days prior to
fledging. Note small mask. Tracing from field sketch by A. Engilis, Jr.

to fledging, bill with maxilla slate-gray and mandible underside whitish-
cream. Tongue pink.

Fledgling

Of the three chicks observed, one fledged. The fledgling resembled the
adults, but its mask was smaller and body plumage paler brown, with less
rufous and cinnamon. The small mask was bounded by the dorsal ridge
of the bill, distal corner of the eye, and the bill just below the gape, and
did not extend onto the chin (Fig. 1). The whitish auricular patch merged
ventrally with the darker throat, rather than being sharply outlined. (This
description was corroborated on 30 August 1994, when a juvenile Poo-
uli being fed by its parents was observed to have a mask “less distinct
than the adults, perhaps more gray in color” M. Reynolds and T Snet-
singer, pers. comm.).

Upperparts. — Crown and nape gray-brown, merging with brown back.
Back, scapulars, wing coverts medium brown tinged with buff. Rump
brown with cinnamon edging, wings and tail dark brown without rufous
edging. We did not record obvious wing bars (paler tips of median or
greater wing coverts or both), but these could have been indistinct and
not noticed or absent altogether.

Underparts. — Inter-ramal space black, chin whitish buff, with some
black feathering disconnected to the mask. Throat dingy white. Upper
breast gray washed with beige. Belly beige. Vent white. Under-tail coverts
cinnamon. Flanks like back, distally becoming cinnamon.

Bill — Maxilla slate gray, with pale tip. Mandible whitish, with dark
edge where the ramphotheca meets the skin. Iris blackish. Gape flange
smaller than nestlings’, bright yellow, with red spot distally at the corner.

Engilis et al. • PLUMAGES AND NEST OF THE POO-ULI 61 1

Inside of mouth reddish, with black patch on the roof of the mouth. Legs
and feet similar to adults.

Comparison of Observed Birds with Museum Specimens

The holotype and paratype were virtually identical in appearance. Cas-
ey and Jacobi (1974) described the paratype as differing from the holotype
in (1) its slightly smaller size, (2) somewhat larger mask “slightly mixed
and stippled with buff, especially on chin,” (3) “upperparts throughout,
including crown, duller grayish brown,” with less cinnamon, and (4) man-
dible tipped “only slightly lighter in color, washed with gray rather than
shell pink.” Photos of the fresh specimens, held by AMNH (one pub-
lished in Casey and Jacobi 1974), show the paratype’s mandible tip as
faintly pale and leg color as darker than that of the holotype. LeCroy
(pers. comm.) found that the stippling in the mask was caused by pale
feathers intermixed with dark ones and that the mandible is now dark
throughout.

The adults and fledgling we observed resembled in most respects both
the holotype and paratype. In both male and female, the bill was dark
throughout in comparison with the pale-tipped mandible of both speci-
mens. The adult male differed most, in its facial pattern of greater contrast
and underparts paler, as described above. The adult female differed from
the specimens perhaps in having the wings and tail with more cinnamon
wash than the paratype (but not the holotype?). The fledgling differed
from the specimens in its smaller mask, pale buffy breast, and slate-gray
mandible, which in the two specimens was dark with a pale tip. Exami-
nation of the two specimens for molt revealed that (1) the' holotype either
was just completing body molt, with a few new feathers growing on the
crown and mask, or was not molting and instead was replacing feathers
lost through wear and (2) the paratype showed no evidence of molt and
appeared to be in fresh plumage. For the holotype, the distal-most greater
wing covert feather on the right wing was buff rather than brown. Flight
and tail feathers on both specimens showed little wear.

We add here to the description of the Poo-uli specimens (Casey and
Jacobi 1974). We attempted with difficulty to determine the number of
flight feathers. There appeared to be nine primaries and nine secondaries
for both specimens, and a total of 12 rectrices for the holotype, as in other
Hawaiian honeycreepers. We examined flight and tail feathers of both the
holotype and paratype for evidence of feather lice, which occasionally
are found on other drepanidines, but saw none. Some of the black feathers
in the mask of the holotype had brownish centers; the paratype was not
examined for this character. The holotype weighed 25.5 g (T. Casey, pers.
comm.).

612

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

We examined the previously dissected, pickled carcasses of the two
specimens. We discovered that the holotype was an immature male by
presence of the right testis measuring <1 mm, together with the epidy-
dimis leading posteriorly. Sex of the paratype could not be determined
by inspection.

Nests

Both nests were open cups and built among the stems of leaf-bearing
ohia-lehua {Metrosideros polyrnorpha Gaud.) branchlets (Frontispiece).
Distance from the nest base (minus nest tail) to the juncture of the sup-
porting stems measured 60 and 70 mm for nests #1 and #2, respectively.
Leafy portions of the stems framed the nests on at least one side and
covered the cup of the nests; however, the height and extent of leafy
cover was not measured, and cover was cut away from the nests upon
collection. For both nests, the branch which joined the supporting stems
measured 14.5 mm in diameter below the juncture. Supporting stems in-
corporated into the frame of the nests were <10 mm diameter and num-
bered 7 and 6 for nests #1 and #2, respectively.

Dimensions (mm) upon collection for nests #1 and #2 were: (1) outer
diameter, 180 by 130 and 180 by 140; (2) outer depth 90 and 110; (3)
inner diameter, 70 by 60 and 85 by 60; and (4) inner depth 50 and 40.
The body of both nests was constructed of bare twigs of pukiawe (Sty-
phelia tameameae [Cham. & Schlechtend.] F. v. Muell.) with coarse moss-
es filling the spaces between the twigs. Mosses identified from both nests
were Homaliodendron flabellatum (Sm.) Fleisch., Thuidium plicatum
Mitt., Trachypodopsis auriculata (Mitt.) Fleisch., with nest #1 containing
Aerobryopsis wallicia (Dozy & Molk.) Fleisch. and nest #2 Floribimdaria
floribunda (Dozy & Molk.) Fleisch. Leaves and stems of graminoids and
dicots accounted for <5% of this filling. For the inner 15 mm of nest
wall, fern rootlets < 1 mm thick replaced pukiawe twigs as the structural
frame, with the amount of moss decreasing toward the interior. To the
internal surface of this lining was added graminoid fiber, perhaps Uncinia
uncinata (L. Fil.) Kukenth., 1 mm in thickness. The resulting lining was
an open network of fiber. For nest #1, which had been abandoned by the
parents three months prior to collection, mosses of the nest body had
expanded into the cup, and there were fewer fern rootlets and graminoid
fibers. Whether the lack of rootlets and fiber in the first nest is due to
differences in original construction, or removal of those materials by the
parents or by other birds, is not known. Neither nest contained deposits
of fecal matter, supporting the observation that parents removed all feces
(Kepler et al. 1996). The nests did not smell of “drepanidine odor” (Pratt
1992) upon collection nor when examined nine years later.

Engilis et al. • PLUMAGES AND NEST OF THE POO-ULI

613

Nest #1 contained eggshell fragments, the largest of which was 9 by 7
mm and, judging from its curvature, may have been from the blunt end
of the egg. Though the fragments appeared weathered, they still showed
fine, dense, brown-gray speckling against a whitish background.

While the Berlese extraction yielded large numbers of arthropods, no
ectoparasitic insects turned up (S. Swift, pers. comm.). Seven diptera lar-
vae and one adult were extracted from nest #1. Most of the mites asso-
ciated with the nest remain unidentified.

DISCUSSION

Plumage. — Can the age and sex of the two museum specimens now
be determined? Given differences in plumage and soft-part colors ob-
served among the sexes and fledgling at the nests, we believe that both
the holotype and paratype were hatch-year birds in first basic (post-ju-
venal) plumage. Dissection confirmed that the holotype was an immature
male. Both specimens match most closely the putative adult female, with
black mask larger and breast grayer than the fledgling. Both specimens
differ from the adults observed at the nest, but resembles the fledgling,
in having the mandible not all dark. Both specimens least resemble the
putative adult male. We think it very unlikely that either could have been
an adult male, especially because adult male drepanidines show little vari-
ation within species (Jeffrey et al. 1993, Pratt et al. 1994).

Color of the mandible is important in our determination of the age of
the specimens. Unfortunately, the mandible color of the paratype now
does not match that in the description and photos of Casey and Jacobi
(1974). They described the mandible of the freshly collected paratype as
dark with the tip “only slightly lighter in color.” The mandible is now
dark throughout. The photos show the original description to be correct
and that the change occurred post mortem. The slightly darker legs of the
paratype in comparison with the holotype evident in Casey and Jacobi
(1974: fig. 1) may indicate darkening of the leg color post fledging. How-
ever, leg color in life or shortly after death was not noted as different
between the two specimens (Casey and Jacobi 1974), and may have either
been overlooked, or changed post mortem, or be an artifact of photog-
raphy. Legs and feet of the observed fledgling were recorded as being
similar to those of the adults. This issue can be resolved by future field
work.

The specimens fit a presumed plumage sequence for Poo-uli as follows.
During the first prebasic molt, Juvenal plumage in either sex gives way
within a few months of fledging to a female-like first basic plumage, as
in many drepanidines. Meanwhile, the mandible color changes from pale
to black, with the last vestige of Juvenal coloration being the pale tip.

614

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

which ultimately darkens. Other drepanidines show a similar change in
bill color from light to dark, e.g., Palila {Loxioides bailleui; Jeffrey et al.
1993) and Akiapolaau {Hemignathus munrov, Pratt et al. 1994). Thus,
both specimens show a larger mask than the observed fledgling but ap-
parently still retained the pale-tipped mandible in life. The holotype shows
other, less certain signs of first prebasic plumage: a few facial feathers in
sheath indicating last stage of molt and a pale brown feather in the greater
wing coverts possibly retained from Juvenal plumage, as with other dre-
panidines (Fancy et al. 1993; Jeffrey et al. 1993; Pratt et al. 1994; Lind-
sey, unpubl. data). For the paratype, the mask shows some pale, perhaps
old Juvenal, feathers. First prebasic molt is likely to occur in the summer
and fall, following breeding, as with other well-studied drepanidines (Jef-
frey et al. 1993, Ralph and Fancy 1994). Evidence for seasonality in molt
for Poo-uli is lacking. However, the two nests we observed would have
or did fledge young in April and June. We add here the above-mentioned
observation by M. Reynolds and T. Snetsinger (pers. comm.) of a Poo-
uli Juvenile on 30 August 1994. The holotype and paratype could have
completed prebasic molt prior to being collected in September 1973.

Freed et al. (1987) tentatively characterized sexual chromatism for Poo-
uli as monochromatic. We can now modify that to monochromatic with
males brighter. In most drepanidines in which the plumage is monochro-
matic or monochromatic with males brighter, female and first basic plum-
ages approach the bright adult male plumage, whereas Juvenal plumages
remain distinctly cryptic. Further, in many monochromatic species with
males brighter, females show variability, with some females more similar
to males than others, e.g., Palila (Jeffrey et al. 1993) and Maui Alauahio
(Paroreomyza montana', H. and P. Baker, pers. comm.). While Poo-uli are
arguably cryptically colored, their most distinctive plumage feature — the
black mask with black bill in the center, highlighted behind by the gray
crown and white auricular patch — may be a strong intra- and interspecific
optical signal (sensu Hailman 1977). Viewed head on, Poo-uli provide a
striking, unmistakable, and unforgettable image (Fig. 2).

Many drepanidines show a dark loral patch. This patch may have
evolved to form an extensive mask in the Poo-uli. Two other, allopatric
drepanidines, the Akekee (Loxops caeruleirostris) and Hawaii Creeper
{Oreomystis mana) have independently evolved smaller masks. In all dre-
panidines with dark lores, including Poo-uli, Juveniles show either gray
or whitish lores, affirming the importance of dark lores as an adult social
signal. Thus, we argue that the mask of the Poo-uli is not a radical de-
parture from the many drepanidine plumage patterns, but instead is note-
worthy mainly in degree. If the mask is indeed a strong optical signal,
then compared with other monochromatic drepanidines with brighter male

Engilis et al. • PLUMAGES AND NEST OF THE POO-ULI

615

Fig. 2. Poo-uli in first basic or adult female plumage showing bold black mask. Pho-
tograph by A. Engilis, Jr.

plumage, Poo-uli also show a convergence of female and first basic plum-
ages with bright male plumage. The difference from other drepanidines
is the expression of this character, albeit smaller and grayer, in juvenal
plumage as well. Besides the black mask, a likely optical signal and
important, ubiquitous field-mark is the cinnamon rump seen as the bird
flies away. We encourage other field observers to determine the extent of
variation in adult, juvenal, and first basic plumages.

While questioning the systematic position of the Poo-uli, Pratt (1992)
claimed “The colors and pattern of the Poo-uli are unlike that of any
previously known Hawaiian honeycreeper,” and “Thus, plumage color
and pattern provide no basis for inclusion of Melamprosops phaeosoma
among the Drepanidinae.” His mis-statement that adult coloration of Poo-
uli was gray and white was based on verbal pers. comm, from one of us
that we did not have an opportunity to catch in review. Above, we inter-
pret the Poo-uli’s black mask as homologous with the black lores or mask
of other drepanidines. Brown coloration in immature or adult plumages
is shared by five other historical Hawaiian honeycreepers (Akepa, Loxops
coccineus; Apapane, Himatione sanguinea; Greater Koa-Finch, Rhoda-

616

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

canthis palmerv, Kakawahie, Paroreomyza flammea\ Ula-ai-hawane, Cir-
idops anna).

Why are Poo-uli brown, an unusual color for a Hawaiian honeycreeper?
Mountainspring et al. (1990) document substrate-restricted foraging by
Poo-uli on branches where they glean or forcibly excavate invertebrates
from bark, lichens, and moss mats. Mountainspring et al. (1990) specu-
lated that Poo-uli also foraged on the ground, partly suggested “by the
bird’s drab color and stout pedal morphology.” They went on to compare
Poo-uli coloration with that of ground-foraging antbirds. Poo-uli have
seldom been seen at ground level (Mountainspring et al. 1990), but this
could result from the difficulty of potentially observing such behavior in
the dense understory that prevails in its range. Nevertheless, comparison
with other branch-foraging specialists is also appropriate. Passerines spe-
cialized for taking insects from bark or epiphytes include members of the
Furnariidae and Troglodytidae in Costa Rica (Sillett 1994), Certhiidae in
North American alder rainforests (Stiles 1978), and Paradisaeini (Corvi-
dae) in New Guinea (Pratt and Stiles 1985). All share brown plumage
(except adult male and some female birds of paradise) and stout pedal
morphology. Thus, brown plumage of Poo-uli may have evolved inde-
pendently as cryptic coloration associated with foraging for invertebrates
on branches.

Eggshells and nestlings. — In background color and spotting color and
pattern, the Poo-uli eggshells resembled eggs of other drepanidines, e.g..
Common Amakihi {Hemignathus virens) and Palila. The Poo-uli egg-
shells, together with eggs of most drepanidines, differ in color from eggs
of most cardueline finches in having the background whitish rather than
tinged with blue or green (Newton 1972). The Poo-uli nestlings share
gray down on the head and a pinkish-red gape with nestlings of nine
other drepanidines: Akikiki {Oreomystis bairdi), Anianiau {Hemignathus
parvus), Apapane, Crested Honeycreeper (Palmeria dolei). Common
Amakihi, Kauai Amakihi {Hemignathus stejnegeri), liwi {Vestiaria coc-
cinea), Maui Alauahio, and Palila (Eddinger 1970; Berger 1981; van Rip-
er 1987; H. and P. Baker, F. Duvall, Jr., pers. comm.; T. Pratt, pers. obs.).
The yellow gape flange of Poo-uli chicks matched the flanges of these
species, which vary from yellow to cream. However, the red spot in the
corner of the gape flange and the dark area observed on the Poo-uli
nestling’s palate (which may have been due to shading and therefore
possibly irrelevant) have not been mentioned for the other species.

Nests. — The Poo-uli nests we collected are similar to open cup nests
of other drepanidine and cardueline species. In gross stucture and place-
ment, the two Poo-uli nests resembled nests we have collected and ex-
amined of such widely divergent Hawaiian honeycreepers as Apapane,

Engilis et al. • PLUMAGES AND NEST OF THE POO-ULI

617

Common Amakihi, Crested Honeycreeper, liwi, Maui Alauahio, and Maui
Parrotbill {Pseudonestor xanthophn>s) in the same habitat on Maui and
for the honeycreeper nests studied on Kauai by Eddinger (1970). The
Poo-uli nests differed from those of other honeycreepers primarily in the
species of plant materials and in size. However, the materials were of
plant species common near the nest site, and the nests seemed to fit well
within the size gradient of drepanidine nests relative to bird body size.
In their loose structure and moss matrix, the Poo-uli nests most closely
resembled nests of the slightly larger Crested Honeycreeper. The Poo-uli
nests differed most from the smaller Maui Alauahio nests in being less
finely and compactly woven and in building materials. Consistent com-
position between the two Poo-uli nests and their unique component plant
materials may be the result of their being built in the same locality by
the same pair of birds. A larger sample of nests would probably reveal
variability in Poo-uli nest construction and placement. The Poo-uli nests
resemble the description of the “commonest type” of nest for cardueline
finches in Europe (Newton 1972:175). This type was described as “rather
bulky and made of various flexible materials, often with a base of twigs
and bents, a main structure of grass and moss, and a lining of hairs and
rootlets.” The Poo-uli nests differ from this description in their near ab-
sence of graminoid leaves and hair, materials rare in the bird’s habitat.

Implications for conservation. — We draw attention to the following
new aspects of Poo-uli natural history that bear on its conservation. In-
formation on age and sex differences in plumage characters will help
investigation of the species’ demography. However, because our conclu-
sions are based on very few individuals, other field workers should at-
tempt to further explore variability in characters used for ageing and sex-
ing Poo-uli.

We note the low wing/tarsus ratio of 2.74 and 2.88 and short tail of
38.0 and 36.5 mm, for the holotype and paratype, respectively (measure-
ments from Casey and Jacobi 1974). This ratio and tail length are the
smallest for the ten historic honeycreeper species from Maui (from mea-
surements in Amadon 1950). The low wing/tarsus ratio, the short rounded
wings, and extremely short tail of the Poo-uli may be clues to its mobility.
Poo-uli are capable of short flights only and may be confined to home
ranges smaller than those of most Hawaiian honeycreepers.

ACKNOWLEDGMENTS

We thank the Hawaii Dept, of Land and Natural Resources and U.S. Fish and Wildlife
Service for supporting our study and for aggressively pursuing re.storation of the Poo-uli;
Tom Hauptman for many memorable chopper flights; Allen Allison, Patrick Ching, Betsy
Harrison-Gagne, Larry Katahira, Jim Krakowsky, Joan Suther, and Rick Vetter for assistance

618

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

in observing Poo-uli at their nests; Phil Bruner and Allen Allison for collecting inactive
Poo-uli nests; Michelle Reynolds and Tom Snetsinger for sharing their observations of Poo-
uli; Mary LeCroy for examining the paratype; Warren Wagner and William Hoe assisted by
Clyde Imada for identifying angiosperm and bryophyte components, respectively, of nests;
Sabina Swift and Gordon Nishida for identifying arthropods extracted from nests; Helen
and Paul Baker for showing T Pratt their large collection of Maui Alauahio nests; Cheryl
Tarr and Rob Eleischer for sharing preliminary findings on molecular evolution in Poo-uli;
Allen Allison, Carla Kishinami, and staff at Bishop Museum for facilitating research at their
collection; Patrick Ching for permission to reproduce his painting; Paul and Helen Baker,
Tonnie Casey, Pern Duvall, Jr., Steve Pancy, Jim Jacobi, and Mary LeCroy for reviewing
drafts of this article.

LITERATURE CITED

Amadon, D. 1950. The Hawaiian honeycreepers (Aves, Drepaniidae). Bull. Am. Mus. Nat.
Hist. 95:157-262.

Baldwin, P. H. and T. L. C. Casey. 1983. A preliminary list of foods of the Poo-uli.
'Elepaio 43:53-56.

Berger, A. J. 1981. Hawaiian birdlife. 2nd edition. Univ. Hawaii Press, Honolulu, Hawaii.
Bock, W. J. 1978. Tongue morphology and affinities of the Hawaiian honeycreeper Me-
lamprosops phaeosoma. Ibis 120:467—479.

Casey, T. L. C. and J. D. Jacobi. 1974. A new genus and species of bird from the Island
of Maui, Hawaii (Passeriformes: Drepanididae). Occ. Pap. Bernice P. Bishop Mus. 24:
216-226.

Eddinger, C. R. 1970. A study of the breeding behavior of four species of Hawaiian
Honeycreepers (Drepanididae). Ph.D. diss., Univ. of Hawaii, Honolulu, Hawaii.
Engilis, a., Jr. 1990. Pield notes on native forest birds in the Hanawi Natural Area Reserve,
Maui. 'Elepaio 50:67-72.

Pancy, S. G., T. K. Pratt, G. D. Lindsey, C. K. Harada, A. H. Parent, Jr., and J. D.
Jacobi. 1993. Identifying sex and age of Apapane and liwi on Hawaii. J. Field Ornithol.
64:262-269.

Freed, L. A., S. Conant, and R. C. Fleischer. 1987. Evolutionary ecology and radiation
of Hawaiian passerine birds. Trends Ecol. Evol. 2:196—203.

Hailman, j. P. 1977. Optical signals: animal communication and sight. Indiana Univ. Press,
Bloomington, Indiana.

James, H. F. and S. L. Olson. 1991. Descriptions of 32 new species of birds from the
Hawaiian Islands: Part II. Passeriformes. Ornith. Monogr. 45:1-88.

Jeffrey, J. J., S. G. Fancy, G. D. Lindsey, P. C. Banko, T. K. Pratt, and J. D. Jacobi.

1993. Sex and age identification of Palila. J. Field Ornithol. 64:490-499.

Kepler, C. B., T. K. Pratt, A. M. Ecton, A. Engilis, and K. M. Fluetsch. 1996. Nesting
behavior of the Poo-uli. Wilson Bull. 108:620—638.

Mountainspring, S., T. L. C. Casey, C. B. Kepler, and J. M. Scott. 1990. Ecology,
behavior, and conservation of the Poo-uli (Melamprosops phaeosoma). Wilson Bull.
102:109-122.

Newton, I. 1972. Finches. William Collins Sons, Glasgow, Scottland.

Pratt, H. D. 1992. Is the Poo-uli a Hawaiian honeycreeper (Drepanidinae)? Condor 94:
172-180.

Pratt, T. K., S. G. Fancy, C. K. Harada, G. D. Lindsey, and J. D. Jacobi. 1994. Iden-
tifying sex and age of Akiapolaau. Wilson Bull. 106:421-430.

AND E. W. Stiles. 1985. The influence of fruit size and structure on composition

of frugivore assemblages in New Guinea. Biotropica 17:314-321.

Engilis et al. • PLUMAGES AND NEST OE THE POO-ULI

619

Ralph, C. J. and S. G. Fancy. 1994. Timing of breeding and molting in six species of
Hawaiian honeycreepers. Condor 96; 151-161.

Scott, J. M., S. Mountainspring, E L. Ramsey, and C. B. Kepler. 1986. Forest bird
communities of the Hawaiian Islands: their dynamics, ecology, and conservation. Stud.
Avian Biol. 9.

Sillett, T. S. 1994. Foraging ecology of epiphyte-searching insectivorous birds in Costa
Rica. Condor 96:863-877.

Stiles, E. W. 1978. Avian communities in temperate and tropical alder forests. Condor 80:
276-284.

VAN Riper, C. 1987. Breeding ecology of the Hawaii Common Amakihi. Condor 89:85-
102.

COLOR PLATE

Publication of the frontispiece painting has been made possible by an endowment estab-
lished by George Miksch Sutton.

Wilson Bull., 108(4), 1996, pp. 620-638

NESTING BEHAVIOR OF THE POO-ULI

Cameron B. Kepler,' Thane K. Pratt,^ A. Marie Ecton,^
Andrew Engilis, Jr.,"* and Kimberly M. Fluetsch^

Abstract. — We describe two sequential nestings of a pair of Poo-uli (Melamprosops
phaeosoma), a Hawaiian honeycreeper nearing extinction. Similarities to nesting of most
other honeycreepers included: nest site in ohia lehua (Metrosideros polymorpha Gaud.)
canopy; breeding in March through June; monogamous breeding system with the putative
male helping build the nest, feeding the putative female throughout each nesting event, and
feeding the chicks, but not incubating or brooding; and complete nest sanitation. Notable
differences were the paucity of songs and calls by the parents and inclusion of snails in the
diet of nestlings. Clutch size was probably two eggs for both nests. High winds, rain, or
both influenced parental behavior; the female stayed longer on the nest and took shorter
recesses in poor weather. Weather did not affect rates at which the male fed the female on
the nest; however, the feeding rate increased from the egg to the chick stage probably
because food was passed on to the chicks. At nest #2, parents fed young chicks (<14 days
old) more often in good than in poor weather; data were insufficient for old chicks. Weather
is usually poor throughout the year in the relictual range of the Poo-uli and is likely to
impact nesting success. The first nest failed in poor weather. The second fledged a single
young 21 days old. Diet of nestlings appeared to consist of a higher proportion of insect
larvae than that of older birds, which are reported to eat mostly snails. Received 12 Dec.
1994, accepted 27 May, 1996.

Few endangered birds are closer to extinction than the Poo-uli {Melam-
prosops phaeosoma), a monotypic species and genus of Hawaiian hon-
eycreeper (Fringillidae; Drepanidini). Since its discovery on Haleakala
Volcano, Maui Island in 1973 (Casey and Jacobi 1974), the Poo-uli pop-
ulation has fallen from several hundred to fewer than 10 birds today, and
it is extinct at the type locality (Scott et al. 1986; Engilis 1990; Moun-
tainspring et al. 1990; J. Simon and M. Reynolds, pers. comm.).

Why is the Poo-uli disappearing? Past research has been sporadic and
underfunded; consequently the life history and population ecology of the
bird are poorly understood. Field work has also been hampered by logis-
tical difficulties, inhospitable conditions, and the bird’s low population
density and lack of vocal activity. Nevertheless, scant data (Casey and
Jacobi 1974, Baldwin and Casey 1983, Engilis 1990, Mountainspring et

' National Biological Service, P.O. Box 44. Hawaii National Park, Hawaii 96718. Present address: Na-
tional Biological Service, Southeast Research Station, Warnell Sch. Forest Res., Univ. of Georgia, Athens,
Georgia 30602-2152.

2 NBS, Hawaii Field Station, P.O. Box 44, Hawaii National Park, Hawaii 96718.

3 NBS, P.O. Box 44, Hawaii National Park, Hawaii 96718. Present address: Biology Dept., University of
Miami, P.O. Box 24918, Coral Gables, Florida 33124-0421.

^ Bernice P. Bishop Museum, P.O. Box 19000-A, Honolulu. Hawaii, 96817. Present address: Ducks Un-
limited. Inc., 3074 Gold Canal Dr., Rancho Cordova, California 95670.

’NBS, P.O. Box 44, Hawaii National Park, Hawaii 96718. Present address: 6018 Royal Creek, San
Antonio, Texas 78239.

620

Kepler et al. • POO-ULI NESTING BEHAVIOR

621

al. 1990) and conjecture based upon biology of other honeycreepers (Ke-
pler et al. 1984; Scott et al. 1986; van Riper et al. 1986; Engilis 1990;
Mountainspring et al. 1990; Atkinson et al. 1995) implicate habitat dam-
age by feral pigs {Sus scrofa), predation by and competition with non-
native small mammals, increased risk to avian disease below 1800 m
elevation, and the untested hypothesis that the bird’s molluscan prey base
is also dwindling. The poo-uli’s substrate-restricted foraging for arthro-
pods and molluscs in bark and epiphytes (Mountainspring et al. 1990)
implies ecological specialization vulnerable to environmental change
brought about by the invasion of non-native organisms. So far, recovery
efforts begun in 1990 have focused not on the poo-uli, but on successful
habitat restoration through pig removal and exclusion. Recently, the Na-
tional Biological Service, funded and otherwise supported by other agen-
cies (see Acknowledgments), initiated a program for research and resto-
ration of the Poo-uli.

In 1985-1986, Kepler and Engilis studied aspects of the ecology of
endangered Maui birds, including the Poo-uli (Mountainspring et al.
1990) at Hanawi Natural Area Reserve. In 1986, they discovered and
monitored two active nests of a pair of Poo-uli. Our purpose is to describe
events at these nests in as much detail as possible, because (1) this is the
only information on Poo-uli reproduction, (2) recovery efforts, in the field
and in captivity, will benefit from knowledge of the natural history of the
species, and (3) the Poo-uli may go extinct, leaving no further record.
We compare behavior of the Poo-uli with that of other Hawaiian hon-
eycreepers and mainland cardueline finches, from which the honeycreep-
ers are descended (James and Olson 1991). We also discuss how this
information may help the species’ survival.

STUDY SITE AND METHODS

The two nests were within about 30 m of each other along a small, eastern tributary
ravine of the east fork of Hanawi Stream at 1800 m elevation. Nest #1 was situated on a
small ridge crest, more exposed to prevailing trade winds than nest #2, which was on the
east flank of the same ridge about 15 m above the ravine floor. Both were within 100 m of
a headwall of the Hanawi gulch. Vegetation at the site was Mixed Shrub Montane Wet
Forest (Jacobi 1989) with mean canopy height of 13 m and crown cover averaging 60%
and dominated by ohia lehua {Metrosideros polymorpha Gaud.) (Mountainspring et al.
1990). Damage to vegetation by feral pigs appeared slight, with the under story largely
intact. Rainfall, brought year-round predominantly by NE trade winds, was estimated to
exceed 3 m per annum.

Nest #1 was built in a canopy ohia lehua 15 m tall. Vegetation surrounding the nest tree
included ( 1 ) a subcanopy kolea tree (Myrsine sp.) and pukiawe shrub (Styphelia tomeiameiae
[Cham. & Schlechtend.j E v. Muell.) where the male often fed the female, and (2) a dense
understory of shrubs and ferns used as cover by the birds when approaching or leaving the
nest area. The nest was located in a secondary, horizontal branch in the lower crown, 8 m

622

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 1

Dates of Observation and Nest Stages

Dates

Stage

Nest #1

5-6 March

Nest construction, courtship

17-20 March

Eggs

24-26 March

Eggs hatch 25 March; nestling

31 March to 3 April

Nestling; first seen April 2

7 April

Nestling

Nest fails 8-14 April

Nest #2

16 April

Nest construction

9-14 May

Eggs hatch 11, 13 May; nestlings

19-22 May

Nestlings

29-31 May

Nestlings, one fledges on 31 May

1 June

Fledgling

above ground, and was incorporated into live twigs and small branchlets a few cm below
live foliage. The nest site was exposed to some direct sunlight in the morning and was
sheltered from trade winds, but it swayed in an arc of ca 1 m in SE winds >13 km h
The globular, open cup nest was composed of sticks, mosses, and plant fiber (Engilis et al.
1996). Two other, inactive and unidentified nests occupied foliage above and below on the
same branch. During nest construction, the male infrequently visited a fourth, triangular nest
4 m up in a 5-m ohia lehua sapling within 10 m of the nest tree. Based on the construction
and location of this nest, we believe it was built by a non-native Red-billed Leiothrix
(Leiothrix lutea).

Nest #2 was also built in an ohia lehua tree surrounded by similar vegetation. Nearby
pukiawe and kanawao (Broussaisia arguta Gaud.) shrubs provided nest material. This nest
was placed in the tree in a position very similar to nest #1. The nest was only 8 m above
ground and sheltered from NE trades, being situated in the SE (140°), uphill portion of the
crown, and 5-10 m lower than the crowns of nearby ohia lehua. It was, however, exposed
to SE winds which caused the nest branch to sway 1-2 m. No other nests were noted in
the tree.

Only two Poo-uli were observed tending the nests. These care-givers, likely the same two
birds at both nests, were recognized by plumage characters (Engilis et al. 1996). We assume
the brightly colored bird was the male and the drab bird was the female. Viewed closely,
neither showed lesions of active or past infection from avian pox that might have influenced
their behavior.

We studied both Poo-uli nests for periods of one to five days, from nest construction until
fledging or failure (Table 1). We observed Poo-uli at the nests from a distance of 40 m (nest
#1) and 15 m (nest #2) through binoculars, spotting telescope (Bausch & Lomb 30X, nest
#1) or Questar telescope (SOX, nest #2) from under a tarp shelter. A creek separated ob-
servers from both nest trees, and when flowing vigorously it prevented us from hearing
Poo-uli vocalizations, especially at the more distant nest #1. On most days, weather per-
mitting, observers watched the nest continuously from 08:30 to 17:00 (all times are Hawaii
Aleutian Times). We did not approach the nest trees while nests were active. Our presence
did not appear to influence the birds' behavior at nest #1, but may have done so at the
clo.ser nest #2 (see below). From the observation points, we could usually view parental

Kepler et cil. • POO-ULI NESTING BEHAVIOR

623

behavior at the nests, but we could not see the nest contents until the nestlings were old
enough to reach above the nest rim. We recorded: duration of behaviors at the nest to the
nearest 10 sec, vocalizations, and feedings and other behaviors off the nest. We mapped the
birds’ movements to and from the nest.

At times, heavy rain or fog prevented accurate observation; these data were omitted from
analyses. We estimated heights and distances to the nearest meter. We recorded percent
cloud cover, rain scores (0 = none; 1 = mist; 2 = drizzle; 3 = light rain; 4 = downpour),
and wind scores by the Beaufort scale. Once inactive, both nests were collected and depos-
ited at the B. R Bishop Museum, Honolulu (Engilis et al. 1996).

The Questar enabled us to identify some prey items brought to nest #2. We identified
prey items as (1) caterpillars, for larvae colored other than white or pink; (2) pale larvae,
for larvae colored white or pinkish, which likely were bark-dwelling coleoptera or lepidop-
tera, (3) beetles; (4) succineid snails (Succineidae); and (5) snails, for unidentified snails.
Mountainspring et al. (1990) described foraging observations in vicinity of the nests.

We assigned observations to the incubation stage at nest #1 prior to 12:00 on 25 March
and at nest #2 prior to 10:00 on 1 1 May; observations afterwards were assigned to the
nestling stage. We categorized nestlings as young (<14 days) or old, based on the assump-
tion that they were partly feathered, thermoregulating, and required less brooding by the
parents at about 14 days age or older.

Time on the nest is the length of the visit to the nest; time off is the time from when the
bird left the nest until it returned. Preliminary models using stepwise linear regression
indicated that both wind and rain significantly affected time spent by the female on and off
the nest. We created a combined weather variable that was coded as “poor” whenever winds
exceeded 8 km h"' (Beaufort scale 2) or rain occurred (rain score >1) or both; otherwise,
we coded weather as “good.” The models also showed that nest number significantly af-
fected time spent by the female on and off the nest, being longer for both at nest #2.
Nearness of observers to nest #2 may have caused the female to hesitate leaving or returning
to the nest. Consequently we analyzed nests separately for time spent by the female on and
off the nest.

We compared rates of the male feeding the female at both nests combined and of parents
feeding chicks at nest #2 (better visibility) using a test of comparison for two Poisson
processes (Cox and Lewis 1978:225). Sample units were daily rates calculated separately
for good and poor weather. Values of P < 0.05 were considered statistically significant.

RESULTS

Behavior

Nestbuilding. — Nest #1 was under final construction by one or both
Poo-uli when discovered and first observed at 13:15-16:37 and 07:24-
14:00 on 5 and 6 March, respectively. At that time, we did not record
observations systematically. In the vicinity of the nest tree, the birds
moved through the subcanopy of the forest at 5-12 m. They were excep-
tionally active for Poo-uli, moving quickly in the subcanopy, pausing at
times to preen, forage or gather nest material such as moss from ohia
lehua branches. When arriving in the nest tree, the birds flew quickly to
the nest; when leaving, they often dropped vertically from the nest, dash-
ing away above the undergrowth, then ascending trees distant from the
nest. Both birds visited the nest with about equal frequency; however, we

624

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

did not determine if one or both performed nest construction. The male
was observed singing repeatedly at and near the nest and while courting
the female. Dense fog frustrated further observation on the second day.
Once incubation began, we did not observe either parent taking building
materials to the nest and assume that nest construction had ceased.

Nest #2 was also under final construction when discovered on 16 April.
At nest site #1 at 12:10-16:00 on 14 April, we detected the male and
female foraging and giving whistles and chit-chit calls 18 times, but did
not hear song or observe nest-building. The pair was associated tightly,
and we observed courtship feeding. We again visited the site at 10:55-
14:50 on 16 April and observed the Poo-uli carrying material to a new
nest at 12:00-13:24. Though both parents were present, the female was
only once seen carrying material to the nest (moss collected near the
ground), while the male was seen carrying material to the nest nine times
(six times with twigs and three with moss). Twigs were collected three
times from a pukiawe shrub, and moss was gathered three times, <1 m
from the ground, in a kanawao shrub. We did not see either bird actually
build the nest. The pair was silent during construction of this nest. Once
incubation began, we observed the female add new material to either nest
only once.

Courtship. — We observed courtship during nest-building only at nest
#1, at 13:45-13:50 on 6 March. The male was detected singing and dis-
playing to the female 12mupinal5m ohia lehua tree distant from the
nest. While the female stood still, the male circled her, wing-flicked, and
delivered six songs in about 30 sec. The female then flew into the nest
tree, with the male following and singing in flight. The female moved
close to the nest; the male joined her and continued circling and singing
eight songs in about 30 sec. The female then returned to her previous
location in the distant tree, the male following and singing. Singing and
chasing continued, screened from view.

Egg stage. — We observed nest #1 for 24.3 h in six days and nest #2
for 11.5 h in three days during the egg stage. At nest #1, egg(s) were
laid either during 8-17 March (by 18 March the female incubated con-
tinuously) or on about 10 March, assuming an incubation period of 16
days and hatch date of 25 March. At nest #2, eggs were laid on about 26
and 27 April, assuming hatch dates of 1 1 and 13 May (see below). Laying
of the second clutch followed 13—19 days after failure of the preceding
brood. Clutch size was not determined but is assumed to be two, because
at nest #1 we saw only one nestling and watched the female eating an
egg, nest #2 contained two chicks, and when collected the nests contained
neither remains of other eggs nor chicks.

Kepler et al. • POO-ULI NESTING BEHAVIOR

625

Only the female incubated at both nests. Behaviors of the incubating
female included inactivity, shifting position in the nest, breast pumping
motions as she settled on eggs, preening, adjusting nest material, accept-
ing food from the male, and manipulating objects inside the nest (prob-
ably turning the egg). She was also observed resting with eyes closed for
a few seconds on nest #2 during the day. The incubating female crouched
low in the nest with her head tilted upward so that her eyes peered just
over the rim of the nest; at times she crouched so low that she was not
visible. Occasionally, she would turn and face a different direction. Both
she and her mate approached and left the nest tree quickly and deliberately
from several favorite routes. They usually arrived at the nest by flying
first into the nest tree, then hopping towards the nest. They usually de-
parted from the nest directly, not via the nest tree.

The female recessed to defecate, to be fed by the male, to forage, and
to perform other activities. Both sexes often wing-flicked in vicinity of
the nest, and the female bill-wiped on branches while approaching the
nest. The female sometimes recessed only to defecate copious white feces,
which she did in the nest tree or from nearby vegetation; she then returned
immediately to the nest. The male usually consorted with the female when
she recessed and was observed feeding her during recesses, either in the
nest tree or in nearby vegetation.

The female was fed by the male both on or off the nest. She solicited
feeding by wing-fluttering or -quivering, and rarely by vocalizations au-
dible to observers. At nest #2, we observed these feedings in better detail
during the chick stage: when on the nest and anticipating the male’s ap-
proach, she would point her bill up and begin bill-clapping with increas-
ing frequency as the male neared. She was rarely heard giving a faint
two-note call prior to the male’s arrival, but this vocalization was hard to
hear and could have gone unnoticed most of the time. All observations
of food transferal were of regurgitation rather than of carrying and trans-
ferring food in the bill. The male delivered boli of food into the female’s
gaping mouth in the same way that he later fed the chicks. During feed-
ings, the male perched on the same level or above or below her. The
female often left the nest shortly before or as the male approached; by
what cues she detected his approach are not known, perhaps by sight or
by faint chit-chit calls rarely heard by us. We believe that the male fed
the female during most recesses, because recesses were usually too short
for the female to forage profitably and because on many brief recesses
the male and female were seen consorting in dense cover, where they
could not be further observed.

During the egg stage, the male was usually seen visiting the vicinity
of the nest before or during the female’s recesses or when he fed the

626

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

female on the nest. Twice, he chased the female in the nest tree. During
a heavy rain he loafed, preened, and head-scratched under the shelter of
a branch in the nest tree. Though he sometimes foraged in the vicinity of
the nest tree, his usual long absences followed by his arrival and imme-
diate feeding of the female suggest that he foraged mostly beyond view
of the nest.

Nestling stage — parental behavior. — We observed nest #1 for 46.2 h
during seven days and nest #2 for 66 h during 1 1 days of the nestling
stage. We believe we observed hatching at both nests. At nest #1, we first
saw a chick on 1 April. However, we suspect hatching occurred much
earlier, perhaps on 25 March, when at 11:44, the female, incubating but
fidgeting, hopped onto the nest rim and extracted from the nest cup a
“flesh-colored object” 4 cm long and “flaccid,” which she immediately
consumed. The observer questioned whether she had eaten one of her
own eggs. We doubt that the object was a hatched eggshell, because it
didn’t look like one, and other honeycreepers discard shells away from
the nest (T. Pratt, pers. obs.). We also doubt it was a food item, because
the female’s last feeding was 17 min earlier, or a fecal sac of a small
chick. Also, from that day onward, we noted that the female more fre-
quently directed her attention to the interior of the nest. We assume that
the observer had witnessed the female eating the contents of a broken
egg or a dead chick.

At nest #2, the first hatching probably occurred on 1 1 May. The day
previous, the female incubated uneventfully. On the morning of 1 1 May,
she frequently interrupted incubation and directed her attention to the nest
interior, and for the first time the male was seen checking the contents of
the nest. Based on the parents’ behavior, we believe that hatching occurred
at about 10:00. At 11:02, the female may have fed a chick, and by the
end of the day one definite feeding was observed. The second chick may
have hatched on 13 May. On four occasions from 09:45 to 1 1:40 on 13
May, the female flicked objects that looked like eggshells (or fecal sacs?)
from the nest. Afterwards, the male and female were observed simulta-
neously feeding chicks at two locations in the nest. We first observed a
chick on 19 May. We could not determine the initial brood size at either
nest, but observed only one chick at nest #1 and two chicks at nest #2,
once the chicks began to lift their head above the nest rim.

Only the female brooded. Behaviors of the female at the nest were
similar between the egg and nestling stages, but now included a behavior
we call “nest-treading,” feeding and grooming the chicks, nest sanitation
and maintenance, drinking water drops from twigs, and an occasional
brief nap. Nest-treading involved the brooding female treading the floor
and inside walls, perhaps either to adjust her position and the chick’s or

Kepler et al. • POO-ULI NESTING BEHAVIOR

627

to stretch and enlarge the nest cup. The female fed the chicks with food
she had collected herself, but more often with food given to her by the
male at or away from the nest. Both parents carried food internally, rather
than in their beak. However, during the old chick stage, food was some-
times carried in the beak. Fecal sacs were eaten, flicked over the side of
the nest, carried and discarded away from the nest, or transferred from
the female at the nest to the male to take away. Nest sanitation must have
been efficient, because feces were not observed on the rim of the nests
or found when the nests were collected. Behaviors of the female on recess
did not change between the egg and nestling stages, except that in addition
she discarded fecal sacs.

The male continued the same patterns of activity as during the egg
stage. Though he fed the female on nest #1, he was not observed feeding
or tending the nestling. Absence of observed care-giving by the male
towards the chick at nest #1 may have been due to difficulty in viewing
that nest, rather than an absence of such behavior. At nest #2, beginning
on hatching day, the male frequently fed the female and nestlings and
removed fecal sacs. On the day prior to fledging, the male on a few
occasions delivered food to the nest by carrying a succineid snail in his
mouth rather than internally.

Nestling stage — chick development. — Chicks remained huddled under
their mother throughout the first two weeks in the nest and were usually
seen only when they lifted their heads to feed. At nests #1 and #2, chicks
were first seen, begging, on 2 April and 19 and 21 May (observers absent
15-18 May), respectively, when each attained nine days of age. Our data
on the behavior of older nestlings is sketchy because chicks were ob-
served for only three days during their last week in the nest. Nest #1
failed in a downpour of 350 mm rain during 8-14 April. Nest #2 fledged
one chick on 31 May; fledging may have been delayed by weather, which
was poor the day before fledging. The smaller and by then much weaker
chick was last seen gaping on 29 May and is presumed to have died in
the nest, though its remains were never found. In the two days prior to
fledging, the surviving chick, when unattended, spent most of its time
resting, preening, exercising in short bouts of wing-flapping, and drinking
water off plant material. When fed, it flapped its wings vigorously. We
never heard it call. Besides relying on its parents for nest sanitation, the
chick also defecated over the side of the nest.

On 31 May, the day it fledged, the chick made several excursions to
branches in the immediate vicinity of the nest, at first returning to the
nest to be fed, then leaving the nest for feeding. It was first heard giving
single, infrequent chip notes. Though the female brooded the chick on
five occasions, the chick sometimes resisted by pushing her off the nest.

628

THE WILSON BULLETIN • VoL 108, No. 4, December 1996

Percent (%) EGGS: TIME OFF

50 n
40
30
20
10

Percent (%) YOUNG CHICKS: TIME ON

Weather
□Good n=83
■ Poor n=93

rm

50 -1

40

30

20

10

O.M.O 1-5 5-10 10-20 20-30 >30

Time (min)

Percent (%) OLD CHICKS; TIME ON

Weather
□Good n=19
■ Poor n=19

DCLc

0 1-1,0 1-5 5-10 10-20 20-30 >30

Time (min)

□Good n=48

■ Poor n=22

0.1-1.0 1-5 5-10 10-20 20-30 >30

Time (min)

Fig. 1. Frequency distribution (percent observations) of incubation, brood, and recess
times of the female Poo-uli in good vs poor weather at nests #1 and #2 combined. Poor
weather had winds >8 km h ' or rain or both. Shown are times for three stages: eggs, young
chicks (<14 days old) and old nestlings (^14 days).

Shortly after 16:30, the chick fledged and moved into the canopy of the
nest tree. It was 21 days old.

Fledgling stage. — We observed the parents attending the fledgling dur-
ing 09:00-12:54 on 1 June. The fledgling was located in a 7 m tall
pukiawe tree near the nest tree. It seemed to be alone, loafed most of the
time, and remained in the subcanopy. It was capable of short horizontal
flights. Both parents provided food. Though the parents occasionally gave
chit-chit calls, the chick could not be heard. It wing-quivered while beg-
ging and moved about awkwardly.

Rates of incubation, brooding, and feeding. — Weather affected time
spent on and off the nest by the female during the egg and young chick
(<14 days old) stages (Fig. 1, Table 2). The female at nest #1 spent more

Kepler et al. • POO-ULI NESTING BEHAVIOR

629

Table 2

Time (min) Spent by the female Poo-uli on and off the Nests in Good vs Poor
(Winds > 8 km-h’' or Rain or Both) Weather during Stages of Eggs, Young Chicks
(<14 Days Old), and Old Chicks (>14 Days Old)

Stage

Weather

Nest #1

Nest #2

Time on

Time off

Time on

Time off

N

Mean

SE

Mean

SE

N

Mean

SE

Mean

SE

Egg

Good

41

14.5

2.25

6.3

1.25

18

25.0

2.12

7.7

3.05

Poor

18

25.1

7.43

4.2

2.25

11

24.9

4.38

3.9

1.24

Young chick

Good

51

15.2

1.63

2.7

0.46

47

12.4

2.19

11.5

1.79

Poor

62

21.0

2.17

1.7

0.30

53

25.9

3.02

8.2

1.43

Old chick

Good

19

7.9

2.41

18.2 .

5.18

Poor

25

11.2

3.04

8.6

2.09

time on the nest during poor weather (mean ± SD = 21.8 ± 17.6 min)
than during good weather (14.9 ± 11.5 min; two-way ANOVA, ^2 ,32 =
3.91, P = 0.02). Nest stage (eggs vs young chicks) did not significantly
affect time spent on the nest for nest #1 (F, ,32 = 0.34, P - 0.56), but
the length of recesses was longer when the female was incubating (5.7
± 7.3; F, ,37 = 12.02, P = 0.0007) than when she was brooding young
chicks (2.2 ± 2.7 min). Length of recesses at nest #1 was 4.2 ± 5.33 min
during good weather and 2.2 ±4.11 min during poor weather, but was
highly variable (F2137 = 2.34, P = 0.10).

For nest #2, nest stage and weather affected both time on the nest (two-
way ANOVA, F6 ,37 = 5.86, P = 0.0001) and time off the nest (F^ ijj =
2.11, P — 0.056). Time spent on the nest (mean ± SD) was 25.0 ± 10.1
min for egg, 19.5 ± 18.4 min for young chick, and 9.5 ± 11.9 min for
old chick (>14 days old) stages. Pairwise comparisons of means showed
that differences were not significant between egg and young chick stages,
but significant for young chick and old chick stages (Tukey’s test, P <
0.05). Mean time spent on the nest was 14.1 ± 13.1 min during good
weather and 21.8 ± 18.5 min during poor weather. Mean length of re-
cesses was 6.4 ± 10.5 min for egg, 9.7 ± 10.0 min for young chick, and
13.0 ± 17.0 min for old chick stages. None of the pairwise comparisons
of means was significant (Tukey’s test, P < 0.05). Recess time was 12.4
± 15.3 min during good weather and 7.8 ± 8.8 min during poor weather,
again highly variable.

The daily rate at which the male fed the female while she was on the
nest was not statistically different between good and poor weather during
both the egg stage (0.23 vs 0.66 feedings h“'; 22.0 and 12.1 h, respec-

630

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 3

Numbers of Identified Food Items Transferred from Male to Female or from

Adults to Chicks at Nest #2

Chick

age-*

Number of
feedings

Cater-

pillars

Pale

larvae

Succineid

snails

Other

snails

Beetles

Total

1-4

8

6+

3

0

(8?)

0

9 or 17

10-12

12

17

9

(2?)

0

0

26 or 28

19-21

32

30

17

34

4

1

86

• Age of the oldest chick in days.

lively; Z = 1.687, P = 0.091) and the young chick stage (1.08 vs 0.95
feedings-h“'; 30.5 and 49.5 h, respectively; Z = 0.556, P = 0.582). Rates
of the male feeding the female increased significantly from the egg to
young chick stage (0.38 vs 1.00 feedings h”' incubation or brooding; 34.1
and 79.9 h, respectively; Z = 4.03, P = 0.0001), because the male’s main
purpose for visiting the nest was to feed the chicks directly or via the
female. During the old chick stage, the female spent little time on the
nest, and consequently there were few male-to-female feedings.

Feeding rates of chicks varied with sex of parent and weather at nest
#2. The female fed young chicks at a significantly greater rate than did
the male (1.77 vs 1.03 feedings h"', 43.6 h observation; Z = 2.920, P -
0.004); however, the increased male-to-female feedings were likely passed
on to the chicks by the female, so that the male’s role in providing food
to young chicks could have been the same or greater than the female’s.
For older chicks, feeding rates by female vs male showed no significant
difference, perhaps due to small sample size (1.60 vs 2.24 feedings h“'
in 15.6 h; Z = 1.290, P = 0.197). Parents fed young chicks more often
in good than in poor weather (1.95 vs 1.04 feedings h“' in 17.2 and 26.5
h; Z = 3.257, P = 0.001); data were insufficient for old chicks.

Diet. — We cannot state whether items identified represent the complete
diet, because we could not view or identify most food transferred, and in
cases when we did recognize food items, these were only one or a few
items in each transfer. Most food appeared as indeterminate goop. Most
food items identified were lepidoptera and coleoptera larvae (Table 3).
Molluscs did not appear for certain in the food until the chick was near
fledging, when succineids became an important dietary component.

Interspecific interactions. — The Poo-uli did not actively defend their
nests from approaches by other honeycreepers. We recorded 14 approach-
es to within 2 m of nest #1: 11 by Apapane {Himatione sanguinea), 1 by
either an Apapane or liwi (Vestiaria coccinea), 1 by a Common Amakihi
(Hemignathus virens), and 1 by a Maui Alauahio {Paroreotnyza mon-

Kepler et al. • POO-ULI NESTING BEHAVIOR

631

tana). In nine approaches, the female was incubating or brooding and did
not respond, apart from watching the intruder or crouching lower in the
nest. In two of five instances when the female was off the nest and an
intruder approached, the male or female drove the intruder (Apapane)
away with displays and chasing. When displaying, the Poo-uli crouched
and, with neck extended forward, faced the intruder. In both instances,
we believe the nest was at the egg stage. On 14 April, after nest #1 was
abandoned, Apapane entered it several times, presumably collecting nest-
ing material. Apapane gather building material from nests of other birds
(Eddinger 1970).

Vocalizations

When not breeding, Poo-uli vocalize infrequently (Engilis 1990, Moun-
tainspring et al. 1990). Their calls are inconspicuous and very simple in
structure, consisting mainly of chit (chip, whit, or tch) notes given singly,
in couplets, or in short bursts (see Pratt 1992 for sonogram). Vocalizations
given while nesting are similarly rare and quiet.

Song. — A single male song on 5 March initially alerted us to a possible
nest; yet, only one song was heard from 13:15 to 16:35. Songs were heard
more frequently on 6 March when nest #1 was still under construction,
twice on 17 March early in the egg stage, and not at all later. The male’s
song consisted mainly of paired couplets in iambic pattern speeding up
towards the end and was audible only within 40 m of the bird. The
following song heard during courtship at 13:45-13:50 on 6 March was
typical: “Chit-chit chit-d chit-ter chit (pause) chit-ter chit-ter chit-ter.”

''Chit-Chit" call. — One to many notes; usually two repeated. Frequent-
ly given by male or female during construction of both nests and on three
occasions by the male during the chick stage at nest #2 when he accom-
panied the female on recess. On three occasions on 10 and 12 May, the
male gave chit-chit calls shortly before the female left the nest to join
him. The female gave soft chit-chit calls as he approached nest #2 on 20,
21 May.

Chit nore^.— Single “chit” notes were given by parents while foraging
together, perhaps as an interspecific flocking call, similar to that of Maui
Alauahio.

Alarm call. — During the nestling stage on 2 April at 16:32, the male
gave a series of three-noted calls, “chit, chit, chit,” interspersed between
bouts by single “chit” notes. He was seen perched low in a Myrsine sp.
tree, while the female brooded. Observers wrote that he “was most likely
disturbed by something and giving vocalizations to ward it off,” and they
implied that the something may have been a small mammal.

Whistle call. — Given once (7 April) by the male foraging in heavy rain

632

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

in the vicinity of the brooding female at nest #1 and given on four oc-
casions (14 April) by the male during construction stage of nest #2.

Chatter. — Once during the egg stage at nest #1, the female gave four
note chatter while wing-quivering and being fed by the male 1.5 m from
the nest. Another observer believed that the male gave the chatter call
during feeding.

Nestling calls. — These, if any, were inaudible from the observation
points. On the day the successful chick fledged, it gave single, infrequent
chit notes when alone on the nest.

Fledgling calls. — We heard none. Reynolds and Snetsinger (pers.
comm.) described the calls of a juvenile Poo-uli being fed on 30 August
1994 as “a high-pitched rapid twitter very similar to that of a juvenile
Hawaii Creeper {Oreomystis mana) or Hawaii Akepa {Loxops coccineus
coccineus) being fed.”

Mechanical sounds. — None. Flight was silent, lacking the wing-whir
of some other Hawaiian honey creepers.

DISCUSSION

Nest site. — The two nest sites we studied are important from both an
evolutionary and conservation perspective. Why did the Poo-uli build
their nests in tall ohia lehua? Poo-uli forage in the understory and sub-
canopy at a modal height of 5 m (Mountainspring et al. 1990); the two
nests at 8 m were at the high end of their reported foraging range. Even
more puzzling, why do the other six species of honeycreepers in Maui
rain forests also nest almost exclusively in canopy ohia lehua? Given (1)
the diversity in morphology, behavior, and life history traits evolved by
the Hawaiian honeycreepers (Amadon 1950, Freed et al. 1987) and (2)
the diversity of nest sites used by continental fringillids (Bent et al. 1968,
Newton 1972) and passerine communities generally (Martin 1988), the
uniformity in nest site selection by rain forest drepanidines is unexpected.
While it is beyond the scope of our paper to explore this convergence in
the selection of nest sites, we call attention to historic changes in pre-
dation pressure on nesting birds in Hawaii. Prior to human settlement, all
potential nest predators in these insular forests were birds — rallids, ibises,
raptors, owls, corvids, and drepanidines. Nearly all have vanished from
Maui forests, and instead six species of small mammals have invaded —
a mongoose (Herpestes auropunctatus), a cat (Felis catus), and four ro-
dents {Mas musculus, Rattus exulans, R. norvegicus, R. rattus). Most hon-
eycreeper species have become extinct as direct and indirect consequence
of human settlement, including mammalian predation; the surviving spe-
cies may be nesting in sites relatively safe from the new predators.

Breeding system and parental care. — We have no evidence to show

Kepler et al. • POO-ULI NESTING BEHAVIOR

633

that the Poo-uli defended an all-purpose territory, as no other conspecifics
were seen during the study, and the male did not sing to advertise a
territory. Nor did either parent consistently defend the nests from ap-
proach by other species. Sightings of the parents foraging in the vicinity
of the nest, their relatively weak flight, and the absence of long flights
over the canopy suggest that the pair may have confined their activity to
a home range of only a few hundred meters in radius.

Even with a sample size of one pair, we are tempted to infer that Poo-
uli are principally monogamous, because of the heavy involvement of the
male at all stages of the nesting cycle. Pair-bonding extends at least
through the breeding season, for the pair initiated the second nest with
minimal courtship and no singing. The pair bond may have been rein-
forced by the male feeding the female regularly throughout both nesting
cycles and by the pair consorting during the female’s recesses from the
nest. Monogamy is universal among drepanidines studied to date (Eddin-
ger 1970; van Riper 1980, 1987; Pletschet and Kelly 1990; Morin 1992;
H. Baker and P. Baker, pers. comm.; T. Pratt, unpubl. data; J. Simon, pers.
comm., E. van Gelder, pers. comm.) and carduelines generally (Newton
1972).

Parental care by the male and female resembled that of nine other
drepanidines studied to date and of carduelines generally (Eddinger 1970;
Newton 1972; van Riper 1980, 1987; Pletschet and Kelly 1990; Morin
1992; H. Baker and P. Baker, pers. comm.; T. Pratt, unpubl. data; J. Simon,
pers. comm.; E. van Gelder, pers. comm.). We note the likely increase of
care-giving to older chicks by the male, documented in few other dre-
panidines (Morin 1992), but perhaps common. Both parents performed
nest sanitation throughout the nestling phase, and we found the successful
nest #2 clean of feces. Among drepanidines studied to date (Eddinger
1970; Newton 1972; van Riper 1980, 1987; Pletschet and Kelly 1990;
Morin 1992; H. Baker and P. Baker, pers. comm.; T. Pratt, unpubl. data;
J. Simon, pers. comm.; van Gelder, pers. comm.), only the Laysan Pinch
{Telespiza cantans) and Palila {Loxioides bailleui) give up nest sanitation
in the final week of the nestling stage, allowing the rim of the nest to
become heavily encrusted with feces. Complete (or nearly complete) nest
sanitation by the Poo-uli and other drepanidines that feed their young
principally on invertebrate rather than plant foods is presumably a derived
behavior, as other carduelines are less fastidious (Newton 1972). The
seasonal span of the two Poo-uli nests coincides with peak nesting for
most other drepanidines in Maui rainforests (H. Baker and P. Baker, pers.
comm.; J. Simon, pers. comm.; E. van Gelder, pers. comm.) and on other
islands (Eddinger 1970, Ralph and Pancy 1994).

The male Poo-uli’s role of provisioning food to the nesting female and

634

THE WILSON BULLETIN • Vo/. 108, No. 4, December 1996

to his chicks assumes added importance in climatic conditions on the
species’ relictual geographic range. New weather stations in Poo-uli hab-
itat have recorded annual rainfall ranging from 5-12 m (L. Loope, pers.
comm.). Here, trade wind showers can prevail for weeks. Poor weather
can threaten the eggs and chicks with hypothermia and pit the survival
of progeny against that of parents faced with constraints on foraging time
(Drent 1975). Poor weather delayed the female Poo-uli from leaving the
nest and curtailed her recesses for foraging. Reduced foraging by the
female may have been compensated with provisioning by the male who
continued to feed the female on the nest at the same rate (feeding bouts
per time spent by the female on the nest) in poor weather as in good.
However, the rate at which parents fed the young chick decreased from
good to poor weather. We note the greater importance of wind versus rain
in influencing the female Poo-uli ’s time on and off the nest. Wind can
have a severe effect on egg temperature, incubation, and incubation be-
havior in small passerines (e.g., Morton and Pereyra 1985). However, we
believe that rain could have had a much greater effect on parental behav-
ior than measured by us. Heavy rains prevented us from observing the
nests, and this biased our sampling to “drier” conditions. For example,
we were unable to observe nest #1 through the curtain of rain that may
have caused its failure. Lastly, Cartar and Montgomerie (1987) found that
for female White-rumped Sandpipers (Calidris fuscicollis) incubation
“behavior appears at least to integrate the effects of both present weather
and weather on the previous day.” We could not explore such effects with
the Poo-uli because of our small data set.

Skutch (1976) noted that time spent on the nest was greater for species
in which the incubating bird received food from its mate. He also pointed
out the influence of rain on nesting birds. The slower growth rates of
chicks of tropical birds has been attributed to food limitation via reduced
rate of food delivery by parents (Ricklefs 1976, Martin 1987). The adap-
tive advantage of monogamous male birds provisioning their mate and
young in windy and/or high rainfall environments has received little at-
tention.

Chick development. — The nestling period of 21 days for our Poo-uli
chick is intermediate between 15-21 days for Common Amakihi (van
Riper 1987) and 22-26 days and 23-29 for Laysan Finch (Morin 1992)
and Palila (van Riper 1980; Pletschet and Kelly 1990; T. Pratt, unpubl.
data), respectively. At 25.5 g (N = 1; Engilis et al. 1996), the mass of
the Poo-uli is greater than that of the Common Amakihi but smaller than
that of the Laysan Finch and Palila, suggesting an intermediate nestling
period. Nestlings of European cardueline finches that nest in low bushes
spend fewer days in the nest and leave at an earlier stage of development

Kepler et al. • POO-ULI NESTING BEHAVIOR

635

than do nestlings of carduelines nesting in tall shrubs and trees (Newton
1972). Newton (1972) considered early fledging as an adaptation miti-
gating greater risk to predation. Nestling stage for all four drepanidines
is longer than that for shrub- or tree-nesting carduelines (Newton 1972).
However, the drepanidine chicks fledged at an advanced stage of devel-
opment, capable of level flight for short distances and with flight feathers
and body size close to that of an adult (van Riper 1980, 1987; Pletschet
and Kelly 1990; Morin 1992; T. Pratt, unpubl. data). These differences in
fledging time and development indicate an advantage for a prolonged
nestling period for the Poo-uli, and perhaps other drepanidines in montane
Hawaiian ecosystems.

The second chick apparently hatched two days after the first chick,
suggesting hatching asynchrony. The smaller chick died before the larger
one fledged, suggesting brood reduction. Whether this is a pattern for
second clutches in Poo-uli remains to be determined. Hatching asyn-
chrony and brood reduction occur in Common Amakihi, Laysan Finch,
and Palila (van Riper 1987; Pletschet and Kelly 1990; Morin 1992; T.
Pratt, unpubl. data).

Diet.- — Data on Poo-uli diet are few and tantalizing. Baldwin and Casey
(1983), after painstaking analysis of stomach contents of the only two
specimens, proposed that Poo-uli feed primarily on various small native
lands snails (especially Succineidae), beetles, and proportionately few oth-
er arthropods. Mountainspring et al. (1990) reported observations of Poo-
uli feeding on insect larvae and succineid snails; they postulated that
insect larvae might be a dietary component more important than proposed
by Baldwin and Casey. If our data are representative, which they may
not be because of observational bias, we confirm that Poo-uli feed exten-
sively on succineid snails. However, we observed lepidoptera and cole-
optera larvae being fed to nestlings at any age in greater proportion than
succineid snails. Poo-uli appear to conform with most passerines by feed-
ing caterpillars and other insect larvae to their young.

Vocalizations. — Our data, corroborated by observations of others
(Mountainspring et al. 1990), show Poo-uli to be the quietest of all dre-
panidines. We heard the male sing only during courtship and construction
at nest #1. At the time of our study, Poo-uli densities were very low, and
we did not observe this focal pair interacting with conspecifics. How
greater population densities and encounters among birds affected rates of
vocalizations is unknown. The song and chit-chit call are both diagnostic
and useful for detecting Poo-uli. However, the species’ rarity and infre-
quent vocalizing render conventional censusing ineffective (Scott et al.
1986).

Implications for recovery. — We found nothing in the nesting biology

636

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

of this pair of Poo-uli to indicate problems for reproduction or population
recruitment. Of significance may be the birds’ placing their nests in the
foliage of tall ohia lehua trees. We presume this location to be less haz-
ardous than sites in tree cavities, subcanopy trees and shrubs, or near or
on the ground, where nests might be encountered more often by non-
native, mammalian nest predators. We observed Rattus rattus below the
nest tree. This notorious enemy of insular birds (Atkinson 1985) thrives
in high population density in the study area (Sugihara, in press). Whether
the nest sites we observed are typical remains to be determined. Other
factors that may help prevent detection of Poo-uli nests by mammalian
predators are (1) complete nest sanitation; (2) the absence of odor at the
nests, relative to other drepanidine nests (Pratt 1992); and (3) infrequent
vocalizations at the nest. Nevertheless, we emphasize that reduction of
small mammal populations is crucial to lessening the threat of nest pre-
dation for the Poo-uli (Kepler et al. 1984).

The long nestling period and the potential of no more than two young
fledging would seem to handicap Poo-uli. However, Maui Alauahio and
Maui Parrotbill {Pseudonestor xanthophrys), two other sympatric insec-
tivorous honeycreepers sharing these life history characteristics and the
same windy and rainy habitat, have far larger geographic ranges and
population sizes. We suspect that factors such as decreasing food avail-
ability, habitat disturbance by feral pigs, and predation by non-native
mammals may be more important to the Poo-uli’s decline than vulnera-
bility arising from the species’ nesting behavior.

ACKNOWLEDGMENTS

We thank the Hawaii Dept, of Land and Natural Resources, U.S. Fish and Wildlife Ser-
vice, National Park Service, and The Nature Conservancy of Hawaii for supporting our
study and for aggressively pursuing restoration of the Poo-uli; Tom Hauptman for many
memorable helicopter flights; Allen Allison, Patrick Ching, Betsy Harrison-Gagne, Larry
Katahira, Jim Krakowsky, Joan Suther, and Rick Vetter for assistance with nest observations;
Phil Bruner and Allen Allison for collecting inactive Poo-uli nests; John Simon, Michelle
Reynolds, and Tom Snetsinger for sharing their observations on Poo-uli; Cheryl Tarr and
Rob Fleischer for sharing preliminary findings on molecular evolution in Poo-uli; Allen
Allison, Carla Kishinami, and staff at Bishop Museum for facilitating research at their
collection; Kim Berlin for discussions and guidance on avian incubation; Greg Brenner,
Steve Fancy, and Jeff Hatfield for statistical guidance; Paul and Helen Baker, Tonnie Casey,
Fern Duvall, Jr., Steve Fancy, Jeff Hatfield, Marie Morin, and Jim Jacobi for review and
discussion of drafts of this paper.

LITERATURE CITED

Amadon, D. 1950. The Hawaiian honeycreepers (Aves, Drepaniidae). Bull. Am. Mus. Nat.
Hi.st. 95:157-262.

Atkinson, C. T, K. L. Woods, R. J. Dusek, L. S. Sileo, and W. M. Iko. 1995. Wildlife

Kepler el al. • POO-ULI NESTING BEHAVIOR

637

disease and conservation in Hawaii: pathogenicity of avian malaria (Plasmodium relic-
turn) in experimentally infected liwi (Vestiaria coccinea). Parasitology. 11 1:559—569.

Atkinson, I. A. E. 1985. The spread of commensal species of Rattus to oceanic islands
and their effects on island avifaunas. Pp. 35—81 in Conservation of Island Birds. (P. J.
Moors, ed.). Inter. Council for Bird Preservation Technical Publication No. 3.

Baldwin, P. H. and T. L. C. Casey. 1983. A preliminary list of foods of the Poo-uli.
'Elepaio 43:53-56.

Bent, A. C. 1968. Lite histories of North American cardinals, buntings, towhees, finches,
sparrows and their allies (O. L. Austin, ed.). U.S. Nat. Mus. Bull #237, parts 1-3.
Washington, D.C.

Cartar, R. V. AND R. D. Montgomerie. 1987. Day-to-day variation in nest attentiveness
of White-rumped Sandpipers. Condor 89:252-260.

Casey, T. L. C. and J. D. Jacobi. 1974. A new genus and species of bird from the Island
of Maui, Hawaii (Passeriformes: Drepanididae). Occ. Pap. Bernice P. Bishop Mus. 24:
216-226.

Cox, D. R. AND P. A. W. Lewis. 1978. The statistical analysis of series of events. Chapman
and Hall, London, England.

Drent, R. H. 1975. Incubation. Pp. 333-420 in Avian Biology Vol. 5. (D. S. Earner and
J. R. King, eds.). Academic Press, New York, New York.

Eddinger, C. R. 1970. A study of the breeding behavior of four species of Hawaiian
Honeycreepers (Drepanididae). Ph.D. diss., Univ. of Hawaii, Honolulu, Hawaii.

Engilis, a., Jr. 1990. Field notes on native forest birds in the Hanawi Natural Area Reserve,
Maui. 'Elepaio 50:67-72.

, T. K. Pratt, C. B. Kepler, A. M. Ecton, and K. M. Fluetsch. 1996. Description

of adults, nestling, fledgling, and nest of the Poo-uli. Wilson Bull. 108:607-619.

Freed, L. A., S. Conant, and R. C. Fleischer. 1987. Evolutionary ecology and radiation
of Hawaiian passerine birds. Trends Ecol. Evol. 2:196-203.

Jacobi, J. D. 1989. Vegetation maps of the upland plant communities on the islands of
Hawaii, Maui, Molokai, and Lanai. Cooperative National Park Resources Study Unit,
University of Hawaii, Honolulu, Technical Report #68.

James, H. F. and S. L. Olson. 1991. Descriptions of 32 new species of birds from the
Hawaiian Islands: Part II. Passeriformes. Ornith. Monogr. 46:1-88.

Kepler, C. B., T. Burr, C. B. Cooper, D. Dunatchik, J. Medeiros, J. M. Scott, M. Ueoka,
and W. Wong. 1984. Maui-Molokai forest birds recovery plan. U.S. Fish and Wildlife
Service.

Martin, T. E. 1987. Food as a limit on breeding birds: a life-history perspective. Ann. Rev.
Ecol. Syst. 18:453-487.

. 1988. Processes organizing open-nesting bird assemblages: competition or nest

predation? Evolutionary Ecol. 2:37-50.

Morin, M. P. 1992. The breeding biology of an endangered Hawaiian honeycreeper, the
Laysan Finch. Condor 94:646—667.

Morton, M. L. and M. E. Pereyra. 1985. The regulation of egg temperatures and atten-
tiveness patterns in the Dusky Flycatcher (Empidonax oberholseri). Auk 102:25-37.

Mountainspring, S., T. L. C. Casey, C. B. Kepler, and J. M. Scott. 1990. Ecology,
behavior, and conservation of the Poo-uli (Melamprosops phaeosoma). Wilson Bull.
102:109-122.

Newton, I. 1972. Finches. William Collins Sons, Glasgow, Scotland.

Pletschet, S. M. and J. F. Kelly. 1990. Breeding biology and nesting success of Palila.
Condor 92:1012-1021.

638

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

Pratt, H. D. 1992. Is the Poo-uli a Hawaiian Honeycreeper (Drepanidinae)? Condor 94:
172-180.

Ralph, C. J. and S. G. Eancy. 1994. Timing of breeding and molting in six species of
Hawaiian honeycreepers. Condor 96:151-161.

Ricklefs, R. E. 1976. Growth rates of birds in the humid New World tropics. Ibis 118:
179-207.

Scott, J. M., S. Mountainspring, E L. Ramsey, and C. B. Kepler. 1986. Forest bird
communities of the Hawaiian Islands: their dynamics, ecology, and conservation. Stud.
Avian Biol. 9:1-431.

Skutch, a. E 1976. Parent birds and their young. Univ. Texas Press, Austin, Texas.
SuGiHARA, R. T. Relative abundance and diets of black and Polynesian rats in two Hawaiian
rain forests. Pac. Sci. (in press)

VAN Riper, C. 1980. Observations on the breeding of the Palila Psittirostra bailleui of
Hawaii. Ibis 122:462-475.

, S. G. VAN Riper, M. L. Goff, and M. Laird. 1986. The epizootiology and eco-
logical significance of malaria in Hawaiian land birds. Ecol. Mongr. 56:327-344.
. 1987. Breeding ecology of the Hawaii Common Amakihi. Condor 89:85-102.

Wilson Bull., 108(4), 1996, pp. 639-649

POPULATION DENSITY, VOCAL BEHAVIOR, AND
RECOMMENDED SURVEY METHODS FOR
BICKNELLS THRUSH

Christopher C. Rimmer,' Jonathan L. Atwood,^

Kent P. McFarland,' and Laura R. Nagy'-^

Abstract. — We studied territorial and vocal behavior of Bicknell’s Thrushes (Catharus
bicknelli) on Mt. Mansfield, Vermont, during June to September 1992 and in June 1993 to
1995. Birds sang and called consistently throughout the day during early— mid June. Later
in the season, songs were given infrequently, with vocalizations (mostly call notes) being
concentrated at dawn and dusk. Spot-mapping of vocalizing males yielded breeding density
estimates of 36-52 pairs/40 ha in 1992, 50-59 pairs/40 ha in 1993, 55-65 pairs/40 ha in
1994, and 45—53 pairs/40 ha in 1995. Other less labor-intensive techniques, including fixed-
width point counts and fixed-width transects, generally resulted in lower density estimates.
We recommend that future presence-absence surveys for BicknelFs Thrush in the north-
eastern United States be concentrated from 1-20 June. Surveys from late June through mid-
September should be attempted only if observers are able to be on-site before dawn or after
sunset. Received 26 Jan. 1996, accepted 8 May 1996.

BicknelFs Thrush {Catharus bicknelli), recently classified as a distinct
species from the Gray-cheeked Thrush {Catharus minimus) (Ouellet 1993,
American Ornithologists’ Union 1995), historically nested from the Gulf
of St. Lawrence, Gaspe Peninsula, Magdalen Islands, and Nova Scotia
south to the mountains of New England and New York (Wallace 1939).
Recent studies have suggested that the breeding range of this bird has
been reduced in both the United States and Canada (Ouellet 1993; At-
wood et al. 1996; J. Marshall, pers. comm.). The breeding habitat of
BicknelFs Thrush consists predominantly of dense, stunted coniferous for-
est dominated by balsam fir {Abies balsamea) and red spruce {Picea rub-
ens) (Wallace 1939, Atwood et al. 1996). In the northeastern United
States, this vegetation type is restricted to mountaintops higher than ap-
proximately 915 m elevation (Wallace 1939). Not only are most of these
areas somewhat inaccessible geographically, but the habitat itself is dif-
ficult to work within, being characterized by nearly impenetrable, dense
stands of conifers that are often located on rugged, steep slopes. These
logistic challenges have hindered attempts to estimate the current popu-
lation size of BicknelFs Thrush or to clarify its distributional limits in the
United States.

In this study we examine territorial and vocal behavior of BicknelFs

' Vermont Institute of Natural Science, RR 2, Box 532, Woodstock, Vermont 05091.

^ Manomet Observatory for Con.servation Sciences, P.O. Box 1770, Manomet. Massachusetts 02345.
Present address: Dept, of Biological Sciences, Univ. of Arkansas, Fayetteville, Arkansas 72701.

639

640

THE WILSON BULLETIN • Vo/. 108, No. 4, December 1996

Thrush in the context of potential techniques that might be used in future
population and distributional surveys. In particular, we compare density
estimates obtained through labor-intensive spot-mapping of countervo-
calizing, territorial males (Kendeigh 1944, Robbins 1970, Williams 1936)
with estimates derived from fixed-width line transects (Emlen 1974, Ken-
deigh 1944) and point counts at fixed-radius circular plots (Anderson
1970, Anderson and Shugart 1974). Additionally, we describe studies of
the influence of date and time-of-day on Bicknell’s Thrush vocal behavior.
On the basis of these results, we recommend a standard methodology for
future work aimed at assessing the distribution and population status of
this poorly known songbird.

METHODS

We established an irregularly-shaped, 8.8 ha study plot on the eastern portion of Mt.
Mansfield, Chittenden Co., Vermont (44°32'N, 72°49'W), at approximately 1150 m eleva-
tion. Dominant woody vegetation consisted of balsam fir interspersed with limited amounts
of red spruce, white birch {Betula papyrifera cordifolid) and mountain ash (Sorbus spp.).
Boundaries of the plot were determined primarily by topography and the location of two
pre-existing trails (Long Trail and Amherst Trail). The eastern edge was defined by a nearly
vertical cliff face covered with dense balsam fir forest; the western edge followed a major
ridgeline that separated suitable Bicknell’s Thrush habitat from extremely stunted (<0.5 m
high), “krummholz” vegetation located immediately to the west.

Because the plot’s dense vegetation and rugged topography made it extremely difficult to
leave the established trail system, we did not establish a standard grid system such as usually
forms the basis for spot-mapping studies (Franzreb 1981b). Instead, we used compass bear-
ings and tape measurements to map the locations of known “vantage points throughout
the plot. Compass bearings and distance estimates were then taken from these locations for
all Bicknell’s Thrushes that were seen or heard; these data were later transferred to a master
map of the area.

Spot-mapping was conducted on 12 dates in 1992 (11 — 18, 27—30 June). To validate our
1992 data and to assess changes in density between years, we also spot-mapped on 8 dates
in 1993 (8-10, 16, 17, 23, 24, 29 June), 14 dates in 1994 (5-9, 14-17, 20, 22, 23, 29, 30
June), and 16 dates in 1995 (1, 5-8, 12-14, 16, 19, 20, 22, 23, 27 June). Observations made
throughout the day were used to estimate the number of territorial males, here assumed to
represent breeding pairs. Only the locations of stationary birds were included. Simultaneous
registration of two or more vocalizing birds was used as the primary basis for discrimination
between adjacent individuals (Robbins 1970). 3Ve mapped positions of both singing and
calling thrushes, because our observations showed that both types of vocalization were used
to indicate territorial status. Although we obtained evidence of occasional calling by females,
we are confident that the great majority of calls registered in each year were given by males.

To assess the utility of other less labor-intensive methods for estimating breeding densities
of Bicknell’s Thrush, we conducted fixed-width line transects and point counts made at
fixed-radius circular plots on 10 dates (11 — 13, 15—18, 27—29 June) in 1992. Transects and
point counts, each of which required approximately 1 h to complete, were simultaneously
conducted three times per day: morning (beginning approximately 1 hour after sunrise),
mid-day (beginning approximately 8 h after sunrise), and evening (beginning approximately
30 min before sunset). Only spot-mapping data from 1992 were used to compare with other
census techniques.

Rimmer et al. • BICKNELL’S THRUSH SURVEY

641

The transect line, approximately 1 . 1 km in total length, followed existing trails that passed
through the study plot and formed its western boundary. A single observer walked slowly
along the transect and recorded all Bicknell’s Thrushes that were seen or heard within 30
m on either side of the trail. No recorded calls or “spishing” were used to elicit responses.
The extremely stunted vegetation immediately west of the study plot’s western border was
deemed unsuitable for Bicknell’s Thrushes; consequently, this section of the transect (0.5
km in length) was limited to a strip only 30 m wide, located entirely east of the transect
line. A total of 5.1 ha were included in the area sampled by the fixed-width line transect
method.

A total of seven 30-m radius circular plots >100 m apart were also located in the study
area. The positions of these plots were determined by trail and terrain considerations and
not by vegetation characteristics or observations of Bicknell’s Thrushes. Boundaries of cir-
cular plots were marked with flagging tape. A single observer recorded all Bicknell’s Thrush-
es heard or seen during a 5-min period at each plot; recorded calls or “spishing” were not
used to elicit vocalization. This basic technique was modified on four dates (11, 15, 18, 27
June) to include the use of recorded playbacks as a means of eliciting Bicknell’s Thrush
response. After completion of a standard 5-min count, a second count of 5-min duration
was conducted in which a tape recording of calls and songs was played for the initial 3
min. A total of 2.0 ha were included in the area sampled by the circular plot method.

To examine the effects of date and time-of-day on Bicknell’s Thrush vocal behavior,
counts of all songs and calls heard during 5-min periods were recorded at 30-min intervals
on 14 dates (11—13, 15—18, 27—29 June; 6, 14, 29 July; 16 September) in 1992. The first
of these counts on each date occurred approximately 20 min before sunrise; the last ap-
proximately 90 min after sunset. Vocalization counts were conducted near the “Octagon”
on Mt. Mansfield (elevation approximately 950 m); 7-10 Bicknell’s Thrush territories were
estimated to be within hearing distance of this site.

RESULTS

Density estimates. — Spot-mapping of vocalizing males yielded density
estimates for Bicknell’s Thrush of 36-52 pairs/40 ha in 1992, 50-59 pairs/
40 ha in 1993, 55-65 pairs/40 ha in 1994, and 45-53 pairs/40 ha in 1995.
The territories of 8, 11, 12, and 10 pairs were located entirely within the
borders of the plot in 1992, 1993, 1994, and 1995, respectively. The total
number of territories on the plot was estimated at 11.5 in 1992, 13.0 in
1993, 14.25 in 1994, and 11.75 in 1995. Independent evaluation of our
data by an individual experienced in spot-mapping but unfamiliar with
the plot yielded estimated totals of 11.75 territories in 1992, 13.25 terri-
tories in 1993, 12.5 territories in 1994, and 10.0 territories in 1995. Max-
imum density values were obtained by including percentages of territories
estimated to be located within the boundaries of the study plot. Minimum
density values were calculated by excluding all “partial” territories from
consideration.

Mean densities calculated from point counts conducted without play-
back elicitation were consistently lower than the maximum estimate de-
rived from spot-mapping but were comparable to the results of spot-map-
ping when edge territories were excluded (Table 1). Mean values obtained

642

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 1

Density Estimates for Breeding Bicknell’s Thrushes Based on Circular Plots and

Line Transects

Time of day

Survey technique

Morning

Mid-day

Evening

Circular plots (no playback) (N = 10)

36 ± 28.4^

8 ± 13.9

32 ± 25.6

65-97%*’

15-22%

58-86%

Circular plots (with playback) (N = 4)

48 ± 23.3

66 ± 78.2

35 ± 19.6

87-130%

120-178%

64-95%

Line transects (N = 10)

23 ± 14.1

5 ± 7.4

18 ± 11.7

42-62%

9-14%

33-49%

” Mean density (pairs/40 ha) ± 1 SD.

Mean density expressed as percent of estimates calculated from spot-mapping results (37-55 pairs/40 ha).

in the morning (36 pairs/40 ha) were slightly higher than those obtained
in the evening (32 pairs/40 ha). Mean mid-day results (8 pairs/40 ha)
greatly underestimated densities calculated through spot-mapping. The
modal density obtained for both morning and evening point counts over
the 10 days of the study was 20 pairs/40 ha. Bicknell’s Thrushes were
not detected within any of the circular study plots during 10% of morning
counts, 70% of mid-day counts, and 20% of evening counts.

Use of playback recordings increased the density estimates obtained
through point counts, but, in general, the results were still less than the
maximum estimate derived through spot-mapping (Table 1). Morning and
evening counts resulted in mean density estimates of 48 pairs/40 ha and
35 pairs/40 ha, respectively. Because of a single count in which playback
recordings stimulated an unusually large number of territorial interactions,
the mean density estimate based on mid-day counts (66 pairs/40 ha) ex-
ceeded the maximum estimate based on spot-mapping. Using playback
elicitation, the modal densities obtained on morning and evening counts
throughout the study period were 40 pairs/40 ha and 20 pairs/40 ha, re-
spectively. All point counts (morning, mid-day, and evening) that made
use of playback recordings successfully detected at least one Bicknell s
Thrush on each of the four days surveyed using this technique.

Line transects produced the lowest density estimates (Table 1). Results
based on morning transects (23 pairs/40 ha) were higher than those ob-
tained in the evening (18 pairs/40 ha). Transect data collected at mid-day
underestimated (5 pairs/40 ha) actual densities as calculated through spot-
mapping. Bicknell’s Thrushes were not detected on 10% of morning tran-
sects, 50% of mid-day transects, and 10% of evening transects.

Vocal behavior. — At the time of our preliminary visit to Mt. Mansfield

Ritnmer et ai • BICKNELL’S THRUSH SURVEY

643

on 2 June 1992, Bicknell’s Thrushes sang and called frequently through-
out the day, although no quantitative data were collected. According to
Green Mountain Club summit caretakers stationed on the mountain, sing-
ing Bicknell s Thrushes had been evident for at least one week prior to
this date. In 1993, 1994, and 1995 several individuals were heard calling
sporadically during the mornings of our initial site visits that ranged from
25 to 27 May. During mid-June (11—18 June) 1992, Bicknell’s Thrushes
called and sang consistently throughout the day, with only 18% of the
total vocalizations (N = 7998) being restricted to the first and last 2-h
periods of the day (Fig. 1). On 15 June, under full moon conditions,
Bicknell’s Thrushes were silent from 03:00 until the first song was given
at 03:57. Of 238 5-min counts conducted during mid-June between 04:30
and 21:30, we failed to record any Bicknell’s Thrush vocalizations in only
25 (10.5%) instances.

By late June (27-29 June) the frequency and consistency of Bicknell’s
Thrush vocalizations declined sharply, and songs, in particular, were given
only rarely (Fig. 1). Of the total vocalizations (N = 807) recorded on
these dates, 28% occurred during the first and last 2-h periods of the day.
We failed to record a single Bicknell’s Thrush vocalization during 60
(58.8%) of 102 5-min counts made during late June.

There appeared to be some resurgence in Bicknell’s Thrush vocal ac-
tivity, including both songs and calls, during July (6, 14, 29 July). Of the
total vocalizations (N = 1650) detected on these dates, 31% were given
during the first and last 2-h periods of the day (Fig. 1). We failed to
record a single Bicknell’s Thrush vocalization during 51 (50.0%) of 102
5-min counts made during July.

No vocalization samples were obtained during August. On 16 Septem-
ber, Bicknell’s Thrushes called inconsistently, and gave no songs during
the sampling periods (Fig. 1). Of the total vocalizations (N = 537) re-
corded on this date, 20% were given during the first and last 2-h periods
of the day. We failed to record any Bicknell’s Thrush vocalizations during
15 (51.7%) of 29 5-min counts on 16 September.

DISCUSSION

Surveys for Bicknell’s Thrush, aimed at determining presence or ab-
sence, or at estimating population size, are hindered by significant logistic
challenges. In the United States, known breeding habitat occurs almost
exclusively on mountaintops of more than approximately 915 m elevation
(Wallace 1939; Atwood et al. 1996). Of over 500 such high elevation
sites in New York, Vermont, New Hampshire, Maine and Massachusetts
that potentially support Bicknell’s Thrush, most are accessible only by
hiking, and many lack any established trail system. In addition to its

PERCENT OF TOTAL VOCALIZATIONS

644

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

JUNE 1 1

■ songs (N=694)
□ calls (N=713)

1,1-1, in,a,l

■ JUNE 27

■ songs (N=37)

□ calls (N=484)

■n[

nnll

1 1 1 1

■

JUNE 15

■ songs (N=517)
□ calls (N=429)

JULY 6

JUNE 1 6

■ songs (N=760)
□ calls (N=405)

'n^„n

■ligpg- .iMfl

JUNE 17

■ songs (N=736)
□ calls (N=4n)

40

30 -
20 -
10 -
0

JUNE 18

■ songs (N=373)
□ calls (N=547)

SEPT 16

■ songs (N=1 11)
□ calls (N=544)

■nnRu

■ songs (N=0)

□ calls (N=537)

J]

in

5 7 9 11 13 15 17 19 21

TIME (HRS)

Fig. 1. Daily and seasonal variation in Bicknell’s Thrush vocalizations. Axes of all
graphs on same scale as shown for 18 June.

frequent occurrence on difficult-to-access peaks, Bicknell’s Thrush nest-
ing habitat, which generally consists of nearly impenetrable thickets of
stunted coniferous forest located on steep, rugged terrain, is very difficult
to work within. Wallace (1939) commented that “only a freak ornithol-

Rimmer et al. • BICKNELL’S THRUSH SURVEY

645

ogist would think of leaving the trails [on Mt. Mansfield] for more than
a few feet [due to] the discouragingly dense tangles” of vegetation. Fur-
thermore, the time period in which surveys for Bicknell’s Thrush may be
effectively conducted is constrained by potentially harsh, high-elevation
weather conditions coupled with a brief breeding season. For example,
Davenport (1903) described Mt. Mansfield’s weather during June 1902 as
being in whole or part rainy . . . with high winds often making it im-
possible to keep one’s footing in the open on the summit.”

As a result of our study and data included in Wallace’s (1939) classic
life history work, we recommend the following protocol for conducting
presence-absence surveys for Bicknell’s Thrush. Sites in New England
and New York should ideally be visited between 1—20 June, during which
time songs and calls are consistently given throughout the day. In northern
parts of the species’ Canadian breeding range, surveys should be con-
ducted several days later than this (Ouellet, pers. comm.). Although Bick-
nell’s Thrush may be present on its United States breeding grounds as
early as 17 May (Rimmer and McFarland, unpubl. data), the possibility
of northward-bound migrants cannot be excluded until early June (Wal-
lace 1939). Because vocal activity becomes more sporadic later in the
season, especially during the middle portion of the day, surveys conducted
from late June through mid-September should be attempted only if the
observer is able to be present on site at dawn or dusk, and is familiar
with the call notes and song of Bicknell’s Thrushes. Especially during
this time period, we recommend use of playback recordings of songs and
calls as a means of eliciting vocal response.

Our vocalization studies failed to show any strong effects of varying
weather conditions on singing or calling behavior. Bicknell’s Thrushes
vocalized consistently in all but the most severe weather during early and
mid June. We thus found no evidence to support Wallace’s (1939) asser-
tion that “this mist-loving bird has acquired a well-founded reputation
for singing more in damp or rainy weather than on clear, sunshiny days.”
We do, however, recommend that field observers avoid conditions of
strong winds (>15-20 knots), moderate to heavy precipitation, and cold
temperatures, especially in combination. Under such conditions, Bick-
nell’s Thrush vocalize less frequently, and vocalizations are more difficult
to detect.

Although Wallace (1939) suggested that Bicknell’s Thrush begins its
southward, fall migration in mid to late September, the extent of post-
breeding dispersal prior to the onset of actual migration is unknown. At
the present time, we tentatively suggest that records obtained in late Au-
gust and early-mid September in areas of suitable breeding habitat prob-
ably represent birds on breeding or natal sites. JJowever, such observa-

646

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

tions may also include wandering individuals, and therefore may be of
less value as documentation of breeding distribution than records obtained
earlier in the summer.

We believe that surveys directed toward actual population censuses of
Bicknell’s Thrushes throughout their breeding range are impractical, given
the extensive areas which must be visited and the difficulties associated
with field work on these sites. Instead, we recommend that future popu-
lation estimates be based on (a) calculation, using satellite or aerial pho-
tography, of the extent of suitable habitat, and (b) density estimates ob-
tained from representative sites selected throughout the species’ breeding
range. We further recommend that a standardized monitoring program be
established on a subset of northeastern United States peaks for detection
of population trends. This scheme should include sites located throughout
the species’ range, characterized by varying amounts of subalpine spruce-
fir habitat and with differing degrees of isolation from other high elevation
areas.

Line transects and point counts both underestimated the densities of
Bicknell’s Thrushes as compared with the maximum density estimate cal-
culated from spot-mapping of territorial males. Point counts conducted in
the morning and evening without playback elicitation yielded results that
best approximated the minimum density estimate obtained through spot-
mapping. Point counts that made use of playback recordings frequently
resulted in inflated density estimates relative to the minimum value ob-
tained through spot-mapping. Our results indicated that fixed-width line
transects were the least effective method of estimating the breeding den-
sity of Bicknell’s Thrushes.

Because we were attempting to evaluate techniques that could provide
consistent results even when used by relatively inexperienced observers,
we rejected the variable-width transect and variable-width circular plot
survey methods that have generally been superior to fixed-width tech-
niques (DeSante 1981; Edwards et al. 1981; Franzreb 1981a, b). Variable-
width methods require observers to estimate distances to unseen, vocal-
izing birds; we found the volume of Bicknell’s Thrush vocalizations to
be so variable that even trained observers were sometimes unable to es-
timate accurately distances beyond approximately 30 m. We believe that
unlimited distance point counts (e.g., Blondel et al. 1981) provide the
most feasible means to obtain abundance indices and assess population
trends of Bicknell’s Thrushes on a rangewide basis. These types of count
eliminate difficulties of distance estimation, are much less labor intensive
than spot mapping, and are well suited for use in a single habitat type
over multiple years (Blondel et al. 1981).

Based on our field experience with Bicknell’s Thrushes, we believe that

Rimmer et al. • BICKNELL’S THRUSH SURVEY

647

use of tape recorded playbacks may increase the efficiency of detecting
the species, when present, and may lead to more accurate counts, as sug-
gested for other species by Marion et al. (1981) and Falls (1981). We
recommend that future censuses of Bicknell’s Thrushes be conducted as
standardized series of unlimited distance point counts. Counts at each
point should include an initial 5- or 10-min period of listening, followed
by a 1-min taped broadcast of songs and calls and an additional 4-min
listening period. Thrushes detected during both portions of each point
count should be tallied separately, to compare the relative efficiency of
both methods and to allow comparability with other censuses that may
not employ tapes.

The density estimates we calculated from spot-mapping data on Mt.
Mansfield in 1992 to 1995 were higher than those reported frorri two
other similar studies of this species in New Hampshire. Morse (1979)
recorded Bicknell’s Thrush densities of 22 pairs/40 ha on Mt. Osceola,
while Sabo (1980) estimated only 4 pairs/40 ha on Mt. Moosilaukee.
However, our data may not be strictly comparable to those of other studies
due to differences in sampling frequency and technique (Rimmer, unpubl.
data), seasonal timing of censuses (Morse 1979), plot size (Morse 1979,
Sabo 1980), and elevational and habitat gradients over which sampling
occurred (Morse 1979; Sabo 1980; Rimmer and McFarland, unpubl. data).
We believe that our spot-mapping data, from which two independent ob-
servers calculated similar numbers of territories in 1992-1995, closely
approximated actual Bicknell’s Thrush densities on the study plot. Mar-
chant (1981) reported that edge-effects in spot-mapping studies tended to
inflate density estimates by incorrectly including 10-27% of “edge clus-
ters” that do not strictly belong to the plot. Discounting 27% of partial
territories in this study results in maximum estimates of 48 pairs/40 ha
in 1992, 57 pairs/40 ha in 1993, 62 pairs/40 ha in 1994, and 55 pairs/40
ha in 1995. We believe these to be more accurate than our minimum
values calculated by excluding all partial territories.

Wallace (1939) reported that Bicknell’s Thrush territories on Mt. Mans-
field “may apparently cover an acre or more”. Assuming densely-packed
territories of about 0.6 ha (1.5 ac) in size, Wallace’s suggestion would
yield density estimates of approximately 65 pairs/40 ha. This is similar
to our maximum estimates and may reflect unusually high densities on
Mt. Mansfield, which we believe supports >250 pairs. Further study of
Bicknell’s Thrush territorial behavior, based on observation of color-band-
ed or radio-marked birds, is clearly warranted, as are studies of possible
geographic differences in breeding density and habitat selection.

ACKNOWLEDGMENTS

We thank Dan Lambert, Karin Fischer, Andrew Ingersoll, and Jamie Christian for their
invaluable field assistance, and Jim Goetz for assistance with data transcription. Rick Paradis

648

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

of the University of Vermont, William Kemp and Rob Apple of the Mt. Mansfield Co., and
the summit caretakers of the Green Mountain Club all provided logistical support. Trevor
Lloyd-Evans independently plotted spot-mapping data for each year. Joe Marshall and Henri
Ouellet provided useful discussions of Bicknell’s Thrush biology. We thank Henri Ouellet,
Walter Ellison, and an anonymous reviewer for helpful comments on the manuscript. This
project was supported financially by the U.S. Fish and Wildlife Service, the National Fish
and Wildlife Foundation, the William P. Wharton Trust, the Vermont Monitoring Coopera-
tive, and the members and trustees of the Vermont Institute of Natural Science and Manomet
Observatory for Conservation Sciences.

LITERATURE CITED

American Ornithologists’ Union. 1995. Fortieth supplement to the American Ornithol-
ogists’ Union check-list of North American birds. Auk 112:819-830.

Anderson, S. H. 1970. The avifaunal composition of Oregon white oak stands. Condor
72:417-423.

AND H. H. Shugart. 1974. Habitat selection of breeding birds in an east Tennessee

deciduous forest. Ecology 55:828—837.

Atwood, J. L., C. C. Rimmer, K. P. McFarland, S. H. Tsai, and L. R. Nagy. 1996.
Distribution of Bicknell’s Thrush in New England and New York. Wilson. Bull. 108:
650-661.

Blondel, J., C. Ferry, and B. Frochot. 1981. Point counts with unlimited distance. Pp.
414-420 in Estimating numbers of terrestrial birds (C. J. Ralph and J. M. Scott, eds.).
Allen Press Inc., Lawrence, Kansas.

Davenport, E. B. 1903. Birds observed on Mt. Mansfield and in Stowe Valley in 1902.
Wilson Bull. 10:77-86.

DeSante, D. F. 1981. A field test of the variable circular-plot censusing technique in a
California coastal scrub breeding bird community. Pp. 177-185 in Estimating numbers
of terrestrial birds (C. J. Ralph and J. M. Scott, eds.). Allen Press Inc., Lawrence,
Kansas.

Edwards, D. K., G. L. Dorsey, and J. A. Crawford. 1981. A comparison of three avian
census methods. Pp. 170—176 in Estimating numbers of terrestrial birds (C. J. Ralph
and J. M. Scott, eds.). Allen Press Inc., Lawrence, Kansas.

Emlen, j. T. 1974. An urban bird community in Tucson, Arizona: derivation, structure,
regulation. Condor 76:184-197.

Falls, J. B. 1981. Mapping territories with playback: an accurate census method for song-
birds. Pp. 86-91 in Estimating numbers of terrestrial birds (C. J. Ralph and J. M. Scott,
eds.). Allen Press Inc., Lawrence, Kansas.

Franzreb, K. F. 1981a. The determination of avian densities using the variable-strip and
fixed-width transect surveying methods. Pp. 139-145 in Estimating numbers of terres-
trial birds (C. J. Ralph and J. M. Scott, eds.). Allen Press Inc., Lawrence, Kansas.

. 1981b. A comparative analysis of territorial mapping and variable-strip transect

censusing methods. Pp. 164-169 in Estimating numbers of terrestrial birds (C. J. Ralph
and J. M. Scott, eds.). Allen Press Inc., Lawrence, Kansas.

Kendeigh, S. C. 1944. Measurement of bird populations. Ecol. Monogr. 14:67-106.
Marion, W. R., T. E. O’Meara, and D. S. Maehr. 1981. Use of playback recordings in
sampling elusive or secretive birds. Pp. 81-85 in Estimating numbers of terrestrial birds
(C. J. Ralph and J. M. Scott, eds.). Allen Press Inc., Lawrence, Kansas.

Marchant, j. H. 1981. Residual edge effects with the mapping bird census method. Pp.
488-491 in Estimating numbers of terre.strial birds (C. J. Ralph and J. M. Scott, eds.).
Allen Press Inc., Lawrence, Kansas.

Rimmer et al. • BICKNELL’S THRUSH SURVEY

649

Morse, D. H. 1979. Habitat use by the Blackpoll Warbler. Wilson Bull. 91:234-243.
OUELLET, H. 1993. Bicknell’s Thrush: taxonomic status and distribution. Wilson Bull. 105:
545-572.

Robbins, C. S. 1970. An international standard for a mapping method in bird census work
recommended by the International Bird Census Committee. Aud. Field Notes 24:722-
726.

Sabo, S. R. 1980. Niche and habitat relations in subalpine bird communities of the White
Mountains of New Hampshire. Ecol. Monogr. 50:241-259.

Wallace, G. J. 1939. Bicknell’s Thrush, its taxonomy, distribution, and life history. Proc.
Boston Soc. Nat. Hist. 41:211-402.

Williams, A. B. 1936. The composition and dynamics of a beech-maple climax community.
Ecol. Monogr 6:317-408.

Wilson Bull., 108(4), 1996, pp. 650-661

DISTRIBUTION OF BICKNELL’S THRUSH IN
NEW ENGLAND AND NEW YORK

Jonathan L. Atwood,' Christopher C. Rimmer^,

Kent R McFarland,^ Sophia H. Tsai,' and
Laura R. Nagy'-^

Abstract. — We conducted presence-absence surveys for Bicknell’s Thrush {Catharus
bicknelli) in Marine, Massachusetts, New Hampshire, New York, and Vermont during the
1992—1995 breeding seasons. The species was found at 234 sites, of which 225 (96%) were
dominated by varying mixtures of balsam fir {Abies balsamea) and red spruce {Picea rub-
ens). Ninety-one percent of the occupied sites were ^915 m (3000 ft) in elevation. Size of
occupied habitat patches was generally small; 73% of occupied areas delimited by the 915
m elevation contour were less than 1000 ha in extent. A logistic regression model using
independent variables describing vegetation, elevation, land area S:915 m located within 1
km of a site, and latitude successfully predicted thrush presence. There was no conclusive
evidence of widespread population declines of BicknelTs Thrush in the United States; we
found the species at 63 of 73 sites (86%) known to have been occupied prior to 1992.
However, the restricted breeding distribution and narrow habitat requirements of BicknelTs
Thrush in the United States suggest that it is vulnerable to habitat loss and degradation, and
that continued efforts to document the species’ status and ecology are warranted. Received
26 Jan. 1996, accepted 18 May 1996.

BicknelTs Thrush {Catharus bicknelli), until recently considered a sub-
species of the Gray-cheeked Thrush {Catharus minimus) (Ouellet 1993,
American Ornithologists’ Union 1995), breeds from southern Quebec and
the Maritime Provinces south to the higher elevations of New England
and New York (Wallace 1939, Ouellet 1993). Suitable nesting habitat of
this species in the United States has been described as dense forests of
balsam fir {Abies balsamea) and red spruce {Picea rubens) occurring near
tree-line (Wallace 1939). In Canada, the species also occurs at lower el-
evations, and has been documented in regenerating clearcuts and coastal
areas where structure of the spruce-fir habitat approximates that found in
the United States at higher elevations (Ouellet 1993; E. Nixon, unpubl.
data).

Concern recently has been raised that BicknelTs Thrush has disap-
peared from portions of its historic range, especially in Canada (J. T.
Marshall, pers comm.; E. Nixon, unpubl. data), and various factors seem
to pose likely threats to the species. In its breeding range, habitat deg-
radation caused by acid precipitation (Vogelmann 1982, Schreiber and
Newman 1988), replacement of high elevation coniferous forests by de-

' Manomet Observatory for Conservation Sciences, P.O. Box 1770, Manomet, Massachusetts 02345.

2 Vermont Institute of Natural Science, RR2 Box 532, Woodstock, Vermont 05091-9720.

’ Present address: Dept, of Biological Sciences, Univ. of Arkansas, Fayetteville, Arkansas 72701.

650

Atwood ei al. • BICKNELL’S THRUSH DISTRIBUTION

651

ciduous tree species as a result of global warming (Davis and Botkin
1985, Rodenhouse 1991), or habitat loss caused by development of ski
resorts or communications facilities could conceivably impact it. On its
wintering range, apparently restricted to the Greater Antilles (Wallace
1939, Ouellet 1993), Bicknell’s Thrush may be threatened by deforesta-
tion (Arendt 1992; Wunderle and Waide 1993; Rimmer and McFarland,
unpubl. data).

The Committee on the Status of Endangered Wildlife in Canada cur-
rently is considering a proposal to designate the bird as “threatened” in
Canada (E. Nixon, unpubl. ms), and in the United States Bicknell’s
Thrush was listed as a Category 2 candidate species under the Endangered
Species Act in 1994 (U.S. Fish and Wildlife Service 1994). Rosenberg
and Wells (1995) identified Bicknell’s Thrush as the top priority for con-
servation concern among neotropical migrant birds in the northeastern
United States.

Few empirical data exist by which to evaluate the current status of
Bicknell’s Thrush, and various aspects of the species’ life history have
caused it to be exceptionally difficult to study (Rimmer et al. 1996). In
this paper we present the results of surveys for Bicknell’s Thrush con-
ducted in Maine, Massachusetts, New Hampshire, New York, and Ver-
mont from 1992-1995. Where possible, we compare these data with his-
toric distributional information. Finally, we present a preliminary quan-
titative model for assessing the probability of occurrence of Bicknell’s
Thrush in the United States portion of its breeding range.

METHODS

We solicited survey volunteers from various sources. Cooperators were instructed to col-
lect data in a standardized manner. Each cooperator was provided with a tape recording of
Bicknell’s Thrush songs and call notes obtained from the Cornell Laboratory of Ornithol-
ogy’s Library of Natural Sounds. We emphasized censusing at locations where the bird was
previously recorded and at areas located above 915 m elevation. Three hundred eighty
localities were visited during June and early July 1992-1995, usually within 3 h of sunrise
or sunset. Bicknell’s Thrush was determined to be present on the basis of clearly identified
vocalizations or observation of a territorial bird responding to broadcast tape recordings.
Dominant habitat at each site [VEG] was subjectively categorized as (1) spruce-fir forest,
(2) mixed hardwood-coniferous forest, or (3) northern hardwood forest. Elevation above sea
level at each survey location [ELEV] was approximated to the nearest 100 ft.

Given the limited nature of these surveys and the variable levels of expertise among
observers, we cannot certainly conclude that Bicknell’s Thrushes were absent from locations
where we failed to encounter them, especially at areas that were visited only on a single
date (N = 80). However, during 213 visits to sites where Bicknell’s Thrushes were known
to occur and which were surveyed on two or more dates, the species was missed in only
20 instances (9%), suggesting that when appropriate sampling protocols are used (Rimmer
et al. 1996), even single visits are likely to detect the species when it is pre.sent.

In addition to these surveys, we also incorporated relevant data collected from 1992—

652

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

1995 as part of more intensive studies of bird populations and forest habitats in Maine (35
sites; J. M. Hagan, unpubl. data) and Vermont (15 sites; C. C. Rimmer, unpubl. data).
Although these projects did not use playbacks of Bicknell’s Thrush vocalizations, nonethe-
less we believe that the multiple visits made to each site by trained observers familiar with
the songs and calls of all local breeding species warrant inclusion of these data in our
analysis.

All surveyed localities were manually digitized from 1:24,000 USGS topographic maps
using Arcinfo GIS software, as were all 915 m (3000 ft) contour lines in the study region.
For each point in the resulting data set, we calculated an index of latitude [EAT], the amount
of high elevation land (^915 m) occurring in a 1-km diameter circle centered on the survey
point [KMl], and the amount of high elevation land occurring in a 10 km diameter circle
centered on the survey point [KMIO]. KMl was considered to reflect habitat availability in
the immediate vicinity of the survey point, while KMIO provided a representation of the
survey point’s location relative to the regional distribution of high elevation land. Addition-
ally, Arcinfo was used to calculate the area of high elevation polygons, as delineated by
the 915 m elevation contour, that were occupied by Bicknell’s Thrushes. Area estimates
were not corrected for topography.

Using the SAS procedure LOGISTIC, we performed a stepwise logistic regression (Mills
et al. 1993, Ak^akaya et al. 1995) with the response variable defined as thrush presence (1)
or absence (0). Independent variables (VEG, ELEV, LAT, KMl, and KMIO) were added
and removed at a significance level of P = 0.05. A total of 392 sites for which complete
and unambiguous data were available were used in this analysis.

We also conducted road-based surveys from 18—21 June 1993 in the north Maine woods
from T. 11 R. 17 east to Garfield Township, from T. 11 R. 13 southeast to T. 7 R. 11, and
from Garfield Township southwest to T. 9 R. 8. These surveys consisted of 294 5-min stops
spaced at intervals of approximately 1 km. Recordings of Bicknell’s Thrush vocalizations
were broadcast at each point to elicit responses from any birds that might be present. All
road-based point surveys were conducted prior to 11:00 under good weather conditions;
lakes and recent clear cuts were excluded. However, because data from these surveys were
not recorded in a geographically explicit format, we did not include these results in calcu-
lating the quantitative habitat suitability function described above.

RESULTS

A total of 430 localities (251 sites >915 m elevation, 87 sites from
610-915 m, 10 sites from 305-610 m, and 82 sites <305 m), were sur-
veyed for Bicknell’s Thrushes. The species was found at 234 of these
sites (Fig. 1). A detailed list of documented sites of its occurrence
(through 1995) is available on request from the senior author. The species
was found at 234 sites, of which 225 (96%) were dominated by varying
mixtures of balsam fir and red spruce. Ninety-one percent of the occupied
sites were >915 m (3000 ft) in elevation. Road-based surveys in the north
Maine woods failed to encounter bicknelli at any of 294 point localities.

Ninety-nine high elevation polygons, arbitrarily defined as areas de-
marcated by the 915 m elevation contour, were occupied by Bicknell’s
Thrushes. Of these, 73 (74%) were relatively small in size, being char-
acterized by <1000 ha of land >915 m elevation (Fig. 2). The mean area
of distinct high elevation polygons occupied by bicknelli was 1046 ha

Atwood et al. • BICKNELL’S THRUSH DISTRIBUTION

653

Fig. 1. (A) Documented occurrence of Bicknell’s Thrush within its United States breed-

ing range. (B, inset): Sample localities where Bicknell’s Thrush was not encountered. (C)
Location of high elevation (>approximately 900 m) land in New England and New York.

654

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

>-

U

g

a

LAND AREA > 915 ra ELEVATION (ha)

Fig. 2. Frequency distribution of area estimates of high elevation land occupied by
Bicknell’s Thrush in the United States.

(SD = 2006; range 1.5 ha (Little Bigelow Mtn., Maine)- 130,020.1 ha
(Mt. Washington, New Hampshire).

Approximately 155,295 ha of land >915 m elevation were identified
in New England and New York; areas >793 m (2600 ft) that were cat-
egorized by Miller- Weeks and Smoronk (1993) in vegetation classes like-
ly to be occupied by Bicknell’s Thrushes totalled 99,188 ha in western
Maine, the Adirondack region of New York, New Hampshire, and Ver-
mont (Table 1). Neither of these values should be construed as actual
measurements of the extent of Bicknell’s Thrush habitat. Nonetheless,
these results do suggest that the maximum land area potentially occupied
by the species in the United States is probably in the range of 100,000-
150,000 ha.

Stepwise logistic regression yielded the following index of habitat sui-
tablity:

P = 1/(1 + exp(-(3o - (P.-VEG) - (p2-ELEV)

- (Pa-KMl) - (P4-LAT))),

where P is the probability of occurrence, VEG, ELEV, KMl, and LAT
are the data values from each site, and px are the associated regression
coefficients (Table 2). Based on the goodness of fit statistics (Chi-square
for covariates: log likelihood statistic = 289.8 with 4 df, P — 0.0001;
score statistic = 221.8 with 4 df, P = 0.0001), the model is highly sig-
nificant. Sites occupied by Bicknell’s Thrushes had a mean habitat suit-

Anx’ood et al. • BICKNELL’S THRUSH DISTRIBUTION

655

Table 1

Estimates of Approximate Maximum Extent of Bicknell’s Thrush Habitat in New

England and New York

Area a 915 m (ha)“

Area a 793

m (ha)*’

Total Mean

SD

Spruce-fir

Total sample

Maine

Western Maine

18,332

168

498

32,366

45,458

Other

6471

196

639

na

0

Massachusetts

302

151

180

na

0

New Hampshire

54,480

436

1572

34,154

88,632

New York

Catskills

17,305

247

530

na

0

Adirondacks

41,674

344

1050

28,393

79,205

Vermont

16,731

164

359

4275

45,858

“Estimates of land area (uncorrected for topography) a915 m (3000 ft) elevation.

'* Estimates of land area a793 m (2600 ft) elevation. Spruce-fir estimates based on values for “Spruce-Fir Slope” and
“Balsam Fir" categories provided by Miller- Weeks and Smoronk (1993). “Total Sample” = total amount of photographed
land area included in analysis by Miller- Weeks and Smoronk. na = data not available.

ability value of 0.831 (SD = 0.19); sites where we failed to document
the species had a mean value of 0.221 (SD = 0.29).

However, because these data were used to calculate the habitat function,
they cannot be considered an independent validation of the model’s pre-
dictive power. To validate use of this approach to predict occupancy of a
site by Bicknell’s Thrushes we randomly subdivided the overall data set
into two equal parts (N = 196), recalculated a new habitat suitability
function based on one part (A) and then applied the resultant model to
the second part (B) of the data. Using this smaller sample size, the vari-
ables KMl and KM 10 both failed to meet the stepwise selection criterion.

Table 2

Results of Stepwise Logistic Regression Predicting Bicknell’s Thrush Presence in

New England and New York"'

Variable

Regression coefficient

Standard error

Wald Chi-square

Probability

INTERCEPT

-36.6017

9.4148

15.11

0.0001

ELEV

0.00151

0.000325

21.59

0.0001

VEG

-1.9502

0.4069

22.97

0.0001

LAT

0.00000691

0.00000185

14.16

0.0002

KMl

0.0200

0.00678

8.72

0.003 1

“Variables listed in order of entry into model. lOKM failed to meet the Q.D-S significance level for inclusion.

656

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

PROBABILITY OF OCCUPANCY

Fig. 3. Frequency distribution of habitat suitability values (probability of occurrence)
among subset (N = 196) of total dataset, based on logistic regression model derived from
a different data subset (N = 196). Habitat suitability index varies from 0 (no probability of
occurrence) to 1 (certain probability of occurrence).

SO the model included only ELEV, VEG, and EAT as independent vari-
ables. Eighty-nine percent of sites in subset B where Bicknell’s Thrushes
were observed had predicted occupancy values >0.60 based on the model
derived from data subset A, and 85% of unoccupied sites had values
<0.60 (Fig. 3).

Two sites (Quoddy Head State Park and Boot Head, Maine) where
Bicknell’s Thrushes were reported were poorly accounted for by the over-
all regression model (Pearson residuals = 7.50 and 7.38, respectively).
Both areas are located coastally near sea level and represent the only low
elevation sites in the United States where the species was documented
during 1992-1995. The sighting from Quoddy Head State Park occurred
on 4 Jul 1993; the species was not subsequently seen at this site during
multiple visits from 1993-1995. At Boot Head, Bicknell’s Thrush was
reported on 12 Jun 1993, but later surveys during 1993, as well as mul-
tiple visits in 1994 and 1995, failed to locate the species. If correctly
identified, the two 1993 records may involve transient rather than breed-
ing individuals.

We found 98 historic (pre-1992) breeding sites for Bicknell’s Thrush
in Maine, Massachusetts, New Hampshire, New York, and Vermont
through literature review and coirespondence with active field ornithol-
ogists. Seventy-three of these localities were surveyed from 1992-1995;

Atwood et id. • BICKNELL’S THRUSH DISTRIBUTION

657

Table 3

Summary of 1992-1995 Bicknell’s Thrush Surveys at Historic (Pre-1992) Sites in

New England and New York

State

Historic sites
(total)

Historic sites
visited

Present

Absent

Maine

5

5

5

0

Massachusetts

2

2

0

2

New Hampshire

26

20

16

4

New York

33

16

16

0

Vermont

32

30

26

4

TOTAL

98

73

63

10

Bicknell’s Thrushes were documented at 63 sites (86%) (Table 3). Historic
sites where we failed to find bicknelli included: Massachusetts — Mt.
Greylock (3 visits), Saddleball Mtn. (3 visits); New Hampshire — Mt. Pe-
migewasset (1 visit), Mt. Monadnock (1 visit), Mt. Sunapee (1 visit), and
North Moat Mtn. (1 visit); Vermont — Glebe Mtn. (Magic Mtn.) (1 visit).
Green Peak (Mt. Aelous) (3 visits), Mt. Ascutney (2 visits), and Molly
Stark Mtn. (3 visits). At Mt. Greylock, the only historic locality where
Bicknell’s Thrush numbers have been recorded over an extended period
of time (29 years between 1938-1993), population estimates suggest a
gradual, long-term decline that culminated in the species’ disappearance
from the site in 1973 (Fig. 4; Veit and Petersen 1993).

DISCUSSION

In the United States, Bicknell’s Thrush regularly breeds only at the
higher elevations of Maine (730-1280 m, but see below). New Hampshire
(850-1460 m). New York (880-1430 m), and Vermont (820-1250 m).
Because its obligate habitat, subalpine spruce-fir forest (Wallace 1939,
Rimmer et al. 1996), is generally restricted to mountaintops surrounded
by large areas of northern hardwoods or mixed hardwood— conifer stands,
the distribution of bicknelli in the United States is extremely patchy at
the landscape level. Furthermore, of high elevation (>915 m) regions
known to be occupied by the species, few exceed 1000 ha in area, sug-
gesting that much of the range of Bicknell’s Thrush in the United States
is limited to relatively small fragments of suitable habitat. We did not
encounter the species in low elevation, regenerating clearcuts as has been
reported in Canada (Ouellet 1993; E. Nixon, unpubl. ms).

The status of Bicknell’s Thrush in coastal Maine is problematical. The
species has been documented at various low elevation, coastal areas of
Quebec (Ouellet 1993, Gauthier and Aubry 1995), New Brunswick (Er-

658

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Fig. 4. Decline and local extinction of Bicknell’s Thrush on Mt. Greylock, Massachu-
setts.

skine 1992), Nova Scotia (Allen 1916, Erskine 1992, Wallace 1939), and
Prince Edward Island (Erskine 1992), lending credence to the possibility
of its breeding in coastal Maine. However, we obtained only two coastal
records during our study, and both were at sites where the species could
not be confirmed during multiple subsequent visits by experienced ob-
servers. Wallace (1939) reported that Gross, Pettingill, and other orni-
thologists never detected Bicknell’s Thrush during their intensive studies
of the avifauna of coastal Maine. At the present time, we do not feel that
available data warrant inclusion of coastal Maine within the breeding
range of Bicknell’s Thrush, although more field work in this area is cer-
tainly desirable.

Lack of detailed historic data makes it difficult to evaluate whether
populations of bicknelli in the United States have declined in recent years.
The species has disappeared from its principal site of historic occurrence
in Massachusetts (Mt. Greylock), where the population numbered ap-
proximately 5-10 pairs in the early 1900s. Based solely on presence-
absence determinations, we found no clear evidence that Bicknell s
Thrush has declined in Maine, New Hampshire, New York, or Vermont.

However, although these data do not substantiate recent concerns that
Bicknell’s Thrushes may be showing serious rangewide declines in the
United States, neither have we refuted this possibility. We believe that
the distribution and ecology of the species places it in a precarious po-

Atwood et al. • BICKNELL’S THRUSH DISTRIBUTION

659

100%-.

75%-

50%-

25% -

0%

NY

(Adirondacks)

28393 ha

4275 ha

NH

(western)

34154 ha

^ MODERATE
□ HEAVY

32366 ha

STATE

Fig. 5. Condition of high elevation (>793 m) spruce-fir forests occurring in New York’s
Adirondack region, Vermont, New Hampshire, and western Maine. Based on data presented
by Miller- Weeks and Smoronk (1993). Mortality classifications; Moderate = 11—30% stand-
ing dead trees; Heavy = >30% standing dead trees. Total area of “Spruce-Fir Slope” and
“Balsam Fir” cover types occurring at elevations >793 m (2600 ft) provided above each
bar (Miller- Weeks and Smoronk 1993).

sition that warrants further monitoring efforts. Acid precipitation, global
warming, and other complex biotic and abiotic factors have been postu-
lated as potentially impacting the geographically limited, high elevation
spruce-fir ecosystem (Weiss and Millers 1988) that is required by Bick-
nell’s Thrushes breeding in the United States. In fact. Miller- Weeks and
Smoronk (1993) found, based on data collected 1985-1986, that most
areas of high elevation spruce-fir habitat in New England and the Adi-
rondack region of New York showed extensive levels of tree mortality
(Fig. 5). The causes of this spruce-fir mortality are uncertain, and no
information is available concerning its possible impacts on the biology
of Bicknell’s Thrushes or other birds that breed at high elevations. Finally,
if the species’ poorly-known wintering grounds are, in fact, restricted to
the Carribean (Wallace 1939, Ouellet 1993), then the extensive defores-
tation that has occurred in this region (Arendt 1992; Rimmer and Mc-
Farland, unpubl. data) would also be expected to adversely impact the
population.

Based on these reasonably postulated threats, Bicknell’s Thrush should
be ranked as one of the most potentially threatened species of Neotropical
migrant songbirds in the United States (Reed 1989, Rosenberg and Wells
1995). Further research should attempt to clarify details of its current

660

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

breeding and wintering distribution, calculate population sizes based on
remote sensing data combined with regionally-explicit density estimates,
evaluate levels of population interchange among birds breeding on iso-
lated mountain peaks, and assess the impacts of differences in habitat
quality on Bicknell’s Thrush occupancy and reproductive success.

ACKNOWLEDGMENTS

We especially thank the many survey and data entry volunteers who contributed their
time and energy. Although too many to list individually, this project would have been
impossible without their help. C. Darmstadt, R. Harrison, T. Nunan, W Richardson, D.
North, and N. Lamous all contributed field assistance. J. Marshall and H. Ouellet provided
useful discussion of Bicknell’s Thrush biology. Environmental Systems Research Institute,
Inc. provided Arcinfo and ArcView GIS software, and Stacie Grove patiently explained its
intricacies. This project was supported financially by the U.S. Lish and Wildlife Service,
the National Lish and Wildlife Loundation, the Wharton Trust, the Plumsock Lund, and the
trustees and members of Manomet Observatory for Conservation Sciences and the Vermont
Institute of Natural Science.

LITERATURE CITED

Akcakaya, H. R., M. A. McCarthy, and J. Pearce. 1995. Linking landscape data with
population viability analysis: management options for the Helmeted Honeyeater. Biol.
Conser. 73:169-176.

Allen, E. C. 1916. Annotated list of birds of Yarmouth and vicinity, southwestern Nova
Scotia. Nova Scotia Inst, of Sci. 14:67-95.

American Ornithologists’ Union. 1995. Lortieth supplement to the American Ornithol-
ogists’ Union Check-list of North American birds. Auk 112:819-830.

Arendt, W. j. 1992. Status of North American migrant landbirds in the Caribbean region:
a summary. Pp. 143-171 in Ecology and conservation of neotropical migrant landbirds
(J. M. Hagan and D. W. Johnston, eds.). Smithsonian Institution Press, Washington,
D.C.

Davis, M. B. and D. B. Botkin. 1985. Sensitivity of cool-temperate forests and their fossil
pollen record to rapid temperature changes. Quaternary Research 23:327-340.

Erskine, a. j. 1992. Atlas of breeding birds of the Maritime provinces. Nimbus Publishing
Ltd., Nova Scotia.

Gauthier, J. and Y. Aubry. 1995. The breeding birds of Quebec. Environment Canada.

Miller-Weeks, M. and D. Smoronk. 1993. Aerial assessment of red spruce and balsam
fir condition in the Adirondack region of New York, the Green Mountains of Vermont,
the White Mountains of New Hampshire, and the mountains of western Maine 1985-
1986. Eorest Service Northeastern Region, U.S. Dept, of Agriculture NA-TP-16-93.

Mills, L. S., R. J. Eredrickson, and B. B. Moorhead. 1993. Characteristics of old-growth
forests associated with northern Spotted Owls in Olympic National Park. J. Wildl. Man-
agement 57:315-321.

Ouellet, H. 1993. Bicknell’s Thrush: taxonomic status and distribution. Wilson Bull. 105:
545-572.

Reed, J. M. 1989. A system for ranking conservation priorities for Neotropical migrant
birds based on relative susceptibility to extinction. Pp. 524-536 in Ecology and con-
servation of neotropical migrant landbirds (J. M. Hagan and D. W. Johnston, eds.).
Smithsonian Institution Press, Washington, D.C.

Atn’ood et al. • BICKNELL’S THRUSH DISTRIBUTION

661

Rimmer, C. C., J. L. Atwood, K. P. McFarland, and L. R. Nagy. 1996. Population density,
vocal behavior, and recommended survey methods for Bicknell’s Thrush. Wilson Bull
108:639-649.

Rodenhouse, N. L. 1991. Potential effects of climatic change on a neotropical migrant
landbird. Conserv. Biol. 6:263-272.

Rosenberg, K. V. and J. V. Wells. 1995. Final job report: importance of geographic areas
to neotropical migrant birds in the Northeast. Prepared for U.S. Fish and Wildlife Ser-
vice, Region 5; July 1995.

ScHREiBER, R. K. AND J. R. Newman. 1988. Acid precipitation effects on forest habitats:
implications for wildlife. Conservation Biology 2:249-259.

U.S. Fish and Wildlife Service. 1994. Endangered and threatened wildlife and plants:
animal candidate review for listing as endangered or threatened species; proposed rule.
Fed. Reg. 59:58982-59028.

Veit, R. R. and W. R. Petersen. 1993. Birds of Massachusetts. Massachusetts Audubon
Society, Lincoln, Massachusetts.

Vogelmann, H. W. 1982. Catastrophe on Camel’s Hump. Natural History 91:8-14.

Wallace, G. J. 1939. Bicknell’s thrush, its taxonomy, distribution, and life history. Proc.
Boston Soc. Nat. Hist. 41:211-402.

Weiss, M. J. and I. Millers. 1988. Historical impacts on red spruce and balsam fir in the
northeastern U.S. Pp. 271-277 in Proceedings of the effects of atmospheric pollution
on spruce and fir in the eastern United States and the Federal Republic of Germany.
Burlington, Vermont.

WuNDERLE, J. M. AND R. B. Waide. 1993. Distribution of overwintering Nearctic migrants
in the Bahamas and Greater Antilles. Condor 95:904-933.

Wilson Bull., 108(4), 1996, pp. 662-672

MIGRATION ROUTES OF THE WESTERN SANDPIPER

Robert W. Butler,' - Francisco S. Delgado,'*

Horacio de la Cueva,'* Victor Pulido,^ and
Brett K. Sandercock^

Abstract. — We report the locations of 97 sightings of over 15,000 Western Sandpipers
(Calidris mauri) color banded in Peru, Panama, Mexico, British Columbia, and Alaska.
Ninety-five sightings were made in states and provinces along the Pacific Coast of Central
America and North America. One sandpiper banded in British Columbia and one from Peru
were seen east of the Rocky Mountains. We propose that most Western Sandpipers migrate
from Central and South American winter quarters along the Pacific Coast of North America.
We also propose that most post-breeding Western Sandpipers use a trans-Pacific route that
bypasses southeast Alaska and makes landfall in southern British Columbia. Western Sand-
pipers that spend the winter on the Atlantic coast of the USA and in the Caribbean fly a
trans-continental route beginning from the Pacific Coast of North America. Received 27
Sept. 1995, accepted I April 1996.

Most Arctic breeding shorebirds stop at several staging sites to rest and
feed while migrating to and from their winter quarters. A generally held
view is that staging sites contribute to the reproductive fitness of individ-
ual shorebirds (Davidson and Evans 1988, Alerstam and Lindstrom 1990,
Ens et al. 1994) and the degradation of these sites will result in population
declines (Davidson and Evans 1986, Davidson and Piersma 1992). As a
beginning, an assessment of this hypothesis requires a description of mi-
gratory routes. For even the most abundant shorebird species, however,
migratory routes and staging sites are not well known (Morrison 1984).

The Western Sandpiper {Calidris mauri) is the most numerous shore-
bird on the Pacific Coast of North America. Between 250,000 and one
million individuals are present on single days in San Francisco Bay, Grays
Flarbor, the deltas of the Fraser, Stikine, Fox, and Copper rivers, and
Redoubt and Kachemak bays during spring migration from mid-April
through mid-May (Page et al. 1979, Butler 1994, Gill et al. 1994, O’Reilly
and Wingfield 1995). Single-day counts at these sites are 10- to 15-fold
smaller during the southward migration from late June to mid-September
(Senner 1977, Page et al. 1979, Butler 1994). Radio-tagging studies by
Iverson et al. (1996) confirmed that many of the same individuals used

' Pacific Wildlife Research Centre, Canadian Wildlife Service, 5421 Robertson Road, RR 1, Delta, British
Columbia V4K 3N2, Canada.

2 Department of Biological Sciences, Simon Fraser Univ., Burnaby, British Columbia V5A 1S6, Canada.

3 C. Meliton Martin, N 32-94, Chitre, Herrera, Republic of Panama.

••Centro de Investigacion Cientifica y de Educacion Superior de Ensenada, km 107 Carretera, Tijuana-
Ensenada, B. C. Mexico.

* Programa de Conservacion y Desarrollo Sostenido de Humedales, Los Eucaliptos 285 Camacho, La
Molina, Lima 12, Peru.

662

Butler et al. • SANDPIPER MIGRATION

663

these sites during spring migration. Despite its abundance, the migratory
pathways of the Western Sandpiper are incompletely known. It is clear
that most Western Sandpipers migrate along the Pacific Coast of the Unit-
ed States and Canada (Gill 1979, Gill et al. 1994). Senner and Martinez
(1982) used data from regional bird accounts and recoveries of banded
birds to postulate that a small number of Western Sandpipers use an
interior route through the Great Plains east of the Rocky mountains to
and from Alaska’s North Slope and that some Western Sandpipers travel
north in spring along the Pacific Coast and return to wintering quarters
by routes that cross portions of the continental interior. Migration routes
south of the United States of America were not described.

The purpose of this paper is to report sightings of banded Western
Sandpipers in the Americas up to February 1996 and to relate them to
migration routes proposed by Senner and Martinez (1982).

STUDY AREAS AND METHODS

Peru, Panama, and Mexico represent the respective southern, central, and northern por-
tions of the winter range of the Western Sandpiper on the Pacific Coast of the Americas
(American Ornithologists’ Union 1983), whereas British Columbia is a staging site used by
large numbers of Western Sandpipers in spring and autumn (Butler et al. 1987, Butler 1994).
Most Western Sandpipers were captured in mist nets and fitted with colored darvic plastic
“flag” bands to the legs at Paracas near Pisco, Peru (13°42'S, 76°13'W), near Chitre, Re-
public of Panama (7°58'N, 80°26'W), near Ensenada, Mexico (31°52'N, 116°37'W), on
Sidney Island, BC (48°40'N, 123°30'W), and at Safety Sound near Nome, AK (64°20'N,
164°56'W) (Figs. 1, 2).

About 300 Western Sandpipers were banded with yellow flags in Peru between December
1990 and May 1991, about 14,000 were banded with red over white flags in Panama between
January 1989 and April 1995, 100 were banded with red over yellow flags in Mexico in
February and March 1993, 553 were banded with white flags in British Columbia in July
and August 1990—1994, and 1271 were banded with green flags in Alaska in June and July
1994-1995. We relied on birdwatchers and biologists to send us sightings of banded birds,
which created a bias in the reporting locations since sighting rates will depend on the
intensity of search. Ten juveniles had miniature (0.8 g) radio transmitters glued to the feath-
ers of the back (Warnock and Warnock 1993) near Nome, Alaska on 11 July 1994. Daily
scans with a hand-held Yagi antenna and radio receiver were made every kilometer along
dykes near the high tide line on the Fraser River delta, British Columbia from 14 July to
10 August 1994. The search area was along Boundary Bay and southern Roberts Bank
where large numbers of Western Sandpipers roost during high tide (Butler 1994).

There is the possibility that some birds that lost a flag were identified with the wrong
country. We received two reports of birds with only red flags out of about 14,000 that were
banded. Red flags are the designated color for Chile where Western Sandpipers are very
rare (Modinger et al. 1986), and none has been banded (M. Sallaberry, pers. comm.). These
also could have been Mexican-banded sandpipers that lost a yellow flag. Sightings of yellow
flags from Peru were made before red and yellow flags were used in Mexico. Some white
flags attached to sandpipers in Panama had become stained yellow when the birds were
recaptured a few years later. All birds in Canada were banded with white flags and colored

664

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Fig. 1. Location of sightings of Western Sandpipers banded in Peru, Mexico, British
Columbia, and Alaska up to 28 February 1996. Some sightings are omitted for clarity. Stars
indicate banding site, squares indicate locations of sightings during winter (Oct.-Feb.), tri-
angles are during northward migration (March-May), and circles are during southward mi-
gration ( July-Sept.).

Butler et al. • SANDPIPER MIGRATION

665

Fig. 2. Location of Western Sandpipers banded in Panama up to 28 February 1996.
Symbols as in Fig. 1 .

666

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 1

Number of Sightings and Recoveries by 28 February 1996 of Western Sandpipers
Banded at Five Sites Along the Pacific Coast

Banding

site

Peru

Panama

Mexico

CA

OR

USA

WA TX

AZ

KS

BC

AK

Total

Peru

0

0

1

0

0

1

0

0

1

2

5

Panama

0

—

2

11

7

1

0

0

0

18

16

55

Mexico

0

0

—

4

1

0

0

0

0

4

1

10

BC

0

0

0

1

0

0

0

0

1

3

1

6

Alaska

0

4

2

3

5

1

0

1

0

5

—

21

Total

0

4

4

20

13

2

1

1

1

31

20

97

bands to prevent confusion. We discarded any records when the observer was in doubt about
the color of a flag or if the country of origin was in doubt.

RESULTS

Ninety-seven sightings of flag-banded Western Sandpipers were re-
ported by 29 February 1996 (Table 1). Five sightings were from sand-
pipers banded in Peru, 55 from Panama, 10 from Mexico, six from British
Columbia, and 21 from Alaska (Table 1). No flag-banded birds from the
other banding sites were seen in Peru, four were seen in Panama, four in
Mexico, 38 in continental US, 31 in British Columbia, and 20 in Alaska
(Table 1). Two of the 97 sightings were made east of the Rocky Moun-
tains; one in Kansas and one in Texas (Table 1, Fig. 1). The Kansas record
was from a juvenile male Western Sandpiper banded on Sidney Island,
BC on 2 August 1990 and last seen there 3 August. It was subsequently
captured 2200 km to the southeast at Quivira National Wildlife Refuge,
Kansas on 25 August 1990. The Texas record was of a sandpiper banded
in Peru and seen near Corpus Christi, Texas on 16 December 1989. The
remaining 95 sightings were made west of the Sierra Madre Mountains
of Mexico, the Rocky Mountains of the US, and the Coast Range of
Canada (Figs. 1, 2).

Western Sandpipers banded in Panama were seen on the northward and
southward migrations only in Pacific Coast states and provinces of North
America (Fig. 2). The migration route through the western US and Canada
includes many coastal and inland sites along the west side of the Sierras
and Cascade Mountains from the Gulf of California to the Strait of Geor-
gia, British Columbia (Fig. 2). Sightings of marked birds away from the
coast include Salton Sea, San Leandro and China Lake, California, and
Sauvie Island and Salem, Oregon. From southern British Columbia
through Alaska, sightings were confined to coastal locations.

Butler et al. • SANDPIPER MIGRATION

667

At the close of the breeding season. Western Sandpipers gathered in
flocks on mudflats near the breeding grounds at Nome where they gained
about 3 g of mass prior to departure (B. Sandercock, unpubl. data). An
adult female and two juvenile sandpipers of unknown gender that were
equipped with radio-transmitters were detected on the Fraser River delta
28, 23 and 29 days after release, respectively. The respective average
speeds of these individuals over the 4000 km distance between Nome and
the Fraser River delta were 142, 173, and 137km/d, respectively. These
records coupled with sightings of flag-banded birds, indicate that Western
Sandpipers from at least as far north as Nome, Alaska, migrate south
along the Pacific Coast.

DISCUSSION

Many species of shorebirds use different migratory routes in spring and
autumn (Morrison 1984, Alerstam 1990, Gratto-Trevor 1994). Censuses
of shorebirds in the western US indicate that nine sites hold over 100,000
Western Sandpipers on a single day during spring migration. They are
San Francisco Bay and Humboldt Bay, California, Grays Harbor, Wash-
ington, the Fraser River delta, British Columbia, and the Stikine, Copper
river deltas. Redoubt Bay, Kachemak Bay, Alaska (G. W. Page, W. D.
Shuford, J. E. Kjelmyr, and L. E. Stenzel, unpubl. data). On the southward
migration, flocks are much smaller at all sites and most notably in south-
central and southeast Alaska, which is largely avoided by Western Sand-
pipers (Senner 1977; R. Gill, C. Iverson, and M. A. Bishop, pers. comm.;
this study). This contrasts with sites in southern British Columbia and
many western states in the US where many more sites are used in autumn
than in spring migration (Paulson 1993).

The migration of the Western Sandpiper is more compressed in time
and space in spring than in fall. Many of the large mudflats along the
Pacific Coast of North America support larger numbers of Western Sand-
pipers on the northward migration than on the southward migration (Page
et al. 1979; Butler 1994; Senner 1977; G. West, pers. comm.), and many
small mudflats avoided in spring are used by southbound migrants. As a
result, population censuses on large mudflats are greater during the north-
ward migration than the southward migration. This phenomenon con-
founds our understanding of the southward migration. Gill (1979) re-
ported two sightings of Western Sandpipers marked at Nelson Lagoon in
the northern Gulf of Alaska and Prince William Sound and two sightings
of birds marked on the Yukon River delta at Nelson Lagoon. He proposed
that the migration route of juvenile Western Sandpipers was south along
the Alaska and British Columbia coast. However, some Western Sand-
pipers might also make a trans-Pacific flight from western and southern

668

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Alaska that bypasses southeast Alaska and makes landfall in southern
British Columbia. Large numbers of post-breeding Western Sandpipers
gather on intertidal mudflats near their breeding grounds in western Alas-
ka and on the Alaskan Peninsula (Gill and Jorgenson 1979, Connors 1984,
Gill and Handel 1990, this study). Individuals caught at this time near
Nome were about 3 g heavier than their breeding masses (B. Sandercock,
unpubl. data), and mass gains occurred among post-breeding Western
Sandpipers on the Alaskan Peninsula (R. E. Gill, Jr., pers. comm.), sug-
gesting they were preparing for a long flight. Banded individuals on the
southward migration were seen in southern Alaska and on the Alaskan
Peninsula but not southeastern Alaska (Gill 1979; Figs. 1, 2). The species
occurs in very small numbers in southeast Alaska at this time (M. A.
Bishop, R. E. Gill, C. Iverson, pers. comm.). Maximum single day cen-
suses of adults and juveniles using the Fraser River delta during their
southward migration were about 21,000 and 45,000 birds, respectively,
which is far below the 500,000 individuals counted in spring (Butler
1994), but many other coastal and interior sites are also used mostly only
during the southward migration.

Senner and Martinez (1982) proposed that most Western Sandpipers
used a Pacific Coast route to and from the breeding grounds in western
Alaska. Our study confirms their hypothesis for the Pacific route and
extends it over 8000 km south of California to Peru (Table 1, Figs. 1, 2).
Western Sandpipers banded in Peru, Panama, and northern Mexico were
sighted along the Pacific Coast of the Americas from northern Mexico to
southwestern Alaska (Figs. 1, 2).

Evidence for the existence of a Great Plains route beginning in northern
Alaska (Senner and Martinez 1982) is less convincing than the Pacific
Coast route. Individuals banded near Nome and Kotzebue, Alaska have
been sighted and recovered along the Pacific Coast (Fig. 1; D. Schamel,
pers. comm.), but the route taken by populations that breed along the
North Slope of Alaska are unknown. The species is considered as a rare
breeding species in northern Alaska (Johnson and Herter 1989, Wilson
1994), and small numbers of Western Sandpipers have been seen in spring
and fall migration on the northeast coast of Alaska and western Canadian
arctic (Martell et al. 1984, Johnson and Herter 1989). The species is rarely
seen in Alberta. Birdwatchers there recorded an individual on 20 May
1990, four individuals on 2 July 1993, and one individual on 6 May 1995
at Beaverhill Lake Bird Observatory near Edmonton between 1986 and
1995 (G. Holroyd and D. Ross, pers. comm.). None has been caught in
James Bay, the Canadian Prairies, or North Dakota despite extensive
banding there (D. Lank, R. I. G. Morrison, C. L. Gratto-Trevor and G.
Beyersbergen, pers. comm.). This evidence suggests that the Canadian

Butler et al. • SANDPIPER MIGRATION

669

prairies and adjacent parts of the US are north of the main migration
route across the North American continent or that Western Sandpipers fly
over the Great Plains on their way to and from northern Alaska.

A diagonal migration across the North American continent proposed
by Senner and Martinez (1982) is the most plausible route for populations
that spend the winter along the Atlantic and Caribbean coasts. This route
was supported by recoveries and from sightings of banded sandpipers in
this study. The most probable migration route occurs along a wide front
from southern Alaska and the north Pacific Coast through continental US.
The Western Sandpiper is an abundant autumn shorebird species on the
south coast and in the southern interior of British Columbia, but it is
uncommon farther north in the interior of the province (Campbell et al.
1990). In Idaho, the species is widespread in fall but not in spring (Oring
1962, Taylor et al. 1992), and it is abundant in mid- western US (Martinez
1979, Neil 1992, Skagen and Knopf 1993). The species is a regular winter
resident on the southern Atlantic Coast of the USA and in the Caribbean
(Wilson 1994, Rice 1995).

The migration routes of Western Sandpipers through Central America
and northern South America are poorly known. Hundreds of thousands
of Western Sandpipers spend the winter on the Pacific coasts of Mexico
(Morrison et al. 1992, 1994) and Panama (Butler et al. 1992b). We assume
that the migration route south of Mexico follows the Pacific Coast but
records are lacking. The sighting of a Peruvian banded bird in Texas in
December is enigmatic, given that the species exhibits strong winter site
fidelity (Smith and Stiles 1979, Rice 1995).

Evidence to support Wilson’s (1994) contention that migration is more
leisurely in fall than in spring is equivocal. Fifty-eight adult spring mi-
grants carrying 0.8 g mass radio transmitters flew the 3200 km distance
between San Francisco Bay and Copper River delta at an average speed
of 422 km/d (Iverson et al. 1996). This is 1. 3-3.0 times faster than the
speed we found for autumn migrants flying 4000 km between Nome and
the Fraser River delta. However, some of this time might have been spent
accumulating mass in preparation for migration. Once migration was un-
derway, individuals appeared to travel at similar speeds: the length of
stay at Pacific Coast staging sites is about three days in spring and autumn
(Butler et al. 1987, Iverson et al. 1996).

The trans-Pacific flight of the Western Sandpiper likely evolved during
the last glacial period about 15,000 BP. During the height of glaciation,
sea level along the Pacific Coast was about 100 m lower than today. Much
of the present day continental shelf off Washington, Oregon, and Cali-
fornia was not covered by water. In Alaska, Beringia extended over 1000
km wide in places, providing tundra habitats for many shorebirds. Be-

670

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

tween the breeding grounds in Beringia and the beaches of continental
US lay an immense sheet of ice requiring flights of about 3000 km by
shorebirds migrating to winter quarters in Central and South America.
The river deltas in British Columbia and Alaska are less than 9000 years
old, and so their use by Western Sandpipers is a relatively recent phe-
nomenon. It is unclear why shorebirds use these deltas in spring but not
fall but the answer may lie in the nature of favorable winds for migration.
In spring, winds along the British Columbia and Alaska coasts are fre-
quently from the southeast and in the direction of the migration, whereas
late summer winds are generally out of the northwest (Favorite et al.
1976). We hypothesize that Western Sandpipers migrate northward along
the coast and southward across the Pacific to take advantage of favorable
winds.

ACKNOWLEDGMENTS

Many people assisted us while banding sandpipers, including S. Cullen, S. Gonzalez, L.
Imbeau, N. Imbeault, G. Kaiser, M. Lemon, P. O’Hara, K. O’Reilly, Y. Sandoval, C. Schmidt,
C. Schuppli, P. Shepherd, C. St. Pierre, T Sullivan, N. Warnock, and S. Wyshynski. We also
thank the many people who sent us their sightings of flag-banded birds and observations,
including C. Gratto-Trevor, G. Holroyd, J. Lyons, D. Ross, N. Tsipora, and staff at Manomet
and Point Reyes bird observatories. Gary Page generously allowed us to use shorebird
census data of California, Oregon, and Washington between 1988 and 1992 belonging to
Point Reyes Bird Observatory. Bob Gill and Doug Schamel shared their knowledge of the
species in Alaska. Our field studies in Nome and Sidney Island, British Columbia, were
permitted by the Sitnasauk Native Corporation and BC Ministry of Parks, respectively. The
manuscript was improved by comments from R. E: Gill, C. Hitchcock, D. Lank, G. Mor-
rison, N. Warnock, and an anonymous referee. This project was funded by the Canadian
Wildlife Service (CWS), CWS Latin American Program, CWS/NSERC Chair in Wildlife
Ecology at Simon Eraser Univ. (SEU), graduate fellowships from Queens’ Univ. and SFU,
Northern Studies Training Program, Frank M. Chapman, John K. Cooper, and Jennifer Rob-
inson memorial awards, and Centro de Investigacion Cientifica y de Educacion Superior de
Ensenada.

LITERATURE CITED

Alerstam, T. 1990. Bird migration. Cambridge Univ. Press, Cambridge, U.K.

AND A. Lindstrom. 1990. Optimal bird migration; the relative importance of time,

energy and safety. Pp. 331-351 in Bird migration (E. Gwinner, ed.). Springer, Berlin,
Germany.

American Ornithologists’ Union. 1983. Check-list of North American birds, 6th ed. A.
O. U., Washington, D.C.

Butler, R. W. 1994. Distribution and abundance of Western Sandpipers, Dunlins and
Black-bellied Plovers in the Fraser River estuary. Pp. 18-23 in Abundance and distri-
bution of birds in estuaries in the Strait of Georgia (R. W. Butler and K. Vermeer, eds.).
Can. Wildl. Serv. Occas. Pap. No. 83, Ottawa, Canada.

, G. W. Kaiser, and G. E. J. Smith. 1987. Migration, chronology, length of stay,

sex ratio, and weight of Western Sandpipers, (Calidris mauri) on the south coast of
British Columbia. J. Field Ornithol. 58:103-111.

Butler et al. • SANDPIPER MIGRATION

671

R. I. G. Morrison, and F. S. Delgado. 1992b. The distribution of fish-eating,
wading, and raptorial birds in the Gulf of Panama, October 1991. Can. Wildl. Serv.
Progr. Notes No. 198, Ottawa, Canada.

Campbell, R. W., N. K. Dawe, I. McTaggart-Cowan, J. M. Cooper, G. W. Kaiser, and
M. C. E. McNall. 1990. The birds of British Columbia. Roy. Brit. Col. Mus., Victoria.

Connors, P. G. 1984. Ecology of shorebirds in the Alaskan Beaufort littoral zone. Pp. 403—
416 in The Alaskan Beaufort Sea (P. W. Barnes, D. M. Schell, and E. Reimnitz, eds.).
Academic Press, New York.

Davidson, N. C. and P. R. Evans. 1986. The role and potential of man-made and man-
modified wetlands in the enhancement of the survival of overwintering shorebirds. Col.
Waterbirds 9:176-188.

AND . 1988. Prebreeding accumulation fat and muscle protein by arctic-

breeding shorebirds. Proc. Internat. Ornithol. Congr. 19:342-352.

AND T. PiERSMA. 1992. The migration of Knots: conservation needs and implica-
tions. Wader Study Grp. Bull. 64, Suppl: 198-209.

Ens, B. J., T. Piersma, and J. M. Tinbergen. 1994. Towards predictive models of bird
migration schedules: theoretical and empirical bottlenecks. Netherl. Instit. Ocean Re-
search Rep. 1994-5, Texel, Netherlands.

Favorite, E, A. J. Dodimead, and K. Nasu. 1976. Oceanography of the subarctic Pacific
region, 1960-71. Internat. North Pac. Fish. Comm. Bull. 33:187.

Gill, R. E. Jr. 1979. Shorebird studies in western Alaska, 1976-1978. Wader Study Grp.
Bull. 25:37-40.

AND C. M. Handel. 1990. Tbe importance of subarctic intertidal habitats to shore-

birds: a study of the central Yukon Kuskokwim delta, Alaska. Condor 92:709-725.

AND P. D. Jorgenson. 1979. A preliminary assessment of timing and migration of

shorebirds along tbe northcentral Alaska Peninsula. Stud. Avian Biol. 2:113-123.

R. W. Butler, T. Mundkur, P. S. Tomkovich, and C. M. Handel. 1994. Conser-
vation of North Pacific shorebirds. Trans. 59th N. Am. Wildl. and Natur. Resour. Conf.,
Washington D.C. Pp. 63-78.

Gratto-Trevor, C. L. 1994. Confirmation of elliptical migration in a population of Semi-
palmated Sandpipers. Wilson Bull. 106:78—90.

Iverson, G. C., S. E. Warnock, R. W. Butler, M. A. Bishop, and N. Warnock. 1996.
Spring migration of Western Sandpipers (Calidris niauri) along the Pacific Coast of
North America: a telemetry study. Condor 98:10-21.

Johnson, S. R. and D. R. Herter. 1989. The birds of the Beaufort Sea. BP Exploration
(Alaska) Inc., Anchorage, Alaska.

Martell, a. M., D. M. Dickinson, and L. M. Casselman. 1984. Wildlife of the Mackenzie
Delta region. Boreal Instit. North. Studies Occas. Pap. No. 15, Univ. Alberta, Edmonton,
Canada.

Martinez, E. E 1979. Shorebird banding at Cheyenne Bottoms Waterfowl Management
Area. Wader Study Grp. Bull. 25:41-42.

Modinger, B. a., G. M. Holman, and M. B. Morales. 1986. Gula de campo de las aves
de Chile. Editorial Universitaria, Santiago, Chile.

Morrison, R. I. G. 1984. Migration .systems of some New World shorebirds. Pp. 125-202
in Behavior of marine animals, Vol. 6. (J. Burger and B. L. Olla, eds.). Plenum Press,
New York, New York.

R. K. Ross, AND J. Guzman. 1994. Aerial surveys of Nearctic shorebirds wintering

in Mexico: preliminary results of surveys of the southern half of the Pacific coast states
of Chiapas to Sinaloa. Can. Wildl. Serv. Progr. Notes No. 206, Ottawa, Canada.

672

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

, , AND S. Torres M. 1992. Aerial surveys of Nearctic shorebirds in Mexico:

some preliminary results. Can. Wildl. Serv. Progr. Notes No. 201, Ottawa, Canada.

Neil, R. L. 1992. Recent trends in shorebird migration for north-central Texas. South-
western Nat. 37:87-88.

O’Reilly, K. M. and J. C. Wingfield. 1995. Spring and autumn migration in Arctic shore-
birds: same distance, different strategies. Amer. Zool. 35:222-233.

Oring, L. W. 1962. Observations of the birds of southeastern Idaho. Murrelet 43:40-50.

Page, G. W., L. E. Stenzel, and C. M. Wolfe. 1979. Aspects of occurrence of shorebirds
on a central California estuary. Pp. 15-32 in Shorebirds in marine environments (L A.
Pitelka, ed.). Stud. Avian Biol. No. 2, Cooper Omithol. Soc., Allen Press.

Paulson, D. 1993. Shorebirds of the Pacific Northwest. Univ. Wash. Press, Seattle, Wash-
ington.

Rice, S. M. 1995. Residency rates, annual return rates and population estimates of Semi-
palmated and Western Sandpipers at the Cabo Rojo Salt flats, Puerto Rico. Unpubl.
MS. thesis, Univ. Puerto Rico.

Senner, S. E. 1977. The ecology of Western Sandpipers and Dunlins during spring migra-
tion through the Copper-Bering River Delta system, Alaska. Unpubl. MSc. thesis, Univ.
Alaska, Eairbanks.

AND E. E Martinez. 1982. A review of Western Sandpiper migration in interior

North America. Southwest Natural. 27:149-159.

Skagen, S. K. and E L. Knopf. 1993. Toward conservation of midcontinental shorebird
migrations. Conserv. Biol. 7:533-541.

Smith, S. M. and E G. Stiles. 1979. Banding studies of migrating shorebirds in north-
western Cost Rica. Pp. 41-47 in Shorebirds in marine environments (E A. Pitelka, ed.).
Stud. Avian Biol. No. 2, Cooper Ornithol. Soc. Allen Press.

Taylor, D. M., C. H. Trost, and B. Jamison. 1992. Abundance and chronology of migrant
shorebirds in Idaho. Western Birds 23:49-78.

Warnock, N. and S. E. Warnock. 1993. Attachment of radio-transmitters to sandpipers:
review and methods. Wader Study Grp. Bull. 70:28—30.

Wilson, H. E. 1994. The Western Sandpiper, Calidris mauri. The birds of North America,
No. 90 (A. Poole and E Gill, eds.). Philadelphia Acad. Natur. Sci. and Amer. Ornithol.
Union.

Wilson Bull., 108(4), 1996, pp. 673-684

BREEDING BEHAVIOR AND REPRODUCTIVE
SUCCESS OF CERULEAN WARBLERS IN
SOUTHEASTERN ONTARIO

Catherine J. Oliarnyk' and Raleigh J. Robertson'

Abstract. — Little is known about the breeding biology of the Cerulean Warbler (Den-
droica cerulea), a species declining throughout much of its range. However, life history
information can provide important insight into the vulnerability of a rare species to habitat
disturbance. We studied the breeding behavior of Cerulean Warblers at three different sites
in southeastern Ontario through the breeding season, from early May to late July 1995, and
at one of the above sites in 1994. Twenty-seven nests, including three renests, were located
within 27 territories. Average territory size was 1.04 ha (N = 18). Although both males and
females displayed at potential nest sites, building was performed only by females and took
five to six days. Nest trees were predominantly sugar maple {Acer saccharum) or oak (Ouer-
cus spp.), with an average height of 17.7 m and an average diameter at breast height (DBH)
of 40.2 cm. Average nest height was 1 1.8 m. There were no incidents of brood parasitism
by Brown-headed Cowbirds (Molothrus ater) and nest loss due to predation was low (14%).
Incubation, performed only by females, lasted 11 to 12 days. Clutch size ranged from two
to five eggs, with a modal clutch size of 5 (N = 6). During the 10 to 11 day nestling period,
both males and females fed the chicks equally. Reproductive success in this population was
high in both years. Twenty of the 27 pairs successfully fledged young, with a mean of 3.2
fledglings per successful nest. Record increases, and the high reproductive success of Ce-
rulean Warblers in the Frontenac Axis area of the Canadian Shield, may be the result of
reforestation of agricultural lands abandoned at the turn of the century. Received 28 Nov.
1995, accepted 15 April 1996.

The Cerulean Warbler (Dendroica cerulea), like many other Neotrop-
ical migrants, is a species of conservation concern. Robbins et al. (1992)
found that between 1966 and 1988, Cerulean Warblers declined at an
average rate of 3.4% per year, the greatest decline for any species of
warbler during that period. Currently listed as vulnerable in Canada
(McCracken 1992), its dependence on large tracts of mature deciduous
forest for successful breeding may make the Cerulean Warbler especially
sensitive to continuing forest fragmentation and isolation (Adams and
Barrett 1976, Temple 1986, Robbins et al. 1992, Parker 1994). Concurrent
with this overall decline, several authors have suggested that Cerulean
Warblers are undergoing a northeastern expansion of their range as ag-
ricultural lands abandoned at the turn of the century are allowed to suc-
ceed to mature forest (Ouellet 1967, Laughlin and Kibbe 1985, Andrle
and Carroll 1988, Hamel 1992, Maurer 1994). This may be the case in
Ontario.

The northernmost portion of the Cerulean Warbler’s range occurs in

‘ Dept, of Biology, Queen’s Univ., Kingston, Ontario K7L 3N6, Canada.

673

674

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

the mixed deciduous forests from the Bruce Peninsula to eastern Ontario,
Canada, with a concentration on the Frontenac axis, at the northeast tip
of Lake Ontario (Cadman et al. 1987). Local records suggest a gradual
increase in the number of Cerulean Warblers in the Kingston area. Since
sporadic sightings began in the 1930s, the birds have been found annually
since 1961 (Weir 1989).

While interest in the Cerulean Warbler appears to be growing, details
of its life history, necessary for establishing management and conserva-
tion programs, are incomplete. In this paper, we provide descriptions of
the breeding behavior, nesting chronology, habitat use, and breeding suc-
cess of a population of Cerulean Warblers in southeastern Ontario.

METHODS

During the spring and summer of 1995, we monitored a population of Cerulean Warblers
at three 9-ha study sites near the Queen’s Univ. Biological Station (QUBS), Lake Opinicon,
Leeds/Frontenac Counties, Ontario, Canada (44°30'N; 76°23'W). The sites were selected as
areas where Cerulean Warblers had been recorded with greatest regularity and highest den-
sity based on breeding bird surveys (R. Weir, pers. comm.). In 1994, a less intensive,
preliminary study was conducted at one of the sites. Nests located in 1994 are included in
the results on nesting success and nest-tree species use; all other results are based on data
collected in 1995 alone.

The physiography of the region around QUBS consists of shallow till soils with rock
ridges and undulating topography (Chapman and Putnam 1984). The area was initially
cleared for agriculture in the early 1800s, but the amount of forest cover has increased since
the early settlement period, as a result of abandonment of farmland and reduction of timber
harvesting (Keddy 1994). The forest at all three sites is now characterised by second-growth
mixed deciduous forest, dominated by sugar maple, with components of ironwood, oak, elm,
ash, hickory, basswood, and birch (see Table 1 for scientific names of trees).

All three sites were contained within a region of extensive forest cover interspersed with
lakes, roads, and some small agricultural fields — an area which is more or less contiguous
with extensive forests on the Canadian Shield to the north without significant interruptions
of agriculture or urban development. Consequently, limits of forest patch size were not easily
determinable either from topographic maps or aerial photos, and any attempt to define forest
area would be arbitrary at best. Two of the sites were located about 500 m apart, while the
third site was separated from the other two by approximately 15 km of forest and open
water (Fig. 1). Because of small sample sizes, observations from all three sites were pooled
for analyses.

To determine territory size, territorial boundaries were mapped in early May using play-
back (Falls 1981). We used a Sony Professional Walkman in combination with a single
Sony SRS-77G, 25 watt speaker (audible to a distance of about 100 m) to broadcast the
song of an unfamiliar conspecific to a territorial male. This method stimulates territorial
males to sing, and draws them out to the boundaries of their territory. A boundary point is
established either when the bird no longer approaches or a neighbor approaches the playback
speaker. Each of the study plots consisted of a 6.25-ha grid marked with flagging tape every
25 m and represented only a sample of the total population in the area. Territories with
boundaries at lea.st 50% within a 6.25-ha grid were drawn on a map (Fig. 1). Territorial
disputes, involving bouts of countersinging or fighting, were also mapped and used to con-

Oliarnyk and Robertson • BREEDING CERULEAN WARBLERS

675

£

CN

d

r-;

NO

ON

O

q

4

+ 1

+ 1

+ 1

NO

d

(N

d

+1

a

Q 2

00

(N

CN

“ c

4

d

d

o

(N

cn

ON

m

i?

o

m

d

d

(N

d

. ^

+1

+1

+ 1

q

00

q

q

+1

tA £

(N

—

CN

rn

■ Z: 3

Q i:

4.3

2.6

O

(N

CN

q

rn

00

•O

QC

U

<N

CN

q

U

E

»ri

d

<N

ON

d

CQ

q

r-;

a:

3:

+ 1

+1

+1

4

00

CN

+1

<

CQ

Q

NO

cn

(N

m

m

CN

On

d

d

z

<

-3

D

o:

UJ

4

m

m

NO

NO

U

(N

>n

q:

0

u.

Nesl
ighl (m

d

+1

d

+l

CN

+ 1

q

00

7.0

q

00

14.0

d

+1

H

<u

ON

00

Z

U

S

UJ

d

4

— .*

u

oa

u

<

<

£X

a

rt

H

vq

NO

On

NO

c/)

E

o

q

d

LU

o

in

o

o

2

+1

+1

+ 1

d

d

d

NO

+1

ft

r" 00

CN

u>

O

'C

iq

00

ON

c/2

u

00

(N

d

d

H

c/2

UJ

H

00

cn

CN

r-

U

Z

CN

<

X

U

•2

a

3

C

C

5-

a

o

c

0

Nj

g

-5:

c

*3

C3

'/3

o

u

3

Cj

a

5}

U»5

Tree spec

Q

(n

5

jj

“5.

S

s

s

"3

9

?

-C)

s

O)

Co

o

§

>*

u

O

O

X

S

3

S

B

13

c

3

<D

E

c3

e

u

0

3

0

O

S3

O

■o

0

o

3

o

u

3

C

u

c

3

O

'0

"E

u

O

cd

00

3

(Z)

t

3

CQ

T3

U

a;

w

5

(U

E

<

>

O

ui
+ 1

676

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

N

□

E]

El

□

N

R

Fig. 1. Map of 18 Cerulean Warbler territories in 1995 at three sites (A = Hahn, B =
Bedford Mills, and C = Lindsay Lake) near Lake Opinicon, Ontario (site A was studied in
both 1994 and 1995). Outline shows the boundary of each 6.25-ha study site.

hrm borders between neighboring males. To increase confidence in the boundaries, five
males were banded with a Canadian Wildlife Service metal band and two color bands in
early June (at least one per site). Males were captured using aerial mist nets raised to a
level of approximately 10 m, accompanied by playback and model presentation. All sub-
sequent observations of these males occurred within the previously recorded boundaries.

Territories were visited at least twice a week for periods of 30 to 60 minutes to search
for nests for the duration of the breeding period, or until the nest was located. If a nest
failed, the territory was monitored for renesting. Nests were monitored every second day to
determine the length of the nest building, incubation, and nestling periods. Incubation period
was considered to be the time between the first prolonged visit to the nest by the female
(>20 min) and the first observed feeding visit. Behavior at the nest, such as feeding rates
and duration of parental visits, was recorded during 60-min nest watches between 07:00
and 08:00 EST on days 3, 7, 10, 14, 17, and 20 of the nesting cycle. On the estimated day
of fledging, nests were monitored starting at 05:00 EST in order to count the number of
chicks leaving the nest. For nests that were not clearly observable from the ground, in
addition to monitoring the nest on the day of fledging, the nest tree was climbed on day 7
or 8 of the nestling period to give a maximum estimate of the number of young likely to
fledge. In 1994, six trees were climbed during the incubation period, and this information
was used to determine clutch size. Trees were climbed using ropes, a leather climbing belt,
and occasionally, climbing spurs. Once the observer was at nest height, nest contents were
examined using an extendible mirror pole to determine brood size and degree of brood
parasitism by Brown-headed Cowbirds (Molothrus ater).

Oliarnyk and Robertson • BREEDING CERULEAN WARBLERS

677

Parasitism by Brown-headed Cowbirds has been implicated in reducing reproductive suc-
cess in certain forest songbirds (Brittingham and Temple 1983), including Cerulean Warblers
in northwestern Ohio (Mayfield 1977). To determine the density of cowbirds in the area of
the study sites, we recorded each visual or acoustic identification of Brown-headed Cowbirds
throughout the breeding season.

We recorded nest-tree species and measured five characteristics at each nest; tree height,
nest height, diameter at breast height (DBH), distance of nest from the trunk, and distance
of nest tree to the nearest canopy gap. We defined a canopy gap as an area where the height
of the vegetation changed abruptly from average canopy height to ground level, such as a
dirt road, rock outcrop, field, or open water. Heights were measured using a rangefinder and
DBH was measured using a diameter tape.

RESULTS

Territorial behavior and pair formation. — Males began to arrive in the
area during the first week of May. Before pairing, males moved around
their territories singing from high, often exposed, perches. The male song
has been described as zray, zray, zray, zree with the final zree syllable
sliding upwards slightly in pitch (Peterson 1980). Variations of this song
occurred between individuals (R. Woodward, unpubl. data). Females did
not sing, but both males and females used a soft, high-pitched, metallic
“chip” note which is used as a contact and alarm call. The chip note of
the male was shown on a sonagram to be slightly lower in frequency than
that of the female, but the difference is difficult to distinguish by ear (C.
Oliarnyk, unpubl. data).

Territories ranged from 0.38 ha to 2.4 ha, with a mean size (±SE) of
1.04 ± 0.16 ha (N = 18). Territories covered most of the area of the
study grids, and boundaries of several territories abutted one another (Fig.

1) . Males defended territories with bouts of counter-singing with neigh-
boring males at the boundaries of their territory. Eight times, physical
disputes were observed between males, in which one male would either
supplant another from a perch (N = 6) or make contact in mid-air (N =

2) .

Females arrived the second week of May, and pairs spent the first
couple of days foraging close together. During this courtship period, sev-
eral pairs were observed in short chase flights initiated by the male. Fe-
males were also sometimes involved in disputes between neighboring
pairs (N = 5).

Nest construction. — Nest-site selection and construction began one to
two days after the arrival of the female on a territory. Both males and
females were seen displaying at potential nest sites by spreading their
wings and fanning their tails in combination with nest-shaping motions.
This behavior was usually performed with the mate nearby, and often the
observing mate would replace the displaying individual and repeat the

678

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

display (N = 5). Only two copulations were observed, one at a nest, the
other about 75 m from a nest.

Once a nest site was selected, interactions between males and females
were infrequent. Usually they were seen together less than once every
half hour. Nest building was performed entirely by the female, and lasted
5-6 days, not including the egg-laying period. The range of dates for
initiation of construction for known first brood nests was 18 May to 24
May (N = 12) and for renests was 31 May to 4 June (N = 3).

Nests were open, shallow cups resembling a knot on a branch when
viewed from the ground. The exteriors were formed from bark strips,
grasses, and other plant fibres. Linings were made of spider webs, fine
bark strips and soft plant material. Lichen and birch bark “shingles”
usually covered parts of the exterior of the nest. Five nests were collected
and measured and had the following dimensions (mean ± SE): outside
diameter 6.5 ± 1.2 cm, inside diameter 5.0 ± 0.6 cm, outside depth 3.3
± 0.7 cm, and inside depth 3.0 ± 0.9 cm.

Nest location. — In 1995, 17 nests, including two renests, were located
in 15 of the 18 mapped territories. Adults were observed feeding fledg-
lings in two of the three territories where nest sites were not determined.
Therefore, we assume that nesting efforts were made in most, if not all,
of the 18 territories. One female abandoned her first nest during the early
part of the incubation period, possibly due to the presence of an Eastern
Wood-Pewee (Contopus virens) nest 1.7 m from it in the same tree. In
1994, ten nests were located, one of which was a renest. In all three
instances of renesting, the second nest was constructed using material
from the previous nest which was carried to the new site. No second
broods were found in either 1994 or 1995.

Seventy percent of nests were located within ten meters of a territorial
boundary, and over half (59%) were within thirty meters (mean distance
± SE = 33.4 ± 4.7 m, N = 27) of a canopy gap (see Fig. 1). In a random
sample of points taken from within a territory, only 23% were within 30
m of a gap. It should be noted that in only one case did a nest occur near
a gap between differing, adjacent habitat types. Most were contained with-
in patches of otherwise continuous forest. The majority of nests were
found in the lower third of the canopy (71%), had no other vegetation
between the nest and ground cover (understory gap) (59%), and were
found on the outer edge of a lateral branch (77%). The most commonly
used tree species for nesting were those with the greatest overall basal
area; sugar maple (N = 18) and oak spp. (N = 6) (Table 1). Sugar maple
was the dominant species in all three sites, and oaks were the second
most abundant species group with mean total basal areas of 199.58
m^-ha ' and 128.71 m^-ha ‘ respectively.

Oliarnyk and Robertson • BREEDING CERULEAN WARBLERS

679

Clutch size. — Clutch size ranged from two to five eggs (mean ± SE =
3.8 ± 0.22, N = 6), with five being the modal clutch size. We found no
Brown-headed Cowbird eggs in any of the six nests examined before
hatching, and no cowbird chicks being fed either during the nestling pe-
riod or after fledging. Cowbirds were recorded only two or three times
at each of the three sites and only during the period between 12 May and
12 June.

Incubation. Incubation was performed by the female and lasted be-
tween 11 (N = 5) and 12 (N = 3) days. The first incubating female was
found on 30 May, and the last day of observed incubation was 23 June.
The mean (±SE) total incubation time per hour was 50 ± 5.3 min (N =
18 h, 9 nests). Males were twice observed feeding females on the nest
during incubation (N = 20 h, 9 nests). Contact was maintained by call
notes from incubating females in response to their mate’s song. Eemales
also sometimes gave call notes when they left the nest.

Nestlings. — The nestling period lasted between 10 (N = 6) and 11 (N
= 4) days. Eemale attentiveness remained high for the first one to three
days after hatching, with the female spending a mean (±SE) of 49.2 ±
4.7 (N = 5 h, 5 nests) min per hour brooding the nestlings. Time spent
at the nest by the female declined progressively as the nestlings matured
(Fig. 2), with females spending less time brooding the nestlings and more
time standing near, or on, the edge of the nest. Males spent minimal time
at the nest, usually just long enough to feed the young. Both males and
females fed the young as soon as they hatched, with males and females
feeding approximately equally (mean feedings/h ± SE; males = 3.5 ±
1.8, females = 3.5 ± 3.7, N = 17 h, 8 nests).

On the day of fledging, activity and noise from the parents and chicks
were high, with both the male and female chipping rapidly and continu-
ously, and the chicks begging for food. Several pairs also showed alarm
at our presence for the first time on fledging day. This activity made
locating already fledged chicks relatively easy, however, direct observa-
tion of the chicks after the day of fledging was difficult. Fledglings used
all parts of the canopy, were often hidden under leaves, and their call was
very difficult to isolate. In addition, chicks responded quickly to parental
alarm calls. On two occasions, when a male gave a chip note, nestlings
ceased all begging calls and movement and did not resume until the male
began to sing again. In begging behavior the fledgling crouched down
low on the branch, fluttered its wings, and gave high-pitched, burry call
notes. All young had fledged by 4 July. One pair was observed feeding
young 19 days after fledging. Adults were observed feeding chicks within
territorial boundaries at least 10 days after fledging in 50% of the suc-
cessful territories. If parents were observed feeding fledged young within

680

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

0 5 10 15 20 25

Days After Start of Incubation

Lig. 2. Time (mean ± SE) spent at the nest by female Cerulean Warblers during the
nesting cycle, based on 35 one-hour watches at nine different nests. Time 0 is the start of
incubation and vertical dotted line indicates the start of the nestling period. Sample size is
indicated in brackets above each point.

the bounds of a territory, they were assumed to belong to the territorial
pair; however, family groups are likely quite vagile. For example, one
banded male was observed feeding a fledgling approximately 100 m from
the edge of his territory, within the territory of an unpaired male.

Nesting success. — Twenty of 27 pairs successfully fledged young over
the two years of the study. No significant difference in reproductive suc-
cess (successful fledging of at least one chick) existed between years
(Fisher Exact Test, P = 0.65, df = 1). The mean (±SE) number of young
in nests examined before fledging was 3.6 ± 0.2 (N = 8). This is slightly
higher than the number based on observations of fledging from the ground
3.2 ± 0.2 (N = 20).

Nine nests failed (five in 1995, four in 1994). Eour nests were likely
depredated. Two of these nests failed during the nestling stage and al-
though no evidence of nestlings was found, almost the entire bottom of
both nests was torn out, suggesting predation. The other two nests failed
early in the incubation stage. In all cases of suspected predation the fe-
male was probably preyed on as well. No subsequent observations of
females were recorded on these territories from the time of nest failure
to the end of the breeding season in mid-July. Males on these territories

Oharnyk and Robertson • BREEDING CERULEAN WARBLERS

681

also resumed behavior characteristic of unpaired males, such as contin-
uous singing from high, exposed perches for extended periods of time. It
is not known what preyed on these nestlings and females, but Blue Jays
{Cyanocitta cristata) and Red-shouldered hawks (Buteo lineatus) were
sometimes heard in the area. Gray squirrels (Sciurius carolinensis) were
observed in trees containing nests — at nest height. On one occasion a
black rat snake {Elaphe obsolete!) was observed in a nest tree above nest
height, approximately one hour after the nestlings had fledged.

The reason for the abandonment of the other five failed nests is un-
known although, as mentioned earlier, interspecific interference may have
played a role. Females from three of these territories renested, but the
status of the other two was not determined.

DISCUSSION

Despite increasing data demonstrating declining trends in Cerulean
Warbler numbers, this local population appears to be relatively successful.
Low instances of nest predation and brood parasitism suggest that, in this
area at least. Cerulean Warbler numbers are not limited by poor repro-
ductive success.

Based on results from our study plots. Cerulean Warbler density in this
area may be as high as 96 prs km^^, assuming uniform conditions. Across
Ontario, Cerulean Warblers were recorded in only 6% of the 10 X 10 km
breeding bird atlas squares (Cadman et al. 1987). In areas where Ceru-
leans did occur, 80% of abundance estimates are of fewer than 1 1 pairs
per atlas square. Based on these results, the density of birds around Kings-
ton appears to be much higher than in the rest of Ontario. This may be
related to the higher proportion of forested land in the area around Kings-
ton relative to the more southerly parts of Ontario. However, it should be
noted that an inconspicuous, forest interior bird such as the Cerulean
Warbler may occur over a greater range or at a higher density than is
indicated by breeding bird surveys (Cadman et al. 1987, Terborgh 1989).
The density of birds, and consequently territory size, differed between
the three sites; 1.12 prs ha~', 1.28 prs ha“', and 0.48 prs ha ' at sites A,
B, and C respectively. Hamel (unpubl. data) has suggested that Cerulean
Warblers seem to breed in loose “colonies”. Since all three sites are the
same forest type, the difference in density between the adjacent sites, B
and C, may reflect this colonial nature, with site C acting as a spillover
area for the more saturated site B.

The mean nest height of 11.8 m falls within the 5 to 18 m range of
nest heights cited in the literature (Saunders 1900, Harrison 1984). Maple,
oak, and elm were the three most commonly cited species of nest tree
across the range of the Cerulean Warbler (Saunders 1900, Bent 1963 and

682

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

references therein, Ouellet 1967, Bull 1976, Laughlin and Kibbe 1985,
Peck and James 1987, Andrle and Caroll 1988). In this study, nests were
most often located in sugar maple and oak spp., the dominant tree species
at all three sites. Since insufficient information is provided on dominant
tree species in the literature, it is not possible to determine whether Cer-
uleans use trees in proportion to occurrence, throughout their range.

The high proportion of nests associated with understory and canopy
gaps, has also been noted by other authors (Bent 1963, Harrison 1984,
Peck and James 1987). These apparent associations with gaps may be an
important component of Cerulean Warbler nest-site selection. Examina-
tion of Fig. 1, and a comparison of nest location versus random locations
within a territory, suggests that nests may be associated with gaps such
as water, paths, or open areas in greater proportion than their availability
within the territory. It is unlikely that this result is due to observer bias
in locating easily observable nests associated with gaps, since nests were
located in 83% of the territories. However, detailed analyses of habitat
data are necessary before any definitive conclusions can be made about
the role of gaps in nest-site selection.

The mean clutch size of four is the same as the mean clutch size of
nests located in southern Ontario (N = 36) (Peck and James 1987). The
modal clutch size of five, however, is a new record for Ontario (Peck and
James 1987). In 35 nests in New York state, the average clutch size was
again four eggs (51%), with nests also containing three eggs (26%) and
five eggs (23%) (Bull 1976). The lack of brood parasitism by Brown-
headed Cowbirds was noteworthy for the Ontario range. In southwestern
Ontario, the incidence of Brown-headed Cowbird parasitism of Cerulean
Warbler nests is as high as 17.9% (Peck and James 1987).

The incubation period of 11 to 12 days determined in this study is
similar to Harrison’s (1975) speculation that the incubation period is 12
to 13 days.

Cerulean Warbler habitat has been described as mature, riparian forest
(Bull 1976, Lynch 1981, Laughlin and Kibbe 1985). The area of maturing
second-growth forest, dotted with numerous lakes and streams, north of
Kingston approaches this description. Forests in this area are also com-
positionally similar to forests present before extensive post-settlement al-
teration (Keddy 1994); therefore, the recorded northeastern expansion of
the species may be a reoccupation of habitat within the species range
prior to European settlement. Unfortunately, no records of Cerulean War-
blers in this area exist prior to 1930 to confirm this theory.

The establishment and success of this northern population of Cerulean
Warblers over the past two decades is likely the result of reforestation of
areas rendered unsuitable by European colonization in the northern range

Oliarnyk and Robertson • BREEDING CERULEAN WARBLERS

683

of the Cerulean Warbler. With forested areas continuing to be cleared and
maintained for agriculture, and local populations being eliminated in the
southern portion of its breeding range, maturing second-growth forest in
the north may provide an important source area for Cerulean Warbler
populations in the future.

ACKNOWLEDGMENTS

We give special thanks to Clive Goodinson, Brent Gurd, Andre Patry, and Becky Whit-
tham for their assistance with field work; Brent Gurd, Kelvin Conrad, Paul B. Hamel, Bart
Kempenaers, Wally Rendell, and an anonymous reviewer for helpful suggestions on an
earlier version of this manuscript. This study was funded by a grant from the Eastern Ontario
Model Forest Program and a Natural Science and Engineering Research Council operating
grant to R. J. Robertson. Logistical support was provided by the Queen’s Univ. Biological
Station.

LITERATURE CITED

Adams, D. L. and G. W. Barrett. 1976. Stress effects on bird-species diversity within
mature forest ecosystems. Am. Midi. Nat. 96:179-93.

Andrle, R. E and J. R. Carroll. 1988. The atlas of breeding birds in New York State.
Cornell Univ. Press, Ithaca, New York.

Bent, A. C. 1963. Life histories of American wood warblers. U.S. Nat. Mus. Bull. 203,
Washington, D.C.

Brittingham, M. C. and S. A. Temple. 1983. Have cowbirds caused forest songbirds to
decline? BioSci. 33:31-35.

Bull, J. 1976. The Birds of New York State. Cornell Univ. Press, Ithaca, New York.

Cadman, M. D., P. E j. Eagles, and E M. Hellenier. 1987. Atlas of the breeding birds
of Ontario. Univ. of Waterloo Press, Waterloo, Ontario.

Chapman, L. J. and D. E Putnam. 1984. Physiography of southern Ontario. Ontario Geo-
logical Survey, Toronto, Ontario.

Falls, J. B. 1981. Mapping territories with playback: an accurate census method for song-
birds. Studies Avian Biol. 6:86-91.

Hamel, P. B. 1992. Cerulean Warbler, Dendroica cerulea. Pp. 385—400 in Migratory non-
game birds of management concern in the Northeast (K. J. Schneider and D. M. Pence,
eds.). U.S. Dept. Inter., Fish and Wildl. Serv., Newton Corner, Massachusetts.

Harrison, H. H. 1975. A field guide to birds’ nests. Houghton Mifflin, Boston, Massachu-
setts.

. 1984. Wood warblers’ world. Simon and Schuster, Inc. New York, New York.

Keddy, C. 1994. Forest history of eastern Ontario. Information Report No. 1. Eastern
Ontario Model Forest Program, P.O. Bag 2111, Kemptville, Ontario.

Laughlin, S. B. and D. P. Kibbe. 1985. The atlas of breeding birds of Vermont. Univ.
Press of New England, Hanover, New Hampshire.

Lynch, J. M. 1981. Status of the Cerulean Warbler in the Roanoke River basin of North
Carolina. Chat 45:29-35.

Maurer, B. A. 1994. Geographical population analysis: tools for the analysis of diversity.
Blackwell Scientific Publications, Cambridge, Massachusetts.

Mayreld, H. 1977. Brown-headed Cowbirds: agent of extermination? Am. Birds 3:1 12.

McCracken, J. D. 1992. Status report on the Cerulean Warbler (Dendroica cerulea) in
Canada. COSEWIC, Ottawa, Ontario.

684

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

OuELLET, H. 1967. The distribution of the Cerulean Warbler in the province of Quebec,
Canada. Auk 84:272—274.

Parker, T. A. 1994. Habitat, behavior and spring migration of the Cerulean Warbler in
Belize. Am. Birds 48:70-73.

Peck, G. K. and R. D. James. 1987. Breeding birds of Ontario: nidiology and distribution.
Vol. 2: Passerines. Life Sciences Miscellaneous and distribution. Vol. 2: Passerines.
Life Sciences Misc. Publications, Royal Ontario Museum, Toronto, Ontario.

Peterson, R. T. 1980. A field guide to the birds of eastern and central North America.

Houghton Mifflin Company, Boston, Massachusetts.

Robbins, C. S., J. W. Eitzpatrick, and P. B. Hamel. 1992. A warbler in trouble: Dendroica
cerulea. Pp. 549-562 in Ecology and conservation of Neotropical migrant landbirds (J.
M. Hagan, III and D. W. Johnson, eds.). Smithsonian Inst. Press, Washington, D.C.
Saunders, W. E. 1900. Nesting habits of the Cerulean Warbler. Auk 17:358-362.

Temple, S. A. 1986. Predicting impacts of habitat fragmentation on forest birds: a com-
parison of two models. Pp. 301-304 in Wildlife 2000: modeling habitat relationships
of terrestrial vertebrates (J. Verner, M. L. Morrison, and C. J. Ralph, eds.). The Univ.
of Wisconsin Press, Madison, Wisconsin.

Terborgh, j. 1989. Where have all the birds gone? Princeton Univ. Press, Princeton, New
Jersey.

Weir, R. D. 1989. Birds of the Kingston region. Kingston Field Naturalists, Kingston,
Ontario.

Wilson Bull., 108(4), 1996, pp. 685-696

GRIT-USE PATTERNS IN NORTH AMERICAN
BIRDS: THE INFLUENCE OF DIET, BODY
SIZE, AND GENDER

James P. Gionfriddo' - and Louis B. Best'

Abstract. — We investigated avian grit use by examining the gizzard contents of 1440
birds collected from 12 states. Grit was present in gizzards of 62 of 90 species and varied
greatly in number and mean particle size. Gizzards of granivorous birds contained more grit
particles than those of insectivores, omnivores, and frugivores. Grit particle characteristics
(mean size, shape, and surface texture) did not differ among birds consuming different diets.
Mean grit size increased linearly with the common logarithm of the bird body mass. Within
avian species, grit-use patterns did not differ by gender. Grit use is widespread among birds,
and diet strongly influences the amount of grit used by birds. Received 18 Oct. 1995,
accepted 10 March 1996.

Best and Gionfriddo (1991) characterized grit use by 22 species of
North American birds. The impetus for that work was the relevance of
grit use to avian mortality caused by ingestion of granular pesticides used
to control corn rootworms and other agricultural pests (Best and Fischer
1992). Birds included in that study were limited to species that commonly
use midwestern cornfields during the breeding season, when pesticides
usually are applied. The present research examined grit use by a larger
number of avian species, over a wider geographical area, and included
many birds collected during the nonbreeding season. Our objectives were
to survey grit use by a broad range of avian species and to examine the
influence of diet, body size, and gender on the amounts and characteristics
of grit used by birds.

METHODS

We obtained birds opportunistically from a variety of sources including road kills, colli-
sions with windows and other objects, hunter harvests, and research projects. Birds were
collected year-round and from 12 (mostly midwestern) states. We removed the gizzards from
all collected birds and preserved them in ethanol. Later, each gizzard was sliced in half and
its contents were flushed into a petri dish and examined under a stereomicroscope. We
separated all grit particles from the other gizzard contents and excluded particles <0.1 mm
in size because (1) the size distributions of grit in the gizzards of most species showed that
as grit size decreased toward 0. 1 mm, the number of particles per size class declined abruptly
(Best and Gionfriddo 1991), and (2) we felt that particles <0.1 mm probably repre.sented
soil ingested incidentally during foraging. We then systematically counted the grit and char-
acterized particles on the basis of size, shape, and surface texture. The longest and shortest
dimensions of each particle in the 1440 gizzards we examined were measured to the nearest

' Dept, of Animal Ecology, Iowa State Univ., Ames, Iowa 5001 I.

^ Present address: Dept, of Forestry and Natural Resources, Purdue Univ., West Lafayette, Indiana 47907.

685

686

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

0.1 mm with a digital caliper or an ocular micrometer in the microscope. We used the
average of these two values as a measure of particle size and their ratio as a shape index
value. Shape index values were >1.0, with 1.0 representing a somewhat spherical shape
and larger values representing more oblong shapes. We characterized the surface texture of
each grit particle in 499 of the gizzards by using a classification system developed by
petrologists to describe mineral grains (fig. 1 in Best and Gionfriddo 1991). The five surface-
texture categories were angular, sub-angular, sub-rounded, rounded, and well-rounded. An
overall mean surface-texture value was calculated for each bird by assigning particles in the
angular category a value of 1, those in the sub-angular category a value of 2, etc.

For comparisons of grit use based on diet, we classified each bird as an insectivore,
granivore, omnivore, frugivore, or carnivore, taking into account the month in which the
bird was collected and following the classification of DeGraaf et al. (1985). We classified
Ring-necked Pheasants (scientific names of avian species are in Table 1) collected during
the non-breeding season, however, as granivores (rather than herbivores, as suggested by
DeGraaf et al. [1985]) because most of their non-breeding season diet consists of seeds
(e.g., Dalke 1937, Ferrel et al. 1949, Korschgen 1964). Carnivores and frugivores were
excluded from some analyses because of small sample sizes.

Except for the diet-based comparisons (in which data from species with similar diets were
combined), our analyses of grit use were limited to those species for which the contents of
at least five gizzards were examined. This limitation was imposed because of the great
intraspecific variation in the number of grit particles per gizzard, a variable for which there
is only one value per gizzard. For other variables, such as grit size, shape, or surface texture,
each gizzard yielded as many values per variable as there were grit particles in the gizzard.

Bird body masses were obtained from Dunning (1993). For dimorphic species used in
intraspecific comparisons by gender (i.e., species with ^5 birds of each gender included in
the sample), we assigned the gender-specific masses given by Dunning (1993). For the other
dimorphic species, we assigned to each bird the mean of Dunning’s values for males and
females.

Analysis of variance (ANOVA) was used to identify differences in grit use among birds
consuming different types of foods. When ANOVA detected differences among dietary
groups, we used Student-Newman-Keuls (SNK) multiple comparison tests to identify spe-
cifically which groups differed. Using data only from those species for which we examined
the contents of ^5 gizzards, we calculated Pearson product-moment correlations to examine
if the mean number of grit particles per gizzard was related to grit occurrence (the percentage
of gizzards containing grit) or mean grit size. We used regression analysis to examine the
relationship between mean grit size and bird body size. Two-tailed r-tests were used to
examine if mean grit counts, sizes, shape index values, or surface-texture values differed by
gender. Unless otherwise indicated, a significance level of P ^ 0.05 was used for statistical
tests.

RESULTS

Gizzard contents of birds representing 90 species and 10 orders were
examined. Grit was present in gizzards of 62 species (9 orders). Fre-
quencies of occurrence of grit in gizzards and grit counts per gizzard
varied greatly among species. Species with low frequencies of occurrence
of grit (and low grit counts) generally were insectivores, whereas those
with high frequencies of occurrence (and high grit counts) usually were
granivores. Among the 35 avian species for which >five gizzards were

Gionfriddo and Best • GRIT USE BY BIRDS

687

examined, frequencies of grit occurrence in gizzards ranged from 0 to
100%, and mean grit counts per gizzard ranged from 0 to 281 (Table 1).
Gizzards of Ring-necked Pheasants, American Tree Sparrows, and House
Sparrows had the highest frequencies of occurrence of grit and also gen-
erally had the highest grit counts. Low frequencies of occurrence and grit
counts were recorded for Eastern Kingbirds, Cedar Waxwings, Barn Swal-
lows, Dickcissels, Common Yellowthroats, Yellow-rumped Warblers, and
Northern Orioles. Grit occurrence and mean grit counts were correlated
(r = 0.502). Intraspecific variation in grit counts was substantial: standard
deviations typically exceeded mean values (Table 1).

Mean grit counts per gizzard differed among birds consuming different
foods (/^3,i432 = 59.01, P < 0.001). Gizzards of granivores contained more
grit particles than those of insectivores, omnivores, and frugivores (SNK
test, P < 0.05) (carnivores were excluded because of small sample size).
The adjusted mean grit size (see below), mean shape index value, and
mean surface-texture value did not differ among granivores, insectivores,
and omnivores (size: 7^2 ^69 = 2.41, P = 0.090; shape: ^2,959 = 0.691, P
= 0.501; surface texture: ^2,490 = 2.455, P = 0.087). (Carnivores and
frugivores were excluded because of small sample sizes.

Regression analysis indicated that mean grit size was related to bird
body size, increasing linearly with the log^o) of the body mass (Fig. 1).
To permit examination of the relationships between mean grit size and
other variables (with the effects of bird body size partitioned out), we
adjusted the mean grit size for each species by adding the species’ residual
value to the overall mean grit size of the sample (Steel and Torrie 1980:
251). For most (22) of the 33 grit-using species for which we examined
^five gizzards, there was no detectable relationship between the adjusted
mean grit size and the number of grit particles per gizzard. In 1 1 species,
however, a significant (P ^ 0.01) negative correlation was found between
these variables.

Among the 17 species for which we examined >5 gizzards for each
gender (Table 1), mean grit counts differed (P < 0.05) intraspecifically
by gender only in Ring-necked Pheasants (t = 3.40, 31 df, P = 0.002)
and Red-headed Woodpeckers {t = 2.45, 16 df, P = 0.026). In both
species, gizzards of females contained more grit than those of males.
Gender comparisons of adjusted mean grit sizes, mean shape index values,
and mean surface-texture values detected few differences. Because sur-
face-texture values were calculated only for 499 birds, sample sizes per-
mitted gender comparisons within only five species: Northern Bobwhite,
Brown-headed Cowbird, Red-winged Blackbird, Vesper Sparrow, and
House Sparrow. In Brown-headed Cowbirds, females used larger (r =
2.97, 87 df, P = 0.004), less oblong (r = 2.72, 87 df, P = 0.008), and

Table 1

Grit Use by Wild Birds‘

688

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

o

u

q

m

q

q

q

1

q

00

—

q

q

q

q

q

q

q

1

00

-;

q

q

><

w

cn

rn

(ri

(N

rn

cn

CO

CO

(N

CO

(N

CO

CN

CO

(N

ri

CN

CO

(N

CN

CO

CN

CO

'U

V

q

00

00

q

r-

OO

00

q

q

1

q

q

q

00

CO

00

q

q

q

q

00

q

*o

C

ri

(N

c4

p— ■

CO

(N

(N

"■

CN

CN

E

q

00

O)

cn

cq

OS

CO

CN

(N

q

q

CO

CO

00

q

CO

r--

CO

m

00

q

E

(N

"*

(N

CN

d

**“

ri

d

d

d

d

d

d

d

d

d

c

2

00

o

CN

in

O

CO

o

(N

O

o

00

in

(N

o

'tt

X

o

o

r-

o

o

c^>

00

o

'S

00

m

X

in

Os

Os

CN

U

S

-rt

o

00

X

X

CO

X

CN

OS

lo

O

(n

X

OS

CO

(N

in

o

X

1^

X

in

r-

CN

00

in

o

a

Os

Tf

in

’'t

— '

*—

—

(N

00

in

"tt

00

CO

+1

+ !

+ 1

+ 1

+ 1

+1

+1

+1

+1

+ 1

+1

+ !

+ 1

+ 1

+1

+ 1

+1

+ 1

+1

+ 1

+1

o

+ 1

+1

+1

+1

OS

00

os

O

CO

in

os

Ov

os

CN

CO

—

CN

— ■

m

00

X

CO

—

CO

(N

—

CN

00

o

""

<n

CN

o

o

o

<n

O

00

(N

Os

in

CO

O

O

CO

X

00

o

o

CO

o

£

o

o

OS

Os

o

X

in

cs

(N

OS

00

in

(N

o

00

CO

in

OS

<N

o

Os

o

—

—

N E

o

5 ^

r^Tj-»noo»nooNt^ONmONO'^o

(NCNCN^OCN^CN

'sO m

r- so in Tt
Tt —

Tt —

— CN

0 0 0-^00-

^OOOp-h-OOO-O

o o o

■5

"5

c

•2

-c

Q.

C3

a>

SI

d-

T)

<u

c

OJj

5

s

•2

-5

"S

SJ

s

c

OC

.2
c

cd

C/5 dj

.ti
oj x:

Cu ^
X)
T3 O

1) CQ

0 c

D u

c a;

1 X

M t;
•E o

a: z

o

o

>

.s

'C

T3

S

a

-«

u

5

>■

S

;j

C3

g

to

2

2

2

5

2

to

2

o

o

U

c O
§ 3
^ u
D

^ u

O *0

O “

to

*2

-2

S-

2

to

Cl ^

C 2

CX) C

2 2 C'

:3 2 2

5 Co

S S 3

u. ^

u g o
u ^ c

D. V

■o i;

o -o 3;
P .i= -

■0^1
S c 00

^ c/5 U

D 03 C3

o; UJ QQ

to

lb

!3

-5

§■

5

2

CL

to

2

•2

2

2

-2

-2

2

2

-2

to

2

2

o

U

S

3

O ‘C
2

^ n
o 2

3 3

is ^

G

^ 3
-Cl b-

S ^

3 ^
CO 3

3 U
00 ^
C J=

- M

% 2

« £
> E-

to

•2

2

I

00

C -3
3 3
3 r

I-

u u

>^ w
CO o

u *u iS F
^ E u <u
CQ < U I

CO 5
3 ^

*— C("

3 3

3 3

00 JS
. S'
IJ ^
^ c

■E

« o

2 ac

H =

^ CO

4-. 3

E 'S
E

X <

,c<i

3

's c::

2
"2

tX)
2

2

2

2 ^

2

•2

2

§

“25

^ 2
.y

IS

^ "3
'S Q

^ I

E ■3'
Oo C)

it

—

u s_.

c3

C

<u

t= ^
V

a D

o ^

3 O
IJJ X

cn

L.

to

to

2

Q.

O

t

cd

CL

CO

(U w

Xi n

t- ^

If

TJ O
V ZZ
D- ^

E >-
2 =
i I

= E
<u o
>- U

3

2

3

c-i

3

O

fc

CO

a

00

^ I

i « x:
c« D. eo
D. OO C
C/0 c
00 cd
X c >
O O C3
LL. OO 00

Gionfriddo and Best • GRIT USE BY BIRDS

689

-j

CQ

<

Q

UJ

D

z

H

Z

O

u

a ^

■= n

C/3 -E

4> p

ON

On

so

00

0

On

1

—

Os

q

^ E

d

d

d

d

d

d

d

—

d

— '

c

CJ

CN

m

00

(N

0

SO

0

0

*

'■B

CnJ

0

(N

i

CN

CM

0

—

0

0

00

CM

0

0

00

CM

—

10

—

00

Os

so

cn

(N

(n

c/3

+1

+ 1

+1

+1

+1

+ 1

+1

+1

+ 1

0

+ 1

+ 1

+ 1

+1

0

CM

(N

CM

0

in

On)

0

0

so

P>M

«

(n

<N

1/

J

0

0

yr)

SO

0

in

0

m

r-

CM

sO

0

c^

00

0

CM

r-

00

r-

in

in

U

L. ^

”■

j2
5 ^

r<S CN ro ro r<S (N (N

CT\0\— <00000000
— — (N — — (N(N— ^

P q ON CT;

ro ro (N (N

p p <N O
— ’ — ’ oi (N

oooomiooofN — otNONi^m

(NtN (NfNCN 00 ■'tr^

0000000000-00

Q

— o

.>5

>3

s

o

a

c

> N/

■€ ,5.

c <0

"I

5^ t

•i.

^ D.
O C/D

o.^

on H

-c 5

c3 cQ
C o
C 'u
cd (L)

> E

03 E
00 <!

.C

C

'ij

to

=3

Cl.

s

Oc

>3

*>«

o

o

a.

3

3

I

3

3

2 -
00 Q
c<3 “

3

O

t

cd
CL

• ^ Un

D. U
D. O.
.— (/:
x: V
U >

2 ^

C3 _
^ a

C -S
a s
-a

2 ^

•3 ^

03

^ . .s
s. ^

^ Co b

fc - ^
a. ^ E

c^.S£ u
Os:
^ ^ t:

^ I

J Q Z

3

3

05

>3

3

Cl,

V

0£) ^
C .0

w,

§ o

CQ e

•2 a

^ <2
-s; 00
Cc. ^

S 3

^ I

Sc^

T3 ^
P .Oi

O:

•2 2
s -s
'> o

s 5

XI

cj

—

CO

S) -5

1 °
£ Z

■a

V

00

,S c

> OJ

CS 3

3 ^

*0
3 .b

O-g
d) o

u

T3
1)
■o
ca
u

S ^
aC >

o

u m

Si a

4> 4)

^ Q

. 00
^ <*-

ON ®
On c/3

— C

0 I
1 1

1 E
E °

O CJ

5 s
1 =

S u

ffl S
-- -o Co

M S3 r S'

-g s I 2

4i s

’5b C9 3

^ ’c5

W-l CC _ • >

M -

Sou

_c
<0 3

Is

4) a.
Q. n

crt o

J £

*" c

>. O

51

S O

■o TD
_3 4>
O «

C CO

^ O

*2 .E c
td c O

N I'i.

5b 'u
c_ r y
o 00 w

■ij •—

00 o

!9 .

rt w
(J ^

55

> o

^ -yi

S

t: E - c !::

S 3 X ca

^ c 4> > 4)

1» U dJ .2 1>

X J= « J= X

H H C/3 [— H

690

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

MEAN BODY SIZE (g)

Fig. 1. Relationship between mean grit size and mean bird body mass. Body masses
were obtained from Dunning (1993). Each point represents a species for which the contents
of at least five gizzards were examined.

more angular (t = 2.38, 43 df, P = 0.022) grit than males. Female Ring-
necked Pheasants and Horned Larks used smaller grit than males (pheas-
ants: t = 2.48, 31 df, P = 0.019; larks: r = 2.10, 28 df, P = 0.045).

DISCUSSION

The most commonly proposed function of avian grit use is the facili-
tation of mechanical grinding and pulverization of food in the gizzard
(Meinertzhagen 1964, Ziswiler and Farner 1972). Grit also may provide
supplementary minerals, especially calcium (e.g., Korschgen 1964, Norris
et al. 1975, Turner 1982). Avian grit use often varies with such factors
as the bird’s diet (e.g., Mott et al. 1972, Norris et al. 1975, Alonso 1985,
Bishton 1986), age (Bartonek 1969, Verbeek 1970, Alonso 1985, Mayoh
and Zach 1986), body size (Best and Gionfriddo 1991), and gender and

Gionfriddo and Best • GRIT USE BY BIRDS

691

reproductive status (e.g.. Harper 1964, Kopischke and Nelson 1966, May
and Braun 1973, Pinowska and Krasnicki 1985), as well as with the
availability of suitable grit particles (Bump et al. 1947, Tindall 1973,
Norris et al. 1975).

Diet is a major factor influencing avian grit use. Grit generally is found
in the gizzards of most species that eat plant material (Earner 1960, Mei-
nertzhagen 1964) and many that eat insects (e.g., Barlow et al. 1963,
Jenkinson and Mengel 1970, Barrentine 1980, Mayoh and Zach 1986).
Relatively large amounts of grit often are associated with diets consisting
of hard, coarse materials, especially seeds and other plant parts (Meiner-
tzhagen 1954, Earner 1960). Insectivores often use less grit than herbiv-
orous or granivorous birds (Mott et al. 1972, Bishton 1986, Hogstad
1988), and frugivores typically use little grit (Meinertzhagen 1954, 1964).
The amounts of grit needed by insectivores and granivores may depend
on the hardness of the items consumed. The digestion of soft-bodied
insect larvae, for example, may require relatively little grit, whereas the
breakdown of hard-bodied insects (e.g., adult coleopterans) may require
large amounts (Pinowska 1975, Gionfriddo and Best 1995). Grit-sized
hard insect fragments (e.g.. Bird and Smith 1964, Jenkinson and Mengel
1970, Mott et al. 1972) or hard seeds (e.g.. Beer and Tidyman 1942,
Sharp and McClure 1945, Lewin and Lewin 1984) are sometimes retained
in the gizzard where they serve as grit substitutes by aiding in the me-
chanical grinding of softer foods. When such grit substitutes are present
in gizzards, less grit is required. Hard insect parts or hard seeds were
present in some of the gizzards we examined, and therefore, in some
cases, our grit counts may underestimate the birds’ need for grit.

Our finding no differences in adjusted mean grit size, mean grit shape,
and mean grit surface texture among birds consuming different diets sug-
gests that different foods may not require grit with different physical
characteristics for adequate digestion. The characteristics of grit particles
in birds’ gizzards are influenced by avian preferences and aversions (Best
and Gionfriddo 1994), the availability of grit particles with different char-
acteristics (Bump et al. 1947, Tindall 1973, Norris et al. 1975), and rates
of breakdown and passage of grit from the gizzards (Lienhart 1953,
Korschgen et al. 1965, Vance 1971, Norris et al. 1975, Gionfriddo and
Best 1995).

The large amounts of grit we observed in gizzards of Ring-necked
Pheasants, American Tree Sparrows, and House Sparrows were not sur-
prising because these species feed mainly on seeds on the ground. The
relatively low grit counts and frequencies of occurrence of grit in gizzards
of Eastern Kingbirds, Barn Swallows, and Cedar Waxwings were ex-
pected because these species feed on insects or fruits, aerially or in trees.

692

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

The Other species that tended to use little or no grit (Common Yellow-
throat, Yellow-rumped Warbler, Northern Oriole) also glean insects in
trees or shrubs. The finding that Dickcissels used little grit on their North
American breeding grounds is consistent with Zimmerman’s (1963) report
that wintering Dickcissels in central America used relatively large
amounts of grit, but that breeding birds in Illinois used little. Zimmerman
(1963) attributed the difference to the Dickcissels’ heavier use of insects
in North America and seeds in central America. Field studies of other
avian species that shift diets seasonally also have documented correspond-
ing shifts in grit use (Bishton 1986, Hogstad 1988, Gionfriddo and Best
1995). Investigation of seasonal patterns of grit use by additional species
that shift diets seasonally would further clarify the relationship between
grit use and diet.

The observed log-linear relationship between mean grit size and bird
body mass is similar to that reported for 19 avian species collected mainly
in the Midwest during the breeding season (Best and Gionfriddo 1991).
Augmenting the earlier sample by including birds representing many ad-
ditional avian taxa, geographical locations, and seasons did not substan-
tially alter the linear model describing this relationship. The lack of a
strong relationship between mean grit size and mean grit count for most
(22 of 33) species we tested differs from the results of other field (Myr-
berget et al. 1975, Norris et al. 1975, Alonso 1985, Best and Gionfriddo
1991) and laboratory (McCann 1939, Smith 1960, Gionfriddo and Best
1995) studies which found that the larger the grit particles, the fewer were
present in gizzards. The current finding is consistent, however, with the
results of a diet experiment with captive House Sparrows (Gionfriddo and
Best 1995).

Grit use by males and females often is reported to be similar (e.g.,
Alonso 1985, Norman and Brown 1985, Garcher and Carroll 1991, Gion-
friddo and Best 1995). When differences exist, they sometimes are due
to differences in body size (Rajala 1958, Pulliainen 1979, Norman and
Mumford 1985) or to females’ increased calcium requirements during egg
laying (Sadler 1961, Harper 1964, Taylor 1970). Females may greatly
increase their consumption of grit during the egg-laying period (Pinowska
and Krasnicki 1985), and/or they may selectively consume calcium-rich
grit particles (Harper 1964, Korschgen 1964, Kopischke and Nelson
1966). Such reproduction-related changes in grit use (and gender differ-
ences that result from them) may be short-lived (Pinowska and Krasnicki
1985), however, and therefore probably would not have been detectable
in our study.

In addition to demonstrating the importance of diet and body size as
factors affecting avian grit use, our study shows that grit use is widespread

Gionfriddo and Best • GRIT USE BY BIRDS

693

among birds, both taxonomically and geographically. Moreover, although
the collection of birds for this research was not limited geographically or
temporally to areas or seasons of application of granular pesticides, our
results nonetheless have important implications for the design and use of
such products. Granular pesticides are applied to millions of hectares of
agricultural crops in North America each year (U.S. Dept. Agric. 1992).
Many of these pesticides are acutely toxic to birds (Balcomb et al. 1984,
Hill and Camardese 1984), and one potential route of avian exposure is
consumption of granules as a source of grit. Our results indicate that, if
this is an important route of avian exposure to granular pesticides, then
in pesticide-treated areas granivorous birds (including such important
game species as the Ring-necked Pheasant and Northern Bobwhite) may
have a relatively high risk of exposure because they use more grit than
birds that eat other foods. Our results confirm that altering granule size
to reduce avian risk is not likely to be effective. The wide range of grit
sizes used by birds, and the extensive interspecific overlap in grit sizes
result in there being no “safe” granule size that would be infrequently
used by birds, given the upper and lower limits to pesticide granule size
imposed by other factors, such as human safety and ease of application.

ACKNOWLEDGMENTS

We thank K. L. Andersen, K. K. Aulwes, B. M. Ballard, C. Best, J. D. Best, N. Best, J.
R. Gionfriddo, J. M. Grady, D. S. Huntrods, L. D. Igl, A. L. Linville, F. D. Zenitsky, and
G. D. Zenitsky for assisting with laboratory work and data tabulation. We are grateful to
those who provided or collected birds for our study, especially J. J. Dinsmore, B. J. Giesler,
L. D. Igl, D. W. DeGeus, D. S. Huntrods, and G. M. Booth. We appreciate the cooperation
of state and federal agencies in granting permission for collection of birds. We thank C. E.
Braun, J. J. Dinsmore, E. E. Klaas, J. L. Sell, K. C. Shaw, D. W. White, and C. R. Blem
for reviewing earlier drafts of the manuscript and offering helpful suggestions. Funding was
provided by Miles, Inc., Rhone-Poulenc, American Cyanamid, and Dow Elanco. This is
Journal Paper No. J- 161 81 of the Iowa Agriculture and Home Economics Experiment Sta-
tion, Ames, Project 2168.

LITERATURE CITED

Alonso, J. C. 1985. Grit in the gizzard of Spanish Sparrows (Passer hispaniolensis). Vo-
gelwarte 33:135-143.

Balcomb, R., R. Stevens, and C. Bowen II. 1984. Toxicity of 16 granular insecticides to
wild-caught songbirds. Bull. Environ. Contam. Toxicol. 33:302—307.

Barlow, J. C., E. E. Klaas, and J. L. Lenz. 1963. Sunning of Bank and Cliff swallows.
Condor 65:438-440.

Barrentine, C. D. 1980. The ingestion of grit by nestling Barn Swallows. J. Field Ornithol.
51:368-371.

Bartonek, j. C. 1969. Build-up of grit in three pochard species in Manitoba. Wilson Bull.
81:96-97.

694

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Beer, J. and W. Tidyman. 1942. The substitution of hard seeds for grit. J. Wildl. Manage.
6:70-82.

Best, L. B. and D. L. Fischer. 1992. Granular insecticides and birds: factors to be con-
sidered in understanding exposure and reducing risk. Environ. Toxicol, and Chem. 1 1 :
1495-1508.

AND J. P. Gionfriddo. 1991. Characterization of grit use by cornfield birds. Wilson

Bull. 103:68-82.

AND . 1994. Effects of surface texture and shape on grit selection by House

Sparrows and Northern Bobwhite. Wilson Bull. 106:689-695.

Bird, R. D. and L. B. Smith. 1964. The food habits of the Red-winged Blackbird, Agelaius
phoeniceus, in Manitoba. Can. Field-Nat. 78:179—186.

Bishton, G. 1986. The diet and foraging behaviour of the Dunnock Prunella modularis in
a hedgerow habitat. Ibis 128:526—539.

Bump, G., R. W. Darrow, E C. Edminster, and W. E Crissey. 1947. The Ruffed Grouse:
life history, propagation, management. New York State Conserv. Dept., Albany, New
York.

Dalke, P. L. 1937. Food habits of adult pheasants in Michigan based on crop analysis
method. Ecology 18:199-213.

DeGraaf, R. M., N. G. Tilghman, and S. H. Anderson. 1985. Foraging guilds of North
American birds. Environ. Manage. 9:493-536.

Dunning, J. B., Jr., ed. 1993. CRC handbook of avian body masses. CRC Press, Boca
Raton, Florida.

Earner, D. S. 1960. Digestion and the digestive system. Pp. 411-467 in Biology and
comparative physiology of birds. Vol. 1 (A. J. Marshall, ed.). Academic Press, New
York, New York.

Ferrel, C. M., H. Twining, and N. B. Herkenham. 1949. Food habits of the Ring-necked
Pheasant (Phasianus colchicus) in the Sacramento Valley, California. Calif. Fish and
Game 35:51-69.

Garcher, M. L. and j. P. Carroll. 1991. Comparative diet study of male and female
Wilson’s Phalaropes in North Dakota. J. Pennsylvania Acad. Sci. (Suppl.) 64:190.

Gionfriddo, J. P. and L. B. Best. 1995. Grit use by House Sparrows: effects of diet and
grit size. Condor 97:57-67.

Harper, J. A. 1964. Calcium in grit consumed by hen pheasants in east-central Illinois. J.
Wildl. Manage. 28:264-270.

Hill, E. E and M. B. Camardese. 1984. Toxicity of anticholinesterase insecticides to birds:
technical grade versus granular formulations. Ecotoxicol. Environ. Safety 8:551—563.

Hogstad, O. 1988. Foraging pattern and prey selection of breeding Bramblings Fringilla
montifringilla. Fauna Norv. Ser. C, Cinclus 1 1 :27-39.

Jenkinson, M. a. and R. M. Mengel. 1970. Ingestion of stones by goatsuckers (Capri-
mulgidae). Condor 72:236—237.

Kopischke, E. D. and M. M. Nelson. 1966. Grit availability and pheasant densities in
Minnesota and North Dakota. J. Wildl. Manage. 30:269-275.

Korschgen, L. j. 1964. Foods and nutrition of Missouri and midwestern pheasants. Trans.
North Am. Wildl. and Nat. Resour. Conf. 29:159-180.

, G. D. Chambers, and K. C. Sadler. 1965. Digestion rate of limestone force-fed

to pheasants. J. Wildl. Manage. 29:820—823.

Lewin, V. AND G. Lewin. 1984. The Kalij Pheasant, a newly established game bird on the
island of Hawaii. Wilson Bull. 96:634-646.

Lienhart, R. 1953. Recherches sur le role des cailloux contenus dans le gesier des oiseaux
granivores. Bull. Soc. Sci. Nancy 12:5-9.

Gionfriddo and Best • GRIT USE BY BIRDS

695

May, T. a. and C. E. Braun. 1973. Gizzard stones in adult White-tailed Ptarmigan {La-
gopus leucurus) in Colorado. Arctic and Alpine Res. 5:49-57.

Mayoh, K. R. and R. Zach. 1986. Grit ingestion by nestling Tree Swallows and House
Wrens. Can. J. Zool. 64:2090-2093.

McCann, L. J. 1939. Studies of the grit requirements of certain upland game birds. J.
Wildl. Manage. 3:31-41.

Meinertzhagen, R. 1954. Grit. Bull. Br. Omithol. Club 74:97-102.

. 1964. Grit. Pp. 341-342 in A new dictionary of birds (A. L. Thomson, ed.).
McGraw-Hill, New York, New York.

Mott, D. E, R. R. West, J. W. De Grazio, and J. L. Guarding. 1972. Eoods of the Red-
winged Blackbird in Brown County, South Dakota. J. Wildl. Manage. 36:983-987.

Myrberget, S., C. Norris, and E. Norris. 1975. Grit in Norwegian Lxigopus spp. Nor-
wegian J. Zool. 23:205-212.

Norman, E I. and R. S. Brown. 1985. Gizzard grit in some Australian waterfowl. Wildfowl
36:77-80.

AND L. Mumford. 1985. Studies on the Purple Swamphen, Porphyria porphyria,
in Victoria. Australian Wildl. Res. 12:263-278.

Norris, E., C. Norris, and J. B. Steen. 1975. Regulation and grinding ability of grit in
the gizzard of Norwegian Willow Ptarmigan (Lagopus lagopus). Poult. Sci. 54:1839-
1843.

PiNOWSKA, B. 1975. Pood of female House Sparrows (Passer domesticus L.) in relation to
stages of the breeding cycle. Polish Ecol. Stud. 1:211-225.

AND K. Krasnicki. 1985. Quantity of gastroliths and magnesium and calcium con-
tents in the body of female House Sparrows during their egg-laying period. Zesz. Nauk.
Filii UW, 48, Biol. 10:125-130.

F*ulliainen, E. 1979. Grit intake of the Capercaillie, Tetrao urogallus, in the northern
Finnish taiga in autumn. Aquilo Sen Zool. 19:45-47.

Rajala, P. 1958. The choice of grinding stones by Capercaillie, Black Grouse, and ptar-
migan in the light of enclosure experiments. Suomen Riista 12:89-93.

Sadler, K. C. 1961. Grit selectivity by the female pheasant during egg production. J. Wildl.
Manage. 25:339-341.

Sharp, W. M. and H. E. McClure. 1945. The pheasant in the sandhill region of Nebraska.
Pp. 203-233 in The Ring-necked Pheasant and its management in North America (W.
L. McAtee, ed.). Am. Wildl. Inst., New York, New York.

Smith, R. E. 1960. The influence of size and surface condition of grit upon the digestibility
of feed by the domestic fowl. Can. J. Anim. Sci. 40:51-56.

Steel, R. G. D. and J. H. Torrie. 1980. Principles and procedures of statistics: a biomet-
rical approach, 2nd ed. McGraw-Hill, New York, New York.

Taylor, T. G. 1970. The provision of calcium and carbonate for laying hens. Pp. 108-1 17
in Proc. Univ. Nottingham Fourth Nutrition Conf. for Feed Manufacturers (H. Swan
and D. Lewis, eds.). J. and A. Churchill, London, U.K.

Tindall, A. R. 1973. Gastroliths in Norwegian grouse (Lxigopus lagopus). Acta Physiol.
Scand., Suppl. 396:66.

Turner, A. K. 1982. Timing of laying by swallows (Hirundo rustica) and Sand Martins
(Riparia riparia). J. Anim. Ecol. 51:29—46.

U.S. Department of Agriculture. 1992. Agricultural chemical u.sage: 1991 field crops
survey. U.S. Dep. Agric., Econ. Res. Serv., Washington, D.C.

Vance, D. R. 1971. Physical and chemical alterations of grit consumed by pheasants. J.
Wildl. Manage. 35:136-140.

Verbeek, N. a. M. 1970. Breeding ecology of the Water Pipit. Auk 87:425-451.

696

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Zimmerman, J. L. 1963. The bioenergetics of the Dickcissel, Spiza americana. Ph.D. diss.,
Univ. Illinois, Urbana, Illinois.

ZiswiLER, V. AND D. S. EARNER. 1972. Digestion and the digestive system. Pp. 343-430 in
Avian biology. Vol. 2 (D. S. Earner, J. R. King, and K. C. Parkes, eds.). Academic
Press, New York, New York.

Wilson Bull., 108(4), 1996, pp. 697-711

RED-COCKADED WOODPECKER NESTING SUCCESS,
FOREST STRUCTURE, AND SOUTHERN FLYING
SQUIRRELS IN TEXAS

Richard N. Conner, D. Craig Rudolph,

Daniel Saenz, and Richard R. Schaefer

Abstract. For several decades general opinion has suggested that southern flying squir-
rels (Glaucomys volans) have a negative effect on Red-cockaded Woodpeckers (Picoides
borealis) through competition for cavities and egg/nestling predation. Complete removal of
hardwood trees from Red-cockaded Woodpecker cavity tree clusters has occurred on some
forests because southern flying squirrel abundance was presumed to be associated with the
presence and abundance of hardwood vegetation. In some locations, southern flying squirrels
have been captured and either moved or killed in the name of Red-cockaded Woodpecker
management. We determined southern flying squirrel occupancy of Red-cockaded Wood-
pecker cavities in loblolly (Pinus shortleaf {P. echinata) pine habitat (with and with-

out hardwood midstory vegetation) and longleaf pine (P. palustris) habitat (nearly devoid
of hardwood vegetation) during spring, late summer, and winter during 1990 and 1991.
Flying squirrel use of Red-cockaded Woodpecker cavities was variable and was not related
to presence or abundance of hardwood vegetation. Woodpecker nest productivity was not
correlated with flying squirrel use of woodpecker cavities within clusters. In addition, we
observed six instances where Red-cockaded Woodpeckers successfully nested while flying
squirrels occupied other cavities in the same tree. Our results suggest that complete removal
of hardwoods from woodpecker cluster areas in loblolly and shortleaf pine habitat may not
provide benefits to the woodpeckers through reduction of flying squirrel numbers. Reduction
of hardwood midstory around cavity trees, however, is still essential because of the wood-
pecker’s apparent innate intolerance of hardwood midstory foliage. Received 3 Nov. 1995,
accepted 21 Mar. 1996.

The Red-cockaded Woodpecker {Picoides borealis) is a cooperative
breeder that lives in family groups composed of a breeding pair and fre-
quently one to several helpers (Ligon 1970, Walters et al. 1988, Walters
1990). The activities of the group center around a cluster of cavity trees
composed of living pines that contain one to several cavities and cavity
starts. Cavities are excavated into the heartwood of pines that typically
are infected with red heart fungus (Phellinus pini) (Conner and Locke
1982, Hooper 1988, Hooper et al. 1991, Rudolph et al. 1995). Cavity
excavation in Texas requires an average of 1.8 y in loblolly pines {Pinus
taeda), 2.4 y in shortleaf pines {P. echinata), and 6.3 y in longleaf pines
{P. palustris) (Conner and Rudolph 1995). Pines selected for cavities in
Texas usually exceeded 90 years of age (Conner and O’Halloran 1987,

Wildlife Habitat and Silviculture Laboratory (Maintained in cooperation with the College of Forestry,
Stephen F Austin State University), Southern Re.search Station, U.S.D.A. Forest Service, Nacogdoches,
Texas 75962.

697

698

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Rudolph and Conner 1991). The Red-cockaded Woodpecker is a keystone
species of the fire-climax, southern pine ecosystems in that they are the
primary species to excavate cavities in an otherwise cavity-barren envi-
ronment relative to hardwood ecosystems (Conner and Rudolph 1995).
Thus, the cavities that take Red-cockaded Woodpeckers a long time to
create tend to be in relatively high demand by other cavity-using species
(Dennis 1971, Rudolph et al. 1990b, Loeb 1993).

As cavities near completion, Red-cockaded Woodpeckers peck shallow
excavations, termed resin wells, around their cavity entrances. Continued
pecking at resin wells causes a copious flow of resin down the bole of
the pine (Ligon 1970). Woodpeckers also scale loose bark from the bole
of the cavity tree and nearby pines. Although bark scaling and resin flow
usually deters climbing by rat snakes {Elaphe obsoleta) (Jackson 1974,
Rudolph et al. 1990a), the resin barrier does not deter southern flying
squirrels (Glaucomys volans) which are frequent users of cavities with
unenlarged entrances that are also preferred by Red-cockaded Woodpeck-
ers (Rudolph et al. 1990b, Loeb 1993).

Past studies have indicated a negative association between Red-cock-
aded Woodpeckers and the density of hardwood midstory and understory
(Hopkins and Lynn 1971, Van Balen and Doerr 1978, Hovis and Labisky
1985, Conner and Rudolph 1989, Loeb et al. 1992). This has lead to
widespread management programs that remove all hardwood vegetation
from Red-cockaded Woodpecker cavity tree cluster areas (Conner and
Rudolph 1991b). Although the negative effect of hardwood vegetation on
Red-cockaded Woodpeckers is well documented, the mechanism that
causes this negative relationship is poorly understood. One proposed
mechanism for the hardwood effect is that southern flying squirrels, a
potential competitor for Red-cockaded Woodpecker cavities, are depen-
dent on hardwood midstory foliage. However, flying squirrels appear to
prefer hardwood vegetation primarily as understory cover and as a food
source (Bendel and Gates 1987). Contrary to popular belief, southern
flying squirrels may avoid areas with dense midstory foliage because it
interferes with flight paths between boles of larger pines (Bendel and
Gates 1987). The influence of plant species composition and midstory
and understory foliage densities in pine forests on southern flying squirrel
abundance is not fully understood. To date, no published studies have
demonstrated that southern flying squirrels have a negative effect on Red-
cockaded Woodpecker populations, yet management programs that in-
clude removal of southern flying squirrels from cavities and euthanasia
(Gaines et al. 1995) are becoming more widespread.

Several species of woodpeckers enlarge Red-cockaded Woodpecker
cavity entrance tunnels by excavation and use the cavities (Conner et al.

Conner et al. • WOODPECKERS AND FLYING SQUIRRELS

699

1991, Neal et al. 1992). Some of these species, e.g., Pileated (Dryocopus
pileatus) and Red-bellied (Melanerpes carolinus) woodpeckers, are con-
sidered to be primarily associated with hardwood forests (Reller 1972,
Conner et al. 1975). Metal plates that restrict the entrance diameter of
Red-cockaded Woodpecker cavities (Carter et al. 1989) have been devel-
oped for placement over enlarged cavities in hopes that some currently
unsuitable cavities can be rehabilitated and on unenlarged cavities to pre-
vent enlargement. Although these plates may prevent further damage by
larger species of woodpeckers, they will not deter the use of cavities by
southern flying squirrels or other small species of woodpeckers which
prefer smaller entrances.

The competitive impact of southern flying squirrels on Red-cockaded
Woodpeckers is largely hypothetical. If a detrimental impact is occurring,
it may be exacerbated in small declining Red-cockaded Woodpecker pop-
ulations such as those in eastern Texas that are also stressed by other
factors such as isolation and forest fragmentation (Conner and Rudolph
1989, 1991a).

Our objectives were to (1) determine the availability and use of Red-
cockaded Woodpecker cavities during the nesting, late-summer, and win-
ter seasons, (2) evaluate southern flying squirrel use of cavities in relation
to species composition and structure of vegetation, and (3) explore pos-
sible negative effects of southern flying squirrels on Red-cockaded Wood-
pecker breeding success.

STUDY AREAS AND METHODS

The study was conducted on the Angelina (62,423 ha; 31°15'N, 94°15'W) and Davy
Crockett (65,329 ha; 31°2I'N, 95°07'W) National Forests from March 1990 to October 1991.
We examined 1 1 Red-cockaded Woodpecker cavity clusters in open longleaf pine habitat,
10 clusters in loblolly-shortleaf pine habitat with all hardwood vegetation removed in the
cluster area, and seven clusters in loblolly-shortleaf pine habitat with extensive hardwood
vegetation present during 1990. We suspected that different seasons of the year may impose
varying levels of competition for cavities. The breeding season (spring) is likely to be a
season of potentially elevated competition, and competition at that time can decrease breed-
ing success. The late summer season may also be a critical period because new young have
fledged and are searching for cavities for nocturnal roost sites. We sampled cavity occupants
during winter to examine the possibility that thermal stress during the colder months may
lead to increased demand for cavities.

We climbed approximately 230 cavity trees using Swedish climbing ladders and examined
them for occupancy during spring (April to May) of 1990 and 1991, late summer (August
to October) of 1990 and 1991, and winter of 1990-1991 (December 1 990 to February 1991).
Only a few cavity trees were not climbed because of safety factors during each climbing
season. Such trees were typically small-diameter, old, inactive cavity trees that were pri-
marily hollow shells and whose cavities were of no use to Red-cockaded Woodpeckers for
cavities. We lowered a small, high intensity light into each cavity chamber, examined con-
tents with an oval mechanics mirror mounted on an extendable handle, and identified and

700

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

counted cavity occupants. Often more than one southern flying squirrel was present in a
cavity. When this occurred, a coat hanger wire (with the end bent around to prevent injury)
was placed into the cavity and flying squirrels lifted out for counting. We used presence of
chewed pine needles and fresh flying squirrel feces as an indicator of flying squirrel use.
Unchewed pine needles in an enlarged cavity indicated use by fox squirrels (Sciurus niger).
Cavity trees that were being simultaneously used by both southern flying squirrels and Red-
cockaded Woodpeckers (in two different cavities) were examined closely during the wood-
pecker nesting season to determine woodpecker fledging success and during other seasons
to detect cavity usurpation by flying squirrels. We measured the entrance diameters of cav-
ities and monitored cavities with restrictors in each cluster studied. Based on previous studies
(Rudolph et al. 1990b), cavities were divided into those suitable for Red-cockaded Wood-
pecker use (entrance diameters <7 cm in diameter) and those too enlarged to be acceptable
to Red-cockaded Woodpeckers (entrances >7 cm in diameter).

We measured reproductive success of Red-cockaded Woodpeckers in each cluster by
determining the number of young fledged from nest cavities. Young were counted at 8, 20,
and 23 days of age in each nest tree. Clusters were visited within a week of fledging to
determine how many of the nestlings observed on day 23 successfully fledged. We also
visited each cluster during August and September to determine number of surviving young.
We made dawn and dusk visits to each woodpecker group to verify roost locations, band
woodpeckers, and determine number of members in each family group during each climbing
season. We also determined the number of Red-cockaded Woodpeckers roosting outside of
cavities in the open.

Vegetation measured in each cavity tree cluster (six points per cluster) included basal area
of overstory and midstory pines and hardwoods using a one-factor metric prism, height of
midstory and understory vegetation using a clinometer, canopy closure using a 4-cm di-
ameter by 12 cm hollow tube, and foliage density of vegetation from the ground to 1 m,
and 1 m to 2 m using a density board as described by MacArthur and MacArthur (1961).
By spring 1991, the seven clusters which had a well-developed hardwood midstory during
the first year of the study had received midstory treatment. All hardwood vegetation was
removed from these seven clusters during the 1990-1991 winter giving them the same
structural appearance as the 10 clusters in loblolly— shortleaf habitat that were initially with-
out hardwoods.

Lor each cavity tree cluster during each season we calculated the percentage of unenlarged
cavities occupied by southern flying squirrels and those occupied by Red-cockaded Wood-
peckers. Analysis of variance and Duncan’s multiple range test were used to test for differ-
ences in flying squirrel and Red-cockaded Woodpecker use of cavities among habitat treat-
ments during each season {P = 0.05). We related fledging success with the proportion of
unenlarged cavities (all unenlarged and available unenlarged cavities) occupied by flying
squirrels with Spearman correlations (r^, P = 0.05).

RESULTS

Vegetation characteristics of cavity tree clusters. — Vegetation within
the three treatments differed distinctly during the first year of the study
(1990). Red-cockaded Woodpecker cavity tree clusters in longleaf pine
habitat were nearly devoid of any hardwood vegetation or understory and
midstory foliage except for grasses and forbs (Table 1 ), and the absence
of hardwoods extended well beyond the boundaries of cluster areas. This
was not the case in clusters located in loblolly-shortleaf pine habitat.

Conner et al. • WOODPECKERS AND FLYING SQUIRRELS

701

Table 1

Vegetative Characteristics (Mean ± SD) of Red-cockaded Woodpecker Cluster
Areas where Cavity Occupants were Monitored in Longleaf Pine, Loblolly-
Shortleaf Pine with no Midstory, Loblolly-Shortleaf Pine with Midstory Present
(Pre-treatment 1990) and Midstory Removed (Post-treatment 1991) on the
Angelina and Davy Crockett National Forests in eastern Texas

Habitat variable

Longleaf

pine

(N = II)

Loblolly-

shortleaf

no

midstory
(N = 10)

Loblolly-

shortleaf

with

midstory
(N = 7)

Loblolly-
shortleaf
midstory
removed
(N = 7)

Overstory pine basal area"" (mV
ha)

14.0 (6.6)*’

12.7 (3.5)*’

10.9 (4.1)*’

13.6 (3.7)*’

Overstory hardwood basal area
(m-/ha)

0.1 (0.2)*’

0.0 (0.0)*’

1.0 (4.3)'=

0.0 (0.3)*=

Midstory pine basal area (m^/ha)

0.2 (0.5)*’

0.5 (1.2)*’

1.8 (3.0)'=

0.3 (0.9)*’

Midstory hardwood basal area
(m^/ha)

0.0 (0.0)*’

0.1 (0.3)*’

1.7 (1.7)'=

0.2 (0.6)*=

Canopy closure (%)

46.3 (26.1)*’

51.5 (16.9)*’’’

55.8 (19.8)'=

53.5 (15.8)*’'=

Overstory height (m)

23.1 (2.0)*’

24.8 (2.8)'’

27.0 (3.7)'*

25.3 (1.5)'=

Midstory height (m)

6.7 (5.4)*’

3.3 (5.4)'’

12.4 (4.5)“

3.3 (7.3)'=

Understory height (m)

1.9 (0.9)*’

1.6 (0.6)'’

2.0 (0.8)*’

1.5 (0.5)'=

Foliage density 0-1 m (cmVmh

0.2 (0.2)*’

0.3 (0.1)*’

0.3 (0.2)*=

0.5 (0.7)'=

Foliage density 1-2 m (cmVmh

0.1 (0.1)*’^

0.1 (0.1)*’

0.2 (0.1)“

0.1 (0.2)'=“

Common letters indicate nonsignificant differences (ANOVA, Duncan's multiple range test [P = 0.05]).

Clusters in habitat where hardwoods had recently been removed were
quite similar to longleaf habitat in the actual cluster area (Table 1), how-
ever, a virtual wall of hardwood midstory foliage was encountered at the
edges of each cluster where midstory removal and thinning of overstory
pines had ceased. Clusters that had not yet received hardwood removal
treatment still had substantial hardwoods in the overstory, midstory, and
understory (Table 1). During the winter of 1990-1991, hardwoods and
midstory trees in the seven untreated clusters were removed, changing
those clusters into a vegetative condition similar to the loblolly-shortleaf
clusters that had received midstory treatment prior to the study (Table 1).

Faunal use of Red-cockaded Woodpecker cavities. — A variety of ver-
tebrates and invertebrates were observed using Red-cockaded Woodpeck-
er cavities during the study. Although observed in cavities infrequently,
American Kestrels {Falco sparverius). Eastern Screech-Owls (Otus asio),
Pileated Woodpeckers, Wood Ducks {Aix sponsa), and fox squirrels typ-
ically used cavities which had both the entrance and cavity chamber en-
larged. Eastern Screech-Owls were observed in three cavities with en-
trances <7 cm in diameter, but the entrances of these three cavities had

702

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

been slightly enlarged and were between 6.5 and 7 cm in diameter. Red-
bellied Woodpeckers are known to conflict with Red-cockaded Wood-
peckers over cavities (Neal et al. 1992, Kappes and Harris 1995) but were
observed using unenlarged cavities only once during spring 1991 and on
four occasions during winter.

Mud-daubers (Sphecidae) were typically found in inactive cavities.
Their mud chambers were tolerated or pecked off when Red-cockaded
Woodpeckers began to use a cavity containing mud-dauber nests. The
presence of mud-daubers or their nests did not appear to interfere with
Red-cockaded Woodpecker use of cavities. However, the presence of pa-
per wasps (Vespidae), particularly large nests, and honey bees {Apis mel-
lifera) did prevent Red-cockaded Woodpecker use of cavities. Broad-
headed skinks (Eumeces laticeps), five-lined skinks {E. fasciatus), and
gray tree frogs {Hyla versicoloAchrysoscelis) were observed occasionally
within inactive enlarged and unenlarged cavities.

Southern flying squirrel use of woodpecker cavities. — Red-cockaded
Woodpeckers preferred unenlarged cavities (Table 2); they used cavities
with greatly enlarged entrances (S:7 cm) in only two instances, both dur-
ing late summer 1990. As previously noted by Rudolph et al. (1990b)
and Loeb (1993), southern flying squirrels also prefer entrance diameters
<7 cm. Thus, the southern flying squirrel exhibited extensive overlap in
cavity use with Red-cockaded Woodpeckers; it was observed in relatively
high numbers and also used primarily unenlarged cavities (Table 2). In
most clusters, however, empty unenlarged and enlarged cavities were
available throughout the year for either Red-cockaded Woodpeckers or
flying squirrels to use (Tables 2, 3).

Southern flying squirrel use of Red-cockaded Woodpecker cavities dur-
ing the woodpecker breeding season (spring) was high, but dwindled
greatly by late summer during both 1990 and 1991 (Table 2). The number
of cavities used by Red-cockaded Woodpeckers was somewhat higher
during late summer than during the breeding season. Red-cockaded
Woodpeckers were present in greater numbers in the late summer because
young woodpeckers had recently fledged from nest cavities and many
were now roosting in cavities.

We detected very few significant differences in the percentage of unen-
larged cavities used by Red-cockaded Woodpeckers and southern flying
squirrels among habitat treatments (Table 3). During spring 1990 southern
flying squirrels used unenlarged Red-cockaded Woodpecker cavities at a
higher frequency in longleaf pine habitat than in loblolly-shortleaf habitat
where hardwood midstory vegetation was absent (Table 3). During spring
1991 the percentage of empty unenlarged cavities in loblolly-shortleaf
pine habitat without midstory was significantly lower than in longleaf

Conner et al. • WOODPECKERS AND ELYING SQUIRRELS

UJ

p

>

<

U

ai

LU

u

UJ

a.

Q

O

O

Q

LL!

Q

<

O

0
u

1

Q

LD

qc

"S'

u

H

lU

S

<

u

Al

u

tj

z

<

oi

H

Z

UJ

>-

>

<

Q

Q

Z

<

<

z

□

w

o

z

<

UJ

r

i-

z

0

t/5

H

Z

U

S

(-

<

LU

Pi

H

h-

<

5

<

X

u

UJ

OS

1

f-

(N
UJ

Q

oa lu
< S
H ^

_3

Z

UJ

Q

Z

<

"S'

LU

H

UJ

<

Q <

u

V

UJ

U

z

<

Z Q
UJ UJ
Q
<

U

o
o

D

UJ

O

os

<

_J

z

UJ

z

D

u.

O

Q
UJ
CL

D

z o

O'

UJ

CO

c/5

<

X

UJ

H

z

O'

UJ

H

I/)

<

-J

UJ

ai.

5

a

00

0
z
>
Uu
z

UJ

1

D

O

00

UJ

0^

O

uu

-J

<

z

o

<

Z

w
X
u
o

Q QC

z U

— *

IT)

VO

VO

p

SO

VO

so

VO

u

22

c; 00

it

m

IT)

II

O Q

II

r-

^ >1

o

r-

V

ON

<N

(N

fo

-- II

CO

Z

2

z

fN

2

2

CN

CN

m

so

(N

VO

fS

(N

22 II

o :::::

= II °

18

0

!C; lo

z

z

Z

00

C; <>1
00
<N

(N

CN

On

VO

(N

(N

r-

C:

(N

CN

r-

so

(N

0\

(N

10 O

vD — .

11 CN o (N

o

CM

(N

5$ m

(N —

— O lO

O 5

(N

o o o

O C:^ On

o Ci Tf

o 5 ^ o

so ,, o C! o

z

z

z

ON

(N

5^

VO

in

VO

II (N

VO

00

II

28

9S ^
\o ^

II 2

VO

2

z

z

Os

SO lo —

^ ^ so

c/5

o

V

V

N

CL

’c^

•o

0

JJ

0

’a.

f-

E

c/5

•o

(U

TD

o

cd

ON

Os

CJ

0

u

Oi)

c

1

"O

V

'u

CL

CC

oo

a>

t

<u •::

u

lU

c/5

13

t

’3

u

<U

c/5

13

fc

’3

c/5

u

(U

c/5

13

fc

’3

C/J

<u

N u

u

O"

c/5

o

cr

O

o

O’

O U

<u

c/5

o

(U

C/)

V

c/5

7s

a

a

Cl

L..

9- a

"O

o

*3

•o

1^

0

(U

•o

r,

o

3

o-

o

c

c/5

cd

C/5

c

C/1

Cd

<u

c/5

•o

Ua

Cd

c/5

■o

<u

0)

cz

o

(U

"O

(U

"O

t

!5

‘>

ON

cd

’3

'>

Cd

'3

>

On

c3

ON

cr

Cd

ON

a*

Cd

On

O

*“

o

C4

o

ON

u

c/5

o

D.

E

UJ

V c

"S Z'

Qi U-

^ —
D. W)

E .E
W a

cz)

U M ^

^ .E CL
u Z* E
0^ u. UJ

a o
6 ^

■o

<u

T3
3
j<:

u
o

(U o

E -a
00

o

c

"oj

fc

c/} On

D On

^ d
> -S

— C/5

•o 0) 4>

fc -z:

'3

*>

o

"O

cd

'5

cr

Cd

Os

cr

c/5

o

ON

o

c/5

W)

E

o

0

u

Ofi

3

E

o

c

■a

(U

>>

E

UJ

a;

E

703

704

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 3

Percentage of Unenlarged Cavities within Each Cluster Occupied by Red-cockaded
Woodpeckers by Southern Flying Squirrels, or Empty (Mean ± SD) in Loblolly-
Shortleaf Pine Habitat without Hardwood Midstory Vegetation (N = 10),
Loblolly-Shortleaf Pine Habitat with Hardwood Midstory Vegetation (Pre- and
Post-removal, N = 7), and Longleaf Pine Habitat (N = 1 1) in Eastern Texas

Habitat treatment

Loblolly-shortleaf

Loblolly-shortleaf

Loblolly-shortleaf

with midstory

post midstory

Longleaf

without midstory

(pre-removal)

removal

Variable

.f*

SD

X

SD

X

SD

X

SD

Spring 1990

Red-cockaded (%)

56.4*’

26.3

43.5*’

24.6

50.3*’

33.7

Flying squirrel (%)

16.9*’

18.8

38.8*’^

28.0

44.8'

31.0

Empty (%)

14.7*’

16.6

14.6*’

18.3

3.8*’

8.6

Spring 1991

Red-cockaded (%)

48.2*’

17.2

29.5*’

14.7

35.5*’

26.3

Flying squirrel (%)

41.6*’

14.8

38.4*’

15.0

28.5*’

20.9

Empty (%)

7.2*’

13.0

27.2^

15.1

30.9'

14.7

Summer 1990

Red-cockaded (%)

64.0*’

34.8

60.8*’

25.8

46.9*’

30.7

Flying squirrel (%)

7.0*’

9.5

14.4*’

14.1

6.^

11.2

Empty (%)

25.0*’

23.8

21.4*’

24.5

46.3*’

28.4

Summer 1991

Red-cockaded (%)

60.4*’

17.5

49.7*’

24.9

58.8*’

20.1

Flying squirrel (%)

7.8*’

10.4

23.0*’

23.2

19.6*’

13.2

Empty (%)

26.2*’

15.1

25.0*’

10.3

21.6*’

21.3

Winter 1990-1991

Red-cockaded (%)

62.8*’

25.2

38.2'’

12.6

47.1*’'

26.3

Flying squirrel (%)

16.3*’

19.3

25.7*’

12.5

19.4*’

15.2

Empty (%)

18.9*’

21.2

27.8*’

11.4

28.1*’

22.9

• Common superscript letters following means indicate nonsignificant differences among habitat treatments (ANOVA.
Duncan's multiple range test, P = 0.05).

pine or loblolly-shortleaf pine from which midstory had been recently
removed (Table 3). Southern flying squirrel use of Red-cockaded Wood-
pecker cavities was not related to the presence or absence of hardwood
midstory (Table 3). Thus, treatment specific and annual use of cavities
by flying squirrels appears to be minimally affected by hardwood mid-
story abundance.

Southern flying squirrel use of cavities in the loblolly-shortleaf habitat
with midstory was greater than their use of loblolly— shortleaf habitat with-
out hardwood vegetation during spring 1990, although not significantly

Conner et al. • WOODPECKERS AND FLYING SQUIRRELS

705

Table 4

Number of Red-cockaded Woodpeckers Roosting in the Open during Spring and Late
Summer 1990 and 1991 in Loblolly-Shortleaf Pine Habitat without Hardwood
Midstory Vegetation (N = 10 Clusters), with Hardwood Midstory Vegetation (Pre-
AND Post-hardwood Removal, N = 7 Clusters), and Longleaf Pine Habitat (N = 1 1

Clusters) in Eastern Texas

Season

Habitat treatment

Loblolly-shortleaf
without midstory

Loblolly-shortleaf
with midstory

Longleaf

No. woodpeckers

No. woodpeckers

No. woodpeckers

Spring 1990

I

u

6

Summer 1990

3

5“

4

Spring 1991

I

3*’

0

Summer 1991

2

2b

2

“ Pre-midstory removal treatment within these clusters.

‘’Post-midstory removal treatment within cluster areas completed during winter 1990-1991.

SO (Table 3). However, the percentage of unenlarged cavities used by
flying squirrels remained the same during spring 1991 even though hard-
wood midstory vegetation had been removed (Table 3). Flying squirrel
use of unenlarged cavities increased in the loblolly-shortleaf area without
midstory even though no habitat alteration occurred (Table 3). Both the
percentage of cavities used by flying squirrels and the abundance of flying
squirrels counted in Red-cockaded Woodpecker cavities decreased be-
tween spring and late summer in 1990 and 1991 (Table 2, 3). We did not
make a detailed survey of the crowns of nearby pines and hardwoods in
the woodpecker cluster areas, but strongly suspect that flying squirrels
were spending the hot, late summers in leaf nests rather than woodpecker
cavities, as also observed by Muul (1974).

During winter, the percentage of unenlarged cavities and available
unenlarged cavities used by southern flying squirrels was relatively sim-
ilar in all habitat treatments (Table 3). Empty unenlarged cavities were
readily available in clusters in all habitat treatments during winter, sug-
gesting that cavity availability did not create a competitive problem for
Red-cockaded Woodpeckers during winter.

Extra-cavity roosting by woodpeckers. — Extra-cavity roosting as de-
scribed by Hooper and Lennartz (1983) is a possible indicator of insuf-
ficient cavity availability for Red-cockaded Woodpeckers. In general, very
few Red-cockaded Woodpeckers were observed roosting in the open (Ta-
ble 4). Typically, when Red-cockaded Woodpeckers roosted in the open,
there were empty cavities available within their cluster areas. Spring 1990
in the longleaf pine habitat appeared to be exceptional in this regard.

706

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

With the exception of longleaf pine habitat in spring 1990, Red-cockaded
Woodpeckers appeared to roost in the open more during late summer than
during the breeding season (Table 4). Flying squirrels were very abundant
during spring 1990 in the longleaf habitat (Table 2) and empty cavities
were few, suggesting that a few Red-cockaded Woodpeckers may have
been forced temporarily to roost in the open. Many recently fledged young
Red-cockaded Woodpeckers did not roost in cavities during the late sum-
mer. Because many unenlarged empty cavities were available for these
woodpeckers to use during late summer, roosting in the open appears to
be voluntary and may have been in response to the typical high air tem-
peratures during August and September.

Red-cockaded Woodpecker fledging success. — We examined Red-cock-
aded Woodpecker fledging success to explore the possibility that inter-
actions with southern flying squirrels reduced woodpecker nest produc-
tivity. Because southern flying squirrel use of woodpecker cavities was
uniformly high over all habitat treatments and years, our ability to eval-
uate the influence of squirrel use of cavities on fledging success through
comparisons across habitats was limited.

Fledging success was slightly higher in loblolly— shortleaf habitat with
hardwood vegetation (pre-hardwood removal) than in the loblolly-short-
leaf habitat without hardwood vegetation during 1990 (Fig. 1). Fledging
success remained somewhat higher in these cluster areas in 1991 (post-
treatment) even though the hardwood vegetation had been removed prior
to the 1991 breeding season. Longleaf pine habitat, relatively devoid of
hardwood vegetation, and often considered the premiere habitat for the
woodpecker, had a slightly lower fledging success than either loblolly-
shortleaf habitats during both 1990 and 1991 (Fig. 1). Excluding nests
where eggs failed to hatch, we failed to detect any significant differences
in fledging success among habitat treatments (Kruskal-Wallis approx-
imation, = 1 -42, P — 0.49).

We compared the proportion of all unenlarged cavities used by flying
squirrels and the proportion of available unenlarged cavities (open cavities
not used by Red-cockaded Woodpeckers) that contained flying squirrels
with woodpecker fledging success (Fig. 1). During both 1990 and 1991
we observed no relationship between southern flying squirrel occupancy
and habitat condition (abundance of hardwood midstory) or Red-cock-
aded Woodpecker fledging success. Red-cockaded Woodpecker fledging
success per habitat treatment during the two breeding seasons (N = 6)
was not correlated with the percentage of all unenlarged cavities occupied
by southern flying squirrels (r, = -0.08, P = 0.87) or the percentage of
available unenlarged cavities (those not used by Red-cockaded Wood-
peckers) occupied by southern flying squirrels = -0.46, P = 0.35). If

Conner et al.

WOODPECKERS AND ELYING SQUIRRELS

707

WITH MIDSTORY WITHOUT MIDSTORY

NOW REMOVED

HABITAT TREATMENT

Fig. 1. Comparisons of Red-cockaded Woodpecker fledging success with the proportion
of all unenlarged Red-cockaded Woodpecker cavities occupied by southern flying squirrels
(Glaucomys volans, G. V.) and available unenlarged cavities (those not used by Red-cock-
aded Woodpeckers) occupied by southern flying squirrels in loblolly-shortleaf pine habitat
with hardwood midstory present (pre- and post-hardwood removal), loblolly-shortleaf pine
habitat without hardwood midstory throughout the study, and longleaf pine habitat during
the 1990 and 1991 breeding seasons on the Angelina and Davy Crockett National Forests
in eastern Texas.

708

THE WILSON BULLETIN • VoL 108, No. 4. December 1996

clusters are treated as the sample unit (N = 44), fledging success is still
not correlated with either the percentage of all unenlarged cavities oc-
cupied by flying squirrels (r^ = —0.18, P = 0.23), or the percentage of
available unenlarged cavities occupied by flying squirrels {r^ = -0.11, P
= 0.48).

Potential for flying squirrel predation on woodpeckers. — During our 2
year study we observed 6 instances where Red-cockaded Woodpeckers
nested and produced young in cavities while southern flying squirrels
were occupying other cavities in the same pine tree. In only one instance
were eggs lost (to unknown causes), but the woodpeckers renested and
successfully fledged young from the same cavity. Young fledged suc-
cessfully from all five of the other nest cavities.

DISCUSSION

Competition between flying squirrels and woodpeckers. — Our obser-
vations in eastern Texas suggest a minimal competitive impact of southern
flying squirrels on Red-cockaded Woodpeckers. Because we did not mea-
sure woodpecker fledging success over a wide range of flying squirrel
abundance, however, our results may not be definitive. Competition from
southern flying squirrels in Texas is likely transient and occurs as isolated
events during ecological “bottle-necks.” If such competition occurs at all
in eastern Texas, the effects are subtle rather than overwhelming. The
effect of southern flying squirrels on any healthy woodpecker population
is probably minimal to non-existent.

Specifically, we have not seen (1) a relationship between woodpecker
fledging success and flying squirrel use of cavities, (2) Red-cockaded
Woodpeckers forced to roost in the open because of a squirrel-caused
shortage of unenlarged cavities, or (3) regular squirrel predation on Red-
cockaded Woodpecker eggs and young even when both woodpeckers and
flying squirrels occupied the same cavity tree.

Relationships among woodpeckers, squirrels, and hardwood vegeta-
tion,— We did not observe a strong relationship between southern flying
squirrel abundance and presence of hardwood vegetation. Flying squirrels
were common in cavities in longleaf pine habitat with almost no hard-
wood vegetation. This finding, however, does not negate the necessity to
reduce hardwood vegetation within woodpecker cluster areas. Past studies
have clearly demonstrated the negative effects of excessive hardwood
midstory on woodpecker populations (Van Balen and Doerr 1978, Hovis
and Labisky 1985, Conner and Rudolph 1989, Loeb et al. 1992). Thus,
we strongly urge that reduction (not elimination) of hardwood vegetation
within Red-cockaded Woodpecker cluster areas be continued.

Our results indicate that complete or partial removal of all hardwoods

Conner et al. • WOODPECKERS AND ELYING SQUIRRELS 709

will likely not affect the use of Red-cockaded Woodpecker cavities by
southern flying squirrels. What our study suggests is that southern flying
squirrels are not the cause of harmful effects on Red-cockaded Wood-
peckers associated with the presence of hardwood vegetation within their
cluster areas. As we have suggested before (Conner and Rudolph 1991b),
Red-cockaded Woodpeckers may have an innate avoidance of areas with
extensive hardwood vegetation as a result of their adaptation to the south-
ern fire-climax pine ecosystem. A selective advantage may accrue for
Red-cockaded Woodpecker pairs that avoid habitat with abundant hard-
wood vegetation because such areas may support greater numbers of other
species of woodpeckers that can easily out-compete Red-cockaded Wood-
peckers for cavities or destroy the cavities they excavate. Another possible
reason why Red-cockaded Woodpeckers have an aversion to hardwoods
is that they may provide predators access to cavities (Walters 1990).

We saw no negative effect of southern flying squirrels on Red-cockaded
Woodpeckers, nor have any other studies demonstrated such an effect.
We strongly discourage removal and euthanasia of southern flying squir-
rels in woodpecker clusters because of the complete lack of evidence that
it would benefit Red-cockaded Woodpeckers. If removal of southern fly-
ing squirrels is deemed necessary, it should be based on site-specific data
that statistically demonstrates a severe competitive problem. In such in-
stances, control of flying squirrels should last only as long as the wood-
pecker population is small and vulnerable to sudden extirpation.

ACKNOWLEDGMENTS

We thank J. R. Walters, R. G. Hooper, E C. James, J. D. Ligon, S. C. Loeb, B. Parresol,
and J. E Taulman for constructive comments on an early draft of the manuscript. Partial
funding was provided by a Challenge Cost Share Agreement (#19-90-008) with the Resource
Protection Division, Texas Parks and Wildlife Dept.

LITERATURE CITED

Bendel, P. R. and j. E. Gates. 1987. Home range and microhabitat partitioning of the
southern flying squirrel (Glaucomys volans). J. Mammal. 68:243-255.

Carter, J. H. Ill, J. R. Walters, S. H. Everhart, and P. D. Doerr. 1989. Re.strictions for
Red-cockaded Woodpecker cavities. Wildl. Soc. Bull. 17:68-72.

Conner, R. N., R. G. Hooper, H. S. Crawford, and H. S. Mosby. 1975. Woodpecker
nesting habitat in cut and uncut woodlands in Virginia. J. Wildl. Manage. 39:144-150.

AND B. A. Locke. 1982. Fungi and Red-cockaded Woodpecker cavity trees. Wilson

Bull. 94:64-70.

AND K. A. O’Halloran. 1987. Cavity-tree selection by Red-cockaded Woodpeckers

as related to growth dynamics of southern pines. Wilson Bull. 99:398-412.

AND D. C. Rudolph. 1989. Red-cockaded Woodpecker colony status and trends on

the Angelina, Davy Crockett, and Sabine National Forests. U.S. Dept. Agric., For. Serv.
Res. Pap. 50-250.

710

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

and . 1991a. Eorest habitat loss, fragmentation, and Red-cockaded Wood-
pecker populations. Wilson Bull. 103:446—457.

and . 1991b. Effects of midstory reduction and thinning in Red-cockaded

Woodpecker cavity tree clusters. Wildl. Soc. Bull. 19:63—66.

and . 1995. Excavation dynamics and use patterns of Red-cockaded Wood-
pecker cavities: relationships with cooperative breeding. Pp. 343-352 in Red-cockaded
Woodpecker: recovery, ecology and management (D. L. Kulhavy, R. G. Hooper, and
R. Costa, eds.). College of Forestry, Stephen E Austin State Univ., Nacogdoches, Texas.

, , D. L. Kulhavy, and A. E. Snow. 1991. Causes of mortality of Red-

cockaded Woodpecker cavity trees. J. Wildl. Manage. 55:531-537.

Dennis, J. V. 1971. Species using Red-cockaded Woodpecker holes in northeastern South
Carolina. Bird-Banding 42:79—87.

Gaines, G. D., K. E. Franzreb, D. H. Allen, K. S. Laves, and W. L. Jarvis. 1995. Red-
cockaded Woodpecker management on the Savannah River site: a management/research
success story. Pp. 81-88 in Red-cockaded Woodpecker: recovery, ecology and man-
agement (D. L. Kulhavy, R. G. Hooper, and R. Costa, eds.). College of Forestry, Ste-
phen E Austin State Univ., Nacogdoches, Texas.

Hooper, R. G. 1988. Longleaf pines used for cavities by Red-cockaded Woodpeckers. J.
Wildl. Manage. 52:392-398.

AND M. R. Lennartz. 1983. Roosting behavior of Red-cockaded Woodpecker clans

with insufficient cavities. J. Field Ornithol. 54:72—76.

, , and H. D. Muse. 1991. Heart rot and cavity tree selection by Red-

cockaded Woodpeckers. J. Wildl. Manage. 55:323-327.

Hopkins, M. L. and T. E. Lynn, Jr. 1971. Some characteristics of Red-cockaded Wood-
pecker cavity trees and management implications in South Carolina. Pp. 140-169 in
The ecology and management of the Red-cockaded Woodpecker (R. L. Thompson, ed.).
Bur. Sport Fish and Wildl. and Tall Timbers Res. Stn., Tallahassee, Florida.

Hovis, J. A. and R. E Labisky. 1985. Vegetative associations of Red-cockaded Woodpecker
colonies in Florida. Wildl. Soc. Bull. 13:307—314.

Jackson, J. A. 1974. Gray rat snakes versus Red-cockaded Woodpeckers: predator-prey
adaptations. Auk 91:342-347.

Kappes, j., Jr. and L. D. Harris. 1995. Interspecific competition for Red-cockaded Wood-
pecker cavities in the Apalachicola National Forest. Pp. 389-393 in Red-cockaded
Woodpecker: recovery, ecology and management (D. L. Kulhavy, R. G. Hooper, and
R. Costa, eds.). College of Forestry, Stephen F. Austin State Univ., Nacogdoches, Texas.

Eicon, J. D. 1970. Behavior and breeding biology of the Red-cockaded Woodpecker. Auk
87:255-278.

Loeb, S. C. 1993. Use and selection of Red-cockaded Woodpecker cavities by southern
flying squirrels. J. Wildl. Manage. 57:329—335.

^ W. D. Pepper, and A. T. Doyle. 1992. Habitat characteristics of active and aban-
doned Red-cockaded Woodpecker colonies. So. J. Appl. For. 16:120-125.

MacArthur, R. H. and J. W. MacArthur. 1961. On bird species diversity. Ecology 42:
594-598.

Muul, I. 1974. Geographic variation in the nesting habits of Glaucomys volans. J. Mammal.
55:840-844.

Neal, J. C., W. G. Montague, and D. A. James. 1992. Sequential occupation of cavities
by Red-cockaded Woodpeckers and Red-bellied Woodpeckers in the Ouachita National
Forest. Arkansas Acad. Sci. 46:106-108.

Reller, a. W. 1972. Aspects of behavioral ecology of Red-headed and Red-bellied wood-
peckers. Am. Midi. Nat. 88:270—290.

Conner et al. • WOODPECKERS AND FLYING SQUIRRELS 7 1 1

Rudolph, D. C., H. Kyle, and R. N. Conner. 1990a. Red-cockaded Woodpeckers vs rat
snakes: the effectiveness of the resin barrier. Wilson Bull. 102:14-22.

•, R. N. Conner, and J. Turner. 1990b. Competition for Red-cockaded Woodpecker
{Picoides borealis) roost and nest cavities: the effects of resin age and cavity entrance
diameter. Wilson Bull. 102:23-36.

Cavity tree selection by Red-cockaded Woodpeckers in relation
to tree age. Wilson Bull. 103:458-467.

’ ~ Schaefer. 1995. Red-cockaded Woodpecker detection of red

heart infection. Pp. 338-342 in Red-cockaded Woodpecker: recovery, ecology and man-
agement (D. L. Kulhavy, R. G. Hooper, and R. Costa, eds.). College of Forestry, Ste-
phen E Austin State Univ., Nacogdoches, Texas.

U.S. Fish and Wildlife Service. 1985. Red-cockaded Woodpecker recovery plan. U.S.
Fish and Wildl. Serv., Atlanta, Georgia.

Van Balen, J. B. and P. D. Doerr. 1978. The relationship of understory vegetation to
Red-cockaded Woodpecker activity. Proc. Annu. Conf. Southeast. Assoc. Fish and
Wildl. Agencies 32:82—92.

Walters, J. R. 1990. Red-cockaded Woodpeckers: a ‘primitive’ cooperative breeder. Pp.
69-101 in Cooperative breeding in birds (P. B. Stacey and W. D. Koenig, eds.). Cam-
bridge Univ. Press, London, United Kingdom.

, P. D. Doerr, and J. H. Carter, III. 1988. The cooperative breeding system of the
Red-cockaded Woodpecker. Ethology 78:275-305.

Wilson Bull., 108(4), 1996, pp. 712-727

HABITAT-USE PATTERNS IN COOPERATIVE AND
NON-COOPERATIVE BREEDING BIRDS; TESTING
PREDICTIONS WITH WESTERN SCRUB-JAYS

D. Brent Burt

Abstract. — I propose a method to test extensions of models concerning the maintenance
of cooperative breeding systems that examines patterns of habitat use relative to the distri-
bution of habitat components among territories. I analyzed habitat use and behavioral time
budget data for a Texas population of the Western Scrub-Jay (Aphelocoma califomica). As
a non-cooperative population, one of two habitat-use patterns was expected: (1) specialist
habitat-use patterns in an abundant, widespread habitat type, with little variation among
territories in habitat composition or (2) generalist habitat-use patterns with the potential of
significant variation in habitat composition among territories. These jays show a combination
of habitat-use patterns supporting both predictions. The only resources that males utilize as
a specialist, tall oak trees during sentinel behavior, are fairly widespread and would not be
considered a limiting resource. In the remaining habitat categories, Texas populations of
Western Scrub-Jays act as generalists, using the habitat in relation to its availability, even
though variation in habitat composition among territories is considerable. Variation among
individuals within a sex was observed but could not be explained using various demographic
and ecological correlates. Additional detailed habitat use data when used in a comparative
framework can aid determination of subtle ecological differences among populations of
Western Scrub-Jays and allow closer examination of intrinsic and extrinsic ecological models
concerning the evolution and maintenance of cooperative breeding systems in this group.
Received 27 Aug 1995, accepted 8 April 1996.

Habitat use and the concept of ecological constraints have played a
major role in the development of theories concerning evolution and main-
tenance of cooperative breeding in birds (Brown 1987, Koenig et al.
1992). Specifically, ecological constraints may serve as both intrinsic and
extrinsic reasons for delayed dispersal, setting the stage for the helping
behavior seen in cooperative breeding systems. Comparative studies of
the genus Aphelocoma have been particularly instructive in testing the
role ecological constraints play in the evolution of social systems because
the genus exhibits extensive geographic variation in social systems and
habitat use (Brown 1974, Fitzpatrick and Woolfenden 1986, Peterson and
Burt 1992). The success of the comparative method rests on an even and
complete sampling of taxa in the study group. To that end, this study
documents the habitat use patterns of a non-cooperative population of
Western Scrub-Jays (A. califomica, see American Ornithologists’ Union
1995), in central Texas.

Mu.seum of Natural History and Dept, of Systematics and Ecology, Dyche Hall, Univ. of Kansas, Law-
rence, Kansas 66045-2454. Present address; Dept, of Biology, Stephen F. Austin State Univ., Box 13003—
SFA Station, Nacogdoches, Texas 75962-3003.

712

D. B. Burt • SCRUB-JAY HABITAT USE

713

Using behavioral time-budget data, it is possible to test whether each
major habitat component is used relative to its availability for each major
behavioral category (i.e., individuals act as generalists) or whether certain
behaviors are more likely to occur in specific habitats (i.e., individuals
act as specialists). When behavioral-habitat specialization patterns are
seen, we can then study in more detail the importance and distribution of
this target of specialization and see if it is possibly a limiting resource
critical in the determination of habitat quality. This study examines habitat
composition of territories and habitat use patterns and determines (1) the
generalist/specialist habitat use status of this population, (2) the avail-
ability and distribution of different habitat components and the presence
of potentially limiting resources, and 3) how specific behaviors are related
to each habitat component. As a non-cooperative population (Burt 1992),
one of two habitat use patterns is expected (see discussion). First, the
population may show specialist habitat use patterns, but only if in an
abundant, widespread habitat type with little change in habitat composi-
tion among territories. Alternatively, the population may show generalist
habitat-use patterns with significant variation in habitat composition
among territories with individuals using each habitat component in rela-
tion to its availability. In the latter case no target of specialization is
expected.

STUDY AREA AND METHODS

I conducted this study in oak-juniper woodland in Kerrville-Schreiner State Park, Kerr-
ville, central Texas. The park’s dominant woody vegetation was Texas live oak (Quercus
virginiana fusiformis), Texas red oak (Q. shumardii texana), and Ashe Juniper {Juniperus
ashei) (Miller and Lamb 1985). I partitioned the study area into 231 quadrats, 33.3 m^ in
area. In each quadrat, I characterized the woody vegetation composition in two ways: bio-
diversity or structural diversity. I characterized biodiversity into four categories by percent-
age of living oak (LO), dead oak (DO), living Juniper (LJ), and dead juniper (DJ). I also
characterized structural diversity into four categories by percentage of isolated living (IL),
isolated dead (ID), dense living (DL), and dense dead (DD) patches of trees. I calculated
percentages by actual counts of trees in each quadrat when possible and by visual estimates
to the nearest five percent in dense vegetation. Because percentage estimates of vegetation
are not comparable between quadrats of different vegetational density, I multiplied each
percentage by the density of vegetation in that quadrat. The resulting number is the stan-
dardized, relative abundance of each vegetation type which can then be compared to the
relative abundance all other quadrats. The density of vegetation in each quadrat was mea-
sured from aerial photographs using the transect method. Evenly spaced transect lines were
drawn in each quadrat of the aerial photograph, and the length of line crossing trees was
divided by total line length (Avery 1985; 87).

I used mist nets to capture 66 jays and marked them with U.S. Fish and Wildlife Service
leg bands and a unique combination of three colored metal bands. I aged adult jays as first-
year birds or older based on plumage characters (Pitelka 1945). During the breeding season,
jays were sexed by presence or absence of a brood patch. Individuals captured at other times

714

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

were sexed by behaviors and vocalizations. I detennined territory boundaries by observing
territorial conflicts and home range use and then mapped each on aerial photographs. I
calculated vegetational composition for each territory each year by averaging numbers of
each vegetation type for each quadrat in each territory.

I collected focal sampling data from 7 June to 6 July 1989 on four breeding pairs and
from 14 June to 29 June 1990 on five breeding pairs. All behaviors were recorded to the
nearest second in 15-min sampling sessions or until the bird was lost from sight. Sessions
where the bird’s behavior was affected by observer presence were excluded from analysis.

I categorized behaviors as sentinel, foraging, inactive/preening, territorial, or other (begging,
caching, sunning, etc.). Sentinel behavior included any behavior engaged in when an indi-
vidual was perched on a tall exposed branch, and its designation overrides concurrent be-
haviors such as preening or territorial calling. I recorded the following data in addition to
behavior; general weather, time of day, substrate used (same vegetation categories as in
quadrat sampling), location in territory, the presence/absence of other jays, and food type
eaten (when possible). I collected data from 07:00 to 20:00 CST using a stop watch, tape
recorder, and 10 X 40 binoculars. An effort was made to observe marked individuals for
equal amounts of time, evenly spaced throughout the time of day and season. Individuals
to be sampled were predetermined each day to avoid observer biases that could result in
simply observing the first bird seen.

For each individual in each year, I calculated percentages of time spent in each behavior
in each main vegetation type listed above. These observed values were then compared to
the availability of this habitat in the individual’s territory. Time spent on substrates other
than woody vegetation (i.e., buildings, fences, the ground) was excluded from analyses.
Data were checked for normality (Lilliefors’ test) and homogeneity of variances (F-max
test), and then either paired comparison r-tests, f'-tests (comparisons of groups with unequal
variance), Mann-Whitney U tests, or two-way ANOVAs were used to examine statistical
significance where appropriate (Sokal and Rohlf 1981, Neave and Worthington 1988). Per-
centage data were arcsine transformed for parametric tests. All tests are two-tailed. Bonfer-
roni’s correction is applied to probability values for multiple tests that share portions of the
data set. Statistical values and associated probabilities are given for all comparisons, but
because of the conservative nature of these tests when dealing with small sample sizes,
trends for values approaching significance are considered for potential biological importance.

Methodological limitations. — Two approaches to smdying habitat use and behavioral time
budgets were considered for this study. One approach is to follow many individuals and
have fewer sampling sessions per individual. This method might reduce the variance among
individuals for each behavioral category, but it is impractical if behaviors vary temporally
(i.e., within a day, month, season). This variation would require equal sampling through
time for all individuals. Behavioral patterns show temporal variation in the Florida Scrub-
Jay (A. coerulescens-, DeGange 1976). The approach in this study was to follow fewer
individuals for more sampling periods, more or less regularly spaced through time (day and
season). The difficulties with this approach are limited sample sizes for statistical tests and
the potential for large among-individual variation.

RESULTS

General habitat use. — This first analysis examines whether each sex
uses each habitat component relative to its availability when all behaviors
are examined together. Bonferroni’s corrected probabilities needed for sta-
tistical significance in this test are P = 0.05/2, or P = 0.025. When
considering biodiversity habitat categories (Fig. lA), males used living

D. B. Burt • SCRUB-JAY HABITAT USE 715

A. BIODIVERSITY HABITAT TYPE USE

B. STRUCTURAL DIVERSITY HABITAT TYPE USE

Fig. 1. Observed and expected time percentages birds spent in each biodiversity (A)
and structural diversity (B) habitat category in 1989.

juniper (LJ) significantly less than expected on the basis of its availability
in 1989 (t = 7.7, P = 0.0045), and possibly in 1990 (t = 3.03, P =
0.039), but used living oak (LO), dead juniper (DJ), and dead oak (DO)
in proportion to their availability (1989: LO, t = 1.06, P - 0.37; DJ, t
= 2.0, P = 0.14; DO, t = 2.8, P = 0.068 df = 3. 1990: LO, r = 1.4, P
= 0.22; DJ, t' = \. 2, P = 0.29; DO, r' - 1.8, P = 0.14), although dead
oak approaches a statistically significant increase in use for 1989. Females
use each habitat in proportion to its availability (1989: LJ, t = 0.45, P =
0.68; LO, t = 0.58, P = 0.60; DJ, t - 0.0086, P = 0.94; DO, t = 0.66,
P = 0.56. 1990: LJ, t = 0.73, P = 0.50; LO, t = 0.0084, P = 0.99; DJ,
r' = 1.3, P = 0.27, DO, L = 0.19, P = 0.86).

Relative to woody vegetation structural diversity, males in 1989 used
isolated living (IL) habitat significantly more than expected {t = 22.9, P
= 0.0002) and dense living (DL) significantly less than expected (/ =

716

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

50

40

LiJ

1 30

LU

O 20
OC
lU
Q.

1 0

0

Fig.

5.5, P = 0.012) (Fig. IB). Females had a similar pattern for 1989 (IL, t
= 5.2, P = 0.014; DL, t = 2.8, P = 0.067) although this pattern was not
statistically significant. Neither use of isolated living nor dense living
habitats were different from expected in 1990 for males (IL, U = 14.5,
P > 0.2; DL, U = \6, P > 0.2) or females (IL, U = 16.5, P > 0.2; DL,
U = 12.5, P > 0.2). Use of isolated dead (ID) habitat was greater than
expected by males in both years (1989, t = 2.45, P = 0.092; 1990, t =
2.45, P = 0.070) but not by females in either year (1989, t = 0.72, P =
0.52; 1990, r = 0.61, F = 0.57). In each year, use of dense dead (DD)
habitat did not differ from expected for either males (1989, t = 0.25, P
= 0.82; 1990, t' = 0.98, P = 0.38) or females (1989, t' = 1.95, P =
0.15; 1990 t' = 1.15, P = 0.31).

Behavioral time budget. — In this analysis, the pair studied only in 1990
was excluded to create a balanced two-way ANOVA for comparisons of
each of the three main behavioral categories (Fig. 2). The ANOVA tests
differences between sexes and years for each behavior. For the preening/
inactive category no differences existed between sexes (F, ,2 = 0.11, P
> 0.50), or between years (F, ,2 = 0.044, P > 0.75), and no interaction
effect existed (F, ,2 = 0.38, P > 0.50). Males spent more time in sentinel
behavior than did females (F, ,2 = 8.6, P = 0.012) but no significant
difference existed between years (F, ,2 = 1.6, F = 0.23) nor was an
interaction effect seen (F,,|2 = 0.63, P > 0.25). Foraging data had sig-
nificant heteroscidacity and therefore a test of equality of means (Games
and Howell method, Sokal and Rohlf 1981) was performed in lieu of a
two-way ANOVA. Unplanned comparisons among all means showed no

BEHAVIORAL TIME BUDGETS

□ FORAGING

El SENTINEL
□ PREENIING/INACTIVE

MALE

k\N\S\S'
\^//////
k N \ \ \ \ N '

\////^//
k S \ S N X S '

\//^////

k \ \ N S X X '

k X X X X X X '

\^//////
k X X X X X X '

\////^//
.. X X X X X X ■

I.XXXXXX-
k X X X X X X '

k X X X X X X '

\///////
kxxxxxx'
\/////^/
k X X X X X X '

' V V V V X -

. j u

FEMALE

2. Percentages of time spent in the three main behavioral categories in 1989.

D. B. Burt • SCRUB-JAY HABITAT USE

717

A. PREENING/INACTIVE, BIODIVERSITY HABITAT TYPE

80

UJ

s

l-

ui

0
oc

UJ

01

60-

40-

20-

B. PREENING/INACTIVE, STRUCTURAL DIVERSITY HABITAT TYPE

Fig. 3. Observed and expected percentages of time birds spent in each biodiversity (A)
and structural diversity (B) habitat category in 1989 while preening/inactive.

significant differences in time spent foraging between either sexes or years
at P < 0.05.

Behavioral differences in habitat use. — The next series of analyses ex-
amine whether each sex uses each habitat component relative to its avail-
ability during each specific behavior. Bonferroni’s corrected probabilities
needed for statistical significance in these tests are P = 0.05/6, or P =
0.0083. For both years, neither males nor females spent more or less time
than expected preening or inactive in any category of habitat classified
by either biodiversity or structural diversity (Fig. 3). Four values do,
however, approach P = 0.05 in the biodiversity categories (male LJ, 1989,
t = 2.55, P = 0.084, 1990, t = 2.05, P = 0.11; male and female D.I
1989, U = 16, P = 0.05) and four values approach significance for the

718

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

UJ

LU

O

oc

LU

Q.

80

60

40

20

0

LIVING OAK LIVING JUNIPER DEAD OAK DEAD JUNIPER

A. SENTINEL, BIODIVERSITY HABITAT TYPE

LU

§

I-

LU

O

OC

LU

Q.

B. SENTINEL, STRUCTURAL DIVERSITY HABITAT TYPE

— [ ■ — — 1 I I

ISOLATED LIVING ISOLATED DEAD DENSE LIVING DENSE DEAD

Fig. 4. Observed and expected percentages of time birds spent in each biodiversity (A)
and structural diversity (B) habitat category in 1989 while sentinel.

Structural diversity categories (male DL 1990, t — 2.98, P — 0.041; female
IL 1989, r = 3.14, P = 0.052; male and female DD 1989, t - 3.46, P
= 0.04).

For sentinel time in biodiversity categorized habitats (Fig. 4A), males
spent less time in living juniper in 1989 (r = 6.28, P = 0.0082) and
possibly in 1990 {U = 25, P = 0.01). Females also may have spent less
sentinel time in living juniper in both years (1989, (7 = 16, P = 0.05;
1990, U = 25, P = 0.01). A pattern of more than expected use of dead
oak during sentinel behavior is suggested, but only the value for males
in 1990 is significant W = 7.611, P = 0.0016). In all other categories
and years, observed percentages do not differ from those expected based
on habitat availability. Sentinel time in structural diversity categories (Fig.
4B) indicates males, and possibly females, used isolated living trees less
than expected in 1990 (male, t = 4.62, P = 0.01; female, U = 25, P =

D. B. Burt • SCRUB-JAY HABITAT USE

719

A. FORAGING, BIODIVERSITY HABITAT TYPE

B. FORAGING, STRUCTURAL DIVERSITY HABITAT TYPE

m

60-

Z 40-
UJ
O
oc

UJ

Q. 20-

s \ N \

s S S \

y / ^

V V \ \

^ S S N

S S N

T

s s \ s

s \ s s

V N \ \

I

□

MALE

□

EXPECTED

H

FEMALE

ISOLATED LIVING ISOLATED DEAD DENSE LIVING DENSE DEAD

Fig. 5. Observed and expected percentages of time birds spent in each biodiversity (A)
and structural diversity (B) habitat category in 1989 while foraging.

0.01) but not 1989. Increased use of isolated dead vegetation for both
sexes in each year is suggested by the data, but values are not significant.
All other comparisons for sentinel time in structural diversity categories
do not differ from expected.

Foraging time does not differ from expected for either sex in either
year in any biodiversity categories of vegetation (Fig. 5A). Five values
approach significance in some of the rarer habitat categories (male and
female DO 1990, t' = 4.50, P = 0.01 1 and f = 3.59, P = 0.023; female
DO 1989, t = 2.72, P = 0.073; male DJ 1990, (7 = 21, P = 0.1; female
DJ 1989, t/ = 16, P = 0.05). Foraging time in structural diversity cate-
gories of vegetation shows males, and possibly females, used isolated
living vegetation more than expected in 1989 (male, t = 7.78, P = 0.0044;
female, t = 2.43, P = 0.094) but not in 1990 (Fig. 5B). Males, and
possibly females, also used the fairly rare category of isolated dead trees

720

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

less than expected in 1990 (male, t = 5.63, P = 0.0049; female, t = 2.79,
P = 0.049) but not in 1989. Three other values approach significance in
rare habitats (male and female DD 1990, t = 3.55, P = 0.024 and t =
2.77, P = 0.05; female DD 1989, / = 5.074, P = 0.015).

DISCUSSION

Habitat use and the concept of ecological constraints are central to
many theories concerning the evolution and maintenance of cooperative
breeding in birds. Specialization in habitat use in long-lived birds and the
resulting potential to saturate this habitat is central to both the habitat
saturation and marginal habitat models. The habitat saturation model
states that suitable breeding habitat slots become filled, forcing young
individuals to delay dispersal (Selander 1964, Brown 1974). The marginal
habitat model builds on this concept by adding an additional constraint,
namely scarcity of habitats of marginal quality. Scarcity of marginal hab-
itats reduces the possibility of individuals dispersing and roaming as non-
breeding floaters (Verbeek 1973, Koenig and Pitelka 1981, Emlen 1982,
Woolfenden and Fitzpatrick 1984, Fitzpatrick and Woolfenden 1986). One
prediction of the marginal habitat model is a high proportion of territories
of high-quality habitat relative to those of marginal quality in cooperative
breeding populations. As an alternative to these extrinsic constraint mod-
els, the benefits-of-philopatry (BOP) model stresses the importance of
intrapopulational variation in territory quality and is not dependent on
complete saturation of either breeding or marginal/floating habitats but is
instead based upon the intrinsic decision making processes of young birds
regarding their dispersal options relative to the quality of their natal ter-
ritory (Stacey and Figon 1987, 1991). According to this model, individ-
uals born in high quality natal territories choose to delay dispersal and
remain at home, thereby increasing their chances of inheriting the natal
territory or occupying another nearby territory of equal quality. These
individuals also may avoid increased chances of mortality associated with
breeding or floating in lower quality habitats. Koenig et al. (1992) intro-
duced an elegant, more inclusive model, which more formally distinguish-
es between extrinsic and intrinsic factors influencing an individual’s de-
cision on whether to delay natal dispersal. This delayed dispersal thresh-
old model identifies five parameters that are many times jointly involved
in the probability of an individual delaying dispersal. A complete expla-
nation of this model is not possible here, however, in regard to this paper,
one of its parameters is the distribution of territory quality as modeled in
either the marginal habitat or BOP models.

An extension of current models. — I believe the logic outlined in each
of these models can be extended to predict specific habitat use patterns

D. B. Burt • SCRUB-JAY HABITAT USE

721

EKpectations:

f

Habitat Specialist

1. Habitat Rare/Patchy

2. Low Uariation in Terr.
Composition/quality

Cooperatiue Breeders

t

Specialist

Micro-habitat Specialist

1. Limiting Resource
Determines Quality

2. High Uariation in Terr.
Composition/ Quality

EKpectations:

f

Specialist

1. Habitat lUidespread

2. Low Uariation in Terr.
Composition/Quality

Non-cooperatiue Breeders
I

Generalist

1. Micro-habitat Used
Relatiue to Huailabilty

2. High Uariation in Terr.

Composition

Fig. 6. Model of habitat-use expectations for cooperative and non-cooperative breeding
populations.

for cooperative and non-cooperative breeding populations (see Fig. 6).
One might expect cooperative populations to show restricted, habitat spe-
cialist patterns of habitat use in one of two ways. First, a population may
show a strict requirement to live in a habitat characterized by a certain
vegetation assemblage. This habitat-use inflexibility would lead to de-
mographic conditions favoring the evolution of cooperative breeding only
if the habitat in question is either rare in comparison to other assemblages
in the same geographic region or is very patchily distributed. With such
broad habitat-use specificity, variation among usable territories in habitat
composition and quality may be small. In extreme cases, all habitat patch-
es in an area are either acceptable or unacceptable for breeding or floating,
with no intermediaries. These predictions fit the habitat saturation model
and, depending on the degree of habitat specificity (i.e., does marginal
habitat even exist for the species), the marginal habitat model. The second
habitat-specialist pattern to be expected from a cooperative population is
the required use of a particular aspect of the habitat, a target of special-
ization, that serves as a limiting resource while other aspects of habitat
composition have reduced importance and may vary independently
among territories. In this case, general habitat composition may vary
greatly but be of little importance. The presence of the limiting resource
is the crucial feature of successful breeding in this situation and is the
currency by which territory quality is measured. Examples of such targets

722

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

of specialization in cooperative species include cavities for both Green
Woodhoopoe {Phoeniculus purpureas, Ligon and Ligon 1988) and Red-
cockaded Woodpecker {Picoides borealis, Kulhavy et al. 1995) for roost-
ing and nesting, and granary trees for acorn storage by Acorn Wood-
peckers (Melanerpes formicivorus, Koenig and Mumme 1987). If territory
quality varies greatly and is tightly correlated with reproductive success,
as predicted by the BOP model, a cooperative breeding system may
evolve as a queue of non-dispersing individuals builds to fill the small
number of high quality breeding territories. In this latter case, it is nec-
essary to demonstrate that individuals utilize the limiting resource dis-
proportionately to its availability in the territory.

Non-cooperative populations also would be expected to show one of
two habitat-use patterns (Fig. 6). First, non-cooperative populations also
may show specialist habitat-use patterns but only in a super-abundant,
widespread habitat type. In this case, territory composition should not
change much among territories. Alternatively, the population may show
generalist habitat-use patterns with significant variation in habitat com-
position among territories. In the latter case, it is expected that individuals
use each habitat component in relation to its availability and no target of
specialization is expected.

Measuring territory quality has proven difficult and is flawed in many
studies because reproductive success is used as the primary currency
which fails to distinguish between habitat quality and the quality of in-
dividuals (Koenig et al. 1992). However, I believe documentation of hab-
itat-use patterns allows closer examination of the mechanisms leading to
changes in breeding systems. This approach must first classify populations
as habitat specialist/generalist relative to other populations based on
knowledge of the distribution of each major habitat component across
territories and detailed habitat-use patterns of these major components.
Then, given the two predictions concerning patterns of habitat-use for
both cooperative and non-cooperative populations and information on
population variation in breeding systems, we can examine the models
concerning the ecological bases of cooperative breeding systems more
closely.

Behavioral patterns in biodiversity habitat categories. — Texas Western
Scrub-Jays use the biodiversity categories of woody vegetation as gen-
eralists with one exception: males use living juniper less than expected
with the excess time transferred to dead and living oaks (Fig. lA). The
behavioral explanation for this deviation is the large fraction of time
males engage in sentinel behavior and the relatively poor visibility attri-
butes of (usually short) living juniper and the good visibility attributes of
(usually tall) oak trees. Females also use these habitat categories in this

D. B. Burt • SCRUB-JAY HABITAT USE

723

pattern, but because so little of their time is spent in sentinel behavior,
deviations from general use of this habitat type are not as large (Figs. 2,
4A). Males also may have used living juniper while preening/inactive less
than expected (Fig. 3A). The habitat in which this behavior occurs may
be influenced by the habitat in which the behavior preceding it occurred.
Individuals may simply preen wherever they find themselves and, in gen-
eral, males spent little time in living juniper. Values approaching signif-
icance in differential use of several of the rarer biodiversity categories
should not be given much weight because of the potential for sampling
srror in categories where behaviors appear also to be rare (e.g., foraging
in DO and DJ, Fig. 5A).

Behavioral patterns in structural diversity categories. — Observed val-
ues deviated from expected values in more structural diversity categories
than in biodiversity categories (Fig. IB). As noted above, males and fe-
males used isolated living vegetation more than expected at the expense
of dense living trees in 1989 but not in 1990. Similarly males, and pos-
sibly females, used isolated living vegetation while foraging more than
expected in 1989 but not 1990 (Fig. 5B). This pattern could have been
due to a food resource shift between years; however, no data exist to test
this possibility. The two most frequently captured prey items were large
katydids (probably Microcentrum sp.) and walking sticks (probably Me-
gaphasma sp.), but their distribution among the different habitat catego-
ries is unknown. Deviations from expected in habitat use of preening/
inactive behavior seem to mirror those of foraging behavior in females
but not in males (Fig. 3B). Again, if individuals preen wherever they find
themselves, foraging would influence where the preening/inactive behav-
ior occurred in females more than in males, because females spent a large
fraction of their time foraging, while males were influenced by both for-
aging and sentinel behaviors (Fig. 2).

Males also may have used isolated dead vegetation more than expected
in both years (Fig. IB). Both sexes increased use of this habitat for sen-
tinel behavior (Fig. 4B); however, not all deviations from expected are
close to statistical significance. Much of the isolated dead habitat is equiv-
alent to the dead oak category of the biodiversity categorization, and a
functional relationship of isolated dead trees to sentinel behavior probably
exists. Both males and females also used isolated living vegetation for
sentinel behavior less than expected in 1990 but not 1989. This difference
is not observed in general habitat use (Fig. IB), and reasons for this
deviation while in sentinel behavior are not apparent.

Habitat use patterns of other scrub- jay populations. — Peterson and
Vargas (1992) provide a thorough analysis of the diversity of habitat types
used by birds in the scrub-jay species complex. Scrub-jays use a wide

724

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

range of habitats including: oak, juniper, pinyon and desert woodlands;
riparian brush; oak-palmetto scrub; pine-oak, alpine pine-spruce, tropical
thorn forests; and mangrove swamps. This level of variation in habitat
use is that which occurs across populations. Levels of variation in habitat
use within populations is highly population specific. Florida, coastal Cal-
ifornia, and Great Basin populations are very habitat specific. This spec-
ificity was hypothesized to be related to the importance of either acorns
or pinyon seeds in the diet of these populations. Baja California and
southern Mexico populations are more general in habitat use, and these
habitats frequently have no obvious replacements for acorns or pinyon
seeds. Koenig et al. (1992) discuss the importance of mast production
relative to differences in social systems in scrub-jays. In the non-coop-
erative California population of Western Scrub-Jays, years of poor acorn
abundance result in increased reproductive failure, adult mortality and
territory abandonment. Oaks are distributed in patches in California and
floaters typically move freely among breeding territories in search of areas
of high acorn abundance (Carmen 1988). Concerning the Florida Scrub-
Jay, mast production is more stable in Florida and this predictable, evenly
distributed, resource is easily defended and floaters are not tolerated in
breeding territories. Koenig et al. (1992) also speculate that differences
in mast production partially explains differences in jay use of the optimal,
recently burned oak patches and unoccupied, dense, unburned patches of
oaks.

Presence of other jay species affects range of habitat use in Florida
(Blue Jays, [Cyanocitta cristata]), central California (Steller’s Jays [C.
stelleri]), and New Mexico/Arizona (Gray-breasted Jays) but does not
limit use in southern Mexico (Magpie-Jays [Calocitta spp.]). However,
the diversity of habitats utilized by Scrub-Jays in Baja California cannot
be solely explained by the absence of other jay species because this pop-
ulation uses habitats that are not used on the mainland by any jay species
(Peterson and Vargas 1992).

Habitat use and predictions of models for the Texas population. — In
most habitat-use categories examined in this study, the central Texas pop-
ulation of Western Scrub-Jays appears to use its habitat as a generalist.
The only clear exception to this statement is the apparent specialization
by males in using oak, particularly isolated snags, for sentinel behaviors.
As a non-cooperative population, one of two habitat-use patterns was
expected (1) specialist habitat-use patterns in an abundant, widespread
habitat type, with little variation among territories in habitat composition
(matching the habitat saturation or marginal habitat models), or (2) gen-
eralist habitat-use patterns with the potential of significant variation in
habitat composition among territories (as predicted under the BOP

D. B. Burt • SCRUB-JAY HABITAT USE

725

model). Texas Western Scrub-Jays show an interesting combination of
habitat-use patterns supporting both predictions. The only resources that
males utilize as a specialist, oak trees during sentinel behavior, are fairly
widespread and would not be considered a limiting resource. In the re-
maining habitat categories Texas Western Scrub-Jays act as generalists,
using the habitat in relation to its availability, even though variation in
habitat composition among territories is considerable. Mast production,
an important resource to many jay populations, was not measured in this
study because of the season in which data were collected.

Areas in need of future research. — Comparisons between cooperative
and non-cooperative populations using more detailed habitat-use studies,
when combined with demographic and phylogenetic information, will al-
low closer examination of the various models of the evolution and main-
tenance of cooperative breeding in birds. It is clear that habitat constraints
are important in the Florida Scrub-Jay and examining habitat use in their
relict scrub habitat may reveal why very dense, unburned scrub and other
habitats are inadequate for maintaining populations (Woolfenden and Fitz-
patrick 1984). This subject is the focus of a detailed habitat-use study
currently underway (R. Curry, pers. commun.) This type of detailed study
also would be most valuable for examining correlations between ecolog-
ical constraints and cooperative breeding in the southern Mexico popu-
lation of the Western Scrub-Jay. This population appears to use a wide
range of habitat types which do not appear saturated (Burt and Peterson
1993). A detailed habitat-use study might identify specific microhabitats
that limit where successful territories can be maintained. These micro-
habitats may be found to exist within several of the more broadly defined
habitat types. As shown in this study, habitat-use patterns can be used to
test predictions concerning the role ecological constraints play in deter-
mination of individual dispersal patterns and, potentially, the evolution of
breeding systems.

ACKNOWLEDGMENTS

I thank the Texas Dept, of Fish and Wildlife for permission to conduct this study in the
Kerrville State Park. Special thanks go to Tim Hufstedler and the rest of the park staff for
their cooperation and a.ssistance. Terry Maxwell allowed me to apply for a banding sub-
permit under his permit number. Financial assistance was provided by grants from the Amer-
ican Museum of Natural History (Chapman Grant), Univ. of Kansas Museum of Natural
History (Panorama Society grants), Texas Ornithological Society, the Univ. of Kansas Dept,
of Systematics and Ecology, and the Univ. of Arizona Research Training Group for the
Analysis of Biological Diversification. Richard Johnston and Phil Humphrey were particu-
larly helpful in my attempts to find financial support. My committee members, Ken Armi-
tage, Mike Gaines, Doug Siegel-Causey, and Richard Johnston (Chair) provided guidance
and careful consideration of early drafts of this manu.script as did Robert Curry, Ronald
Mumme, Daniel Papaj, and Glen Woolfenden, members of the Nathaniel Goss Society

726

THE WILSON BULLETIN • Vol. JOS, No. 4, December 1996

(Univ. of Kansas), and two anonymous reviewers. Persons who were generous with their
time in sharing ideas and/or field work include Tristan Davis, Sue Fairbanks, John Koprows-
ki. Brad Livezey, Terry Maxwell, and Mike Stokes. I also thank Kim Burt and my family
for their continued support.

LITERATURE CITED

American Ornithologists’ Union. 1995. Fortieth supplement to the American Ornithol-
ogists’ Union Checklist of North American Birds. Auk 1 12. (in press)

Avery, T. E. 1985. Interpretation of aerial photographs, fourth ed. Burgess Publ. Co.,
Minneapolis, Minnesota.

Brown, J. L. 1974. Alternate routes to sociality in jays — with a theory for the evolution
of altruism and communal breeding. Am. Zool. 14:63—80.

. 1987. Helping and communal breeding in birds: ecology and evolution. Princeton

Univ. Press, Princeton, New Jersey.

Burt, D. B. 1992. Comparative Scrub Jay behavioral ecology. M.A. thesis, Univ. of Kansas,
Lawrence, Kansas.

AND A. T. Peterson. 1993. Biology of the cooperative-breeding Scrub Jays {Aphel-

ocoma coerulescens) of Oaxaca, Mexico. Auk 110:207—214.

Carmen, W. J. 1988. Behavioral ecology of the California Scrub Jay {Aphelocoma coeru-
lescens californica): a noncooperative breeder with close cooperative relatives. Ph.D.
diss., Univ. of Calif., Berkeley, California.

DeGange, a. R. 1976. The daily and annual time budget of the Florida Scrub Jay. M.A.
thesis, Univ. of South Florida, Tampa, Florida.

Emlen, S. T. 1982. The evolution of helping behavior. 1. An ecological constraints model.
Am. Nat. 119:29-39.

Fitzpatrick, J. W. and G. E. Woolfenden. 1986. Demographic routes to cooperative breed-
ing in some New World jays. Pp. 137-160 in Evolution of animal behavior (M. Nitecki
and J. Kitchell, eds.). Univ. of Chicago Press, Chicago, Illinois.

Koenig, W. D. and R. L. Mumme. 1987. Population ecology of the cooperatively breeding
Acorn Woodpecker. Princeton Univ. Press, Princeton, New Jersey.

AND F. A. PiTELKA. 1981. Fcological factors and kin selection in the evolution of

cooperative breeding in birds. Pp. 261—280 in Natural selection and social behavior:
recent research and new theory (R. D. Alexander and D. Tinkle, eds.). Chiron Press,
New York, New York.

, , W. J. Carmen, R. L. Mumme, and M. T. Stanback. 1992. The evolution

of delayed dispersal in cooperative breeders. Quart. Rev. Biol. 67:1 1 1-150.

Kulhavy, D. L., R. G. Hooper, and R. Costa (eds.). 1995. Red-cockaded Woodpecker:
recovery, ecology, and management. Center for Applied Studies in Forestry, Nacog-
doches, Texas.

Ligon, D. j. and S. H. Ligon. 1988. Territory quality: key determinant of fitness in the
group-living Green Woodhoopoe. Pp. 229-253 in The ecology of social behavior (C.
N. Slobodchikoff, ed.). Academic Press, San Diego.

Miller, H. A. and S. H. Lamb. 1985. Oaks of North America. Naturegraph Publishers
Inc., Happy Camp, California.

Neave, H. R. and P. L. Worthington. 1988. Distribution-free tests. Unwin Hyman Ltd.,
Boston, Massachusetts.

Peterson, A. T. and D. B. Burt. 1992. Phylogenetic history of social evolution and habitat
use in the Aphelocoma jays. Anim. Behav. 44:859—866.

AND N. Vargas. 1992. Ecological diversity in Scrub Jays (Aphelocoma coerules-

D. B. Burt • SCRUB-JAY HABITAT USE

727

cens). Pp. 309-317 in Selected faunistic groups of Mexico (T. P. Ramamoorthy, R. Bye,
A. Lot, and J. Fa, eds.). Oxford Univ. Press, New York, New York.

PiTELKA, F. A. 1945. Pterylography, molt, and age determination of American jays of the
genus Aphelocoma. Condor 47:229-260.

Selander, R. K. 1964. Speciation in wrens of the genus Campylorhyncus. Univ. Calif.
Publ. Zool. 74:1-224.

SOKAL, R. R. AND F. J. Rohlf. 1981. Biometry. 2nd ed. Freeman Press, San Francisco,
California.

Stacey, P. B. and J. D. Ligon. 1987. Territory quality and dispersal options in the Acorn
Woodpecker, and a challenge to the habitat saturation model of cooperative breeding.
Am. Nat. 130:654-76.

AND . 1991. The benefits-of-philopatry hypothesis for the evolution of co-

operative breeding: variation in territory quality and group size effects Am Nat 137"
831-846.

Verbeek, N. a. M. 1973. The exploitation system of the yellow-billed magpie. Univ. Calif.
Publ. Zool. 99:1-58.

Woolfenden, G. E. and j. W. Fitzpatrick. 1984. The Florida Scrub Jay: demography of
a cooperative-breeding bird, Monogr. Pop. Biol. No. 20. Princeton Univ. Press, Prince-
ton, New Jersey.

Wilson Bull.. 108(4), 1996, pp. 728-739

NUTRITIONAL VALUE OF WINTER FOODS FOR
WHOOPING CRANES

Jay T. Nelson, ' R. Douglas Slack,^ and
George F. Gee'

Abstract. — We measured metabolizable energy and digestibility of Whooping Crane
{Grits americana) winter foods (blue crab [Callinectes sapidus]), common Rangia clam
(Rangio cuneata), wolfberry fruit (Lycium carolinianum [wolfberry]), and live oak acorn
(Ouercus virginiana [acorn])] with feeding trials to captive-reared Whooping Cranes. Ap-
parent metabolizable energy coefficients (expressed as %) were for crab (34.1), Rangia clam
(75.0), wolfberry (44.8), and acorn (43.2). Digestion coefficients for protein were lower for
plant foods (48.9 and 53.4) than for animal foods (69.4 and 75.2). Digestion coefficients
for total lipid differed among foods: highest and lowest lipid digestibility was for acorn
(87.2) and wolfberry (60.0), respectively. We also determined total energy and percent pro-
tein and lipid of the four foods and stout razor clam (Tagelus plebeius)\ gross energy was
2-5 X higher for acorn and wolfberry on a dry-weight basis than for blue crab and stout
razor clam. Crude protein was 2-3 X higher for blue crab than for wolfberry and stout razor
clam. Wolfberry ranked the highest of five foods for metabolic energy and total lipid nutrient
availability per kg of food ingested, and blue crab ranked highest for crude protein avail-
ability. Received I December 1995, accepted 10 April 1996.

Investigators have documented foods eaten by Whooping Cranes on
their Texas Coastal wintering ground (Aransas National Wildlife Refuge
[ANWR]). They have determined that Whooping Cranes rely on blue
crabs {Callinectes sapidus), stout razor clams {Tagelus plebeius), wolf-
berries {Lycium carolinianum), and acorns {Ouercus virginiana) for their
energy and nutrient needs (Stevenson and Griffith 1946; Allen 1952,
1954; Shields and Benham 1969; Uhler and Locke 1970; Blankinship
1976; Hunt and Slack 1987, 1989). In order to determine if winter food
resources are adequate, managers should understand metabolizable ener-
gies, nutrient digestibilities, and nutritional values of these foods. Our
objectives were to determine (1) metabolizable energy and nutrient di-
gestibility coefficients for winter Whooping Crane foods, by feeding them
to captive Whooping Cranes, and (2) the energy and nutritional content
of these foods.

METHODS

We conducted feeding trials with captive-reared Whooping Cranes in spring 1994 at
Patuxent Environmental Science Center (PESC), Laurel, Maryland. Because of the endan-

' National Biological Service, Patuxent Environmental Science Center, 1 1510 American Holly Drive,
Laurel. Maryland 20708.

- Dept, of Wildlife and Fisheries Sciences, Texas A&M Univ., College Station. Texas 77843.

' Present addre.s,s: Hawaii Field Station. P.O. Box 44, Bldg 344, Hawaii National Park, Hawaii 96718.

728

Nelson et al. • WHOOPING CRANE DIET

729

gered status of Whooping Cranes (U.S. Eish and Wildl. Serv. 1994), we used captive-reared
birds in our digestibility studies. Cranes were raised on pelleted diets, could not be induced
to eat whole foods voluntarily, and unacceptable risks were associated with forced feeding.
The amount of wild foods we were able to collect influenced the amount and length of time
food was fed to individual birds. The amount of time captive birds could be maintained
under experimental conditions was also limited because we were concerned about behavioral
problems and risk of injury. Therefore, treatment diets were mixed with control diets and
fed to captive Whooping Cranes during short time periods. Wild foods were provided in
pelleted feeds at a 30% level of substitution because of concerns that captive cranes might
be adversely affected by eating feeds with higher levels of wild foods and might reduce
feed consumption at levels higher than 30% (Muztar et al. 1977). Thus, we were unable to
conduct validation trials using 100% wild foods.

We housed four subadult (1-yr-old) and one adult (3-yr-old) Whooping Cranes in adjacent
3.4 X 2.7 X 3.1-m indoor pens at the PESC during feeding trials from 2 March to 27 April
1994. Connected to each indoor pen was a 9.1 X 2.9 X 2.9-m outdoor runway. Each' indoor
pen was equipped with a gravity feeder, water bucket, and a bowl for granite grit. Eloors
were covered with smooth rubber matting to collect excreta. Cranes were allowed to move
between indoor and outdoor runways on days when they were not fed study diets and were
housed indoors at night. Indoor photoperiod was maintained at 10.5D:13.5L, and indoor
temperature ranged from 13°C to 25°C. Cranes were weighed to the nearest 100 g with a
spring scale (1) when first moved to the Propagation Building, (2) six days after the initial
move, and (3) when the study was completed. Daily at 06:00, we fed 1000 g each of blue
crab, Rangia clam, acorn, and reference feeds and 400 g of wolfberry feed.

We collected acorns, blue crabs, wolfberries, and Rangia clams from Whooping Crane
foraging areas on the ANWR Blackjack Peninsula and Matagorda Island, Texas, and adja-
cent coastal areas between 8 October 1993 and 17 February 1994 (Labuda and Butts 1979,
Stehn 1994b, U.S. Fish and Wildl. Serv. 1994). Acorns were also collected at College
Station, Texas, to supplement acorns collected at the ANWR. Whooping Cranes prefer stout
razor clams (Blankinship 1976), but we could not collect sufficient numbers; instead, we
collected the less preferred Rangia clam for our feeding trials. Foods were stored at — 20°C
until they were dried at 55°C for 24-36 h and ground in a Wiley or Hammer mill to pass
through a 20-mesh screen. Study diets were prepared by combining 30% dry weight of blue
crab, Rangia clam, wolfberry, or acorn with 70% commercial crane breeder feed. We added
bentonite (0.5%) to each diet as a hardening agent and an inert tracer, chromic oxide (0.5%),
to determine metabolizable energy and nutrient digestibility coefficients (Karasov 1990).
Moistened diets were mixed in a Hobart food mixer, formed into pellets (0.48-cm diameter)
using a Hobart food processor, and air dried and stored in plastic bags at 5°C until feeding.

Apparent metabolizable energy coefficients (MEC*) for test feeds were determined using
the equation of Karasov (1990): MEC* = [GE, - (%T|/%Te)GEJ/GE,,; where GEj and GE,.
equal, respectively, the gross energy content (cal g'' dry mass) of feed (intake) and excreta,
and %T, and %T^ equal, respectively, the percent of chromic oxide tracer in feed and excreta.
Apparent digestible energy coefficients for test feeds (DEC*) were determined by subtracting
percent gross energy of uric acid in excreta (% uric acid multiplied by 2730 cal g ') (Lide
1994) from total gross energy of excreta (GEJ and substituting energy excreta (minus uric
acid energy) into the equation for MEC*. Individual MEC*s for test ingredients were de-
termined by equation 2: MEC*, = '®%o X (MEC*f — [0.7 X MEC*J); where MEC*, equals
the MEC* for the test ingredient, MEC*, equal the MEC* for the test feed (feed with test
ingredient added), and MEC*^ equal the MEC* for the reference feed (indicator-marked
crane breeder feed) (Wilson and Poe 1985). Apparent digestible energy coefficients for test
ingredients (DEC*,) were determined by substituting energy excreta (minus uric acid energy)

730

THE WILSON BULLETIN • Vo/. 108, No. 4, December 1996

into the equation for MEC*j, where DEC*f equal the DEC* for the test feed and DEC*^
equal the DEC* for the reference feed: DEC*; = (100/30) X [DEC*f — (0.7 X DEC*,)].

Apparent dry matter, crude protein, and total lipid digestibility coefficients for test feeds
and reference feed (ADCf) were determined using an index equation by Lloyd et al. (1978):

ADCf = 100 -

(% Indicator in feed) (% Nutrient in feces)

100 X X

% Indicator in feces % Nutrient in feed

on the basis of the ratios of indicator in feed and feces and nutrient in feces and feed.
Apparent digestibility coefficients for test ingredients (ADCj) were determined by substitut-
ing apparent dry matter, crude protein, and total lipid digestibility coefficients for test feeds
and reference feed into equation 2. Digestible protein for test feeds was determined after
subtracting uric acid nitrogen in excreta from total Kjeldahl nitrogen in excreta (Rotter et
al. 1989).

MEC*s for wild foods with low digestibilities have been shown to be low and in error
for waterfowl fed mixed diets (Karasov 1990). However, Muztar et al. (1977) found that
apparent dry matter digestibility and apparent metabolizable energy values were higher at
the 30% level of substitution than at 10%, 20%, or 40% for alfalfa and five species of
aquatic plants and that 30% substitution agreed most closely with regression methods to
predict digestibility. Our model for determination of MEC* is based upon the assumption
that relations between MEC* values are additive and that there are no synergisms or asso-
ciative effects for MEC* and nutrient digestibilities due to mixing feed ingredients (Cho et
al. 1982, Wilson and Poe 1985).

Diets were fed following a Latin Square design with five, four-day feeding trials. We fed
cranes an unmarked breeder diet during a three-day rest period between trials to allow them
to excrete all indicator-marked food. Whooping Cranes were held indoors during feeding
trails except for four, 30— 45-min periods starting at 06:00, 10:00, 14:00, and 18:00 when
they were moved outdoors while we collected excreta from the rubber matting and cleaned
floors. Eecal samples were collected separately for each bird and for each collection period
from indoor pens; samples for each bird and collection period were pooled. Samples col-
lected on days three and four of each feeding trial were used to estimate metabolizable
energy and digestibility coefficients. We used change in indicator concentration on the first
day indicator-marked diets were fed to evaluate rate of food passage (length of time un-
marked feed was retained in the gut).

We collected samples of blue crabs, Rangia clams, wolfbenies, acorns, and stout razor
clams from six different sites on the ANWR (Stehn 1994b, U.S. Eish and Wildl. Serv.
1994). Each collection site was 1-10 ha depending on the relative density and distribution
of foods collected, and one food type was collected. We collected nine samples of each
food from points on transect lines located at random within each collection area. Sample
collections of food items included (1) acorns — 8-20 of varying sizes from >:five plants from
Dagger Point and (2) along East Shore Road to the west of Sundown Bay, (3) stout razor
clam.s — 10-20 individuals 3-4 cm long from Cedar Lake, Matagorda Island, (4) Rangia
clams — 8-10 individuals 3-5 cm long from Indian Head Point, St. Charles Bay, (5) wolf-
berries — >50 berries from >10 different plants from Sundown Bay, and (6) blue crabs —
2—3 crabs (carapace width >10 cm) from Long Lake.

We multiplied average nutrient values for foods by the appropriate digestibility coeffi-
cients and ranked foods for available nutrient on a dry-weight basis. Digestion coefficients
for Rangia clam have been used in calculations for stout razor clam, assuming that digest-
ibilities for these clam species are similar. Ripe acorns collected from Dagger Point had
fallen to the ground and were scorched during a prescribed burn several days before they

Nelson et al. • WHOOPING CRANE DIET

731

were collected and represented the type and quality of acorns Whooping Cranes ate (Hunt
1987). We used nutrient values of these acorns in food quality calculations.

Nutritional analyses were conducted using fresh excreta samples (lipid analysis) and dried
ground excreta, feed, and food samples. Digestibility calculations for excreta and feed nu-
trient are expressed on a dry-matter basis corrected to standard drying time (3 h) and
temperature (125 C) (Pomeranz and Meloan 1987). Calculations for food nutrient and total
energy and nutrient availability of whole foods are expressed on a dry-weight basis.

Gross energy was determined using a Parr micro-bomb adiabatic calorimeter. Gross en-
ergy for Rangia clam was too low to be accurately determined by bomb calorimetry, and
energy values were determined by multiplying percent crude protein by 4000 cal g ' and
percent total lipid by 9000 cal-g ■ and adding the results. Total nitrogen was determined by
the micro-Kjeldahl method (Helrich 1990). Crude protein was calculated by multiplying
total Kjeldahl nitrogen by 6.25. Total lipids were determined by chloroform/methanol ex-
traction for samples homogenized 4 min in a mechanical homoginizer (Folch et al. 1957).
Total lipid in freshly homogenized foods was 6% less than for dried ground foods, and total
lipid in fresh excreta samples was 4% less than for dried ground excreta samples. Drying
foods and excreta at low temperatures and grinding did not lower lipid yield, but enhanced
lipid yield compared to extraction of fresh samples. Uric acid was determined colorimetri-
cally for dried excreta samples (Marquardt 1983). Chromium was determined by atomic
absorption spectrophotometry after excreta samples were ashed in a muffle furnace and
digested in nitric acid (Helrich 1990). Ash was determined by combusting dried samples
for 4 h at 500°C in a muffle furnace (Helrich 1990). Total phenols were determined color-
imetrically using a gallic acid standard for dried acorn and wolfberry samples extracted 30
min in 70% aqueous acetone (Singleton and Rossi 1965, Hagerman 1988). We determined
chitin in blue crabs by sequential acid and alkali digestions (Black and Schwartz 1950).

Each bird was considered the unit of replication for statistical analysis of metabolizable
energy and digestibility coefficients. Metabolizable energy and nutrient digestibility coeffi-
cients were analyzed for ranked data by three-way Kruskal-Wallis ANOVA (GEM proce-
dure, SAS System) to test for the effects of study feed. Whooping Crane, and feeding trial
on each variable. Kruskal-Wallis and Wilcoxon tests (SAS System) were conducted using
ranked data to test for nutrient differences between feed ingredients and foods collected
from Whooping Crane foraging areas. We compare nutrients of foods collected from Whoop-
ing Crane foraging areas by Kruskal-Wallis tests. Tukey’s means comparison procedure was
used for the separation of means when ANOVA results were significant. Means differences
are reported at P < 0.05.

RESULTS

Food consumption rarely exceeded 200 g day"', indicating that suffi-
cient feed was provided to study birds. Study birds maintained body
weight during feeding trials from 0-2.3% of their body weight at six days
after the initial move. Differences of nutrient levels between test ingre-
dients used in feeds and nutrients in food samples (for the same foods)
were less than 10% but were significant in several instances (Table 1). At
difference levels of 10%, nutrients in test ingredients were representative
of nutrients in foods eaten by wild Whooping Cranes.

Chromic oxide indicator in fecal samples collected at 18:00 EST on
day 1 of each feeding trial was 937 X higher than indicator in fecal sam-
ples collected prior to first exposure to indicator-marked study feeds at

732

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

UJ

X

H

S

O

Qi.

LI.

Q

LD

H

U

LU

-J

.J

o

U

C/5

Q

o

O

U.

Q

Z

<

!/)

H

Q

UJ ^

cc <
o CQ

H

00

LU

H

1^

o

Lu

LU

[:

<

>-

Q Q
K. ^

LU

J o

J ^

. , O 0^
CQ H uj

< Q

< 3

i ^

LU

H z

H

O

Qi

CU

[Xl

Q

_)

<

H
<
D Z

ai

u ^
- 00
X z
00 <

0- <

LU

u

a:

LU

D-

<

m

U

<

<

<

<

<

a\

aC

ON

On

ON

00

ON

a^

o

(N

00

ON

m

(N

O

O

CN

—

o

nC

d

d

d

d

d

d

d

d

d

+1

+1

+1

+1

+1

+1

+1

+1

+1

ON

lO

o

00

in

00

ON

yn

CNl

(N

—

CO

(n

(N

d

yn

d

d

d

d

rn

d

in

<

< CQ

CQ <

CQ

<

z~s

Os OS

os OS

Os

OS

m

00

in

cn CN

m —

sO

CN

o

O

o o

O CO

—

d

d

d

d d

d d

d

d

+1

+1

+1

+1 +1

+1 +1

+1

+1

CN

so

in

— Tt

in so

Os

00

in

CN

00

d

d

» —I

sd Os

00

— *

— —

CO

<

<

<

<

<

m

<

CQ

5^

O^

O^

On

On

OS

Os

r7

CO

(N

CO

CO

r-

o

—

—

o

o

in

o

o

d

d

d

d

d

d

d

d

d

+1

+ !

+1

+1

+1

+1

+1

+ 1

+1

Tf

CN

o

r-

r-

__

CO

os

NO

SO

CO

m

00

q

O

in

(N

CN

(N

so

sd

os

d

cd

OS

Os

m

<

ffl

m

<

m

CD

<

<

<

in

m

in

in

in

in

in

in

m

00

CO

m

CN

00

in

r-

in

CO

CN

in

00

O

o

o

o

o

o

d

d

d

d

d

d

d

d

d

+1

+1

+1

+1

+1

+1

+1

+1

+1

NO

os

o

NO

CO

o

o

p'

Os

00

(N

in

CO

>— ■

o

r-

m

cn

Os

in

Os

d

CN

d

00

d

d

d

(N

CN

d

—

9

>■

o

LU

Z

PJ

to

to

O

at;

a

o

o

cS

o

■o

C O

*0 o
(U u

>- o

W) ^

<z

£

'o

cu

§

u

a;

2

OD

0^)

Ti

&

Q

lac

W)

c

0

0-

C

c

.2

.2

"O

c3

5

■5

CQ

"S

V

<L)

s.-

1)

V

:i:

c

2

t

<U

u

Oi)

c

0

~o

d

u

u

OX)

£

J

OX)

£

to

(U

c

X

<U

0/5

c

0

c

d

F

cn

3

H

d

CQ

For each food and within columns, means with a different letter are significantly different (P < 0.05), values are means ± one SE. The numbers in parentheses indicate sample sizes.

Nelson et al. • WHOOPING CRANE DIET

733

07:00 on day 1. Indicator in excreta at 18:00 on day 1 was not signifi-
cantly different from indicator in excreta at 14:00 on day 1 (1 -tailed t-
test, t — —.6165, P > 0.25). This rapid passage rate indicates that the
2-day adjustment period was adequate to eliminate non-study feed prior
to sample collection on day 3 for excreta nutrient analysis.

The metabolizable energy coefficient for Rangia clam was significantly
higher than that for acorns, wolfberry, and blue crab (Table 2). Protein
digestibility for blue crab was higher than that for acorns and wolfberry.
Lipid digestibility was lower for wolfberry than for Rangia clam and
acorns. There was no significant effect of feeding trial on metabolizable
energy and digestibility coefficients, and Whooping Cranes did not differ
in digestion efficiency.

There were significant differences in nutrients among foods (Table 3).
Gross energy ranged from 21.430 kJ g-' for wolfberry to 0.326 kJ g-' for
Rangia clam; crude protein ranged from 41.89% for blue crab to 1.44%
for Rangia clam. Total lipid ranged from 13.4% for wolfberry to 0.2%
for Rangia clam, and ash ranged from 96.9% for Rangia clam to 2.4%
for acorns from Dagger Point.

Wolfberry ranked highest of the five foods for metabolizable energy
and total lipid nutrient availability per kg food ingested (Table 4). Blue
crab ranked highest of the five foods for crude protein availability.

DISCUSSION

Protein digestibility coefficients obtained for Whooping Cranes for
acorns (48.9) and wolfberry (53.4) are lower than those for Rangia clam
(69.4) and blue crab (75.2); and are similar to the lower digestibilities for
plant protein when compared to animal protein (Karasov 1990). Total dry-
matter digestibility of Rangia clam was lower compared to the other
foods, reflecting the high percentage of ash (96.87%) in Rangia clam.
High metabolizable energy, protein, and lipid digestibility coefficients are
expected for shellfish where muscle comprises most of the readily di-
gested dry material (Karasov 1990); the MEC* for Rangia cuneata (75.0)
was similar (72—73) to the MEC* for intertidal polychaeta {Pseudonereis
variegata), black mussels {Choromytilus meridionalis), and limpet (Pa-
tella granularis) (Hockey 1984, Karasov 1990).

The MEC* for whole live oak acorn (43.2) was lower than that for
white oak acorn meat (Quercus alba) (66.0) consumed by Ruffed Grouse
(Bonasa umbellus) (Servello et al. 1987), and that for pin oak (Q. palus-
tris) acorn meat (55.3) and red oak (Q. rubra) acorn meat (57.3) con-
sumed by Northern Bobwhites (Colinus virginianus) (Robel et al. 1979).
Non-digestible cellulose and hemicellulose together comprise approxi-

Table 2

Apparent Metabolizable and Digestible Energy, Dry Matter, Crude Protein, and Total Lipid Digestibility Coefficients {%)•

734

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

U

5 =

W

<

s

<

m

n

n

w

£

£

in

in

in

in

in

d

m

2

q

q

SJ u

o

o

q

q

1

o

-2 ^

d

d

d

d

d

1

V

in

d

in

3

E

+1

+1

+1

+1

+1

ca

+1

+ 1

+1

+ !

+1

rj

z

C 3
£§ ^

NO

ON

q

(N

q

p

C

(N

q

d

r-

00

4

cn

in

d

X

NO

ni

NO

ON

d

r-

r-

NO

u

yj

u

N

Q

'35

Z

Q

4)

<

El

n

U

U

Q

U

a

C/5

_2 J2

d

On'

£

£

CD

ca

<

CQ

U

E

w

-J

u yj

d

io

d

in

y)

o

O SD

o

m

o

in

»n

in

in

in

W

z

u]

I

N T3

d

d

d

d

d

1

>5

q

q

q

q

o

u o

W DO

3 2

+1

+1

+1

+1

+1

c

V

£

P

Tj-

liM

in

d

■■5

Cu

c

w ^

NO

in

+1

+1

+ 1

+1

+1

ys

4>

J

<

On

q

00

NO

q

00

q

4

q

?

00

oq

in

o

4>

h-

O

H

d

in

r-

cn

d

d

.C

ON

in

NO

V

«

a.

s"

u

in

in

CQ

U

<

m

c

&

s

£

£

oi'

£

yi

v.

ON

d

d

d

d

C

4)

m

U

<

u

j

2 «
u ^

q

ON

NO

q

"o

c

X

<

d

d

d

d

'C.

b.

in

in

in

in

in

E

Urn

41 DO

+1

+1

+1

+1

+1

O

3

3 C

+i

u

Q

o

00

in

On

q

C

o

“d

u

'£

O

<U

(n

+1

ro

+ 1

d

+1

in

+1

d

+ 1

4)

x:

H

ui

d

z

u

1 1 .947

72.96

43.59

41.89

5.28

6.92

■ »

CN

m

CN

q

q

C/5

H

iA

o

cn

cn

d

00

d

4J

c

o

00

5

00

00

o

+1

m

ID

OS

Cu

UJ

Q

<

in

<

Q

<

U

y5

c

•3

P «

d CQ

s

s

£

00

£

fl

D

in

in

d

cc

ON

d

<

<

U

ca

<

CQ

<

E

4>

b.

.J

03

<

OS

u

o

d

04

d

o

d

m

d

d

o

d

1

3

m

in

in

in

in

««

yj

H

x"

fl

+1

+!

+1

+1

+1

+1

4>

C/2

^ =

o

00

NO

00

q

m

■q

q

S

<

m

<N

o

q

in

in

d

d

in

f3

>

q

NO

d

ON

d

d

•2

UJ

5

+1

+1

+I

+ 1

+1

<0

o

a:

D

cs

as

q

q

q

m

d

H

c/5

O

in

in

r--

sd

r--

00

ON

NO

in

r-

V

Q

U

<

a.

c

Ja C

s

£

£

£

£

O .3

m

d

in

d

•o

4)

<

<

<

ca

<

ca

<

4J

.if

■■5

H

Z

UJ

U

OS

H 0

flu

u

S w

0 Of)

o

d

+1

ON

On

d
+ 1

d

+1

r-

d

+1

in

d

+1

NO

d

+1

in

1

m

in

in

in

in

“

Cu

jB

On

q

(n

n

q

q

4)

■o

o

o

d

id

o

Q

00

d

ri

4

rn

d

4)

2

d

(N

'c

z

<

—

4>

+1

+1

+1

+ 1

+1

DO

c

3

CN

in

NO

ON

2

o

CQ

Q

S

Q

m

U

Gv

(N

d

d

d

E^i

in

ON

d

d

oi

O

r-

On

NO

00

2

O 0

00

u

Sos

4>

fl 2

o -c
c/2

4>

in

d

00

m

d

o

Cd

m

d

m

c

4>

b.

>-

o

OS

d

+1

+I

d

+1

d

+1

d

+I

d

+1

1

£

UJ

> ^

NO

00

fN

NO

in

00

z

r--

—

q

q

q

Tf

>Y

■■O

u:

o

(N

4

iri

in

00

CO

d

Ui

c/2

D

0

c

y)

os

0

D

00

u

<U

c

c

0

b>

3

O

c

'S

c/t

3

o

3

N

3

jD

2

a!

s

(U

3

(/)

<u

00

w

S

s

‘5

w

0

u

Q.

<U

T3

?

■g

!§-

3

!£

?

y5

C

«

4>

S

Component

>5

W)

u

o

c

c/5

c/5

o

V

u

2

c/5

'o

c/5

<

2

D.

U

T3

2

u

•g

‘a

B

c

o

JC

a

3

i2

_c

3

U

5

u

Q

u

U

(2

b.

o

Means within rows with a different letter are significantly different (P < 0.05), values are means ± one SE. The numbers in parentheses indicate sample sizes.

Nelson et al. • WHOOPING CRANE DIET

735

Table 4

Total Energy and Nutrient Availability (Dry-Weight Basis)

Grams nutrieni available per
kg food ingested

Food

Metabolizable
energy (kJ g-')

Crude protein
(g-kg“')

Total li
(gkg-

Wolfberry fruit

9.601

104

80

Live oak acorn

8.121

22

31

Blue crab

4.074

315

35

Stout razor clam

3.298

100

13

Rangia cuneata

0.245

10

2

mately 27% of the total dry matter of whole live oak acorns (Short 1976),
and the lower MEC* for whole live oak acorns was expected.

Tannins in foods inhibit protein digestibility (Marquardt and Ward
1979, Robbins et al. 1987, Johnson et al. 1993) and contribute to lower
energy utilization in avian species (Servello and Kirkpatrick 1989, John-
son et al. 1993). Total phenolics are lower for acorns of the white oak
group (includes live oak) than for acorns of the red oak group (Servello
and Kirkpatrick 1989). However, total phenolics (>5%) in live oak acorns
collected from the ANWR is high enough to affect protein digestibility
and energy utilization by Whooping Cranes, as indicated by the low
MEC* and protein digestibility for acorns compared to other foods.

Acorns, wolfberry, blue crab, stout razor clam, and Rangia clam were
markedly different in nutrient composition (dry-weight). We divided
Whooping Crane foods into two categories, (1) high energy-low protein
and (2) low energy-high protein. Wolfberry and acorns are high in caloric
content but lower in protein. Blue crab and stout razor clam are lower in
calories, but have moderate to high protein levels. Rangia clam is low in
energy and protein and is a suboptimal energy and nutrient resource for
Whooping Cranes. Approximately 30X more Rangia clam, would have
to be eaten than wolfberry and blue crabs to achieve comparable intake
of metabolizable energy and protein.

Seasonal availability, relative size of food items, food density, and nu-
tritional value must be considered when evaluating natural foods for
Whooping Cranes. Acorns constitute a high-energy localized food re-
source. However, availability of acorns may be short compared to blue
crab, wolfberry, and stout razor clam (Hunt 1987, Bishop et al. 1987,
Stehn 1994b). Blue crab and stout razor clam provide 3-5 X more di-
gestible crude protein than wolfberry and acorns, are larger per unit cap-
ture, but are less localized and may require greater time for search and

736

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

processing. Wolfberry is small, and because plants are scattered, energy
benefits for wolfberry may be offset by higher energy expenditure and
lower rates of energy capture by foraging Whooping Cranes (pers. obs.).

Although we evaluated food quality on a dry-weight basis, moisture
content among foods varied by as much as 438%. Variance in moisture
of this magnitude may be significant if Whooping Cranes are limited in
the amount of food they can consume and if feeding efficiency is affected
by food availability. Acorns collected from Dagger Point averaged 17.4%
moisture compared to 41.2% moisture for acorns collected along East
Shore Road. On a wet-weight basis, metabolizable energy of acorns from
Dagger Point was 6.699 kJ g“' compared to 4.699 kJ g“' for acorns from
East Shore Road. Wolfberry ranked highest of all foods for metabolizable
energy on a dry-weight basis. However, on a wet-weight basis, wolfberry
provided only 2.275 kJ g“' compared to 6.699 kJ g"' for burned acorns.

Under some conditions of food availability. Whooping Cranes on the
ANWR may have difficulty in meeting maintenance requirements and
building energy reserves needed for spring migration (Iverson and Vohs
1982, Krapu et al. 1985). Eoods may have been less available during fall
and winter of 1993-1994, because acorns and blue crabs were not com-
mon, although wolfberry was abundant into January (Stehn 1994b).
Whooping Cranes rarely fed on clams during winter 1993-1994 because
refuge clam populations were lower in recent years (Stehn 1994b).
Whooping Cranes were more dispersed in 1993-1994, and movements
were less predictable, suggesting possible shortage of blue crab (Stehn
1994b). Cranes also migrated late during spring 1994, and an unprece-
dented 15 cranes remained on the refuge and Matagorda Island until early
May (Stehn 1994b). Whooping Crane mortality was also higher than nor-
mal: three adults and five juveniles disappeared between 29 November
1993-16 February 1994 (Stehn 1994b). Counts in late 1994 included 131,
down five from the spring departure count of 136 (Stehn 1994a). In spring
1993, a record 46 pairs nested; however, only 28 pairs initiated nesting
in spring 1994 of a possible 45 known adult pairs (Stehn 1994a).

Multiple factors contribute to Whooping Crane mortality; predation,
collisions with power lines, disease, and habitat conditions on the breed-
ing ground (Brown et al. 1987, Carton et al. 1989, Kuyt et al. 1992, U.S.
Fish and Wildl. Serv. 1994). It is possible, however, as indicated by the
high over- winter mortality for 1993-1994, the late spring migration, lower
number of returning Whooping Cranes, and low number of pairs that
nested spring 1994, that food shortage on the ANWR was a contributing
factor to low reproduction and high Whooping Crane mortality from late
fall of 1993 to fall of 1994. Conditions of food shortage on the ANWR,
similar to those observed during winter 1993-1994, are of concern if the

Nelson et al. • WHOOPING CRANE DIET

737

observed higher mortality and low reproductive success are related to
lowered fitness caused by limited winter foods and the inability to assem-
ble required energy reserves for migration and breeding.

ACKNOWLEDGMENTS

We especially appreciate the help of the caretaker staff at the Endangered Species Re-
search Branch, PESC, that contributed to the success of working with captive Whooping
Cranes. We also thank the staff at the ANWR for providing valuable logistic and advisory
support. K. L. Risenhoover and D. M. Gatlin generously provided laboratory space and
equipment at the Texas A&M Univ., Fish Nutrition Laboratory and Habitat Laboratory in
the Dept, of Wildlife and Fisheries Sciences. The project was funded by the U.S. Fish and
Wildlife Service.

LITERATURE CITED

Allen, R. P. 1952. The Whooping Crane. Natl. Aud. Soc. Res. Rep. No. 3, New York,
New York.

. 1954. Additional data on the food of the Whooping Crane. Auk 71:198.

Bishop, M. A., H. E. Hunt, and R. D. Slack. 1987. Activity patterns of Whooping Cranes
wintering on the Aransas National Wildlife Refuge, Texas. Pp. 167-171 in Proc. 1985
Crane Workshop (J. C. Lewis, ed.). Platte River Whooping Crane Maintenance Trust,
Grand Island, Nebraska.

Black, M. M. and H. M. Schwartz. 1950. The estimation of chitin and chitin nitrogen in
crawfish waste and derived products. Analyst 75:185-189.

Blankinship, D. R. 1976. Studies of Whooping Cranes on the wintering grounds. Pp. 197-
206 in Proc. Int. Crane Workshop (J. C. Lewis, ed.). Oklahoma State Univ. Publ. and
Printing, Stillwater, Oklahoma.

Brown, W. M., R. C. Drewien, and E. G. Bizeau. 1987. Mortality of cranes and waterfowl
from powerline collisions in the San Luis Valley, Colorado. Pp. 128-136 in Proc. 1985
Crane Workshop (J. C. Lewis, ed.). Platte River Whooping Crane Maintenance Trust,
Grand Island, Nebraska.

Cho, C. Y., S. j. Slinger, and H. S. Bayley. 1982. Bioenergetics of salmonid fishes:
energy intake, expenditure and productivity. Comp. Biochem. Physiol. 73:25-41.

Folch, j., M. Lees, and G. H. Sloane Stanley. 1957. A simple method for the isolation
and purification of total lipides from animal tissues. J. Biol. Chem. 226:497-509.

Carton, E. O., R. C. Drewien, W. M. Brown, E. G. Bizeau, and P. H. Hayward. 1989.
Survival rates and population prospects of Whooping Cranes at Grays Lake NWR. Final
Rep. Fish and Wildlife Dept., Univ. Idaho. Prepared for U.S. Fish and Wildl. Serv.,
Albuquerque, New Mexico.

Hagerman, a. E. 1988. Extraction of tannin from fresh and preserved leaves. J. Chem.
Ecol. 14:453-461.

Helrich, K. (ed.). 1990. Official methods of analysis of the Association of Official Ana-
lytical Chemists. 15th ed. Assoc. Off. Anal. Chem., Washington, D.C.

Hockey, P. A. R. 1984. Growth energetics of the African Black Oystercatcher Haenicitopus
moquini. Ardea 72:1 11-117.

Hunt, H. E. 1987. The effects of burning and grazing on habitat u.se by Whooping Cranes
and Sandhill Cranes on the Aransas National Wildlife Refuge, Texas. Ph.D. diss., Texas
A&M Univ., College Station, Texas.

AND R. D. Slack. 1987. Winter foods of the Whooping Crane based on stomach

738

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

content analyses. Pp. 217—218 in Proc. 1985 Crane Workshop (J. C. Lewis, ed.). Platte
River Whooping Crane Maintenance Trust, Grand Island, Nebraska.

AND . 1989. Winter diets of Whooping and Sandhill Cranes in south Texas.

J. Wildl. Manage. 53:1150-1154.

Iverson, G. C. and P. A. Vohs, Jr. 1982. Estimating lipid content of Sandhill Cranes from
anatomical measurements. J. Wildl. Manage. 46:478—483.

Johnson, W. C., L. Thomas, and C. S. Adkjsson. 1993. Dietary circumvention of acorn
tannins by Blue Jays: implications for oak demography. Oecologia 94:159-164.

Karasov, W. H. 1990. Digestion in birds: chemical and physiological determinants and
ecological implications. Stud. Avian Biol. 13:391-415.

Krapu, G. L., G. C. Iverson, K. J. Reinecke, and C. M. Boise. 1985. Pat deposition and
usage by Arctic-nesting Sandhill Cranes during spring. Auk 102:362-368.

Kuyt, E., S. j. Barry, and B. W. Johns. 1992. Below average Whooping Crane production
in Wood Buffalo National Park during drought years 1990 and 1991. Blue Jay 50:225-
229.

Labuda, S. E., Jr. and K. O. Butts. 1979. Habitat use by wintering Whooping Cranes on
the Aransas National Wildlife Refuge. Pp. 151-157 in Proc. 1978 Crane Workshop (J.
C. Lewis, ed.). Colorado State Univ. Printing Serv., Fort Collins, Colorado.

Lide, D. R. (ed.). 1994. CRC handbook of chemistry and physics. 75th ed. CRC Press,
Boca Raton, Florida.

Lloyd, L. E., B. E. McDonald, and E. W. Crampton. 1978. Fundamentals of nutrition.
Second ed. W. H. Freeman and Co., San Francisco, California.

Marquardt, R. R. 1983. A simple spectrophotometric method for the direct determination
of uric acid in avian excreta. Poult. Sci. 43:2106—2108.

AND A. T. Ward. 1979. Chick performance as affected by autoclave treatment of

tannin-containing and tannin-free cultivars of fababeans. Can. J. Anim. Sci. 5:781-789.

Muziar, a. j., S. j. Slinger, and J. H. Burton. 1977. Metabolizable energy content of
freshwater plants in chickens and ducks. Poult. Sci. 56:1893-1899.

POMERANZ, Y. AND C. E. Meloan. 1987. Food analysis: theory and practice. Second ed.
Van Nostrand Reinhold, New York, New York.

Robbins, C. T, T. A. Hanley, A. E. Hagerman, O. Hjeljord, D. L. Baker, C. C. Schwartz,
AND W. W. Mautz. 1987. Role of tannins in defending plants against ruminants: re-
duction in protein availability. Ecology 68:98-107.

Robel, R. j., a. R. Bisset, T. M. Clement, Jr., A. D. Dayton, and K. L. Morgan. 1979.
Metabolizable energy of important foods of Bobwhites in Kansas. J. Wildl. Manage.
43:982-987.

Rotter, B. A., A. A. Frohlich, R. G. Rotter, and R. R. Marquardt. 1989. Research
note: estimation of apparent protein digestibility using uric acid-corrected nitrogen val-
ues in poultry excreta. Poult. Sci. 68:327—329.

Servello, E a. and R. L. Kirkpatrick. 1989. Nutritional value of acorns for Ruffed
Grouse. J. Wildl. Manage. 53:26-29.

, , and K. E. Webb. 1987. Predicting metabolizable energy in the diet of

Ruffed Grouse. J. Wildl. Manage. 5 1 :560-567.

Shields, R. H. and E. L. Benham. 1969. Farm crops as food supplements for Whooping
Cranes. J. Wildl. Manage. 3:81 1-817.

Short, H. L. 1976. Composition and squirrel use of acorns of black and white oak groups.
J. Wildl. Manage. 40:479-483.

Singleton, V. L. and J. A. Rossi. Jr. 1965. Colorimetry of total phenols with phosphom-
olybdic-phosphotungstic acid reagents. Am. J. Enol. Viticult. 16:144-158.

Stehn, T. 1994a. Aransas Refuge. Unison Call 6(2):4-5.

Nelson et al. • WHOOPING CRANE DIET

739

. 1994b. Whooping Cranes during the 1993-1994 winter. Aransas National Wildl
Refuge Rep., Austwell, Texas.

Stevenson, J. O. and R. E. Griffith. 1946. Winter life of the Whooping Crane. Condor
48:160-178.

Uhler, E M. and L. N. Locke. 1970. A note on the stomach contents of two Whooping
Cranes. Condor 72:246.

U.S. Fish and Wildl. Serv. 1994. Whooping Crane recover plan. U.S. Fish and Wildl.
Serv., Albuquerque, New Mexico.

Wilson, R. R and W. E. Poe. 1985. Apparent digestible protein and energy coefficients of
common feed ingredients for Channel Catfish. Prog. Fish-Cult. 47:154_158.

Wilso>2 Bull., 108(4), 1996, pp. 740-747

TERRITORIES AND CACHING-RELATED BEHAVIOR
OF RED-HEADED WOODPECKERS WINTERING IN A

BEECH GROVE

Paul F. Doherty, Jr., Thomas C. Grubb, Jr., and
C. L. Bronson

Abstract. — We describe caching and related behavior of Red-headed Woodpeckers (Me-
lanerpes erythrocephalus) wintering in a beech grove during a mast year and relate territorial
behavior and territory size to territory-specific mast abundance. We found no difference
between territories of adults and juveniles in either territory size or abundance of mast.
Rates of caching and social interaction decreased over the course of the winter. Received
Oct. 1995, accepted 8 Mar. 1996.

Red-headed Woodpeckers (Melanerpes erythrocephalus) are larder
hoarders during the fall and winter (Bent 1939, Kilham 1983). In autumn,
these birds aggregate and establish singular winter territories at sites of
high mast production (Smith and Scarlett 1987). Each territorial bird se-
questers mast in one or a few larder trees which are then defended both
inter- and intraspecifically. Red-headed Woodpeckers are known to store
mainly acorns and beechnuts and an occasional insect (Hay 1887, Agers-
borg in Beal 1911, Kilham 1983). While acorn storing has been described
for Red-headed Woodpeckers in Maryland, Louisiana, and Florida by
Kilham (1983), MacRoberts (1975), and Moskovits (1978), respectively,
no quantitative data exist on the storage of beech mast by this species.
The present study describes caching and related behavioral patterns of
Red-headed Woodpeckers wintering in a beech grove during a mast year
and relates territorial behavior and territory size to territory specific mast
abundance. We also searched for differences in behavior and territories
between juvenile and adult woodpeckers.

METHODS

From 1 Nov. 1992 to 19 Mar. 1993 we ob.served 14 Red-headed Woodpeckers (6 juveniles
and 8 adults) for a total of 160 h in a woodlot located in Morrow County, Ohio. The woodlot
was dominated by mature American beeches (Fagus grandifolia), red maples (Acer rubrum),
and sugar maples (A. saccharum). Other tree species at this site were white oak (Quercus
alba), red oak (Q. rubra), ash (Fraxinus sp.), ironwood (Carpinus sp.), and hickory (Carya
sp.). Beech was the only tree species experiencing a “mast year.” We used focal bird
techniques to gather behavior data. We watched each focal bird for 10-min periods. For the
first 240 sec of this period, we noted which category of activity the bird was engaged in
every 10 sec (N = 25). The five mutually exclusive categories were (1) lookout-the bird
was perched and alert, (2) flight — the bird was in flight, (3) bipedal locomotion — the bird

Behavioral Ecology Group, Dept, of Zoology, The Ohio State Univ., Columbus, Ohio 43210-1293.

740

Doherty et al. • RED-HEADED WOODPECKER CACHING

741

was moving over a tree, (4) peck-the bird was actively pecking, and (5) preen-the bird
was preening. In addition to these five categories, we also recorded whether the bird was
actively caching or engaged in any type of visual and/or auditory signaling or locomotory
interaction. If the bird was caching, we noted whether the bird was on a trunk, limb, branch,
or a twig, each of the last three substrate categories being an offshoot of the previous one.’
Visual displays consisted of an agonistic pose (Kilham 1983) or head bobbing. The display
vocalization we witnessed has been described as “quirr” by Kilham (1958b). A cha.se was
defined as one bird flying after another.

Over 10-mm periods, we recorded the number of caches a focal bird made and every
(previously-numbered) tree it visited. Temperature (°C), wind velocity (m/sec, Velometer Jr.,
Alnor Instrument Company, Niles, Illinois), solar radiation (mW/cm^; Solar Meter, Dodge
Products, Houston, Texas), and any precipitation were recorded for each observation period.

Juveniles could be distinguished from adults by plumage. Although the birds were not
individually marked, we were able to identify individuals by idiosyncratic plumage patterns
and behavior, as did Kilham (1958a).

We estimated the size of the beech crop by sampling fallen mast as follows: over a two-
day period in early December, we gathered mast and leaf litter from below each beech tree
in every bird’s territory. For each tree, we gathered all leaf litter and mast within a 1660-
cm- circular area located half way between the trunk and the outer edge of the canopy in
each of the four cardinal directions. The samples were bagged and later sorted to determine
number of beechnuts. Infertile nuts were excluded from the count.

Using a compass and a range finder, we constructed a map of all the marked trees in the
study area. Each territory was delineated as the minimum-area polygon that included all the
trees visited by the same bird. We drew each polygon to connect the positions of trunks,
not canopy boundaries, a procedure that underestimated territory sizes. The polygons were
then digitized and territory areas calculated.

The individual bird was the primary sampling unit, and all observations of the same
individual were averaged before being analyzed. The data met the requirements for para-
metric tests.

To increase degrees of freedom when general linear models were employed, the individual
bird was included as a factor. General linear models were performed in a stepwise fashion
with the individual bird, time of day, day of winter (1 Nov. = day 1), average temperature,
average wind speed, and average solar radiation included as independent variables. Only
the independent variables retained in the model are reported. All statistical calculations were
performed using Minitab (Anonymous 1991) or Systat (Wilkinson 1992) software.

RESULTS

Probable mortality between November and March was low (7%). The
only bird disappearing during the study was a juvenile, and its tenitory
was not usurped or occupied by any other Red-headed Woodpecker.

We could find no differences in the wind speed, temperature, or solar
radiation associated with adult and Juvenile territories (Table 1). There
was no significant difference in size between territories of juveniles and
adults (Table 1; Fig. 1). Neither the number of beechnuts per territory
nor beechnut density was significantly different between the two wood-
pecker age classes (Table 1). We could find no difference between adult
and juvenile birds in either the number of beech trees or the total number
of trees defended (Table 1 ). Power analyses of these tests demonstrated

742

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 1

Mean ± SD for Variables Associated with All Red-headed Woodpecker, Juvenile
Red-headed Woodpecker and Adult Red-headed Woodpecker Territories^

Variable

Total
(N = 14)

Juvenile
(N = 6)

Adult
(N = 8)

p

Sample

sizes

from

power

anal-

ysis

Wind speed (m/sec)

0.9

-F

0.2

0.9

± 0.1

0.9

-F

0.3

0.75

23

Temperature (°C)

1.4

-F

2.2

0.5

± 2.5

2.0

-F

2.1

0.27

938

Solar radiation (mW/cm^)

4.1

-F

3.6

3.7

± 1.6

4.3

-F

1.2

0.48

316

Territory size (ha)

0.04

-F

0.03

0.03

± 0.03

0.05

-F

0.03

0.28

236

Number of beech trees

4.8

-F

2.7

4.2

± 2.4

5.3

-F

2.9

0.47

127

Total trees

8.9

-F

3.8

8.3

± 2.3

9.4

-F

4.8

0.49

75

Beechnuts/territory*’

24.5

-F

13.4

19.3

± 15.0

28.3

-F

11.6

0.23

127

Beechnuts/m^

62.3

-F

26.4

62.6

± 31.4

62.0

-F

24.8

0.97

75

Interspecific interactions/min

0.23

-F

0.13

0.21

± 0.09

0.2

-F

0.2

0.73

137

“ P-values are significance levels of f-tesls between juveniles and adults. The listed sample sizes from power analysis are
those that would be required to detect a difference between Juvenile and adult territories at a = 0.05 and P = 0.05 given
a 25% detectibiiity threshold and the current coefficient of variation.

’’ In thousands.

that detecting a significant difference with a = 0.05 and P = 0.05 would
require sample sizes of 23 to 316 (Table 1). Thus, our non-significant
differences were probably not due to small sample sizes alone. There was
some tendency for juvenile territories to be on the edge of the woodlot
(Fisher’s exact test, P — 0.09).

Although Red-headed Woodpeckers sometimes collect mast from the
ground (TCG, pers. obs.), all of the birds we watched cached beechnuts
gathered only from trees. Often, newly-collected beechnuts were broken
on an “anvil”, any portion not eaten immediately was stored. Although
harvested corn fields were nearby, we never saw corn being gleaned and
cached. The birds usually cached beechnuts on a trunk (30.9%), limb
(43.9%), or branch (24.3%) and seldom on a twig (1.2%). There was no
difference between caching rates of juveniles and adults, but over the
months, the caching rate did decline for all birds (Fig. 2). A general linear
model with the individual bird as a factor to account for the variation
among individuals, and with day of winter as a covariate, showed that
day of winter had a significant negative association with caching rate (t
= -3.48, df = 135, P = 0.001).

There was no statistical difference between activity budgets of adults
and juveniles. The birds spent most of their time looking about (69.9 ±
9.6%), with the rest of their time being divided among pecking (15.6 ±
8.7%), flying (8.6 ± 2.9%), bipedal locomotion (5.8 ± 2.4%), and preen-

Doherty er al. • RED-HEADED WOODPECKER CACHING

743

corn

field

field

corn

N

10m

t

Fig. 1. Study area and Red-headed Woodpecker territories. A and J denote territories of
adults and juveniles, respectively.

ing (0.2 ± 0.1%). A general linear model, with time spent looking about
as the dependent variable, and day of winter and individual bird as factors
showed that the percentage of time looking about increased as the winter
progressed {t = 4.13, df = 135, f* < 0.000).

The territories we observed were very small, about 0.05 ha (Table 1).
Across the mosaic of territories, we witnessed 50 agonistic interactions
involving Red-headed Woodpeckers. Nineteen of these were between
Red-headed Woodpeckers, 13 were with Blue Jays (Cyanocitta cristata),

1 1 consisted of a Red-headed Woodpecker chasing mixed-species forag-
ing flocks, six occurred with individual Carolina Chickadees {Parus car-
olinensis). Tufted Titmice (P. bicolor), white-breasted Nuthatches {Sitta
carolinensis), or Downy Woodpeckers {Picoides pubescens), and once we
witnessed a Red-headed Woodpecker chase off a fox squirrel (Sciurus
niger). We could find no difference between juveniles and adults in the
rate of interspecific interactions (Table 1). Although Red-bellied Wood-
peckers (Melanerpes carolinus) resided in the woods just north of the
study site, we rarely saw them and never witnessed them interacting with

744

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Month

Lig. 2. Mean ± SD number of beechnuts cached per minute by Red-headed Wood-
peckers wintering in a beech grove. N = 14.

M. erythrocephalus. Even though they are larger. Red-bellied Woodpeck-
ers are socially subordinate to Red-headed Woodpeckers (e.g., Williams
and Batzli 1979).

Behavior during agonistic interactions consisted of only a visual display
(1.7%), vocalization with or without a visual display (72.1%), or chasing
with or without a vocalization (26.2%). The frequency of interactions
decreased over the course of the winter. A general linear model performed
with percentage of 10-min recording periods lacking any social interaction
as the dependent variable, individual bird as a factor, and day of winter
as a covariate showed that the percentage of time devoid of social inter-
action increased with time {t = 3.10, df = 135, P = 0.002).

DISCUSSION

The average territory size of 0.04 ha was considerably smaller than
that reported by Kilham (1958b, 0. 1-0.2 ha), MacRoberts (1975, 0.8-1. 2)
or Moscovits (1978, 0.97 ha), although Moskovits did have a few terri-
tories as small as 0.04 ha. MacRoberts (1975) hypothesized that territory
size is highly compressible and negatively correlated with mast produc-
tion. This would seem to be so, but, unfortunately, other studies (including
MacRoberts 1975) have no data on mast production (but see Smith and

Doherty et al. • RED-HEADED WOODPECKER CACHING 745

Scarlett 1987). T.C.G. does have unpublished records pertaining to this
point from two consecutive winters (1983-1985) when an index of beech
mast was taken in the same woodlot with similar methods. During 1983,
beech mast density was estimated at 231.1 nuts/m^ and Red-headed
Woodpecker density at 2.4 ± 0.2 birds/ha. The following year when beech
mast was estimated at 57 ±7.5 nuts/m^, the woodpecker density was 3.6
± 0.3 birds/ha. Such an increase in Red-headed Woodpecker density dur-
ing a lower mast year contradicts both MacRoberts’ (1975) assertion and
Smith and Scarlett s (1987) data. Other potentially causal factors, such as
differing mast levels in neighboring woods and/or differing woodpecker
reproduction in previous summers could account for this disparity.

Although Kilham (1958a) thought that adults held smaller, more easily
defended and more desirable” areas, we could not support this assertion,
nor could Moskovits (1978) find any difference between adult and juve-
nile territory sizes. There may have been some tendency for juvenile
territories to be on the edge of the woodlot, a trend apparent in Kilham’s
(1958a) study area in Maryland where all juvenile territories bordered an
old field (Kilham 1958a). Kilham (1958a) thought there were more dead
trees for roosts and mast storage in the territories of adults. Such a dis-
parity might account for the distribution of adult and juvenile territories
in his study area, because the east side of his study area (where more of
the adults were located) had many more dead locust trees. In our study
area, there were few dead trees; the Red-headed Woodpeckers roosted in
holes in dead limbs of living trees. We could find no difference in either
the total number of trees or number of beech trees defended by juvenile
and adult birds.

If, indeed, the territories of juveniles tend to be concentrated on the
edge of woodlots, another possible explanation is that the winter micro-
climate on the edge of a wood is more severe than in the interior, with
juveniles forced to take the lesser quality habitat. However, we could find
no difference in the wind speed, temperature, or solar radiation associated
with juvenile or adult territories that might support this idea.

Another way in which edge (juvenile) territories could be inferior re-
lates to the number of intruders. Often, intruding birds were chased from
territory to territory until they left the study area. Each successive wood-
pecker would be alerted by an intruder’s interactions with other territory
holders. The only Red-headed Woodpecker that would have no warning
would be the initial bird, and so those occupying edge territories may
have suffered more from the effects of intrusions. However, in compen-
sation, owners of edge territories would benefit from fewer intraspecific
interactions than owners of central territories. Since 19 of the 50 inter-

746

THE WILSON BULLETIN • Vol. 108, No. 4. December 1996

actions we observed (38%) were intraspecific, such an advantage could
accrue to individuals on the edge of the territorial mosaic.

Kilham’s (1958a) and MacRoberts’ (1975) reports on Red-headed
Woodpeckers in Maryland and Louisiana suggest that the birds closely
monitor the acorn crop when selecting winter habitat. Willson’s (1970)
and Roller’s (1972) observations in Illinois found them less dependent on
mast crops and acting more as generalists by also feeding on insects. The
local distribution and activity budgets of the birds we studied suggested
that they were almost totally dependent on the stored beech mast crop for
winter food.

MacRoberts (1975) reported that Red-headed Woodpeckers retrieved
acorns from the ground. Only twice did we see a Red-headed Woodpecker
on the ground and never in conjunction with mast collection. Even at the
end of our study period, there were still beechnuts in the canopy, so there
appeared to be no need for the woodpeckers to descend to the ground for
food.

In our study area, all of the nuts an individual bird cached were ap-
parently taken from within its territory. MacRoberts (1975) had similar
findings, but in Kilham’s (1958a) and Moscovits’ (1978) studies, nuts
were brought from a distance, sometimes from communal gathering areas
as far as 100 m away from an individual’s territory. Perhaps, in their
study areas the distributions of cache sites and cachable food did not
overlap to the extent evident at our study site.

The Red-headed Woodpeckers spent most of the late fall and early
winter caching nuts. Time spent looking about and the percentage of time
with no interactions increased as the winter progressed. Also, the rate of
caching decreased over the winter as the birds became more sedentary
and focused their behavior at a favorite site, only leaving to intercept
intruders. Thus, the birds appeared to spend less time caching and more
time guarding their caches once their territories had been established and
stocked with provisions. Even within their extremely small territories, the
birds spent most of their time in a “core” area centered on their storage
area, as reported by MacRoberts (1975).

The Red-headed Woodpecker is a very aggressive bird in the winter
when it is defending its territories. We found it to dominate every inter-
action, as did Moskovits (1978). In MacRoberts’ study (1975), the Red-
headed Woodpeckers rarely trespassed, and territory boundaries were hard
to delineate. Such was not the case in this study. Territory lines were
readily defined. As soon as a bird ventured into another bird’s territory,
an interaction ensued.

Kilham (1958b) found that Red-headed Woodpeckers defended their
entire winter home ranges both intra- and interspecifically, while Mac-

Doherty et al. • RED HEADED WOODPECKER CACHING

747

Roberts (1975) determined that the species defended only those areas
immediately around cache sites. As in Kilham’s study (1958a), the terri-
tories we observed were very small and more easily defended than those
of MacRoberts’ (1975) birds.

Kilham (1958a), Reller (1972), MacRoberts (1975) and Moskovits
(1978) all commented on the prevalence of interactions with the Red-
bellied Woodpecker. We did not find Red-bellied Woodpeckers to be ma-
jor competitors.

Most of the interactions consisted of vocalizations. This was probably
the most efficient means of communicating, since chasing was probably
more expensive energetically and a visual display would probably have
had a lower probability of being received. We did find the number of
interactions decreased as the winter passed, as did Moskovits (1978).

ACKNOWLEDGMENTS

We thank Randy Dettmers, Tim Kloth, Volodya Pravosudov, and Rodney Tienerand for
help with the field work. Walter Riggs granted us permission to work on his property. This
manuscript was improved by the comments of D. Ingold and D. E Tombach.

LITERATURE CITED

Anonymous. 1991. Minitab reference manual. Minitab Inc., State College, Pennsylvania.
Bent, A. C. 1939. Life histories of North American woodpeckers. U.S. Natl. Mus. Bull
174.

Beal, E E. L. 1911. Foods of woodpeckers of the United States. Biol. Surv., U.S. Dept.
Agr. Bull. No. 37.

Hay, O. P. 1887. The Red-headed Woodpecker a hoarder. Auk 4:193-196.

Kilham, L. 1958a. Sealed-in winter stores of Red-headed Woodpeckers. Wilson Bull. 70:
107-1 13.

. 1958b. Territorial behavior of wintering Red-headed Woodpeckers. Wilson Bull.

70:347-358.

. 1983. Life history studies of woodpeckers of eastern North America. Nuttall Or-
nithological Club, Cambridge, Massachusetts.

MacRoberts, M. H. 1975. Food storage and winter territory in Red-headed Woodpeckers
in northwestern Louisiana. Auk 92:382-385.

Moskovits, D. 1978. Winter territorial and foraging behavior of Red-headed Woodpeckers
in Florida. Wilson Bull. 90:521-535.

Reller. A. W. 1972. Aspects of behavioral ecology of Red-headed and Red-bellied Wood-
peckers. Am. Midi. Nat. 88:270-290.

Smith, K. G. and T. Scarlett. 1987. Mast production and winter populations of Red-
headed Woodpeckers and Blue Jays. J. Wildl. Manage. 51:459-467.

Wilkinson, L., M. Hill, J. P. Welna, and G. K. Birkenbenel. 1992. Systat. Systat, Ev-
anston, Illinois.

Williams, J. B. and G. O. Batzli. 1979. Competition among bark-foraging birds in central
Illinois: experimental evidence. Condor 82:122—132.

Willson, M. F. 1970. Foraging behavior of some winter birds of deciduous woods. Condor
72:169-174.

Wilson Bull, 108(4), 1996, pp. 748-759

SEASONAL ABUNDANCE OF MIGRANT BIRDS
AND FOOD RESOURCES IN PANAMANIAN
MANGROVE FORESTS

Gaetan Lefebvre and Brigitte Poulin

Abstract. — We studied temporal variation in abundance of Nearctic-Neotropical mi-
grants, particularly the Northern Waterthrush (Seiurus noveboracensis), Prothonotary War-
bler (Protonotaria citrea), and American Redstart {Setophaga ruticilla) in two black man-
grove sites of central Panama from September 1993 through May 1995. The two sites, on
the Caribbean and the Pacific coasts, differ importantly in annual rainfall, tide amplitude,
and seasonal invertebrate abundance. Most migrant species varied temporally in abundance
with the opposite pattern at each site, suggesting mid-winter movements correlated with
abundance of food resources. Because of their wide geographic distribution and their par-
ticular response to hydrographic factors, mangroves are likely to have a temporally com-
plementary role in sustaining migrant populations throughout the Neotropics. However, vari-
ations in migrant numbers reported in other Neotropical habitats could also reflect large-
scale movements by migrants. Occurrence of mid-winter (facultative) migration has been
documented mostly for the Palearctic- African migratory system, and needs to be investigated
in the Nearctic-Neotropical realm for proper conservation of migratory species. Received
17 Nov. 1995, accepted 16 April 1996.

During the non-breeding period, Nearctic-Neotropical migrants gener-
ally use a succession of habitats. Movements between habitats are known
to occur during the autumn and spring migrations, but most migrant pop-
ulations are presumed to be spatially stable between these two periods
(Rappole et al. 1993). Nonetheless, large variations in migrant abundance
suggest that these movements occur throughout the non-breeding period
in several Neotropical habitats (Galindo et al. 1963; Morton 1980; Emlen
1980; Hilty 1980; Johnson 1980; Orejuela et al. 1980; Greenberg 1984,
1992a; Blake and Loiselle 1992; Lefebvre et al. 1992; Sherry and Holmes
1996; Wunderle 1995). Although movements by migratory landbirds dur-
ing the nonbreeding period have never been investigated in detail, a few
lines of evidence suggest that they could be related to food abundance
(e.g., Terrill and Ohmart 1984, Greenberg 1992a, Lefebvre et al. 1994b).

Mangrove forests are known to support high numbers of insectivorous
migrants in several Neotropical regions (Russell 1980, Lynch 1989, Wun-
derle and Waide 1993, Lefebvre et al. 1994b). In particular, the Northern
Waterthrush (scientific names in Table 1), the Prothonotary Warbler, and
the American Redstart have been observed in mangroves of Mexico (Hut-
to 1980, Lynch 1989), Panama (Morton 1980), Colombia (Russell 1980),

Smithsonian Tropical Research Institute, P.O. Box 2072, Ancon. Republic of Panama. (Present address:
Station Biologique de la Tour du Valet, Le Sambuc, 13200 Arts. France.)

748

Lefehvre and Poulin • MIGRANTS IN MANGROVES

749

Table 1

Number of Observations of Each Migrant Species at Two Mangrove Sites

Species

Galeta

(Atlantic)

Juan Diaz
(Pacific)

Prothonotary Warbler (Protonotaria citrea)

200

222

Northern Waterthrush (Seiurus noveboracensis)

190

181

American Redstart (Setophaga ruticilla)

8

20

Black-and-white Warbler (Mnioltila varia)

13

7

Red-eyed Vireo {Vireo olivaceus)

17

1

Yellow Warbler (Dendroica petechia)

4

1 1

Chestnut-sided Warbler (D. pensylvanica)

9

0

Summer Tanager (Piranga rubra)

8

0

Tennessee Warbler (Vermivora peregrina)

5

2

Eastern Wood-pewee (Contopus virens)

4

2

Acadian Flycatcher (Empidonax virescens)

3

0

Canada Warbler (Wilsonia canadensis)

0

3

Blackpoll Warbler (D. striata)

1

1

Northern Oriole (Icterus galbula)

2

0

Golden- winged Warbler (Vermivora chrysoptera)

0

1

Blackburnian Warbler (D. fusca)

1

0

Bay-breasted Warbler (D. castanea)

0

1

Cerulean Warbler (D. cerulea)

0

1

Worm-eating Warbler (Helmitheros vermivorus)

1

0

Venezuela (Lefebvre et al. 1994b), and the West Indies (Arendt 1992,
Wunderle and Waide 1993). Variation in species abundance and the birds’
short length of stay (Lefebvre et al. 1992, 1994a) suggest that at least
some individuals of these species undertake large movements during the
nonbreeding period.

Mangroves are widely distributed in the Neotropics and often support
more migrants than nearby terrestrial habitats (Hutto 1980, Lynch 1989,
Wunderle and Waide 1993, Poulin et al. 1994). One possible explanation
is that flooded habitats such as mangroves are buffered from the effect
of rainfall seasonality and the resulting marked fluctuations in arthropod
availability (Morton 1980, Lefebvre et al. 1994b, Sherry and Holmes
1996). Food available in mangroves is, however, likely to vary geograph-
ically and with latitude through the differing salinity and inundation pat-
terns at each locality (Duke 1990, Lefebvre and Poulin, in press). There-
fore, migrant species that use mangroves extensively as a wintering hab-
itat could move between mangrove forests to take advantage of these
temporal variations in food resource.

In this paper we compare the seasonal abundance of Nearctic-Neotrop-
ical migrants from two Panamanian black mangrove forests experiencing

750

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

different hydrographic conditions. In particular, we attempt to answer the
following questions; Do migrant populations show large variation in
abundance during the nonbreeding season? Are the two mangrove sites
used differentially through time by migrants? Is variation in migrant abun-
dance related to the abundance of food resources at each site?

STUDY AREAS AND METHODS

Study areas. — We carried out this study in two coastal mangrove forests 65 km apart in
central Panama from September 1993 to May 1995. These were Juan Diaz (9°00'N,
79°04'W) on the Pacific coast and Galeta (9°20'N, 79°09'W) on the Caribbean coast. Each
study site was 4.5 ha of black mangrove (Avicennia germinans L.) with red (Rhizophora
mangle L.) and white {Laguncularia racemosa L.) mangroves comprising around 30%.
Annual rainfall in Galeta (3244 mm) averages nearly twice that in Juan Diaz (1786 mm).
The two sites also differ importantly regarding mean tidal range, with a value of 24 cm for
Galeta and 395 cm for Juan Diaz. Both sites, however, experience a dry season of similar
intensity from January to April which coincides with the period of lowest high tides (Le-
febvre and Poulin, in press).

Bird abundance. — Migrant abundance was evaluated during observation sessions twice
monthly from September 1993 through May 1994. To eliminate any short-term temporal
bias, the two sites were sampled on two consecutive days. At each site, observations were
carried out at 1 8 spot counts located 50 m apart along two transects. Starting at sunrise, we
surveyed each spot for 10 min, and noted every new bird seen within 25 m.

Invertebrate abundance. — Abundance of invertebrates, the main food resource available
to passerines in mangroves (Lefebvre et al. 1994b), was estimated during each observation
session. The first 2 m of vegetation were swept with a standard insect net in late morning
during 15 min. Sweep-netted invertebrates were preserved in 70% ethanol and sorted using
a dissecting scope. Each invertebrate was identified to order or family and categorized as
small (<0.5 mm) or large (>0.5 mm). Sub-adult forms of insects were assigned either to
eggs or larvae without taxonomic distinction.

Bird diet. — We determined diets of migrants through regurgitation sessions during two
non-breeding seasons. We sampled sites alternately twice monthly so that regurgitations
were collected during each calendar month at each site. Twelve mist-nests (10 m X 3 m,
32-mm mesh) were operated from dawn until early afternoon. Some 452 migrants from 14
species were captured and administered tartar emetic following the method of Poulin and
Lefebvre (1995). Diet samples were preserved and sorted using the same methodology as
for sweep-net samples.

Food abundance. — Using data from sweep-net and emetic samples, a monthly index of
food abundance was calculated as follows:

-Vj

Index of food abundance = 2^ p~

",

where p, is the proportion of a specific arthropod group (taxa/size) in migrants’ diet, .x, is
the number of arthropods from that group sampled with sweep net during a specific date,
and //, is the total number of arthropods from that group sampled with sweep net. Accord-
ingly, this index reaches a maximal value when several arthropod groups extensively taken
by the birds simultaneously show a high abundance.

Statistical analyses. — We used Spearman’s correlation rank coefficient to compare tem-
poral abundance in migrants between the two sites. We compared abundance of each ar-

Ufebvre and Poulin • MIGRANTS IN MANGROVES

751

thropod group (taxa/size) between the two sites with contingency tables using G-tests All
18 sweep-net samples collected at each site were combined since they were independent
samples without replicate. To determine whether changes in bird abundance between sites
were correlated with changes in food abundance between sites, we subtracted bird and food
abundances at one site (Juan Diaz) from the other (Galeta) for each sampling date throughout
the nonbreeding period and calculated Spearman correlations between the two data sets
Most data were pooled by month in presentation of figures, but all analyses were run on
data collected twice each month.

RESULTS

Bird abundance. — Overall, 466 and 453 migrants from 15 and 13 spe-
cies were observed at Galeta and Juan Diaz, respectively (Table 1). Nine
species representing over 95% of all observations were sampled at the
two sites. Both migrant communities were dominated by the Prothonotary
Warbler and the Northern Waterthrush (Table 1). Most species showed
large variations in abundance with a different temporal trend at each site
(Fig. 1). Abundances of the Northern Waterthrush at Galeta and Juan Diaz
were inversely correlated from October through March {r, = -0.713, df
= 10, P < 0.01), with higher numbers on the Pacific coast in early winter
and on the Caribbean coast in late winter. The Prothonotary Warbler was
abundant at both sites from September through January only. In Galeta,
its abundance was high in September and gradually decreased during
winter, whereas in Juan Diaz it was high and stable from October through
January. Abundances of the American Redstart at Galeta and Juan Diaz
were negatively correlated from November through March (r, = -0.792,
df = 9, P < 0.05), showing an inverse pattern to the one of the Northern
Waterthrush. The remaining 16 migrant species were sampled only oc-
casionally and were clumped in “other” migrant species (Table 1). Their
abundances at Juan Diaz and Galeta were inversely correlated from Sep-
tember through May {r, = -0.566, df = 13, P < 0.05). These migrants
were sampled mostly during autumn migration at Juan Diaz and at the
end of the wintering period at Galeta (Fig. 1 ).

Estimation of food resources. — Over 400 diet samples were collected
(218 from Juan Diaz and 196 from Galeta), from which 3689 items were
identified and assigned to one of 34 invertebrate groups (taxa/size) (Table
2). Over 98% of the invertebrates taken by migrants were from taxa also
sampled by sweep net. Some 78% of these invertebrates were present in
at least half of the 36 sweep-net samples. In Galeta, migrants fed mostly
on small beetles, small ants, large insect larvae, and small spiders, where-
as in Juan Diaz small snails, beetles, and ants were the most common
invertebrates in migrants’ diet (Table 2).

Abundance of food resources. — Between-site comparisons in the index
of food abundance show that food resources were more abundant in Juan

Number of birds observed

752

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Fig. 1. Monthly abundance of migrant species at the two study sites.

Diaz from September to December (wet season) and more abundant in
Galeta from January to April (dry season) (Fig. 2). Actually, several in-
vertebrate groups (taxa/size) showed a significantly different abundance
between the wet and the dry seasons, with an opposite trend at each site
(Table 2). In Juan Diaz, nine invertebrate groups representing 53% of all
invertebrates sampled were significantly more abundant during the wet
season. In Galeta, 13 invertebrate groups representing 63% of all inver-
tebrates sampled were significantly more abundant during the dry season.

Relationships between migrant and food abundance. — Concurrent
changes in numbers of Northern Waterthrushes and in the index of food
abundance were positively correlated between the two sites (r^ = 0.5772,
df = \5, P < 0.05). No such relation was found for the Prothonotary
Warbler {r, = -0.1545, df = 1 1, F = 0.94) or the American Redstart (r,
= -0.0168, df = 1 1, F = 0.96). Between-site variations in abundance of

Lefebvre and Poulin • MIGRANTS IN MANGROVES

753

Table 2

Seasonal Variation in Sweep-Net Samples and Proportion of Each Invertebrate-

prey Group in the Migrant’s Diet

Taxa

Size

Sweep net

WS vs DS“

Number of items in migrants’
regurgitations (%)

Galeta

Juan Diaz

Galeta

Juan Diaz

Gastropoda (snails)

Small

***

=

65 (4.3)

1089 (49.8)

Gastropoda (snails)

Large

-F

-F-F

1 (0.1)

Pseudoscorpionida

Small

=

=

10 (0.7)

5 (0.2)

Araneida (spiders)

Small

•f + -F

**

127 (8.4)

41 (1.9)

Araneida (spiders)

Large

**

=

53 (3.5)

9 (0.4)

Isopoda

Small

=

1 (0.1)

13 (0.6)

Decapoda (crabs, shrimps)

Small

7

7

5 (0.3)

2 (0.1)

Decapoda (crabs, shrimps)

Large

7

7

3 (0.2)

4 (0.2)

Chilopoda (centipedes)

Large

=

7

1 (0.0)

Thysanura (bristletails)

Large

7

7

1 (0.0)

Odonata

Large

=

=

7 (0.5)

Orthoptera

Small

=

***

5 (0.3)

4 (0.2)

Orthoptera

Large

-t--F +

***

14 (0.9)

7 (0.3)

Dermaptera (earwigs)

Large

7

7

1 (0.0)

Heteroptera (true bugs)

Small

-F-F-l-

=

15 (1.0)

9 (0.4)

Heteroptera (true bugs)

Large

=

***

8 (0.5)

1 (0.0)

Homoptera (plant bugs)

Small

-F-F-F

=

4 (0.3)

88 (4.0)

Homoptera (plant bugs)

Large

-F

=

2 (0.1)

4 (0.2)

Coleoptera (beetles)

Small

+ -F-F

**

351 (23.4)

385 (17.6)

Coleoptera (beetles)

Large

=

***

13 (0.9)

9 (0.4)

Lepidoptera

Small

-F-F-F

=

1 (0.0)

Diptera (flies)

Small

**

**

12 (0.8)

15 (0.7)

Diptera (flies)

Large

=

=

1 (0.1)

2 (0.1)

Hymenoptera (ants)

Small

-F-F-F

=

294 (19.6)

321 (14.7)

Hymenoptera (ants)

Large

-F-F-F

-F-F-F

19 (1.3)

17 (0.8)

Hymenoptera (wasps)

Small

-F-F-F

=

87 (5.8)

76 (3.5)

Hymenoptera (wasps)

Large

-F-F-F

**

18 (1.2)

2 (0.1)

Insect eggs

Small

-F-F-F

-F-F-F

72 (4.8)

59 (2.7)

Insect eggs

Large

=

=

2 (0.1)

Insect larvae

Small

-F + -F

-F-F-F

42 (2.8)

13 (0.6)

Insect larvae

Large

=

=

266 (17.7)

7 (0.3)

Fishes

Large

7

7

5 (0.3)

Frogs

Large

7

7

1 (0.0)

Lizards

Large

7

7

1 (0.1)

" Comparison (G-tests) in numbers of sweep-netted invertebrates from the wet (WS; September through December) to
the dry (DS; January through April) season. Increa.se: + (P < 0.05), -H- (P < O.OI), + + + (F < 0.001); no significant
change: =; decrea.se: * (F < 0.05), ** (F < 0.01) »** (F < 0.001); ?: not sampled.

754

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Fig. 2. Monthly variation in the index of food abundance at the two study sites.

Other migrant species was positively correlated with between-site varia-
tions in the index of food abundance = 0.7815, df = 15, P < 0.001).

DISCUSSION

The two mangrove sites were used by a similar number of migrant
birds. However, all species, with the exception of the Prothonotary War-
bler, showed an opposite abundance pattern at the two sites. We do not
pretend that these variations were related to bird movements from one
mangrove site to another (no bird was captured at both sites), but our
data suggest that the two mangrove sites have a temporally complemen-
tary role in sustaining migrant populations during their stay in central
Panama.

Because of the differing tide and inundation patterns, rainfall season-
ality affects differently the invertebrate fauna at the two mangrove sites

Lefebvre and Poulin • MIGRANTS IN MANGROVES

755

(Lefebvre and Poulin, in press). As a result, invertebrate abundance was
higher in Pacific mangroves during the first part of the wintering period
(wet season) and higher in Caribbean mangroves during the second part
of the wintering period (dry season). This pattern was found in several
groups of arthropods, including the ones frequently taken by the migrant
species.

Between-site variations in abundance of Northern Waterthrushes were
synchronised with between-site variations in food abundance. This species
used extensively one mangrove only when food resources were propor-
tionally more abundant than at the other, and its abundance at any site
was low when food resources were proportionally less abundant.

Within and between site variations in abundance of Prothonotary War-
blers were not correlated to abundance of food resources. This species
was present only during the first part of the nonbreeding season when
resources were relatively stable at both sites. However, the Prothonotary
Warbler showed a different abundance pattern at the two sites, potentially
reflecting differences in food abundance. In Juan Diaz, where resources
were more abundant, the Prothonotary Warbler showed a high and uni-
form abundance. The departure of most individuals in February occurred
when food abundance decreased sharply. In Galeta, where food resources
were less abundant, abundance of Prothonotary Warblers decreased
throughout the nonbreeding period.

The abundance pattern of the American Redstart suggests movements
among mangrove patches, but these were not correlated with food abun-
dance. This was especially notable at Juan Diaz, where bird abundance
increased drastically during the dry season when resource abundance was
low. The few regurgitation samples collected for this species show that
the American Redstart, in contrast to other wood warblers, feeds exten-
sively on small homopterans and insect eggs (76% of all food items
taken). Small homopterans were more abundant in Juan Diaz than in
Galeta (total in sweep-net samples: 422 vs 713, G = 75.4, df = 1, P <
0.001), even considering only the dry season when homopteran abun-
dance increased in Galeta but remained stable in Juan Diaz (274 vs 377,
G = 16.4, df = 1, P < 0.001). Small insect eggs were one of the few
arthropod groups more abundant in the dry than the wet season at Juan
Diaz.

The other migrant species showed large variations in abundance with
an opposite pattern at each site. These species were predominantly found
on the Pacific side during southward migration and on the Caribbean side
during northward migration, which was positively correlated with varia-
tion in food resources at the two sites.

Large variations in migrant abundance occur in Panamanian man-

756

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

groves. Our data suggest that they are related to mid-winter movements
associated with variation in food availability. The proximity of the At-
lantic and Pacific coasts in Panama allowed the sampling of two man-
grove sites experiencing different hydrographic conditions within a short
range, but the variations observed in migrant abundance could conceal
bird movements on a larger geographical scale. In coastal mangroves of
Venezuela, the same warblers species exhibit large variations in abun-
dance during the non-breeding period (Lefebvre et al. 1992). Because of
their wide geographic distribution in the Neotropics and their particular
response to hydrographic factors, mangroves are probably more likely to
support large scale movements of migratory birds than other Neotropical
habitats. However, and regardless of the fact that few studies have inves-
tigated seasonal variation in migrant abundance throughout the nonbreed-
ing season, important changes in migrant numbers have been reported in
grassland (Greenberg 1992a), pine forest (Emlen 1980), lowland second-
growth forest (Blake and Loiselle 1992), lowland dry forest (Orejuela et
al. 1980, Greenberg 1984, Sherry and Holmes 1996), lowland wet forest
(Greenberg 1984), second-growth low mountain wet forest (Wunderle
1995), and low mountain wet forest (Johnson 1980) habitats. While these
variations are often assumed to be associated with mortality or local hab-
itat shifts, they could potentially reflect bird movements on a wide geo-
graphic distance. Nearctic-Neotropical migrants are more numerous and
exploit a greater variety of winter habitats in the northern Neotropics
(Terborgh and Faaborg 1980, Pashley and Martin 1988, Rappole et al.
1993) and are less numerous and more selective in their habitat choice
farther south (Pearson 1980, Bosque and Lentino 1987, Wunderle and
Waide 1993). Several studies carried out in the West Indies (Emlen 1980,
Sherry and Holmes 1996, Wunderle 1995) and Central America (Galindo
et al. 1963, Blake and Loiselle 1992, Greenberg 1992a) have reported a
decrease in migrant abundance over the non-breeding season, whereas
some habitats in South America support an increasing number of migrants
throughout winter (Johnson 1980, Hilty 1980, this study). Considering
the important changes occurring in food resources between the wet and
the dry seasons in most Neotropical habitats (Poulin et al. 1992, Sherry
and Holmes 1996), quality of the foraging microhabitat is likely to vary
during the nonbreeding season. A migration in mid-winter could then
represent an advantageous strategy for several migrant birds.

These migrations during the non-breeding period could correspond to
facultative migrations which are directly in response to changes in envi-
ronmental conditions and may or may not occur in any given year (Bert-
hold 1975, Terrill 1990). Such migrations have been observed in long
distance Palearctic insectivorous migrants (Gwinner et al. 1988, Wood

Lefebvre and Poulin • MIGRANTS IN MANGROVES

757

1979) and short distance Nearctic insectivorous migrants (Terrill and
Ohmart 1984) experiencing an important decrease in food resources. Cage
experiments have demonstrated that between fall and spring migrations,
some insectivorous long distance migrants react to food shortage by night
activity {Zugunnihe), similar to that observed during the obligatory mi-
gration phase (Terrill and Ohmart 1984, Gwinner et al. 1988). Actually,
several lines of evidence suggest a much higher potential for extensive
winter movement by migrants than has generally been considered to be
the case (Curry-Lindahl 1981, Gwinner 1990, Terrill 1990).

Examples of winter site tenacity by migrants are numerous in the Neo-
tropics (see Rappole and Warner 1980). However, the proportion of in-
dividuals exhibiting site tenacity within a species often varies among hab-
itats (Greenberg 1984, Sherry and Holmes 1996, Wunderle 1995) and
years (Emlen 1980, Greenberg 1992a, Wunderle 1995), potentially re-
flecting mid-winter migration by some individuals in response to a de-
crease in foraging microhabitat quality. Understanding seasonal changes
in population and resource use is critical in providing efficient manage-
ment policies for migratory birds (Greenberg 1992b, Sherry and Holmes
1995). Occurrence of multiple stage migrations has been documented,
mostly for the Palearctic-African migratory system, and need to be in-
vestigated in the Nearctic-Neotropical realm for proper conservation of
migrant species.

ACKNOWLEDGMENTS

The Natural Sciences and Engineering Research Council of Canada (NSERC) supported
this study. We are grateful to the Smithsonian Tropical Research Institute for providing
logistical support and to INRENARE (Instituto de recursos naturales renovables) for issuing
the research permit to work in mangroves. We thank Stanley Rand, Neal Smith, and Russell
Greenberg for providing helpful comments on this manuscript.

LITERATURE CITED

Arendt, W. J. 1992. Status of North American migrant landbirds in the Caribbean region:
a summary. Pp, 143-171 in Ecology and conservation of Neotropical migrant landbirds
(J. M. Hagan and D. W. Johnston, eds.). Smith.sonian Institution Press, Washington,
DC.

Berthold, P. 1975. Migration: control and metabolic physiology. Pp. 77-128 in Avian
biology. Vol. 5 (D. S. Earner, J. R. King, and K. C. Parkes, eds.). Academic Press. New
York, New York.

Blake, J. G. and B. A. Loiselle. 1992. Emits in the diets of Neotropical migrant birds in
Costa Rica. Biotropica 24:200-210.

Bosque, C. and M. Lentino. 1987. The passage of North American migratory land birds
through xerophytic habitats on the western coast of Venezuela. Biotropica 19:267—273.
Curry-Lindahl, K. 1981. Bird migration in Africa. Academic Pre.ss. New York, New York.
Duke, N. C. 1990. Phonological trends with latitude in the mangrove tree Avicennia marina.
J. Ecol. 78:1 13-133.

758

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

. 1992. Mangrove floristic and biogeography. Pp. 63-100 in Coastal estuarine stud-
ies. Tropical mangrove ecosystems (A. I. Robertson and D. M. Alongi, eds.). American
Geophysical Union, Washington, D.C.

Emlen, J. T. 1980. Interactions of migrant and resident landbirds in Llorida and Bahama
pinelands. Pp. 133—143 in Migrant birds in the Neotropics: ecology, behavior, distri-
bution and conservation (A. Keast and E. S. Morton, eds.). Smithsonian Institution
Press, Washington, D.C.

Galindo, P, E. Mendez, and A. J. Adames. 1963. Banding of migrant thrushes in Almi-
rante, Panama. Bird-Banding 34:202-209.

Greenberg, R. 1984. The winter exploitation systems of Bay-breasted and Chestnut-sided
Warblers in Panama. Univ. Calif. Publ. Zool. 116:1-107.

. 1992a. Eorest migrants in non-forest habitats on the Yucatan Peninsula. Pp. 273-

286 in Ecology and conservation of Neotropical migrant landbirds (J. M. Hagan and

D. W. Johnston, eds.). Smithsonian Institution Press, Washington, D.C.

. 1992b. The nonbreeding season: introduction. Pp. 175-177 in Ecology and con-
servation of Neotropical migrant landbirds (J. M. Hagan and D. W. Johnston, eds.).
Smithsonian Institution Press, Washington, D.C.

Gwinner, E. 1990. Bird migration: physiology and ecophysiology. Springer- Verlag, Berlin.

, H. ScHWABL, AND I. Schwabl-Benzinger. 1988. Effects of food-deprivation on

migratory restlessness and diurnal activity in the Garden Warbler Sylvia borin. Oecol-
ogia 77:321-326.

Hilty, S. L. 1980. Flowering and fruiting periodicity in a premontane rain forest in Pacific
Colombia. Biotropica 12:292-306.

Hutto, R. L. 1980. Winter habitat distribution of migratory land birds in western Mexico,
with special reference to small, foliage-gleaning insectivores. Pp. 181-203 in Migrant
birds in the Neotropics: ecology, behavior, distribution and conservation (A. Keast and

E. S. Morton, eds.). Smithsonian Institution Press, Washington, D.C.

Johnson, T. B. 1980. Resident and North American migrant bird interactions in the Santa-
Marta highlands. Northern Colombia. Pp. 239-247 in Migrant birds in the Neotropics:
ecology, behavior, distribution and conservation (A. Keast and E. S. Morton, eds.).
Smithsonian Institution Press, Washington, D.C.

Lefebvre, G. and B. Poulin. 1996. Bird communities in Panamanian mangroves: potential
effect of physical and biotic factors. J. Trop. Ecol. (in press).

, , AND R. McNeil. 1992. Abundance, feeding behavior and body condition

of Nearctic warblers wintering in Venezuelan mangroves. Wilson Bull. 104:400-412.

, , and . 1994a. Spatial and social behaviour of Nearctic warblers

wintering in Venezuelan mangroves. Can. J. Zool. 72:757-764.

, , and . 1994b. Temporal dynamics of mangrove bird communities

in Venezuela with special reference to migrant warblers. Auk 1 1 1:405-415.

Lynch, J. E 1989. Distribution of overwintering Nearctic migrants in the Yucatan Penin-
sula, 1: general patterns of occurrence. Condor 91:515-544.

Morton, E. S. 1980. Adaptations to seasonal changes by migrant land birds in Panama
Canal Zone. Pp. 437-453 in Migrant birds in the Neotropics: ecology, behavior, distri-
bution and conservation (A. Keast and E. S. Morton, eds.). Smithsonian Institution
Press, Washington, D.C.

Orejuela, j. E., R. j. Rait, and A. Humberto. 1980. Differential use by North American
migrants of three types of Colombian forests. Pp. 253-264 in Migrant birds in the
Neotropics: ecology, behavior, distribution and conservation (A. Keast and E. S. Mor-
ton, eds.). Smithsonian Institution Press, Washington, D.C.

Pashley, D. N. and P. R. Martin. 1988. The contribution of Christmas Bird Count to

Lefebvre and Poulin • MIGRANTS IN MANGROVES

759

knowledge of winter distribution of migratory warblers in Neotropics Am Birds 42-
1164-1176.

Pearson, D. L. 1980. Bird migration in Amazonian Ecuador, Peru and Bolivia. Pp. 273-
283 in Migrant birds in the Neotropics; ecology, behavior, distribution and conservation
(A. Keast and E. S. Morton, eds.). Smithsonian Institution Press, Washington, D.C.

Poulin, B. and G. Lefebvre. 1995. Additional information on the use of emetic in deter-
mining the diet of tropical birds. Condor 97:897-902.

’ ’ and R. McNeil. 1992. Tropical avian phenology in relation to abundance

and exploitation of food resources. Ecology 73:2295-2309.

’ ’ and . 1994. Diet of land birds from northeastern Venezuela. Con-

dor 96:354-367.

Rappole, J. H., E. S. Morton, T. E. Lovejoy, and J. L. Ruos. 1993. Aves migratorias
nearticas en los Neotropicos. Conservation and Research Center, National Zoological
Park, Smithsonian Institution, Front Royal, Virginia.

AND D. W. Warner. 1980. Ecological aspects of migrant bird behavior in Veracruz,
Mexico. Pp. 353-392 in Migrant birds in the Neotropics: ecology, behavior, distribution
and conservation (A. Keast and E. S. Morton, eds.). Smithsonian Institution Press,
Washington, D.C.

Russell, S. M. 1980. Distribution and abundance of North American migrants in lowlands
of Northern Colombia. Pp. 249-252 in Migrant birds in the Neotropics: ecology, be-
havior, distribution and conservation (A. Keast and E. S. Morton, eds.). Smithsonian
Institution Press, Washington, D.C.

Sherry, T. W. and R. T. Holmes. 1995. Summer versus winter limitation of populations:
what are the issues and what is the evidence? Pp. 85-120 in Ecology and management
of Neotropical migratory birds: a synthesis and review of critical issues (T. E. Martin
and D. M. Finch, eds.). Oxford Univ. Press, New York, New York.

AND . 1996. Winter habitat quality, population limitation, and conservation

of Neotropical-Nearctic birds. Ecology 77:36-48.

Terborgh, j. W. and j. R. Faaborg. 1980. Factor affecting the distribution and abundance
of North American migrants in the eastern Caribbean region. Pp. 145-155 in Migrant
birds in the Neotropics: ecology, behavior, distribution and conservation (A. Keast and
E. S. Morton, eds.). Smithsonian Institution Press, Washington, D.C.

Terrill, S. B. 1990. Food availability, migratory behavior, and population dynamics of
terrestrial birds during the nonreproductive season. Stud. Avian Biol. 13:438-443.

AND R. D. Ohmart 1984. Facultative extension of fall migration by Yellow-rumped

Warblers (Dendroica coronata). Auk 101:427-438.

Wood, B. 1979. Changes in numbers of over-wintering Yellow Wagtails, MotaciHa flava,
and their food supplies in a west African savanna. Ibis 121:228-231.

WuNDERLE, J. M., Jr. 1995. Population characteristics of Black-throated Blue Warblers
wintering in three sites on Puerto Rico. Auk 112:931-946.

AND R. B. Waide. 1993. Distribution of overwintering Nearctic migrants in the

Bahamas and Greater Antilles. Condor 95:904-933.

Wilson Bull., 108(4), 1996, pp. 760-770

EFFECTS OF CONSERVATION RESERVE PROGRAM
FIELD AGE ON AVIAN RELATIVE ABUNDANCE,
DIVERSITY, AND PRODUCTIVITY

Kelly F. Millenbah, Scott R. Winterstein,

Henry Campa III, Ly T. Furrow, and
Richard B. Minnis

Abstract. — Introduced grass dominated Conservation Reserve Program (CRP) fields
were monitored in summer 1992 in Gratiot County, Michigan, to determine the relationship
between field age and avian relative abundance, diversity, and productivity. Younger CRP
fields (1-2 years old), best described as a combination of forbs and bare ground, had the
greatest diversity and relative abundance of avian species. Older CRP fields (3-5/6 years
old) were a combination of grasses and deep litter cover and had the greatest avian pro-
ductivity. We recommend that after 3-5 growing seasons CRP fields be manipulated to
provide a variety of successional stages to maintain simultaneously high avian relative abun-
dance, diversity, and productivity. Received 6 Nov. 1995, accepted I May 1996.

Specialization and intensification of agricultural practices have contrib-
uted to dramatic declines in wildlife populations over the last 60 years
(USDA 1972). These land-use practices have decreased the availability
of cover types (i.e., number of grasslands and wetlands) available for
many farmland wildlife species (Berner 1988). In the Midwest, agricul-
tural practices that adversely impact wildlife are commonly used because
the production of quality wildlife habitat provides reduced economic re-
turns to farmers compared to commodity production (Langer 1979). To
alleviate excess commodity production and to combat the effects of past
agricultural practices, the federal government initiated land retirement
programs beginning in the 1930s. These programs provided varying
amounts and qualities of wildlife habitat (Berner 1988). The most recent
land retirement program is the Conservation Reserve Program (CRP). The
CRP provisions of the 1985 Food Security Act (Farm Bill) provide eco-
nomic incentives to farmers to remove highly erodible and environmen-
tally sensitive cropland from production for 10 years. Perceived benefits
of this program include curtailing soil erosion and excess commodity
production and creating wildlife habitat.

The CRP offers a unique opportunity to view the successional dynam-
ics of grasslands and associated changes in wildlife populations for 10
years. Avian communities should provide insight into the quality of hab-
itat provided by the CRP because avian species are excellent indicators
of habitat quality and respond quickly to environmental changes (Graber

Michigan State Univ., Dept, of Fisheries and Wildlife, East Lansing, Michigan 48824.

760

Millenbah et al. • EFFECTS OF CRP ON AVIAN SPECIES

761

and Graber 1976). Determining which age or ages of CRP fields support
the greatest relative abundance, diversity, and productivity of avian spe-
cies provides a framework on which management recommendations can
be based. Once created and maintained, these grassland habitats should
increase the above-mentioned avian variables in a landscape dominated
primarily by less diverse agricultural monocultures. The objective of this
study was to determine the relationship of various age classes of CRP
fields to avian relative abundance, diversity, and productivity by quanti-
fying vegetation structure and composition of CRP fields and the asso-
ciated changes in avian variables.

STUDY AREA AND METHODS

Nineteen 6.5-20 ha study sites were selected in Gratiot County, Michigan, in 1992. Each
CRP field varied in age from 1-6 years (1-4 [N = 3 in each], 5 [N = 6], and 6 years [N
= 1]) and was planted to introduced grasses and legumes (alfalfa [Medicago sativa], orchard
grass [Dactlyis glomerata], timothy grass [Phleum pratense], and clover [Trifolium spp.])
(Millenbah 1993). None of the fields in this study was mowed, grazed, burned, or disced
since contract initiation.

Data collected in 1992 represent a time-specific analysis of different age CRP fields.
Rather than use a cohort approach, where a sample of fields would be examined annually
for a number of years, representatives from different age classes were observed within a
year. Using this approach, two assumptions were made. First, it was assumed that all fields
selected for observation received similar treatment. That is, all fields were planted in a
similar manner to a similar mix of grasses and legumes, and no field had been disturbed
since contract initiation. Second, it was assumed that weather had affected fields similarly.
However, if there were differences among fields, for either assumption, it is likely that they
were distributed randomly across age classes. Therefore, because both assumption were
likely met, the time-specific approach provides an adequate representation of the changes
in the measured variables as CRP fields age.

Vegetation structure and composition data were collected along six permanent 100 m
long, systematically established transects in each field, with six sampling points per line.
Sampling occurred every 20 m along each transect in May (peak avian breeding season).
Horizontal cover was assessed 4 m from a Robel pole (Robel et al. 1970). Maximum height
of living and standing dead vegetation was recorded at each sampling point. Percentage
canopy cover of live and dead vegetation; percentage canopy cover of grasses, forbs, and
woody vegetation; percentage litter cover; and percentage bare ground were assessed using
a 50 by 50 cm sampling frame (modified from Daubenmire 1959). Frequency of all her-
baceous species occurring within the sampling frame and litter depth were recorded at each
.sampling point.

Bird censuses were conducted in each field to determine relative species abundance and
diversity. Counts were made from 1 May- 15 August every two weeks. Censuses were made
from transects spaced 50 m apart along the long axis of the field. Censuses were conducted
from sunrise to three hours after sunrise. Counts were not made if it was raining or if wind
speed exceeded 16 kph. Observers walked slowly along transect lines making frequent stops
to scan for birds. All birds seen or heard were recorded. Only those birds observed within
25 m of either side of the transect were included in the census to minimize double counting.
Relative bird abundance was reported as bird.s/ha.

Entire fields were searched for nests to quantify nesting success on different age classes

762

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

of CRP fields. Nine fields representing three age classes (1 [N = 3], 4 [N = 2], and 5 years
old [N = 4]) were searched in mid-May and mid-June. Searches were conducted with 3-6
observers walking 1-2 m apart until fields had been completely traversed. Nests were re-
visited every 2-3 days until young had fledged or the nest was determined to be abandoned
or destroyed. Incidentally discovered nest were also monitored.

Comparisons of vegetation variables and avian relative abundance and diversities among
age classes within sampling periods and nesting success among age classes were made using
Kruskal-Wallis (KW) one-way analysis-of-variance (Siegel 1956). A KW multiple compar-
ison test (Miller 1980) was used to determine which field age classes were significantly
different if there was a significant (a = 0.10) KW result. Because there was only one 6-year-
old field, this field was combined with 5-year-old fields for all statistical tests (hereafter
referred to as 5/6) on avian relative abundance and diversity and vegetation variables.

Avian diversities were calculated using the Shannon- Weaver diversity index. Nest success
(nests surviving from initiation of egg laying to fledging) was calculated using the Mayfield
(1961) method.

Friedman’s two-way analysis-of-variance (Siegel 1956) was used to test for differences
among age classes for avian relative abundance and diversities over the census period (May-
August). Blocking was done on the census period. A Friedman’s multiple comparison test
(Miller 1980) was used to determine which age classes were significantly different over the
census period if there was a significant (a = 0.10) Friedman’s result.

Friedman’s two-way analysis of variance also was used to test for differences in relative
frequencies of dominant plant species among fields within an age class. The five plant
species with the greatest relative frequencies on each field were included in the analysis.
Failure to reject the null hypothesis would suggest that an age class of CRP field could be
described by the dominant plant species present.

Principal components analysis (PCA) was used to examine the relationship between field
age, vegetation characteristics, and avian relative abundance and diversities. Because mea-
sured vegetation variables may be related, PCA was used to reduce the number of variables
to a few independent variables. The new variables, or principal components, were linear
combinations of the original vegetation variables. The linear combinations maximized the
variance in the data and could be used to identify the original variables most significant in
describing a particular age of CRP field. Analysis was done using a correlation matrix.

RESULTS

Eighty-two plant species were identified on 19 CRP fields. Individual
fields within each age class differed (Friedman, P < 0.10) in the relative
frequencies of plant species identified. Therefore, a particular age of CRP
field could not be described by the dominant plant species present. These
findings preclude any meaningful test of changes in dominant plant spe-
cies composition as fields age. Similarly, we observed no consistent suc-
cessional replacement of dominant plant species as fields aged (Millenbah,
unpubl. data).

Vegetation characteristics that differed significantly (KW, P < 0.10)
among age classes included horizontal cover, percent total canopy, percent
live canopy, percent dead canopy, percent grass canopy, percent forb can-
opy, percent litter cover, percent bare ground, and litter depth (Table 1).
However, none of the significant differences was consistently related to

Table 1

Mean (SE) Vegetation Characteristics on Different Age Classes of Conservation Reserve Program (CRP) Fields

Millenbah et al. • EFFECTS OF CRP ON AVIAN SPECIES

763

rT

q

P

q

q

q

P

— *

d

d

(N

CN

d

cd

cd

d

d

d

'w'

s_,.

o

Iri

ffi

CQ

u

CQ

CQ

CQ

CQ

CQ

<

<

<

<

<

<

<

CQ

<

CO

(N

q

q

(N

q

q

q

d

CO

d

d

so

in

d

d

d

CN

in

sO

in

CN

in

q

d

CN

OS

q

q

so

q

o

(N

V

d

sd

»n

d

CN

00

d

d

d

d

w-

■w

■s_^

u

CQ

CQ

CQ

CQ

(N

u

CQ

CQ

CQ

<

<

CQ

<

00

q

00

•q

o

oo

q

r-

iri

d

CN

CO

d

cd

00

d

mi

CN

so

00

in

CN

d

2

m

wo

CO

q

00

CN

d

00

SD

q

eg

d

d

uo

in

d

d

CN

V

cd

d

cd

"o

u

OJD

cn

ffi

u

CQ

CQ

<

CQ

<

<

<

CQ

CQ

CQ

CQ

<

(N

CN

q

q

q

—

q

q

q

rn

d

d

cd

in

CN

d

d

CN

d

»n

V

m

y— V

y— S

/.—V

y<-V

1-^

q

,^—v

N

..—v

s,

vO

Os

q

q

in

'

d

r-*

•q

q

00

d

d

d

d

d

.—1

d

cd

CN

d

•w-

N— ^

N— ^

cT

u

CQ

CQ

<

<

CQ

CQ

<

<

<

<

<

q

CN

CO

q

00

On

q

00

CN

CO

d

00

sd

d

d

d

cd

00

r--

CN

»n

s

s

/— s.

/—> V

s

V,

q

00

q

q

OS

q

o

CN

Os

d

d

d

cd

d

d

d

d

5

N—

CQ

<

CQ

<

<

<

<

<

<

<

q

Q)

q

OS

—

o

q

q

q

r-i

(N

d

uo

d

cd

sd

d

d

mi

cd

CN

<r)

»n

(N

CN

(N

CO

e

6

"O

TD

c

C

_0

£

0

*w

2

o

c

cd

c

a>

V5

•o

S

W)

•c

y

b.

f3

.C

O

u

1)

>

o

o

2

W)

(D

>

(U

>

(U

>

•n

c3

<U

TD

canopy*’

£

>>

a

0

c

cd

Q

£i

?o

a

0

c

cd

o

>-*

a

o

c

cd

o

JZ

>%

C.

0

c

cd

o

50

a

o

c

cd

a

50

t-

V

>

o

u

a

-o

c

o

o

u

00

?

(J

s

D-

0

0

“

c/3

•o

u

<U

c

0

N

'u

2?

.£?

w

eS

<u

>

d

cd

(U

Q

C/3

Cd

u

a

-D

k«

O

d

0

1

1)

d

<u

u

cd

PQ

"O

u

2

O

X

‘S

a:

’S

:i:

d

a.

E

o

o

V

Cl.

X> ot)

E ^
o ot)
o c
O

2 £

w -o o
- 00

li s

c f=

flj c ■€
> 00
tl! o'

764

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

2.0n r2.0

Time (years)

Lig. 1 . Diagrammatic representation of changes in vegetation structure and composition
on a Conservation Reserve Program field over the first six growing seasons.

field age. General patterns for horizontal cover, height of live and dead
vegetation, percent total canopy, percent live canopy, percent dead can-
opy, and percent forb canopy suggest an increase (but not necessarily
significantly) from 1- to 4-year-old fields, with the exception of a decrease
in these variables during the third growing season (Fig. 1). These vege-
tation characteristics all decreased from the fourth to 5/6 growing seasons.
Percent grass canopy increased through the fourth growing season, de-
creasing in the 5/6 growing season. Percent litter cover increased with
field age, stabilizing after the third growing season. Litter depth was great-
est during the third growing season, subsequently decreasing through the
5/6 growing season, whereas percent bare ground decreased from 1- to
5/6-year-old fields.

The first three principal components accounted for 81.5% of the vari-
ance in the vegetation variables (Fig. 2). Principal component one (PC
1), explaining 35.2% of the variance, represents a successional change in
vegetation attributes from greater percent total and live canopy to greater
percent litter cover and litter depth. Percent total canopy included percent
live canopy and percent dead canopy but not percent litter cover. The
second principal component (PC 2), explained 32.5% of the variance and
represents a successional change from greater percent grass cover to great-
er percent forb cover. Principal component three (PC 3), explaining 13.7%
of the variance, represents a successional change from greater percent
bare ground to greater percent litter cover.

Thirty-two avian species were encountered on the CRP fields. The most
common species encountered were Red-winged Blackbirds {Agelaius

Millenbah et al. • EFFECTS OF CRP ON AVIAN SPECIES

765

for different age CRP fields. Age classes correspond to the appropriate number located at
each point.

phoeniceus). Song Sparrows {Melospiza melodia). Bobolinks {Dolichonyx
oryzivorus), and Sedge Wrens (Cistothorus platensis).

No differences (KW, P > 0.10) were detected in bird diversities within
a census count among age classes. However, mean avian diversities were
different (Friedman, P = 0.04) for the entire census period among age
classes (Table 2), with 1 -year-old fields having significantly greater di-
versities than 5/6-year-old fields.

Only the periods of 16-31 May and 1-15 August showed differences
(KW, P = 0.06) in avian relative abundance among age classes, with 5/6-
year-old fields having significantly lower relative abundance than 4-year-
old fields for the period 1-15 August. However, a KW multiple compar-
ison test did not detect differences among age classes for the period 16-
31 May. Mean avian relative abundances were different for the entire
census period among age classes (Friedman, P = 0.01; Table 2). Five/

766

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 2

Mean Avian Relative Abundance (Birds/ha) and Diversity (Shannon-Weaver) on
Different Age Conservation Reserve Program (CRP) Fields^

Age class^*

N

Relative
abundance (SE)

Diversity (SE)

I

3

4.24 (LOl)AB

1.37 (0.07^

2

3

3.72 (0.63>^B

1.36 (0.10)^®

3

3

3.30 (0.50)''B

1.28 (0.08)A®

4

3

4.67 (0.57)-'

1.18 (0.04)'^®

5/6

7

2.12 (0.19)®

1.15 (0.06)®

" Within columns,
Miller 1980).

Five- and 6-year-

means with the same

•old fields combined.

letter are not significantly different (Friedman’s

multiple comparison, a = 0.10,

six-year-old fields had significantly lower relative abundance than 4-year-
old fields. Avian relative abundances decreased from 1- to 5/6-year-old
fields with an increase on 4-year-old fields.

We found 166 active nests in three age classes of CRP fields. Nesting
species monitored included Red-winged Blackbird, Vesper Sparrow
(Pooecetes gramineus). Sedge Wren, Northern Harrier (Circus cyaneus).
Mallard (Anas platyrhynchos). Ring-necked Pheasant (Phasianus colchi-
cus), and unidentified sparrow species. Red-winged Blackbirds were the
primary nesting species observed with 83.1% of the monitored nests.

No difference (KW, P - 0.29) was found in percent successful nests
among 1-, 4-, and 5-year-old fields. However, nests on older fields (4-
and 5-year-old) had greater probabilities (SMayfieid ~ 0.283 and 0.293, re-
spectively) of surviving from initial egg laying to fledging than 1 -year-
old fields (SMayfieid = 0.138). Mean number of active nests (nests have at
least one egg or young within the monitoring period) on 1-, 4-, and
5-year-old fields was 10, 22, and 23, respectively.

DISCUSSION

Although we could not distinguish age classes of CRP fields by the
presence of particular dominant plant species (i.e., alfalfa, orchard grass),
they could be described by gross structural characteristics (i.e., grass can-
opy, forb canopy). Fields in this study may be described as in a gradient
from greater forb cover and bare ground in the youngest fields to greater
grass and litter cover in the oldest (Fig. 1). Forb canopy cover was greater
on younger fields due to initial seed mixtures and the natural invasion of
other plant species, with many annual forbs responding to soil distur-
bance. Because of the recent disturbance of planting, bare ground was
more dominant than litter cover on younger fields.

Millenhah et al. • EFFECTS OF CRP ON AVIAN SPECIES

767

Two-year old CRP fields, however, were best described by moderate
forb canopy and litter cover, with a greater total/live canopy than bare
ground (Fig. 1). Although the initial seed mixtures of younger CRP fields
contained both alfalfa and orchard grass, orchard grass takes several years
to establish, whereas alfalfa is noted for its quick establishment (J. Swan-
son, Gratiot County SCS, pers. commun.). Alfalfa, however, has a rela-
tively short life cycle and begins to die out after two growing seasons.
Total/live canopy would be greater on younger fields than litter cover/
depth because a substantial litter layer has not yet developed. However,
a litter layer was evident, and was proportionally greater than the amount
of bare ground present.

Three-, 4-, and 5/6-year-old fields can be described by grass canopy
and litter cover, with a decreasing total/live canopy (Fig. 1). As alfalfa
begins to die back after the second growing season, grasses begin to
dominate sites. The decline in forb cover after the first two growing sea-
son may be explained by the loss of alfalfa and other forbs and the en-
croachment of grasses. Basu et al. (1978) stated that vegetation on a
legume dominated field undergoes successional changes, eventually be-
coming grass-dominated and sparser. As grasses began to dominate and
out compete existing forb species, forb cover decreased. Also, as CRP
fields age, a litter layer develops decreasing the amount of bare ground
and serves as a mechanical barrier to grass development (Rice and Parenti
1978). This barrier decreases productivity of plants present, growth is
isolated to surviving clumps, and total/live canopy cover becomes less
dense.

It has long been accepted that vegetation structural complexity is as-
sociated with avian community structure (MacArthur and MacArthur
1961, Cody 1968). Typically, both avian species diversity and density
increase with increasing habitat complexity (May 1982). This may only
be expressed on established habitat types (i.e., forests, old fields) and may
not be valid for newly established habitats (such as 1 -year-old CRP
fields).

Habitat complexity is likely not the primary or only factor affecting
influx of birds onto 1 -year-old CRP fields. Younger fields may meet a
variety of habitat needs including feeding and nesting habitat not available
on alternative vegetation types present in the landscape, such as agricul-
tural fields. Although nesting did occur on younger CRP fields, produc-
tivity was less than that observed on older fields. Younger CRP fields
may provide some suitable habitat for nesting, but it is likely that amount
and quality of nesting habitat is limited. Roseberry and Klimstra (1984)
reported that areas dominated by annual weeds (forb canopy) may provide

768

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

inferior nesting cover because of the lack of dead grass stems for nest
construction.

Older CRP fields did not have the same high avian relative abundances
as younger fields. Burger et al. (1990) suggested that CRP fields in Mis-
souri do not provide quality nesting cover for Northern Bobwhites {Col-
inus virginianus) until the third year after establishment. Therefore, it may
be possible that a greater quality and availability of nesting habitat was
provided on older CRP fields in Michigan, thus supporting greater nesting
compared to younger CRP fields.

Cody (1985) stated that avian species composition, or diversity, varies
with vegetation structure following a disturbance, thereby creating a di-
verse array of avian species. Species diversity observed on CRP fields
supports this. One-year-old CRP fields, newly disturbed by planting, sup-
ported the greatest diversity of avian species. As fields aged and became
less disturbed diversities declined.

While many factors may be responsible for increasing avian produc-
tivity as fields age, this increase may be an artifact of the most dominant
nesting species encountered. Red-winged Blackbirds which nest in a va-
riety of locations with highly variable structural attributes (Granlund
1991). The conspicuous locations of Red-winged Blackbird nests allowed
for easier detection. It is likely that nests of other species were missed
due to the density of the vegetation. Results may only represent patterns
in Red- Winged Blackbird nesting and not productivity of the entire CRP
avian community.

Several studies have suggested that grasslands established with seed
mixtures similar to planted CRP fields generally did not maintain struc-
tural qualities for more than seven years (Higgins et al. 1987). Distur-
bances on 3- to 5-year intervals enhance avian production by more than
100% (Kirsch 1974, Kirsch et al. 1978). There is general recognition (e.g.,
Schenck and Williamson 1991) that controlled, periodic treatments to re-
vitalize cover by fire, grazing, or mowing may be necessary for the long-
term maintenance of wildlife habitat in grassland ecosystems. Our results
indicate perturbations may be necessary to maintain the greatest avian
relative abundances, diversities, and productivity on CRP fields after 3-
5 growing seasons. Further studies need to be completed to assess fully
changes in vegetation attributes and avian diversity, relative abundance,
and productivity on fields >6 years old.

Many types of disturbances such as mowing, burning, or discing may
create the desired changes in the vegetation. Regardless of the form of
perturbation, it should be accomplished in as short a period of time as
possible and scheduled to minimize the disruptive effects to nesting wild-
life. Although information is available on effects of disturbance practices

Millenbcih et til. • EFFECTS OF CRP ON AVIAN SPECIES

769

on grassland birds (e.g., Kirsch et al. 1978), CRP fields are a unique
vegetation type in Michigan s agricultural landscape, and little is known
about effects of disturbances on avian species using CRP fields. Addi-
tional information is needed on maintenance and rejuvenation methods
for planted CRP grasslands and responses of avian species to these man-
agement practices.

ACKNOWLEDGMENTS

The Michigan Agricultural Experiment Station (NC-203 regional research project). Fed-
eral Aid in Wildlife Restoration Project W-127-R (administered by the Michigan Dept, of
Natural Resources, Wildlife Division), Michigan Chapters of Pheasant Forever, and the
Frank M. Chapman Foundation funded this study. Earlier versions of this manuscript were
reviewed by L. Best, L. W. Burger, A. H. Farmer, and B. Leopold. Special thanks are
extended to all cooperating landowners who allowed research to be conducted on their
properties. Assistance in data collection was provided by interns S. Miller, J. DeDoes, and
K. O’Brien and volunteers J. Fierke, M. Smith, and D. Hyde.

LITERATURE CITED

Basu, P. K., H. R. Jackson, and V. R. Wallen. 1978. Alfalfa decline and its cause in
mixed hay fields determined by aerial photography and ground survey. Can. J. Plant
Sci. 58:1041-1048.

Berner, A. H. 1988. The 1985 farm act and its implications for wildlife. Pp. 437—465 in
Audubon wildlife report 1988/1989 (W. J. Chandler and L. Labate, eds.). The National
Audubon Society, New York, New York.

Burger, L. W. Jr., E. W. Kurzejeski, T. V. Dailey, and M. R. Ryan. 1990. Structural
characteristics of vegetation in CRP fields in northern Missouri and their suitability as
bobwhite habitat. Trans. N. Am. Wildl. Nat. Res. Conf. 55:74-83.

Cody, M. L. 1968. On the methods of resource division in grassland bird communities.
Am. Nat. 102:107-147.

. 1985. Habitat selection in birds. Academic Press, Inc., Orlando, Florida.

Daubenmire, R. F. 1959. A canopy coverage method of vegetational analysis. Northwest
Sci. 33:43-64.

Graber, j. W. and R. R. Graber. 1976. Environmental evaluations using birds in their
habitat. Biol. Notes No. 97. Illinois Nat. Hist. Surv., Urbana, Illinois.

Granlund, j. G. 1991. Red-winged Blackbird. Pp. 494-495 in The atlas of breeding birds
of Michigan (R. Brewer, G. A. Peek, and R. J. Adams, Jr., eds.). Michigan State Univ.
Press, East Lansing, Michigan.

Higgins, K. F, D. E. Nomsen, and W. A. Wentz. 1987. The role of the Conservation
Reserve Program in relation to wildlife enhancement, wetlands, and adjacent habitat in
the northern Great Plains. Pp. 99-104 in Impacts of the Conservation Reserve Program
in the Great Plains (J. E. Mitchell, ed.). USDA For. Serv. Gen. Tech. Rep. RM-158.
Kirsch, L. M. 1974. Habitat management considerations for prairie chickens. Wildl. Soc.
Bull. 2:124-129.

, H. F. Duebbert, and A. D. Kruse. 1978. Grazing and haying effects on habitats

of upland nesting birds. Trans. N. Am. Wildl. Nat. Resour. Conf. 43:486-497.

Fanger, L. L. 1979. An economic perspective of the effects of federal con.servation policies
on wildlife habitat. Trans. N. Am. Wildl. Nat. Resour. Conf. 50:200—210.

770

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

MacArthur, R. H. and J. W. MacArthur. 1961. On bird species diversity. Ecology 42:
594-598.

May, P. G. 1982. Secondary succession and breeding bird community structure: patterns
of resource utilization. Oecologia 55:208—216.

Mayfield, H. M. 1961. Nesting success calculated from exposure. Wilson Bull. 73:255-
261.

Millenbah, K. E 1993. The effects of different age classes of fields enrolled in the Con-
servation Reserve Program in Michigan on avian diversity, density, and productivity.
M. S. thesis, Michigan State Univ., East Lansing, Michigan.

Miller, R. P. 1980. Simultaneous statistical inference. Second ed. Springer- Verlag, New
York, New York.

Rice, E. L. and R. L. Parenti. 1978. Causes of decreases in productivity in undisturbed
tall grass prairie. Am. J. Botany 65:1091-1097.

Robel, R. j., j. N. Briggs, A. D. Dayton, and L. C. Hulbert. 1970. Relationships between
visual obstruction measurements and weight of grassland vegetation. J. Range Manage.
23:295-298.

Roseberry, j. L. and W. D. Klimstra. 1984. Population ecology of the bobwhite. Southern
Illinois Univ. Press, Carbondale, Illinois.

ScHENCK, E. W. AND L. L. WILLIAMSON. 1991. Conservation Reserve Program effects on
wildlife and recreation. Pp. 37-42 in The Conservation Reserve — yesterday, today, and
tomorrow (L. A. Joyce, J. E. Mitchell, and M. D. Skold, eds.). USDA For. Serv. Gen.
Tech. Rep. RM-203.

Siegel, S. 1956. Nonparametric statistics for behavioral sciences. Mc-Graw-Hill Book Co.,
New York, New York.

USDA. 1972. Final report: Conservation Reserve Program — summary of accomplishments
1956-1972. USDA-ASCS, Washington, D.C.

Wilson Bull, 108(4), 1996, pp. 771-775

FEMALE BUNTINGS FROM FIYBRIDIZING
POPULATIONS PREFER CONSPECIFIC MALES

Myron C. Baker

Abstract. I captured five female Indigo Buntings {Passerina cyanea) and 12 female
Lazuli Buntings {P. amoena) from a hybridizing population in Wyoming and tested them
in the laboratory for their sexual display responsiveness to male traits (live male and broad-
cast vocalizations) of the two species. The females responded with more copulation solici-
tation displays when exposed to conspecific male traits than when exposed to heterospecific
male traits. Consistent preference of female buntings for conspecific male characteristics in
these choice tests, together with previous results, suggests that hybrid pairs may form when
females are faced with a choice of mating heterospecifically or not at all. Received 14 Nov.
1995, accepted 20 April 1996.

Populations of Indigo Buntings {Passerina cyanea) and Lazuli Bunt-
ings {P. amoena) overlap and hybridize in the Great Plains of North
America (Rising 1983). Although hybrids are easily found in some pop-
ulations, mating is non-random with positive assortment being the general
rule among the pure phenotypes of the two species (Emlen et al. 1975;
Baker, unpubl. data). Non-random mating in the overlap populations
could have a number of proximate explanations. Mate preferences could
be established in allopatric populations and emigrants from these areas
into sympatric populations may retain their preferences when encounter-
ing heterospecifics, at least in initial encounters. A previous experiment
gave results consistent with this hypothesis (Baker and Baker 1990). On
the other hand, continuing exposure of an immigrant female to the visual
and vocal courtship displays of heterospecific males could perhaps alter
her preferences. An extenuating circumstance might be the availability of
male mates of the alternative phenotypes.

An earlier experiment (Baker 1994) exposed females from allopatry to
heterospecific males during the time they were coming into reproductive
condition upon photostimulation. This exposure was restricted to 25 days
and occurred in cages in the laboratory. Such exposure to displaying
males had no significant effect on female preference, as indicated by an
assay of copulation solicitation displays, for conspecific males and vo-
calizations. It is likely that females in sympatric populations experience
a longer and more intense level of exposure to heterospecific courtship
than was achieved in the laboratory study, and this prolonged exposure
might be effective in altering female preferences. If so, such altered pref-
erences of females could lead to the formation of observed hybrid pair-

Biology Dept., Colorado State Univ., Fort Collins, Colorado 80523.

771

772

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

ings. The present experiment was conducted to address the question of
whether Lazuli and Indigo Bunting females from a natural population
consisting of a mix of phenotypes have significant preferences for con-
specific male traits when given a choice of conspecific versus hetero-
specific males and vocalizations in a controlled experiment.

METHODS

Seventeen female buntings (12 Lazuli, 5 Indigo) were captured in a hybridizing population
in NE Wyoming. Captures were made 18—27 July 1989 of females with fully developed
brood patches, indicating that they were reproductive. All the subjects were captured during
the territorial period and all had both Lazuli and Indigo males in nearby territories. Some
hybrid males also defended territories in the neighborhood. None of the birds was banded,
however, so the identity of the mate was not certain.

The plumage characteristics of the females were evaluated by the criteria of Emlen et al.
(1975). This character index judges three areas of female plumage: throat/breast, wing bars,
and back/rump. A pure Indigo would score 0 in all three regions for a total score of 0, and
a pure Lazuli would score 1 in all regions for a total score of 3. Intermediates in any region
were given a score of 0.5. There was some subjectivity involved in applying the scoring
system, so I considered scores of 0 and 0.5 as Indigo and of 2.5 and 3 as Lazuli.

The females were housed in separate cages in a single room in the laboratory. They were
fed a seed mixture of canary grass and millet, turkey starter, vitamins, grit, and water.
Starting 30 July, the photoperiod was gradually reduced to 8:16 (L:D) by 9 October and
retained on this cycle until 22 January 1990 when the cycle was switched overnight to 16:
8 (L : D). On 22 Lebruary, the females each received one silastic implant ( 10 mm long, 1.47
mm inside diameter, Dow Corning) subcutaneously in the breast region. Each implant con-
tained 17-beta estradiol (Sigma Biochemical) and was sealed at both ends with silastic
adhesive (Dow Corning). This treatment is necessary to bring females into full reproductive
readiness in the laboratory.

Starting 5 Lebruary, each subject in her home cage was habituated to testing circumstances
by being placed in an experimental chamber containing an empty cage and a loudspeaker.
An observation booth one meter away was equipped with a video camera, one-way glass,
and a cassette recorder for playing stimulus tapes over the loudspeaker. Each female had
nine habituation sessions of 30 min at a rate of one session every other day.

In the experiment, a test session consisted of placing a caged female into the observation
chamber together with a stimulus male occupying the adjacent cage and playing vocaliza-
tions, matched with his species, from the loudspeaker located on the side of the male cage
away from the female. Previous studies indicated that both male plumage characteristics and
vocalizations play significant roles in the expression of female preference as indicated by
the copulation display assay. Pour Lazuli and two Indigo males were used as stimuli. All
were captured in allopatric populations and maintained together in acoustic isolation from
the females. The males did not vocalize during female testing, although they hopped around
in the cage. Pollowing 3 min of initial silence, the loudspeaker broadcast a sequence of 1
min of male songs at a rate of one song each 15 sec, 1 min of “tseep” calls (a courtship
call, Thompson and Rice 1970, Thompson 1976) at a rate of one tseep every 10 sec, 1 min
of male songs (different from those in the first min), 1 min of “tseep” calls, 1 min of male
songs (also different from the earlier songs), and a final minute of “tseep” calls. The “tseep”
call, similar in the two species, is an important component of courtship vocal behavior and
increa.ses female responsiveness in the laboratory assay (Baker and Baker 1988).

Prom 3-12 March, half the females were tested first with Lazuli males and vocalizations

Baker • MATE PREFERENCES IN BUNTINGS

773

r

>. 10

CO

Conspp 1 Conspp 2 Hetspp 1 Hetspp 2

Treatment

Fig. 1. Mean number of copulation solicitation displays (+1 SEM) elicited from female
buntings by conspecific versus heterospecific males and their recorded vocalizations. Hori-
zontal lines above histogram bars connect pairs of treatments that were significantly differ-
ent.

and then with Indigo males and vocalizations. The other half the females were tested with
the order of stimuli reversed. No subject was tested twice on the same day, and there was
at least 1-2 days between successive tests on a female. With completion of these first test
sessions, the stimulation of the subjects by conspecific stimuli was considered one treatment,
and the stimulation by heterospecific stimuli was considered a second treatment. A second
set of tests followed the first and again presented Lazuli and Indigo stimuli to each subject.
In this set, the stimulus male presented to each female differed from the one she had
experienced in the first set, and an entirely new group of songs was used for the vocal
stimuli. The same tseep calls were used, however, because of a paucity of good quality
recordings of this vocalization. With completion of this second set of tests, presentation of
conspecific stimuli constituted a third treatment and presentation of heterospecific stimuli a
fourth treatment. I used repeated measures ANOVA and Fisher's LSD for multiple com-
parisons (Winer 1971) for statistical analyses.

RESULTS

In the treatments involving exposure to conspecific males and vocali-
zations, the females exhibited a high level of responsiveness (Fig. 1). The

774

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

average number of copulation solicitation displays elicited by the first
conspecific treatment (Conspp-1) was 8.1 and that by the second conspe-
cific treatment (Conspp-2) was 9.0 (ns, P > 0.05, Fisher’s LSD). The two
treatments presenting heterospecific males and vocalizations elicited low-
er numbers of solicitation displays (Fig. 1). The average number of dis-
plays elicited by the first heterospecific treatment (hetspp- 1 ) was 4.0 and
by the second heterospecific treatment (hetspp-2) was 1.8 (ns, P > 0.05,
Fisher’s LSD).

In comparing conspecific to heterospecific treatments, the number of
solicitation displays elicited by both Conspp-1 and Conspp-2 were greater
than the number elicited by Hetspp-2, and the number elicited by
Conspp-2 was greater than that elicited by Hetspp- 1 (Fig. 1, all Ps <
0.05, Fisher’s LSD). The comparison between Conspp-1 and Hetspp- 1
did not give a significant difference.

Comparing responses of Indigo (N = 5) and Lazuli (N = 12) females
within each treatment revealed that there were no differences in display
production in three of the four treatments (Conspp-1, P = 0.36; Conspp-2,
P = 0.45; Hetspp- 1, P = 0.70; Mann-Whitney U-tests, two tailed, Siegel
1956). In the Hetspp-2 treatment, however. Indigo females averaged more
displays than Lazuli females {P — 0.01, Mann-Whitney U-test, two
tailed). In this treatment, only two of the 12 Lazuli females gave displays
whereas four of the five Indigo females gave displays.

DISCUSSION

The results suggest that prolonged exposure of females to the songs
and plumages of both Indigo and Lazuli bunting males in a natural breed-
ing association does not alter preference for conspecific traits. Although
the exact nature of previous experience of each female is unknown, it is
possible to outline the range of possible exposures the subjects had prior
to testing. These migrants arrived in the study area in late May and early
June. In the immediate neighborhoods of the subjects were territorial Laz-
uli, Indigo, and hybrid males. Given capture in late July, it is likely that
the females were exposed to males of all three types for a minimum of
5-8 weeks. Alternatively, any or all of the subjects could have hatched
in a mixed population and experienced one or more breeding seasons in
such a context. Thus, the experience the females had with males and songs
prior to testing occurred in a natural population and was probably more
intense and longer than was provided in the earlier laboratory experiment
that attempted to alter adult female preferences (Baker 1994). Although
the experimental females were relatively refractory to the potentially in-
fluential presence of alternative species’ male phenotypes, it is possible

Baker • MATE PREFERENCES IN BUNTINGS

775

that more extreme amounts of exposure could affect changes in female
preference.

The general inference that can be made from these results is that when
given a choice between conspecific and heterospecific males and vocali-
zations, female buntings prefer conspecific male traits. This implies fur-
ther that heterospecific pairings in natural populations may result from
females making the best of a bad job, choosing to mate heterospecifically
rather than not mate at all. A female with a territory retains the possible
option of extra-pair fertilization (Westneat 1990) with neighboring con-
specifics while having the benefit of a nest site, feeding area, and other
advantages of territorial residence. A similar pattern of female behavior
has been observed in Fairy-Wrens (Malurus splendens), (Brooker and
Rowley 1995), in which social pairing by a female appears to allow the
opportunity for matings with high quality neighboring males.

ACKNOWLEDGMENTS

Chris Goulart and Jean Boylan assisted with the fieldwork and with maintenance of the
captive birds. Capture was authorized by state and federal permits (Wyoming 1989-16,
Federal PRT-694924). Financial support was from the National Science Foundation (BNS-
87-06526). Helpful comments were received from David Westneat and an anonymous re-
viewer.

LITERATURE CITED

Baker, M. C. 1994. Does exposure to heterospecific males affect sexual preferences of
female buntings (Passerina)? Anim. Behav. 48:1349-1355.

AND A. E. M. Baker. 1988. Vocal and visual stimuli enabling copulation behavior

in female buntings. Beh. Ecol. Sociobiol. 23:105-108.

AND . 1990. Reproductive behavior of female buntings: isolation mecha-
nisms in a hybridizing pair of species. Evolution 44:332-338.

Brooker, M. and I. Rowley. 1995. The significance of territory size and quality in the
mating strategy of the Splendid Fairy-Wren. J. Anim. Ecol. 64:614-627.

Emlen, S. T, J. D. Rising, and W. L. Thompson. 1975. A behavioral and morphological
study of sympatry in the Indigo and Lazuli buntings of the Great Plains. Wilson Bull.
87:145-179.

Rising, J. D. 1983. The Great Plains hybrid zones. Current Ornithol. 1:131-157.

Siegel, S. 1956. Nonparametric statistics for the behavioral sciences. McGraw-Hill, New
York, New York.

Thompson, W. L. 1976. Vocalizations of the Lazuli Bunting. Condor 78:195-207.

AND J. O. Rice. 1970. Calls of the Indigo Bunting, Passerina cyanea. Z. Tierpsy-

chol. 27:35-46.

Westneat, D. F. 1990. Genetic parentage in the Indigo Bunting: a study using DNA fin-
gerprinting. Behav. Ecol. Sociobiol. 27:67-76.

Winer, B. J. 1971. Statistical principles in experimental design. McGraw-Hill, New York,
New York.

Wilson Bull., 108(4), 1996, pp. 776-782

SURVEYS OF PUERTO RICAN SCREECH-OWL
POPULATIONS IN LARGE-TRACT AND
FRAGMENTED FOREST HABITATS

Keith L. Pardieck,' J. Michael Meyers,^ and
Michelle Pagan^

Abstract. — We conducted road surveys of Puerto Rican Screech-Owls (Otus nudipes)
by playing conspecific vocalizations in secondary wet forest and fragmented secondary
moist forest in rural areas of eastern Puerto Rico. Six paired surveys were conducted bi-
weekly beginning in April. We recorded number of owl responses, cloud cover, wind speed,
moon phase, and number of cars passing during 5-min stops at 60 locations. Owls responded
in similar numbers {P > 0.05) in both habitat types. Also, we detected no correlation of the
number of owls with cloud cover, wind speed, moon phase, or passing cars. Received 29
Dec. 1995, accepted 15 May 1996.

The endemic Puerto Rican Screech-Owl {Otus nudipes) is common
throughout forested areas of mainland Puerto Rico but is thought to be
extirpated on adjacent islands such as Culebra and Vieques (Raffaele
1989). Although Puerto Rican Screech-Owl populations may have de-
clined in the early 1900s when many of the islands’ forests were cleared
(Wiley 1986a), Wiley (1986b) reports that the owl population recovered
as forest cover increased and as trees grew large enough to provide suit-
able nesting and roosting cavities.

Unfortunately, Puerto Rican Screech-Owl populations have seldom
been studied (Recher 1970, Snyder et al. 1987, Rivera-Milan 1995). Since
Eastern Screech-Owl {Otus asio) abundance is positively correlated with
amount of woodland habitat available (Nowicki 1974, Cink 1975, Smith
et al. 1987), we hypothesized that Puerto Rican Screech-Owls occupy
fragmented forests at lower densities, if at all. Our objective was to de-
termine the relative abundance of owls in fragmented and large forests
of eastern Puerto Rico.

METHODS

Study areas (each —2500 ha) were in secondary subtropical wet forest (large tract) of the
Luquillo Mountains at elevations from 90-455 m and in secondary subtropical moist forest

' National Biological Service, Patuxent Environmental Science Center, Puerto Rico Research Group, P.O.
Box N, Palmer, Puerto Rico 00721, Present address: National Biological Survey, Inventory and Moni-
toring, 12100 Beech Forest Rd,, Laurel, Maryland 20708,

2 National Biological Service, Patuxent Environmental Science Center, Puerto Rico Research Group, P.O,
Box N, Palmer, Puerto Rico 00721. Pre.sent Address: National Biological Service, D. B. Wamell School
of Forest Resources, The Univ. of Georgia. Athens, Georgia 30602-2152.

■’ Univ. of Puerto Rico, Dept, of Biology, Rio Piedras, Puerto Rico 00931. Present Address: 8803 Enfield
Ct. #24. Laurel, Maryland 20708.

776

Pardieck et al. • PUERTO RICAN SCREECH-OWLS

777

(fragmented) near Naguabo, Puerto Rico (18°18'N, 65°30'W), at elevations from 22-224 m
(Ewel and Whitmore 1973). The forests receive 233-392 cm of rainfall annually and have
an average temperature of 25°C (Brown et al. 1983).

Percent cover of general habitat types was visually estimated by two observers to the
nearest 10% within approximately 300 m of each owl survey point. Average percent cover
in large-tract forest was 80% intermediate forest, 15% young forest, 4% scrub (storm blow-
down areas), and 1% residential. Average percent cover at survey points in the fragmented
forest was 27% intermediate forest, 9% young forest, 19% scrub, 34% pasture, and 11%
residential.

To estimate relative owl populations, we conducted road surveys using vocalization play-
back to elicit owl responses. In each study area (12 km apart), we established three 3.6-km
road surveys with 10 points per survey site (points 0.4 km apart). Distance between survey
points was determined by research in the summer of 1990 and from home ranges determined
for two radio-marked Puerto Rican Screech-Owls (Gannon et al. 1993). Roads with heavy
traffic (>3 cars per 5-min point count) at night were avoided. Survey sites were selected
by lottery from suitable roads (no heavy traffic) near Naguabo (Highway 971, 972, and 974)
and in the Luquillo Mountains (Highway 988, 9966, and 191). Survey sites were paired
between study areas, i.e., paired points were sampled simultaneously throughout the study.

We conducted six surveys during the owls’ 1991 breeding season — five biweekly surveys
beginning in late April and one in July. Surveys began 30 min after sunset (about 19:20)
and ended about midnight. We randomly selected the survey order and rotated the order on
subsequent samplings so that each point was sampled twice during early (19:20-21:00),
mid- (21:00-22:30), and late evening (22:30-24:00).

At each point, we conducted a 5-min survey consisting of visual and aural observations.
Playback overlapped the first 2 min of observation and was followed by 3 min of silence.
The first minute of playback consisted of 30 sec of owl trills followed by 30 sec of “wild”
trills or cackles (van der Weyden 1974). This sequence was repeated during the second
minute after turning the speaker to the opposite direction and perpendicular to the edge of
the road. We used Marantz® (model PMD 430) and Sony® (model TCM-5000EV) cassette
recorders with amplified external speakers (Realistic® Minimus-0.6, 2 W) to broadcast owl
calls. (Use of brand names does not constitute government endorsement.) Speaker volumes
were equal and audible to us to 160 m. To test observer variability, two persons recorded
owl responses independently and simultaneously at each point in the large-tract forest. Both
observers were equally trained and experienced (Kepler and Scott 1981). At each point we
recorded ( 1 ) number of initial owl responses within 5, 1-min periods, (2) cloud cover (nearest
10%), (3) wind speed (Beaufort scale), (4) moon phase, and (5) number of cars passing the
point in 5 min.

We analyzed the data with r-tests (SAS Institute, Inc. 1988) that compared the mean
differences (d) of variables between paired points. Differences were considered significant
at P < 0.05. Pearson correlation analysis (SAS Institute, Inc. 1988) was used to determine
associations between number of owl responses with wind speed, moon phase, cloud cover,
and number of cars. Weather conditions were different between the study areas during the
surveys. More cloud cover (d — 17%, P — 0.0001) and higher winds (Beaufort scale, d —
1.7, p = 0.0001) were present during surveys in large-tract forest. However, more cars {d
= 0.8, P = 0.0001) were present during surveys in fragmented forest. Because of these
differences the association of weather and number of cars with number of owl responses
was analyzed separately for each study area. Data were normalized using the square root of
X (x = cloud cover and wind speed) and the square root of fy + 0.5] (y = number of owls,
moon phase, and number of cars) transformations.

778

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

RESULTS

Paired mean difference {d) of owl responses by study areas was not
different from 0 (P = 0.53) among survey dates; therefore, data for all
dates were pooled. No differences in owl responses were found between
study areas (P = 0.31). In the large-tract forest, a total of 200 owl re-
sponses {x = 1.10 per point per survey) was found compared to 173 owl
responses (x = 0.96 per point per survey) in fragmented forest (Table 1).
Observers conducting simultaneous point surveys for owls in large tract
forest did not detect owls in different numbers (overall x for observer 1
= 1.11 owls/point and x for observer 2 = 1.15 owls/point, P = 0.60).
Cloud cover, wind speed, and number of cars were not associated with
number of vocal responses of owls at either study area (r ranged from
—0.054 to 0.12 and P from 0.12 to 0.97). In large-tract forest, the number
of vocal responses of owls was weakly correlated with moon phase (r =
0.15, P = 0.04). No correlation was found between moon phase and
number of vocal responses by owls in fragmented forest (r = 0.08, P =
0.22). Of a total of 373 aural and visual responses, 18.5% were recorded
in the first minute, 4.8% in the second, 31.6% in the third, 22.0% in the
fourth, and 23.1% in the fifth. Only, 4.8% of total contacts were visual,
all occurring after the first survey minute.

DISCUSSION

Puerto Rican Screech-Owls responded similarly in fragmented habitat
(36% forest cover) and large tract habitat (95% forest cover) which may
indicate similar owl densities. This is contrary to studies on Eastern
Screech-Owls in Michigan (Nowicki 1974) and Kansas (Cink 1975)
which indicated that abundance was positively associated with forested
habitat. Similar numbers of owls in fragmented and large-tract forests
suggests that restoration of the Puerto Rican Screech-Owl may be possible
in former ranges where forest habitat is fragmented (e.g., Vieques Island)
but where nesting cavities (artificial or natural) and food resources are
adequate.

Traditionally, the Vieques owl population has been listed as the O. n.
newtoni subspecies (Bond 1956). However, lack of a specimen from Vie-
ques and the paucity of information, in general, on the Puerto Rican
Screech-Owl indicate that owls on this island were part of the geograph-
ically closer and more abundant race, O. n. nudipes (R. Banks, pers.
commun.; J. Marshall, pers. commun.). In addition, there is some uncer-
tainty as to the existence of two separate owl races (R. Banks, pers.
commun.; Biaggi 1970). No attempts to introduce owls from Puerto Rico
to the Virgin Islands should be undertaken while this species taxonomic
status is in question.

Pardieck et al. • PUERTO RICAN SCREECH-OWLS

779

Table 1

Number of Puerto Rican Screech-Owls and Mean Difference (d) between
Unfragmented Tropical Wet Forest and Fragmented Tropical Moist Forest

Total number of owls“

Unfragmented

Fragmented

forest

forest

d

7

3

0.67'’

5

3

0.33

11

7

0.67

6

7

-0.17

9

7

0.50

4

11

-1.17

11

1

1.67

5

4

-0.17

9

4

0.83

2

8

-1.00

8

5

0.33

11

7

0.67

10

7

0.50

6

5

0.17

3

3

0.00

10

2

1.33

6

8

-0.33

11

0

1.83

4

3

0.17

8

2

1.00

5

4

0.17

3

4

-0.17

2

2

0.00

10

17

-1.17

5

9

-0.67

5

3

0.33

12

4

1.33

1

8

-1.17

10

1 1

-0.17

1

14

-2.17

Total 200

173

* Total number for six surveys of 5 min each.

’’ Mean difference per p>oint per survey; paired ?-tesl of mean difference (owls counted for point by survey for unfrag-
mented forest minus owls counted for point by survey for fragmented forest) of transformed data is / = 1.31. df = 29. P
= 0.20.

Future research, prior to screech-owl restoration efforts, should include
surveys of former and current ranges to assess (1) the status of extant
owl populations, (2) the availability of nesting and roosting cavities, and
(3) the availability of food resources. Since these owls readily use nest

780

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

boxes. Otherwise suitable habitat can be supplemented with artificial cav-
ities (Wiley 1985, 1986a). Our data further suggests that Puerto Rican
Screech-Owls may be able to exist in moist and wet areas with fragmented
forest cover of approximately 36%. But the extent to which Puerto Rican
Screech-Owls use fragmented forest must also be determined. Is frag-
mented forest preferred owl habitat or does it serve only as a population
sink for individuals dispersing from large-tract areas?

Smith et al. (1987) and Gerhardt (1991) reported that Eastern Screech-
Owl and Mottled Owl (Ciccaba virgata) responses were negatively as-
sociated with wind speed, whereas Carpenter (1987) suggested that sub-
zero temperatures had inhibitory effect on Eastern Screech-Owl calling.
Unlike Eastern Screech-Owls (Smith et al. 1987, Ritchison et al. 1988,
Carpenter 1987) and Mottled Owls (Gerhardt 1991), Puerto Rican
Screech-Owl responses may be weakly associated with moonlight. This
weak association was probably caused by the absence of full and new
moon phases during our sampling periods. In Puerto Rico, the lack of
association of owl responses to all weather conditions discussed above,
except moonlight, may be related to the mild temperature fluctuations of
the tropical environment.

Puerto Rican Screech-Owls responded well to conspecific vocalization
playback, with contacts occurring throughout the 5-min period. However,
the lower numbers of responses in the first 2 min suggests that playback
may interfere with the observers aural detection ability. In addition, after
a 5-min survey, the accumulated number of initial owl responses had not
stabilized. Therefore, five minutes per point may not be long enough to
provide data for quantitative estimates of the owl population. Some stud-
ies suggest at least a 10-min sample period when surveying Eastern
Screech-Owls and other raptors, especially when trying to determine pop-
ulation densities (Mosher et al. 1990, De Geus and Bowles 1991).

ACKNOWLEDGMENTS

We thank P. Pagan and N. Lopez for their kind support and use of their vehicle. We are
also grateful for the helpful comments provided by reviewers W. J. Arendt, C. Henny, D.
G. Smith, R. B. Waide, and J. W. Wiley. We thank the U.S. Dept, of Agriculture Lorest
Service for providing access to study areas in the Caribbean National Lorest. In addition,
we are indebted to the U.S. Dept, of Interior, National Biological Service, for the use of
field and office equipment throughout the study.

LITERATURE CITED

Biaggi, V. 1970. Las aves de Puerto Rico. Editorial Universitaria, Univ. Puerto Rico, Ma-
nuel Pareja Publ., Barcelona, Spain.

Bond, J. 1956. Check-list of birds of the West Indies. Acad. Nat. Sci. Phil., Philadelphia,
Pennsylvania.

Pardieck et al. • PUERTO RICAN SCREECH-OWLS

781

Brown, S., A. E. Lugo, S. Silander, and L. Liegel. 1983. Research history and oppor-
tunities in the Luquillo Experimental Forest. USDA For. Serv. Gen. Tech. Rep. SO-44.

Carpenter, T. W. 1987. Effects of environmental variables on responses of Eastern Screech-
Owl to playback. Pp. 277-280 in Biology and conservation of northern forest owls (R.

W. Nero, R. J. Clark, R. J. Knapton, and R. H. Hamre, eds.). USDA For. Serv. Gen.
Tech. Rep. RM-142.

CiNK, C. L. 1975. Population densities of Screech Owls in northeastern Kansas. Kansas
Ornithol. Soc. 26:13-16.

De Geus, D. W. and j. B. Bowles. 1991. Relative abundance of Eastern Screech-Owls in
a south-central Iowa township. J. Iowa Acad. Sci. 98:91-92.

Ewel, j. j. and j. L. Whitmore. 1973. The ecological life zones of Puerto Rico and the
U.S. Virgin Islands. USDA For. Serv. Res. Pap. ITF-18. Inst. Trop. For., Rio Piedras,
Puerto Rico.

Gannon, M. R., K. Pardieck, M. R. Willig, and R. B. Waide. 1993. Movement and home
range of the Puerto Rican Screech-Owl (Otus midipes) in the Luquillo Experimental
Forest. Carib. J. Sci. 29:174-178.

Gerhardt, R. P. 1991. Response of Mottled Owls to broadcast of conspecific calls. J. Field
Ornithol. 62:239-244.

Kepler, C. B. and J. M. Scott. 1981. Reducing bird count variability by training observers.
Pp. 68-75 in Estimating numbers of terrestrial birds (C. J. Ralph and J. M. Scott, eds.).
Stud. Avian Biol. No. 6.

Mosher, J. A., M. R. Fuller, and M. Kopeny. 1990. Surveying woodland raptors by
broadcast of conspecific vocalizations. J. Field Ornithol. 61:453-461.

Nowicki, T. 1974. A census of Screech Owls {Otus asio) using tape recorded calls. Jack-
Pine Warbler 52:98-101.

Raffaele, H. a. 1989. A guide to the birds of Puerto Rico and the Virgin Islands. Princeton
Univ. Press, Princeton, New Jersey.

Recher, H. F. 1970. Population density and seasonal changes of the avifauna in a tropical
forest before and after gamma irradiation. Pp. E69-E86 in A tropical rain forest: a
study of irradiation and ecology at El Verde, Puerto Rico (H. T. Odum and R. F. Pigeon,
eds.). Natl. Inf. Tech. Serv., Springfield, Virginia.

Rivera-Milan, F. F. 1995. Distribution and abundance of raptors in Puerto Rico. Wilson
Bull. 107:452-462.

Ritchison, G., P. M. Cavanagh, J. R. Belthoff, and E. J. Sparks. 1988. The singing
behavior of Eastern Screech-Owls: seasonal timing and response to playback of con-
specific song. Condor 90:648-652.

SAS Institute, Inc. 1988. SAS/STAT® User’s Guide, release 6.03 edition. SAS® Institute,
Inc., Cary, North Carolina.

Smith, D. G., A. Devine, and D. Walsh. 1987. Censusing screech owls in southern Con-
necticut. Pp. 255-267 in Biology and conservation of northern forest owls (R. W. Nero,
R. J. Clark, R. J. Knapton, and R. H. Hamre, eds.). USDA For. Serv. Gen. Tech. Rep.
RM-142.

Snyder, N. F. R., J. W. Wiley, and C. B. Kepler. 1987. The Parrots of Luquillo: natural
history and conservation of the Puerto Rican Parrot. Western Foundation Vertebrate
Zoology, Los Angeles, California.

VAN DER Weyden, W. J. 1974. Vocal affinities of the Puerto Rican and Vermiculated Screech
Owls (Otus nudipes and Otus gucitemalae). Ibis 1 16:369-372.

Wiley, J. W. 1985. Bird conservation in the United States Caribbean. Pp. 107-160 in Bird
conservation 2 (S. A. Temple, ed.). The Univ. of Wisconsin Press, Madison, Wisconsin.

782

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

— . 1986a. Status and conservation of forest raptors in the West Indies. Birds Prey
Bull. 3:57-70.

— . 1986b. Habitat change and its effects on Puerto Rican raptors. Birds of Prey Bull.
3:51-56.

Wilson Bull., 108(4), 1996, pp. 783-796

THE USE OF COASTAL AGRICULTURAL FIELDS IN
VIRGINIA AS FORAGING HABITAT BY SHOREBIRDS

Stephen C. Rottenborn

Abstract. I studied temporal abundance patterns and use of cover types by shorebirds
foraging in coastal croplands on the eastern shore of Virginia from March 1991 through
February 1992. A total of 21,254 shorebirds of 21 species was observed foraging in agri-
cultural croplands. Shorebird abundance reached a peak during spring migration and was
lower during fall and winter. Shorebird species richness was highest in fall and lower in
spring and winter. Some species appeared to use fields primarily as alternate foraging habitat
when preferred intertidal habitats were covered by high tides, whereas other species foraged
in fields regardless of tidal height. Most species showed highly significant positive associ-
ations with plowed fields (bare earth) and negative associations with herbaceous vegetation
>10 cm tall. Associations with vegetated cover of <10 cm varied seasonally, tending to be
positive in spring and negative (or with no association) in fall and winter for most species.
Because most shorebirds foraged on the cover type providing the least cover from predators
(plowed fields), the observed cover associations may reflect differences in predator detection
or foraging efficiency among the cover types. Received 20 Nov. 1995, accepted 13 April
1996.

Most research on habitat use and foraging strategies of shorebirds has
focused on the use of intertidal habitats, where they may choose different
foraging sites and strategies depending on tidal conditions (cf Burger et
al. 1977, Connors et al. 1981). When preferred intertidal habitats are
inundated by high tides, shorebirds may forage temporarily in alternate
areas, including agricultural fields (Atkinson 1976, Gerstenberg 1979,
Page et al. 1979). Previous studies of the use of agricultural croplands by
shorebirds have dealt primarily with species that prefer upland habitats
and rarely move to intertidal areas (Fuller and Youngman 1979, Barnard
and Thompson 1985, Milsom et al. 1985) or were conducted in inland
regions where tidal regime had no influence on shorebird distribution and
behavior (Ohmart et al. 1985). In California, a number of shorebird spe-
cies move from intertidal areas to pastures during high tides (Gerstenberg
1979, Page et al. 1979), and some of the habitat features associated with
the birds’ use of coastal pastures have been described (Colwell and Dodd
1995). Despite the potential importance of cropland as alternate foraging
habitat for shorebirds (Ohmart et al. 1985), the use of coastal cropland
by foraging shorebirds has not been described. The objectives of this
study were (1) to determine the species composition and temporal abun-
dance patterns of shorebirds foraging in coastal cropland on the Eastern

Center for Conservation Biology, Dept, of Biological Sciences, Stanford Univ., Stanford, California
94305-5020.

783

784

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Shore of Virginia, and (2) to determine patterns of cover use by these
shorebirds.

STUDY AREA AND METHODS

I conducted this study on the southern tip of the Eastern Shore of Virginia, within an
area of approximately of 170 km^. The study area is bounded to the east by extensive tidal
salt marsh, mud flats, and barrier islands and to the west by narrow sandy beaches (with
several tidal creeks) along Chesapeake Bay. Large numbers of shorebirds forage in these
tidal areas, especially at low tide when beaches and mudflats are exposed. Previous obser-
vations indicated that individuals of some shorebird species moved from these intertidal
habitats to agricultural croplands to forage during high tide.

In order to determine the relationship between tidal height and the abundance of certain
shorebird species in agricultural fields, I conducted four surveys of a small number of fields
in the study area (6-8 fields/survey). The days on which these censuses were conducted,
29 May, 12 and 31 August, and 2 November 1991, were chosen so that the timing of high
and low tides varied among surveys. Such variation in the timing of high and low tides
allowed me to distinguish possible relationships between shorebird abundance and time of
day from relationships between shorebird abundance and tidal height. During these surveys,
I visited each field at one- to two-h intervals between dawn and dusk, recording the number
of individuals of each species in each field during each round of censuses. All fields were
visited the same number of times on a given date, but the number of census rounds varied
from seven to nine among different survey dates. I divided the species observed in fields
into two groups based on my previous experience with shorebirds in intertidal habitats
around the study area and life history information (Stout 1967, Hayman et al. 1986): field
specialists, thought to forage primarily in fields and rarely visiting intertidal habitats, and
field exploiters, species that regularly forage in intertidal habitats during low tide but oc-
casionally forage in fields during high tide. Eor each field, the abundance of shorebirds in
each of these groups was determined for the censuses conducted closest to the times cor-
responding to high and low tide at Cape Charles, Virginia (U.S. Department of Commerce
1990). Paired comparisons r-tests were used to test the null hypothesis that the abundance
of shorebirds in these fields did not differ between high tide and low tide (Sokal and Rohlf
1981).

In order to determine seasonal patterns of shorebird abundance and the use of various
cover types by shorebirds foraging in coastal croplands, I conducted thirty censuses of
croplands (comprising 4598.8 ha) in the study area from March 1991 through Eebruary
1992. For the purpose of determining the relative availability of cover types, I defined a
field as a continuous plot with homogeneous cover of bare earth or herbaceous vegetation
of the same height class (<10 cm or >10 cm) not divided by a paved road or by woody
vegetation. Because high and low tides were separated by approximately 6.0-6. 5 h, I began
each census 4 h before high tide so that all fields would be surveyed during the highest 2/3
of the tidal cycle. Each census lasted approximately 8 h. To randomize the tide height during
my observation of any particular field, I began each census at a location chosen randomly
from a pool of 20 points along the census route. Censuses were not conducted if visibility
was hampered by fog or precipitation.

During each census, I drove along the census route, stopping at each field to scan for
shorebirds. I counted shorebirds from one or more points at the edge of each field using
8.5 X binoculars. For larger fields and for fields with taller vegetation, I scanned from several
points along the edge of each field, using a 22X spotting scope as necessary. Scanning along
rows of vegetation and listening for the birds’ calls (given frequently by foraging flocks)

Rottenborn • AGRICULTURAL FIELD USE BY SHOREBIRDS

785

aided in the detection ot shorebirds. Eor large or widely dispersed flocks, I made replicate
counts ot shorebirds and recorded the mean of the original and replicate counts. Most of
the shore^rds observed on croplands were actively foraging, with only a small proportion
roosting. Observations made during the four dawn-to-dusk surveys and at other times sug-
gested that very few shorebirds, if any, used these fields for roosting without spending some
time foraging. Therefore, all shorebirds recorded during censuses were assumed to be for-
aging in the fields.

I was able to corroborate the accuracy of counts for cover types of <10 cm vegetation
and plowed fields by locating three fields with each of these cover types that contained large
flocks of foraging shorebirds, counting the birds from the edge of each field, and then
walking through the fields to flush and count every bird. In all six cases, the error in these
replicate counts was less than two percent. I was unable to corroborate directly my counts
for >10 cm cover, but on one occasion a disturbance in a field of tall vegetation flushed
the shorebirds within the field, confirming my previous count.

During each census, I recorded the type and height of cover in each field. I recognized
three cover types in these croplands: plowed (bare earth with no vegetation), herbaceous
vegetation <10 cm tall, and herbaceous vegetation >10 cm tall. No distinction was made
between row crops (which included cotton, soybeans, potatoes, string beans, and several
other crops) and mat crops (mostly cereals). Eields with weedy herbaceous vegetation al-
lowed to grow in the interval between the harvest of one crop and seeding of another were
also included in this study, although I excluded fields that lay fallow for the entire study
period. All crop types reached heights exceeding 10 cm, so vegetation height (<10 cm or
>10 cm) was not dependent on crop type. I did not include any pastureland in this study,
as very little was present within the study area.

The area of each field was measured with a planimeter from field-checked 7.5-min to-
pographic maps. For each census, I summed the areas of all fields with a particular cover
type to determine the total availability of each cover type in the study area. I then deter-
mined, for each census, the number of individuals of each shorebird species expected to
occur in fields with each cover type if shorebird distribution over the study area were
independent of cover type. These numbers were the products of the total number of observed
individuals of each species and the proportion of fields having each of the three cover types.

I summed the numbers of shorebirds observed and expected on each cover type over all
censuses in each season in order to determine the total number of Individuals of each species
that were observed and expected on each cover type by season. In pooling these numbers
by season, I assumed that patterns of cover use did not differ significantly among different
censuses within the same season.

For the purposes of this study, I delimited seasons a posteriori by examining abundance
patterns of species known to occur in coastal Virginia as summer or winter residents or as
northbound (spring) or southbound (fall) migrants (Kain 1987). I defined summer as the
period from early June until mid-July, during which the only species observed in the study
area was a local breeder. Fall lasted from the arrival of the first southbound migrants in late
July until the departure of the last southbound transients (i.e., species that do not overwinter
in Virginia) from the study area in mid-November. Winter extended from mid-November
until late March. Spring began with the first arrival of northbound migrants in early April
and ended in late May. Because I observed only two individuals of one locally breeding
species during summer, analyses presented here are restricted to fall, winter, and spring.
Sampling intensity differed among seasons due to differences in season length; four censuses
were conducted during summer, 10 during fall, 1 1 during winter, and five during spring.

For the species represented by at least 30 individuals during a given season, I u.sed a
test to test the null hypothesis that shorebird distribution was independent of cover type. I

786

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

performed these tests separately for each season. In a few cases, the expected number of
individuals on both plowed and <10 cm vegetation was less than five; in such cases, the
numbers of shorebirds observed and expected were pooled for these cover types, allowing
for a valid test (Sokal and Rohlf 1981). Each total value was the sum of three “component
X"” values, one for each of the three cover types. If the total x^ exceeded the critical value
(rejecting the null hypothesis), then each “component x^” was examined individually to
determine which of the shorebird-cover associations were significant. If a “component x^”
for a single cover type was so high that it exceeded the critical value required for
significance for the entire test, then the association between the species and the cover type
in question was deemed significant. This method provided a conservative approach to as-
sessing individual species-cover associations. The nature of each significant association was
identified by determining whether the number of individuals observed on a cover type was
greater than or less than the number expected on that cover type (indicating a positive or
negative association, respectively).

Standard y^ tests assume independence of individuals, an assumption that may be violated
in species, such as shorebirds, that frequently move in flocks. In this study, I was rarely
able to determine the number of independent groups of shorebirds that selected a field for
foraging. However, to treat the entire shorebird assemblage in each field as a single unit
composed of interdependent individuals may unnecessarily reduce the number of observa-
tions available for analysis if, indeed, the shorebirds individually display their own cover
preferences. In order to determine whether cover use may have been affected by social
interactions, I examined the results of the x^ tests a posteriori. If flocking, rather than se-
lection of a desired cover type, had been the primary influence on the birds’ choice of fields,

1 would have expected a preponderance of insignificant, weak, or inconsistent shorebird-
cover associations. However, the consistent, strong cover associations observed in this study
indicate that social interactions did not mask actual patterns of cover preference or avoid-
ance, supporting the validity of the x^ tests as they were conducted. I conducted an analysis
of variance (in SYSTAT, Wilkinson 1990) to test the null hypotheses that mean species
richness and shorebird abundance per census did not differ among the seasons.

RESULTS

Relationship between shorebird abundance and tidal height. — During
the four dawn-to-dusk surveys, those species thought to prefer foraging
in intertidal habitats (field exploiters) were most abundant on croplands
during the higher portion of the tidal cycle and were scarce or absent
during the lower portion (Fig. 1). During these surveys, such species
included Semipalmated Plover, Black-bellied Plover, Whimbrel, Willet,
Ruddy Turnstone, Semipalmated Sandpiper, Least Sandpiper, and Dunlin
(scientific names in Table 1). Abundance of these shorebirds in fields at
high tide and low was significantly higher during censuses conducted near
high tide on 29 May {t = 2.670, df = 5, P < 0.05), 12 August {t =
3.940, df = 6, P < 0.01), 31 August {t = 4.255, df = 7, P < 0.01), and

2 November (r = 2.449, df = 6, P < 0.05). According to Fig. 1, relatively
few field exploiters were present on censuses conducted more than two
hours preceding or following high tide, although peak abundance did not
always coincide with the census conducted nearest high tide (e.g., 29
May). Peak shorebird abundance occurred at different times during the

Rottenborn • AGRICULTURAL FIELD USE BY SHOREBIRDS

787

2 NOVEMBER

Fig. 1. Plots of shorebird abundance vs time of day. Solid circles represent individuals
of “field exploiter” species, while open circles represent “field specialist” species (see
“Methods” for descriptions of these two groups). Labels on the lower axis indicate hours
after midnight. Solid and open arrows along the lower axis signify the times of high and
low tides, respectively.

different surveys, indicating that these dynamics were related to tidal
height and not time of day.

In contrast, the abundance of species thought to forage almost exclu-
sively in fields (field specialists), including Killdeer, American Golden-

Table 1

Abundance of Shorebirds Foraging in Agricultural Fields by Species and Season

788

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

C

H o

f3

CN

On

o

m

o

00

7t

r--

1

2

1

4

4

4

1

1

1

1

"o

3

00

CN

r^i

cn

ON

NO

NO

in

<n

m (n —

"O

ON

On

—

O

■G

CN

6

Z

>

'■5

NO

IT) O O
00 O
(N r-

— . CM

O <N O O O O

r^OOOOOOOOO —

o

o

CM

m On Gv ON 00 m O
On sD in in ON
O ON CN 00
(N —

(n Tj* — (N — rN
in m <n (n —

o —
in
in
in

TtCN(N — (NOOOOO — O

NO in CN o cn no in

nd CN 00 in —

m CN

OOOOCNOOO — o

o —
o
in

NOininONNO^’^cncnmcN^ — ^OpppOOO

t^cNONTtcN — ddddoooddddoooo

Om-^OOOOONONONDONNO

•rfTi*-^fncNm»— — ^ —

o

o

CN

in

CN

CN

O

m

5

S:

i 2

S.

u

> <3

g

5

u 2

« g

O ^

5 -c
is ^

5 ^

S. 2

•2 5

:g

cs
U

Cu C

■o is

(U '3;

c

^ =
3

ca Q

0, g

•o §
u

M U

D

Cl (u

E 2

^ 2

D

Q. ^
O

a s:
TD o
c ■

c3
00

.1^

C

:2

"5

u

-S

"5

.g.

>1

g

o

Cj

^ O

S

-S ^

s;

o

-C)

s

'>2

53

oi -2
s:

S'

2 s CL

o

"S Q

«

« u

E =

IS

a. ^

E o

u s:

c/5 1/3

JJ Cl
D. E

-O Cl
C O
w C:

03 ^

U 4-«

O ^
o —
<0 >

0, >

c

2

o

O

c

03

O

'u

(U

E

<

>1

"3 fe
U .9-
^ a.

u *0
(U c
a.

73 73
C 0)

F ^

.2 ^
R ^

r;

o

s

•2

"o

R

CQ h.

u CA
O OD
CL O

£P S

R

&

'i)

cs

-c

Cl

L5

.g

s

5U

S C!

•S 'c:

K CS

:g

Cj

o

>1

tS

■s

I

O

00

R

•S

"S

00

o

00

R

R

CL

73

C

03

cn

73

C

03

(U
yi

O. o

D J

^ §

= ^

>H (U
L-

L. rj

^ E

S3 .E

P x:

6 ^

o

u CL

Q. C

S3

C C/5 «
C^ >' =

„ S3

^ 53 _

S3 LX T3

1) O
J C/5 Oi

w

r o

(U ^
Q. <D

’S. .E-

73 C

E' cn

CO

00 c

- i

2 E

S3 o

QQ U

Rottenborn • AGRICULTURAL FIELD USE BY SHOREBIRDS

789

Fig. 2. Number of individual shorebirds foraging in agricultural fields during each cen-
sus (Mar. 1991 -Feb. 1992).

Plover, Upland Sandpiper, and Buff-breasted Sandpiper, fluctuated little
during each of the four dawn-to-dusk surveys (Fig. 1). There were no
significant differences in the abundance of these species in fields between
high and low tides on 12 August (t - 1.016, df = 6, P > 0.05), 31
August (t = 0.429, df = 7, P > 0.05), and 2 November (r = 1.000, df
= 6, P > 0.05). None of these species was observed during the 29 May
survey.

Species composition and seasonal abundance patterns. — During the
study of seasonal abundance patterns and cover use, I recorded a total of
21,254 individuals of 21 species of shorebirds foraging on croplands (Ta-
ble 1). Shorebirds were observed foraging in agricultural fields on 24 of
30 censuses; no shorebirds were observed on two censuses in June and
one each in April, July, January, and February. Only two individuals of
one species (Willet) were observed during the four censuses in summer.
Total shorebird species richness reached a peak of 18 species in fall,
compared to 10 in spring and six in winter. This pattern also held for the
mean number of species/census, being highest in fall {x = 8.4 ± 6.9
[SD]), intermediate in spring {x = 6.2 ± 2.7), and lowest in winter {x =
2.0 ± 3.0). In contrast, mean abundance/census was highest during spring
(x = 2300.2 ± 2320.8 individuals/census) and lower in fall (x ^ 555.1
± 536.1) and winter (x = 381.8 ± 663.7). ANOVAs confirmed that mean
species richness/census (P = 24.5, P < 0.001) and mean shorebird abun-
dance/census (F = 4.83, P < 0.02) did differ among seasons. In addition.

790 THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

CENSUS DATE

Eig. 3. Number of shorebird species foraging in agricultural fields during each census
(Mar. 1991-Eeb. 1992).

there was substantial variation in shorebird species richness and abun-
dance within seasons as well.

The four species which were recorded on the most censuses (Black-
bellied Plover, Killdeer, Semipalmated Plover, and Dunlin) were also the
most abundant species, accounting for 94.3% of all the individuals ob-
served. Most species were represented by relatively few individuals. Only
1 1 species were observed on more than four censuses, and only six spe-
cies were represented by more than 90 individuals.

Whereas most species foraged in agricultural fields in only one or two
seasons. Black-bellied Plover, Killdeer, Dunlin, and Short-billed Dow-
itcher were present during spring, fall, and winter. There were no species
recorded year-round, as the only species present during summer (Willet)
was not recorded in fall or winter. Of the remaining eight species repre-
sented by at least 30 individuals, four occurred strictly as fall transients
(Pectoral Sandpiper, American Golden-Plover, Buff-breasted Sandpiper,
and Upland Sandpiper), three were present only as spring and fall tran-
sients (Semipalmated Plover, Semipalmated Sandpiper, and Ruddy Turn-
stone), and one was present only during fall and winter (Western Sand-
piper).

Because few of the fields were irrigated and rainfall was generally
lower than average during the study period (U.S. Dept, of Commerce
1991 ), standing water was rarely present in the fields. As a result, species
that require open water, such as Greater Yellowlegs, Lesser Yellowlegs,

Rottenborn • AGRICULTURAL FIELD USE BY SHOREBIRDS

791

and Solitary Sandpiper, were poorly represented on surveys. All other
shorebirds usually foraged on relatively dry substrates, quite different
from the saturated intertidal habitats with which the “field exploiter”
species are more commonly associated.

Cover use. — A total of 19 species foraged on plowed fields, 17 species
on fields with <10 cm cover, and seven species on fields with >10 cm
cover. Although the number of species recorded on plowed fields was
only slightly higher than the number on fields with <10 cm cover, far
more individuals were observed foraging on plowed fields than on the
other cover types. A total of 17,714 shorebirds, or 83.3% of all birds
recorded during the study, foraged on plowed fields. In contrast, 3432
individuals (16.2% of total) foraged on fields with <10 cm vegetation
and 108 birds (0.5% of total) on fields with >10 cm vegetation.

Of the species sufficiently numerous for analysis, all consistently
showed highly significant positive assocations with plowed fields and
negative assocations with fields with >10 cm cover except for Willet in
spring and Upland Sandpiper in fall (Table 2). No species seemed actively
to avoid plowed fields, and none was shown to prefer fields with >10
cm cover. The data for fields with <10 cm cover suggest that use of this
cover type varied not only among species but also among seasons within
a species. The proportion of species analyzed for cover use that showed
positive associations with fields of <10 cm cover decreased from 0.71 in
spring to 0.09 in fall and to 0.00 in winter. Conversely, the proportion of
species negatively associated with fields of <10 cm cover increased from
0.14 in spring to 0.45 in fall and 1.00 in winter. For no species was the
association with fields of <10 cm cover consistent among all seasons.

DISCUSSION

Agricultural fields on Virginia’s Eastern Shore seem to be important
foraging areas for migrating and overwintering shorebirds. A few species,
such as Killdeer, American Golden-Plover, Upland Sandpiper, and Buff-
breasted Sandpiper, are known to prefer such habitats to intertidal areas
(Stout 1967, Hayman et al. 1986). These field specialists were never seen
moving from fields to intertidal areas, and their numbers did not seem to
vary with tidal height. Most of the species observed in this study, how-
ever, are known to prefer intertidal habitats when they are available. These
field exploiters use fields as alternate foraging sites when mudflats and
beaches are inundated by high water (Goss-Custard 1969, Page and Whi-
tacre 1975, Gerstenberg 1979, Page et al. 1979).

Data from the four dawn-to-dusk surveys indicated that the abundance
of field exploiters on cropland was high only within approximately two
hours of high tide. Therefore, the surveys conducted to determine sea-

792

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Table 2

Cover Associations of Shorebirds Loraging in Agricultural Lields

Total No. observed/expected on each cover type"

ob-

Species served Plowed <10 cm >10 cm

Spring

Black-bellied Plover

4664

3271/1323.1

1383/

415.7

-b-b

10/2925.2 --

Semipalmated Plover

2822

2439/

694.2

-b-b

233/

209.4

ns

0/1918.4 --

Willet

68

19/

16.8

ns

27/

7.9

-b-b

22/

43.3 -

Ruddy Turnstone

51

38/

10.0

-b-b

13/

4.7

-b +

0/

36.3 --

Semipalmated Sandpiper

502

483/

108.7

-b-b

19/

43.2

—

0/

350.1 --

Dunlin

3252

2596/1150.3

-b-b

656/

313.2

-b-b

0/1788.5 --

Short-billed Dowitcher

138

108/

45.1

-b-b

30/

12.1

-b-b

0/

80.8 —

Pall

Black-bellied Plover

2033

1970/

527.2

-b-b

61/

388.6

—

2/1 1 17.2 --

American Golden-Plover

57

50/

8.7

-b-b

4/

4.3

ns

3/

44.0 --

Semipalmated Plover

1299

1122/

246.1

-b-b

132/

97.4

-b

45/

955.2 —

Killdeer

869

829/

242.8

-b-b

38/

177.7

—

2/

448.5 --

Upland Sandpiper*’

32

7/

5.2

ns

4/

2.5

21/

24.3 ns

Semipalmated Sandpiper

58

58/

8.3

+ -b

0/

4.6

ns

0/

37.2 --

Western Sandpiper

34

33/

9.1

-b-b

1/

5.2

ns

0/

19.6 --

Pectoral Sandpiper

90

89/

15.4

-b-b

0/

5.9

-

1/

68.7 --

Dunlin

959

959/

300.8

-b-b

0/

262.6

—

0/

395.6

Short-billed Dowitcher

53

53/

16.1

-b-b

0/

11.8

-

0/

24.1

Buff-breasted Sandpiper*"

31

28/

4.2

+ -b

3/

2.0

0/

24.8 --

Winter

Black-bellied Plover

1285

1012/

287.8

-b-b

273/

755.5

—

0/

241.7 --

Killdeer

169

145/

28.8

-b-b

24/

103.4

—

0/

36.8 --

Dunlin

2700

2177/

577.5

-b-b

523/1679.7

—

0/

442.8 —

Short-billed Dowitcher

42

42/

7.1

+ -b

0/

22.4

—

0/

12.5

" ns. P > 0.05; + and ++, significant positive associations {P < 0.05. P < O.OOl respectively), - and — , significant
negative associations {P < 0.05, P < 0.001 respectively).

^ For these species, ob.served and expected numbers were pooled for plowed fields and fields with <10 cm to achieve
expected values sufficiently high for testing.

sonal abundance and cover use patterns probably underestimated the use
of croplands by field exploiters, as roughly half of the fields in each
survey were visited 2-4 h preceding or following high tide. Because there
was no apparent relationship between the times at which fields were vis-
ited and the cover types present in fields, this observation should have
no effect on analyses of cover use.

Shorebird abundance in agricultural fields showed substantial variation
among censuses. Fluctuations in abundance within (rather than among)
seasons may have reflected differences in tidal height during the census
periods. More shorebirds may have foraged in fields during higher peak

Rottenhorn • AGRICULTURAL FIELD USE BY SHOREBIRDS

793

tides, when intertidal habitats were inundated for long periods, than dur-
ing periods when high tide levels failed to inundate intertidal flats com-
pletely. In addition, variation in shorebird abundance within seasons may
have been magnified by pulses of migration or regional movements that
augmented or reduced shorebird abundance within the study area.

More individuals foraged in fields during spring migration than in any
other season. This result may simply reflect larger numbers of birds pass-
ing through the study area in the shorter period of spring migration than
during the more protracted fall migration. The compressed spring move-
ment may also increase the shorebirds’ need for rapid energy intake at
staging areas (Pitelka 1979). If shorebirds benefit energetically from ag-
ricultural fields during high tide, then they should forage in fields during
those periods when rapid energy intake is most critical (e.g., spring mi-
gration).

Agricultural fields were used by few species during winter, but of those
present, most foraged in fields in high numbers. Because the threat of
starvation to shorebirds may be quite high during winter (Goss-Custard
1979), alternate foraging areas such as coastal agricultural fields may
enhance the chances of survival of overwintering shorebirds.

The type and structure of cover on which shorebirds forage may influ-
ence both predation risk and foraging efficiency. Despite the abundance
of avian predators in the study area during winter, spring, and especially
fall (pers. obs., Sutton 1992), most of the shorebird species in this study
showed an overwhelming preference for plowed fields while avoiding
dense vegetation.

Nearly all of the shorebirds observed were foraging in flocks, reducing
the risk of depredation for any one individual during a predation attempt
(Page and Whitacre 1975). Thus, flocking may have allowed the birds to
forage relatively safely on plowed fields as they do on open mudflats and
beaches. Flocking may also enhance foraging efficiency by allowing birds
to share vigilance (Powell 1974, Metcalfe 1989). Sharing vigilance re-
duces the time that any one individual must spend looking for predators,
thus increasing foraging time. Because vigilance sharing requires close
contact among individuals and is inhibited in tall cover (Metcalfe 1984),
vigilance sharing may not have been as profitable on vegetated fields as
on plowed fields.

Foraging in plowed fields may also have been more efficient than on
fields with >10 cm vegetation due to problems associated with foraging
in dense vegetation. Because taller or more dense vegetation is thought
to reduce the search area available to visual hunters, such as plovers
(Fuller and Youngman 1979), or to inhibit locomotion and prey extrac-
tion, foraging on plowed fields may have been more efficient than on

794

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

fields with >10 cm vegetation. On fields with <10 cm cover, vegetation
may not have been thick enough to impede efficient foraging in those
species exhibiting positive associations with this cover type. Wind, ex-
treme temperatures, and drought have been shown to reduce prey avail-
ability to a greater extent on bare earth than on vegetated cover (Burton
1974, Murton and Westwood 1974, Evans 1976, Shrubb 1988). However,
given the year-round preference for plowed fields shown by most of the
species in this study, these factors (and their seasonal variation) seem to
have had little effect on the birds’ preference for this cover type.

Although associations between most shorebird species and fields of
bare earth or >10 cm vegetation were consistent among seasons, the use
of fields with <10 cm cover showed substantial seasonal variation. Most
significant associations with this cover type were positive in spring but
negative in fall and winter. It is possible that the large numbers of shore-
birds present in spring and the need to obtain energy rapidly during that
season forced some shorebirds to forage on non-preferred cover types
(e.g., <10 cm vegetation). Alternately, prey densities on these agricultural
croplands may have been higher in spring than in other seasons, allowing
efficient foraging on both plowed fields and fields with <10 cm cover
during spring.

Because the periods of spring and fall migration require long-distance
migrants to accumulate large energy reserves (Davidson 1984, Myers et
al. 1987), and the threat of starvation to wintering shorebirds may be high
(Goss-Custard 1979), coastal agricultural croplands may be important for-
aging areas for shorebirds during these periods. Where croplands are pres-
ent near important shorebird staging areas, field management for foraging
shorebirds may increase the value of the overall staging areas to these
birds. Based on the results of this study, management regimes that provide
an ample supply of plowed fields during periods of peak shorebird abun-
dance might be most beneficial to shorebirds. Further research on the
effects of field size, shape, or proximity to intertidal areas and of agri-
cultural practices (e.g., plowing vs no-till farming, pesticide use, and tim-
ing of plowing and crop rotation) on shorebird abundance may facilitate
the management of coastal agricultural lands for use by nonbreeding
shorebirds.

ACKNOWLEDGMENTS

I am grateful to B. Watts, R. Beck, and M. Byrd for their help with study design and
manuscript preparation. The manuscript also benefited from comments by C. Blem, A.
Launer, J. Reaser, T Wong, D. Shuford, and S. Melvin. I thank the Augusta and Monticello
bird clubs and the College of William and Mary for their support of this study.

Rottenborn • AGRICULTURAL FIELD USE BY SHOREBIRDS

795

LITERATURE CITED

Atkinson, J. M. S. 1976. Inland wintering and urban roosting by redshanks. Bird Study
23:51-55. ^

Barnard, C. J. and D. B. A. Thompson. 1985. Gulls and plovers. Columbia Univ. Press,
New York, New York.

Burger, J., M. A. Howe, D. C. Hahn, and J. Chase. 1977. Effects of tide cycles on habitat
selection and habitat partitioning by migrating shorebirds. Auk 94:743-758.

Burton, P. J. K. 1974. Feeding and the feeding apparatus in waders. British Museum of
Natural History, London, U.K.

Colwell, M. A. and S. L. Dodd. 1995. Waterbird communities and habitat relationships
in coastal pastures of northern California. Cons. Biol. 9:827-834.

Connors, P. G., J. P. Myers, C. S. W. Connors, and F. A. Pitelka. 1981. Interhabitat
movements by Sanderlings in relation to foraging profitability and the tidal cycle Auk
98:49-64.

Davidson, N. C. 1984. Identification of refueling sites by studies of weight changes and
fat deposition. Pp. 68-78 in Shorebirds and large waterbirds conservation (P. R. Evans,
H. Hafner, and P. L’Hermite, eds.). Commission of the European Communities.

Evans, P. R. 1976. Energy balance and optimal foraging strategies in shorebirds: some
implications for their distributions and movements in the non-breeding season. Ardea
64:1 17-39.

Fuller, R. J. and R. E. Youngman. 1979. The utilisation of farmland by golden plovers
wintering in southern England. Bird Study 26:37-46.

Gerstenberg, R. H. 1979. Habitat utilization by wintering and migrating shorebirds on
Humboldt Bay, California. Pp. 33—40 in Shorebirds in marine environments (F. A.
Pitelka, ed.). Allen Press Inc., Lawrence, Kansas.

Goss-Custard, j. D. 1969. The winter feeding ecology of the redshank Tringa totanus.
Ibis 111:338-356.

. 1979. Effect of habitat loss on the numbers of overwintering shorebirds. Pp. 167-

177 in Shorebirds in marine environments (F. A. Pitelka, ed.). Allen Press Inc., Law-
rence, Kansas.

Hayman, P, j. Marchant, and T. Prater. 1986. Shorebirds: an identification guide to the
waders of the world. Houghton-Mifflin Company, Boston, Massachusetts.

Kain, T. 1987. Virginia’s birdlife. Virginia Society of Ornithology, Virginia Avifauna No. 3.

Metcalfe, N. B. 1984. The effects of habitat on the vigilance of shorebirds: is visibility
important? Anim. Behav. 32:981-985.

. 1989. Flocking preferences in relation to vigilance benefits and aggression costs

in mixed-species shorebird flocks. Oikos 56:91-98.

Milsom, T. R, R. S. Holditch, and J. B. A. Rochard. 1985. Diurnal use of an airfield and
adjacent agricultural habitats by Lapwings Vanellus vanellus. J. Appl. Ecol. 22:313-
326.

Murton, R. K. and N. j. Westwood. 1974. Some effects of agricultural change on the
English avifauna. Brit. Birds 67:41-67.

Myers, J. P, R. I. G. Morrison, P. Z. Antas, B. A. Harrington, T. E. Lovejoy, M.
Sallaberry, S. E. Senner, and A. Tarak. 1987. Conservation strategy for migratory
species. Am. Sci. 75:19-26.

Ohmart, R. D., B. W. Anderson, and W. C. Hunter. 1985. The influence of agriculture
on waterbird, wader, and shorebird use along the lower Colorado River. Pp. 1 17—122
in Riparian ecosystem and their management: reconciling conflicting uses (R. R. John-

796

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

son, C. D. Ziebell, D. R. Patton, B. E Ffolliott, and R. H. Hamre, eds.). USDA For.
Serv. Gen. Tech. Rep. RM-120.

Page, G., L. E. Stenzel, and C. M. Wolfe. 1979. Aspects of the occurrence of shorehirds
on a central California estuary. Pp. 15-32 in Shorehirds in marine environments (F. A.
Pitelka, ed.). Allen Press Inc., Lawrence, Kansas.

AND D. F. Whitacre. 1975. Raptor predation on wintering shorehirds. Condor 77:

73-83.

Pitelka, F. A. 1979. Introduction: the Pacific Coast shorehird scene. Pp. 1-1 1 in Shorehirds
in marine environments (F. A. Pitelka, ed.). Allen Press Inc., Lawrence, Kansas.

Powell, G. V. N. 1974. Experimental analysis of the social value of flocking hy starlings
(Sturnu.s vulgaris) in relation to predation and foraging. Anim. Behav. 22:501-505.

Shrubb, M. 1988. The influence of crop rotations and field size on a wintering lapwing V.
vanellus population in an area of mixed farmland in West Sussex. Bird Study 35:123-31.

SoKAL, R. R. AND F. J. Rohlf. 1981. Biometry (second edition). W. H. Freeman and Com-
pany, New York.

Stout, G. D. 1967. The shorehirds of North America. Viking Press, New York.

U.S. Dept, of Commerce. 1990. Tide tables 1991 — high and low water predictions, east
coast of North and South America. National Oceanic and Atmospheric Administration,
Washington, D.C.

. 1991. Climatological data: Virginia. National Oceanic and Atmospheric Admin-
istration, Washington, D.C.

Wilkinson, L. 1990. SYSTAT: the system for statistics. SYSTAT, Inc., Evanston, Illinois.

Wilson Bull.. 108(4), 1996, pp. 797-802

SHORT COMMUNICATIONS

Extra nest site occupancy by Tree Swallows: Do floaters avoid nest sites near settled
pairs?— Tree Swallows (Tachycineto bicolor) are secondary hole-nesters for whom suitable
cavities are a limiting resource (Holroyd 1975, Stutchbury and Robertson 1985). Despite
nest site shortage, more than one cavity may be occupied by a single breeding pair (Rendell
and Robertson 1994). Because Tree Swallows are single-brooded and generally monoga-
mous (Robertson et al. 1992, but see Hussel 1983), the second site often goes unused.
Occupancy of extra nest sites (ENSs) is therefore puzzling.

In this study we tested the hypothesis that ENS occupancy by Tree Swallows results from
avoidance by floaters of nest sites near settled pairs. To date, most explanations of this
beha\ior have assumed that a differential cost to defending the second site exists,. and all
have focused on benefits arising from its successful defense (Kendeigh 1941, Harris 1979,
Robertson and Gibbs 1982, Muldal et al. 1985, Einch 1990, Dunn and Hannon 1991, Rendell
and Robertson 1994). An alternative explanation is that floaters avoid or reject nest sites
too close to already settled pairs. Residents may then occupy the extra sites without any
increase in their total defense costs. That is, an ENS may be nothing more than a convenient
perch from which to observe the primary site. The “avoidance hypothesis” sees ENS oc-
cupancy as an artifact of floater behavior rather than as resulting from the resident pair’s
defense and predicts that floaters will not make use of the nest sites in question even if the
residents’ defense radius is reduced. The hypothesis was noted in passing by Muldal et al.
(1985) but has never been tested.

The avoidance hypothesis gains credibility from the observation that floaters in some Tree
Swallow populations reject nest boxes placed in wooded areas or near forest edges (Harris
1979, Rendell and Robertson 1990, but see Erskine and McLaren 1976, Peterson and Gau-
thier 1985). The selectivity may reflect a risk at those nest boxes of egg destruction and
nest site usurpation by House Wrens (Troglodytes aedon. Rendell and Robertson 1990) or
a need for an open area nearby for foraging (Munro and Rounds 1985). If Tree Swallows
are prepared to reject an otherwise suitable nest box because of its proximity to a forested
area, might they also do so because of proximity to breeding conspecifics?

Eloaters rejecting nest sites may lose breeding opportunities. Estimates of adult annual
mortality in Tree Swallows range from 40-60% (Chapman 1955, Houston and Houston
1987) so delaying reproduction is risky. If Tree Swallows do avoid or reject nest sites near
settled pairs, failure to do so must entail costs similar in magnitude to those of wren activity.
This seems unlikely. Anecdotal accounts of Tree Swallows raising broods in nest boxes 1
m apart (Harris 1979) and even of two females laying in the same nest box (Quinney 1983,
Muldal et al. 1985, Rendell 1992) suggest that floaters will breed near conspecifics rather
than not breed at all.

To test the avoidance hypothesis, we arranged nest boxes in pairs and visually separated
half those pairs with sheets of chipboard. Early in the season, breeding Tree Swallows defend
their nest sites by spending the majority of their time perched at or near their nest holes
(Leffelaar and Robertson 1984, Stutchbury and Robertson 1987a). Our visual barriers there-
fore made defense of a second box more difficult. Floaters locate unoccupied nest sites from
the air and would be able to see both the unoccupied ne.st box and its proximity to an
occupied box. If floaters avoid nest sites because of their proximity to already settled birds
then barriers should have no effect. But, if defense of the extra nest box is important, then
floaters should settle more often where barriers are pre.sent, i.e., we should observe two
pairs of Tree Swallows breeding in adjacent nest boxes when barriers are present. As argued

797

798

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Forest

Eig. 1 . Diagram of the Tree Swallow grid showing the arrangement of nest boxes (rect-
angles) and barriers (short vertical lines separating alternate nest box pairs) and the settle-
ment pattern a day before the first egg was laid. Rings around cells or individual nest boxes
indicate the nest boxes occupied by a pair of Tree Swallows and are not intended to delineate
accurately the area defended. The dashed line ring around four nest boxes (two cells) circles
the nest boxes occupied by a bigamous male; each female occupied one cell. Note that two
control cells were occupied by one Tree Swallow pair (top right on the diagram) and that
two pairs settled at each of four experimental cells.

above, we felt that the avoidance hypothesis, while plausible, was improbable and predicted
that floaters would settle adjacent to a breeding pair more readily when a visual barrier was
in place.

To avoid confusion between pairs of breeding Tree Swallows and paired nest boxes, we
hereafter refer to the latter as cells.

Methods. — In late April 1991, we erected a grid of 44 nest boxes, arranged in 22 pairs
(cells), on a 1.6 ha hayfield in Leed’s County, Ontario (Fig. 1). Boxes were placed 1.5 m
above the ground on aluminum poles. The two nest boxes of a cell were 8 m apart. Since
Tree Swallows from this population normally defend a territory around a nest site of radius
15 m (Robertson and Gibbs 1982), we expected defense of a complete cell (two nest boxes)
by each breeding pair. The shortest distance between nest boxes from different cells was 23

SHORT COMMUNICATIONS

799

Table I

Number of Control and Experimental Paired Nest Boxes (Cells) at Which Both
Nest Sites Were and Were Not Occupied by a Single Breeding Pair of Tree

Swallows

Barrier No barrier

Extra nest site occupation maintained 7 I j

Extra nest site occupation not maintained 4 o

m, which is greater than the normal defense radius, so defense of more than one cell by a
breeding pair was not expected at either experimental or control cells.

At alternate cells down each row of the grid, we erected sheets of chipboard (1.2 X 2.4
m) between the nest boxes such that, from each nest box, the other, 8 m away, could not
be seen. Chipboard sheets were visual barriers which, when present, made defense of both
nest boxes in a cell more difficult for a breeding pair. Birds could not perch on the barrier
because we put a tight line, strung with plastic drinking straws, about 1 cm above the top
of each barrier. We refer to cells with chipboard obstructing the view of the adjacent box
as barrier cells. The remaining 1 1 cells were control cells. Nothing obstructed the view of
Tree Swallows perched on nest boxes at control cells.

Erecting barriers after settlement might have caused resident abandonment, so we put
them up at the same time as the nest boxes, prior to Tree Swallow settlement. We could
not detect any difference in any index of settlement date between control and barrier cells
and, therefore, concluded that the presence of barriers did not affect the birds’ nest box
choice decisions (Mitchell 1992).

Residents were captured using mist nets and box traps (Cohen and Hayes 1984, Stutch-
bury and Robertson 1986) and marked with unique acrylic wing paint codes. Following
banding, settlement and residency were determined using periodic 3-5 min watches at in-
dividual nest boxes and by regular examination of nest box contents.

Results. — Many Tree Swallow pairs initially occupied more than one cell. The number
of breeding pairs then continued to increase throughout the breeding season until 28 active
nests (eggs or nestlings present) were spread across the 22 cells and at least one breeding
pair of Tree Swallows was present at every cell on the grid. We therefore had to pick a date
on which to compare ENS occupancy frequencies at control and barrier cells. Costs and
benefits of a second nest site may change following laying, so we decided to compare those
frequencies on 13 May, a day prior to the laying of the first egg on the grid, at which time
the grid held 25 active nests. Figure 1 shows the grid settlement pattern on 13 May. Oc-
cupancy of both nest boxes of a cell by a single breeding pair of Tree Swallows was
significantly less common when a barrier .separated the two boxes than at controls (one-way
Fisher exact test, N = 22, m = 11,4, f = 0, P = 0.04) (Table 1). In fact, there were no
cases where the two nest boxes of a cell were occupied by separate pairs of Tree Swallows
except where a barrier separated those boxes.

The test is conservative since maintenance of two nest boxes broke down only at barrier
cells; in addition, one Tree Swallow pair was still occupying two control cells (four boxes)
and a bigamous male was also occupying two control cells, with each female occupying a
single cell, at this time (Fig. 1). The bigamous two-cell occupation remained stable for the
duration of the breeding season. The two-cell occupation ended when the original female
started laying at one cell, whereupon a new Tree Swallow pair settled at the second cell.

800

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

This was the twenty-sixth active nest. No Tree Swallow was occupying more than one barrier
cell at that time.

Approximately four weeks after the initial settling, females settled at the unused nest
boxes at one barrier and one control cell following the hatching of the original females’
clutches. These may have been lone females or may have been paired with the original
resident male. Only the new females fed the nestlings.

We observed several resident females, at both barrier and control cells, adding nesting
material to both nest boxes of cells involved in ENS occupancy. In each instance, though,
eggs were laid in only one of the boxes. Where pairs occupied extra nest sites, both residents
spent time at each nest box and both responded aggressively to intruders, although it was
not possible to determine to which box a threat was perceived.

Discussion. — The lower frequency of ENS occupancy when adjacent nest boxes were
separated by a visual barrier permits rejection of the avoidance hypothesis. When the ability
of a resident to defend a second nest box was experimentally reduced, floaters settled and
bred in boxes 8 m from existing pairs. Floaters may prefer to nest at greater distances from
conspecifics, but this preference is not a sufficient explanation for ENS occupancy. It is fair
to characterize ENS occupancy as ENS defense.

We emphasize that our result cannot simply be an outcome of barriers causing settling
floaters to be unaware of their neighbors or residents to be unaware of a nearby nest box.
While territory defense is accomplished through time spent perched at a nest site, floaters
searching for nest sites do so by flying through a colony repeatedly (Stutchbury and Rob-
ertson 1987b) and thus should be aware of a neighbor’s presence. The ability of seven Tree
Swallow pairs to maintain two nest boxes despite the presence of a barrier demonstrates
that the second box’s presence is recognized by the resident pair and that defense is possible.

Rendell and Robertson (1994) found that male Tree Swallows, and possibly females,
preferred territories on which a second nest box could easily be defended (because of the
proximity of adjacent nest boxes). Mitchell (1992) detected no evidence of preference for
control cells by Tree Swallows in his study, but the rapid pace of settlement and our small
sample size could have masked such a preference. If a preference did exist, it was not
sufficient to induce settling Tree Swallow pairs to reject cells consistently where barriers
were present. Therefore, our conclusion remains unchanged: Avoidance by floaters is an
insufficient explanation for ENS occupancy by Tree Swallows.

Using model intruders, Robertson and Gibbs (1982) concluded that Tree Swallows defend
a circular territory around a primary nest box and occupy additional nest boxes which
happen to lie within that radius. That hypothesis has since been refuted by Rendell and
Robertson (1994) who found that Tree Swallows occupied nest sites up to 56 m apart. These
findings raise a question that has thus far not been addressed: When a resident Tree Swallow
is observed responding aggressively to an intruder, how can one determine whether the
resident perceives a threat to its primary nest site or to an ENS? Our study succeeds in
rejecting the avoidance hypothesis without using observations of aggressive behavior. How-
ever, we suspect that distinguishing between two remaining hypotheses will require such
observations.

Rendell and Robertson (1994) argued that ENS defense by pairs of Tree Swallows results
from an intersexual conflict of interest. Males defend an ENS in hopes of becoming polyg-
ynous. Females then defend the same extra site to ensure male monogamy. Rendell and
Robertson (1994) could not convincingly reject an alternative hypothesis that Tree Swallow
pairs cooperate in defense of an ENS as a site for renesting should their first attempt fail.
These hypotheses reflect fundamentally different views of territorial behavior. Not only does
the explanation for ENS defense differ, but the perceived relationship between members of
the breeding pair differs.

SHORT COMMUNICATIONS

801

It aggressive behavior can be unequivocally assigned to defense of an ENS, then the two
explanations can be distinguished by looking at which sex male and female Tree Swallows
defend a second site against. If a polygyny-monogamy conflict is involved, females should
respond aggressively to intruding females but ignore or actively encourage intruding males,
since settlement by a second male removes the risk of polygyny for the female. Male Tree
Swallows pattern of aggression should be just the opposite. If defense is cooperative, both
male and female Tree Swallows should defend an ENS against intruders of either sex.

This distinction was not possible in our study, in which paired nest boxes were only 8 m
apart. However, the comparison would be worthwhile when the two nest sites are more
widely separated, as in some of the cases reported by Rendell and Robertson (1994) and
where floaters have been captured, sexed, and individually marked.

Acknowledgments. — Stan Teeple generously allowed us to use one of his hayfields as a
study site. Beth MacDougall, Melanie Sharman, Lisa Venier, and Linda Webster assisted
with grid construction, and Grace Erb with banding and observations. Andrew Chek, Floyd
Connor, Kelvin Conrad, Peter Dunn, Frank Phelan, Susan Meek, Laurene Ratcliffe, and
Linda Whittmgham provided useful criticism of the design and interpretation of this study.
The Queen’s Univ. Biological Station provided logistic support. The study was funded by
NSERC and SEED grants to RJR. JSM was supported during preparation of the manuscript
by an NSERC Postgraduate Scholarship.

LITERATURE CITED

Chapman, L. B. 1955. Studies of a Tree Swallow colony. Bird-Banding 26:45—70.

Cohen, R. R. and D. J. Hayes. 1984. A simple unattached nest-box trapping device. N.
Am. Bird-Bander 9: 10-1 1.

Dunn, P. O. and S. J. Hannon. 1991. Intraspecific competition and the maintenance of
monogamy in tree swallows. Behav. Ecol. 2:258-266.

Erskine, a. j. and W. D. McLaren 1976. Comparative nesting biology of some hole-
nesting birds in the Cariboo Parklands, British Columbia. Wilson Bull. 88:61 1-620.
Finch, D. M. 1990. Effects of predation and competitor interference on nesting success of
House Wrens and Tree Swallows. Condor 92:674-687.

Harris, R. N. 1979. Aggression, superterritories, and reproductive success in tree swallows.
Can. J. Zool. 57:2072-2078.

Holroyd, G. L. 1975. Nest site availability as a factor limiting population size of swallows.
Can. Field Nat. 89:60-64.

Houston, M. I. and C. S. Houston. 1987. Tree Swallow banding near Saskatoon, Sas-
katchewan. N. Am. Bird-Bander 12:103-108.

Hussel, D. j. T. 1983. Age and plumage color in female Tree Swallows. J. Field Ornithol.
54:312-318.

Kendeigh, S. C. 1941. Territorial and mating behavior of the house wren. Illinois Biol.
Monogr. 18:1-120.

Leffelaar, D. and R. J. Robertson. 1984. Do male tree swallows guard their mates?
Behav. Ecol. Sociobiol. 16:73-79.

Mitchell, J. S. 1992. Multiple nest site defen.se in the Tree Swallow {Tachycinetci hicolor).

B.Sc. thesis, Kingston: Queen’s Univ., Kingston, Ontario.

Muldal, a., H. L. Gibbs, and R. J. Robertson. 1985. Preferred nest spacing of an obligate
cavity-nesting bird, the Tree Swallow. Condor 87:356—363.

Munro, H. L. and R. C. Rounds. 1985. Selection of artificial nest sites by five sympatric
passerines. J. Wildl. Manage. 49:264—276.

Peterson, B. and G. Gauthier. 1985. Nest site u.se by cavity-nesting birds of the Cariboo
Parkland. British Columbia. Wilson Bull. 97:319-331.

802

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Quinney, T. E. 1983. Tree Swallows cross a polygyny threshold. Auk 100:750-754.
Rendell, W. B. 1992. Peculiar behavior of a subadult female Tree Swallow. Wilson Bull.
104:756-759.

AND R. J. Robertson. 1990. Influence of forest edge on nest-site selection by Tree

Swallows. Wilson Bull. 102:634-644.

AND . 1994. Defense of extra nest-sites by a cavity nesting bird, the Tree

Swallow (Tachycineta bicolor). Ardea 82:273—285.

Robertson, R. J. and H. L. Gibbs. 1982. Superterritory in Tree Swallows: a reexamination.
Condor 84:313-316.

, B. J. Stutchbury, and R. R. Cohen. 1992. Tree Swallow. The Birds of North

America, No. 1 1 (A. Poole, P. Stettenheim, and F. Gill, eds.). Academy of Natural
Sciences, Philadelphia; The American Ornithologists’ Union, Washington, D.C.
Stutchbury, B. J. and R. J. Robertson. 1985. Floating populations of female Tree Swal-
lows. Auk 102:651-654.

AND . 1986. A simple trap for catching birds in nest boxes. J. Field Omithol.

57:64-65.

AND . 1987a. Do nest building and first egg dates reflect settlement patterns

of females? Condor 89:587-593.

AND . 1987b. Behavioral tactics of subadult female floaters in the tree swal-
low. Behav. Ecol. Sociobiol. 20:413-419.

Jeremy S. Mitchell, Dept, of Biological Sciences, Simon Fraser Univ., Burnaby, British
Columbia V5A 1S6-, Raleigh J. Robertson, Dept, of Biology, Queen’s Univ., Kingston,
Ontario K7L 3N6. Received 29 Feb. 1996, accepted 10 May 1996.

Wilson Bull., 108(4), 1996, pp. 802-804

Swainson’s Warblers nesting in early serai pine forests in East Texas. — Swainson’s
Warbler (Limnothlypis swainsonii) breeds locally throughout the southeastern United States
on the south Atlantic and Gulf coastal plains, in the southern Appalachians, and on the
southern Piedmont Plateau (Meanley 1966). They select areas with low, dense understories
(Eddleman et al. 1980, Meanley 1971). Meanley (1945, 1966, 1969) described several hab-
itats of Swainson’s Warblers in the eastern coastal plains, and they also have nested in dense
rhodendron-laurel thickets in the southern Appalachians (Brooks and Legg 1942). However,
there is little information on the specific habitat features which may influence the distribution
and abundance of this species, especially in the western portion of its breeding range. In
this paper, I report on the first described use of early serai pine forests by Swainson’s
Warblers nesting in eastern Texas. I compare habitats selected in this study to those previ-
ously described and suggest possible relationships between habitat selection and abundance
patterns.

I surveyed for Swainson’s Warblers from April through June 1992 on the San Jacinto
Ranger District, Sam Houston National Forest, San Jacinto County, Texas (95°07'W,
30°30'N). The district consists of approximately 24,000 ha of pine, pine-hardwood, and
bottomland hardwood forest of various age classes. The district is managed for timber
production, recreation, wildlife, and some oil and mineral extraction. Timber is managed
primarily in even-age stands ranging from 4-40 ha on 70-year rotations. Harvest methods
include thinning, clearcuts, seed-tree cuts, and shelterwood cuts. Large areas regularly are

SHORT COMMUNICATIONS

803

prescribed burned. I used playback recordings to detect territorial males, primarily along
maintained roads, old logging roads, and trails which transect much of the district. Males
captured in mist nets were banded with unique combinations of colored leg bands and a
U.S. Fish and Wildlife Service aluminum leg band. I located territories and determined
whether each male was paired. I considered an individual interacting with another bird
within the territory without hostile behavior to be paired. I returned to each territory 2-4
times during the breeding season. Stand ages were determined using the Continuous Inven-
tory of Stand Condition database maintained by the U.S.D.A. Forest Service.

I monitored 38 territories during the 1992 season. Eleven (28.9%) territories were in
loblolly pine {Pinus taeda) plantations (age range = 3-18 yrs; mean age = 14.0 ± 4.2 yrs)
with dense understories composed primarily of yaupon {Ilex vomitoria), Smilax spp.. Vi-
burnum spp., and various sapling hardwood species. Sixteen (42.1%) were in sites previously
logged to remove pine, infested with southern pine beetles {Dendroctonus frontalis). These
sites ranged from 1—2 ha in size and were dominated by a dense growth of early successional
shrub species and sapling trees. The remaining 1 1 (28.9%) territories were in relatively
mature stands of loblolly pine with a sparse overstory and patches of dense understory.
Eight of these territories were in mesic sites near streams or ephemeral forest swamps. The
remaining three were in relatively xeric upland stands with a mixed pine-hardwood overstory
and dense yaupon understory.

I confirmed nine paired males: three in plantations, five in logged sites, and one in a
mature stand. I found a probable Swainson’s Warbler nest in an active territory in a pine
plantation on 21 May 1992. The nest closely resembled photographs by Graves (1992) and
possessed the bulky outer layer of dead hardwood leaves characteristic of the species (Mean-
ley 1969). The nest was empty and appeared to have been abandoned or predated since no
eggs or young were observed during six visits I made the following three weeks. I observed
an adult Swainson’s Warbler foraging approximately 2 m from the nest during the first visit.
The nest was approximately 1 m high and supported by a tangle of Smilax spp. suspended
in pine saplings. I observed one pair feeding fledged young in a logged site located within
the territory of the male. The young were still incapable of sustained flight and were more
than 20 m within the logged site.

Early serai pine forest previously has not been described as potential breeding habitat for
Swainson’s Warblers. They generally use areas with a low, dense deciduous understory
(Meanley 1971). River floodplain forest with dense understories of cane {Arundinaria gi-
gantea), sweet pepperbush (Clethra alnifolia), or scrub palmetto (Sabal minor) have been
identified as important breeding habitats on the upper and Gulf coastal plains (Meanley
1971). Swainson’s Warblers in the southern Appalachian Mountains use dense rhodendron-
laurel thickets in mature hardwood communities (Brooks and Legg 1942). Eddleman et al.
(1980) found several territories in areas dominated by early successional tree species, such
as sweetgum (Liquidamhar styraciflua) plantations, and in late old field habitats with dense
shrubs. These habitats possess a dense understory structure comparable to the pine planta-
tions and logged sites used by Swainson’s Warblers in eastern Texas. The understory foliage
diversity patterns of the.se different habitats would probably be relatively similar despite
differences in plant species composition, suggesting vegetative structural patterns may be
an important factor influencing habitat selection and distribution of Swainson’s Warblers.

Man-made areas of early serai forest, such as pine plantations and logged sites, may
represent new habitat options for breeding Swainson’ Warblers. There is no evidence to
suggest that such habitat resulting from normal forest succession was not used in addition
to historically abundant canebrake habitat. Cane was formerly a dominant understory feature
of the lowlands of eastern and southeastern Texas, but with the introduction of domestic
livestock, it has become relatively rare (Correll and Johnston 1970). Swainson’s Warblers

804

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

appear common on the San Jacinto Ranger District, despite the lack of cane habitat. The
paired individuals and nesting activity observed in pine plantations and logged sites suggests
such areas may be important breeding habitat.

Apparent flexibility of habitat use by Swainson’s Warblers may explain its relatively stable
population in contrast to the decline of the Bachman’s Warbler {Vermivora bachmani). The
Bachman’s Warbler may have been a cane specialist, and the probable extinction of the
species parallels the decline of cane stands throughout the southeastern United States (Wid-
mann 1897, Remsen 1986). The more generalist strategy of the Swainson’s Warbler may
have enabled it to continue exploiting alternative habitats such as dense thickets in forest
openings to maintain population levels.

Acknowledgments. — I thank the staff of the San Jacinto Ranger District, Sam Houston
National Forest, for assistance during this project. L. J. Carmical provided access to the
district and Forest Service records. Comments from R. E. Brown, D. K. Carrie, R. N. Conner,
J. Nelson, and D. C. Rudolph improved an earlier draft of this manuscript.

LITERATURE CITED

Brooks, M. and W. C. Lego. 1942. Swainson’s Warbler in Nicholas County, West Virginia.
Auk 59:76-86.

CORRELL, D. S. AND M. C. JoHNSTON. 1970. Manual of the vascular plants of Texas. George
Banta Co., Inc., Menasha, Wisconsin.

Eddleman, W. R., K. E. Evans, and W. H. Elder. 1980. Habitat characteristics and man-
agement of Swainson’s Warbler in southern Illinois. Wildl. Soc. Bull. 8:228-233.
Graves, G. R. 1992. A case of aggregated nest placement and probable polygyny in the
Swainson’s Warbler. Wilson Bull. 104:370—373.

Meanley, B. 1945. Notes on Swainson’s Warbler in central Georgia. Auk 62:395-401.

. 1966. Some observations on habitats of the Swainson’s Warbler. Living Bird 5:

151-165.

. 1969. Pre-nesting and nesting behavior of the Swainson’s Warbler. Wilson Bull.

81:246-257.

. 1971. Natural history of the Swainson’s Warbler. North Am. Fauna 69:1-90.

Remsen, J. V., Jr. 1986. Was Bachman’s Warbler a bamboo specialist? Auk 103:216-219.
WiDMANN, O. 1 897. The summer home of Bachman’s Warbler no longer unknown. Auk
14:305-309.

N. Ross Carrie, Dept, of Wildlife and Fisheries Sciences. Texas A&M Univ., College Sta-
tion, Texas 77843. (Present address: HQ JRTC & Ft. Polk, AFZX-PW-EC, Ft. Polk, Loui-
siana 71459-7100). Received 21 Aug. 1995, accepted 2 March 1996.

Wilson Bull., 108(4), 1996, pp. 804-807

Measurements of Snail Kite eggs from central Florida. — The Snail Kite (Rostrhamus
sociabilis) is a raptor with a disjunct distribution among several lake and everglade wetlands
in central and south Florida (Sykes et al. 1995). The objectives of the present study were
to collect measurements of kite eggs, delimit egg size variation among wetlands, and de-
termine if egg size was correlated with clutch size, hatching success, fledging success,'and
breeding chronology of kites in central Florida.

Methods. — I visited Snail Kite nests every 1-2 weeks during 1991 at East Lake Toho-

SHORT COMMUNICATIONS

805

Table 1

Linear Measurements and Calculated Volumes of Snail Kite Eggs from Central

Florida”

Site

Variable

N

^ ± SD

Range

East Lake Tohopekaliga

Length

15

44.27 ±

1.16 mm

41.9-47.2 mm

Breadth

15

36.56 ±

0.74 mm

35. 1-37.8 mm

Volume

15

30.20 ±

1.58 cm3

26.33-32.37 cm3

Lake Tohopekaliga

Length

254

44.33 ±

1.74 mm

40.0-48.4 mm

Breadth

254

36.38 ±

0.90 mm

32.3—38.5 mm

Volume

254

29.97 ±

2.27 cm3

22.67-34.77 cm3

St. Johns Marsh

Length

57

44.46 ±

1.53 mm

40.2—47.5 mm

Breadth

57

36.41 ±

1.22 mm

33.6-40.1 mm

Volume

57

30.14 ±

2.87 cm3

24.07-37.64 cm3

Lake Kissimmee

Length

129

43.90 ±

1.64 mm

40.5-49.1 mm

Breadth

129

35.91 ±

1.11 mm

33.2-38.5 mm

Volume

129

28.94 ±

2.61 cm3

22.77-35.61 cm3

Lake Okeechobee

Length

250

44.21 ±

1 .35 mm

41 .0—48.6 mm

Breadth

250

36.21 ±

0.98 mm

33.7-39.4 mm

Volume

250

29.61 ±

2.21 cm3

24.67-38.28 cm3

Total

Length

705

44.22 ±

1.57 mm

40.0—49.1 mm

Breadth

705

36.24 ±

1.01 mm

32.3-40.1 mm

Volume

705

29.67 ±

2.38 cm3

22.67-38.28 cm3

“Egg volume = 0.51 X length X breadthL

pekaliga. Lake Tohopekaliga, Lake Kissimmee, Lake Okeechobee, and an impounded marsh
at the headwaters of the St. Johns River (hereafter St. Johns Marsh). Eggs were numbered
with an indelible felt-tip pen according to deposition sequence when known. Length and
breadth measurements were made with a Tajima dial caliper (model dial- 15) to the nearest
0.1 mm. Egg volume was estimated using the equation of Hoyt (1979): volume = 0.51 X
length X breadth^. Relative egg volume used in some statistical analyses was calculated
after Arnold (1991): individual egg volume minus the mean egg volume of the clutch
divided by the standard deviation of egg volume for the clutch.

All computations were preformed using SAS Institute, Inc. software. Length, breadth, and
volume were found to be normally distributed (Shapiro-Wilk test, P > 0.05; SAS Institute
1990). A plot of the residuals versus the predicted values from an ANOVA model using
PROC GLM (SAS Institute 1990; yielded a random scatter that suggested homogeneity of
variance for the three variables. Thus, parametric statistics were u.sed in all subsequent
analyses. Mensural variables were determined for both complete and incomplete clutches.
However, only data from complete clutches were u.sed when comparing egg size with clutch
size, hatching order, hatching success, and fledging success. Logistic regression in PROC
GENMOD (SAS Institute 1996) was u.sed to analyze hatching and fledging success since
these data are binomial. Hatch date was a convenient index for breeding chronology and
was determined as the week of hatching for the first egg in each clutch.

Results. — I calculated mean, standard deviation, and range of length, breadth, and volume
of 705 eggs from 260 complete and incomplete clutches of Snail Kites (Table 1). The relation

806

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

of mensural variables was length = 0.81 breadth (Pearson’s r^704 = 0.52, P < 0.0(X)1). No
significant difference (ANOVA, F = 1.97, df = 704, P = 0.10) was detected in length
among the five wetlands. However, significant differences were found in breadth (ANOVA/
Fisher’s LSD test, F = 5.58, df = 704, P = 0.0002) and volume (ANOVA/Fisher’s LSD
test, F = 4.87, df = 704, P = 0.0007). Eggs from Lakes Kissimmee and Okeechobee
consistently exhibited the smallest mean breadth and volume (Table 1).

No significant difference was found in length (ANOVA, F = 1.70, df = 46, P = 0.19),
breadth (ANOVA, F = 0.15, df = 46, P = 0.86), or relative volume (ANOVA, F = 0.60,
df = 46, P = 0.55) among the first, second, or third egg hatched. There also was no
significant difference (ANOVA, F = 1.41, df = 238, P = 0.24) in mean relative egg volume
among clutches of one (N = 3), two (N = 52), three (N = 169), and four (N = 8) eggs.
Thus, egg size was similar regardless of laying order or clutch size.

Relative volume was not correlated with hatch date at East Lake Tohopekaliga (r^,3 =
0.05, F = 0.63, P = 0.44), Lake Tohopekaliga (H252 0.002, F = 0.60, P = 0.44), Lake

Okeechobee (r^248 “ 0.006, F = 1.52, P = 0.22), and St. Johns Marsh (r^jj = 0.001, F =

O. 04, P = 0.84), except at Lake Kissimmee (r^i27 = 0.059, F = 5.18, P = 0.02). Thus,
volume generally did not exhibit intraseasonal variation at most wetlands. Finally, relative
volume was not correlated with hatching success (P = 0.88, df = 155) or fledging success
(P = 0.99, df = 227) of kite nestlings.

Discussion. — The mensural values of my study are similar to the average length (44.6
mm, N = 317 eggs) and breadth (36.1 mm) reported by Sykes (1987). Although Snail Kite
eggs exhibited differences in breadth and volume among my study sites, I found little
correlation with clutch size, laying sequence, and hatching date. Kites are similar to other
avian species whose egg sizes do not correlate with laying sequence (Pulliainen and Saari
1993), clutch size (Ojanen et al. 1981, Jover et al. 1993), or laying date (Arnold 1991, Potti
1993, Pulliainen and Saari 1993).

An increase in egg size during the laying sequence has been suggested as a strategy by
females to adjust their egg contribution to the initial nestling size disadvantage due to
asynchronous hatching (reviewed by Magrath 1992). However, size and volume of Snail
Kite eggs were not correlated with laying sequence and fledging success. Kites neither
support the brood reduction strategy (reviewed by Magrath 1992) that egg size decreases
with laying order nor the bet-hedging strategy (reviewed by Slagsvold et al. 1984) that egg
size increases with laying order. Because egg volumes within a clutch and among different
clutch sizes were similar, the competitive ability of Snail Kite nestlings results mainly from
hatching asynchrony and the ability of parents to secure adequate snails to feed their young.
Magrath (1992) suggested hatching asynchrony was more important than egg mass in de-
termining hatch size hierarchies and resultant early nestling survivorship. Significant differ-
ences in egg breadth and volumes among wetlands suggest local effects on egg sizes but
not the fledging success of Snail Kites during the single breeding season of my study. These
differences warrant further study to establish if a relationship exists between habitat quality
and condition of females with nestling survival during more stressful years.

Acknow ledgments. — Funding for this study was provided by Section 6 funding from the
U.S. Fish and Wildlife Service and the Nongame Wildlife Program of the Florida Game
and Fresh Water Fish Commission. S. T. Schwikert assisted me with the data collection. S.
B. Linda provided statistical consultation. 1 thank M. F. Delany, S. A. Nesbitt, D. A. Wood,

P. W. Sykes, C. R. Blem, and an anonymous referee for reviewing an earlier draft of the
manuscript. This research was part of study number 7520 of the Bureau of Wildlife Re-
search.

SHORT COMMUNICATIONS

807

LITERATURE CITED

Arnold, T. W. 1991. Intraclutch variation in egg size of American Coots. Condor 93:19-27.

Hoyt, D. F. 1979. Practical method of estimating volume and fresh weight of bird eggs
Auk 96:73-77. ‘

JOVER, L., X. Ruiz, AND M. Gonzalez-Marti'n. 1993. Significance of intraclutch egg size
variation in the Purple Heron. Ornis Scand. 24:127-134.

Magrath, R. D. 1992. Roles of egg mass and incubation pattern in establishment of hatch-
ing hierarchies in the Blackbird (Turdus merula). Auk 109:474-487.

Ojanen, M., M. Orell, and R. A. Vaisanen. 1981. Egg size variation within clutches:
effects of ambient temperature and laying sequence. Ornis Fenn. 58:93-108.

PoTTi, J. 1993. Environmental, ontogenetic, and genetic variation in egg size of Pied Fly-
catchers. Can. J. Zool. 71:1534-1542.

PULLIAINEN, E. AND L. Saari. 1993. Egg size of the Dotterel Charadrius morinellus in
Finland. Omis Fenn. 70:44-46.

SAS Institute. 1990. SAS procedures guide, version 6, third edition. SAS Institute, Inc.,
Cary, North Carolina.

. 1996. SAS/STAT software changes and enhancements through release 6.11. SAS

Institute, Inc., Cary, North Carolina.

Slagsvold, T, T. Sandvik, G. Rofstand, 0. Lorentsen, and M. Husby. 1984. On the
adaptive value of intraclutch egg-size variation in birds. Auk 101:685-697.

Sykes, P. W., Jr. 1987. Some aspects of the breeding biology of the Snail Kite in Florida.
J. Field Ornithol. 58:171-189.

, J. A. Rodgers, Jr., and R. E. Bennetts. 1995. Snail Kite (Rostrhamus sociabilis).
in The birds of North America, no. 171 (A. Poole and F. Gill, eds.). Acad. Nat. Sci.
Phil., Philadelphia, Pennsylvania, and Amer. Ornithol. Union, Washington, D.C.

James A. Rodgers, Jr., Wildlife Research Laboratory, Florida Game and Fresh Water Fish
Commission, 4005 South Main Street, Gainesville, Florida 32601. Received 14 Feb. 1996.
accepted 10 May 1996.

Wilson Bull., 108(4), 1996, pp. 807-808

The Andean Flamingo in Brazil. — Bege and Pauli (1990a, b) recorded the Andean
Flamingo (Phoenicoparrus andinus) for the first time in Brazil, based on an emaciated
juvenile found in May 19, 1989 in Erval Velho (27°13'S, 5I°23'W), midwest of Santa
Catarina State that had been banded in Chile. The specimen is now housed in the Museu
Nacional, Rio de Janeiro (MN 36.548). Antas (1990) also recorded a subadult bird foraging
at the Lagoa do Peixe (31°20'S, 51°05'W), southeastern Rio Grande do Sul, along with the
Chilean Flamingo (Phoenicopterus chilensis).

The Museu do Seminario Cora^ao de Jesus, in Corupa, Santa Catarina, houses a juvenile
specimen (MSCJ 220) obtained in 1952 at Jaragua do Sul (26°28'S, 49°06'W), northeastern
Santa Catarina, which was wrongly identified by Sick et al. ( 198 1 ) as an American Flamingo
(Phoenicopterus ruber). This specimen is a well preserved, mounted skin. To Rio Grande
do Sul we have an additional record based upon a color photograph of three adult birds in
Lagoa do Peixe, taken in the fall of 1992 (A. Hoffmann, pers. comm.). It was published

808

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

without proper identification by Porto (1992). At the time, many flocks of Llamingos were
recorded and photographed by Hoffmann.

Silva and Caye (1992) reported the Andean Llamingo as accidental in Rio Grande do Sul
and Bornschein (1992) also considered as such the occurrence of this species in Brazil.
However, it is quite likely that the Andean Flamingo migrates to Rio Grande do Sul with
more frequency than previously supposed, given the current number of records (fide Hoyo
et al. 1992, La Pena 1992). It is worth thus to stress the care in identifying Flamingos in
southern Brazil, where Phoenicopterus chilensis is a regular species, mainly in the winter
(Belton 1994).

Acknowledgments. — We thank Jorge B. Nacinovic. Jorge Baldo, Alejandro R. Giraudo,
and Dante M. Teixeira for confirmation identification of the flamingo in a photograph pub-
lished in the magazine Globo Rural, and J. B. Nacinovic also for text revision and trans-
lation. J. Baldo also sent us a color photograph with the details of the bills of Phoenicoparrus
andinus, P. jamesi and Phoenicopterus chilensis, which helped us identify the specimen in
the museum in Santa Catarina.

LITERATURE CITED

Antas, P. T. Z. 1990. Novos registros para a avifauna do Rio Grande do Sul. Pp. 80-81
in VI Encontro Nacional de Anilhadores de Aves, Pelotas.

Bege, L. a. R. and B. T. Pauli. 1990a. Two birds new to the Brazilian avifauna. Bull.
B.O.C. 110:93-94.

AND . 1990b. Primer reporte de Phoenicoparrus andinus en Brasil. Volante

Migratorio 14:6.

Belton, W. 1994. Aves do Rio Grande do Sul: distribui^ao e biologia. Sao Leopoldo, Ed.
UNISINOS.

Bornschein, M. R. 1992. Nova ocorrencia de Phoenicoparrus andinus para o Brasil. In:

II Congresso Brasileiro de Ornitologia, Campo Grande, r. 55.

Hoyo, J. del, A. Elliot, and J. Sargatal (eds.). 1992. Handbook of the birds of the world
vol. 1 : ostrich to ducks. Lynx Edicions, Barcelona, Spain.

La Pena, M. R. de. 1992. Guia de aves argentinas. Segunda edicidn (incluye nidos y
huevos), Tomo 1. Buenos Aires, L.O.L.A. (Literature of Latin America).

Porto, A. 1992. Passageiros do Sol. Globo Rural, No. 82:46—54.

Sick, H., L. A. do Rosario, and T. R. de Azevedo. 1981. Aves do Estado de Santa Catarina.
Sellowia, ser. zool.. No. 1:1-51.

Silva, E and C. E. Caye. 1992. Lista de Aves: Rio Grande do Sul. Porto Alegre, Pontificia
Universidade Catolica.

Marcos R. Bornschein and Bianca L. Reinert, Museu de Historia Natural “Capdo da
Imbuia”, Rua Prof. Benedito Conceigdo 407, Curitiba (PR), Brazil, 82810-080. Received 9
Oct. 1995, accepted 21 March 1996.

Wilson Bull., 108(4), 1996, pp. 809-812

ORNITHOLOGICAL LITERATURE

Edited by William E Davis, Jr.

Avian biochemistry and molecular biology. By Lewis Stevens. Cambridge University
Press, Cambridge, U.K. 1996:272 pp. $50.00 (cloth).-A synopsis on the first page of this
book says it is “. . . the only comprehensive and up-to-date survey of avian biochemistry
and molecular biology available.” The author has written a clear and concise coverage of
this area. The text covers such aspects as protein and amino acid metabolism, nutrition,
hpids, carbohydrates, avian hormones, and metabolic adaptation. Part 2 of the book is de-
voted to the avian genome and its expression” and summarizes this material not available
elsewhere.

The text is nicely organized and cleanly presented. The figures are nicely done and
relevent to the account. There were few, if any, typographical errors. The references are
fairly thorough and up-to-date, although I suspect United States authors are a bit under-
represented in the review. I was impressed that the author appreciated correct nomenclature
and taxonomy and made an effort to use it in a book that essentially is about molecular
biology. It frustrates me to see manuscripts about birds written by people who are not
sufficiently informed to recognize appropriate nomenclature.

This text is well worth the money for those interested in the cellular-molecular level of
ornithology. — C. R. Blem.

Orioles, blackbirds, and their kin: a natural history. By Alexander E Skutch. Ulus,
by Dana Gardner. The University of Arizona Press, Tucson, Arizona. 1996:291 pp. $50.00
(cloth), $21.95 (paper). — Who says there are no more heroes to emulate? Mine easily could
be Alexander F. Skutch. At time when study of whole organisms is considered passe by
some. Dr. Skutch has produced a masterful account of the natural history of a very important
group — the blackbirds (and after his 90th birthday!). This book, his twenty-sixth, guides us
to through a detailed, but never boring, account of how the orioles, blackbirds, and their
kin go through their lives. Some accounts were taken, “with slight modifications” from
Skutch’s Life histories of Central American birds, published in 1954 by The Cooper Orni-
thological Society, and from a few other miscellaneous published sources. The material has
been updated with references cited from the scientific literature through 1994, and new
accounts were added.

Skutch writes clearly, cleanly, and with a lot of content. To my taste he adds just enough
anthropomorphism to be interesting, but not unscientific. The book includes historical vi-
gnettes, personal asides, and insightful comments — all done in an economical, interesting
manner. The sections on cowbirds alone are worth the price of the book. Beginning birders
will enjoy the book and learn a great deal of fundamental ornithology. Advanced students
will find much useful information and introduction to specific references. The reference
section is not (nor was it intended to be) a comprehensive coverage of the blackbirds and
their kin.

Gardner’s black-and-white sketches are excellent, the binding and covers attractive, and
the book is well worth the money (particularly the paperback version). All-in-all this is a
wonderful book. It stands as an example of the way in which bird books of this genre should
be done. — C. R. Blem.

809

810

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Finding birds in southeast Arizona. By the Tucson Audubon Society Publications Com-
mittee. Tucson Audubon Society, Tucson, Arizona. 1995:347 pp., color photographs, maps,
bar-graphs. $16.95 (wire-O binding).

A birder’s guide to southeastern Arizona. By Richard Cachor Taylor. American Bird-
ing Association, Inc., Colorado Springs, Colorado. 1995:x + 342 pp., black & white pho-
tographs, drawings, maps, bar-graphs. $16.95 (wire-O binding). Here we have the two new-
est bird finding guides to some of the hottest hot spots in North America — the canyons,
deserts, and oases of southeastern Arizona. I am writing this review from the point of view
of a New Englander who has visited southeastern Arizona three times chasing birds, mam-
mals, herptiles, and butterflies. When I next go back, should I take along both of these
books or will one suffice?

Physically, the books are nearly identical — same height, width, and thickness and same
“wire-O” spiral binding with wrap-around back cover. By the way, who ever thought up
these wrap-around covers? They are awful. If they did not have maps printed inside them,
I would cut off the wrap-around flaps in an instant.

Every good bird finding guide should give the reader help in deciding when to visit, what
to wear, and where to stay. Both of these guides offer sound advice on these subjects. Books
of this genre should also provide a general introduction to biomes and climate, and in this
regard the Audubon Society book is superior, providing color photographs of typical habi-
tats. The Audubon Society guide also offers nine pages of timely advice to the would-be
visitor to Mexico and this is a definite plus.

The main body of text in these guides is in two or three parts and includes maps of,
directions to, and discussions of the best birding locales; seasonal bar-graphs; and comments
on selected species. First, the birding areas: As a three-peater, I have my favorite spots in
southeastern Arizona, so I chose four — Madera Canyon, the Chiracahua Mountains, Mount
Lemmon, and the Patagonia-Sonoita Creek Preserve — to compare coverage between the two
books. “Finding Birds in Southeastern Arizona” covers these four areas with 26 pages of
text and six maps; “A Birder’s Guide to Southeastern Arizona” uses 57 pages and nine
maps. Everything about the ABA guide is superior in this crucial section of the text. The
maps are clearer, and nothing is more important than a clear map! Taylor’s text offers keener
insights into all that is new and wonderful along the way, and although there is an unnec-
essarily greater emphasis placed on rarities, his are the words I would choose to follow. Not
only are there a lot more pages of description of these sights, there are considerably more
words per page in the ABA guide and so the reader has more than 2.5 times the text in
Taylor’s book, at least on these sites.

Second, the seasonal bar-graphs: In a word, they are excellent in both books. To the
serious bird student, these bar-graphs are gold mines. I must admit I favor simple, straight-
forward bar-graphs and for this reason I prefer those in the Audubon Society guide; too
much “going on” in the ABA bar-graphs.

Third, discussions of selected species: The ABA guides always contain an enlightening
chapter on “Specialties,” those species that for one reason or another birders most want to
see. A Birder’s Guide to Southeastern Arizona offers up a long list of specialties, along
with information on locating and identifying them. Finding Birds in Southeastern Arizona
includes annotations on all of the birds of the region, not just the “most wanted,” and this
is noteworthy and commendable.

Now, back to the original question. Do I buy both books and pack them along to Arizona
or do I choose between the two? Personally, I like to have as many compact nature guides
along as I can fit in a small box in the middle of the front seat of my rental car. I would
get them both, since each has outstanding features lacking in the other. However, if I ab-

ORNITHOLOGICAL LITERATURE

81 1

solutely had to choose, Richard Taylor’s “A Birder’s Guide to Southeastern Arizona” would
have pride of place in my luggage.— Brian E. Cassie.

Barn Owls: predator-prey relationships and conservation. By Iain Taylor. Cam-
bridge University Press, Cambridge. 1994:xvi + 304 pp., 46 photographs, 28 drawings, 81
graphs and tables, 2 maps, 2 appendices. $37.95 (cloth).— The Barn Owl (Tyto alba) ranges
across the warmer regions of six continents and perhaps has been as well studied as any
owl species in the world. The author of the present study has chosen to present the ecology,
biology, and conservation of the Barn Owl primarily through an investigation of the rela-
tionships between this bird and its prey. Taylor’s study area was in southern Scotland and
with the help of co-workers and local farmers, he managed to keep track of all of the Barn
Owls in his 1600 km^ main study site, as well as those in two smaller replicate study, areas.
His studies, conducted from 1978-1992, as well as field research studies from other parts
of the world, especially in New Jersey and Utah, form the basis of this book.

After a short introduction to the Barn Owl and the author’s research area and methods,
the book presents chapters on distribution, diet, foraging behavior, ecology and behavior of
the prey, prey selection and foraging habitats, ranging and roosting behavior, molt, breeding
seasons, nest sites, courtship and eggs, production of young, dispersal, mortality, population
size and regulation, and conservation. Two conventions that will make using this book
particularly easy for researchers are the use of headings on the right-hand pages and the
inclusion of a summary as each chapter’s conclusion. If only all authors and editors took
the time to include these! There is a 365-reference bibliography and the illustrations and
graphs are well done.

Of course, not everyone who picks up lain Taylor’s book is or will be an owl biologist.
Most, I assume, will be interested in owls in general and will want to know if the biology
and ecology presented are sound and if the book is readable. They will want to read the
book, not merely “use” it. For all of them (and you), owl fanciers and research biologists
alike, the book is very highly recommended. Taylor’s discussions are first-rate: well re-
searched, well written, and thought provoking. When there are suspect references in the
literature, Taylor questions them. When research is lacking into some aspect of the Bam
Owl’s life, Taylor points this out and offers suggestions to future researchers. This book is
an outstanding model for comprehensive, well structured, and enjoyable scientific litera-
ture.— Brian Cassie.

The Megapodes Mecapodudae. By Darryl N. Jones, Rene W. R. J. Dekker, and Cees S.
Roselaar, illus. by Ber van Perlo. Oxford University Press, New York. 1995:262 pp., 8 color
plates, 24 range maps, 18 numbered text figs., 5 tables. $60 (cloth). — This family of galli-
form-like birds is unique in that all members use some form of naturally occurring heat to
incubate their eggs, earning them the vernacular names of “thermometer or incubator
birds. Some species build mounds and use heat from the decomposition of damp organic
matter, while others u.se burrows and geothermal heat or solar-heated volcanic sand. The
chicks are extremely precocious, digging their way out of incubation mounds or burrows
without assistance, capable of flight on the day of hatching, and receiving no subsequent
parental assistance.

This third volume in Oxford University Press’ series on bird families of the world provides
a thorough treatment of this fascinating family which is restricted to the Indo-Australia and

812

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Pacific Islands area. The monograph is divided into two parts, the first nine chapters deal
with aspects of the biology and conservation of megapodes; the second part includes ac-
counts of the seven genera and 22 species. Chapter two deals with the rather problematical
taxonomy of the group. Here megapodes are treated as the sister group of all other galli-
formes instead of a sister group of cracids. The authors also conclude that brush-turkey
genus Talegalla is not as closely related to Alectura and Aepypodius as had previously been
considered and this is used as justification for changing the common name of the three
Talegalla species from brush-turkey to talegalla. They describe three new subspecies in the
genus Megapodius in which they include 13 species (the number recognized in various
taxonomic schemes over the years has varied from three to 19). They also changed the name
of Megapodus species from scrubfowl to megapode. I consider both of these name changes
ill-advised. I object to using genus names as common names because, unless you are a
Greek and Latin scholar, they are not as descriptive as English names ought to be. The use
of megapode for a subset of the megapodes (the title of the monograph) provides lots of
room for confusion.

The third chapter deals with megapode distribution, biogeography, and speciation, and
the fourth discusses behavior and demonstrates the link between reproductive behavior and
other behaviors of the megapodes. Chapters 5-8 discuss aspects of the breeding biology of
megapodes including adaptations in embryo physiology and mating strategies. Chapter nine
discusses the grim problems associated with island living, habitat destruction, and production
of large, nutritious eggs in areas overpopulated with humans — particularly in the context of
the breakdown in traditional tribal constraints on over-exploitation of megapode eggs.

The species accounts are thorough, starting with nomenclature and descriptions of plum-
ages, weights and measurements which often include tabular data, range and range maps,
previously unpublished sonagrams to augment vocalization descriptions, behavior and ab-
breviated references (with full references in the References section). I found the abbreviated
reference section at the end of each species account, although redundant, an excellent feature
since it facilitates looking up references for a particular species. The range maps are excel-
lent, and demarcate ranges of subspecies but many contain historical data and hence do not
represent current distribution. Eor some, however, historical information is identified, and
for the Malleefowl Leipoa ocellata present and past distributions are given. The plates are
excellent, depicting immature plumages and chicks where appropriate, and female plumages
in the few dimorphic species. The plates are accompanied by a mini text which is very
helpful when perusing or comparing species.

This is a thorough treatment of a fascinating family of birds. The list of more than 800
references, many of them from the 1980s and 1990s, a period of renewed interest and intense
research on megapodes, includes references from a wide variety of journals in several lan-
guages. Aside from my quibbles about the name changes mentioned above, my only criti-
cism of the book is the level of redundancy, perhaps inevitable in a multi-authored, com-
plicated monograph. We read, for example, that the heat in megapode mounds results from
microbial respiration in at least four places. Nevertheless, this is an excellent monograph,
the product of exhaustive research, that should be in the library of anyone interest in Aus-
tralasian birds or in the range of adaptation in birds. — William E. Davis, Jr.

Wilson Bull., 108(4), 1996, pp. 813-824

PROCEEDINGS OF THE SEVENTY-SEVENTH
ANNUAL MEETING

John A. Smallwood, Secretary

The Seventy-seventh annual meeting of the Wilson Ornithological Society was held
Thursday, 11 April, through Sunday, 14 April, 1996 at the Grand Hotel, Cape May, New
Jersey, in joint session with the New Jersey Audubon Society. The local committee, chaired
by Sheila Lego, was composed of Joan Walsh, Pete Dunne, Tom Parons, Kathy lozzo, Vince
Elia, Bill Seng, Pat Sutton, Marleen Murgitroyde, Fred Mears, Bill Glaser, and Louise
Zemaitis. The meeting was sponsored by the Cape May Bird Observatory, New Jersey
Audubon Society.

The Council met from 13:17 to 19:02 on Thursday, 11 April in the Grand Ballroom of
the Grand Hotel. At that time there were 128 registrants. On Thursday evening there was
an informal reception for the conferees and guests at Blackbeard’s, a libatory and culinary
establishment in the Grand Hotel.

The opening session on Friday convened in the Grand Ballroom at 8:30 with welcoming
remarks from Joan Walsh of the Cape May Bird Observatory, New Jersey Audubon Society,
and from WOS president Keith Bildstein.

The scientific program included 53 contributed papers and 15 contributed posters, which
were organized into four paper sessions, two poster sessions, a symposium on Raptor Mi-
gration and Ecology chaired by John C. Kricher, and a workshop on Teaching Ornithology,
organized by Edward H. Burtt. The workshop also included field demonstrations on record-
ing bird song and on censusing birds. The evening program on Friday featured Pete Dunne,
director of Cape May Bird Observatory, who delivered a well-received presentation entitled
“Small-headed Flycatcher. Seen Yesterday. Didn’t Leave His Name.” This presentation was
followed by the first business meeting of the Wilson Ornithological Society, which in turn
was followed by another informal reception for members and guests at the above-mentioned
Blackbeard’s. Field trips on Friday, Saturday, and Sunday morning included forays to Higbee
Beach and to other local birding hot spots, and the extended field trip on Sunday included
a boat cruise to Back Bay.

The attendees enjoyed a ninety-minute social gathering prior to the annual banquet, which
was held in the Grand Ballroom of the Grand Hotel. After a grand dinner. President Bildstein
delivered a few brief remarks to the Society. The following awards also were presented:

EDWARDS PRIZE (for the best major article in volume 107 of The Wilson Bulletin)

Mary H. Clench and John R. Mathias, “The avian cecum: a review.” Wilson Bull., 107
(1):93-121.

LOUIS AGASSIZ FUERTES AWARD

Paul M. Nealen, “Design and function of the Carolina Wren (Thryolhorus ludovicianus)
song system.”

MARGARET MORSE NICE AWARD

Susan R. Blackshaw, “A study of wintering Loggerhead Shrikes (Lanius ludoviciunu.s)
in Texas and/or Florida.”

813

814

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

PAUL A. STEWART AWARDS

Paul Doherty, “Avian responses to forest fragmentation: determining the economic cost
of maintaining minimum viable populations.”

Andrew Dolby, “Benefits of mixed-species flocking for Downy Woodpeckers and White-
breasted Nuthatches: a removal experiment.”

Kimberley J. Lernie, “Effects of electromagnetic fields generated by powerlines on avian
reproduction and development.”

Joseph J. Nocera, “The effects of predation on the movement of Common Terns {Sterna
hirundo) and Arctic Terns (Sterna paradisaea) in Nova Scotia.”

Liana V. Pravosudova, “The effect of forest fragmentation on social structure of the Tufted
Titmouse, Parus bicolor (Paridae, Aves).”

ROGER TORY PETERSON TRAVEL AWARD

Sara Morris, “Pall songbird migration in Maine: factors affecting the likelihood of stop-
over.”

David J. Ziolkowski, Jr., “Coordination of female nest attentiveness with male song output
in the House Wren.”

ALEXANDER WILSON PRIZE (for best student paper)

Sara Morris, “Pall songbird migration in Maine: factors affecting the likelihood of stop-
over.”

Selection committee for the Edwards Prize-Charles Blem (chair), Tom Haggerty, Bret
Whitney, and Ted Davis; for the Puertes, Nice, and Stewart Awards-Daniel Klem, Jr., chair,
Judith M. Rhymer, and Richard B. Stiehl; for the Peterson Travel Award-John Kricher,
chair, Keith Bildstein, and Ted Davis.

Poliowing the award presentations, the guest speaker, Helen Hayes of the American Mu-
seum of Natural History, gave a wonderful slide presentation titled “Desperately Seeking
Roseates.”

FIRST BUSINESS MEETING

The first business meeting was called to order by President Bildstein at 20:35 on Priday,
12 April, in the Grand Ballroom. Secretary Smallwood presented a synopsis of Thursday’s
Council meeting, and introduced Pirst Vice President Edward H. Burtt, Jr., who discussed
his proposal to establish a lectureship for the annual meetings in honor of Margaret Morse
Nice. Vice President Burtt suggested the presentation could be an opening plenary lecture
or one scheduled at another prominent place in the program. This lectureship would provide
the honorary guest speaker a longer format in which to describe scientific inquiry within
the overall context of a lifetime of research and personal experience. Secretary Smallwood
continued, describing the Council’s interest in encouraging high-quality abstracts from Latin
America to compete for the Wilson Ornithological Society awards. Because many of these
abstracts are written by authors not fluent in English, Jon Barlow generously pledged to
oversee those translations. The Council approved a motion to allow the editor to select a
recipient of a new award to recognize meritorious service to those who contribute to pub-
lishing The Wilson Bulletin. Through the efforts of webmeister Janet Hinshaw, the Wilson
Ornithological Society now has a home page on the Internet. As of 15 March, 1996, mem-
bership in the WOS stood at 2487, including 512 life members. The Council reelected Editor
Blem for another term, and expressed its great appreciation. The Council also thanked Ernest

ANNUAL REPORT

815

T Willoughby and the Undergraduate Outreach Committee for its work on teaching orni-
thology (see the report and committee members, below). In response to the invitation by
Jerome A. Jackson, the WOS Council approved a motion to join and send delegates to the
North American Banding Council, whose purpose is to facilitate an increase in the quality
o data submitted to the Bird Banding Laboratory. No resolutions were received by the
retiring Resolutions Committee during the previous year, and a new Resolutions Committee
shall be appointed by President Bildstein. Next year’s meeting will be hosted by the Division
of Biology at Kansas State University from 17-20 April, 1997, in Manhattan, at the invi-
tation of local chair John L. Zimmerman. The 1998 annual meeting will be held in St Louis
from 24-29 March, hosted by the University of Missouri College at St. Louis, with Bette
Loiselle the local chair. Although not yet finalized, this is being planned as a joint meeting
with the American Ornithologists’ Union, the Cooper Ornithological Society, and possibly
the Raptor Research Foundation and the Colonial Waterbird Society. Council gratefully
accepted the invitation from local chair Herbert T. Hendrickson to host the 1999 annual
meeting at the University of North Carolina at Greensboro. The secretary then asked those
assembled to stand in recognition of the following members who have died since we last
met: Andrew J. Berger (Kailua, HI), Mrs. W. P. Cottrile (Jackson, MI), John Farrand, Jr.
(Summit, NJ), James M. Hartshorne (Ithaca, NY), Thomas A. Imhof (Birmingham, AL),
Herbert W. Kale (Casselberry, FL), C. N. Mason (Washington, DC), Robert A. McCabe
(Madison, WI), John T. Ricks (Huntington, NY), Carol S. Roesler (Darien, CT), V/alter
Spofford (Portal, AZ), Gustav A. Swanson (Ft. Collins, CO), Arthur C. Taylor (Appleton,
WI), and Gerard F. Van Tets (Lyneham, Australia).

The treasurer’s report was then presented by Doris Watt.

Charles Blem presented the editor’s report.

William (Ted) Davis, chair, presented the report of the nominating committee, which also
included Mary H. Clench and Jerome A. Jackson: President, Keith L. Bildstein; First Vice-
president, Edward H. Burtt, Jr.; Second Vice-president, John C. Kricher; Secretary, John A.
Smallwood; Treasurer, Doris J. Watt; Members of Council for 1997-1999, Peter C. Frederick
and Danny J. Ingold.

Vice-president Burtt introduced members to the newly formed Mesoamerican Society for
Biology and Conservation, which will publish a news bulletin and sponsor annual confer-
ences in Mesoamerica.

The meeting was adjourned at 20:58.

SECOND BUSINESS MEETING

The second business meeting was called to order by President Bildstein at 11:07 on
Saturday, 13 April, in Room 127 of the Grand Hotel, at which time he asked Secretary
Smallwood to read aloud the single resolution to be considered:

COMMENDATION

WHEREAS the Wilson Ornithological Society held its annual meeting in scenic Cape
May, New Jersey, at the invitation of the New Jersey Audubon Society,

RECOGNIZING that the Committee on the Scientific Program, under the adept direction
of John C. Kricher, arranged and managed an exemplary schedule of oral and poster pre-
sentations, which included an innovative workshop on teaching ornithology, organized by
Edward H. Burtt, Jr., and

RECOGNIZING that the Committee on Local Arrangements, through the efforts of Sheila
Lego, Joan Walsh, and other members of the New Jersey Audubon Society’s Cape May
Bird Observatory, provided an excellent conference venue with comfortable accommoda-

816

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

tions, interesting guest speakers of national renown, exciting bird watching opportunities,
and, through mechanisms not yet revealed, outstanding local weather, and
WHEREAS the conferees found this meeting informative and enjoyable,

THEREFORE BE IT RESOLVED that the Wilson Ornithological Society commend the
Committee on the Scientific Program, the Committee on Local Arrangements, and the local
sponsors for a most successful and rewarding meeting in Cape May.

The motion to accept this resolution was made by Herbert T. Hendrickson, seconded by
Clait Braun, and enthusiastically passed by acclamation.

The President recalled the report of the Nominating Committee to the floor, asking for
any additional nominations. As there were none, Phillips B. Street moved and Clait Braun
seconded that the nominations be closed, and, by acclamation, it was so. Further, Herbert
T. Hendrickson moved and Richard N. Conner seconded that the candidates for all offices
be elected unanimously, and again by acclamation, this too came to pass.

The President invited David Blockstein, chairman of the Ornithological Council, to ad-
dress the members assembled. Dr. Blockstein reviewed the issues of current interest to the
Ornithological Council, including improvements to the process of issuing federal scientific
permits, the relocation of the National Biological Service (which administers the Bird Band-
ing Laboratory) into the United States Geologic Survey, reauthorization of the Endangered
Species Act, and a revision of the guidelines on the research use of wild birds.

The members in attendance were treated to a slide show by John Zimmerman that pre-
viewed the beautiful Kansan scenes we would behold next year at the 1997 WOS meeting
in Manhattan. This meeting will be the initial site for the Nice Lecture, and a symposium
on grassland birds is being planned. President Bildstein presented an update on the plans to
meet in St. Louis, Missouri, in 1998, and in Greensboro, North Carolina, in 1999.

President Bildstein then called on those assembled for any announcements, but none were
offered. Adjournment occurred at 1 1 :32.

REPORT OF THE TREASURER
1 July 1995 to 30 June 1996

GENERAL EUNDS

Balance Forward

$ 89,713.18

Receipts

Regular and Sustaining Memberships $ 33,226.00

Student Memberships 4,505.00

Family Memberships 176.00

Total Dues

Subscriptions $ 25,919.16

Contributions from Authors for Page Charges 9,549.26

Back Issues 664.00

Total Income from the Publications

Contributions to The Van Tyne Library $ 355.00

Contributions to the Student Membership Endowment 0.00

Contributions to the Wilson Award Endowment 0.00

Contributions to the General Endowment (Life, Patrons) ... 375.00

Contributions to the Roger Tory Peter.son Travel Fund 0.00

Contributions to the General Endowment 498.00

Unrestricted Contributions 458.00

$ 37,907.00

$ 36,132.42

ANNUAL REPORT

817

Total Contributions

Royalties

Interest from Endowments

Interest from Checking Account
Dividends from Dreyfus Account (Reinvested)

List Rental

Miscellaneous

OSNA Adjustment

TOTAL RECEIPTS

Disbursements
Bulletin Publication

June 1995

September 1995

December 1995

March 1996

Editor’s expenses

Total Publication Costs

OSNA Expenses

Secretary’s Expenses

Treasurer’s Expenses

Treasurer’s Bond

The Elock

Editor’s Honorarium

Incorporation Eee

Awards Committee Phone

Meeting Costs/Symposium Speakers’ Airfare ...

Advertisement (Allen Press)

CPA (Tax Piling)

Van Tyne Library (Back Issues)

AAZN Dues

Miscellaneous (Western Union, Refunded Later)

Total Operating Expenses

Organizational Awards

Ornithological Council Contribution

Total Philanthropies

TOTAL DISBURSEMENTS

Transfer to Mellon Account

Ending Balance

$ 19,464.22

16.904.22

22.141.22
19,397.82

5,168.29

$ 13,948.00
79.50
81.63
0.00
259.52
2,000.00
5.00
66.75
2,148.10
0.00

425.00
40.00

100.00
850.00

$ 2,206.00
500.00

$ 1,686.00
42.00
24,820.25
540.73
841.80
1,034.00
1,313.09
-587.99

$103,729.30

$ 83,075.77

$ 20,003.52

$ 2,706.00

$105,785.29
$ 22,813.00
$ 64,844.19

CASH ACCOUNTS

Pirst Source Bank Checking Account 1 April 1996 $ 45,408.95

Dreyfus Liquid Assets 1 March 1996 19,435.24

Total Cash on Hand $ 64,844.19

Van Tyne Library Accounts Starting Balance $ 2,857.19

Receipts S 1,171.10

Expenses 1 , 1 23.42

Ending Balance $ 2,904.87

818

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

DESIGNATED ACCOUNTS

Sutton Color Plate Fund (Endowment Principal $55,727.99)

1995 Balance $ 500.20

1995 Earnings' $ 2,173.39

Funds Disbursed for Color Plates 1995-96 2,654.00

1996 Balance 19.59

TOTAL ENDOWMENT EUNDS

1989 Market Value $386,992.00

1990 Market Value $372,063.00

1991 Market Value $423,698.00

1992 Market Value $430,258.00

1993 Market Value $487,786.00

1994 Market Value $425,155.00

1995 Market Value (June 1995) $555,600.00

1996 Market Value (March 1996) $647,600.00

' Based on 3.9% interest on Mellon account.

editor’s report — 1995

In 1995, 213 manuscripts (112 major papers, 101 short communications) were received
by the Wilson Bulletin 1 editorial office. This is 32 more than in 1994. Of these, approxi-
mately 61% were rejected. The time between receipt of manuscript from the author(s) and
our return of the manuscript with referee comments nearly always has been less than three
months. A few manuscripts required slightly more than four months for a decision. There
is a modest backlog of manuscripts, because we have reset deadlines one month earlier. The
average time between receipt of a manuscript and its appearance in The Wilson Bulletin in
1995 was almost always less than a year, as in past issues. Frontispiece articles sometimes
require longer to appear in print, depending upon backlog of such papers. We are using
e-mail in the editorial process to the degree that some costs have been reduced and the
speed of response for computer users often is more prompt.

I am grateful to the editorial board — Kathy G. Beal, R. N. Conner, Tom Haggerty, and
J. A. Smallwood — for their timely, skilled evaluations of many of the manuscripts submitted
to the journal. Assistant Editors Leann Blem and Albert E. Conway are responsible for the
consistency of style and format, and for making arcane prose more readable. I thank them
for their efforts. Kathy G. Beal deserves special praise for continuing to assemble the index
for The Wilson Bulletin. This is a tedious task and the entire society benefits from her
careful work. Leann Blem has provided much of the manual labor that keeps the editorial
office running and she catches many of the small errors that can plague a publication.
Virginia Commonwealth University Department of Biology supports the editorial process
and the running of the office in numerous ways. As always, I remain open to suggestions
as how to improve the service we provide the readers and authors, and invite you to make
your opinions known to me.

C. R. Blem, Editor

The reports of the standing committees are as follows:

ANNUAL REPORT

819

REPORT OF THE MEMBERSHIP COMMITTEE

The current members of the WOS membership committee are Jim Ingold at Louisiana
State University, Mark Woodrey at the Mississippi Museum of Natural Science, Mary
Clench at the University of Texas, and myself, at Montclair State University in New Jersey.

Mary Clench joined the committee in 1995, offering to contact nonrenewing members to
invite them back into our flock. This proved to be a formidable task: about 350 members
did not renew in 1995, and about 280 in 1994. Mary has been concentrating on the most
recent drop-outs. Mark Woodrey continues to send membership invitations to authors who
publish in The Wilson Bulletin, and Jim Ingold contacts nonmembers who make presenta-
tions at the annual meetings.

Dave Cimprich, who formerly oversaw the itinerary of the WOS membership poster, has
retired. During the past year I brought the poster to the WOS meeting in Williamsburg, the
Raptor Research meeting in Duluth, and the AOU meeting in Cincinnati. I’m not sure how
effective of a recruitment device the poster has been at OSNA-member societal meetings,
where a large percentage of the participants already are familiar with the WOS. The poster
may have a greater impact at regional gatherings, and at meetings of highly motivated bird-
people, such as members of Audubon and the American Birding Association. I am currently
seeking a new membership committee person who will have the time and energy to imple-
ment an aggressive traveling schedule for the poster. Suggestions will be appreciated.

I received during the past year 14 letters from people interested in joining us. I sent to
each a personal letter of welcome and our membership brochure. I recently revised the
brochure; the editorial suggestions forwarded to me by Janet Hinshaw were particularly
helpful. Both the scanned artwork and the brochure’s layout are incorporated into a
WordPerfect© document, so that laser printing produces a camera-ready copy, and future
edits may be made easily. At this writing the new brochures are at the printers. I plan to
bring a supply of them to Cape May.

John A. Smallwood, Chair

REPORT OF THE UNDERGRADUATE OUTREACH COMMITTEE

The Committee on Undergraduate Outreach was established in June, 1991, under Presi-
dent Richard C. Banks, to help stimulate an interest in ornithology among undergraduate
students, and to help maintain and focus that interest so as to stimulate students to continue
studies in ornithology. The Committee has explored ways to increase interest of undergrad-
uate students in ornithology both as a focus for future postgraduate studies, and as an
avocation for those who seek careers in fields other than biological sciences.

Members of the Committee during 1995-1996 were (alphabetically): Albert R. (Jay)
Buckelew, Jr., Bethany College, WV; Edward H. Burtt, Jr., Ohio Wesleyan University, OH;
Danny J. Ingold, Muskingum College, OH; Dale Kennedy, Albion College, MI; John C.
Kricher, Wheaton College, MA; Lynn A. Mahaffy, Randolph-Macon College, VA; Barnaby
Marsh IV, Cornell University, NY; Dan A. Tallman, Northern State University, SD; Ernest
J. Willoughby, St. Mary’s College of Maryland; W. Herbert Wilson, Colby College, ME.
Of these members, Buckelew was appointed to the Committee by President Keith Bildstein
in October, and Ingold resigned in February.

The Committee has completed its analysis of the returns on its questionnaire distributed
to teachers of college undergraduates in November, 1993, for the purpose of helping us to
decide what to do to promote education in ornithology. A copy of the complete analysis
has been submitted to the secretary. The Committee will be discussing these results in its
sessions at the annual meeting.

820

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

The Committee decided last year to organize an ornithology teaching workshop for this
year’s annual meeting. Dr. Burtt and Dr. Kricher took charge of organizing and scheduling
that workshop in the program.

Work on the Committee’s guide to graduate degree programs in ornithology continues.
The process of organizing and keyboarding the information from many institutions that
responded to our request has taken longer than we supposed it would. At the same time,
the Committee has received several requests for copies when they become available.

Dan Tallman continues to supervise our electronic information exchange.

The Committee is working on a project to produce a series of videotape interviews with
eminent ornithologists. Such tapes could be lent through the WOS library. Lynn Mahaffy
is chairing a subcommittee to draft a series of questions for such interviews.

Ernest J. Willoughby, Chair

REPORT OF THE JOSSELYN VAN TYNE MEMORIAL
LIBRARY COMMITTEE

Once again at year’s end, thanks in large part to the constant hard work of Janet Hinshaw
and her helpers (operating from the Bird Division Library, University of Michigan Museum
of Zoology, our WOS Library’s “home base”), we can report satisfactory progress and
many worthwhile services, both to our membership and to the science of ornithology. Our
special thanks again to Pat Ahrens, secretary for the Bird Division.

Donations to the library during the year included 30 books, 57 reprints, 241 journal issues,
19 reports, and two microfilm dissertations; the reprints and journal issues, for whatever
reasons, show a sharp drop-off from last year. The number of members and institutions
contributing, however, remained about the same, 23, namely: S. Conant, S. Emslie, A.
Feduccia, A. Gaunt, S. Goodman, G. Hall {The Wilson Bulletin review copies), J. Hinshaw,
J. Jackson, L. Kiff (for The Peregrine Fund), N. Klein, S. Latta, F. Lohrer, P. Lowther, H.
Mayfield, H. McClure, R. Payne, D. Pence, T. Root, J. Ryder, C. Shipman, J. Spendelow,
Station Ornithologique Suisse, and P. Street. One can’t help noting, after many years, the
absence of the late Andy Berger’s name as a leading donor.

Gifts to other institutions were as follows: 32 journal issues to Grand Valley State Uni-
versity, 52 journal issues to Hawk Mountain Sanctuary, 128 journal issues to The Peregrine
Fund, four journal issues to Station Ornithologique Suisse, and 83 journal issues to the
Zoological Museum, University of Moscow.

Loan transactions totaled 98, to 60 people, the almost twofold increase being partly due
to our providing copies to various individuals working on Birds of North America ac-
counts— certainly justifiable even if our rules had to be relaxed a bit.

The total of publications regularly received in our library, 290 titles from 199 organiza-
tions, comprise the following: 127 exchanges ( 192 journals, books, and reprints), 51 gifts
(67 publications), and 21 subscriptions (31 titles). These figures show a continuing upward
trend.

During the year, we sold 52 books for $2535.05 ($1022 in cash; $1513.05 in trade for
credit with Buteo Books), and 72 journal issues for $132.00: a total of $2677.05. These
items are from the accumulating backlog of member contributions which, though surplus
duplicates, have a very large cash value.

This cash accrues to replenish our New Book Fund. In this past year, a total of $1414.55
was spent from this fund: $701.38 for 36 books, monographs, tapes, and records; $292.17
for back issues of journals; and $421.00 for 12 journal subscriptions. Such purchases are
adding very substantially to the overall value of our holdings.

ANNUAL REPORT

821

One could hardly look at our figures and not be very much encouraged by the general
picture. All we can do is once more thank the many who have been involved, while urging
more members to use our library and to contribute what they can to its support.

We would encourage everyone to view our web page, which can be found at http://
www.ummz.lsa.umich.edu/birds/wos.html. There is a link to the University of Michigan’s
on-hne library catalogue (MIRLYN) which provides listings of the catalogued books and
journals in the Van Tyne library.

William A. Lunk, Chair

PAPER SESSIONS

Richard C. Banks, National Biological Service, National Museum of Natural History, Wash-
ington, DC, “The name of the Lawrence’s Flycatcher.”

Jon C. Barlow and G. Cooke, Dept, of Ornithology, Royal Ontario Museum, Toronto, ON,
Canada, Variation in song in circum-Caribbean peppershrikes: subspecies identity of the
Isla Margarita population.”

Mark Fink, Dept, of Wildlife and Fisheries Sciences, Texas A&M Univ., College Station,
TX, “Effect of edge on nest predation within Golden-cheeked Warbler habitat.”

Jeanette Bider, Univ. of Arkansas, Fayetteville, AR, “Use of microhabitats by woodcreepers
(Dendrocolaptidae) and flycatchers (Tyrannidae) in tropical broadleaf forest remnants.”

C. T. Baril, Dept, of Zoology, Univ of Toronto, and Dept, of Ornithology, Royal Ontario
Museum, Toronto, ON, Canada, “Geographic variation in Vireo huttoni territorial song.”

Leonard Reitsma, Benjamin Steele, Sherman Burson, and Peter Hunt, New England Institute
for Landscape Ecology, “Habitat selection and socioecology of Louisiana (Seiurus mo-
tacilla) and Northern Waterthrushes (S. novaboracensis) overwintering in Puerto Rico,
West Indies.”

Daniel S. McGeen, Auburn Hills, MI, “Atoms, ions, and the warbler.”

J. M. Utter, B. Farrell, F. Arengo, R. Drummond, C. Lindner, C. Smith, and C. Safina,
Purchase College, SUNY, Purchase, NY, “Overwinter decline in multiflora rose fruit on
mockingbird territories.”

Eric C. Atkinson, Hawk Mountain Sanctuary Association, Kempton, PA, Pam Dugger and
Christina Swindall, Boise State Univ., Boise, ID, “Year-round habitat association of
Black-billed Magpies, Horned Larks, and Western Meadowlarks on the Snake River Birds
of Prey National Conservation Area.”

Barbara J. Bowen, Dept, of Biology, Central Connecticut State Univ., New Britain, CT,
“Comparisons of predation rates on artificial and natural avian nests in open, edge, and
forested habitats.”

Catherine M. Devlin, Rutgers Univ., New Brunswick, NJ, “The Eastern Wild Turkey (Mele-
agris gallopavo silvestris) in the southern pinelands of New Jersey.”

Winli Lin, Univ. of San Diego, San Diego, CA, “Parental behavior and possible brood
division in Whimbrel (Numenius phaeopiis).”

Sylvia L. Halkin, Dept, of Biological Sciences, Central Connecticut State Univ., New Brit-
ain, CT, “Recording and analysis of bird vocalizations.”

W. Herbert Wilson, Colby College, Waterville, ME, “Teaching the fundamentals of bird
vocalizations: a laboratory on computer-aided analysis.”

Alex Dowling, Dept, of Wildlife and Fisheries Sciences, Texas A&M Univ., College Station,
TX, “Impacts of habitat fragmentation on Prothonotary Warblers in east Texas.”

Mark R. Ryan, School of Natural Resources, Univ. of Missouri, Columbia, MO, “Teaching
optimal foraging behavior: an experiment on group vs. solitary foraging.

822

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Ernest J. Willoughby, St. Mary’s College of Maryland, St. Mary’s City, MD, “Field study
of migratory behavior.’’

Edward H. Burtt, Jr., Dept, of Zoology, Ohio Wesleyan Univ., Delaware, OH, “Teaching
identification of birds through censusing habitats that differ in the extent of human im-
pact.”

Doris J. Watt, Biology Dept., Saint Mary’s College, Notre Dame, IN, “A laboratory exercise
in armchair biology.”

Bruce G. Peterjohn, John R. Sauer, and William A. Link, National Biological Service,
Patuxent Wildlife Research Center, Laurel, MD, “Mapping change in bird distributions
from breeding bird survey data.”

John R. Sauer, William A. Link, and Bruce G. Peterjohn, National Biological Service,
Patuxent Wildlife Research Center, Laurel, MD, “Estimating population change from the
North American Breeding Bird Survey.”

Andrew B. T. Smith, Univ. of Ontario, ON, Canada, “Changes in the abundance of forest
birds in Algonquin Park: 1952—1995.”

Kurtis L. Dean and David L. Swanson, Univ of South Dakota, Vermillion, SD, “Seasonal
variation in density and diversity of neotropical migrants at stopover sites in the northern
great plains.”

Deanna K. Dawson, Patuxent Wildlife Research Center, Laurel, MD, and Lonnie J. Darr,
Montgomery County Dept, of Environmental Protection, Rockville, MD, “Land-use plan-
ning for area-sensitive forest birds.”

Eric T. Liknes, Kurtis L. Dean, and David L. Swanson, Dept, of Biology, Univ. of South
Dakota, Vermillion, SD, “Differential timing of migration of sex/age classes in Ruby-
crowned Kinglets.”

Joseph R. Jehl, Jr., Hubbs-Sea World Research Institute, San Diego, CA, “Leaving town:
how Eared Grebes prepare to migrate.”

Ronnie E. Stout, North Dakota State Univ., Fargo, ND, “Fall Red-necked Grebe migration
behavior and molt in the Great Lakes region.”

Chao-Chieh Chen and R. B. Hamilton, Louisiana State Univ., Baton Rouge, LA, “Analysis
of searching movements of insectivorous migratory songbirds on the Chenier Plain of the
Gulf Coast.”

Sara R. Morris, Cornell Univ., Ithaca, NY, “Fall songbird migration in Maine: factors af-
fecting the likelihood of stopover.”

Jeffrey P. Dugay and Petra S. Wood, West Virginia Cooperative Fish and Wildlife Research
Unit, NBS, West Virginia Univ., Morgantown, WV, “The management effects of clearcuts
and two-age timber harvests on nongame birds in West Virginia.”

John A. Smallwood, Dept, of Biology, Montclair State Univ., Upper Montclair, NJ, and
Peter D. Smallwood, Dept, of Biology, Univ. of Pennsylvania, Philadelphia, PA, “Decoy
holes and monitoring rates: applied conservation biology for secondary cavity nesting
species.”

Bradley D. Ross and Richard H. Yahner, Dept, of Ecology, Pennsylvania State Univ., Uni-
versity Park, PA, “Integrating habitat and bird variables using geographic information
systems and extensive wildlife data sets.”

Bruce W. Baker, Brian S. Cade, Warren L. Mangus, National Biological Service, Fort Col-
lins, CO, Janet L. McMillen, Dept, of Zoology, Ohio State Univ., Columbus, OH, and E
Joshua Dein, National Biological Service, Madison, WI, “Multi-scale evaluation of a
suitability model for Sandhill Crane nesting habitat.”

Robert W. Russell, National Oceanographic and Atmospheric Association, Seattle, WA,
“Boundary-layer convergence lines: aerial corridors for soaring birds.”

ANNUAL REPORT

823

B. G. Murray, Jr., Dept, of Biological Sciences, Rutgers Univ., Piscataway, NJ, “On the
meaning of ‘reproductive success’ and other terms.”

Peter W. C. Paton, Univ. of Rhode Island, Kingston, RI, and Ron Flores, U.S. Fish and
Wildlife Service, Ninigret NWR, Charlestown, RI, “Effects of the North Cape barge oil
spill on the distribution of birds in coastal Rhode Island.”

Richard N. Conner, D. Craig Rudolph, Southern Research Station, U.S. Forest Service,
Nacogdoches, TX, and Robert N. Coulson, Dept, of Entomology, Texas A&M Univ.,
College Station, TX, Relationships between southern pine beetle population level and
bark-beetle caused mortality of Red-cockaded Woodpecker cavity trees.”

D. Craig Rudolph and Richard N. Conner, Southern Research Station, U.S. Forest Service,
Nacogdoches, TX, Red-cockaded Woodpeckers, ecosystem management, and silvicul-
ture.”

Douglas W. White, Albion College, Albion, MI, “Effect of egg burial on nest vandalism
by House Wrens.”

L. Scott Johnson, David Ziolkowski, Dept, of Biology, Towson State Univ., Towson, MD,
Krishna Hannam, and William A. Searcy, Dept, of Biology, Univ. of Miami, Miami, FL,
“Coordination of female nest attentiveness with male song output in the House Wren.”

E. Dale Kennedy and Douglas W. White, Albion College, Albion, MI, “Competition be-
tween Bewicks Wrens and House Wrens.”

Gregory S. Keller, School of Forest Resources, Pennsylvania State Univ., University Park,
PA, “Seasonal distribution and community composition of avifauna in isolated deciduous
forest patches.”

Ernest J. Willoughby and Christopher J. Lindsay, St. Mary’s College of Maryland, St. Mary’s
City, MD, “Significance of the odd prenuptial molt of American Goldfinch.”

Todd J. Underwood and Roland R. Roth, Dept, of Entomology and Applied Ecology, Univ.
of Delaware, Newark, DE, “Estimating Wood Thrush population demographics by con-
stant effort mistnetting.”

Jake Stein, Univ. of Tennessee at Martin, Martin, TN, “Effects of sex, time of day, and age
on Carolina Chickadee weights.”

David M. Whalen, Center for Conservation Biology, College of William & Mary, Williams-
burg, VA, “Fall migration of Northern Saw-whet Owls at the southern tip of the Delmarva
Peninsula.”

Keith L. Bildstein and Laurie J. Goodrich, Hawk Mountain Sanctuary, Kempton, PA, “Link-
ing raptor migration watch-site counts to mainstream ecology and conservation: 60 years
of science at Hawk Mountain Sanctuary.”

Paul Kerlinger, New York, NY, “Economics of hawk watching.”

David F. Brinker, Maryland Dept, of Natural Resources, Annapolis, MD, “Migratory move-
ments of Northern Saw-whet Owls: what do we really know?”

Jacques Ibarzabal, Ecole de la mer des Jeunes Explos., Beauport, PQ, Canada, and Jean-
Pierre L. Savard, Canadian Wildlife Service, Sainte-Foy, PQ, Canada, “Migration mon-
itoring in Tadoussac, Quebec.”

Robert W. Russell, National Oceanic and Atmospheric Administration, Seattle, WA, “Water-
crossing tactics of migrating hawks at Cape May Point.”

Richard P. Gerhardt, Miguel Angel Vasquez, and Paula M. Harris, The Peregrine Fund, Inc.,
Boise, ID, “Observations and food habits of nesting Great Black Hawks in El Peten,
Guatemala.”

POSTERS

J. A. Dick, R. D. James, and J. C. Barlow, Dept, of Ornithology, Royal Ontario Mu.seum,
Toronto, ON, Canada, “Current status of transient Rufous Hummingbirds in Ontario.”

824

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Alix Dowling, Dept, of Wildlife and Eisheries Sciences, Texas A&M Univ., College Station,
TX, “Impacts of habitat fragmentation on Prothonotary Warblers in East Texas.”

Jessica R. Eberhard, Princeton Univ., Princeton, NJ, “The evolution of nest-building in
parrots.”

H. T. Hendrickson, Univ. of North Carolina at Greensboro, Greensboro, NC, “History of
winter populations of Purple Pinch in eastern North America.”

Peter Hunt, Mascoma Lake Bird Observatory, Enfield, NH, “Winter distribution of the sex
and age classes of the Yellow-rumped Warbler.”

Michael J. Justice, Dept, of Psychology, Dominican College, Orangeburg, NY, and Teresa
C. Justice, Dept, of Psychology, Iona College, New Rochelle, NY, “Random mating by
size in Northern Mockingbirds.”

T. Greg King, Mark A. Howell, Brian R. Chapman, Karl V. Miller, and Sandra S. Chapman,
School of Forest Resources, Univ. of Georgia, Athens, GA, “The effects of growing-
season vs. dormant-season prescribed fire on neotropical migrant birds.”

Brook Lauro, St. John’s Univ., Jamaica, NY, “The foraging ecology of Sooty Oystercatchers
(Haematopus fuliginosus) nesting at rocky shores in Australia.”

Lynn A. Mahaffy and Kendall Malone, Randolph-Macon College, Ashland, VA, “Some
effects of mosquito control on submerged aquatic vegetation utilized by waterfowl.”

R. McLain, Dept, of Entomology and Applied Ecology, M. Parcells, Dept, of Animal and
Food Sciences, and R. Roth, Dept, of Entomology and Applied Ecology, Univ. of Dela-
ware, Newark, DE, “Determining sex of Wood Thrush by flow cytometry.”

John O’Reilly and Roland R. Roth, Univ. of Delaware, Newark, DE, “Within-season
changes in nest location and mate by female Wood Thrushes.”

Jeanette Rilling, Frederick Terranova, and Terry L. Master, East Stroudsburg Univ., East
Stroudsburg, PA, “Habitat selection and population status of the Louisiana Waterthrush
(Seiurus motacilla) in the Delaware Water Gap National Recreation Area.”

Michel Robert and Pierre Laport, Canadian Wildlife Service, Sainte-Foy, PQ, Canada, “Yel-
low Rail distribution, habitat and conservation along the St. Lawrence River, southern
Quebec.”

James A. Sedgwick, National Biological Service, Fort Collins, CO, “Some aspects of Wil-
low Flycatcher population dynamics.”

Jeffrey A. Spendelow, Patuxent Environmental Science Center, NBS, Laurel, MD, James
M. Zingo, Massachusetts Cooperative Fish and Wildlife Research Unit, NBS, Univ. of
Massachusetts, Amherst, MA, David A. Shealer, Dept, of Biology, Rutgers Univ., Pis-
cataway, NJ, and Grey W. Pendleton, Patuxent Environmental Science Center, NBS, Lau-
rel, MD, “Growth and fledging of Roseate Terns in exceptionally ‘good’ and ‘poor’ years
of overall productivity.”

Wilson Bull.. 108(4), 1996, pp. 825-847

INDEX TO VOLUME 108, 1996

By Kathleen G. Beal

This index includes references to genera, species, authors, and key words or terms. In
addition to avian species, references are made to the scientific names of all vertebrates
mentioned within the volume and other taxa mentioned prominently in the text. Common

names are as they appear in the volume unless otherwise specified. Reference is made to
books reviewed, and announcements as they appear in the volume.

abundance

effects of field age on, 760-770
in riparian zones in the Cascade Moun-
tains, Oregon, 280-291
Accipiter cooperii, 236
Acrobatornis fonsecai gen. nov., sp. nov.,
397-433

(Frontispiece), 434-448
Acrocephalus arundinaceus, 513
luscinia, 246-267
Actitis macularia, 191, 535, 537
addendum, 204
Aegolius acadicus, 123-128
funereus, 127

Aerodramus vanikorensis, 247, 258
Agelaius [assimilis], 372-374
phoeniceus, 184, 372-374, 543, 573-583,
687, 689, 764, 766, 768
Agosia chrysogaster, 377
Aimophila aestivalis, 190, 474
Aix sponsa, 62, 69, 701
Akekee, see Loxops caeruleirostris
Akepa, see Loxops coccineus
Akiapolaau, see Hemignathus munroi
Akikiki, see Oreomystis bairdi
Alauahio, Maui, see Paroreomyza montana
Alces alces, 171

Alden, Peter, review by, 390—391, 391-393
Allen, Paul E., Breeding biology and natural
history of the Bahama Swallow, 480—
495

Alophoixus phaeocephalus, 171
Amakihi, Common, see Hemignathus virens
Kauai, see Hemignathus stejnegeri
Amazila rutila, 168
Amazilia versicolor, 9
violiceps, 228-245
Amazona vittata, 159—163, 164—166
Amblycercys holosericeus, 225

Anas acuta, 69
clypeata, 187
crecca, 187
cyanoptera, 187
discors, 145
fulvigula, 187-189
platyrhynchos, 68, 69, 187, 557, 766
rubripes, 187
strepera, 145, 187

Andrews, Brenda J., Marie Sullivan, and J.
David Hoerath, Vermilion Flycatcher
and Black Phoebe feeding on fish,
377-378

Ani, Groove-billed, see Crotophaga sulciros-
tris

Anianiau, see Hemignathus parvus
Ankney, C. Davison, see Leafloor, James O.,

John E. Thompson, and

announcement

The atlas of southern African birds, pub-
lication information, 603
Anolis sp., 176

Anthony, Robert G., Gregory A. Green, Eric
D. Forsman, and S. Kim Nelson, Avi-
an abundance in riparian zones of
three forest types in the Cascade
Mountains, Oregon, 280-291
Anthracothorax nigricollis, 9
Antshrike, Barred, see Thamnophilus dolia-
tus

Anumbius annumbi, 375, 447
spp., 447

Anous stolidus, 317-334
tenuirostris, 332
An.ser albifrons, 154-159
anser, 155, 369

Apapane, see Himatione sanguinea
Aphelocoma californica, 712-727
ultramarina, 236, 724

825

826

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Aplonis opaca, 246—267
Aptenodytes patagonicus, 78
Aquila adalberti, 354
Archilochus alexandri, 228—245, 542
colubris, 168, 498
Ardea cocoi, 526

herodias occidentalis, 342, 354, 355
Arenaria interpres, 330, 783-796
Arendt, Wayne J., see Meyers, J. Michael,

, and Gerald D. Lindsey

Arnold, Keith A., see Dickinson, Vanessa

M., and

Asio otus, 127
Asthenes anthoides, 420
baeri, 397-433
dorbignyi, 397-433
luizae, 447

patagonica, 397-433, 445
spp., 397-433, 434-448
Atlapetes rufinucha, 223
Atwood, Jonathan L., Christopher C. Rim-
mer, Kent P. McLarland, Sophia H.
Tsai, and Laura R. Nagy, Distribution
of Bicknell’s Thrush in New England
and New York, 650—661
Atwood, Jonathan L., see Rimmer, Christo-
pher C., , Kent P. McEarland,

and Laura R. Nagy
awards and grants

Biological Research Station of the Ed-
mund Niles Huyck Preserve, 603
North American Bluebird Society, 602
Aythya affinis, 380-381, 556-566
ferina, 556, 563
fuligula, 561, 563
marila, 380-381, 563

Baker, Myron C., Female buntings from hy-
bridizing populations- prefer conspe-
cific males, 771-775

Baltosser, William H., Nest attentiveness in
hummingbirds, 228-245
Bananaquit, see Coereba flaveola
Barrow, Mark V., Jr., review by, 200-201
Barth, Robert H., Jr., see Whitney, Bret M.,
Jo.se Fernando Pacheco, Paulo Sergio

Moreira da Fonseca, and

Bartramia longicauda, 783-796
bear, black, see Ursus americanus
Becard, Black-capped, see Pachyramphus
marginatus

Chestnut-crowned, see Pachyramphus
castaneus

Rose-throated, see Pachyramphus aglaiae
Bedell, Paul A., Evidence of dual breeding
ranges for the Sedge Wren in the cen-
tral great plains, 115-122
behavior
avoidance

of cabbage fields by Chen caerulescens,
369-371

breeding

cooperation in Wilsonia citrina, 382-
384

of Charadrius montanus in Colorado,
28-35

of Dendroica cerulea, 673-684
caching

in wintering Melanerpes erythrocephal-
us, 740-747

courtship

of Dendroica chrysoparia, 591—592
drinking

by Pinicola enucleator, 186—187
feeding

of radio tagged Aythya affinis, 556-566

on fish by Pyrocephalus rubinus and
Sayornis nigricans, 377-378

foraging

of sympatric Chen caerulescens, Anser
albifrons, and Branta canadensis,
154-159

use of coastal agricultural fields by
shorebirds, 783-796
mate choice

of female Passerina cyanea and Passer-
ina amoena in hybridizing pop-
ulation, 771-775

nesting

of Melamprosops phaeosoma, 620—638
singing

effect of removal of mate on female
Cardinalis cardinalis, 550-555

vocal

of Catharus bicknelli, 639—649
Best, Louis B., see Gionfriddo, James P, and

Best, Louis B., .see Mitchell, Mary Crowe,

, and James P. Gionfriddo

Bevier, Louis R., (ed.). The atlas of breeding
birds of Connecticut, 595-596

INDEX TO VOLUME 108

827

Bildstein, Keith L., reviews by, 593-594
594

Bittern, Yellow, see Ixobrychus sinensis
Blackbird, Brewer’s, Red-winged, see Age-
laius phoeniceus

Bleier, William J., see Homan, H. Jeffrey,

George M. Linz, , and Robert

B. Carlson

Blem, C. R., reviews by, 197, 198, 393, 809
Block, William M., Michael L. Morrison,
and M. Hildegard Reiser, (eds.), The
Northern Goshawk: ecology and
management, reviewed, 593-594
Bobolink, see Dolichonyx oryzivorus
Bobwhite, Northern, see Colinus virginianus
Bodsworth, Fred, Last of the curlews, re-
viewed, 393—394
body mass

changes during nesting and brood rearing
in female Bucephala clangula, 61—
71

of fall migrant Clangula hyemalis, 567-
572

Bombus sp., 236

Bombycilla cedrorum, 126, 186, 381-382,
687, 688, 691
garrulus, 125, 186
spp., 125

Bonasa umbellus, 129-136, 733
Bornschein, Marcos R., and Bianca L. Rei-
nert, The Andean Flamingo (Phoeni-
coparrus andinus) in Brazil, 807-808
Bos taurus, 236

Bradley, Patricia, Birds of the Cayman Is-
lands, reviewed, 198
Branta canadensis, 154-159
leucopsis, 155

Brauning, Daniel W., review by, 596—597
breeding
biology

of Anous stolidus in Hawaii, 317-334

of Jabiru mycteria in Venezuela, 524-
534

of Parus montanus in northeastern Si-
beria, 80-93

of Polyborus [Caracara] plancus, 516—
523

of Tachycineta cyaneoviridis, 480—495
cooperative

in Wilsonia citrina, 382-384

range

evidence of dual distribution for Cisto-
thorus plaatensis, 115-122
Brodhead, Michael J., see Tomer, John S.,
and

Bronson, C. L., see Doherty, Paul E, Jr.,

Thomas C. Grubb, Jr., and .

Brumfield, Robb T, and J. V. Remsen, Jr.,
Geographic variation and species lim-
its in Cinnycerthia wrens of the An-
des, 205-227
Bubo virginianus, 127
Bubulcus ibis, 171, 342
Bucephala clangula, 61-71
Buckelew, Albert, Jr., and George A. Hall,
The West Virginia breeding bird at-
las, reviewed, 596—597
Bulbul, Yellow-bellied, see Alophoixus
phaeocephalus
bumblebee, see Bombus sp.

Bunting, Indigo, see Passerina cyanea
Lazuli, see Passerina amoena
Painted, see Passerina ciris
Buphagus africanus, 170
erythrorhynchus, 170

Burt, D. Brent, Habitat-use patterns in co-
operative and non-cooperative breed-
ing birds: testing predictions with
Western Scrub-Jays, 712-727
Buteo albicaudatus, 521
jamaicensis, 357—368
lineatus, 357-368, 681
playpterus, 498

Butler, Robert W., Francisco S. Delgado,
Horacio de la Cueva, Victor Pulido,
and Brett K. Sandercock, Migration
routes of the Western Sandpiper, 662—
672

Cacholote, Brown, see Pseudoseisura lopho-
tes

Rufous, .see Pseudoseisura cristata
Cacique, Yellow-billed, see Amblycercus
holosericeus
Calidris alba, 191
alpina, 783-796
bairdii, 788
canutus, 788
fuscicollis, 634
mauri, 662—672, 788, 792
melanotos, 788, 790, 792

828

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

minutilla, 786, 788
pusilla, 190, 191, 783-796
Calocitta spp., 724
Calothorax lucifer, 228
Calypte anna, 241
costae, 228, 241

camel, see Camelus dromedarius

Camelus dromedarius, 171

Campa, Henry, III, see Millenbah, Kelly E,

Scott R. Winterstein, , Ly T.

Eurrow, and Richard B. Minnis
Campy lopterus hemileucurus, 241
Campy lorhynchus rufinucha, 168
Canastero, Austral, see Asthenes anthoides
Cipo, see Asthenes luizae
Creamy-hreasted, see Asthenes dorhignyi
Patagonian, see Asthenes patagonica
Short-hilled, see Asthenes haeri
Canis familiaris, 295
latrans, 156, 292, 299
capybara, see Hydrochaeris hydrochaeris
Caracara, Black, see Daptrius ater

Crested, see Polyborus [Caracara] plancus
Red-throated, see Daptrius americanus
Yellow-headed, see Milvago chimachima
caracara, see Milvago spp.

Cardinal, Northern, see Cardinalis cardinalis
Cardinalis cardinalis, 314, 550—555, 573-
583, 689

Carduelis pinus, 125, 186, 282, 283, 289
tristis, 125, 182-186

Carib, Purple-throated, see Eulampis jugu-
laris

Carloss, Michael, see Johnson, William R,

Frank C. Rohwer, and

Carlson, Robert B., see Homan, H. Jeffrey,
George M. Linz, William J. Bleier,
and

Carpodacus mexicanus, 125
Carrie, N. Ross, Swainson’s Warblers nest-
ing in early serai pine forests in east
Texas, 802-804
Casmerodius albus, 526
Cassie, Brian E., reviews by, 810-81 1,811
cat, civet, see Civettictis civetta

domestic, see Felis domesticus, Felis cat-
tus

Catbird, Gray, see Dumetella carolinensis
Catchpole, C. K., and P J. B. Slater, Bird

song: biological themes and varia-
tions, reviewed, 600-601
Catharus bicknelli, 639-649, 650-661
fuscescens, 190, 542

guttatus [Hylocichla guttata], 53, 313,
542, 688
minimus, 639

ustulatus, 168, 280-291, 542
Catoptrophorus semipalmatus, 783-796
cattle, domestic, see Bos taurus
cavity

use of artificial snags by woodpeckers,
449-456
census methods

for Catharus bicknelli, 639—649
for populations of Limnothlypis swain-
sonii in Jamaica, 94-103
Centeno, Marco V., see Siegel, Rodney B.,
and

Certhia americana, 280—291
Chaetura pelagica, 542
Chantler, Phil, and Gerald Driessens, Swifts,
reviewed, 198

Chapman, Brian R., see Kilgo, John C.,

Robert A. Sargent, , and Karl

V. Miller

Chapman, Brian R., see Moorman, Christo-
pher E., and

Charadrius alexandrinus, 292-301
montanus, 28-35
semipalmatus, 191, 783-796
vociferus, 191, 688, 783-796
Chat, Yellow-breasted, see Icteria virens
Chelydra serpentina, 190-192
chemoreception

in Sturnus vulgaris, 36—52
Chen caerulescens, 154-159, 369—371
canagica, 155
rossii, 154

Chickadee, Black-capped, see Parus atricap-
illus

Carolina, see Parus carolinensis
Chestnut-backed, see Parus rufescens
Mountain, see Parus gambeli
chicken, see Gallus gallus
Chiffchaff, see Phylloscopus collybita
chipmunk, eastern, see Tamias striatus
Chlorostilbon alice, 2
aureoventris, 2, 4
mellisugus, 1—27

INDEX TO VOLUME 108

829

olivaresi sp. nov., 1-27 (Erontispiece)
poortmanni, 2
ricordii, 18
stenura, 2

Cholorceryle amazona, 168

Chondestes grammacus, 689

Christie, David A., see Winkler, Hans,

, and David Nurney

Chrysemys picta, 191
Ciccaba virgata, 780
Ciconia maguari, 526
Cinnycerthia [fulva], 205-227
[olivascens], 205-227
peruana, 205-227
unirufa, 218
Circus cyaneus, 766
Ciridops anna, 616
Cistothorus palustris, 125
platensis, 115-122, 765, 766
Civettictis civetta, 369
Clangula hyemalis, 567-572
Clark, Larry see Mason, J. Russell, and

Clark, Larry, Trigeminal repellents do not
promote conditioned odor avoidance
in European Starlings, 36—52
Cleptornis marchei, 246—267
coachwhip, western, see Masticophis flagel-
lum testaceus

coati, white-nosed, see Nasua narica
Coccothraustes [Hesperiphona] vespertinus,
125, 236, 280-291
Coccyzus americanus, 498, 688
Coereba flaveola, 412
Colaptes auratus, 125, 126, 542
Colinus virginianus, 129-136, 687, 688,
693, 733, 768
colony-site

use by Quiscalus quiscula, 104—1 14
Coluber constrictor flaviventris, 314
Columba leucocephala, 354
livia, 688

Columbina passerina, 168, 573-583
talpacoti, 168
community

breeding Neotropical migrants in riparian
forests of varying widths, 496—506
seasonal populations in Micronesia, 246—
267

competition

effect of Glaucomys volans on nesting
success in Picoides borealis, 697-
711

Conebill, Tamurugo, see Conirostrum ta-
marugense

Conirostrum cinereum, 268, 272
tamarugense, 268-279
Conner, Richard N., and Daniel Saenz,
Woodpecker excavation and use of
cavities in polystyrene snags, 449-
456

Conner, Richard N., D. Craig Rudolph, Dan-
iel Saenz, and Richard R. Schaefer,
Red-cockaded Woodpecker nesting
success, forest structure, and southern
flying squirrels in Texas, 697-7 1 1
Conner, Richard N., review by, 389
conservation

effects of field age on abundance, diver-
sity, productivity, 760—770
of Conirostrum tamarugense in northern
Chile, 268-279

Contopus borealis, 282, 283, 287
sordidulus, 378-380
sp., 168

Virens, 379, 498, 503, 542, 678, 749
Cooper, Jerry A., Birdfinder: a birder’s guide
to planning North American trips, re-
viewed, 598

Coot, Eurasian [European], see Fulica atra
Coquette, White-crested, see Lophornis
adorabilis
Corvus albus, 328
brachyrhynchos, 129—136, 462, 688
corax, 282, 283, 284
ossifragus, 573—583
Coturnicops novoboracensis, 119
Coturnix coturnix, 129-136, 294, 535—539
Cowbird, Brown-headed, see Molothrus ater
Giant, .see Scaphidura oryzivora
coyote, see Canis latrans
Craig, Robert J., Seasonal population sur-
veys and natural history of a Micro-
nesian bird community, 246—267
Crane, Whooping, see Grus americana
Cranioleuca albiceps, 223-224, 397-433
curtata, 401, 418
gutturata, 418
marcapatae, 223

830

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

meulleri, 446
pallida, 401, 412

pyrrhophia, 401, 402, 418, 445, 446
spp., 397-433, 434-448
sulphurifera, 417
vulpina, 418

Creeper, Brown, see Certhia americana
Hawaii, see Oreomystis mana
Crotophaga sulcirostris, 168
Crow, American [Common], see Corvus
brachyrhynchos
Fish, see Corvus ossifragus
Pied, see Corvus albus
Cuckoo, Striped, see Tapera naevia

Yellow-billed, see Coccyzus americanus
Curson, David R., Christopher B. Goguen,
and Nancy E. Mathers, Nest-site re-
use in the Western Wood-Pewee,
378-380

Custer, Christine M., Thomas W Custer, and
Daniel W. Sparks, Radio telemetry
documents 24-hour feeding activity
of wintering Lesser Scaup, 556—566
Custer, Thomas W., see Custer, Christine M.,

, and Daniel W. Sparks

Cutler, Norma G., see McIntyre, Judith W,
and

Cyanocitta cristata, 130, 133, 462, 681, 688,
724, 743

stelleri, 282, 283, 284, 724
Cyclarhis gujanensis, 412
Cynanthus latirostris, 228-245
Cyr, Andre, see Dunn, Erica H., Jacques
Larivee, and

da Fonseca, Paulo Sergio Moreira, see Whit-
ney, Bret M., Jose Fernando Pacheco,

, and Robert H. Barth, Jr.

dace, longfin, see Agosia chrysogaster
Dacnis, Blue, see Dacnis cayana
Dacnis cayana, 412

Dallman, Matthew, see Smith, Robert, and

Daptrius americanus, 171
ater, 171-175

Davis, William E., Jr., and Jerome A. Jack-
son (eds.). Contributions to the his-
tory of North American Ornithology,
reviewed, 198-200

Davis, William E., Jr., reviews by, 193—194,
194, 196-197, 197, 197-198, 198,

385, 385-386, 393-394, 594-595,
595-596, 597-598, 598
deer, brocket, see Mazama americana
gray brocket, see Mazama gouazoubira
DeGraaf, R. M., and T. J. Maier, Effect of
egg size on predation by white-footed
mice, 535-539

Dekker, Rene W. R. J., see Jones, Darryl N.,

, and Cees S. Roselaar

de la Cueva, Horacio, see Butler, Robert W,

Francisco S. Delgado, , Victor

Pulido, and Brett K. Sandercock
Delgado, Francisco S., see Butler, Robert

W, , Horacio de la Cueva,

Victor Pulido, and Brett K. Sander-
cock

Dendrocygna autumnalis, 145
bicolor, 137-150

Dendroica caerulescens, 57, 59, 98, 101,
178-180, 467-479, 542
castanea, 542, 749
cerulea, 673—684, 749
coronata, 542, 687, 688, 692
chrysoparia, 591-592
discolor, 98, 313
dominica, 498
fusca, 542, 749
magnolia, 168, 542
occidentalis, 280—291
palmarum, 101
pensylvanica, 537, 542, 749
petechia, 167, 167, 472, 475, 542, 749
striata, 749
tigrina, 542
virens, 542, 588-591

Dick, James A., and Ross D. James, Rufous
crown feathers on adult male Tennes-
see Warblers, 181-182
Dickcissel, see Spiza americana
Dickinson, Vanessa M., and Keith A. Ar-
nold, Breeding biology of the Crest-
ed Caracara in south Texas, 516-
523

Didelphis virginianus, 130, 313, 462
diet

food availability and preferences of breed-
ing Dendrocygna bicolor, 137-150
grit-use patterns in North American birds,
685-696

INDEX TO VOLUME 108

831

nutritional value of winter foods for Grus
americana, 728-739

of Glaucidium gnoma and Aegolius acad-
icus in west-central Montana, 123-
128

Diglossa carbonaria, 225
dispersal

of juvenile Egretta thula and Nycticorax
nycticorax, 342-356
distribution

of Catharus bicknelli in New England and
New York, 650-661

diversity

effects of field age on, 760-770
dog, domestic, see Canis familiaris
Doherty, Paul E, Jr., Thomas C. Grubb, Jr.,
and C. L. Brown, Territories and
caching-related behavior of Red-
headed Woodpeckers wintering in a
beech grove, 740-747
Dolichonyx oryzivorus, 543, 765
Dove, Collared, see Javanese Turtle, see
Streptopelia bitorquata
Mourning, see Zenaida macroura
Rock, see Columba livia
White-throated Ground, see Gallicolumba
xanthonura

Dowitcher, Short-billed, see Limnodromus
griseus

Driessens, Gerald, see Chantler, Phil, and

drinking

by Pinicola enucleator, 186—187

Droege, Sam, see Price, Jeff, , and

Amy Price

Dryocopus pileatus, 282, 283, 287, 453,
699, 701

Duck, American Black, see Anas rubripes
Mottled, see Anas fulvigula
Tufted, see Aythya fuligula
Wood, see Aix sponsa
Dumetella carolinensis, 535, 542
Dunlin, see Calidris alpina
Dunn, Erica H., Jacques Larivee, and Andre
Cyr, Can checklist programs be used
to monitor populations of birds re-
corded during the migration season?,
540-549

Dunn, Jon L., see Graves, Gary R., Michael
A. Patten, and

Dunne, Pete, Before the echo: essays on na-
ture, reviewed, 197-198
Eagle, Bald, see Haliaeetus leucocephalus
Spanish Imperial, see Aquila adalberti
Eberhard, Jessica R., Nest adoption by
Monk Parakeets, 374-377
ecology
nesting

of Acrobatornis fonsecai, 434-448
of Tyrannus forficatus, 302-316
Ecton, A. Marie, see Engilis, Andrew, Jr.,
Thane K. Pratt, Cameron B. Kepler,

, and Kimberly M. Fluetsch

Ecton, A. Marie, see Kepler, Cameron B.,
Thane K. Pratt, , Andrew En-

gilis, Jr., and Kimberly M. Fluetsch

egg

effect of type on depredation of ground
nests, 129-136

Egret, Cattle, see Bubulcus ibis
Great, see Casmerodius albus
Snowy, see Egretta thula
Egretta caerulea, 342
thula, 342-356

Elaenia, Gray, see Myiopagis caniceps
Elanoides forficatus, 498, 503
Elaphe obsoleta, 573, 575, 681, 698
obsoleta lindheimeri, 592
Elgood, J. H., et al.. The birds of Nigeria,
reviewed, 203-204

Emerald, Chiribiquete, see Chlorostilbon
olivaresi, sp. nov. (Frontispiece)
Cuban, see Chlorostilbon ricordii
versicolor, see Amazilia versicolor
Empidonax difficilis, 282, 283, 284, 289,
379

flaviventris, 542
hammondii, 280-291
minimus, 168, 542
sp., 168
traillii, 184

virescens, 190, 302, 496—506, 749
Engilis, Andrew, Jr., .see Kepler, Cameron
B., Thane K. Pratt, A. Marie Ecton,

, and Kimberly M. Fluetsch

Engilis, Andrew, Jr., Thane K. Pratt, Cam-
eron B. Kepler, A. Marie Ecton, and
Kimberly M. Fluetsch, Description of
adults, eggshells, nestling, fledgling,
and nest of the Poo-uli, 607-619

832

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Equus zebra, 171
Eremophila alpestris, 688, 690
errata, 192, 394, 603

Erwin, R. Michael, John G. Haig, Daniel B.
Stotts, and Jeff S. Hatfield, Dispersal
and habitat use by post-fledging ju-
venile Snowy Egrets and Black-
crowned Night-Herons, 342-356
Estades, Cristian E, Natural history and con-
servation status of the Tamarugo
Conebill in northern Chile, 268-279
Estrilda astrild, 442
Eucemes faciatus, 702
laticeps, 702

Eudyptes chrysocome, 72-79
chrysolophus, 78
Eugenes fulgens, 229, 236
Eulampis jugularis, 238, 239, 240, 241
Euphonia, Chestnut-bellied, see Euphonia
pectoralis

Violaceous, see Euphonia violacea
Euphonia pectoralis, 412
violacea, 412

Fairy- Wren, see Malurus splendens
Falco mexicanus, 151-153
sparverius, 236, 701
Falcon, Prairie, see Falco mexicanus
Fantail, Rufous, see Rhipidura rufifrons
Felis catus, 632
domesticus, 450

Ficken, Millicent S., see Thumser, Nina N.,

Jeffrey D. Darron, and

Finch, House, see Carpodacus mexicanus
Laysan, see Telespiza cantans
Zebra, see Poephila guttata
Firewood-gatherer, see Anumbius annumbi
Flamingo, Andean, see Phoenicoparrus an-
dinus

Greater, see Phoenicopterus ruber
Flaspohler, David J., Nesting success of the
Prothonotary Warbler in the upper
Mississippi River bottomlands, 457-
466

Flicker, Northern, see Colaptes auratus
Flowerpiercer, Carbonated, see Diglossa car-
bonaria

Fluetsch, Kimberly M., see Engilis, Andrew,
Jr., Thane K. Pratt, Cameron B. Ke-
pler, A. Marie Ecton, and

Fluetsch, Kimberly M., see Kepler, Cameron

B., Thane K. Pratt, A. Marie Ecton,

Andrew Engilis, Jr., and

Flycatcher, Acadian, see Empidonax vires-
cens

Ash-throated, see Myiarchus cinerascens
Brown-crested, see Myiarchus tyrannulus
Dusky-capped, see Myiarchus tuberculifer
Great Crested, see Myiarchus crinitus
Hammond’s, see Empidonax hammondii
La Sagra’s, see Myiarchus sagrae
Least, see Empidonax minimus
Olive-sided, see Contopus borealis
Pacific Slope, see Empidonax difficilis
Scissor-tailed, see Tyrannus forficatus
Vermilion, see Pyrocephalus rubinus
Western, see Empidonax difficilis
Willow, see Empidonax traillii
Yellow-bellied, see Empidonax flaviven-
tris

Yellow-olive, see Tolmomyias sulphures-
cens

flycatcher, see Empidonax sp.
food availability

for breeding Dendrocygna bicolor in Lou-
isiana ricefields, 137-150
Forsman, Eric D., see Anthony, Robert G.,

Gregory A. Green, , and S.

Kim Nelson

Foss, Carol R., (ed.). Atlas of breeding birds
of New Hampshire, reviewed, 195-
196

Fournier, Michael A., and James E. Hines,
Nest sharing by a Lesser Scaup and
a Greater Scaup, 380-381
fox, gray, see Urocyon cinereoargenteus
red, see Vulpes vulpes
swift, see Vulpes velox
Fregata aquila, 330
magnificens, 332
minor, 326, 330, 332, 333
Frigatebird, Ascension, see Fregata aquila
Great, see Fregata minor
Magnificent, see Fregata magnificens
frog, gray tree, see Hyla versicolor, Hyla
chrysoscelis

Fruit-Dove, Mariana, .see Ptilinopus rosei-
capilla

Fulbright, Timothy E., see Nolte, Kenneth

R., and

Fulica atra, 577

INDEX TO VOLUME 108

833

Furrow, Ly T, see Millenbah, Kelly E, Scott
R. Winterstein, Henry Campa III,
, and Richard B. Minnis
Fussell, John O., Ill, A hirder’s guide to
coastal North Carolina, reviewed,
196-197

Gadwall, see Anas strepera
Gallicolumha xanthonura, 246-267
Gallinago gallinago, 788
Gallus gallus, 129-136
Garrido, Orlando, and Arturo Kirkconnell,
Taxonomic status of the Cuban form
of the Red-winged Blackbird, 372-
374

Garrido, Orlando H., see Kirkconnell, Ar-
turo, George E. Wallace, and

Gawlik, Dale E., and R. Douglas Slack,
Comparative foraging behavior of
sympatric Snow Geese, Greater
White-fronted Geese, and Canada
Geese during the non-breeding sea-
son, 154-159

gecko, see Sphaerodactylus sp.

Gee, George E, see Nelson, Jay T, R. Doug-
las Slack, and

Gehlbach, Frederick R., The Eastern
Screech-Owl: life history, ecology
and behavior in the suburbs and
countryside, reviewed, 201-202

genus nova

Acrobatornis fonsecai gen. nov., sp. nov.,
397-433

geographic variation

of Cinnycerthia wrens of the Andes, 205-
227

Geothlypis trichas, 97, 102, 167, 168, 687,
688, 692

Gionfriddo, James R, and Louis B. Best,
Grit-use patterns in North American
birds: the influence of diet, body size,
and sex, 685-696

Gionfriddo, James R, see Mitchell, Mary
Crowe, Louis B. Best, and

Glaucidium gnoma, 123—128

Glaucomys volans, 449, 450, 452, 697—7 1 1

Gnatcatcher, Blue-gray, see Rolioptila caeru-
lea

Goguen, Christopher B., see Curson, David
R., , and Nancy E. Mathews

Golden-Rlover, American, see Rluvialis
dominica

Goldeneye, Common, see Bucephala clan-
gula

Goldfinch, American, see Carduelis tristis
Gonzaga, Luiz R, see Racheco, Jose Fernan-
do, Bret M. Whitney, and

Gonzalez, Jose A., Breeding biology of the
Jabiru in the southern Llanos of Ven-
ezuela, 524-534

Goodbred, Catherine O’Neill, and Richard
T. Holmes, Factors affecting food
provisioning of nestling Black-throat-
ed Blue Warblers, 467-479
Goodwin, Clive E., A bird-finding guide to
Ontario, reviewed, 597-598
Goose, Barnacle, see Branta leucopsis
Canada, see Branta canadensis
domestic, see Anser anser
Emperor, see Chen canagica
Graylag, see Anser anser
Greater White-fronted, see Anser albi-
frons

Ross’, see Chen rossii
Snow, see Chen caerulescens
gopher, northern pocket, see Thymomas tal-
poides

Crackle, Common, see Quiscalus quiscula
Great-tailed, see Quiscalus mexicanus
grants

see awards and grants
Grassquit, Blue-black, See Volatinia ujacar-
ina

Graves, Gary R., Censusing wintering pop-
ulations of Swainson’s Warblers: sur-
veys in the Blue Mountains of Ja-
maica, 94-103

Graves, Gary R., Michael A. Fatten, and Jon
L. Dunn, Comments on a probable
gynandromorphic Black-throated
Blue Warbler, 178-180
Graveteiro, Rink-legged, see Acrobatornis
fonsecai gen. nov., sp. nov. (Frontis-
piece)

Graytail, Double-banded, see Xenerpestes
minlosi

Equatorial, see Xenerpestes singularis
Green, Gregory A., see Anthony, Robert G.,

, Eric D. Forsman, and S. Kim

Nelson

834

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Gregory, Mark, see Koenen, Marcus, T, Da-
vid M. Leslie, Jr., and

Griffin, Curtice R., see Megyesi, Jennifer L.,
and

Grosbeak, Black-headed, see Pheucticus me-
lanocephalus

Evening, see Coccothraustes vespertinus
Pine, see Pinicola enucleator
Rose-breasted, see Pheucticus ludovici-
anus

Ground-Dove, Common, see Columbina
passerina

Ruddy, see Columbina talpacoti
Grouse, Ruffed, see Bonasa umbellus
Grubb, Thomas C., Jr., Doherty, Paul E, Jr.,

, and C. L. Brown

Grus americana, 728—739
Gull, Herring, see Lams argentatus
Laughing, see Lams atricilla
Ring-billed, see Lams delawarensis
Gygis alba, 332
habitat

change and nesting success in Sterna an-
tillamm and Charadrius alexandri-
nus, 292—301

large tract and fragmented forest use by
Otus nudipes, 776-782
use by juvenile Egretta thula and Nyctico-
rax nycticorax, 342-356
use in cooperative and non-cooperative
breeding birds, 712—727
use of coastal agricultural fields by shore-
birds, 783-796

use of forest gaps by breeding Dendroica
virens, 588-591

Haig, John G., see Erwin, R. Michael,

, Daniel B. Stotts, and Jeff S.

Hatfield

Halcyon chloris, 246-267
Haliaeetus leucocephalus, 156
Hall, George A., review by, 195-196
Hall, George A., see Buckelew, Albert, Jr.,
and

Hamerstrom, Frances, My double life: mem-
oirs of a naturalist, 194-195
Harper, Lee H., see Karwowski, Kenneth, J.

Edward Gates, and

Harrier, Northern, see Circus cyaneus
Hatfield, Jeff S., see Erwin, R. Michael,

John G. Haig, Daniel B. Stotts, and

Hawk, Broad-winged, see Buteo platyptems
Cooper’s, see Accipiter cooperii
Red-shouldered, see Buteo lineatus
Red-tailed, see Buteo jamaicensis
White-tailed, see Buteo albicaudatus
Helmitheros vermivoms, 98, 101, 168, 176,
749

Hemignathus munroi, 614
parvus, 616
stejnegeri, 616

virens, 265, 616, 617, 631, 634, 635
Hemithraupis mficapilla, 412
Hermit, Little, see Phaethornis longuema-
reus

Heron, Great White, see Ardea herodias oc-
cidentalis

Little Blue, see Egretta caemlea
White-necked, see Ardea cocoi
Herpestes auropunctatus, 632
Himatione sanguinea, 265, 615, 616, 630,
631

Hines, James E., see Fournier, Michael A.,
and

Himndo mstica, 168, 557, 687, 688, 691
Hodges, Malcolm E, Jr., and David G. Kre-
mentz. Neotropical migratory breed-
ing bird communities in riparian for-
ests of different widths along the Al-
tamaha River, Georgia, 496-506
Hoerath, J. David, see Andrews, Brenda J.,

Marie Sullivan, and

Holmes, Richard T, see Goodbred, Cathe-
rine O’Neill, and

Holt, Denver W., and Leslie A. Leroux, Di-
ets of Northern Pygmy-Owls and
Northern Saw-whet Owls in west-
central Montana, 123-128
Holt, Denver W, review by, 201—202
Hohman, William L., Timothy M. Stark, and
Joseph L. Moore, Food availability
and feeding preferences of breeding
Fulvous Whistling-Ducks in Louisi-
ana ricefields, 137-150
Homan, H. Jeffrey, George M, Linz, Wil-
liam J. Bleier, and Robert B. Carlson,
Colony-site and nest-site use by
Common Crackles in North Dakota,
104-1 14

INDEX TO VOLUME 108

835

Honeycreeper, Crested, see Palmeria dolei
Honeyeater, Micronesian, see Mysomela
rubrata

Howell, Steve N. G., and Sophie Webb, A
guide to the birds of Mexico and
northern Central America, reviewed,
391-393

Hummingbird, Allen’s, see Selasphorus sas-
in

Anna’s, see Calypte anna
Black-chinned, see Archilochus alexandri
Broad-billed, see Cynanthus latirostris
Broad-tailed, see Selasphorus playtcercus
Calliope, see Stellula calliope
Cinnamon, see Amazila rutila
Costa’s, see Calypte costae
Lucifer, see Calothorax lucifer
Magnificent, see Eugenes fulgens
Ruby-throated, see Archilochus colubris
Rufous, see Selasphorus rufus
Scaly-breasted, see Phaeochroa cuvierii
Violet-crowned, see Amazilia violiceps
Violet-headed, see Klais guimete
White-eared, see Hylocharis leucotis
hybrid

male preference in female Passerina cy-
anea and Passerina amoena, 771-
775

Hydrochaeris hydrochaeris, 173
Hyla fasciatus, 702
versicolor, 702
Hylocharis leucotis, 241
Hylocichla mustelina, 498, 536, 542
Icteria virens, 167, 168
Icterus cucullatus, 236
galbula, 168, 543, 687, 689, 692, 749
Ictinia mississippiensis, 498
liwi, see Vestiaria coccinea
information for authors, 395-396, 604-605
Ixobrychus sinensis, 246-267
Ixoreus naevius, 280-291
Jabiru mycteria, 524-534
Jabiru, see Jabiru mycteria
Jackson, Jerome A., review by, 386—387,
387-388

Jackson, Jerome A., see Davis, William E.,
Jr., and

James, Ross D., see Dick, James A., and

Jay, Blue, see Cyanocitta cristata
Gray, see Perisoreus canadensis
Gray-breasted, see Aphelocoma ultramar-
ina

Stellar’s, see Cyanocitta stelleri
Western Scrub, see Aphelocoma califor-
nica

Johnsgard, Paul A., Arena birds: sexual se-
lection and behavior, reviewed, 193—
194

Johnsgard, Paul A., This fragile land. A nat-
ural history of the Nebraska sand-
hills, reviewed, 389-390
Johnson, William P, Frank C. Rohwer, and
Michael Carloss, Evidence of nest
parasitism in Mottled Ducks, 187-
189

Jones, Darryl N., Rene W. R. J. Dekker, and
Cees S. Roselaar, The megapodes
Megapodiidae, reviewed, 811-812
Junco, Dark-eyed, see Junco hyemalis
Junco hyemalis, 125, 280-291, 543
Kakawahie, see Paroreomyza flammea
Karron, Jeffrey D., see Thumser, Nina N.,

, and Millicent S. Ficken

Kemp, Alan, The hornbills, reviewed, 386-
387

Kepler, Cameron B., see Engilis, Andrew,

Jr., Thane K. Pratt, , A. Marie

Ecton, and Kimberly M. Fluetsch
Kepler, Cameron B., Thane K. Pratt, A. Ma-
rie Ecton, Andrew Engilis, Jr., and
Kimberly M. Fluetsch, Nesting be-
havior of the Poo-uli, 620-638
Kerlinger, Paul, How birds migrate, re-
viewed, 594

Kestrel, American, see Falco sparverius
Kilgo, John C., Robert A. Sargent, Brian R.
Chapman, and Karl V. Miller, Nest-
site selection by Hooded Warblers in
bottomland hardwoods of South Car-
olina

Killdeer, see Charadrius vociferus
King, David I., Carnivory observed in the
Cedar Waxwing, 381-382
Kingbird, Cassin’s, see Tyrannus vociferans
Eastern, see Tyrannus tyrannus
Thick-billed, see Tyrannus crassirostris
Kingfisher, Amazon, see Cholorceryle ama-
zona

Collared, see Halcyon chloris

836

THE WILSON BULLETIN • Vol. 108. No. 4, December 1996

Kinglet, Golden-crowned, see Regulus satra-
pa

kingsnake, see Lampropeltis spp.
Kirkconnell, Arturo, George W. Wallace,
and Orlando H. Garrido, Notes on the
status and behavior of the Swainson’s
Warbler in Cuba, 175—178
Kirkconnell, Arturo, see Garrido, Orlando,
and

Kite, American Swallow-tailed, see Elano-
ides forficatus

Mississippi, see Ictinia mississippiensis
Snail, see Rostrhamus sociabilis
Kiviat, Erik, American Goldfinch nests in
purple loosestrife, 182—186
Klais guimeti, 241

klipspringer, see Oreotragus oreotragus, 171
Knopf, Fritz L., and Jeffery R. Rupert, Re-
production and movements of Moun-
tain Plovers breeding in Colorado,
28-35

Knot, Red, see Calidris canutus
Koa-Finch, see Rhodacanthis palmeri
Koenen, Marcus, T, David M. Leslie, Jr.,
and Mark Gregory, Habitat changes
and success of artificial nests on a al-
kaline flat, 292-301

Krementz, David G., see Hodges, Malcolm
E, Jr., and

Kricher, John C., review by, 193-194
Lagothris lagotricha cana, 172
Lampropeltis spp., 313
Larivee, Jacques, see Dunn, Erica H.,

, and Andre Cyr

Lark, Horned, see Eremophila alpestris
LaRoe, Edward T, III, see Noss, Reed E,

, and J. Michael Scott

Larus argentatus, 557
atricilla, 190
delawarensis, 296

Leafloor, James O., John E. Thompson, and
C. Davison Ankney, Body mass and
carcass composition of fall migrant
Oldsquaws, 567-572

Lefebvre, Gaetan, and Brigitte Poulin,
Abundance of migrant birds in rela-
tion to food resources in Caribbean
and Pacific mangroves of Panama,
748-759

Leontopithecus chrysomelas, 430

Leroux, Lislie A., see Holt, Denver W, and

Leslie, David M., Jr., see Koenen, Marcus,

T, , and Mark Gregory

Levi, Peter, Edward Lear: a biography, re-
viewed, 598—600

Limnodromus griseus, 788, 790, 792
Limnothlypis swainsonii, 94—103, 175-178,
498, 802-804

Lindsey, Gerald D., see Meyers, J. Michael,

Wayne J. Arendt, and

Linz, George M., see Homan, H. Jeffrey,

, William J. Bleier, and Robert

B. Carlson
lizard, see Anolis sp.

Lockwood, Mark W, Courtship behavior of
Golden-cheeked Warblers, 591-592
Lophornis adorabilis, 241
Loxioides bailleui, 614, 616, 631, 634, 635
Loxops caeruleirostris, 614
coccineus, 615, 632

MacDougall-Shackleton, Scott, review by,
600-601

Magpie, Black-billed, see Pica pica
magpie-jay, see Calocitta spp.

Mahan, Carolyn G., see Yahner, Richard H.,
and

Maier, T. J., see DeGraaf, R. M., and

Mallard, see Anas platyrhynchos
Malurus splendens, 775
Mango, Black-throated, see Anthracothorax
nigricollis

Margarornis bellus, 426
spp., 397—433, 446
squamiger, 397—433, 446
Martin, Purple, see Progne subis
Marzluff, John M., and Mary McFadzen, Do
standardized brood counts accurately
measure productivity?, 151-153
Mason, J. Russell, and Larry Clark, Avoid-
ance of cabbage fields by Snow
Geese, 369-37 1

Masticophis flagellum testaceus, 314
Mathews, Nancy E., see Curson, David R.,

Christopher B. Goguen, and

Mazama americana, 173
gouazoubira, 171-175
McClure, H. Elliot, Stories I like to tell: an
autobiography, reviewed, 385-386

INDEX TO VOLUME 108

837

McElroy, David B., and Gary Ritchison, Ef-
fect of mate removal on singing be-
havior and movement patterns of fe-
male Northern Cardinals, 550-555
McFaden, Mary, see Marzluff, John M., and

McFarland, Kent R, see Atwood, Jonathan

L., Christopher C. Rimmer, ,

Sophia H. Tsai, and Laura R. Nagy
McFarland, Kent R, see Rimmer, Christo-
pher C., Jonathan L. Atwood,

, and Laura R. Nagy

McIntyre, Judith W., and Norma G. Cutler,
Annotated bibliography of the loons,
Gaviidae, reviewed, 194
Meadowlark, Eastern, see Sturnella magna
Western, see Sturnella neglecta
Megapode, Micronesian, see Megapodius
laperouse

Megapodius laperouse, 246-267
Megyesi, Jennifer L., and Curtice R. Griffin,
Breeding biology of the Brown Nod-
dy on Tern Island, Hawaii, 317—334
Melamprosops phaeosoma, 607-619, 620-638
Melaneipes carolinus, 450, 454, 699, 702, 743
erythrocephalus, 687, 688, 740-747, 743
formicivorus, 236, 722
hoffmanni, 25
rubrucapillus, 25
uropygialis, 236
Meleagris gallopavo, 129-136
Melospiza georgiana, 184, 543
lincolnii, 543

melodia, 125, 184, 340, 543, 688, 765
Mephitis mephitis, 130, 462
methods

survial of radio-collared nestling Amazo-
na vittata, 159-163

Metopothrix aurantiacus, 397-433, 446, 447
spp., 397-433, 434, 435, 446
Meyers, J. Michael, New nesting area of
Ruerto Rican Rarrots, 164-166
Meyers, J. Michael, see Rardieck, Keith L.,

, and Michelle Ragan

Meyers, J. Michael, Wayne J. Arendt, and
Gerald D. Lindsey, Survival of radio-
collared nestling Ruerto Rican Rar-
rots, 159-163

Microtus montanus, 125, 126
pennsylvanicus, 125, 126
spp., 123-128

migration

routes of Calidris mauri, 662-672
Millenbah, Kelly E, Scott R. Winterstein,
Henry Campa III, Ly T. Furrow, and
Richard B. Minnis, Effects of Con-
servation Reserve Rrogram field age
on avian relative abundance, diversi-
ty, and productivity, 760-770
Miller, Karl V., see Kilgo, John C., Robert
A. Sargent, Brian R. Chapman, and

Milvago chimachima, 173, 174
spp., 173

Mimus polyglottos, 236, 314, 573-583
mink, see Mustela vison
Minnis, Richard B., see Millenbah, Kelly E,
Scott R. Winterstein, Henry Campa

III, Ly T. Furrow, and

Mitchell, Jeremy S., and Raleigh J. Rob-
ertson, Extra nest site occupancy by
Tree Swallows: do floaters avoid
nest sites near settled pairs?, 797-
802

Mitchell, Mary Crowe, Louis B. Best, and
James R Gionfriddo, Avian nest-site
selection and nesting success in two
Florida citrus groves, 573-583
Mniotilta varia, 98, 101, 168, 190, 536, 542,
749

Mockingbird, Northern, see Mimus poly-
glottos

Moldenhauer, Ralph R., see Regelski, Daniel
J., and

Molothrus ater, 184, 381, 457, 543, 592,
676, 677, 687, 689

mongoose, see Herpestes auropunctatus
monkey, wooly, see Lagothrix logotricha
cana

Monroe, Burt L., Jr., The birds of Kentucky,
reviewed, 197

Moore, Joseph L., see Hohman, William L.,

Timothy M. Stark, and

Moorman, Christopher E., and Brian R.
Chapman, Nest-site selection of Red-
shouldered and Red-tailed hawks in a
managed forest, 357-368
moose, see Alces alces
Morrison, Michael L., see Block, William
M., , and M. Hildegard Reiser

838

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

mouse, see Peromyscus spp.

deer, see Peromyscus maniculatus
white-footed, see Peromyscus leucopus
Mus musculus, 632
Mustela spp., 130
vison, 462

Myiarchus cineracens, 236
crinitus, 498, 542
sagrae, 490
tuberculifer, 168
tyrannulus, 236
Myiopagis caniceps, 412
Myiopsitta monachus, 374-377, 584-588
Myzomela rubrata, 246-267
Nagy, Laura R., see Atwood, Jonathan L.,
Christopher C. Rimmer, Kent P
McLarland, Sophia H. Tsai, and

Nagy, Laura R., see Rimmer, Christopher C.,
Jonathan L. Atwood, Kent P. Mc-

Larland, and

Nasua narica, 173
natural history

of Conirostrum tamarugense in northern
Chile, 268-279

of Micronesian community, 246—267
of Tachycineta cyaneoviridis, 480-495
Nelson, Jay T, R. Douglas Slack, and
George L Gee, Nutritional value of
winter foods for Whooping Cranes,
728-739

Nelson, S. Kim, see Anthony, Robert G.,
Gregory A. Green, Eric D. Lorsman,

and

Neotropical migrants

breeding communities in riparian forests
of different widths, 496-506
Nerodia sipedon, 191
nest

adoption

by Myiositta monachus, 374-377
artificial

success for Sterna antillarum and Cha-
radrius alexandrinus, 292-301
use of polystyrene snags by woodpeck-
ers, 449—456
attentiveness

in hummingbirds, 228-245
description

of Melamprosops phaeosoma, 607-619

parasitism

in Anas fulvigula, 187-189
reuse

by Contopus sordidulus, 378-380
sharing

by Aythya affinis and Aythya marila,
380-381

site

extra occupancy by Tachycineta bicolor,
797-802

new area for endangered Amazona vit-
tata, 164-166

selection by Wilsonia citrina in a bot-
tomland hardwoods, 53-60
selection in Buteo lineatus and Buteo
jamaicensis, 357—368
selection in Limnothlypis swainsonii,
802-804

selection in two Llorida citrus groves,
573-583

selection in Tyrannus forficatus, 302-
316

use by Quiscalus quiscula in North Da-
kota, 104-114

use of purple loosestrife by Carduelis
tristis, 182-186

structure

of Acrobatornis fonsecai, 434-448
nesting

behavior of Melamprosops phaeosoma,
620-638

changes in body mass of female Bucepha-
la clangula, 61—71

of Limnothlypis swainsonii in east Texas,
802-804

success in two Llorida citrus groves, 573-
583

success of Picoides borealis in south Tex-
as, 697-7 1 1

success of Protonotaria citra in Mississip-
pi River bottomland, 457-466

nestling

description of Melamprosops phaeosoma,
607-619

eaten by Bombycilla cedrorum, 381—382
food provisioning in Dendroica caerules-
cens, 467-479

Night-Heron, Black-crowned, see Nyctico-
rax nycticorax

INDEX TO VOLUME 108

839

Noddy, Black, see Anous tenuirostris
Brown, see Anous stolidus
Nolle, Kenneth R., and Timothy E. Ful-
bright. Nesting ecology of Scissor-
tailed Flycatchers in south Texas,
302-316

Noss, Reed E, Edward T. LaRoe III, and J.
Michael Scott, Endangered ecosys-
tems of the United States: a prelimi-
nary assessment of loss and degra-
dation, reviewed, 389
Numenius phaeopus, 786, 788
Nunrey, David, see Winkler, Hans, David A.
Christie, and

Nuthatch, Red-breasted, see Sitta canadensis
White-breasted, see Sitta carolinensis
Nycticorax nycticorax, 332, 342-356, 557
Oldsquaw, see Clangula hyemalis
Oliarnyk, Catherine J., and Raleigh J. Rob-
ertson, Breeding behavior and repro-
ductive success of Cerulean Warblers
in southeastern Ontario, 673-684
Onychognathus nabouroup, 171
Oporornis formosus, 101, 168, 176, 190,
498

Philadelphia, 335, 542
tolmiei, 168, 190, 282, 283, 284
opossum, Virginia, see Didelphis virgini-
anus

Oreomystis bairdi, 616
mana, 614, 632

Oriole, Hooded, see Icterus cucullatus
Northern, see Icterus galbula
Otus asio, 127, 553, 701, 776, 778, 780
kennicottii, 127
nudipes, 776—782
Ovenbird, see Seiurus aurocapillus
Owl, Barn, see Tyto alba
Barred, see Strix varia
Boreal, see Aegolius funereus
Great Horned, see Bubo virginianus
Long-eared, see Asio otus
Mottled, see Ciccaba virgata
Northern Spotted, see Strix occidentalis
caurina

Saw-whet, see Aegolius acadicus
Oxpecker, Red-billed, see Buphagus african-
us

Yellow-billed, see Buphagus erythrorhyn-
chus

Pacheco, Jose Fernando, Bret M. Whitney,
and Luiz P. Gonzaga, A new genus
and species of Furnariid (Aves: Fur-
nariidae) from the cocoa-growing re-
gion of southeastern Bahia, Brazil,
397-433

Pacheco, Jose Fernando, see Whitney, Bret

M., , Paulo Sergio Moreira da

Fonseca, and Robert H. Barth, Jr.
Pachyramphus aglaiae, 168
castaneus, 412
marginatus, 412

Pagan, Michelle, see Pardieck, Keith L., J.

Michael Meyers, and ■

Palmeria dolei, 616, 617

Parakeet, Monk, see Myiopsitta monachus

parasitism

new host species for Protocalliphora (Dip-
tera: Calliphoridae), 189-190
Pardieck, Keith L., J. Michael Meyers, and
Michelle Pagan, Surveys of Puerto
Rican Screech-Owl populations in
large-tract and fragmented forest hab-
itat, 776-782

Paroreomyza flammea, 616
montana, 614, 616, 617, 631, 636
Parrott, Puerto Rican, see Amazona vittata
Parrotbill, Maui, see Pseudonestor xantho-
phrys

Parula americana, 335—341, 496-506, 542
pitiayumi, 412

Parula, Northern, see Parula americana
Tropical, see Parula pitiayumi
Parus atricapillus, 125, 282, 283, 287
bicolor, 743

carolinensis, 449, 450, 454, 743
gambeli 90, 91, 125
major, 474, 507, 513, 514
montanus, 80-93, 507
Rufescens, 280—291

Passer domesticus, 124, 125, 481, 490, 687,
688, 691, 692
montanus, 247

Passerculus sandwichensis, 472, 543, 688,
689

Passerella iliaca, 688
Passerina amoena, 771-775
ciris, 167, 168
cyanea, 168, 689, 771—775

840

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Patten, Michael A., see Graves, Gary R.,

, and Jon L. Dunn

peccary, see Tayassu tajacu and Tayassu pe-
cari.

Peck, Robert McCracken, review by, 598-
600

Penguin, Humboldt, see Spheniscus hum-
boldti

Jackass, see Spheniscus demersus
Macaroni, see Eudyptes chrysolophus
Magellanic, see Spheniscus magellanicus
Rockhopper, see Eudyptes chrysocome
Peppershrike, Rufous-browed, see Cuclarhis
gujanensis

Peres, Carlos A., Ungulate ectoparasite re-
moval by Black Caracaras and Pale-
winged Trumpeters in Amazonian
forests, 170—175

Perisoreus canadensis, 282, 283, 289
Peromyscus leucopus, 535-539.
maniculatus, 125, 126
spp., 462

pewee, see Contopus sp.

Phacellodomus dorsalis, 421
rufifrons, 401, 424, 434-448
sibilatrix, 397-433
spp., 415, 421, 423, 441, 446
Phaeochroa cuvierii, 241
Phaethon rubricauda, 332
Phaethornis longuemareus, 241
Phainopepla nitens, 1 19
Phainopepla, see Phainopepla nitens
Phasianus colchicus, 685-696, 766
Pheasant, Ring-necked, see Phasianus col-
chicus

Pheucticus ludovicianus, 543
melanocephalus, 236
Phoebe, Black, see Sayornis nigricans
Eastern, see Sayornis phoebe
Phoenicoparrus andinus, 807-808
chilensis, 807

Phoenicopterus ruber, 485, 807
Phoeniculus purpureus, 722
Phylloscopus collybita, 340
Pica pica, 1 7 1
Picoides borealis, 697-71 1
pubescens, 449-456, 743
villosus, 280-291, 450, 454, 490
pig, feral, see Sus scrofa

Pigeon, White-crowned, see Columba leu-
cocephala

Pinicola enucleator, 186—187
Pintail, Northern, see Anas acuta
Pipilo fuscus, 236
Piranga ludoviciana, 282, 283, 284
olivacea, 543
rubra, 236, 498, 749
Platyrinchus mystaceus, 168, 224, 225
Plover, Black-bellied, see Pluvialis squata-
rola

Mountain, see Charadrius montanus
Semipalmated, see Charadrius semipal-
matus

Snowy, see Charadrius alexandrinus
plumage

of Melamprosops phaeosoma, 607—619
probable gynandromorphic Dendroica
caerulescens, 178-180
rufous crown feathers on adult male Ver-
mivora peregrina, 181-182
Plushcrown, Orange-fronted, see Metopo-
thrix aurantiacus
Pluvialis dominica, 783-796
squatarola, 191, 783-796
Pochard, European, see Aythya ferina
Poephila guttata, 535-539
Polioptila caerulea, 496-506
Polyborus [Caracara] plancus, 516—523,
530, 532

Poo-uli, see Melamprosops phaeosoma
Pooecetes gramineus, 58, 543, 687, 689
population

density of Catharus bicknelli, 639—649
growth of Myiopsitta monachus in the
United States, 584-588
of Micronesian community, 246—267
use of checklist to monitor trends during
migratory season, 540-549
Poulin, Brigitte, Lefebvre, Gaetan, and

Power, Dennis M., (ed.). Current ornitholo-
gy, Vol. 1 2, reviewed, 393
Pravosudov, Vladimir V., and Elena V. Pra-
vosudova. The breeding biology of
the Willow Tit in northeastern Sibe-
ria, 80-93

Pravosudova, Elena V., see Pravosudov,

Vladimir V., and

Pratt, Thane K., see Engilis, Andrew, Jr.,

INDEX TO VOLUME 108

841

, Cameron B. Kepler, A. Ma-
rie Ecton, and Kimberly M. Fluetsch
Pratt, Thane K., see Kepler, Cameron B.,
, A. Marie Ecton, Andrew En-
gilis, Jr., and Kimberly M. Fluetsch
predation

effects of egg type on depredation of
ground nests, 129-136
of eggs by Peromyscus leucopus, 535-539
of shorebirds by Chelydra serpentina,
190-192

Price, Amy, see Price, Jeff, Sam Droege, and

Price, Jeff, Sam Droege, and Amy Price,
The summer atlas of North American
birds, reviewed, 594-595
proceedings

seventy-seventh annual meeting, 813-824
Procyon lotor, 130, 133, 313, 451, 462, 538
productivity

accuracy of standardized counts, 151-153
effects of field age on, 760-770
Progne subis, 482

Protonotaria citrea, 449, 450, 454, 457-466,
496-506, 535, 748-759
Pruett-Jones, Stephen, see Van Bael, Sun-
shine, and

Pryor, Gregory S., Observations of shore-
bird predation by snapping turtles in
eastern Lake Ontario, 190-192
Pseudonestor xanthophrys, 617, 636
Pseudoseisura cristata, 447
lophotes, 375, 376, 443, 444
spp., 441

Psophia crepitans, 172
leucoptera, 171-175
spp., 173
viridis, 172

Ptilinopus roseicapilla, 246—267
Pulido, Victor, see Butler, Robert W., Fran-
cisco S. Delgado, Horacio de la Cue-

va, , and Brett K. Sandercock

Pygmy-Owl, Northern, see Glaucidium gno-
ma

Pyrocephalus rubinus, 302, 377-378
Quail, Common, see Coturnix coturnix
Japanese, see Coturnix coturnix
Quiscalus mexicanus, 313
quiscula, 104-114, 462, 543, 689

racer, yellow-bellied, see Coluber constrictor
flaventris

racoon, see Procyon lotor
Rail, Yellow, see Coturnicops novoboracen-
sis

Rattus exulans, 632
norvegicus, 632
rattus, 632, 636

Raven, Common, see Corvus corax
Fan-tailed, see Corvus rhipidurus
Redstart, American, see Setophaga ruticilla
Reed- Warbler, Great, see Acrocephalus
arundinaceus

Nightingale, see Acrocephalus luscinia
Regelski, Daniel J., and Ralph R. Molden-
hauer, Discrimination between re-
gional song forms in the Northern Pa-
rula, 335-341

Regulus satrapa, 280-291, 542
Reinert, Bianca L., see Bomschein, Marcos
R., and

Reiser, M. Hildegard, see Block, William
M., Michael L. Morrison, and

Remsen, J. V., Jr., see Brumfield, Robb T,

and

reproduction

of Charadrius montanus in Colorado, 28-
35

success of Dendroica cerulea, 673-684
Revels, Mia, Eight new host species for the
parasitic blow fly genus Protocalli-
phora (Diptera: Calliphoridae), 189-
190

Rhipidura rufifrons, 246-267
Rhodacanthis palmeri, 616
Riggs, Michael R., see Zicus, Michael C.,
and

Rimmer, Christopher C., Jonathan L. At-
wood, Kent P. McFarland, and Laura
R. Nagy, Population density, vocal
behavior, and recommended survey
methods for Bicknell’s Thrush, 639-
649

Rimmer, Christopher C., see Atwood, Jona-
than L., , Kent P. McFarland,

Sophia H. Tsai, and Laura R. Nagy
Ritchi.son, Gary, see McElroy, David B., and

842

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Robertson, Raleigh J., see Mitchell, Jeremy
S., and

Robertson, Raleigh J., see Oliarnyk, Cathe-
rine J., and

Robin, American, see Turdus migratorius
Clay-colored, see Turdus grayii

Rodgers, James A., Jr., Measurements of
Snail Kite eggs from central Florida,
804-807

Rohwer, Frank C., see Johnson, William R,
, and Michael Carloss

Roselaar, Cees S., see Jones, Darryl N.,
Rene W. R. J. Dekker, and

Rostrhamus sociabilis, 804—807

Rottenborn, Stephen C., The use of coastal
agricultural fields in Virginia as for-
aging habitat by shorebirds, 783—
796

Rowley, Peter, The chronicles of the Row-
leys, reviewed, 197

Rudolph, D. Craig, see Conner, Richard N.,

, Daniel Saenz, and Richard

R. Schaefer

Rupert, Jeffrey R., see Knopf, Fritz L., and

Sabrewing, Violet, see Campylopterus hem-
ileucurus

Saenz, Daniel, see Conner, Richard N., and

Saenz, Daniel, see Conner, Richard N., D.

Craig Rudolph, , and Richard

R. Schaefer

Saltator coerulescens, 168
Saltator, Grayish, see Saltator coerulescens
Sandercock, Brett K., see Butler, Robert W.,
Francisco S. Delgado, Horacio de la

Cueva, Victor Pulido, and

Sanderling, see Calidris alba
Sandpiper, Baird’s, see Calidris bairdii
Buff-breasted, see Tryngites subruficollis
Least, see Calidris minutilla
Pectoral, see Calidris melanotos pusilla
Semipalmated, see Calidris
Solitary, see Tringa solitaria
Spotted, .see Actitis macularia
Upland, .see Bartramia longicauda
Western, see Calidris mauri
White-rumped, see Calidris fuscicollis
Sapsucker, Yellow-bellied, see Sphyrapicus
varius

Sargent, Robert A., see Kilgo, John C.,

, Brian R. Chapman, and Karl

V. Miller

Sayornis nigricans, 377-378
phoebe, 542
spp., 302

Scaphidura oryzivora, 173
Scaup, Greater, see Aythya marila
Lesser, see Aythya affinis
Schaefer, Richard R., see Conner, Richard
N., D. Craig Rudolph, Daniel Saenz,
and

Schoeniophylax phryganophila, 447
Sciurus carolinensis, 130, 462, 681
niger, 462, 700, 701, 743
spp., 462

Scott, J. Michael, see Noss, Reed F, Edward

T. LaRoe III, and

Screech-Owl, Eastern, see Otus asio
Puerto Rican, see Otus nudipes
Western, see Otus kennicottii
Scytalopus spp., 412

Searcy, William A., and Ken Yasukawa, Po-
lygyny and sexual selection in Red-
winged Blackbirds, reviewed, 600
Seedeater, White-collared, see Sporophila
torqueola

Seiurus aurocapillus, 97, 102, 168, 176, 177,
265, 537, 542
motacilla, 97—98, 102
noveboracensis, 167, 168, 542, 748-759
Selasphorus platycercus, 239, 241
rufus, 239, 280-29 1
sasin, 241

Setophaga ruticilla, 98, 101, 168, 176, 474,
542, 748-759

Shoveler, Northern, see Anas clypeata
shrew, see Sorex spp.

masked, see Sorex cinereus
vagrant, see Sorex vagrans
Siegel, Rodney B., and Marco V. Centeno,
Neotropical migrants in marginal
habitats on a Guatemalan cattle
ranch, 166-170
Siptornis spp., 415
striaticollis, 401

Siskin, Pine, see Carduelis pinus
Sitta canadensis, 282, 283, 289, 290
carolinensis, 743

INDEX TO VOLUME 108

843

skink, broad-headed, see Eumeces laticeps
five-lined, see Eumeces fasciatus
skunk, striped, see Mephitis mephitis
Skutch, Alexander E, Orioles, blackbirds,
and their kin: a natural history, re-
viewed, 809

Slack, R. Douglas, see Gawlik, Dale E., and

Slack, R. Douglas, see Nelson, Jay T,

, and George E Gee

Slagsvold, Tore, Dawn and dusk singing of
male American Robins in relation to
female behavior, 507-515
Slater, R J. B., see Catchpole, C. K., and

Smith, P. William, review by, 203—204
Smith, Robert, and Matthew Dallman, For-
est gap use by breeding Black-throat-
ed Green Warblers, 588-591
snake, black rat, see Elaphe obsoleta
northern water, see Nerodia sipedon
rat, see Elaphe obsoleta
Texas rat, see Elaphe obsoleta lindheimeri
Snipe, Common, see Gallinago gallinago
Softtail, Plain, see Thripophaga fusciceps
Russet-mantled, see Thripophaga berlep-
schi

Striated, see Thripophaga macroura
Sorex cinereum, 126
spp., 126
vagrans, 125, 126

Spadebill, White-throated, see Platyrinchus
mystaceus

Sphaerodactylus sp., 176

Sparks, Daniel W., see Custer, Christine M.,

Thomas W. Custer, and

Sparrow, American Tree, see Spizella arbo-
rea

Bachman’s, see Aimophila aestivalis
Chipping, see Spizella passerina
Eurasian Tree, see Passer montanus
Fox, see Passerella iliaca
House, see Passer domesticus
Lark, see Chondestes grammacus
Lincoln’s, see Melospiza lincolnii
Savannah, see Passerculus sandwichensis
Song, see Melospiza melodia
Swamp, see Melospiza georgiana
Vesper, see Pooecetes gramineus

White-throated, see Zonotrichia albicollis
species nova

Acrobatornis fonsecai gen. nov. sp. nov.,
397-433

Chlorostilbon olivaresi sp. nov., 1-27
Spheniscus demersus, 72-79
humboldti, 72-79
magellanicus, 72-79
Sphyrapicus varius, 542
Spinetail, Chotoy, see Schoeniophylax phry-
ganophila

Light-crowned, see Cranioleuca albiceps
Marcapata, see Cranioleuca marcapatae
Rufous-breasted, see Synallaxis erythro-
thorax

Rusty-backed, see Cranioleuca vulpina
Scaled, see Cranioleuca meulleri
Speckled, see Cranioleuca gutturata
Sulphur-bearded, see Cranioleuca sulphur-
ifera

Spiza americana, 314, 689, 692
Spizella arborea, 687, 689, 691
passerina, 543, 689
Sporophila torqueola, 168
squirrel, see Sciurus spp.
fox, see Sciurus niger
gray, see Sciurus carolinensis
red, see Tamiasciurus grahamensis
southern flying, see Glaucomys volans
Stark, Timothy M., see Hohman, William L.,

, and Joseph L. Moore

Starling, European, see Sturnus vulgaris
Micronesian, see Aplonis opaca
Pale-winged, see Onychognathus nabour-
oup

status

of Limnothlypis swainsonii in Cuba, 175-
178

of Phoenicoparrus andinus in Brazil, 807—
808

Stelgidopteryx serripennis, 168
Stellula calliope, 239, 241
Sterna antillarum, 292-301
forsteri, 190
fuscata, 332

Stevens, Lewis, Avian biochemistry and
molecular biology, reviewed, 809
Stiles, E Gary, A new species of Emerald
hummingbird (Trochilidae, Chloro-
stilbon) from the Sierra de Chiribiquete,

844

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

southeastern Colombia, with a review
of the C. mellisugus complex, 1-27
Stork, Maguari, see Ciconia maguari
Stotts, Daniel B., see Erwin, R. Michael,

John G. Haig, , and Jeff S.

Hatfield

Streptopelia bitorquata, 247
Strix occidentalis caurina, 280
varia, 127

Sturnella magna, 543
neglecta, 689

Sturnus vulgaris, 36-52, 481, 490, 513, 688
Stutchbury, Bridget J., see Taroff, Scott A.,
and

Sullivan, Marie, see Andrews, Brenda J.,

, and J. David Hoerath

survival

of radio-collared nestling Amazona vitta-
ta, 159-163
Sus scrofa, 621, 636

Swallow, Bahama, see Tachycineta cyaneo-
viridis

Barn, see Hirundo rustica
Golden, see Tachycineta euchrysea
Mangrove, see Tachycineta albilinea
Northern Rough-winged, see Stelgidopte-
ryx serripennis
Tree, see Tachycineta bicolor
Swift, Vaux’s, see Chaetura vauxi, 282, 283,
287, 289

Swiftlet, Island, see Aerodramus vanikorensis
Synallaxis erythrothorax, 168
Tachycineta albilinea, 168, 481
bicolor, 480-495, 797-802
cyaneoviridis, 480-495
euchrysea, 485
spp., 480

Tachyphonus cristatus, 412
tamarin, golden-headed lion, see Leontopi-
thecus chrysomelas
Tamias striatus, 130, 537
Tamiasciurus hudsonicus, 130
Tanager, Blue-gray, see Thraupis episcopus
Flame-crested, see Tachyphonus cristatus
Green-headed, see Tangara seledon
Palm, see Thraupis palmarun
Red-necked, see Tangara cyanocephala
Rufous-headed, see Hemithraupis ruficap-
illa

Sayaca, see Thraupis sayaca

Scarlet, see Piranga olivacea
Summer, see Piranga rubra
Western, see Piranga ludoviciana
Tangara cyanocephala, 412
seledon, 412

tapaculo, see Scytalopus spp.

Tapera naevia, 442
tapir, Baird’s, see Tapirus bairdii
Brazilian, see Tapirus terrestris
Tapirus bairdii, 173
terrestris, 171

Taroff, Scott A., and Bridget J. Stutchbury,
A case of cooperative breeding in the
Hooded Warbler, 382—384
taxonomy

of Acrobatornis fonsecai gen. nov., sp.
nov., 397-433

of Chlorostilbon mellisugus complex, 1-
27

of Cinnycerthia wrens of the Andes, 205-
227

of Cuban form of Agelaius phoeniceus,
372-374

Tayassu pecari, 173
tajacu, 173

Taylor, Iain, Barn Owls: predator-prey rela-
tionships and conservation, 81 1
Taylor, Richard Cachor, A birder’s guide to
southeastern Arizona, reviewed, 810-
811

Teal, Blue-winged, see Anas discors
Cinnamon, see Anas cyanoptera
Green-winged see Anas crecca
technique

use of checklists to monitor population
trends during migration, 540-549
Telespiza cantans, 633, 634, 635
Temby, I. D., see Emison, W. B., C. M.

Beardsell, and .

Tern, Common, Forster's, see Sterna forsteri
Least, see Sterna antillarum
Sooty, see Sterna fuscata
White, see Gygis alba
territory

of wintering Melanerpes erythrocephalus,
740-747

Thamnophilus doliatus, 168
Thompson, John E., see Leafloor, James O.,
, and C. Davison Ankney

INDEX TO VOLUME 108

845

Thornbird, Chestnut-backed, see Phacello-
domus dorsalis

Little, see Phacellodomus sibilatrix
Thrasher, Brown, see Toxostoma rufum
Long-billed, see Toxostoma longirostre
Thraupis episcopus, 168
palmarus, 412
sayaca, 412

Thripophaga berlepschi, 421
fusciceps, 397-433, 445
macroura, 421, 445
spp., 415, 417, 421, 445
Thrush, Bicknell’s, see Catharus bicknelli
Gray-cheeked, see Catharus minimus
Hermit, see Catharus guttatus
Swainson’s, see Catharus ustulatus
Varied, see Ixoreus naevius
Wood, see Hylocichla mustelina
Thumser, Nina N., Jeffrey D. Karron, and
Millicent S. Eicken, Interspecific
variation in the calls of Spheniscus
penguins, 72-79
Thymomas talpoides, 126
Tit, Great, see Parus major
Willow, see Parus montanus
Titmouse, Tufted, see Parus bicolor
Todirostrum cinereum, 168
Tody-Flycatcher, Common, see Todirostrum
cinereum

Tolmomyias sulphurescens, 168
Tomer, John S., and Michael J. Brodhead,
(eds.), A naturalist in Indian territory:
the journals of S. W. Woodhouse,
1849-1850, reviewed, 200-201
Towhee, Canyon, see Pipilo fuscus
Toxostoma longirostre, 579
rufum, 542, 573-583

Treerunner, Beautiful, see Margarornis bel-
lus

Pearled, see Margarornis squamiger
Tringa flavipes, 190, 191, 788, 790
melanoleuca, 788, 790
solitaria, 788, 791

Troglodytes aedon, 449, 457-466, 542, 553,
688, 797
solstitialis, 222
troglodytes, 280-291, 542
Tropic-bird, Red-tailed, see Phaethon rubri-
cauda

Trumpeter, Gray-winged, see Psophia crepi-
tans

Green-winged, see Psophia viridis
Pale-winged, see Psophia leucoptera
White-winged, see Psophia leucoptera
Tryngites subruficollis, 783-796
Tsai, Sophia H., see Atwood, Jonathan L.,
Christopher C. Rimmer, Kent P.

McFarland, , and Laura R.

Nagy

Tucson Audubon Society, Finding birds in
southeast Arizona, reviewed, 810-
811

Turdus grayi, 168

migratorius, 125, 507-515, 542, 688
Turkey, Wild, see Meleagris gallopavo
Turnstone, Ruddy, see Arenaria interpres
turtle, painted, see Chrysemys picta
snapping, see Chelydra serpentina
Tyrannus crassirostris, 236
forficatus, 302-316
spp., 302

tyrannus, 312, 313, 378, 474, 687, 688,
691

vociferans, 236
Tyto alba, 127

Ula-ai-hawane, see Ciridops anna
Urocyon cinereoargenteus, 130
Ursus americanus, 130
Van Bael, Sunshine, and Stephen Pruett-
Jones, Exponential population growth
of Monk Parakeets in the United
States, 584-588
Veery, see Catharus fuscescens
Venilornis fumigatus, 168
Vermivora bachmani, 804
celata, 181
chrysoptera, 749

peregrina, 168, 181 — 182, 542, 749
ruficapilla, 181, 471, 474, 542
Vestiaria coccinea, 616, 617, 631
Vireo bellii, 168
flavifrons, 190, 498
gilvus, 282, 283, 284, 542
griseus, 496—506

olivaceus, 412, 496—506, 542, 749
philadelphicus, 542
Vireo, Bell's, see Vireo bellii
Philadelphia, see Vireo philadelphicus
Red-eyed, see Vireo olivaceus

846

THE WILSON BULLETIN • Vol. 108, No. 4, December 1996

Warbling, see Vireo gilvus
White-eyed, see Vireo griseus
Yellow-throated, see Vireo flavifrons
vocalization

comparison of vocal repertoires of Sphen-
iscus spp., 72-79

dawn and dusk singing in male Turdus
migratorius, 507-515
discrimination between song types in Pa-
rula americana, 335-341
effect of mate removal on singing behav-
ior of Cardinalis cardinalis, 550-
555

Volatinia jacarina, 168
vole, see Microtus spp.

meadow, see Microtus pennsylvanicus
montane, see Microtus montanus
Vulpes velox, 28, 31, 33
vulpes, 130

Wallace, George E., see Kirkconnell, Arturo,

, and Orlando H. Garrido

Warbler, Bachman’s, see Vermivora bach-
mani

Bay-breasted, see Dendroica castanea
Black-and-white, see Mniotilta varia
Black-throated Blue, see Dendroica caeru-
lescens

Black-throated Green, see Dendroica vi-
rens

Blackburnian, see Dendroica fusca
Blackpoll, see Dendroica striata
Canada, see Wilsonia canadensis
Cape May, see Dendroica tigrina
Cerulean, see Dendroica cerulea
Chestnut-sided, see Dendroica pensylvanica
Golden-cheeked, see Dendroica chryso-
paria

Golden-winged, see Vermivora chrysop-
tera

Hermit, see Dendroica occidentalis
Hooded, see Wilsonia citrina
Kentucky, see Oporornis formosus
MacGillivray’s, see Oporornis tolmic
Magnolia, see Dendroica magnolia
Mourning, see Oporornis Philadelphia
Nashville, see Vermivora ruficapilla
Orange-crowned, see Vermivora celata
Palm, see Dendroica palmarum
Prairie, see Dendroica discolor
Prothonotary, see Protonotaria citrea

Swainson’s, see Limnothlypis swainsonii
Tennessee, see Vermivora peregrina
Virginia’s, see Vermivora virginiae
Wilson’s, see Wilsonia pusilla
Worm-eating, see Helmitheros vermivorus
Yellow, see Dendroica petechia
Yellow-rumped, see Dendroica coronata
Yellow-throated, see Dendroica dominica
Waterthrush, Louisiana, see Seiurus mota-
cilla

Northern, see Seiurus noveboracensis
Waxbill, Barbed, see Estrilda astrild
Waxwing, see Bombycilla spp.

Bohemian, see Bombycilla garrulus
Cedar, see Bombycilla cedrorum
weasel, see Mustela spp.

Webb, Sophie, see Howell, Steve N. G., and

Wheatley, Nigel, Where to watch birds in
South America, reviewed, 390-391
Whimbrel, see Numenius phaeopus
Whistling-Duck, Black-bellied, see Dendro-
cygna autumnalis
Lulvous, see Dendrocygna bicolor
White, Mel, A birder’s guide to Arkansas,
reviewed, 385

White-Eye, Bridled, see Zosterops conspi-
cillatus

Golden, see Cleptornis marchei
Whitney, Bret M., Jose Eernando Pacheco,
Paulo Sergio Moreira da Fonseca,
and Robert H. Barth, Jr., The nest and
nesting ecology of Acrobatornis fon-
secai (Furnariidae), with implication
for intrafamilial relationships, 434-
448

Whitney, Bret M., see Pacheco, Jose Fer-
nando, , and Luiz P. Gonzaga

Willet, see Catoptrophorus semipalmatus
Wilsonia canadensis, 542, 749
citrina, 53-60, 101, 168, 176, 190, 383-
384, 498

pusilla, 282, 283, 284

Winkler, Hans, David A. Christie, and David
Nurney, Woodpeckers, reviewed,
387-388

Winterstein, Scott R., see Millenbah, Kelly

E, , Henry Campa III, Ly T.

Furrow, and Richard B. Minnis
Wolfe, David F. G., Opportunistic winter

INDEX TO VOLUME 108

847

water acquisition by Pine Grosbeaks,
186-187

Wood-Pewee, Eastern, see Contopus virens
Western, see Contopus sordidulus
Woodhoopoe, Green, see Phoeniculus pur-
pureus

Woodpecker, Acorn, see Melanerpes formi-
civorus

Downy, see Picoides pubescens
Gila, see Melanerpes uropygialis
Hairy, see Picoides villosus
Hoffman’s, see Melanerpes hoffmanni
Ivory-billed, see Xiphorynchus flavigaster
Pileated, see Dryocopus pileatus
Red-bellied, see Melanerpes carolinus
Red-cockaded, see Picoides borealis
Red-crowned, see Melanerpes rubrucapil-
lus

Red-headed, see Melanerpes erythroce-
phalus

Smoky-brown, see Venilomis fumigatus
Wren, House, see Troglodytes aedon
Long-billed, see Cistothorus palustris
Mountain, see Troglodytes solstitialis
Rufous-naped, see Campylorhynchus ru-
finucha

Sedge, see Cistothorus platensis

Sepia-brown, see Cinnycerthia peruana
Winter, see Troglodytes troglodytes
Xenerpestes minlosi, 397-433, 446
singularis, 397-433
spp., 397-433, 434, 435, 446
Xenops rutilans, 412
Xenops, Streaked, see Xenops rutilans
Xiphorhynchus flavigaster, 168
Yahner, Richard H., and Carolyn G. Mahan,
Effects of egg type on depredation of
artificial ground nests, 129-136
Yasukawa, Ken, see Searcy, William A., and

Yellowlegs, Greater, see Tringa melanoleuca
Lesser, see Tringa flavipes
Yellowthroat, Common, see Geothlypis tri-
chas

zebra, mountain, see Equus zebra
Zenaida macroura, 313, 314, 581, 582, 688
Zicus, Michael C., and Michael R. Riggs,
Change in body mass of female Com-
mon Goldeneyes during nesting and
brood rearing, 61-71

Zimmerman, John L., review by, 389—390
Zonotrichia albicollis, 543
leucophrys, 553

Zosterops conspicillatus, 246—267

This issue of The Wilson Bulletin was published on 31 December 1996.

PUBLISHED

VOLUME 108

BY THE WILSON ORNITHOLOGICAL SOCIETY
1996 QUARTERLY

EDITOR: CHARLES R. BLEM
EDITORIAL BOARD: KATHY G. BEAL

RICHARD N. CONNER
THOMAS M. HAGGERTY
JOHN A. SMALLWOOD
INDEX EDITOR: KATHY G. BEAL
ASSISTANT EDITORS: LEANN BLEM

ALBERT E. CONWAY

EVIDENCE OF NEST PARASITISM IN MOTTLED DUCKS -

William P. Johnson, Frank C. Rohwer, and Michael Carloss 187

EIGHT NEW HOST SPECIES FOR THE PARASITIC BLOW FLY GENUS PROTOCALUPHORA (DIPTERA;

CALLIPHORIDAE) Mia Revels 189

OBSERVATIONS OF SHOREBIRD PREDATION BY SNAPPING TURTLES IN EASTERN LAKE ONTARIO

Gregory S. Pryor 190

ORNITHOLOGICAL LITERATURE 193

NUMBER 2

MAJOR PAPERS

GEOGRAPHIC VARIATION AND SPECIES LIMITS IN CINNYCERTHIA WRENS OF THE ANDES

Robb T. Brumfield and J. V. Remsen, Jr. 205

NEST ATTENTIVENESS IN HUMMINGBIRDS William H. Baltosser 228

SEASONAL POPULATION SURVEYS AND NATURAL HISTORY OF A MICRONESIAN BIRD COMMUNITY — .

Robert J. Craig 246

NATURAL HISTORY AND CONSERVATION STATUS OF THE TAMARUGO CONEBILL IN NORTHERN CHILE

Christian F. Estades 268

AVIAN ABUNDANCE IN RIPARIAN ZONES OF THREE FOREST TYPES IN THE CASCADE MOUNTAINS,

OREGON Robert G. Anthony, Gregory A. Green, Eric D. Forsman, and S. Kim Nelson 280

HABITAT CHANGES AND SUCCESS OF ARTIRCIAL NESTS ON AN ALKALINE FLAT

Marcus T. Koenen, David M. Leslie, Jr., and Mark Gregory 292

NESTING ECOLOGY OF SCISSOR-TAILED FLYCATCHERS IN SOUTH TEXAS

Kenneth R. Nolte and Timothy E. Eulbright 302

BREEDING BIOLOGY OF THE BROWN NODDY ON TERN ISLAND, HAWAII

Jennifer L. Megyesi and Curtice R. Grijfin 317

DISCRIMINATION BETWEEN REGIONAL SONG FORMS IN THE NORTHERN PARULA

Daniel J. Regelski and Ralph R. Moldenhauer 335

DISPERSAL AND HABITAT USE BY POST-FLEDGING JUVENILE SNOWY EGRETS AND BLACK-CROWNED

NIGHT-HERONS R. Michael Erwin, John G. Haig, Daniel B. Stotts, and Jejf S. Hatfield 342

NEST-SITE SELECTION OF RED-SHOULDERED AND RED-TAILED HAWKS IN A MANAGED FOREST

Christopher E. Moorman and Brian R. Chapman 357

SHORT COMMUNICATIONS

AVOIDANCE OF CABBAGE FIELDS BY SNOW GEESE J. Russell Mason and Larry Clark 369

TAXONOMIC STATUS OF THE CUBAN FORM OF THE RED-WINGED BLACKBIRD

Orlando Garrido and Arturo Kirkconnell 372

NEST ADOPTION BY MONK PARAKEETS Jessica R. Eberhard 'ilA

VERMILION FLYCATCHER AND BLACK PHOEBE FEEDING ON FISH

Brenda J. Andrews, Marie Sullivan, and J. David Hoerath 377

NEST-SITE REUSE IN THE WESTERN WOOD-PEWEE

David R. Curson, Christopher B. Goguen. and Nancy E. Mathews 378

NEST SHARING BY A LESSER SCAUP AND A GREATER SCAUP

Michael A. Eournier and James E. Hines 380

CARNIVORY OBSERVED IN THE CEDAR WAXWING — David /. King 381

A CASE OF COOPERATIVE BREEDING IN THE HOODED WARBLER

Scott A. Tarof and Bridget J. Stutchbury 382

385

ORNITHOLOGICAL LITERATURE

NUMBER 3

MAJOR PAPERS

A NEW GENUS AND SPECIES OF FURNARIID (aVES: FURNARIIDAE) FROM THE COCOA-GROWING REGION
OF SOUTHEASTERN BAHIA, BRAZIL

Jose Fernando Pacheco, Bret M. Whitney, and Luiz Gonzaga

THE NEST AND NESTING ECOLOGY OF ACROBATORN/S FONSECA! (FURNARIIDAE), WITH IMPLICATIONS
FOR INTRAFAMILIAL RELATIONSHIPS

Bret M. Whitney, Jose Fernando P acheco, Paulo Sergio Moreira da Fonseca, and

Robert H. Barth, Jr.

WOODPECKER EXCAVATION AND USE OF CAVITIES IN POLYSTYRENE SNAGS

Richard N. Conner and Daniel Saenz

NESTING SUCCESS OF THE PROTHONOTARY WARBLER IN THE UPPER MISSISSIPPI RIVER BOTTOMLANDS

D(jv/c/ J. Flaspohler

FACTORS AFFECTING FOOD PROVISIONING OF NESTLING BLACK-THROATED BLUE WARBLERS

- - Catherine O’Neill Goodbred and Richard T. Holmes

BREEDING BIOLOGY AND NATURAL HISTORY OF THE BAHAMA SWALLOW Paul E. Allen

NEOTROPICAL MIGRATORY BREEDING BIRD COMMUNITIES IN RIPARIAN FORESTS OF DIITERENT WIDTHS
ALONG THE ALTAMAHA RIVER, GEORGIA

Malcolm F. Hodges, Jr. and David G. Krementz

DAWN AND DUSK SINGING OF MALE AMERICAN ROBINS IN RELATION TO FEMALE BEHAVIOR

Tore Slagsvold

BREEDING BIOLOGY OF THE CRESTED CARACARA IN SOUTH TEXAS

Vanessa M. Dickinson and Keith A. Arnold

BREEDING BIOLOGY OF THE JABIRU IN THE SOUTHERN LLANOS OF VENEZUELA

— Jose A. Gonzalez

EFFECT OF EGG SIZE ON PREDATION BY WHITE-FOOTED MICE R. M. DeGraaf and T. J. Maier

CAN CHECKLIST PROGRAMS BE USED TO MONITOR POPULATIONS OF BIRDS RECORDED DURING THE
MIGRATION SEASON? Erica H. Dunn, Jacques Larivee, and Andre Cyr

EFFECT OF MATE REMOVAL ON SINGING BEHAVIOR AND MOVEMENT PATTERNS OF FEMALE NORTHERN
CARDINALS David B. McElroy and Gary Ritchison

RADIO TELEMETRY DOCUMENTS 24-HOUR FEEDING ACTIVITY OF WINTERING LESSER SCAUP

Christine M. Custer, Thomas W. Custer, and Daniel W. Sparks

BODY MASS AND CARCASS COMPOSITION OF FALL MIGRANT OLDSQUAWS

James O. Leafloor, John E. Thompson, and C. Davison Ankney

AVIAN NEST-SITE SELECTION AND NESTING SUCCESS IN TWO FLORIDA CITRUS GROVES

Mary Crowe Mitchell, Louis B. Best, and James P. Gionfriddo

SHORT COMMUNICATIONS

EXPONENTIAL POPULATION GROWTH OF MONK PARAKEETS IN THE UNITED STATES

Sunshine Van Bael and Stephen Pruett-Jones

FOREST GAP USE BY BREEDING BLACK-THROATED GREEN WARBLERS

Robert Smith and Matthew Dallman

COURTSHIP BEHAVIOR OF GOLDEN-CHEEKED WARBLERS Mark W. Lockwood

ORNITHOLOGICAL LITERATURE —

397

434

449

457

467

480

496

507

516

524

535

540

550

556

567

573

584

588

591

593

NUMBER 4

MAJOR PAPERS

DESCRIPTION OF ADULTS, EGGSHELLS, NESTLING, FLEDGLING, AND NEST OF THE POO-ULl

Andrew Engilis, Jr., Thane K. Pratt, Cameron B. Kepler, A. Marie Ecton,

and Kimberly M. Eluetsch 607

NESTING BEHAVIOR OF THE POO-ULi Cameron B. Kepler, Thane K. Pratt, A. Marie Ecton,

Andrew Engili.s, Jr., and Kimberly M. Eluetsch 620

POPULATION DENSITY, VOCAL BEHAVIOR, AND RECOMMENDED SURVEY METHODS FOR BICKNELL’S

THRUSH Christopher C. Rimmer, Jonathan L. Atwood, Kent P. McEarland,

and Laura R. Nagy 639

DISTRIBUTION OF BICKNELL’S THRUSH IN NEW ENGLAND AND NEW YORK

Jonathan L. Atwood, Christopher C. Rimmer, Kent P. McEarland,

Sophia H. Tsai, and Laura R. Nagy 650

MIGRATION ROUTES OF THE WESTERN SANDPIPER Robert W. Butler, Erancisco S. Delgado,

Horacio de la Cueva, Victor Pulido, and Brett K. Sandercock 662

BREEDING BEHAVIOR AND REPRODUCTIVE SUCCESS OF CERULEAN WARBLERS IN SOUTHEASTERN ON-
TARIO - Catherine J. Oliarnyk and Raleigh J. Robertson 61 'i

GRIT-USE PATTERNS IN NORTH AMERICAN BIRDS: THE INFLUENCE OF DIET, BODY SIZE, AND GENDER

James P. Gionfriddo and Louis B. Best 685

RED-COCKADED WOODPECKER NESTING SUCCESS, FOREST STRUCTURE, AND SOUTHERN FLYING SQUIR-
RELS IN TEXAS Richard N. Conner, D. Craig Rudolph, Daniel Saenz,

and Richard R. Schaefer 697

HABITAT-USE PATTERNS IN COOPERATIVE AND NON-COOPERATIVE BREEDING BIRDS: TESTING PRE-
DICTIONS WITH WESTERN SCRUB-JAYS D. Brent Burt 712

NUTRITIONAL VALUE OF WINTER FOODS FOR WHOOPING CRANES

Jay T. Nelson, R. Douglas Slack, and George E. Gee 728

TERRITORIES AND CACHING-RELATED BEHAVIOR OF RED-HEADED WOODPECKERS WINTERING IN A

BEECH GROVE Paul F. Doherty, Jr., Thomas C. Grubb, Jr., and C. L. Bronson 740

SEASONAL ABUNDANCE OF MIGRANT BIRDS AND FOOD RESOURCES IN PANAMANIAN MANGROVE

FORESTS Gaetan Lefebvre and Brigitte Poulin 748

EFFECTS OF CONSERVATION RESERVE PROGRAM FIELD AGE ON AVIAN RELATIVE ABUNDANCE, DI-
VERSITY, AND PRODUCTIVITY Kelly F. Millenbah, Scott R. Winterstein,

Henry Campa III, Ly T. Furrow, and Richard B. Minnis 760

FEMALE BUNTINGS FROM HYBRIDIZING POPULATIONS PREFER CONSPECIFIC MALES

Myron C. Baker 11 1

SURVEYS OF PUERTO RICAN SCREECH-OWL POPULATIONS IN LARGE-TRACT AND FRAGMENTED FOR-
ESTS — Keith L. Pardieck, J. Michael Meyers, and Michelle Pagan 776

THE USE OF COASTAL AGRICULTURAL FIELDS IN VIRGINIA AS FORAGING HABITAT BY SHOREBIRDS

Stephen C. Rottenborn 783

SHORT COMMUNICATIONS

EXTRA NEST SITE OCCUPANCY BY TREE SWALLOWS: DO FLOATERS AVOID NEST SITES NEAR

SETTLED PAIRS? Jeremy S. Mitchell and Raleigh J. Robertson 797

SWAINSON'S WARBLERS NESTING IN EARLY SERAL PINE FORESTS IN EAST TEXAS

N. Ross Carrie 802

MEASUREMENTS OF SNAIL KITE EGGS FROM CENTRAL FLORIDA James A. RodgerS, Jr. 804

THE ANDEAN FLAMINGO IN BRAZIL . Marcos R. Bomschein and Bianca L. Reinert 807

ORNITHOLOGICAL LITERATURE : 809

PROCEEDINGS OF THE SEVENTY-SEVENTH ANNUAL MEETING 813

825

INDEX

Thk Wilson Bulletin

Editor Chahi.ks R. Bi.km

Department of Biology
Virginia Commonwealth University
816 Park Avenue
Richmond, Virginia 23284-2012

Editorial Board KaI'HV G. Bkai,

RlCltAKI) N. CoNNKli
Thomas M. Ha(;(;kh'ia
John A. Smai.i.wood

Review Editor Wii.i.iam E. Davis, Jh.

127 East Street

Foxboro, Massachusetts 02035

Assistant Editors Lkann Bi.km Index Editor Kathv G. Bkai,

Ai.iiKliT E. Conway 616 Xenia Avenue

Yellow Springs, Ohio 45387

Su(;(;kstions to Authors

See Wilson Bulletin, 108:604-605, 1996 for more detailed “Information for Authors.”
Manuscripts intended for publication in The Wilson Bulletin should be submitted in triplicate,
neatly typewritten, double-spaced, with at least 3 cm margins, and on one side only of good
quality white paper. Do not submit xerographic copies that are made on slick, heavy paper. Tables
should be typed on separate sheets, and should be narrow and deep rather than wide and shallow.
Follow the AOU Check-list (Sixth Edition, 1983) insofar as scientific names of U.S., Canadian,
Mexican, Central American, and West Indian birds are concerned. Abstracts of major papers
should be brief but quotable. In both Major Papers and Short Communications, where fewer than
5 papers are cited, the citations may be included in the text. Follow carefully the style used in
this issue in listing the literature cited; otherwise, follow the “CBE Style Manual” (AIBS, 1983).
Photographs for illustrations should have good contrast and be on glossy paper. Submit prints
unmounted and attach to each a brief but adequate legend. Do not write heavily on the backs of
photographs. Diagrams and line drawings should be in black ink and their lettering large enough
to permit reduction. Original figures or photographs submitted must be smaller than 22 X 28 cm.
Alterations in copy after the type has been set must be charged to the author.

Notick ok Chanck ok Addkkss

If your address changes, notify the Society immediately. Send your complete new address to
Ornithological Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

The permanent mailing address of the Wilson Ornithological Society is: c/o The Museum of
Zoology, The University of Michigan, Ann Arbor, Michigan 48109. Persons having business with
any of the officers may address them at their various addresses given on the back of the front
cover, and all matters pertaining to the Bulletin should be sent directly to the Editor.

Mkmrkrship iNyUlRIKS

Membership inquiries should be sent to Dr. John Smallwood, Dept, of Biology, Montclair Slate
Univ., Upper Montclair, New Jersey 07043.

CONTENTS

MAJOR PAPERS

DESCRIPTION OF ADULTS, EGGSHELLS, NESTLING, FLEDGLING, AND NEST OF THE POO-ULI

Andrew Engilis, Jr., Thane K. Pratt, Cameron B. Kepler, A. Marie Ecton,

and Kimberly M. Fluetsch

NESTING BEHAVIOR OF THE POO-ULI Cameron B. Kepler, Thane K. Pratt, A. Marie Ecton,

Andrew Engilis, Jr., and Kimberly M. Fluetsch

POPULATION DENSITY, VOCAL BEHAVIOR, AND RECOMMENDED SURVEY METHODS FOR BICKNELL’S

THRUSH Christopher C. Rimmer, Jonathan L. Atwood, Kent P. McFarland,

and Laura R. Nagy

DISTRIBUTION OF BICKNELL’S THRUSH IN NEW ENGLAND AND NEW YORK

— Jonathan L. Atwood, Christopher C. Rimmer, Kent P. McFarland,

Sophia H. Tsai, and Laura R. Nagy

MIGRATION ROUTES OF THE WESTERN SANDPIPER Robert W. Butler, Francisco S. Delgado,

Horacio de la Cueva, Victor Pulido, and Brett K. Sandercock

BREEDING BEHAVIOR AND REPRODUCTIVE SUCCESS OF CERULEAN WARBLERS IN SOUTHEASTERN ON-
TARIO Catherine J. Oliarnyk and Raleigh J. Robertson

GRIT-USE PATTERNS IN NORTH AMERICAN BIRDS: THE INFLUENCE OF DIET, BODY SIZE, AND GENDER

James P. Gionfriddo and Louis B. Best

RED-COCKADED WOODPECKER NESTING SUCCESS, FOREST STRUCTURE, AND SOUTHERN FLYING SQUIR-
RELS IN TEXAS Richard N. Conner, D. Craig Rudolph, Daniel Saenz,

and Richard R. Schaefer

HABITAT-USE PATTERNS IN COOPERATIVE AND NON-COOPERATIVE BREEDING BIRDS: TESTING PRE-
DICTIONS WITH WESTERN SCRUB-JAYS D. Brent Burt

NUTRITIONAL VALUE OF WINTER FOODS FOR WHOOPING CRANES

Jay T. Nelson, R. Douglas Slack, and George F. Gee

TERRITORIES AND CACHING-RELATED BEHAVIOR OF RED-HEADED WOODPECKERS WINTERING IN A

BEECH GROVE Paul F. Doherty, Jr., Thomas C. Grubb, Jr., and C. L. Bronson

SEASONAL ABUNDANCE OF MIGRANT BIRDS AND FOOD RESOURCES IN PANAMANIAN MANGROVE
FORESTS Gaetan Lefebvre and Brigitte Poulin

EFFECTS OF CONSERVATION RESERVE PROGRAM FIELD AGE ON AVIAN RELATIVE ABUNDANCE, DI-
VERSITY, AND PRODUCTIVITY Kelly F. Millenbah, Scott R. Winterstein,

Henry Campa III, Ly T. Furrow, and Richard B. Minnis

FEMALE BUNTINGS FROM HYBRIDIZING POPULATIONS PREFER CONSPECIFIC MALES

Myron C. Baker

SURVEYS OF PUERTO RICAN SCREECH-OWL POPULATIONS IN LARGE-TRACT AND FRAGMENTED FOR-
ESTS Keith L. Pardieck, J. Michael Meyers, and Michelle Pagan

THE USE OF COASTAL AGRICULTURAL FIELDS IN VIRGINIA AS FORAGING HABITAT BY SHOREBIRDS

Stephen C. Rottenborn

SHORT COMMUNICATIONS

EXTRA NEST SITE OCCUPANCY BY TREE SWALLOWS: DO FLOATERS AVOID NEST SITES NEAR

SETTLED PAIRS? Jeremy S. Mitchell and Raleigh J. Robertson

SWAINSON’S WARBLERS NESTING IN EARLY SERAL PINE FORESTS IN EAST TEXAS

N. Ross Carrie

MEASUREMENTS OF SNAIL KITE EGGS FROM CENTRAL FLORIDA James A. Rodgers, Jr.

THE ANDEAN FLAMINGO IN BRAZIL Marcos R. Bomschein and Bianca L. Reinert

ORNITHOLOGICAL LITERATURE

PROCEEDINGS OF THE SEVENTY-SEVENTH ANNUAL MEETING

Index -

607

620

639

650

662

673

685

697

712

MCZ ERNST MAYR LIBRARY

1

III

II I I I II I

III

III III

2

344 118

6-

5 98£;