Full text of "The Wilson bulletin"

See other formats

HARVARD UNIVERSITY

Ernst Mayr Library
of the Museum of
Comparative Zoolop'y

-e

Tlie WIson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY
VOL. Ill, NO. 1 MARCH 1999 PAGES 1-156

(ISSN (KU3-5643)

THE WILSON ORNITHOLOGICAL SOCIETY
FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President— Edward H. Burtt, Jr., Department of Biology, Ohio Wesleyan University, Delaware, Ohio
43015; E-mail: EHBurtt@cc.owu.edu.

First Vice-President— John C. Kricher, Biology Department, Wheaton College, Norton, Massachusetts
02766; E-mail: JKricher@wheatonma.edu.

Second Vice-President— William E. Davis, Jr., College of General Studies, 871 Commonwealth Ave.,
Boston University, Boston, Massachusetts 02215; E-mail: WEDavis@bu.edu.

Editor— Robert C. Beason, Department of Biology, State University of New York, 1 College Circle,
Geneseo, New York 14454; E-mail; WilsonBull@geneseo.edu.

Secretary— John A. Smallwood, Department of Biology, Montclair State University, Upper Montclair,
New Jersey 07043; E-mail: Smallwood@satum.montclair.edu.

Treasurer— Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana 46556;
E-mail: DWatt@saintmarys.edu.

Elected Council Members — Peter C. Frederick and Danny J. Ingold (terms expire 1999), Charles R. Blem
and Sara R. Morris (terms expire 2000), Jonathan L. Atwood and James L. Ingold (terms expire

2001).

Membership dues per calendar year are: Active, $21.00; Student, $15.00; Family, $25.00; Sustaining,
$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 m the University
of Michigan Museum of Zoology, was established in concurrence with the University of Michigan m
1930. Until 1947 the Library was maintained entirely by gifts and bequests of books, reprints, and orni-
thological magazines from members and friends of the Society. Two members have generously established
a fund for the purchase of new books; members and friends are invited to maintain the fund by regular
contribution. The fund will be administered by the Library Committee. William A. Lunk, University of
Michigan is Chairman of the Committee. The Library currently receives over 200 periodicals as gifts and
in exchange for The Wilson Bulletin. For information on the library and our holdings, see the Society’s
web page at http://www.ummz.lsa.umich.edu/birds/wos.html. 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: Josselyn Van Tyne Memorial Library, Museum of Zoology, The
University of Michigan, 1 109 Geddes Ave., Ann Arbor, MI 48109-1079 USA. Contributions to the New
Book Fund should be sent to the Treasurer.

THE WILSON BULLETIN
(ISSN 0043-5643)

THE WILSON BULLETIN (ISSN 004.^-5643) i.s published quarterly in March. June, September, and December by
the Wilson Ornithological Society. 810 East 10th Street. Lawrence, KS 66044-8897. The .subscription price, both in the
United States and elsewhere, is S40.00 per year. Periodicals postage paid at Lawrence. KS. POSTMASTER: Send address
changes to THE WILSON BULLETIN. P.O. Box 1897, Lawrence, KS 66044-8897.

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.

All articles and communications for publications, books and publications for reviews should be addressed to the Editor.
Exchanges should be addressed to The Josselyn Van Tyne Memorial Library. Museum of Zoology. Ann Arbor, Michigan

48 1 09

Subscriptions, changes of addre.ss and claims for undelivered copies should be .sent to the OSNA, P.O. Box 1897.
Lawrence. KS 66044-8897. Phone: (785) 843-1221; FAX; (785) 843-1274.

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 JOURNAL OF ORNITHOLOGY
Published by the Wilson Ornithological Society

VOL. Ill, NO. 1 MARCH 1999 PAGES 1-156

Wilson Bull., 111(1), 1999, pp. 1—6

ANNUAL SURVIVAL RATES OF FEMALE HOODED
MERGANSERS AND WOOD DUCKS IN
SOUTHEASTERN MISSOURI

KATIE M. DUGGER,' 23 BRUCE D. DUGGER,' ^ AND LEIGH H. FREDRICKSON'

ABSTRACT. — Successful conservation and management, particularly of harvested species, relies on accurate
estimates of population demographics. In addition, estimates of survival and longevity allow more accurate
modeling of evolutionary life-history trade-offs within and between species. We estimated survival rates for box
nesting female Hooded Mergansers {Lophodytes cucullatus) and Wood Ducks (Ai.x sponsa) in southeastern
Missouri during 1987-1997 and 1987-1993, respectively. Hooded Merganser survival rates varied annually and
ranged from 0.42-1.0 (jc = 0.66 ± 0.04). Wood Duck survival did not vary significantly over time and averaged
0.63 (± 0.02). Mean annual survival rates and capture probabilities were similar for the two species (x^ = 0.49,
df = \, P > 0.05; = 0.02, df = 1, P > 0.05). Annual variation in Hooded Merganser survival rates was an

important component of this species’ population ecology, but was not related to winter weather conditions,
harvest rates, breeding season rainfall, or nesting parameters. Our female Wood Duck survival rates were higher
than survival estimates for other adult females in the north-central subpopulation, but were comparable to some
estimates for adult females that breed in southern and mid- Atlantic states. Received 12 May 1998, accepted 5

Sept. 1998.

Estimates of annual survival are important
for comparative studies of life-history strate-
gies between species (Krementz et al. 1989)
and for modeling population demographics for
conservation and management. Annual sur-
vival rate estimates are available for certain
geese (tribe Anserini) and dabbling ducks
(tribe Anatini) traditionally important to hunt-
ers (Johnson et al. 1992). Except for the Com-
mon Eider {Somateria molUssima-, Krementz
et al. 1996), survival estimates based on mod-
em survival estimation procedures are com-
pletely lacking for most seaducks (tribe Mer-
gini), despite increased harvest pressure in re-

‘ Gaylord Memorial Laboratory, School of Natural
Resources, Univ. of Missouri, Puxico, MO 63690.

^ Present address: Cooperative Wildlife Research
Laboratory, Southern Illinois Univ., Carbondale, IL
62901.

^ Corresponding author; E-mail:
tinamou @ midwest.net

cent years (U.S. Fish and Wild. Serv., unpubl.
data).

Hooded Mergansers {Lophodytes cuculla-
tus) are among the least studied of all Mergini
(Dugger et al. 1994). Attempts to estimate sur-
vival rates are difficult because of their low
harvest rate, secretive behavior, year-round
occurrence in low densities, and preference
for forested wetlands. However, Hooded Mer-
gansers nest in man-made boxes, and capture
of these females can provide mark-recapture
samples large enough to estimate survival
rates (Dugger et al. 1994).

Over much of their range Hooded Mergan-
sers occur sympatrically with Wood Ducks
{Aix sponsa, tribe Anatini; Livezey 1986) pro-
viding an opportunity for comparison of sur-
vival estimates. The two species are similar in
body size (HM, 540-725 g; WD, 530—680 g;
Bellrose and Holm 1994), both nest in tree
cavities, and both rely on forested wetlands

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

during the breeding season. However, these
species differ considerably in many aspects of
their biology and might be expected to exhibit
differences in annual survival rates. Hooded
Mergansers are carnivorous, forage by diving,
and exhibit delayed maturation (first breed at
^2 years old; Dugger et al. 1994). In contrast.
Wood Ducks are omnivorous, forage near the
water’s surface, and most breed as yearlings
(Bellrose and Holm 1994). Based on the dif-
ference in age at first breeding and phylogeny
(Krementz et al. 1997), we predict that Hood-
ed Mergansers experience higher annual sur-
vival rates than Wood Ducks (Ricklefs 1973,
Wittenberger 1979). In this paper we estimate
annual survival rates of box nesting female
Hooded Mergansers and Wood Ducks in
southeastern Missouri and compare mean an-
nual survival rates of these two species.

STUDY AREA AND METHODS

The study was conducted on the Duck Creek Wild-
life Conservation Area (WCA) and Mingo National
Wildlife Refuge (NWR) in southeastern Missouri.
These adjoining areas comprise Mingo Swamp, the
largest contiguous block of bottomland hardwood for-
est in Missouri (1 1,174 ha). For a detailed description
of the habitat types available in Mingo Swamp see
Heitmeyer and coworkers (1989). Approximately 85
boxes were available to nesting females in all years of
the study on Mingo NWR, and 100-120 nest boxes
were available for nesting birds on Duck Creek WCA
during 1987-1993, and most nest boxes were equipped
with predator guards to reduce predation by raccoons
(Procyort lotor). After the nesting season in 1993, ap-
proximately 50 boxes were removed on Duck Creek
WCA and the Wood Duck nesting study was termi-
nated. Capture of Hooded Mergansers continued on
both Mingo NWR and Duck Creek WCA through
1997.

Nest boxes were cleaned, repaired when necessary,
and filled with wood chips before each nesting season.
Hooded Mergansers and Wood Ducks were captured
in nest boxes between 1 February and 15 August each
year during 1987-1997 and 1987-93, respectively. We
checked boxes at 2-4 week intervals and captured
nesting females of both species during the third week
of incubation. Unmarked females were banded with
U.S. Fish and Wildlife Service leg bands, and band
numbers of previously marked birds were recorded.
All breeding females of both species captured in nest
boxes were u.sed in our analysis, including Wood
Ducks that bred as yearlings. Although both Hooded
Mergan.ser and Wood Duck ducklings were web-
tagged in boxes at hatching, .sample sizes of known-
age birds were too small to analyze by age class.

We constructed mark-recapture matrices for Hooded
Mcrgan.sers ( I 1 years) and Wood Ducks (7 years) and

TABLE 1. The number of Hooded Merganser
nesting attempts and nest success in southeastern Mis-
souri 1987-1997.

Year

Nest attempts*

Nest success
rate*’ (%)

1987

88.2

1988

84.4

1989

62.2

1990

77.1

1991

73.0

1992

35.6

1993

39.6

1994

51.5

1995

73.7

1996

58.1

1997

56.0

“ Total number of nests initiated by Hooded Mergansers in southeast Mis-
souri.

^ Hooded Merganser nest success (number of successful nests/total num-
ber of nest attempts).

used Jolly-Seber mark-recapture models for open pop-
ulations to estimate survival for both species (Pollock
et al. 1990). Program JOLLY computes point esti-
mates, their associated variances, goodness-of-fit tests,
and likelihood ratio tests for five open population mod-
els (Pollock et al. 1990). We used model goodness-of-
fit tests and likelihood ratio tests between models to
select the model that provided the best fit for each data
set (Pollock et al. 1990). We then compared Wood
Duck and Hooded Merganser mean annual survival
rates and capture probabilities using the program Con-
trast (Hines and Sauer 1989) and the methods de-
scribed by Sauer and Williams (1989). All analyses
were performed on an IBM computer under DOS.

Our analysis suggested time dependent variation
was an important component of Hooded Merganser
survival, so we attempted to identify factors that might
be correlated with merganser annual survival rates. We
correlated Hooded Merganser survival estimates for
each year with annual harvest (USFWS, unpubl. data),
winter weather conditions (rainfall, temperature) in the
Mississippi Alluvial Valley, and rainfall in Mingo
Swamp from March through September. We also cor-
related Hooded Merganser survival with the number
of nesting attempts and nest success on Mingo NWR
and Duck Creek WCA (Table 1 ). These reproductive
variables might be expected to index local Hooded
Merganser nesting density, and therefore represent the
potential for permanent emigration from our study
sites.

Although specific wintering areas for birds in our
population arc largely unknown, the Mississippi Al-
luvial Valley is the closest region with suitable win-
tering habitat, and females nesting in Mingo Swamp
have been recovered from this region (Dugger et al.
1994). We used mean daily temperature and monthly
rainfall totals recorded at two sites in each of three
states (Arkansas, Mississippi, and Louisiana) October

Du}i}>er et al. • MERGANSER AND WOOD DUCK SURVIVAL

TABLE 2. Annual survival estimates and associated capture probabilities for female Hooded Mergansers
and Wood Ducks nesting in southeastern Missouri.

Survival probability (SE)

Capture probability (SE)

Year

Hooded Merganser

Wood Duck

Hooded Merganser

Wood Duck

1987

0.58 (0.17)

1988

0.86 (0.16)

0.78 (0.19)

0.71 (0.07)

1989

1.00 (0.25)

0.43 (0.13)

0.77 (0.05)

1990

0.42 (0. 1 1 )

0.37 (0.12)

0.55 (0.05)

1991

0.97 (0.15)

0.53 (0.12)

0.48 (0.05)

1992

0.54 (0.12)

0.69 (0.13)

0.51 (0.06)

1993

0.55 (0.12)

0.59 (0.13)

0.66 (0.08)

1994

0.52 (0.13)

0.71 (0.14)

1995

0.45 (0.12)

0.65 (0.15)

1996

0.66 (0.16)

Mean

0.66 (0.04)

0.63 (0.02)

0.60 (0.05)

0.61 (0.04)

^ Recapture data from these years were available to estimate constant survival rale for Wood Ducks.

through January for each year (National Climatic Data
Center) to index winter habitat conditions. We used
total rainfall during March through September collect-
ed at Advance, Missouri (National Climatic Data Cen-
ter) to index local habitat conditions during the breed-
ing and post-breeding season when Hooded Mergan-
sers are present in Mingo Swamp.

RESULTS

We used individual capture histories of 151
Hooded Merganser and 512 Wood Duck fe-
males to estimate annual survival. Model A,
from Program JOLLY, with time-dependent
capture probabilities and survival rates pro-
vided the best fit for the Hooded Merganser
data (x^ = 19.61, df = 12, P > 0.05). Model
A also fit the Wood Duck data (x^ = 17.52,
df = 12, P > 0.05) as did Model B, a reduced
parameter model with constant survival rates
and time-dependant capture probability (x^ =
18.41, df = 16, P > 0.05). The likelihood ra-
tio test between Models A and B (x^ = 0.90,
df = 4, P > 0.05) suggested Model B provid-
ed the most parsimonious fit for the Wood
Duck data. Jolly-Seber models estimate sur-
vival through sample k-2, and capture proba-
bility for samples 2 through k-l (model A) or
k (model B). Thus, we had 9 estimates of an-
nual survival and capture probability for
Hooded Mergansers (1987-1995) and 6 esti-
mates of capture probability (1988-1992) for
Wood Ducks with a single estimate of con-
stant survival (Table 2). Hooded Merganser
survival rates ranged from a low of 0.42 in
1990 to a high of 1.0 in 1989 with a mean of
0.66 (± 0.04 SE) with 95% confidence limits

of 0.59-0.73 (Table 2). Wood Duck annual
survival was 0.63 (2; 0.02) with 95% confi-
dence limits of 0.59—0.68 (Table 2). The mean
annual survival of Hooded Mergansers was
not significantly different than Wood Duck an-
nual survival (x^ = 0.49, df = 1, P > 0.05).

Capture probability for Hooded Mergansers
varied annually from a low of 0.37 in 1990 to
a high of 0.78 in 1988 (Table 2). Wood Duck
capture rates also exhibited annual variation,
ranging from 0.51 in 1992 to 0.77 in 1989
(Table 2). Mean capture probabilities did not
differ between species ~ 0.02, df = 1, P
> 0.05). In addition, we observed no signifi-
cant correlation between Hooded Merganser
survival rates and annual harvest, winter
weather conditions (temperature and rainfall),
breeding season rainfall, or Hooded Mergan-
ser nesting parameters (all P > 0.05).

DISCUSSION

Hooded Merganser and Wood Duck annual
survival estimates from our study were gen-
erally higher than those reported for other
duck species (Johnson et al. 1992). However,
because our estimates were for birds using
nest boxes with some protection from preda-
tory raccoons, comparisons with other species
must be made with caution. Nevertheless,
Hooded Mergansers in our study had substan-
tially lower mean annual survival rates than
the Common Eider (x = 0.87; Krementz et al.
1996), the only other member of Mergini for
which estimates are available. Yearly survival
rates for Hooded Mergansers during 1988,

THE WILSON BULLETIN • Vol. HI, No. 1, March 1999

1989, and 1991 were comparable with those
for Common Eiders (Krementz et al. 1996);
rates for other years were substantially lower.
Wood Duck survival in our study was higher
than or comparable to other estimates avail-
able for female Wood Ducks (Johnson et al.
1986, Nichols and Johnson 1990, Kelley
1997), including estimates from a South Car-
olina box-nesting population (T = 0.55, Hepp
et al. 1987). Wood Duck females are strongly
philopatric to nest sites (nearly 100%), so
mark-recapture survival estimates that include
a measure of capture site fidelity (nest boxes
in this case) can be comparable to band re-
covery estimates (Hepp et al. 1987). Whether
our high survival rates reflect general regional
differences in survival of eastern Wood Ducks
(Nichols and Johnson 1990) or a survival ben-
efit associated with box-nesting remains un-
clear. Wood Ducks breeding in Missouri are
included in the “north-central” sub-popula-
tion of Bowers and Martin (1975) and Kelley
(1997), but exhibit survival rates much higher
than band recovery estimates for adult females
in this region (Kelley 1997). Our Wood Duck
females exhibited survival rates most similar
to adult females in the southern population of
the Mississippi Fly way {x = 0.61; Kelley
1997) and the mid- Atlantic population of the
Atlantic Flyway (T = 0.63; Kelley 1997).

Mean survival did not differ between
Hooded Mergansers and Wood Ducks, al-
though in three of nine years Hooded Mer-
ganser survival was higher than the constant
rate estimated for Wood Ducks (Table 2). This
is inconsistent with life-history theory which
predicts that birds with delayed maturation
should experience higher annual survival
(Wittenberger 1979), but consistent with anal-
yses showing survival rates are correlated
with body size and breeding latitude in wa-
terfowl (Arnold 1988). Maybe more important
than the comparison of mean survival rates
was our observation that annual variation was
an important component of Hooded Mergan-
ser, but not Wood Duck, survival. Differences
in diet, foraging method, and habitat require-
ments may make Hooded Mergansers more
.sensitive to local fluctuations in food resourc-
es or water conditions during reproduction or
winter, with increa.sed mortality or emigration
during years when habitat conditions are poor.
None of the harvest, breeding sea.son rainfall.

density-dependent factors, or winter weather
variables we investigated were significantly
related to variation in Hooded Merganser sur-
vival. However, we did not quantify food
availability on the breeding grounds directly
and because information on migration patterns
and winter site fidelity for Hooded Mergan-
sers is lacking we may not have compared our
Hooded Merganser survival rates with the
most appropriate winter or breeding season
weather conditions. Very little information ex-
ists concerning Hooded Merganser foraging
ecology or foraging habitat characteristics
(Dugger et al. 1994), consequently, determin-
ing the climatic factors that index habitat con-
ditions throughout the Hooded Merganser an-
nual cycle will be difficult.

We believe our survival estimates for both
species are unbiased, but some behavioral re-
sponses such as permanent emigration from
the study area cannot be distinguished from
“deaths” by Jolly-Seber models and can re-
sult in negatively biased survival estimates
(Pollock et al. 1990). Wood Ducks are strong-
ly philopatric (Hepp et al. 1987), but data are
lacking to estimate philopatry for Hooded
Mergansers or to make direct quantitative
comparisons with Wood Ducks. We believe
that Hooded Mergansers are strongly philo-
patric to general nesting areas (e.g., Zicus
1990), but not as philopatric as Wood Ducks
to specific nesting boxes. Factors that might
have caused Hooded Merganser females to
have left the study site or chosen not to nest
in boxes in subsequent years (low nest success
or high breeding density) were not correlated
with annual survival as we might expect if
permanent emigration were common. Further-
more, in Minnesota, distances moved by
Hooded Mergansers between nesting sites
each year were not related to nest success (Zi-
cus 1990), suggesting that variation in nest
success does not affect philopatry. Habitat
conditions in Mingo Swamp could have af-
fected Hooded Merganser use of boxes, but it
is unlikely that these effects would be per-
manent. Finally, we do not believe that nest
boxes were limiting for Hooded Mergansers
or that competition for nest sites led to higher
permanent emigration by Hooded Mergansers
from Mingo Swamp. Hooded Merganser pop-
ulations were substantially lower than Wood
Duck populations each year (Dugger 1991)

Dui>}>er el al. • MERGANSER AND WOOD DUCK SURVIVAL

and the annual number of Hooded Merganser
nest attempts remained low in relation to box
availability. In addition. Hooded Mergansers
initiate nests on average 3-4 weeks earlier
than Wood Ducks, thereby reducing the po-
tential for nest site competition between the
two species (Dugger et al. 1994).

Overall mean Hooded Merganser survival
in this study was lower than a previous esti-
mate (Dugger et al. 1994), and this earlier
analysis did not detect any significant annual
variation in survival rates. As more data were
collected and added to the analysis, estimate
precision increased, mean annual survival de-
creased, and annual variation became an im-
portant component of Hooded Merganser sur-
vival. The increased variation in annual sur-
vival estimates and changes in the long-term
mean associated with additional years of study
on Hooded Mergansers illustrate the impor-
tance of long-term data sets. Continued band-
ing effort and more information concerning
Hooded Merganser habitat use, foraging ecol-
ogy, and age-specific survival rates are needed
to understand the factors affecting annual var-
iation in survival of this species.

ACKNOWLEDGMENTS

We thank L. Bollman, R Blum, J. H. Gammonley,
J. R. Kelley, M. Shannon, J. Ware, and J. S. Wortham
for assistance with data collection and the staffs at
Duck Creek WCA and Mingo NWR for access to their
nest boxes. Comments provided by P. Blums, J. Gam-
monley, R. Kennamer and an anonymous reviewer
greatly improved earlier drafts of this manuscript. This
research was funded by Gaylord Memorial Laboratory,
Patuxent Wildlife Research Center, and the Missouri
Cooperative Wildlife Research Unit at the University
of Missouri, Columbia. This is a contribution from the
Missouri Agricultural Experiment Station, Journal Se-
ries Number 12, 771.

LITERATURE CITED

Arnold, T A. 1988. Life histories of North American
game birds; a reanalysis. Can. J. Zool. 66:1906-
1912.

Bellrose, F. C. and D. j. Holm. 1994. Ecology and
management of the Wood Duck. Stackpole Books,
Harrisburg, Pennsylvania.

Bowers, E. E and E W. Martin. 1975. Managing
Wood Ducks by population units. Trans. N. Am.
Wildl. Nat. Resour. Conf. 40:300-324.

Dugger. B. D., K. M. Dugger, and L. H. Fredrick-
son. 1994. Hooded Merganser (Lophodytes cu-
cullatus). In The birds of North America, no. 98
(A. Poole and F Gill, Eds.). The Academy of Nat-

ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Dugger, K. M. 1991. Nesting parameters and popu-
lation dynamics of box-nesting female wood
ducks in Mingo Swamp. M.S. thesis. Univ. of
Missouri, Columbia.

Heitmeyer, M. E., L. H. Fredrickson, and G. F.
Krause. 1989. Water and habitat dynamics of the
Mingo Swamp in southeastern Missouri. Fish
Wildl. Res. 6:1-26.

Hepp, G. R., R. T. Hoppe, and R. A. Kennamer. 1987.
Population parameters and philopatry of breeding
female Wood Ducks. J. Wildl. Manage. 51:401 —
404.

Hines, J. E. and J. R. Sauer. 1989. Program CON-
TRAST— a general program for the analysis of
several survival or recovery rate estimates. Fish
Wildl. Tech. Report 24:1-7.

Johnson, D. H., J. D. Nichols, and M. D. Schwartz.
1992. Population dynamics of breeding waterfowl.
Pp. 446-485 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.

Johnson, F. A., J. E. Hines, F. Montalbano, III, and
J. D. Nichols. 1986. Effects of liberalized harvest
regulations on Wood Ducks in the Atlantic Fly-
way. Wildl. Soc. Bull. 14:383-388.

Kelley, J. R., Jr. 1997. Wood duck population mon-
itoring initiative; final report. Atlantic Flyway
Council, Mississippi Flyway Council, and U.S.
Fish and Wildlife Service, Laurel, Maryland.

Krementz, D. G., R. j. Barker, and J. D. Nichols.
1997. Sources of variation in waterfowl survival
rates. Auk 114:93-102.

Krementz, D. G., J. E. Hines, and D. F Caithamer.
1996. Survival and recovery rates of American
Eiders in eastern North America. J. Wildl. Man-
age. 60:855-862.

Krementz, D. G., J. R. Sauer, and J. D. Nichols.
1989. Model-based estimates of annual survival
rate are preferable to observed maximum lifespan
statistics for use in comparative life-history stud-
ies. Oikos 56:203—208.

Livezey, B. C. 1986. A phylogenetic analysis of recent
anseriform genera using morphological characters.
Auk 103:737-754.

Nichols, J. D. and F A. Johnson. 1990. Wood Duck
population dynamics: a review. Pp. 83-105 in
Proc. 1988 North Am. Wood Duck Symp. (L. H.
Fredrickson, G. V. Burger, S. P. Havera, D. A.
Grabcr, R. E. Kirby, and T. S. Taylor. Eds.). St.
Louis, Missouri.

Pollock, K. H., J. D. Nichols, C. Brownie, and J. E.
Hines. 1990. Statistical inference of capture-re-
capture experiments. Wildl. Monogr. 107; 1—97.

Ricklefs, R. E. 1973. Fecundity, mortality and avian

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

demography. Pp. 366—435 in Breeding biology of
birds (D. S. Earner, Ed.). National Academy of
Sciences, Washington, D.C.

Sauer, J. R. and B. K. Williams. 1989. Generalized
procedures for testing hypotheses about survival
or recovery rates. J. Wildl. Manage. 53:137-142.

WiTTENBERGER, J. E 1979. A model for delayed repro-
duction in iteroparous animals. Am. Nat. 114:
439-446.

Zicus, M. C. 1990. Nesting biology of Hooded Mer-
gansers using nest boxes. J. Wildl. Manage. 54:
637-643.

Wilson Bull: 111(1), 1999, pp. 7-14

COMPARATIVE NEST SITE HABITATS IN SHARP-SHINNED AND

COOPER’S HAWKS IN WISCONSIN

DALE R. TREXEL,' ROBERT N. ROSENFIELD,^ ^ JOHN BIELEFELDT\ AND

EUGENE A. JACOBS^

ABSTRACT. — From an analysis of nest site habitat data at 24 Sharp-shinned Hawk (Accipiter striatus) and
52 Cooper’s Hawk (A. cooperii) nests in Wisconsin, we conclude that Cooper’s Hawks tend to nest in stands
with lower densities of taller and larger trees than do Sharp-shinned Hawks, and that Cooper s Hawks also tend
to nest in sites with a greater proportion of hardwood cover than Sharp-shinned Hawks. Significant interspecific
differences were found in combined habitat types (hardwoods, mixed conifer-hardwoods, and conifer plantations)
for nest tree height and nest tree DBH (diameter at breast height); nest height; nest height relative to tree height;
canopy height; canopy cover; tall shrub density; tree density; and mean DBH. Nest sites of the two species were
similar in terms of understory canopy cover, ground cover, low shrub index, understory tree density, basal area,
distance to nearest forest opening, and distance to water. We detected few significant intraspecific differences in
nest site habitat, and these only in the Cooper’s Hawk. Received 23 Oct. 1997, accepted 4 Nov. 1998.

Although the Sharp-shinned Hawk {Accip-
iter striatus) and the Cooper’s Hawk (A. coop-
erii) breed sympatrically in many parts of the
United States and southern Canada, their nest
site habitats have been compared in only four
published quantitative studies. These conge-
ners are sometimes assumed to partition nest-
ing habitat by way of interspecific competition
and/or predation (Siders and Kennedy 1996).
With one exception in Missouri (Wiggers and
Kritz 1991), these studies were conducted in
the western United States (Oregon: Reynolds
et al. 1982, Moore and Henny 1983; New
Mexico: Siders and Kennedy 1996).

Such geographically restricted results may
be difficult to extrapolate to other areas of
sympatry because of regional differences in
vegetational composition and structure. Each
of the previously published comparisons of
these hawks’ nest site habitats was derived
from upland forests with relatively homoge-
neous vegetation, principally montane conifer
forests in New Mexico and Oregon, and co-
nifer plantations or oak-hickory forests in
Missouri. Our study area (Fig. 1) was the state

‘ Dept, of Ecology, Evolution, and Behavior, 100
Ecology Building, 1987 Upper Buford Circle, Univ. of
Minnesota, St. Paul, MN 55108.

^ Dept, of Biology, Univ. of Wisconsin, Stevens
Point, WI 5448 1 .

^ Park Planning, Racine County Public Works Di-
vision, Sturtevant, WI 53177.

■* Linwood Springs Research Station, 1601 Brown
Deer Lane, Stevens Point, WI 54481.

’Corresponding author; E-mail: rrosenfi@uwsp.edu

of Wisconsin (145,000 km^). The ecologically
diverse set of available woodland nesting hab-
itats on this statewide scale includes boreal
conifer forests (plus conifer swamps of boreal
affinity over much of the state), conifer plan-
tations, mixed conifer-hardwood forests, pure-
ly deciduous woodlands on upland and low-
land sites, and highly fragmented or urban
woodlands (Rosenfield et al. 1996) as well as
extensive forests. For further details on Wis-
consin forests see Curtis (1959).

Potentially conflicting results among past
studies may also limit their utility in unstudied
areas of sympatry. In New Mexico, for ex-
ample, Siders and Kennedy (1996) found sig-
nificant differences between Sharp-shinned
Hawks and Cooper’s Hawks in the majority of
nest site variables tested, while in Oregon,
both Reynolds and coworkers (1982) and
Moore and Henny (1983) found few discern-
ible differences in nest site characterisitics be-
tween these accipiters. Furthermore, Siders
and Kennedy (1996) have suggested that in-
terpretations of previous results may be ham-
pered by small sample sizes, especially for
Sharp-shinned Hawks (n < 18 nests in prior
studies), and by possible biases in nest search
methods or methods of selecting search areas.

We compare habitat at 24 Sharp-shinned
Hawk nests and 52 Cooper’s Hawk nests in
Wisconsin, 1980-1994, all discovered by un-
biased means. Previous comparative work on
nest site habitats of these two hawks has em-
phasized interspecific differences within rela-
tively uniform habitat types. We expand this

THE WILSON BULLETIN • VoL 111, No. I. March 1999

FIG. 1. Distribution by county of nest sites sampled for Accipiter cooperii (circled) and A. striatus (not
circled) in Wisconsin.

emphasis to include intraspecific similarities
as well as interspecific differences across hab-
itat types (i.e., combined habitats) at a land-
scape scale. Intraspecific nest site features
held in common across habitat types may aid
land management agencies in assessing and
conserving a range of usable breeding habitats
for Sharp-shinned and Cooper’s hawks. Our
results seem timely and pertinent to the recent
Birds in Forested Landscape project for North
America (Cornell Lab of Ornithology), which
focuses in part on the nesting habitats of these
two hawks, and is designed to develop man-
agement and conservation strategies on their
behalf (Anonymous 1997).

METHODS

Ne.st locations. — Nest site locations were considered
unbia.sed if they were discovered by one of two meth-
ods: ( 1 ) incidental or random locations obtained by
cooperators during any activity other than .searching
for accipiter nests, and (2) locations resulting from
Cooper's Hawk density studies in which objectively

drawn study areas were completely searched regardless
of their perceived suitability for nesting and without
foreknowledge of current or historical nest sites on
these areas. By these methods, we located Cooper’s
Hawk nests in Wisconsin (see Fig. 1) on 52 widely
separated, independent nesting areas, as defined in Ro-
senfield and Bielefeldt (1992, 1996). All 24 Sharp-
shinned Hawk nests occurred in independent nesting
areas; therefore each was included in our analyses.

Data collection and analyses. — Habitat measure-
ments (Table 1 ) were made postfiedging at each nest
site within a 0.04 ha circular plot centered on the nest
tree following the technique of James and Shugart
(1970) as modified by Titus and Mosher (1981).

All variables were tested for normality with Lillie-
fors test, further statistical analyses were performed on
SYSTAT (Wilkinson 1992). /-tests were used exclu-
sively to examine interspecific differences among sev-
en habitat varables that exhibited normal distributions
in combined habitats (i.e., tree height, tree DBH, nest
height, nest percent, canopy height, total canopy, and
mean DBH; Table 2). We used the Mann-Whitney La-
test for all other inferential comparisons because all
other variables were not normally distributed.

To examine inter- and intraspecific differences and

Trexel et al. • ACCIPITER NEST SITE HABITATS

TABLE 1. Habitat variables and measurement tecliniques at Accipiter strialus and A. coopcrii nest sites in
Wisconsin.

Variable

Description

Tree height
Tree DBH
Nest height
Nest percent
Canopy height
Total canopy
Deciduous can.
Coniferous can.
Understory can.
Ground cover
Shrub density
Shrub index
Tree density
Under, dens.
Basal area
Mean DBH
Dist. to water
Dist. to open.

Height (m) of nest tree (Haga altimeter)

Diameter (cm) at breast height of nest tree
Height (m) of nest (meter tape or Haga altimeter)

(Nest height/Tree height) X 100

Mean height (m) of five canopy trees in study plot (Haga altimeter)

Percent of area over study plot occluded by overstory foliage'’

Percent of area over plot (not of total canopy) occluded by deciduous overstory foliage''
Percent of area over plot occluded by evergreen overstory foliage"

Percent of area over plot occluded by understory foliage"

Percent of ground in plot covered by ground-layer foliage"

Index of tall shrubs < 3 cm DBH and ^ shoulder height* **

Index of low shrubs < 3 cm DBH between knee and shoulder height**

Number of canopy trees ^ 9 cm DBH per hectare
Number of understory trees ^ 9 cm DBH per hectare
mV ha of canopy trees

Mean DBH (cm) of canopy trees in study plot

Distance (m) to nearest permanent water source (pacing or USGS 7.5 min. quadrangles)
Distance (m) to nearest forest opening > 5 ha (pacing or USGS 7.5 min. quadrangles)

“ 40 ocular tube readings.

** Sum of four plot radii.

similarities among habitat types, we separated our nest
site samples into three categories based on trees pre-
sent within the 0.04 ha plot. We first divided nest site
samples between those occurring within conifer plan-
tations and those not in plantation habitats. (While de-
ciduous trees occurred in some conifer plantations, no
nest sites occurred in hardwood plantations.) We then
divided non-plantation nest sites into those situated in
pure hardwood stands (where no trees within the study
plot were conifers) and those in mixed conifer-hard-
wood stands. In keeping with our statewide sample,
these three habitat categories should be construed as
physiognomic types that do not necessarily exhibit oth-
er internal similarities in vegetational attributes. In
hardwood stands, for example, dominant or prominent
tree species might include oaks (Quercus spp.), maples
(Acer spp.), aspen (Populus spp.), and other species of
varied ages, management histories, and moisture re-
gimes. Mixed woodlands might include lowland co-
nifers such as tamarack (Lxirix larcina) and black
spruce (Picea rnariana) or upland conifers such as
pines (Pinus spp.) as well as deciduous species. For
compositional variety among nest tree species (and sci-
entific names) see Table 3.

Because there was only one pure hardwood site used
by Sharp-shinned Hawks, we examined interspecific
differences only within conifer plantations and mixed
conifer-hardwood habitats. Likewise, we could only
test for intraspecific differences between mixed and
plantation habitats among Sharp-shinned Hawk nest
sites. We used the nonparametric Kniskall-Wallis test
to examine interspecific differences among the three
habitat types used by Cooper’s Hawks. Because of the
number of multiple univariate comparisons (Table 2),

we calculated that an alpha of 0.001 was the appro-
priate level of significance for both inter- and intraspe-
cific inferences (Sokal and Rohlf 1981).

RESULTS

The majority of the 18 nest site variables
compared for Sharp-shinned and Cooper’s
hawks in combined habitats showed statisti-
cally significant interspecific differences (Ta-
ble 2). Nest tree DBH in conifer plantations,
and nest height and canopy height in mixed
conifer-hardwood stands were significantly
different between species across uncombined
habitat types (Table 2).

Of the 18 variables examined only four ex-
hibited significant intraspecific differences
across habitat types, and only in Cooper’s
Hawk (Table 2). Although intraspecific nest
site selection itself might vary among habitats,
we speculate that these statistical differences
instead are attributable to inherent vegetation-
al contrasts among habitat types as circum-
scribed here. In the most transparent example,
cross-habitat intraspecific differences in per-
centages of coniferous and deciduous canopy
covers at Cooper’s Hawk nest sites in hard-
wood stands versus pine plantations (Table 2)
are a predictable outcome of our habitat cat-
egories. The more interesting result of intra-

THE WILSON BULLETIN • Vol. Ill, No. I. March 1999

TABLE 2. Mean values
Wisconsin.

of habitat variables measured

at Accipiter striatus and A.

cooperii nest sites in

Variable

A. striatus (mean ± SE)

Combined

Mixed

Plantation

(n = 24)“

(n = 11)

(» = 12)

Tree height (m)

15.1 ± 0.6

14.8 ± 0.6

15.8 ± 1.0

Tree DBH (cm)

23.8 ± 1.4

26.2 ± 2.3

21.4 ± 1.6

Nest height (m)

9.1 ± 0.6‘>

7.9 ± 0.7”

10.4 ± 0.8

Nest percent (%)

59.7 ± 2.7'’

54.1 ± 4.3”

65.5 ± 2.7

Canopy height (m)

15.2 ± 0.6

15.0 ± 0.6

15.8 ± 0.9

Total canopy (%)

76.5 ± 2.3

74.3 ± 2.5

80.4 ± 3.3

Deciduous can. (%)

15.2 ± 4.4

24.3 ± 7.7

3.8 ± 1.7

Coniferous can. (%)

61.3 ± 5.6

50.0 ± 8.1

76.7 ± 3.6

Understory can. (%)

32.7 ± 5.8

42.7 ± 8.7

19.8 ± 6.2

Ground cover (%)

39.0 ± 6.2

53.4 ± 8.0

22.9 ± 7.8

Shrub density

61.6 ± 9.3

73.4 ± 17.5

50.3 ± 9.2

Shrub index

90.3 ± 13.7

79.0 ± 18.4

98.1 ± 21.9

Tree density (trees/ha)

1071 ± 95'=

914 ± 96

1037 ± 13U

Under, dens, (trees/ha)

334 ± 71”

375 ± 120”

231 ± 57

Basal area (m’/ha)

28.9 ± 3.0

22.7 ± 2.7

37.5 ± 3.9^'

Mean DBH (cm)

17.6 ± 0.9^'

16.9 ± 1.2

19.1 ± 1.3'=

Dist. to water (m)

260 ± 71

88 ± 28

440 ± 120

Dist. to open, (m)

58.9 ± 16.2

72.7 ± 26.2

50.8 ± 21.8

* P s 0.()0I.

** P s 0.00()5.

“ Combined data for A. striaius includes one hardwood nest site in addition to mixed and plantation nest sites.
Missing data at one nest site (n = 23 combined, n = 10 mixed).

Missing data at one nest site (n = 23 combined, ;i = 1 1 plantation).

specific analyses may lie in the variables that
did not differ significantly across habitats,
such as nest tree height, nest height, canopy
height, and mean tree DBH — each of which
differed between species (see Discussion).

For combined habitats. Cooper’s Hawks
nested in a wider array of tree species than
Sharp-shinned Hawks (Table 3). This varia-
tion however, occurred mostly within hard-
wood sites; within mixed stands and conifer
plantations Sharp-shinned Hawks used a
greater variety of tree species. Of the conif-
erous nest trees used by Cooper’s Hawks {n
= 29), only Pinus was represented in this
sample, while Sharp-shinnned Hawks {n =
23) used five genera. With only one exception
[a Cooper’s Hawk nest in a white ash (Frax-
inus americana)], both species consistently
used conifers for nesting in mixed sites where
both hardwoods and conifers were present in
the canopy. For both species, nest trees in co-
nifer plantations were all conifers, despite the
presence of canopy-level hardwoods in 60%
of Cooper’s Hawk and 42% of Sharp-shinned
Hawk plantation sites.

DISCUSSION

Our comparative analyses of nest site hab-
itat at 52 Cooper’s Hawk and 24 Sharp-
shinned Hawk nests in Wisconsin did not pro-
vide data on nest site use relative to avail-
ability, and we cannot contend that numbers
of nests in our three habitat categories are nec-
essarily proportional to use of these habitat
types. Nevertheless, our sample involves in-
dependent nests discovered by unbiased
means on a statewide scale in compositionally
diverse woodland habitats: upland and low-
land sites; coniferous, hardwood, and mixed
forests; urban and rural woodlands of varied
sizes; and both managed and unmanaged for-
ests including conifer plantations. Thus we
suggest that our data set provides a reasonably
thorough and representative sample of the
range of nest site habitats used by these hawks
in Wisconsin.

If interspecific differences in nest site char-
acteristics of these congeners occur on a finer
within-habitat scale, as some prior work has
indicated (Siders and Kennedy 1996), then

I'rexel el al. • ACCIPITER NEST SITE HABITATS

TABLE 2. Extended.

A. cooperii (mean ± SE)

Interspecific differences

Inira.specific

differences

Combined

Hardwood

Mi.xed

Plantation

Combined Mixed Plantation

A. siriatus A. cooperii

(/7 = 22)

(n = 10)

in = 20)

19.1 ± 0.6

20.5 ± 1.0

17.5 ± 0.8

18.1 ± 0.8

32.6 ± 1.2

36.2 ± 1.6

29.7 ± 3.1

30.0 ± 1.6

** *

13.1 ± 0.4

13.3 ± 0.7

13.2 ± 0.7

12.9 ± 0.5

** *

69.8 ± 1.4

66.1 ± 2.9

75.0 ± 1.9

72.3 ± 2.1

19.5 ± 0.5

20.9 ± 0.9

18.6 ± 0.6

18.3 ± 0.7

** *

84.9 ± 1.3

86.3 ± 2.1

79.3 ± 3.4

86.3 ± 1.8

54.9 ± 4.8

86.1 ± 2.1

46.3 ± 7.3

24.8 ± 5.9

30.0 ± 4.5

0.1 ±0.1

33.0 ± 6.2

61.5 ± 5.5

37.8 ± 3.6

48.8 ± 5.1

30.3 ± 6.0

29.5 ± 6.1

47.8 ± 3.0

53.6 ± 4.2

54.5 ± 5.7

38.1 ± 5.2

30.0 ± 4.4

26.2 ± 5.4

38.6 ± 15.8

29.9 ± 6.4

71.5 ± 8.4

65.5 ± 10.7

60.0 ± 15.2

84.0 ± 16.8

623 ± 48

438 ± 38

623 ± 103

826 ± 87

307 ± 28

340 ± 40

383 ± 68

233 ± 44

31.6 ± 2.8

27.4 ± 3.4

24.0 ± 2.0

39.9 ± 5.8

25.6 ± 0.9

27.4 ± 1.5

22.4 ± 1.7

25.2 ± 1.4

320 ± 56

412 ± 87

277 ± 111

468 ± 115

56.7 ± 8.6

86.8 ±17.4

33.9 ± 9.6

39.1 ± 7.3

differences might also exist on a coarser scale
among more broadly defined and heteroge-
neous habitat types. Such differences might
furthermore emerge on a landscape scale
among woodland habitats in general.

The variables we measured are not inde-
pendent indicators of interspecific differences
in nest site habitat; many of them seem to be
related to stand age or successional stage. Tree
age was not measured in this study, but it ap-
pears that Cooper’s Hawks tended to use older
stands with a lower density of taller and larger
trees. Sharp-shinned Hawks, on the other
hand, tended to use younger stands with a
higher density of smaller, shorter trees. Reyn-
olds and coworkers (1982) and Moore and
Henny (1983) also have suggested that differ-
ences in accipiter nest site habitat are corre-
lated with stand age or successional stage,
with Cooper’s Hawks using older stands than
Sharp-shinned Hawks.

Interspecific differences in combined habi-
tats seldom seem the result of contrasting pro-
portions of habitats used on the intraspecific
level. The lower percent coniferous canopy in
combined Cooper’s Hawk habitats versus
Sharp-shinned Hawk habitats (30% vs 61%, F
< 0.0005; Table 2) appears to be the result of
the disproportionate number of Cooper’s

Hawk sites in hardwoods (42% of 52 nests)
compared to the one Sharp-shinned Hawk at
a hardwood site (4% of 24 nests). The differ-
ence in deciduous canopy cover between Coop-
er’s Hawks versus Sharp-shinned Hawks in
combined habitats (55% vs 15%, P < 0.0005;
Table 2) also seems to be a result of contrast-
ing proportions of habitat use.

In addition to having proportionally more
nests in hardwood stands. Cooper’s Hawks
nested in conifer plantations that had substan-
tially greater deciduous canopies than those
used by Sharp-shinned Hawks (25% vs 4%;
Table 2). In mixed conifer-hardwood habitats
the deciduous canopy cover percentage again
was greater for Cooper’s Hawks than for
Sharp-shinned Hawks (46% vs 24%; Table 2).
Although neither of these within-habitat dif-
ferences was statistically significant, they are
clearly consistent with a significant difference
in deciduous canopy in combined habitats. On
a landscape-scale continuum from Wiscon-
sin’s northern coniferous forests (plus conifer
plantations) to mixed and southern deciduous
woodlands (see Curtis 1959), nest habitat thus
appears to be comprised more of deciduous
sites or elements for the Cooper’s Hawk and
coniferous elements for the Sharp-shinned
Hawk, albeit with considerable overlap in

TABLE 3. Nest tree species used by Accipiter striatus and A. cooperii in Wisconsin.

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

^ o

(N'^'^CN(N(N(NCN'^CO’^(N

lO — <N CNl --

fNj —

SO fO

IT)

(N (N

:§

-2

-Q

llg

Ol ^

■Si

■a

a 2

a s

ii K

S 8

b ^

W) V5

3 ca

Ci.

■S

<U dJ

^ p

cd 5

Cu <

00— '-H-rt— 100— -00 00

(N 'Tf ro

■b" (N ro (N (N

■b" —

(N —

m (N (N

-S5

.5^

^ «3

m ^ ^

a
• S oo
Q.

2 ^
o o
o 5
on Z

Q
<0

-Cl
' '>)

Q -ii
^ -Cl

8 ^

2 •-

au.

4^ cd
O (A

-2 -3

m oQ

'S'

■S

■2

c/5

1(4%) 11(46%) 12(50%) 24 22 (42%) 10 (19%) 20 (38%)

Trexel et al. • ACCIPITER NEST SITE HABITATS

mixed forests. This divergence seems unap-
parent in western montane environments
(Reynolds 1983, Fischer 1986).

Siders and Kennedy (1996) also found that
Cooper’s Hawks used significantly taller nest
trees with greater diameters and nest sites with
lower tree densities than did Sharp-shinned
Hawks. However, they reported that Sharp-
shinned Hawk nest sites had significantly
higher basal areas and canopy closures than
did those of Cooper’s Hawks. Reynolds and
coworkers (1982) found, as we did that Coop-
er’s Hawks had greater nest heights in eastern
Oregon, and used habitats at lower tree den-
sities in northwestern Oregon than did Sharp-
shinned Hawks. As did Siders and Kennedy
(1996), they found that Sharp-shinned Hawk
nest sites had greater canopy closure than
those of Cooper’s Hawks. Moore and Henny
(1983) noted that Cooper’s Hawk nests were
significantly higher than those of Sharp-
shinned Hawks, but again in contrast to our
results they found that Sharp-shinned Hawk
nest sites had significantly higher canopy clo-
sure than sites used by Cooper’s Hawks.

It seems that significantly higher tree den-
sities at Sharp-shinned Hawk nest sites would
usually lead to significantly greater canopy
closure, as reported for mostly coniferous hab-
itats in New Mexico and Oregon, but this was
not the case for combined habitats in Wiscon-
sin. Our results show lesser canopy closure in
association with greater tree densities at nest
sites of Sharp-shinned versus Cooper’s hawks
(Table 2). This seeming contradiction is prob-
ably the result of disproportional use of broad-
leaved hardwood forests (vs needle-leaved co-
nifer forests) by Cooper’s Hawks and conse-
quent effects of leaf surface on measures of
canopy closure. Overstory canopy measures
might also be influenced by lower foliage den-
sities (e.g., tamarack) or strongly conical
growth forms (e.g., black spruce) in some nest
tree species used by Sharp-shinned Hawks
(Table 3).

In Missouri, Wiggers and Kritz (1991) used
the most similar set of habitat measures and
techniques for analyzing those measures, yet
they reported no significant differences in nest
site characteristics for these two accipiters.
However, they divided their nest sites into
habitat types differently than we did and were
able to make interspecific comparisons only

for pine dominated habitat (“> 50% of over-
story trees were pines”). Still, with small data
sets (Table 2) and the same alpha level (0.001)
we detected significant differences in nest
height and average canopy height for nest
sites in mixed conifer-hardwoods, and in nest
tree DBH in conifer plantations. Wiggers and
Kritz (1991) reported significant intraspecific
differences between pine habitat and hard-
wood habitat for Cooper’s Hawks; had they
combined these habitats they might have
found overall interspecific differences as we
did. Their ability to detect significant differ-
ences may also have been hampered by the
fact that 87% of their nests were located by
searching habitat (especially coniferous habi-
tat) that was assumed a priori to be suitable
for one or both species (Siders and Kennedy
1996). Consequently, 92% of nests in conifers
(n = 50) and 77% of all nests (n = 60) were
situated in pine plantations of similar age and
vegetational structure.

In Wisconsin there appear to be numerous
interspecific differences in nest site habitats of
Cooper’s and Sharp-shinned hawks. Such in-
terspecific differences, within and across di-
vergent habitat types, may provide guidance
in identifying and managing the respective
nesting habitats of these birds, one or both of
which have been listed as species of conser-
vation concern in several midwestern states
(Rosenfield et al. 1991, Rosenfield and Bie-
lefeld! 1993). Many of the nest tree and nest
site variables differing significantly between
species (tree heights, densities, diameters, and
coniferous components) are routinely and eas-
ily estimated measures of woodland habitats
among resource managers.

Intraspecific analyses of nest site variables
across habitat types may also be useful to
management and conservation. Significant in-
traspecific differences among habitats in the
Cooper’s Hawk would seem to portray the
breadth of acceptable nesting habitat(s). Var-
iables that do not differ intraspecifically
across habitat types (e.g., nest tree height or
mean DBH of nest site trees) may serve as
focal points for managers in identifying po-
tentially usable nesting habitats, whether or
not these features actually provide proximate
cues to nest site use for the birds themselves.

We examined habitat characteristics only at
the nest tree level and in a small area (0.04

THE WILSON BULLETIN

Vol. Ill, No. 1, March 1999

ha) immediately surrounding the nest. We did
not deal with other habitats used by these ac-
cipiters such as hunting areas or non-breeding
habitats. Recent studies of nest site habitat in
the Cooper’s Hawk in North Dakota (M. Nen-
neman, pers. comm.) suggest that existing an-
alyses of breeding habitats from disparate ar-
eas and woodland types may not be general-
izable to other regions. Management impli-
cations drawn from our Wisconsin data should
therefore be cautious.

ACKNOWLEDGMENTS

We gratefully acknowledge the field assistance of T.
Doolittle, M. Fuller, J. Partelow, and M. Thwaits, as
well as the computer and plant ecology knowledge of
D. Hillier. Advice on statistical analyses and interpre-
tation was provided by M. Bozek, W. Gould, and B.
Rogers. K. Beal, P. Kennedy and especially K. Kritz
provided helpful reviews of this paper. Primary fund-
ing was provided by the Wisconsin Department of Nat-
ural Resources, the University of Wisconsin-Stevens
Point, and the Society for Tympanuchus Cupido Pin-
natus. Ltd. The Personnel Development Committee at
the University of Wisconsin-Stevens Point provided
support for publication and sabbatical leave to RNR.

LITERATURE CITED

Anonymous. 1997. Birds in forested landscapes. Wild-
lifer 281:21-44.

Curtis, J. T. 1959. The vegetation of Wisconsin: an
ordination of plant communities. Univ. of Wis-
consin Press, Madison.

Fischer, D. L. 1986. Daily activity patterns and habitat
use of coexisting Accipiter hawks in Utah. Ph.D.
diss., Brigham Young Univ., Provo, Utah.

James, F. C. and H. H. Shugart, Jr. 1970. A quanti-
tative method of habitat description. Audubon
Field Notes 24:727-736.

Moore, K. R. and C. J. Henny. 1983. Nest site char-

acteristics of three coexisting accipiter hawks in
northeastern Oregon. Raptor Res. 17:65-76.

Reynolds, R. T. 1983. Management of western conif-
erous forest habitat for nesting accipiter hawks.
USDA For. Serv. Gen. Tech. Rep. RM-102:1— 7.

Reynolds, R. T, E. C. Meslow, and H. M. Wight.
1982. Nesting habitat of coexisting Accipiter in
Oregon. J. Wildl. Manage. 46:124—138.

Rosenfield, R. N. and J. Bielefeldt. 1992. Natal dis-
persal and inbreeding in the Cooper’s Hawk. Wil-
son Bull. 104:182-184.

Rosenfield, R. N. and J. Bielefeldt. 1993. Cooper’s
Hawk {Accipiter cooperii). In The birds of North
America, no. 75 (A. Poole and E Gill, Eds.). The
Academy of Natural Sciences, Philadelphia, Penn-
sylvania; The American Ornithologists’ Union,
Washington, D.C.

Rosenfield, R. N. and J. Bielefeldt. 1996. Lifetime
nesting area fidelity in male Cooper’s Hawks in
Wisconsin. Condor 98:165-167.

Rosenfield, R. N., J. Bielefeldt, R. K. Anderson,
AND J. M. Papp. 1991. Status reports: accipiters.
Natl. Wildl. Fed. Sci. Tech. Serv. 15:42—49.

Rosenfield, R. N., J. Bielefeldt, J. L. Affeldt, and
D. J. Beckmann. 1996. Urban nesting biology of
Cooper’s Hawks in Wisconsin. Pp. 41-44 in Rap-
tors in human landscapes (D. M. Bird, D. E. Var-
land, and J. J. Negro, Eds.). Academic Press, Lon-
don, U.K.

Siders, M. S. and P. L. Kennedy. 1996. Forest struc-
tural characteristics of accipiter nesting habitat: is
there an allometric relationship? Condor 98:123-
132.

SoKAL, R. R. AND F. J. Rohlf. 1981. Biometry, second
ed. W. H. Freeman and Co., San Francisco, Cali-
fornia.

Titus, K. and J. A. Mosher. 1981. Nest-site habitat
selected by woodland hawks in the central Ap-
palachians. Auk 98:270-281.

Wiggers, E. P. and K. j. Kritz. 1991. Comparison of
nesting habitat of coexisting Sharp-shinned and
Cooper’s hawks in Missouri. Wilson Bull. 103:
568-577.

Wilkinson, L. 1992. SYSTAT: the system for statis-
tics. SYSTAT Inc., Evanston, Illinois.

Wilson Bull., 111(1), 1999, pp. 15-21

MADAGASCAR FISH-EAGLE PREY PREFERENCE AND

FORAGING SUCCESS

JAMES BERKELMAN,"^ JAMES D. FRASER,' AND RICHARD T. WATSON^

ABSTRACT. — We investigated Madagascar Fish-Eagle (Haliaeenis vociferoides) foraging ecology to deter-
mine prey preference and the effect of fish abundance on fish-eagle foraging rates and foraging success. We
observed fish-eagle foraging behavior at nine lakes in western Madagascar from May to August 1996. We
sampled the fish population at each lake using gill nets and recorded fish weights and species. Introduced tilapia,
Oreochromis spp. and Tilapia spp., made up the majority of both the gill net (66.3%) and fish-eagle catch
(64.7%) in simiku" proportion, suggesting that the fish-eagle is an opportunistic predator. Consequently, replace-
ment of native fish species by exotics probably has not been detrimental to the island’s fish-eagle population.
Male fish-eagle foraging success was positively correlated (P < 0.001) with number of fish species, suggesting
that fish species diversity may affect fish-eagle foraging effectiveness. Received 24 July 1997, accepted 2 Nov.
1998.

Prey availability influences breeding area
selection (Swenson et al. 1986), breeding den-
sity (Dzus and Gerrard 1989), reproductive
success (Grubb 1995), and date of breeding
(Hansen 1987) in Bald Eagles {Haliaeetus
leucocephalus) and productivity in White-
tailed Eagles {Haliaeetus albicilla\ Helander
1985). It also affects distribution and density
of Bald Eagles at wintering sites (Griffin and
Baskett 1985, Sabine and Klimstra 1985,
Keister et al. 1987, Hunt et al. 1992b) and
migratory stopovers (Fraser et al. 1985, Ben-
netts and McClelland 1991).

Although prey availability is clearly impor-
tant to Haliaeetus eagles, there has been little
research aimed at quantitatively determining
prey abundance and its effects on prey selec-
tion, foraging rates, and foraging success in
the genus. Steenhof (1976), Mersmann
(1989), and Hunt and coworkers (1992a) used
gill nets to inventory relative fish abundance
and determined that the most frequently netted
fish species made up the greatest proportion
of the Bald Eagle’s diet. Wintering Bald Ea-
gles in New Mexico fed most frequently on
big game carrion when it was the most abun-

‘ Dept, of Fisheries and Wildlife Sciences, Virginia
Polytechnic Inst, and State Univ., Blacksburg, VA
24061-0321.

^ The Peregrine Fund, 566 West Flying Hawk Lane,
Boise, ID 83709.

^ Present address: Dept, of Wildlife Ecology, Univ.
of Wisconsin, 226 Russell Labs, 1630 Linden Dr.,
Madison, WI 53706-1598.

■* Corresponding author; Email:
jberkelman@facstaff.wisc.edu

dant prey source (Grubb 1984). There is a
positive relationship between prey abundance
and foraging success of wintering Bald Eagles
both between locations (Stalmaster and Plett-
ner 1992) and between years (Brown 1993).
Knight and Knight (1983) found a negative
correlation between search time and relative
prey abundance of Bald Eagles wintering in
Washington, but Mersmann (1989) did not
find a correlation between Bald Eagle forag-
ing rates and gill net catch rates on the north-
ern Chesapeake Bay.

Langrand and Meyburg (1989) and Raza-
findramanana (1995) have documented fish
species eaten by Madagascar Fish-Eagles
{Haliaeetus vociferoides), but there has been
no previous attempt to quantitatively assess
the eagle’s diet. The objectives of this study
were (1) to describe the diet and foraging be-
havior of the Madagascar Fish-Eagle at lakes
in western Madagascar, (2) to determine fish-
eagle prey preference, and (3) to determine
whether fish-eagle foraging rates and foraging
success are dependent on prey abundance.

STUDY AREA AND METHODS

We observed Madagascar Fish-Eagle foraging ecol-
ogy from 22 May to 4 August 1996 at nine lakes in
the Tsiribihina, Manambolo, and Beboka river drain-
ages between the Bongolava escarpment and the Mo-
zambique Channel in western Madagascar (Table 1 ).
We selected lakes that we felt would offer the best
conditions for viewing eagles throughout the day from
among 32 lakes with resident Madagascar Fish-Eagle
pairs that we studied in 1995 (Berkelman 1997).

We observed fish-eagle foraging behavior through-
out daylight hours from 06:00 to 18:00 (GMT + 3 h)
for six or seven days at each lake. We recorded both

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

TABLE 1. Lakes included in Madagascar fish-eagle foraging ecology study in western Madagascar, May-
August, 1996.

Lake

Coordinate.s

Dates observed

Ambereny

18°

55' S,

44°

23' E

22-28 May

Bejijo

19°

13' S,

44°

32' E

30 May-5 June

Ankazomena

19°

42' S,

45°

23' E

8-15 June

Asonjo

19°

50' S,

45°

26' E

16—23 June

Ampamandrika

19°

46' S,

44°

34' E

27 June— 3 July

Befotaka

19°

1' S, 44° 24' E

7-12, 20 July

Masama

18°

50' S,

44°

27' E

13-19 July

Bevoay

19°

9' S, 44° 24' E

22-28 July

Tsiandrora

18°

58' S,

44°

38' E

30 July— 4 August

observer-time, the number of hours spent observing
eagles, and eagle-time, the product of observer-time
and the number of eagles observed during each hour,
tor each lake. At the three lakes that had more than
one resident fish-eagle pair, we randomly selected one
of the pairs for observation during the period. We
watched from an inflatable kayak or from the shore
using 10 X 50 binoculars and a 15-45 X spotting
.scope on a rifle mount. We concentrated on the resi-
dent eagle pair while also noting the behavior of any
other eagles that we could see, including immatures
and other adults that were associated with the resident
pair. We distinguished adult fish-eagle sexes by the
smaller size and higher pitched vocalizations of the
males. We distinguished adults from immatures by
their vocalizations and by the completely white plum-
age on the tails and the faces of the adults (Langrand
and Meyburg 1989).

We recorded prey searches when eagles flew low
over the water looking down (Stalmaster and Plettner
1992) and kills when they picked up a fish from the
water. We al.so noted instances of fish-eagles scaveng-
ing dead fish from the shoreline or pirating fish from
Black Kites (Milvu.s migrans). We identified fish to
species whenever possible. If we could not identify the
fish while the eagle was in flight, we looked for prey
remains on the ground beneath the eagle’s feeding
perch.

We .set two monofilament gill nets for 3 h at each
lake starting at 06:00-06:15. The gill nets had a foam
core float rope and a lead core bottom rope, were 0.91
m deep by 45.7 m long, and were divided into three
15.2 m panels of 2.5, 3.8, and 5.1 cm mesh size. We
attached floats to the first net and .set it parallel to the
shore in water about 0.9 m deep. We set the second
net on the bottom, parallel to the shore, in water about
1 .8 m deep. Thus, we sampled fish from among the
first and second 0.9 m of the water column. If the lake
was less than 1.8 m deep, we .set the .second net in the
deepest water within 200 m of where we had set the
first net. We placed nets adjacent to the nest or, where
we did not find a nest (/; = I ), adjacent to a frequently
used perch. We believed that the.se sites were repre-
sentative of fish-eagle foraging areas becau.se we ob-

served the eagles forage most frequently within 300 m
of the nest at eight of the nine lakes.

We identified each fish caught in the gill nets using
keys (Arnoult 1959, Kiener 1963, Glaw and Vences
1994) and weighed it to the nearest g and measured
total fish length to the nearest cm. We combined the
data for the three days that we sampled each lake (nine
hours total) and calculated total number of fish caught,
total weight (kg) of fish catch, average fish weight (g),
and number of species. We did not include fish that
weighed over 1.5 kg in these calculations because we
did not see fish-eagles capture larger fish.

We conducted the test of equal proportions to
determine it fish-eagle use of fish species was different
from expected use based on gill net samples using SAS
on an IBM compatible computer (PROC EREQ, SAS
Institute Inc. 1990). We excluded unidentified prey
from this analysis. After finding a significant {P <
0.05) overall difference, we tested the hypothesis of no
difference between use and availability of each fish
species, following Marcum’s and Loftsgaarden’s
(1980) technique. We calculated Spearman correlation
coefficients between fish-eagle foraging variables and
fish variables (PROC CORR, SAS Institute Inc. 1990).
For all analyses, we used an overall confidence level
of a — 0.05 and a confidence level of a.lk, where k
was the number of significance values calculated, fol-
lowing the Bonferroni approach (Miller 1966).

RESULTS

Foraging behavior. — There were extra
adult Madagascar Fish-Eagles associated with
three pairs and immatures with another three
of the nine resident pairs that we studied. Al-
together, we observed 1 1 adult males, 10 adult
females, and 3 immatures.

Hunting methods were similar to those used
by other .sea eagles (Brown 1980, Love 1983,
Stalmaster 1987). The fish-eagles we observed
hunted from perches and either stooped di-
rectly from a perch or searched low over the
water, generally returning to perch within 5

Herkelimin et al. • MADAGASCAR FISH-EAGLE FORAGING

TABLE 2. Male Madagascar Fish-Eagle foraging
at nine lakes in western Madagascar, May-Aiigust,
1996.

Variable

.X (n = 11)

Searches

36.1

8.0

Kills

5.9

1.7

Searches/hour/eagle

0.68

0.15

Kills/hour/eagle

0.10

0.03

Kills/search

0.16

0.04

TABLE 3. Number of Hsh, total fish weight, av-
erage fish weight, and number of hsh species caught
in gill nets at nine lakes occupied by Madagascar Fish-
Eagles in western Madagascar, May-August, 1996.

Variable

Range

Number of fish

30.1

7.3

4-66

Total weight, kg

4.6

1.6

0.2-15.9

Average weight, g

139.0

23.7

55.3-269.3

Number of species

3.9

0.4

2-6

min of leaving. When striking, the eagles en-
tered the water feet first at a low angle and
only took fish that were at or just below the
surface.

We watched eagles for 669.5 h observer-
time and 1030.98 h eagle-time, including
490.25 h (47.6%) male eagle-time, 526.0 h
(51.0%) female eagle-time, and 14.73 h
(1.4%) immature eagle-time. We recorded 67
occurrences of eagles obtaining fish, including
60 (89.6%) occasions when they captured fish
in open water, 3 (4.4%) when they scavenged
dead fish from the shoreline, and 4 (6.0%)
when they stole fish from Black Kites. We
also recorded 32 occurrences of eagles eating
fish or delivering fish to their mates when we
did not see an eagle obtain the fish. On one
occasion we observed an eagle eating a do-
mestic duckling (Anas sp.). We did not see the
eagle capture the duckling, but the local peo-
ple claimed that the same eagle pair had killed
domestic ducklings and turkey (Meleagris sp.)
poults at the same lake on several occasions
in 1996.

Of the 67 occasions when we saw eagles
obtain fish, the eagles were adult males on 53
(79.1%) occasions, adult females on 13
(19.4%), and an immature on 1 (1.5%) occa-
sion. Nine (69.2%) of 13 adult females that
we saw capture fish were not nesting at the
time. The other four (31.8%) were incubating
eggs. All 32 occasions on which we saw ea-
gles eating or delivering fish but did not see
them catch the fish involved adult male ea-
gles. All four instances of piracy from kites
occurred at the same lake and involved two
cooperating adult male eagles associated with
the same territory. In each case, the eagles ha-
rassed a kite until it dropped its fish, which
one of the eagles then retrieved.

Foraging rates and fish abundance. — Male

fish-eagle kills/search was positively correlat-
ed with number of fish species caught in gill
nets (p = 0.909, P < 0.001). There were no
other significant correlations between fish-ea-
gle foraging rates (Table 2) and fish variables
(Table 3). We only analyzed male foraging be-
cause we rarely saw females forage.

The Madagascar Fish-Eagle search rate
peaked in the early morning and again, at a
higher level, in the early afternoon (Fig. lA).
Foraging success, expressed as the proportion
of prey searches that resulted in kills, was
highest before 10:00 and after 16:00 but lower
between these times (Fig. IB).

Dietary preference. — Of the 99 observed
occurrences of fish-eagles capturing, carrying,
or eating fish, we were able to identify 68
(68.7%) either to species or to a closely re-
lated group of species (Table 4). We were un-
able to identify eagle-caught tilapia to species
or to distinguish between the closely related
Oreochromis and Tilapia genera. In our gill
net samples, we caught 271 fish of 12 species,
including four species of tilapia and eight oth-
er species. The total weight of the catch at all
nine lakes was 41.1 kg.

We combined all native fish species into a
single group because our catches of each spe-
cies were too small to analyze separately (Ta-
ble 4). The proportions of fish species differed
significantly between the fish-eagle catch and
the gill net catch (x^ = 41.97, df = 4, P =
0.001). The 95% confidence limits for the dif-
ference between the proportion used and the
proportion available suggested that fish-eagles
catch Ophicephalus striatus in greater propor-
tion, Cyprinus carpio in lesser proportion, and
tilapia, Heterotis niloticus, and native species
in equal proportion to their relative abun-
dance.

THE WILSON BULLETIN • Vol. Ill, No. 1. March 1999

(1)

jfl

0.3 .

Hour of day

FIG. 1. Madagascar Fish-Eagle prey
by time of day at nine lakes in western

searches per hour per eagle (A) and prey searches
Madagascar, May-August, 1996.

resulting in kills (B)

DISCUSSION

Foraging behavior. — Our observation that
4.4% of fish taken by Madagascar Fish-Eagles
were scavenged was lower than scavenging
rates that Mersmann (1989) and Brown (1993)
reported for Bald Eagles (25% and 7.7%, re-
spectively) but comparable to the 4% reported
by Stalmaster and Plettner (1992). We did not
.see fish-eagles take floating dead fish from the
surface in open water, but it is possible that
some of the fish that we observed eagles catch
from a distance were dead fish floating below
the water surface.

We are unaware of previous reports of pi-

racy or capture of avian prey by the Mada-
gascar Fish-Eagle. It is unclear why we ob-
served four instances of piracy at one of the
lakes and none at the other eight lakes. We
saw numerous other fish-eating birds at all the
lakes, including Black Kites, herons, storks,
anhingas, and cormorants.

Although we observed adult male eagles in-
cubating eggs and tending nestlings, it appears
that the male does most of the foraging for the
pair, at least during early nesting. The four
instances in which we observed incubating fe-
males catch fish occurred near the nest when
males were not present. The most advanced

Berkelmcm et al. • MADAGASCAR FISH-EAGLE FORAGING

TABLE 4. Fish caught (number and % of total) and number of lakes where fish were caught (out of nine)
by Madagascar Fish-Eagles and in gill nets in western Madagascar, May-August, 1996.

Fish species“, family

Fish-eagle catch

Gill nei catch

No. of
Hsh

% of
fish

No. of
lakes

No, of

Hsh

% of
fish

No. of
lakes*’

Exotic

Tilapia'-', Cichlidae

64.7

183

67.5

Heteroti.'! niloticu.s, Osteoglossidae

10.3

6.3

Cyprinus carpio, Cyprinidae

2.9

12.2

1 (2)

Ophicephalus striatus, Channidae

14.7

0.4

1 (6)

Native

Megalops cyprinoicles, Megalopidae

2.9

6.6

6 (1)

Arius madagascariensis, Ariidae

4.4

5.2

Glossogobius giuris, Gobiidae

0.7

Ambassis gymnocephalus, Ambassidae

0.7

Scatophagus tetracanthus, Scatophagidae

0.4

TOTAL

100

271

100

“ Unidentified fish that fish-eagles caught (n = 31) were excluded.

^Number in parentheses represents additional lakes where each fish species was known to be present either from 1995 gill net sampling or from fish
catches of local fishermen.

Tilapia species included in order of decreasing gill net catch: Oreocliromis macrochir, Tilapia zilii. O. mossamhicus, and O. niloticus.

nesting attempt we observed had a 2-3 week
old downy chick, so we were unable to doc-
ument whether female fish-eagle foraging
rates change as nesting progresses.

Foraging rates and fish abundance. — The
strong positive correlation between fish-eagle
kills/search and number of fish species may
indicate that the eagles forage most effectively
at lakes that have the highest fish species di-
versity. In a previous study, Berkelman (1997)
found that fish species diversity, along with
shoreline perch density, was one of the best
predictors of fish-eagle lake use, lending fur-
ther support to the importance of fish species
diversity. However, the strength of the corre-
lation between foraging success and fish spe-
cies diversity may be related to the low range
of variability in number of fish species caught
(2 to 6) at lakes in this study.

The early morning peak in search rate that
we observed also was reported for Madagas-
car Fish-Eagles by Razafindramanana (1995)
and for Bald Eagles (Steenhof et al. 1980,
Mersmann 1989) and Ospreys (Pandion hali-
aetus; Flemming and Smith 1990). This peak
may result from hunger after fasting overnight
or from eagles taking advantage of greater fish
availability and calmer weather during the
early morning hours. The early afternoon
search rate peak may reflect eagles that have
digested the morning food and are hungry
again. Whitfield and Blaber (1978) observed

a midday foraging peak in African Fish-Ea-
gles (Haliaeetus vocifer) and suggested that
the eagles were taking advantage of thermals
at this time, but the Madagascar Fish-Eagles
that we observed foraged mostly low over the
water from a perch. Daily weather patterns
varied little during the study, so differences
among eagle pairs in foraging rates and suc-
cess probably were not related to weather.

Dietary preference. — The results suggest
that Madagascar Fish-Eagles prefer Ophice-
phalus striatus to other fish and avoid Cypri-
nus carpio. Ophicephalus striatus was the
largest fish species that we saw fish-eagles
capture. We estimated the largest ones caught
by fish-eagles to be between 1 and 1.5 kg.
This species is a predatory fish that was intro-
duced to Madagascar in 1978 (Reinthal and
Stiassny 1991). In field experiments. Bald Ea-
gles selected large fish more often than small-
er fish during the breeding season, but not dur-
ing the non-breeding season (Jenkins and
Jackman 1995); Madagascar Fish-Eagle pref-
erence for O. striatus may reflect the eagle’s
greater energy requirements during the breed-
ing season. Fish-eagles showed no preference
for Heterotis niloticus, another large intro-
duced fish species, but it was present in only
three (33.3%) of the nine lakes. Cyprinus car-
pio may not be used because this species feeds
on the bottom of lakes (Scott and Crossman
1973) where it is difficult for eagles to catch.

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

The results also may indicate biases in the
fish abundance data. Ophicephalus striatus, a
visually orienting predator, may be better at
avoiding entanglement in gill nets than the
other fish species. All of the C. carpio abun-
dance data were from a single lake where we
caught 33 individuals. This lake was so shal-
low (0.9 m) that both gill nets extended to the
bottom and consequently were more likely to
catch bottom-dwelling fish such as C. carpio.

The use and relative abundance data for ti-
lapia, the most abundant fish in all of the
lakes, were similar (64.7% of identifiable fish-
eagle catch and 67.5% of gill net catch). Ti-
lapia were introduced to Madagascar for aqua-
culture in the 1950s (Kiener 1963) and have
since spread to most bodies of freshwater
throughout the island. The predominance of
tilapia in the fish-eagles’ diet in this study
suggests that the Madagascar Fish-Eagle is an
opportunistic predator that catches whatever
prey species are most abundant. Thus, the
marked change in species composition of
Madagascar’s freshwater fish fauna resulting
from exotic species introductions (Loiselle
1993, Reinthal et al. 1995) probably has not
been detrimental to the island’s fish-eagle pop-
ulation. The positive relationship between
fish-eagle foraging success and number of fish
species suggests that the fish-eagle population
may be sensitive to declines in fish species
diversity.

ACKNOWLEDGMENTS

We thank C. Razafimahatratra, G. Tohaky, E Paul,
and M. Philbert for technical a.ssistance. Thanks to Y.
Rakotonirina for his driving and mechanical skills.
Thanks to J. Rajesy, R. Rabarisoa, and M. Razafin-
drakoto for administrative and logistical support. We
thank C. A. Haas, J. J. Ney, R. G. Oderwald, D. E
Stauffer, and R. Thorstrom for comments on the man-
uscript. We received funding from the National Geo-
graphic Society, the Raptor Research Foundation, The
International Osprey Foundation, the World Nature As-
•sociation, the American Museum of Natural History,
and the Cooper Ornithological Society. We conducted
this study under the auspices of The Peregrine Fund’s
Madaga.scar Fish-Eagle and Wetland Conservation
Project in Madaga.scar.

LITERATURE CITED

Arnoult, j. 1959. Pois.son des eaux donees. Faune de
Madaga.scar 10. ORSTOM and CNRS, Paris,
France.

Bknnktt.s, R. E. and B. R. McCi.klland. 1991. Dif-

ferences in the distribution of adult and immature
Bald Eagles at an autumn concentration in Mon-
tana. Northwest Sci. 65:223-230.

Berkelman, j. 1997. Habitat requirements and forag-
ing ecology of the Madagascar Fish-Eagle. Ph.D.
diss., Virginia Polytechnic Inst, and State Univ.,
Blacksburg.

Brown, B. T. 1993. Winter foraging ecology of Bald
Eagles in Arizona. Condor 95:132—138.

Brown, L. 1980. The African Fish Eagle. Bailey Bros,
and Swinfen Ltd, Folkstone, U.K.

Dzus, E. H. AND J. M. Gerrard. 1989. Interlake var-
iations of Bald Eagle, Haliaeetiis leucocephalus,
populations in north-central Saskatchewan. Can.
Field-Nat. 103:29-33.

Flemming, S. P. and P. C. Smith. 1990. Environmental
influences on Osprey foraging in northeastern
Nova Scotia. J. Raptor Res. 24:64-67.

Fraser, J. D., L. D. Frenzel, J. E. Mathisen, and M.
E. Shough. 1985. Seasonal distribution of sub-
adult Bald Eagles in three Minnesota habitats.
Wilson Bull. 97:365-366.

Glaw, E and M. Vences. 1994. A fieldguide to the
amphibians and reptiles of Madagascar including
mammals and freshwater fish, second ed. Moos
Dmck, Leverkusen and Farbo, Cologne, Germany.

Griffin, C. R. and T. S. Baskett. 1985. Food avail-
ability and winter range sizes of immature and
adult Bald Eagles. J. Wildl. Manage. 49:592-594.

Grubb, T. G. 1984. Winter activity of Bald Eagles Hali-
aeetus leucocephalus at Navajo Lake, New Mex-
ico. Southwest. Nat. 29:335-342.

Grubb, T. G. 1995. Food habits of Bald Eagles breed-
ing in the Arizona desert. Wilson Bull. 107:258-
274.

Hansen, A. J. 1987. Regulation of Bald Eagle repro-
ductive rates in southeast Alaska. Ecology 68-
1387-1392.

Helander, B. 1985. Reproduction of the White-tailed
Sea Eagle Haliaeetiis alhicilla in Sweden. Ho-
larct. Ecol. 8:21 1-227.

Hunt, W. G., J. M. Jenkins, R. E. Jackman, C. G.
Thelander, and a. T. Gerstell. 1992a. Foraging
ecology of Bald Eagles on a regulated river. J.
Raptor Res. 26:243-256.

Hunt, W. G., B. S. Johnson, and R. E. Jackman.
1992b. Carrying capacity for Bald Eagles winter-
ing along a northwestern river. J. Raptor Res '’b'
49-60.

Jenkins, J. M. and R. E. Jackman. 1995. Field exper-
iments m prey selection by resident Bald Eagles
m the breeding and non-breeding season. J. Field
Ornithol. 65:441-446.

Kelster, G. R, Jr., r. g. Anthony, and E. J. O'Neill.
1987. Use of communal roosts and foraging areas
by Bald Eagles wintering in the Klamath Basin.

J. Wildl. Manage. 51:415-420.

Kiener, A. 1963. Poissons, peche et pisciculture a
Madaga.scar. Centre Technique Forestier Tropical
24, Nogent-sur-Marne, France.

Knight, S. K. and R. L. Knight. 1983. Aspects of

Berkeltmm el cil. • MADAGASCAR FISH-EAGLE FORAGING

food finding by wintering Bald Eagles. Auk 100:
477-484.

Langrand, O. and B.-U. Meyburg. 1989. Range, sta-
tus, and biology of the Madagascar Sea Eagle
Haliaeetiis vociferoides. Pp. 269-277 in Raptors
in the modern world (B.-U. Meyburg and R. D.
Chancellor, Eds.). World Working Group for Birds
of Prey, Berlin, Germany.

Loiselle, P. V. 1993. Madagascar update. Aquatic Sur-
vival 2:1-3.

Love, J. A. 1983. The return of the Sea Eagle. Cam-
bridge Univ. Press, Cambridge, U.K.

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

Mersmann, T. j. 1989. Eoraging ecology of Bald Ea-
gles on the northern Chesapeake Bay with an ex-
amination of techniques used in the study of Bald
Eagle food habits. M.S. thesis, Virginia Polytech-
nic Inst, and State Univ., Blacksburg.

Miller, R. 1966. Simultaneous statistical inference.
McGraw-Hill Book Company, New York.

Razafindramanana, S. 1995. Contribution a I’etude
de la biologie de Haliaeetus vociferoides, Desmur
1845 (Pygargue de Madagascar): reproduction et
domaine vital. Memoire de Diplome d’Etudes Ap-
profondies, Univ. of Antananarivo, Madagascar.

Reinthal, P. N. and M. L. J. Stiassny. 1991. The
freshwater fishes of Madagascar: a study of an
endangered fauna with recommendations for a
conservation strategy. Conserv. Biol. 5:231-243.

Reinthal, P. N., J. Sparks, and K. Riseng. 1995. Ecol-

ogy, vulnerability, and temporal changes in fresh-
water fish communities. P. 20 in Environmental
change in Madagascar (B. D. Patterson, S. M.
Goodman, and J. L. Sedlock, Eds.). Eield Museum
of Natural History, Chicago, Illinois.

Sabine, N. and W. D. Klim.stra. 1985. Ecology of
Bald Eagles wintering in southern Illinois. Trans.
111. State Acad. Sci. 78:13-24.

SAS Institute Inc. 1990. SAS user’s guide: statistics,
version 6, fourth ed. SAS Institute Inc., Cary,
North Carolina.

Scott, W. B. and E. J. Crossman. 1973. Ereshwater
fishes of Canada. Eish. Res. Board Can. Bull. 184:
1-966.

Stalmaster, M. V. 1987. The Bald Eagle. Universe
Books, New York.

Stalmaster, M. V. and R. G. Plettner. 1992. Diets
and foraging effectiveness of Bald Eagles during
extreme winter weather in Nebraska. J. Wildl.
Manage. 56:355-367.

Steenhof, K. 1976. The ecology of wintering Bald
Eagles in southeastern South Dakota. M.Sc. the-
sis, Univ. of Missouri, Columbia.

Steenhof, K., S. S. Berlinger, and L. H. Eredrick-
SON. 1980. Habitat use by wintering Bald Eagles
in South Dakota. J. Wildl. Manage. 44:798-805.

Swenson, J. E., K. L. Alt, and R. L. Eng. 1986. Ecol-
ogy of Bald Eagles in the greater Yellowstone
ecosystem. Wildl. Monogr. 95:1-46.

Whitfield, A. K. and S. J. M. Blaber. 1978. Eeeding
ecology of piscivorous birds at Lake St. Lucia:
part I, diving birds. Ostrich 49:185-198.

Wilson Bull, 111(1), 1999, pp. 22-29

THE RELATIONSHIP BETWEEN SPOTTED OWL DIET AND
REPRODUCTIVE SUCCESS IN THE SAN BERNARDINO
MOUNTAINS, CALIFORNIA

RICHARD B. SMITH,' - M. ZACHARIAH PEERY,' ^ ^ r. j. GUTIERREZ,' AND

WILLIAM S. L AH AYE'-*

ABSTRACT. — We analyzed the breeding season diets of California Spotted Owls {Strix occidentalis occi-
dentalis) in the San Bernardino Mountains from 1987 through 1991 to estimate the relative importance of
individual prey species to owl reproduction. We identified a total of 8441 prey remains from 109 unique terri-
tories, which represents the largest collection of prey remains from a single Spotted Owl population. Dusky-
footed woodrats (Neotoma fuscipes) and Jerusalem crickets {Stenopelmatus fuscus) were the most frequently
consumed taxa (42.2% and 20.7%, respectively), but dusky-footed woodrats dominated Spotted Owl diets by
biomass (74.0%). Spotted owls consumed primarily mammals by both frequency (66.4%) and biomass (95.3%).
After excluding territories with less than 20 prey remains, we compared the diets of 24 nonnesting, 24 unsuc-
cessfully nesting, and 58 successfully nesting pairs of Spotted Owls from 56 unique territories; estimated diet
along a large elevational gradient; and controlled for interterritorial and annual variation in diet. A significant
relationship existed between reproductive status and the percent biomass of woodrats in Spotted Owl diets where
successful nesters consumed a greater percent biomass of woodrats (x = 81.8) than nonnesters (.v = 74.1) but
not unsuccessful nesters (x = 75.5). Unsuccessful nesters and nonnesters did not consume a significantly different
percent biomass of woodrats. The percentage of woodrat biomass in Spotted Owl diets increased with elevation
but did not differ among territories or years. We hypothesized that breeding Spotted Owls were able to meet
the increased energetic demands associated with producing young by consuming primarily large, energetically
profitable prey such as woodrats. Received 6 May 1998, accepted 21 Oct. 1998.

The Spotted Owl (Strix occidentalis) preys
on a wide range of vertebrate and invertebrate
taxa, but primarily on a few species of small
mammals (Gutierrez et al. 1995). The distri-
bution of these small mammals has an impor-
tant influence on the owl’s home range size
(Carey et al. 1992, Zabel et al. 1995), habitat
use patterns (Carey et al. 1992, Carey and
Peeler 1995, Zabel et al. 1995, Ward et al.
1998), and demographic rates (Franklin 1997,
Ward et al. 1998). In particular, prey abun-
dance positively influences Spotted Owl re-
production. Ward and coworkers (1998) found
that dusky-footed woodrats (Neotoma fusci-
pes) were more abundant in the territories of
breeding Northern Spotted Owls (S. o. cauri-
na) than in the territories of nonbreeding owls.
Although this difference was not statistically

' Dept, of Wildlife, Humboldt State Univ., Areata,
CA 95521.

2 Present Address; P.O. Box 266, Mattituck, NY
I 1952.

’ Present Address: 927 Lincoln Way, San Francisco,
CA 94122.

^ Present Address; PO. Box 523, Big Bear City, CA
92314.

'Corresponding author; E-mail:
mz.pl @axc. humboldt.edu

significant, the authors suggested that high
variation in woodrat abundances resulted in
low statistical power for rejecting the hypoth-
esis of no difference between breeders and
nonbreeders. Franklin (1997) showed that the
distribution of woodrat habitat explained a
large amount of interterritorial variation in
Northern Spotted Owl reproductive success.
In addition, studies of other strigids demon-
strate convincingly that reproduction for most
owl species is limited, at least in part, by prey
availability (see Verner et al. 1992 for a re-
view).

Assuming that Spotted Owl reproductive
success is determined in part by food avail-
ability, food habit studies based on prey re-
mains from egested pellets can be used to ex-
amine the relative importance of individual
prey species for reproduction. Such studies
have shown that successful breeders consume
a greater proportion of large prey than unsuc-
cessful breeders (Barrows 1985, 1987;
Thradkill and Bias 1989; White 1996), al-
though Ward and coworkers (1998) found no
difference. These studies have been based
upon relatively few prey items or owl terri-
tories which has resulted in (1) data being
pooled among territories or (2) data from a

Smith et ill. • SPOTTED OWL DIET AND REPRODUCTION

particular territory being considered indepen-
dent from year to year. The first approach
weights all territories equally regardless of the
number of prey items collected. For example,
a breeding territory from which one prey item
was collected is given the same weight as a
breeding territory with hundreds of prey
items. The second approach results in pseu-
doreplication and can dramatically inflate the
probability of rejecting the null hypothesis
(Hurlbert 1984) that breeding and nonbreed-
ing Spotted Owls have the same diets. This is
especially likely because Spotted Owl diets
exhibit interterritorial variation (Laymon
1988).

Because previous diet studies were based
on relatively few territories sampled from
large, open populations, inferences about the
relationship between diet patterns and the
breeding ecology of the Spotted Owl are lim-
ited. We have been studying the entire popu-
lation of California Spotted Owls (S. o. occi-
dentalis) in the San Bernardino Mountains of
southern California since 1987. It is the larg-
est subpopulation within the southern Califor-
nia owl metapopulation (Noon and McKelvey
1992, LaHaye et al. 1994) and occupies a di-
verse array of habitats along a large elevation
gradient. This allowed us to estimate diet over
a range of ecological conditions and reliably
evaluate its relationship with reproduction. In
this paper we compare the diets of nonnesting,
unsuccessfully nesting, and successfully nest-
ing Spotted Owls throughout the breeding sea-
son (March through October), enumerate the
food habits of individuals of this population,
and compare our results to other food habit
studies of the California Spotted Owl.

STUDY AREA AND METHODS

The San Bernardino Mountains Study Area was lo-
cated approximately 140 km east of Los Angeles, Cal-
ifornia and encompassed 1890 km^ with elevations
ranging from 800-3500 m. Mean annual precipitation
ranged from less than 20 cm to more than 100 cm and
was strongly influenced by elevation and topography
(Minnich 1988). The vegetation was diverse, ranging
from Mojave Desert scrub (Vasek and Barbour 1988)
at lower elevations to alpine (Major and Taylor 1988)
on San Gorgonio Mountain. Most Spotted Owls oc-
cupied conifer dominated forest (Sawyer et al. 1988,
Thome 1988) between 1000 m and 2500 m elevation.

Owl survey methods. — We surveyed the study area
for Spotted Owls following methods described by
Franklin and coworkers (1996) during the breeding

season (March-October), 1987-1991. We conducted
1113 nocturnal surveys during which we spent 15 min
at an individual point or walked along designated
routes, using vocal imitations of Spotted Owl calls to
survey the forested habitat within the study area. Sur-
vey points and routes were placed so that all of this
potentially suitable habitat was surveyed each year. We
conducted 1659 diurnal surveys at territories that were
known to be occupied in order to locate nests, locate
roosts, collect regurgitated pellets, and assess repro-
ductive status. We conducted initial diurnal surveys for
each territory in March or April and conducted follow-
up diurnal surveys every 3 to 5 weeks. We collected
pellets by thoroughly searching areas underneath Spot-
ted Owl nests and roosts during most diurnal surveys.
We assessed nesting status by feeding owls live mice
{Mus musculus) during diurnal surveys (Franklin et al.
1996). We considered an owl pair not to have nested
if one member of the pair ate or cached four consec-
utive mice during a single diurnal survey prior to 31
May. Although only one formal survey was conducted
to assess nesting status, multiple surveys were con-
ducted at each territory to band and resight owls. If a
nest was located during one of these surveys, the owls
at that particular territory were of course considered to
be nesting. When a nest was located, we used the same
method to locate fledglings or to determine nest fail-
ure. Owls that did not take sufficient mice were not
included in the study.

Quantifying Spotted Owl diets. — We dissected all
collected pellets and isolated all identifiable prey re-
mains. Identifiable remains included skulls (birds and
mammals), mandibles (mammals, reptiles, and inver-
tebrates), legs (birds and invertebrates), claws (inver-
tebrates), and bills (birds). Remains collected during
each diurnal survey were enumerated to the lowest
possible taxonomic level, and the highest count was
taken as the total number of prey items for that survey.
If, for example, pellets collected during a single survey
contained one woodrat skull, two left mandibles, and
three right mandibles, three woodrats were considered
to be present. The percent frequency of each taxa
(Marti 1987) was then calculated for each territory
pooled among years. The percent biomass (Marti
1987) was calculated for each territory using the mean
prey weights in Table 1 . The mean percent frequency
and biomass among territories were used as estimates
of the overall diet composition. This approach weights
all territories equally so that overall percentages were
not biased towards territories with many prey remains
(Swanson et al. 1974).

Statistical analyses. — Statistical analyses were per-
formed only on the biomass data 1988-1991 because
of small sample sizes in 1987 and because we felt
biomass more accurately represented the energetic im-
portance of each prey taxa.

We used a mixed-model ANOVA approach (Proc
MIXED of pc SAS, version 6.12; Littel et al. 1996) to
estimate variation in Spotted Owl diet, where each diet
variable was considered as the dependent variable in a
separate model. We treated reproductive status (non-

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

TABLE 1.

Diet of Spotted Owls

in the San Bernardino Mountains, California.

Mean prey
weight (g)

Source^

% of prey
by frequency
(n = 8,169)

% of prey
by biomass
(707,193 g)

Mammals

Neotonia fu.scipe.s

173.7

42.2

74.0

Thomoniy.'i hottae

124.4

7.4

10.4

GUiiicomy.s sahrinus

121.5

2.1

3.0

Peromy.'icu.s spp.

22.1

11.3

4.0

Microtus spp.

60.2

1,2

1.8

1.3

Sylvilagus spp.

538.3

0.2

1.0

Sciurus griseii.s

785.0

0,1

1.0

D ipodomy.s me ni a ini

42.1

0.1

<0.1

Scapaiui.s latimanus

55.2

0.4

0.3

Spennophilu.s lateralis

153.4

0.1

0.2

Tamius merriami

67.6

1,2

0.1

Antrozous pallidiis

21.5

<0.1

Eptesiciis fuscus

14.0

<0.1

Lasiiirius cinereus

25.5

0.1

<0.1

My Otis spp.

4.9

0.1

<0.1

Perognathus spp.

14.4

0.1

<0.1

Sore.x ornatus

5.1

0.5

<0.1

Subtotal

66.4

95.3

Birds

64.4

4.3

3.5

Invertebrates

Stenopelnuit us fuscus

2.0

20.7

0.9

Prionus californicus

2.0

3.6

0.1

Scorpionida

2.0

2.7

0.1

Unidentified

2.0

2.1

0.1

Subtotal

29.1

1.2

Reptiles

Sceloporus spp.

10.0

0.1

<0.1

■' Sources were ( I ) specimens from the San Bernardino Mountains in the Museum of Vertebrate Zoology at University of California, Berkeley collections
( specimens from nearest available location to the San Bernardino Mountains in the Museum of Vertebrate Zoology at University of California. Berkeley
collections. (3) Foreman et al. ( 1984; biomass of birds divided by the number of birds in tables 12-14), (4) estimated based on prey size.

nesters, unsuccessful nesters, and successful nesters)
and year as fixed effects, and elevation (at the center
of activity for each territory) as a covariate. We u.sed
territory as a random blocking factor to estimate var-
iation in diet among territories and control for non-
independence of samples within territories. Two-way
interactions between reproductive status and the other
main effects were included in the model. Main effect
by territory interactions were assumed to be nonexis-
tent which resulted in all effects being tested over the
residual mean square (Newman et al. 1997). Signifi-
cant tixed effects were further tested with /-tests on
least squares means using sequential Bonferonni ad-
justments on critical values (Rice 1989). This proce-
dure involves testing each comparison, starting with
the most significant and stopping at the first nonsig-
nificant comparison, using a/{ \ + k - i) as the critical
value where k is the number of post-hoc compari.sons
and /' is the number of the comparison.

Proper use of an ANOVA model requires a normally
distributed dependent variable and equality of varianc-
es among treatment levels (Zar 1984). Because the per-
cent biomass of woodrats was a proportional variable.

and hence formed a binomial distribution, we used the
square root-arcsine transformation to normalize the
data (Zar 1984). Normality was then tested with the
Shapiro- Wilk statistic (Zar 1984). Equality of variances
among reproductive groups was tested with an F-test
(Zar 1984).

By using diet as the dependent variable we do not
imply that diet is a function of reproductive status; the
mixed modeling approach simply provides a conve-
nient way of examining variation in diet in relation to
other variables (including reproduction). We believe
that this approach is justified because ( 1 ) the goal of
the study was to determine if diet differed among non-
nesting, unsuccessfully ne.sting, and successfully nest-
ing Spotted Owls, and (2) it is appropriate to analyze
a correlative relationship with a linear model because
the resultant F-statistic provides a test of the null hy-
pothesis that the multiple correlation coefficient R is
zero (Zar 1984).

RESULTS

Spotted Owl diets.~We iijentified a total of
8441 prey items from 109 unique Spotted Owl

Smilh el al. • SPOTTED OWL DIET AND REPRODUCTION

TABLE 2. The effect of

reproductive status, elevation, and year on the percent biomass

of woodrats in

Spotted Owl diets in the San

Bernardino

Mountains.

Results are from

a mixed-model ANOVA (Littel ct al.

1996) where reproductive status and year

were fixed

effects, elevation

was a covariatc, and

territory^' was a

random blocking factor effect.

Effecl

ncir

(Jdf

Reproductive status

3.65

0.04

Elevation

34.93

<0.01

Year

1.87

0.15

Reproductive status*year

1.17

0.35

Reproductive status*elevation

3.16

0.06

“Territory was not significant (Z = 1.43, P = 0.15).

territories. This represents, to the best of our
knowledge, the largest collection of prey
items recorded from a single Spotted Owl
population. When estimating the overall diet
composition, we excluded territories from
which fewer than 20 prey remains were col-
lected during the entire study period (this re-
sulted in a subsample of 8,169 prey remains
from 71 territories). Dusky-footed woodrats
and Jerusalem crickets (Stenopelmatus fuscus)
were the most common taxa by frequency
(42.2% and 20.7%, respectively). White-foot-
ed mice (Peromyscus spp.) and northern pock-
et gophers (Thomomys bottae) were less com-
mon by frequency (11.3% and 7.4%, respec-
tively). No other taxa contributed more than
4.3% to the total number of prey items (Table
1).

Dusky-footed woodrats dominated Spotted
Owl diet by biomass (74.0%). followed by
pocket gophers (10.4%). Dusky-footed wood-
rats were the largest (173.7 g) of the common
prey items. Western grey squirrels (Sciurus
griseus) and cottontails (Sylvilagus spp.) were
larger, but represented only 2.0% of the total
biomass. No other taxa contributed more than
4.0% to the total biomass consumed (Table 1).
Mammals contributed 66.4% and 95.3% to the
total number of prey items and the total bio-
mass, respectively. Invertebrates contributed
29.1% to the total number of prey items, but
only 1.2% to the total biomass.

Variation in Spotted Owl diets. — For statis-
tical analyses, we excluded territories repre-
sented by fewer than 20 prey items in any one
year. In doing so, we retained a large sample
size (106 samples from 56 unique Spotted
Owl territories; 24 nonnesters, 24 unsuccess-
ful nesters, and 58 successful nesters) and
were able to estimate the percent biomass of

woodrats with reasonable precision on a ter-
ritory by territory basis (mean CV = 0.08,
maximum CV = 0.27). The percent biomass
of pocket gophers was estimated with consid-
erably less precision (mean CV = 0.50, max-
imum CV = 1.00). In addition, the percent
biomass of pocket gophers was highly and
negatively correlated with the percent biomass
of woodrats (r = —0.67, P < 0.01, n = 106).
For these reasons, we did not model the per-
cent biomass of pocket gophers statistically.

Although the percent biomass of woodrats
was not distributed normally (W = 0.93, P <
0.01) the variance in the percent biomass of
woodrats did not differ between the least var-
iable (successful nesters) and the most vari-
able (unsuccessful nesters) reproductive
groups (7^23 57 = 1.46, P > 0.05). Because AN-
OVA is sensitive to heterogeneity of variances
among treatment levels but robust to depar-
tures from normality (Hicks 1993), we as-
sumed the data met the assumptions of the
model.

A significant relationship existed between
reproductive status and the percent biomass of
woodrats in Spotted Owl diets (Table 2). Suc-
cessful nesters consumed a greater percent
biomass of woodrats (x = 81.8 ± 1.5 SE) than
nonnesters (x = 74.1 ± 2.4; ^0017.36 ^ 2.49, P
= 0.017) but not unsuccessful nesters (.v =
75.5 ± 2.4; 02.3,36 = 2.08, P - 0.044), al-

though the difference between successful and
unsuccessful nesters was nearly significant.
Unsuccessful nesters and nonnesters did not
consume a different percent biomass of wood-
rats (t() 050.36 “ 0.04, P > 0.05). The percentage
of woodrat biomass in Spotted Owl diets did
not differ among territories but increased with
elevation (Table 2; Fig. 1 ). The interaction be-
tween elevation and reproductive status was

20 -I ^ 1—

1000 1200 1400

1600 1800 2000
ELEVATION (m)

2200 2400 2600

FIG. 1 . The relationship between elevation and the percent biomass of woodrats in the diets of Spotted Owls
in the San Bernardino Mountains, California, 1988-1991 (h = 0.000085, R~ = 0.26, df = 104, F < 0.01).

not significant (Table 2), indicating that the
percent biomass of woodrats increased with
elevation at a constant rate for nonnesters, un-
successful nesters, and successful nesters. The
percentage of woodrat biomass in the owl’s
diet did not differ among years and no inter-
action existed between year and reproductive
status (Table 2).

DISCUSSION

For an owl pair to produce young, females
must acquire sufficient fat reserves prior to
nesting, and males need to provide sufficient
amounts of food to the female and nestlings
during the nesting period (Hirons 1985).
Dusky-footed woodrats are relatively large
and may have provided an energetically prof-
itable food source that enabled Spotted Owls
to reproduce successfully. This idea is sup-
ported by Ward and coworkers (1998) who
showed that selection of dusky-footed wood-
rats provided an indirect benefit to Spotted
Owl fitness by reducing the amount of habitat
needed to reproduce successfully. Further, Or-
ians and Pearson (1979) predict that central

place foragers such as Spotted Owls (Carey
and Peeler 1995) can increase the rate of en-
ergy return to the central place (nest), and
hence fitness, by consuming large prey. By
consuming large prey, the male can minimize
the number of flights from the point of capture
to the nest, allowing more time for hunting.

Since the diet of unsuccessful nesters was
more similar to nonnesters than successful
nesters, consuming woodrats may be more im-
portant for incubation and brooding than for
nest initiation. This is supported by the fact
that the energetic cost of egg production is
relatively small compared to the amount of
energy needed to provision the female and
young during the nesting period (Ward et al
1998).

Our results support Barrows (1985, 1987),
Thrailkill and Bias (1989), and White (1996)
who found that Spotted Owls that fledged
young consumed a higher proportion of large
prey than Spotted Owls that did not fledge
young. Despite the sample size limitations of
these previous studies, it appears that a true
difference in diet, particularly in terms of prey

Smith et al. • SPOTTED OWL DIET AND REPRODUCTION

size, exists between breeding and nonbreeding
Spotted Owls. The importance of prey size
has also been observed for Black-shouldered
Kite (Elanus caeruleus; Slotow and Perrin
1992) and Common Kestrel {Falco tinnuncu-
lus; Korpimaki 1986) reproduction, suggest-
ing that this may be a common pattern for
raptorial species.

It is possible that the difference between
successful nesters and nonnesters was statis-
tically significant, but not biologically signif-
icant. An 8% difference in the percent bio-
mass of woodrats is relatively small when
compared to the additional amount of food a
male must procure to provision a female and
even just one nestling. Based upon Ward’s and
coworkers’ (1998) energetic calculations, a
male must increase the amount of food he pro-
cures by 276% in order to provision the fe-
male and a Juvenile from egglaying to fledg-
ing. However, the dietary difference observed
between breeders and nonbreeders may be
only one of the factors that allows an owl pair
to produce young. For example, the variation
in diet caused by breeding status may simply
reflect higher prey availabilities in the terri-
tories of breeding and nonbreeding owls (see
below).

The percent biomasses of the prey species
presented here are not necessarily estimates of
the total amount of prey taken because it is
unlikely that we collected all regurgitated pel-
lets. Hence we cannot evaluate the effect of
food supply on reproduction. It is possible that
even though successfully nesting owls con-
sumed a greater percent biomass of woodrats,
they obtained less total biomass from wood-
rats than nonnesters or unsuccessful nesters.
However, this seems unlikely because breed-
ing Spotted Owls need to take considerably
more total prey to feed their juveniles.

Without measuring prey abundances, we
cannot be certain if the relationship between
Spotted Owl diet and reproduction was the re-
sult of differences in prey availability or prey
selection among territories. Spotted Owls se-
lect dusky-footed woodrats more than would
be expected based on availability in both
northern and southern California (Hedlund
1996, Ward et al. 1998). Further, Barrows
(1987) found that Spotted Owl pairs shift from
large to small prey after nest failure and sug-
gested that this change in prey selection was

in response to reduced energy requirements. If
Barrow’s hypothesis was true for our popula-
tion, unsuccessful nesters should have had an
intermediate percentage of woodrat biomass
in their diets. Since this did not occur, we con-
sider it more likely that the difference in the
percent biomass of woodrats between owls
that fledged young and owls that did not
fledge young was the result of differences in
prey availability among territories. Optimal
foraging theory (see Pyke et al. 1977 for a
review) predicts, and raptor field studies show
(Korpimaki 1986, Steenhof and Kochert
1988), that when the density of primary prey
is high within the landscape, the percentage of
that prey in the predator’s diet is also high
(i.e., a functional response). Hence, if wood-
rats occurred at higher densities in the terri-
tories of successfully reproducing owls
(Franklin 1997, Ward et al. 1998) one would
expect these owls to consume a greater per-
centage of woodrats.

Bull and coworkers (1989) found that male
Great Gray Owls (S. nebulosa) preferentially
consumed small prey items at the point of
capture and brought large prey items back to
the nest. Although this presents a potential
source of bias, we do not believe that it affects
the conclusions of this study. For nesting
owls, our pellets were probably biased to-
wards males because we (1) often located
male roosts and (2) females frequently flew
away from the nest to egest pellets (personal
observation). If Bull’s and coworkers’ (1989)
findings are true for Spotted Owls, we may
have underestimated the percent biomass of
woodrats in the diets of nesting owls. This
bias, however, would have decreased the dif-
ference between owls that fledged young and
owls that did not fledge young.

Another potential bias is that pellets may
not represent a random sample of the owl’s
diet. In particular, single prey-item Bam Owl
{Tyto alba) pellets are more likely to contain
large prey than expected by chance (Yom-Tov
and Wool 1997). If this was tme for our Spot-
ted Owl population, the percent frequency and
biomass of relatively large prey such as wood-
rats, pocket gophers, and northern flying
squirrels {Glaucomys sabrinus) would be pos-
itively biased, while the percent frequency and
biomass of relatively small prey such as
white-footed mice would be negatively bi-

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

ased. However, these biases should be consis-
tent for nonnesting, unsuccessfully nesting,
and successfully nesting owls and should not
affect the relationship between reproductive
status and diet that we observed.

Although some similarities existed between
Spotted Owl diets in the San Bernardino
Mountains and other localities in southern
California, some marked differences existed
as well. For example, while the percent fre-
quencies of woodrats and pocket gophers in
Spotted Owl diets reported in this study were
similar to those reported by Barrows (1980,
1987), white-footed mice comprised a consid-
erably smaller percentage (11.3% versus
40.0%) and invertebrates a greater percentage
(29.1% versus 18.0%) of the total prey items
in this study. The percent biomasses of wood-
rats, pocket gophers, and most other prey taxa
were remarkably similar to those reported by
Barrows (1980, 1987).

In the Sierra Nevada, woodrats are the
Spotted Owl’s primary prey at low elevations
but are almost completely replaced by flying
squirrels at high elevations (Verner et al.
1992). In the San Bernardino Mountains,
woodrats increased in Spotted Owl diets with
elevation and flying squirrels were uncommon
(2.1%). Flying squirrels are probably more
abundant in the upper mixed-conifer and red
fir {Abies magnified) zones of the Sierra Ne-
vada (see Waters and Zabel 1995) than they
are in the San Bernardino Mountains where
they reach the southern edge of their range
(Hall and Kelson 1959). The San Bernardino
Mountains apparently lack a relatively large
alternative prey species at higher elevations
and it is possible that most of the other prey
species decrease in elevation as well. Wood-
rats are an important part of the Spotted Owl’s
diet in southern California, both in terms of
biomass and reproduction, and we recommend
that future management of forested habitat in
this region promote high woodrat densities.

ACKNOWLEDGMENTS

Wc thank M. Seamans, P. Carlson. A. Franklin, P.
Ward, E. Forsman, and an anonymous reviewer for
many u.seful comments on earlier drafts. We thank the
many dedicated people who provided field assistance
including A. Kirn, D. Call. D. Cousins, D. Roberts, E.
Pausch. G. Caulkins. J. Gronski, J. Stephenson, J.
Schmid. K. Norgaard. M. Hollister, R. Long, R. Tan-
ner. S. Bailey, S. Stroich. and K. Dimcheff. We thank

B. Stein of the Museum of Vertebrate Zoology at the
University of California, Berkeley for access to mam-
mal collections. Funding was provided by the San Ber-
nardino National Forest, Region 5 USDA-Forest Ser-
vice, Southern California Edison, California Depart-
ment of Fish and Game, Snow Summit Ski Coopera-
tion, San Bernardino County Audubon Society, the
California State Resource Agency, Pacific Southwest
Forest and Range Experiment Station, Bear Mountain,
Ltd., the Army Corps of Engineers, and Salad King
Inc.

LITERATURE CITED

Barrows, C. W. 1980. Feeding ecology of the Spotted
Owl in California. Raptor Res. 14:73-78.
Barrows, C. W. 1985. Breeding success relative to
fluctuations in diet for Spotted Owls in California.
U.S.D.A. Pacific Northwest Forest and Range Ex-
periment Station. Gen. Tech. Rept. PNW-GTR-
185:50-54

Barrows, C. W. 1987. Diet shifts in breeding and non-
breeding Spotted Owls. Raptor Res. 21:95-97.
Bull, E. L., M. G. Henjum, and R. S. Rohweder.
1989. Diet and optimal foraging of Great Grey
Owls. J. Wildl. Manage. 53:47—50.

Carey, A. B., S. P. Horton, and B. L. Biswell. 1992.
Northern Spotted Owls: influence of prey base and
landscape character. Ecol. Monogr. 62:223-250.
Carey, A. B and K. C. Peeler. 1995. Spotted Owls:
resource and space use in mosaic landscapes. J.
Raptor Res. 29:223-239.

Forsman, E. D., E. C. Meslow, and H. M. Wight.
1984. Distribution and biology of the Spotted Owl
in Oregon. Wildl. Monogr. 87:1—64.

Franklin, A. B. 1997. Factors affecting temporal and
spatial variation in Northern Spotted Owl popu-
lations in northwest California. Ph.D. diss., Col-
orado State Univ., Fort Collins.

Franklin, A. B., D. R. Anderson, E. D. Forsman, K.

P Burnham, and F F Wagner. 1996. Methods
for collecting and analyzing demographic data on
the Northern Spotted Owl. Stud. Avian Biol 17-
12-26.

Gutierrez, R.J., A. B. Franklin, and W. S. LaHaye.
1995. Spotted Owl {Strix occidenlalis). In The
birds of North America, no. 179 (A. Poole and F
Gill, Eds.). The Academy of Natural Sciences,
Philadelphia, Pennsylvania; The American Orni-
thologists’ Union, Washington, D.C.

Hall, R. and K. R. Kelson. 1959. The mammals of
North America. Ronald Press Co., New York
Hedlund, C. D. 1996. Relative proportion of prey spe-
cies m the diet of California Spotted Owls Stri.x
occicIenlali.K occiclentalis in the San Gabriel
Mountains of Southern California. M.Sc. thesis,
California State Polytechnic Univ., Pomona.

Hicks, C. R. 1993. Fundamental concepts in the design
ot experiments. Saunders College Publishing
New York.

H.RONs, G. j. M. 1985. The importance of body re-

Smith et ill. • SPOTTED OWL DIET AND REPRODUCTION

serves for successful reproduction in the Tawny
owl (Strix aluco). J. Zool. (Lond.) 1:1-20.

Hurlbert, S. H. 1984. Pseudoreplication and the de-
sign of ecological field experiments. Ecol. Mon-
ogr. 54: 1 87—2 1 I .

Korpimaki, E. 1986. Diet variation, hunting habitat,
and reproductive output in the Kestrel Falco lin-
niinculus in the light of optimum diet theory. Or-
nis Fennica 63:84-90.

LaHaye, W. S., R. J. Gutierrez, and H. R. Akqa-
KAYA. 1994. Spotted Owl metapopulation dynam-
ics in southern California. J. Anim. Ecol. 63:775-
785.

Laymon, S. D. 1988. The ecology of the Spotted Owl
in the central Sierra Nevada, California. Ph.D.
diss., Univ. of California, Berkeley.

Littel, R. C., G. a. Milliken, W. W. Stroup, and R.
D. WoLFiNGER. 1996. SAS system for mixed mod-
els. SAS Institute, Cary, North Carolina.

Major, J. and D. W. Taylor. 1988. Alpine. Pp. 601-
678 in Terrestrial vegetation of California (M. G.
Barbour and J. Major, Eds.). California Native
Plant Society, Sacramento, California.

Marti, C. D. 1987. Raptor food habits studies. Pp. 67—
81 in Raptor management techniques manual (B.
A. G. Pendleton, B. A. Millsap, K. W. Cline, and
D. M. Bird, Eds.). National Wildlife Federation,
Washington, D.C.

Minnich, R. a. 1988. The biogeography of fire in the
San Bernardino Mountains of California. Univ. of
California Press, Los Angeles.

Newman, J. A., J. Bergelson, and A. Grafen. 1997.
Blocking factors and hypothesis tests in ecology:
is your statistics text wrong? Ecology 78:1312-
1320.

Noon, B. R. and K. S. McKelvey. 1992. Stability
properties of the Spotted Owl metapopulation in
Southern California. U.S.D.A. Pacific Southwest
Research Station. Gen. Tech. Rep. PSW-GTR-
133:187-206.

Orians, G. H. and N. E. Pearson. 1979. On the theory
of central place foraging. Pp. 155-177 in Analysis
of ecological systems (D. J. Horn, R. D. Mitchell,
and G. R. Stairs, Eds.). Ohio Univ. Press, Colum-
bus.

Pyke, G. H., R. H. Pulliam, and E. L. Charnov. 1977.
Optimal foraging: a selective review of theory and
tests. Quart. Rev. Biol. 52:137-154.

Rice, W. R. 1989. Analyzing tables of statistical tests.
Evolution 43:223-225.

Sawyer, T. O., D. A. Thornburgh, and T. R. Griffin.
1988. Mixed evergreen forest. Pp. 359-382 in

Terrestrial vegetation of California (M. G. Bar-
bour and J. Major, Eds.). California Native Plant
Society, Sacramento, California.

Slotow, R. and M. R. Perrin. 1992. The importance
of large prey for Blackshouldered Kite reproduc-
tion. Ostrich 63:180-182.

Steenhof, K. and M. N. Kochert. 1988. Dietary re-
sponses of three raptor species to changing prey
densities in a natural environment. J. Anim. Ecol.
57:37-48.

Swanson, G. A., G. L. Krapu, J. C. Bartonek, J. R.
Serie, and D. H. Johnson. 1974. Advantages in
mathematically weighting waterfowl food habits
data. J. Wildl. Manage. 38:302-307.

Thorne, R. E 1988. Montane and subalpine forests of
the Transverse and Peninsular Ranges. Pp. 537-
558 in Terrestrial vegetation of California (M. G.
Barbour and J. Major, Eds.). California Native
Plant Society, Sacramento, California.

Thrailkill, j. and M. A. Bias. 1989. Diets of breeding
and nonbreeding California Spotted Owls. J. Rap-
tor Res. 23:39-41 .

Vasek, F. C. and M. G. Barbour. 1988. Mojave De-
sert scrub vegetation. Pp. 835-867 in Terrestrial
vegetation of California (M. G. Barbour and J.
Major, Eds.). California Native Plant Society, Sac-
ramento, California.

Verner, j., R. j. Gutierrez, and G. I. Gould, Jr.
1992. The California Spotted Owl: general biol-
ogy and ecological relations. U.S.D.A. Pacific
Southwest Research Station. Gen. Tech. Rep.
PSW-GTR- 133:55-77.

Ward, J. R, Jr., R. J. Gutierrez, and B. R. Noon.
1998. Habitat selection by Northern Spotted Owls:
the consequences of prey selection and distribu-
tion. Condor 100:79-92.

Waters, J. R. and C. J. Zabel. 1995. Northern flying
squirrel densities in fir forests of northeastern Cal-
ifornia. J. Wildl. Manag. 59:858—866.

White, K. 1996. Comparison of fledging success and
sizes of prey consumed by Spotted Owl in north-
western California. J. Raptor. Res. 30:234-236.

Yom-Tov, Y. and D. Wool. 1997. Do the contents of
Barn Owl pellets accurately represent the propor-
tion of prey species in the field? Condor 99:972—
976.

Zabel, C. J., K. McKelvey, and J. P. Ward, Jr. 1995.
Influence of primary prey on home-range size and
habitat-use patterns of Northern Spotted Owls
(Strix occidentaUs caurina). Can. J. Zool. 73:433-
439.

Zar, j. H. 1984. Biostatistical analysis. Second cd.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

Wilson Bull., 111(1), 1999, pp. 30-37

FOOD, FORAGING, AND TIMING OF BREEDING OF THE BLACK

SWIFT IN CALIFORNIA

MANUEL MARIN* 2

ABSTRACT. — The nestling diet and breeding seasonality of the Black Swift {Cypseloides niger) were studied
in southern California 1990-1992. The peak (40%) of egg laying was in mid-June and the peak of fledging
(60%) was mid- to late August {n = 87 nests). Winged ants comprised 91% (n = 1179 prey items, 10 boluses)
of nestling diet. Three main prey size classes were found; 6, 8, and 13 mm. Food bolus mass increased and
number of trips per day to feed the nestlings decreased with nestling age. The parents made short and long
foraging trips during early morning hours and long trips from early to late afternoon. Short trips were observed
only during the first half of the nestling period. During the last half of the nestling period, parent swifts made
a single foraging trip per day that lasted about 12 hrs. Perhaps the short foraging bouts are for feeding the
young, whereas the long foraging bouts are not only for feeding the young but also for parental energy storage.
The single foraging bout, during the mid- and late nestling period, might also serve to store fat for migration
by the adults. Received 13 Feb. I998\ accepted 24 Oct. 1998.

The Black Swift {Cypseloides niger) is a
member of the subfamily Cypseloidinae
which consists of 12-13 species, most of
which are tropical or subtropical in their
breeding distribution. The Black Swift is
found locally in the West Indies, Middle
America, and north through much of western
North America to southeastern Alaska (Bent
1940; AOU 1957, 1983). For a species with
such a wide latitudinal distribution, quantita-
tive data on diet and timing of breeding are
rare. Most of what is known about the Black
Swift is limited to breeding and distributional
records (e.g., Vrooman 1901, 1905; Michael
1927; Dixon 1935; Knorr 1961; Foerster
1987; Foerster and Collins 1990; Stiles and
Negret 1994). This reflects the difficulty of
studying this species because of its aerial life
style and its usually inaccessible nest sites.
Here 1 present new information on diet and
timing of breeding of this species.

The Black Swift is a summer breeding vis-
itor to western North America, and like many
migratory species there, has a restricted breed-
ing season. Furthermore, it has a proportion-
ately large egg, a single-egg clutch, is single
brooded, and has a long incubation and nest-
ling period (Marin 1997). These factors

' Dept, of Biological Science.s and Museum of Nat-
ural .Science, I 19 Foster Hall, Louisiana State Univ.,
Baton Rouge, LA 70803; E-mail; Zomari@LSU.edu
^ Present address; Western Foundation of Vertebrate
Zoology. 439 Callc San Pablo, Carmarillo, CA 93012;
E-mail; MMA@wfv/..org

should constrain variation in the timing of
breeding of the Black Swift.

Swifts catch airborne insects and ballooning
spiders (Lack and Owen 1955, Whitacre
1991). During the reproductive season, breed-
ing cypseloidine swifts accumulate insects and
arachnids in the back of the throat continuing
into the esophagus and bind them with saliva
to produce a sticky assortment of insects. This
insect conglomerate or food bolus is produced
exclusively to feed the nestling(s) and have
never been reported outside the breeding sea-
son. Alive or dead these insects are complete,
making them ideal for identifying and quan-
tifying diets. Hespenheide (1975) pointed out
that one difficulty in analyzing bird diets is
knowing which is more important: the number
or the size of food items. Some diet studies
of New World swifts have addressed both pa-
rameters [e.g., Whitacre (1991)]; however,
most have emphasized only one or they had
small sample sizes (e.g., Collins and Landy
1968, Hespenheide 1975, Foerster 1987, Bull
and Beckwith 1993, Marm and Stiles 1993).
Other authors have considered only the num-
ber of prey items (e.g., Beebe 1949; Rathbun
1925; Rowley and Orr, 1962, 1965; Marin and
Stiles 1992). One problem in quantifying swift
diets is the source of prey samples, either
stomach contents or food boluses. Stomach
contents of adults are prey items that the adult
bird has fed upon, whereas food boluses are
prey fed to nestlings. Thus, examination of
stomach contents versus food boluses might
produce different results. Because my focus

Mcuin • DIET AND BREEDING OE BLACK SWIFT IN SOUTHERN CALIFORNIA

was on the diet of Black Swift nestlings, 1
analyzed only food boluses.

METHODS

Most data were gathered during a study of the
breeding biology of the Black Swift in the San Jacinto
Mountains, Riverside Co., California. The study site is
at 1500 m elevation, and the surrounding area is mon-
tane forest. The study site was visited 40 times from
1990 to 1992, between the months of May and Sep-
tember. Observations were made in blocks of 4 to 7
hours, from 05:00 to 12:00, 12:00 to 19:00, or 19:00
to 23:00 PST, throughout the breeding season. For
more detailed information on the study site and distri-
bution of visits see Marin (1997) and references there-
in.

Data on nestling diet were obtained from regurgi-
tated food boluses from adult swifts captured upon ar-
rival at the nest. The boluses were weighed immedi-
ately (to nearest 0.1 g; Pesola scale) and placed in a
vial containing alcohol. Insects in the boluses were
counted and measured to the nearest 0.1 millimeter
with a micrometer under a microscope in the lab.

Ten boluses (n = 1179 prey items) were collected
from different adults on different dates and years
throughout the study to minimize nest disturbance.
Seven boluses were collected in 1991 and three during
1992. Bolus mass and time of collection were recorded
for nine of them. I measured the length of 15 randomly
selected individual prey items per prey species per bo-
lus, and I used the average length as the mean of that
species in the specific food bolus. Insects were sorted
to morphospecies and identified to families using Bor-
ror and Delong (1970), Borror and White (1970), and
Powell and Hogue (1979).

To assess timing of breeding, I combined field nest
data (n = 20) with archived nest and egg data cards
(n = 67; see Acknowledgments for list of sources),
along with museum study skins and the literature. To
determine length of the breeding season, I used egg
laying, hatching, and fledgling periods. I restricted anal-
yses of museum egg data cards to two areas in Cali-
fornia: mountains (San Jacinto area) and coast (Santa
Cruz Co.). I compared the estimated hatching date
from the museum egg data cards to my own field data
on hatching dates gathered at San Jacinto to look for
potential date discrepancies between actual and esti-
mated data. The incubation stage given in the egg data
cards (e.g., fresh, commenced, advanced, etc.) was ex-
trapolated using the known days of the incubation
stage from San Jacinto (24 days; Marin 1997). I can-
died 16 eggs at San Jacinto and determined that no
egg of this species could be blown without signifi-
cantly damaging or destroying the shell by day 16-18.
Thus, any museum egg specimen of this species was
unlikely to have been collected beyond 18 days of in-
cubation. I estimated the duration of each stage visible
through candling as: “fresh” (0-4, 5 days) “veins and
small embryo” (5-10 days), “embryo” (11-14 days),
and “large embryo” (>15 days). These data were used

TABLE 1.

Contents of food boluses (n = 10) of

Cyp.seloides niger from San Jacinto, California.

Order

Family

Number

Isoptera

Hodotermitidae

Hemiptera

Pentatomidae

Coreidae

Miridae

Reduviidae

Nabidae

Homoptera

Cicadidae

Cicadellidae

Aphidae

Neuroptera

Hemerobiidae

Myrmeleonitidae

Corydalidae

Family?

Coleoptera

Buprestidae

Lepidoptera

Pyralidae

Family?

Diptera

Family?

Hymenoptera

1074

Formicidae

1074

Arachnidea

TOTAL

1179

to infer incubation stage (from the data cards) and to
estimate egg laying dates. Using the known incubation
and fledgling period (24 and 48 days, respectively;
Marin 1997), the estimated laying, hatching, and fledg-
ing dates were compared and then combined with field
data. The dates were separated by month, and each
month was subdivided into early, middle, and late.

Rainfall data were gathered for each month for the
years of study (1990-1992) and also 1963-1973 from
the Idyllwild Fire Department weather station about
1.5 km from study area (National Oceanic and At-
mospheric Administration 1991, 1992).

I was able to identify adults individually because
they were already banded by C. Collins and K. Foers-
ter because this was the main study site that Foerster
{ 1987) used for his MS thesis work.

RESULTS

Nestling diet and foraging. — All boluses
were composed of one predominant prey spe-
cies, suggesting that the birds had fed on
swarming species or highly localized prey.
Winged ants were the majority of prey items
(91%; Table 1). In 10 boluses the average pro-
portion of female winged ants was 79.5 %

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

1 I

0-10 ■ ■

qI Imbj-I

1 2 3 4 5 6 7 8 9 10 11 12 13 14

LENGTH (mm)

FIG. 1. Frequency distribution of prey sizes taken by the Black Swift {Cypseloides niger) in southern
Calitornia. Data are from 10 food boluses (/; = 1154 prey items); prey size categories are; 0.5— 1.5 mm = I;
1.6-2. 5 mm = 2; 2. 6-3. 5 mm = 3; etc.

(range 20-100%; n = 1179 prey items). Sex-
ual dimorphism in ants accounted for prey
size differences among boluses. Male ants
were smaller and ranged 2-6 mm, whereas fe-
males ranged 6-14 mm. Sometimes a bolus
contained a small number of female ants but
those ants were the largest prey. Thus, female
winged ants had the highest volume/prey in
all boluses. The average measurable prey size
was 7.4 mm (range 1.8-14.5 mm; n = 1 154).
Three main size classes were found in the bo-
luses: 6 mm (33%), 8 mm (39%), and 13 mm

(12.2%; Fig. 1). Two length categories (6 and
8 mm) made up the bulk of the samples. Be-
cause only a few prey taxa were represented
in the diet, size frequencies for all prey spe-
cies follow a similar pattern.

Nestling age and bolus mass were positive-
ly correlated (r" = 0.93, P < 0.001, /? = 10;
Fig. 2). No correlation was found between
mean prey size per bolus and chick age (r- =
0.08, P > 0.05) or mass per bolus (r^ = 0.04,
P > 0.05, ?i = 10). No individual prey item
was weighed, but female winged ants were

•• • •

0 10 20 30 40

AGE OF YOUNG (days)

FIG. 2. Relation.ship between bolu.s mass and nestling age (F = 0.93; P < 0 001 n

= 10).

Mcuin • DIET AND BREEDING OE BLACK SWIFT IN SOUTHERN CALIFORNIA

FIG. 3. Time of day young were fed versus their age. None of the adult birds were observed feeding young
between the two feeding clusters or during early morning after 30 days.

undoubtedly the heaviest prey items because
they were the largest. After day 30 I never saw
any adults feed young in the morning; they
were fed mainly late in the evening (Fig. 3).
Other nestlings not included in the analysis
because they were inaccessible and of un-
known age, but at least 30 days old, were also
observed being fed between 18:30—20:00. The
overall pattern seems to be that as age in-
creased, feeding rate decreased, but bolus
mass increased (Figs. 2, 3).

Data gathered from adults feeding young at
the nest showed two clusters of feeding times:
between 8:30 and 12:30 and after 18:30 (Fig.
3). The birds usually left the cave at about 05:
30. In two instances, however, some departed
earlier unnoticed because of the darkness.
This implies that they spent 3-7 hours search-
ing for food for the first feeding bout. For the
second bout, the birds were away from the
nest longer: 6-8 hrs (Fig. 3). I never observed
nestlings being fed between 12:30-18:30. I
did not gather data late in the evening or at
night during the early nestling stage so late
arrivals and feeding at that stage are possible.
If the intervals between feeding bouts were
consistent through the season, then older nest-
lings often waited more than 12 hrs between
meals when they were well grown or more
than 30 days of age.

Breeding season. — Hatching dates were the

main variable I used to compare the timing of
breeding between a mountain site (San Jacinto
area, San Bernardino Co., California) and a
coastal site (Santa Cruz area, Santa Cruz Co.,
California). The San Jacinto data were pri-
marily from my observations, whereas the
Santa Cruz data were taken from museum nest
and egg data cards. I found no significant dif-
ference between the coastal and mountain
sites in timing of hatching (Fisher’s exact Test
(2-tail): P > 0.05). Therefore, I concluded that
it was safe to pool both field and museum data
for coastal and interior southern California.
Most eggs were laid during mid-June (40%),
with 30% during late June (Fig. 4). The ear-
liest laying date was estimated to be 18 May,
from an egg set collected in 1960 near Santa
Cruz, California. The latest date for egg laying
on the data cards was estimated to be 12 July
1921, from the same site as the earliest date.
Some of the observed variation might be due
to inter-year differences, which are difficult to
evaluate with the present data. Nevertheless,
most eggs (81%) had an estimated laying date
in June. The earliest estimated date for hatch-
ing was about 1 1 June (same nest as above)
and the latest date was about 5 August from
the same site. In total, 89% of the hatching
dates were in July; 24% of the estimated
hatching dates were during the first 10 days
of July and 53% during the middle third of

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

FIG. 4. Distribution of estimated dates of egg laying (open bars), hatching (black bais), and fledging (stipled
on bars) of Black Swifts in southern California.

July. Sixty percent of the young were esti-
mated to fledge during mid- to late August
and 29% during the first 10 days of Septem-
ber. The highest proportion of fledging (78%)
was estimated to occur between late August
and early September (Fig. 4).

DISCUSSION

Contrary to most, if not all tropical cypse-
loidines, the southern California populations
of the Black Swift breeds during the dry sea-
son. For the small, tropical cypseloidine
swifts, rainfall itself can be as important a
stimulus as food in initiating breeding. Mois-
ture is needed to keep the appropriate condi-
tions for nest “growth” and maintenance
(Marm and Stiles 1992). Many Black Swifts
did not build a nest at all, instead layed eggs
directly on ledges, especially in the coastal
sites (Man'll 1997). This might be related to
the lack of the proper nesting materials (moss-
es and liverwoths).

The breeding season of the Black Swift in
southern California is spread over 4.5-5
months (Foerster 1987, Marin 1997). Lack
(1954, 1968) observed that breeding in most
species of birds is timed to occur when food
is most abundant, especially in temperate re-
gions.

In the western United States, Chapman
( 1 954) noted that ants swarmed from May

through September and that the peak of ant
swarming was July. The observed peak of egg
hatching in Black Swifts was also July coin-
ciding with the peak of ant swarming (Fig. 5).
These data support Holroyd’s and Jalkotzy’s
(in Campbell et al. 1990) suggestion that the
breeding of the Black Swift in southwestern
Canada was timed to the swarming of flying
ants (Hymenoptera). In the western U.S. ants
swarm in large numbers on mountain and
ridge tops for several days (Chapman 1954).
The peak time of ant swarming observed by
Chapman (1954) was from 07:00 to 14:00 and
coincides with the first period of shorter feed-
ing bouts in the Black Swifts (Fig. 3).

Foerster (1987) reported average prey sizes
from two boluses (n = 289 prey items) as 9.9
and 10.2 mm, slightly larger than my aver-
ages. He did not report sizes smaller than 7
mm or larger than 13 mm. Foster (1987) spec-
ulated on possible size selection by the swifts;
however, I observed 46.5% percent of prey
items below and above those categories (n =
1179 prey items, 10 boluses). The . data from
this and other studies (Collins and Landy
1968, Foerster 1987) suggest that prey items
given to the nestlings are selected not by size
but by insect taxon. This is probably a con-
sequence of feeding on insect swarms.

The mam diet of Black Swift nestlings at
San Jacinto was winged ants, which have a

Marin • DIET AND BREEDING OE BLACK SWIFT IN SOUTHERN CALIFORNIA

C/3

100

□ ANT SWARMING
■ SWIFT HATCHING

JUN

JUL AUG

MONTH

SEP

FIG. 5. Frequency of Black Swift hatching and ant swarming in southern California. Data for ant swarming
are from Chapman (1954).

high fat content. The large preponderance of
winged ants in the nestling diet is similar that
of other cypseloidine swifts (Whitacre 1991).
The percent fat per dry weight in alate ants
ranges from 23.8 to 59.5% in females and
from 3.3 to 9.6% in males (Taylor 1975, Red-
ford and Dorea 1984). A nestling of any bird
species fed a diet rich in energy could accu-
mulate large amounts of subcutaneous fat. Be-
fore fledging the young Black Swift accumu-
lates much visible subcutaneous fat and at-
tains up to 148% of adult body mass; it reach-
es adult mass at day 15-16 of the nestling
period (Marin 1997). The limited inter-year
sampling by Foerster (1987) and myself sug-
gests that the swifts at San Jacinto, during the
breeding season, may specialize in exploiting
local concentrations of 2-3 ant species (Cam-
ponotus spp.). Winged ants are a temporarily
superabundant, patchy, and ephemeral, but
lipid-rich food source. Other important prey
items included Hemiptera and Homoptera (Ta-
ble 1).

From scattered observations, (e.g., Michael
1927, Smith 1928, Bent 1940, Collins 1998,
Collins and Peterson 1998) there is a general
agreement that Black Swift nestlings are fed
at long intervals, primarily early in the morn-
ing and late in the afternoon or at night. My
data corroborate those conclusions (Fig. 3).
The alternation of long and short foraging

trips resembles the strategy of energy expen-
diture described for foraging and food deliv-
ery in pelagic seabirds. Charurand and Wei-
merskirch (1994) and Weimerskirch and co-
workers (1994) showed that long trips were
primarily for parental food storage as well as
nestling food gathering, whereas short trips
were used to deliver food to the nestlings. Al-
though the duration of seabirds’ trips is days,
instead of hours as in swifts, they might well
serve analogous purposes.

Like seabirds, the Black Swift might gain
weight on the long trips and lose it overnight.
Black Swifts have a high metabolic rate and
lose on average 7.9 % of body mass overnight
(Marin, unpubl. data). Thus, the need for the
long foraging bout is in accordance with the
energy storage hypothesis (Chaurand and
Weimerskirch 1994, Weimerskirch et al.
1994). Black Swift migration occurs imme-
diately after the nestlings fledge; other species
of swifts (e.g., Chaetura spp.) stay a few
months after breeding, probably to store some
energy for migration (Marm 1997). Accord-
ingly, this long single foraging bout might
also serve to store energy for migration, par-
ticularly during the later part of the breeding
season.

As the Black Swift nestlings increase in
age, it seems that the adults feed them only
late at night. Quantitative data on feeding

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

rates or number of trips per day with respect
to nestling age are scarce; however, declines
in the number of trips per day with nestling
age have been reported in other swift species
(Malacame et al. 1992, Oniki et al. 1992).

Lack (1954, 1968) suggested that seabirds
with long nestling periods and single egg
clutches were energy limited. This energy lim-
itation was in food finding, food delivery, or
both. Some swifts, particularly cypseloidines,
have life history parameters similar to procel-
lariiform seabirds (Lack and Lack 1951; Lack
1956, 1968; Marin and Stiles 1992; Marin
1993). For example, the Black Swift rears a
single, slowly growing nestling, which sug-
gests a constraint in either finding or deliver-
ing food. Because Black Swift nestlings re-
quire a highly specialized diet in order to have
an initial fast growth and acquire a size larger
than the adult quickly. 1 predict that the major
constraint is food finding.

ACKNOWLEDGMENTS

I thank R. Corado, L. Lyon. D. MacLean, J. Schmitt.
C. Sumida. and W. Wehtje for their companionship and
field help. This paper benefited from comments by C.
T. Collins. K, Naoki. A. G. Navarro S.. J. V. Remsen.
A. Styring. D. Whitacre. and an anonymous reviewer.

I thank the following museums for access to specimens
and egg card data: R Sweet. M. LeCroy, and E Vuil-
leumier. American Museum of Natural History, New
York; N. Johnson and C. Cicero, Museum of Verte-
brate Zoology University of California, Berkeley; R.
Prys-Jones and M. Walters, Natural History Museum.
Tring (ex-British Museum of Natural History); R Col-
lins, Santa Barbara Museum of Natural History, Santa
Barbara; L. Kiff, C. Sumida, R. Corado, and W. We-
htje, Western Foundation of Vertebrate Zoology, Cam-
arillo, The Western Foundation of Vertebrate Zoology
helped with funds for the field work and with a small
grant to cover some preparation costs, and AVINET
provided some field equipment. Financial assistance
was also provided by a Collection Study Grant of the
American Museum of Natural History. I am grateful
to all those egg and bird collectors who seldom receive
recognition for what they did.

LITERATURE CITED

American Ornithologi.sts' Union. 1957. Check-list
of North American birds, fifth ed. American Or-
nithologists' Union. Port City Press, Baltimore,
Maryland.

American Ornithologi.sts' Union. 1983. Check-list
of North American birds, sixth ed. Allen Press,
Lawrence, Kan.sas.

Beebe. W, 1949, The swifts of Rancho Grande, north

central Venezuela, with special reference to mi-
gration. Zoologica 34;53-62.

Bent, A. C. 1940. Life histories of North American
cuckoos, goatsuckers, hummingbirds, and their al-
lies. U.S. Nat. Mus. Bull. 176:1-506.

Borror, D. L. and D. M. DeLong. 1970. An intro-
duction to the study of insects. Holt, Rinehart and
Winston, New York.

Borror, D. L. and R. E. White. 1970. A field guide
to the insects. Houghton Mifflin Co., Boston,
Massachusetts.

Bull, E. L. and R. C. Beckwith. 1993. Diet and for-
aging behavior of Vaux's Swifts in northeastern
Oregon. Condor 95:1016-1023.

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.
Vol. 11, nonpasserines. Royal British Columbia
Museum, Victoria.

Chapman, J. A. 1954. Swarming of ants on western
United States mountain summits. Pan-Pacific En-
tomol. 30:93-102.

Chaurand, T. and H. Weimerskirch. 1994. The reg-
ular alternation of short and long foraging trips in
the Blue Petrel Halohaeno caerulea'. a previously
undescribed strategy of food provisioning in a pe-
lagic seabird. J. Anim. Ecol. 63:275-282.

Collins, C. T. 1998. Food delivery and chick provi-
sioning in cypseloidine swifts. Bull. Brit. Orni-
thol. Club 1 18:108-1 12.

Collins, C. T. and M. J. Landy. 1968. Breeding of
the Black Swift in Veracruz, Mexico. Bull. South.
Calif. Acad. Sci. 67:266-268.

Collins, C. T. and B. M. Peterson. 1998. Nocturnal
chick provisioning by Black Swifts. West Birds
29:227-228.

Dixon, J. S. 1935. Nesting of the Black Swift in Se-
quoia National Park. Condor 37:265-267.

Foerster, K. S. 1987. The distribution and breeding
biology of the Black Swift {Cypseloides niger) in
southern California. M.Sc. thesis. California State
Univ., Long Beach.

Foerster. K. S. and C. T. Collins. 1990. Breeding
distribution of the Black Swift in southern Cali-
fornia. West. Birds 21:1-9.

Hespenheide, H. a. 1975. Prey characteristics and
predator niche width. Pp. 158-180 in Ecology and
evolution of communities (M. L. Cody and J. M.
Diamond, Eds.). Belknap Press, Cambridge, Mas-
sachusetts.

Knorr, O. a. 1961. The geographical and ecological
di,stribution of the Black Swift in Colorado. Wil-
son Bull. 73:155-170.

Lack. D. 1954. The natural regulation of animal num-
bers. Clarendon Press, Oxford, U.K.

Lack, D. 1956. Swifts in a tower. Methuen and Co
London. U.K.

Lack, D. 1968. Ecological adaptations for breeding in
birds. Methuen and Co., London, U.K

Lack D^ AND E. Lack. 1951. The breeding biology of
the Swift Apu.s' apus. Ibis 93:501-546.

Mann • DIET AND BREEDING OE BLACK SWIFT IN SOUTHERN CALIFORNIA

Lack, D. and D. E Owen. 1955. The food of the swift.
J. Anim. Ecology. 24:120-136.

Malacarne, G., M. Cucco, and G. Orecchia. 1992.
Nest attendance, parental roles and breeding suc-
cess in the Pallid Swift (Apus pallidus). Vogel-
warte 36:203-210.

Marin, M. 1993. Patterns of distribution of swifts in
the Andes of Ecuador. Avocetta 17:117-123.

Marin, M. 1997. Some aspects of the breeding biology
of the Black Swift. Wilson Bull. 109:290-306.

Marin, M. and E G. Stiles. 1992. On the biology of
five species of swifts (Apodidae, Cypseloidinae)
in Costa Rica. Proc. West. Found. Vertebr. Zool.
4:287-351.

Marin, M. and E G. Stiles. 1993. Notes on the bi-
ology of the Spot-fronted Swift. Condor 95:479-
483.

Michael, C. W. 1927. Black Swifts nesting in Yosem-
ite National Park. Condor 24:89-97.

National Oceanic and Atmospheric Administra-
tion. 1991. Climatological data California. 95(6-
9): 1-28.

National Oceanic and Atmospheric Administra-
tion. 1992. Climatological data California. 96(6—
9): 1-28.

Oniki, Y, E. O. Willis, and M. M. Willis. 1992.
Chaetura andrei ( Apodiformes, Apodidae): as-
pects of nesting. Ornitol. Neotrop. 3:65-68.

Powell, J. E. and C. L. Hogue. 1979. California in-
sects. Univ. of California Press, Berkeley.

Rathbun, S. F. 1925. The Black Swift and its habits.
Auk 42:497-516.

Redford, K. H. and j. G. Dorea. 1984. The nutri-

tional value of invertebrates with emphasis on ants
and termites as food for mammals. J. Zool.
(Lond.) 203:385-395.

Rowley, J. S. and R. T. Orr. 1962. Nesting and feed-
ing habits of the White-naped Swift. Condor 64:
361-367.

Rowley, J. S. and R. T. Orr. 1965. Nesting and feed-
ing habits of the White-collared Swift. Condor 67:
449-456.

Smith, E. 1928. Black Swifts nesting behind a water-
fall. Condor 30:136-138.

Stiles, F. G. and A. J. Negret. 1994. The nonbreeding
distribution of the Black Swift: a clue from Co-
lombia and unsolved problems. Condor 96:1091 —
1094.

Taylor, R. L. 1975. Butterflies in my stomach. Or:
insects in human nutrition. Woodbridge Press,
Santa Barbara, California.

Vrooman, a. G. 1901. Discovery of the egg of the
Black Swift Cypseloides niger borealis. Auk 18:
394-395.

Vrooman, A. G. 1905. Discovery of a second egg of
the Black Swift. Condor 7:176—177.

Weimerskirch, H., O. Chastel, L. Ackermann, T.
Chaurand, F. Cuenot-Chaillet, X. Hindermey-
ER, AND J. Judas. 1994. Alternate long and short
foraging trips in pelagic seabird parent. Anim. Be-
hav. 47:472-476.

Whit ACRE, D. F. 1991. Studies of the ecology of the
White-collared Swift (Streptoprocne zonaris) and
White-naped Swift (S. semicollaris), and of pat-
terns of adaptation among the swifts (Aves: Apod-
idae). Ph.D. diss., Univ. of California, Davis.

Wilson Bull., 111(1), 1999, pp. 38-45

FACTORS THAT INFLUENCE TRANSLOCATION SUCCESS IN THE

RED-COCKADED WOODPECKER

KATHLEEN E. FRANZREB'

ABSTRACT— To restore a population that had declined to 4 individuals by late 1985, 54 Red-cockaded
Woodpeckers (Picoides borealis) were translocated at the Savannah River Site in South Carolina between 1986
and 1995. Translocation success was evaluated by sex, age, and distance between the capture and release site.
For moves involving females, the presence of a resident male and the status of the male (breeder, inexperienced,
or helper) also was assessed. Of the factors I evaluated, only the distance of the move was statistically significant
with increasing success associated with increasing distance. The presence of a resident male at the female’s
release site led to no more success than releasing the female concunently with a male; nor did the male’s status
appear to play a significant role in female translocation success. Overall, 31 of 49 (excluding nestlings) trans-
located birds remained at or near the release site for at least 30 days, resulting in a success rate of 63.2%. Of
the birds that were successfully translocated, 51.0% had reproduced by July 1996. Received 2 March 1998
accepted 15 Oct. 1998.

Endemic to the open pine woodlands of the
South, Red-cockaded Woodpeckers {Picoides
borealis) are cooperative breeders whose
groups usually consist of a breeding pair and
often one or more helpers, usually male off-
spring (U.S. Fish and Wildlife Service 1985).
A series of cavity trees occupied by such a
group is referred to as a cluster. These cavities
are used year round for night roosting and as
nest sites during the breeding season (Steirly
1957). Since 1970 the species has been con-
sidered Federally endangered primarily be-
cause of widespread habitat loss, which has
fragmented the original population into many
subunits, some quite small and/or isolated
(U.S. Fish and Wildlife Service 1985). One
such small population occupies the Savannah
River Site in South Carolina.

By late 1985 the number of Red-cockaded
Woodpeckers had dwindled to one breeding
pair and two single males (DeFazio et al.
1987), and the Forest Service began intensive
management to prevent extirpation on the site
(Gaines et al. 1995). With the nearest known
Red-cockaded Woodpecker population 32 km
away, natural recruitment of and colonization
by new individuals was considered unlikely.
Because Red-cockaded Woodpeckers prefer
older, live pine trees for constructing their
cavities (Steirly 1957, Jackson et al. 1979,
Conner and O’Halloran 1987, Rudolph and
Conner 1991) and few trees of sufficient age

' Southern Re.scarch Station-USDA Forest Service,
Dept, of Forest Re.sourees, Clemson Univ., Clemson,
.SC 29634—1003; E-mail: KFRANZR@clcmson.edu

and diameter were available. Forest Service
personnel installed 305 artificial cavities (see
Allen 1991 for details on artificial cavity con-
struction and installation). Other management
activities have included: (1) restricting cavity
access by other larger woodpecker species
with metal “restrictor” plates (Carter et al.
1989), (2) removing southern flying squirrels
{Glaucomys volans) encountered while moni-
toring cavities and squirrel nest boxes, and (3)
improving habitat quality by controlling the
hardwood midstory vegetation that causes
woodpeckers to abandon their cavities (Con-
ner and Rudolph 1989, Costa and Escano
1989, Hooper et al. 1991, Loeb et al. 1992).

In an effort to stabilize and eventually in-
crease the population at the site, the Forest
Service began a program of translocating
woodpeckers from populations outside of and
within the site. The objectives were to in-
crease the number of breeding pairs, bolster
the overall population size, and minimize po-
tential adverse genetic consequences arising
from small population size (Allen et al. 1993,
Gaines et al. 1995). Here I assess the results
of 10 years of Red-cockaded Woodpecker
translocations at the Savannah River Site to
determine the variables most likely to contrib-
ute to successful translocations, an important
strategy in the recovery of small, isolated pop-
ulations.

STUDY AREA AND METHODS

Study «m^_The Savannah River Site lies within
the Upper Coastal Plain physiographic region in Ai-
ken, Allendale, and Barnwell counties in South Caro-

Fnmzreh • RED-COCKADED WOODPECKER TRANSLOCATIONS

lina. By the early 195()s, most of the site was in ag-
ricultural use or had been harvested for timber. In
1951, the Department of Energy (DOE) acquired
80,269 ha of contiguous land to develop the area as a
nuclear production facility. Under an interagency
agreement, the Savannah River Natural Resource Man-
agement and Research Institute (U.S. Department of
Agriculture, Forest Service) has managed the natural
resources on the site for DOE since 1952. Today ap-
proximately 69.000 ha on the site are in pine stands
(Workman and McLeod 1990), most of which are less
than 50 years old although there are some residual old-
er pine trees. The area managed for the woodpecker
contains 3 1,970 ha of pine forest consisting of longleaf
(Pimis paliisiris; Yl .1% of the pine acreage), loblolly
(P. taedo: 45.4%), slash {P. elliotti; 13.4%), and other
(0.2%) pines in addition to pine-hardwoods (3.3%; G.
Gaines, unpubl. data).

Field methods. — Beginning in 1980, Red-cockaded
Woodpeckers at the Savannah River Site were banded
with a U.S. Fish and Wildlife Service aluminum leg
band and with a unique color plastic leg band combi-
nation for field identification. Birds were banded as
nestlings on the site, when first captured as adults, or
just prior to release if they were from an offsite pop-
ulation.

Blood samples were taken from adults at the Savan-
nah River Site to determine the relatedness of individ-
uals of unknown heritage and level of genetic hetero-
zygosity (Stangel et al. 1992, Haig et al. 1993). The
results helped to provide the genealogical pedigree and
verification of parentage needed to determine which
individuals should be matched for mating.

Individual translocations either provided a mate for
an established breeding bird (e.g., to replace a lost
mate) or established a new pair in unoccupied territory.
If a translocated bird remained in the vicinity of the
release site for at least 30 days, I regarded the release
as successful. Preference was given to abandoned clus-
ters that were more than 1 km from other active clus-
ters to minimize interference by other Red-cockaded
Woodpeckers.

Translocated individuals and pairs were introduced
into groups with an unpaired, resident bird or into
abandoned clusters. Trapping, transportation, and re-
lease followed the methods described by DeFazio and
coworkers (1987) and Allen and coworkers (1993).
The age, group, sex, status (such as helper or breeding
female), previous breeding experience, distance from
the capture to release site, and location of capture site
were recorded for each bird translocated. Transloca-
tions involved moving an independent subadult (one
year or less in age) or adult female to a bachelor male,
moving an unpaired female and/or male to unoccupied
habitat, moving a family unit (mated pair and nest-
lings) to unoccupied habitat, and cross-fostering nest-
lings.

Observations of translocated birds lasted approxi-
mately 8-30 hrs per bird the week immediately fol-
lowing release and were repeated at least once per
week during the breeding season and once per month

during the non-breeding season to monitor each bird’s
status. If a translocated bird could not be relocated, a
thorough search was made in clusters within approxi-
mately 0.8 km. For birds captured on the site, the
search included previous roost trees even if they were
beyond 0.8 km of the release site.

Analytical methods. — How a bird responded to
translocation (e.g., stayed at release site, returned
home, disappeared), whether or not it eventually re-
produced in the vicinity of the release site, and the
number of fledglings produced was recorded for each
bird. Because the distance between the capture and re-
lease site was found to influence the results, the data
were examined separately for moves of various dis-
tances 1 km, 19-23 km, 182-483 km).

To evaluate if sex of the translocated bird affected
the outcome of a move, translocation success was
compared for all males to all females, adult males to
adult females, and subadult males to subadult females.
To determine if there was a period of time shortly after
fledging when younger subadult females were more
likely to remain at the release site, the translocation
success of subadult females 5—7 months of age was
compared to those 7—12 months old.

Distance between the capture and release site was
evaluated by examining translocation success for short
1 km), moderate (19—23 km), and long (182-483
km) distance moves. There are two subpopulations of
Red-cockaded Woodpeckers at the Savannah River
Site. Moves within either subpopulation were no more
than 7 km. The two subpopulations are separated by
about 19 km. Hence, translocations onsite between the
two subpopulations involved distances of 19-23 km.
All offsite populations were at least 182 km from the
Savannah River Site. Moves from offsite were done
for 7 of 10 years between 1986 and 1995. Capture sites
on the Savannah River Site were monitored to deter-
mine if released birds returned home. Similar moni-
toring was not undertaken at offsite populations be-
cause they were too far from the Savannah River Site
to check routinely.

To determine if the presence of a resident male af-
fected the translocation success of a female, I com-
pared responses of females who were moved to clus-
ters with a resident male (regardless of his reproduc-
tive experience) versus a “co-move" in which a male
(captured in a separate cluster) was translocated si-
multaneously with a female to a new site. To evaluate
the possible influence of distance from the capture to
release site, the translocation success for females
moved to resident males and those moved with a male
were compared with respect to distance.

The possible effect of male status (breeder, helper,
or inexperienced) on female relocation success was ex-
amined for females: ( I ) moved to a resident male. (2)
moved simultaneously with a male, and (3) for both
situations combined. An “inexperienced" male had no
known experience as a breeder or helper. Female trans-
location succe.ss with respect to male status was seg-
regated further by distance moved.

All statistical comparisons were made using Fi.sher's

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

Exact Test (Sokal and Rohlf 1995) with the level of
significance defined as P < 0.05 and executed with
SAS (version 6.12, Windows 95, IBM-compatible;
SAS Institute 1990). Fisher’s Exact Test for contingen-
cy tables was used because in most cases cell frequen-
cies were too small to support standard y- tests. Unless
otherwise noted, the results for the five translocated
nestlings are excluded from these comparisons.

RESULTS

From 1986 to 1995, 54 Red-cockaded
Woodpeckers were translocated, at first from
populations off the site, but later from onsite
as their numbers increased. Beginning in
1986, 21 birds were taken from offsite popu-
lations. 7 females, 1 male, and 5 nestlings
from the Francis Marion National Forest in
South Carolina (about 192 km away); 5 fe-
males from the Apalachicola National Forest
in Florida (483 km away); 1 female from Fort
Bragg in North Carolina (266 km away); and
2 females from the Carolina Sandhills Nation-
al Wildlife Refuge in South Carolina (182 km
away). Offsite locations were selected because
they contained relatively large numbers of
Red-cockaded Woodpeckers. Thirty-three
birds were translocated within the site begin-
ning in 1987.

Of the 24 subadult females, 6 remained at
the release site and bred, 1 died after remain-
ing more than 30 days, 6 moved to clusters
near the release site and bred with nearby
males, 1 returned to the capture site, and 10
disappeared. Of the 10, 4 remained at the re-
lease site for more than 30 days, I was chased
away by other Red-cockaded Woodpeckers,
and another reappeared five months later ap-
proximately 20 km away and became the
breeding female in that cluster. Of 9 adult fe-
males, 2 remained at the release site and bred,

5 moved to nearby clusters and bred, 1 re-
turned to her original cluster, and 1 remained
at the release site but did not breed.

Of 10 subadult males, 2 stayed at the re-
lease site and bred, 2 moved to a nearby clus-
ter and bred, 1 remained at the release site for
four months then disappeared, 4 disappeared
soon after release, and 1 returned home after
30 days. Two adult males remained where re-
leased or close by and bred, 1 disappeared in
less than two days, and 3 returned home im-
mediately.

Five nestlings were relocated. The first
three nestlings were moved with their parents

to the Savannah River Site from the Francis
Marion National Forest in 1988 and later died
from parental neglect (Allen et al. 1993). The
other two nestlings were fostered in 1987;
both successfully fledged after being placed in
a Red-cockaded Woodpecker nest. The female
disappeared after five months, and the male
became a breeder in a nearby cluster and
eventually produced two fledglings.

Ten of the 49 birds (excluding the 5 nest-
lings) that were moved consisted of pairs of
subadult females and subadult males moved
concurrently. Overall, 31 of 49 (63.2%) adults
and subadults remained at or near the release
site for at least 30 days after release and 25
(51.0%) eventually reproduced (Table 1). The
number of birds represented in the various
combinations of moves segregated by age and
cluster status are shown in Table 1.

Translocation was successful for 61.8% of
subadults (21 of 34) and 66.7% of adults (10
of 15; Table 2). There were no significant dif-
ferences in success measured either by the
number that stayed or by the number that re-
produced for adult males compared to adult
females for short, moderate, or long distance
moves (Table 2; Fisher’s Exact Tests: all P >
0.05). Nor was there a difference in success
of subadult males compared to subadult fe-
males for any of the distance classes (Fisher’s
Exact Tests: all P > 0.05).

Of 189 fledglings produced from 1986-
1996, 104 (55.0%) had at least one parent that
had been translocated. The number of fledg-
lings excludes the young produced by birds
that were translocated but did not remain in
the vicinity of the release site to breed (Table
2).

Translocation success of younger subadult
females (5-7 months of age) did not differ
significantly from those that were older (7-12
months of age; Fisher’s Exact Test: P > 0.05
for all comparisons). There were no short dis-
tance moves involving younger subadult fe-
males.

Because sex and age did not appear to in-
fluence success (Table 2), I pooled the data
and tested for a distance effect. Translocated
birds were more likely to stay with increasing
distance from their capture .site: 25.0% suc-
cess for translocations less than 7 km, 71.4%
for 19-23 km, and 81.3% for 182-483 km
moves. The distance a bird was moved had a

h'nmzreh • RED-COCKADED WOODPECKER TRANSL(3CATIONS

TABLE 1. Succes.s of translocated Red-cockaded Woodpeckers by age and type of move
River Site (1986-1995).

at the Savannah

Translocated

Number of birds

Stayed > 30 days

Reproduced

Translocated to resident male:

Adult female

Subadult female

Translocated to unoccupied cluster:

Adult breeding male

Subadult male

Adult female

Adult male/adult female-'

Adult niale/sLibadult female*’

Subadult male/subadult female'-

Adult male/adult female/nestings'* (family unit)

Fostered nestlings

Total adults and subadults

31 (63.2%)

25 (51.0%)

Total including nestlings

33 (61.1%)

26 (48.1%)

^ In one move, the male remained; in the other, only the female remained.

In one move, neither bird remained; in the other, both remained.

In three cases, neither the male nor female remained; in two ca.ses, only the male remained; in three ca.ses, only the female remained; in two cases,
both birds remained.

Only the female remained.

highly significant effect on whether the bird
remained more than 30 days (Fisher’s Exact
Test: P = 0.01), but was not significant for
birds that eventually reproduced (Fisher’s Ex-
act Test: P = 0.12; Table 3). Birds moved a
short distance were more likely to return home
[41.7% (n = 12) for short versus 4.8% (n =
21) for moderate distance moves; Fisher’s Ex-
act Test: P = 0.02J. There was no significant
difference in the success rate of a bird moved
a moderate versus a long distance (Fisher’s
Exact Test: P > 0.05 stay, P > 0.05 repro-
duce, n — 21 and 16, respectively). Nor was
there a significant difference in rate of return
for males versus females moved a short (Fish-
er’s Exact Test: P > 0.05) or moderate dis-
tance (Fisher’s Exact Test: P > 0.05).

Eighteen of 22 females (81.8%) that were
moved to resident males were successful
(stayed), whereas 5 of 10 females (50.0%)
succeeded that were moved concurrently with
a male. Of the 10 co-moves, 3 females re-
mained after the male left and 2 males stayed
even though the female departed. In two cas-
es, both male and female remained. In one of
the three instances when both members of the
co-move left, the female left first and in the
other two cases it could not be determined
which of the birds was the first to leave. Of
the six cases in which the male left, the female

remained behind in three of them. Although
moving a female to a site where a male al-
ready was established was thought to be ad-
vantageous, the success rate was not signifi-
cantly different from situations in which the
female was moved simultaneously with a
male for either moderate (Fisher’s Exact Test:
P > 0.05 for stay, P > 0.05 for reproduce, n
= 14) or long distance moves (Eisher’s Exact
Test: P > 0.05 for stay, P > 0.05 for repro-
duce, n = 14). Nor was there a significant
difference in success of females moved either
to a resident male or with a male when trans-
locations of moderate and long distances were
combined (Fisher’s Exact Test: P > 0.05 stay,
P > 0.05 reproduce, n = 28). No short dis-
tance moves of a female to a resident male
were undertaken.

Because female success was not influenced
by whether the male already was on the re-
lease site or whether he was moved simulta-
neously with her, these data were pooled. Fe-
males had a success rate of 87.5% (seven suc-
cesses in eight cases) if the male involved was
an experienced breeding male, 40.0% (/; = 5)
if he was a helper, and 73.7% (/? = 19) if the
male was inexperienced. Because there were
only four short distance moves and none of
these involved a breeder male, the effect of
male status could not be assessed for females

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

TABLE 2. Translocation success by sex and age of Red-cockaded Woodpeckers with respect to distance
moved at the Savannah River Site (1986-1995).^

Di.stance moved

No. fledglings

Sex/age class

< 7 km

19-23 km

182-483 km

Total

produced

(No./sex-age

class)

Adult females:

29 (3.2)

Number translocated

Number stayed (%)

0 (0.0%)

3 (100.0%)

5 (100.0%)

8 (88.9%)

Number reproduced (%)
Adult males:

0 (0.0%)

3 (100.0%)

4 (80.0%)

7 (77.8%)

4 (0.7)

Number translocated

Number stayed (%)

2 (50.0%)

0 (0.0%)

2 (33.3%)

Number reproduced (%)
Subadult females:

2 (50.0%)

0 (0.0%)

2 (33.3%)

58 (2.4)

Number translocated

1 1

Number stayed (%)

1 (33.3%)

7 (63.6%)

8 (80.0%)

16 (66.7%)

Number reproduced (%)
Subadult males:

1 (33.3%)

5 (45.4%)

6 (60.0%)

12 (50.0%)

20 (2.0)

Number moved

Number stayed (%)

0 (0.0%)

5 (83.3%)

5 (50.0%)

Number reproduced (%)

All females:

0 (0.0%)

4 (66.7%)

4 (40.0%)

87 (2.6)

Number translocated

Number stayed {%)

1 (25.0%)

10 (71.4%)

13 (86.7%)

24 (72.7%)

Number reproduced (%)

All males:

1 (25.0%)

8 (57.1%)

10 (66.7%)

19 (57.6%)

24 (1.5)

Number translocated

Number .stayed (%)

2 (25.0%)

5 (71.4%)

0 (0.0%)

7 (43.8%)

Number reproduced (%)
Nestlings:

2 (25.0%)

4 (57.1%)

0 (0.0%)

6 (37.5%)

2 (0.4)

Number moved

Number stayed (%)

2 (40.0%)

Number reproduced (%)
Total includes nestlings

Total excludes nestlings

1 (20.0%)

1 (20.0%)
54

104'^

102“^

“ None of the results from Fisher’s Exact Tests was significant at A* < 0.05.

h — = not applicable; no tests of this type were made.

fledJtgTprSucedl" translocated; total figure includes

moved a short distance. For all moves, there
was no significant difference in female suc-
cess when comparing breeder, helper, or in-
experienced males (Fisher’s Exact Test: all P
> 0.05).

DISCUSSION

The first reported Red-cockaded Wood-
pecker translocations involved a 1981 relo-
cation of 12 birds from 5 groups at the Fort
Stewart Army Base to St. Catherines Island,
both in Georgia (Odum et al. 1982). Five of
these birds survived at least eight months and
two produced one fledgling in 1981. In 1984
and 1986, two pairs and one single male were
moved from private land to the St. Marks Na-

tional Wildlife Refuge and adjacent Ochlock-
onee River State Park in Florida in an attempt
to enhance the three active groups at the re-
lease site (Reinman 1995). One female re-
mained and nested successfully for four con-
secutive years, one male returned to the cap-
ture site, one male died, and the fate of the
other birds is unknown. Other translocations
have been conducted to establish a group at a
site occupied by a single bird (Allen et al.
1993) and to establish new groups (Rudolph
et al. 1992, Allen et al. 1993).

Working with data collated from 143 Red-
cockaded Woodpecker translocations under a
wide range of circumstances, Costa and Ken-
nedy (1994) found various definitions of

Fnmzreh • RED-COCKADED WOODPECKER TRANSLOCATIONS

TABLE 3. Effect of distance between

capture and release

site on number of successful

translocations of

Red-cockaded Woodpeckers at the Savannah River Site (1986-

1995).

Distance moved

Number of birds < 7 km

19-23 km

182-483 km

Total

Moved 1 2

49“

Stayed (%) 3 (25.0%)

15 (71.4%)

13 (81.3%)

31 (63.3%)

Reproduced (%) 3 (25.0%)

12 (57.1%)

10 (62.5%)

25 (51.0%)

Returned home (%) 5 (41.7%)

1 (4.8%)

6 (7.5%)

Fisher's exact test: P value

Distance moved

Stayed

Reproduce

Returned home

All distances

0.01

0.12

Short vs moderate distance

0.01

0.15

Moderate vs long distance

Short vs moderate distance

0.70

1.00

0.02

“ Excludes nestlings.
^ — = not available.

translocation success ranging from “interacted
well” to “fledged young.” They noted suc-
cessful moves 66% of the time for subadult
females {n = 44) and 58% of the time for
adult females (n = 33). My study showed an
overall female success rate of 67% for sub-
adults and 89% for adults. However, Costa’s
and Kennedy’s results are difficult to compare
to mine because they contain a variety of cri-
teria for translocation success. Moreover, in
my study there was no significant difference
in the success rate based on age (subadult ver-
sus adult) for either females or males when
considering the distance of the move. Some of
my comparisons involve small sample sizes
and it is possible that a larger data set may
have revealed some significant differences.
Additional work is needed to explore more
fully any possible differences in success rate
based on age of the bird.

My study showed a greater tendency for
birds being moved a moderate (19-23 km) or
long (182-483 km) distance to remain at the
release site and reproduce than birds that were
moved short distances (< 7 km). Because
there were no moves between 7-19 km, it is
not known at what distance the success would
equal that of moves more than 19 km. There-
fore, at the present time, it is recommended
that translocations involve distances of at least
7 km (preferably more) between the capture
and release sites to discourage homing by the
birds.

DeFazio and coworkers (1987), Hess and
Costa (1995), and Reinman (1995) suggest

that the most successful translocations of fe-
males are those in which the release site con-
tains established single males — a finding sup-
ported by earlier translocations of 16 females
at the Savannah River Site (Allen et al. 1993).
The success rate for translocations to areas
that contained single established males was
63.2% for Costa and Kennedy (1994) and
81.0% for my study. However, I found that
when the release site contained a resident
male, female success was no greater than
when a female was moved concurrently with
a male for moderate and long distances.

Costa and Kennedy (1994) recommend us-
ing a two level standardized definition of suc-
cess. One level reflects primary evidence of
breeding (e.g., copulation, etc.) and the other
that the bird has become attached to the site
(e.g., roosting in a cluster, etc.). For any trans-
location effort to succeed, the first major hur-
dle is for the bird to remain at the release site.
In my study, the presence of a translocated
bird at the release site after 30 days was con-
sidered evidence that the bird had accepted the
site and was likely to breed once a suitable
mate became available. Because disease and
predation may prevent some of these birds
from surviving long enough to reproduce, the
use of breeding as the criterion of transloca-
tion success may be overly conservative. If
producing at least one fledgling is used to
measure translocation success, then 51.0% of
the translocated birds in this study were suc-
cessful. The success rate was 63.2% if defined

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

as the number of birds remaining near the re-
lease site for at least 30 days.

Translocations at the Savannah River Site
have played an instrumental role in restoring
the Red-cockaded Woodpecker population
from 4 individuals in 1985 to 99 individuals
(56 adults and 43 young-of-the-year) and 19
breeding pairs in 1996 (Franzreb 1997). Clear-
ly, the use of translocations as a management
tool has been an integral part in the recovery
of this nearly extirpated population.

ACKNOWLEDGMENTS

Thi.s research was funded by the Department of En-
ergy (DOE), Savannah River Site, and its cooperation
is gratefully acknowledged. G. Gaines, J. Blake, and
R. Hooper provided important discussion and com-
ments. I thank R. Conner, J. Reinman, T. Engstrom,
M. Reed, R Doerr, B. Wigley and R. Hooper for their
insightful reviews. In addition, P. Jackson (DOE) and
Savannah River Natural Resource Management and
Research Institute staff (especially J. Irwin, J. Blake,
E. LeMaster, and W. Jarvis) provided support through-
out the course of this work. M. Lennartz deserves spe-
cial credit for involvement through 1990. I am grateful
to the numerous hard-working research field support
staff, C. Dachelet, D. Allen, K. Laves, J. Edwards, P.
Johnston, D. Ussery, and K. Shinn, as well as wildlife
biologists, foresters, and technicians at the donor pop-
ulation forests for their outstanding efforts on behalf
of this project. And I thank statistician W. Pepper who
was instrumental in providing statistical data analysis.

LITERATURE CITED

Allen, D. H. 1991. An insert technique for construct-
ing artificial Red-cockaded Woodpecker cavities.
U.S.D.A. Southeast. For. Res. Sta. Gen. Tech.
Rep. SE-73;1-19.

Allen, D. H., K. E. Franzreb, and R. E. E Escano.
1993. Efficacy of translocation strategies for Red-
cockaded Woodpeckers. Wildl. Soc. Bull. 21:155-
159.

Carter, J. H., Ill, J. R. Walters, S. H. Everhart,
AND P. D. Doerr. 1989. Restrictors for Red-cock-
aded Woodpecker cavities. Wildl. Soc. Bull. 17-
68-72.

Conner, R. N. 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.

Conner. R. N. and D. C. Rudolph. 1989. Red-cock-
aded Woodpecker colony status and trends on the
Angelina, Davy Crockett, and Sabine National
Forests. U.S.D.A. .South. Res. Sta. Res. Pap. SO-
250:1-15.

Co.STA, R. AND R. E. E E.SCANO. 1989. Red-cockaded
Woodpecker: status and management in the south-

ern region in 1986. U.S.D.A. For. Serv. Tech. Pub.
R8-TP 12:1-71.

Costa, R. and E. Kennedy. 1994. Red-cockaded
Woodpecker translocations 1989-1994: state-of-
our-knowledge. Pp. 74-81 in Proc. Annu. Conf.
Am. Zoo Aquarium Assoc., Atlanta, Georgia.

DeFazio, j. T, Jr., M. A. Hunnicutt, M. R. Lennartz,
G. L. Chapman, and J. A. Jackson. 1987. Red-
cockaded Woodpecker translocation experiments
in South Carolina. Proc. Annu. Conf. Southeast.
Assoc. Fish Wildl. Agen. 41:311-317.

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 management (D. L. Kulhavy, R. G.
Hooper, and R. Costa, Eds.). Center for Applied
Studies, College of Forestry, Stephen E Austin
State Univ., Nacogdoches, Texas.

Franzreb, K. E. 1997. Success of intensive manage-
ment of a critically imperiled population of Red-
cockaded Woodpeckers in South Carolina. J. Field
Omithol. 68:458-470.

Haig, S. M., J. R. Belthoff, and D. H. Allen. 1993.
Examination of population structure in Red-cock-
aded Woodpeckers using DNA profiles. Evolution
47:185-194.

Hess, C. A. and R. Costa. 1995. Augmentation from
the Apalachicola National Forest: the develop-
ment of a new management technique. Pp. 385-
388 in Red-cockaded Woodpecker: recovery,
ecology and management (D. L. Kulhavy, R. G.
Hooper, and R. Costa, Eds.). Center for Applied
Studies, College of Forestry, Stephen F. Austin
State Univ., Nacogdoches, Texas.

Hooper, R. G., D. L. Krusac, and D. L. Carlson.
1991. An increase in a population of Red-cock-
aded Woodpeckers. Wildl. Soc. Bull. 19:277-286.

Jackson, J. A., M. R. Lennartz, and R. G. Hooper.
1979. Tree age and cavity initiation by Red-cock-
aded Woodpeckers. J. For. 77:102-103.

Loeb, S. C., W. D. Pepper, and A. T. Doyle. 1992.
Habitat characteristics of active and abandoned
Red-cockaded Woodpecker colonies. South. J.
Appl. For. 16:120-125.

Odum, R. R., J, Rappole, J. Evans, D. Charbonneau,
AND D. Palmer. 1982. Red-cockaded Woodpecker
relocation experiment in coastal Georgia. Wildl.
Soc. Bull. 10:197-203.

Reinman, J. P 1995. Status and management of Red-
cockaded Woodpeckers on St. Marks National
Wildlife Refuge 1980-1992. Pp. 106-111 in Red-
cockaded Woodpecker: recovery, ecology and
management (D. L. Kulhavy, R. G. Hooper, and
R. Co.sta, Eds.). Center for Applied Studies, Col-
lege of Forestry, Stephen F. Austin State Univ.,
Nacogdoches, Texas.

Rudolph, D. C. and R. N. Conner. 1991. Cavity tree
■selection by Red-cockaded Woodpeckers in rela-
tion to tree age. Wilson Bull. 103:458-467.

Franzreh • RED-COCKADED WOODPECKER TRANSLOCATIONS

Rudolph, D. C., R. N. Conner, D. K. Carrie, and R.
R. Schaefer. 1992. Experimental reintroduction
of Red-cockaded Woodpeckers. Auk 109:914-
916.

SAS Institute Inc. 1990. SAS procedures guide, ver-
sion 6. Third ed. SAS Institute, Cary, North Car-
olina.

SoKAL, R. R. AND E J. Rohlf. 1995. Biometry. Third ed.,
W. H. Freeman and Co., San Francisco, California.
Stangel, P. W., M. R. Lennartz, and M. H. Smith.
1992. Genetic variation and population structure

of Red-cockaded Woodpeckers. Conserv. Biol. 6:
283-292.

Steirly, C. C. 1957. Nesting ecology of the Red-cock-
aded Woodpecker in Virginia. Atl. Nat. 12:280-292.

U.S. Fish and Wildlife Service. 1985. Recovery plan
for the Red-cockaded Woodpecker. Region 4, At-
lanta, Georgia.

Workman, S. W. and K. W. McLeod. 1990. Vegeta-
tion of the Savannah River Site: major community
types. U.S. Dept, of Energy, Savannah River Site,
SRO-NERP- 19: 1-137.

Wilson Bull., 1 1 1(1). 1999, pp. 46-55

BANDING RETURNS, ARRIVAL PATTERN, AND SITE-FIDELITY

OF WHITE-EYED VIREOS

S. L. HOPP,'^ A. KIRBY, 2 and C. A. BOONE^

ABSTRACT. — We present nine years of return data for individually color-banded White-eyed Vireos and
describe patterns of arrival and territory use. Of all opportunities for annual return. 48.3% of males and 50% of
females were resighted. Most males arrived between 17 and 30 April, with a median arrival date of 24 April,
while most females arrived between 21 April and 1 May with a median arrival date of 26 April; males arrived
significantly earlier than females. Older males arrived significantly earlier than younger, as has been reported
for several other species. The arrival dates for individual males were consistent across years; an individual's
arrival date in one year reliably predicted its arrival date in the next year. Thus, the timing of arrival co-varied
with three factors: sex, age, and individual. Nearly all males remained faithful to previous territories, although
some shifted so that the new territory overlapped the old. Aspects of our data and those of others suggest our
return rates are likely a low estimate of survivorship for the species; the actual survival rate is probably higher.
Received 3 May 1998, accepted 23 Sept. 1998.

Recent studies have strongly indicated that
populations of many species of migratory
birds in North America are declining. This
recognition has invigorated research efforts
aimed at documenting various aspects of pop-
ulation dynamics among different species and
determining the factors underlying these de-
clines (Terborgh 1980, Lovejoy and Oren
1981, Hagan and Johnston 1992, DeGraaf and
Rappole 1995, Rappole 1995, Sauer et al.
1996). These studies are useful in establishing
current population status for various species,
for monitoring relationships between popula-
tion status and ecological and demographic
factors, and for identifying future research and
conservation goals. Several approaches have
been employed, including point counts of
breeding populations, mist-netting of birds at
migration stopovers, breeding bird surveys
(BBS), and studies of banded populations of
birds on both wintering and breeding grounds
(Askins et al. 1990, Payne and Payne 1990,
Bibby et al. 1992).

Among the groups of birds in decline are
migratory species in the family Vireonidae
(Robbins et al. 1989). Two species of vireo
are federally endangered in the United States:
the Black-capped Vireo (Vireo atricapillm)
found in Texas and Oklahoma (Grzybowski et

' Dept, of Ecology and Evolutionary Biology, Univ.
of Arizona. Tuc.son. AZ 85721.

^ P.O. Box 193, Independence, OR 97351.

' Mu.seum of the Middle Appalachians, 123 Palmer
Avc., .Saltville, VA, 24370.

^ Correspt)nding author; E-mtiil: SHOPP@u.tirizona.cdu

al. 1986, USFWS 1991), and a race of Bell’s
Vireo in California, the Least Bell’s Vireo (V.
bellii pusillus; USFWS 1986, Franzreb 1989).
Several other species of vireos have declined
in numbers, as determined by the USGS
Breeding Bird Survey (BBS; Robbins et al.
1989, Sauer and Droege 1992). A number of
studies have addressed different aspects of
population dynamics in various vireo species
including migration patterns (Remsen et al.
1996, Woodrey and Chandler 1997, Woodrey
and Moore 1997), population structure (Grzy-
bowski 1991), aspects of breeding (Graber et
al. 1985, Grzybowski et al. 1994, Marvil and
Cruz 1989, Barber and Martin 1997), and win-
tering ecology (Greenberg 1992; Greenberg et
al. 1993, 1995). Studies of these types are im-
portant for gauging long-term changes in mi-
gratory vireo populations and the factors that
affect them (Holmes et al. 1989; Lynch 1989
1992).

One of these Nearctic migrant species is the
White-eyed Vireo (Vireo griseus), a small
passerine that occupies secondary deciduous
habitat, thickets, and forest-edge in the eastern
United States. Its winter range extends across
the southern US from Texas to South Caroli-
na, south through the West Indies, and along
the eastern coast of Mexico (Barlow 1980,
Hopp et al. 1995). The northern subspecies,
V. g. noveboracemis, is fully migratory. The
southeastern subspecies, V. g. griseus, has
been reported to migrate (Barlow 1980). How-
ever, Bradley (1981) reported that a popula-
tion near Gainesville, Florida was sedentary,

Hopp el ill. • VIREO RETURNS

with individuals remaining through the winter.
The remaining subspecies, V. g. micrus, V. g.
bermudianus, V. g. maynardi, and V. g. per-
quisitor, are apparently nonmigratory (Barlow
1980, Hopp et al. 1995). Analyses of the data
from the BBS suggest that White-eyed Vireos
have been declining in parts of their breeding
range (Robbins et al. 1989, Sauer and Droege
1992). Efforts to document these declines and
determine factors that influence population
structure and territory use are important for
monitoring long-term changes in vireo popu-
lations.

In the present study we report on returns of
banded White-eyed Vireos for nine years in
southwestern Virginia. Banded White-eyed
Vireos have been reported to show site fidelity
in their breeding range (Hopp et al. 1995) and
on winter territories (Rappole and Warner
1980). We used this fidelity to territories to
directly measure return rates of banded indi-
viduals. In addition we report on the pattern
of arrival in the spring and outline the use of
territories by individuals during the breeding
season.

METHODS

Most birds were banded on two study sites in Wash-
ington and Smyth counties in southwestern Virginia.
The two sites, each about 35 ha and located 600-800
m above sea level comprised tracts of secondary de-
ciduous growth, typically favored by this species (Con-
ner et al. 1983, Graber et al. 1985, Hopp et al. 1995).
Both areas contained 12-16 contiguous White-eyed
Vireo territories. Other birds were banded in the same
counties at smaller tracts of 3—12 ha, each with 1-6
territories; 2 of these smaller areas were within 5 km
of the main study area in Washington County. For all
of these study areas the habitat type was mixed, with
approximately 60-80% of the areas constituting suit-
able habitat for the vireos. We attempted to locate and
band birds at all smaller appropriate tracts within about
15 km of the site in Washington County. The principal
habitat types in this region are deciduous forest or
open pastureland; the presence of habitat appropriate
for White-eyed Vireos is limited and typically exists
in small areas supporting only a few individuals. All
but one male in our study shared at least one territorial
boundary with another male; most birds shared bound-
aries with several other males.

We captured most birds soon after their arrival in
late April or early May. Males were easily taken in
mist nets as they approached tape playbacks of con-
specific song. Our attempts to capture females by in-
tercepting approaches to nests clearly disrupted nesting
activities, so we discontinued those attempts. On a few
occasions females were captured with males, if they

followed them into nets during playbacks. All birds
were banded under permit with a unique configuration
of USFWS aluminum and plastic color bands.

We began checking For arrivals in known territories
and neighboring suitable habitat in the first week ol'
April. Monitoring consisted of listening for singing
males in known territories for at least 15 minutes; often
we used recorded song to attract males. When detect-
ed, males were followed to determine whether they
were banded, and to identify the color(s) and config-
uration for banded individuals. Study areas were
checked daily until mid-May; unoccupied territories
were then checked at least twice weekly until mid-
June. We also checked several other known popula-
tions of birds within 15 km of the study areas to po-
tentially detect dispersed birds. We were able to as.sess
returns without recapture by identifying the color-band
configurations. Because the males of this species are
vocally prolific, it was easy to locate newly arrived
males. The arrival dates used in this analysis are re-
stricted to subjects for which we spent at least 1 5 min-
utes in the area on the day prior to their first detection,
i.e., we were confident they were not present on the
previous day. Determining arrival dates for females
was more difficult because they are behaviorally cryp-
tic, and because they sometimes changed locations af-
ter a day or two. Most reliably, finding males allowed
us to locate females. Whenever a male was located,
we observed him long enough to determine whether a
female was also present. In most cases, changes in the
males’ behavior was indicative of pairing status; un-
paired males usually sang at high rates from high, ex-
posed perches, while paired males typically spent more
time in lower areas and sang at lower rates (Hopp et
al. 1995). Because newly paired males stay close to
females it was relatively easy to assess pairing status.
Because females were mostly unbanded, determining
whether a particular female was a new arrival or had
moved from a neighboring male’s territory was not
possible. Arrival dates for females could be unambig-
uously determined by assessing the total number of
females on the entire study site, with changes between
successive days indicating new arrivals. Arrival dates
for females used in this analysis are only those we
could definitively determine to be new arrivals.

We considered an individual’s tenitory to be the to-
tal area the bird was observed to occupy throughout
the course of a season without outside influence either
by other birds or by the researchers. Site fidelity was
then defined as use of an area that overlapped a terri-
tory from the previous year by at least 50%. Approx-
imately two-thirds of the males also served as focal
subjects for studies of vocal communication, and hence
provided more complete data on territory use.

We also examined the encounter records for White-
eyed Vireos from the Bird Banding Laboratory (BBL)
through 1996. These records include encounters within
the same 10' block (prior to 1958) and document in-
dividuals encountered outside the original 10' block
where they were banded. These records potentially
provide information about the dispersal of both hatch-

THE WILSON BULLETIN

Vol. Ill, No. I. March 1999

TABLE 1. Returns of male White-eyed Vireos
known to be at least one year old when banded.

Banding

year

Number

banded

Returns i

in years following banding

One

Two

Three

Four

Five

Six

1985

1986

1987

1 1

1988

1 I

—

1989

—

1990

—

1991

—

Total

Percent

^ Calculations based on percent returns relative to banding year.

ing year (HY) and adult (after hatching year; AHY)
birds. To assure that we were assessing dispersal rather
than migratory movements, we considered only birds
with both banding and encounter dates between 16
May and 16 September inclusive. These dates are well
within the average spring arrival and fall departure
dates for this species throughout most of its range (see
appendix 1 in Hopp et al. 1995).

RESULTS

Returns of banded birds. — During the
breeding seasons of 1985-1991, we banded
74 adult male birds used in this analysis. We
also banded 5 females and 42 nestlings/fledg-
lings, most of which were known to have sur-
vived until they were capable of sustained
flight. Returns of males for each year from
1986—1994 individually and combined are
given in Table 1. The percentage returns are
based on combined data from nine years,
computing the number that return relative to
the banding year. Because these subjects were
all banded as adults (AHY), these percentages
give an indication of survivorship in the suc-
cessive years following banding. The two in-
dividuals surviving five seasons after banding
were thus at least six years old. Our oldest
bird (not included in this analysis) arrived as
an AHY bird in 1990 and was still alive in
the summer of 1998, making him at least 9
years old.

Calculating the probability of return given
that a bird was alive in the previous year, 40
of 74 (54.1%) returned the year following
banding, 19 of 40 (47.5%) returned the second
year, 5 of 19 (26.3%) returned the third year,
3 of 5 (60.0%) returned the fourth, 2 of 3
(66.7%) returned the fifth year, and none of 2

10 20 30 10 20 30 9

April May June

Date

FIG. 1. Box plot of arrival dates for all females
and males. Males are further divided into various age
clas.ses (see text for explanations). For each plot the
total box encompasses the second and third quartiles
for each distribution, with the central vertical line
showing the median arrival date. Horizontal lines show
the 1 0—90% range of dates, and triangles indicate in-
dividual arrival dates in the outlying 10% ranges.

returned the sixth year following banding.
Over the nine year period, there were 143 op-
portunities for return and 69 documented re-
turns. Thus the percentage of male White-
eyed Vireos returning in any year, given the
bird was known to be alive the previous year,
was 48.3%. Of the five females banded, three
returned in the year following banding and
one individual returned in the second and third
years following banding. Of the ten opportu-
nities for female returns, five (50.0%) were
resighted. We did not recover any of the birds
banded as nestlings or fledglings.

Male arrival pattern. — We obtained 90 ar-
rival dates from 63 individuals. The combined
dates ranged from 12 April to 11 June with a
median arrival date of 23 April (Fig. 1). Most
birds arrived during the two weeks between
16 April and 30 April (75 of 90: 83.3 %).
There were six arrival dates prior to 1 6 April,
with three of these from the same banded in-
dividual in successive years. Eight arrival
dates were after 30 April, five ol these in early
May (2 May, and two each on 5 May and 6
May) and three were well outside the distri-

Hopp et al. • VIREO RETURNS

bution of other arrivals — 28, 29 May and 1 1
June. These three extremely late unhanded in-
dividuals likely represent relocations rather
than arrival dates (see below).

Female arrival pattern. — We obtained 37
arrival dates for females. Of these only 4 were
from returning, banded individuals, the re-
mainder from unbanded individuals. The com-
bined dates ranged from 17 April to 9 May,
with a median arrival date of 26 April (Fig.
1). Most of the arrival dates fell in the ten day
period between 21 April and 1 May (28 of 37:
75.7%). Two of the four earliest arrival dates
(17 and 20 April) were from returning banded
individuals. Because females are cryptic in
their plumage and behavior and because we
assessed their arrival primarily by pairing with
males, our sampling was likely biased toward
the earlier arrival dates. Despite this, the me-
dian arrival date for females was later than all
of the individual male arrival date categories
(Mann- Whitney Test: U = 3.43, P < 0.001;
see Fig. 1). We typically observed several
males on territory before any females arrived.
In several years, most male territories were
occupied before any females were seen.

Age-related male returns. — We recorded 49
arrival dates from returning banded males and
41 from unbanded males that were later band-
ed. Of the unbanded birds, 21 were of un-
known age. The remaining 20 arrival dates
were from unbanded birds that occupied ter-
ritories previously occupied by non-returning
banded individuals. Given the high site fidel-
ity for returning individuals, this second cat-
egory of unbanded replacement birds possibly
were second-year adults (SY; first breeding
season) which were treated separately in our
analysis of age-related returns.

Figure 1 shows the distributions of arrival
dates for various age-related classes of males.
Because the sample sizes in the three oldest
age categories were too small to permit statis-
tical comparison across all age categories, we
combined all dates of birds returning after
three years into one category (four years or
older, n = 8). To avoid pseudoreplication, in
analysis we averaged dates for individuals that
contributed more than one arrival date to this
combined category, yielding an effective sam-
ple size of 5. A comparison of the four male
groups, previously unbanded birds, first re-
turns, second returns, and third + returns, was

significant (Kruskal-Wallis Test: H = 1 1.27, P
< 0.05). In post hoc (Bonferroni) pairwise
comparisons, the only significant difference
was between the new (unbanded) arrivals and
the first return year (Z = 2.77, P < 0.05),
indicating that the gain in arrival is between
the first (banding) year and the first return
year. In a more conservative version of this
comparison, we compared all returning, band-
ed birds to all unbanded birds. This compar-
ison was also significant, showing arrival
dates of all banded birds to be earlier than
arrival dates of all unbanded birds (Mann-
Whitney Test: U = 3.66, P < 0.001). In a
more direct test, we compared successive
dates from individuals for whom we obtained
arrival dates in two consecutive years (n = 32
pairs). This comparison was only marginally
significant when using all consecutive arrival
pairs (Wilcoxon Test: Z = 1.465, P = 0.074).
When this analysis was restricted to consec-
utive-year pairs starting with the first (un-
banded) year, the comparison was significant
(Wilcoxon Test: Z = 1.80, P < 0.05, n — 14),
showing that birds arrived earlier in the sec-
ond of these two consecutive years, and re-
inforcing the finding that the gain in arrival
dates is apparent only between the banding
and subsequent year. For the 32 pairs of con-
secutive year arrival dates, we also found that
the arrival date of individuals was a significant
predictor of its arrival date in the subsequent
year (r = +0.616, P < 0.001). Thus, while a
portion of the variability seen in arrival dates
can be attributed to age, a substantial portion
can be attributed to individual-specific differ-
ences in arrival, with early and late arrivers
remaining early and late arrivers respectively
across seasons (Fig. 2).

Site fidelity. — Of the returning banded
males, 67 of 69, or 97% of returns were to
their previous territory. The two individuals
observed to move to non-overlapping territo-
ries between years both remained within 800
m of their original territory. Several birds en-
larged their territories in subsequent years, oc-
cupying areas that included their previous ter-
ritory. Seven individuals disappeared during
the course of a season; three returning birds
and four birds in the year they were banded.
None of these seven birds was seen in sub-
sequent years. One of the three returning birds
lost his territory to human habitat clearing

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

Individual Return Dates in Successive Years

FIG. 2. Scatter plot showing return dates obtained
from individual males in successive years. Line depicts
best-fitting regression line; squares denote 2 overlap-
ping data points.

during the winter. He was seen in adjacent and
nearby territories for only five days following
his arrival. Two of the four birds that disap-
peared during the year they were banded re-
appeared later in the season. One of these ar-
rived on 22 April in a territory previously oc-
cupied by a non-retuming banded individual.
He was banded on 24 April, disappeared on 2
May, then reappeared on the same territory 7
June where he remained for the season.

For females, all returns were to the same
study area where they were first banded, but
none to the same territory; none paired with
the same male in more than one year. The
nearest movement was to an adjacent territory.
The farthest moved three territories away, i.e.,
two intervening territories from the previous
year, a distance of about 750 meters.

Territory use. — Territory use in this species
is a typical “type A” territory (Nice 1941),
with males aggressively defending territories
against other males, and with the area used
for mating, nesting, and feeding for both
adults and young during the breeding season.
The defense of territory is usually by a series
of behavioral displays, including singing, and
less frequently by direct contact (see descrip-
tion by Bradley 1980). Most territorial en-

counters occurred early in the season, prior to
the arrival and pairing with females; following
the onset of nesting activities these territorial
encounters were rare.

Individual males could easily be tracked
during the season, and adults remained on ter-
ritory into September; the latest date we re-
corded a bird on territory in autumn was 6
October. During the last few weeks in the fall
the adults increased their singing rate and
again became quite responsive to tape-record-
ed song.

Female use of territory was less easy to ob-
serve. On several occasions we observed two
females within the same territory, always ear-
ly in a season. In no case did we observe in-
teractions between females indicating territory
defense. On one occasion we observed two
females within a few centimeters of each other
with no overt behavioral response.

As with females, HY birds were difficult to
track carefully during a season. Typically,
banded young could be found in their natal
territories for 4-5 weeks; often birds from the
same brood were found together. After this
time, banded HY birds began to disperse from
their natal territories mid-August through ear-
ly September, often being found in other ter-
ritories within the study site. A few HY birds
would defend small territories, responding ag-
gressively to tape-recordings with approach
and/or singing. Most HY birds left the region
on migration prior to the onset of banded
AHY bird departures.

USFWS encounter data. — There were 81
total White-eyed Vireo encounter records
from the Bird Banding Laboratory. Of these,
22 were both banded and encountered within
the 16 May to 16 September period, all of
these encountered during the breeding season
in the first year after banding. Of these 22, 18
were AHY birds, three were HY, and one was
local status (incapable of sustained flight; pre-
sumably nestling). Of the 18 adults, 14 were
encountered in the same 10' block in which
they were banded. Of the four adults encoun-
tered outside their initial banding 10' block,
three were encountered in the next 10' block
of latitude (approximately 19 km on average)
and one was encountered in the adjacent 10'
block of longitude (approximately 15 km). Of
the three HY birds, one was encountered in
the original block, and the other two were en-

Hopp et ill. • VIREO RETURNS

countered one 10' block of latitude and one
10' block of longitude respectively from
where originally banded. The local-status bird
was encountered in the same 10' block where
it was initially banded. Two other records de-
serve mention. One HY bird was banded on
17 August and encountered approximately
155 km away in the following year on 14
May. The other, a local bird, was banded on
12 June and encountered the following 13
May, approximately 190 km from initial band-
ing location. As both of these birds were
banded and encountered in states near the
northern edge of the species’ range (Massa-
chusetts and Maryland respectively), these en-
counters in the middle part of May likely rep-
resent an encounter on their breeding ground
rather than a point in their migration paths.

DISCUSSION

Returns of banded birds. — Our overall per-
centage returns of 48.3% for males and 50%
for females gives an indication of the baseline
returns expected for this species. These can be
compared to return rates reported for Black-
capped and Bell’s vireos. In a long-term study
of Black-capped Vireos, Grzybowski (1991)
reported male returns of 65% for a study pop-
ulation of approximately 250 birds. Perhaps
more directly comparable were his return rates
for approximately 130 males from scattered
study locations. For these he reported a return
rate of 58%, slightly higher than in our study.
For females, he reported 52% and 41% en-
countered for larger and more scattered study
groups, respectively. Similarly, for Bell’s Vir-
eos, Greaves and Labinger (1998) reported
male returns of 62.5% and 58.6% for two sep-
arate study areas, and 57.1% and 58.8% for
females from the same two study groups. Two
notable differences between the studies of
both these species and our study are the sam-
ple sizes and the study areas. Grzybowski
(1991) reported his percentage returns on
samples of approximately 250 and 130 males
in the main and scattered study areas. The
study by Greaves and Labinger (1998) re-
ported on returns from a smaller number ot
males: approximately 40 and 30. The study
areas, however, were considerably larger than
in our study. Many researchers have noted th.it
the percentages of birds re-encountered in
banding studies is directly related to the size

of the study population, with larger popula-
tions affording a higher re-encounter percent-
age, and that site fidelity is greater in larger
study groups (e.g.. Temple and Cary 1988,
Payne and Payne 1990, Grzybowski 1995).
Applied to this study, then, our reported return
rates likely represent a low return estimate for
the species. The data presented here provide
a baseline rate of returns for adult White-eyed
Vireos. Because the population densities in
this region are relatively low (Price et al.
1995), comparative studies in other areas of
this species’ range are needed to determine
which aspects of the data observed here are
shared in other regions, and what factors
might affect return rates and survival of
White-eyed Vireos.

Arrival patterns. — Our finding that older
birds arrive sooner than younger birds in
spring is consistent with similar reports in oth-
er species (e.g., Nolan 1978, Bedard and
LaPointe 1984, Hill 1989, Morton and Der-
rickson 1990). An earlier arrival could provide
either a longer potential breeding season or
better chances of obtaining a mate (see Mpller
1990). However, the variation we saw was
modest; the only significant difference was
approximately 2 days gained between the first
(banding) year and the first return year, a pat-
tern seen with both the between-subject and
within-subject comparisons. It’s difficult to ar-
gue strongly for a significant pairing advan-
tage, particularly because most males arrive
before most females. We have found that the
arrival date itself is not a significant predictor
of either pairing date or probability of obtain-
ing a mate. Rather, the age of the individual
is more likely a factor, i.e., females are more
likely to pair with older (returning) males, re-
gardless of aiTival date (SLH, unpubl. data).
However, the relation between arrival date and
reproductive success in White-eyed Vireos is
unknown.

Several studies have identified reasons for
earlier arrival dates by older birds, and several
of these might pertain to White-eyed Vireos
(see Ketterson and Nolan 1983, Woodrey and
Chandler 1997 for summaries). First, older
birds might winter further north than younger
birds. However, the site fidelity of White-eyed
Vireos to winter territories (Rappole and
Warner 1980) argues against this. Second, old-
er birds might leave earlier for northern mi-

52 THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

gration in the spring. Third, older birds might
be better at finding their territory once they
have arrived on the breeding grounds. Finally,
it is possible that older birds travel faster dur-
ing spring migration, either through superior
navigational skills or more efficient foraging
while enroute.

Our finding that the arrival date for an in-
dividual in a given year predicted its arrival
in the subsequent year is a pattern that to our
knowledge has not been previously reported
for any species. The variation seen as a result
of individual differences accounts for more of
the arrival date variability than that seen for
age related return dates. The factors outlined
above for age related returns may also be used
to explain this result, but with individual rath-
er than age related differences in these abili-
ties being applied to individuals initiating
spring migration from a common wintering
location. Alternatively, this pattern could re-
sult from individuals commencing from a
wide range of geographic origins, with more
northerly-wintering individuals arriving earli-
er. The relationship between the winter and
breeding locations of individuals is unknown.

Philopatry and territory use. — Male White-
eyed Vireos show a high degree of territory
fidelity. Some researchers have distinguished
between site fidelity, where birds return to the
study area but to a different territory, and ter-
ritory fidelity, where birds return to their pre-
vious territory (Greenwood 1980, Greenwood
and Harvey 1982, Holmes and Sherry 1992).
The extent to which birds move between ter-
ritories in consecutive years has been called
breeding dispersal by Greenwood (1980). In
our study, the attachment to particular terri-
tories is remarkably high and the two tenden-
cies, site and territory fidelity, appear essen-
tially the same. While perhaps uncommon,
three lines of evidence may indicate that
breeding dispersal movements occur in White-
eyed Vireos. First, on two occasions we ob-
■served individuals disappear only to reappear
in the same territory later within the same sea-
son. While we did not locate these birds on
other study areas, they obviously relocated for
at least a portion of that season. Grzybowski
(1995) reported that male Black-capped Vir-
eos sometimes sequentially occupied two non-
contiguous territories, and this might also oc-
cur in White-eyed Vireos. Second, records

from the BBL show four relocations of adult
White-eyed Vireos between breeding seasons,
with distances of between 15 and 20 km for
each. While these records are limited, they
nevertheless show that adult White-eyed Vir-
eos are known to relocate between breeding
seasons. Third, we observed three arrival
dates of individual males in late May and ear-
ly June, well outside of the migration dates
known for this species (Hopp et al. 1995,
Remsen et al. 1996). While circumstantial,
these late dates likely reflect instances of re-
locating individuals rather than first-time ar-
rivals. Taken together, these lines of evidence
suggest that at least some portion of adult
birds tend to disperse from one breeding ter-
ritory to another, both within and between sea-
sons. This also suggests that our return per-
centages represent a low estimate of survival
for the species.

We observed that HY birds depart from the
breeding grounds before AHY birds. This ear-
lier departure of young may be related to an
unusual pattern of molt described for White-
eyed Vireos, where juveniles exhibit a partial
replacement of primaries prior to fall migra-
tion, perhaps facilitating migration in HY
birds (Lloyd-Evans 1983). This pattern of
molt, however, may vary geographically
(George 1973). Whether departures of HY
birds precede adults in other parts of their
range is unknown. The timing of fall migra-
tion between HY and AHY White-eyed Vir-
eos is not significantly different in the south-
ern United States. Woodrey and Moore (1997)
found that AHY and HY White-eyed Vireos
did not differ in their distribution of arrivals.
This timing pattern is in contrast to Red-eyed
Vireos (V. olivaceus) whose adults depart sig-
nificantly earlier than the young, and the two
age-classes apparently migrate at different
rates (Woodrey and Chandler 1997). Perhaps
while young White-eyed Vireos depart earlier,
the adults minimize the timing differences by
the time the two groups reach the southern
United States. Alternatively, our observed ear-
lier departure of HY birds may signal the on-
set of a pre-migration dispersal by yoiing
birds, followed by their actual southerly mi-
gration some time later.

Our failure to recover any SY birds banded
as young may stem from several factors. First,
it is possible that young birds tend not to re-

Hopp et cil. • VIREO RETURNS

turn to the natal area. For many species, a dis-
persal of young provides an effective mecha-
nism for avoiding inbreeding (Greenwood and
Harvey 1982, Davis and Howe 1992), al-
though some young of other vireo species re-
turn to their natal area (see below). Alterna-
tively, since White-eyed Vireos inhabit a suc-
cessional window of habitat it may be in the
best interest of young birds to disperse to oth-
er areas, because the optimum successional
window will eventually close in their natal
area. Young birds may disperse from their na-
tal areas to locate suitable breeding habitat for
the following spring (e.g.. Brewer and Harri-
son 1975). Second, it is possible that the com-
bination of a modest number of birds banded
and relatively small study areas simply pre-
cluded our re-encounter with young. In the
BBL encounter records there were six report-
ed recoveries of banded HY birds in later
years; two were encountered in the original
10' block, two were encountered 15—20 km of
initial banding location, and two were en-
countered 150-200 km away from the original
banding block. These records are equivocal in
suggesting a modal pattern of natal dispersal
for the species. Instead the pattern appears
complex, with distance depending perhaps on
other factors, such as habitat size and com-
position, bird densities, or hatching time of
year. In Black-capped Vireos, young birds
show a gradient of natal dispersal to at least
21 km (Grzybowski 1991). For this species,
the return percentages were directly related to
size of the study area, with fewer returns in
smaller areas (Grzybowski 1995). In Bell’s
Vireos, natal dispersal is complex, with some
young returning to their natal area, and others
dispersing up to 300 km (Kus 1995, Greaves
and Labinger 1998). Greaves and Labinger
(1998) found that early and late (before and
after 15 June, respectively) cohorts of HY
Bell’s Vireos exhibit significantly different de-
grees of natal philopatry, with earlier young
returning at higher rates. For both Black-
capped and Bell’s vireos, reported return per-
centages of young from various studies were
15-27% (Greaves 1987; Grzybowski 1991,
1995; Greaves and Labinger 1998). Our lack
of returns provides us with little information
on natal dispersal and survival, and instead
raises questions about dispersal dynamics of
this species.

ACKNOWLEDGMENTS

We thank J. C. Barlow, R. Greenberg, J. A. Grzy-
bowski, T. Keller, B. Kingsolver, K. Kliinkiewicz, B.
McClellan, T. Melcer, C. Qualls, and two anonymous
reviewers for comments on earlier versions of this
manuscript. We thank the Bird Banding Laboratory for
providing encounter records for White-eyed Vireos.
We thank T. Lloyd-Evans and K. Klimkiewicz for per-
mission to use individual encounter records. We grate-
fully acknowledge Mr. and Mrs. L. R Collins, III for
their kind permission to conduct our field work on
their land. We thank G. McLaren, D. Kegley and C.
Carlin for field assistance. The work was partially sup-
ported by a Mednick Research fellowship and two
Mellon Faculty Development grants to SLH, and a J. J.
Munay research award to CAB. This research was
conducted while SLH was on the faculty of Emory and
Henry College in Emory, Virginia.

LITERATURE CITED

Askins, R. a., j. F. Lynch, and R. Greenberg. 1990.
Population declines in migratory birds in eastern
North America. Curr. Ornithol. 7:1-57.

Barber D. R. and T. E. Martin. 1997. Influence of
alternate host densities on Brown-headed Cowbird
parasitism rates in Black-capped Vireos. Condor
99:595-604.

Barlow, J. C. 1980. Patterns of ecological interactions
among migrant and resident vireos on the winter-
ing grounds. Pp. 79-107 in Migrant birds in the
Neotropics: ecology, behavior, distribution and
conservation (A. Keast and E. S. Morton, Eds.).
Smith.sonian Institution Press, Washington, D.C.
Bedard, J. and G. LaPointe. 1984. Banding returns,
arrival times, and site fidelity in the Savannah
Sparrow. Wilson Bull. 96:196—205.

Bibby, C. j., N. D. Burgess, and D. A. Hill. 1992.
Bird census techniques. Academic Press, London,
U.K.

Bradley, R. A. 1980. Vocal and territorial behavior in
the white-eyed vireo. Wilson Bull. 92:302-31 1.
Bradley, R. A. 1981. Song variation within a popu-
lation of White-eyed Vireos (Vireo gri.sens). Auk
98:80-87.

Brewer, R. and K. G. Harrison. 1975. The time of
habitat selection by birds. Ibis 117:521-522.
Conner, R. N., J. G. Dickson, B. A. Locke, and C.
A. Segelquist. 1983. Vegetation characteristics
important to common .songbirds in East Texas.
Wilson Bull. 95:349-361.

Davis. G. J. and R. W. Howe. 1992. Juvenile dispersal,
limited breeding sites, and the dynamics of meta-
populations. Thcor. Pop. Biol. 41:184-207.
DeGraaf, R. and j. H. Rappole. 1995. Neotropical
migratory birds. Cornell Univ. Press, Ithaca, New
York.

Franzreb, K. E. 1989. Endangered status and strate-
gies for protection of the Least Bell's Vireo (Vireo
helm pusiUus) in California. West. Birds 1 8:43-49.

THE WILSON BULLETIN

Vol. Ill, No. I, March 1999

George. W. G. 1973. Molt of juvenile White-eyed Vir-
eos. Wihson Bull. 85:327-330.

Graber, J. W., R. R. Graber, and E. L. Kirk. 1985.
Illinois birds: vireos. 111. Nat. Hist. Surv. Biol.
Notes 124:1-38.

Greaves, J. M. 1987. Nest-site tenacity of Least Bell’s
Vireos. West. Birds 18:50-54.

Greaves, J. M. and Z. Labinger. 1998. Site tenacity
and dispersal of Least Bell’s Vireos. Trans. West.
Sect. Wildl. Soc. 33:18-23.

Greenberg, R. 1992. Forest migrants in non-forest
habitats on the Yucatan Peninsula. Pp. 273-286 in
Ecology and conservation of neotropical migrant
landbirds (J. M. Hagan, III and D. W. Johnston,
Eds.). Smithsonian Institution Press, Washington,
DC.

Greenberg, R., D. K. Niven, S. L. Hopp, and C.
Boone. 1993. Fmgivory and coexistence in a res-
ident and a migratory vireo on the Yucatan pen-
insula. Condor 95:990-999.

Greenberg, R., M. S. Foster, and L. Marquez-Val-
DELAMAR. 1995. The role of the White-eyed Vireo
in the dispersal of Biirsera fruit on the Yucatan
peninsula. J. Trop. Ecol. 11:619-639.

Greenwood, P. J. 1980. Mating systems, philopatry,
and dispersal in birds and mammals. Anim. Be-
hav. 28:1 140-1 162.

Greenwood, P J. and P. H. Harvey. 1982. The natal
and breeding dispersal of birds. Annu. Rev. Ecol.
Syst. 13:1-21.

Grzybowski, j. a. 1991. Survivorship, dispersal and
population structure of Black-capped Vireos at the
Kerr Wildlife Management Area, Texas. Resour.
Protection Div., Texas Parks Wildlife Dept., Aus-
tin.

Grzybowski, J. A. 1995. The Black-capped Vireo
(Vireo alricapillus). In The birds of North Amer-
ica, no. 181 (A. Poole and F. Gill, Eds.). The
Academy of Natural Sciences, Philadelphia; The
American Ornithologists' Union, Washington,
D.C.

Grzybowski, J. A., R. B. Clapp, and J. T. Marshall.
1986. Hi.story and current status of the Black-
capped Vireo in Oklahoma. Am. Birds 34:742-
745.

Grzybowski, J. A., D. J. Tazik, and G. D. Schnell.
1994. Regional analysis of Black-capped Vireo
breeding habitats. Condor 96:512-544.

Hagan, J. M. and D. W. Johnston. 1992. Ecology and
con.servation of Neotropical migrant landbirds.
Smith.sonian Institution Press, Washington, D.C.

Hill, G. E. 1989. Late spring arrival and dull nuptial
plumage: aggression avoidance by yearling
males? Anim. Behav. 37:665-673.

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

Holmes, R. T. and T. W. Sherry. 1992. Site fidelity
of migratory warblers in temperate breeding and
Neotropical wintering areas: implications for pop-

ulation dynamics, habitat selection, and conser-
vation. Pp. 563-575 in Ecology and conservation
of neotropical migrant landbirds. (J. M. Hagan, III
and D. W. Johnston, Eds.). Smithsonian Institution
Press, Washington, D.C.

Hopp, S. L., A. Kirby, and C. A. Boone. 1995. White-
eyed Vireo (Vireo griseiis). In The birds of North
America, no. 168 (A. Poole and F. Gill, Eds.). The
Academy of Natural Sciences, Philadelphia; The
American Ornithologists’ Union; Washington,
DC.

Ketterson, E. D. and V. Nolan, Jr. 1983. The evo-
lution of differential bird migration. Curr. Orni-
thol. 1:357-402.

Kus, B. E. 1995. Status of the Least Bell’s Vireo and
southwestern Willow Flycatcher at Marine Corps
Base Camp Pendleton, California, in 1995. Envi-
ronmental and Natural Resources Office, Marine
Corps Base Camp Pendleton, California.

Lloyd-Evans, T. L. 1983. Incomplete molt of juvenile
White-eyed Vireos in Massachusetts. J. Field Or-
nithol. 54:50-57.

Lovehoy, T. E., Ill AND D. C. Oren. 1981. The mini-
mum critical size of ecosystems. Pp. 7-12 in For-
est island dynamics in man-dominated landscapes
(R. L. Burgess and D. M. Sharpe, Eds.). Springer-
Verlag, New York.

Lynch, J. F. 1989. Distribution of overwintering Ne-
arctic migrants in the Yucatan peninsula, I: gen-
eral patterns of occurrence. Condor 91:515-544.

Lynch, J. F. 1992. Distribution of overwintering Ne-
arctic migrants in the Yucatan peninsula, II: use
of native and human-modified vegetation. Pp.
178-196 in Ecology and conservation of neotrop-
ical migrant landbirds (J. M. Hagan, III and D. W.
Johnston. Eds.). Smithsonian Institution Press,
Washington, D.C.

Marvil, R. E. and a. Cruz. 1989. Impact of Brown-
headed Cowbird parasitism on reproductive suc-
cess of the Solitary Vireo. Auk 106:476-480.

M0LLER, A. P. 1990. Male tail length and female mate
choice in the monogamous swallow Hirundo rus-
tica. Anim. Behav. 39:458-465.

Morton, E. S. and K. C. Derrickson. 1990. The bi-
ological significance of age-specific return sched-
ules in breeding Purple Martins. Condor 92: 1040-
1050.

Nice, M. M. 1941. The role of territory in bird life.
Am. Midi. Nat. 26:441-487.

Nolan, V., Jr. 1978. The ecology and behavior of the
Prairie Warbler (Dendroica discolor). Ornithol.
Monogr. 26:1-595.

Payne, R. B. and L. L. Payne. 1990. Survival esti-
mates ot Indigo Buntings: comparisons of banding
recoveries and local observations. Condor 92'
938-946.

Price, J., S. Droege, and A. Price. 1995. The summer
atlas of North American birds. Academic Press,
London, U.K.

Rappole, j. H. 1995. The ecology of migrant birds: a

Hopp el al. • VIREO RETURNS

Neotropical perspective. Smithsonian Institution
Press, Washington, D.C.

Rappole, J. H. and D. W. Warner. 1980. Ecological
aspects of migrant bird behavior in Veracruz,
Mexico. Pp. 353-393 in Migrant birds in the Neo-
tropics: ecology, behavior, distribution and con-
servation (A. Keast and E. S. Morton, Eds.).
Smithsonian Institution Press, Washington, D.C.

Remsen, J. V, Jr., S. W. Cardiff, and D. L. Ditt-
MANN. 1996. Timing of migration and status of
vireos (Vireonidcie) in Louisiana. J. Eield Orni-
thol. 67:119-140.

Robbins, C. S., J. S. Sauer, R. S. Greenberg, and S.
Droege. 1989. Population declines in North
American birds that migrate to the Neotropics.
Proc. Natl. Acad. Sci. USA 86:7658-7662.

Sauer, J. R. and S. Droege. 1992. Geographic pat-
terns in population trends of Neotropical migrants
in North America. Pp. 26-42 in Ecology and con-
servation of Neotropical migrant landbirds (J. M.
Hagan, III and D. W. Johnston, Eds.). Smithsonian
Institution Press, Washington, D.C.

Sauer, J. R., G. W. Pendleton, and B. G. Peterjohn.
1996. Evaluating causes of population change in

North American insectivorous songbirds. Con-
serv. Biol. 10:465-478.

Temple, S. A. and J. R. Cary. 1988. Modeling dy-
namics of habitat-interior bird populations in frag-
mented landscapes. Conserv. Biol. 2:340-347.

Terborgh, j. W. 1980. The conservation status of Neo-
tropical migrants: present and future. Pp. 21-30 in
Migrant birds in the Neotropics: ecology, behav-
ior, distribution and conservation (A. Keast and E.
S. Morton, Eds.). Smithsonian Institution Press,
Washington, D.C.

USFWS (United States Fish and Wildlife Service).
1986. Endangered and threatened wildlife and
plants; determination of endangered status for the
Least Bell’s Vireo. Federal Register 51:16474—
16483.

USFWS (United States Fish and Wildlife Service).
1991. Black-capped Vireo {Vireo atricapillus) re-
covery plan. Austin, Texas.

WOODREY, M. S. AND C. R. CHANDLER. 1997. Age-
related timing of migration: geographic and inter-
specific patterns. Wilson Bull. 109:52—67.

WooDREY, M. S. AND E R. MooRE. 1997. Age-related
differences in the stopover of fall landbird mi-
grants on the coast of Alabama. Auk 114:695-707.

Wilson Bull., 111(1), 1999, pp. 56-60

RESPONSE OF BROWN-HEADED NUTHATCHES TO THINNING

OF PINE PLANTATIONS

MICHAEL D. WILSON^ - AND BRYAN D. WATTS'

ABSTRACT. — Brown-headed Nuthatches (Sitta pusilla) reached their highest abundance within loblolly pine
(Finns taecki) plantations in the hrst year after thinning and declined in subsequent years. Commercial thinning
of plantations resulted in a reduction of canopy cover, hardwood basal area, and understory density. Overall,
the detection rates of nuthatches were low (19% of points surveyed) and no nuthatches were detected in stands
before thinning. Nuthatches were more than three times as likely to be detected within survey points containing
snags compared to those that did not. However, snag density did not vary signihcantly between stand ages.
These patterns suggest that nuthatch distribution within stands may be influenced by snag distribution but that
distribution among stands may be determined by the density and height of understory vegetation. Received 13
March 1998, accepted 15 Sept. 1998.

Prior to European settlement of North
America, the Southeastern Coastal Plain was
characterize<J by old-growth pine forests that
covered more than 24 million ha (Croker
1979). This ecosystem was maintained by
low-intensity ground fires caused by lightning
strikes (Komarek 1964, 1974) and indigenous
people (Bartram 1791, Ware et al. 1993). Fires
occurred over vast areas at approximately 3-
5 year intervals (Chapman 1932, Krusac et al.
1995) and maintained forests with an open
midstory and dense ground cover of forbs and
grasses (Platt et al. 1991).

Land clearing for agriculture, harvesting of
longleaf pine (Pinus palustris) for the naval
stores industry, and the suppression of wild-
fires severly reduced the extent of the south-
eastern pine ecosystem by the early 1800s
(Ashe 1894, 1915; Pinchot and Ashe 1897).
Currently, natural stands of longleaf pine are
restricted to only about 1% of their former
range (Ware et al. 1993).

Brown-headed Nuthatches {Sitta pusilla)
are among a small group of species including
the Red-cockaded Woodpecker (Picoides ho-
realis) and the Bachman’s Sparrow (Aimophi-
la aestivalis) that are endemic to the south-
eastern pine ecosystem (Jackson 1988). The
Red-cockaded Woodpecker and the Bach-
man’s Sparrow have experienced significant
population declines within the southeast re-
gion (Lennartz and Henry 1985, Dunning
1993); however, both have benefited from

' Center F*or Con.servation Biology, College of Wil-
liam and Mary. Williamsburg, VA 23187-8795.
^Corresponding author; E-mail: mdwils@mail.wm.edu

management practices that produce a habitat
structure similar to the historic southeastern
pine ecosystem (Gobris 1992, Plentovich et al.
1998).

The Brown-headed Nuthatch has also ex-
perienced a contraction of its former range
(Jackson 1988), and according to data from
the U.S. Fish and Wildlife Service’s Breeding
Bird Survey has been declining at a rate of
more than 1.5% per year throughout much of
the Southeast (Sauer et al. 1997). Very little
is known about the ecology and habitat re-
quirements of the Brown-headed Nuthatch
and even less is known about how current for-
est management practices may affect its dis-
tribution. The purpose of this paper is to pre-
sent some information on the use of pine plan-
tations by Brown-headed Nuthatches relative
to stand age and commercial thinning.

METHODS

This study was conducted in managed loblolly pine
(Finns taeda) plantations in eastern North Carolina
(approximately 35° 50' N, 77° 00' W). These planta-
tions are managed for pulpwood and sawtimber pro-
duction on a 30—35 year rotation. After canopy clo-
sure, the plantations are thinned twice before final har-
vest. Thinnings reduce the number of trees, open the
forest canopy, and allow for growth of understory veg-
etation.

We selected stands that repre.sented seven different
ages and relation to thinning: (1)9-11 year old stands
with closed canopies, (2) 13—16 year old stands within
one year after the first commercial thinning, (3) l6-f8
year old stands that were three years after first thin-
ning, (4) 19-21 year old stands that were 5 years after
first thinning, (5) 22-26 year old stands that were with-
in I year after .second (hinning, (6) 28-29 year old
stands that were 3 years after second thinning, and (7)
30—35 year old stands that were 5 years after second

VV/7.sv;/( and Walls • BROWN-HEADHD NUTHATCH HABn'AT USE

thinning. Six replicate stands (each > 24 ha) were se-
lected for each stand type. Within each stand age,
stands were chosen to minimize variation in planted
stocking level and basal area of pine. Stands within
each type were separated by at least 500 m.

Seven minute, rtxed-radiiis (50 m) point counts were
used to measure the density and frequency of occur-
rence of Brown-headed Nuthatches within study plan-
tations. Four point counts were established within each
stand and distributed evenly between edge and interior
locations. Edge points were positioned 50 m from the
stand edge such that the plot perimeter was tangential
to the stand edge. For all stands, edge points were
positioned on stand edges that were adjacent to log-
ging roads. Interior points were positioned 150 m from
the stand edge. Stands were surveyed three times be-
tween 1 June and 4 July 1997. Surveys were initiated
0.5 hr after sunrise and concluded within four hours.

The vegetation was sampled within all point count
plots to determine ( 1 ) vegetation changes across the
growing period, (2) vegetation responses to thinning,
and (3) relationships between nuthatch distribution and
vegetation. Linear transects were used for vegetation
sampling parallel to the long, regularly distributed can-
opy openings created by row thinning. The length of
vegetation transects was standardized to 25 m and the
width varied between 4 and 7 m to accommodate var-
iation in thinned and non-thinned longitudinal rows
within stands. Four vegetation transects were estab-
lished within each point count and equally distributed
between thinned and non-thinned rows.

Habitat data were collected at two levels within tran-
sects. Counts of all large woody plants (> 8 cm dbh)
and dead standing stems (snags) by type (hardwood vs
pine) and stem diameter class (8-23, 24—38, > 38 cm
dbh) were made over the entire 25 m transect. Pine
and hardwood basal areas were estimated using the
midpoint dbh for the two smaller diameter classes and
38 cm for the larger class (few trees were larger than
38 cm dbh). Additional information was collected
within 2 X 2 m quadrats established at opposite ends
of each transect. Information collected included can-
opy cover (measured in four cardinal directions of a
compass by convex densiometer) and canopy height
(measured using a clinometer), groundcover height and
counts of all stems, shrubs, and saplings (> 0.5 m in
height and < 8 cm dbh). Counts were summed to rep-
resent total groundcover density (stem.s/m^).

A Kruskal-Wallis test was used to test for the influ-
ence of stand age on all habitat variables except for
counts of snags. Because of the many zero values for
counts of both Brown-headed Nuthatches and snags
(i.e., data were distributed as a negative binomial), fre-
quency of occurrence values were used to assess pat-
terns among stand types. The relationship between nut-
hatches and habitat variables was assessed at the level
of the point count using Kendall’s rank correlation.
Nuthatches were not detected in 9-10 year old stands,
so this stand age was eliminated from all analy.ses and
used only for descriptive purposes.

RESULTS

Stand age had a significant influence (Krus-
kal-Wallis test: df = 5, P 0.01) on all hab-
itat variables measured except the density of
snags (Table 1). Canopy height and ground
cover height were positively related to stand
age whereas pine density was negatively re-
lated to stand age. All other significant vari-
ables increased with stand age but were also
influenced by commercial thinning.

Detection rates for Brown-headed Nut-
hatches within pine plantations were relatively
low. Nuthatches were detected in 15 of 42
(35.7%) pine stands included in the study and
32 of 168 (19%) individual point counts sur-
veyed. Stand age had a significant influence
on the detection of Brown-headed Nuthatches

(XNates Correction = 12-3, df = 5, P < 0.05; Fig.
1). No nuthatches were detected in forest
patches prior to first thinning. The number of
points where nuthatches were detected was
greatest in the year immediately following
thinning and declined with time after thinning.
Using survey points as statistical units, nut-
hatches were significantly associated with
habitat variables that were directly influenced
by thinning events. For example, nuthatch
abundance was negatively correlated with
canopy cover (Kendall t = —0.12, 77 = 144,
P < 0.03), hardwood density (t = —0.14, n
= 144, P < 0.02), and basal area of hard-
woods (t = —0.13, n = 144, P < 0.02). In
addition, nuthatch abundance was positively
correlated with groundcover density (t =
0.19, n = 144, P < 0.001). Nuthatch density
was not significantly correlated with canopy
height (t = —0.04, n = 144, P > 0.05), pine
density (t = 0.04, n — 144, P > 0.05) or pine
basal area (t = —0.02, n = 144, P > 0.05).

Although stand type did not have a signif-
icant influence on the number of survey plots
containing standing snags, and snags did not
appear to result from thinning, nuthatches
were positively correlated with standing snags
(T = 0.15, 77 = 144, P < 0.009). In fact, nut-
hatches were over three times more likely to
be detected within survey plots containing
standing snags (12 of 32 plots, 37.5%) com-
pared to plots that did not (13 of 122 plots,
11.6%; = 7.35, df = 1, P < 0.007).

DISCUSSION

It is generally thought that partially rotted
wood is a prerequisite for cavity excavation

TABLE 1. Habitat characteristics of 7 stand types selected for study (« — 6 each). Median values are presented for all habitat variables except for snags where
frequency of occurrence is presented.

THE WILSON BULLETIN • Vol. 111. No. 1. March 1999

— — — oo — lO — —
cqqqooooo
dcddddddd

VVVVVVAVV

I DJ
O' >s

— VO r- rj (N

^ O O O O On (N
(N On r- (N ID —

m O'

O On O On O' cD
P O O »D NO CN

(N ON On fN CN

Dj O (N O NO
CD (N O <N O
(N 00 so —

^ o
d (N

ON Tf O 00 o

— ‘ no d — d

n 00 — — ON

rD ID
nO CD

^ ON o p o o

On d — - d

— 00 CD O

fD —

00 00
— 00

ID p O CD
NO d ^
— On CD
O

^ o
d

— ID
nO i

Tf «D

r- 00 — ^

p O 00 O
d d »D '

•o

eg eg

'S =5

C/) (N

-■ r*

eg o
6

flj rM

■n E
.S E

o B

D. c/)
u- ^ •
O ^

=tt •

Of)

‘S

-C

Vi eg

w (U
u

>> e^

o c/;

>1 S

CL T3
C3 ^

« rt
C

<U C3
■V

■o -o
O O
O O

■a •2

U U cu ix j: H

c .2?

g' T3 J=
C u u
(U o OJ
3 > >

cr O o

<U O CJ
^"0-0
^ c c

00 O 3

eg o o
c ^ ^

c/5 o a

by Brown-headed Nuthatches, and the major-
ity of cavities reported have been located in
snags (McNair 1984). In general the popula-
tion density of cavity-nesting birds is posi-
tively related to snag density (Cunningham et
al. 1980, O’Meara 1984, Raphael and White
1984). In Florida, a large percentage of the
variation in the density of cavity-nesting birds
(including Brown-headed Nuthatches) was ex-
plained by snag density and dispersion (Land
et al. 1989). Snag density has been shown to
be lower in pine plantations than in natural
stands (McComb et al. 1986), and was low in
the plantations we surveyed. Brown-headed
Nuthatches were significantly more likely to
be detected within survey points that con-
tained snags. The possibility that snag density
may serve to limit overall nuthatch density
within loblolly pine plantations requires fur-
ther investigation. Because nuthatch density
was influenced by thinning and snag density,
snag density alone does not explain nuthatch
distribution among pine stands.

Brown-headed Nuthatches exhibited a rapid
response to thinning. Nuthatches were not de-
tected within pine plantations prior to the first
thinning but reached their highest densities
within the first year after thinning. This re-
sponse suggests that thinning activities may,
in some way, enhance habitat structure for
nuthatches. Thinning activities were shown to
reduce canopy cover, reduce the density and
basal area of hardwoods, and increase ground-
cover density.

Although the importance of canopy cover
to the use of pinelands by Brown-headed Nut-
hatches has not been explored, Engstrom and
coworkers (1984) reported that nuthatch abun-
dance declined as the density of midstory
hardwoods increased. This result is consistent
with our observations that nuthatches were
less common in years after thinning, as the
density and basal area of hardwoods in-
creased.

The effects of burning hardwoods on stand
use by Brown-headed Nuthatches is similar to
that of thinning. Nuthatches used (45 nut-
hatches/km^) mature longleaf pine stands that
were regularly burned in Florida (Repenning
and Labisky 1985), but not stands with well
developed understories (Hirth et al. 1991).
Nuthatch density decreased with time follow-
ing burning (Engstrom et al. 1984, Wilson et

Wilson and Watts • BROWN-HEADED NUTHATCH HABITAT USE

FIG. 1. Detection frequency (# of point counts) of Brown-headed Nuthatches in commercially thinned pine
plantations in eastern North Carolina (n = 24 point counts for each stand type). Frequency distribution between
stand ages was significantly different from an even distribution (x^vaies correction = 12.3, df = 5, P < 0.05).

al. 1995), sirnilar to the decline in nuthatches
we observed following thinning.

One possible explanation for the inverse re-
lationship between the density of understory
vegetation and numbers of Brown-headed
Nuthatches is that vegetation may obscure po-
tential cavity locations. Brown-headed Nut-
hatch cavities are frequently excavated in rel-
atively low positions; usually below 3.66 m
(n = 309; McNair 1984). Most (68%) cavities
were located in tree stumps (McNair 1984)
suggesting that the potential for increasing
cavity height may be limited. Regeneration of
understory and groundcover vegetation was
rapid in the current study such that low cavity
positions could be obscured quickly, but the
impact on patch use by Brown-headed Nut-
hatches is unknown.

ACKNOWLEDGMENTS

We thank T. Melchiors for administrative support
and assistance through all aspects of the field work and
for providing editorial comments on the manuscript.
We also thank B. Barber and J. Hughes for educating
us about forest tracts and silvicultural techniques and
for assistance in selecting study sites. D. Bradshaw, G.
Levandoski, and B. Piccolo provided valuable field as-
sistance, and T. Rafiq helped with data management.
This study was funded through a cooperative agree-
ment between the Weyerhaeuser Company and the
College of William and Mary.

LITERATURE CITED

Ashe, W. W. 1894. The forests, forest lands, and forest
products of eastern North Carolina. N.C. Geol.
Surv. 5:1-128.

Ashe, W. W. 1915. Loblolly or North Carolina pine.

N.C. Geol. Surv. and Econ. Surv. 24:1—176.
Bartram, W. 1791. Travels through North and South
Carolina, Georgia, east and west Florida. Dover
Publications, New York.

Chapman, H. H. 1932. Is the longleaf type a climax?
Ecology 13:328-334.

Conner, R. N. and D. C. Rudolph. 1991. Effects of
midstory reduction and thinning in Red-cockaded
Woodpecker cavity clusters. Wildl. Soc. Bull. 19:
63-66.

Croker, T. C., Jr. 1979. Longleaf pine. J. For. Hist.
23:32-43.

Cunningham, J. B., R. P. Balda, and W. S. Gaud.
1980. Selection and use of snags by secondary
cavity-nesting birds of the ponderosa pine forest.
U.S. For. Serv. Res. Pap. RM-222:1-15.

Dunning, J. G. Jr. 1993. Bachman's Sparrow in The
birds of North America, no. 38 (A. Poole. P. Set-
tenheim, and F. Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists' Union, Washington,
D.C.

Engstrom, R. T, R. L. Crawford, and W. W. Baker.
1984. Breeding bird populations in relation to
changing forest structure following fire exclusion:
a 15-year study. Wilson Bull. 96:437-450.
Gobris, N. M. 1992. Habitat occupancy during the
breeding season by Bachman's Sparrow at Pied-

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

mont National Wildlife Refuge in central Georgia.

M. Sc. Thesis, Univ. of Georgia, Athens.

Hirth, D. H., L. D. Harris, and R. E Noss. 1991.

Avian community dynamics in a peninsular Flor-
ida longleaf pine forest. Fla. Field. Nat. 19:33-48.
Jackson, J. A. 1988. The southeastern pine forest eco-
system and its birds: past, present, and future. Pp.
119-159 in Bird conservation 3 (J. A. Jackson,
Ed.). Univ. of Wisconsin Press, Madison.
Komarek, E. V., Sr. 1964. The natural history of light-
ning. Proc. Tall Timbers Fire Ecol. Conf. 3:139-
183.

Komarek, E. V., Sr. 1974. Effects of fire on temperate
forests and related ecosystems: southeastern Unit-
ed States. Pp. 251-278 in Fire and ecosystems (T.
T. Kowalski and C. E. Ahlgren, Eds.). Academic
Press, New York.

Krusac, D. L., j. M. Dabney, and J. J. Petrick. 1995.
An ecological approach to recovering the Red-
cockaded Woodpecker on southern National For-
ests. Pp. 61-66 in Red-cockaded Woodpecker: re-
cover, ecology and management (D. L. Kulhavy,
R. G. Hooper, and R. Costa, Eds). Center for Ap-
plied Studies in Forestry, College of Forestry, Ste-
phen E Austin State Univ., Nacogdoches, Texas.
Land, D., W. R. Marion, and T. E. O’Meara. 1989.
Snag availability and cavity nesting birds in slash
pine plantations. J. Wildl. Manage. 53: 1 165-1171.
Lennartz, M. R. and V. G. Henry. 1985. Recovery
plan for the Red-cockaded Woodpecker. U.S. Fish
and Wildlife Service, Atlanta, Georgia.

McComb. W. C., S. a. Bonney, R. M. Sheffield, and

N. D. Cost. 1986. Snag re.sources in Florida — are
they sufficient for average populations of primary
cavity nesters? Wildl. Soc. Bull. 14:40-48.

McNair, D. B. 1984. Clutch-size and nest placement
in the Brown-headed Nuthatch. Wilson Bull. 96:
296-301.

O’Meara, T. E. 1984. Habitat-island effects on the avi-
an community in cypress ponds. Proc. Annu.
Conf. Southeast. Assoc. Fish W'ildl. Agencies 38:
97-110.

Platt, W. J., J. S. Glitzenstein, and D. R. Streng.
1991. Evaluating pyrogenicity and its effects on
vegetation in longleaf pine savannahs. Proc. Tall
Timbers Fire Ecol. Conf. 17:143-161.

PiNCHOT, G. AND W. W. AsHE. 1897. Timber trees and
forests of North Carolina. N.C. Geol. Surv. 6:1-
227.

Plentovich, S., j. W. Tucker, Jr., N. R. Holler, and
G. E. Hill. 1998. Enhancing Bachman’s Sparrow
habitat via management of Red-cockaded Wood-
peckers. J. Wild. Manage. 62:347-354.

Raphael, M. G. and M. White. 1984. Use of snags
by cavity-nesting birds in the Sierra Nevada.
Wildl. Monogr. 86:1-66.

Repenning, R. W. and R. F. Labisky. 1985. Effects of
even-age timber management on bird communi-
ties of the longleaf pine forest in northern Florida.
J. Wildl. Manage. 49:1088-1098.

Sauer, J. R., J. E. Hines, G. Gough, I. Thomas, and
B. G. Peterjohn. 1997. The North American
breeding bird survey results and analysis, version
96.4. Patuxent Wildlife Research Center, Laurel,
Maryland. URL = http://www.mbr.nbs.gov/bbs/
bbs.html

Ware, S., C. Frost, and P D. Doerr. 1993. Southern
mixed hardwood forest: the former longleaf pine
forest. Pp. 447-494 in Biodiversity of the south-
eastern United States (W. H. Martin, S. G. Boyce,
and A. C. Echternacht, Eds.). John Wiley and
Sons, Inc., New York.

Wilson, C. W., R. E. Masters, and G. A. Bukenho-
FER. 1995. Breeding bird response to pine-grass-
land community restoration for Red-cockaded
Woodpeckers. J. Wildl. Manage. 59:56-67.

Wilson Bull., 111(1), 1999, pp. 61-69

DIFFERENCES IN MIGRATORY TIMING AND ENERGETIC
CONDITION AMONG SEX/AGE CLASSES IN MIGRANT
RUBY-CROWNED KINGLETS

DAVID L, SWANSON,'^ ERIC T. LIKNES,' AND KURTIS L. DEAN'

ABSTRACT. — Ruby-crowned Kinglets, Kegulus calendula, are small temperate zone passerine migrants that
breed in conifer forests of Canada and the western United States and winter along the west coast and southern
United States into Mexico. Previous studies have revealed that male kinglets precede females during spring
migration and that females precede males during fall migration at various sites in eastern North America. We
used mist net capture data to document sex/age structure, fat loads, and morphometries of kinglets passing
through southeastern South Dakota during spring and fall migrations from 1992-1995. Males migrated signifi-
cantly earlier than females in spring; median passage dates differed by at least 8 days. These data are consistent
with U.S. Geological Survey Bird Banding Laboratory records for central and eastern North America, which
indicate that passage of males before females during spring migration is a widespread phenomenon. Bird Banding
Laboratory data also indicate that males winter significantly farther north than females. We tested whether the
differential winter distribution of the sexes could account for the differential pattern of spring migration and
found that differential winter distributions do not fully account for the differential timing of spring migration
between the sexes. Sex-specific migration rates do not differ over the entire spring migration route, so differences
in the onset of migration apparently contribute to differential spring migration. Males migrated significantly later
in the fall than females, with median dates in southeastern South Dakota differing by 7 days. We also examined
whether differences in energetic condition were associated with differences in migratory timing. Visible fat
scores did not differ between males and females in spring, although males had a significantly higher mean
condition index (mass/wing chord). Sexes within the same age class did not differ in visible fat scores in fall,
but adults tended to be fatter than juveniles, significantly so in some cases. No significant differences in condition
index were apparent among sex/age classes in fall, although the adult-juvenile difference approached significance
{P < 0.09), with adults having higher values. This tendency toward elevated fat and condition index in adults
relative to juveniles in fall is consistent with more efficient foraging by adults than by inexperienced juveniles
at migratory stopover sites, although differences in predation pressure or migratory routes and destinations might
also be involved. Received 12 May 1998, accepted 24 Oct. 1998.

Differential timing of migration of sexes
an(j/or age classes has been reported for nu-
merous birds (see Gauthreaux 1982, Ramos
1988 for reviews). For passerine birds, males
generally precede females in spring, but the
timing of migration in fall is more complex
and varied between sexes or age classes (see
Hall 1981, Gauthreaux 1982, Francis and
Cooke 1986, Ramos 1988, Nolan and Ketter-
son 1990, Hussell 1991, Winker and Rappole
1992). Earlier passage of males than females
in spring presumably relates to advantages in
territory acquisition that accrue to those in-
dividuals arriving early on the breeding
grounds, regardless of whether this early ar-
rival results from sexual differences in win-
tering areas, departure dates, or rates of mi-
gration (Myers 1981, Chandler and Mulvihill

' Dept, of Biology, Univ. of South Dakota, Vermil-
lion, SD 57069-2390.

^ Corresponding author; E-mail:
dlswanso(g) sunflowr.usd.edu

1990). Differential timing of fall migration
may result from differences in departure dates
or rates of migration as a result of ecological
or physiological factors (e.g., juvenile devel-
opment, parental responsibilities, molt, fat de-
position), and/or from differential distribution
of age/sex classes throughout the wintering
range (Gauthreaux 1982, Ketterson and Nolan
1976, Morton 1984, Prescott and Middleton
1990, Woodrey and Chandler 1997).

Ruby-crowned Kinglets, Regulus calendu-
la, are small passerines breeding in coniferous
or mixed coniferous-deciduous forests over
most of northern and western North America
and wintering throughout the western and
southern United States, Mexico, and south to
Guatemala (American Ornithologists’ Union
1983). The winter range is apparently restrict-
ed to regions with a relatively mild winter cli-
mate (Lepthein and Bock 1976, Root 1988).
Differential timing of migration in males and
females has been reported from Iowa (Crim
1976) and Ontario (Fairfield and Shirokoff

THE WILSON BULLETIN

Vol. Ill, No. I, March 1999

1978), but migration timing of sexes in the
Great Plains has not been previously reported.
In addition, male kinglets apparently winter
farther north, on average, than females (Fair-
field and Shirokoff 1978), but the extent to
which differences in winter distribution affect
differential migration is unknown. No previ-
ous study has been attempted to define the rel-
ative roles of geographic origin, departure
dates, and rates of migration in shaping dif-
ferential migration patterns in kinglets.

We used capture data from spring and fall
migration periods from 1992-1995 in south-
eastern South Dakota to document patterns of
migratory movement in Ruby-crowned King-
lets and to determine whether timing of mi-
gration and energetic condition differed be-
tween sexes in either season. We compared
these results with banding data from the U.S.
Geological Survey Bird Banding Laboratory
(BBL) for central and eastern North America
(east of 107° W longitude) to determine if pat-
terns of migration between sexes in south-
eastern South Dakota are consistent with mi-
gratory patterns in Ruby-crowned Kinglets in
eastern North America. We also tested the hy-
pothesis that differential winter distributions
could fully account for differential migratory
timing.

METHODS

Collection of capture data. — Banding was conduct-
ed during spring and fall migration from 1992-1995
at four deciduous woodland study sites in southeastern
South Dakota. Two sites were located in riparian hab-
itats along the Missouri River (42° 45' N, 97° 00' W;
and 42° 46' N, 97° 07' W), one study site was in a
riparian woodland along the Big Sioux River (42° 45'
N, 96° 37' W), and one study site included both ripar-
ian and upland woodlands along Brule Creek (42° 55'
N. 96° 46' W), a tributary of the Big Sioux River. Ri-
parian habitats along the Missouri River consisted of
deciduous forest dominated by cottonwood (Populu.t
deltoide.s). boxelder (.Acer nef>undo), American elm
(Ulmu.s- americana), mulberry (Moru.s alba), and dog-
wood (Cannes spp.). The Big Sioux River site was
dominated by boxelder, silver maple (Acer .sacchari-
nunt), and cottonwood. The riparian forest at the Brule
Creek site consisted mainly of boxelder and American
elm, while the upland forest was dominated by bur oak
(Quercu.s macrocarpa) with American elm and hack-
berry (Celti.s occidentali.s) also present. The Missouri
River study sites have a generally west-east orienta-
tion, while the Big Sioux River and Brule Creek sites
are oriented north-south. Extensive deciduous wood-
lands in this area of .South Dakota, and indeed over

most of the northern Great Plains, are mainly restricted
to river courses (South Dakota Ornithologists’ Union
1991).

Kinglets were captured using mist nets during spring
(15 April-26 May) and fall (5 September-2 Novem-
ber) migration periods. Capture dates were chosen to
coincide with the major migratory movements for
Neotropical migrants, not kinglets specifically, so the
distribution of capture effort did not cover the kinglet
migration evenly. However, the bulk of the kinglet mi-
gration through South Dakota occurs during our cap-
ture periods (South Dakota Ornithologists’ Union
1991). From 15 April through 26 May in spring and
from 5 September to 5 October in fall, nets were
opened daily and cycled on a 4 day rotation among
the four study sites so that each study site was sampled
every 4 days. Later in the fall, nets were not opened
every day and most capture effort was concentrated at
the Brule Creek study site. We erected 2-7 mist nets
(10 m X 2.6 m, 30 mm mesh) each day, depending on
weather and available personnel. Net placement at the
various study sites was consistent among seasons and
years. We opened the nets at sunrise and nets remained
open until approximately 1 1 :00 CST during both
spring and fall migration. Upon capture we measured
the bird’s mass to the nearest 0.1 g with an Ohaus
Model LS200 balance, unflattened wing chord to the
nearest 0.1 mm, and visible fat on a scale of 0-5
(Helms and Drury 1960). Fat scoring was performed
by all three authors, but we regularly checked each
other on individual birds to ensure that we were scor-
ing fat similarly. During fall migration birds were aged
by skull ossification as “after hatching-year’’ (AHY,
adults) or “hatching-year’’ (HY, juveniles), although
this method may be inaccurate after 1 October because
some hatching-year birds may complete ossification af-
ter this date (Leberman 1970, Pyle et al. 1987). Retrix
shape was not an effective measure of age in our pop-
ulation because most birds showed an intermediate
condition (Pyle et al. 1987). Sex was determined by
plumage differences in both seasons. Recaptures were
very rare and were not included in our analyses.

We used banding data from the BBL from 1986-
1995 to determine winter distributions and timing of
spring migration in males and females. Data were
grouped into 5° latitude zones for comparisons. Only
banding records east of 107°W longitude were used
in these comparisons to eliminate populations in west-
ern North America that breed at lower latitudes and
winter at higher latitudes than populations in central
and eastern North America. We considered records
from 1 December-29 February as winter records, and
from 1 April-31 May as spring records. Banding data
were provided as number of captures for consecutive
5 day periods.

To determine whether differences in winter distri-
bution of male and female Ruby-crowned Kinglets are
sufficient to account for spring migration patterns, we
u.sed a method developed by Chandler and Mulvihill
(1990) to calculate a predicted pattern of differential
migration based on winter distributions of the sexes.

Swanson et al. • DIFFERENTIAL MIGRATION IN KINGLETS

TABLE 1. Winter (December-Eebruary) abundance of male and female Ruby-crowned Kinglets in eastern
and central North America from BBL records and results of Goodness of Fit tests for equal abundance at
each latitude.

L.ititude“ (°N)

Males

Females

% Males

Xc^

Significance

< 25

50.0

0.04

P > 0.05

25-29° 59'

51.7

0.02

P > 0.05

30-34° 59'

169

133

56.0

4.06

P < 0.05

35-39° 59'

76.1

18.25

P < 0.001

“ Latitudes north of 40° N had too few records (n = 5) to adequately assess distribution of sexes.

Chandler and Mulvihill (1990) assumed a linear
change in the percentage of males in the population
from the northern to the southern extent of the win-
tering range for their winter distribution; they predict-
ed a linear change in the percentage of males in the
population throughout migration from the first appear-
ance to the last appearance of migrating males. They
then used linear regression to calculate the predicted
percentage of males throughout the migratory period
and compared this regression against a linear regres-
sion of the observed percentage of males over the mi-
gratory period.

We used BBL data for wintering Ruby-crowned
Kinglets to calculate a predicted differential pattern of
spring migration between the sexes. The predicted per-
centage of males in the migratory population was cal-
culated using the percentage of males in wintering
populations from BBL data for 5° latitude zones over
the entire kinglet wintering range in eastern and central
North America (Table 1). We then divided the spring
migratory period for males into an equal number of
intervals (4 intervals at 5 days each) and assumed that
the percent males during each successive 5-day inter-
val (from first to last appearance) should equal the per-
cent males in consecutive 5° latitude zones from north
to south.

Statistics. — Median passage dates for sexes were
compared by median tests (Zar 1996). Because capture
effort was not evenly distributed over the entire kinglet
migration period, median tests may be misleading. To
control for uneven capture effort, we compared pas-
sage dates for sexes by Kolmogorov-Smirnov test for
frequency distributions (Sokal and Rohlf 1995) as a
function of cumulative net hours over all years pooled.
Morphometries, condition index [mass/wing chord
(Winker 1995)], and visible fat scores for males and
females in spring were compared by Student’s r-test or,
if sample variances were unequal (determined by F-
test), Mann-Whitney test. In fall, morphometries, con-
dition index, and visible fat scores of sex/age classes
were compared by one-way ANOVA, or by Kruskal-
Wallis test if sample variances differed (F-test). Be-
cause skull ossification provides unreliable age dis-
tinction after 1 October in kinglets (Leberman 1970),
fall comparisons of sex/age classes were conducted
twice, excluding data after 30 September for birds with
completely ossified skulls (aged AHY) and excluding
all data (HY and AFIY) after 30 September. If signif-

icant differences among sex/age classes were detected
by ANOVA or Kruskal-Wallis tests, Student’s r-tests
or Mann-Whitney tests were used to identify which
means differed with a sequential Bonferroni procedure
employed for a-level adjustment to protect against in-
flated type 1 error rates in multiple comparisons (Rice
1989). This procedure involved six comparisons (k =
6) among the four sex/age classes for each variable
measured (mass, wing chord, fat score, and condition
index). For a-level adjustment, P values from individ-
ual r-tests for each comparison were ranked from
smallest (P,) to largest (Pg). The smallest P value was
considered significant only if it was less than a/k (0.05/
6 in this case). If Pj was significant, then Pj was con-
sidered significant only if less than a/k — 1, P3 was
considered significant only if less than a/k — 2, and
so on, until the equality Pj ^ a/(l-t-/: — i) was not met
(Rice 1989). Once a P value was found to be not sig-
nificant, all larger P values for that comparison were
also considered nonsignificant.

For winter BBL data, distributions of males and fe-
males in 5° latitude zones were compared by 2-tailed

Goodness of Fit test with a null hypothesis of equal
distribution (1:1) in each latitude zone. Bird Banding
Laboratory data for timing of spring migration in
males and females were compared by Kolmogorov-
Smirnov test for frequency distribution as a function
of cumulative captures during the migratory season
pooled over all years (1986-1995). To determine
whether winter distributions of sexes could fully ac-
count for differential migration in spring, predicted and
observed percentages of males in the migratory pop-
ulation were compared by ANCOVA. Because neither
the change in the percentage of males in the wintering
population as a function of latitude nor the percentage
of males in the migratory population through the
spring migratory period varied in a linear fashion (Fig.
1), data for percent males were arcsine transformed
prior to ANCOVA. In addition, we conducted Fisher’s
exact tests on observed versus predicted percent males
for each 5 day interval over the spring migration of
male kinglets. Statistical comparisons of mass, wing
chord, condition index, and visible fat scores were con-
ducted with Number Cruncher Statistical System (Ver-
sion 4.1, Kaysville, Utah). Other statistical tests were
performed with SAS (PC Version 6.03, SAS Institute
1988).

THE WILSON BULLETIN

Vo/. Ill, No. J, March 1999

FIG. 1 . Observed and predicted percentages of
male Ruby-crowned Kinglets in the population during
successive 5-day intervals over the spring migratory
period as measured by capture data (first captures
only). Predicted values were calculated based upon
U.S. Geological Survey Bird Banding Laboratory data
from 1985-1995 for wintering distributions of male
and females kinglets east of 107° W longitude.

RESULTS

Southeastern South Dakota is not a part of
the breeding range for Ruby-crowned Kinglets
(South Dakota Ornithologists’ Union 1991),
so all birds in our samples were transients.
Median passage dates for all years pooled for
male and female kinglets in spring were 23
April and 1 May, respectively (Fig. 2). For fall
migration, median passage dates for all years
pooled were 29 September for females and 6
October for males (Fig. 2). Median tests in-
dicated that males migrated significantly ear-
lier (x- = 45.74, df = \, P < O.OOl) than
females in spring and significantly later (y^ =
5.30, df = \, P < 0.01) than females in fall.

Total net hours ( 1 net hour = 1 net open
for I h) were 2,081 in spring and 1,369 in fall.
Our capture effort increased gradually over
the spring migratory period (as days became
longer) and similarly decreased gradually over
the early part of fall migration (5 September-
5 October). Capture effort was reduced after
5 October and only 12% of all fall net hours
occurred after this date. This reduced capture
effort late in fall migration may have reduced
the relative number of captures of late mi-
grating kinglets, mostly males, so the differ-

Females -

(44)

Males -

(60)

SPRING

10 Apr 20 Apr 30 Apr 10 May 20 May 30 May

Date

FIG. 2. Distribution of capture dates (first captures
only) for male and females Ruby-crowned Kinglets
during spring and fall migration. Box plots display the
median (vertical line within box), the interquartile
range (extent of box), the 10th-90th percentile range
(extent of horizontal lines), and the data points falling
below the 10th or above the 90th percentiles (dots).
For spring females, the median passage date and the
lower interquartile boundary were coincident on 1
May. The numbers included in parenthesis for males
and females represent sample sizes for each season.

ence in median passage dates between sexes
in fall that we report is probably conservative.
Because capture effort was not even through-
out migration periods, we tested whether fre-
quency distributions of sexes differed as a
function of cumulative capture effort. Again,
these tests indicated that males migrated sig-
nificantly earlier in spring (Kolmogorov-Smir-
nov test: D = 0.673, n = 104, P < 0.001) and
significantly later in fall (D = 0.425, n = 132,
P < 0.001) than females.

Data from the BBL indicated that at all lat-
itudes males migrated significantly earlier in
spring than females (Table 2). Median passage
dates differed between the sexes by 10-15
days in the different 5° latitude zones. In win-
ter, BBL data indicated that males had a sig-
nificantly higher relative abundance than fe-
males at 30-35° N (y^ = 4.06, df = 1, P <

Swanson et al. • DIFFEREN I’lAL MIGRATION IN KINGLETS

TABLE 2. Median migration dates (5-day intervals) for male and female Ruby-crowned Kinglets in eastern
and eentral North America from BBL records for April and May 1985-1996. Males migrated signiheantly earlier
(Kolmogorov-Smirnov test: F < O.OOl) than females at all latitudes. Sample sizes are given in parentheses.

Latitude-' (°N) Males Females I) F

30-34° 59'

6-10 April (69)

16-20 April (115)

0.333

< 0.001

35-39° 59'

16-20 April (275)

1-5 May (319)

0.566

< 0.001

40-44° 59'

21-25 April (1647)

6-10 May (1691)

0.502

< 0.001

45-50°

26-30 April (230)

6-10 May (297)

().45()

< O.OOl

® 5° latitude zones to the south and north of those above had loo few total records 27) to adequately quantity medians tor each sex.

0.05) and 35-40° N (x“ - 18.25, df = 1, P <
0.001), but sexes were evenly distributed
south of 30° N latitude (Table 1). Thus, male
Ruby-crowned Kinglets winter farther north
than females.

ANCOVA revealed that slopes for equa-
tions describing observed and predicted per-
cent males in the spring migrant population
differed significantly (F, 4 = 9.46, P = 0.037).
Least-squares regression yielded the following
equations;

arcsin Predicted % Males = 65.76 —
5.22(Time Interval) (n = 4, P- = 0.99, P <
0.001)

arcsin Observed % Males = 111.05 —
19.54(Time Interval) (n = 4, P- = 0.89, P =

0.05)

where Time Interval refers to successive 5-
day intervals (1-4) from first (15 April) to last
(2 May) capture of migratory males. For 20-
24 April a higher percentage of males was
captured than predicted (Fisher’s exact test, P
— 0.039) and for 30 April-4 May a lower
percentage of males was captured than pre-
dicted (Fisher’s exact test, P = 0.04); signifi-
cant differences were not detected for other
intervals. These data indicate that migratory
passage of the sexes during the early season
is more biased toward males than expected on
the basis of their wintering distribution.

Rates of migration can also be estimated for
males and females by dividing the distance
between two successive 5° latitude zones (as-
suming 1° equals 1 1 1 km) by the number of
days difference between median passage dates
for each zone as determined from BBL data
(Ellegren 1990, Woodrey and Chandler 1997).
Estimating migration rates in this manner
shows that both males and females increase
migration speed as they near their breeding

grounds. For latitudes south of our study sites
migration rates for males and females were 74
and 56 km day', respectively, but for all lat-
itudes overall migration speed was 83 km
day' for both sexes.

Male kinglets were significantly heavier
than females (t,02 = 5.76, P < 0.001) in
spring. Visible fat scores did not differ signif-
icantly between males and females in spring,
but spring males had significantly longer
wings (^102 = 9.49, P < 0.001) and signifi-
cantly higher condition index (Mann- Whitney
test: Z102 = 3.20, P = 0.001) than spring fe-
males (Table 3). When age classes were
pooled by sex in fall, males were significantly
heavier than females ((,30 = 4.80, P < 0.001)
and had significantly longer wings (r,3o
= 11.79, P < 0.001). Neither visible fat nor
condition index differed significantly between
males and females in fall when age classes
were pooled.

In fall, when AHY data after 30 September
were excluded from comparisons, AHY males
(^38 = 3.39, P = 0.002) and HY males (t,, =
3.93, P < 0.001) were both significantly
heavier than HY females. Similarly, when all
data after 30 September were excluded, AHY
males were significantly heavier than HY fe-
males (^24 = 3.035, P = 0.006). Mass did not
vary significantly among other sex/age classes
in fall (Table 3). Visible furcular fat scores
were significantly greater in AHY females
than in juveniles in fall when AHY data after
30 September were excluded from compari-
sons (Mann-Whitney test; Z41, = 3.14, P =
0.002; Z35 = 2.99, P = 0.003 for HY females
and HY males, respectively). Likewise, when
all data after 30 September were excluded
from analyses, AHY females carried signifi-
cantly more furcular fat than HY females
(Mann-Whitney test: Z34 = 2.840, P = 0.005).

THE WILSON BULLETIN • Vol. HI, No. I, March 1999

TABLE 3. Means (± SD) of mass, wing chord, fat class, and condition index (mass/wing chord) in migrant
Ruby-crowned Kinglets. Age classes (AHY = adults, HY = hatching-year) refer to fall migrants.

Sex/age class n Mass (g) Wing chord (nini) Furcular fat Abdominal fat Condition index

Spring males

6.6

-h

0.6

57.7

-F

1.3

2.2

-F

1.1

2.1

-F

1.4

1.15

-F

0.10

Spring females

6.0

-h

0.4

55.0

-F

1.6

2.1

-F

0.9

2.1

-F

1.0

1.09

0.07

Eall males

AHY

6.5

-F

0.6

58.5

-F

1.5

3.2

-F

0.8

3.2

-F

0.8

1.1 1

-F

0.10

HY (before 1 Oct.)"'

6.3

-F

0.4

57.6

-F

1.4

1.6

-F

1.1

1.6

-F

1.1

1.08

-F

0.06

HY (all data)*’

6.3

-F

0.4

57.6

-F

1.2

1.9

-F

1.0

1.8

-F

1.0

1.09

-F

0.07

All males'’

6.4

-F

0.4

57.7

-F

1.4

2.0

-F

1.1

2.1

-F

1.0

1.10

-F

0.07

Eall females

AHY

6.1

-F

0.4

55.1

-h

1.4

2.9

-F

0.7

2.8

-F

0.7

1.12

-F

0.07

HY (before 1 Oct.)"

5.9

-F

0.4

54.8

-F

1.2

2.0

-F

0.8

2.0

-F

0.7

1.08

-F

0.06

HY (all data)*’

5.9

-F

0.3

54.7

-F

1.4

2.0

-F

0.8

2.1

-F

0.8

1.08

-F

0.06

All females'

6.0

-F

0.3

54.9

-F

1.3

2.3

-F

0.8

2.3

-F

0.8

1.10

-F

0.06

“ Data for HY birds prior to 1 October only.

^ Data for HY birds over the entire fall season.

^ Includes birds with completely ossified skulls after 30 September that were of indeterminate age.

Visible abdominal fat scores were significant-
ly greater in AHY females than in HY birds
when AHY data after 30 September were ex-
cluded from comparisons (Z35 = 3.11, P =
0.002; Z49 = 2.84, P = 0.005 for HY males
and HY females, respectively). When all data
after 30 September were excluded from com-
parisons, AHY females had significantly more
abdominal fat than HY females (Z34 = 2.695,
P = 0.007). The same trend occurred for AHY
males, although differences were nonsignifi-
cant (Table 3). Neither furcular nor abdominal
fat varied significantly between males and fe-
males within the same age class. No signifi-
cant differences among age/sex classes were
detected for condition index in fall, although
when male and female data for each age class
were pooled the adult-juvenile difference ap-
proached significance = 1.74, P > 0.05
when AHY data after 30 September were ex-
cluded; ^43 = 1.85, P > 0.05 when all data
after 30 September were excluded).

Wing chord did not differ significantly
among age classes in fall, but males of both
age classes had significantly longer wings
than females when AHY data after 30 Sep-
tember were excluded (P < 0.001) and when
all data after 30 September were excluded (P
^ 0.005, Table 3). Test statistics for between
sex wing chord comparisons were; AHY
males vs. AHY females = 4.35), AHY
males vs. HY females = 5.59), HY males
vs. AHY females (/„ = 5.59), and HY males
vs. HY lemales = 7.94) for comparisons

with AHY data after 30 September excluded.
For comparisons with all data after 30 Sep-
tember excluded, between sex wing chord
comparison test statistics were: AHY males
vs. AHY females (r.g = 4.35), AHY males vs.
HY females (r24 = 5.99), HY males vs. AHY
females (6g = 3.23), and HY males vs. HY
females (U4 = 4.59).

DISCUSSION

Male Ruby-crowned Kinglets migrated ear-
lier in spring than females in both southeast-
ern South Dakota and across eastern North
America (Crim 1976, Fairfield and Shirokoff
1978, BBL data). Bird Banding Laboratory
data indicate that median passage dates for the
latitudes of our study sites (approximately
43° N) in eastern North America are 21-25
April for males and 6—10 May for females.
These dates agree very closely with median
passage dates for male and female kinglets
from Iowa, which occurred from 20-25 April
for males and from 5—10 May for females
(Crim 1976). The median date for males in
southeastern South Dakota was similar to me-
dian dates in Iowa and to median dates de-
rived from BBL data for 40-45° N latitude in
central and eastern North America (Table 2).
The median date for females in our study is
several days earlier than the median date from
Iowa and from the median date derived from
BBL data for the latitudes of our study sites.
However, capture effort was not quantified in
Crim (1976) and was not available for BBL

Swanson et al. • DIFFERENTIAL MIGRATION IN KINGLETS

data, so median passage dates may not be di-
rectly comparable.

Earlier spring passage of males than fe-
males is a common pattern among passerines,
presumably because of the advantages that
early arrival provides to the sex establishing
territory, which among passerines is usually
the male (Gauthreaux 1982, Francis and
Cooke 1986). Postponing arrival on the breed-
ing grounds until after males have established
territories might be beneficial to females if
they compete for mates (Francis and Cooke
1986). Moreover, because temperatures and
food availability increase throughout spring in
northern latitudes, late arrival on the breeding
grounds might also benefit females by provid-
ing more favorable conditions for breeding.
Consistent with this latter argument, male
kinglets were larger (i.e., heavier and longer
wings), had higher condition index, and were
more cold tolerant (Swanson and Dean 1999)
than females in spring.

Because juvenile Ruby-crowned Kinglets
may show complete skull ossification as early
as 1 October (Leberman 1970), documenta-
tion of differential migration patterns of age
classes in fall is problematic. In addition, we
found retrix shape (Pyle et al. 1987) to be un-
reliable for aging kinglets in our population.
Thus, we were unable to compare passage
dates for age classes during fall migration.
However, males migrated significantly later
than females during fall migration in south-
eastern South Dakota. This pattern of males
migrating later in the fall than females is con-
sistent with the observations of Fairfield and
Shirokoff (1978) for Ruby-crowned Kinglets
from Ontario.

Prescott (1980) found that adult female
kinglets were fatter than adult males in fall in
New Jersey, but sexes did not differ in fat
scores during fall migration in our study.
Adult kinglets in New Jersey (Prescott 1980)
had a higher percentage of individuals with
either no fat or heavy fat than juveniles, while
juveniles had higher percentages of individu-
als with intermediate fat loads. In our study
males were heavier than juvenile females in
fall, but not adult females. Moreover, adults
in our study carried more fat than juveniles in
fall, significantly so for females. Woodrey and
Moore (1997) reviewed several possible ex-
planations for elevated fat levels in adults

compared to juveniles during migratory stop-
over. These include: ( 1 ) less efficient foraging
by juveniles at stopover sites because of in-
experience or lower social status, (2) juveniles
carry less fat to increase mobility for escape
from predators because they may be more vul-
nerable to predation, and (3) possible differ-
ences in migratory routes or destinations af-
fecting fattening.

Proximate factors regulating differential
timing of migration among sex/age classes in-
clude differences in geographic origin, timing
of the onset of migration, and rates of migra-
tion (Chandler and Mulvihill 1990). Fairfield
and Shirokoff (1978) analyzed North Ameri-
can banding data from 1972-1975 and found
that male kinglets winter farther north than fe-
males on average. Bird Banding Laboratory
data from 1986-1995 also indicate that male
kinglets winter farther north than females (Ta-
ble 1) so different geographic origins un-
doubtedly contribute to differential migration
of the sexes in the spring. However, passage
of sexes during the early portion of spring mi-
gration is more biased toward males than ex-
pected on the basis of the differential winter-
ing distributions. Sample sizes from BBL data
for 1986-1995 for wintering populations of
kinglets in eastern and central North America
were relatively small, particularly south of
30° N latitude, even though kinglets regularly
winter south of this latitude (AOU 1983, In-
gold and Wallace 1994). Bird Banding Labo-
ratory data indicate approximately equal num-
bers of males and females at southern latitudes
in the wintering range, whereas females
should outnumber males in the southern part
of the wintering range, assuming equal sex ra-
tios on breeding grounds. This could influence
calculation of predicted percent males during
the latter part of the spring migration of males,
but should not affect comparisons during the
early portion of the migration. Because a
higher percentage of males was observed than
predicted based on wintering distribution dur-
ing the early part of spring migration, the
small sample size from southern portions of
the wintering range should not influence our
conclusions that wintering distributions do not
fully account for differential migration of the
sexes in spring.

This suggests that males either migrate at a
faster rate or initiate migration earlier than fe-

THE WILSON BULLETIN

Vol. Ill, No. 1, March 1999

males. Calculation of spring migratory rates
from BBL data revealed that male kinglets mi-
grate at a faster pace than females south of
45° N latitude, but that rates over the entire
migratory range do not differ between sexes.
Thus, elevated rates of migration in males rel-
ative to females may contribute to their early
arrival at our study sites, but not at higher
latitudes. This suggests that the onset of
spring migration is earlier in male Ruby-
crowned Kinglets than in females and that this
difference contributes substantially to differ-
ential spring migration of the sexes in this
species.

For fall migration, where all sex/age classes
presumably initiate migration from breeding
grounds, differential timing of migration
should reflect differences in onset or rate of
migration, assuming that sex/age classes ex-
hibit similar migration routes and that little
postbreeding dispersal away from breeding
sites occurs prior to southward migration.
Fairfield and Shirokoff (1978) suggested that
differential migration of the sexes in fall re-
sults from earlier departure of females from
breeding grounds for two possible reasons.
First, females are smaller and presumably less
hardy than males, so they might depart before
food availability and temperatures decline
markedly in the fall. However, cold tolerance
of male and female kinglets during fall mi-
gration through South Dakota did not differ
significantly (Swanson and Dean 1999). Sec-
ond, females might depart earlier than males
because they winter farther south and, there-
fore, must travel farther. If we assume that all
sex/age classes initiate migration from breed-
ing grounds, then differential migration of
sex/age classes might result not only from dif-
ferences in departure, but also from differences
in rates of migration (Chandler and Mulvihill
1990). Because Fairfield and Shirokoff (1978)
did not directly test whether early departure
or rates of migration differed between the sex-
es in fall, their suggestion must remain ten-
tative. In our study males migrated later than
females in fall but comparisons among age
classes were confounded because juvenile
kinglets could not safely be distinguished
from adults after 1 October (Leberman 1970).

ACKNOWLEDGMENTS

This work was supported, in part, by a grant from
the U..S. Fish and Wildlife Service Neotropical Migra-

tory Bird Program to DLS. We thank the many vol-
unteers who helped us with mist-netting, prominent
among them were M. Dutenhoffer, J. Martin, and D.
Weinacht. We also thank K. Olmstead for help with
statistics, and J. Ingold, M. Woodrey, and S. Morris
for helpful comments on an earlier version of this man-
uscript. J. Ingold also provided access to hard-to-find
literature. Finally, we thank K. Klimkiewicz of the
BBL for providing access to banding records for king-
lets.

LITERATURE CITED

American Ornithologists’ Union. 1983. Checklist of
North American birds, sixth ed. Allen Press,
Lawrence, Kansas.

Chandler, C. R. and R. S. Mulvihill. 1990. Inter-
preting differential timing of capture of sex clas-
ses during spring migration. J. Field Ornithol. 61:
85-89.

Crim, G. B. 1976. Kinglets in Iowa. Iowa Bird Life
76:45-47.

Ellegren, H. 1990. Autumn migration speed in Scan-
dinavian Bluethroats Luscinia .s. svecica. Ringing
& Migr. 11:121-131.

Fairfield, D. M. and P. A. Shirokoff. 1978. Migra-
tory patterns and winter distribution in the Ruby-
crowned Kinglet (Regiilus calendula). Blue Bill
(suppl.) 25:22-25.

Francis, C. M. and F Cooke. 1986. Differential tim-
ing of spring migration in wood warblers (Paru-
linae). Auk 103:548-556.

Gauthreaux, S. a., Jr. 1982. The ecology and evo-
lution of avian migration systems. Avian Biol. 6:
93-168.

Hall, G. A. 1981. Fall migration patterns of wood
warblers in the southern Appalachians. J. Field
Ornithol. 52:43-49.

Helms, C. W. and W. H. Drury, Jr. 1960. Winter and
migratory weight and fat field studies on some
North American buntings. Bird-Banding 3 1 : 1-40.
Hussell, D. j. T. 1991. Fall migrations of Alder and
Willow flycatchers in southern Ontario. J. Field
Ornithol. 62:260-270.

Ingold, J. L. and G. E. Wallace. 1994. Ruby-
crowned Kinglet (Regains calendula), no. 1 19. In
The birds of North America (A. Poole and F. Gill,
Eds.). The Academy of Natural Sciences, Phila-
delphia, Pennsylvania; The American Ornitholo-
gists’ Union, Washington, D.C.

Ketter.son, E. D. and V. Nolan, Jr. 1976. Geographic
variation and its climatic correlates in the sex ratio
of eastern-wintering Dark-eyed Juncos (.lunco
hyemalis hyemalis). Ecology 57:679-693.
Leberman, R. C. 1970. Pattern and timing of skull
pneumatization in the Ruby-crowned Kinglet-.
Bird-Banding 41: 121-124.

Lepthein, L. W. and C. E. Bock. 1976. Winter abun-
dance patterns of North American kinglets. Wil-
•son Bull. 88:483-485.

Morton, M. L. 1984. Sex and age ratios in wintering
White-crowned Sparrows. Condor 86:85-87.

Swanson el al. • DIFFERENTIAL MIGRATION IN KINGLETS

Myers, J. P. 1981. A test of three hypotheses for lat-
itudinal segregation of the sexes in wintering
birds. Can. J. Zool. 59:1527-1534.

Nolan, V., Jr. and E. D. Ketterson. 1990. Timing of
autumn migration and its relation to winter distri-
bution in Dark-eyed Juncos. Ecology 71:1267-
1278.

Prescott, D. R. C. and A. L. A. Middleton. 1990.
Age and sex differences in winter distribution ot
American Goldfinches in eastern North America.
Omis Scand. 21:99-104.

Prescott, K. W. 1980. Weight, fat class, and wing
measurements of Ruby-crowned Kinglets during
migration. Inland Bird Banding 52:1—7.

Pyle. P, S. N. G. Howell, R. P. Yunick, and D. F.
DeSante. 1987. Identification guide to North
American passerines. Slate Creek Press, Bolinas,
California.

Ramos, M. A. 1988. Eco-evolutionary aspects of bird
movements in the northern Neotropical region.
Proc. Int. Omithol. Congr. 19:251-293.

Rice, W. R. 1989. Analyzing tables of statistical tests.
Evolution 43:223-225.

Root, T. 1988. Atlas of wintering North American
birds. Univ. of Chicago Press, Chicago, Illinois.

SAS Institute. 1988. SAS user’s guide. SAS Institute,
Inc., Cary, North Carolina.

SOKAL, R. R. AND F. J. Rohle. 1995. Biometry, third
ed. Freeman, New York.

South Dakota Ornithologists’ Union. 1991. The
birds of South Dakota, second ed. Northern State
Univ. Press, Aberdeen, South Dakota.

Swanson, D. L. and K. L. Dean. 1999. Migration-
induced variation in thermogenic capacity in mi-
gratory passerines. J. Avian Biol. 30:In press.

Winker, K. 1995. Autumn stopover on the Isthmus ol
Tehuantepec by woodland Nearctic-Neotropic mi-
grants. Auk 112:690—700.

Winker, K. and J. H. Rappole. 1992. The autumn pas-
sage of Yellow-bellied Flycatchers in South Texas.
Condor 94:526-529.

WOODREY, M. S. AND C. R. CHANDLER. 1997. Age-
related timing of migration: geographic and inter-
specific patterns. Wilson Bull. 109:52-67.

WooDREY, M. S. AND F. R. MooRE. 1997. Age-related
differences in the stopover of fall landbird mi-
grants on the coast of Alabama. Auk 1 14:695—
707.

Zar, j. H. 1996. Biostatistical analysis, third ed. Pren-
tice Hall, Upper Saddle River, New Jersey.

Wilson Bull., 111(1), 1999, pp. 70-75

SCALE-DEPENDENT HABITAT SELECTION BY AMERICAN
REDSTARTS IN ASPEN-DOMINATED FOREST FRAGMENTS

NAVJOT S. SODHI,' 23 CYNTHIA A. PASZKOWSKI,' AND SHANNON KEEHN'

ABSTRACT. — We examined scale-dependent site occupancy of American Redstarts {Setophaga ruticilla) in
forest fragments (2-140 ha in area) dominated by mature trembling aspen (Populus tremuloides) in central
Alberta, Canada. Vegetation within territories, both adjacent to and away from nests, differed from vegetation
in unoccupied fragments and within occupied fragments outside of territories. Territories contained higher den-
sities of willow (Salix sp.), a taller shrub-layer, and lower densities of trembling aspen than other sites. Willow
was the most frequently used plant species for nesting and foraging. Our results indicate that even within mature
forest patches, American Redstarts select disturbed areas dominated by early successional plant species. Received
6 March 1998, accepted 30 Aug. 1998.

Numerous researchers have investigated
habitat selection by American Redstarts {Se-
tophaga ruticilla), but debate continues re-
garding the preferred age of forest stands used
for breeding. Although most studies reported
that American Redstarts prefer early succes-
sional forests (Martin 1960, Collins et al.
1982, DeGraaf 1991, Westworth and Telfer
1993, Huffman 1997), they have also been
commonly found in mid- to late successional
stands (Bond 1957, Ficken and Ficken 1967,
Crawford et al. 1981, Morgan and Freedman
1986, Thompson and Capen 1988). Scale-de-
pendent habitat requirements have been poor-
ly documented for the American Redstart. Be-
cause abiotic and biotic factors may affect a
species differently at various spatial scales
(Bock 1987, Wiens et al. 1987), some docu-
mented differences in habitat selection may be
due to scale effects. Thus, a holistic under-
standing of a species’ habitat use patterns may
require that data be collected and analyzed at
several spatial scales (e.g., patch, territory,
and nest site; Bergin 1992, Steele 1992, Kelly
1993). Considering that many forest patches
are heterogeneous in age and structure (e.g.,
woody vegetation may increase in age and
height from edge to interior), the investigation
of scale-dependent habitat use patterns of the
American Redstart could clarify the species’
requirements and reconcile some of the incon-
sistencies in reported habitat preferences.

' Dept, of Biological Sciences, Univ. of Alberta, Ed-
monton, AB. Canada T6G 2E9.

^ Prc.sent address: Dept, of Biological Sciences, Na-
tional Univ. of Singapore, Kent Ridge, Singapore
1 19260: Email: dbsns(®mis.edu.sg
' Corresponding author.

We studied scale-dependent habitat selec-
tion of the American Redstart in forest frag-
ments dominated by mature trembling aspen
{Populus tremuloides) at the northwestern
edge of the species’ range in central Alberta,
Canada. We asked: (1) do American Redstarts
occupy available forest fragments based on
particular habitat characteristics (vegetation
structure and composition), and (2) within an
occupied fragment, do American Redstarts se-
lect territories and nest sites based on habitat
characteristics?

STUDY AREA AND METHODS

The study was conducted around the Meanook Bi-
ological Research Station (54° 37' N, 1 13° 20' W) near
Athabasca, Alberta, between 9 May and 6 July 1994.
Upland forests in this region are primarily a mosaic of
trembling aspen and white spruce (Picea glauca)
stands which was historically created and maintained
by fire (Rowe 1972). We examined eight forest frag-
ments dominated by mature (>80 yr old) trembling
aspen interspersed with patches of willow {Salix sp.)
and alder (Alnus spp.). A fragment was defined as a
wooded area separated from other wooded areas by
more than 30 m of cropland or pasture (see Villard et
al. 1995). Eorest fragments varied from 2 to 140 ha
(areas were calculated from 1:30,000 aerial photo-
graphs using a Placon® digital planimeter). All frag-
ments were located within a 9-km^ area and repre-
sented the range of fragment sizes available in the
landscape (for details see Sodhi and Paszkowski 1997).
The selection of fragments was constrained by various
factors including stand age, access, presence of cattle,
and landowner cooperation.

Each fragment was flagged into 100 X 100 m grids
and was surveyed three times to locate redstart terri-
tories. During surveys, we walked along the flagged
gridlines and visually located all redstarts (primarily
by following singing males). Locations of redstarts,
along with characterization of individual plumage pat-
terns (see below) were recorded on a gridded map.

Sodhi et al. • REDSTART HABITAT SELECTION

Unoccupied fragmenls were surveyed again in late
June by playing American Redstart territorial songs I'or
2 min at 50-m intervals and waiting 5 min tor a re-
sponse. No new individuals were detected during these
surveys. Individual males were identified based on lo-
cation of ten itory and on patterns of plumage colora-
tion on the breast, abdomen, neck, and head (Sherry
and Holmes 1989, Lemon et al. 1992). Because second
year male redstarts can differ from older males in hab-
itat selection (e.g.. Sherry and Holmes 1989, Hunt
1996), we collected habitat data only for males in full
adult plumage.

The composition of vegetation was measured within
territories, outside of territories, adjacent to the nest
tree, and in unoccupied habitat fragments. We sampled
in early to mid-July after the breeding season. Vege-
tation plots were located at random with the following
exceptions: plots a,ssociated with nest trees were dic-
tated by nest location and plots associated with teiri-
tories were always placed at least 10 m inside a bound-
ary. A total of 88 vegetation plots were sampled: 33
within 13 territories, 17 outside of temtories but within
4 occupied fragments, 17 immediately around 17 nest
trees, and 21 plots in 4 unoccupied fragments.

We sampled vegetation using a modification of the
circular sample-plot method (James and Shugart
1970). For each of the four site types, we recorded
four sets of information: (1) tree (woody plants more
than 1.75 m tall) species, number, height (using a cli-
nometer), and diameter at breast height (using a dbh
measuring tape) within a 22.4 m diameter circular plot,
(2) shrub (woody plant less than 1.75 m tall) species,
number, and maximum height within a 10 m diameter
circular plot. (3) presence/absence of canopy (above 5
m) at 20 random points within the 22.4 m diameter
plot using a 4 cm diameter ocular tube, and (4) pres-
ence/absence of ground cover at the same 20 random
points using the ocular tube. The last two sets of in-
formation gave a measure of relative canopy and
ground cover, respectively.

We used Principal Components Analysis (PCA) to
compare vegetation composition within and among the
four site-types (CANOCO; ter Braak 1991) tor all 88
plots. Principal Components Analysis is recommended
and frequently used for analysis of habitat data (Don-
caster et al. 1996, Hunt 1996). The 27 vegetation var-
iables recorded for each plot (Table 1) were square-
root transformed to approximate a normal distribution.
A univariate analysis, Kruskal-Wallis test, was used to
determine if PCA scores for the first two ordination
axes differed among site types. If the Kruskal-Wallis
test was significant (P < 0.05), Multiple Comparisons
tests were performed to determine which characteris-
tics differed (Siegel and Castellan 1988). All statistical
tests were done using STATVIEW version 4.1 on a
Macintosh-compatible computer.

RESULTS

Redstarts occurred in four forest fragments
(2, 50, 107 and 140 ha in area) but were ab-

sent from four others (4, 6, 9 and 32 ha). Prin-
cipal Components Analysis based on vegeta-
tion characteristics clearly grouped and sepa-
rated unoccupied plots from plots within ter-
ritories (Fig. 1). Loadings of vegetation
variables on PCA Axis 1 indicated that unoc-
cupied fragments and areas outside of terri-
tories within occupied fragments contained
larger trees and higher densities of trembling
aspen, accompanied by prickly wild rose
(Rosa acicularis), red-osier dogwood {Cornus
stolonifera), and honeysuckle (Lonicera spp.)
(Table 1). Areas within territories were char-
acterized by high densities of willow (both
trees and saplings), cherry (Prunus sp.), sas-
katoon berry (Amelanchier alnifolia) and bal-
sam poplar {Populus balsamifera) saplings, as
well a tall shrub-layer (loadings on Axis 1,
Table 1). Scores on Axis I differed signifi-
cantly among site types {KW = 42.51, df =
3, P < 0.001). There was no significant dif-
ference between nest sites and plots within
territories (Multiple Comparisons test: P >
0.05), but both differed significantly from
plots in unoccupied fragments and plots out-
side of territories within occupied fragments
(Multiple Comparisons tests: P < 0.05). Plots
in unoccupied fragments and those outside of
territories in occupied fragments had similar
PCA I scores (Multiple Comparisons test: P
> 0.05).

Nest sites were associated with high den-
sities of white birch {Betula papyrifera, both
trees and saplings), willow, cherry, and bal-
sam poplar saplings, as well as Viburnum
spp., which contributed to a tall shrub layer
(high scores and loadings on Axis II). Nest
sites were also clustered on PCA Axis II {KW
= 17.79, df = 3, P < 0.001). On this axis,
only scores for nest sites differed significantly
from scores for plots from unoccupied frag-
ments (Multiple Comparisons tests, P < 0.05).
However, there was similarity in vegetation
characteristics in some plots within territories
and those at nest sites (Fig. 1).

DISCUSSION

Principal Components Analysis and subse-
quent univariate analyses indicate that red-
starts did not establish territories randomly
within or among forest patches, but based
their occupancy on vegetation characteristics.
A high abundance of willow was a particular-

THE WILSON BULLETIN • Vol. HI, No. I, March 1999

TABLE 1. Loadings of vegetation variables on
PC II).

the first and second Principal Components axes (PC I and

Variables

PC 1

PC 11

Eigenvalues

0.42

0.12

Variation explained (%)

42.4

11.1

No. trees (woody plants > 1.75 ni tall)

Willow’ (Sali.x sp.)

-121

-289

Trembling aspen (Populit.s tremuloides)

338

-654

Balsam poplar {P. halsamifera)

White birch (Betula papyrifera)

-101

556

Alder (Alnu.s spp.)

195

White spruce {Piceti glauca)

133

Tree sizes

Mean tree height (m)

-95

-68

Mean dbh (mm)

203

318

No. shrubs (woody plants 0.5-1.75 m tall)

Red-osier dogwood (Cornu.'i .stolonifera)

224

106

Wild gooseberry (Rihe.s o.xyacanthoide.'i)

-55

103

Wild red raspberry (Ruhii.s idaeu.s)

-229

Honeysuckle (Lonicera spp.)

183

-418

Prickly wild rose {Rosa acicidaris)

996

Snowberry (Symphoricarpos spp.)

203

Caragana sp.

-74

-121

Viburnum spp.

522

Primus spp.

-223

448

Saskatoon berry (Amelanchier alnifolia)

-139

-5

Maple sapling {Acer sp.)

-11

Balsam poplar sapling

-122

266

Willow .sapling

-158

380

Trembling a.spen sapling

196

White birch sapling

-119

769

Alder sapling

142

172

Shrub height

Mean maximum shrub height (mm)

-206

563

Cover'

Canopy

-97

177

Ground

-37

^ See Methods for estimation of these variables.

ly distinctive feature of areas within territories
and around nest sites. Willow was the most
frequently used foraging site (mean ± SE =
65.8 ± 8.9% of observation time, n ~ 13
males for which we had at least 10 min of
observations) and nesting site (13 of 20 nests;
Sodhi and Paszkowski, unpubl. data). Areas
within territories and nest sites were charac-
terized not only by willows, but also by a tall
shrub layer in general. American Redstarts
have been found to be positively associated
with shrub cover in Minnesota (Huffman
1997). The extra cover and structural com-
plexity provided by taller shrubs may have
protected nests from predation or parasitism
and thus enhanced reproductive success (Mor-

ris and Lemon 1987). Nest sites also support-
ed higher densities of white birches, which
have wider leaves than willows and may im-
prove nest cover. In our study area, only 18%
{n = 22) of the nests were parasitized by
Brown-headed Cowbirds (Molothrus ater) and
only 18% {n = 22) were depredated (Sodhi
and Paszkowski, unpubl. data). Both nest par-
asitism and depredation rates at our sites were
relatively low compared to other studies (e.g.,
Freedman 1929, Sherry and Holmes 1992)!
The structural complexity of understory veg-
etation in earlier successional stages could of-
fer appropriate nest protection for redstarts.

Certain plant species, most notably trem-
bling aspen, were underrepresented on sites

Sodhi et al. • REDSTART HABITAT SELECTION

400

300

200

(/)

100 --

-100

-200

-300 -I—
-250

ONest Site (n = 17)

□ Within Territories (n = 33)
^Outside Territories (n = 17)

• Unoccupied Fragments (n = 21)

-200 -150 -100 -50 0 50 100 150 200 250

PC I Score

FIG. 1 . Comparison of sites occupied and unoccupied by American Redstarts based on the first two Principal
Component scores of plot vegetation. The Principal Components are based on 27 vegetation variables collected
from 88 plots. Note: one plot near a nest was the only plot dominated by conifers and it contained little
undergrowth, therefore it is distinct from all other plots.

occupied by redstarts. The lack of use of hab-
itat dominated by common, and even domi-
nant, tree and shrub species at both the scale
of the forest patch and the territory may be
related to constraints imposed on redstart for-
aging behavior by the leaf and branch mor-
phology of these plant species (Holmes et al.
1978, Sedgewick and Knopf 1992). The ex-
clusion and inclusion of certain shrub species
on redstart territories could also be an artifact
of their co-occurrence with overstory domi-
nants, especially trembling aspen. The canopy
forming woody species associated with habitat
used by redstarts, i.e., willow and white birch,
are all species typically found on wetter sites
in central Alberta (Rowe 1972). Thus, it is
also possible that redstarts were choosing lo-
cations in this relatively dry landscape that
had higher soil moisture levels and were
therefore more productive in terms of plant
and invertebrate biomass (Adams and Morri-
son 1993). Our surveys showed that even
within stands dominated by mature trembling

aspen, redstarts prefered areas that supported
more willows, which are typically early suc-
cessional species. The presence of thickets of
willow and other shrubs in these stands might
be maintained by periodic localized flooding
resulting from vernal soil saturation following
heavy snow cover or from beaver activity (So-
dhi and Paszkowski, pers. obs.).

American Redstarts are abundant in har-
vested aspen forests in Minnesota (Hoffman
1997). Other studies also show that redstarts
attain maximum densities in stands in early
successional stages of different forest types
(DeGraaf 1991, Westworth and Telfer 1993,
Hunt 1996). It might be argued, based on pat-
terns reported for some other passerines
(Vickery et al. 1992), that although redstarts
occur at high densities in early successional
forests, they reproduce poorly here compared
to mid- to late successional forests. We found
that 90.6% (n = 32) of older adult redstart
males were paired in our study area in 1994
(Sodhi, unpubl. data); this figure is similar to

THE WILSON BULLETIN

Vol. ///, No. /, March 1999

values for older males in various successional
forests and in a continuous forest tract in New
Hampshire (Sherry and Holmes 1989, Hunt
1996). Assuming that male pairing success is
correlated with reproductive success, pockets
of disturbed, productive early successional
vegetation nested within mature forest stands
may offer good breeding conditions for red-
starts.

Every avian habitat can be represented as a
spatially based hierarchy that ranges from the
level of landscape to nest site. The behavioral
decision-making processes behind habitat use
can be related to these hierarchical habitat
units (Kolasa 1989, Kotliar and Wiens 1990).
Studies of scale-dependent habitat selection
have revealed that the decision-making pro-
cess varies among species and can sometimes
operate at multiple scales within a single spe-
cies (Bergin 1992). For example. Black-
throated Blue Warblers (Dendroica ccientles-
cens) were more selective at the habitat patch
(or stand) level than at the territory level
(Steele 1992). In contrast, habitat selection in
Dusky Flycathers {Empidoncix oberholseri)
was primarily based on nest-site characteris-
tics rather than on territorial features (Kelly

1993) . In our study, habitat selection by
American Redstarts operated most clearly at
the territory level. Males defended areas of
forest with a distinct plant composition and
structure. Habitat selection at the territorial
level was manifested, in turn, at the next high-
er spatial scale, as males did not occupy frag-
ments that contained insufficient appropriate
vegetation. At the finest spatial scale, evi-
dence for nest site selection was present but
relatively weak. Nest sites were not signifi-
cantly different from other locations within
territories, possibly because the willows and
other tall shrubs used as foraging sites also
offered good nest protection.

Some que.stion remains as to whether birds
actually differentiate between potential terri-
tories ba.sed on the environmental parameters
that researchers perceive to be important
(Morse 1989). Most authors agree however,
that, at some level, territory establishment is
a behavioral response to certain vegetation
characteristics (Maurer and Whitmore 1981,
Smith and Shugart 1987, Parrish and Sherry

1994) . We concur that American Redstarts in
central Alberta establish territories within a

definable type of vegetation that appears to
offer appropriate conditions for successful for-
aging and nesting. Based on our research, as
well as previous studies, redstarts appear to
prefer early successional forest stands or dis-
turbed sites embedded within older stands.
However, with correlative data, it remains
possible that redstarts choose sites based on
soil moisture or food availability, which are
in turn associated with particular vegetation
characteristics. In light of reported population
declines for the species in parts of its range
(Sauer et al. 1996), the value of such habitats
should be assessed on a regional basis and
their use by redstarts integrated into conser-
vation strategies.

ACKNOWLEDGMENTS

We thank S. Jamieson for field assistance. This
study was supported by the Alberta Sports, Recreation,
Parks and Wildlife Foundation, Canadian Circumpolar
Institute, Natural Sciences and Engineering Research
Council of Canada (grant to CAP), and James Ander-
son McAfee Postdoctoral Fellowship (to NSS). We
thank two anonymous reviewers for making construc-
tive comments on an earlier draft.

LITERATURE CITED

Adams. E. M. and M. L. Morrison. 1993. Effects of
forest stand structure and composition on Red-
breasted Nuthatches and Brown Creepers. J.
Wildl. Manage. 57:616-629.

Bergin, T. 1992. Habitat selection by the Western
Kingbird in western Nebraska: a hierarchical anal-
ysis. Condor 94:903-91 1.

Bock, C. 1987. Distribution-abundance relationships
of some Arizona land birds: a matter of scale?
Ecology 68:124-129.

Bond, R. R. 1957. Ecological distribution of breeding
birds in upland forests of southern Wisconsin.
Ecol. Monogr. 27:351-384.

Collins, S. L., E C. James, and P. G. Ris.ser. 1982.
Habitat relationships of wood warblers (Parulidae)
in northern central Minnesota. Oikos 39:50-58.
Crawford, H. S., R. G. Hooper, and R. W. Titter-
INGTON. 1981. Songbird population response to
silvicultural practices in central Appalachian hard-
woods. J. Wildl. Manage. 45:680-692.

DeGraaf, R. M. 1991. Breeding bird assemblages in
managed northern hardwood forests in- New Eng-
land. Pp. 153-171 in Wildlife and habitats in man-
aged landscapes (J. E. Rodiek and E. G. Bolen,
Eds.). Island Press, Covelo, California.

Donca.ster, C. R, T. Micol, and S. P. Jensen. 1996.
Determining minimum habitat requirements in
theory and practice. Oikos 75:335-339.

Ficken, M. S. and R. W. Ficken. 1967. Age-specific
differences in the br'eeding behavior and ecology

Sodhi et al. • REDSTART HABITAT SELECTION

of the American Redstart. Wilson Bull. 79:188-
199.

Freedman, H. 1929. The cowbird: a study of social
parasitism. Charles C. Thomas, Springfield, Illi-
nois.

Holmes, R. T, T. W. Sherry, and S. E. Bennett.
1978. Diurnal and individual variability in the for-
aging behavior of American Redstarts. Oecologia
36:141-149.

Huffman, R. D. 1997. Effects of residual overstory on
bird use and aspen regeneration in aspen harvest
sites in Tamarac National Wildlife Refuge, Min-
nesota. M.Sc. thesis. West Virginia Univ., Mor-
gantown.

Hunt, R D. 1996. Habitat selection by American Red-
starts along a successional gradient in northern
hardwoods forest: evaluation of habitat quality.
Auk 113:875-888.

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

Kelly, J. P. 1993. The effect of nest predation on hab-
itat selection by Dusky Flycatchers in limber pine-
juniper woodland. Condor 95:83—93.

Kolasa, J. 1989. Ecological systems in hierarchical
perspective: breaks in community structure and
other consequences. Ecology 70:36—47.

Kotliar, N. B. and j. a. Wiens. 1990. Multiple scales
of patchiness and patch structure: a hierarchical
framework for the study of heterogeneity. Oikos
59:253-260.

Lemon, R. E., D. M. Weary, and K. J. Norris. 1992.
Male morphology and behavior correlates with re-
productive success in the American Redstart (Se-
tophaga ruticilla). Behav. Ecol. Sociobiol. 29:
399-403.

Martin, N. D. 1960. An analysis of bird populations
in relation to forest succession in Algonquin Pro-
vincial Park, Ontario. Ecology 41:126-140.

Maurer, B. A. and R. C. Whitmore. 1981. Foraging
of five bird species in two forests with different
vegetation structure. Wilson Bull. 93:478—490.

Morgan, K. and B. Freedman. 1986. Breeding bird
communities in a hardwood forest succession in
Nova Scotia. Can. Field-Nat. 100:506—519.

Morris, M. M. J. and R. E. Lemon. 1987. American
Redstart nest placement in southwestern New
Brunswick. Can. J. Zool. 66:212—216.

Morse, D. H. 1989. American warblers. Harvard Univ.
Press, Cambridge, Massachusetts.

Parrish, J. D. and T. W. Sherry. 1994. Sexual seg-
regation by American Redstarts wintering in Ja-
maica: importance of resource seasonality. Auk
111:38-49.

Rowe, J. F. 1972. Forest regions of Canada. Depart-

ment of Environment. Can. For. Serv. Publ. 1300:
1-171.

Sauer, J. R., S. Schwartz, B. G. Peterjohn, J. E.
Hines, and S. Droege. 1996. The North American
breeding bird survey. Version 95.1. Patuxent
Wildlife Research Center, Laurel, Maryland. URL
= http://www.nbs. go v./bbs/bbs.html.

Sedgewick, j. a. and F L. Knopf. 1992. Describing
Willow Flycatcher habitats: scale perspectives and
gender differences. Condor 94:720-733.

Siegel, S. and N. J. Castellan, Jr. 1988. Nonpara-
metric statistics for the behavioral sciences. Mc-
Graw-Hill Book Company, New York.

Sherry, T. W. and R. T. Holmes. 1989. Age-specific
social dominance affects habitat use by breeding
American Redstarts (Setophaga ruticilla): a re-
moval experiment. Behav. Ecol. Sociobiol. 25:
327-333.

Sherry, T. W. and R. T. Holmes. 1992. Population
fluctuation in a long-distance neotropical migrant:
demographic evidence for the importance of
breeding season events in the American Redstart.
Pp. 431-442 in Ecology and conservation of neo-
tropical migrant landbirds (J. M. Hagan and D. W.
Johnston, Eds.). Smithsonian Institution Press,
Washington, D.C.

Smith, T. M. and H. H. Shugart. 1987. Territory size
variation in the Ovenbird: the role of habitat struc-
ture. Ecology 68:695-704.

Sodhi, N. S. and C. A. Paszkowski. 1997. The pairing
success of male Black-and-white Warblers, Mni-
otilta varia, in forest fragments and a continuous
forest. Can. Field-Nat. 111:457-458.

Steele, B. B. 1992. Habitat selection by breeding
Black-throated Blue Warblers at two spatial
scales. Ornis Scand. 23:33-42.

TER Braak, C. j. E 1991. Program CANOCO, version
3.2. Agriculture Mathematics Group, DLO, AC
Wageningen, The Netherlands.

Thompson, F. R., Ill and D. E. Capen. 1988. Avian
assemblages in serai stages of a Vermont forest.
J. Wildl. Manage. 52:771-777.

Vickery, P. D., M. L. Hunter. Jr., and J. W. Wells.
1992. Is density an indicator of breeding success?
Auk 109:706-710.

ViLLARD, M-A., G. Merriam. and B. a. Maurer.
1995. Dynamics in subdivided populations of neo-
tropical migratory birds in fragmented temperate
forest. Ecology 76:27—40.

Wiens, J. A.. J. T. Rotenberry, and B. Van Horne.
1987. Habitat occupancy patterns of North Amer-
ican shnibstcppe birds: the effect of spatial scale.
Oikos 48:132-147.

Westworth, D. a. and E. S. Telfer. 1993. Summer
and winter bird populations associated with five
age-classes of aspen forest in Alberta. Can. J. For-
est. Res. 23:1830-1836.

Wilson Bull., 1 1 1(1), 1999, pp. 76-83

FEMALE MATE CHOICE IN NORTHERN CARDINALS: IS THERE
A PREFERENCE FOR REDDER MALES?

L. LAREESA WOLEENBARGER* ^

ABSTRACT. — 1 tested whether female Northern Cardinals iCardinalis cardinalis) associated with redder
males in two laboratory experiments, one using males with unaltered plumage and the other using males with
plumage altered by a lightened or reddened treatment. Females exhibited no preference for redder males. Given
the long duration over which pair formation can occur in natural populations and the importance of territory
quality to reproductive success, a female may choose a mate based on other morphological characteristics or
aspects of his territory rather than on only red coloration. Received 27 May 1998, accepted 22 Oct. 1998.

Ornate plumage characteristics in birds are
generally assumed to have arisen through sex-
ual selection and, specifically, through female
mating preferences or competitive interactions
between males (Darwin 1871, Andersson
1994). A growing number of studies has dem-
onstrated that females prefer males with
brighter coloration (Hill 1990, Sastre et al.
1994, Sundberg 1995) as well as with other
exaggerated plumage characteristics, such as
longer crests (Jones and Hunter 1993) and
longer tails (Andersson 1982, 1992; Mpller
1988, 1992; Evans and Hatch well 1992).

Northern Cardinals {Cardinalis cardinalis)
are highly dichromatic. Males vary from or-
ange to scarlet red. Male coloration varies
both in overall hue and in evenness of breast
coloration. In contrast, females are primarily
a light brown but have small and variable
amounts of red in the crest and breast; all fe-
males also have red coloration on the wings
and tail.

In addition to being highly dichromatic,
cardinals are territorial and socially monoga-
mous. Males feed females substantially during
nest construction, a time coinciding with egg
production (Kinser 1973). Males also feed
nestlings at higher rates than do females (Fil-
liater and Breitwi.sch 1997, but see Linville et
al. 1998). Because coloration derived from ca-
rotenoids is dependent on diet in birds (Good-
win 1950, Hill 1992), red coloration in car-
dinals could signal information to the female

' Section of Neurobiology and Behavior, Cornell
Univ., Ithaca, NY 14853-2702.

^ Present address: Dept, of Biology, Univ. of Mary-
land. College Park. MD 20742;

E-mail: LW 1 37@umail.umd.edu

regarding a male’s foraging abilities, age, or
overall phenotypic quality.

In this study, I address whether female pref-
erences may account for the maintenance of
red coloration in male Northern Cardinals. I
used two laboratory experiments to test
whether female cardinals preferred to associ-
ate with redder males during the breeding sea-
son.

METHODS

The experiments were conducted on the Cornell
University campus, Ithaca, New York, from mid Feb-
ruary through April 1995. Males and females used in
the experiments were captured between 14 January
and 16 April 1995 using baited traps and mist nets at
6 sites within Tompkins County, New York (42° N,
76° W). Individuals were marked uniquely with one
color band. Because previous studies have demonstrat-
ed that red color bands can influence behavior of in-
dividuals in other species (Burley et al. 1982, Hagan
and Reed 1988, Metz and Weatherhead 1991), I did
not use red or orange color bands.

Prior to trials males were housed in two indoor flight
aviaries (3.6 X 3.6 X 3.0 m) with skylights and incan-
descent lighting synchronized to dawn and dusk. Fe-
males were housed in a similar third aviary (5.2 X 3.6
X 3.0 m). Aviary rooms were both visually and acous-
tically isolated from each other. The temperature of
each aviary ranged between 10 and 12° C. Food and
water were provided ad libitum.

Measuring coloration. — To measure male colora-
tion, I used methods described elsewhere (Wolfenbar-
ger in press). Briefly, I used the color chip series of
the Methuen Handbook of Color (Kornerup 1967) to
quantify breast coloration of males. The color chip se-
ries provides a measure of three components of color:
hue, tone, and intensity. The hue component ranked
color on a scale from yellow (5) to intense red (11).
The tone component quantified the amount of black
present 1 1 (all black) to 6 (none)], the intensity com-
ponent indicated the degree of saturation of pigment
[from little ( 1 ) to complete (8)|. I used a grid that
divided the breast into eight 1 X 4 em rectangles and

Wolfenhcirfier • RED COLORATION AND FEMALE PREFERENCE

TABLE 1. Male color and morphology differences in

male cardinals used

in natural trials.

Relative color

score

Lower .f ± SD (n)

Higher .f ± .SD (h)

Z (Cf

Bright breast hue

73.9 ± 7.0 (8)

81.7 ± 4.3 (8)

2.81 (0.005)

Dull breast hue

65.9 ± 6.4 (8)

71.9 ± 4.0 (8)

2.81 (0.005)

Bright breast intensity

62.9 ± 0.9 (3)

63.2 ± 0.6 (3)

Dull breast intensity

61.1 ± 3.1 (7)

62.7 ± 0.9 (7)

2.64 (0.008)

Morphological traits (in mm):

Tarsus length

24.3 ± 0.7 (10)

24.1 ± 1.0 (10)

0.82 (> 0.05)

Crest length

36.8 ± 3.5 (10)

35.8 ± 1.9 (10)

0.46 (> 0.05)

Tail length

99.9 ± 5.0 (10)

98.4 ± 3.7 (10)

0.46 (> 0.05)

Black bib length

19.8 ± 4.6 (10)

20.2 ± 4.3 (10)

0.65 (> 0.05)

Black bib width

21.8 ± 1.1 (10)

22.0 ±1.1 (10)

0.46 (> 0.05)

“Z-value is from Wilcoxon matched pairs signed rank test.

^ In 7 trials, male scores for bright breast intensity were identical.

placed it immediately posterior to the black bib and
measured coloration in each of these regions. Because
male cardinals often had clumps of orange, yellow, or
even tan feathers interspersed within a background of
red feathers, I recorded the highest (“Bright” breast
score) and lowest color score (“Dull” breast score) for
hue, tone and intensity in each of the 8 regions. For
all of these components a higher score indicates a red-
der or brighter color. Among the birds I used, there
was greater variation among males in the color of dull
regions relative to bright regions (Wolfenbarger, in
press). I tested the repeatability of the color scoring
method by using specimens in the Cornell Vertebrate
Collections, and found the method to be highly re-
peatable (Wolfenbarger 1996).

For analyses of the association between measures of
female association and male coloration, I summed each
component of color (i.e., hue, tone, intensity) for the
eight regions and used separate scores for the bright
and dull color measurements. In this experiment, all
males except two exhibited the maximum tone scores
possible for both bright and dull breast measurements;
therefore, the relationship between color tone and fe-
male association was not tested. Similarly, there was
very little variation in bright breast intensity. There-
fore, I used three variables; bright breast hue, dull
breast hue, and dull breast intensity to characterize
male coloration.

Although other studies have created a composite
score from color chips (Linville et al. 1998), 1 used
these components separately for two reasons. Hue and
intensity were not consistently correlated in this pop-
ulation of cardinals (Wolfenbarger 1996), and combin-
ing these scores would result in identical scores for
males that actually had different hue and intensity
scores. Second, the relative importance of hue, inten-
sity and tone for color perception of cardinals is not
known. Any weighting of these components becomes
problematic for the interpretation of negative results
since one obvious alternative would be that the weight-
ing factor might be incorrect.

Morphological measurements. — 1 measured the fol-

lowing on males: tarsus length, tail length, crest length,
maximum length of black bib (after Mpller 1987), and
width of black bib. All were measured to the nearest
0.1 mm except tail length which was measured to the
nearest 0.5 mm.

Experiment I: natural plumage trials. — Ten trials
were conducted to test whether females spent more
time with males having higher natural plumage scores.
Males and females in each trial were captured at least
4 km apart to reduce the possibility that males and
females had interacted prior to the experiment. Males
and females were tested in the order that they were
captured from the field so that males spent similar
lengths of time in captivity (approximately 7 days for
natural trials and 18 days for manipulated trials), but
within a trial males and females had spent similar
amounts of time in captivity. No apparent change in
behavior was associated with when males and females
were captured or how long they spent in captivity.

In each trial males differed primarily in hue scores.
Within a trial males differed significantly in bright and
dull breast hue (Wilcoxon matched-pairs test; Bright
hue; Z = 2.81, P < 0.05; Dull hue: Z = 2.81. P <
0.05; Table 1). There were also significant but small
differences in dull breast intensity between males in a
trial (Z = 2.64, P < 0.05; Table 1). The lack of dif-
ferences in bright breast intensity (Table 1 ) reflects the
limited natural variation of breast intensity in male car-
dinals. 1 analyzed bright and dull breast measurements
separately because male rankings based on bright and
dull measurements were not identical.

Trials were conducted in a rectangular experimental
aviary (5.0 X 1.4 X 1.75 m) divided by netting into
three main compartments. Males were placed at op-
posite ends of the aviary and a female was placed in
the center compartment (2.5 X 1.4 X 1.75 m). The
female compartment was divided into three equal areas
so that females could associate with either male or
spend time in a middle area where she could not in-
teract visually with either male. Opaque baniers within
the female area were used to divide the compartment.
These prevented a male from observing a female in-

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

TABLE 2. Male color and morphology differences in

male cardinals used in

manipulated trials.

Lightened male

Reddened male

X ± SD (n)

^±50 (n)

Z (P)“

Color score before manipulation:

Bright breast hue

75.7 ± 7.4 (10)

80.5 ± 6.5 (10)

1.89 (0.059)

Dull breast hue

68.9 ± 4.9 (10)

70.9 ± 3.5 (10)

1.99 (0.046)

Bright breast intensity

63.0 ± 0.7 (10)

63.1 ± 0.9 (10)

0.30 (> 0.05)

Dull breast intensity

61.5 ± 3.1 (10)

62.3 ± 0.9 (10)

0.34 (> 0.05)

Color score after manipulation^:

Bright breast hue

72.9 ± 5.4 (9)

86.1 ± 4.3 (9)

2.69 (0.0072)

Dull breast hue

66.6 ± 6.4 (8)

77.3 ± 4.8 (8)

2.52 (0.011)

Bright breast intensity

62.8 ± 0.9 (5)

63.8 ± 0.7 (5)

0.68 (> 0.05)

Dull breast intensity

61.2 ± 0.9 (9)

63.0 ± 0.8 (9)

2.69 (0.0072)

Morphological traits (in mm):

Tarsus length

24.0 ± 0.9 (10)

24.4 ± 0.8 (10)

0.83 (> 0.05)

Crest length

35.4 ± 2.5 (10)

37.1 ± 2.9 (10)

1.17 (> 0.05)

Tail length

99.7 ± 4.3 (10)

98.6 ± 4.5 (10)

0.35 (> 0.05)

Black bib length

20.4 ± 4.1 (10)

19.6 ± 4.8 (10)

0.56 (> 0.05)

Black bib width

21.7 ± 0.9 (10)

22.1 ± 1.3 (10)

0.82 (> 0.05)

“ Z-value is from Wilcoxon matched pairs signed rank test.

Males that have identical scores are not included in means of color scores after manipulation.

teracting with the male on the opposite side, but did
not restrict female movement. I considered a female as
interacting with a male when she spent time in the area
adjacent to the male compartment as opposed to the
center area. To minimize the possibility of side pref-
erences, all aspects of the male compartments as well
as the female compartment were symmetrical. A single
perch was provided in each male compartment and
three perches in the female compartment, one in each
area. Food and water were provided ad libitum in each
male compartment and in the middle area of the female
compartment.

Each trial lasted 29 hours. For each trial a female
was placed in the experimental aviary for a one hour
acclimation period without the presence of males. A
male was introduced to each side of the aviary and the
experiment proceeded for 26 hours. At the end of 26
hours, the positions ot the two males were switched
and the female interacted with males for an additional
two hours.

Females were videotaped during four different ob-
servation periods: ( 1 ) for one hour with no males pres-
ent (Pretrial Period), (2) for the first two hours (0 Hour
Period) after introduction of the males, (3) from 24-
26 hours after introduction of the males (24 Hour Pe-
riod), and (4) from 26-28 hours of the trial (Reversed
Period). In sum, each female was taped for 6 hours
with the two males present and for one hour prior to
the introduction of males. After completion of a trial,
males and females were returned to their respective
aviaries.

E.xperiment 2: manipulated plumaf’e trials. — The 30
individuals used in the natural plumage experiment
were also used in the manipulated plumage experi-
ment. but within each experiment individuals were
used only once. Males and females for each trial were
unfamiliar with each other because they were housed

in different aviaries prior to the trial and were intro-
duced in novel combinations for the two experiments.
For each pair of males in a manipulated trial, one was
randomly assigned to a reddened plumage treatment
and the other to a lightened plumage treatment. Trials
proceeded as in experiment one. At the conclusion of
the manipulated trials, all individuals were released at
the original site of capture.

Plumage manipulation methods. — Prior to the ma-
nipulation, males were anesthetized with 1.5 mg/kg of
Midazolam injected into the pectoralis muscle. This
dosage induced a state of light anesthesia: males closed
their eyes but opened them in response to external
stimuli such as having a wing extended. No mortality
was associated with using the anesthesia.

For the “reddened” treatment, a mixture of 1 part
Divina 20 Volume Creme Developer and 3 parts Clai-
rol Professional Hi Power Tint 670® was placed on the
feathers for 25 minutes. A “lightened” treatment con-
sisted of applying Clairol Professional 7th Stage
Creme Hair Lightener mixed with Divina 20 Volume
Creme Developer in a 1:1 mixture to the feathers for
25 minutes. In both treatments males were rinsed and
were dried with a hair dryer. The males were alert
within an hour of the injection and were released back
into an aviary within 2 hours. Trials were conducted
at least four days after the males were manipulated.

Manipulated plumage coloration scores. — Although
treatments were assigned randomly, plumage hue
scores before the manipulation were significantly high-
er for males in the reddened treatment group (Wilcox-
on matched-pairs test: Natural bright hue: Z = 1 .89, P
= 0.06; Natural dull hue: Z = 1.99, P < 0.05; Table
2). There was no significant difference in the intensity
scores prior to the manipulation (Natural bright inten-
sity. Z = 0.30, P > 0.05, Natural dull intensity: Z =
0.034, P > 0.05; Table 2). After plumage manipula-

\V<>lfenhari>er • RED COLORATION AND FEMALE PREFERENCE

tions, differences between line seores for reddened and
lightened treatment groups were significant (Manipu-
lated bright hue: Z = 2.69. P < {).05, Manipulated dull
hue: Z = 2.52, P < ().()5; Table 2). Males in the red-
dened treatment had signihcantly higher intensity for
the dull breast measurement but not for the bright mea-
surement (Manipulated bright intensity: Z = 0.68, P >
0.05; Manipulated dull intensity: Z = 2.69, P < 0.05;
Table 2). Little variation in manipulated tone seores
e.xisted: all except two indi\'iduals had the maximum
possible tone scores. Reflectance spectra (from 280-
750 nm) of manipulated plumage were within the
range of natural variation in plumage reflectance that
occurs at these wavelengths (Wolfenbarger, unpubl.
data). As in experiment one. dull and bright breast
measurements were analyzed separately.

Analyses of experiments. — For the periods video-
taped, the time a female spent in each area of the ex-
perimental aviary was measured either during the trial
via a monitor connected to one camera or after the trial
from the videotapes. The monitor was located in a
room visually and acoustically isolated from the ex-
perimental aviary. For each sampling period. 1 deter-
mined the number of trials in which the female spent
more time with the male having the higher color score
for each period videotaped (0 Flour. 24 Hour, Re-
versed). I used the 0 Hour and 24 Hour samplings to
assess female preferences and the pretrial and reversed
sampling periods to assess whether females had site
preferences in the aviary.

For the 0 Hour and 24 Hour periods, I used a one-
tailed binomial test to determine whether the number
of trials in which a female spent more time with the
redder/brighter male was significantly greater than ex-
pected by chance (50%; Conover 1980). The test sta-
tistic T refers to the number of trials in which the
female spent more time with the male with the higher
color score (Conover 1980).

Because males were randomly assigned to trials,
there were trials in which hue or intensity scores were
identical. I eliminated these from analyses because nei-
ther male was redder or brighter using my measure-
ments. I also eliminated .sampling periods in which the
difference in amount of time spent with males was less
than 2 minutes (i.e.. the female showed no preference
for a particular male).

Among trials, differences in male coloration scores
varied widely (range in differences; bright breast = 0-
21. dull breast hue = 0-24, dull breast intensity = 0-
8). Females may exhibit strong preferences (as mea-
sured by time spent with male) when differences be-
tween male coloration are large; whereas females may
spend equal amounts of time with males whose color
scores are similar. 1 used regression analysis to test
whether the magnitude t)f the difference in color was
related to the difference in the time a female spent with
a particular male. In particular, I tested whether there
was a positive relationship between the difference in
color scores between the two males and the difference
in the amount of time spent between the two males in
the 0 Hour and 24 Hour .sampling periods (Wilkinson

(n = 8) (n = 8) (n = 7)

FIG. 1. Number of natural trials in which female
spent more time with male having lower or higher col-
or score during the 0 Hour Sampling period (Binomial
test: all P > 0.05).

et al. 1992). Because the predicted difference in the
amount of time spent with males of the same color is
zero, the regression line was forced through the origin.

I tested whether females exhibited two types of po-
tential site preferences; ( 1 ) a general preference among
females for either the east or west side of the experi-
mental aviary or (2) the likelihood that an individual
female stayed on the same side of the aviary between
consecutive sampling periods. I used the pretrial and
reversed periods to assess the consistency of females’
preferences for a particular side or male. Two-tailed
binomial tests were used to determine whether females
remained on the same side of the aviary between sub-
sequent sampling periods in more than half of the trials
(Conover 1980). I used a sequential Bonferroni ad-
justment for multiple comparisons (Rice 1989) because
I used both dull and bright breast measures in analyses.
A Wilcoxon matched pairs signed rank test was used
to compare moiphological variables between paired
males (Wilkinson et al. 1992).

RESULTS

Natural and manipulated plumage experi-
ment.— Of the 59 observation periods during
the two experiments, females spent an equal
amount of time (±2 min) with both males in
only 7 periods. During the remaining 52 pe-
riods, females spent an average of 45.8 (SE =
4.0) minutes more with one male than the oth-
er (range = 12-110 min, total possible = 120
min).

In the 0 Hour and 24 Hour sampling peri-
ods, females were as likely to associate with
the relatively dull males as with brighter, red-
der males (Fig. 1, binomial test: 0 Hour: T
Bright hue = 3, P > 0.05, n— 8; T Dull hue
= 3, P > 0.05, n = 8; T Dull Intensity = 4,
P > 0.05, n = 7; 24 Hour: T Bright hue = 5,
P > 0.05, n= 9; T Dull hue = 3, P > 0.05,

THE WILSON BULLETIN

Vol. Ill, No. 1, March 1999

FIG. 2. Number of manipulated trials in which fe-
male spent more time with males having lower or
higher color .scores during the 0 Hour Sampling period
(Binomial test; all P > 0.05).

/2 = 9; T Dull Intensity — 6, P > 0.05, n =
7). Similarly, in the manipulated plumage tri-
als where color differences between males
were greater, females did not consistently as-
sociate with males having higher coloration
scores (Fig. 2; binomial test: 0 Hour: T Bright
hue = 3, P > 0.05, n = 9; T Dull hue = 4,
P > 0.05, n = 8; T Dull intensity = 2, P >
0.05, n = 9; 24 Hour: T Bright hue = 3, P >
0.05, n = 9; T Dull hue = 4, P > 0.05, n =
8; T Dull intensity = 2, P > 0.05, n = 9).

The power of binomial tests at the critical
value is 95% at sample sizes of 8 and 9, and
70% with a sample size of 7. However, the
power of all of these tests at the P-values of
the results is less (25—36%) because of the
small effect observed in the experiments. Re-
sults from at least 150 trials would be needed
to hnd significant differences with such a mi-
nor effect (Conover 1980).

Magnitude of color differences and female
association. — Within trials of both experi-
ments, males varied widely in color differenc-
es, and it is possible that females only asso-
ciated with redder or brighter males when col-
or differences were large. If so, then the dif-
ference between coloration of males should be
positively related to the difference in time a
female spent with males within a trial. In con-
trast to this prediction, there were no positive
or significant relationships between color
score differences and differences in time spent
with males in a trial for the natural or manip-
ulated experiment during the 0 Hour and 24
Hour periods. In fact, all of the slopes were
near zero or negative (range of h - -0.36 to
0.08, n = 10, all P > 0.05). Therefore as color

differences between males increased, females
did not spend a greater amount of time with
redder or brighter males.

Female behavior in the experimental avi-
ary.— Males and females rapidly adjusted to
the experimental aviary and females spent the
majority of time during trials on the three
perches rather than on the floor or netting.
Typically within the first 15 minutes after
males were introduced, females had visited
each of the side compartments. Singing oc-
curred in 7 of the 10 natural trials and 8 of
the 10 manipulated trials, but I was unable to
determine from the videotapes which male (or
female) sang. In any one trial singing occurred
for less than 5 minutes. When adjacent to a
male compartment, females sat on perches for
relatively long times but regularly interrupted
these periods by flights back and forth be-
tween and within the two side male compart-
ments. Such flights also occurred in the pre-
trial periods.

Site preferences by females. — Females did
not spend more time consistently on the east
or west side of the experimental aviary (Wil-
coxon test, two-tailed: natural trials: Pretrial Z
= 1.13, P > 0.05, n — 10, 0 Hours Z = 0.06,
P > 0.05; 24 Hours Z = 0.05, P > 0.05; Re-
versed Z = 1.07, P > 0.05; manipulated trials:
Pretrial Z = 0.41, P > 0.05; 0 Hours Z = 1.38,
P > 0.05; 24 Hour Z = 1.07, P > 0.05; Re-
versed Z = 0.97, P > 0.05). Between the pre-
trial and 0 H periods, females did not exhibit
a tendency to remain on the same side of the
experimental aviary (binomial test, natural tri-
als: T= 3, P > 0.05, n = 8, manipulated trials:
T = 5, n = 10, P > 0.05). In the 0 H and 24
H periods, most females remained on the
same side (binomial test, natural trials: T = 5,
n = 10, P > 0.05; manipulated trials: T = 8,
n = 10, P = 0.01) indicating that females
were likely to associate with the same male in
the 0 H and 24 H sampling periods. However,
in the 24 H and the reversed sampling periods,
females again spent the majority of time on
the same side (binomial test, natural trials: T
= 7, /? = 9, P = 0.02; manipulated trials: T
= 6, n = 8, P = 0.035), indicating that fe-
males did not consistently associate with the
same male once the males’ positions were re-
versed.

Morphological variables. — In the trials us-
ing natural plumage coloration, males with the

Wol/enharffer • RED COLORATION AND FEMALE PREFERENCE

higher color score did not have a longer tarsus
length, crest length, tail length, or black bib
size (Table 1; Wilcoxon rank tests: all P >
0.05). Likewise, males in the reddened and
lightened groups did not differ significantly in
morphological measurements (Table 2, Wil-
coxon rank tests, all P > 0.05).

DISCUSSION

During the majority of observation periods,
females spent significantly more time with
one male. However, the two experiments pro-
vided no evidence to support the hypothesis
that female Northern Cardinals prefer males
with redder or brighter coloration. In trials us-
ing natural plumage coloration, females were
as likely to spend more time with males hav-
ing low color scores as with those having high
color scores (Fig. 1). Similarly in the manip-
ulated trials, where average plumage differ-
ences between males were greater than in the
natural trials, females still did not spend more
time with redder or brighter males (Fig. 2).

These results differ from other studies of
female preference in passerines. Under aviary
conditions analogous to the conditions I used,
females of many species showed a preference
for brightly colored males (Johnson 1988, Hill
1990, Enstrom 1993, Johnson et al. 1993,
Saetre et al. 1994, Sundberg 1995; but see
Rohwer and Rpskaft 1989, Alatalo et al. 1990,
Butcher 1991). Although bright plumage col-
ors do not seem important in gallinaceous spe-
cies, females show preferences for other male
ornaments (Zuk et al. 1992, Buchholz 1995,
Ligon and Zwartjes 1995, Mateos and Carran-
za 1995).

The individuals adapted well to captivity
and appeared to be in breeding condition. The
experiments were conducted at a seasonally
appropriate time when pair formation occurs
in central New York (pers. obs.). Behaviors
associated with pair formation such as singing
and slow flight displays occurred in the hous-
ing aviaries and during experimental trials,
suggesting that males were responding to fe-
male presence.

Assigning female preference based on the
total time spent with males has become a stan-
dard method in avian studies of female choice
(Burley et al. 1982, Burley 1986, Hill 1990,
Enstrom 1993, Johnson et al. 1993, Sundberg
1995). Observations of pair formation in the

field indicate that male and female cardinals
interact extensively, consequently time spent
with males should be a reasonable indication
of mate preference (Kinser 1973).

While males in this study appeared to ex-
hibit some behaviors associated with pair for-
mation, female behavior was more ambigu-
ous. Females flew toward the male area re-
gardless of whether the male was present or
not. The side of the aviary on which females
spent more time changed from the pretrial to
0 hour sampling period, indicating that fe-
males did not immediately establish a prefer-
ence for one side of the experimental aviary
and may have responded to the addition of the
males. However, females preferred the same
side of the aviary during the 24 hour and re-
versed sampling periods, even though the
males had switched sides. This suggests that
after 24 hours individual males did not strong-
ly influence where females spent more time.

With the sample sizes used, the power of
the binomial test is high at the critical value.
It is noteworthy that other choice experiments
of this design have found significant female
preferences with similar samples sizes (rang-
ing from 7 to 21 females; Hill 1990, Enstrom
1993, Saetre et al. 1994, Sundberg 1995). This
suggests that if a preference for red color ex-
ists, it is weak in comparison to color pref-
erences found in other passerines.

The dietary basis of red coloration in male
cardinals suggests that females could acquire
information about a male’s foraging abilities
or ability to defend resources using coloration.
Male color and his absolute effort in feeding
nestlings are not related in cardinals (Linville
et al. 1998), emphasizing the limitation of us-
ing red coloration alone to assess mates. How-
ever, females mated to brighter males fed nest-
lings less (Linville et al. 1998), suggesting
that there may be advantages to pairing with
brighter males.

My results provided no evidence for female
preference for brighter or redder male plum-
age coloration as a single criterion for mate
choice; however, the possibility remains that
female cardinals assess mates using a combi-
nation of factors, including coloration, as has
been found in other species (Zuk et al. 1990;
Omland 1996a, b; Scheffer et al. 1996; Mpller
et al. 1998). The prolonged opportunities for
direct interactions between males and females

THE WILSON BULLETIN

Vol. Ill, No. 1, March 1999

prior to pairing (Kinser 1973) suggests that
females may be able to use other male char-
acteristics in mate choice, such as song or
courtship behaviors. If females assess male
condition, multiple ornaments or traits may
provide more accurate information either
through redundancy or because some orna-
ments are unreliable indicators of condition
(Mpller and Pomiankowski 1993). It also re-
mains possible that females assess male col-
oration in some circumstances but not other.
For example, females may use different cri-
teria when choosing social mates and when
choosing extra-pair mates.

Lastly, territorial resources may influence
mate choices, but coloration may be relatively
more important in mediating competition be-
tween males. Redder male cardinals acquire
territories with denser vegetation density, and
pairs on these territories produce more off-
spring, most likely because of reduced nest
predation (Wolfenbarger in press). Given the
importance of territory quality, females may
directly assess a male’s territory during pair
formation rather than relying on plumage col-
oration alone for mate choices.

ACKNOWLEDGMENTS

For assistance in designing and conducting these ex-
periments, I thank S. Barker, C. Brown, L. Carbone,

R. Cocroft, S. Emlen, M. Martino, J. McCarty, K.
Reeve, J. Roth, P. Sherman, S. Smedley, and D. Wink-
ler. The manuscript was improved by comments from

S. Emlen, J. McCarty, P. Sherman, D. Winkler, and two
anonymous reviewers. The work was conducted in ac-
cordance with federal (PRT-792153), NY state
(LCP94-681), and Cornell University's Institutional
Animal Care and Use Committee (94-129) regulations.
This work was supported by a Dissertational Improve-
ment Grant from the NSF (IBN-9321801 ).

LITERATURE CITED

Alatalo, R. V., A. Lundberg, and J. Sundberg. 1990.
Can female preference explain sexual dichroma-
tism in the Pied Flycatcher, Ficedula hypoleuca?
Anim. Behav. 39:244-252.

Andersson, M. 1982. Female choice selects for ex-
treme tail length in a widowbird. Nature (Lond.)
299:818-820.

Andersson, M. 1994. Sexual selection. Princeton
Univ. Press, Princeton, New Jersey.

Andersson, S. 1992. Female preferences for long tails
in lekking Jackson's Widowbirds: experimental
evidence. Anim. Behav. 43:379-388.

Buchholz, R. 1995. Female choice, parasite load and

male ornamentation in Wild Turkeys. Anim. Be-
hav. 50:929-943.

Burley, N. 1986. Comparison of the band-colour pref-
erences of two species of estrildid finches. Anim.
Behav. 34:1732-1741.

Burley, N., G. Krantzberg, and P. Radman. 1982.
Influence of colour-banding on the conspecific
preferences of Zebra Finches. Anim. Behav. 30:
444-455.

Butcher, G. S. 1991. Mate choice in female Northern
Orioles with a consideration of the role of the
black male coloration in female choice. Condor
93:82-88.

Conover, W. J. 1980. Practical non-parametric statis-
tics, second edition. John Wiley and Sons, Inc.,
New York.

Darwin, C. 1871. The descent of man, and selection
in relation to sex. Princeton Univ. Press, Prince-
ton, New Jersey.

Enstrom, D. a. 1993. Female choice for age-specific
plumage in the Orchard Oriole: implications for
delayed plumage maturation. Anim. Behav. 45:
435-442.

Evans, M. R. and B. J. Hatchwell. 1992. An exper-
imental study of male adornment in the Scarlet-
tufted Malachite Sunbird: II. The role of the elon-
gated tail in mate choice and experimental evi-
dence for a handicap. Behav. Ecol. Sociobiol. 29:
421-427.

Filliater, T. S. and R. Breitwisch. 1997. Nestling
provisioning by the extremely dichromatic North-
ern Cardinal. Wilson Bull. 109:145-153.

Goodwin, T. W. 1950. Carotenoids and reproduction.
Biol. Rev. 25:391-413.

Hagan, J. M. and J. M. Reed. 1988. Red color bands
reduce fledging success in Red-cockaded Wood-
peckers. Auk 105:498-503.

Hill, G. E. 1990. Female House Finches prefer col-
ourful males: sexual selection for a condition-de-
pendent trait. Anim. Behav. 40:563-572.

Hill, G. E. 1992. Proximate basis of variation in ca-
rotenoid pigmentation in male House Finches.
Auk 109:1-12.

Johnson, K. 1988. Sexual selection in Pinyon Jays I:
female choice and male-male competition. Anim.
Behav. 36:1038-1047.

Johnson, K., R. Dalton, and N. Burley. 1993. Pref-
erences of female American Goldfinches {Car-
cluelis iri.stis) for natural and artificial male traits.
Behav. Ecol. 4:138-143.

Jones, I. L. and F. M. Hunter. 1993. Mutual .sexual
selection in a monogamous seabird. Nature
(Lond.) 362:238-239.

Kinser, G. W. 1973. Ecology and behavior of the car-
dinal Richmomlemi cardinaUs (L). in southern In-
diana. Ph.D. diss., Indiana Univ., Bloomington.

Kornerup, a. 1967. Methuen handbook of color. Thiru
ed. Methuen, London, U.K.

Ligon, j. D. and P. W. Zwartjes. 1995. Ornate phm
age of male Red Jungle Fowl does not influeiu

Wolfenbarger • RED COLORATION AND FEMALE PREFERENCE

mate choice by females. Anim. Behav. 49:117-
125.

Linville, S.U., R. Breitwisch, and A. J. Schilling.
1998. Plumage brightness as an indicator of pa-
rental care in Northern Cardinals. Anim. Behav.
55: 119-127.

Mateos, C. and J. Carranza. 1995. Female choice
for morphological features of male Ring-necked
Pheasants. Anim. Behav. 49:737-748.

Metz, K. J. and P. J. Weatherhead. 1991. Color
bands function as secondary sexual traits in male
Red-winged Blackbirds. Behav. Ecol. Sociobiol.
28:23-27.

Moller, a. P. 1987. Variation in badge size in male
House Sparrows Passer domesticus: evidence for
status signalling. Anim. Behav. 35:1637—1644.

Moller, a. P. 1988. Female choice selects for male
sexual tail ornaments in the monogamous swal-
low. Nature (Lond.) 332:640—642.

Moller, A. P. 1992. Female swallow preference for
symmetrical male sexual ornaments. Nature
(Lond.) 357:238-240.

Moller, A. P. and A. Pomiankowski. 1993. Why have
birds got multiple sexual ornaments? Behav. Ecol.
Sociobiol. 32:167-176.

Moller, A. P, N. Saino, G. Taramino, P. Galeotti,
AND S. Ferrario. 1998. Paternity and multiple sig-
naling: effects of a secondary sexual character and
song on paternity in the Bam Swallow. Am. Nat.
151:236-242.

Omland, K. E. 1996a. Female Mallard mating pref-
erences for multiple male ornaments. 1. Natural
variation. Behav. Ecol. Sociobiol. 39:353-360.

Omland, K. E. 1996b. Female Mallard mating prefer-
ences for multiple male ornaments. II. Experimen-
tal variation. Behav. Ecol. Sociobiol. 39:361—366.

Rice, W. R. 1989. Analyzing tables of statistical tests.
Evolution 43:223-225.

Rohwer, S. and E. Roskalt. 1989. Results of dyeing
male Yellow-headed Blackbirds solid black: im-
plications for the arbitrary identity badge hypoth-
esis. Behav. Ecol. Sociobiol. 25:39-49.

S/ETRE, G.-P, S. Dale, and T. Slagsvold. 1994. Fe-
male Pied Flycatchers prefer brightly coloured
males. Anim. Behav. 48:1407-1416.

Scheffer, S. J., G. W. Uetz, and G. E. Stratton.
1996. Sexual selection, male morphology, and the
efficacy of courtship signalling in two wolf spi-
ders (Araneae: Lycosidae). Behav. Ecol. Socio-
biol. 38:17-23.

SuNDBERG, J. 1995. Female Yellowhammers {Emheriza
citrinella) prefer yellower males: a laboratory ex-
periment. Behav. Ecol. Sociobiol. 37:275-282.

Wilkinson, L., M. Hill, S. Miceli, P. Howe, and E.
Vang. 1992. SYSTAT version 5.2. SYSTAT, Inc.,
Evanston, Illinois.

WoLFENBARGER, L. L. 1996. Fitness effects associated
with red coloration of male Northern Cardinals
(Cardinalis cardinalis). Ph.D. diss., Cornell Univ.,
Ithaca, New York.

WOLFENBARGER, L. L. In press. Red coloration of male
northern cardinals correlates with mate quality
and territory quality. Behav. Ecol.

ZuK, M., J. D. Ligon, and R. Thornhill. 1992. Effects
of experimental manipulation of male secondary
sex characters on female mate preference in Red
Jungle Fowl. Anim. Behav. 44:999-1006.

ZuK, M., R. Thornhill, J. D. Ligon, K. Johnson, S.
Austad, S. Ligon, N. Thornhill, and C. Costin.
1990. The role of male ornaments and courtship
behavior in female choice of Red Jungle Fowl.
Am. Nat. 136:459-473.

Wilson Bull., 111(1), 1999, pp. 84-88

FRUIT SUGAR PREFERENCES OF HOUSE FINCHES

MICHAEL L. AVERY,' 3 CARRIE L. SCHREIBER,' - AND DAVID G. DECKER'

ABSTRACT. — In a series of choice tests, we determined the relative preferences of House Finches (Carpo-
dacus mexicamis) for equicaloric aqueous solutions of hexoses (1:1 mixture of fructose and glucose) and sucrose.
At 2% (m/v), birds consumed each sugar solution equally and in amounts similar to plain water. Consumption
of hexose but not sucrose increased at 4% sugar concentration. At 6% and 10%, finches displayed consistent,
strong preferences for the hexoses over sucrose. In other passerine species, strong hexose preference has been
linked to the absence of sucrase, the enzyme needed for digestion of sucrose. Fecal sugar readings from the
House Finches, however, indicated approximately equal assimilation of hexose and sucrose, so the hexose pref-
erence apparently is not due to sucrase deficiency. Rather, energetics may determine the finches’ sugar prefer-
ences: hexoses are rapidly processed because the 6-carbon sugars are readily assimilable whereas sucrose must
first be hydrolyzed. Received 22 Jem. 1998, accepted 30 Aug. 1998.

Physiology imposes major constraints on
the digestion of sugars by some fruit-eating
birds. These constraints in turn affect species’
food selection behavior. Species of Stumidae
(e.g., European Starling, Sturnus vulgaris) and
Turdidae (e.g., American Robin, Turdus mig-
ratorius) are unable to digest sucrose because
they lack the enzyme sucrase needed to hy-
drolyze sucrose into 6-carbon sugars, glucose
and fructose that can be assimilated (Martinez
del Rio and Stevens 1989, Karasov and Levey
1990). Ingestion of high concentrations of su-
crose by these species produces osmotic di-
arrhea and, in extreme cases, death (Martinez
del Rio et al. 1988, Brugger and Nelms 1991).
Consequently, in feeding and drinking trials
starlings and robins learn to avoid sucrose
(Schuler 1983, Martinez del Rio et al. 1988,
Brugger 1992).

Although Cedar Waxwings (Bombycilla
cedrorum) can digest sucrose, in choice tests
they also prefer hexoses to sucrose (Martinez
del Rio et al. 1989, Avery et al. 1995). Wax-
wings exhibit very rapid gut passage rates
(Levey and Grajal 1991). As a result, sucrose
is not in the gut long enough to be completely
hydrolyzed and is therefore inefficiently as-
similated relative to hexose sugars (Martinez
del Rio et al. 1989).

In the Icteridae and Emberizidae, two fru-

' U.vS. Department of Agriculture, Animal and Plant
Health In.spcction .Service. National Wildlife Research
Center. Florida Field Station, 2820 East University Av-
enue, Gainesville, FL .12641.

’ Present addre.ss: 716 NE 4th Avenue, Gainesville,
FL .12601.

'Corresponding author; E-mail: dwrc-ffs@afn.org

givorous species, the Yellow-winged Cacique
{Cacicus melanicterus) and the Yellow-breast-
ed Chat (Icteria virens) preferred 15% (by
mass) hexose solution over sucrose solution
and displayed relatively inefficient sucrose di-
gestion (Martinez del Rio and Restrepo 1993).
Conversely, Red-winged Blackbirds (Agelaius
phoeniceus) and Common Crackles {Quisca-
lus quiscula), granivorous icterids, preferred
sucrose solutions to water but did not distin-
guish between 0.175M and 0.35M hexose so-
lutions and water (Martinez del Rio et al.
1988).

The House Finch (Carpodacus mexicanus)
is primarily granivorous (Martin et al. 1951)
but feeds opportunistically on cultivated fruit
(Tobin and DeHaven 1984, Avery et al. 1992).
To our knowledge the sugar preferences of
House Finches and other Fringillidae have not
been evaluated. Responses of House Finches
to fruit sugars are pertinent to the develop-
ment of high-sucrose fruit cultivars for poten-
tially reducing bird damage to fruit crops
(Brugger et al. 1993, Darnell et al. 1994).
Thus, our objectives were (1) to document
House Finch consumption of sucrose and hex-
ose in equicaloric aqueous solutions across a
range of sugar concentrations typically found
in cultivated fruit and (2) to measure fecal
sugar to determine relative digestion of su-
crose and hexoses.

METHODS

House Finches were from a captive population
maintained at the Florida Field Station of the U.S.
Dept, of Agriculture’s National Wildlife Research Cen-
ter m Gainesville, Florida. We maintained birds on a
mixed .seed diet supplemented three days/week with

Aven’ et cil. • HOUSE FINCH SUGAR PREFERENCES

apples and lettuce. Testing occurred during October—
November 1995. After testing, birds were returned to
their home cages.

We removed birds from communal enclosures (2 X
1.5 X 2.2 m) and placed them into individual, visually
isolated test cages (45 cm on a side) in a roofed out-
door aviary. To acclimate the birds, we offered plain
water tinted with red food coloring in clear glass tubes
(8 mm diameter) 4-5 days before testing. We fixed two
tubes, 5 cm apart, to the front of each cage. During
acclimation, we measured water consumption after 6
h and 24 h daily to determine baseline fluid intake and
to accustom the birds to disturbances.

We prepared test solutions by dissolving 20, 40, 60,
or 100 g of sucrose or hexose sugars (Sigma Chemical
Company, St. Louis, Missouri) in 1 L of distilled wa-
ter. The hexose solution contained equal amounts of
fructose and glucose. We then conducted separate tests
at each of 4 sugar concentrations (m/v): 2%, 4%, 6%,
and 10%. Tests lasted 4 days and there were 6 birds/
group. One hexose tube and one sucrose tube, 5 cm
apart, were available during each test. For each cage,
we first randomly determined the position of the su-
crose tube and then alternated sucrose and hexose po-
sitions daily. We removed maintenance food and water
at 08:00 and presented the tubes with sugar solutions
from 09:00 until 15:00. Maintenance food and water
were then returned to the cages.

We measured the amount of solution missing from
each tube to the nearest mm, and then converted to
amount of sugar (g) ingested for analyses. We assessed
sugar consumption in a 3-way analysis of variance,
with sugar concentration as the independent factor, and
repeated measures over sugar type and days. We used
Tukey’s HSD test (Steel and Torrie 1980) to isolate
differences {P < 0.05) among means.

To determine relative digestion by finches of sucrose
and hexoses, we analysed fecal sugar with a hand-held
refractometer (Hainesworth 1974, Brugger et al. 1993).
We offered six birds a 10% (m/v) agar-sucrose mixture
(Avery et al. 1995) for 6 hours and offered similar
food made with hexose (equal amounts of glucose and
fructose) for 6 hours the next day. We measured three
fresh defecations from each bird with each sugar treat-
ment, and compared mean values in a paired r-test
against a null hypothesis of no difference between sug-
ars. Refractometer readings are expressed as degrees
Brix which corresponds to the percentage of sugar pre-
sent in the sample on a mass : mass basis (Bolten et al.
1979).

RESULTS

Total sugar consumption varied (F, 20 =
22.77, P < 0.001) with concentration. Sugar
ingestion at 6% (mean ± SE, x = 1.07 ±0.12
g/bird) and 10% (jc = 1.66 ±0.11 g/bird) ex-
ceeded that at 2% (x = 0.03 ± 0.01 g/bird)
and 4% (T = 0.29 ± 0.08 g/bird). Overall,
hexose consumption (0.67 ± 0.07 g/bird) ex-
ceeded (F,2o = 93.55, P < 0.001) sucrose

FIG. 1 . Mean consumption of hexoses and sucrose
by House Finch groups (6 birds/group) exposed to two
tubes of aqueous sugar solutions for 4 days, 6 hours
per day. Vertical bars denote 1 SE. Note that the y axis
is logarithmic.

consumption (0.09 ± 0.02 g/bird). Finches
consumed less sugar (F^^eo = 2.83, P = 0.046)
on day 1 (0.61 ± 0.17 g/bird) than on days
2-4 (mean consumption 0.79-0.83 g/bird).

Across the range of test concentrations,
finches responded differently (F3 20 = 24.78, P
< 0.001) to the two types of sugars (Fig. 1).
Sucrose consumption was consistently low
(mean consumption 0.02-0.22 g/bird) and did
not differ from hexose consumption at 2%
(0.02 ± 0.01 g/bird) and 4% (0.22 ± 0.07 g/
bird). Hexose consumption increased (P <
0.05) substantially, however, at 6% (1.01 ±
0.11 g/bird) and at 10% (1.43 ± 0.10 g/bird).

The interaction between type of sugar and
test day affected consumption (F^^o ~ 16.59,
P < 0.001). Sucrose consumption did not dif-
fer across the 4 test days, and on day 1, mean
sucrose consumption (0.22 ± 0.10 g/bird)
equalled hexose consumption (0.39 ± 0.10 g/
bird). Hexose consumption increased thereaf-
ter and averaged 0.73 to 0.79 g/bird on days
2-4.

The 3-way interaction (Fg^,,, = 5.31, P <
0.001) reflected differing daily consumption
patterns of the two sugar types as sugar con-
centration varied (Fig. 2). At 2%, consump-
tion of both types of sugar remained low
throughout the test. At 4%, mean hexose con-
sumption increased each day but not suffi-
ciently to achieve statistical significance (F >
0.05). At 6%, mean hexose consumption in-
creased (F < 0.05) from day 1 (0.66 ± 0.21

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

Day 1 Day 2 Day 3 Day 4

FIG. 2. Mean daily consumption of hexoses (open bars) and sucrose (solid bars) by House Finches (6 birds/
trial) exposed to two tubes of aqueous sugar solutions for 6 hours each day. Vertical bars denote 1 SE.

g/bird) to day 4 (1.25 ± 0.22 g/bird). On day
1, finches consumed equal amounts of hexose
and sucrose at the 10% level, but consumption
diverged {P < 0.05) on day 2 and remained
so through day 4.

During pretest days, hourly consumption of
water averaged 0.22 ± 0.10 and 0.33 ± 0.07
ml/bird for the 2% and 4% groups, respec-
tively, similar to their total consumption of
2% (0.25 ± 0.03 ml/bird) and 4% (1.19 ±
0.30 ml/bird) sugar solutions. Fecal sugar
analysis from six birds revealed no difference
(/ = 1.63, P > 0.05) between sugars. Hexose
readings averaged 4.2 ± 0.7° Brix compared
to an average of 2.8 ± 0.3° Brix for sucrose.

DISCUSSION

In the range of concentrations we tested,
preference for hexose over sucrose has not
previously been demonstrated, even in species
lacking sucrase. Rejection of sucrose by Eu-
ropean Starlings and American Robins oc-
curred at concentrations in excess of 10%
(Schuler 1983, Martmez del Rio et al. 1988,
Brugger 1992). Other species are either indif-
ferent (domestic hen, Kare and Medway 1959;

Rock Dove, Columba livia, Duncan 1960;
Common Raven, Corpus corax, Harriman and
Fry 1990) or prefer sucrose (Common Crack-
le, Red-winged Blackbird; Martmez del Rio et
al. 1988). In choice tests, hummingbirds pre-
fer sucrose and reject fructose (Stiles 1976),
but when fructose is offered alone, humming-
birds consume it at a rate no different from
sucrose. Other nectarivorous species also se-
lect sucrose preferentially over equimolar
fructose and glucose solutions (Downs and
Perrin 1996).

The sugar solutions we offered appeared
alike to us and their relative positions were
switched daily. At 2%, it appeared that finches
did not distinguish dilute sugar solutions from
plain water; consumption was low and re-
mained so throughout the trial. Finches re-
sponded to sugar at the 4% level, and mean
consumption of hexose increased steadily
across the 4-day trial while sucrose consump-
tion remained low. At 6%, hexose consump-
tion increased markedly over that at 2% and
4%, while sucrose consumption did not differ
from that at lower concentrations. Discrimi-
nation between sugars was more rapid at 10%,

Avcrv et al. • HOUSE FINCH SUGAR PREFERENCES

as finches decisively selected hexose over su-
crose after one trial. The birds apparently
tracked the position of the hexose tube
through a nonvisual cue. The mechanism by
which they discriminated hexose from sucrose
is unclear, but the rapidity of the discrimina-
tion increased with sugar concentration.

We hypothesize that finches chose hexoses
in response to an increased rate of energy gain
relative to sucrose solutions during the 6-h
drinking trials. Birds are sensitive to differ-
ences in rates of energy assimilation (Witmer
1994), and the extra step, hydrolysis of the
sucrose molecule required for sucrose diges-
tion imposes a constraint on the potential rate
of energy assimilation. In our choice tests,
finches responded facultatively and selected
the more energetically efficient food source.

Martinez del Rio and coworkers (1988) pre-
dicted that granivores should have high su-
crase activity and prefer, or at least tolerate,
sucrose. This follows from the facts that malt-
ose is the major constituent of complex car-
bohydrates found in seeds, granivorous spe-
cies show high intestinal maltase activity, and
the activity of sucrase seems to vary with that
of maltase and isomaltase (Martinez del Rio
1990, Martinez del Rio et al. 1995). Although
House Finches are basically granivorous, they
strongly favored moderate hexose sugar so-
lutions over sucrose (Figs. 1, 2). We did not
determine intestinal enzyme activity directly,
but fecal sugar analyses indicated that the
preference for hexoses was not because of ab-
sence of sucrase. House Finches prefer hexose
sugars but are “sucrose tolerant” granivores,
consistent with the hypothesis of Martinez del
Rio and coworkers (1988). Comparative stud-
ies of House Finches and other granivores will
help to define more clearly the physiological
basis underlying their food selection behavior.

Development of high-sucrose fruit cultivars
could represent one nonlethal component of
an integrated plan to manage bird damage to
berry crops (Brugger et al. 1993, Darnell et
al. 1994). Such an approach will most likely
be effective against species such as the Eu-
ropean Starling and American Robin that lack
sucrase and are thus unable to digest sucrose.
For sucrose tolerant species such as the House
Finch, elevated sucrose concentrations in fruit
will probably not reduce crop damage unless
alternative food sources are readily available.

Rather, because of inefficient energy assimi-
lation from sucrose ingestion, sucrose tolerant
species might compensate by increasing fruit
consumption, thereby causing greater damage
(Avery et al. 1995).

ACKNOWLEDGMENTS

We thank K. L. Roca and C. C. McClester for caring
for the House Finches. Our paper was improved by the
constructive comments of J. R. Belthoff, K. E. Brug-
ger, D. J. Levey, C. Martinez del Rio, and two anon-
ymous reviewers.

LITERATURE CITED

Avery, M. L., J. W. Nelson, and M. A. Cone. 1992.
Survey of bird damage to blueberries in North
America. Proc. East. Wildl. Damage Control
Conf. 5:105-1 10.

Avery, M. L., D. G. Decker, J. S. Humphrey, A. A.
Hayes, and C. C. Laukert. 1995. Color, size, and
location of artificial fruits affect sucrose avoidance
by Cedar Waxwings and European Starlings. Auk
112:436-444.

Bolten, a. B., P. Feinsinger, H. G. Baker, and 1.
Baker. 1979. On the calculation of sugar concen-
tration in flower nectar. Oecologia 41:301-304.
Brugger, K. E. 1992. Repellency of sucrose to captive
American Robins (Turdus migratorius). J. Wildl.
Manage. 56:794-799.

Brugger, K. E. and C. O. Nelms. 1991. Sucrose
avoidance by American Robins {Turdus migrato-
rius): implications for control of bird damage in
fruit crops. Crop Prot. 10:455-460.

Brugger, K. E., P. Nol, and C. I. Phillips. 1993. Su-
crose repellency to European Starlings: will high-
sucrose cultivars deter bird damage to fruit? Ecol.
Applic. 3:256—261.

Darnell, R. L., R. Cano-Medrano, K. E. Koch, and
M. L. Avery. 1994. Differences in sucrose me-
tabolism relative to accumulation of bird-deterrent
sucrose levels in faiits of wild and domestic Vac-
cinium species. Physiol. Plant. 92:336—342.
Downs, C. T. and M. R. Perrin. 1996. Sugar prefer-
ences of some southern African nectarivorous
birds. Ibis 138:455-459.

Duncan, C. J. 1960. The sense of taste in birds. Ann.
Appl. Biol. 48:409-414.

Hainesworth, E R. 1974. Food quality and foraging
efficiency: the efficiency of sugar assimilation by
hummingbirds. J. Comp. Physiol. 88:425-431.
Harriman, a. E. and E. G. Fry. 1990. Solution ac-
ceptance by Common Ravens (Con-us cora.x) giv-
en two-bottle preference tests. Psychol. Rep. 67:
19-26.

Karasov, W. H. and D. j. Levey. 1990. Digestive sys-
tem trade-offs and adaptations of frugivorous pas-
serine birds. Physiol. Zool. 63:1248—1270.

Kare, M. R. and W. Medway. 1959. Discrimination
between carbohydrates by the fowl. Poultry Sci.
38:1119-1 127.

THE WILSON BULLETIN • Vol. HI, No. I, March 1999

Levey, D. J. and A. Grajal. 1991. Evolutionary im-
plications of fruit-processing limitations in Cedar
Waxwings. Am. Nat. 138:171-189.

Martin, A. C., H. S. Zim, and A. L. Nelson. 1951.
American wildlife and plants. McGraw-Hill Book
Co., Inc., New York.

Martinez del Rio, C. 1990. Dietary and phylogenetic
correlates of intestinal sucrase and maltase activity
in birds. Physiol. Zool. 63:987-1011.

Martinez del Rio, C., K. E. Brugger, J. L. Rios, M.
E. Vergara, and M. Witmer. 1995. An experi-
mental and comparative study of dietary modu-
lation of intestinal enzymes in European Starlings
(Sturnu.'i vulgari.s). Physiol. Zool. 68:490-511.

Martinez del Rio, C., W. H. Karasov, and D. J. Lev-
ey. 1989. Physiological basis and ecological con-
sequences of sugar preferences in Cedar Wax-
wings. Auk 106:64-71.

Martinez del Rio, C. and C. Restrepo. 1993. Eco-
logical and behavioral consequences of digestion
in frugivorous animals. Vegetatio 107/108:205-
216.

Martinez del Rio, C. and B. R. Stevens. 1989. Phys-
iological constraint on feeding behavior: intestinal
membrane disaccharidases of the starling. Science
243:794-796.

Martinez del Rio, C., B. R. Stevens, D. E. Daneke,
AND P. T. Andreadis. 1988. Physiological corre-
lates of preference and aversion for sugars in three
species of birds. Physiol. Zool. 61:222—229.

Schuler, W. 1983. Responses to sugars and their be-
havioural mechanisms in the starling {Sturnus vul-
garis L.). Behav. Ecol. Sociobiol. 13:243-251.

Steel, R. G. D. and J. H. Torrie. 1980. Principles and
procedures of statistics. Second ed., McGraw-Hill
Book Co., New York.

Stiles, E G. 1976. Taste preferences, color preferences
and flower choice in hummingbirds. Condor 78:
10-26.

Tobin, M. E. and R. W. DeHaven. 1984. Repellency
of methiocarb-treated grapes to three species of
birds. Agric. Ecosyst. Environ. 11:291-297.

Witmer, M. C. 1994. Contrasting digestive strategies
of frugivorous birds. Ph.D. diss., Cornell Univ.,
Ithaca, New York.

Wilson Bull., 111(1), 1999, pp. 89-99

HIERARCHICAL COMPARISONS OF BREEDING BIRDS IN OLD-
GROWTH CONIFER-HARDWOOD FOREST ON THE
APPALACHIAN PLATEAU

J. CHRISTOPHER HANEY'

ABSTRACT. — I compared relative abundances of breeding birds in old-growth forest (>300 years old) to
surrounding landscapes using data from the Breeding Bird Census (BBC) and Breeding Bird Atlas (BBA).
Eleven study plots (148 ha total) were established in relict, presettlement hemlock-white pine-northern hardwood
(Tsiiga canodensis-Pinus strobus) forest on the northern Appalachian Plateau, Pennsylvania. Of 56 breeding
species recorded in old-growth forest, 34% were either uncommon (:S25% of BBA blocks) or rare (^10% of
BBA blocks) in adjacent landscape units. A species accumulation curve indicated that about 40 species recurred
in old-growth habitat. This avian community included species less likely to occur in oldgrowth, forest interior
species showing a statistically neutral relationship to oldgrowth, and habitat specialists more likely to reside in
oldgrowth than in the landscape at large. The last group included several taxa linked to structural features of
oldgrowth elsewhere in North America: Hairy Woodpecker (Picoides villosus). Red-breasted Nuthatch (Sitta
canadensis). Brown Creeper (Certhia americana). Winter Wren (Troglodytes troglodytes). Golden-crowned
Kinglet (Regidus satrapa), Empidona.x flycatchers, and several species of arboreal Dendroica warblers. Received
14 July 1998, accepted 4 Nov. 1998.

Old-growth forests possess unique ecolog-
ical characteristics that can exert profound in-
fluence on some bird populations and com-
munities (Hunter et al. 1995, Dellasala et al.
1996). Ecological importance of oldgrowth to
birds is poorly known in much of North
America, largely because late successional
forest outside the Pacific Northwest now oc-
curs only in relict patches (Davis 1996). Ide-
ally, the role of old-growth forest in facilitat-
ing avian diversity could be best evaluated by
comparing species occurrences within entire
landscapes made up of many different habi-
tats.

What is the best way to evaluate bird dis-
tributions and abundances over multiple spa-
tial scales? Results of studies on species oc-
currences at any one scale may conflict with
results at alternative scales (Conroy and Noon
1996). In avian ecology, this concern may find
expression as a tradeoff among within-habitat
(a), between-habitat ((3), and landscape (y) di-
versity (Whittaker 1977, Wiens 1989). For
avian conservation, management actions at lo-
cal scales must be weighed against their con-
sequences at broader scales in order to opti-
mize benefits of land use (Flather 1996). A
hierarchical framework is the method usually

‘ Ecology and Economics Research Dept., The Wil-
derness Society, 900 17th Street, NW, Washington,
D.C. 20006; E-mail: Jchris_haney(3)tws.org

recommended to address such scale depen-
dency (Kotliar and Wiens 1990).

I used a landscape hierarchy to evaluate
bird distribution and abundance in old-growth
forest, once a widespread vegetation type in
eastern North America. Local species’ occur-
rences in old-growth conifer-hardwood forest
were compared to occurrences in the sur-
rounding landscape using data from the
Breeding Bird Census and Breeding Bird At-
las. Three questions were posed: (1) can a
metric be devised to compare bird species oc-
currences across different spatial scales, (2)
does old-growth forest harbor birds deter-
mined independently to be uncommon or rare
in larger landscape units, and (3) which indi-
vidual species are more likely to occur in old-
growth than in the landscape as a whole?

METHODS

Study area. — Bird communities in oldgrowth were
studied within three permanent forest reserves in Penn-
sylvania: Cook Forest State Park. Heart's Content, and
Tionesta Scenic and Research natural areas in the Al-
legheny National Forest (41° 2()'-41° 42' N, 78° 56'-
79° 15' W). Tract sizes of old-growth habitat varied
from 1000 ha at Tionesta to 60 ha at Heart's Content;
each of these reserves is embedded within much larger
contiguous tracts of younger managed forest. Land-
scape fragmentation is greater in Cook Forest where
developed and agricultural lands virtually surround this
3000 ha reserve (which includes some 200 ha of old-
growth in three sites). All reserves are located on the
northern Appalachian Plateau (212Ga: Allegheny High

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

Plateau Subsection, Northern Unglaciated Allegheny
Plateau Section; Keys et al. 1995), a region character-
ized by broad, flattened ridges (500-700 m) and dis-
sected by deep, V-shaped valleys. Higher precipitation
and greater cloud cover create a cooler, more humid
climate compared to adjacent regions (Whitney 1990).

Each reserve possesses relict stands of true old-
growth, stands with extreme ecological maturity (Lev-
erett 1996). The forest consists of hemlock- white pine-
northem hardwoods (Nichols 1935), a cover type that
most closely resembles USDA Forest Service CISC
(continuous inventory of stand condition) types 4 and
8, Society of American Foresters forest type code 22
(Eyre 1980), and International Classification of Eco-
logical Communities I.B.8.N.b. 150 (USDA 1997).
Canopy dominants include eastern hemlock (Tsuga
canadensis), eastern white pine (Pinas strohu.'i), yellow
birch (Betula alleghaniensis), black birch (Betula len-
ta), sugar maple (Acer saccharum), red maple (Acer
ruhrum), and American beech (Fagus grandifolia\
Whitney 1990).

Understories are generally sparse with little herba-
ceous ground cover. Canopy trees at each site are 300-
530 years old (Hough and Forbes 1943; Abrams and
Orwig 1996; Stable 1996; C. Nowack, pers. comm.),
and so are beyond the threshold (275 years) at which
unique structure begins to develop in this community
type (Tyrrell and Crow 1994). Typical of forest in pre-
settlement condition, most stands have ecological at-
tributes that are rare or absent in younger, managed
forests (e.g., 57 metric tons of coarse woody debris
ha '; >Vs of stand basal area in trees >70 cm diameter
at breast height; Haney, unpubl. data).

Since the late 1800s conifer-hardwood forest in the
eastern U.S. has been fragmented into isolated blocks,
markedly reduced in area, and converted into cover
types dominated by younger, shade intolerant hard-
woods. On the northern Appalachian Plateau, the old-
growth hemlock-hardwood forest once covered 2.4
million ha (Bjorkblom and Larson 1977), but today it
is reduced to no more than 20% of its presettlement
extent (Whitney 1990, Abrams and Ruffner 1995).

Data collection. — I used the Breeding Bird Census
(BBC; Hall 1964) to evaluate bird species occurrences
within old-growth habitat. Breeding Bird Census meth-
odology is u.sed primarily to assess local population
density by counting the number of breeding territories
on a few ha. Five BBC plots (15-18 ha each) were
located in Cook Forest, two ( 10-1 2 ha) in Heart’s Con-
tent, and four (12 ha) in Tionesta. Individual plots
within sites at each reserve ranged from 200 m to 2500
m apart. Breeding birds were counted during eight or
more visits to each plot during May and June 1994
using .standard protocols (Hall 1964, Lowe 1993).
Each plot was visited on a different day, usually within
a few minutes of sunrise. Two visits were made at
dusk. Each visit la.sted about two hours, which resulted
in a census speed of about 9 min ha ' and is compa-
rable to speeds deemed appropriate for relatively open
forests (Engstrom and James 1984).

On each visit, an observer walked slowly along a

flagged census line through the plot, delineating all
bird territories on grid maps. Birds were detected both
visually and acoustically, but most detections were
acoustic. The census line was configured to place the
observer no more than 50 m from any part of the plot
so as to reduce detection bias from acoustic attenuation
(Schieck 1997). Numbers of territories were then cal-
culated from grid maps using standard spot-mapping
procedures (Hall 1964).

The Pennsylvania Breeding Bird Atlas project
(1983-1989) was a grid-based survey using techniques
developed originally in Britain and Ireland, with stan-
dards modified for the northeastern U.S. (Laughlin
1982). Atlas projects are used primarily for broad map-
ping of avian distributions and rely upon a network of
volunteer field ornithologists to document breeding ev-
idence at three levels of certainty (“possible,” “prob-
able,” and “confirmed”). In Pennsylvania, the basic
sampling units consisted of 7.5' U.S. Geological Sur-
vey topographic maps divided into six equal-size
blocks formed longitudinally by 3.75' intervals and
latitudinally by 2.5' intervals (Brauning 1992). Atlas
efforts were undertaken in both summer and winter
within known “safe” dates for nesting activity of all
species. Based on previous theoretical and empirical
work, blocks were considered adequately covered if
75-80% of the expected species were found, 10-20
hours of survey effort were expended, or 70 or more
species were recorded.

Community level analyses. — To test whether sam-
pling effort was adequate for characterizing the total
species complement (5„,3J of the old-growth bird com-
munity, I conducted two analyses on the area curve of
cumulative species richness (5). The shape of species
accumulation curves depends on the order in which
samples are added, a feature not modeled well with
parametric methods (Bunge and Fitzpatrick 1993). I
used non-parametric routines to randomize sample or-
der (PISCES 1.2 software, Windows 95 version; Hen-
derson and Seahy 1997). For greater resolution in con-
structing the species accumulation curve, I first sub-
sampled the BBC data at a scale of 3 ha. From each
and all of the 1 1 original study plots, I randomly se-
lected and ordered 3 ha subplots and scored bird spe-
cies occurrences and territorial densities using methods
identical to those used in the original large plots.

Accurate estimation of is possible only if the
species accumulation curve is derived from a homoge-
nous community (Henderson and Seahy 1997). I first
compared the mean randomized curve (1000 itera-
tions) with a curve expected if all individual birds re-
corded over all the samples were assigned randomly
to individual samples (Colwell and Coddington 1994).
If the expected curve (Coleman et al. 1982) rises more
sharply from its origin, then heterogeneity is greater
than can be explained by chance. Such a result could
indicate that the samples were a combination of dis-
tinct bird communities or derived from different hab-
itats (Flather 1996).

Asymptotic models of species accumulation curves
are usually appropriate for homogenous communities

Haney • OLD-GROWTH BIRD COMMUNITY

(Henderson and Seahy 1997). I calculated using a
non-parametric maximum likelihood estimater (Raa-
ijmakers 1987) in which sampling is assumed to be
complete when the asymptotic estimate is equal to or
less than the observed. This procedure was applied in-
crementally to larger combinations of randomly shuf-
fled 3 ha subplots (1000 iterations each) until the
“stopping rule” indicated that sampling of the old-
growth bird community was sufficient.

Species level analyses. — I used incidence (frequency
in a set of samples; Wright 1991) as the metric to
compare individual species’ occurrences in oldgrowth
to their occurrence in landscape units. Incidence in
old-growth samples was calculated by dividing the
number of plots containing each species by 11. For
Pennsylvania and the northern Appalachian Plateau, I
used the proportion of BBA blocks recording that spe-
cies for each of the two landscape divisions. The atlas
program covered a total of 4928 and 2027 BBA blocks
state and province wide, respectively (Brauning 1992).
Species recorded in less than 25% and 10% of BBA
blocks in either landscape division were considered
uncommon and rare, respectively.

I compared incidence in oldgrowth (Iqg) to inci-
dence statewide (1st) and province-wide (I^p) with the
normal deviate, Z, where:

z = (I, - y/(i X (1 - I)(1/N, + i/N2))°'5,

and I and 1 — I are the joint probabilities of the com-
bined incidences in the two sample proportions of find-
ing and not finding that species, respectively (Snedecor
and Cochran 1980). I used Pearson’s product moment
correlation to test whether incidence was related to the
natural log of population density (number of breeding
territories). Log transformations on population density
were used to smooth variances in data composed of
whole integers (Snedecor and Cochran 1980). Values
of test statistics were considered significant at P <
0.05 unless otherwise indicated.

RESULTS

Community composition. — Fifty-six species
were recorded in 148 ha of old-growth forest
across the 11 study plots (Table 1). Thirteen
species were found in only one plot (incidence
value = 0.091). Another species, Downy
Woodpecker (Picoides pubescens), was re-
corded in two plots, but less than one full ter-
ritory was recorded in each plot. Without con-
sidering these 14 species, a recurring comple-
ment of 42 species was identified in which full
breeding territories were established in two or
more of the 1 1 study plots (Table 2).

The observed species accumulation curve
(Fig. 1) did not differ from the curve expected
in a homogenous community (x^ = 0.25, P >
0.05, df = 27). Thus, this analysis gave no
indication that more than one bird community

was being sampled. The estimated asymptotic
value (38.6) for species richness fell below the
observed value (39) after 1000 randomiza-
tions of 28 3-ha subplots. This level of effort
corresponded to 84 ha (57%) of the total area
actually sampled in this study.

Species groups. — About one-third of all
species recorded as breeders in old-growth co-
nifer-hardwood were either uncommon or rare
over broad spatial scales (Table 1). Nineteen
species (34%) were more likely to occur in
oldgrowth than in the landscape unit consist-
ing of the entire state. Sixteen species (29%)
were more likely to occur in oldgrowth than
in the landscape unit of the northern Appala-
chian Plateau. Fifteen individual species were
more likely to occur in oldgrowth than in the
landscape at both state and province levels
(Table 2). Red-shouldered Hawk (Buteo linea-
tus) and Barred Owl (Strix varia) were more
likely to occur in oldgrowth than in the land-
scape unit consisting of the entire state but not
the northern Appalachian Plateau.

Seventeen species were less likely to occur
in old-growth forest than in the landscape at
large (Table 2). This group included perma-
nent resident, habitat generalists [e.g., Amer-
ican Crow (Corvus brachyrhynchos)] as well
as some Neotropical migrants with more spe-
cific habitat preferences [e.g., cavity-nesting
Great Crested Flycatcher {Myiarchus crini-
tus)\.

No species showing negative association
with oldgrowth (Table 2) was rare at the state
level, and none of the species in this group
was either rare or uncommon at the level of
the physiographic province (Table 1). Only
one species. Black-throated Blue Warbler
(Dendroica caerulescens), was uncommon at
the state level (15% of BBA blocks). Most
species negatively associated with oldgrowth
were very widespread within broad landscape
units, occurring in 50-90% of the BBA
blocks.

Based on statistical criteria, 10 species were
neither more nor less likely to occur in old-
growth than in at least one of the larger land-
scape units (Tables 1 and 2). All species in
this group rely upon forest interior habitat, in-
cluding the raptors Red-shouldered Hawk and
Barred Owl, and Neotropical migrant song-
birds such as Red-eyed Vireo (Vireo oliva-

TABLE 1. Relative occurrence of breeding birds in old-growth hemlock-white pine-hardwood forest on the northern Appalachian Plateau, Pennsylvania. Incidence
(I) of each species in oldgrowth (proportion of study plots, n = 11) is compared to incidence across the entire state and within the physiographic province [proportion
of Breeding Birds Atlas (BBA) plots recording the species; n = 4928 and 2027 blocks, respectively].

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

0.2
a. r3

<5:

^ H

O O

in lo

o o o

o o

A A

VO o
in ov

d d
I

ON ro

' rn

CN ro VO Tf vD

— ' m (N 00 vO

' d ■ ■

o o

-vt (N in ro
in in ^ _

— ' 00

00 in in

<n m
I I

-It

in vD
vq C'
d 00 ri
I

■'tr^mvoinin-rj-
O 00 cn ^ (N
r-'-^ddr^vDodm

- I r I

in in (N
‘ o o
d d
A

ov in 00 o ro in in

OV O' VO ov (N — ^

0 —

o (N in

O' 00 O

— ' O' d 00 in rn

— VO 00 O' O'
_ _ _ (N| 00 O'

o in

d •sf

oommmO'cnt^O't^r'

^0(Nin'^m''^^ino

cNvodiNdirioddinin

I rsi I (N I

o m in C'

O (N m
n) m 00 (N

d d d d

•'ifO'inmmvooovo
O— 'in^t^vovoo
t^rooovoooinooo
dddddddd

'It VO O
m in It

O' It
m VO O'

O
-It

cNvqvqovO'(Nvoo'

dddddddd

00 tn

d d

O' CN (N O'

O' O' 00 —I

O 00 O' O'

o o o o

Or)— ;o^(Nt''.— -oovDO'inooo<nrnr':0^0;-Hooooooor'iooo(NO'0'^cN

oooooooodddddddddddddddddddddddd

— '■^minrrmno'
O'O'vot^ini^t^ooo
oorrn^nr)— 'O'
ddddddddd

in — -H00(Nrivonr)m
'fO'O' — oooomoooot^
inoooq — — VO — — r)

dddddddddd

O' <N VO

o 00 m

(N vt n

00 VO 00

— m —

d d d

£0

2 '3

A 5

“O

u«

-C

c/5

a d

? '3

A >

^ s

’■2 u,
OX) B

c ^

C 3
3 1/3

3 rv
3 ct

X ^
s .JJ
o

5u s;

cx -i:

a *0
x:

cn oC

V ^

1-1 W

3 <U

V. ^

I °

>> ^
C >,
^ .b

O 3

Q X

1-1

- 2
E

§■ E
o 2
C‘ 5
Q U

U (U
<D (D

O ^

a ^
V -6

.>1

j-i

? t

'S.

o .

B £
S Si

D c/5
rz CT3
CL W

O u

>> (U

E o

4-*

•O 3

D O

"B tL
X3

_ g

, ^ S

S- 3 ^
S: o 25

^ ^ -S

^ E 2
u o
-o s:

JC 1) Q

o ^

c/5

C3 (U U

pi' ^ >v
pL, t- Cd

(U •-5

^ ti
c/5 C3

(U 3

<U t
J O CQ

■2

■S d

^ r
-3 ^

"5 ^ r-

^11

^ =3

(u n

C4l U
3

'g -S
S "S
Jr*

t; ^

2 a

c/5 (U
C3 (U
D »-)

i I

E p

c e -p

oo a ^
r ^ ^

b rSi 2

*- Q; ^

^ W <3

^ -a

^ (U ■***

^ OOCJ
,2 3 0

2? 2
p ^ t?

d T3 3

fc ^ I

^ V S

s ^

2 ^

1 i
2l

a -2
■s; Cl

2 -2

S 2

G O
d 2

■CS -i;

o 1)
3 2

2 ^
jz 2

4-1 C

^ O

0) P
X ^

.2 J=

^ 3

Q. 1/3

V 2

2; E
- o 5

C4-( cy T3

H03a:^cQ^Ocna:^<um

•S >

TABLE 1. CONTINUED

Haney • OLD-GROWTH BIRD COMMUNITY

a."

O. <T3

<a:

r<P

1 °

CTv

1 ^

r-;

—

r-J

I I

I I I

C^l

1 ^

"=t

c^i

I I

I I I

I r I

»n

C^l

"“V

-V

“2

y-S ^

2 d

§2

C JV

2 §

5 Q

c to
"S u

Q -B

2 I
02 >1

<3 -ii

-s: a

-S-f

^ .=0

•t c
> Q,

1) "5

0-
u £

V j=

•a t

(U O

01 Z

“ m

OJ ^

S Q

U ■>-'

rt .

•o ■£ -a
<L> d (U

:h ^ s

c/5 O

' d 5-

3 - ■£

C O I

c ^

S 2P

— d

•D

<j (L)

S! ^

^ (U

■e 3

b Ui

2 «

s 3

^ !3

13 ^

s ^

Cl B

1- ^
(U ?

^ s

^ d

5^ ^

i'l

^ 2 o

CO 2
^ ^3

t: ^

d 5

-o -2

V ^

OeC 52

d u

cj .-

?: O

ScB

o ^

^ t;

u t;

0{)

w -- qj c-#

iH « c «

m CQ ix pq

=: xi
c ^

O XJ
(U
ID
O
O

^ c:

CJ o
'D 5

?3 ^
00 ^
c* d

2 ^

^ £
o

'U
00
C3

.c
'Z' ^

^ S

S
E

U X ^ 1^

~ lU
C3 «

c •!:/

tb ^

^ s

Q. ^
^ o
00 t

C C3

■a a

C C/P

m ^

o '5.

.°‘^ a

V 3
3 U

5U O

■^'5

^ TO

o u

c -o
3 U
■—I "3
■n ^
3

^ 'S

S. a

s ;3
■2 ^
^ <—
^ O

2 c

V c

d u

Q CO

-9 ^

^ Test statistic (normal deviate) on the difference between proportions: positive values of Z indicate a greater incidence of that species in oldgrowth, negative values indicate lower incidence in oldgrowth.

^ Unrecorded during the state s 5 year BBA project. From 1993-1995, an irruption of the elm spanworm Ennomus subsignarius affected much of Pennsylvania’s forests. Several birds influenced by geometrid outbreaks were
recorded during these breeding seasons for the first time: Bay-breasted Warbler, Blackpoll Warbler (Dendroica striata) and Evening Grosbeck {Coccothraustes vespertinus).

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

TABLE 2. Number of territories and population densities (number territories/ 1 0 ha) of the core community
of breeding birds in all plots {n = 1 1 ) of old-growth hemlock-white pine-hardwood forest on the northern
Appalachian Plateau, Pennsylvania.

Territorial density

Association: Species“ Total territories Mean SE

Positive old-growth affinity:

Hairy Woodpecker (Picoides villosus)

12.5

0.50

0.42

Acadian Llycatcher (Empidonax virescens)

24.5

1.27

1.57

Red-breasted Nuthatch (Sitta canadensis)

0.40

0.36

Brown Creeper (Certhia americana)

1.58

0.71

Winter Wren {Troglodytes troglodytes)

30.5

1.73

0.98

Golden-crowned Kinglet (Regulus satrapa)

0.10

0.23

Swainson’s Thrush (Catharus ustulatus)

2.08

2.42

Hermit Thrush {Catharus guttatus)

36.5

1.90

1.49

Blue-headed Vireo {Vireo solitarius)

4.57

2.26

Magnolia Warbler {Dendroica magnolia)

185

11.10

2.70

Black-throated Green Warbler {Dendroica virens)

176

9.41

5.36

Blackburnian Warbler (Dendroica fusca)

410

23.94

8.25

Pine Warbler {Dendroica pinu.s)

0.17

0.25

Dark-eyed Junco {Junco hyemalis)

4.09

0.92

Purple Pinch {Carpodacus purpureas)

0.72

0.52

Neutral oldgrowth affinity:

Red-shouldered Hawk {Buteo lineatusf

2.5

0.09

0.13

Barred Owl (Strix varia)'°

4.5

0.18

0.24

Yellow-bellied Sapsucker {Sphyrapicus varius)

0.14

0.26

Pileated Woodpecker {Dryocopus pileatus)

5.5

0.25

0.31

Common Raven {Corvus corax)

1.5

0.07

0.19

Black-capped Chickadee {Poecile atricapillus)

1.14

0.62

Red-eyed Vireo {Vireo olivaceus)

110

5.22

3.06

Hooded Warbler {Wilsonia citrina)

0.87

1.40

Scarlet Tanager {Piranga olivacea)

42.5

2.23

1.08

Chipping Sparrow {Spizella passerina)

1.00

0.63

Negative old-growth affinity:

Mourning Dove {Zenaida macroura)

0.09

0.21

Ruby-throated Hummingbird {Archilochus coluhris)

0.06

0.11

Least Llycatcher {Empidonax minima.^)

3.5

0.18

0.53

Great Crested Llycatcher {Myiarchus crinitus)

1.5

0.08

0.18

Blue Jay {Cyanocitta cristata)

10.5

0.45

0.35

American Crow {Corx’us brachyrhynchos)

1.5

0.06

0.14

Tufted Titmouse {Baeolophus hicolor)

0.11

0.19

White-breasted Nuthatch {Sitta carolinensis)

7.5

0.27

0.31

Wood Thrush {Hylocichla mustelina)

1.5

0.09

0.25

American Robin (Tiirdus migratorius)

4.5

0.20

0.40

Cedar Waxwing (Bombycilla cedrorum)

0.10

0.23

Black-throated Blue Warbler {Dendroica caerulescens)

1.08

1.46

American Redstart {Setophaga ruticilla)

7.5

0.43

1.28

Ovenbird (Seiurus aurocapillus)

0.46

0.95

Common Yellowthroat {Geothlypis tricha.'i)

0.06

0.14

Rose-breasted Grosbeak (Pheucticus ludovicianus)

0.24

0.44

Brown-headed Cowbird {Molothrus ater)

0.32

0.58

Does not include species found only in one plot or for which less than one full territory was recorded (Table 1).
^ Positively associated with oldgrowth at landscape level of entire state but not at level of physiographic province.

Hane\ • OLD-GROWTH BIRD COMMUNITY

FIG. 1. Mean randomized accumulation curve
(1000 iterations) of species richness in the bird com-
munity of old-growth conifer-hardwood forest on the
northern Appalachian Plateau, Pennsylvania. Horizon-
tal line indicates the putative asymptote of species
richness as a function of area sampled. The asymptote
was identified with a maximum likelihood estimator.

cells). Hooded Warbler (Wilsonia citrina), and
Scarlet Tanager (Piranga olivacea).

Incidence versus population density.- — Bird
species that were rare or uncommon at land-
scape levels typically had low population den-
sities locally as well. Incidence explained
slightly more than 80% of the variation in the
natural log of population size as assessed by
territorial density (Fig. 2).

DISCUSSION

Old-growth affinities. — At both province
and state levels, more than one-third of bird
species were more likely to occur in old-
growth conifer-hardwood forest than in the
broader landscape. These species included
some of Pennsylvania’s rarest breeding birds.
Yellow-bellied Flycatcher (Empidonax flavi-
ventris) and Swainson’s Thrush (Catharus us-
tulatus) are listed in the state as threatened
and candidate-rare, respectively (D. A. Gross,
pers. comm.). Pending investigation of specif-
ic habitat preferences for individual species,
the 15 birds in this group (Table 2) are best
regarded as old-growth associates rather than
old-growth obligates. Nevertheless, it is no-
table that these species have diverse habitat
affinities, including conifer [e.g.. Red-breasted
Nuthatch {Sitta canadensis)], hardwood [e.g..
Blue-headed Vireo (Vireo solitarius)], and
mixed forest cover types [e.g.. Hairy Wood-
pecker (Picoides villosus)].

Several taxa identified as old-growth asso-
ciates in this study have been linked repeat-
edly to late successional forest elsewhere
throughout North America. Hairy Woodpeck-

>-
in -t->

.2? m 2
o c
0) 0)

Q. -O

c 0

l4_ O

o :::
ro

3 -2

c a.

— o
a.

-4

0 0.2 0.4 0.6 0.8 1.0

Species landscape incidence

FIG. 2. Relationship between natural log of pop-
ulation density (number of breeding territories) and
landscape incidence of all bird species (n = 56) found
in old-growth conifer-hardwood forest of the northern
Appalachian Plateau, Pennsylvania. Landscape inci-
dence is based on the proportion of old-growth study
plots (n = 11; Table 1 ) in which the species was re-
corded. Log of population density (y) is related to
landscape incidence (x) by: y = 4.74 lx — 3.513, (r^
= 0.805, P < 0.001).

ers rely on mature forests with large snags and
logs (Anthony et al. 1996, Shackelford and
Conner 1998). Across broad geographic do-
mains, Red-breasted Nuthatch, Brown Creep-
er (Certhia americana). Winter Wren {Trog-
lodytes troglodytes), and Golden-crowned
Kinglet (Regulus satrapa) form a recurring
group that exploits coniferous cover and com-
plex structure typical of older natural forests
(DeGraaf et al. 1992, Hansen et al. 1995,
Schieck et al. 1995, Anthony et al. 1996, Del-
lasala et al. 1996).

In other cases, regional counterparts of gen-
era exhibited a common tendency to prefer
mature forest. Acadian Flycatcher {Empidon-
ax virescens) showed a greater likelihood of
occurring in mesic old-growth forest relative
to adjacent landscapes (Table 1), similar to
Hammond’s {E. hammondii) and Pacific-slope
{E. difficilis) flycatchers in Oregon and British
Columbia (Hansen et al. 1995, Schieck et al.
1995). Like their western congeners Hermit
{Dendroica occidentalis) and Townsend’s
warblers {D. townsendi; Hansen et al. 1995,
Schieck et al. 1995), several species of arbo-
real Dendroica warblers were far more likely
to occur in old-growth hemlock-white pine-
northem hardwood forest than in adjacent

THE WILSON BULLETIN • VoL III, No. I, March 1999

landscape units (Tables 1 and 2). Populations
of Blackburnian Warblers (D. fitsca) achieve
particularly high densities in the oldest coni-
fer-hardwood forests of this region (Haney
and Schaadt 1996: fig. 6.1).

Two species, Red-shouldered Hawk and
Barred Owl, were more likely to occur in old-
growth than across the state as a whole (Table
1). Red-shouldered Hawks depend on mature
forests with large trees for nest sites (Titus and
Mosher 1981, Moorman and Chapman 1996).
Barred Owls exhibit greater territorial occu-
pancy and breeding propensity in this region’s
old-growth forest (Haney 1997).

Despite directional biases in comparisons of
incidence across spatial scales (see Sampling
adequacy), negative associations of bird spe-
cies with oldgrowth may have had biological
causes. Mourning Dove {Zenaida macroura).
Blue Jay (Cyanocitta cristata), American
Crow, and Brown-headed Cowbird (Moloth-
rus ater. Table 2) all typically exploit land-
scapes with extensive anthropogenic distur-
bance (Martin 1988, Hoover and Brittingham
1993, Seitz and Zegers 1993, Rodenhouse et
al. 1995). Consequently, they would be less
expected to occur in mature tracts of reserved
forest. Least Flycatcher {Empidonax mini-
mus), Black-throated Blue Warbler, American
Redstart {Setophaga ruticilla), and Rose-
breasted Grosbeak (Pheucticus ludovicianus)
more commonly exploit the deciduous habi-
tats (Sherry and Holmes 1988, Steele 1993,
Yahner 1993) typical of younger, regenerating
forest now prevalent in this region (Alerich
1993). Ovenbirds (Seiurus aurocapillus) were
probably scarce because of their preference
for heavy ground cover (Burke and Nol 1998),
a microhabitat virtually absent in the old-
growth forest studied here.

Sampling adequacy. — Deletion of periph-
eral species and analysis of the asymptote on
the species accumulation curve gave similar
values for total species: 42 and 39 species,
respectively. I conclude that census effort was
adequate for characterizing the avian com-
munity in old-growth hemlock-white pine-
northern hardwood forest on Pennsylvania’s
northern Appalachian Plateau. Numerical an-
alyses indicated that the community sampled
was in fact homogenous and an asymptotic
limit to species richness (5’, „;,;<) was achieved
with little more than half the sampling effort

actually undertaken (Fig. 1). Support for sam-
pling adequacy is reassuring because limited
amounts and local distributions of eastern old-
growth forest often preclude obtaining larger
sample sizes and greater sample dispersion in
this scarce habitat type.

Use of bird species richness {S or S^^) to
evaluate avian habitat can be problematic un-
less studies account for: (1) “core” members
of the avifauna, (2) quantity and quality of
sampling effort, (3) number of habitat types
within areas, and (4) proximity of other hab-
itats (Remsen 1994, Elphick 1997). The BBC
method itself purposefully discounts non-ter-
ritory holders, thereby eliminating nonbreed-
ing species. I also established study plots
within interiors of old-growth forest so as to
avoid inflating or confounding species rich-
ness caused by proximity of different habitat
patch types (Flather 1996).

Although BBC and BBA methods have dis-
tinct purposes and can have different quanti-
ties and qualities of observer effort, several
factors facilitated comparisons of data from
the two techniques in this study. First, species
occurrence data from both methods were com-
parable by developing a common incidence
metric. Second, both methods rely to some ex-
tent upon a measure of saturation in the cu-
mulative number of species recorded in order
to guage whether sampling is adequate. Third,
both methods had similar levels of observer
effort as measured by survey duration. All
BBCs took 16.7—20.1 h to complete versus an
average of 17 field-h per atlas block (Brauning
1992).

The BBC method’s reliance on three or
more records to score territorial occupancy,
however, is more restrictive in tallying species
occurrences than the BBA method. The latter
includes “possible,” “probable,” and “con-
firmed” categories of breeders, and is there-
fore likely to include more species per unit
effort. Greater numbers of species may also
be detected with the BBA method because of
the substantially larger areas covered (poten-
tially hundreds or thousands of ha per block
versus the tens of ha in most BBCs).

As a consequence of differences in the
scope of effort between BBC and BBA meth-
ods, comparisons of incidence values (I) for
individual species (Table 1) may be biased
against detecting greater occurrence (and to-

Haiie\ • OLD-GROWTH BIRD COMMUNITY

wards detecting lower occurrence) in old-
growth habitat than in the landscape at large.
Findings of positive old-growth association by
individual species (Table 2) are more robust
as a result. Negative and neutral associations
with oldgrowth should be interpreted cau-
tiously because more liberal listing of species
under BBA methodology could elevate rela-
tive incidence values at state and province lev-
els, thereby leading to false conclusions that
no differences in species occurrences existed
across spatial scales (Type II error).

Hierarchical comparisons. — Although
scale is viewed as essential for interpreting
distributional data in birds (Lacy and Bock
1986), logistical constraints and methodolog-
ical inconsistencies often prevent hierarchical
or multi-tiered approaches. Comparing local
density of bird populations to density in a re-
gion as large as an entire state is impossible
because the BBC method requires large in-
vestments in time for limited spatial coverage.
Proportions are easy to derive from virtually
any kind of sample, however, and a metric
based on incidence enabled direct comparison
of species occurrences in BBC plots and atlas
blocks (Table 1).

Comprehensive coverage in Pennsylvania’s
atlas program also enabled more reliable com-
parisons of birds in oldgrowth to the wider
landscape: all blocks, including those on the
state’s borders, were censused (Brauning
1992). Synoptic coverage allows evaluation of
the likely impacts of potential actions on
groups of bird species within a wider context.
It would be easy to scale down from the eco-
physiographic province or state levels used in
this study to some smaller region of interest
(e.g., county, national forest, watershed). Al-
ternatively, BBA data from adjacent states
could be aggregated to examine individual
species occurrences across even larger land-
scape units. This spatial flexibility should en-
able better evaluation of potential consequenc-
es of local management prescriptions on the
regional distributions of birds.

I used an incidence metric as a reasonable
proxy for population size (Fig. 2). Several re-
searchers have documented a general relation-
ship between abundance and range size in
birds (Bock and Ricklefs 1983, Lacy and
Bock 1986, Mauer and Hey wood 1993). This
relationship may not indicate the existence of

a particular ecological hypothesis (Wright
1991). Nevertheless, the generality that spe-
cies with sparse distributions also have low
population densities was confirmed in this
study by documenting regional scarcity in
several bird species that use a rare and very
local habitat type.

ACKNOWLEDGMENTS

I thank B. Allison, J. Cheek, L. Hepfner, R. Kauf-
mann, J. Lydic, C. Schaadt, J. Seachrist, J. Smreker, S.
Weilgosz, S. Wetzel, and R. Williams for their help in
conducting the Breeding Bird Censuses and vegetation
surveys; D. DeCalesta, C. Nowack, J. Palmer, S. Stout,
C. Schlentner, L. Lentz, J. Sowl, and D. Wright for
facilitating logistic anangements; and Cook Forest
State Park and the USD A Forest Service for research
access. Financial support was provided by the Wilder-
ness Society, Dodge Foundation, Johnson and John-
son, Sweet Water Trust, William P. Wharton Trust,
Pennsylvania Wild Resource Conservation Fund,
Pennsylvania Game Commission, Center for Rural
Pennsylvania, DuBois Educational Foundation Fund
for Academic Excellence, and Pennsylvania State Uni-
versity Research and Development funds.

LITERATURE CITED

Abrams, M. D. and D. A. Orwig. 1996. A 300-year
history of disturbance and canopy recruitment for
co-occurring white pine and hemlock on the Al-
legheny Plateau, USA. J. Ecol. 84:353—363.
Abrams, M. D. and C. M. Ruffner. 1995. Physio-
graphic analysis of witness-tree distribution
(1765-1798) and present forest cover through
north central Pennsylvania. Can. J. For. Res. 25:
659-668.

Alerich, C. L. 1993. Forest statistics for Pennsylva-
nia-1978 and 1989. U.S. Dept. Agric. For. Serv.
Res. Bull. NE- 126: 1-244.

Anthony, R. G., G. A. Green, E. D. Forsman, and
S. K. Nelson. 1996. Avian abundance in riparian
zones of three forest types in the Cascade Moun-
tains, Oregon. Wilson Bull. 108:280—291.
Bjorkblom, j. C. and R. G. Larson. 1977. The Tio-
nesta scenic and research natural areas. USDA
For. Serv. Gen. Tech. Rept. NE-31:1— 24.

Bock, C. E. and R. E. Ricklefs. 1983. Range size and
local abundance of some North American song-
birds: a positive coirelation. Am. Nat. 122:295-
299.

Brauning, D. W, (Ed.). 1992. Atlas of breeding birds
in Pennsylvania. Univ. of Pittsburgh Press. Pitts-
burgh. Pennsylvania.

Bunge, J. and M. Fitzpatrick. 1993. Estimating the
number of species: a review. J. Am. Stat. Assoc.
88:364-373.

Burke, D. M. and E. Nol. 1998. Influence of food
abundance, nest-site habitat, and forest fragmen-
tation on breeding Ovenbirds. Auk 115:96-104.

THE WILSON BULLETIN • Vol. HI, No. I. March 1999

Coleman, M. D., M. D. Mares, M. R. Willig, and
Y.-H. Hsieh. 1982. Randomness, area, and species
richness. Ecology 63:1121-1133.

Colwell. R. K. and J. A. Coddington. 1994. Esti-
mating terrestrial biodiversity through extrapola-
tion. Phil. Trans. Royal Soc. (B) 345:101-118.

Conroy, M. J. and B. R. Noon. 1996. Mapping of
species richness for conservation of biological di-
versity: conceptual and methodological issues.
Ecol. Appl. 6:763-773.

Davis, M. B. 1996. Extent and location. Pp. 18-32 in
Eastern old-growth forests: prospects for rediscov-
ery and recovery (M. B. Davis, Ed.). Island Press,
Washington, D.C.

DeGraaf, R. M., M. Yamasaki. W. B. Leak, and J.
W. L.anier. 1992. New England wildlife: manage-
ment of forested habitats. US DA Eor. Serv.,
Northeast For. Exp. Station, Gen. Tech. Rept. NE-
144:92.

Dellasala, D. a., j. C. H.agar, K. A. Engel, W. C.
McComb. R. L. Fairbanks, and E. G. Campbell.
1996. Effects of silvicultural modifications of tem-
perate rainforest on breeding and wintering bird
communities. Prince of Wales Island, southeast
Alaska. Condor 98:706-721.

Elphick, C. S. 1997. Correcting avian richness esti-
mates for unequal sample effort in atlas studies.
Ibis 139:189-190.

Engstrom, R. T. and E C. James. 1984. An evaluation
of methods used in the Breeding Bird Census.
Am. Birds 38:19-23.

Eyre, E H. (Ed.). 1980. Forest cover types of the Unit-
ed States and Canada. Society of American For-
esters, Washington, D.C.

Feather, C. H. 1996. Fitting species-accumulation
functions and asse.ssing regional land use impacts
on avian diversity. J. Biogeogr. 23:155-168.

Hall, G. A. 1964. Breeding bird censuse.s — why and
how. Audubon Field-Notes 18:413-416.

Haney, J. C. 1997. Spatial incidence of Baired Owl
iStrix varia) reproduction in old-growth forest of
the Appalachian Plateau. J. Raptor Res. 31:241-
252.

Haney, J. C and C. P. Schaadt. 1996. Functional roles
of eastern old growth in promoting forest bird di-
versity. Pp. 76-88 in Eastern old-growth forests:
prospects for rediscovery and recovery (M. B. Da-
vis, Ed.). Island Press, Washington, D.C.

Hansen, A. J.. W. C. McComb, R. Vega, M. G. Ra-
phael, AND M. Hunter. 1995. Bird habitat rela-
tionships in natural and managed forests in the
west Cascades of Oregon. Ecol. Appl. 5:555-569.

Hender.son, P. a. and R. M. H. Seahy. 1997. Species
diversity and richness, version 1.2. PISCES Con-
servation Ltd., Lymington, U.K.

Hoover, J. P. and M. C. Brittingham. 1993. Regional
variation in cowbird parasitism of Wood Thrush-
es. Wifson Bull. 105:228-238.

Hough, A. E and R. D. Forbes. 1943. The ecology
and silvics of Pennsylvania high-plateau forests.
Ecol. Monogr. 13:299-320.

Hunter, J. E.. R. J. Gutierrez, and A. B. Franklin.
1995. Habitat configuration around Spotted Owl
sites in northwestern California. Condor 97:684-
693.

Keys, J. E., Jr., W. H. McNab, and C. A. Carpenter
(Eds.). 1995. Ecological units of the eastern Unit-
ed States — first approximation. USDA, Forest
Service. Atlanta, Georgia.

Kotliar, N. B. and j. a. Wiens. 1990. Multiple scales
of patchiness and patch structure: a hierarchical
framework for the study of heterogeneity. Oikos
59:253-260.

Lac^ . R. C. AND C. E. Bock. 1986. The correlation
between range size and local abundance of some
North American birds. Ecology 67:258-260.

Lalighlin, S. B. (Ed.). 1982. Proceedings of the north-
eastern breeding bird atlas conference. Vermont
Institute of Natural Sciences, Wood.stock, Ver-
mont.

Leverett, R. 1996. Definitions and history. Pp. 3—17
in Eastern old-growth forests: prospects for redis-
covery and recovery (M. B. Davis, Ed.). Island
Press, Washington. D.C.

Lowe, J. D. 1993. Resident bird counts. 1992. J. Field
Ornithol. 64( Supplement ):3-5.

Martin, T. E. 1988. Habitat and area effects on forest
bird assemblages: is nest predation an influence?
Ecology 69:74-84.

Maurer, B. A. and S. G. Heywood. 1993. Geographic
range fragmentation and abundance in Neotropical
migratory birds. Conserv. Biol. 7:501-509.

Moorman, C. E. and B. R. Chapman. 1996. Nest-site
selection of Red-shouldered Hawks in managed
forest. Wilson Bull. 108:357-368.

Nichols. G. E. 1935. The hemlock-white pine-north-
ern hardwood region of eastern North America.
Ecology 16:403-422.

Raaijmakers. j. G. W. 1987. Statistical analysis of the
Michaelis-Menten equation. Biometrics 40:119-
129.

Remsen, j. V., Jr. 1994. Use and misuse of bird lists
in community ecology and conservation. Auk
1 I 1 :225-227.

Rodenhouse, N. L., L. B. Best, R. J. O'Connor, and
E. K. Bollinger. 1995. Effects of agricultural
practices and farmland structures. Pp. 269—293 in
Ecology and management of Neotropical migra-
tory birds (T. E. Martin and D. M. Finch. Eds.).
Oxford Univ. Press, Oxford. U.K.

Schieck, j. 1997. Biased detection of bird vocaliza-
tions affects comparisons of bird abundance
among forested habitats. Condor 99:179-190.

.Schieck. J., K. Lertzman, B. Nyberg, a'nd R. Page.
1995. Effects of patch size on birds in old-growth
montane forests. Conserv. Biol. 9:1072-1084.

Seitz, L. C. and D. A. Zegers. 1993. An experimenlal
study of nest predation in adjacent deciduous, co-
niferous and successional habitats. Condor 95:
297-304.

Shackelford. C. E. and R. N. Conner. 1998. Wood-

Hanex • OLD-GROWTH BIRD COMMUNITY

pecker abundance and habitat use in three forest
types in eastern Texas. Wilson Bull. 109:614-629.

Sherry, T. W. and R. T. Holmes. 1988. Habitat selec-
tion by breeding American Redstarts in response
to a dominant competitor, the Least Flycatcher.
Auk 105:350-364.

Snedecor, G. W. and W. G. Cochran. 1980. Statistical
methods, seventh ed. Iowa State Univ. Press,
Ames.

St.ahle, D. W. 1996. Tree rings and ancient forest his-
tory. Pp. 321-343 in Eastern old-growth forests:
prospects for rediscovery and recovery (M. B. Da-
vis, Ed.). Island Press, Washington, D.C.

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

Titus, K. and J. A. Mosher. 1981. Nest-site habitat
selected by woodland hawks in the Central Ap-
palachians. Auk 98:270-281.

T'i rrell, L. E. and T. R. Crow. 1994. Structural char-

acteristics of old-growth hemlock-hardwood for-
ests in relation to stand age. Ecology 75:370-386.

USDA. 1997. Guidance for conserving and re.storing
old-growth forest communities on national forests
in the Southern Region. U.S. Department of Ag-
riculture, Forest Service, Southern Region, Atlan-
ta, GA. Forestry Rept. R8-FR 62:35-38.

Whitney, G. G. 1990. The history and status of hem-
lock-hardwood forests of the Allegheny Plateau.
J. Ecol. 78:443-458.

Whittaker, R. H. 1977. Evolution of species diversity
in land communities. Evol. Biol. 10:1-67.

Wiens, J. A. 1989. The ecology of bird communities:
foundations and patterns. Cambridge Univ. Press,
Cambridge, U.K.

Wright, D. H. 1991. Correlations between incidence
and abundance are expected by chance. J. Bio-
geogr. 18:463-466.

Yahner, R. H. 1993. Effects of long-term forest clear-
cutting on wintering and breeding birds. Wilson
Bull. 105:239-255.

Wilson Bull.. 111(1), 1999, pp. 100-104

EFFECTS OF WIND TURBINES ON UPLAND NESTING BIRDS IN
CONSERVATION RESERVE PROGRAM GRASSLANDS

KRECIA L. LEDDY,' 3 KENNETH E HIGGINS,^^ and DAVID E. NAUGLE'^

ABSTRACT. — Grassland passerines were surveyed during summer 1995 on the Buffalo Ridge Wind Resource
Area in southwestern Minnesota to determine the relative influence of wind turbines on overall densities of
upland nesting birds in Conservation Reserve Program (CRP) grasslands. Birds were surveyed along 40 m fixed
width transects that were placed along wind turbine strings within three CRP fields and in three CRP fields
without turbines. Conservation Reserve Program grasslands without turbines and areas located 180 m from
turbines supported higher densities (261.0—312.5 males/ 100 ha) of grassland birds than areas within 80 m of
turbines (58.2-128.0 males/100 ha). Human disturbance, turbine noise, and physical movements of turbines
during operation may have distrurbed nesting birds. We recommend that wind turbines be placed within cropland
habitats that support lower densities of grassland passerines than those found in CRP grasslands. Received 9
Sept. 1997, accepted 5 Oct. 1998.

Technological advances that have reduced
the cost of electricity generated from wind-
plants have enabled the wind-power industry
to expand from California into the eastern
United States and Canada (Nelson and Curry
1995). Wind power has received strong public
support as an alternative energy source despite
the potential threats that the presence of wind
turbines may pose to avian species. Recent re-
search has indicated that raptor mortality from
collisions with wind turbines varies greatly
from no mortality (Higgins et al. 1996; Us-
gaard et al., in press) to substantial mortality
(Orloff and Flannery 1992). In addition to di-
rect mortality from collisions, research also
has indicated that waterfowl, wading bird, and
raptor densities near turbines were lower com-
pared to densities in similar habitats away
from turbines (Winkelman 1990; Pedersen
and Poulsen 1991; Usgaard et al., in press).
The influence of wind turbines on grassland
nesting passerine species has not been previ-
ously measured.

Recent construction of the first windplant
facility in the midwestern United States pro-

' Dept, of Wildlife and Fisheries Sciences, Box
2I40B, South Dakota State Univ., Brookings, SD
57007.

^ South Dakota Cooperative Fish and Wildlife Re-
.search Unit, USGS-BRD, South Dakota State Univ.,
Box 2140B, Brookings, SD 57007.

’ Pre.sent address; Natural Re.sources Conservation
Service, RR 1 Box 740, Webster, SD 57274.

Present address: College of Natural Resources,
Univ. of Wisconsin-Stevens Point, Stevens Point, WI
5448 1 .

' Corresponding author.

vided a unique opportunity to study the effects
of wind turbines on grassland nesting passer-
ines. Several midwestern grassland passerine
species have declined in abundance (Johnson
and Schwartz 1993) in response to agricultural
tillage, grazing, and invasive woody species
that have destroyed or degraded most of the
remaining grasslands (Kantrud 1981, Castrale
1985). Although Conservation Reserve Pro-
gram (CRP; Young and Osborn 1990) grass-
lands provide habitat for grassland nesting
birds (Johnson and Schwartz 1993, Igl and
Johnson 1995, Johnson and Igl 1995, King
and Savidge 1995, Millenbah et al. 1996), the
potential impact of wind turbines in CRP
fields could negate those benefits. The objec-
tive of this study was to determine whether
density of upland nesting passerines in CRP
grasslands was influenced by the presence of
wind turbines. We hypothesized that bird den-
sity in CRP grasslands would not differ in re-
lation to distance from wind turbines.

STUDY AREA AND METHODS

Study area. — The Buffalo Ridge Wind Resource
Area (WRA) in southwestern Minnesota is located
along a 100 km segment of the Bemis Moraine near
Lake Benton, Minnesota. Elevation is 546-610 m.
Wind turbines cover 32 km^ of the 293 km^ Buffalo
Ridge WRA. Additional lands within, the Buffalo
Ridge WRA have been leased as future wind-turbine
development sites. The windplant contains 73 opera-
tional wind turbines that are arranged in 10 turbine
strings, with 3-20 turbine.s/string. Turbines are 91-183
m apart within strings. Turbines (model KVS-33; KE-
NETECH Windpower, Inc.), which operate at wind
speeds of 14-104 km/h, consist of a 33 m diameter
rotor mounted on a 37 m tubular tower.

100

Lcdd\ et cd • EFFECTS OF WIND TURBINES ON BIRDS

101

Upland grassland bird nesting habitat within the
Buffalo Ridge WRA consisted primarily of CRP grass-
lands, mostly planted with a mixture of smooth brome
(Bronius inermis) and alfalfa (Medica^’o saliva) or
switchgrass (Panicmn vir^atum). Habitats surrounding
CRP grasslands were agricultural lands dominated by
corn {Zea mays) and soybeans {Glycine max) with
smaller areas of haylands, pasturelands, and scattered
woodlands near farmsteads and in ravines.

Methods. — Bird survey transects were placed along
wind-turbine strings within three CRP fields and in
three CRP fields without turbines (i.e., control; Feddy
1996). We selected CRP fields that were 7-8 years of
age to minimize effects of field age on diversity and
density of avain species (Millenbah et al. 1996). Visual
obstruction readings (Robel et al. 1970, Higgins and
Barker 1982) did not differ between CRP grasslands
with and without turbines, indicating that vegetation
structure in experimental and control fields was similar
(Feddy 1996). Transects were surveyed weekly in ran-
dom order from 15 May to 1 July 1995. Multiple sur-
veys of a single transect were averaged into one bird
density to avoid pseudoreplication (Hurlbert 1984). Six
40-m fixed width transects (Wakeley 1987) paralleling
each turbine string were used per field. One transect
ran directly underneath turbine strings. Two additional
transects on each side of the turbine string paralleled
the string at distances of 40 and 80 m; the sixth tran-
sect was placed 180 m from the turbine string. Tran-
sects varied in length according to field size and were
placed at least 30 m from field borders and wetlands
to minimize bias associated with edges (Arnold and
Higgins 1986. Reese and Ratti 1988). One transect was
established at a random location in each of the three
control CRP fields without turbines.

Inconsistencies among surveys attributable to peri-
odic bird inactivity (Skirvin 1981, Verner and Ritter
1986) were minimized by conducting surveys from
sunrise to 10:00 CST We recorded all birds seen or
heard while walking transects at 1.0- 1.5 km/h (Mikol
1980, Wakeley 1987); only perched and/or singing
males were used in statistical analyses. Flushed birds
seen leaving transects were counted (Burnham et al.
1980), whereas birds seen entering transects or flying
overhead were not counted. Surveys were not con-
ducted during heavy rain or high winds (^20 km/h;
Ralph et al. 1993). Birds were surveyed in CRP fields
with turbines when turbines were operational and non-
operational because turbines began operating during
surveys when wind speeds reached 14—20 km/h. We
compared surveys that were conducted during opera-
tional and non-operational periods to determine wheth-
er noise produced during turbine operation biased sur-
veys.

An index of total breeding bird density was calcu-
lated by dividing the number of perched and/or singing
males by transect area. Percent species composition
was calculated by dividing the number of perched and/
or singing males of a particular species by the total
number of males. Species richness was defined as the
number of species (Koford et al. 1994).

Analysis of covariance (SAS 1989) was used to de-
termine whether bird density across transects was re-
lated to noise produced during wind turbine operation.
We used turbine operational status (i.e., running versus
idle) to determine whether the slope of bird densities
differed. Analysis of variance (ANOVA; SAS 1989)
was used to determine whether bird density in CRP
grasslands without turbines differed from that in CRP
grasslands containing turbines. An ANOVA also was
used to determine whether bird density was related to
distance from wind turbines. A Least Significant Dif-
ference Multiple Comparisons test was used to deter-
mine where differences in bird density occurred among
transects.

RESULTS

Ten upland grassland bird species occurred
in CRP grasslands with and without turbines
(Table 1). Bobolinks (Dolichonyx oryzivorus).
Red-winged Blackbirds (Agelaius phoeni-
ceus), and Savannah Sparrows (Passerculus
sandwichensis) comprised 74.5% of the birds
in CRP grasslands with turbines (Table 1).
Bobolinks, Sedge Wrens (Cistothorus platen-
sis), and Savannah Sparrows comprised
80.0% of the individuals in CRP fields with-
out turbines (Table 1).

Mean bird densities from surveys conduct-
ed while wind turbines were operational {x =
4.7 ± 0.88 SE) and non-operational {x = 5.4
± 0.94 SE) were pooled because slopes of
bird densities among transects did not differ
(F = 0.39, 1,30 df, P > 0.05). Total bird den-
sity was lower in CRP grasslands containing
turbines than in CRP grasslands without tur-
bines (F = 17.36, 6,14 df, P = 0.001; Table
2). Bird density was lower (F = 12.37, 1,10
df, P = 0.006) in the 0 and 40 m transects
compared to density in transects 80 m or more
from turbines (Table 2). Bird density also was
lower (F= 13.10, l,10df, F = 0.001) in tran-
sects within 80 m of the turbines compared to
180 m from turbines (Table 2). Bird density
180 m from turbines did not differ (F = 0.10.
1,10 df, P > 0.05) from that in CRP grass-
lands without turbines (Table 2). A linear re-
lationship existed (r- = 0.746, n = 18, F <
0.001) between bird density and transect dis-
tance from turbines (Fig. 1).

DISCUSSION

Conservation Reserve Program grasslands
without turbines and areas located 180 m from
turbines supported mean densities of grassland
birds that were four times higher than those

102

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

TABLE 1. Number (/?) and percent (%) composition of breeding

grassland birds

in Conservation Reserve

Program grasslands with and without turbines at the Buffalo Ridge Wind Resource Area, Minnesota, May— July

1995.

Turbines

No turbines

Species

Bobolink {Dolichonyx oryzivorus)

139

36.6

32.0

Red-winged Blackbird (Agelaius phoeniceus)

22.4

4.7

Savannah Sparrow (Passerculus sandwichensis)

15.5

22.0

Common Yellowthroat (Geothlypis trichas)

9.5

4.7

Dickcissel (Spiza americana)

5.8

1.3

Le Conte’s Sparrow (Ammodramus leconteii)

2.6

Brown-headed Cowbird (Molothriis ater)

2.4

4.0

Western Meadowlark (Strunella neglecta)

1.3

0.7

Grasshopper Sparrow (Ammodramus savannarum)

1.1

2.7

Sedge Wren (Cistothorus platensis)

0.3

26.0

Clay-colored Sparrow (Spizella pallida)

0.7

Unknown

2.3

1.3

Total species

in grasslands nearer to turbines. Three of four
species that composed at least 74.5% of the
bird community composition (Bobolink, Sa-
vannah Sparrow, Sedge Wrens) in CRP fields
with and without turbines are area-sensitive
species (Herkert 1994a, b; Swanson 1996) that
require large tracts of tall, dense vegetation for
nesting (Wiens 1969, Herkert 1994a). Minor
differences in overall bird species richness
and composition were likely related to subtle
structural differences in grassland stand types.
Leddy and coworkers (in press) found that
Clay-colored Sparrows (Spizella pallida) and
Sedge Wrens using CRP grasslands on the
Buffalo Ridge WRA preferred dense stands of
switchgrass while Dickcissels (Spiza ameri-
cana) and Bobolinks usually used stands of
smooth brome and alfalfa.

TABLE 2. Species richness and mean density of
upland grassland birds/100 ha at varying distances
from wind turbines in Conservation Reserve Program
grasslands at the Buffalo Ridge Wind Resource Area,
Minnesota, May-July 1995.

Breeding males

Transect

Species

richness

Mean

density^

0 m

58.2 A

26.3

40 m

66.0 A

17.1

80 m

128.0 B

19.6

180 m

261.0 C

12.0

CRP Control

312.5 C

15.7

Little evidence has been found linking avi-
an mortality to collisions with wind turbines
on the Buffalo Ridge WRA (Higgins et al.
1996). Although wind turbines may not di-
rectly cause mortality, the presence of wind
turbines may indirectly affect local grassland
bird populations by decreasing the area of
grassland habitat available to breeding birds.
Comparison of bird density and species rich-
ness among transects indicated that bird use
of grasslands 1 80 m from turbines was similar
to that in CRP fields without turbines (Table
2). Although research in the Netherlands also

FIG. 1. Linear relationship (Density = 32.30 +
1.22 X Distance; H = 0.746) between breeding bird
density (males/100 ha) and distance (0-180 m) from
wind turbines in Conservation Reserve Program grass-
lands at the Buffalo Ridge Wind Resource Area in
southwestern Minnesota, May-July 1995.

•' Means deni)ted by the same letter do not differ (F £ O.O.S).

Leckh et al • EFFECTS OF WIND TURBINES ON BIRDS

103

has indicated that the presence of turbines has
prevented waterfowl and wading bird species
from using otherwise suitable habitat (Win-
kelman 1990, Pedersen and Poulsen 1991),
mechanisms inhibiting birds from exploiting
grasslands near turbines have not yet been
identified. In addition to human disturbance
and noise, the physical movements of the tur-
bines when they are operating may have dis-
turbed nesting birds. Maintenance trails be-
tween turbines that are driven daily may have
further decreased the availiability of grassland
habitat adjacent to turbines.

Construction of windplants within mid-
western grassland habitats may soon become
an additional source of habitat degradation as
demands for wind generated power increase.
Current grazing and tillage practices on many
privately owned lands that are less conducive
to grassland bird production increase the im-
portance of remaining grasslands to prairie
nesting birds (Johnson and Schwartz 1993;
Johnson and Igl 1995; Leddy et al., in press).
Until additional research is conducted, we rec-
ommend that wind turbines be placed within
cropland habitats that support lower densities
of grassland passerines than those found in
CRP grasslands (Leddy et al., in press). We
also recommend that additional research be
conducted in other geographic regions where
wind generated power is currently used to fur-
ther assess possible effects of wind turbines
on avian habitats.

ACKNOWLEDGMENTS

We thank L. D. Flake, D. H. Johnson, and anony-
mous referees for reviews of our manuscript. We also
thank R D. Evenson for assisting with statistical anal-
yses and V. J. Swier and R. G. Osborn for conducting
field work. Project funding was provided by Kenetech
Windpower, Inc., and the South Dakota Cooperative
Fish and Wildlife Research Unit, in cooperation with
South Dakota Department of Game, Fish and Parks,
Natural Resources Conservation Service, Wildlife
Management Institute and South Dakota State Univer-
sity.

Mention of trade names does not constitute any en-
dorsement, guarantee, or warranty of any trademark
proprietary product by the authors. South Dakota State
University, the Department of Wildlife and Fisheries
Sciences, or the South Dakota Cooperative Fish and
Wildlife Research Unit (United States Geological Sur-
vey, Biological Resources Division).

LITERATURE CITED

Arnold, T. W. and K. F. Higgins. 1986. Effects of
shrub coverages on birds of North Dakota mixed-
grass prairie. Can. Field-Nat. 100:10-14.

Burnham, K. R, D. R. Anderson, and J. L. Laake.
1980. Estimation of density from line transect
sampling of biological populations. Wildl. Mon-
ogr. 72:1-202.

Castrale, j. S. 1985. Responses of wildlife to various
tillage conditions. Trans. N. Am. Wildl. Nat. Re-
sour. Conf. 50:142-156.

Herkert, j. R. 1994a. Breeding bird communities of
midwestern prairie fragments: the effects of pre-
scribed burning and habitat-area. Nat. Areas J. 14:
128-135.

Herkert, J. R. 1994b. The effects of habitat fragmen-
tation on midwestern grassland bird communities.
Ecol. Appl. 4:461-471.

Higgins, K. F and W. T. Barker. 1982. Changes in
vegetation structure in seeded nesting cover in the
Prairie Pothole Region. U.S. Fish Wildl. Serv.
Spec. Sci. Rep. 242:1-26.

Higgins, K. F, R. E. Usgaard, and C. D. Dieter.
1996. Monitoring seasonal bird activity and mor-
tality at the Buffalo Ridge Windplant, MN. KE-
NETECH Windpower, Inc., Cooperative Fish and
Wildlife Research Unit, South Dakota State Univ.,
Brookings, South Dakota.

Hurlbert, S. H. 1984. Pseudoreplication and the de-
sign of ecological field experiments. Ecol. Mon-
ogr. 54:187-211.

Igl, L. D. and D. H. Johnson. 1995. Dramatic increase
of Le Conte’s Sparrow in Conservation Reserve
Program fields in the northern Great Plains. Prairie
Nat. 27:89-94.

Johnson, D. H. and L. D. Igl. 1995. Contributions of
the Conservation Reserve Program to populations
of breeding birds in North Dakota. Wilson Bull.
107:709-718.

Johnson, D. H. and M. D. Schwartz. 1993. The Con-
servation Reserve Program: habitat for grassland
birds. Great Plains Res. 3:273-295.

Kantrud, H. a. 1981. Grazing intensity effects on the
breeding avifauna of North Dakota native grass-
lands. Can. Field-Nat. 95:404-417.

King, J. W. and J. A. Savidge. 1995. Effects of the
Conservation Reserve Program on wildlife in
southeast Nebraska. Wildl. Soc. Bull. 23:377-385.

Koford, R. R., j. B. Dunning, Jr., C. A. Ribic, and
D. M. Finch. 1994. A glossary for avian conser-
vation biology. Wilson Bull. 106:121 — 137.

Leddy, K. L. 1996. Effects of wind turbines on non-
game birds in Conservation Reserve Program
grasslands in southwestern Minnesota. M.S. the-
sis, South Dakota State Univ.. Brookings.

Leddy, K. L., K. F. Higgins, and D. E. Naugle. In
press. The importance of Conservation Reserve
Program fields to breeding grassland birds at Buf-
falo Ridge, Minnesota. S.D. Acad. Sci.

Mikol, S. a. 1980. Field guidelines for using transects

104

THE WILSON BULLETIN • Vol. Ill, No. /, March 1999

to sample nongame bird populations. U.S. Fish
and Wildl. Serv. Bio. Serv. Prog. OBS-80— 58:1 —
26.

Millenbah, K. E, S. R. WiNTERSTEiN, H. Campa, III,
L. T. Furrow, and R. B. Minnis. 1996. Effects of
Conservation Reserve Program field age on avian
relative abundance, diversity, and productivity.
Wilson Bull. 108:760-770.

Nelson, H. K. and R. C. Curry. 1995. Assessing avi-
an interactions with windplant development and
operation. Trans. N. Am. Wildl. Nat. Resour.
Conf. 60:266-277.

Orloff, S. G. and a. W. Flannery. 1992. Wind tur-
bine effects on avian activity, habitat use, and
mortality in (the) Altamont Pass and Solano
County Wind Resource Areas, 1989-1991. Ala-
meda, Contra Costa and Solano Counties, Cali-
fornia Energy Commission, Biosystems Analysis,
Inc., Tiburon, California.

Pedersen, M. B. and E. Poulsen. 1991. Avian re-
sponses to the implementation of the Tjaereborg
Wind Turbine at the Danish Wadden Sea. Dan.
Wildtundersogelser 47:1-44.

Ralph, C. J., G. R. Geupel, P. Pyle, T. E. Martin,
AND D. E DeSante. 1993. Handbook of field
methods for monitoring landbirds. Gen. Tech.
Rep. PSW-GTR- 144: 1-41.

Ree.se, K. P. and J. T. Ratti. 1988. Edge effect: a con-
cept under scrutiny. Trans. N. Am. Wildl. Nat. Re-
sour. Conf. 53:127-136.

Robel, R. j., j. N. Briggs, A. D. Dayton, and L. C.
Hulbert. 1970. Relationships between visual ob-

struction measurements and weight of grassland
vegetation. J. Range Manage. 23:295-297.

SAS Institute, Inc. 1989. SAS/STAT user’s guide,
version 6, fourth ed. SAS Institute, Inc. Cary,
North Carolina.

Skirvin, a. a. 1981. Effect of time of day and time
of season on number of observations and density
estimates of breeding birds. Stud. Avian Biol. 6:
271-274.

Swanson, D. A. 1996. Nesting ecology and nesting
habitat requirements of Ohio’s grassland-nesting
birds: a literature review. Ohio Fish Wildl. Rep.
13:1-60.

UsGAARD, R. E., D. E. Naugle, K. E Higgins, and R.
G. Osborn. In press. Effects of wind turbines on
nesting raptors at Buffalo Ridge in southwestern
Minnesota. S.D. Acad. Sci.

Verner, j. and L. V. Ritter. 1986. Hourly variation
in morning point counts of birds. Auk 103:117—
124.

Wakeley, j. S. 1987. Avian line-transect methods.
Section 6.3.2, U.S. Army Corps Eng. Wildl. Re-
sour. Manage. Manual. Tech. Rep. EL-87-5: 1—21.

Wiens, J. A. 1969. An approach to the study of eco-
logical relationships among grassland birds. Or-
nithol. Monogr. 8:1-93.

Winkelman, E. 1990. Impact of the wind park near
Urk, Netherlands, on birds: bird collision victims
and disturbance of wintering fowl. Int. Ornithol.
Cong. 20:402-403.

Young, E. C. and C. T. Osborn. 1990. Costs and ben-
efits of the Conservation Reserve Program. J. Soil
Water Conserv. 45:370—373.

Wilson Bull., 111(1), 1999, pp. 105-1 14

AVIAN USE OF PURPLE LOOSESTRIFE DOMINATED HABITAT
RELATIVE TO OTHER VEGETATION TYPES IN A LAKE HURON

WETLAND COMPLEX

MICHAEL B. WHITT,' HAROLD H. PRINCE,' AND ROBERT R. COX, JR.-

ABSTRACT. — Purple loosestrife (Lythriini salicaria), native to Eurasia, is an introduced perennial plant in
North American wetlands that displaces other wetland plants. Although not well studied, purple loosestrife is
widely believed to have little value as habitat for birds. To examine the value of purple loosestrife as avian
breeding habitat, we conducted early, mid-, and late season bird surveys during two years (1994 and 1995) at
258 18-m (0. 1 ha) fixed-radius plots in coastal wetlands of Saginaw Bay, Lake Huron. We found that loosestrife-
dominated habitats had higher avian densities, but lower avian diversities than other vegetation types. The six
most commonly observed bird species in all habitats combined were Sedge Wren (Cislolhorus plaiensis). Marsh
Wren (C. palustris). Yellow Warbler (Deiulroica petechia). Common Yellowthroat (Geothylpis trichas). Swamp
Sparrow (Melospiza georf>iana), and Red-winged Blackbird (Agelaius phoeniceus). Swamp Spanow densities
were highest and Marsh Wren densities were lowest in loosestrife dominated habitats. We observed ten breeding
species in loosestrife dominated habitats. We conclude that avian use of loo.sestrife warrants further quantitative
investigation because avian use may be higher than is commonly believed. Received 27 May 1998, accepted 26
Aug. 1998.

Purple loosestrife (Lythrum salicaria) is an
exotic, broad-leaved, herbaceous perennial
that is common in North American freshwater
wetland habitats north of 35° N latitude
(Thompson 1989). Loosestrife is native to
Eurasia where it occurs in freshwater marshes,
open stream margins, and alluvial floodplains;
it invades similar habitats in North America
(Thompson 1989). Common plant associates
of loosestrife in North American wetland hab-
itats such as cattails (Typha spp.), reed canary
grass (Phalaris anmdinacea), sedges (Carex
spp.), and rushes (Jimcus spp.) closely resem-
ble its associates in Eurasian wetlands
(Thompson et al. 1987). Loosestrife out com-
petes and partially or completely replaces na-
tive emergent vegetation (Thompson 1989).
Loosestrife often pioneers in disturbed areas
such as drainage ditches (Wilcox 1995) and
displaces moist-soil species such as smart-
weeds {Polygonum spp.) and millets (Panicum
spp.) on mudflats (Thompson et al. 1987).
Species of wetland plants become distributed
along a wetland gradient and are good indi-

' Michigan State Univ., Dept, of Fisherie.s and Wild-
life, East Lansing, Ml 48824.

-Northern Prairie Wildlife Research Center. 8711
37th Street SE, Jamestown, ND 58401.

^ Present address: Svoboda Ecological Resources,
2477 Shadywood Road. Excelsior, MN 55331;

E-mail: michaeLwhitt@hotmail.com
^ CoiTesponding author.

cators of long-term hydrology and other abi-
otic factors (Keddy and Reznicek 1985). Wet-
land vegetation types generally grade from
forested wetland to shrub-scrub, to wet mead-
ow, to strand (or mudflat), to emergent marsh,
and finally, to open water (Cowardin et al.
1979, Keddy and Reznicek 1985). Loosestrife
occupies zones near the strand including
emergent and wet meadow zones.

Avian use of loosestrife is not well studied
(Thompson et al. 1987). Prince and Flegel
(1995) found no records in the literature of
loosestrife as avian food or nesting habitat in
Lake Huron wetlands. In New York wetlands.
Rawinski and Malecki (1984) observed that
Marsh Wrens (Cistothorus palustris) preferred
cattail for nesting, whereas Red-winged
Blackbirds (Agelaius phoeniceus) preferred
loosestrife for nesting. Rawinski and Malecki
(1984) also noted that Black-crowned Night-
herons (Nycticorax nycticorax) roosted in
loosestrife, and Pied-billed Grebes {Podilyin-
hus podiceps) nested in one- and two-year-old
emergent loosestrife stands. Kiviat (1996)
found 15 American Goldfinch (Carduelis tris-
tis) nests in loosestrife during a 23-year study
of birds in the Hudson Valley. Swift and co-
workers (1988) observed Least Bitterns {l.xo-
hrychus exilis) and other birds in Hudson Riv-
er wetlands that consisted of cattail, river bul-
rush (Scirpus fluviatilis), loosestrife, and com-
mon reed (Phragmites australis).

105

106

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

PaJustrine®P

buttonbush

{Cephalanthus ocadentalis)
silky dogwood
{Cortms amonum)
sedges
(Carer spp.)

sedges blue-joint reedgrass

{Carex spp.) (Ca/amagrostis canadensis)
blue-joint reedgrass red-osier dogwood

(Calamagroslis canadensis) (Cormis slolonifera)
red-osier dogwood purple loosestrife

{Cormis stolonifera) {Lyihrum salicaria)

blue-joint reedgrass blue-joint reedgrass

{Calamagroslis canadensis) {Calamagroslis canadensis)
sedges purple loosestrife

(Carer spp.) {Lyihrum salicaria)

cattail

{Typhaspp.)

bulrush

{Scirpus amencana)

Seasonal « C

Saturated => B

Saturated = B

Semi-
permanent = F

Inicrmittenly
exposed « G

Intermittenly
exposed = G

PSSIC

PEMl/SSlB

PEMI/SSIB

PEMIB

Senib-shrub

Wet meadow/

Wet

scrub-shrub

scrub-shrub/

meadow

loosestrife

PEMIB

PEMIF

PEMIG

Wet

Inland

Coastal

Meadow/

cattail

bulrush

loosestrife

FIG. 1. Characteristics of surveyed vegetation types in Saginaw Bay wetlands, 1994-1995, based on Na-
tional Wetlands Inventory (Cowardin et al. 1979) classification system. PSSIC = Palustrine, broad-leafed de-
ciduous scrub-shrub, and seasonally flooded; PEMI/SSIB = Palustrine, persistent emergent/broad-leafed decid-
uous scrub-shrub, and saturated; PEMIB = Palustrine, persistent emergent, and saturated; PEMIF = Palustrine,
persistent emergent, and semi-permanently flooded; PEMIG = Palustrine, persistent emergent, and intermittently
exposed (Cowardin et al. 1979).

Minnesota established the hrst statewide
loosestrife control program in 1987 with the
goal of broadening public awareness, con-
ducting inventories, developing control meth-
ods, and initiating control work (Skinner et al.
1994). Minnesota has spent $US 1.75 million
since the beginning of the program (Skinner,
pers. comm.). Other state and federal agencies
also have spent considerable money and effort
to control loosestrife, in part, because wildlife
values of this plant are widely regarded to be
limited. Methods of control have included use
of chemicals, water manipulation, mowing,
tillage, planting robust mudflat species such as
Japanese millet (Thompson 1989), and, most
recently, biological control using insects (Ma-
lecki et al. 1993).

Our objective was to compare avian use of
vegetation zones dominated by loosestrife
with other wetland zones where loosestrife
was absent or not dominant. Comparison of
avian breeding species richness, density, and
diversity is a necessary hrst step to assess the
value of loosestrife-dominated habitats to
birds, and ultimately to evaluate costs and
benehts of loosestrife control.

METHODS

We conducted field work during 1994 and 1995 in
Bay, Tuscola, and Huron counties adjacent to Saginaw
Bay, Lake Huron, Michigan. Saginaw Bay comprises
the majority of remaining wetland habitat on Lake Hu-
ron because unsuitable shore morphology (e.g., cliffs)
prohibited wetland formation, and development pres-
sures (mostly agricultural) eliminated presettlement
wetland habitats (Prince and Flegel 1995). Although
this area has experienced a 50% overall wetland loss
(Dahl 1990), 70% of inland wetlands and 99% of lake-
plain prairies have been drained and converted to other
uses (Comer 1996). Most existing Saginaw Bay wet-
lands are disturbed by adjacent urban and agricultural
development, diking, and exotic flora and fauna.

We surveyed birds on 18-m fixed-radius plots in
eight vegetation types based on hydrology and plant
form and structure: scruh-shrub, wet meadow/scrub-
shrub, wet rneadow/scrub-shrub/loosestrife, wet mead-
ow, wet meadow/loo.sestrife, inland cattail, coastal cat-
tail, and coastal bulrush (Sc/r/JZ/.v spp.). Our habitat
classifications were based on Cowardin and coworkers
(1979); dominant plants had greater than 30% cover
(Fig. 1). We used a split class (e.g., broad-leafed de-
ciduous scrub-shrub/persistent emergent; National
Wetlands Inventory) to classify two vegetation types
because scattered shrubs of at least 30% cover were
present. We separated cattail sites into coastal and in-
land because hydrologies differed; coastal sites were

Whitt et id. • AVIAN USE OF PURPLE LOOSESTRIFE

107

TABLE 1. Mean cover (Robel) height (cm) ±
type in Saginaw Bay wetlands, 1994-1995.

SE and

mean water depth (cm)

± SE by period and vegetation

Period 1

Period

Period 3

Year and site^

Cover height

Water depth

Cover height

Water depth

Cover height

Water depth

1994 SS

27.8 ±

4.3

27.5 ± 0.9

55.0

•+-

5.9

12.1 ± 3.2

75.7

-h

8.1

25.4 ± 2.9

1995 SS

35.8 ±

3.4

24.4 ± 0.8

68.8

-h

4.3

14.4 ± 0.7

74.6

-h

5.5

1.9 ± 0.6

1994 WM/SS

—

62.8

-h

4.2

saturated

56.7

-h

3.5

saturated

1995 WM/SS

24.4 ±

4.1

0.2 ± 0.2

67.8

7.8

saturated

100.8

-H

8.7

saturated

1994 WM/SS/LS

—

107.5

1 1.0

saturated

101.1

-h

3.9

saturated

1995 WM/SS/LS

30.0 ±

8.5

saturated

48.9

4.2

saturated

87.8

H-

9.5

0.5 ± 1.7

1994 WM/LS

—

84.4

-+-

8.8

saturated

106.7

-h

8.2

2.2 ± 0.4

1995 WM/LS

38.9 ±

4.8

saturated

63.3

3.3

saturated

87.8

-h

6.4

saturated

1994 WM

—

81.7

5.1

0.3 ± 0.2

83.9

-h

2.6

1.8 ± 0.8

1995 WM

30.6 ±

2.6

saturated

61.1

-E

4.3

saturated

68.3

-t-

6.7

saturated

1994 IC

80.0 ±

3.2

29.9 ± 2.0

70.7

-h

4.1

21.0 ± 1.7

1 10.6

-h

7.2

26.4 ± 1.4

1995 IC

41.4 ±

6.6

17.1 ± 3.4

52.4

9.0

13.3 ± 2.5

94.8

H-

6.4

7.7 ± 1.9

1994 CC

—

71.2

-H

7.5

22.0 ± 2.6

144.3

-1-

8.2

31.6 ± 2.4

1995 CC

60.3 ±

4.7

9.1 ± 1.4

82.2

-h

6.3

12.5 ± 1.5

127.9

8.3

22.2 ± 1.7

1994 CB

—

11.3

-h

3.4

33.5 ± 1.4

34.7

5.5

36.8 ± 1.2

1995 CB

—

8.9

-+■

2.7

23.7 ± 0.8

33.2

3.6

30.1 ± 1.1

“ Vegetation types: SS = scrub-shrub, WM/SS = wet meadow/scrub-shrub. WM/SS/LS = wet meadow/scrub-shrub/loosestrife. WM/LS = wet meadow/
loosestrife. WM = wet meadow. IC = inland cattail. CC = coastal cattail. CB = coastal bulrush.

Dashes ( — ) indicate insufficient or lack of data.

intermittently exposed, whereas inland sites were semi-
permanently flooded by groundwater and precipitation.

Sampling periods were divided into an early season
during the second and third weeks of May, a mid-
season during the first and second weeks of June, and
a late season during the last week of June and first
week of July. We conducted surveys between sunrise
and 10:00 EST. Surveys were not conducted if sus-
tained winds exceeded 24 km/h or during heavy rain.

We selected plots using the following protocol: first,
an azimuth was determined that traversed the habitat.
The center of the first plot was placed at least 18 m
from the outer boundary of the vegetation on that az-
imuth. The center of the next plot was 70 m from the
first plot on the same azimuth. This procedure was
continued until observers surveyed three or more plots
or reached a different vegetation type. If fewer than 3
plots were established on the first azimuth, we estab-
lished a second azimuth, approximately perpendicular
to the first azimuth, that traversed the vegetation type
and permitted plot placement at least 70 m from other
plots. Plots were set on this azimuth in the same man-
ner as on the first azimuth. Plots were placed in dif-
ferent locations at the same site among time periods
to avoid resampling the same plots and recounting the
same nests. Coastal bulrush plots were not surveyed
during the first periods of each year because they
lacked structure; new vegetative growth was not yet
established and the previous year's growth was elimi-
nated by ice action. Neither did we survey three veg-
etation types (wet meadow/scrub-shrub/loosestrife, wet
meadow, wet meadow/loosestrife) during the first pe-
riod of 1994. We surveyed 258 plots in 8 wetland hab-
itats.

Observers waited 5 min for normal bird activity to

resume after arriving at a survey plot. We recorded all
birds seen or heard on plots during a 7-min observa-
tion period. We recorded flying birds if their flight
originated or terminated within the plot and we tallied
individual birds only once. We played tape-recorded
calls (Peterson 1990) of five secretive species [Amer-
ican Bittern (Botauru.s lentiginosus). Least Bittern,
King Rail (Rcdlus elegcm.s)^ Virginia Rail {R. limicola),
and Sora (Porzana Carolina)] during the last 3 min
using portable cassette recorders (Johnson et al. 1981,
Marion et al. 1981, Johnson and Dinsmore 1986). We
played calls for 25-30 sec followed by 10 sec of si-
lence. We measured water depth and vertical cover 4
m from the plot center at 0°. 120°, and 240° (Table 1).
Observers measured vertical cover to the nearest 10
cm using a 2-m Robel pole placed at plot center and
viewed while maintaining eye level 1 m above the wa-
ter surface or ground level and looking back toward
plot center (Higgins et al. 1994). Workers returned to
plots later that day and searched the innermost 13-m
radius (0.05 ha) portion for nests. A bird species was
designated as breeding when nests or flightless young
were observed in one or more periods or when adults
were observed in two of three periods (Brown and
Dinsmore 1986). A nest verified breeding status when
eggs, young, or strong evidence of use such as egg
shell fragments, down, or fecal sacs were pre.sent. We
considered predated nests as breeding evidence when
prey species could be determined. We also tallied spe-
cies as breeding if they were ob.served within the sam-
pled vegetation type but outside of plot boundaries on
two of three visits.

We tallied breeding species richness (i.e., number of
breeding species) for each vegetation type. We calcu-
lated avian diversities for each plot using the Shannon-

108

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

TABLE 2. Distribution of breedin

ig birds by vegetation type in Saginaw Bay wetlands.

1994-

1995.

Breeding

status based on observation of adults

on at least two of three visits, or a nest or fli

ghtless

young

on at least one

\isit.

Species

SS“

WM/SS“

WM/SS/LS“ WM/LS“

WM"*

tea

CC“

CB“

Pied-billed Grebe^

American Bittern

Least Bittern

Canada Goose

Wood Duck*’

American Black Duck

Mallard

X X

Blue-winged Teal

RedheaeP

Northern Harrier'’

Ring-necked Pheasant

Virginia Rail

X X

Sora

Common Moorhen/American CooL

Forster's Tern'’

Black Tern'’

Northern Flicker

Eastern Wood-pewee

Willow Flycatcher

Great Crested Flycatcher

Eastern Kingbird

Tree Swallow

Sedge Wren

Marsh Wren

Gray Catbird

Yellow Warbler

Common Yellowthroat

Rose-breasted Grosbeak

Savannah Sparrow

Song Sparrow

Swamp Sparrow

X X

Bobolink

Red-winged Blackbird

X X

Yellow-headed Blackbird

Brewer's Blackbird

Brown-headed Cowbird

Baltimore Oriole

American Goldfinch

Total

9 5

^ Vegetation types: SS = scmb-shnib. WM/SS =

wet meadow/.scrub-.shrub, WM/SS/LS = wet meadow/scrub-shrub/loose.strife, WM/LS =

wet meadow/

loosestrife. WM = wet meadow. 1C = coastal cattail. CB = coa.stal bulrush.

^.Species observed within the sampled vegetation type but not on plots.

American Coot and Common Moorhen were grouped together because these species were most often observed by call only and their calls are difficult
to distinguish.

Weiner diversity index. Density was the number of
birds (both sexes) observed on a plot multiplied by 10
to obtain density per hectare.

We used ANOVA (PROC GLM; SAS 1990; SAS
6. 1 2 for Windows) to as.sess fixed effects of vegetation,
period, year, tmd their interactions on avian density and
diversity. Residuals were normally distributed, but var-
iiinces were not homogeneous because we never ob-
served some species in one or more habitats (resulting

in means and variances of zero). However, the overall
F-statistic from ANOVA is robust to violations in as-
sumptions of homogeneous variances (Sokal and Rohlf
1981 ). Early-period observations were eliminated from
all analyses because of missing data. We considered
plots as the experimental units becau.se we decided a
priori to restrict our inference to Saginaw Bay wet-
lands. We used a = 0.05 for all statistical comparisons.
We initially analyzed fully specified models (all main

Whitt et cil. • AVIAN USE OF PURPLE LOOSESTRIFE

109

etfects and interactions included). We fitted eacli mod-
el using a backward, stepwise procedure by eliminat-
ing non-significant {P > 0.05) effects, beginning with
highest-order interactions. Thus, our final models in-
cluded only significant effects or interactions, and
main effects or interactions contained in significant
higher-order interactions. We used Fisher's protected
least significant difference test to isolate differences
antong least-square means (LSMEANS, SAS 1990) for
significant effects in the ANOVA (Milliken and John-
son 1984). We compared density and diversity of birds
in loosestrife-dominated vegetation types (wet mead-
ow/scrub-shrub/loosestrife and wet meadow/loose-
strite) to those in other vegetation types using orthog-
onal contrasts (PROC GLM; SAS 1990), and estimated
least-square means using estimate statements (PROC
GLM; SAS 1990). We developed similar models for
abundance of the six most commonly observed bird
species: Sedge Wren (Cistothoriis platensis). Marsh
Wren. Yellow Warbler (Dendroica petechia). Common
Yellowthroat (Geothlypis trichas). Swamp Sparrow
(Meiospiza georgiana), and Red-winged Blackbird.

Standard errors reported are for least-square means
(SAS 1990). Because multiple comparison of means
with heterogenous variances may be misleading (Sokal
and Rohlf 1981), we further examined comparisons of
non-zero means to means of zero using confidence in-
tervals. For each mean of zero, we constructed a 90%
upper confidence limit after assigning the highest stan-
dard deviation associated with any mean in the model.
We then compared 90% lower confidence intervals for
nonzero means to 90% upper confidence intervals for
zero means; we considered failure of these intervals to
overlap as statistically significant. Resulting confi-
dence intervals for zero means are likely overestimat-
ed, yielding a con.servative comparison. We note in
tables instances where confidence interval compari-
.sons did not corroborate multiple comparisons using
Fisher's least significant difference test.

RESULTS

As the season progressed water depths at
coastal sites (coastal cattail and coastal bul-
rush) increased and those at inland sites de-
creased while vertical cover generally in-
creased at all sites (Table 1). We surveyed 258
plots and observed 39 breeding bird species
in Saginaw Bay wetland habitats (Table 2).
Six breeding species were observed in the
sampled vegetation type, but not on survey
plots: Pied-billed Grebe, Wood Duck (Aix
sponsa). Redhead (Aythya americana). North-
ern Harrier {Circus cyaneus), Forster’s Tern
(Sterna forsteri), and Black Tern (Chlidonias
niger). We also observed 10 species breeding
in loosestrife dominated habitats (Table 2).

Marsh Wren (n = 20), Swamp Sparrow (/;
= 16), and Red-winged Blackbird {u = 21)

were the most commonly observed nests on
all plots (Table 3). We observed Mallard,
Blue-winged Teal (Anas discors), Virginia
Rail, and Red-winged Blackbird nests while
traversing between plots in loosestrife-domi-
nated vegetation zones, but not on the plots.

Avian density and diversity. — Our final
model indicated that avian density differed
only in relation to vegetation (ANOVA: F =
14.45, df = 7, 181, P < 0.001; Table 4). Avi-
an density was higher (orthogonal contrast: F
= 8.87, df = 1, 181, P = 0.003) in loosestrife-
dominated vegetation types [46.9 ± 3.8 (SE)
birds/haj than in other vegetation types (34.7
± 1.6). Avian diversity also differed only in
relation to vegetation (ANOVA: F = 12.76,
df = 7, 181, P < 0.001; Table 4). Avian di-
versity was lower (orthogonal contrast: F =
4.74, df = 1, 181, P = 0.03) in loosestrife-
dominated vegetation types (0.42 ± 0.08) than
in other vegetation types (0.60 ± 0.03). Ef-
fects of year, period, and all interactions were
not significant (P > 0.05 for all tests) for both
avian density and diversity. Scrub-shrub con-
tained the highest bird species diversity and
wet meadow/loosestrife and coastal bulrush
the lowest (Table 4).

Species abundance. — The vegetation X pe-
riod X year interaction was significant (AN-
OVA: F = 2.34, df = 7, 157, P = 0.03) in
our initial Sedge Wren model. Thus, vegeta-
tion related differences in Sedge Wren abun-
dance were not consistent among periods and
years (Table 5). Within periods and years,
Sedge Wren abundance did not differ (orthog-
onal contrasts: P > 0.05 for all tests) between
loosestrife dominated vegetation types and
other vegetation types.

Marsh Wren abundance differed among
vegetation types (ANOVA: F = 30.72, df =
7, 181, P < 0.001; Table 5). Marsh Wren
abundance was lower (orthogonal contrast: F
— 10.73, df = 1, 181, P = 0.001) in loose-
strife-dominated vegetation types (0 ± 1.8)
than in other vegetation types (6.2 ± 0.7).
Yellow Warbler abundance differed among
vegetation types, but differences were not
consistent between mid- and late periods (AN-
OVA: vegetation X period interaction, F =
2.08, df = 7, 173, P = 0.048; Table 5). The
interaction was due to significantly higher (P
< 0.001) numbers of Yellow Warblers ob-
served in late period scrub-shrub compared

110

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

TABLE 3. Number of nests and percent
commonly observed bird species were found

of plots within vegetation types where nests
in Saginaw Bay wetlands, 1994—1995.

of the three most

Vegetation type“

Marsh Wren

Swamp Sparrow

Red-winged Blackbird

Total plots*’

4 (9%)

13 (30%)

WM/SS

2 (6%)

WM/SS/LS

2 (12%)

WM/LS

4 (27%)

1 (4%)

3 (11%)

8 (23%)

3 (8%)

5 (14%)

12 (17%)

Total nests

—

“ Vegetation types: SS = scrub-shrub. WM/SS = wet meadow/scrub-shrub, WM/SS/LS = wet meadow/scrub-shrub/loosestrife, WM/LS = wet meadow/
loosestrife. WM = wet meadow. 1C = inland cattail. CC = coastal cattail. CB = coastal bulrush.

Includes early, mid-, and late season surveys.

with mid-period scrub-shrub (Table 5). Yellow
Warbler abundance did not differ (orthogonal
contrast: P > 0.05 for both tests) between
loosestrife-dominated and other vegetation
types in either period. Common Yellowthroat
abundance differed among vegetation types
(ANOVA: F = 6.04, df = 7, 181, P < 0.001;
Table 5). Common Yellowthroat abundance
did not differ (orthogonal contrast: F = 1.20,
df = 1, 181, P > 0.05) between loosestrife-
dominated and other vegetation types.

Swamp Sparrow abundance differed among
vegetation types (ANOVA: F = 39.03, df =
7, 180, P < 0.0001; Table 5) and between
periods (ANOVA: F = 6.88, df = 1, 180, P
= 0.009). Swamp Sparrow abundance was
higher during the late period (19.1 ± 1.1
birds/ha) compared with the mid-period (15.2
± 1.1 birds/ha). Swamp Sparrow abundance
was higher (orthogonal contrast: F = 133.06,
df = 1, 180, P < 0.001) in loosestrife-domi-
nated vegetation types (36.0 ± 2.0) than in
other vegetation types (10.8 ± 0.8). Swamp
Sparrows accounted for 95% and 65% of the
overall avian density at wet meadow/loose-
strife and wet meadow/scrub-shrub/loosestrife
plots, respectively. Abundance of Red-winged
Blackbird differed among vegetation types,
but differences were not consistent between
mid- and late periods (ANOVA: vegetation X
period interaction, F = 2.14, df = 7, 173, P
= 0.04; Table 5). The interaction was due to
significantly higher (P < 0.001) numbers of
Red-winged Blackbirds observed in mid-pe-
riod scrub-shrub compared with late period
scrub-shrub (Table 5). Red-winged Blackbird
abundance did not differ (orthogonal con-

trasts: P > 0.05 for both tests) between loose-
strife dominated and other vegetation types in
either period.

DISCUSSION

Weller and Spatcher (1965), Kantrud and
Stewart (1984), and Burger (1985) concluded
that plant form and structure, rather than tax-
onomic composition, play key roles in habitat
selection by marsh-nesting birds. The struc-
ture of loosestrife consists of stout, wood-like
persistent growth and herbaceous new growth,
similar to shrubs. Overall, species richness in
loosestrife was slightly lower than that in oth-
er vegetation types except coastal bulrush (Ta-
ble 2). Scrub-shrub habitat contained the high-
est breeding species richness and diversity,
but these values may be explained in part by
the location of scrub-shrub as an ecotone be-
tween forest and emergent wetland. Several
scrub-shrub breeding birds were not wetland-
dependent species but instead birds of forest
edge and gaps such as Northern Flicker (Co-
laptes aiiratus', Moore 1995), Eastern Wood-
pewee (Contopus virens; McCarty 1996),
Great Crested Flycatcher (Myiarchus crinitus;
Lanyon 1997), and Brown-headed Cowbird
(Molothrus ater\ Lowther 1993).

Swamp Sparrow nests were most abundant
in vegetation types where loosestrife was
dominant (Table 3). Reinert and Golet (1986)
determined that breeding Swamp Sparrows
principally required shallow standing water,
low (<1.5 m) dense cover, and elevated song-
posts, similar to our loosestrife-dominated
sites. Swamp Sparrows constructed nests us-
ing fine-stemmed sedges and grasses anchored

Whitt et al. • AVIAN USE OF PURPLE LOOSESTRIFE

TABLE 4. Mean avian density (no. /ha) ± SE. avian diversity (Shannon-Weaver) ± SE, and number of
seeond and third period plots by {n) vegetation type in Saginaw Bay wetlands, 1994-1995.

Vegetation type*'

Density'-'

Diversity'-'

51.33 ± 3.4 A

1.05 ± 0.08 A

WM/SS

38.95 ± 4.3 BC

0.63 ± 0.10 B

WM/SS/LS

44.62 ± 5.2 ABC

0.59 ± 0.12 B

WM/LS

49.17 ± 5.4 AB

0.22 ± 0.12 C

39.52 ±4.1 BC

0.62 ± 0.09 B

41.74 ± 3.9 ABC

0.74 ± 0.09 B

36.27 ± 2.6 C

0.56 ± 0.06 B

0.5 ± 4.2 D

0 C

“ Vegetation types: SS = .scrub-shrub, WM/SS = wet meadow/scrub-shrub, WM/SS/LS = wet meadow/,scrub-shrub/loo.sestrife. WM/LS = wet meadow/
loosestrife. WM = wet meadow, 1C = inland cattail, CC = coastal cattail. CB = coastal bulrush.

I’ Excludes early period surveys becau.se of missing data.

Means within columns followed by the same letter do not differ (F > 0.05) as determined by ANOVA and Fisher's lea.st significant difference.

in persistent loosestrife stalks. We also ob-
served Mallard, Blue-winged Teal, Virginia
Rail, and Red-winged Blackbird nests at our
loosestrife-dominated sites, and found Amer-
ican Bittern, Sedge Wren, Yellow Warbler,
Common Yellowthroat, and American Gold-
finch breeding based on our criteria. Pied-
billed Grebe (Rawinski and Malecki 1984),
Least Bittern (Swift et al. 1988), Red-winged
Blackbird (Rawinski and Malecki 1984), and
American Goldfinch (Kiviat 1996) were ob-
served nesting in loosestrife habitats previous
to this study.

Rawinski and Malecki (1984) observed that
Marsh Wrens preferred cattail habitats, but
Red-winged Blackbirds preferred loosestrife
habitats. We also found that nesting Marsh
Wrens used cattail habitats, but we observed
Red-winged Blackbird nests most frequently
in scrub-shrub zones (Table 5). Inconsisten-
cies in vegetation type, period, and year ef-
fects (i.e., significant three-way interaction) on
Sedge Wren abundance may reflect this spe-
cies’ variable breeding site selection (Table 5).
Bums (1982) observed that Sedge Wrens
show little site fidelity; this characteristic may
be due to the ephemeral nature of wet mead-
ow habitats (Kroodsma and Verner 1978). We
believe that Sedge Wren abundance may de-
cline as loosestrife increases in wet meadow
canopies. We observed greater areal cover of
loosestrife at the wet meadow/loosestrife site
compared with the wet meadow/scrub-shrub/
loosestrife site and Sedge Wren abundance
was significantly higher in two of four sam-
pling periods at the site with less loosestrife
(Table 5).

The avian diversity in loosestrife dominated
habitats was lower on average than that of
other wetland habitats that we surveyed, in-
dicating uneven distributions of fewer species.
We found higher avian densities in loosestrife-
dominated habitats compared to other vege-
tation types, although Swamp Sparrows com-
prised the majority of overall density in loose-
strife habitats. Swamp Sparrows accounted for
59% of the overall wet meadow density.
Swamp Sparrow densities reported in other
studies ranged up to 8.78 individuals/ha
(Mowbray 1997) and are considerably lower
than our densities in several vegetation types.
We observed a significant increase in Swamp
Sparrow density between mid- and late peri-
ods, which may be explained, in part, by the
addition of juveniles from early nests (Peck
and James 1987, Beaver 1991, Mowbray
1997). Swamp Sparrows prefer open wetlands
of sedges, grasses (i.e., wet meadow), and cat-
tail during the breeding season (Beaver 1991,
Mowbray 1997). Principally, loosestrife oc-
curs in the wet meadow, strand, and emergent
portions of a typical wetland profile, which
are the areas where Swamp Sparrows reach
their highest abundance (Beaver 1991, Mow-
bray 1997).

Nesting female and young Swamp Spar-
rows satisfy their high protein requirements
by consuming invertebrates. Wetherbee
(1968) determined that 88% of Swamp Spar-
row diets during spring and early summer
consisted of insects. Arroll (1995) found that
aquatic invertebrate abundance in loosestrife
in central Washington was similar to that in
cattail and bulrush. Arroll (1995) found only

112

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

TABLE 5
in Saginaw

Mean density (no./ha) ±
Bay wetlands, 1994-1995.

SE of the six most commonly observed bird species by vegetation type

Vegetation type^

Sedge Wren

Marsh Wren*’

1994

1995

Period 2’’

Period S*"

Period 2*’

Period 3^

0 c

0 B

0 c

WM/SS

6.7 ± 1.6 B

3.0 ± 1.6 BC

0 B

10.0 ± 1.9 A

0 c

WM/SS/LS

10.0 ± 1.9 AB

0 C

6.7 ± 2.2 A^

3.3 ± 2.2 B

0 c

WM/LS

0 C

6.7 ± 2.2 B-^

0 B

0 c

13.3 ± 2.2 A

13.3 ± 1.6 A

3.3 ± 1.6 AB

3.3 ± 1.6 B

0.5 ± 1.9 C

0 C

0 B

15 .6 ± 1.8 B

0 C

0 B

20.6 ± 1.2 A

0 C

0 B

0.5 ± 2.0 C

“ Vegetation types: SS = scrub-shrub. WM/SS = wet meadow/scrub-shrub. WM/SS/LS = wet meadow/scrub-shrub/loosestrife, WM/LS = wet meadow/
loosestrife. WM = wet meadow. 1C = inland cattail. CC = coastal cattail, CB = coa.stal bulrush.

^ Means within columns followed by the same letter do not differ (P > 0.05) as determined by ANOVA and Fisher's least significant difference.

Fisher's least significant difference multiple comparisons were not corroborated by 90% confidence interval comparisons with WM/SS/LS and IC (see
METHODS).

Fisher's least significant difference multiple compari.son were not corroborated by 90% confidence interval comparisons with WM/SS and WM/LS (see
METHODS).

' Fisher's lea.st significant difference multiple comparison were not corroborated by 90% confidence interval comparisons with WM/LS (see METHODS).

Fisher's lea.st significant difference multiple comparisons were not corroborated by 90% confidence interval compari.son with WM/SS. WM/LS, and
CB (.see METHODS).

8 Fisher's lea.st significant difference multiple comparisons were not corroborated by 90% confidence interval compari.son with WM/SS. WM/LS. and
CB (see METHODS).

nine statistically significant results in 1 1 1 in-
dividual comparisons of aquatic invertebrates
associated with macrophyte stems (using stem
vacuum), sediment (using sediment core), and
the water column (using activity traps). Of the
four statistically different comparisons involv-
ing loosestrife, two showed higher Diptera
and Ostracoda abundance in cattail compared
with loosestrife, and two showed higher Co-
pepod abundance in loosestrife compared with
cattail (Arroll 1995). Thus, invertebrate food
items during the breeding season do not ap-
pear limiting in loosestrife habitat, although
quantitative data from the Northeast are need-
ed.

Loosestrife is an anathema to wetland man-
agers because it often replaces seed-producing
mudflat species managed to attract waterfowl.
Water level manipulations such as early sea-
son drawdowns encourage loosestrife estab-
lishment (Thompson 1989). Loosestrife forms
dense stands that are difficult for some bird
species to negotiate and this may be especially
true for larger birds such as waterfowl or spe-
cies that walk on the ground such as bitterns
and rails. Our study demonstrates that loose-
strife may provide suitable habitat for some
passerines.

Many researchers have observed that habi-
tat diversity leads to faunal diversity in wet-

lands (Weller and Spatcher 1965, Weller and
Fredrickson 1974, Weller 1978, Kantrud and
Stewart 1984, Burger 1985). The highest avi-
an density, diversity, and productivity in
marshes occurs where emergent vegetation is
interspersed 1:1 with open water (Weller and
Spatcher 1965, Weller and Fredrickson 1974,
Fredrickson and Reid 1988). Wetland man-
agers manipulate vegetative interspersion in
marshes using artificial drawdowns, muskrat
management, and other means (Fredrickson
and Reid 1988). Kaminski and Prince (1981)
observed increased waterfowl density and di-
versity coincident with increased abundance,
biomass, and diversity of macroinvertebrates
in manipulated emergent wetland habitat. Our
loosestrife sites contained few openings. We
suspect that manipulated loosestrife habitat (to
create interspersion) could result in higher
bird diversity.

Loosestrife was widespread in Saginaw Bay
coastal wetlands and dominated canopies at
several sites. Although diversity was low,
loosestrife provided nesting and brood rearing
habitat to birds in Saginaw Bay wetlands
where alternative habitat choices were avail-
able. Some species, such as Marsh Wren, may
be disadvantaged as loosestrife displaces other
plant forms (e.g., cattail and bulrush). Swamp
Sparrows may prefer loosestrife habitat where

Whin et III. • AVIAN USE OF PURPLE LOOSESTRIFE

113

TABLE 5. Extended.

Yellow Warbler

Common

Yellowihroat*’

Swamp

Sparrow^’

Red-winged Blackbird

Period 2*'

Period .2*’

Period 2*’

Period 3*’

9.3

± 1.0 A

14.7 ± 1.0 A

2.7 ± 0.6 B'

1 1.3 ± 1.8 C

17.3 ± 1.8 A

6.0

± 1.8 A?

3.3

± 1.3 B

2.0 ± 1.2 B

5.3 ± 0.8 A

21.5 ± 2.3 B

0 C

0 B

2.8

± 1.4 B

1.7 ± 1.6 B

1.5 ± 1.0 BC

27.8 ± 2.8 B

7.1 ± 2.7 Bf

1.7

± 2.9 AB

0 B

0 C

44.2 ± 2.9 A

0 C

0 B

1.4 ± 0.8 BC

21.2 ± 2.2 B

6.7 ± 2.4 B'

4.2

± 2.0 AB

0.7

± 1.0 B

0 B

0 C

9.6 ± 2.1 C

5.7 ± 1.9 BC

4.4

± 2.4 AB

0 B

0.2 ± 0.5 C

1.5 ± 1.4 D

1.8 ± 1.4 BC

1.2

± 1.4 B

0 B

0 C

0 D

0 C

0 B

nest-building materials (fine-stemmed grasses
and sedges) are available. We conclude that
avian use of loosestrife warrants further quan-
titative investigation because avian use may
be higher than is commonly believed.

ACKNOWLEDGMENTS

The Michigan Agricultural Experiment Station and
the Michigan Department of Natural Resources pro-
vided funding. T. M. Burton and J. Burley, both of
Michigan State University, provided advice and guid-
ance. L. D. Igl and D. H. Johnson, both of Northern
Prairie Wildlife Research Center (NPWRC) in James-
town, North Dakota, W. Scharf, E. Kiviat, and two
anonymous reviewers provided comments on drafts. D,
H. Johnson, W. E. Newton, and G. A. Sargeant (all of
NPWRC) provided statistical advice. L. A. Jagger and
D. Ford assisted in field observations.

LITERATURE CITED

Arroll, S. G. 1995. Effects of purple loosestrife
(Lythnim salicaria) and its control on wetlands in
central Washington state. Ph.D. diss., Univ. of
Washington, Seattle.

Beaver, D. L. 1991. Swamp Sparrow. Pp. 486—487 in
The atlas of breeding birds of Michigan (R. Brew-
er, G. A. McPeek, and R. L. Adams, Jr., Eds.).
Michigan State Univ. Press, East Lansing.

Brown, M. and J. J. Dinsmore. 1986. Implications of
marsh size and isolation for marsh bird manage-
ment. J. Wildl. Manage. 50:392-397.

Burger, J. 1985. Habitat selection in temperate marsh-
nesting birds. Pp. 253-281 in Habitat selection in
birds (M. L. Cody, Ed.). Academic Press, Orlan-
do, Florida.

Burns, J. T. 1982. Nests, territories, and reproduction
of Sedge Wrens {Cistothonis platen.'ii.s). Wilson
Bull. 94:338-349.

Comer, P. J. 1996. Wetland trends in Michigan since
1800: a preliminary assessment. U.S. Environ-
mental Protection Agency, Water Division and
Wildlife Division, Michigan Department of Nat-

ural Resources. Michigan Natural Features Inven-
tory, Lansing.

CowARDiN, L. M., V. Carter, F C. Golet, and E. T.
LaRoe. 1979. Classification of wetlands and deep-
water habitats of the United States. U.S. Dept. In-
terior, FWS/OBS-79/3 1:1-103.

Dahl, T. E. 1990. Wetland losses in the United States:
1780s to 1980s. U.S. Dept. Interior, U.S. Fish
Wildl. Serv., Washington, D.C.

Fredrickson, L. H. and F. A. Reid. 1988. Preliminary
considerations for manipulating vegetation. U.S.
Fish Wildl. Serv., leaflet 13.4.9.

Higgins, K. F, J. L. Oldemeyer, K. J. Jenkins, G. K.
Clambey, and R. F. Harlow. 1994. Vegetation
sampling and methods. Pp. 567-591 in Research
and management techniques for wildlife and hab-
itats, fifth ed. The Wildlife Society, Bethesda, Mary-
land.

Johnson, R. R., B. T. Brown, L. T. Haight, and J. M.
Simpson. 1981. Playback recordings as a special
avian censusing technique. Stud. Avian Biol. 6:
68-75.

Johnson. R. R. and J. J. Dinsmore. 1986. The use of
tape-recorded calls to count Virginia Rails and So-
ras. Wilson Bull. 98:303-306.

Kaminski, R. M. and H. H. Prince. 1981. Dabbling
duck and aquatic macroinvertebrate responses to
manipulated wetland habitat. J. Wildl. Manage.
45:1-15.

Kantrud, H. a. and R. E. Stewart. 1984. Ecological
distribution and crude density of breeding birds
on prairie wetlands. J. Wildl. Manage. 48:426-
437.

Keddy, P. a. and a. a. Reznicek. 1985. Vegetation
dynamics, buried .seeds, and water level fluctua-
tions on the shorelines of the Great Lakes. Pp. 33-
58 in Coastal wetlands (H. H. Prince and F M.
DTtri, Eds.). Lewis Publishers, Chelsea. Michi-
gan.

Kiviat, E. 1996. American Goldfinch nests in purple
loosestrife. Wilson Bull. 108:182-186.

Kroodsma, D. and j. Verner. 1978. Complex singing
behavior among Cistothoni.s wrens. Auk 95:703-
716.

114

THE WILSON BULLETIN • Vol. Ul, No. I, March 1999

Lanyon, W. E. 1997. Great Crested Flycatcher (Myiar-
chus crinitiis). In The birds of North America, no.
300 (A. Poole and E Gill, Eds.). The Academy of
Natural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
D.C.

Lowther, P. E. 1993. Brown-headed Cowbird (Mol-
othriis ater). In The birds of North America, no.
47 (A. Poole and F. Gill, Eds.). The Academy of
Natural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Malecki, R. a., B. Blossey, S. D. Hight, D. Schroe-
DER, L. T. Kok, and J. R. Coulson. 1993. Bio-
logical control of purple loosestrife. BioScience
43:680-686.

Marion, W. R., T. E. O’Meara, and D. S. Maehr.
1981. Use of playback recordings in sampling elu-
sive or secretive birds. Stud. Avian Biol. 6:81-85.

McCarty, J. P. 1996. Eastern Wood-pewee {Contopu.s
virens). In The birds of North America, no. 245
(A. Poole and F. Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Milliken, G. a. and D. E. Johnson. 1984. Analysis
of messy data, vol. 1: designed experiments. Van
Nostrand Reinhold Co., New York.

Moore, W. S. 1995. Northern Flicker (Colapte.s' aii-
ratiis). In The birds of North America, no. 166
(A. Poole and E Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Mowbray, T. B. 1997. Swamp Sparrow (Melo.spiza
georgiana). In The birds of North America, no.
279 (A. Poole and E Gill, Eds.). The Academy of
Natural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
D.C.

Peck, G. and R. James. 1987. Breeding birds of On-
tario: nidiology and distribution. Vol. 2. Passer-
ines. Royal Ontario Museum, Toronto.

PETER.SON, R. T. 1990. Peterson's held guides, eastern/
central bird songs. Houghton Mifflin Co., Boston,
Massachusetts.

Prince, H. H. and C. S. Flegel. 1995. Breeding avi-

fauna of Lake Huron. Pp. 247—272 in The Lake
Huron ecosystem: ecology, hsheries and manage-
ment (M. Munwar, T. Edsall, and J. Leach, Eds.).

5. P.B. Academic Publishing, Amsterdam, The
Netherlands.

Rawinski, T. j. and R. A. Malecki. 1984. Ecological
relationships among puiple loosestrife, cattail, and
wildlife at the Montezuma National Wildlife Ref-
uge. N.Y. Fish Game J. 31:81-87.

Reinert, S. E. and E C. Golet. 1986. Breeding ecol-
ogy of the Swamp Sparrow in a southern Rhode
Island peatland. Northeast Sect. Wildl. Soc. 1986:
1-13.

SAS Institute. 1990. SAS/STAT user’s guide. Version

6, fourth ed. Volume 2. SAS Institute, Cary, North
Carolina.

Skinner, L. C., W. J. Rerdall, and E. L. Fuge. 1994.
Minnesota’s purple loosestrife program: history,
hndings, and management recommendations.
Minnesota Dept. Nat. Resour. Spec. Publ. 145:1 —
36.

SOKAL, R. R. AND E J. Rohlf. 1981. Biometry, second
ed. W. H. Freeman and Company, New York.
Swift, B. L., S. R. Orman, and J. W. Ozard. 1988.
Response of Least Bitterns to tape-recorded calls.
Wilson Bull. 100:496-499.

Thompson, D. Q. 1989. Control of purple loosestrife.

U.S. Fish Wildl. Serv., leaflet 13.4.11.

Thompson, D. Q., R. L. Stuckey, and E. B. Thomp-
son. 1987. Spread, impact, and control of puiple
loosestrife. U.S. Fish Wildl. Serv., Fish Wildl.
Res. 2:1-55.

Wetherbee, D. K. 1968. Southern Swamp Sparrow.

U.S. Natl. Mus. Bull. 237:1475-1490.

Weller, M. W. 1978. Management of freshwater wet-
lands for wildlife. Pp. 267-284 in Freshwater wet-
lands (R. E. Good, D. E Whigham, and R. L.
Simpson, Eds.). Academic Press, New York.
Weller, M. W. and C. E. Spatcher. 1965. Role of
habitat in the distribution and abundance of marsh
birds. Iowa State Univ. Agr. Home Econ. Exp.
Sta., Spec. Rept. 43:1-31.

Weller, M. W. and L. H. Fredrickson. 1974. Avian
ecology of a managed glacial marsh. Living Bird
12:269-291.

Wilcox, D. A. 1995. Wetland and aquatic macrophytes
as indicators of anthropogenic disturbance. Nat.
Areas J. 15:240-248.

Short Communications

Wilson Bull., 111(1), 1999, pp. 115-116

Bald Eagle Predation on Common Loon Chick

James D. Paruk,'-^ Dean Seanfieldr and Tara Mack^

ABSTRACT. — We report predation of a Common
Loon (Gavia immer) chick by an adult Bald Eagle
(Haliaeetiis leucocephaliis) in northern Wisconsin. Re-
ceived 27 Feb. 1998, accepted 26 Sept. 1998.

Common Loon (Gavia immer) chicks may
be vulnerable to aerial and underwater pred-
ators, including Bald Eagle (Haliaeetus leu-
cocephalus). Common Raven (Corvus corax).
Herring Gull (Larus argentatus), snapping
turtle (Chelydra serpentina), northern pike
(Esox niger), and muskellunge (Esox masqui-
nongy, Yonge 1981, McIntyre 1988, J. Wilson
and M. Meyer, pers. comm.). However, few
observers have actually observed such pre-
dation (McIntyre 1988). Here we describe the
first documented observation of a Bald Eagle
killing a loon chick.

On 26 July 1996, while collecting data on
parental effort on a pair of color-marked adult
Common Loons at the Turtle Flambeau Flow-
age in north central Wisconsin (46° 0' N,
90° 10' W), we observed an adult Bald Eagle
capture a 15 day-old loon chick. Loon chicks
are most prone to predation during the first
two weeks after hatching, but continue to be
vulnerable up to 4-5 weeks of age (Yonge
1981, McIntyre 1983, pers. obs.). At 10:00
CST an adult eagle circled the territory sev-
eral times before perching near the top a lone,
8 m tall tamarack (Larix laricina) on a small
island (0.1 ha). An adult eagle(s) had been
observed within the loons’ territory on 2 pre-
vious occasions (19, 22 July). When the rest-
ing adult loons saw the eagle, they wailed and

' Dept, of Biological Sciences, Idaho State Univ.,
Pocatello, ID 83209.

^ Earthwatch, 680 Mt. Auburn St., Watertown, MA
02272.

^ Sigurd Olson Env. Institute, Northland College,
Ashland, WI 54806.

Corresponding author; E-mail; parujame@isu.edu

tremoloed several times. The chick, initially
with the adults, disappeared from our sight,
apparently beneath some overhanging alder
(Alnus sp.) along the island’s edge. By 10:20
the adult loons stopped vocalizing; the male
resumed foraging dives, and the female re-
mained alert on the surface. However, at
10:30, two non-resident loons intruded on the
established pair’s territory, triggering aggres-
sive interactions for 10-12 min until the in-
truding pair left the area. The territorial adults
remained alert on the surface with the chick
still out of sight.

At 10:58 the eagle swooped down and
grabbed something offshore near the island
(because of the density of the alder we could
not positively identify it). The adult loons im-
mediately started giving 3-note wails and 3-
note tremolos, and the male let out several
yodels. At 1 1 :00 the eagle attempted to fly
with the loon chick (now visible) in its talons,
but had trouble becoming airborne in the alder
thicket. Thirty seconds later, the eagle attained
flight without the chick in its talons and re-
turned to its previous perch in the tamarack.
It remained there for 10 min before it flew
from the area. The adult loons continued to
wail and tremolo for 20 min while swimming
around the island. They remained in the area
until 12:25, then swam about 300 m to the
other side of their territory. We searched for
the chick 3 hours later and found it dead, with
puncture wounds to the head and a crushed
skull. We skinned and mounted the chick as a
study specimen, but did not record its body
mass.

ACKNOWLEDGMENTS

We thank Earthwatch and Biodiversity Research In-
stitute for financial support, and both Earthwatch and
Loon Watch for providing volunteers. Jeff Wilson pro-
vided logistical and moral support throughout the field
sea,son. Additionally, we thank D. Gorde, T. Gerstell

115

116

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

and two anonymous reviewers for improving the man-
uscript.

LITERATURE CITED

McIntyre, J. W. 1983. Nurseries: a consideration of
habitat requirements during the early chick-rear-
ing period in Common Loons. J. Field Omithol.
54:247-253.

McIntyre, J. W. 1988. The Common Loon: spirit of
northern lakes. Univ. of Minnesota Press, Min-
neapolis.

Yonge, K. S. 1981. The breeding cycle and annual
production of the Common Loon (Govia immer)
in the boreal forest region. M.S. thesis, Univ. of
Manitoba, Winnipeg.

Wilson Bull., 111(1), 1999, pp. 116—117

Territorial takeover in Common Loons {Gavia immer)

James D. Paruk'

ABSTRACT. — Breeding Common Loons (Gavia
immer) are well known for vigorously defending their
territory from conspecihcs. Territory holders are not
previously known to be supplanted by loons during the
breeding season. I observed a pair of adult Common
Loons displace a resident pair from their territory
shortly after a territorial conflict; the takeover coincid-
ed with the death of the resident loons’ chick caused
by an adult Bald Eagle (Haliaeetus leucocephalu.s).
Received 27 Feb. 1998, accepted 7 Sept. 1998.

Common Loons (Gavia immer) are philo-
patric, but territory switching occurs infre-
quently (Piper et al. 1997; Evers, Reaman,
Kaplan, and Paruk, unpubl. data). Our under-
standing of territory switching in Common
Loons remains largely unknown (Piper et al.
1997). In 1995—1997, while coordinating
studies of parental effort and social flocking
in Common Loons at the Turtle Flambeau
Flowage in northern Wisconsin (46° 00' N,
90° 10' W), I observed a territory takeover.

The Turtle Flambeau Flowage is a large im-
poundment (5798 ha) that contains 24-26
loon territories. Territories are generally well
delineated by coves or islands, but there are
several places where the presence of small is-
lands makes it difficult to distinguish individ-
ual territorial boundaries.

On 26 July 1996, at 10:30 CST a pair of
unbanded loons entered the territory (Long Is-

'Dcpt. of Biological Sciences, Idaho State Univ., Po-
catello, ID 83209; E-mail: parujame@isu.edu.

land, LI) of an established color-banded pair.
All four birds moved behind an island and
were out of sight for 2 min. A territorial bird
(sex could not be determined because of an
obstruction) “surface rushed” one of the in-
truders driving it onto an island with repeated
bill thrusts. At 10:42, the territorial pair was
reunited and the intruders were no longer in
sight.

At 10:58, an adult Bald Eagle (Haliaeetus
leucocephalus) killed the resident pairs’ 15
day-old chick (Paruk et al. 1999). At this age,
loon chicks are still dependent upon adults for
food and protection from predators (Dulin
1988, McIntyre 1988). The LI pair remained
near the location where the eagle killed the
chick until 12:25, at which time they swam to
the other side of their territory. I searched the
territory for the banded loons from 15:25-
16:00, but did not observe them. Instead, I
observed two unbanded loons in the territory.

On 27 July, I observed two unbanded,
paired adults in the LI territory. The original
pair was not observed on 27 July, but on 28
July an assistant spotted them several hundred
meters south of their former territory.
Throughout the rest of the summer, until 18
August, several observations (n = 9) con-
firmed that the original territorial pair had
been supplanted by two unbanded individuals.
The unbanded birds were observed foraging,
resting and preening in the LI territory until
the end of the observation period on 18 Au-
gust.

Prior to the takeover, a pair of unbanded

SHORT COMMUNICATIONS

loons engaged the territorial pair on four con-
secutive days (22-25 July), with several ago-
nistic encounters. On both 24 and 25 July the
LI pair left their chick and engaged in ritual-
ized behavior (jerk diving, facing away) with
conspecifics for 20 and 26 min respectively.
Whether these were the same individuals that
took over the territory is unknown. It seems
likely that there was a territorial dispute prior
to the death of the LI pair’s chick and the
subsequent takeover.

Conmion Loons do not typically abandon
their territory after the loss of a chick, al-
though they will often show less aggression
towards conspecifics and may wander more
frequently than loon pairs with chicks (Evers,
pers. comm.; pers. obs.). Thus, it is unlikely
that the resident LI pair simply abandoned
their territory after the loss of their chick.

In 1997, the former LI pair remained to-
gether and occupied a new territory 400 m
south of their original territory. Two unbanded
loons nested in the former LI territory. Zack
and Stutchbury (1992) proposed that non-
breeders are likely to acquire territories they
visit frequently and Piper and coworkers
(1997) proposed that mid- to late seasonal
movements observed in loons may be partly
explained by their searching for new or un-
occupied territories (reconnaissance hypothe-
sis). The lack of distinct physical barriers sep-
arating loon territories and the high number of
nonbreeders present on the Turtle Flambeau
Flowage (Belant 1989, pers. obs.) may result
in higher intrusion rates and more interterri-
torial interaction, lowering territorial stability
(Strong and Bissonette 1988; Belant 1991;
Piper et al. 1997; Evers et al., unpubl. data).
The timing of the observed supplanting/take-
over supports the reconnaissance hypothesis
for loon movements during mid- to late sum-

1 17

mer, and suggests that Common Loons may
actively engage in territory acquisition during
the breeding season. To what extent this take-
over was precipitated by the death of the res-
ident pair’s chick remains unknown.

ACKNOWLEDGMENTS

The behavioral project, of which this observation
was a part, was funded by Earthwatch and Biodiversity
Research Institute. I am particularly indebted to ail the
volunteers for their assistance in gathering the data, but
e.specially so to R Hart, M. Lockman, T. Mack, D.
Seefeldt, A. Turpen and M. Wiranowski. J. Wilson pro-
vided logistical and moral support throughout the in-
vestigation. I also thank T. Ford, T. Gerstell, J. Mc-
Intyre and an anonymous reviewer for their comments
and suggestions on this manuscript.

LITERATURE CITED

Belant, J. L. 1989. Common Loon productivity and
brood habitat use in northern Wisconsin. M.S. the-
sis, Univ. of Wisconsin, Stevens Point.

Belant, J. L. 1991. Territorial activities of Common
Loons on single-pair lakes. Passenger Pigeon. 54;
115-118.

Dulin, G. S. 1988. Pre-fledging feeding behavior and
sibling rivalry in Common Loons. M.S. thesis.
Central Michigan Univ., Mt. Pleasant.

McIntyre, J. W. 1988. The Common Loon: spirit of
northern lakes. Univ. of Minnesota Press, Min-
neapolis.

Paruk, j. D., D. Seanfield, and T. Mack. 1999. Bald
Eagle predation on Common Loon chick. Wilson
Bull. 111:116-117.

Piper, W. H., J. D. Paruk, D. C. Evers, M. W. Meyer,
K. B. Tischler, M. Klich, and J. J. Hartigan.
1997. Local movements of color-marked Common
Loons. J. Wildl. Manage. 61:1253-1261.

Strong, P. and J. Bissonette. 1988. Territorial activ-
ities of Common Loons on multiple-pair lakes. Pp.
19-24 in Papers from 1987 conference on loon
research and management (P. Strong, Ed.). North
American Loon Fund, Meredith, New Hampshire.
Zack, S. and B. J. Stutchbury. 1992. Delayed breeding
in avian social systems: tlie role of territory quality
and “floater” tactics. Behaviour 123:195-219.

118

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

Wilson Bull., 1 1 1(1), 1999, pp. 118-119

Courtship Behavior of the Buff-necked Ibis (Theristicus caudatus)

Nathan H. Rice

ABSTRACT — Buff-necked Ibis (Theristicus cau-
datus) courtship displays include “Bill Popping” and
grasping twigs, behaviors reported for other species of
ibis. Received 4 Dec. 1997, accepted 20 Aug. 1998.

The courtship displays of the ibises (Thres-
kiornithidae) are poorly known and little doc-
umented, including those of the South Amer-
ican Buff-necked Ibis (Theristicus caudatus-,
del Hoyo et al. 1992, Hancock et al. 1992).
Here, 1 present field observations of apparent
courtship displays by this species made at
Parque Nacional San Luis (22° 40' S, 57° 21'
W), depto. Concepcion, on 25 October 1996,
in the semihumid forest/savanna region of
northern Paraguay.

At approximately 09:00 I heard and ob-
served two Buff-necked Ibises at the top of a
dead tree, approximately 30 m above the
ground making snapping noises with their
bills. The presumed male and female would,
in turn, grasp dead twigs (2-3 cm diameter)
with their bills and release them, never break-
ing the twigs from the branch. Occasionally,
the birds would make low grunts. This con-
tinued for about 3 min until the birds abruptly
stopped, faced each other, pointed their bills
vertically (similar to photo in del Hoyo et al.
1992:480) and gave a loud squawking call.
Then they slapped their bills together horizon-
tally, making a sound similar to two hollow
pieces of bamboo hitting each other, and flew
away together to the northwest.

About 30 min later, approximately 1.5 km
northwest of the first site, I heard the same
snapping sound and quickly located a pair of
ibises, perhaps the same birds, displaying in a
similar manner. Again, the behavior lasted
about 3 min and ended with the birds slapping
their heads together and flying away. On 30

Natural History Museum. Univ. of Kansas,
Lawrence, K.S 6604.5;

E-mail: nricc@falcon.cc.ukans.edu

October 1997, a male (testes 18 X 11 mm,
KU #88342) and female (ovary 20 X 14 mm,
largest ovum 3X3 mm, oviduct convoluted
3 mm, specimen deposited in Museo Nacional
Historia Natural del Paraguay) were collected
that may be the pair I observed earlier. Based
on the specimen gonad sizes, the birds were
in breeding condition.

The behaviors I observed resemble court-
ship and breeding displays of other ibises (del
Hoyo et al. 1992, Hancock et al. 1992). Un-
mated male ibises use Bill Popping (Hancock
et al. 1992) when soliciting females. This be-
havior involves the bird snapping its gaped
bill shut, occasionally making a popping
sound. Some ibis grab a twig and shake it dur-
ing this behavior. The initial “snapping” noise
that I heard may have been a product of Bill
Popping by the male. Although the pair I ob-
served never removed twigs from the tree,
they did grab sticks during the encounter — a
further indication that this was the Bill Pop-
ping behavior described in Hancock and co-
workers (1992).

Males will also respond to females entering
their territories with a ritualized form of Spar-
ring display (Hancock et al. 1992). Sparring
behavior consists of one bird, in this case the
male, lunging at the other. The female will flee
and not fight back. Both of the behavioral se-
ries I observed ended with the birds confront-
ing one another (i.e., slapping bills together)
and then departing. Perhaps this was a modi-
fication of the Sparring behavior described in
Hancock et al. (1992).

ACKNOWLEDGMENTS

Director O. Romero of Departamento del Inventario
Biologico Nacional and Museo Nacional Historia Nat-
ural del Paraguay helped in innumerable ways. C. Eox,
director of Dirreccion de Barques Nacionales y Vida
Silvestre graciously granted permits for the work at
San Luis. A. L. Aquino, Director of CITES, kindly
provided logistical help in getting to San Luis. I thank
San Luis park guard, A. Acosta, for accommodating

SHORT COMMUNICATIONS

1 19

us. Special thanks to R. Faucett, M. Robbins and J.
Simmons for field assistance. This manuscript was im-
proved by comments from A. T. Peterson, M. Robbins
and two anonymous reviewers. This work was partially
funded by a Panorama Society grant from the Univer-
sity of Kansas Natural History Museum, and by the
generous support of the Ornithology Interest Group at
the University of Kansas.

LITERATURE CITED

DEL Hoyo, J., a. Elliot, and J. Sargatal. (Eds.y
1992. Handbook of the birds of the world, vol. 1.
Lynx Edicions, Barcelona, Spain.

Hancock, J. A., J. A. Kushlan, and M. P. Kahl. 1992.
Storks, ibises, and spoonbills of the world. Aca-
demic Press, San Diego, California.

Wilson Bull, 111(1), 1999, pp. 119-121

Habitat Use by Masked Ducks Along the Gulf Coast of Texas

James T. Anderson ‘ and Thomas C. Tacha'-^

ABSTRACT. — We counted 47 Masked Ducks {No-
monyx dominiciis) in seven flocks during the fall and
winter of 1992-1993 on 1009 64.75-ha plots in the
Coastal Plains of Texas. Among the three wetland sub-
classes used by Masked Ducks, bird densities were
higher on lacustrine littoral aquatic-bed rooted vascular
and lacustrine littoral aquatic-bed floating vascular
than palustrine scrub-shrub broad-leaved deciduous
wetlands. These wetlands provide important habitat
even though they are not the most abundant wetlands
in the region. Received 23 June 1998, accepted 25
Aug. 1998.

Masked Ducks (Nomonyx dominicus) are
small, scarce, and reclusive inhabitants of
wetlands throughout eastern South America
and north into Texas and Florida (Johnsgard
and Carbonell 1996, Lockwood 1997, Todd
1997). Little ecological data exist for this spe-
cies anywhere, but particularly at the northern
extent of its range. Appropriate habitat has
been subjectively defined as overgrown
swamps and marshes, where aquatic plants
like water hyacinth (Eichornia crcissipes) and
water lilies (Nymphciceae spp.) occur (Johns-
gard and Carbonell 1996, Todd 1997). Our ob-

‘ Caesar Kleberg Wildlife Research Institute, Cam-
pus Box 218, Texas A&M Univ.— Kingsville, Kings-
ville, TX 78363.

^ Present Address: Wildlife and Fisheries Program,
Division of Forestry, West Virginia Univ., P.O. Box
6125, Morgantown, WV 26505-6125.

^ Deceased.

■* Corresponding author.

Jective was to quantify habitat use by Masked
Ducks in the Coastal Plains of Texas.

The study area covered 5.5 million ha from
Galveston Bay, Texas south to the Rio Grande
River (Anderson et al. 1996, 1998). The re-
gion is dominated by coastal prairie and sandy
plains in the southeast, and rice fields and
coastal marsh in the northeast (Anderson et al.
1996). Palustrine and estuarine wetlands (Co-
wardin et al. 1979) are the most abundant of
the wetland systems (Muehl et al. 1994).

We conducted ground based surveys of all
wetlands located on 512 quarter-sections
(64.75-ha plots) in 1991-1992 and 1009 in
1992-1993 (Anderson et al. 1996, 1998). Sur-
veys for Masked Ducks on wetlands were
conducted during September, November, Jan-
uary, and March. Wetlands were classified ac-
cording to Cowardin and coworkers (1979).
Surveys were part of a larger project address-
ing waterbird habitat use (Anderson 1 994, An-
derson et al. 1996), waterbird abundance (An-
derson et al. 1998), and wetland abundance
(Muehl et al. 1994).

We compared densities (no./ha) of Masked
Ducks among wetland types on which they
occurred using ANOVA and Scheffe's proce-
dure as the mean separation technique with a
= 0.05 (SAS Institute Inc. 1988). We included
in the analysis all wetlands of a type on which
Masked Ducks were observed (Anderson et
al. 1996). We compared microsite habitat use
in wetlands with two-way contingency tables
and a G-test (Sokal and Rohlf 1995). Count

120

THE WILSON BULLETIN • Vol. HI, No. 1. March 1999

periods were considered independent because
counts were at least two months apart, wet-
lands were dynamic (Muehl et al. 1994), and
the number of birds varied among count pe-
riods (Anderson et al. 1996, 1998). All
Masked Duck density data were rank trans-
formed (Conover and Iman 1981, Potvin and
Roff 1993) because of the large number of
wetlands that had no Masked Ducks. Data
were back transformed for presentation.

We did not observe any Masked Ducks dur-
ing 1991-1992. During 1992-1993, we count-
ed 47 Masked Ducks (September 6; Novem-
ber 4; January 34; March 3) in 7 flocks in 4
separate basins. Masked Ducks occupied 0.3%
of quarter-sections surveyed during 1992-
1993. All observations were made in the
coastal and other crop strata of the area re-
ferred to as the Texas Mid-coast (Anderson et
al. 1996, 1998). Masked Duck flocks averaged
6.7 birds (SE = 3.16; range 1-25). Sixty-four
percent (n = 22) of undisturbed Masked
Ducks were observed feeding.

Masked Duck densities (no./ha) on lacus-
trine littoral aquatic-bed rooted vascular (x =
0.93; SE — 0.52) and lacustrine littoral aquat-
ic-bed floating vascular (x = 0.40; SE = 0.40)
wetlands were not different, but densities on
both were greater than densities on palustrine
scrub-shrub broad-leaved deciduous (x =
0.16; SE = 0.15) wetlands (ANOVA: F =
10.23; df = 2, 199; P < 0.001). Masked
Ducks did not occur on the other 79 wetland
subclasses that were surveyed. Masked Ducks
were equally likely to occur in open water
(43%) and in emergent vegetation microsites
within these three wetland types (57%; G-test:
G = 0.2; P > 0.05).

Masked Ducks occupied wetlands that av-
eraged 8.25 ha (SE = 1.94) in area. All wet-
lands were seasonally or semipermanently
flooded with fresh water and had emergent
vegetation interspersed with open water [i.e.,
cover type two (Stewart and Kantrud 1971)J.
Rooted vascular vegetation on occupied wet-
lands was primarily yellow lotus {Nelumbo lu-
tea), but yellow waterlily (Nuphar mexicana)
was also present. Floating vascular wetlands
were dominated by water hyacinth. Scrub-
shrub vegetation was primarily huisache (Aca-
cia .smallii) and sesbania (Sesbania drummon-
dii).

Although few Masked Ducks were ob-

served, it was apparent that they prefered wet-
lands with abundant vegetation, particularly
aquatic-bed and scrub-shrub wetlands. No
Masked Ducks were observed in emergent
wetlands, as has often been stated (Johnsgard
and Carbonell 1996, Lockwood 1997, Todd
1997). Masked Ducks also are known to occur
in flooded rice fields in Venezuela (Gomez-
Dallmeier and Cringan 1990), but none were
observed in the Texas rice fields we surveyed.

It is interesting to note that Masked Ducks
were not found on smaller (palustrine) aquat-
ic-bed wetlands, which are more common
than lacustrine littoral aquatic-bed wetlands in
the area (Muehl et al. 1994). Their absence
from these wetlands may be related to pref-
erence for larger (lacustrine) wetlands, which
provide greater habitat diversity, increased
protection from predators, and more food re-
sources (Anderson et al. 1996). However,
Weller (1968) and Todd (1997) indicated that
Masked Ducks can use smaller wetlands than
other stiff-tailed ducks because they can take-
off vertically like dabbling ducks.

Previously, no specific information existed
on Masked Duck densities (Johnsgard and
Carbonell 1996). Anderson and coworkers
(1998) estimated 3817 Masked Ducks oc-
curred in coastal Texas during January 1993,
but only 354 during March 1993. Masked
Ducks are not as abundant or wide-spread as
other waterfowl species in the study area and
are rare throughout their range (Johnsgard and
Carbonell 1996). The presence of Masked
Ducks in Texas may be a temporary phenom-
enon (Johnsgard and Hagemeyer 1969, Blan-
kenship and Anderson 1993) or they may al-
ways be present, but seldom seen, as a result
of their rarity, secretive nature, and the pre-
ponderance of private property in Texas.

ACKNOWLEDGMENTS

We appreciate the landowners for allowing us access
to their properties. We thank G. T. Muehl for assisting
with data collection. S. L. Beasom, E S. Guthery, R.
Kennamer, and an anonymous referee reviewed the
manuscript. Funding was provided by the Caesar Kle-
berg Foundation for Wildlife Conservation and' the
Texas Prairie Wetlands Project.

LITERATURE CITED

Anderson, J. T. 1994. Wetland use and selection by
waterfowl wintering in coastal Texas. M.S. thesis,
Texas A&M Univ.— Kingsville, Kingsville.

SHORT COMMUNICATIONS

121

Anderson, J. T, T C. Tacha, G. T. Muehl, and D.
Lobpries. 1996. Wetland use by waterbirds that
winter in coastal Texas. Natl. Biol. Serv. Inf. Tech.
Rep. 8: 1-40.

Anderson, J. T, G. T. Muehl, and T. C. Tacha. 1998.
Distribution and abundance of waterbirds in coast-
al Texas. Bird Pop. 4:1-15.

Blankenship, T. L. and J. T. Anderson. 1993. A large
concentration of Masked Ducks (Oxyura domini-
ca) on the Welder Wildlife Refuge, San Patricio
County, Texas. Bull. Tex. Ornithol. Soc. 26:19-
21.

Conover, W. J. and R. L. Iman. 1981. Rank transfor-
mations as a bridge between parametric and non-
parametric statistics. Am. Stat. 35:124-129.

CowARDiN, L. M., V. Carter, F. C. Golet, and E. T.
Laroe. 1979. Classification of wetlands and deep-
water habitats of the United States. U.S. Fish
Wildl. Serv. FWS/OBS-79/3 1 : 1-103.

Gomez-Dallmeier, F. and a. Cringan. 1990. Biology,
conservation and management of waterfowl in
Venezuela. Editorial Ex Libris, Caracas, Venezue-
la.

Johnsgard, P. a. and M. Carbonell. 1996. Ruddy
Ducks and other stifftails: their behavior and bi-
ology. Univ. of Oklahoma Press, Norman.

JOHN.SGARD, P. A. AND D. Hacjemeyer. 1969. The
Masked Duck in the United States. Auk 86:691-
695.

Lockwood, M. W. 1997. A closer look: Masked Duck.
Birding 29:386-390.

Muehl, G. T, T. C. Tacha, and J. T. Anderson. 1994.
Distribution and abundance of wetlands in coastal
Texas. Texas J. Ag. Nat. Resour. 7:85-106.

PoTviN, C. AND D. A. Roff. 1993. Distribution-free
and robust statistical methods: viable alternatives
to parametric statistics? Ecology 74:1617-1628.

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

SoKAL, R. R. and F. j. Rohlf. 1995. Biometry, third
edition. W. H. Freeman and Company, New York.

Stewart, R. E. and H. A. Kantrud. 1971. Classifi-
cation of natural ponds and lakes in the glaciated
prairie region. U.S. Fish Wildl. Serv., Resour.
Publ. 92:1-23.

Todd, F. S. 1997. Natural history of the waterfowl. Ibis
Publishing, Vista, California.

Weller, M. W. 1968. Notes on some Argentine ana-
tids. Wilson Bull. 80:189-212.

Wilson Bull, 111(1), 1999, pp. 121-123

Gizzarii Contents of Piping Plover Chicks in Northern Michigan

Francesca J. Cuthbert,*-^ Brian Scholtens,^ Lauren C. Wemmer,’ and Robyn McLain^-"*

ABSTRACT. — The diet of Piping Plovers (Char-
adrius melodiis) is not well known and information on
diet requirements will enhance food resource assess-
ment and identification of suitable habitat for this rare
species. Discovery of four dead Piping Plover chicks
at Grand Marais, Michigan, allowed us to examine
their digestive tracts for identifiable prey. Gizzard con-
tents represented 16 families in 6 orders of freshwater
and terrestrially occurring insects confirming behav-
ioral observations that plover chicks opportunistically
capture insects in shallow water and along shorelines.
The most commonly taken orders were Hymenoptera,

' Dept, of Fisheries and Wildlife, Univ. of Minne-
sota, St. Paul, MN 55108.

^ Dept, of Biology, College of Charleston, Charles-
ton, SC 29424.

^ Univ. of Michigan Biological Station, Pellston, Ml
49769.

^ Present address: 5005 Elderben y Dr., Reading, PA
19606.

^ Corresponding author;

E-mail: cuthbOOl (3)maroon. tc.umn.edu

Coleoptera, and Diptera. Received 6 May 1998, ac-
cepted 30 Aug. 1998.

Little is known about the diet or foraging
behavior of the Piping Plover (Charadrius
melodus) during any part of its annual cycle.
Federal threatened and endangered status
(U.S. Fish and Wildlife Service 1985) and
sensitivity to human disturbance preclude col-
lection of birds for stomach content analysis
and require use of nondisruptive techniques to
sample food while plovers are present. Be-
cause food availability is critical to shorebird
reproductive success, migration, and over-
winter survival (Howe 1983, Helmers 1992),
assessment of food resources is an important
component of conservation efforts for this
species. Direct observations of food prefer-
ence and foraging ecology are needed to im-

122

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

prove assessment of food resources and allow
identification of areas appropriate for critical
habitat designation and for reintroduction ef-
forts (U.S. Fish and Wildlife Service 1988).

The Piping Plover is a visual rather than
tactile feeder, capturing invertebrates moving
on the beach surface. Information on diet has
been derived from gizzard contents (Bent
1929), direct observation of feeding (Cairns
1977), sampling of organisms present in the
habitat (Whyte 1985, Nordstrom 1990, Loe-
gering 1992, Nordstrom and Ryan 1996) and
fecal analysis (Nicholls 1989, Shaffer and La-
porte 1994); however, studies of prey actually
consumed by Piping Plovers have been re-
ported only for marine environments. General
diet for the species is described as freshwater
and marine invertebrates washed up on the
shore and terrestrial invertebrates (Haig 1992).
Insects appear to be a major dietary compo-
nent in most or all habitats occupied by Piping
Plovers throughout the year.

While monitoring plover nests in Michigan
during 1996 and 1997, we salvaged carcasses
of four chicks and examined gizzard contents.
We believe this is the first direct information
on diet reported for Piping Plovers from the
Great Lakes population.

STUDY AREA AND METHODS

In 1996 three pairs of Piping Plovers nested near the
mouth of the Sucker River east of Grand Marais,
Michigan (46° 40' N, 85° 56' W) on the shore of Lake
Superior. Two chicks from one of these pairs disap-
peared at six days of age. The carcass of one was dis-
covered approximately 1.5 weeks later and the other
2.5 weeks later. The fresh carcass of the third chick
from this brood was discovered a few hours after its
disappearance at 19 days of age. In 1997, a fourth
Piping Plover chick carcass was found in the same
general area. This chick disappeared from its brood
when it was one week old and was found three days
later. The digestive tract from each chick was removed,
cut open and Hushed with 70% ethyl alcohol. Only
gizzards yielded identifiable samples. The contents of
the four gizzards were examined with a dissecting mi-
croscope and identified to family by an entomologist
(Scholtens) familiar with in.sects of the region. Num-
bers of individual prey were estimated.

RESULTS AND DISCUSSION

Parts of adult and larval insects were the
only prey identified in the gizzards. Prey rep-
resented 16 families in 6 orders; Hymenoptera
(32%), Coleoptera (29%), Diptera (28%), He-

miptera and Homoptera (10%), and Ephem-
eroptera (1%). Based on the natural history of
these families in northern Michigan, they can
be characterized as inhabiting shoreline/wet
sand (Dolichopodidae, Ephydridae), shallow
water/wet sand (Corixidae, Dytiscidae, Hali-
plidae), beach vegetation/sand surface (Aphi-
didae, Braconidae, Carabidae, Cicadellidae,
Curculionidae, Ichneumonidae) and general
shoreline habitat (Superfamily Chalcidoidea,
Chironomidae, Formicidae, Muscidae, un-
identified Ephemeropteran family). The only
previous information related to the prey of the
Great Lakes Piping Plover was found in a
study of invertebrates present within National
Park lands being assessed as potential reintro-
duction sites (Nordstrom 1990, Nordstrom
and Ryan 1996). Nordstrom (1990) found 48
families of 9 orders of insects and 1 family of
arachnid on the shore of Lake Superior in Pic-
tured Rocks National Lakeshore (approxi-
mately 20 km west).

In the Great Lakes region. Piping Plovers
nest on wide sandy beaches and forage along
the water line of Lake Michigan and Lake Su-
perior. Birds occasionally glean insects from
beach vegetation and at some sites forage
along the edges of creeks and shallow beach
ponds. At the Grand Marais site, adults and
chicks were observed foraging primarily
along the edge of the Sucker River and in
shallow pools of water and wet depressions in
the sand along the river. Aquatic insects in the
chick gizzards are consistent with this obser-
vation, and the presence of terrestrial, phy-
tophagous insects indicates that chicks also
gleaned insects from beach vegetation. Pres-
ence of aquatic algae-eating beetles (Halipli-
dae) in the gizzards suggests that chicks pick-
ed insects from algae on the river edges.
While Nordstrom and Ryan (1996) reported a
predominance of Dipterans in the Lake Su-
perior habitat they sampled, we found pre-
dominantly Hymenopterans and Coleopterans
in the chick gizzards. We identified two fam-
ilies (Corixidae and Dytiscidae) in the giz-
zards that were not reported by Nordstrom
(1990). Members of both families inhabit
streams, ponds, and stagnant pools associated
with beaches.

Given the constraints on disturbance and
collecting, opportunistic discovery of dead
plovers and subsequent study of their diges-

SHORT COMMUNICATIONS

123

live tracts may contribute information vital to
understanding the diet of this endangered spe-
cies. For example, insects gleaned from veg-
etation, algal mats, and the water surface
probably would not be revealed by traditional
methods (e.g., sticky traps) used to sample in-
vertebrates in the habitat. The information on
foraging behavior and prey selection that both
gizzard and fecal analyses provide is needed
to increase accuracy of methods used to sam-
ple invertebrates from the habitat. It is impor-
tant to note that all three methods of quanti-
fying plover food resources (gizzard analysis,
fecal analysis, and sampling from habitat)
may be greatly affected by the time samples
are obtained because of temporal variation in
insect abundance. Fecal analysis offers the ad-
vantage that numerous samples can be col-
lected from the same individuals to reveal
temporal patterns of prey selection; however,
this method underestimates soft-bodied inver-
tebrates (Shaffer and Laporte 1994). Because
gizzard contents have undergone less diges-
tion, they are presumably less biased in this
regard, but this has not been confirmed.

ACKNOWLEDGMENTS

Funds and permission to monitor and salvage plo-
vers were provided by the Endangered Species Pro-
gram, Michigan Department of Natural Resources and
the Michigan Field Office and Region 3 Endangered
Species Office, U.S. Fish and Wildlife Service. Addi-
tional support was received from the University of
Michigan Biological Station and the University of
Michigan Research Experiences for Undergraduates
Program (National Science Foundation Grant #94-
0512). We thank employees of Whitefish Point Bird
Observatory and Grand Marais landowners for their
cooperation with Piping Plover monitoring efforts.

LITERATURE CITED

Bent, A. C. 1929. Charadrius melodus. U.S. Natl.

Mus. Bull. 146:236-246.

Cairns, W. E. 1977. Breeding biology and behavior of
the Piping Plover {Cluiradriiis melodus) in south-
ern Nova Scotia. M.S. thesis, Dalhousie Univ.,
Halifax, Nova Scotia.

Haig, S. M. 1992. Piping Plover {Charadrius melo-
dus). In The birds of North America, no. 2 (A.
Poole, P. Stettenheim, and E Gill, Eds.). The
Academy of Natural Sciences, Philadelphia, Penn-
sylvania; The American Ornithologists’ Union,
Washington, D.C.

Helmers, D. L. 1992. Shorebird management manual.
Western Hemisphere Shorebird Reserve Network.
Manomet. Massachusetts.

Howe, M. 1983. Breeding ecology of North American
shorebirds: patterns and constraints. Proc. West.
Hemisphere Waterfowl Waterbirds Symp. 1 :95-
100.

Loegering, J. P. 1992. Piping Plover breeding biology,
foraging ecology and behavior on Assateague Is-
land National Seashore, Maryland. M.S. thesis,
Virginia Polytechnic Inst, and State Univ., Blacks-
burg.

Nicholes, J. L. 1989. Distribution and other ecological
aspects of Piping Plovers (Charadrius melodus)
wintering along the Atlantic and Gulf Coasts.
M.S. thesis. Auburn Univ., Auburn, Alabama.

Nordstrom, L. H. 1990. Assessment of habitat suit-
ability for reestablishment of Piping Plovers in the
Great Lakes National Lakeshores. M.S. thesis,
Univ. of Missouri, Columbia.

Nordstrom, L. H. and M. R. Ryan. 1996. Invertebrate
abundance at occupied and potential Piping Plover
nesting beaches: Great Plains alkali wetlands vs.
the Great Lakes. Wetlands 16:429—435.

Shaffer, F. and P. Laporte. 1994. Diet of Piping Plo-
vers on the Magdalen Islands, Quebec. Wilson
Bull. 106:531-536.

U.S. Fish and Wildlife Service. 1985. Determination
of endangered and threatened status for the Piping
Plover. Federal Register 50:50726-50734.

U.S. Fish and Wildlife Service. 1988. Recovery plan
for Piping Plovers breeding on the Great Lakes
and Northern Great Plains. U.S. Fish and Wildlife
Service. Twin Cities, Minnesota.

Whyte, A. J. 1985. Breeding ecology of the Piping
Plover (Charadrius melodus) in central Saskatch-
ewan. M.S. thesis, Saskatchewan Univ.. Saska-
toon.

124

THE WILSON BULLETIN • Vol. HI, No. I, March 1999

Wilson Bull., II Id), 1999, pp. I24-I28

Nesting of Four Poorly-Known Bird Species on the Caribbean Slope of

Costa Rica

Bruce E. Young'-^'^ and James R. Zook-

ABSTRACT — We describe the nests of four species
of birds from the Caribbean slope of Costa Rica. A
Great Potoo (Nyctibius grand is, nest previously un-
known from Mesoamerica) nest was nothing more than
a crevice in a high branch of a large tree, similar to
those reported in South America. A nest of the Torrent
Tyrannulet [Serpophaga cinerea) was found along a
river at 35 m elevation, much lower than previous
breeding reports for this normally montane species.
Also, we confirm systematists’ predictions that the
Tawny-chested Llycatcher (Aplianotriccus capilalis), a
species of near-threatened conservation status, is a sec-
ondary cavity nester. Finally, we report on the second
known nest of the Sooty-faced Finch (Lysurus crassi-
rostris) from montane forest. Received 20 May 199H,
accepted 7 Oct. 1998.

RESUMEN. — Describimos los nidos de cuatro es-
pecies de aves de la vertiente del caribe en Costa Rica.
Un nido de Nyctibius grandis (nido era anteriormente
desconocido en Mesoamerica) era nada mas que una
grieta en una rama alta de un arbol grande, muy pa-
recido a los nidos de la misma especie encontrados en
America del Sur. Describimos un nido de Serpophaga
cinerea, el cual fue encontrado en un rio a los 35
msnm, mucho mas bajo que los otros registros de esta
especie del bosque montano. Describimos unas obser-
vaciones que indican que el Aphanotriccus capitalis,
una especie ligeramente amenazada desde la punta de
vista conservacionista, anida secondariamente en los
huecos dentro de los arboles. Finalmente, presentamos
la segunda descripcion del nido de Lysurus crassiros-
tris, el cual fue encontrado en un bosque montano.

Despite decades of intensive ornithological
study, the nests and eggs of a number of Cen-
tral American bird species are poorly known
(Skutch 1954, 1960; Stiles and Skutch 1989).
Especially enigmatic are species inhabiting

' Organizacion para Estudios Tropicales, Apartado
676-2050. .San Pedro de Montes de Oca, Costa Rica.

^Apartado 182-4200. Naranjo de Alajuela, Costa
Rica.

' Current Address: Latin America and Ctiribbean Di-
vision. The Nature Conservancy, 4245 North Fairfax Dr..
Arlington. Virginia 22203; E-mail: byoung@tnc.org
■* Corresponding author.

the relatively inaccessible habitats of the hu-
mid Caribbean slope. The lack of reproductive
information about these species hinders phy-
logenetic studies of the relationships among
avian lineages, studies of intraspecific varia-
tion, general analyses of reproductive behav-
ior, and the development of management prac-
tices for conserving avian biodiversity. Here
we describe the first Mesoamerican nest of the
Great Potoo (Nyctibius grandis), the first low-
land nest of the Torrent Tyrannulet {Serpo-
phaga cinerea), the first nest of the Tawny-
chested Flycatcher (Aphanotriccus capitalis),
and the second nest of the Sooty-faced Finch
(Lysurus crassirostris).

STUDY AREA

Our observations were made in the 45,000 ha La
Selva-Braulio Canillo National Park reserve complex
in Heredia Province, northeastern Costa Rica. The re-
serve complex, the largest protected elevational tran-
sect in Central America, extends from montane rain
forest surrounding the Barva Volcano at 2,900 m down
to lowland wet forest at the La Selva Biological Sta-
tion at 35 m elevation on the Caribbean slope (Timm
et al. 1989).

Observations were made in 1997 during routine bird
monitoring activities at La Selva (10° 26' N, 83° 59'
W) and at a remote campsite at 1070 m elevation in
Braulio Carrillo National Park (10° 16' N, 84° 5' W).
Annual rainfall at La Selva averages 3962 mm, with a
relative dry period between January and March during
most years (Sanford et al. 1994). Although precipita-
tion data are .scarce for higher elevations, annual rain-
fall may average over 5000 mm at the 1070-m site,
where clouds frequently bathe the premontane rain for-
est in mist (Hartshorn and Peralta 1988). Average can-
opy height varies from 28-38 m at La Selva to 22-36
m at 1070 m (Lieberman et al. 1996).

NEST DESCRIPTIONS AND DISCUSSION

Nyctibius grandis. — The Great Potoo is one
of three Central American species of the ge-
nus Nyctibius, the only genus in the exclu-
sively Neotropical family Nyctibiidae. Al-
though difficult to observe in daylight because
it roosts motionless on canopy branches, the

SHORT COMMUNICATIONS

125

FIG. 1. Adult and nestling Great Potoo (Aycr/Wi/A-
grandis) on the branch of a large Hernatulia didyman-
tha tree. La Selva Biological Station, Costa Rica, April

1997.

Great Potoo is readily detectable at night by
its characteristic calls (Perry 1979, Slud
1979). It ranges from southern Mexico to
southeastern Brazil and central Bolivia (AOU

1998, Howell and Webb 1995).

On 5 February, a visitor to the station (K.
McGowan) found a single Great Potoo
perched 34 m high on the branch of a 44 m
tall Hernandia didyrnantha (Hemandiaceae)
tree. The tree was located 500 m inside of old
growth forest, at the edge of a large treefall
gap on a steep hill. The bird was perched on
an upward sloping section of an S-shaped
branch approximately 20 cm in diameter.
Thereafter on daily visits, we observed a bird
in exactly the same position.

On 4 April, we observed for the first time
a fully feathered chick on the branch in front
of the adult. The chick was paler in coloration
than the adult and about one quarter the size
(Fig. 1). On subsequent days, the chick moved
between a hidden position under the adult’s
breast feathers and the branch immediately in
front of the adult. The adult remained almost
motionless and never moved along the branch.
We last saw the birds on 20 April. Despite
extensive searches on all neighboring branch-
es in the following days and months, neither
adult nor young was seen again.

The motionless adult we saw was probably

incubating an egg initially and then brooding
a chick, although we could not determine the
hatching date. Although collections of adults
on eggs in Brazil and observations of N. gri-
seus in Costa Rica indicate that males incu-
bate during the day (Skutch 1970, Sick 1993),
we could not identify the sex of the individual
(or individuals) we saw because sexes are
similar in outward appearance in Great Potoos
(Land and Schultz 1963, Wetmore 1968). No
nest was visible and, from our vantagepoint
30 m from the tree and level with the nest,
only a slight crevice in the branch was visible.
Although the location appeared precarious, re-
ports from Brazil suggest that a notch in the
nest branch can securely hold an egg (Sick

1993) .

At a nest in Venezuela, a chick remained
with its parent for a month, and then alone for
almost another month before growing to a size
greater than two-thirds that of the adult and
dispersing (Vanderwerf 1988). This observa-
tion suggests that the chick we observed,
which never attained half the size of the adult,
did not survive. In addition, the wings did not
appear sufficiently developed for sustained
flight. Despite an extensive ground search, we
found no evidence of its having fallen from
the nest. The chick may have been taken by
an arboreal predator such as a monkey {Cebus
capucinus, Ateles geojfroyi, or Alouatta pal-
liata), tayra (Eira barbara), or Collared For-
est-Falcon (Micrastur sernitorquatus), all
common in the area. Even though the attempt
was probably unsuccessful, its daily survivor-
ship rate in the (presumed) egg and chick
stage of 98.7% is substantially higher than the
93% rate measured for understory cup-nesting
birds the same year in the same area (B.
Young, unpubl. data).

Serpophaga cinerea. — The Torrent Tyran-
nulet is a conspicuous resident of highland
rivers from Costa Rica to Venezuela and Bo-
livia (AOU 1998). Their cup-shaped nest at-
tached to vegetation above rivers is well
known (Skutch 1960). The elevational distri-
bution of the species is variously described as
250-2500 m in different parts of its range
(Meyer de Schauensee and Phelps 1978, Hilty
and Brown 1986, Stiles and Skutch 1989, Rid-
gely and Gwynne 1993, Ridgely and Tudor

1994) .

We first detected Torrent Tyrannulets on the

126

THE WILSON BULLETIN • Vol. HI. No. I, March 1999

Sarapiqui River in 1994 in Chilamate, 5 km
west of La Selva at an elevation of 40 m. In
March 1997, we began seeing individuals
along the same river where it passes through
La Selva at 35 m elevation. These were the
first observations of the species in La Selva
in four decades of ornithological investigation
(Levey and Stiles 1994). During the first week
of April, a pair of tyrannulets began construc-
tion of a nest 0.5 m above the water level in
shrub vegetation growing in the middle of a
small island in the river, which is approxi-
mately 80 m wide and 1 m deep at this point.
The nest was typical for the species in being
cup shaped, supported by vertical branches,
with feathers and moss woven into the struc-
ture (Skutch 1960). We monitored the nest ev-
ery 2-5 days until 22 April when two eggs
were found in the nest. During each visit, two
birds were active near the nest, adding mate-
rial and adjusting its structure. On our next
visit, on 30 April, the nest had disappeared
and the birds were not present.

Despite its conspicuousness on rivers, pre-
vious reports of the species occurring in the
lowlands are of scattered observations of in-
dividuals as low as 100 m elevation (Hilty and
Brown 1986, Ridgely and Gwynne 1993, Rid-
gely and Tudor 1994). The Torrent Tyrannu-
lets we observed may have strayed from their
higher elevation habitat because of a hydro-
electric project in progress higher up the Sar-
apiqui River.

Aphanotriccus capitalis. — The Tawny-
chested Flycatcher occurs in second growth
and disturbed forest in Nicaragua and Costa
Rica (AOU 1998). Nests of both this species
and its congener are apparently undescribed
(Lanyon and Lanyon 1986), and both have
near threatened conservation status as a result
of their small, fragmented ranges (Collar et al.
1994). We provide two observations to sug-
gest the species nests in either cavities or
crevices in trees.

On 23 April, a group of birdwatchers spot-
ted a pair of Tawny-chested Flycatchers build-
ing a nest in the hollow of a dead branch stub
in an otherwise live Alchornea costaricense
(Euphorbiaceae) tree. On 30 April, we again
observed two birds carrying fine nesting ma-
terial, including moss, to the hollow for about
30 min. The birds apparently abandoned the
attempt, as we never saw them there again.

The rectangular hollow was oriented slightly
upward, 8 cm deep and 25 X 8 cm wide, 1.5
m up in the 36 cm dbh tree. The tree was in
a small, shady clearing 10 m away from a
small (0.25 ha) patch of second growth forest.

A local naturalist guide, E. Castro, reported
finding an active nest of this species in a hol-
low section of a 30 m diameter clump of
Asian bamboo {Guadua sp.) in a patch of sec-
ond growth forest 2 km north of La Selva.
Castro reported seeing adults carrying food to
the nest and later feeding a fledgling in the
vicinity. The nest was 5.9 m above the
ground, 13 cm in diameter (the diameter of
the bamboo), and entered through a 5 cm high
by 2 cm wide teardrop-shaped opening. The
bottom of the opening was 3 cm above a node,
leaving a shallow area for the nest. This find-
ing confirms predictions that Aphanotriccus
builds nests in crevices based on its phylo-
genetic closeness to LMthrotriccus and Cne-
motriccus, two genera known to build nests in
crevices (Lanyon 1986, Lanyon and Lanyon
1986). This crevice nesting habit may aid in
the conservation of the species; Guadua bam-
boo is widely introduced in the region for use
in supporting banana trees on plantations.

Lysurus crassirostris. — The Sooty-faced
Finch occurs in dense vegetation in wet, mid-
elevation forest in Costa Rica and Panama
(AOU 1998). Its one congener, the Olive
Finch (L. castaneiceps), occurs in humid mon-
tane forests along the coastal Andes moun-
tains from Colombia to Peru (Ridgely and Tu-
dor 1989).

On 7 May we discovered two adult Sooty-
faced Finches entering and exiting a nest 2.1
m high attached to the side of a large Sorn-
mera sp. (Rubiaceae) tree (dbh = 46 cm) next
to our camp at 1070 m, several kilometers in-
side undisturbed forest. The nest tree was lo-
cated at the side of a little used trail at the
edge of a stream crossing. The nest itself was
woven into a thick epiphyte mat on a section
of the trunk directly above the 3 m wide
stream. The nest was a bulky, covered dome
with a side entrance. The nest was almost en-
tirely constructed of fresh moss with a lining
of thin, black, stringy fungal rhizomorphs and
strips of dried bamboo (Chusquea sp.) leaves.
The trunk of the nest tree was covered with
the same moss as was used to construct the

SHORT COMMUNICATIONS

127

nest, causing the nest to be fairly inconspic-
uous.

On 1 1 May, after observing the adults
spending long periods of time in the nest, we
exainined the eggs. The two eggs were whit-
ish and speckled with lavender. The speckles
were densest around the thick ends of the
eggs. Assuming the adults were incubating
and that the female had finished laying, the
clutch size for this nest was two, typical for
birds of humid tropical forests (Skutch 1985).

This nest was very similar to the one other
nest described for the species, although the
latter was built into the side of a fern stem
and apparently was not associated with water
(Barrantes 1994). These two nests were sim-
ilar to the single nest of the congeneric Olive
Finch described from Ecuador (Schulenberg
and Gill 1987). All nests of the two species
were bulky and dome-shaped, constructed pri-
marily of mosses on the outside, cryptically
situated in moss-dominated vegetation, and, in
two cases, located over moving water. The
nests of the two species differ in that the Olive
Finch nest was built on the side of a rock in-
stead of a tree and had a lining of dry leaves
instead of rhizomorphs. The eggs differ sub-
stantially in that the eggs of the Olive Finch
were immaculate white (Schulenberg and Gill
1987) compared with the spotted eggs found
in both Sooty-faced Finch nests (Barrantes
1994). Interestingly, eggs in two collections
attributed to the Olive Finch are spotted
(Schulenberg and Gill 1987), suggesting that
the Ecuadorian discovery of all white eggs
may have been atypical for the group.

ACKNOWLEDGMENTS

We thank R. Tenorio of the Area de Conservacion
Cordillera Volcanica Central and the Organization for
Tropical Studies for facilitating our studies. We are
grateful for the field observations of R. Alvarado, A.
Downs, Y. Hernandez, and A. Hies, and to K. Mc-
Gowan for finding the potoo. R. Thiele photographed
the potoo. J. Blake and two anonymous reviewers pro-
vided many useful comments to improve the manu-
script. Financial support for our studies was generously
provided by H. M. S. Ambassador Michael Jackson
and the British Embassy in Costa Rica (for work in
Braulio Carrillo) and by the Chiquita Brands Company
(for work in La Selva).

LITERATURE CITED

American Ornithologists’ Union. 1998. Check-li.st
of North American birds, seventh ed. American
Ornithologists’ Union, Washington, D.C.

Barrantes, G. 1994. First description of the nest and
eggs of the Sooty-faced Finch. Wilson Bull. 106:
574.

Collar, N. J., M. J. Crosby, and A. J. Stattersfield.
1994. Birds to watch 2: the world list of threat-
ened birds. Smithsonian Press, Washington, D.C.

Hartshorn, G. S. and R. Peralta. 1988. Preliminary
description of primary forests along the La Selva-
Volcan Barva altitudinal transect. Pp. 281-296 in
Tropical rainforests: diversity and conservation (F.
Alameda and C. M. Pringle, Eds.). California
Academy of Sciences, San Francisco.

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

Howell, S. N. G. and S. Webb. 1995. The birds of
Mexico and northern Central America. Oxford
Univ. Press, New York.

Land, H. C. and W. L. Schultz. 1963. A proposed
subspecies of the Great Potoo, Nyctibius grandis
(Gmelin). Auk 80:195-196.

Lanyon, W. E. 1986. A phylogeny of the thirty-three
genera in the Empidonax assemblage of tyrant fly-
catchers. Am. Mus. Novit. 2846:1-64.

Lanyon, W. E. and S. M. Lanyon. 1986. Generic sta-
tus of Euler’s Flycatcher: a morphological and
biochemical study. Auk 103:341—350.

Levey, D. J. and F. G. Stiles. 1994. Birds: ecology,
behavior, and taxonomic affinities. Pp. 217-228 in
La Selva: ecology and natural history of a Neo-
tropical rain forest (L. A. McDade, K. S. Bawa,
H. A. Hespenheide, and G. S. Hartshorn, Eds.).
Univ. of Chicago Press, Chicago, Illinois.

Lieberman, D., M. Lieberman, R. Peralta, and G. S.
Hartshorn. 1996. Tropical forest structure and
composition on a large-scale altitudinal gradient
in Costa Rica. J. Trop. Ecol. 84:137—152.

Meyer de Schauensee, R. and W. H. Phelps. Jr.
1978. A guide to the birds of Venezuela. Princeton
Univ. Press, Princeton, New Jersey.

Perry, D. R. 1979. The Great Potoo in Costa Rica.
Condor 81:320-321.

Ridgely, R. S. and j. a. Gwynne, Jr. 1993. Guia de
las aves de Panama. ANCON, Panama City, Pan-
ama.

Ridgely, R. S. and G. Tudor. 1989. The birds of
South America, vol. 1 . Univ. of Texas Press, Aus-
tin.

Ridgely, R. S. and G. Tudor. 1994. The birds of
South America, vol. 2. Univ. of Texas Press, Aus-
tin.

Sanford, R. L., Jr., P. Paaby. J. C. Luvall, and E.
Phillips. 1994. Climate, geomorphology, and
aquatic systems. Pp. 19-33 in La Selva: ecology
and natural history of a Neotropical rain forest (L.
A. McDade, K. S. Bawa, H. A. Hespenheide, and
G. S. Hartshorn, Eds.). Univ. of Chicago Press,
Chicago, Illinois.

Schulenberg, T. S. and F. B. Gill. 1987. First de-
scription of the nest of the Olive Finch, Lysurus
castaneiceps. Condor 89:673-674.

128

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

Sick, H. 1993. Birds in Brazil. Princeton Univ. Press,
Princeton, New Jersey.

Skutch, a. E 1954. Life histories of Central American
birds. Pac. Coast Avifauna 31:1-448.

Skutch, A. E 1960. Life histories of Central American
birds 11. Pac. Coast Avifauna 34:1—593.

Skutch, A. E 1970. Life history of the Common Po-
too. Living Bird 9:265-280.

Skutch, A. E 1985. Clutch size, nesting success, and
predation on nests of Neotropical birds reviewed.
Ornithol. Monogr. 36:575-594.

Slud, P. 1979. Calls of the Great Potoo. Condor 81:
322.

Stiles, E G. and A. E Skutch. 1989. A guide to the
birds of Costa Rica. Cornell Univ. Press, Ithaca,
New York.

Timm, R. M., D. E. Wilson, B. L. Clauson, R. K.
LaVal, and C. S. Vaughan. 1989. Mammals of
the La Selva-Braulio Carrillo Complex, Costa
Rica. N. Am. Fauna 75:1-162.

Vanderwerf, E. a. 1988. Observations on the nesting
of the Great Potoo {Nyctibiu.s grandis) in central
Venezuela. Condor 90:948—950.

Wetmore, a. 1968. The birds of the Republic of Pan-
ama. Smithson. Misc. Collect. 150(2): 1—605.

Wilson Bull. 111(1), 1999, pp. 128-130

Sexual Dimoq^hism in the Song of Sumichrast’s Wren

Monica Perez-Villafana,' Hector Gomez de Silva and
Atahualpa DeSucre-Medrano^

ABSTRACT. — We report on a song-like vocaliza-
tion of female Sumichrast’s Wren (Hylorchilus sumi-
chrasti). The female song is a series of similar sylla-
bles, all at the same low pitch, that varies in length.
Thus, it differs strongly from the rich and complex
songs of male Sumichrast’s Wrens and of most other
wrens. Received 27 Feb. 1998, accepted 25 Aug. 1998.

There is increasing evidence that female
song is not as rare in birds as previously
thought (Langmore 1998). Among most
wrens, females have songs that are similar to
and sometimes combine with those of their
mates in antiphonal duets (Skutch 1940,
1960). Until recently, the only reported case
of strong sexual difference in wren songs was
in the southern House Wren {Troglodytes ae-
don) among which the females give a simple
twittering and/or a short trill, at least in Costa
Rica and Panama, generally countersinging

' Calle 1537-3, Col. San Juan de Aragon, Seccion
6, C.P. 07918, Mexico, D.E, Mexico.

^ Instituto de Ecologia, UNAM, Apartado Po.stal 70-
275, Ciudad Univ., UNAM, C.P. 04510, Mexico, D.E,
Mexico.

’ UNAM Campus Iztacala, Laboratorio de Zoologfa,
Apartado Postal 314, Tlalnepantla, Estado de Mexico,
Ctkligo Postal 54500, Mexico.

* Corresponding author;

E-mail: hgomez@nosferatu. ecologia. imam. mx

with the males (Chapman 1929; Skutch 1940,
1953). Distinct female songs have more re-
cently been recorded in other Troglodytes
wrens — one population of northern House
Wren (Johnson and Kermott 1990) and So-
corro Wren {T. sissonii; Howell and Webb
1995). During fieldwork on the life-history of
Sumichrast's Wren {Hylorchilus sumichrasti)
in Cerro de Oro, Oaxaca (18° 02' N, 96° 15'
W; Perez-Villafana 1997), we recorded the
previously unknown song of a female H. sum-
ichrasti.

Sumichrast’s Wren is sexually monomor-
phic in plumage. The birds we observed were
not color-banded; however, during direct ob-
servation of the members of a single pair from
March to July 1994, we realized that the pre-
viously unrecorded song was always made by
the bird that did not emit the more complex
song described by Howell and Webb (1995)
and Gomez de Silva (1997). By analogy with
other wrens that have strong sexual differenc-
es in song (in which the female’s song is the
simpler one), and from the birds’ behaviors,
we concluded that this previously unrecorded
song was the song of the female. We subse-
quently have heard this “female song’’ at dif-
ferent points along a 738 m transect at Cerro
de Oro, and throughout the range of Sumi-
chrast’s Wren: 2 km south of Amatlan (18° 50'

SHORT COMMUNICATIONS

129

kHz

1 Seconds

FIG. 1. Songs of female (A) and male (B) Sumichrast’s Wren (Hylorchilus sumichrasti). Recorded by S. N.
G. Howell 2 km south of Amatlan, Veracruz.

N, 96° 55' W), Agua Escondida (18° 32' N,
96° 47' W), OaxacaA^eracruz border on the
road to San Juan del Rio (17° 32' N, 95° 44'
W), and 2 km south of Bethania (17° 56' N,
96° 01' W).

The distinction between songs and calls is
sometimes unclear. In general, calls comprise
one or two syllables whereas songs are longer
vocalizations comprising multiple syllables
(Langmore 1998). The “female song” of
Sumichrast’s Wren is a simple phrase consist-
ing of a single repeated syllable (fundamental
< 2 kHz). In this respect, it resembles the
main song of the Cactus Wren (Campy lorhyn-
chus brunneicapillus) rather than the rich and
complex songs of male Sumichrast’s and of
most other wrens. The female song had 4-22
such syllables per song. The pause between
the first and second syllables, and to a lesser
extent the pause before the last syllable, are
the longest. This simple song of uniform fre-
quency contrasts with the males’ complex

songs which spans a range of frequencies and
contains syllables of variable form (Fig. 1, see
also other sonograms of male songs in Gomez
de Silva 1997 and Atkinson et al. 1993).

Females sing less frequently than males.
Along the transect at Cerro de Oro, censused
twice a month between April 1994 and March
1995, only 28.5-75% as many females were
recorded singing per morning as males. Fe-
males sometimes countersang or sang at the
same time as males. In the one focal pair, the
female countersang with the male 46.2% of
the time. Females usually sang with their bod-
ies held upright and their tails pointing down-
ward, the same position as singing males in
Cerro de Oro.

Nonantiphonal female songs may be more
widespread in wrens than previously thought.
The few records may be due to a scarcity of
detailed observation. Carmona (1989) ob-
served that female Canyon Wrens (Catherpes
rnexicanus) produce a vocalization that is dif-

130

THE WILSON BULLETIN • Vol. Ill, No. 1, March 1999

ferent from the male’s but gave no details
about the vocalization. A systematic survey of
female songs (or their general absence) in
wrens appears to be an interesting possibility
for research.

ACKNOWLEDGMENTS

We are grateful to T. Altamirano of the Museo de
las Ciencias Biologicas Iztacala, for partial support of
M. P. V.’s fieldwork in Cerro de Oro. S. N. G. Howell
kindly provided valuable comments and information
as well as recorded the songs and made the sonagrams
shown here. L. F. Baptista kindly made available his
sound analysis laboratory at the California Academy
of Sciences and offered helpful comments. The sug-
gestions of three anonymous reviewers helped improve
the manuscript.

LITERATURE CITED

Atkinson, P. W., M. Whittingham, H. Gomez de Sil-
va Garza, A. M. Kent and R. T. Maier. 1993.
Notes on the conservation, ecology and taxonomic
status of Hylorchilus wrens. Bird Conserv. Int. 3:
75-85.

Carmona, R. 1989. Contribucion al conocimiento de
la historia natural de Catherpes mexicanus (Trog-
lodytidae: Aves) en la Reserva Ecologica del Ped-
regal de San Angel, Mexico, D.F. B.Sc. thesis,
UNAM — Campus Iztacala, Mexico, D.F.

Chapman, F. M. 1929. My tropical air castle. Appleton,
New York.

Gomez de Silva G., H. 1997. Comparative analysis of
the vocalizations of Hylorchilus wrens (Troglo-
dytidae). Condor 99: 981—984.

Hardy, J. W. and D. J. Delaney. 1987. The vocali-
zations of the Slender-billed Wren {Hylorchilus
sumichrasti): who are its close relatives? Auk 104:
528-530.

Howell, S. N. G. and S. Webb. 1995. A guide to the
birds of Mexico and northern Central America.
Oxford Univ. Press, Oxford, U.K.

Johnson, L. S. and L. H. Kermott. 1990. Structure
and context of female song in a north-temperate
population of House Wrens. J. Field Ornithol. 61:
273-284.

Langmore, N. E. 1998. Functions of duet and solo
songs of female birds. Trends Ecol. Evol. 13:136—
140.

Perez-Villafana, M. 1997. Contribucion al conoci-
miento de la historia de vida de Hylorchilus sum-
ichrasti (Aves: Troglodytidae) en el norte del Es-
tado de Oaxaca. B.Sc. thesis, UNAM — Campus
Iztacala, Mexico, D.F.

Skutch, a. F. 1940. Social and sleeping habits of Cen-
tral American wrens. Auk 57:293—312.

Skutch, A. F. 1953. Life history of the Southern House
Wren. Condor 55:121—149.

Skutch, A. F. 1960. Life histories of Central American
birds II. Cooper Ornithological Society, Berkeley,
California.

Wilson Bull, 111(1), 1999, pp. 130-132

An Incident of Female-Female Aggression in the House Wren

Tom Alworth' and Isabella B. R. ScheibeF

ABSTRACT. — In this paper we describe one ex-
ample of female-female aggression in the House Wren
{Troglodytes aedon). An intruding female usurped the
resident female and paired with the resident male.
House Wrens are known for committing infanticide as
well as puncturing and removing eggs of conspecifics
and other species. These behaviors have been mainly
attributed to resident and floating males, but we sug-
gest that females may also be responsible. Received 22
.Inly I99H. accepted 3 Nov. 1998.

' The E. N. Huyck Pre.serve and Biological Research
Station, Rensselaerville, NY 12147.

^ Dept, of Biological Sciences, Univ. at Albany, Al-
bany, NY 12222.

’ Corresponding author;

E-mail: is5()4l @cnsunix. albany.edu

In many passerine bird species, males es-
tablish breeding territories in the spring,
which they defend against intruders. This
form of sexual competition among males has
been recognized as one of the driving forces
behind mating patterns and parental care (Da-
vies 1991, Andersson 1994). Aggression
among females has received much less atten-
tion, although it has recently been shown to
be more common among birds than initially
assumed (Lenington 1980, Leffelaar and Rob-
ertson 1985, Searcy 1986, Martin et al. 1990,
Slagsvold 1993, Hansson et al. 1997, Liker

SHORT COMMUNICATIONS

131

and Szekely 1997). Female-female aggression
may affect several aspects of mating systems
and parental care, for example, maintaining
monogamy (Slagsvold 1993) or reducing har-
em size (Hurly and Robertson 1985). Female-
female aggression in Red-winged Blackbirds
{Agelaius phoeniceus’, Beletsky 1996) as well
as Lapwings (Vanellus vonellus; Liker and
Szekely 1997) is strongest early in the breed-
ing season when females first settle on the ter-
ritories. One evolutionary force behind fe-
male-female aggression in polygynous mating
systems is the conflict between females for the
male’s parental investment (Slagsvold and Li-
fjeld 1994). However, female-female aggres-
sion is not limited to polygynous mating sys-
tems (Slagsvold 1993). Here we report an in-
cident of female-female aggression in the
House Wren {Troglodytes aedon).

We have been studying a population of
House Wrens on the E. N. Huyck Preserve
and Biological Research Station in Rensse-
laer ville, New York since 1992 and have color
banded all individuals since 1995. On 17 May
1997, 07:15 EST we observed a fight between
two females that lasted for 30 minutes. Fe-
male A, who had been paired to the resident
male of the territory since 12 May was chased
in circles both in the air and on the ground by
female B. Female B had been the resident fe-
male of the same territory in 1996 but was
paired with a different male. The 1996 male
was not seen in 1997. The fight included chas-
es by female B with occasional aggressive in-
teractions that included bodily contact and
pecking. Neither female vocalized during the
encounter. The resident male was perched and
visible during the whole fight; he sang but did
not participate in the fight. Female B eventu-
ally usurped the territory from female A,
paired with the resident male, and took over
the nest that was close to completion. Female
A was not seen again during the 1997 breed-
ing season. We suggest that female B was
probably fighting for the territory rather than
for the resident male. One of the most suc-
cessful males in our study population, who
was polygynous in 1996 and 1997, occupied
the adjacent territory and was at that time un-
paired. Female B did not pair with this un-
mated male but returned to the territory with
which she was familiar.

Many researchers strongly suspect that res-

ident birds and probably non-resident floaters
as well (Johnson and Kermott 1993) routinely
enter territories not their own and kill and/or
remove eggs or young from nests (Belles-lsles
and Pieman 1986, 1987; Quinn and Holroyd
1989; Kermott et al. 1991). Until recently
these birds have been assumed to be male
(Quinn and Holroyd 1989, Kermott et al.
1991), but it now appears that residents need
to be concerned about intruding females as
well. This and other observations of female-
female aggression (Freed 1986, Johnson and
Searcy 1996) demonstrate that female House
Wrens may play an equally important role as
the males in the selection of nest sites, terri-
torial defense, and intraspecific aggression.

ACKNOWLEDGMENTS

We thank K. R Able for his continuing support, and
all the landowners including the E. N. Huyck Preserve
for allowing us to put up our nest boxes. We are grate-
ful to our field assistants. This paper significantly ben-
efitted from the review and comments of L. S. John-
son. This study has been funded by the E. N. Huyck
Preserve and Biological Research Station (to T A. and
1. S.) and the New York State Museum Biological Sur-
vey (to I. S.)

LITERATURE CITED

Andersson, M. 1994. Sexual selection. Princeton
Univ. Press, Princeton, New Jersey.

Belles-Isles, J. C. and J. Picman. 1986. House Wren
nest destroying behavior. Condor 88:190-193.
Belles-Isles, J. C. and J. Picman. 1987. Suspected
intraspecific killing by House Wrens. Wilson Bull.
99:497-498.

Beletsky, L. 1996. The Red-winged Blackbird. Aca-
demic Press, Ltd., London, U.K.

Davies, N. B. 1991. Mating systems. Pp. 263-293 in
Behavioral ecology: an evolutionary approach (J.
R. Krebs and N. B. Davies, Eds.). Blackball Sci-
entific Publications, London, U.K.

Freed, L. A. 1986. Territory takeover and sexually se-
lected infanticide in tropical House Wrens. Behav.
Ecol. Sociobiol. 19:197-206.

Hansson, B., S. Bensch, and D. Hasselquist. 1997.
Infanticide in Great Reed Warblers: secondary fe-
males destroy eggs of primary females. Anim. Be-
hav. 54:297-304.

Hurly, T. A. and R. J. Robertson. 1985. Do female
Red-winged Blackbirds limit harem size? A re-
moval experiment. Auk 102:205-209.

Johnson, L. S. and L. H. Kermott. 1993. Why is
reduced male parental assistance detrimental to
the reproductive success of secondary female
House Wrens? Anim. Behav. 46:1 1 1 1-1 120.
Johnson, L. S. and W. A. Searcy 1996. Female at-

132

THE WILSON BULLETIN • Vol. Ill, No. I. March 1999

traction to male song in House Wrens (Trof^lo-
dytes aedon). Behaviour 133:357-366.

Kaufman, J. H. 1983. On the dehnitions and functions
of dominance and territoriality. Biol. Rev. 58:1-
20.

Kermott, L. H., L. S. Johnson, and M. S. Merkle.
1991. Experimental evidence for the function of
mate replacement and infanticide by males in a
north-temperate population of House Wrens. Con-
dor 93:630-636.

Leffelaar, D. and R. J. Robertson. 1985. Nest usur-
pation and female competition for breeding op-
portunities by Tree Swallows. Wilson Bull. 97:
221-224.

Lenington, S. 1980. Female choice and polygyny in
Red-winged Blackbirds. Anim. Behav. 28:347—
361.

Liker, a. and T. Szekely. 1997. Aggression among
female Lapwings, Vanellu.s vanellu.s. Anim. Be-
hav. 54:797-802.

Martin, K., S. J. Hannon, and S. Lord. 1990. Fe-
male-female aggression in White-tailed Ptarmigan
and Willow Ptarmigan during pre-incubation pe-
riod. Wilson Bull. 102:532—536.

Quinn, M. S. and G. L. Holroyd. 1989. Nestling and
egg destruction by House Wrens. Condor 91:206-
207.

Searcy, W. A. 1986. Are female Red-winged Black-
birds tenitorial? Anim. Behav. 34:1381-1391.

Slagsvold. T. 1993. Female-female aggression and
monogamy in Great Tits Panes major. Ornis.
Scand. 24:155-158.

Slagsvold, T. and T. J. Lifjeld. 1994. Polygyny in
birds: the role of competition between females for
male parental eare. Am. Nat. 143:59-94.

Wil.son Bull., 111(1), 1999, pp. 132-133

Nest Reuse by Wood Thrushes and Rose-breasted Grosbeaks

Lyle E. Friesen,' - Valerie E. Wyatt,' and Michael D. Cadman'

ABSTRACT. — We report on two instances of nest
reuse by Wood Thrushes (Hylocichia mu.stelina) within
the same breeding season, and three cases of nest reuse
in successive years, two by Wood Thrushes and one
by Rose-breasted Grosbeaks (Pheucticu.s liidovici-
anit.s). In each of the five cases of nest reuse, host
young were successfully fledged in the original nesting
episode and in the .second nesting episode. Although
occasional nest reuse within a single breeding season
has been reported before, our study is the first to doc-
ument reuse of the same nest in successive years by
Wood Thrushes and Rose-breasted Grosbeaks. Re-
ceived 29 April 1998, accepted 4 Oct. 1998.

Open nesting passerines, with the possible
exception of tyrannid flycatchers (Curson et
al. 1996), seldom reuse nests within and be-
tween breeding seasons (Briskie and Sealy
1988). Earlier authorities (Weaver 1949,
Brackbill 1958) were unaware of nest reuse
by Wood Thrushes (Hylocichia mu.stelina).
Roth and coworkers (1996) documented oc-

' Canadian Wildlife Service. 75 Farquhar Street,
Ciuelph, ON, Canada N I H 3N4.

^ Corresponding author:

E-mail: lyle.frie.sen@sympatico.ca

casional nest reuse by Wood Thrushes during
the same breeding season (three cases out of
389 first nests). However, they did not report
any nest reuse between years, describing such
an event as unlikely because nests usually dis-
integrate after the nesting season. We report
on two cases of nest reuse by Wood Thrushes
within a breeding season and three cases of
nest reuse in successive years, two by Wood
Thrushes and one by Rose-breasted Grosbeaks
(Pheucticus lucloviciaitus).

Data presented in this study were gathered
in the course of a larger study of the nesting
success of Wood Thrushes and Rose-breasted
Grosbeaks conducted in 1996 and 1997 in
Waterloo Region, a fragmented agricultural
landscape located in southwestern Ontario
(see Friesen et al., in press for a description
of the landscape). In these two years, 154
Wood Thrush nests and 63 Rose-breasted
Grosbeak nests were found and regularly
monitored to determine their outcome.

Two Wood Thrush nests (one each year)
were reused during the same breeding season,
with young successfully fledging in all four
nesting attempts (see Wyatt 1997 for a de-

SHORT COMMUNICA TIONS

133

tailed account of one of the renests). Neither
of these nesting attempts were parasitized by
Brown-headed Cowbirds (Molothriis citer) al-
though 47% of Wood Thrush nests on our
study sites contained cowbird eggs or young
(Friesen et al., in press). In 1996, 20 days
elapsed between the fledging of the first brood
and the initiation of the second clutch (June
20 to July 10); in 1997, this interval was 13
days (June 25 to July 8). Neither of the orig-
inal nests appeared to have been relined or
refurbished prior to its second use. It is likely
that the same pairs reused each of the nests
but this could not be confirmed because the
birds were not color-banded. Studies of band-
ed birds in Waterloo Region in 1998 showed
that at least half of the pairs attempted two
broods in a nesting season (Friesen, unpubl.
data).

We mapped and marked all of the Wood
Thrush nests found in 1996 (?i = 61) and ob-
served, through visits to the sites the follow-
ing spring prior to the breeding season, that
five (8%) of them survived the winter seem-
ingly intact. Two of these nests were subse-
quently reused in 1997: one nest which
fledged three Wood Thrushes and one cowbird
in 1996, fledged four thrushes and two cow-
birds the following year; the other nest fledged
two thrushes and one cowbird in each of the
years. Neither nest appeared to have been sig-
nificantly renovated in the second year, al-
though both were in poor repair by the time
the young fledged in 1997.

Rose-breasted Grosbeaks are typically sin-
gle brooded in southern Ontario (Friesen et
al., in press) and we found no evidence of nest
reuse within the same breeding season. Three
(12%) of the 24 grosbeak nests we found in
1996 survived the winter. One of these, in
which three young were fledged in 1996, was
reused in 1997 and again fledged three young.

Our results suggest that nest reuse is a con-

sistent, albeit infrequent, breeding strategy.
The reuse of old nests may have resulted from
a shortage of suitable nesting sites although it
seemed to us that apparently suitable alterna-
tive sites were present nearby. It may also be
that the birds reusing nests recognized the lat-
ter as being of high quality, borne out by the
fact that all five nest reuses resulted in fledged
host young. Our study is a reminder to re-
searchers of the importance of monitoring the
status of used nests both within and between
breeding seasons.

ACKNOWLEDGMENTS

We thank our tielci assi.stants D. Dieboldt. E. Mc-
Leish, M. Nighswander. R. Nonas, B. Pollock. A.
Spender, E. Stephens, N. Wessely, and S. Zaheer. Parts
of this project were supported by the Ontario Region
of Environment Canada's Canadian Wildlife Service.
Human Resources Development Canada, Environmen-
tal Youth Corps-Ontario, Long Point Bird Observato-
ry, and the Regional Municipality of Waterloo.

LITERATURE CITED

Brackbill, H, 1958. Nesting behavior of the Wood
Thrush. Wilson Bull. 70:70-89.

Briskie, J. V. AND S. G. Sealy. 1988. Nest reuse and
egg burial in the Least Elycatcher Empidonax inin-
iimis. Can. Eield-Nat. 102:729-730.

CuRSON, D. R.. C. B. Goguen, and N. E. Mathews.
1996. Nest-site reu.se in the Western Wood-Pewee.
Wilson Bull. 108:378-380.

Friesen, L. E., M. D. Cadman, and R. J. MacKav. In
press. Ne.sting success of Neotropical migrant
songbirds in a highly fragmented landscape. Con-
serv. Biol.

Roth, R. R., M. S. John.son, and T. J. Underwood.
1996. Wood Thrush (Hylocichla miistelina). In
The birds of North America, no. 246 (A. Poole
and E Gill, Eds.). The Academy of Natural Sci-
ences, Philadelphia, Pennsylvania; The American
Ornithologists’ Union. Washington. D.C.

Weaver, F. G. 1949. Wood Thrush. U.S. Natl. Mus.
Bull. 196:101-123.

Wyatt, V. E. 1997. Nest reused by Wood Thrush. Ont.
Birds 15(l):36-37.

134

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

Wilson Bull., 111(1), 1999, pp. 134-137

Singing in a Mated Female Wilson’s Warbler

William M. Gilbert'-^ and Adele F. Carroll-

ABSTRACT. — A female Wilson’s Warbler, Wilson-
ia pusilla, was heard singing regularly on the territory
of a male in middle inner-coastal California during ear-
ly April, 1996, and occasionally after mid-April. Based
on their behavior, the resident male and the singing
female were paired. The female averaged about four
songs/min during singing bouts, and was heard about
30% of the time during early April. The female’s song
was high pitched, and did not resemble typical male
“chatter” song. In contrast to the functions of female
song in many tropical and some temperate parulids,
this song seemed to serve as a simple contact vocali-
zation between mates, as call notes might. A single
female song heard in a newly formed pair in 1997
raises the possibility that such songs might function in
pair formation. Received 10 April 1998, accepted 25
Aii^. 1998.

Singing by female birds, while common
and perhaps characteristic among tropical spe-
cies, occurs much less commonly among tem-
perate species (Morton 1996). Tropical and
temperate wood warblers follow this pattern.
Songs of tropical parulid females commonly
are used in “duets” with mates and may func-
tion in pair formation, communication with
mate, and territorial defense (Spector 1992).
To serve such functions, consistency in oc-
currence and stereotypy in form of song seem
to be required. Song in temperate parulid fe-
males, where reported, has typically occurred
in few females within a population, and in
some cases song patterns have varied among
females (e.g., Nolan 1978, Hobson and Sealy
1990). This suggests that song in temperate
parulid females is idiosyncratic, and/or serves
very limited and infrequent functions. To our
knowledge, female song has been reported in
ten temperate parulid species from six genera:
Vermivora, Panda, Dendroica, Setophaga,
Seiurus, and Geothlypis (Spector 1992, Mol-
denhauer and Regelski 1996). Here we report

' 4630 Driftwood Ct., El Sobrante, CA 94803;
E-mail: wmglbrt@aol.com

2 I 147 Fre.sno Ave., Berkeley, CA 94707.

' Corresponding author.

singing in a mated female Wilson’s Warbler,
Wdsonia pusdla.

We first heard the female song on 3 April,
1996, from the territory (ca 0.2 ha) of a color-
banded resident male (the “male”) Wilson’s
Warbler in the Nature Study Area of Tilden
Regional Park, Contra Costa Co., California.
We subsequently heard this unusual (com-
pared with typical male “chatter” song; Fig.
IB) and distinctive song at various times be-
tween 07:15 and 11:40 PST through 18 April.
The high frequency song sounded “sharp”
and “squeaky” (Fig. lA, C). It was delivered
at a rate of 4.0 ± 0.6 songs/min (range = 2-
6, n = 7) within singing bouts (we considered
a singing bout to be continuous singing with
no pause greater than one minute between
songs). On four separate occasions, we ob-
served the beak of a Wilson’s Warbler to open
and move as the female song was heard, con-
firming that the song came from the species.
Behavioral observations (see below) indicate
that the singer was the resident female (the
“female”) in the territory she occupied. This
female also frequently chipped within this ter-
ritory (Fig. 1C).

From 3 through 12 April we located the fe-
male during 30% of our observation time (490
min) based on hearing her song and/or sight-
ing the singing bird. We confirmed that the
male was in her proximity 20% of our obser-
vation time (the male usually did not sing and
often was more difficult to locate). Bouts of
female song that we monitored lasted 11.5 ±

3.0 min (range = 1.5-27, n = 11), and male-
female separation distance was 6.8 ± 1.2 m
(range = 2-20, n = 22). From 13 through 30
April, we located the female just 5% of our
observations time (825 min), and located the
male in her proximity only 2% of that time.
Bouts of female song that we monitored lasted

3.1 ± 1.3 min (range = 1. 5-7.0, n = 4), and
mean male-female separation distance was 4.4
± 1.2 m (range = 0-12, n = 11). Three con-
secutive singing bouts monitored on 18 April

SHORT COMMUNICATIONS

135

FIG. 1. Wilson’s Warblers vocalizations recorded in April, 1996, within a breeding territory at a study site
in Contra Costa Co., California. A. Eight-note song, sung by resident female, with resident male in close
proximity. B. A typical “chatter song” of resident male. C. Three call notes, followed by a 6-note song, by
resident female. Recordings made by W. M. Gilbert using a SONY TCD-D8 recorder and a Sennheiser K3U
microphone. Spectrograms made with Canary 1.2.1 software (Cornell Laboratory of Ornithology) using a Mac-
intosh 7.5 system computer.

had separation intervals of 30 and 44 min. At
no time during our observations did we notice
unusual resident male behavior (compared
with other mated males) that might have elic-
ited singing in the female.

We relied on several contextual clues to de-
termine the sex of the bird singing the unusual
song (the sexes of west coast Wilson’s War-

blers often are indistinguishable in the field),
and that bird’s relationship to the resident
male on whose territory it sang: (1) the male
was mated (although we found no direct evi-
dence of breeding, unpaired males tend to sing
persistently, often from exposed perches,
while this male sang sporadically, often from
undergrowth); (2) we consistently heard the

136

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

unusual song from within the male’s territory,
but not from beyond its borders; (3) about
two-thirds of the time that we heard the un-
usual song, the male was sighted within 25 m
of the singer, and the male and the singer often
foraged in the same tree, in adjacent trees, or
in the same restricted area of undergrowth; (4)
even though the two birds frequently were
close, they never were in conflict; (5) we nev-
er simultaneously sighted nor heard the two
birds at opposite ends of the territory; and (6)
we never consistently sighted a third Wilson’s
Warbler within the male’s territory, as we
would have had the male been mated to a bird
other than the one singing the unusual song.
Based on this evidence, we concluded that the
bird singing the unusual song was a female
mated with the resident male on the territory
where she sang.

On 27 March, 1997, WMG observed a
newly formed (<3 h) pair of Wilson’s War-
blers on a territory adjacent to that in which
we heard the female song in 1996. The new
pair was in view for 41 min and the color-
banded male followed the unbanded female
through vegetation ranging from undergrowth
to tree canopy. Amid continuous chipping
from the pair, WMG heard one song indistin-
guishable by ear from the female songs heard
in 1996. It could not be determined which bird
delivered the song, nor if the female of the
pair was the same bird that sang the female
songs in 1996.

Information on female song in temperate
parulid species is limited. In female Prairie
Warblers (Dendroica discolor), a limited num-
ber of females’ songs heard were all simple
and perhaps “primitive,” highly variable
among individuals, and unlike normal male
songs (although some were identifiable to spe-
cies). These songs were delivered early in the
.season, were heard sporadically, and may
have been delivered by older females display-
ing more male-like behavior (Nolan 1978). In
Yellow Warblers (Dendroica petechia), Hob-
son and Sealy (1990) suggest that female song
can function in intrasexual conflicts within
very dense breeding populations.

The singing we heard from a female Wil-
son’s Warbler(s) occurred early in the season
and possibly came from an older bird(s), as
would be consistent with some findings of No-
lan (1978) for the Prairie Warbler and with

some correlates of female song in other spe-
cies where usually only the male sings (Nice
1943). Other aspects of the singing we heard
appear to differ from what occurs in females
of some other parulids, however. The songs
we heard were not used in duets with mates,
as in many tropical species (Spector 1992).
There was no evidence that they functioned in
intrasexual conflicts within dense populations
(Hobson and Sealy 1990). Finally, we heard
the female consistently for more than a week
in early April, 1996, as opposed to isolated
bouts of singing heard on single days (Nolan
1978).

The female songs we heard in 1996 (as well
as the single song heard in 1997) were inter-
spersed between chip notes and usually deliv-
ered with the resident male close to the fe-
male. This suggests the songs may have func-
tioned to communicate with a mate. The sin-
gle song heard in 1997 (if delivered by a
female and one different from the singer of
1996) introduces the additional possibility that
the song could serve a special communicatory
function during early Wilson’s Warbler pair-
ing. If so, then the persistent female singing
heard in 1996 would have been an abnormal
carry-over of that behavior into the nesting
period. Singing in that female may have re-
flected an abnormal hormonal balance, similar
to effects of testosterone injection in stimu-
lating song in female birds that normally don’t
sing (e.g., Baptista and Morton 1988).

ACKNOWLEDGMENTS

We thank the East Bay Regional Park District for
authorization to conduct research in the Tilden Nature
Area of Tilden Regional Park. We thank D. Kroodsma
and an anonymous reviewer for their comments and
critical review of this paper.

LITERATURE CITED

Baptista, L. E and M. L. Morton. 1988. Song learn-
ing in montane White-crowned SpaiTOWs; from
whom and when. Anim. Behav. 36:1753—1764.
Hobson, K. A. and S. G. Sealy. 1990. Female song
in the Yellow Warbler. Condor 92:259—261.
Moldenhauer, R. R. and D. J. Regelski. 1996. North-
ern Parula (Parula americami). In The birds of
North America, no. 215 (A. Poole and E Gill,
Eds.). The Academy of Natural Sciences, Phila-
delphia, Pennsylvania; The American Ornitholo-
gists’ Union, Washington, D.C.

Morton, E. S. 1996. A comparison of vocal behavior
among tropical and temperate passerine birds. Pp.

SHORT COMMUNICATIONS

137

258-28 1 /■/( Ecology and evolution of acoustic coni-
iminication in birds (D. Kroodsma and E. Miller,
Eds.). Cornell LIniv. Press, Ithaca, New York.

Nice, M. M. 1943. Studies in the life history of the
Song SpaiTow, part 2. Trans. Linnaean Soc. N.Y.
6:1-328.

Nolan, V., Jk. 1978. fhe ecology and behavior ol the
Prairie Warbler Deiulroiia discolor. Ornithol.
Monogr. 26:1-595.

Spector, D. a. 1992. Wt)od-warbler song systems: a
review of paruline singing behaviors. Curr. Orni-
thol. 9:199-243.

Wilson Bull.. 111(1), 1999, pp. 137-139

Laying Time of the Bronzed Cowbird

Brian D. Peer' --^ and Spencer G. Sealy'

ABSTRACT. — We report the first observations of
egg laying by the parasitic Bronzed Cowbird (Molotli-
rus aeneus). Three direct observations and two esti-
mates of laying times were made at two Northern Car-
dinal (Cardincdis cardinalis) nests. Bronzed Cowbirds
laid at 18.2 min ± 1.7 (SE) before sunrise (range 14—
24 min). Laying lasted 5-10 seconds. Although the
parasitic Brown-headed Cowbird (M. ater) and some-
times Shiny Cowbirds (M. bonariensis) also lay before
sunrise, direct observations of laying by other cow-
birds are required before it can be concluded that pre-
sunrise laying is an adaptation for brood parasitism.
Received 9 June 1998, accepted 5 Sept. 1998.

Avian brood parasites that are surreptitious
when parasitizing nests may avoid detection
by their hosts. Indeed, they often lay their
eggs in a matter of seconds; Sealy and co-
workers (1995) found this behavior to be
unique to the diverse groups of brood para-
sites. The parasitic Brown-headed Cowbird
(Molothrus ater) generally lays in the minutes
prior to sunrise and it has been suggested that
laying at this time, presumably when hosts are
less likely to be at their nests, is an adaptation
for brood parasitism (Chance and Hann 1942).
Scott (1991) found that female Brown-headed
Cowbirds lay their eggs an average of 9 min
before sunrise, whereas seven potential host
species all lay their eggs after sunrise. Shiny
Cowbirds (M. bonariensis) and, possibly, the
nonparasitic Bay-winged Cowbird (M. bacbu.s)

' Dept, of Zoology, Univ. of Manitoba, Winnipeg,
MB. R3T 2N2, Canada.

^Present address: 3163 5th St., East Moline, IL
61244; E-mail: bdpcowbird(5> aol.com
^ Corresponding author.

also sometimes lay before sunrise (see Scott
1991), but the data available to Scott (1991)
were insufficient to conclude that sunrise lay-
ing is an adaptation for brood parasitism.
There were no direct observations of laying
for the Bronzed Cowbird (M. aeneus). Carter
(1986) stated only that this brood parasite lays
“during dawn hours”. Here we report, to our
knowledge, the first recorded observations of
laying times for the Bronzed Cowbird.

METHODS

Our observations were made at the Welder Wildlife
Refuge in San Patricio County, Texas (28° 0' N, 97° 5'
W) in 1994. Both Bronzed and Brown-headed cow-
birds were present during the breeding season. After
locating a nest at which a host apparently had not com-
pleted laying, we watched it the following morning
beginning approximately 30 min before sunrise. We
hid far enough away so that hosts or visiting cowbirds
were not disturbed. The nests were observed with bin-
oculars when necessary. Sunrise (SR) times were ob-
tained from the website of the United States Naval
Observatory Astronomical Applications Department
(http://aa.usno.navy.mil/AA/). All times are Central
Standard Time.

RESULTS

Three Bronzed Cowbird laying events were
observed directly, all at Northern Cardinal
{Cardinalis cardinalis) nests. On 30 May
1994 we located a cardinal nest (94-16) con-
taining one cardinal egg. The following morn-
ing, BDP arrived at this nest at 05:14 (SR —
20 min) and found a Bronzed Cowbird egg
that was slimy, suggesting it had been laid re-
cently, plus one cracked cardinal egg. Later
the same day the nest contained two cardinal
eggs plus the cowbird egg. At 05:06 (SR -

138

THE WILSON BULLETIN • Vol. 111. No. I. March 1999

28 min) on 1 June the nest contents were the
same, but by 05:14 a second Bronzed Cow-
bird egg had been laid. Later that day, the
damaged cardinal egg was gone and a third
cardinal egg had been laid. On the morning of
2 June the female cardinal was accidentally
flushed from the nest, and the nest contents
were the same as the day before. At 05:14 (SR
— 20 min) BDP watched a female Bronzed
Cowbird fly directly to the nest, lay an egg,
and fly away in 5-10 s. BDP left momentarily
at 05:15, but upon returning at 05:19, found
four Bronzed Cowbird eggs in the nest. No
bird species is known to lay more than one
egg per day (Sturkie 1976), thus we assumed
a second female had laid an egg in this nest.

Nest 94-22 was found with one cardinal egg
on 14 June 1994. On 15 June a female
Bronzed Cowbird looked into the nest at 5:19
(SR — 14 min), but she did not lay. On 16
June K. Stewart observed two Bronzed Cow-
birds parasitize this nest, one at 05:17 (SR —
16 min) and a second at 05:19 (SR — 14 min).
Neither cardinal was present during the laying
events and both laying bouts lasted 5-10 s. In
addition to the three laying events observed,
we estimated the two other laying times by
taking the midpoints of repeated visits to the
nests (Scott 1991) and found that Bronzed
Cowbirds laid their eggs 18.2 min ± 1.7 SE
before sunrise (range, SR — 14 to 24 min).

DISCUSSION

Like the Brown-headed Cowbird, and
sometimes the Shiny Cowbird (Scott 1991),
Bronzed Cowbirds lay prior to sunrise and,
similar to other brood parasites, they lay rap-
idly (Sealy et al. 1995). It is undoubtedly ad-
vantageous for brood parasites to lay their
eggs when hosts are absent and to lay as rap-
idly as possible (reviewed in Sealy et al.
1995). Indeed, Neudorf and Sealy (1994)
found that hosts of the Brown-headed Cow-
bird at Delta Marsh, Manitoba, that did not
roost on their nests overnight typically arrived
at the nests in the morning after cowbird par-
asitism would have occurred. Female Bronzed
Cowbirds have been observed entering host
nests at various times of the day, but it is un-
known whether eggs were laid (Thurber and
Villeda 1980; T. Brush, pers. comm.). While
it is possible that eggs were laid during these
visits, these females may have been inspecting

nests (see below), or they may have punctured
host eggs (Carter 1986, Peer 1998).

The female Bronzed Cowbird observed vis-
iting a nest prior to sunrise without laying
may have been inspecting this nest to deter-
mine whether it was active and ready to be
parasitized (see also Mayfield 1961, Nolan
1978). This nest was parasitized by two
Bronzed Cowbirds the following morning.
The cowbirds were clearly aware of the nest
beforehand because they flew directly to it.
Similar behavior has been reported for
Brown-headed Cowbirds (Hann 1941, Neu-
dorf and Sealy 1994).

The three cowbird species mentioned above
are the only icterids known to lay before sun-
rise. The nonparasitic Bay-winged Cowbird
may also lay prior to sunrise (see Scott 1991).
Direct observations of laying by the Bay-
winged, Giant {Scaphidura oryzivora), and
Screaming (M. rufoaxillaris) cowbirds are re-
quired before it can be concluded that pre-
sunrise laying in cowbirds is an adaptation for
parasitism.

ACKNOWLEDGMENTS

The Welder Wildlife Refuge provided accommoda-
tion and logistical assistance along with M. L. Peer.
We are grateful to K. Stewart who observed two
Bronzed Cowbirds lay in the same nest in one morn-
ing. Constructive comments by T. Brush, D. Burhans
and two anonymous reviewers improved the manu-
script. This research was supported by a research grant
from the Natural Sciences and Engineering Research
Council of Canada to S.G.S. and a G. A. Lubinsky
Memorial Scholarship from the Department of Zool-
ogy, University of Manitoba to B.D.P.

LITERATURE CITED

Chance, E. P. and H. W. Hann. 1942. The European
Cuckoo and the cowbird. Bird-Banding 13:99-
103.

Carter, M. D. 1986. The parasitic behavior of the
Bronzed Cowbird in south Texas. Condor 88:1 1-
25.

Hann, H. W. 1941. The cowbird at the nest. Wilson
Bull. 53:21 1-221.

Mayfield, H. E 1961. Vestiges of a proprietary inter-
est in nests by the Brown-headed Cowbird para-
sitizing the Kirtland's Warbler. Auk 78:162-166.
Neudorf, D. L. and S. G. Sealy. 1994. Sunrise nest
attentiveness in cowbird hosts. Condor 96:162—
169.

Nolan, V., Jr. 1978. The ecology and behavior of the
Prairie Warhler Deiulroica di.scolor. Ornithol.
Monogr. 26:1-595.

SHORT COMMUNICATIONS

139

Peer, B. D. 1998. An experimental investigation of
egg rejection behavior in the grackles (Quiscatus).
Ph.D. diss., Univ. of Manitoba, Winnipeg.

Scott, D. M. 1991. The time of day of egg laying by
the Brown-headed Cowbird and other icterines.
Can. J. Zool. 69:2093-2099.

Sealy, S. G., D. L. Neudorf, and D. P. Hill. 1995.

Rapid laying by Brown-headed Cowbirds Moloih-
riis ater and other parasitic birds. Ibis 137:76-84.

Sturkie, P. D. 1976. Avian physiology, third ed.
Springer- Verlag, New York.

Thurber, W. a. and a. Villeda. 1980. Notes on par-
asitism by Bronzed Cowbirds in El Salvador. Wil-
son Bull. 92:1 12-1 13.

Wilson Bull., 111(1), 1999, pp. 139-143

Temporal Differences in Point Counts of Bottomland Forest Landbirds

Winston Paul Smith and Daniel J. Twedt-

ABSTRACT. — We compared number of avian spe-
cies and individuals in morning and evening point
counts during the breeding season and during winter in
a bottomland hardwood forest in west-central Mississip-
pi, USA. In both seasons, more species and individuals
were recorded during morning counts than during even-
ing counts. We also compared morning and evening de-
tections for 18 species during the breeding season and
9 species during winter. Blue Jay {Cyanocitta cristata).
Mourning Dove (Zenciida macroura), and Red-bellied
Woodpecker (Melanerpes caroliniLs) were detected sig-
nificantly more often in morning counts than in evening
counts during the breeding season. Tufted Titmouse
IBaeoloplms bicolor) was recorded more often in morn-
ing counts than evening counts during the breeding sea-
son and during winter. No species was detected more
often in evening counts. Thus, evening point counts of
birds during either the breeding season or winter will
likely underestimate species richness, overall avian
abundance, and the abundance of some individual spe-
cies in bottomland hardwood forests. Received 15 Nov.
1997. accepted 20 Aug. 1998.

Improvement and standardization of assess-
ment techniques for monitoring bird popula-
tions has received considerable attention (e.g.,
Ralph et al., 1993, 1995a, b; Hamel et al.
1996). Although most studies of avian popu-
lation assessment techniques have focused on
breeding birds, some have evaluated winter

' United States Department of Agriculture, Forest
Service, Pacific Northwest Re.search Station, Forestry
Sciences Laboratory, 2770 Sherwood Lane — Suite 2A,
Juneau, AK 99801-8545; E-mail: wpaulsmith@aol.com
^ USGS Patuxent Wildlife Research Center. 2524
South Frontage Road, Vicksburg, MS 39180.

^ Corresponding author.

bird populations (Rollfinke and Yahner 1990;
Gutzwiller 1991, 1993a, b). Detecting statis-
tically significant changes in avian popula-
tions may require an extensive monitoring
network (Smith et al. 1993, Hamel et al.
1996). To achieve monitoring objectives using
limited resources, protocols that reduce costs
and maximize efficiency are required (Smith
et al. 1993). Unfortunately, many factors that
influence survey efficiency are beyond the
control of investigators. For example, detec-
tion varies among species, among census tech-
niques (e.g., Grue et al. 1981, Rollfinke and
Yahner 1990), and may be influenced by phys-
ical or biological factors (Gutzwiller 1993a,
b).

If detection probabilities were constant over
time, the efficiency of avian surveys could be
increased by providing a greater window of
opportunity during which surveys could be
conducted. However, most species exhibit diel
and seasonal variation in detectability. Thus,
to optimize sampling effort and reduce sam-
pling variances, monitoring should be focused
on periods when species are most frequently
detected (Gutzwiller 1993a).

To assess optimal periods of detection, in-
vestigators have compared point counts from
different times of the morning during the
breeding season (Shields 1977, Grue et al.
1981, Robbins 1981, Skirvin 1981) or winter
(Gutzwiller 1993a). Only Rollfinke and Yah-
ner (1990), using tran,sect counts, compared
morning counts to evening counts during win-
ter. Although birds are generally assumed to

140

THE WILSON BULLETIN • Vol. Ill, No. /, March 1999

be more detectable during morning than dur-
ing evening, we suspected that some species
were equally detectable during both periods.
If true, monitoring efforts that focused only
on these species could greatly expand the time
during which surveys could be conducted. To
evaluate the differences in detection of avian
species between morning and evening, we
conducted paired morning and evening point
counts during the breeding season and during
winter. In this paper, we report the resulting
estimates of avian species richness, overall
abundance, and abundance of selected species.
We also assess the relationship between de-
tection of individual species and the observed
variation between morning and evening
counts.

STUDY SITE AND METHODS

Point counts were conducted on the 1050 ha Delta
Experimental Forest, near Stoneville, Mississippi
(33° 29' N, 90° 55' W). Surrounded largely by agri-
culture, this bottomland hardwood forest was heavily
logged from 1910 to 1920 with additional research and
commercial harvests continuing though the 1960s.
There was no timber harvest on Delta Experimental
Forests between the early 1960s and the time of this
study.

We conducted morning and evening point counts
during the breeding season (30 May— 12 June 1991; 8—
21 May 1992) and winter (4-14 February 1991; 9-29
January 1992) at 25 stations in each of 4 forest stands.
Forest stands were similar in habitat but were subjeet-
ed to different silvicultural management. We generally
followed standardized protocols for conducting point
counts (Ralph et al. 1993, Hamel et al. 1996) but used
4-min sampling periods instead of 5-min and 20-m
hxed radius circular plots instead of 50-m. We reduced
the sampling period based on species detection curves
from preliminary survey data and we restricted the plot
radius to 2()-m because of a concurrent effort to model
habitat using these same data. Points were visited so
that each visit occurred at a different time during the
3-h periods following sunri.se (morning) and preceding
sun.set (evening). Over the two years of this study, we
made a total of 10 morning and 10 evening visits to
each of the 4 stands during the breeding .season. Within
the same time interval, we made a total of 6 morning
and 6 evening visits to each .stand during winter. De-
tection probability (Gutzwiller 1993a) was e.stimated
for each species as the proportion of total point counts
during which the species was detected. During the
breeding season and during winter we eompared the
number of species and individuals detected during
morning and evening visits using a split plot, repeated
measures analysis of variance; each stand (an experi-
mental design block) was split into morning and even-
ing treatment periods with visits (dates) constituting

the repeated measure. All statistical analysis were per-
formed using the SAS System for Windows (Release
6.11, SAS Institute, Inc., Cary, NC, USA). We sub-
sequently compared the abundance of selected individ-
ual species between morning and evening counts using
the same experimental design. However, individual
species abundances were compared only if the overall
variability of the species allowed detection of at least
0.25 individuals when the power of the test (1 — (3)
was at least 0.80 with a = 0.10 (Hamel et al. 1996).
Furthermore, because we conducted multiple tests
when comparing individual species, we used Bonfer-
roni’s correction which reduced the probability re-
quired for signihcance of these tests to a < 0.006.

RESULTS

We recorded 57 forest landbird species dur-
ing the breeding season and 36 species during
the winter. More species (F, 3 = 383.35, P <
0.01) and total individuals (F, 3 = 597.38, P

< 0.01) were detected in morning counts (x
± SE; 10.05 ± 0.06 species, 11.66 ± 0.08
individuals) than in evening counts (.v ± SE;
7.77 ± 0.07 species, 8.46 ± 0.09 individuals)
during the breeding season. During winter, we
again detected more species (F, 3 = 82.38, P

< 0.01) and total individuals (F, 3 = 26.59, P
= 0.01) in morning counts (T ± SE; 6.12 ±
0.07 species; 9.36 ± 0.14 individuals) than in
evening counts (.v ± SE; 4.44 ± 0.08 species,
6.45 ±0.13 individuals).

During the breeding season, 16 of 57 spe-
cies met our criteria for comparing morning
and evening counts of individual species (Ta-
ble 1). We detected significantly (F, 3 > 46.83,
P < 0.006) more individuals during morning
counts than during evening counts for four
species: Blue Jay {Cyanocitta cristata). Tufted
Titmouse (Baeolophus bicolor). Mourning
Dove (Zenaida macroiira), and Red-bellied
Woodpecker (Melcinerpes carolinus). No sig-
nificant differences (P > 0.01) were detected
between morning and evening counts for the
other 12 species (Table 1). Of the 11 species
eligible for comparison during winter, only
Tufted Titmouse was detected significantly
(F,.3 = 50.6, P < 0.006) more during morning
than during evening. As with the breeding
season, the detection of the remaining species
did not differ significantly (P > 0.01) between
morning and evening counts (Table 1).

Detection probability of individual species
ranged from less than 0.01 to 0.70 during win-
ter and from less than 0.01 to 0.82 during the
breeding sea.son. There was a significant cor-

TABLE 1. Abundance {.x ± SE) and detection probability of individual species from morning and evening point counts conducted during the breeding
and during winter of 1991 and 1992 on Delta Experimental Forest, Stoneville, Mississippi, USA.

SHORT COMMUNICATIONS

(41

2,g

=» ^1

c >>
.2 ~
u ^
2 2
w o

•-H

•o

± SE

+ 1 1

d
! +1

o o

+ 1 O +1

+1 o

»r)

1— H

0-)

•n

C 'H

d o

Tj-

C^l

’t

+ 1

(^1

O')

+ 1

■3-

-i

•n ON 00 CN ^

r- — ' r- NO o CN

o r- o

d d d d d d

OONOOOvO'^CNmOO'O »riOONOO
OinCNsOOfNONOON

dddddddddoddddo

1—1

1-H

,—1

+1 ^1

+1 ■“!

+1 ■°|

+ !

o +1

+1 o

f-H

C^l

"=t

r-*

— *

^1?

5 S

'>5 ^

3 o
S °

^ u
Q 1

o rj “
O D u
Oh Q, <U

■o -a
° o ^

III

_ o
T3 -a o

^ ^ ^
C3 ^

(U (U

-C X)

p (U

-S X
U a

w o
u, ^

^ E

“O
r- (U

^ -S
-S Q
o 2 .S

■SI ^

TJ "6 ^
dj <U O

a: ai Q

o ^

z o

c: 1)

U -c
- u

« c

§■ -2

^I'-S
u b :c’

c/5 ^ w

2^5

H g fS

(U o
o S

— 4 Cd P

CQ U H U >

o S3 ^

2 ^ ^

Q 5 S
^

Oc ^ ^
K §

5 o "5
"3^0
3 S S!

- o S

.3 Ih O
XI ^ JJ

o ^ .3
(i X i>

o V ^

'C (U (U

E e o
a!

3 ^

S ^

S 5i

C -3
o o

u w
? ^
I ■£

■u ^
w

a. >,

I «

S o
? o

O 4=

5 2

a s;
ac -5

3 S

(D C

ID -j-

c h

'■' J

a N

' -o
u.
c3

Q 3

Q, a
oo

00 “O

c S

c/^ Z

r- '•'*

E O

^ u
QQ X

00 ^
••g X
E ^

a ^

141

Morning and evening counts not compared.

142

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

relation between detection probability and
variation in abundance during the breeding
season (r^ = 0.74, P < 0.01) but not during
winter (r^ = 0.32, P > 0.05). Six of 1 1 species
found at our study site throughout the year.
Red-bellied Woodpecker, Downy Woodpecker
{Picoides pubescens). Blue Jay, Tufted Tit-
mouse, Carolina Wren {Thryothorus ludovi-
cicmus), and Northern Cardinal {Cardinalis
cardinalis), had greater detection probabilities
iP < 0.05) during summer (Table 1). Con-
versely, Red-headed Woodpecker {Melaner-
pes erythrocephalus). Northern Flicker (Co-
laptes auratiis) and American Robin (Turdus
migratorius) had greater detection probabili-
ties during winter.

DISCUSSION

During the breeding season, morning point
counts yielded more species and more indi-
viduals than did evening counts. Furthermore,
when significant differences existed for indi-
vidual species, morning counts were consis-
tently higher than evening counts. Forest land-
birds have long been presumed to be more
detectable during early morning than at other
times of the day and many observers restrict
breeding bird censuses to morning hours
(Skirvin 1981). In studies of diel variation
(Shields 1977, Skirvin 1981), more species
and individuals were detected during the ini-
tial 2 h after sunrise than at other times. In
floodplain forests, Robbins (1981) found that,
although the total number of birds recorded
diminished beyond 2 h after sunrise, the num-
ber of species detected remained nearly uni-
form for up to 5 h after sunrise. Although
there have been few comparisons of early
morning and late evening censuses, Grue and
coworkers (1981), working in desert habitats
during the breeding season, found more spe-
cies and individuals during morning point
counts than during evening counts. Our data
provide further empirical evidence to support
presumed temporal differences in avian detec-
tions between morning and evening counts
during the breeding season.

During winter, we also detected more spe-
cies and individuals on morning point counts
than on evening counts. Rollfinke and Yahner
(1990) also reported more species and more
individuals on early morning transects than on
evening tran.sects during winter. Although we

detected only Tufted Titmouse significantly
more on morning counts than on evening
counts during winter, Gutzwiller (1993a)
found that five species had higher detection
probabilities on point counts between 07:00
and 13:45 than at other times of the day. Over-
all, our data suggest that evening point counts
during either the breeding season or winter
will probably underestimate species richness,
overall avian abundance, and the abundance
of at least some species in bottomland hard-
wood forests.

ACKNOWLEDGMENTS

These data were collected while W. P. Smith was a
biologist with the Southern Hardwoods Laboratory,
USDA Forest Service, Southern Research Station,
Stoneville, Mississippi. J. H. McGuiness and T D.
McCarthey provided valuable assistance with point
counts on Delta Experimental Forest.

LITERATURE CITED

Grue, C. E., R. P. Balda, and C. D. Johnson. 1981.
Diurnal activity patterns and population estimates
of breeding birds within a disturbed and undis-
turbed desert scrub community. Stud. Avian Biol.
6:287-291.

Gutzwiller, K. J. 1991. Estimating winter species
richness with unlimited-distance point counts.
Auk 108:853-862.

Gutzwiller, K. J. 1993a. Refining the use of point
counts for winter studies of individual species.
Wilson Bull. 105:612—627.

Gutzwiller, K. J. 1993b. Avian responses to observer
clothing: caveats from winter point counts. Wilson
Bull. 105:628-636.

Hamel, P. B., W. P. Smith, D. J. Twedt, J. R. Woehr,
E. Morris, R. H. Hamilton, and R. J. Cooper.
1996. A land manager’s guide to point counts of
birds in the Southeast. USDA For. Ser. Gen. Tech.
Rep. SO- 120: 1-39.

Ralph, C. J., G. R. Guepel, P. Pyle, T. E. Martin,
AND D. E Desante. 1993. Handbook of field
methods for monitoring landbirds. USDA For. Ser.
Gen. Tech. Rep. PSW-GTR-144, Pacific South-
west Res. Stn., Albany, California.

Ralph, C. J., J. R. Sauer, and S. Droege. (Eds.).
1995a. Monitoring bird populations by point
counts. USDA For. Ser. Gen. Tech. Rep. PSW-
GTR-149, Pacific Southwest Res. .Stn., Albany,
California.

Ralph, C. J., S. Droege, and J. R. Sauer. 1995b.
Managing and monitoring birds using point
counts: .standards and applications. Pp. 161-168
in Monitoring bird populations by point counts (C.
J. Ralph, J. R. Sauer, and S. Droege, Ed.). USDA
For. Ser. Gen. Tech. Rep. PSW-GTR-149, Pacific
Southwest Res. Stn., Albany, California.

SHORT COMMUNICATIONS

143

Robbins, C. S. 1981. EITcct of nine of day on bird
activity. Stud. Avian Biol. 6:273-286.

Rollfinke, F. B. and R. H. Yahner. 1990. Effects of
time ot day and season on winter bird counts.
Condor 92:213-219.

Shields, W. M. 1977. The effect of time of day on
avian census results. Auk 94:380-383.

Skirvin, a. a. 1981. Effect of time of day and time
of season on the number of observations and den-

sity estimates of breeding birds. Stud. Avian Biol.
6:271-274.

Smith, W. R, D. J. Twedi, D. A. Wiedenfeld, P. B.
Hamel, R. P. Ford, and R. J. Cooper. 1993. Point
counts ot birds in bottomland hardwood forests of
the Mississippi Alluvial Valley: duration, mini-
mum sample size, and points versus visits. USDA
For. Sen Res. Paper SO-274, Southern For. Exp.
Stn., New Orleans, Louisiana.

Wilson Bull., 111(1), 1999, pp. 144-156

Ornithological Literature

Edited by William E. Davis, Jr.

THE BIRD COLLECTORS. By Barbara
and Richard Meams. Academic Press, San Di-
ego and London. 1998: xviii + 472 pages,
many unnumbered figures, 4 maps of expe-
dition itineraries. $49.95 (cloth). — This is the
third book by the Mearnses in 10 years. The
first two consisted of biographies of persons
after whom birds have been named; first, for
the Western Palearctic (“Biographies for bird
watchers”, 1988, reviewed in Wilson Bulletin
101:658-659), and then for North America
(“Audubon to Xantus”, 1992, reviewed in
Wilson Bulletin 105:701-702). Richard is a
countryside ranger in Scotland; Barbara is a
professional occupational therapist, but in re-
cent years has spent more and more time on
her avocation, biohistorical research.

Their new book, although biographical in
large part (and thus overlapping slightly with
the first two books), has quite a different bal-
ance. The Mearnses are well aware that the
collecting of bird specimens in the late 20th
Century is subject to much debate, often ac-
rimonious. The arguments and the social and
political pressures of the current anti-collect-
ing camp are phenomena seldom if ever faced
by the historical collectors whose exploits
form most of the book’s subject matter. The
Mearnses make their position plain in a brief
preface; neither of them has ever deliberately
killed a bird (they salvage accidentally killed
birds for the Royal Museum of Scotland).
They point out, however, that anyone who, in
their words, turns “pale at the mere thought
of killing birds” must realize that “anyone
who drives a car, uses products of the petro-
chemical industries, owns a cat, has glass in
the windows of their home, buys paper, or
consumes electricity will be responsible for
killing birds.” This subject is dealt with at
greater length in their Chapter 17, “The im-
portance of old and new bird collections,”
which relies heavily on the important paper
by Rem.sen (1995).

The first three chapters bounce around a
good bit, dealing with reasons for killing birds
other than for museum collections, bird books

and journals, human casualties, labeling and
note-taking, problems in the field, in shipment
of specimens, and in the museum, etc. By the
fourth chapter, the emphasis becomes primar-
ily historical, reviewing collecting and collec-
tors chronologically. But the arrangement is
not strictly chronological as a whole; chapters
on specific kinds of collectors are internally
chronological. Examples include “Bird artists
as collectors,” “Government-sponsored col-
lecting,” “Army officers,” “The medical pro-
fession,” and “Clergymen and missionaries.”
A chapter entitled “The great accumulators”
treats the owners of large private collections,
beginning with the 13th Earl of Derby (1775—
1851) and ending with the notorious Colonel
Richard Meinertzhagen (see Knox 1993). A
chapter on “The professional field collectors”
is a hodgepodge, including scientists such as
Alfred Russell Wallace, whose collections
were indeed sold, and contract collectors
working for museums, such as Rollo Beck.
This chapter is much too short, as major por-
tions of the holdings of several of the large
museums were made by collectors under con-
tract; in South America, for example, some of
the most prolific were Samuel Klages, the
Steinbachs (father and son), M. A. Carriker,
Jr., and more recently William H. Partridge.
Of these, only Carriker is mentioned, thrice;
half of a sentence in the “Professional field
collectors” chapter and half of a paragraph on
his two wives in a chapter on “Women in the
field”!

The Mearnses admit in their Preface that
their “approach has been rather Anglo-cen-
tric,” after which they list by name 17 “great
collectors [who] have been omitted or not
mentioned in detail” — the third reference to
Carriker is the presence of his name in this
list. The relative neglect of some parts of the
world is obvious all through the book, and I’m
not sure this can be wholly excused by their
admission quoted above. They appear to have
been obsessed by the history of collecting in
central Asia, as this is the subject of a 27-page
chapter called “Terra Incognita”, in which IV2

144

ORNITHOLOGICAL LITERATURE

145

pages, a map and three portraits are devoted
to the exploits of General Nicholas M. Prjev-
alsky (of Prjevalsky’s Horse), and slightly
more text plus a map and a portrait to those
of Armand David (of Pere David’s Deer).
Surely most of the South American continent
was at least as “incognita”.

In the “Professional field collectors” chap-
ter, Lord Rothschild is quoted as having stated
that William Doherty (1857-1901) was “un-
questionably the best collector for the last fifty
years.” To anybody who has seen bird skins
from Doherty's expeditions, this statement is
incomprehensible; he never learned to skin
birds for himself, turning them over to Indian
servants, according to the Mearnses. Carnegie
Museum of Natural History has a collection
Doherty made in Kenya six months before his
death in 1901 (Holland 1905). The skins are
mediocre and the original labels bear nothing
but a pencilled sex mark; the data (“10 miles
W of Mombasa, September-October 1900”)
were apparently supplied by Doherty to Dr.
Holland, Director of the museum, who had
purchased the collection. All of this suggests
that the Mearnses were probably correct in
suggesting that Rothschild’s high praise of
Doherty may have been based on his “servic-
es to entomology.”

A chapter on “Women in the field” con-
tains 9 short biographies. Of the women thus
honored, only 3 [Emilie Snethlage, Elizabeth
Kozlova, and Beryl P. (Pat) Hall] contributed
to ornithology primarily through their collect-
ing activities, at a level comparable to that of
most of the males featured in the rest of the
book. Here is where we find the Mearnses’
most appalling omission; the late Maria Koep-
cke (1924-1971), whose name appears as the
only woman in the Preface list of collectors
who “deserve more space than we could give
them”. Maria Koepcke may justifiably be said
to be one of the true pioneers in the 20th Cen-
tury study of the ornithology of Peru, a small
country whose avifauna numbers about IVS
times that of the entire Palearctic. She and her
husband Hans-Wilhelm, trained as a hydrobi-
ologist, founded Casa Humboldt (^^Humboldt
House) in Lima in 1957; this became the con-
venient base for many expeditions throughout
Peru involving scientists from several nations.
Maria conducted avifaunal surveys in areas
ranging from the desert coast to the Amazo-

nian rain forest. Her bibliography lists 29 ti-
tles, plus 12 co-authored with her husband.
She described 3 new species and 1 3 new sub-
species.

An Appendix lists, in sequence of size, the
world’s 69 largest collections of bird speci-
mens, together with the most significant com-
ponents included therein. Unfortunately there
are many errors and omissions in this list,
partly because for some museums the authors’
information was a quarter-century out of date,
based on Banks and coworkers (1973). It is a
pity that they felt no need to enter into cor-
respondence to get more recent figures. Attri-
bution of components is irregular; parts of the
dismantled collections of the Cleveland Mu-
seum of Natural History, for example, are list-
ed for the Field Museum; the Museum of Zo-
ology, University of Michigan; and the Pea-
body Museum, Yale University, but not the
Carnegie Museum of Natural History, which
may have the largest number of former Cleve-
land birds. No components are listed for the
Delaware Museum of Natural History, which
holds one of the most important U.S. collec-
tions of Philippine birds, and much of the for-
mer collections of John E. duPont, George M.
Sutton, and Allan R. Phillips. Numerous other
omissions could be mentioned, such that this
Appendix is not as valuable as it could have
been.

Any heterogeneous work of this sort is
bound to induce comments and corrections. A
few are listed below:

p. 13. The authors mention the extinction of
the endemic flowerpecker {Dicaeum quadri-
color) of Cebu owing to the deforestation of
that Philippine island; in fact, no fewer than
9 endemic forms of Cebu were thought to
have been extirpated (Rabor 1959), although
small remnant populations of a few of these
have subsequently been found,
p. 111. The authors refer to “remote, little-
known islands such as Whitsunday Atoll,
Clarion Island and the Revillagigedo group.”
Clarion is, in fact, one of the islands in the
Revillagigedo group.

p. 149. For a more complete account of the
fate of the Gould Australian collection, see
Meyer de Schauensee 1957.
p. 182. The Ruwenzori mountains are on the
western, not the eastern border of Uganda,
p. 244. British writers should know that the

146

THE WILSON BULLETIN • Vnl. III. No. I. March 1999

authorship of the classical Handbook of Brit-
ish Birds was in the sequence Witherby, Ti-
cehurst, Jourdain and Tucker; the Mearnses
list Jourdain last.

p. 270. A comment would have been appro-
priate after the statement about Prjevalsky’s
first expedition; “there would have been more
[bird specimens] but most were moulting so
nine-tenths of all of the birds shot were dis-
carded.” This atrocious practice has, of course,
proved to be the bane of students of molts and
plumages; I encountered the same kind of
statement in connection with some 19th Cen-
tury British collectors of Philippine birds,
p. 366. Brina Kessel and 1 have been friends
and colleagues for fifty years, but I can’t un-
derstand the rationale for including her in a
book about collectors. The Mearnses mention
that “her first specimen-based research was
for her Ph.D. dissertation on European Star-
lings and involved the preparation of over 500
skins.” True enough, except that the majority
of the starlings were collected and prepared
by Robert W. Dickerman.

1 need hardly say that any reader of this
book will find, as 1 did, that the descriptions
of the exploits of many collectors previously
known to us as little more than names at-
tached to bird species have been brought to
vivid life by the Mearnses. Their book is high-
ly readable, and their attitude toward collect-
ing as fair-minded as one might ask of writers
not directly involved themselves in collecting.

I am indebted to Manuel Plenge for sending
me biographical materials on Maria Koep-
cke.— KENNETH C. PARKES.

LITERATURE CITED

Banks, R. C., M. H. Clench, and J. C. Barlow. 1973.
Bird collections in the United States and Canada.
Auk 90:136-170.

Holland, W. J. 1905. A list of the birds collected near
Mombasa. East Africa, by William Doherty. Ann.
Carnegie Mus. 3:453—463.

Kno.x, a. G. 1993. Richard Mcinert/.hagen — a ca.se of
fraud examined. Ibis 135:320-325.

Mi:yer de Schauensee, R. 1957. On some avian types,
principally Gould's, in the collection of the Acade-
my. Proc. Acad. Nat. Sci. Philadelphia 109:123-246.
Rabor, D. S. 1959. The impact of deforestation on
birds of Cebu. Philippines, with new records for
that island. Auk 76:37-43.

Rem.sen .1. V., Jr. 1995. The importance of continued
collecting of bird specimens to ornithology and
bird conservation. Bird Conserv. Int. 5:145-180.

LIFE OF THE FLYCATCHER. By Alex-
ander E Skutch. Illustrated by Dana Gardner.
University of Oklahoma Press, Norman,
Oklahoma. 1997: xiii + 162 pp., 16 color
plates, 32 black-and-white drawings, 4 tables,
bibliography, index. $40.00. ISBN 0-8061-
2919-0. — Alexander Skutch brings forth a
popular review of the Tyrannidae in Life of
the Flycatcher. This volume continues
Skutch’s Life of. . . series, each reviewing var-
iation in life history within a bird family (e.g.,
Tanager, Pigeon, Woodpecker, Hummingbird).
The Tyrannidae, subject of this current vol-
ume, is a large family (380 species) of subos-
cines limited to the New World. The text is
successful in showing the great degree of di-
versity displayed by this family. If your
knowledge of New World flycatchers is lim-
ited to North American species, the diversity
described for the family in Life of the Fly-
catcher will amaze you. The text may offer
inspiration for evolutionary inquires and di-
rections for further study.

The text begins with an overview of the
family, and continues with chapters on food,
daily life, song, courtship, nest, eggs, young
and breeding success. Skutch’s personal ex-
perience with this group during his long career
in the American tropics is evident. The book
ends with 138 references organized by chap-
ters and an index. The bibliography provides
some direction to seeking further information.

Skutch takes some exception to the illogic
often apparent in birds’ names and uses “fly-
catcher” in common names rather than vari-
ous forms of “tyrant” (as in “marsh-tyrant”
or “tyrannulet”); this practice may delay
tracking down specific species in other texts.
The only error of some note I found was in
the index: Alder Flycatcher was identified as
"" Empidonax minimum." — PETER E.
LOWTHER.

A NEOTROPICAL COMPANION: AN
INTRODUCTION TO THE ANIMALS,
PLANTS, AND ECOSYSTEMS OF THE
NEW WORLD TROPICS. Second edition, re-
vised and expanded. By John Kricher. Prince-
ton University Press, Princeton, New Jersey.
1997: 451 pp., 177 color photographs, 86 line
drawings. $29.95 (cloth). — This sturdy vol-

ORNITHOLOGICAL LITERATURE

147

ume accompanied me for 10,000 miles this
spring, enlightening me during airport vigils,
on long flights, and during hot nights in Gua-
temala and elsewhere. Except for a warped
cover it is still in excellent condition. Despite
many prior trips to the American tropics, I
found this book fascinating reading as well as
a valuable reference to other literature sourc-
es.

This is a series of authoritative essays on
tropical ecology, organized in 14 chapters and
followed by an Appendix, a list of Acronyms,
a Glossary, 35 pages of References, and an
Index that includes every species, genus, and
topic mentioned. The Appendix, “And, Hey,
Let’s be Careful Out There,” should catch the
reader’s eye before (s)he leaves home. If you
are a tropical bookworm you will appreciate
the miniabstracts of his 40 book-length ref-
erences.

Chapters are arranged in a logical sequence,
beginning with textbook descriptions of cli-
mates and ecosystems, and rainforest struc-
ture, diversity, and function. Chapters on evo-
lutionary patterns and coevolution/ecology of
fruit track relationships among animals and
plants from the research of Darwin to inves-
tigators of the mid 1990s. After brief chapters
on the Neotropics as a pharmacy and on living
off the land, are chapters on ecosystems (riv-
ers, mountains, savannas, dry forests, man-
groves, coral reefs) and on Neotropical birds
and mammals, culminating with a lively dis-
cussion on deforestation and biodiversity. A
strong conservation theme teases the reader to
take appropriate action.

Professor Kricher writes in an easy conver-
sational style, tempting the student, the re-
searcher, or the vacationer to read on and on.
His wide field experience in Central and South
America, his extensive knowledge of the trop-
ical literature, his long academic career, and
his gift for writing combine to make this pub-
lication a gem for the tropical explorer. State-
ments in the text are supported by close to a
thousand references to the scientific literature.
Terms defined in the glossary are italicized in
the text. Dozens of delightful line drawings of
birds, mammals, and lesser life forms from the
pen of Ted Davis grace the pages. A new fea-
ture in this edition is a collection of 177 color
photos, all cross-referenced from the text.

I was shocked to read in the conversion ta-

ble (p. xviii) that 1 square mile = 2,590
square km and (p. 35) that 1 cm = 2.5 in.
And I regret to report 9 scientific names of
birds were misspelled one or more times, in-
cluding “Beautiogallus” (3 times). These pre-
ventable accidents aside, the book is a gold-
mine of information that will greatly enrich
one’s tropical experience.

The Neotropical Companion should be in
every high school and college library, in the
travel section of your public library, and in
the carryon luggage of all students and birders
bound for the American tropics. — CHAND-
LER S. ROBBINS.

THE BIRDS OF ST. LUCIA, WEST IN-
DIES. By Allan R. Keith. British Ornitholo-
gists’ Union, c/o The Natural History Muse-
um, Tring, Herts HP23 6AP, United Kingdom.
1997; 176 pp., 40 color plates with captions,
7 text figures, 3 tables, 7 appendices, 14 £
(cloth). — In recent years, the British Ornithol-
ogists’ Union has done the ornithological
community a tremendous service by sponsor-
ing the production of a variety of check-list
style books specializing on unusual or exotic
regions of the world. Compared to Gambia,
Nigeria, Cyprus, The Philippines, or even the
Southern Bahamas, St. Lucia is a tiny area
that supports a fairly small avifauna. Never-
theless, this is a splendid book that provides
both the detailed observational information
one expects from a check-list plus a vast
amount of other information that shows us
why St. Lucia is an important place in the
ornithological world.

The nitty-gritty of any check-list lies in its
species accounts, which in this case cover 162
species that currently occur on St. Lucia or
have reliable records from the past. The author
does these accounts in a spartan 60 pages. To
this can be added short appendices covering
species of uncertain occurrence, where mu-
seum specimens from St. Lucia reside, recov-
eries of banded birds in St. Lucia, a gazetteer,
and the origins of resident breeding birds. A
major appendix ( 1 8 pages) provides more in-
formation on the 3 endemic species and 3 en-
demic sub-species found on the island. The
author seems to have left no leaf, library, or

148

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

museum unturned in gathering all the details
available about St. Lucian birds.

Were the above material all of the book, I
would readily recommend it for anyone plan-
ning a visit to the island, but 1 would question
its value to the general reader. Allan Keith’s
introductory material (9 chapters adding up to
60 pages) does an excellent job of introducing
the island on which the bird records were
made. Chapters include political history, ge-
ology and geography, climate, vegetation, his-
tory of ornithological exploration, migration,
breeding, zoogeography and conservation. All
of this information makes the avifauna of St.
Lucia come much more alive as part of one
island within the West Indian region. Keith
also does an outstanding job of pointing out
the unusual role this island has played in West
Indian conservation, as a result of the com-
bined efforts of a local biologist (the late Ga-
briel Charles) and an English vagrant (Paul
Butler) in developing a conservation plan that
was very successful on St. Lucia and has been
exported elsewhere.

This combination of detailed observational
work in the species accounts with the broad
biogeographic and ecological perspective
found in the introduction allows the reader to
not only find out what species occur on St.
Lucia, but to get a feeling about why it is this
number and how the island compares to its
neighbors. The author’s synthesis of current
scientific thought is exemplary (although a
few sections may need to be revised when the
current work being done on molecular genet-
ics of West Indian birds is finally released).
This is a great way to both find out what is
on St. Lu^ia and to get a bit of a feeling about
why this is so. — JOHN FAABORG.

MUNIAS AND MANNIKINS. By Robin
Restall, illus. by the author. Yale University
Press, New Haven, Connecticut. 1997: 264
pp., 16 colored plates with captions, 64 color
plates with measured drawings, innumerable
text figures. $60.00 — Robin Restall is a life-
long enthusiast of caged finches, whose em-
ployment with an international advertising
agency has taken him to most parts of the
world where he could learn about many of the
species of finches in their native contexts. His

particular interest is with the genus Lonchura,
which now includes all of the birds tradition-
ally referred to as munias, mannikins, and the
Java and Timor Sparrows which were former-
ly considered in the genus Padda.

This is a masterful collection of virtually
every thing known about this assemblage of
estrildine finches. The book starts off with a
review of the taxonomy and relationships of
the genus Lonchura, continues with an over-
view of their natural history, and concludes
with a detailed accounting of all pertinent in-
formation about each individual species (in-
cluding one apparently new, undescribed spe-
cies).

Each species account includes headings for
field characters, status, habitat, habits and be-
haviors, food and feeding, movements, call,
song, courtship and display, breeding, distri-
bution, description (including all known sub-
specific variants), hybrids, conservation, and
a list of relevant references. There are color
plates for each species, showing all of the
most obvious sex, age, and geographic vari-
ants and an even more impressive collection
of color plates of measured drawings that
show the dorsal and ventral views with one
wing extended.

1 found this to be an excellent summary of
information on a group of birds about which
I knew relatively little. Yale University Press
has done a fine job of editing (I found very
few typographic errors), and the color repro-
duction of the plates appears to be first-rate. I
recommend the volume highly. — HERBERT
T. HENDRICKSON.

FOREST PATCHES IN TROPICAL
LANDSCAPES. By John Schelhas and Rus-
sell Greenberg, Eds. Island Press, Washington,
D.C. 1996: 426 pp., maps, tables, black-and-
white figures. $30.00 (paper). — It is obvious
to anyone who studies any aspect of global
tropical ecology that forest fragmentation is
increasing annually. The complex ecological
effects of fragmentation are only now becom-
ing known, an increasing database generated
from the efforts of numerous researchers in
tropical regions around the world. Most trop-
ical ecologists are familiar with the ongoing
study known now as the Biological Dynamics

ORNITHOLOGICAL LITERATURE

149

of Forest Fragments Project, located north of
Manaus, Brazil, but this multi-authored vol-
ume includes data from numerous studies in
other regions (as welt as one chapter dealing
with the BDFFP).

The book is divided into four parts: chang-
ing forests, regional landscapes, human di-
mensions, and management. The introduction,
authored by the editors, provides a concise but
thorough overview of the issues covered in
depth throughout the volume. Birds are the
focus of but one of the 19 chapters, but are
discussed to varying degrees in numerous oth-
er chapters. Most of the chapters deal with
South and Central America though there is
one chapter on Africa, one largely on Indo-
nesia, and one on India. Each chapter is ref-
erenced from the primary literature.

This volume, broad in scope, and excel-
lently edited, is an important resource for
tropical ecologists, particularly those whose
research is focused on biodiversity preserva-
tion and ecologically sound management pol-
icies.—JOHN C. KRICHER.

WHERE TO WATCH BIRDS IN ASIA. By
Nigel Wheatley. Princeton Univ. Press,
Princeton, New Jersey, 1996: 463 pp., 51 line
drawings, 8 figs., 105 maps. $35.00. — This re-
markable book compresses an impressive
amount of birding data from our largest con-
tinent into a relatively compact volume. Asia
(including the island nations of Indonesia and
the Philippines) harbors just under 2,700 bird
species. Considering that various parts of the
continent have been off limits — or inaccessi-
ble— to outsiders for much of our lifetimes, I
was delighted to see how many places have
now been surveyed by keen birders. The au-
thor is to be congratulated on gathering these
scattered data into one volume. The book is
not meant as an encyclopedic reference to all
sites or to all species in a given country and
the author rightly suggests that the informa-
tion may be best used as a first “guiding
light”, a starting point in travel planning. The
book fills this function admirably and by
studying material presented here you can
quickly focus on desirable Asian locations and
the species found there. The format is ap-
pealing and easy to follow, while the text.

composed in a clear 9 point Cheltenham
Light, is remarkably free of typographical er-
rors.

The book is divided into three parts: a gen-
eral introduction, the main text (organized by
countries), and additional suggestions and in-
dexes. The introduction is necessarily brief
but explains the book’s layout and then pro-
vides much useful information relating to
birds and birding in Asia and includes notes
on habitat diversity, bird diversity (at the fam-
ily and species levels), and how Asia com-
pares with other continents (400 species more
than Africa and 400 less than South America).
Conservation is given a separate six paragraph
section where Wheatley is brave enough to
identify human population growth as a major
conservation problem and to call for growth
stabilization, or better yet, “we should aim to
reduce” population growth. Under General
Tips we learn of familiar techniques for good
birding (some Asian birds are very shy) and
of various dangers to travelers. Altitude sick-
ness (in the higher Himalayas and Tibet) is
mentioned but I find the subject treated too
lightly. The recommended one night at 3000
m. (9,843') before ascending to 4500 m.
(14,764') the next night is far too fast for
many people hiking in the Himalayas. Simi-
larly, it is suggested that if one “turns blue
and coughs up pink mucus” one should im-
mediately descend to below 3000 m. (9,843').
Actually one should have descended well be-
fore this grim stage is reached. Keep in mind,
that “descending immediately” from some lo-
cations in Tibet is not possible.

In the main text, forty countries are sum-
marized in varying detail (one paragraph for
the Maldives to 57 pages for Indonesia; North
Asia is not covered). English names follow
those used by James Clements Birds of the
World: A Check List (Fourth edition, 1991 and
Supplements) and is a most useful correlation.
Yet it is a pity that, except for 189 species
given in an appendix, scientific names are not
recorded somewhere in the book.

Each country report follows a standard for-
mat that starts with a general map, followed
by a summary and then short notes on size,
transport within the country, accommodations,
health and safety, climate and timing, habitats,
conservation, special birds, a note on endem-
ics, and how many species one might expect

150

THE WILSON BULLETIN • Vol. HI, No. I, March 1999

to see in a stated time. These sections are
helpful planning tools. Accommodations and
travel within the country are only lightly
touched upon as vast descriptions are avail-
able in various general guides: accommoda-
tions mentioned are usually aimed at the in-
dividual traveling on a limited budget.

Following this general introduction, impor-
tant birding sites are covered in a clear format.
This material starts with an overall note and
sometimes a map (of some 250 sites, 184 are
not mapped) and then moves to a list of en-
demics seen at that particular site, followed by
specialties, others, and finally other wildlife.
Maps are so helpful that I would like to have
seen more. The 51 line drawings (mostly of
excellent quality) sprinkled throughout add vi-
sual appeal but do not enhance the usefulness
of the text. Dropping line drawings to add
maps would make the book more utilitarian
albeit less attractive. Each country section
ends with Additional Information that in-
cludes addresses of local bird clubs and nature
societies, suggested readings, a complete list
of all the country’s endemics, and finally a
paragraph giving near-endemics.

The book nears the end with yet additional
addresses that include various general socie-
ties and clubs dealing with Asia, where one
may obtain trip reports, and of 16 companies
that do birding tours. After a selection of gen-
eral book titles there are three pages of three-
column fine print that give scientific names
that correspond with Clements English names,
and then with other English names used in
Asian bird books (where these differ from the
names used by Clements). The book con-
cludes with two indexes.

With any volume of this magnitude readers
will have varying opinions on the coverage
and the presentation of the material. Perhaps
the most serious omission of the book, to my
mind, is that there is no clue as to the abun-
dance of the species listed. Thus there is no
distinction between a bird that has been seen
once at that site or another that is recorded in
numbers every day. Even a two-tier indication
giving an “r” for very rare species and an
“a” for an abundant species would be helpful.
Similarly, some birds move seasonally and at
times the text gives no idea as to when the
bird might be at the site described. Tickell’s
Leaf Warbler {Phylloscopus affinis), for ex-

ample, is listed for Corbett National Park (In-
dia), but for much of the year this migratory
species is not in the park.

A curious paragraph called “near-endem-
ics” appears at the end of each country sum-
mary. To save space, this section could easily
be dropped or at least the definition of “near-
endemic” tightened. It is hard to see how the
Nepal Fulvetta {Alcippe nipalensis) is a “near-
endemic” when it is listed for Bangladesh,
Bhutan, Burma, Nepal, and northeast India.

As common to many first editions dealing
with this much detail, there are a number of
minor factual errors. For example, the Purple-
rumped Sunbird (India) is listed under “more
or less throughout” while the bird does not
occur in the north. The plural of genus, p.25,
is genera (not genuses). Similarly, Padang en
route to Kirinci-Seblat National Park (Indo-
nesia) is twenty-four hours (not six) by bus
from Berestagi (Brastagi) — twenty-four, that
is, if one is lucky. The Khunjerab Pass (Pak-
istan) is close to 16,000 feet (not 5575 m.,
18,290'). On page 38 we learn that “most of
the pristine forest which is left in the eastern
Himalayas is in . . . Bhutan.” In reality, there
is far more eastern Himalayan forest in Arun-
achal Pradesh than in Bhutan but much of
Arunachal is still off limits to outsiders. Nam-
dapha National Park (India) is accorded “the
greatest altitudinal range of any park in the
world” but the Sagarmatha National Park and
the Makalu-Barun National Park and Conser-
vation Area in Nepal and the Quomolungma
Nature Reserve in Tibet cover substantially
more altitude than does Namdapha.

I fear this book will not sell well in the
Maldives for on page 124 the Maldives dis-
appears as a country only to resurface as a
single paragraph under India. This treatment
would surprise the citizens of this Indian
Ocean nation. Agreed, the Maldives may not
be particularly good for birding, but it has
world class coral reefs and scuba diving and
should be accorded full country status in the
next edition.

These errors, however, are of a minor nature
and do not detract from the importance of this
effort and 1 strongly recommend the book to
anyone who is even remotely thinking about
those nearly 2700 Asian species that are just
waiting to be seen.-ROBERT L. FLEMING,
JR.

ORNITHOLOGICAL LITERATURE

151

THE BIRDS OF SULAWESI. By Derek
Holmes and Karen Phillipps. Oxford Univer-
sity Press, Oxford, U.K. 1997: 86 pp., 20 col-
or plates illustrating 142 species, 22 black-
and-white illustrations, one table. $24.95
(cloth). — The island of Sulawesi, part of the
vast country of Indonesia, has only relatively
recently been frequented by ecotourists and
birders. Sulawesi is perhaps best known for
Torajaland, where the local people practice
elaborate funeral rites and the dead are placed
in cliffside alcoves, commemorated with
unique statues. Sulawesi was of great interest
to Alfred Russel Wallace, as it sits almost
astride “Wallace’s Line,” separating two bio-
geographic realms. The authors of this guide
provide a brief table that compares the avifau-
na of Kalimantan, on one side of Wallace’s
Line, with Sulawesi, on the other side. They
note that of the 380 species found on Sulawesi
and its near neighbor islands (what the authors
call the “Sulawesi region”), 96 are endemic
to the region, and 115 are endemic to Indo-
nesia.

While this guide is admittedly not compre-
hensive, it will prove very useful. The species
treated are those most commonly seen, the il-
lustrations are of good quality, and the text
descriptions are adequate for identification. It
would be helpful if some maps were included
but none are. The authors state that “For the
bird-watcher, Sulawesi is unequalled.” That is
certainly not true. Much of the island is de-
voted to rice farming and birds other than mu-
nias and some herons, are sparse. To get any
real sense of the endemic avifauna one must
visit one or more of the nature reserves and
national parks, where some natural forest sur-
vives. Like much of the rest of Indonesia,
roadside birding is disappointing, with a sur-
prising paucity of birdlife.

The book is indexed and provides a check-
list of resident land birds in the Sulawesi Fau-
nal Region.— JOHN C. KRICHER.

THE WHOOPING CRANE: NORTH
AMERICA’S SYMBOL OF CONSERVA-
TION. By Jerome J. Pratt. Castle Rock Pub-
lishing, Prescott, Arizona. 1996: 171 pp., 46
photographs. $12.95 (paper). — The Whooping
Crane (Grus americcma) is a species that has

been close to extinction throughout my life-
time and has rightly come to symbolize the
conservation movement in North America.
This book is an effort to relate the history of
the species and the efforts made to prevent it
from disappearing completely. The author, Je-
rome Pratt, was one of the charter members
of the Whooping Crane advisory group estab-
lished in 1956, and thus provides an “insid-
er’s” perspective on how the recovery project
has unfolded. This may be the book’s greatest
short-coming. Saving endangered species ap-
pears to have some strong similarities to the
making of sausage; you don’t necessarily want
to know everything that goes into it.

Pratt clearly has some strong opinions on
how the recovery program should have pro-
ceeded, which differ from the way it actually
unfolded. However, I found it difficult in
reading the text to determine precisely what
the points of difference were. Perhaps some-
one with less personal involvement would
have been able to present a more clear ex-
planation of exactly how the differing phi-
losophies interacted in the varying political
environments to determine the decisions that
were made.

I found it troublesome that nowhere in the
text is there a simple graph describing the
numbers of Whooping Cranes extant in the
wild and/or in captivity. Much of the data is
mentioned in the text, but it does not appear
to be complete and it is extremely difficult to
locate. It is clear that there are presently many
more Whooping Cranes than there were in,
say 1954, but exactly how the various man-
agement techniques that have been applied re-
late to the number of birds is not.

There appears to have been only minimal
editing done on the original manuscript and
I found a large number of spelling errors and
other typographic mistakes all over the book.
Most of the time these were merely annoy-
ing, but when numbers were transposed in
years, these mistakes were greatly mislead-
ing.

The strongest feature of this book is its 28-
page bibliography. This contains essentially
all of the relevant literature on Whooping
Cranes and is a must for anyone who wishes
to try writing the definitive history of Whoop-
ing Crane conservation in North America. —
HERBERT T. HENDRICKSON.

152

THE WILSON BULLETIN • Vol. 11], No. 1, March 1999

A PASSION FOR BIRDS. AMERICAN
ORNITHOLOGY AFTER AUDUBON. By
Mark V. Barrow, Jr. Princeton University
Press, Princeton, New Jersey. 1998: 326 pp.,
33 unnumbered text figures. ISBN 0-691-
04402-3. $39.50 (cloth). — Histories of orni-
thology, and especially of American ornithol-
ogy are rare and those by professional histo-
rians of science are rarer still as shown by a
perusal of the bibliography of this excellent
new book by Mark Barrow. Barrow’s A Pas-
sion for Birds does not provide such a com-
plete history of North American ornithology
as this was not his purpose in writing this par-
ticular analysis; ornithologists who are look-
ing for a full history of their field may be
disappointed, but should not be. Rather the
goal of Barrow’s book is to examine the rise
of professionalization in ornithology from the
death of Audubon in 1851 to 1940. It is a most
excellent treatment of this important aspect of
ornithology, one which is well worth the close
attention of everyone interested in this biolog-
ical discipline. A Passion for Birds is a su-
perbly excellent history of science with full
documentation and an exhaustive bibliogra-
phy. Barrow chose to begin his analysis at
1850 because this coincides with Audubon’s
death in 1851 and continues Farber’s analysis
which ended in 1850. Barrow undertook a
most intensive study of archives and the lit-
erature, as demonstrated by his citations, and
presents a tremendous amount of information
on the history of North American ornithology
which is of interest to all ornithologists, pro-
fessional and amateur alike. One does not
have to be a historian of science to enjoy read-
ing A Passion for Birds and to learn much
from it.

Barrow stres.ses three topics in his analysis
of professionalization of North American or-
nithology; these are: (a) collecting and sys-
tematics; (b) the American Ornithologists’
Union founded in New York City in Septem-
ber, 1883; and (c) bird conservation. Quite
clearly, the pathway for most early North
Americans into ornithology, whether they re-
mained amateurs or became professionals, was
via the accumulation of a collection of bird
skins or eggs.

Ornithology during the 19th century was
characterized by inten.se activity amassing col-
lections and describing the diversity of North

American birds, first species and then subspe-
cies, followed by those in the rest of the New
World and finally the Old World. This descrip-
tive work led to two different check-lists of
North American birds, namely by Elliott
Coues and by Robert Ridgway. It was largely
the differences between these check-lists
which led to the founding of the American
Ornithologists’ Union in 1883.

One of the important aspects in the profes-
sionalization of a science, as emphasized by
Barrow, is the founding of a national society
and the publication of a scholarly journal by
that society. This was certainly true in the
course of professionalization of ornithology in
North America even if the initial goals of the
American Ornithologists’ Union were not
quite so noble. Invitations were sent to a small
group of ornithologists with the clear purpose
of establishing a society with the primary goal
of solving the check-list problem and with a
structure which kept control of the society in
the hands of a small group of leading orni-
thologists. The new Union solved splendidly
the classification and nomenclature problems
in short order, publishing the first edition of
its check-list and its code of nomenclature in
1886. Professionalization was further achieved
with the decision to accept the offer to take
over the Bulletin of the Nuttall Ornithological
Club as The Auk.

However, the AOU was less successful in
dealing with other ornithological matters. The
formal hierarchical structure of membership
classes, designed to keep the large masses of
amateur and other ornithologists out of the
running of the Union, caused problems from
the beginning which became more serious as
the decades passed; this archaic system is still
in place as the 20th century draws to a close.
Moreover, the Union did not deal readily with
other ornithological questions such as migra-
tion and bird protection, largely because of the
preoccupation of the leading members with
matters of collecting and systematics.

Barrow discusses in detail the development
of conservation and bird protection, which be-
came a central issue for the Union from its
origin. He showed that although several prom-
inent members of the AOU had central roles
in the development of bird protection in North
America, the Union did little in this area. A
large part of the problem stemmed from con-

ORNITHOLOGICAL LITERATURE

153

flicts between member in favor of collecting
birds for scientific (as opposed to commercial)
purposes and those in favor of bird protection.
Barrow does an excellent job in summiirizing
these arguments and showing their conse-
quences for the Union and for conservation
groups; but he fails to analyze these argu-
ments and therefore does not show that these
two groups were largely arguing past one an-
other. By 1900, it was clear that collecting
birds had nothing to do with the decline and
extinction of avian species. Rather it was clear
that the primary factors were market hunting
and habitat destruction.

The second major aspect, and perhaps the
more difficult one to analyze, in the profes-
sionalization of a science is the balance be-
tween amateurs and professionals. Barrow
states (p. 5) that ornithology is “a classic ex-
ample of an inclusive scientific field” but does
not clarify this concept. He pays most careful
attention to the professional-amateur distinc-
tion in his introduction providing numerous
citations to the literature in this field, but does
not come to a clear resolution because of the
quandaries in defining an amateur (including
whether a single definition would serve equal-
ly well in 1880, 1930, and 1980, and in char-
acterizing particular persons as an amateur or
as a professional. Was Sir William Herschel
(1738-1822) an amateur astronomer when he
discovered the planet Uranus in 1781, a time
when he was still earning his keep as a mu-
sician? Were William Brewster, Elliott Coues,
Margaret M. Nice, or Lord Walter Rothschild
amateurs? Possibly this quandary will never
be settled. Major questions still exist, such as:
How much interest in ornithology does a per-
son have to have to be identified as an amateur
ornithologist? Is every person who maintains
a bird feeding station and possesses a pair of
binoculars and a bird identification guide, an
amateur ornithologist? Perhaps a distinction
should be made between serious amateur or-
nithologists and all others (the “hobby-orni-
thologists” as expressed in German). What
were the roles of amateurs to ornithology
since 1850? Since 1900? Since 1940, the end
of Barrow’s analysis? And is ornithology the
science in which professionals are so outnum-
bered by amateurs (see statement by Frank

Chapman, p. 5)7 1 think not. 1 suspect that
ornithologists have over-emphasized the num-
ber and importance of amateurs in their sci-
ence. Comparative studies are needed. But 1
suspect that there have been and still are more
serious amateur astronomers than serious am-
ateur ornithologists, and that these amateur as-
tronomers have and continue to contribute
more to astronomy than amateur ornitholo-
gists do to ornithology.

Nevertheless, Barrow was able to demon-
strate that by 1940, the termination of his
analysis. North American ornithology had be-
come fully professionalized from its largely
non-professional status in 1850. By 1940, a
good majority of active ornithologists were
professionals based on publishing scholarly
works in avian biology and earning their liv-
ing in a position requiring some involvement
in the study and/or conservation of birds. Fifty
years later, serious amateur ornithologists
have all but disappeared. Clearly profession-
alization in ornithology and other sciences is
dependent on the rise of paid positions in the
field, but this aspect has not been examined
for ornithology in any detail by Barrow or
specifically mentioned in the end notes, pos-
sibly because it is too obvious. Yet, it would
be most interesting to have a detailed analysis
of the employment opportunities for North
Americans in all aspects of avian biology
from teaching and research to conservation
and protection as well as the diversity of pos-
sibilities in government, publishing, and in-
dustry, as well as the changes in these em-
ployment opportunities over the decades.

1 would like to congratulate Mark Barrow
on his excellent analysis of the professionali-
zation of North American ornithology and for
presenting it in the clear and lively style used
in A Passion for Birds. 1 thoroughly enjoyed
reading this book and learned a great deal
about the history of North American orni-
thology and the course of professionalization
of a scientific field. Without any hesitation, 1
can recommend most firmly A Passion for
Birds to everyone with any interest in orni-
thology. 1 would like to thank Keir B. Sterling
for answering urgent questions on the fine
points of American ornithology and for read-
ing the manuscript. — WALTER J. BOCK.

154

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

THE EBCC ATLAS OF EUROPEAN
BREEDING BIRDS: THEIR DISTRIBU-
TION AND ABUNDANCE. W. J. M. Hage-
meijer and M. J. Blair (editors). T. and A. D.
Poyser, London. 1997: cxli + 903 pp.
$88.00. — Twenty-six years in the making, this
book is a monumental work and one of the
three most significant bird books to emerge
from Europe in recent years. Weighing in at
over six pounds (if my bathroom scales are
correct), its 900-plus large format (81/2" X
12W') pages report on the distribution and rel-
ative abundance of birds across the European
continent, ranging from Gibralter north to Ost-
end, and east to Svalbard and Novaya Zemlya
and Franz Josef Land and south again to the
eastern Mediterranean. Significant parts of
Russia and the Ukraine are covered, as is Ma-
deira and the Azores but the Mediterranean
islands administered from North Africa, as
well as Cyprus and the Canary Islands are
omitted. In addition, an important part of the
Western Palearctic, the North African coast, is
also omitted. Within this area some 10,000
plus ornithologists from every European
country obtained presence/absence data on
some 495 species in over 4400 50 X 50 km^,
often with estimates of the order of magnitude
of each population. Coverage is nevertheless
regionally uneven, with large areas of former
USSR countries lacking any data at all and
with Norway, Poland, and parts of four west-
ern countries lacking the quantitative popula-
tion estimates available for most countries.
The area to be covered was only about half
again as large as the conterminous United
States and the number of countries (ca 40, de-
pending on which year one considers) was
about the same as the number of states in the
U.S. but was complicated by the presence of
about 40 different languages and the occa-
sional armed conflict between countries!
These political, language, and cultural differ-
ences meant that a large part of the success of
the project depended on the use of relatively
simple methods that could be adopted and im-
plemented relatively rigorously by observers
of diverse background.

The book is divided into two major sec-
tions, one of 141 (i-cxli) introductory pages,
followed by 903 substantive pages. A Fore-
word by the eminent biogeographer K. H.
Voous, a 3-page Preface, and an 8-page En-

glish language Introduction describe the pur-
pose and structure of the Atlas. Figure 1 is
particularly important in that it documents the
completeness of coverage. I could find no
quantitative figures as to the coverage
achieved, but it looks as if about two-thirds of
the squares received data for at least 75% of
the breeding species expected to be found
there. However, a large chunk of the former
USSR received no data at all and there are
also gaps in Albania and some of the outer
island groups. The remaining squares, partic-
ularly concentrated in eastern Europe, re-
ceived data for fewer than 75% of the breed-
ing species anticipated. What could have been
usefully included here is a political map of the
region: country boundaries are shown but not
identified, leaving readers unfamiliar with the
political geography of Europe to guess which
country is which.

The Atlas is mapped on a Universal Trans-
verse Mercator (UTM) projection, chosen be-
cause it covered the intended area of the Atlas,
was familiar in most European countries, and
was compatible with national map projections
for each country. Each UTM 100 km X 100
km square was subdivided into four smaller
50 X 50 km grid squares to parallel an earlier
botanical Atlas. Because lines of longitude
converge towards the poles, gradual reduction
of the number of 50 X 50 km squares on
northern lines of latitude was necessary. De-
spite this the visual effect on the final Atlas is
very acceptable, with one having to look
closely to find areas where the local density
of dots was not regularly spaced.

Originally intended for 1985-1988, Atlas
field work actually took far longer. Data for
Spain, for example, spanned from 1970 to
1992, those from Finland and from Moldova
spanned 1986—90, those from Georgia were
for 1992, and the few species that were cov-
ered in Azerbaijan came from 1994. For some
countries visitors’ records contributed signifi-
cantly to the information available. A standard
form allowed recording the breeding status of
about 440 species recorded for each square.
Seven classes of information (e.g., distraction
displaying, egg shells found, fledged young,
nests seen) confirmed breeding, another of
seven classes (e.g., pairs observed in suitable
nesting habitat in the breeding season, court-
ship display observed, nest being seen, etc.)

ORNITHOLOGICAL LITERATURE

155

indicates probable breeding, and two catego-
ries (species observed in possible nesting hab-
itat in the breeding season, and seeing males
present in breeding season) indicated possible
breeding. The validity of the data was re-
viewed through a hierarchy of subsequent
checking, ranging from EBCC national and
regional coordinators through species experts
to the authors of the species texts. In addition
to the presence/absence data, the Atlas sought
to include logarithmic population size esti-
mates as semi-quantitative information of
population levels. However, organizers in
some countries (Norway and Poland) refused
entirely to provide such estimates and small
or large parts of several other countries (Ice-
land, France, Italy, Spain) likewise lack such
estimates. Not surprisingly with armed con-
flict there “the project lost contact with Bos-
nia and Serbia, ...” and had to make use of
earlier presence/absence data from that region.
Estimates from border squares between coun-
tries were merged to the higher value.

The main body of the Atlas consists of spe-
cies accounts. Part 1 includes 496 species for
which the mapping data were satisfactory in
quality; a second group of 17 poorly covered
species is covered in a series of briefer ac-
counts. Each species account covers as far as
possible a list of standard topics, including
world distribution, breeding habitat, distribu-
tion and abundance in Europe, recent changes
in status, and migration patterns. The accounts
do not cover breeding biology and conserva-
tion status since these two topics are respec-
tively covered in detail in the standard work
on the region (Cramp et al."s 1977-94 nine-
volume Birds of the Western Palearctic) and
in Tucker and Heath’s (1994) Birds in Eu-
rope— their conservation status. The status of
subspecies is referenced briefly in the species
section. For each species a population size es-
timate is provided as the geometric mean of
the population size in pairs, together with es-
timates of the minimum and maximum pop-
ulation, the year of the estimate, and the pop-
ulation trend in numbers and range size. Pop-
ulation change estimates cover five categories
of decreases or increases of more than 50%,
of 20-50%, and stable (± 20%).

The English introduction is followed by a
translation of the introduction into thirteen
languages, namely Czech, German, Castilian,

French, Finnish, Greek, Hungarian, Italian,
Dutch, Portuguese, Polish, Russian, and
Swedish. Then follow two figures showing the
maximum data quality and the minimum data
quality for each square, for the best coverage
and worst coverage species respectively.
These are, in my view, rather uninformative
figures, given the extreme nature of the out-
liers. A copy of the recording form then pre-
cedes a four page account of the project’s or-
ganization and background. Reading between
the lines, it is evident that two countries, Brit-
ain and the Netherlands, were the major driv-
ers of the project, providing both initial fund-
ing and much in-kind support. Intriguingly, a
significant amount of support came from a
“sponsor a species” campaign in which in-
dividuals, organizations, and commercial
firms undertook to sponsor the cost of analysis
and writing-up species accounts. The 512 spe-
cies required authors from 37 countries, typi-
cally involving people from two countries at
opposite ends of the species range, thus pro-
moting national collaboration and a broad per-
spective on the treatment of the species.

The remainder of the introductory material
includes an 8-page account of the evolution
and history of the European bird fauna by
Jacques Blondel, an acknowledgments sec-
tion, a bibliography of national and major re-
gional bird books for Europe, and a 2-page
introduction to the individual species ac-
counts.

The Atlas maps and accompanying species
accounts form the core of the volume. Each
Atlas map contains insets for six island groups
(Svalbard, Franz Josef Land, Novaya Zemlya,
the Azores, the Madeiran archipelago, and the
Selvagens). However, one is expected to
memorize the identity of each of these insets,
the only key being on page cxl. Within each
map different colored dots distinguish unsur-
veyed squares from absence and mere pres-
ence from estimated abundance, and show
whether breeding was confirmed or probable
or merely possible. Dot size characterizes log-
arithmic abundance (1-9 breeding pairs, 10-
99 pairs, etc., through “more than 100,000
breeding pairs”). For parts of Russia colored
shading indicates extrapolation of presence
from earlier literature was necessary. Most
species accounts in Part 1 also include a graph
and pie chart providing information on pop-

156

THE WILSON BULLETIN • Vol. Ill, No. I, March 1999

ulation size and trends in each of the major
countries, a tally of the total number of Eu-
ropean countries in which the species breeds,
and an estimate of its European population
(excluding Russia and Turkey). Most species
accounts are illustrated by a black and white
drawing of the species by a variety of Euro-
pean artists and each has a list of the species
name in 14 European languages. The accounts
for the 17 irregular or rare breeding species in
Part 2 typically include an illustration and a
short text commenting on its status.

The species accounts are followed by a
short (5-page) account of the conservation sta-
tus of European birds, adapted by Melanie
Heath and Graham Tucker from their 1994
book Birds in Europe — their conservation sta-
tus. A short summary provides an overview
for the European avifauna as a whole. The re-
maining pages are essentially technical mis-
cellania, with a 65-page bibliography and in-
dices of scientific names and indices in the
various languages of the book concluding the
work.

There is an old joke in computer science to
the effect that it pays to be first or third: if
you are first, you get the credit; if you are
third, you get something that actually works.
This volume is undoubtedly in the first cate-
gory. It provides an excellent approximation
to a continent-wide atlas, and one that is in-
finitely superior to the range maps drawn by

guesstimate in earlier regional or European
avifauna and field identification guides. How-
ever, the differing effort afforded by different
countries, the variation in timing of field
work, the differing national perspectives as to
the inclusion of population estimates, and the
major uncertainties about the status of species
within the former USSR, limit the scientific
quality of this work. Nevertheless, the cover-
age of some 4400 50 X 50 km squares by a
network of volunteer observers organized on
essentially a shoestring budget is a stunning
achievement. For a North American audience
this work takes away any excuse for not hav-
ing a Canada/United States/Mexico atlas of
bird distributions, the major gap in our current
spectrum of ornithological resources. For Eu-
ropean biogeographers it creates a tremendous
new database that will undoubtedly fuel in-
novative analyses of large-scale bird distri-
butions on a scale previously impossible. And
for a European Union that is increasingly
moving to becoming a “United States of Eu-
rope” this work provides a remarkable con-
servation resource that should guide pan-Eu-
ropean conservation strategies, at least for
birds, over the next few decades. For orni-
thologists of all nationalities the book is a
beautifully produced volume that will invite
browsing, stimulate comparisons, and provoke
thought for years to come — RAYMOND J.
O’CONNOR.

'Ihis issue of rhe Wilson linllelin was published on I March 1999.

THE WILSON BULLETIN

Editor ROBERT C. REASON

Editorial Board KATHY G. BEAL

Department of Biology

State University of New York

1 College Circle

Geneseo, NY 14454

E-mail: WilsonBull@geneseo.edu

CLAIT E. BRAUN
RICHARD N. CONNER

Review Editor WILLIAM E. DAVIS, JR.

127 East Street

Foxboro, Massachusetts 02035

Editorial Assistants TARA BAIDEME

Index Editor KATHY G. BEAL

JOHN LAMAR
DANTE THOMAS
DORIS WATT

616 Xenia Avenue
Yellow Springs, Ohio 45387

SUGGESTIONS TO AUTHORS

See Wilson Bulletin, 110:152-154, 1998 for more detailed “Instructions to Authors.”
http://www.ummz.lsa.umich.edu/birds/wilsonbull.html
Submit four copies of manuscripts intended for publication in The Wilson Bulletin, 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 (Seventh Edition,
1998) insofar as scientific names of U.S., Canadian, Mexican, Central American, and West Indian birds
are concerned. Abstracts should be brief but quotable. 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 Scientific Style and Format Manual” (AIBS 1994). Photographs for illustra-
tions should have good contrast and be on glossy paper. Submit prints unmounted and provide a brief but
adequate legend for each figure with all captions on a single page. 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 Ornitho-
logical 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.

MEMBERSHIP INQUIRIES

Membership inquiries should be sent to Laurie J. Goodrich, Route 2 Box 301 A, New Ringgold, PA
17960-9445; E-mail: goodrich@haukmountain.org.

CONTENTS

MAJOR PAPERS

ANNUAL SURVIVAL RATES OF FEMALE HOODED MERGANSERS AND WOOD DUCKS IN

SOUTHEASTERN MISSOURI — Katie M. Dugger, Bruce D. Dugger, and Leigh H. Fredrickson 1
COMPARATIVE NEST SITE HABITATS IN SHARP-SHINNED AND COOPER’S HAWKS IN WIS-
CONSIN Dale R. Tre.xel, Robert N. Rosenfield, John Bielefeldt, and Eugene A. Jacobs 1

MADAGASCAR FISH-EAGLE PREY PREFERENCE AND FORAGING SUCCESS

James Berkehnan, James D. Fraser, and Richard T. Watson 1 5

THE RELATIONSHIP BETWEEN SPOTTED OWL DIET AND REPRODUCTIVE SUCCESS IN THE

SAN BERNARDINO MOUNTAINS, CALIFORNIA

Richard B. Smith, M. Zachariah Peery, R. J. Gutierrez, and William S. Lcdiaye 22

FOOD, FORAGING, AND TIMING OF BREEDING OF THE BLACK SWIFT IN CALIFORNIA

— Manuel Marin 30

FACTORS THAT INFLUENCE TRANSLOCATION SUCCESS IN THE RED-COCKADED WOOD-
PECKER Kathleen E. Eranzreb 38

BANDING RETURNS, ARRIVAL PATTERN, AND SITE-FIDELITY OF WHITE-EYED VIREOS

- S. L. Hopp, A. Kirby, and C. A. Boone 46

RESPONSE OF BROWN-HEADED NUTHATCHES TO THINNING OF PINE PLANTATIONS

Michael D. Wilson and Bryan D. Watts 56

DIFFERENCES IN MIGRATORY TIMING AND ENERGETIC CONDITION AMONG SEX/AGE

CLASSES IN MIGRANT RUBY-CROWNED KINGLETS ,

- — David L. Swanson, Eric T. Liknes, and Kurtis L. Dean 61

SCALE-DEPENDENT HABITAT SELECTION BY AMERICAN REDSTARTS IN ASPEN-DOMINAT-
ED FOREST FRAGMENTS Navjot S. Sodhi, Cynthia A. Paszkowski, and Shannon Keehn 70

FEMALE MATE CHOICE IN NORTHERN CARDINALS: IS THERE A PREFERENCE FOR REDDER

MALES? L. Lareesa Wolfenbarger lb

FRUIT SUGAR PREFERENCES OF HOUSE FINCHES '.....

Michael L. Avery, Carrie L. Schreiber, and David G. Decker 84

HIERARCHICAL COMPARISONS OF BREEDING BIRDS IN OLD-GROWTH CONIFER-HARD-

WOOD FOREST ON THE APPALACHIAN PLATEAU J. Christopher Haney 89

EFFECTS OF WIND TURBINES ON UPLAND NESTING BIRDS IN CONSERVATION RESERVE

PROGRAM GRASSLANDS Krecia L. Leddy, Kenneth E. Higgins, and David E. Naugle 100

AVIAN USE OF PURPLE LOOSESTRIFE DOMINATED HABITAT RELATIVE TO OTHER VEGE-
TATION TYPES IN A LAKE HURON WETLAND COMPLEX

— — Michael B. Whitt. Harold H. Prince, and Robert R. Cox, Jr. 105

SHORT COMMUNICATIONS

BALD EAGLE PREDATION ON COMMON LOON CHICK

James D. Paruk, Dean Seanfield, and Tara Mack 1 1 5

TERRITORIAL TAKEOVER IN COMMON LOONS (GAV/A IMMER) James D. Paruk 116

COURTSHIP BEHAVIOR OF THE BUFF-NECKED IBIS (THERISTICUS CAUDATUS)

Nathan H. Rice 1 1 8

HABITAT USE BY MASKED DUCKS ALONG THE GULF COAST OF TEXAS

- James T. Anderson and Thomas C. Tacha 1 19

GIZZARD CONTENTS OF PIPING PLOVER CHICKS IN NORTHERN MICHIGAN

Erance.sca ./. Cuthhert. Brian Scholtens. Uiuren C. Wemnier. and Robxn McLain 121
NESTING OF FOUR POORLY-KNOWN BIRD SPECIES ON THE CARIBBEAN SLOPE OF

COSTA RICA Bruce E. Young and James R. Zook 124

SEXUAL DIMORPHISM IN THE SONG OF SUMICHRAST'S WREN

— Monica Perez-Villafaha, Hector Gomez de Silva G.. and Atahualpa DeSucre-Medrano 128

AN INCIDENT OF FEMALE-FEMALE AGGRESSION IN THE HOUSE WREN ..

- lorn Ahvorth and Isabella B.R. Scheiber 130

NEST REUSE BY WOOD THRUSHES AND ROSE-BREASTED GROSBEAKS

- - Lyle E. Eriesen. Valerie E. Wyatt, and Michael D. Cadman 132

SINGING IN A MATED FEMALE WILSON’S WARBLER

William M. Gilbert and Adele F. Carroll 134

LAYING TIME OF THE BRONZED COWBIRD Brian D. Peer and Spencer G. Sealy 137

TEMPORAL DIFFERENCES IN POINT COUNTS OF BOTTOMLAND FOREST LANDBIRDS

Winston Paul .Smith and Daniel J. Twedt 1 39

144

ORNITHOLOGICAL LITERATURE

JhcWilsonBulktin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. Ill, NO. 2 JUNE 1999 PAGES 157-302

(ISSN (XM3-5643)

c r.

■ * ^

i 0

THE WILSON ORNITHOLOGICAL SOCIETY
FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President Edward H. Burtt, Jr., Department of Biology, Ohio Wesleyan University, Delaware, Ohio
43015; E-mail: EHBurtt@cc.owu.edu.

First Vice-President — John C. Kricher, Biology Department, Wheaton College, Norton, Massachusetts
02766; E-mail: JKricher@wheatonma.edu.

Second Vice-President — William E. Davis, Jr., College of General Studies, 871 Commonwealth Ave.,
Boston University, Boston, Massachusetts 02215; E-mail: WEDavis@bu.edu.

Editor Robert C. Beason, Department of Biology, State University of New York, 1 College Circle,
Geneseo, New York 14454; E-mail: WilsonBull@geneseo.edu.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper Montclair,
New Jersey 07043; E-mail: Smallwood@saturn.montclair.edu.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana 46556;
E-mail: DWatt@saintmarys.edu.

Elected Council Members — Peter C. Frederick and Danny J. Ingold (terms expire 1999), Charles R. Blem
and Sara R. Morris (terms expire 2000), Jonathan L. Atwood and James L. Ingold (terms expire
2001).

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

The Wilson Bulletin is sent to all members not in airears 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 of books, reprints, and orni-
thological magazines from members and friends of the Society. Two members have generously established
a fund for the purchase of new books; members and friends are invited to maintain the fund by regular
contribution. The fund will be administered by the Library Committee. William A. Lunk, University of
Michigan is Chairman of the Committee. The Library currently receives over 200 periodicals as gifts and
in exchange for The Wilson Bulletin. For information on the library and our holdings, see the Society’s
web page at http://www.ummz.lsa.umich.edu/birds/wos.html. 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: Josselyn Van Tyne Memorial Library, Museum of Zoology, The
University of Michigan, 1 109 Geddes Ave., Ann Arbor, MI 48109-1079 USA. Contributions to the New
Book Fund should be sent to the Treasurer.

THE WILSON BULLETIN
(ISSN 0043-5643)

THE WILSON BULLETIN (ISSN 0043-5643) is published quarterly in March. June, September, and December by
the Wilson Ornithological Society, 810 East lOth Street, Lawrence. KS 66044-8897. The subscription price, both in the
United States and elsewhere, is .S40.00 per year. Periodicals postage paid at Lawrence. KS. POSTM.^STER: Send address
changes to THE WILSON BULLETIN, PO. Box 1897, Lawrence. KS 66044-8897.

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.

Subscriptions, changes of address and claims for undelivered copies should be sent to the OSNA. P.O. Box 1897,
Lawrence, KS 66044-8897. Phone: (785) 843-1221; FAX: (785) 843-1274.

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

FRONTISPIECE. Left from top to bottom; Myrmollienila luiemutoiioiu haematonota male, Mvrmoilienila
liaeimitoiiota pyrrhonoUi male, Mynuothenila haematonota pyrrhonota Female, Myrmothenila spodionota spo-
dionota male, Myrmothenila spodionota spodionota female. Right from top to bottom: Mvrmotherula fjeldsaai
male type, Myrmothenila fjeldsaai female, Myrmothenila lencophthalma lencophthalnia male, Mvnnotherida
leiieophthalma lencophthalma female. Water color painting by J. Fjeldsa.

THE WILSON BULLETIN

A QUARTERLY JOURNAL OF ORNITHOLOGY
Published by the Wilson Ornithological Society

VOL. Ill, NO. 2 JUNE 1999 PAGES 157-302

Wilson Bull., 111(2), 1999, pp. 157-165

A NEW SPECIES IN THE MYRMOTHERULA HAEMATONOTA
SUPERSPECIES (AVES; THAMNOPHILIDAE) FROM THE
WESTERN AMAZONIAN LOWLANDS OF ECUADOR AND PERU

NIELS KRABBE,' MORTON L. ISLER,^ PHYLLIS R. ISLER,^

BRET M. WHITNEY, 3 JOSE ALVAREZ A.,^ AND PAUL J. GREENFIELD^

ABSTRACT. — A new species of antwren (Myrmotherula fjeldsaai) closely related to Myrmotherula haema-
tonota is described from the lower tropical zone of eastern Ecuador and immediately adjacent Peru. It primarily
differs from M. h. haematonota by its brown instead of red back in both sexes. New distributional data for
nominate M. h. haematonota shows that it meets the new species north of the Ri'o Marafion, between the Rios
Napo and Pastaza, with no apparently significant physical barrier between them. Received 6 March 1998, ac-
cepted 30 Dec. 1998.

Two species of “stipple-throated” ant-
wrens, Ornate Antwren {Myrmotherula orna-
ta) and White-eyed Antwren {Myrmotherula
leucophthalma) show distinct geographical
variation in back color. In some populations
of each species the back is rufous; in others it
is gray or olive-brown. Separated by large riv-
ers, these populations are not in physical con-
tact and have traditionally been ranked as sub-
species. Here we describe similar variation in
Myrmotherula haematonota, but, because the
two forms are known to be in contact and to

' Zoological Museum, Univ. of Copenhagen, Uni-
versitetsparken 15, DK-2100, Denmark.

- Division of Birds, National Museum of Natural
History, Smithsonian Institution, Washington, D.C.
20560.

^ Museum of Natural Science, 1 19 Foster Hall, Lou-
isiana State Univ., Baton Rouge, Louisiana 70803.

^ Instituto de Investigaciones de la Amazom'a Peru-
ana-IIAP, Avenida A. Quinones km 2.5, Apartado 784,
Iquitos, Peru.

161-162 Calle 6, El Bosque, Quito, Ecuador.

® Present address: Cas. 17-21-791, Quito, Ecuador;
E-mail; NKrabbe@pi.pro.ec
’ Corresponding author.

retain their integrity, we propose to rank them
as species.

Zimmer (1932) defined Myntiotherula hae-
matonota to encompass both the rufous
backed forms that occupy Amazonian low-
lands and the gray backed forms of Andean
foothills. He believed two specimens from the
lowlands of Loreto, Peru, to be intermediate
between these forms. More recently, Hilty and
Brown (1986) and Parker and Remsen (1987)
considered the foothill forms as a distinct spe-
cies, Myrmotherula spodionota (including so-
roria), but did not address the issue of the
apparently intermediate specimens.

METHODS

In 1992 PJG noticed a male specimen with a brown
back in the collection of the Museo Ecuatoriano de
Ciencias Naturales (MECN), labeled as M. h. haema-
tonota, taken at Rfo Bufeo in the lowlands of Pastaza
by R. Olalla on 3 February 1963. This specimen was
referred to Myrmotherula leucophthalma by Ortiz-
Crespo and coworkers ( 1990) and by Ridgely and Tu-
dor (1994).

In 1994 NK tape recorded and collected a male and
a female “stipple-throated” antwren with brown backs
near Pompeya, Napo, Ecuador. These specimens and
the Rfo Bufeo male were compared directly with the

157

158

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

male from the “Mouth of Rfo Curaray” at the Amer-
ican Museum of Natural History (AMNH). The male
is one of two specimens Zimmer (1932) identified as
intermediate between the lowland and foothill forms.
Six additional specimens of the brown backed form
were later collected or located in the Western Foun-
dation of Vertebrate Zoology (WFVZ), Louisiana State
University, Museum of Zoology (LSUMZ), Museum
of Natural History, University of Kansas (MNHUK),
and Museo de Historia Natural de San Marcos
(MUSM). Four records were available only as photo-
graphs. Thus, 12 specimens (4 of them photographic)
were compared: 7 males and 5 females. All males were
compared directly with the Rfo Curaray specimen.
Two females were compared with each other by NK,
three by MLI. Both red backed and brown backed
forms were collected along Rfo Tigre, Loreto, Peru in
1995. An apparently intermediate specimen (not lo-
cated) from an unspecified locality along Rfo Tigre
was described as having the back “dark olive-brown,
some of the feathers of the middle of the upper back
with rufous brown edges” (Hellmayr 1910), suggest-
ing some gene flow between the two forms. This was
the second specimen that Zimmer (1932), without ex-
amining it, had considered intermediate between M. h.
haematonotci and M. h. spodionota.

RESULTS

Specimens of each sex of the brown backed
birds were found to be essentially identical,
suggesting a homogenous population rather
than birds variously intermediate between M.
haematonota and M. spodionota. The grayish
edges of the inner webs of the remiges in the
males (reddish in M. leucophthalma) and the
coloration of the underparts of the female in-
dicate that the brown backed specimens are
more closely related to M. haematonota and
M. spodionota than to M. leucophthalma. An
analysis of vocalizations (unpubl. data) con-
firms this relationship. Nominate M. h. hae-
matonota was found to be parapatric with
brown backed birds in seemingly uniform
habitat (see Fig. 1). Brown backed birds are
uniform in plumage over a large area and thus
clearly represent a valid taxon. They are re-
placed sharply by red-backed birds in similar
habitat, suggesting species rank of the new
taxon, which we propose to name:

Brown-backed Antwren
Myrmotherula fjeldsaai, new species

Holotype. — MECN 6924, adult male ob-
tained by N. Krabbe 16 July 1994 near Rio
Tiputini, 37 road km south-southwest of Pom-
peya, Provincia de Napo, Ecuador; 0° 38' S

76° 26' W, altitude 275 m. Blood sample
(NK 14- 16.7.94) deposited at Zoological Mu-
seum, University of Copenhagen. Vocalization
recordings (LNS 65998) archived at the Li-
brary of Natural Sounds, Cornell Laboratory
of Ornithology.

Diagnosis. — Capitalized names and num-
bers of colors follow Munsell Soil Color Chart
(Kollmorgen Instruments Corp., 1994 edi-
tion). Size, shape, plumage pattern, and col-
oration similar to those of M. haematonota
haematonota, except that the mantle and back
are between Dark Yellowish Brown (10YR3/
4) and Dark Brown (10YR3/3). In M. h. hae-
matonota the back is Dark Red (varying be-
tween 2.5YR4/8 and 2.5YR3/6). The male of
M. fjeldsaai differs from M. h. pyrrhonota by
having brown instead of red back, paler flanks
and tail, and larger and pale buff instead of
pure white spots on tips of median and some
lesser coverts. Female differs from M. h. pyr-
rhonota by having a red back, pale (mostly
white) throat streaked with black, and buffy
brown breast and belly; in M. h. pyrrhonota
the throat is yellow ochre and usually un-
streaked, and the breast is reddish brown (Ta-
ble 1). Male differs from M. spodionota by
having a brown instead of pure gray back;
larger, more buffy, and rounded wing covert
spots, distinctly lighter gray underparts with
darker olive brown sides and flanks, and by
not showing the tendency in many individuals
of M. spodionota to have the white streaks of
the throat that continue onto the breast and
sometimes even the belly. Female differs from
M. spodionota by having a red back and being
considerably paler throughout with whitish,
black-streaked throat and virtually uniform
buffy brown breast and belly. Myrmotherula
spodionota has a somewhat flammulated yel-
low ochre throat, breast and belly (with throat
lightly marked; Table 1). Females differ from
the two known females of M. h. haematonota
from north of the Rfo Maranon by throat color
(Table 1), but this difference falls within the
variation seen in large samples of M. h. hae-
matonota and M. h. amazonica. Myrmotherula
fjeldsaai differs from brown backed forms of
M. leucophthalma by having darker general
coloration and smaller, paler, and decidedly
more rounded wing spots; male has grayish as
opposed to buffy inner webs of remiges, and
usually has smaller throat spots; female has

Krcibhe el cil. • NEW ANTWREN FROM WESTERN AMAZONIA

159

FIG. 1 . Distribution of taxa of the Myrmotherula [haematonota] superspecies in the eastern Ecuador-northern
Peru region. Heavy lines = coast and national boundaries. Dotted line = continental divide. Black stars =
Myrmotherula fjeldsaai. Open circles = M. spodionota (including sororia). Open squares = M. h. haematonota.
Open triangles = M. h. pyrrhonota. Circle surrounding a “U” = species unknown. Identification of locations
discussed in text and type localities (identification of other locations on map available from MLI): 1. Rio Tiputini
(0° 38' S, 76° 26' W), Napo; type locality of M. fjeldsaai. 2. “Sunka 1” (0° 42' S, 75° 51' W), Napo. 3. Tzapino
(ca 01° 1 r S, 77° 44' W), Pastaza. 4. Ri'o Bufeo (ca 02° 12' S, 76° 48' W), Pastaza. 5. Teniente Lopez (ca 02° 32'
S, 76° 14' W), Loreto. 6. San Jacinto (ca 02° 21' S, 75° 43' W), Loreto. 7. Mouth of Rio Curaray (ca 02° 24' S,
74° 04' W, exact location of collecting station uncertain), Loreto. 8. Cocha Hildalgo, left bank Rio Tigre above
mouth of Rio Pacacuro, Loreto. 9. Vicinity of Intuto, right bank Rio Tigre (03° 16' S, 75° 04' W), Loreto. 10.
Rio (Quebrada) Pavayacu, Loreto. 11. Andoas, Loreto. 12. Ri'o Pacacuro, left bank near mouth, Loreto. 13.
Santa Andrea, left bank Ri'o Tigre approximately half way between mouth of Ri'o Pacacuro and Intuto. 14. Nuevo
Manchuria (03° 50' S, 74° 19' W) and Nuevo Tarma (03° 48' S, 74° 21' W), left bank Ri'o Tigre, Loreto. 15.
Puerto Indiana, Loreto. 16. Libertad, Loreto. 17. Chamicuros, Loreto; type locality of M. h. haematonota. 18.
Huampami (200 m), Amazonas. 19. Ri'o Kagka (800 m), Amazonas. 20. Sarayacu, Pastaza; type locality of M.
spodionota. 21. Tigre Playa, Sucumbi'os (MECN 6750 and 6751).

mottled cheeks, whitish-streaked throat and
huffy brown breast, as opposed to uniform
bright buffy yellow cheeks, throat and breast
in M. leucophthalma.

Description of holotype.— Above, including
most of crown, between Dark Yellowish
Brown (10YR3/4) and Dark Brown (10YR3/
3), edge of tail and 1-2 mm wide tips of feath-
ers of back a more reddish Dark Brown
(7.5YR3/3). Wing coverts brownish black
with pale tips, forming three distinct rows of

pale Reddish Yellow (7.5YR8/6) spots. Inner
webs and basal half of outer web of inner
greater secondary coverts Dark Brown
(10YR3/3) to Dark Yellowish Brown (10YR3/
4), inner webs washed with Dark Brown
(7.5YR3/3) on their tips. Primary coverts
blackish brown with minute, barely discern-
ible, Reddish Yellow (7.5YR6/8) tips. Pale
tips of the two alula feathers as large as on
secondary coverts, but lighter, whitish on out-
er web. Median coverts and the largest of the

160

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 1. Diagnosis of plumage differences of females of Myrmotherula fjeldsaai
lations in the M. [haematonota] superspecies.

and neighboring popu-

M. spodionota
« = >10

M. fjeldsaai
ii = 5

M. h. haemainnola from
north of Rio Maranon;
n = 2

M. h. pyrrhonota from
north bank of Rio Napo;

« = 2

Chin & throat

Buffy Yellow to
Yellow Ochre

10 YR8/6 to
10YR6/8

Mostly white;
sides faintly
tinged yellow
brown

Pale Buff

10YR8/2

Buffy Yellow
10YR8/6 to Yel-
low Ochre be-
tween 10YR7/6
and 7.5YR7/6

Streaks on throat

Streaks are mini-
mal, but black-
ish feather bases
show through

Moderate to heavy
blackish streaks
over blackish
feather bases

Usually no streaks,
sometimes a
few fine streaks,
but blackish
feather bases
not apparent

Center of breast to

Yellow Ochre

Buffy Brown to

Reddish-yellow

upper center of

10YR7/6 to

Light Yellowish

7.5YR6/8, suf-

belly

10YR6/8 spots;
lateral feather
edges like sides
giving a spotty
appearance

Brown 10YR7/4
to 10YR6/4 suf-
fused with Light
Olive Brown
2.5Y5/3

fused with Light
Olive Brown
2.5Y5/3

Crown & anterior

Very Dark Grayish

Dark Yellowish

Olive Brown

mantle

Brown 2.5Y3/2

Brown between
10YR4/4 &
10YR4/6

2.5Y4/3

Posterior mantle to
tail coverts

Same as crown

Reddish Brown
2.5YR3/6 to 4/6
to Dark Reddish
Brown

2.5YR4/8

Reddish Brown
2.5YR3/6 to 4/6
to Dark Red-
dish Brown
2.5YR4/8

lesser coverts blackish brown, spots palest on
the distal coverts. The smallest of the lesser
wing coverts mainly Bluish Gray (5PB5/1),
each with a small, black bordered whitish dot
at the tip. Forehead, cheeks, breast, and belly
Bluish Gray (5PB5/1, but slightly lighter),
flanks and vent between Olive Brown (2.5Y4/
4) and Light Olive Brown (2.5Y5/4). Throat
black, feathers with pale shafts and white tips,
forming white, triangular spots pointing an-
teriorly. Inner webs of remiges gray. Twelve
rectrices, tail strongly graduated, with tips of
the three outer rectrices 16, 8 and 4.5 mm
from the tips of the central pair of rectrices.
Body mass 9.8 g. hides grayish brown; bill
blackish with thin gray blue line along cutting
edge; feet gray blue. Skull 100% ossified. No
Bursa Fabricii found. Largest testis 3 X 1 mm.
No fat. Stomach contents: small arthropods
(saved). Netted in tongue of flooded forest in
hill country.

Variation in males. — Specimens from the

Rio Bufeo, the mouth of Rio Curaray, and one
hand held bird photographed near the type lo-
cality in 1994 (C. Canaday photo) were ex-
amined by NK and found to be similar to the
type. The two old specimens were slightly
lighter gray, equivalent to the difference seen
between old and fresh material of M. spo-
dionota. MLI made the same conclusions
when he compared the Rio Curaray specimen
to the Teniente Lopez specimen and to the de-
scription of the Rio Tigre specimen collected
by JA. The backs of the males were slightly
brighter (Dark Yellowish Brown, 10YR3/4)
than the description of the type specimen;
backs of two specimens were more Olive
Brown (between 10YR3/4 and 2.5Y4/4); and
all lacked the reddish brown feather tips de-
scribed for the type specimen.

Additional male specimens examined. —
Myrmotherula fjeldsaai: ECUADOR: Napo,
Rio Bufeo (MECN 2181). PERU: Loreto,
mouth of Rio Curaray (AMNH 255780); Ten-

Krahhe el al. • NEW ANTWREN EROM WESTERN AMAZONIA

161

iente Lopez, Ri'o Conientes (MNHUK uncata-
logued, skull incompletely ossified, collected
July 1993 by Aucca); above Intuto, Rio Tigie
(collected Januaiy 1995 by JA, to be depos-
ited in MUSM). Mynnothenda h. haematon-
ota and M. h. amazonica (Museum abbrevia-
tions aie followed by number of specimens):
PERU: Loreto, Lores, Rio Tigre (collected
January 1995 by JA, to be deposited in
MUSM, 2); Libeitad, S bank Napo (LSUMZ,
2); Puerto Indiana, N bank Amazon (AMNH,
1); Quebrada Vainilla, S bank Amazon
(LSUMZ, 5); Orosa (AMNH, 3); Sai’ayacu
(AMNH, 2); 15 km E Puerto Maldonado
(MNHUK, 1). BOLIVIA: (LSUMZ, 5). BRA-
ZIL: (AMNH, 10). Mynnothenda h. pyrrhon-
ota: ECUADOR: Sucumbios, about 14 km N
Tigre Playa (MECN, 1). PERU: Loreto, N of
Rio Napo, 157 km by river NNE of Iquitos
(LSUMZ, 2); Rio Yanayacu (LSUMZ, 2);
Quebrada Oran (LSUMZ, 2). COLOMBIA:
(AMNH, 1). BRAZIL: (AMNH, 8; USNM,
10). VENEZUELA: (AMNH, 26; USNM, 6).
Mynnothenda spodionota including soroha:
ECUADOR: Napo, San Jose Abajo (AMNH,
5; USNM, 1); above San Jose (ANSP, 1); Avi-
la (ANSP, 1); above Avila (AMNH, 4); Chon-
ta Urcu (ANSP, 1); Morona-Santiago, Cutucii
(AMNH, 1). PERU: (AMNH, 8; LSUMZ, 23).

Female. — A topotypical young female
(MECN 6925) with 2% ossified skull is sim-
ilar to the male above, but slightly lighter and
yellower (between 10YR4/4 and 10YR4/6),
with tail feathers having Strong Brown
(7.5YR4/6) lateral edges. Sides of head like
crown, but with ill-defined buff mottling. Be-
low Light Yellowish Brown (between 2.5Y6/
4 and 10YR6/4), somewhat browner on sides
and flanks, breast with faint pale flammula-
tions. Throat white tinged Pale Yellow
(2.5Y8/3), irregularly but conspicuously
streaked by black edges (but not tips) on some
feathers. Irides gray brown, bill like the type,
feet slaty; body mass 9.0 g.

Another female netted and released at the
type locality in 1994 by NK was similar, but
with whitish irides and with nairower black
edges on the feathers of the whitish throat;
body mass 10.7 g. A third individual netted
and photographed near the type locality in
1994 (C. Canaday photo) was found by NK
to be similar. It had a whitish throat with nar-
row dark feather edges, appeai'ed to have a

fully ossified skull, brown irides, and blue
gray feet; body mass 9.5 g. The foregoing de-
scription of MECN 6925 was found by MLI
to match females taken at “Sunka 1”, Tza-
pino, and San Jacinto, as well as the descrip-
tion and photographs of the specimen collect-
ed upstream from Intuto on the Rio Tigre by
JA, except that central feathers of throats were
white (untinged) and lateral edges of tail
feathers were the same as the color of the back
(rather than reddish brown).

Additional female specimens examined. —
Myrmothenda fjeldsaai: ECUADOR: Napo,
“Sunka 1”, 40 km S Coca (WFVZ 45662,
collected November 1988). Pastaza, Tzapino
(LSUMZ 83109, collected in May 1976 by
Tallman). PERU: Loreto, San Jacinto, upper
Rio Tigre (MNHUK uncatalogued; collected
in July 1993 by Aucca); above Intuto, Rio Ti-
gre (taken by JA in January 1995 and to be
deposited at MHNJP). Mynnothenda h. hae-
matonota and M. h. amazonica'. PERU: Lor-
eto, Libertad, S bank Napo (LSUMZ, 2);
Puerto Indiana, N bank Amazon (AMNH, 1);
the vicinity of Huampami, Amazonas, Peru,
about 200 m elevation (LSUMZ, 1); Quebrada
Vainilla, S bank Amazon (LSUMZ, 6); Orosa
(AMNH, 1); Sarayacu (AMNH, 1); “mouth of
Rio Umbamba” (AMNH, 1). Madre del Dios,
15 km E Puerto Maldonado (MNHUK, 1).
BOLIVIA: (LSUMZ, 5). BRAZIL: (AMNH,
7). Myrmotherula h. pyrrhonota: ECUADOR:
Sucumbios, about 14 km N Tigre Playa
(MECN, 1). PERU: Loreto, Quebrada Oran
(LSUMZ, 1). COLOMBIA: (AMNH, 1).
BRAZIL: (AMNH, 16; USNM, 7). VENE-
ZUELA: (AMNH, 26; USNM, 5). Mynnothe-
nda spodionota including soioria: ECUA-
DOR: Napo, San Jose Abajo (AMNH, 2;
USNM, 1); Morona-Santiago: Chiguaza. Cu-
tucu (ANSP, 1); Zamora-Chinchipe: Zamora
(AMNH, 1). PERU: (AMNH, 4; LSUMZ, 20).

Mensural variation. — Measurements are
given in Table 2. As a whole, measurements
of populations in M. haematonota and M. spo-
dionota do not differ significantly (Table 2).
The only measurement whose ranges do not
overlap are the wing measurements for the
few known specimens of M. h. haematonota
from north of the Amazon and M. spodionota,
although there is only slight overlap of some
wing and tail measurements of M. spodionota
compared to other populations. Using a rule

162

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 2. Measurements of six populations in the Myrmotherula [haematonota] superspecies. Sexes are
combined. Bill is measured from the anterior nostril; wing is unflattened wing chord.

Measurement

M. fjeldsaai

M. h. haematonota
north of Amazon

M. h. haematonota
south of Amazon

M. h. pyrrhonota
Venezuela &
northern Brazil

M. spodionota

Bill Width

n = 7

n = 6

n = S

n = 10

n = \0

Range

3.6-4.0

3. 2-4.0

3.5-4. 1

3. 7-4. 2

3.6-3.9

Mean ± SD

3.79 ± 0.14

3.58 ± 0.34

3.86 ± 0.18

4.05 ±0.18

3.75 ± 0.13

Bill Depth

n — 1

n = 6

n = 8

n = 10

Range

3. 6-3. 9

3.6-4.2

3. 8-4.4

3. 6-4.4

3.8-4.3

Mean ± SD

3.72 ±0.15

3.83 ± 0.22

4.04 ± 0.21

4.03 ± 0.23

4.09 ±0.19

Bill Length

n = 10

n = 1

n = 10

Range

8.4-9. 5

8. 3-8.9

8.4-9.6

8. 8-9. 7

8. 2-9.4

Mean ± SD

8.97 ± 0.32

8.57 ± 0.30

9.07 ± 0.36

9.19 ± 0.31

8.81 ± 0.43

Tarsus Length

n = 8

= 6

n = 8

n = 10

Range

15-16

16-17

15-17

16-17

16-18

Mean ± SD

15.7 ± 0.5

16.2 ± 0.4

15.9 ± 0.6

16.1 ± 0.3

16.8 ± 0.6

Tail Length

n = 10

n = 1

n = 10

/! = 10

Range

33-39

34-37

34-39

34-38

37-40

Mean ± SD

35.8 ± 1.5

35.3 ± 1.1

35.1 ± 1.8

35.9 ± 1.3

38.4 ± 1.2

Wing Chord

n = 10

n = 1

n = 10

/? = 10

n = 10

Range

47-54

47-50

48-53

47-51

51-55

Mean ± SD

49.8 ± 1.9

48.1 ± 1.1

50.3 ± 1.6

49.2 ± 1.4

52.6 ± 1.4

similar to one proposed by Amadon (1949),
comparing means of measurements plus or
minus three standard deviations, differences
between wing and tail measurements of M.
spodionota compared with M. h. haematonota
north of the Amazon again stand out. Another,
larger set of measurements (mean followed by
range) confirms this tendency: the wing (53.8
mm; 51-56) and tail (38.9 mm; 33-43) mea-
surements of M. spodionota (n = 29) com-
pares to wing (51.0 mm; 46-54) and tail
(36.4; 33-41) for all populations of M. h. pyr-
rhonota (n = 86) and wing (52.4 mm; 50-56)
and tail (38.9 mm; 33-43) for all populations
of M. h. haematonota (n = 40). Differences
in bill among these larger samples were found
to be insignificant.

Distribution. — South and west of the Rfo
Napo in the lowlands of eastern Ecuador and
extreme northern Peru. Geographic range ex-
tends from the type locality and from nearby
“Sunka 1 ” in Napo, Ecuador, south to the
central portion of the Rio Tigre, a short dis-
tance upriver from Intuto, Loreto, Peru, ap-
proximately 360 km south-south-east of the
type locality (Fig. 1). A sight record (Willis
1988) from Andoas, Loreto, may also be of
this species.

Habitat. — At the type locality NK observed
M. fjeld.saai in tangled, but fairly open under-

story to lower canopy in humid primary terra
firma forest and into adjacent tongues of var-
zea forest. Neai' Sachacocha in the middle Rio
Tigre, Loreto, Peru, BMW and JA found M.
fjeldsaai in tall, closed canopy (except for
scattered treefalls), teira firme forest with an
abundance of palms in the understory. Known
from about 150 to 300 m elevation.

Vocalizations. — The loudsongs (Isler et al.
1998) of nominate M. haematonota, M. jjeld-
saai, and M. spodionota (Fig. 2) are similar,
but easily distinguishable from those of M.
leucophthalma (Isler, Isler, and Whitney, un-
publ. data). We do not provide a detailed anal-
ysis of differences among the loudsongs illus-
trated in Fig. 2 because the sample size for M.
fjeldsaai is inadequate. However, pace (notes
per sec), an element that appears to vary
among the loudsongs of M. haematonota
complex, has been shown to be one of the
most important characteristics distinguishing
loudsongs of closely related and syntopic
pairs of Thamnophilidae (Isler et al. 1998).

Behavior. — Foraging behavior (NK, BMW,
JA, C. Canaday, G. Rivadaneira, and R. S.
Ridgely, unpubl. data), of M. fjeldsaai appears
to resemble the other dead leaf specialists in
the group as described by Gradwohl and
Greenberg (1984), Rosenberg (1990, 1993,
1997), and Whitney (1994).

Krahhe et ul. • NEW ANTWREN FROM WESTERN AMAZONIA

163

1 0 -

-< -N ^ A -X ^

' ' ' ' \ \

1 1 1

2 “

0 -

1 1 1

0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5 2.0

1 0 -

8 -

6 -

4 -

iiiiiirTd:]. iti ^

1 \ 1

2 “

0 -

1 1 1

0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5 2.0

FIG. 2. Vocalizations of Myrmothenila fjeldsoai and loudsongs representative of neighboring populations of
other taxa in the M. [haematonota] superspecies. A. Myrmotherula fjeldsaai loudsong from the type locality
(recorded by N. Krabbe; LNS 65998). B. Myrmotherula fjeldsaai loudsong from Sachacocha, Loreto, Peru (B.
M. Whitney; to be archived at LNS). C. Myrmotherula h. pyrrhonota loudsong from Quebrada Sucusari, Loreto,
Peru (T. A. Parker, III; LNS 33798). D. Myrmotherula h. haematonota loudsong from Lores, Loreto, Peru (B.
M. Whitney; to be archived at LNS). E. Myrmotherula s. spodionota loudsong from 30 km west of Loreto,
Napo, Ecuador (B. M. Whitney; to be archived at LNS). E, G., and H. Myrmotherula fjeldsaai calls and rattle
from the type locality (N. Krabbe; LNS 65998) — compare with the rattle of M. spodionota presented by Whitney
(1994, fig. 1).

164

THE WILSON BULLETIN • VoL HI. No. 2, June 1999

Taxonomic rank. — The possibility that M.
fjeldsaai and M. haematonota occasionally
hybridize is raised by the Rio Tigre specimen
described by Hellmayr (1910). However, the
fact that the two maintain their integrity, de-
spite being in contact in an area with exten-
sive floodplain dynamics, makes it most prob-
able that they are correctly ranked as full spe-
cies.

Speciation. — In addition to being parapatric
with M. h. haematonota, M. fjeldsaai might
meet with M. h. pyrrhonota and M. spodiono-
ta, but there is no evidence of intergradation.
To the southeast (Fig. 1) the ranges of M.
fjeldsaai and M. h. pyrrhonota appear to be
separated by the middle Rio Napo, but both
may occur farther upstream where the rivers
Napo and Aguarico are narrow enough for
them to cross. To the west, the ranges of M.
fjeldsaai and M. spodionota appear to be sep-
arated altitudinally, with a narrow elevational
band that neither species occupies. Neither
was among the 505 species recorded at Jatun-
sacha, at 450 m during 26 months of field
work (B. Bocham, unpubl. data). Nor has
BMW found either between 400-600 m dur-
ing extensive field work around the village of
Loreto, Napo, Ecuador. Additionally, neither
was among the six species of Myrmotherula
recorded by NK during 1 2 days of field work
at Canelos, Pastaza, 500-700 m. Because the
forest is continuous between their elevational
limits, further field work might show the two
species’ ranges to be in contact.

Conservation. — The known range of Myr-
motheriila fjeldsaai encompasses the Yasuni
National Park (7281 km-) in Ecuador (whence
comes the type). Given that this huge area re-
mains effectively protected, M. fjeldsaai ap-
pears not to be at risk at present.

Etymology. — We take the pleasure of nam-
ing this species in honor of Prof. Jon Fjeldsa
of the Zoological Museum, University of Co-
penhagen. Through his countless publications,
most based on results obtained during field
trips to the most hostile of environments, he
has inspired a large number of biologists to
leave their desks and get into the field. Among
his achievements should also be mentioned
his most recent work (with C. Rahbek) in de-
limiting areas of top priority for conservation
in South America and Africa, work that could
eventually save a number of species from ex-

tinction. Apart from his impressive profes-
sional knowledge and self discipline, his many
achievements were possible only through his
great understanding of human nature, his gen-
erosity and helpfulness, and his unfailing habit
of treating everybody as an equal, legendary
from the jungles of South America to Africa.
The English name refers to the distinctive col-
or of the back.

ACKNOWLEDGMENTS

We are indebted to K. V. Rosenberg, J. M. Bates,
and D. Stotz for fruitful discussions; to E Vuilleumier,
AMNH, for helping NK with fund raising; to M. Mo-
reno, MECN, M. LeCroy, AMNH, and D. Agro, ANSP
for help in the use of collections in their care; to G.
Erisk, Naturhistoriske Riksmuseum, Stockholm, J. V.
Remsen, LSUMZ, R Colston, BMNH, M. B. Robbins
and R. O. Prum, UKMNS, and C. Samida, WEVZ, for
the loan of specimens under their care; to G. Budney,
LNS, for assistance and use of recordings; to the late
T. A. Parker, III, who recorded much of the material
used to compare vocalizations of M. leucophthalma
and M. haematonota-, and to P. Coopmans and J. Row-
lett who supplied additional recordings of M. spodion-
ota. Krabbe’s field work and studies in the USA were
generously funded by ECUAAMBIENTE, Quito; Zoo-
logical Museum, University of Copenhagen; and a
Chapman grant from the AMNH. M. Islet's work at
AMNH was supported by a Chapman collection study
grant. J. M. Bates, J. EJeldsa, G. R. Graves, D. Stotz,
J. P. O'Neill, and an anonymous referee kindly com-
mented on early drafts of the manuscript.

LITERATURE CITED

Amadon, D. 1949. The seventy-five per cent rule for
subspecies. Condor 51:250-258.

Cory, C. B. and C. E. Hellmayr. 1924. Catalogue of
birds of the Americas. Part III. Eield Mus. Nat.
Hist. Zool. Sen 13 (Publ. 223): 1-369.

Gradwohl, j. and R. Greenberg. 1984. Search be-
havior of the Checker-throated Antwren foraging
in aerial leaf-litter. Behav. Ecol. Sociobiol. 15:
281-285.

Hellmayr, C. E. 1910. The birds of the Rio Madeira.
Novit. Zool. 17:257-428.

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

ISLER. M. L., P. R. ISLER, AND B. M. Whitney. 1998.
Use of vocalizations to establish species limits in
antbirds (Passeriformes: Thamnophilidae). Auk
I 15:577-590.

Ortiz-Crespo, E I., P. J. Greenfield, and J. C. Ma-
THEUS. 1990. Aves del Ecuador Continente y Ar-
chipielago de Galdpagos. FEPROTUR and CE-
CIA, Quito, Ecuador.

Parker, T. A., Ill and J. V. Remsen, Jr. 1987. Fifty-

Krahhe et al. • NEW ANTWREN FROM WESTERN AMAZONIA

165

two Amazonian bird species new to Bolivia. Bull.
Brit. Orn. Club 107:94-107.

Ridgely, R. S. and G. Tudor. 1994. The birds of
South America. Vol. 2. The suboscine passerines.
Univ. of Texas Press, Austin, Texas.

Rosenberg, K. V. 1990. Dead-leaf foraging speciali-
zation in tropical forest birds. Ph.D. diss., Loui-
siana State Univ., Baton Rouge.

Rosenberg, K. V. 1993. Diet selection in Amazonian
antwrens. Auk 110:361-375.

Rosenberg, K. V. 1997. Ecology of dead-leaf foraging
specialists and their contribution to Amazonian
bird diversity. Ornithol. Monogr. 48:673—700.

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

Whitney, B. M. 1994. Behavior, vocalizations, and
possible relationships of four Myrmolherula ant-
wrens (Formicariidae) from eastern Ecuador. Auk
1 1 1:469-475.

Willis, E. O. 1988. Behavioral notes, breeding re-
cords, and range extensions for Colombian birds.
Rev. Acad. Colombiana Cienc. Exactas Fisicas
Nat. 16:137-150.

Zimmer, J. T. 1932. Studies of Peruvian birds. III. The
genus Myrmolherula in Peru, with notes on extra-
limital forms. Part 1. Am. Mus. Novit. 523:1-19.

Wilson Bull., 111(2), 1999, pp. 166-180

COMPARATIVE SPRING HABITAT AND FOOD USE BY TWO

ARCTIC NESTING GEESE

SUZANNE CARRIERE,' 2 5 ROBERT G. BROMLEY,'^ AND GILLES GAUTHIER'

ABSTRACT. — The timing of egg laying is generally constrained by female condition, which is partly deter-
mined by the food available to her before laying. Although it was generally believed that geese rely exclusively
on internal nutrient reserves for egg production, spring feeding is intensive in many populations of geese,
significantly adding nutrients necessary for egg production and incubation. We compared the spring feeding
ecology of Greater White-fronted Geese {Anser albifrons frontalis) and Canada Geese {Brantci canadensis hutch-
insii) on a shared nesting ground on the Kent Peninsula, NWT (68° N, 108° W), where pairs feed intensively
from arrival until incubation. Live plant biomass did not significantly increase within specific habitats during
preincubation, but the total available biomass was greater after snow melt because habitats with higher biomass
became available. Live plant biomass available in pond margins (30—60 g/m-) was 4-15 times higher than in
habitats that were available earlier, i.e., mud-flats and hummocks (4-8 g/m-). Before snow melt, both species
shared the 1-20% of the study area that was snow free (max. density 600 pairs/km-), opportunistically used the
only two available habitats, mud-flats and hummocks, and primarily ate (50-70%) tillers of Puccinellia spp.
During snow melt, pairs dispersed, pair density decreased (max. of 40 pairs/km-), and interspecific differences
in habitat and food use appeared. White-fronted Geese used pond margins and ponds more often than Canada
Geese. After snow melt. White-fronted Geese predominantly fed in ponds on Carex spp. and Dupontia fisheri
rhizomes and basal stems; Canada Geese continued feeding opportunistically, pecking leaves in all habitats and
grubbing rhizomes in pond margins and ponds. White-fronted Geese used the grubbing technique more often
than Canada Geese in all habitats and periods. Received 13 Feb. 1998. accepted 17 Nov. 1998.

Energy investment by females in reproduc-
tion is highest during the period of egg for-
mation in birds with precocial young, such as
geese (King 1973). Clutch size and timing of
laying are potentially constrained by female
condition, which is partly determined by the
amount of energy and nutrients available to
her before egg formation (Drent and Daan
1980, Winkler and Walters 1983).

Early nesting is critical for Arctic nesting
geese because of the short summer and the
rapid seasonal decline in components of re-
productive success such as gosling growth and
probability of producing recruits (Barry 1962,
Cooke et al. 1984, Cooch et al. 1991, Sedinger
and Flint 1991, Lindholm et al. 1994). One

' Dept, de Biologic et Centre d’Etudes Nordiques,
Univ. Laval, Quebec, QC, GIK 7P4, Canada.

’ Present address: Wildlife and Fisheries Division,
Dept, of Resources, Wildlife, and Economic Devel-
opment, Government of the Northwest Territories, 600
5102-50 Avenue, Yellowknife, NT XI A 3,S8, Canada.

’ Wildlife and Fisheries Division, Dept, of Resourc-
es, Wildlife, and Economic Development, Government
of the Northwest Territories, 600 5102-50 Avenue,
Yellowknife, NT XI A 3S8, Canada.

Present address: Box 1 177, Yellowknife, NT XI A
2N8, Canada.

^ Corresponding author;

E-mail: suzanne_carriere@gov.nt.ca

Strategy to facilitate early nesting is to carry
nutrient reserves accumulated during migra-
tion to the breeding grounds (Ankney and
Macinnes 1978, Wypkema and Ankney 1979,
Ankney 1984, Budeau et al. 1991, Bromley
and Jarvis 1993, Choiniere and Gauthier
1995). There is, however, an upper limit to the
amount of reserves that can be economically
carried during migration (Lindstrom and Al-
erstam 1992).

In migratory birds, the timing of rapid fol-
licular development (RED) initiation with re-
spect to spring migration may affect the rel-
ative contribution to egg production of nutri-
ents acquired en route versus those acquired
on nesting grounds. Timing of RED initiation
directly determines laying date of the first egg
(reviewed by Rohwer 1992). Because some
female geese in some locations typically nest
soon (3-6 days) after their arrival on the nest-
ing ground {Anser rossii: Ryder 1970; Anser
caerulescens caerulescens: Ankney 1977, An-
kney and Macinnes 1978; Anser canagica:
Thompson and Raveling 1987; Branta berni-
cla bernicla: Spaans et al. 1993), RED is ini-
tiated before arrival, hence the date of nest
initiation is independent of food availability
on the nesting ground (Raveling 1978, Ank-
ney 1984, Newton 1977; but see Prop and de
Vries 1993).

166

Carriere el al. • SPRING FEEDING BY ARCTIC GEESE

167

In other populations most female geese lay
at least 12 days after an'ival on the nesting
grounds, long enough for the completion of
RFD of the first egg [Branta bernicla nigri-
cans: Raveling 1978; B. canadensis: Macln-
nes et al. 1974, Bromley 1984; A. albifrons:
Fox and Madsen 1981, Budeau et al. 1991; A.
caenilescens caerulescens (at La Perouse
Bay): Findlay and Cooke 1982; A. c. atlantica:
Gauthier and Tardif 1991]. In this case, food
availability on the nesting ground can affect
the date of RFD initiation, the date of nest
initiation, and potentially clutch size and con-
stancy of incubation. Most of the energy nec-
essary for egg formation and laying was met
by food on the nesting grounds for A. albi-
frons frontalis (Budeau et al. 1991), B. can-
adensis occidentalis (Bromley and Jarvis
1993), and A. caerulescens atlantica (Choin-
iere and Gauthier 1995). Intermediate cases
may occur where the time available for feed-
ing between arrival on the nesting ground and
laying varies greatly among females and/or
among years within populations (Raveling
1978).

Food availability in the Arctic in spring was
traditionally thought to be so low that, al-
though female geese could feed, they could
not meet their energy requirements for daily
maintenance or egg production (Barry 1962,
Ryder 1970; reviewed by Rohwer 1992). For
some species, body mass of females generally
increases before or during egg production
(Wypkema and Ankney 1979, Budeau et al.
1991, Bromley and Jarvis 1993, Choiniere
and Gauthier 1995) indicating that energy in-
take during these periods could at least meet
requirements for daily maintenance (Ganter
and Cooke 1996, Carriere 1996). Further, var-
iation in timing of nesting can be related to
variation in food availability prior to egg for-
mation (Prop and de Vries 1993). Thus, food
availability and use in the Arctic in spring
clearly is significant.

In the Arctic, food availability during pre-
incubation is highly variable. This variation
hinders interspecific comparisons of feeding
ecology during preincubation because we can-
not differentiate between factors that are site
specific (e.g., weather, snow melt patterns,
plant phenology) and species specific (e.g.,
body size, bill moiphology; Prevett et al.
1985, Fox et al. 1992).

We compared the feeding ecology of Great-
er White-fronted Geese (Anser albifrons fron-
talis) and Canada Geese {Branta canadensis
hutchinsii) during preincubation. These spe-
cies have similai' body mass (White-fronted
Goose 8% > Canada Goose; R.G.B., unpubl.
data), reproductive chronology, and share the
same Arctic spring feeding and nesting
grounds on the Kent Peninsula, NWT, Canada.

Our objectives were to determine (1) how
snow melt affected the availability of feeding
habitats and plant biomass, (2) whether
White-fronted and Canada geese differed in
their use of habitat and food during preincu-
bation on a shared nesting ground, and (3)
how changes in habitat and food availability
resulting from snow melt affected dispersal,
habitat and food use, and the timing of nesting
in these two species of geese.

STUDY AREA AND METHODS

The study was conducted on the Walker Bay Study
Area (68° 22' N, 108° 04' W), southwest Kent Penin-
sula, Northwest Territories (Eig. 1), as part of longterm
studies on the breeding ecology of White-fronted and
Canada geese (see Bromley et al. 1995). The shallow
plain of the river valley is a high density nesting area
for both species of geese (Bromley et al. 1995).

Studies of the feeding ecology during preincubation
were made on an intensive study area positioned to
permit semicontinuous observation of geese in spring
while minimizing disturbance caused by human move-
ments (Eig. 1 ). The area was further divided into two
sites: site A (0.72 km-) was representative of the hab-
itats found all along the river and typical of areas in
which arriving geese concentrated until snow melt; site
B (4.33 km-) was representative of the rest of the gen-
eral study area where nesting typically occurred. Site
A, a raised levee paralleling the river, was drier than
site B. Habitats available to geese were classified from
dry to wet: hummock, mud-fiat, pond margin, and
pond (see plant list per habitat in Carriere 1996). Mud-
fiat habitats were sparsely vegetated flats of exposed
glacial marine sediments (ca 3000 yrs ago; Dyke and
Dredge 1989), largely saline clays and silts. Hummock
habitats were formed by frost heaves (5-30 cm high)
and covered by thin soil where Salix spp. dominated.
Pond margins were the edges of depressions that were
wet from snow melt to early June, had low salinity,
and formed meadows dominated by graminoids and
forbs. Pond habitats were depressions 0.10-1.00 m
deep, inundated at least until early July, and dominated
by hydrophilic forbs. All habitats formed a fine grained
mosaic with patches (i.e., continuous areas of same
habitat) of various shapes. Most patches covered 50-
900 m-, with some mud-flats outside sites A and B
extending more than 1 km-.

General phenology. — We recorded the phenology of

168

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Nests

White-fronted geese

1993

1994

Canada geese

1993 *

1994 ■

Drumlin — —
Lake/Pond ^

T ransect

Tower ■

FIG. 1. Walker Bay Study Area is situated in a valley near Walker Bay on the Kent Peninsula, NWT,
Canada. A section selected for detailed study was divided into 2 sites: site A was snow-free first in spring and
was typical of habitats available along the river banks. Snow melt phenology in site B was typical of the rest
of the study area.

snow melt, goose arrival, and goose dispersal along
four 1 km transects (Fig. 1). At 2-3 day intervals we
visually estimated snow cover (±5%) in 200 X 200 m
areas, situated in pairs, one on each side of the transect
at each of five stations 200 m apart. Thus, we covered

11% of site A (2 stations) and 7% of site B (18 sta-
tions). At each station, we used binoculars to count
and locate all geese that could be identified to species.

Preincubation was divided into three periods defined
by snow melt phenology on site A: (1) before snow

Carhere el al. • SPRING FEEDING BY ARCTIC GEESE

169

melt (snow cover >80%). (2) during snow melt (snow
cover = 20-80%)- — a period of rapid change, and (3)
after snow melt (snow cover <20%).

Nesting phenology. — Nests were located by observ-
ing nest building behavior from towers. Date of the
first egg laid (i.e., date of nest initiation) was estimated
only for nests found during laying by subtracting 1 .3
d per egg (R.G.B., unpubl. data) already laid. Nests
outside site A and B were located during ground sur-
veys of randomly selected 1 km- plots (Bromley et al.
1995). We compared median laying dates for combi-
nations of years and species using Kruskal-Wallis AN-
OVA on Ranks tests, followed by non-parametric mul-
tiple comparison tests using Excel 97 following Daniel
(1978).

Habitat availability. — We measured habitat avail-
ability in the areas at two towers (see below) by map-
ping habitat patches with a 20 X 20 m grid in the field
and using aerial photographs. Availability of a habitat
was defined by the cumulative area of all patches of
that habitat as a proportion of the total snow-free area.
Only snow-free areas were considered accessible to
feeding geese (Hall et al. 1997).

Use and selection of habitats. — We observed goose
pairs throughout preincubation from two towers in
1993 and one tower in 1994. Daily observations were
conducted from one tower at a time depending on the
distribution of geese in the study area. We used the
scan sampling method (Altman 1974), with scans con-
ducted every 2-6 hr during the 24 hr cycle. During a
scan, we observed all pairs present within 400 m of
the observation tower. Sections of sites A and B were
visible from both towers (Eig. 1). Eor each pair, we
noted sex, behavior, and habitat used by each pair
member. Sex was determined using relative neck size
(larger in males), abdominal profile (larger in females
with developing follicles; Owen 1981, Eox et al.
1995), and alert position (male usually standing high-
er). Pairs were recorded as feeding when either the
female or both members were grubbing (on below-
ground plant parts), pecking (on above-ground plant
parts), searching for food (moving with head down),
or drinking.

We calculated habitat use for each scan as the pro-
portion of observed feeding pairs using each available
habitat. Pond and pond margin use data were pooled
as wet habitats to obtain sufficiently large samples for
a C-test (Sokal and Rohlf 1981). For analysis, we
pooled habitat use data within snow melt period and
tower. Habitat use data were assumed to be indepen-
dent within and between scans. This seemed reason-
able for two reasons. First, preliminary observations
on focal individuals indicated that pairs showed little
synchronization in movements between habitat patches
(i.e., no group behavior). Second, because of the ex-
treme patchiness of habitats, pairs could use all avail-
able habitats within 2 hr (minimum time interval be-
tween scans); 58% of Canada Geese (n = 96) and 43%
of White-fronted Geese (/; = 69) individuals observed
in focals used more than I habitat patch within 10 min
(S.C., unpubl. data). The constant movement of pairs

between the observed and unobserved areas during
preincubation minimized pseudoreplication. For ex-
ample, no individual with a coded neck collar (from
1987-1994 banding operations; Bromley et al. 1995)
was observed in more than six different scans (only
30% of collared individuals were observed more than
twice).

For each year and period, interspecific differences
in habitat use were analyzed without reference to hab-
itat availability using G-tests of independence, fol-
lowed, when significant differences were detected, by
pairwise unplanned comparison tests (Sokal and Rohlf
1981 ; using Excel 97).

Habitat use was compared to availability before and
after snow melt, but not during melting, because avail-
ability changed too rapidly, i.e., from 20% to 80%
snow-free area in <7 d. We used X" goodness-of-fit
tests, followed by Bonferroni simultaneous confidence
interval tests (Sokal and Rohlf 1981, Neu et al. 1974;
using Excel 97) to detect sources of significant differ-
ences (design 1 in Thomas and Taylor 1990).

Plant availability within habitats. — We estimated
food plant availability in snow-free habitats by sam-
pling randomly 20 X 20 cm quadrats within habitat
patches in site A and B, 4 times between 17 May-14
June throughout preincubation (1993: n = 120; 1994:
n = 218). We collected all above and below ground
vegetation 0. 1-3 cm deep then froze each sample for
transportation. Ground vegetation below 3 cm was
never thawed and was considered unavailable to geese.
In the laboratory, live above and below ground vege-
tation was sorted by species, dried at 45° C to constant
mass, and weighed. We pooled plant species according
to food plant categories used in the analysis of feces
(see below). Availability of each food plant category
in each habitat was defined as the average proportion
of total dry biomass represented by that category in
each habitat during each snow melt period (Eig. 2).

We analyzed total (pooled species) above and below
ground biomass separately. We first analyzed differ-
ences in total biomass among habitats and sampling
dates for each year separately using Kruskal-Wallis
ANOVA on Ranks tests, followed by multiple com-
parison tests to detect pairwise differences (Daniel
1978). We then pooled sampling dates and analyzed
differences in biomass among habitats and years with
Two Way ANOVA using square-root transformed data
(Sokal and Rohlf 1981).

Diet. — We determined diet using two complemen-
tary methods: (I) microhistological analyses of feces
and (2) detailed observations of feeding techniques.
Additional data were available from the esophageal
content of 16 female Canada Geese collected in 1994.

Throughout preincubation, we collected all feces
that we could assign to an individual bird (i.e., a bird
that was observed defecating). Feces were individually
frozen for transportation, dried in the laboratory, and
analyzed using microhistological techniques (see John-
son 1982). We sampled four slides per feces, with 20
observation fields per slide. We identified most plant
fragments (80-100% per feces) to genus. We grouped

170

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

(/)

CO
CT3

O 20
^ 0
^ 20
^ 40

Pond

I After (20)

Mud-flat

’ ^

# i

' -3

1 1

^ i

■'s/

<tr

J 1

U7 S'

_l L

Hummock

Margin

Pond

FIG. 2. Relative availability of plants in each snow-free habitat before, during, and after snow melt, measured
as the average percent food in dry biomass on the Walker Bay Study Area, NWT, May-June 1993-1994. Sample
sizes are given in parentheses, years pooled. Availability of aboveground plant parts is given above the zero
line and of below ground plant parts below the zero line.

some rare food plants of the same genus or family that
were growing in the same habitat (Table 1, Fig. 2).
The validity of the identification of plant fragments in
feces was assessed by analyzing plant mixtures of
known composition (Holechek and Gross 1982) and
by comparing fecal results to esophageal contents of
female Canada Geese (Carriere 1996). We could not
easily differentiate Sali.x arctica from Potentilla nivea,
nor discriminate Dupontia fisheri and Eriophorum spp.
from some fragments of other Gramineae and Cyper-
aceae, respectively. Potentilla spp. and Eriophorum
spp., however, were relatively rare on the study area
(Fig. 2, see other Dicots and Eriophorum spp.). Some
below ground plant parts could not be identified reli-
ably (Carriere 1996). Feces analyses consequently pro-
vided only an estimate of use of common above
ground food plants on our study area. Coefficients of
variation of the proportion of food plants among slides

within feces ranged from less than 5% for graminoids
to 10—20% for dicots. We defined diet as the average
percent of each food plant (% of fragments) present in
the feces (frequency of occurrence method; Johnson
1982). Diet was determined for each goose species and
snow melt period by pooling years, with individual
feces representing sampling units.

We compared use of each food plant among all pos-
sible combinations of goose species and snow melt
periods using Kruskall-Wallis ANOVA on Ranks tests
followed, if significant, by multiple comparison tests
(Daniel 1978) using Excel 97.

Feeding techniques. — Use of below ground plant
parts was indexed by the proportion of feeding pairs
grubbing in each habitat during scans. Conversely,
pairs were assumed to eat above ground plant parts
while pecking.

We used a Paired ?-test to determine if the propor-

TABLE 1 . Diet of White-fronted and Canada geese before, during, and after snow-melt, from arrival on the nesting grounds to incubation. Walker Bay Study
Area, Kent Peninsula, NWT, 1993—94. “Arriving” females were collected before snow-melt, “pre-laying” females during and after snow-melt, and the incubating
female after snow-melt.

Carriere et ui • SPRING FEEDING BY ARCTIC GEESE

'sO

00 ^00

— sO ^

\n ^ — o

SO so
Ov

< U

•rt n

Urn

n n

u <

—

u IT)

■rt

r^i

—

(■■I

r't

—

(^1

—

X-

c/)

OX)

O ic:

•t: i: t: 'ij

(u C
= 5;
■“ Si
d.^
a

V 2

^ S-

UD "O

■o o

^ S

« -

a,k)O'^O^Q03:GD

"O
3 ■
42

§•

) ^

5 '

■ • :

; ^ Cl ,

-S

^ O.

i= a.

5 t/5

S =>,

“■S3

I Q I
■2

^✓5 V) . .

^ -S
S -SJ-
s£

: <n i
“ o S ^

; s cL S ^

: O S'

O O

o T3 y

171

f Hippiiris sp.: W. vulfturis or //. lelniphylla. Senecio sp.: 5. conges/us, includes rhizomes, young shoots, and leaves.

Unidentified includes fragments of dicots, of Eriophorum spp.. and all rhizomes except those of Etymus sp.. and Hippiiris spp.
' tr = trace.

172

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Date

■O

LIG. 3. Number of individual White-fronted Geese (WL) and Canada Geese (CG) counted daily at each
station on sites A (/; — 2) and B (// = 18; see Lig. 1), from arrival to early incubation, and number of nests
initiated daily by each species on the Walker Bay Study Area, NWT, 1993-1994. Arrows indicate duration of
rapid follicular development (RED) for a first egg layed at the median laying date each year. Shaded areas
indicate periods of snow melt, defined as 80-20% snow cover. Snow melt pattern on site A was used to define
periods for all analyses.

tion of time spent grubbing differed between goose
species, pairing observations within habitat-period-
year combinations. Lor each species, we used linear
regre.ssion to analyze how the proportion (arcsine
square-root transformed in degrees; Sokal and Rohlf
1981) of time spent grubbing changed with the pro-
portion of below ground biomass available in different
habitats, snow melt periods, and years. The short, rapid
melting period in 1994 was combined with the after-
melt period.

Unless otherwise stated, statistical analyses were
performed using SigmaStat® (version 1.0; Jandel Sci-
entific Software 1993) in the PC DOS and Microsoft®
Windows® operating system. Statistical significance
was established at P < 0.05.

RESULTS

Annual phenology. — Upon our airival on
20 May 1993 and 17 May 1994, sites A and
B were 95% and 99% snow covered, respec-
tively. Hummock and mud-flat habitats locat-
ed along the river were the only habitats avail-
able. Relative to 1987-1996 (R.G.B., unpubl.
data), snow melt was early in 1994 and av-
erage in 1993 (Fig. 3). In general, snow melt-
ed first on dry habitats (mud-flat, hummock),
then on wet ones (pond margin, pond), and
earlier on site A than on site B (Fig. 3). The
periods before snow melt were 10 days (1993)

Carriere el al. • SPRING FEEDING BY ARCTIC GEESE

173

and 7 days (1994); melting periods were 7
days (1993) and 3 days (1994), and periods
after snow melt [end of melting to onset of
incubation (median laying date + 4 days)]
were 9 days (1993) and 13 days (1994). The
total length of the preincubation periods on
the nesting ground were thus 26 d in 1993 and
23 d in 1994.

Geese were present along the river (on site
A or outside the intensive study area) upon
our aiTival in 1993 {n = 50-80 pairs) and
1994 {n = 140-150 pairs), but major anivals
occuned aiound 27 May 1 993 and on or prior
to 20 May 1994 (see transect surveys: Fig. 3,
site A). During snow melt, goose numbers de-
creased on site A (and from other areas along
the river) and increased slightly on site B as
geese dispersed from areas along the river to
the rest of the study area (Figs. 1, 3).

We are confident that the pairs observed
during this study were part of the locally nest-
ing population because about Vi of the collared
individuals we found nesting in the study area
{n > 1 1 of each species per year) were ob-
served on site A and other areas along the
river. Individuals collared in other study areas
were never observed on Walker Bay Study
Area (R.G.B., pers. obs.).

The first nests were found on 5 June 1993
and on 28 May 1994. Median laying dates
were 11 June 1993 (White-fronted Goose: n
= 38, Canada Goose: /r = 35) and 2 June
1994 (White-fronted Goose: n = 26, Canada
Goose: n = 21) for both species (Kruskal-
Wallis ANOVA on Ranks: H = 77.4, P <
0.001; medians were different between years
only; Fig. 3). The minimum intervals between
goose arrival and median laying date were 14
days in 1993 and 13 days in 1994.

Use and selection of habitats. — White-
fronted and Canada geese did not differ in
their use of habitats before snow melt in either
year (G-tests of independence: 1993: G.,jj =
0.083, 1 df, F > 0.05; 1994: G,„ = 0.031, 1
df, P > 0.05; Fig. 4A, B). Both species also
used habitats according to their availability
during that period (y^ Goodness-of-fit: all
< 3.08, P > 0.05).

Habitat use differed between species during
snow melt in both years (G-tests of indepen-
dence: 1993: G,jj = 18.14, P < 0.001; 1994:
Gajj = 39.26, P < 0.001). Both species fed in
newly available wet habitats (pond margin

and pond) during snow-melt, but White-
fronted Geese used them more often than did
Canada Geese.

Pairwise compaiison tests showed that in
both years Canada Geese used hummock and
mud-flat habitats significantly more often than
White-fronted Geese and that White-fronted
Geese used wet habitats significantly more of-
ten than Canada Geese (Fig. 4C, D).

Similar interspecific differences in habitat
use were observed after snow melt in both
years (G-test of independence: 1993: G^^j =
23.03, P < 0.001; 1994 sites pooled: G =
69.44, P < 0.001). Pairwise comparison tests
showed that White-fronted Geese used ponds
(1993 and 1994) and pond margins (1994)
significantly more often than Canada Geese
(Fig. 4E, F, G).

In 1993, habitat selection (i.e., use vs avail-
ability) by both species differed after snow
melt (x" Goodness-of-fit: Canada Goose: x^ =

94.8, P < 0.001; White-fronted Goose: yf =
117.9, P < 0.001). Canada Geese preferred
(i.e., use > available) mud-flat, pond margin,
and pond, and avoided hummock habitats
(Fig. 4E). White-fronted Geese preferred pond
and avoided hummock habitat (Eig. 4E). In
1994, most geese fed neai' tower 1 after snow
melt (Fig. 1), where the relative availability of
habitats differed greatly between site A and B
(Fig. 4). Consequently, we analyzed habitat
use in each site separately. Data in dry (mud-
flat/hummock) and wet (pond maigin/pond)
habitats were pooled at site B to obtain suf-
ficient sample sizes for x^ tests. Neai' the river
(site A), Canada Geese used habitats accord-
ing to their availability (x' Goodness-of-fit: x"
= 0.92, P > 0.05; Fig. 4F), while White-front-
ed Geese prefeired pond maigin and pond
habitats and avoided mud-flat habitats (x“ =

49.8, P < 0.001). On site B, Canada Geese
again used habitats according to their avail-
ability (x^ = 0.22, P > 0.05; Fig. 4G) and
White-fronted Geese still prefeired wet habi-
tats (pond margin/pond) and avoided dry ones
(mud-flat/hummock; x‘ = 23.4, P < 0.001).

Diet. — Nine food types were recognized in
feces (Table 1 ). The number of different food
types detected increased as snow melted, as
expected because of the increase in available
habitats. A major food before snow melt was
Puccinellia spp. tillers, which accounted for
52% and 73% of White-fronted and Canada

174

THE WILSON BULLETIN • Vol. HI, No. 2, June 1999

1993 1994

Mud Hummock Mud Hummock

100

Mud Hummock Margin Pond Mud Hummock Margin Pond

I I Canada Geese
El White-fronted Geese
H Availability (expected use)

Mud/Hummock Margin/Pond

FIG. 4. Use (% of feeding pairs) of mud-flat, hummock, pond margin, and pond habitats by White-fronted
(WE) and Canada geese (CG), and availability of these habitats during preincubation on the Walker Bay Study
Area, NWT, 1993-1994. (a-b) Before snow melt 1993 (CG: n = 332; WE: n = 228) and 1994 (CG: /; = 1265;
WE; n = 383). (c-d) During snow melt 1993 (CG: n = 532; WF: n = 137) and 1994 (CG: n = 637; WE: n =
168); habitat availability could not be estimated, (e) After snow melt at site A in 1993 (CG: /; = 120; WF; n
= 87). (f-g) After snow melt in 1994, site A (CG: n = 50; WF; n = 33) and site B (CG; n = 22; WF: n =
29). A -I- indicates use > availability, indicates use < availability and no symbol indicates use =

availability (P > 0.05).

goose diet respectively. After snow melt, more
than 50% of the diets of White-fronted and
Canada geese were composed of Carex spp.
and Diipontia fisheri. For both species, most
(85%) feces contained food plants represen-
tative of at least 2 habitats. Consequently,
changes in diet reflected changes in both hab-
itat use and food use within habitats.

The greater diversity of the diet found in
the feces of Canada Geese relative to esophagi
(Table 1 ) is probably caused by the accumu-
lation of food fragments in the digestive sys-
tem. Nevertheless, analyses of esophageal
content indicated that some Gramineae (frag-
ments not identified at the genus level) could
be overestimated in fecal analyses, whereas

Stellaria humifusa, Diipontia fisheri, and Carex
seeds could be underestimated (see methods,
Cairiere 1996). Young leaves and open buds
of Salix spp. were found only in female Can-
ada Geese collected during incubation.

Available biomass. — Plant biomass varied
among combinations of habitats and sampling
dates for both above ground (Kruskal- Wallis
ANOVA on ranks: 1993; H = 46.0, P <
0.001; 1994: H = 152.1, P < 0.001) and be-
low ground plant parts (1993: H = 25.7, P =
0.012; 1994: H = 58.7, P < 0.001; Fig. 5).
There was little seasonal increase in total dry
biomass within each habitat, except after the
median laying date lor above ground biomass
in pond margins (1994 only) and below

Carriere et al. • SPRING FEEDING BY ARCTIC GEESE

175

21 28 'June 6 14

MUD-FLAT HUMMOCK MARGIN POND

FIG. 5. Live above and below ground dry biomass
(mean ± SE) available to geese in 4 habitats during
preincubation on the Walker Bay Study Area, NWT,
1993-1994. Values with same letter within graphs do
not significantly differ (multiple comparison tests after
a Kruskall-Wallis ANOVA on Ranks test for each year,
above and below ground biomass tested separately, all
tests P < 0.02). Grey areas indicate periods of snow
melt (80-20% snow cover on site A). Number of sam-
pled 20 X 20 cm quadrats per sampling date were:
(1993) mud-flat = 3, 5, 17, 14; hummock = 6, 9, 11,
1 1; pond margin = NA, 3, 6, 6; pond = NA, NA, 3,
26 and (1994) mud-flat = 12, 16, 24, 23; hummock =
10, 16, 16, 16; pond margin = NA, 17, 16, 16; pond
= NA, NA, 16, 20. NA = datum was not available,
under snow.

ground biomass in pond margins (1993) and
pond (1994). Seasonal changes in below
ground biomass were mostly the result of a
gradual deepening of the active layer. We
pooled sampling dates and found differences
in above ground biomass among habitats
(Two Way ANOVA: F = 1 12.1, P < 0.001)
and years (F = 4.25, P = 0.04; habitat-year
interaction, F = 7.25, P < 0.001). Above
ground biomass in pond margins was higher
than in any other habitat, particularly in 1994.
We found a tendency for below ground bio-
mass to be lower in dry (0-1.08 g/m^) than in
wet habitats (1.33-24.4 g/m-; Two Way AN-
OVA: F = 2.61, P = 0.05), but no difference

EIG. 6. Relationship between below ground bio-
mass available and proportion of feeding time spent
grubbing by female White-fronted (WE) and Canada
geese (CG) during preincubation on the Walker Bay
Study Area, NWT, 1993-1994. Each point is the pro-
portion of scanned pairs observed grubbing (arcsine
square-root transformed) in a habitat during a specific
period and year. Closed symbols are Canada Geese,
open symbols are White-fronted Geese. Habitats are
mud-flat (circles), hummock (squares), pond margin
(triangles), and pond (diamonds). Linear regressions:
for WE: y = 0.78 -F 0.91x [0.50-1.32; 6(95%C1)], F
= 18.0, P = 0.001; for CG: y = 0.28 + 0.72x [0.27-
1.17; B(95%CI)], F = 9.9, P = 0.009.

between years {F — 1.53, P > 0.05; habitat-
year interaction, F = 4.09, P = 0.007).

Use of below ground food plants. — Canada
Goose pairs grubbed less often than White-
fronted Geese (Canada Goose: 12.4%, White-
fronted Goose: 56.9%; Paired f-test: t —
— 14.2, P < 0.001; Fig. 6), which grubbed ex-
tensively even in habitats where below-ground
biomass was very low (i.e., in mud-flat/hum-
mock), before widespread snow melt and
thawing of the ground surface. Spatial distri-
bution of below ground vegetation (mostly
Elymus sp. rhizomes) was clumped in these
habitats and available below ground biomass
may have effectively been much higher in
some patches used by foraging individuals. In
all habitats, before and during snow melt,
most of the below ground biomass was frozen
under the active layer and hence was not
available to geese.

There was a significant positive relationship
between the proportion of below ground bio-
mass in different habitat-year-period combi-
nations and the proportion of feeding time

176

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

geese spent grubbing in these habitats (linear
regression: White-fronted Geese: F = 18.0, P
= 0.001; Canada Geese: F = 9.9, P = 0.009;
Fig. 6). This indicates that both White-fronted
and Canada geese modified their feeding be-
havior with changes in below ground food
availability.

DISCUSSION

Arrival, phenology, and timing of laying. —
The minimum interval between peak arrival
and laying was 1-4 day longer than necessary
for rapid follicular development (Alisauskas
and Ankney 1992a), which is typical in most
goose populations during eaily and average
years (Raveling 1978, Fox and Madsen 1981,
Budeau et al. 1991, Gauthier and Tardif 1991,
Bromley and Jarvis 1993). Anival and laying
initation were both earlier in 1994 than in
1993 and peak initiation of laying occurred
about 15 days after peak arrival in 1993, but
apparently only 12-13 days after airival in
1994. The longer interval in 1993 likely was
due to later snow melt that yeai" nest initiation
closely followed snow melt in site B. In con-
trast, during the early snow melt of 1994,
nesting sites became available (i.e., when
snow cover <50% in site B) 5-8 days before
peak nest initiation, indicating that completion
of RFD, rather than snow melt, constrained
nesting that year. These relationships were re-
markably similar in White-fronted and Canada
geese.

The close proximity of prenesting feeding
sites to nesting sites conveys benefits to geese
arriving on the Arctic nesting grounds before
RFD. During both years, pairs of both species
dispersed from early exposed locations near
the river to additional areas exposed late dur-
ing snow melt. This resulted in their using
feeding areas that were increasingly distant
from the river. Such short distance movement
by breeding pairs from feeding areas to nearby
nesting areas has been observed in other
goose populations (Gauthier 1993, Prop and
de Vries 1993). This local dispersal is similar
to the final migratory flight from staging areas
to the breeding grounds in populations where
most females nest soon after arrival in the
Arctic (Ankney and Machines 1978, Wypke-
ma and Ankney 1979). For the latter case,
however, the flight occurs late in RFD (Wyp-
kema and Ankney 1979). Unlike geese nesting

adjacent to feeding sites, this flight necessarily
draws largely upon body reserves during RFD
because long migratory flights would use
more energy than local dispersal. Further-
more, earlier anival on the breeding ground
enables pairs to directly assess nesting con-
ditions (Wypkema and Ankney 1979, Peterson

1992, Rohwer 1992, Ganter and Cooke 1996).

Although little plant growth occurred, food

availability rapidly and greatly increased dur-
ing preincubation. The change in total avail-
able biomass was due to the rapid exposure
of habitats with large plant biomasses during
snow melt, a common phenomenon in tundra
ecosystems (Wielgolaski et al. 1981). This
may play an important role in determining in-
dividual reproductive decisions in many Arc-
tic nesting geese (Gauthier 1993, Prop and de
Vries 1993) because increasing food avail-
ability before and during egg-production
could improve individual condition (as in-
dexed by body mass; Bromley and Jaivis

1993, Choiniere and Gauthier 1995, Ganter
and Cooke 1996). Which components of re-
productive success (laying date, clutch size,
nest attentiveness, condition at hatching, or a
combination of these) may be affected by a
female’s improved body condition depends on
the timing of changes in food availability with
respect to RFD initiation in individual females
(e.g., Bolton et al. 1993, Dalhaug et al. 1996).

Interspecific differences in feeding ecolo-
gy.— Snow melt, and the concomitant changes
in habitat and food availability, enabled us to
detect similarities and differences in resource
use and selection between White-fronted and
Canada geese during preincubation. In both
years, both species shared feeding sites and
had the same average date of nest initiation
(i.e., RFD initiation), hence they could expe-
rience the same changes in the availability of
resources at similar times relative to their re-
productive process.

Both species used habitats opportunistically
before snow melt when pair density was high-
est (up to 600 pairs/km^ in snow-free areas)
and habitat availability lowest (1—20% of
study area was snow free). With onset of snow
melt, both species dispersed and overall pair
density decreased (<40 pairs/km').

Alter snow melt, Canada Geese generally
used habitats more opportunistically than
White-fronted Geese, which consistently se-

Caniere et al. • SPRING FEEDING BY ARCTIC GEESE

177

lected wet habitats with high plant biomass.
Below ground food plants were used more ex-
tensively by White-fronted Geese than by
Canada Geese in all habitats. These differenc-
es are similar to those reported between Bran-
ta canadensis interior and Anser caeridescens
caeridescens on a common staging area on
James Bay (Prevett et al. 1985), but unlike our
study those differences could be explained in
part by segregation of feeding areas.

Although we could not test for density de-
pendent habitat selection because of synchro-
nous changes in habitat availability and in pair
density with snow melt, the evidence indicates
that interspecific differences in the relative
suitability of habitats for feeding exists. Con-
sistent with a release from population density
effects in a constant environment (Rosen-
zweig 1985, Morse 1990), White-fronted
Geese changed from opportunistic to selective
use of habitats with the highest plant biomass.
In contrast, Canada Geese exhibited a rela-
tively weak shift.

We suggest three possible non-exclusive
explanations for the interspecific differences
in habitat use and diet we observed.

1. Pond margin and pond habitats may be
more profitable to White-fronted Geese than
to Canada Geese because of the longer and
apparently more robust bill morphology of the
former (WF culmen is 35.3% and skull is
15.4% longer than CG’s; Bolen and Rylander
1978, Gawlik and Slack 1996, R.G.B., unpubl.
data). A longer bill may enable White-fronted
Geese to be more efficient at grubbing in the
ground, presumably giving them easier access
to resources buried in a frozen and dry sub-
strate. There is a spectrum of feeding tech-
niques used by geese to obtain food (Bolen
and Rylander 1978, Bellrose 1980, Prevett et
al. 1985, Ganter and Cooke 1996) but we still
know little of how morphology affects the rel-
ative efficiency of these techniques among
species.

2. White-fronted and Canada geese may
differ in their food use because their nutrient
requirements for reproduction differ. Protein,
fat, and calcium are the most important nutri-
ents required for egg formation (Robbins
1993, Alisauskas and Ankney 1992a). Geese
generally switch from a simple, carbohydrate-
rich diet before spring migration to a protein-
rich one during spring migration and egg pro-

duction (Mainguy and Thomas 1985, Prevett
et al. 1985, Budeau et al. 1991, Alisauskas
and Ankney 1992b, Bromley and Jarvis 1993,
Gauthier 1993), suggesting that protein is a
limiting nutrient to egg formation in geese
(Krapu and Reinecke 1992). Incubating fe-
males mainly require large fat reserves for
maintenance (Raveling 1979, Le Maho et al.
1981, Boismenu et al. 1992). Nutrient require-
ments for egg formation and incubation there-
fore differ, and whether a female should select
for protein or energy rich food during egg pro-
duction will depend on her initial nutrient re-
serves, and on how requirements are met
through food intake and reserve reallocation.

How females allot nutrients ingested during
preincubation to short (egg production) and
long term (incubation) requirements may be
reflected in their incubation behavior (Brom-
ley 1984, Thompson and Raveling 1987).
White-fronted Geese have cryptic nesting be-
havior, and like Emperor and Giant Canada
{B. c. maxima) geese, rely mostly on stored
nutrients for incubation (mean feeding time
per day = 1-8 min; Thompson and Raveling
1987; R.G.B., unpubl. data). In the central Ca-
nadian Arctic, Canada Geese nest openly, de-
fend their nests, and take frequent recesses to
feed (mean: 40 min per day; Jarvis and Brom-
ley, in press; R.G.B., unpubl. data). Little
feeding during incubation by White-fronted
Geese suggests a greater requirement for en-
ergy rich food prior to incubation. Longer in-
cubation recesses by Canada Geese imply a
greater reliance on foraging to meet nutrient
requirements during incubation, allowing for
more nutrients (e.g., protein) ingested prior to
incubation for egg production (see Thompson
and Raveling 1987). Thompson and Raveling
(1987) suggested that greater incubation atten-
tiveness may be related to larger body size in
geese, because they are vulnerable to different
types of predators and have a greater fasting
endurance (Calder 1974, Boismenu et al.
1992).

3. For Canada Geese, feeding pair density
may not entirely reflect the suitability of the
habitat for feeding during laying (Van Home
1983). Canada Geese typically nest on pond
margins (R.G.B., unpubl. data), and as laying
time approaches pairs may show increasing
tenitorial behavior in some patches of this
habitat. Intraspecific aggressive behavior was

178

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

observed, and could exclude conspecifics
from some patches of pond margins. In con-
trast, White-fronted Geese prefer hummock
habitat to nest in, and their territorial behavior
appears much weaker (R.G.B. and S.C., pers.
obs.).

Potential for competition. — Interspecific
competition for habitat and food in spring is
believed to be minimal in geese because spe-
cies usually segregate either spatially (differ-
ent ranges: Bellrose 1980, Ebbinge et al.
1982, Owen and Black 1990; different feeding
habitats: Fox et al. 1992) or temporally (dif-
ferent timing of breeding: Fox et al. 1992). We
documented differences in the feeding ecolo-
gy of two sympatric goose species that nest in
synchrony and differ mostly in bill moiphol-
ogy, slightly in body mass, and in incubation
behavior. We did not examine niche shifts
with changes in the relative densities of each
species during spring (Madsen and Mortensen
1987) because both species experienced sim-
ilar decreases in pair densities with snow melt.
Nonetheless, we showed that both species
overlapped in habitat use but each differed in
its preferred feeding technique and hence in
its selection of plant parts (“niche comple-
mentarity”; see Nudds 1992). On our study
area, potential for both intra- and interspecific
competition for food is highest before snow
melt, when food availability and diversity are
low, and pair density is high.

Interactions between nutrient requirements,
foraging ability and food profitabilities (hence
habitat selection: Stephens and Krebs 1986) in
goose species with similar body sizes but with
different bill morphologies wanant further
study.

ACKNOWLEDGMENTS

This study was funded by the Government of the
Northwest Territories, Department of Resources, Wild-
life and Economic Development; by a grant from the
National Science and Engineering Research Council of
Canada to G.G.; and by grants to S.C. from the Fonds
pour la Formation des Chercheurs et I’Aide a la Re-
cherche, Ministere de I'Education du Quebec; the Ca-
nadian Wildlife Foundation; the Canadian Wildlife
Service, Government of Canada; Indian and Northern
Affairs, Canada; Ducks Unlimited Canada, Yellow-
knife, NT; and the Centre d’Etudes Nordiques, Univer-
sitc Laval. Quebec. We thank Polar Continental Shelf
Project for transportation and logistical support, and
the Umingmaktok and Ekaluktutiak Hunters’ and Trap-
pers’ Associations for general support. Laboratory sup-

port was provided by the Departement de Biologic,
Universite Laval. We are grateful to C. Brodie, B.
Croft, R. DeKoster, M.-J. Hotte, Y. Richard, and D.
Wilson for their help in the field and to E. Thibeault,
for his help in the laboratory. We thank Drs. J. H.
Bedard, A. D. Fox, J. Huot, E Cooke, E. G. Cooch
and two anonymous reviewers for their comments on
the manuscript.

LITERATURE CITED

Alisauskas, R. T. and C. D. Ankney. 1992a. The cost
of egg laying and its relationship to nutrient re-
serves in waterfowl. Pp. 31-61 in Ecology and
management of breeding waterfowl (B. D. J. Batt,
A. Afton, M. G. Anderson, C. D. Ankney, D. H.
Johnson, J. A. Kaldec, and G. L. Krapu, Eds.).
Univ. of Minnesota Press, Minneapolis.
Alisauskas, R. T. and C. D. Ankney. 1992b. Spring
habitat use and diet of mid-continent adult Lesser
Snow Geese. J. Wildl. Manage. 56:43-54.
Altmann, j. 1974. Observational study of behavior:

sampling methods. Behaviour 49:227—267.
Ankney, C. D. 1977. The use of nutrient reserves by
breeding male Lesser Snow Geese. Can. J. Zool.
55:1984-1987.

Ankney, C. D. 1984. Nutrient reserve dynamics of
breeding and molting Brant. Auk 101:361-370.
Ankney, C. D. and C. D. MacInnes. 1978. Nutrient
reserves and reproductive performance of female
Lesser Snow Geese. Auk 95:459-471.

Barry, T. W. 1962. Effects of late seasons on Atlantic
Brant reproduction. J. Wildl. Manage. 26:19-26.
Bellrose, F. C. 1980. Ducks, geese, and swans of
North America. Stackpole Books, Harrisburg,
Pennsylvania.

Boismenu, C., G. Gauthier, and J. Larochelle. 1992.
Physiology of prolonged fasting in Greater Snow
Geese. Auk 109:51 1-521.

Bolen, E. G. and M. K. Rylander. 1978. Feeding
adaptations in the Lesser Snow Goose (Anser ca-
enilescens). Southwest. Nat. 23:158-161.

Bolton, M, P. Monaghan, and D. C. Houston. 1993.
Proximate determination of clutch size in Lesser
Black-backed Gulls: the roles of food supply and
body condition. Can. J. Zool. 71:273-279.
Bromley, R. G. 1984. The energetics of migration and
reproduction of Dusky Canada Geese (Branta
canadensis occidentalis). Ph.D. diss., Oregon
State Univ., Corvallis, Oregon.

Bromley, R. G., D. C. Heard, and B. Croft. 1995.
Visibility bias in aerial surveys relating to nest
success of arctic geese. J. Wildl. Manage. 59:364-
371.

Bromley, R. G. and R. L. Jarvis. 1993. The energet-
ics ol migration and reproduction of Dusky Can-
ada Geese. Condor 95:193-210.

Budeau, D. a., j. T. Ratti, and C. R. Ely. 1991.
Energy dynamics, foraging ecology, and behavior
of prenesting Greater White-fronted Geese. J.
Wildl. Manage. 55:556-563.

Calder, W. a. 1974. Consequences of body size for

Caniere el al. • SPRING FEEDING BY ARCTIC GEESE

179

avian energetics. Publ. Nullall Ornilhol. Club 15:
86-144.

Carriere, S. 1996. Ecologie de I’alimentation printan-
iere chez des Anserini {Auser albifrons et Brantci
canadensis) nichant en sympatrie dans I’Arctique.
Ph.D. diss., Univ. Laval, Quebec.

Choiniere, L. and G. Gauthier. 1995. Energetics of
reproduction in female and male Greater Snow
Geese. Oecologia 103:379-389.

CoocH, E. G., D. B. Lank, R. E Rockwell, and E
Cooke. 1991. Long-term decline in body size in
a Snow Goose population: evidence of environ-
mental degradation? J. Anim. Ecol. 60:483-496.

Cooke, E, C. S. Eindlay, and R. F. Rockwell. 1984.
Recruitment and the timing of reproduction in
Lesser Snow Geese {Chen caeritlescens caerules-
cens). Auk 101:451-458.

Dalhaug, L., I. M. Tombre, and K. E. Erikstad.
1996. Seasonal decline in clutch size of the Bar-
nacle Goose in Svalbard. Condor 98:42-47.

Daniel, W. W. 1978. Applied nonparametric statistics.
Houghton Mifflin Company, Boston, Massachu-
setts.

Drent, H. R. and S. Daan. 1980. The prudent parent:
energetic adjustment in avian breeding. Ardea 68:
225-252.

Dyke, A. S. and L. A. Dredge. 1989. Quaternary ge-
ology of the northwestern Canadian shield. Pp.
188-214 in Quaternary geology of Canada and
Greenland (R. J. Fulton, Ed.). Geological Survey
of Canada, Ottawa, Ontario.

Ebbinge, B. S., a. M. St. Joseph, P. Prokosch, and
B. Spaans. 1982. The importance of spring stag-
ing areas for Arctic-breeding geese, wintering in
western Europe. Aquila 89:249-258.

Findlay, C. S. and F. Cooke. 1982. Breeding syn-
chrony in the Lesser Snow Goose (Anser caeru-
lescens caendescens): I. Genetic and environmen-
tal component of hatch date variability and their
effects on hatch synchrony. Evolution 36:342—
351.

Eox, A. D., H. Boyd, and R. G. Bromley. 1995. Mu-
tual benefits of associations between breeding and
non-breeding White-fronted Geese Anser albi-
frons. Ibis 137:151-156.

Fox, A. D., H. Boyd, and S. M. Warren. 1992. Spa-
tial and temporal feeding segregation of two Ice-
landic goose species during the spring pre-nesting
period. Ecography 15:289-295.

Fox, A. D. AND J. Madsen. 1981. The pre-breeding
behaviour of the Greenland White-fronted Goose.
Wildfowl 32:48-54.

Ganter, B. and F. Cooke. 1996. Pre-incubation feed-
ing activities and energy budgets of Snow Geese:
can food on the breeding ground influence fecun-
dity? Oecologia 106:153-165.

Gauthier, G. 1993. Feeding ecology of nesting Great-
er Snow Geese. J. Wildl. Manage. 57:216—223.

Gauthier, G. and J. Tardif. 1991. Female feeding and
male vigilance during nesting in Greater Snow
Geese. Condor 93:701—71 1.

Gawlik, D. E. and R. D. Slack. 1996. Comparative
foraging behavior of sympatrie Snow Geese,
Greater White-fonted Geese, and Canada Geese
during the non-breeding season. Wilson Bull. 108:
154-159.

Hall, L. S., P. R. Krausman, and M. L. Morrison.
1997. The habitat concept and a plea for standard
terminology. Wildl. Soc. Bull. 25:173—182.

Holechek, j. L. and B. Gross. 1982. Training needed
for quantifying simulated diets from fragmented
range plants. J. Range Manage. 35:644-647.

Jandel Scientific Software. 1993. SigmaStat®* sta-
tistical software — user’s manual. Jandel Scientific
GmbH, Erkrath, Germany.

Jarvis, R. L. and R. G. Bromley. In press. Incubation
behaviour of Richardson’s Canada Geese (Branta
canadensis butchinsii) on Victoria Island, North-
west Territories, Canada. Can. Wildl. Serv. Occas.
Pap.

Johnson, M. K. 1982. Frequency sampling for micro-
scope analysis of botanical compositions. J. Range
Manage. 35:541-542.

King, J. R. 1973. Energetics of reproduction in birds.
Pp. 78-107 in Breeding biology of birds (D. S.
Farner, Ed.). National Academy of Science, Wash-
ington, D.C.

Krapu, G. L. and K. j. Reinecke. 1992. Eoraging ecol-
ogy and nutrition. Pp. 1-29 in Ecology and man-
agement of breeding waterfowl (B. D. J. Batt, A.
Afton, M. G. Anderson, C. D. Ankney, D. H.
Johnson, J. A. Kaldec, and G. L. Krapu, Eds.).
Univ. of Minnesota Press, Minneapolis.

LeMaho, Y., H. Vu Van Kha, H. Koubi, G. Dewas-
MES, J. Girard, P. Eerre, and M. Gagnard. 1981.
Body composition, energy expenditure, and plas-
ma metabolites in long-temi fasting geese. Am. J.
Physiol. 241:E342-354.

Lindholm, a, G. Gauthier, and A. Desrochers.
1994. Effects of hatch date and food supply on
gosling growth in Arctic-nesting Greater Snow
Geese. Condor 96:898-908.

Lindstrom, a. and T. Alerstam. 1992. Optimal fat
loads in migrating birds: a test of the time-mini-
mization hypothesis. Am. Nat. 140:477-491.

MacInnes, C. D., R. a. Davis, R. N. Jones, B. C.
Lieff, and a. j. Pakulak. 1974. Reproductive ef-
ficiency of McConnell River small Canada Geese.
J. Wildl. Manage. 38:686-707.

Madsen, J. and C. E. Mortensen. 1987. Habitat ex-
ploitation and interspecific competition of moult-
ing geese in east Greenland. Ibis 129:25-44.

Mainguy, S. K. and V. G. Thomas. 1985. Compari-
sons of body reserve buildup and use in several
groups of Canada Geese. Can. J. Zool. 63:1765-
1772.

Morse, D. H. 1990. Eood exploitation by birds: some
current problems and future goals. Stud. Avian
Biol. 13:134-143.

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.

180

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Newton, I. 1977. Timing and success of breeding in
tundra-nesting geese. Pp. 3—26 in Evolutionary
ecology (B. Stonehouse and C. Perrins, Eds.).
Univ. Park Press, Baltimore, Maryland.

Nudds, T. D. 1992. Patterns in breeding waterfowl
communities. Pp. 540—567 in Ecology and man-
agement of breeding waterfowl (B. D. J. Batt, A.
Afton, M. G. Anderson, C. D. Ankney, D. H.
Johnson, J. A. Kaldec, and G. L. Krapu, Eds.).
Univ. of Minnesota Press, Minneapolis.

Owen, M. 1981. Abdominal profile — a condition index
for wild geese in the field. J. Wildl. Manage. 45:
227-230.

Owen, M. and J. M. Black. 1990. Waterfowl ecology.

Chapman and Hall, New York.

Petersen, M. R. 1992. Reproductive ecology of Em-
peror Geese: annual and individual variation in
nesting. Condor 94:383-397.

Prevett, j. R, I. E Marshall, and V. G. Thomas.
1985. Spring foods of Snow and Canada geese at
James Bay. J. Wildl. Manage. 49:558-563.

Prop, J. and J. deVries. 1993. Impact of snow and
food conditions on the reproductive performance
of Barnacle Geese Branta leucopsis. Ornis Scand.
24:1 10-121.

Raveling, D. G. 1978. The timing of egg laying by
northern geese. Auk 95:294-303.

R.aveling, D. G. 1979. The annual cycle of body com-
position of Canada Geese with special reference
to control of reproduction. Auk 96:234-252.
Robbins, C. T. 1993. Wildlife feeding and nutrition.

Academic Press, San Diego, California.

Rohwer, E C. 1992. The evolution of reproductive
patterns in waterfowl. Pp. 486-539 in Ecology
and management of breeding waterfowl (B. D. J.
Batt, A. Afton, M. G. Anderson, C. D. Ankney,
D. H. Johnson, J. A. Kaldec, and G. L. Krapu,
Eds.). Univ. of Minnesota Press, Minneapolis,
Minnesota.

Rosenzweig, M. L. 1985. Some theoretical aspects of
habitat selection. Pp. 517-539 in Habitat selection
in birds (M. L. Cody, Ed.). Academic Press, Lon-
don, U.K.

Ryder, J. P. 1970. A possible factor in the evolution
of clutch size in Ross’s Geese. Wilson Bull. 82:
5-13.

Sedinger, j. S. and P. L. Flint. 1991. Growth rate is
negatively correlated with hatch date in Black
Brant. Ecology 72:496-502.

SoKAL, R. R. AND E J. Rohlf. 1981. Biometry. W. H.
Freeman and Company, New York.

Spaans, B., M. Stock, A. St Joseph, H.-H. Bergmann,
AND B. S. Ebbinge. 1993. Breeding biology of
dark-bellied Brant Geese Branla h. bernicla. Polar
Res. 12:117-130.

Stephens, D. W. and J. R. Krebs. 1986. Foraging the-
ory. Princeton Univ. Press, Princeton, New Jersey.

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

Thompson, S. C. and D. G. Raveling. 1987. Incuba-
tion behavior of Emperor Geese compared with
other geese: interactions of predation, body size,
and energetics. Auk 104:707-716.

Van Horne. B. 1983. Density as a misleading indi-
cator of habitat quality. J. Wildl. Manage. 47:893-
901.

WlELGOLASKl, E E., L. C. Bliss, j. Svoboda, and G.
Doyle. 1981. Primary production of tundra. Pp.
187-226 in Tundra ecosystems: a comparative
analysis (L. C. Bliss, O. W. Heal, and J. J. Moore,
Eds.). Cambridge Univ. Press, Cambridge. U.K.

Winkler, D. W. and J. R. Walters. 1983. The deter-
mination of clutch size in precocial birds. Curr.
Ornithol. 1:329-356.

Wypkkma, C. P. and C. D. Ankney. 1979. Nutrient
reserve dynamics of Lesser Snow Geese staging
at James Bay, Ontario. Can. J. Zool. 57:213-219.

Wilson Bull.. I I 1(2), 1999, pp. 181-187

A TEST OF THE CONDITION-BIAS HYPOTHESIS YIELDS
DIFFERENT RESULTS FOR TWO SPECIES OF
SPARROWHAWKS {ACCIPITER)

EDNA GORNEY,' WILLIAM S. CLARK,^ AND YORAM YOM-TOV'-^

ABSTRACT. — The determination of body condition of birds is important for many field studies. However,
when using trapping methods based on food as a lure, the sample of trapped birds could be biased toward
individuals in poor physical condition. We provide information on body mass, body condition, and sex and age
ratio of Levant Sparrowhawks (Accipiter hrevipes) and Eurasian Sparrowhawks (Accipiter nisns) caught in Elat,
southern Israel, during spring migration. We compared physical condition of birds trapped in baited traps to
physical condition of birds trapped in mist nets (no bait). The body mass and index of physical condition of
migrating Levant Sparrowhawks trapped in baited traps was lower than birds trapped in mist nets. By comparison
no differences were detected in body mass and condition index of migrating Eurasian Sparrowhawks caught by
the different trapping methods. The differences found in condition of Levant Sparrowhawks trapped with and
without food support the predictions of the condition-bias hypothesis; however, data from the Eurasian Sparrow-
hawk do not. The extent to which biases occur may be different even for closely related species. Received 1
Oct. 1998. accepted 7 Jan. 1999.

Using food to trap animals, especially pred-
ators, is probably the most widespread capture
method used by biologists. The capture of
birds of prey involves many methods, most of
which use small birds and rodents as lures
(Clark 1981, Bloom 1987). These trapping
methods are vulnerable to sampling bias be-
cause hungry birds in poor condition are more
likely to overcome their fear of entering traps
compared to birds in good condition (Weath-
erhead and Greenwood 1981). Considerable
support for this condition-bias hypothesis
comes from studies of songbirds (Weather-
head and Greenwood 1981, Dufour and
Weatherhead 1991) and ducks (Greenwood et
al. 1986, Hepp et al. 1986, Reinecke and
Shaiffer 1988, Conroy et al. 1989). We know
of no such studies conducted on birds of prey.

Two species of spaiTOwhawks occur regu-
larly during spring migration at Elat, Israel.
The Levant Spanowhawk {Accipiter brevipes)
breeds in Russia, eastern Europe, and the Bal-
kans. It migrates in large flocks to Africa in
September, returning in May (Cramp and Sim-
mons 1980). In autumn spanowhawks pass
over northern Israel between mid-September
and the beginning of October; up to 41,700
birds were counted during one autumn (Lesh-
em and Yom-Tov 1996). During spring they

' Dept, of Zoology, Tel Aviv Univ., Tel Aviv 69978,
Israel.

- 7800 Dassett Cl., Annandale. VA., 22003.

' Corresponding author; E-mail; yomlov@post.tau.ac.il

return through southern Israel between 20
April and the beginning of May, with up to
49,800 birds counted during one season (Clark
et al. 1986, Shirihai 1996). The Eurasian Spar-
rowhawk {Accipiter nisus) breeds across the
Palearctic region (Newton 1986). In contrast
to the Levant Spanowhawk only a few hun-
dred Eurasian Sparrowhawks pass through Is-
rael during autumn (Dovrat 1986, Leshem and
Yom-Tov 1996) and spring (Clark et al. 1986,
Shirihai 1996).

Physical condition during spring migration
is significant, not only for the successful
completion of the migratory journey, but also
for the timing of amval at the breeding
grounds and the condition of nutrient reserves
required for successful breeding. The breeding
success of birds in the temperate zone typi-
cally declines as the breeding season progress-
es. Body condition and amount of nutrient re-
serves have been shown to affect breeding
success in a variety of species (Korschgen
1977, Moss and Watson 1984, Newton 1986).
McLandress and Raveling (1981) demonstrat-
ed the importance of spring nutrient reserves
for migration as well as for reproduction and
their effect on clutch size in Canada Geese
{Branta canadensis). Price and coworkers
(1988) suggested that the ability to accumu-
late sufficient reserves required for breeding
may prevent the evolution of progressively
earlier breeding dates. Spring migrants anive
in Elat after crossing desert areas where feed-

181

182

THE WILSON BULLETIN • Vol. Ill, No. 2. June 1999

ing opportunities ai'e probably limited. Thus,
it is important to investigate the occunence of
condition biases which may go undetected and
lead to misinteipretation of data collected
from trapped birds (Weatherhead and Green-
wood 1981, Gomey and Yom-Tov 1994).

In conjunction with a spring migration rap-
tor banding project, we captured Eurasian and
Levant sparrowhawks using a variety of trap-
ping methods. Our objective was to test the
condition-bias hypothesis in birds of prey and
to compare the physical condition of spring
migrants to condition of pre-breeding spai-
rowhawks.

METHODS

Migrating sparrowhawks were captured as part of a
raptor banding project during five consecutive spring
migration seasons between March and May in 1984-
1988, in the agricultural areas just north of Elat
(29° 33' N 34° 55' E), Israel (Clark et al. 1986). Birds
were caught in mist nets, bal-chatri traps, and bow nets
(Bloom 1987) commonly used by hawk researchers in
North America and conforming to regulations applied
there. Only bal-chatri and bow-nets were placed on the
ground using food as a lure. Three-meter tall, 4 shelf
mist nets were placed in palm groves along rows of
trees used for roosting by sparrowhawks. The distance
between the lured traps and mist nets was generally at
least 500 m. Occasionally small birds were caught in
mist nets and they may have lured some sparrowhawks
into the nets. Such cases were excluded from the data
analysis when detected.

All birds were measured, weighed to the nearest I
g, and banded before release. Wing chord (straight line
from wrist to wing tip, without stretching or flattening
the wing) was measured with a ruler to the nearest I
mm. Culmen length was measured to 0.1 mm using
Manostat calipers. Age and sex were determined by
plumage, eye color, and size (Cramp and Simmons
1980). Age was determined as either adult (two years
old or more) or yearling (approximately one year old,
i.e., birds in their second calendar year).

We compared the body mass and body condition
index (body mass/wing chord X culmen length) of Le-
vant and Eurasian sparrowhawks that were caught in
mist nets to those that were caught in baited traps (bal-
chatri and bow-nets). Comparisons were conducted on
four groups; adult and yearling females, and adult and
yearling males. Sample sizes vary slightly because we
lack measurements for several individuals. This partic-
ular physical condition index was used because of its
high correlation with fat reserves in migrating Steppe
Buzzards (fiuteo huteo vulpinii.'i; Corney and Yom-Tov
1994). We found similar results however, using the
more commonly employed index of body mass divided
by wing length.

To test whether birds in poorer body condition arc
more likely to be attracted to baited traps than birds in

better condition, we compared for each age and sex
category the body mass and condition index of birds
trapped with baited traps and birds trapped in mist nets
(no bait). We also compared the number of individuals
from each age and sex class that were trapped in baited
and non-baited traps.

We used Abstat and Minitab 6.1.1 programs for
IBM compatable computer for statistical analysis. All
variables were tested for normality. When assumptions
were met we used separate SD /-tests. If normality
assumptions were not met we used Mann-Whitney
tests. All probabilities are two-tailed.

RESULTS

Levant Sparrowhawk. — Although Levant
Sparrowhawks pass through Elat during thi'ee
weeks, we captured 90% of all age and sex
categories within a mean of 9.4 (yearling
males) and 10.5 (adult females) days (means
calculated for 5 years of the study). Although
the passage for this species is concentrated
and rapid, adults were trapped earlier than
yearlings with median capture dates of 24
April for adult males, 25 April for adult fe-
males and 26 April for yearling males and fe-
males. We caught 425 Levant SpaiTowhawks,
404 of which we sexed and aged. Slightly
more yearlings were trapped (52%) than
adults (48%) at Elat. Sex ratio deviated from
the expected 1 : 1 ratio in favor of males in
both adults (61.1% males, 38.9% females; n
= 193; X' = 9.58, P = 0.002), and yearlings
(56.9% males, 43.1% females; n = 211; x' =
3.98, P = 0.046).

Body masses of adult males and females
were significantly larger than those of yeai-
lings (Table 1 ). We found no significant dif-
ferences between condition indices of adult
and yearling birds (Table 1). Thus, much of
the difference in mass appeared to be due to
the smaller size of yearling birds (Table 1).
We also found no significant associations
within age and sex groups between date of
migration and physical condition, nor did the
physical condition of trapped birds vary with
time of day.

Mean condition index of adult females was
significantly higher than condition index of
adult males (Mann-Whitney U = 1862, P =
0.001, two-tailed test). Similarly yearling fe-
males were in better condition than yearling
males (Mann-Whitney U = 1740, P = 0.001,
two-tailed test).

We found no significant differences be-

Gornex el at. • BODY CONDITION OF SPARROWHAWKS

183

crj

IT)

-o

—

• S

IT)

•3

r--

rri

•a

■o

•D

—

-C

•o

Cifj

'■j

01}

—

C/D

ITj

c/5

■o

r-*

■o

—

• ""

■o

-t

Tt-

Oii

■o

—

•o

—

OJ)

-i-

C>0

—

C/5

tT)

—

OX)

c/5

f—

—

OX)

OXJ

9—

■a

■o

C/0

>“ < >■

tween males and females (x" — 0.313, P >
0.05, df = 1 ) nor between adults and yearlings
(X“ = 0.024, P > 0.05, df = 1 ) in proportion
of birds trapped in baited and non-baited
traps. The mean body mass and mean condi-
tion index of adult males and yearling females
trapped in mist nets was higher than the mean
body mass and condition index of birds
trapped in baited traps (P < 0.05; Table 2).
Adult females and yearling males demonstrat-
ed no difference in mean condition index be-
tween both kinds of traps (Table 2).

Eurasian Sparrowhawks. — Most Eurasian
Spanowhawks (90%) pass through Elat in five
weeks. As with Levant Sparrowhawks, adults
were trapped before yearlings. Median capture
date was 14 April for adult males and females,
20 April for yearling females and 22 April for
yearling males. We trapped 72 Eurasian Spar-
rowhawks, with slightly more yearlings (53%)
than adults (47%). Sex ratio among adults (n
= 35; 43% males, 75% females), did not de-
viate from the expected 1:1 ratio (x' = 1.48,
P > 0.05), but did deviate among young birds
in favor of females (n = 37; 27% males, 73%
females; X“ “ 7.81, P = 0.005).

Mean body mass of adult females was sig-
nificantly larger than that of yearling females
(P < 0.05), and their wing chord was signif-
icantly longer (P < 0.01; Table 3). However,
the large body mass of adult females was not
due to size difference alone since their phys-
ical condition index was significantly higher
than that of yearling females (P < 0.05; Table
3). We also found no significant difference in
mean body mass or physical condition index
between adult and yearling males (Table 3).
However, the wing chord of adult males was
significantly longer than that of yearling males
(P < 0.01; Table 3). Females in both age
groups had significantly higher condition in-
dices than males (Mann-Whitney U = 13, P
< 0.001; U = 65, P == 0.04, for adults and
yearlings, respectively). No association was
found between index of physical condition
and between time of day and date of capture.

We found no significant differences in body
mass and condition indices between Eurasian
Spanowhawks trapped in lured traps and in
mist nets. Significantly more females (33 of
48) than males (6 of 18) were trapped in bait-
ed traps (x' = 12.3, P < 0.001, df = 1). A
similar proportion of adults (16 of 35) and

184

THE WILSON BULLETIN • VoL III, No. 2, June 1999

TABLE 2. Comparison of body mass and condition index of Levant Sparrowhawks caught in mist nets (no
lures) and in lured traps, Elat 1984-1988.

Mislnets

Lured traps

Comparison

Mean

Body mass

Adult females

213

219

0.74

>0.05

Adult males

175

167

2.06

0.043

Yearling females

208

197

2.42

0.019

Yearling males

162

157

1.14

<0.05

Mann-Whitney

VT P

Condition index

Adult females

0.065

0.007

0.065

0.007

1 142

>0.05

Adult males

0.064

0.008

0.061

0.006

2568

0.027

Yearling females

0.068

0.005

0.063

0.006

1261

0.008

Yearling males

0.062

0.007

0.060

0.006

2124

>0.05

yearlings (23 of 37) were trapped in baited
traps (x^ = 1.96; P > 0.05, df = 1).

DISCUSSION

Physical condition of migrants trapped with
food as a lure may be subject to biases
( Weatherhead and Greenwood 1981). One of
the prominent differences we found between
the two Accipiter species is that the condition
index of Levant Sparrowhawks captured with
and without food as a lure supports the con-
dition-bias hypothesis, but we found no evi-
dence for a condition-bias for Eurasian Spar-
rowhawks. Thus, the occurrence and extent of
a condition-bias may be different even for
closely related species of approximately the
same body size trapped during the same study
using the same traps. Although few data are
available to determine the cause of this dif-
ference, several aspects of their migration
strategies may be pertinent. Physical condition
of birds on the wintering grounds, and food
availability could differ for the two species. In
addition, they may differ in their tendencies
to hunt during migration. Levant SpaiTow-
hawks migrate mainly in large flocks whereas
Eurasian Sparrowhawks migrate singly
(Cramp and Simmons 1980, Shirihai 1996).
Levant Sparrowhawks might not regularly
hunt in areas like Elat where many birds
would be unlikely to And food. Therefore only
weak individuals would hunt and the species
would demonstrate a condition-bias. On the
other hand, all Eurasian Sparrowhawks may

hunt regularly as has been described from oth-
er migrating populations and species (Rude-
beck 1950, Cochran 1972, Newton 1986) and
thus show no evidence for a condition-bias.

When baited traps are used, age bias can
occur with more younger birds trapped than
older individuals (Nass 1964). Indeed, more
yearling than adult Steppe Buzzards were
trapped in Elat during the same years (Gomey
and Yom-Tov 1994). However, we detected no
age bias in the capturing of the two Accipiter
species.

Migrating Eurasian SpaiTowhawks do not
appear to store extra fat for migration. Moritz
and Vauk ( 1976) found that birds trapped dur-
ing autumn and spring migration at Helgoland
weighed the same as non-migrating individu-
als caught at the same time. Extra fattening is
probably unnecessary because spaiTowhawks
in some regions migrate with their prey spe-
cies, and thus can obtain food along the way
(Rudebeck 1950, Cochran 1972, Newton
1986). The spring migrants in our study
weighed considerably less than birds at other
times of year. The mean body masses of adult
Eurasian Sparrowhawks in our study were 33
g (females) and 14 g (males), lower than the
mean for European birds (Cramp and Sim-
mons 1980) and 20 g lower for both males
and lemales than birds from the former Soviet
Union (Dementiev 1966). The mass diffei ence
between migrants and non-migrants is not the
result of size differences between populations.

Gorney et al. • BODY CONDITION OF SPARROWHAWKS

185

t/5

—

•o

-o

f—

—

COj

'■o

c/2

IT)

0^)

1—

«—

“O

-o

—

Oi)

t/5

c/2

■o

•o

O')

■o

•o

oi:)

—

c/2

—

c/2

C/2

•

■o

Oij

c/2

zz.

-C

—

D-

“O

Mean wing length of adult Eurasian Sparrow-
hawks in our study was the same as reported
by Cramp and Simmons ( 1 980) and Demen-
tiev (1966). The lower body mass of birds
during migration likely reflects reserves used
for migration over deserts where chances for
feeding are probably minimal. A similar com-
parison of body mass of Levant Sparrow-
hawks during migration and at other times of
year was not possible because body mass data
are not available for non-migration periods.
Levant SpaiTowhawk’s wing chord data from
our study are similar to data of Cramp and
Simmons (1980) and Dementiev (1966), also
to data from a 1996 study conducted in Elat
(Clark and Yosef 1997).

Our finding that Eurasian Sparrowhawks
migrating during spring weigh less than birds
at other times of year is not unique to this
species. The mean mass of spring migrating
Steppe Buzzards in Elat was significantly low-
er than for the same species on their wintering
grounds in southern Africa (Gomey and Yom-
Tov 1994). Adult Eleonora’s Falcons {Falco
elenorae) during migration weighed 100 g
less than adults on their breeding grounds
(Cramp and Simmons 1980). Although data
were not available for Levant Spairowhawks,
it is likely that they also lose weight during
migration. In general birds are more likely to
demonstrate considerable mass declines dur-
ing spring migration in Elat because they must
cover large stretches of desert on their way
from Africa. Upon entering the Meditenanean
area north of Elat these species probably begin
replenishing their reserves (Yom-Tov and
Ben-Shahar 1995). Indeed, Levant SpaiTow-
hawks in autumn were not attracted to the
same food in bal-chatri traps that attracted
them during spring migration (E.G., pers.
obs.).

Adults migrate in spring before young birds
in most raptor species probably because of
benefits they receive from early airival to the
breeding tenitories (Newton 1979, Kerlinger
1989). Their earlier passage is undoubtedly
aided by better physical condition. In support
of this, several researchers found larger fat re-
serves in adults than in immature birds (Dunn
et al. 1988, Serie and Shai*p 1989, Alerstam
and Lindstrom 1990, Morton et al. 1990, Gor-
ney and Yom-Tov 1994; but see Alerstam and
Lindstrom 1990). Our data from this study of

186

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

two Accipiter species also demonstrate earlier
trapping dates of adult than yearling birds.
However, we found no difference in physical
condition of Levant Sparrowhawks between
adults and yearlings. In addition, Eurasian
Sparrowhawk adult females were in signifi-
cantly better condition than yearling females;
however, we found no such trend among
males.

ACKNOWLEDGMENTS

We thank K. Titus, B. Millsap and one anonymous
reviewer for their comments on a previous draft of this
paper. We thank Y. Leshem for his help and encour-
agement, B. Millsap for useful comments on the man-
uscript. and N. Paz for editorial comments. We are also
indebted to many tireless field workers of whom we
can mention but a few: M. Britain, K. Duffy, O. Hat-
zofe, Z. Labinger, J. Mason, M. McGrady, C. Mc-
Intyre, R. Parslow, T. Sbocbat, and C. Scbultz. Thanks
to Kibbutz Elot for their generous hospitality. This
study was funded by the Israel Raptor Information
Center of the Society for the Protection of Nature in
Israel, the Inter-University Ecological Fund of the Jew-
ish National Fund, and the Elat Ornithological Center.

LITERATURE CITED

Alkrstam, T. 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- Verlag, Berlin, Germany.
Bloom, P. H. 1987. Capturing and handling raptors.
Pp. 99-124 in Raptor management techniques
manual (B. A. Giron Pendelton, B. A. Millsap, K.
W. Cline and D. M. Bird, Eds.). National Wildlife
Federation, Washington, D.C.

Clark, W. S. 1981. A modified dho-gaza for use at a
raptor banding station. J. Wildl. Manage. 45:
1043-1044.

Clark, W. S., K. Duhfy, E. Gorney, M. McGrady,
AND C. SCHLfLTZ. 1986. Raptor ringing at Elat, Is-
rael. Sandgrouse 7:21-28.

Clark, W. S. and R. Yosef. 1997. Migrant Levant
Sparrowhawks (Accipiter hrevipes) at Elat, Israel:
measurements and timing. J. Raptor Res. 31:317—
320.

Cochran, W. W. 1972. A few days of the fall migra-
tion of a Sharp-shinned Hawk. Haw'k Chalk I I:
39-44.

Conroy, M. J.. G. R. Co.stanzo, and D. B. Stotts.
1989. Winter survival of female American Black
Ducks on the Atlantic coast. J. Wildl. Manage. 53:
99-109.

Cramp. S. and K. E. L. Simmons. 1980. The birds of
the western Palearctic Vol. 2. Oxford Univ. Press,
Oxford, U.K.

Dementiev, G. P. 1966. Birds of the Soviet Union Vol.
I. Israel Program for Scientific Translations. Je-
rusalem. Israel.

Dovrat, E. 1986. Kafr-Qasim cross Samaria raptor
migration survey, autumn. 1985. Torgos 5:28-60.

Dufour, K. W. and P. j. Weatherhead. 1991. A test
of the condition-bias hypothesis using Brown-
headed Cowbirds trapped during the breeding sea-
son. Can. J. Zool. 69:2686—2692.

Dunn, P. O., T. A. May, M. A. McCollough, and M.
A. Howe. 1988. Length of stay and fat content of
migrant Semipalmated Sandpipers in eastern
Maine. Condor 90:824-835.

Ginn, H. B. and D. S. Melville. 1983. Moult in birds.
British Trust for Ornithology, Hertfordshire, U.K.

Gorney, E. and Y. Yom-Tov. 1994. Fat, hydration,
and moult of Steppe Buzzards Buteo buteo vul-
pinus on spring migration. Ibis 136:185-192.

Greenwood, H., R. G. Clark, and P. J. Weather-
head. 1986. Condition bias of bunter-shot Mal-
lards. Can. J. Zool. 64:599-601.

Hepp, G. R., R. j. Blohm, R. E. Reynolds, J. E. Hines,
AND J. D. Nichols. 1986. Physiological condition
of autumn banded Mallards and its relationship to
hunting vulnerability. J. Wildl. Manage. 50:177-
183.

Kerlinger, P. 1989. Flight strategies of migrating
hawks. Univ. of Chicago Press, Chicago, Illinois.

Korschgen, C. E. 1977. Breeding stress of female ei-
ders in Maine. J. Wildl. Manage. 41:360-373.

Leshem, Y. and Y. Yom-Tov. 1996. The magnitude
and timing of migration by soaring raptors, peli-
cans and storks over Israel. Ibis 138:188-203.

McLandress and D. G. Raveling. 1981. Changes in
diet and body composition of Canada Geese be-
fore spring migration. Auk 98:65-79.

Moritz, D. and G. Vauk. 1976. Der Zug des Sperbers
(Accipiter nisiis) auf Helgoland. J. Ornithol. 117:
317-328.

Morton, J. M., R. L. Kirkpatrick, and M. R.
Vaughan. 1990. Changes in body composition of
American Black Ducks wintering at Chinco-
teague, Virginia. Condor 92:598-605.

Moss, R. and a. Watson. 1984. Maternal nutrition,
egg quality and breeding success of Scottish Ptar-
migan Lagopus niutus. Ibis 126:212-220.

Nass, R. D. 1964. Sex and age ratio bias of cannon-
netted geese. J. Wildl. Manage. 28:522-527.

Newton, I. 1979. Population ecology of raptors. T &
A D Poy.ser Ltd., Berkhamstcd, U.K.

Newton, I. 1986. The Sparrowhawk. T & A D Poyser
Ltd., Waterhouses, U.K.

Price, T, M. Kirkpatrick, and S. J. Arnold. 1988.
Directional selection and the evolution of breed-
ing date in birds. Science 240:798-799.

Reinecke, K. j. and C. W. Shaiffer. 1988. A field test
lor dillerences in condition among trapped and
shot Mallards. J. Wildl. Manage. 52:227-232.

Rudebeck, G. 1950. The choice of prey and modes of
hunting ol predatory birds with special reference
to their selective effect. Oikos 2:65-88.

Gonie\ et al. • BODY CONDITION OF SPARROWHAWKS

187

Serie, J. R. and D. E. Sharp. 1989. Body weight and
composition dynamics of fall migrating Canvas-
backs. J. Wildl. Manage. 53:431-441.

Shirihai, H. 1996. The birds of Israel. Academic Press
Ltd., London, U.K.

Weatherhead, P. j. and H. Greenwood. 1981. Age

and condition bias of decoy-trapped birds. J. Field
Ornithol. 52: 10-15.

Yom-Tov, Y. and R. Ben-Shahar. 1995. Seasonal
body mass and habitat selection of some migra-
tory passerines occurring in Israel. Israel J. Zool.
41:443-454.

Wilson Bull., 111(2), 1999, pp. 188-194

THE DEVELOPMENT OF A VOCAL THERMOREGULATORY
RESPONSE TO TEMPERATURE IN EMBRYOS OF THE

DOMESTIC CHICKEN

SHAWN C. BUGDEN' 2 AND ROGER M. EVANS'

ABSTRACT. — We examined the vocal responsiveness of chicken (Callus gallu.s) embryos at the pipped egg
stage to determine if they were able to regulate their thermal environment by soliciting heat from a surrogate
parent. There was no overall effect on vocalizations of exposure to 20° C or 45° C relative to the normal
incubation temperature of 37.8° C. There was, however, a general trend towards increased calling as the time of
hatching approached. There was also some indication that embryos tested in the late stages of hatching (ringing)
vocalized more in the cold, then became relatively silent when rewarmed. When cold-challenged embryos were
given 2 mm of rewamriing (surrogate brooding) in response to their calls body temperature was slightly but
significantly elevated above cold only exposed controls. Unlike previously reported anecdotal evidence sug-
gesting a strong vocal response to cold, our results suggest chicken embryos show only weak incipient vocal
response to temperature that begins to increase late in incubation and becomes fully functional only after hatch-
ing. Unlike other species tested to date, the developmental progression of behavioral and metabolic thermoreg-
ulation appear to be tightly linked in this species. Received 27 Feb. 1998, accepted 5 Oct. 1998.

The embryos of many avian species are ca-
pable of vocalizations prior to hatch (Freeman
and Vince 1974; reviewed in Evans 1988a).
Embryonic vocalizations may facilitate the
transition from incubating eggs to brooding
and feeding young (Impekoven 1973; Temple-
ton 1983; Evans 1988a, b), may function as
antecedents to post-hatching social behavior
(Tuculescu and Griswold 1983), or may func-
tion in soliciting care from parents. The em-
bryos of several species respond to cold by
increasing vocalizations (Evans 1990a; Bug-
den and Evans 1991; Evans et al. 1994, 1995;
Brua et al. 1996). These vocalizations affect
parental behavior and may elicit more atten-
tive incubation (Evans 1989, 1990b, 1992;
Brua 1996).

When cold-induced vocalizations trigger re-
warming by a surrogate parent in altricial pel-
icans and semiprecocial gulls, cold-challenged
embryos can vocally regulate their body tem-
peratures at relatively safe levels (Evans
1990a; Evans et al. 1994, 1995). In these spe-
cies such vocal behavioral thermoregulation
evidently precedes their ability to thermoreg-
ulate endothermically; altricial and semipre-
cocial species show no apparent metabolic re-
sponse to cooling before hatching (Matsunaga

' Dept, of Zoology, Univ. of Manitoba, Winnipeg,
MB R3T 2N2, Canada.

^ Current address: Mowbray Research Station,
McElroy Mouse, 645 Thornhill .St., Morden, MB R6M
IU4, Canada.

et al. 1989, Kuroda et al. 1990). However, in
precocial species there is evidence that grad-
ual cooling of late stage embryos brings about
a small but measurable incipient endothermic
response (Ereeman 1964, Tazawa et al. 1988,
Kuroda et al. 1990). The relationship between
the emergence of endothermy and the timing
of a vocal response to cold in precocial spe-
cies is unknown. We examined this issue in
the precocial domestic chicken (Gallus gal-
lus).

Chicken embryos are capable of vocaliza-
tions 2-3 days prior to hatch, but these vo-
calizations become more frequent during the
final 24 hours before hatching (Gottlieb and
Vandenbergh 1968, Dawes 1981, Tuculescu
and Griswold 1983). In response to the loud
calls (“distress calls”, Collias 1987) of em-
bryos, hens will vocalize or move around on
the nest (Tuculescu and Griswold 1983). Fol-
lowing the maternal response the embryos be-
come silent or emit soft trill calls (“pleasure
calls”, Collias 1987).

Study of the vocal response of chicken em-
bryos to cold has produced variable results.
Early investigation suggested that a pipped
egg that was alternately cooled and warmed
would give “di.stress” calls or “pleasure”
calls in close conespondence to the tempera-
ture changes (Collias 1952). Subsequent stud-
ies have shown decreased vocal activity (Op-
penheim and Levin 1974), increased vocal ac-
tivity (Evans 1988a), and inconsistent vocal

188

Buiiden cind Evans • VOCAL RESPONSE OF CHICKEN EMBRYOS

189

responses (Dawes 1981). Some evidence also
suggests that chicken embryos increase their
rate of vocalization during exposure to high
temperatures (Oppenheim and Levin 1974).

This study was designed to systematically
reassess the vocal response of late stage chick-
en embryos to temperature. We examined the
effects of temperatures both above and below
the normal incubation temperature, and the ef-
fectiveness of call-induced rewarming bouts
in the regulation of temperature during cold
challenge.

METHODS

White Leghorn Chicken eggs were incubated in a
forced air commercial poultry incubator (Perersime
Model no. 1 ) that maintained conditions within a suit-
able range (37.8 ± 0.5° C and 65 ± 5% relative hu-
midity). Only externally pipped eggs, which are known
to be capable of vocalizations (Tuculescu and Gris-
wold 1983), were selected for study.

Effects of continuous chilling and heating. — To ex-
amine the vocal response of embryos to low or high
temperature, pipped eggs were placed singly, pip hole
up, within an environmental chamber that consisted of
a coil of copper tubing surrounded by insulating Styro-
foam. The temperature in the chamber was controlled
by pumping water from controlled water baths
(±0.5° C) through the coil surrounding the egg. Testing
began with a 10 min pre-test at the control temperature
(37.8° C). The coil temperature was then changed to
the experimental (20° C or 45° C) or left at the control
temperature (37.8° C) for 30 minutes. This was fol-
lowed by a 10 min post-test period at 37.8° C.

The body temperature of the embryo was measured
with a thermocouple placed approximately I cm di-
rectly into the pip hole. The thermocouple was sur-
rounded by deep lying portions of the embryo’s body,
away from the outer shell. It was held in place by
porous adhesive tape (“Micropore”) applied to the ex-
terior of the shell. The body temperature of the embryo
and the coil temperature were recorded to the nearest
0.1° C every 30 s by a data logger [Grant Instrument
(Cambridge) Model 1203). Calls with a minimum in-
tensity of 78 dB (2.5 cm from the pip hole, B-fast
scale) were recorded by a microphone set in the plex-
iglass lid of the chamber, connected to a sound oper-
ated relay and an Esterline Angus event recorder.

Vocal regulation of temperature during cold chal-
lenge.— The apparatus was similar to that described
above except that calling of the embryo triggered a
period of rewarming (illustrated in Evans 1990a). The
embryo faced a continuous cold challenge at 20° C un-
til 5 calls were given. The fifth call then triggered a 2
min period of rewarming with water at 37.8° C being
pumped through the coil surrounding the egg. This pe-
riod of rewarming was followed by a return to default
chilling at 20° C until another bout of calling was ini-
tiated. If an embryo called in response to each succes-

sive period of cold challenge it would in elTect be ca-
pable of regulating ambient, and hence body temper-
ature (Evans 1990a, Bugden and Evans 1997). Control
embryos were placed in the same apparatus and held
at a constant 37.8° C throughout. Calls in the control
situation triggered a mock warming bout where the
same timer and pumps were activated as in the cold
challenge situation but the water circulating through
the coil remained at 37.8° C. All temperatures were
recorded as in the first experiment. An Esterline event
recorder recorded both individual calls and the warm-
ing and mock warming bouts. Control and cold chal-
lenge tests lasted for 1 hour on separate samples of
eggs.

To determine when during the pip-to-hatch interval
the voeal response to cold might develop, the timing
of pipping, hatching, and testing were reeorded at 4 h
intervals. Short term exposure of pipped eggs to mod-
erate cold can delay hatching in domestic chickens
(Evans 1990c). Testing of the eggs at 20° C in this
experiment thus could potentially affect the timing of
their hatching and so distort the interpretation of the
developmental onset of the vocal response to cold. To
control for this possibility, the pip-to-hatch intervals of
cold challenged and control embryos were compared
with a separate sample of embryos (untested control
embyros) that were not tested and left to hatch nor-
mally in the incubator. Statistical tests were done with
STATISTIX (version 4.1, Analytical Software, IBM
platform).

RESULTS

Effects of continuous chilling and heat-
ing.— Seven chicken embryos exposed to a 30
minute period of chilling at 20° C experienced
a fall in body temperature of 8.0 ± 0.3° C
(mean ± SE) from 36.9 ± 0.2° C at the start
to 28.8 ± 0.3° C at the end of the exposure
period. These embryos had a mean calling rate
of 10.2 ± 6.8 calls per minute. This result was
skewed by two highly vocal embryos that
were nearly hatched (ringing stage, Ereeman
and Vince 1973) by the end of the test. The
remaining five embryos were completely si-
lent during their exposure to cold, resulting in
a median call rate of 0.0 calls per minute. The
body temperature of embryos held at the con-
trol temperature of 37.8° C shifted by 0.5 ±
0.1° C, from 37.0 ± 0.2° C to 37.5 ± 0.2° C.
None of the 7 control embryos reached the
ringing stage and all were relatively quiet dur-
ing the test period with a mean calling rate of
0.1 ± 0.04 (median of 0.1) calls per minute.

Body temperature of seven embryos ex-
posed to 45° C rose by an average of 4. 1 ±
0.3° C, from 36.3 ± 0.2° C to 40.4 ± 0.4° C.
Call rate averaged 1.2 ± 0.7 calls per minute

190

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

with a median of 0.5 calls per minute. While
2 of these embryos were in the ringed stage,
their call rates (1.7 and 0.5 calls/min) were
similar to the overall average call rate for this
group. A Kruskal-Wallis one way nonpara-
metric ANOVA showed no overall differences ^
in calling rates of chicken embryos at the 3
three temperatures (H = 4.14, P > 0.05, df = m
2). b

Vocal regulation of temperature during 2
cold challenge. — Thirty-three embryos were <
tested under cold challenge experimental con- ^
ditions (20° C) and 34 were tested as warm- °
only (37.8° C) controls. There was no signif- ^
leant difference in the time from pipping to §
the time of hatching in cold challenged, con- ^
trol, and 12 untested control embryos (One <
way ANOVA: 7^2,76 = 1-34, P > 0.05). For ^
additional analyses embryos were grouped ac-
cording to time between testing and hatching.
There were 8, 13, and 12 embryos tested un-
der cold challenge and 13, 12, and 9 embryos
tested as warm only controls for three devel-
opmental categories 0-2.5, 2. 5-7. 5 and >7.5
h before hatching. The vocal response of cold
challenged embryos was greater than that of
the warm only controls in all developmental
categories (Fig. 1). However, a priori two-
sample t-tests between cold experimental and
warm controls showed no significant differ-
ence in any of the developmental categories
(0-2.5: t = 0.93, P > 0.05; 2.5-1. 5\ t = 0.44,

P > 0.05; >7.5: t = 1.63, P > 0.05).

Within 2.5 h of hatching, the mean number
of warming or mock warming bouts increased
in both the cold challenged embryos and in
the warm only controls (Fig. 1 ). Since cold
exposure produced no significant difference in
vocal response, all data were combined for
further comparison of vocal response with re-
spect to time before hatch. Embryos vocalized
significantly more frequently as the time of
hatching approached (Kj'uskal-Wallis ANO-
VA: H = 1 1.10, P < 0.01, df = 2).

The pattern of calling during the 2 min re-
warming and mock rewarming bouts provided
an additional measurement of the response to
temperature (Table 1 ). Because of a 6 s delay
of the chamber to temperature changes, cold
challenged embryos were still experiencing
temperatures well below the incubation/con-
trol temperature of 37.8° C at the start of each
rewarming period. Cold challenged embryos

LIG. 1. Mean (±SE) number of vocally generated
warming (cold challenged) and mock-warming (con-
trol) bouts at three pipped-egg developmental stages
(cold challenge n equals 8, 13, and 12; warm-only con-
trol n equals 13, 12 and 9 for 0-2.5, 2. 5-7. 5 and >7.5
h before hatching respectively).

in the two groups vocalized more often than
controls at this time, significantly so in the
2. 5-7. 5 h age group (Table 1). By the final
minute of rewarming, the cold challenged em-
bryos were almost silent, but this was also
true for most of the control embryos whose
temperatures had not changed. Control em-
bryos that were less than 2.5 hours from hatch
were an exception. These embryos vocalized
at a relatively high rate during the final minute
of the rewarming bout and maintained a call
rate significantly greater than the experimental
embryos.

Body temperatures maintained by experi-
mental embryos during vocal regulation tests
increased by about 1°C as hatching time ap-
proached, but this increase was not statistical-
ly significant (Fig. 1, Table 2). Body temper-
atures of experimentals were significantly
lower than those of control embryos held at
37.8° C (Table 2), reflecting the general low
level of call-induced rewarming periods in the
experimentals. Body temperature of the ex-
perimental embryos by the end of 30 min of
testing was significantly higher (29.9 ±

Bit}>den and Evans • VOCAL RESPONSE OF CHICKEN EMBRYOS

191

TABLE 1. Median number of calls per minute given by chicken embryos in pipped eggs at the start and
end of vocally-generated rewarming (experimental) and mock rewarming (control) bouts. Listed are medians of
calls given per embryo per bout (1st and 3rd quartiles in parentheses).

Stage
(h before
hatch)

Eirst 20 s

Final minute

Experimental

Control

Experimental

Control

<2.5

9.6 (6.1-1 1.8)

6.6 (2.0-17.4)

0.2 (0.0-2.2)

3.1 ( 1.0-8. 7)“

/i = 8

/; = 12

/; = 8

/; = 12

2.5-7.5

9.0 (4.7-13.1)

3.0 (0.0-6.9)^

0.0 (0.0-L4)

0.5 (0.0-2. 3)

n = 12

/; = 1 1

n = 12

/7 = 1 1

>7.5

3.0 (1.7-5. 3)

3.0 (0.0-6.0)

0.0 (0.0-0.5)

;t = 8

n = 1

/t = 8

n = 1

7.03

3.61

1.13

10.67

<0.05

>0.05

<0.01

^ Experimenlals and controls differ significantly (P < 0.05: Mann-Whitney U test).
H. Kruskal-Wallis ANOVA statistic, distributed as x".

0.4° C, all stages combined; t = 2.28, P <
0.05, df = 23) than embryos that were ex-
posed to constant chilling for 30 min in ex-
periment 1 (28.8 ± 0.3° C). This suggests a
slight warming effect of vocalizations in ex-
periment 2.

DISCUSSION

Exposure of chicken embryos to continuous
cold (20° C) and continuous hot (45° C) en-
vironments did not significantly increase their
rates of vocalization. The embryos remained
relatively silent in spite of exposure to envi-
ronmental temperatures that altered body tem-
perature to a level which, if continued, would
be expected to result in death of the embryo
(Webb 1987). While the embryos were clearly
capable of vocalizing, the close correspon-

dence of calling and temperature suggested in
the literature (e.g., Collias 1952) was not ev-
ident. Only two cold challenged embryos, in
the process of ringing prior to hatching,
showed a strong vocal response. The vocal
regulation experiments also showed an in-
crease in vocal response in chilled embryos
that were near to hatching (Fig. 1) but this
trend was also seen in control embryos that
were not exposed to cold. Tuculescu and Gris-
wold (1983) have also noted a general in-
crease in the rate of vocalization in the few
hours just prior to hatching.

Despite increased vocalization rates as
hatching approached, cold challenged chicken
embryos in the vocal regulation appaiatus
were not able to elevate their body tempera-
tures to safe, near normal incubation temper-

TABLE 2. Median body temperature of chicken embryos during experimental and control vocal regulation
test. Listed are medians of the average body temperature maintained during the one-hour test period (1st and
3rd quartiles in parentheses).

Body temperature of embryos (°C)

(h before hatch)

Experimental

Control

<2.5

31.5 (30.0-35.5)

37.8 (37.2-38.0)

<0.002

n = 1-'

n = 7“

2.5-1.5

30.5 (30.0-31.0)

38.1 (37.8-38.3)

<0.001

n = 13

/) = 12

>7.5

30.4 (29.4-30.1)

37.6 (36.9-38.1)

<0.001

n = 12

n = 9

2.79

4.61

>0.05

“ E.xcludes I experimenlal and 6 control embryos that displaced the thermocouple during ringing.
Comparison of experimentals and controls at each stage — Mann-Whitney U test.

H, Kruskal-Wallis ANOVA statistic, distributed as x~-

192

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

ature, at least under the conditions employed
here (Table 2). This contrasts strikingly with
a significant increase in vocal regulatory ca-
pability of hatched domestic chicks given 2
min periods of rewarming in response to cold-
induced calling on the day of hatching (Bug-
den and Evans 1997). The finding that body
temperature after 30 min of testing of vocal
regulation in 20° C cold challenged embryos
was significantly different than cold only ex-
posed embryos in experiment 1 is consistent
with the presence of some incipient effects of
vocally elicited rewarming bouts. Our results
thus raise the possibility that the marginally
higher rate of calling in cold challenged em-
bryos, especially as they neared hatching (Fig.
1 ), may represent the beginnings of a prehatch
vocal thermoregulatory response to cold.

Examination of the patterns of calling (Ta-
ble 1 ) suggests that immediately before hatch
(<2.5 h), rewarming was associated with de-
creased calling in experimental (cold chal-
lenged) embryos, while the calling of control
embryos during the final minute of the 2 min
mock rewarming period continued at a signif-
icantly higher level. This difference is also
suggestive of an incipient vocal response to
temperature by chick embryos during the final
hours before hatching. An increase in vocal
response to temperature as the time of hatch-
ing approaches has also been noted during vo-
cal regulation studies in semiprecocial Ring-
billed Gulls (iMrus delawarensis', Evans et al.
1994) and Herring Gulls {Lams argentatus\
Evans et al. 1995).

The subtle beginnings of the vocal response
to temperature found here parallels the incip-
ient development of endothermy in late stage
chicken embryos. The prehatching endother-
mic response is not robust. It has been noted
only when exposure to cold was limited to
gradual cooling of late stage embryos (Tazawa
et al. 1988). Incipient endothermy is thought
to be limited initially by conductance of O,
through the eggshell (Tazawa et al. 1989) and
then by the embryo’s limited endothermic
power (Tazawa et al. 1988, Whittow and Ta-
zawa 1991). While these physiological results
suggest that incipient endothermy occurs dur-
ing the latter stages of embryonic develop-
ment, the dramatic increase in oxygen con-
sumption at the time of hatching (Kuroda et
al. 1990, Whittow and Tazawa 1991) indicates

that endothermy becomes functional at that
time (Freeman 1971, 1983). Our present re-
sults suggest that there is a similar incipient
vocal thermoregulatory response of late stage
embryos to cold that becomes fully functional
at or soon after hatching (Tuculescu and Gris-
wold 1983, Espira and Evans 1996, Bugden
and Evans 1997). The developmental onset of
vocal and metabolic thermoregulation thus ap-
pear to be closely linked in precocial domestic
chicks.

In contrast to domestic chicks, the vocal
thermoregulatory system of altricial pelicans
and semiprecocial gulls is well developed pri-
or to hatch (Evans 1988a, 1990a; Evans et al.
1994, 1995). Altricial and semiprecocial spe-
cies evidently do not show any endothermic
response before hatch (Matsunaga et al. 1989,
Kuroda et al. 1990, Whittow and Tazawa
1991) suggesting that behavioral and meta-
bolic thermoregulation are developmentally
uncoupled in these species, unlike the appar-
ent linkage in the domestic chicken.

Pelicans and gulls both exhibit asynchro-
nous hatching, and later hatching eggs poten-
tially experience significant levels of incuba-
tion neglect as the parents attend to the chicks
that have already hatched (Evans 1990d, Lee
et al. 1993, Evans et al. 1995). Although
chickens display some level of hatching asyn-
chrony (mean of 15 h in Burmese Junglefowl;
Meijer and Siemers 1994), the chicken is not
known to neglect its eggs and will normally
remain on the nest for the first 12-24 h after
hatching (McBride et al. 1969, Miller 1978,
Meijer and Siemers 1994). A vocal response
to cold thus may not be a functionally useful
behavioral response for chicken embryos. Af-
ter hatching, the situation changes dramatical-
ly when mobile chicks are potentially exposed
to colder ambient temperatures, especially
during foraging bouts (McBride et al. 1969,
Sherry 1981). At that time calling to solicit
brooding warmth becomes an important part
of their behavioral response to cold (Kaufman
and Hinde 1961, McBride et al. 1969, Sheny
1981). Taken together, results to date suggest
that in precocial chickens, vocal and endo-
thermic thermoregulation both show incipient,
but largely nonfunctional, development prior
to hatching and are both turned on rapidly as
the chicks hatch and thermoregulation be-
comes a highly adaptive capability.

Bugden and Evans • VOCAL RESPONSE OF CHICKEN EMBRYOS

193

ACKNOWLEDGMENTS

This study was supported financially by an operating
grant to R.M.E. from the Natural Sciences and Engi-
neering Research Council, Ottawa, Canada. R. Wilson
and R. A. McArthur provided helpful comments on the
manuscript.

LITERATURE CITED

Brua, R. B. 1996. Impact of embryonic vocalizations
on the incubation behaviour of Eared Grebes. Be-
haviour 133:145-160.

Brua, R. B., G. L. Nuechtkrlein, and D. Buitron.
1996. Vocal response of Eared Grebe embryos to
egg cooling and egg turning. Auk 1 13:525-533.
Bugden, S. C. and R. M. Evans. 1991. Vocal respon-
siveness to chilling in embryonic and neonatal
American Coots. Wilson Bull. 103:712-717.
Bugden, S. C. and R. M. Evans. 1997. Vocal solici-
tation of heat as an integral component of the de-
veloping thermoregulatory system in young do-
mestic chickens. Can. J. Zool. 75:1949-1954.
COLLiAS, N. E. 1952. The development of social be-
havior in birds. Auk 69:127—159.

COLLiAS, N. E. 1987. The vocal repertoire of the Red
Junglefowl: a spectrographic classification and the
code of communication. Condor 89:510-524.
Dawes, C. M. 1981. The effects of cooling the egg on
the respiratory movements of the hatching fowl.
Callus g. domesticus, with a note on vocalization.
Comp. Biochem. Physiol. 68A:399— 404.

Espira, a. and R. M. Evans. 1996. Energy savings
from vocal regulation of ambient temperature by
3 day-old domestic chicks. Can. J. Zool. 74:599-
605.

Evans, R. M. 1988a. Embryonic vocalizations as care-
soliciting signals, with particular reference to the
American White Pelican. Proc. Int. Ornith. Congr.
19:1467-1475.

Evans, R. M. 1988b. Embryonic vocalizations and the
removal of foot webs from pipped eggs in the
American White Pelican. Condor 90:721—723.
Evans, R. M. 1989. Egg temperatures and parental be-
havior during the transition from incubation to
brooding in the American White Pelican. Auk
106:26-33.

Evans, R. M. 1990a. Vocal regulation of temperature
by avian embryos: a laboratory study with pipped
eggs of the American White Pelican. Anim. Be-
hav. 40:969-979.

Evans, R. M. 1990b. Embryonic fine tuning of pipped
egg temperature in the American White Pelican.
Anim. Behav. 40:963-968.

Evans, R. M. 1990c. Effects of low incubation tem-
peratures during the pipped egg stage on hatch-
ability and hatching limes in domestic chickens
and Ring-billed Gulls. Can. J. Zool. 68:836—840.
Evans, R. M. 1990d. Terminal egg neglect in the
American White Pelican. Wilson Bull. 102:684—
692.

Evans, R. M. 1992. Embryonic and neonatal vocal

elicitation of parental brooding and feeding re-
sponses in American While Pelicans. Anim. Be-
hav. 44:667—675.

Evans, R. M., A. Whitaker, and M. O. Wiehe. 1994.
Development of vocal regulation of temperature
by embryos in pipped eggs of Ring-billed Gulls.
Auk 1 1 1 :596-604.

Evans, R. M., M. O. Wiebe, S. C. Lee, and S. C.
Bugden. 1995. Embryonic and parental preferenc-
es for incubation temperature in Herring Gulls:
implications for parent-offspring conflict. Behav.
Ecol. Sociobiol. 36:17—23.

Freeman, B. M. 1964. The emergence of the homeo-
thermic metabolic response in the fowl {.Callus
domesticus). Comp. Biochem. Physiol. 13:413-
422.

Freeman, B. M. 1971. Body temperature and ther-
moregulation. Pp. 1115-1151 in Physiology and
biochemistry of the domestic fowl (D. M. Bell and
B. M. Freeman, Eds.). Academic Press, London.
U.K.

Freeman, B. M. 1983. Body temperature and ther-
moregulation. Pp. 365-377 in Physiology and bio-
chemistry of the domestic fowl. Vol. 4 (D. M. Bell
and B. M. Freeman, Eds.). Academic Press, Lon-
don, U.K.

Freeman, B. M. and M. A. Vince. 1974. Development
of the avian embryo. Chapman and Hall, London,
U.K.

Gottlieb, G. and J. G. Vandenbergh. 1968. Ontog-
eny of vocalization in duck and chick embryos. J.
Exp. Zool. 168:307-326.

Impekoven, M. 1973. The response of incubating
Laughing Gulls {Earns alricilla L.) to calls of
hatching chicks. Behaviour 46:94-1 13.

Kaufman, I. C. and R. A. Hinde. 1961. Factors influ-
encing distress calling in chicks, with special ref-
erence to temperature changes and social isola-
tion. Anim. Behav. 9:197-204.

Kuroda, O., C. Matsunaga, G. C. Whittow, and H.
Tazawa. 1990. Comparative metabolic responses
to prolonged cooling in precocial duck {Anas do-
mesticiis) and altricial pigeon {Coluinba doniesti-
ca) embryos. Comp. Biochem. Physiol. 95A:407—
410.

Lee, S. C., R. M. Evans, and S. C. Bugden. 1993.
Benign neglect of terminal eggs in Herring Gulls.
Condor 95:507-514.

Matsunaga, C., P. M. Matiiili, G. C. Whittow, and
H. Tazawa. 1989. Oxygen consumption of Brown
Noddy {Anous stolidu.s) embryos in a quasiequi-
librium stale at lowered ambient temperatures.
Comp. Biochem. Physiol. 93A:707— 710.

McBride, G., 1. P. Parer, and F. Foenander. 1969.
The social organization and behaviour of the feral
domestic fowl. Anim. Behav. Monogr. 2:127-181.

Meijer, T. and I. SiEMERS. 1994. Incubation develop-
ment and asynchronous hatching in junglefowl.
Behaviour 127:309-322.

Miller, D. B. 1978. Early parent-young interaction in

194

THE WILSON BULLETIN • Vol. Ill, No. 2. June 1999

Red Junglefowl: earlobe pecking. Condor 80:503-
504.

Oppenhein, R. W. and H. L. Levin. 1975. Short-term
changes in incubation temperature: behavioral and
physiological effects in the chick embryo from 6
to 20 days. Dev. Psychobiol. 8:103-115.

Sherry, D. E 1981. Parental care and the development
of thermoregulation in Red Junglefowl. Behaviour
76:250-279.

Tazawa, H., H. Wakayama, J. S. Turner, and C. V.
Paganelli. 1988. Metabolic compensation for
gradual cooling in developing chick embryos.
Comp. Biochem. Physiol. 89A: 125-129.

Tazawa, H., G. C. Whittow, J. S. Turner, and C. V.
Paganelli. 1989. Metabolic responses to gradual
cooling in chicken eggs treated with thiourea and
oxygen. Comp. Biochem. Physiol. 92A:619-622.
Templeton, R. K. 1983. Why do Herring Gull chicks
vocalize in the shell? Bird Study 30:73-74.
Tuculescu, R. a. and j. G. Griswold. 1983. Pre-
hatching interactions in domestic chickens. Anim.
Behav. 3 1 : 1-10.

Webb, D. R. 1987. Thermal tolerance of avian embryos:

a review. Condor 89:874-898.

Whittow, G. C. and H. Tazawa. 1991. The early de-
velopment of themioregulation in birds. Physiol.
Zool. 64:1371-1390.

Wilson Bull.. 111(2), 1999, pp. 195-209

BEHAVIOR AND VOCALIZATIONS OF THE CAURA AND THE

YAPACANA ANTBIRDS

KEVIN J, ZIMMER' 2

ABSTRACT. — The first detailed information on the vocalizations (including the first sound spectrograms) and
natural history of the Caura Antbird (Percnostola caurensis) and the Yapacana Antbird (Myrmeciza clLsjuncta)
are presented. The Caura Antbird was studied in the Serrania de la Cerbatana, edo. BoKvar, Venezuela, where
it inhabits humid foothill forest dominated by large rocks. Caura Antbirds specialized in foraging on or beneath
rocks, a behavior unusual among the Thamnophilidae. The Yapacana Antbird was studied at a site along the
south bank of the Rio Ventuari, edo. Amazonas, Venezuela. These antbirds were locally abundant in a specialized
stunted woodland that grows on white sand soils. Based on newly described vocal characters, the closest relatives
of P. caurensis appear to be P. leucostigrna and P. schistacea, whereas M. disjuncta has no apparent close
relatives and probably merits placement in a monotypic genus. Received 9 July 1998, accepted 5 Jan. 1999.

Among the least known members of the
large antbird family Thamnophilidae are the
Caura Antbird {Percnostola caurensis) and
the Yapacana Antbird {Myrmeciza disjuncta).
Both species are nearly endemic to south-
western Venezuela and have remained rela-
tively unobserved by modem field ornitholo-
gists. The most extensive collections of both
species (36 specimens of P. caurensis and 5
specimens of M. disjuncta) reside in the Co-
lecion Omitologia Phelps (COP), Caracas,
Venezuela. Single specimens of each species
collected near Pico Neblina (edo. Amazonas,
Venezuela) in 1984 by Field Museum of Nat-
ural History (FMNH) personnel represent the
only specimens of P. caurensis and M. dis-
juncta collected anywhere since 1972 and
1981 respectively. There is essentially no pub-
lished information on habitat or behavior of
the two species, and nothing is known of their
vocalizations (Ridgely and Tudor 1994).

In February 1998 I observed the habitats
and behaviors and tape-recorded the vocali-
zations of Caura Antbirds in the Senania de
la Cerbatana, edo. Bolivar, Venezuela, and of
Yapacana Antbirds in Yapacana National
Park, edo. Amazonas, Venezuela. This is the
first detailed information on the natural his-
tory and vocalizations of these species and al-
lows a more informed assessment of their pos-
sible generic affinities.

' Los Angeles County Museum of Natural History,
900 Exposition Blvd., Los Angeles, CA 90007.

-Correspondence address: 1665 Garcia Rd., Atas-
cadero, CA 93422; E-mail: kjzsrzC^tcsn.net

STUDY AREAS AND METHODS

I observed Caura Antbirds 10-15 February, 1998 in
the Serrania de la Cerbatana near Hato Las Nieves
(6° 34' N, 66° 12' W), edo. Bolivar, Venezuela (Fig.
1). The Serrania de la Cerbatana rings a large valley
vegetated mostly by a mixture of savanna and tropical
dry forest, transected by narrow bands of gallery forest
along occasional streams, and dotted with groves of
Mauritia palms in poorly drained areas. Three main
rivers, Cano Las Nieves, Rio Agua Fria, and Rio Dan-
ta, drain the valley. The Serrania mountain range rises
dramatically from the valley, and on the southern and
western borders is covered with humid forest. The
south facing slopes of the mountains to the north of
the valley are noticeably drier. The tallest peak in the
chain is Pico Las Nieves at 2080 m. I studied Caura
Antbirds along a 1.5 km trail that began in humid for-
est at 280 m elevation and extended up the side of a
ridge to 400 m.

I studied Yapacana Antbirds 22-27 February 1998
along a trail (hereafter the “Picua Trail”) near the set-
tlement of Picua (4° 5' N, 66° 45' W) in Yapacana Na-
tional Park, edo. Amazonas, Venezuela (Fig. 1). Picua
is a small settlement of Piaroa and Mako Indians, on
the south (left) bank of the Rfo Ventuari at about 150
m elevation. The surrounding area contains a mosaic
of different soil types that support a patchwork of dis-
tinct vegetation types. The Ventuari and its many small
tributaries are flanked by bands (of varying width) of
tall forest (15-25 m) that grow on yellow clay soils
and are seasonally flooded. This varzea forest is char-
acterized by a closed canopy with a fairly open un-
derstory and an abundance of vines and lianas. On
higher banks above the river the varzea grades into a
taller transitional forest with a denser understory.
Much of the area farther removed from the river is
dominated by white sand soils on which grow lower
stature woodlands and grassy savannas. The savannas
range in size from less than 1 ha to about 1 km-, with
scattered shrubs and small trees. Larger savannas in
the area contained stands of Mauritia palms. Large,
isolated patches of red clay soils support tall (>30 m).

195

196

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

tola caurensis) and Yapacana Antbird (Myrmeciza cHs-
juncta): open circles = confirmed sites for P. caurenxis
(confirmation based on specimen records); open
squares = confirmed sites for Myrmeciz.a disjuncta\
open star = site near the base of Pico de la Neblina
where both species have been collected in close prox-
imity. The black circle and #1 indicates the site of the
present study in the Serrania de la Cerbatana, edo. Bo-
livar, Venezuela where P. caurensis was studied. The
black square and #2 indicates the site of the present
study near Picua, in Yapacana National Park, edo.
Amazonas, Venezuela where M. disjuncta was studied.

lush “islands” of humid tropical forest that are not
seasonally flooded (= terra firme forest). Quartzite
dome-like sandstone hills (cerros) and low outcrop-
pings are scattered throughout the region.

Whenever individuals or pairs of P. caurensis or M.
di.sjuncta were located, 1 followed them for as long as
possible, tape-recording as many vocalizations as 1
could and summarizing foraging and other behaviors
on cassette tape. Some of these behaviors were also
documented on videotape. I used tape-playback of
these recordings to assess presence or absence of ant-
birds in places where no spontaneously vocalizing
birds were heard and to determine the limits of terri-
torial boundaries. All measurements included in such
summaries (height above ground, territory size, dis-
tances, times, etc.) are estimates. Terminology for for-
aging behavior follows Remsen and Robinson (1990).

Tape recordings were made with a Sony TCM-50()0
recorder and ,Sennheiser MKH-70 microphone. All re-
cordings have been or will be archived at the Library
of Natural Sounds. Cornell Univ., Ithaca, New York.
Spectrograms were made by Phyllis Isler on a Power
Macintosh 7500 computer using Canary version 1.2.1
(Bioacoustics Research Program, Cornell Laboratory
of Ornithology, Ithaca. New York). Morton Isler com-
piled a comprehensive list of distributional records of
the two species of antbirds as documented by speci-

mens, tape recordings, or photographs (Isler 1997; Pig.
1).

PERCNOSTOLA CAURENSIS

Distribution and habitat. — The Caura Ant-
bird is known only from the western portion
of the “Pantepui” region (Mayr and Phelps
1967) south of the Rfo Orinoco in the Vene-
zuelan states of Bolivar and Amazonas and in
extreme northern Brazil (Fig. 1). Percnostola
caurensis is well represented in museum col-
lections, with 36 specimens (the most recent
collected in 1972) in the Colecion Omitologia
Phelps (Caracas, Venezuela) alone (C. Rodner,
pers. comm.). The most recent substantiated
record was of a male collected in 1984 at 1250
m near the base of Pico Neblina (edo. Ama-
zonas, Venezuela; Willard et al. 1991). The
occurrence of P. caurensis in the Serrania de
la Cerbatana represents a slight range exten-
sion to the northwest (Fig. 1).

The forest along the first 500 m of the trail
was tall (ca 30 m), with an open understory
dominated by slender palms. The terrain was
flat and nearly devoid of large rocks. I could
locate only one territory of P. caurensis along
this portion of the trail, and it abutted the bot-
tom of the hill. The hillside forest beyond 500
m was also fairly open, with an intermittent
canopy of about 20 m. Few trees were larger
than 30 cm dbh, and woody vines were abun-
dant. Large stands of a naiTow-leaved, non-
spiny bamboo (1-2 m in height) occupied
most light gaps. The entire slope was extreme-
ly rocky, with numbers of boulders up to 8 m
tall and 15 m along their longest axis. These
boulders were typically moss and fern cov-
ered, with terrestrial bromeliads, cacti, and
bamboo growing over their tops and in the
crevices (Fig. 2). Many were topped with
small trees, the gnarled roots of which draped
off the sides of the rocks like tendrils, trapping
leaf litter and organic debris. I located 6 pairs
of P. caurensis along about 1 km of trail
through this rocky, hillside forest.

The SeiTania de la Cerbatana was extremely
dry in February 1998. Typical dry season con-
ditions appeared exacerbated by ongoing El
Nino related events. Leaf litter throughout the
forest was extremely dry and many trees had
shed large numbers of leaves. This was par-
ticularly evident in the hillside forest, where
large patches (0.5-1 ha) of deciduous vege-

Zimmer • CAURA AND YAPACANA ANTBIRDS

197

FIG. 2. Rocky hillside forest in the Serrania de la Cerbatana, edo. Bolivar, Venezuela. (A) Relatively open
forest, with a broken canopy of about 20 m. The rocks were 2—3 m tall and 4 ni in diameter. (B) A rocky
alluvial fan along which were located two Caura Antbird territories. Note the highly deciduous state of the
vegetation in this light gap, and the abundance of leaf litter trapped in the roots and vines overtopping the rocks.

tation were conspicuously scattered across the
slope, usually coincident with the rock strewn
alluvial fans at the bottom of ravines.

Morphology. — Soft-part colors were iden-
tical for both sexes. The iris was reddish-
brown, the legs and feet were slate gray (a
shade paler than the bill), and the bill was
blackish. Plumage was as described by Ridge-
ly and Tudor (1994).

Vocalizations. — Caura Antbirds were gen-
erally quiet during my fieldwork, as were
most other species of insectivorous birds.
Dawn choruses were both unremarkable and
short, suggesting a low level of breeding ac-
tivity for most species during the height of the
dry season. There was a sustained rain during
the early morning hours of 13 February; the
two following mornings I noted increased
spontaneous song from Caura Antbirds.

I recorded over 120 loudsongs (as defined
by Isler et al. 1997) and 900 calls from 12

individual antbirds. The loudsong of P. cau-
rensis is a far carrying series of 7-15 modu-
lated and well spaced notes (Fig. 3A). The
first notes are widely spaced, and the terminal
notes are closer together and drop in pitch.
Female songs (Fig. 3B) were similar in patterm
to male songs, but differed in other character-
istics such as mean number of notes, mean
frequency, etc. Females sang less frequently
than males.

In response to tape playback and during ter-
ritorial encounters, both sexes gave loud,
buzzy “zhew” calls at varying levels of fre-
quency modulation (Figs. 3C-E). On a few
occasions, birds involved in tenitoiial dis-
putes uttered sharp “quip” notes (Fig. 3F)
when neighboring birds approached closely.
The most frequently heard vocalization, and
one given by birds startled along the trail, was
an abbreviated loud rattle (Figs. 3G, H), sim-
ilar to the alar'm calls of P. leucostigma and

198

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

H 1

III

^ 1 1 — 1 1

1 1 ^

1 1 1 1

' I T I I 1 1 1—

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4,0 4 5

Time (seconds)

PIG. 3. Spectrograms ot Caura Antbird {Percnosloki caurensi.s) vocalizations: (A) male loudson'g, (B) female
loLidsong, (C, D, E) “zhew” calls, which probably function as aggression calls, at various levels of frequency
modulation, (E) “quip” call, (G) male alarm rattles, and (H) female alarm rattles.

several species of Myrmeciza anlbirds (pers.
obs.). A loud “chikit” (not tape recorded) was
often given by Caura Antbirds at the moment
they took flight. Birds alarmed by my pres-

ence gave this call immediately before flush-
ing, as did birds chasing one another about in
an aggressive boundary dispute.

Behavior. — Caura Antbirds were encoun-

Zimmer • CAURA AND YAPACANA ANTBIRDS

199

tered singly or in pairs, but did not associate
with mixed-species flocks. Mates foraging to-
gether were typically within 15 m of one an-
other, and alarm calls from one bird elicited
an immediate vocal response from the other.
Foraging birds maintained a neaily horizontal
posture, with the head held higher than the
axis of the body. All individuals continuously
raised their tails a few degrees above the plane
of the body and then wagged them slowly
downward in an arc 20-30° below the plane
of the body. This tail movement was most ex-
aggerated when the antbirds clung laterally to
elevated perches or responded to tape play-
back, but was also used during tenestrial for-
aging. Mostly the tail appeared to be slightly
fanned. Tail movements were occasionally ac-
companied by a simultaneous wing-flick.

Caura Antbirds foraged mostly on rocks or
the ground. Both sexes spent long times
creeping over the large boulders, often cling-
ing laterally to nearly vertical rock faces and
going in and out of the numerous crevices (of-
ten for minutes at a time) in the manner of
Slaty Bristlefronts (Meridaxis ater, M. Isler
and P. Isler, pers. comm.). Rock Wrens {Sal-
pinctes obsoletus) or Canyon Wrens {Cather-
pes mexicamis). While creeping about the
rocks, the antbirds frequently probed in the
mosses and small ferns covering the surface,
but spent most of their time inspecting the leaf
litter trapped between the roots and vine tan-
gles of trees overtopping the boulders. Arthro-
pod prey (primarily orthopterans and hemip-
terans, as well as many arthropods too small
to be identified) were gleaned from root and
vine surfaces with quick stabbing motions.
Curled dead leaves were carefully probed with
the bill. Some antbirds inspected dead leaves
without tossing them, others picked up leaves
with the bill before tossing them aside. The
antbirds routinely squeezed themselves into
small spaces between the rock surface and
overlying roots and vines, remaining in these
somewhat “canopied” niches to forage for up
to 60 s. When foraging on rocks, the antbirds
tended to spend most of their time in some-
what protected locations, within the interior of
vine tangles and root masses overtopping the
rocks, within crevices in the rocks, or beneath
rocky ledges and overhangs. The antbirds
moved steadily over open rock faces, pro-
gressing by short hops of 5—10 cm and oc-

casional longer wing-assisted hops or short,
abrupt flights, often to low, overhanging
branches or ledges. They frequently hopped
from a rock up to a low branch or sapling to
scan for 1-5 s before dropping back to the
rock. Birds often dropped 1-5 m from the
rocks to the ground. There, they hopped be-
neath the overhangs, probed in leaf litter and
inspected rock surfaces. Antbirds spent many
minutes inspecting fissures, crevices, and gaps
within and between rocks, often retracing their
routes.

One male observed foraging for more than
30 minutes spent the bulk of this time forag-
ing over, under, and between rocks, probing in
moss and vine tangles. On three occasions the
bird dropped to the ground at the base of large
rocks and spent 1-5 minutes vigorously toss-
ing dead leaves in the manner of a leaftosser
{Sclerurus spp.). Large leaves (many larger
than the bird) were picked up with the bill and
tossed. Smaller leaves were frequently flipped
by inserting the bill beneath the leaf and then
giving a quick upward flaking motion. He also
made an upward sally of 20-25 cm from the
ground to take an unidentified arthropod from
the underside of a green leaf. Other individ-
uals occasionally made similar short upward
sallies to glean prey from overhanging rocks.

One female antbird spent several minutes
hopping ar'ound the periphery of a large emer-
gent swarm of small, winged ants. Although
she picked at the ground several times, it was
not clear whether she was feeding on the ants.
The bird flew off and returned to the ant-
swarm twice. The male antbird was foraging
nearby, but did not attend the swarm.

Territories appeared to be about 150—200 m
in diameter. I witnessed only one territorial
conflict. This was a prolonged encounter be-
tween pairs whose territories bounded a large
rock-slide along the center of a ravine. The
conflict was marked by several advances and
retreats by both pairs. Both members of each
pair countersang at length while gradually ap-
proaching their counterparts. Whenever the
pairs approached to within about 20 m of one
another, they tended to substitute “zhew”
calls for songs. When they were in visual con-
tact or close auditory contact, the pairs ex-
changed harsh calls for a few minutes before
one individual or pair retreated, often with its
rival in close pursuit. These abrupt retreats

200

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

were always preceded by the loud, hard
“chikit” call. The dispute lasted for more than
30 minutes. Eventually the pairs retreated to
opposite ends of the rockslide, periodically
singing.

Most species of passerines besides P. cau-
rensis were encountered in mixed-species can-
opy and mid-level flocks. The apparent ab-
sence of terrestrial or semi-terrestrial antbirds
(other than P. caurensis) or fumariids is note-
worthy, although my failure to detect such
species could be an artifact of depressed song
activity during the dry season. The taller, less-
deciduous forest below the mountains had a
more diverse avifauna.

The presence of large rocks appeared to be
a critical component of P. caurensis habitat.
Of the 7 territories, 6 were in areas with abun-
dant large rocks. The seventh territory was
close enough to the bottom of the hillside to
possibly have included some rocky terrain as
well. The rocks provided microhabitats in
which a diversity of plants flourished, and
these plants, in turn, provided a wealth of po-
tential foraging strata for a small terrestrial in-
sectivore such as P. caurensis. Caura Antbirds
spent more than 80% of their foraging time
on rocks and overtopping vegetation, or on the
ground directly beneath overhanging rocks.
The importance of rocks as foraging sites for
Caura Antbirds may be increased during the
dry season, when favorable moisture-retaining
microclimates are created along rock edges or
crevices. Furthermore, the tendency for over-
topping root masses and vine tangles to trap
leaf litter is accentuated during the dry season,
when many trees drop their leaves. Such ac-
cumulations of organic litter may provide at-
tractive sites for arthropods when the forest is
water-stressed. It is interesting that no other
bird species was observed to exploit the extra
resource dimensions created by the large
rocks. The presence of Guianan Cock-of-Rock
(Rupicola rupicola) in the area is almost cer-
tainly dependent on the availability of large
rocks for nest sites (Snow 1982), but only the
Caura Antbirds seemed to use the rocks as
foraging substrates.

MYRMECfZA DISJUNCTA

Distrihution and habitat. — The Yapacana
Antbird was described in 1945 by H. Fried-
mann from two specimens collected near the

base of Cerro Yapacana (edo. Amazonas, Ven-
ezuela) in April 1931. Five more specimens
were collected from the same general locality
in April-May 1947 (specimens COP). Cerro
Yapacana is an isolated outlier of the western
Tepuis, rising steeply above the Rio Orinoco
to an elevation of 1 340 m. Meyer de Schauen-
see and Phelps (1978:218), perhaps describing
the general habitat surrounding Cerro Yapa-
cana, listed the habitat of M. disjuncta as
“High rain forest at about 100 m in under-
growth and low bushes.” Subsequent to its
description, M. disjuncta has been document-
ed from only three additional sites (Fig. 1). In
March 1981, while working in sandy-belt for-
est near Puerto Inirida, depto. Guainia, Co-
lombia, J. Dunning mist-netted an antbird lat-
er identified from photos as a female Yapa-
cana Antbird (Hilty and Brown 1986; ANJP
specimen 175723, R. Ridgely, pers. comm.;
photo on file at VIREO). In Februaiy 1984 a
single female M. disjuncta was collected at
140 m near the left bank of the Rio Baria on
the Venezuelan-Brazilian border (Willard et
al. 1991). This extended the known range of
M. disjuncta about 350 km south. There is no
published description of the habitat in which
this Yapacana Antbird was collected, but the
nearby base camp was in “tall seasonal rain
forest drained by both black-water and white-
water streams” (Willard et al. 1991). J. Coons
and D. Stejskal (pers. comm.) were the first to
find M. disjuncta near Picua in January 1997.
They reported seeing or hearing several indi-
viduals in savanna woodland on white sand
soils along the Picua Trail.

The only habitat of the Ventuari in which I
found M. disjuncta was what the local people
refer to as “monte cenado.” I found this hab-
itat only on the south bank of the Rio Ventuari
along the Picua trail. This is a stunted, virtu-
ally impenetrable woodland that grows on
fine, compacted white sand soils that are sea-
sonally saturated (Fig. 4). It is similar to the
“savanna woodland” described from Cam-
pamento Junglaven located farther north (up-
stream) along the Ventuari (Zimmer and Hilty
1997) but has a greater density of vines, along
with abundant sawgrass and bamboo scattered
through the understory. The canopy varies
from 6-10 m and is of generally uniform
height with only occasional emergent trees of

Zimmer • CAURA AND YAPACANA ANTBIRDS

201

FIG. 4. (A) Monte cerrado woodland along the Picua Trail, Yapacana National Park, edo. Amazonas, Ven-

ezuela. (B) Interior of the monte cerrado, depicting a typical Yapacana Antbird tenitory.

10-15 m. Few trees in this habitat have trunks
thicker than 10 cm dbh.

A different type of “sandy belt forest” or
savanna woodland occurred on the north bank.
This woodland was partly deciduous, less di-
verse, and even more stunted than the monte
cerrado. The understory was more open and
lacked both bamboo and sawgrass. This forest
grew on coarser, well-drained white sand soils
atop low ridges or rocky outcroppings. Myr-
meciza disjuncta and many other species typ-
ical of the monte cenado were absent from
this scrub woodland.

Using tape playback 1 located at least 24
pairs of M. disjuncta along 1 350 m of the Pi-
cua Trail. No birds were detected farther than
50 m from the trail. Tenitories were evenly
spaced along both sides of the trail and ap-
peared to be no more than 50—75 m in di-
ameter. Near the savanna edge the monte cer-
rado was particularly stunted, with a more
open canopy and more sawgrass in the under-
story. I found only two tenitories of Yapacana

Antbirds along more than 150 m of trail tran-
secting this more grassy woodland.

Morphology. — There has been some con-
fusion in the literature regarding plumage
characters of M. disjuncta. Central to the con-
fusion is the type specimen of M. disjuncta,
an immature male molting into adult plumage
(Friedmann 1945, 1948), which displayed a
combination of adult and immature plumage
characters. Thus, Meyer de Schauensee (1970:
249) wrote that the male has “uppeipaits
blackish-gray, crown and nape tinged brown”
and “sides of head gray, chin white; rest of
undeiparts white, strongly tinged ochraceous
buff; center of abdomen white.” Similarly,
Meyer de Schauensee and Phelps (1978:218)
described the male plumage as having the
“throat and breast white suffused with ochra-
ceous tawny, strongly so on breast and sides
of throat, middle of abdomen white, undertail
coverts dark gray.” They described the female
as differing from the male by “daik ochra-
ceous buff spots on wing coverts and ochra-

202

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

ceous under tail coverts.” This description
was essentially repeated in Hilty and Brown
(1986). Ridgely and Tudor (1994:332) were
much more nearly correct in both their illus-
tration of the male and in the description of
the plumage, but still suggested that the un-
derparts of the male are “tinged with creamy
buff.”

None of the 20 or more male Yapacana
Antbirds that I saw showed any hint of tawny,
ochraceous buff, or creamy buff color on the
underparts, nor did they have the crown or
nape tinged brown. Instead they were uni-
formly dark gray (with an almost steely blue
cast) above (except for the usually concealed
white interscapular patch) and on the sides of
the face, with the chin, throat, breast, belly,
and undertail coverts white. The sides and
flanks were washed with gray, but these areas
were often concealed by the wings. The wing
coverts were blackish, with the lesser and me-
dian wing coverts fringed white. The tail was
blackish and the rectrices were either narrowly
tipped or fringed white. This character is not
mentioned in any of the previous descriptions,
possibly because the narrow white tips/fringes
were lost to wear in the few existing speci-
mens. The legs were pinkish gray, the iris
blackish brown, the maxilla blackish, and the
mandible whitish. Females differed in having
a slightly brownish cast to the upperparts, par-
ticularly the crown and nape, which contrasted
with the gray sides of the face. The underparts
were a bright ochraceous buff (almost pale or-
ange) from the throat to the undertail coverts,
and were only slightly paler on the chin and
upper throat. I did not note pale tips or fringes
on the rectrices of females, nor could I con-
firm the presence of a contrastingly colored
interscapular patch.

All birds that I saw had proportionately
short, broad tails that appeared graduated.
This could have been influenced by molt, but,
if so, all of the individuals in this area were
highly synchronous in the regrowth of their
outer rectrices.

Vocalizations. — J. Coons and D. Stejskal
(pers. comm.) tape recorded some single-note
harsh calls of M. disjimcta during their Janu-
ary 1997 visit to Picua, but did not encounter
singing birds. Yapacana Antbirds were also
generally quiet during the period of my field
work. Even at dawn there was little sponta-

neous singing, suggesting that the level of
breeding activity was low. However, birds
were highly responsive to tape playback,
which often elicited singing from one or more
neighboring pairs in addition to the resident
pair. Both males and females responded vo-
cally to playbacks, although males were much
more aggressive in approaching the speaker.
All types of vocalizations that I recorded were
given by both sexes of Yapacana Antbirds. I
tape recorded over 350 loudsongs and 230
calls from 34 individuals.

The typical loudsong of M. disjimcta con-
sisted of two prolonged, harsh, heavily fre-
quency modulated notes separated by a short
pause, into which were inserted one or two
“pip” notes (Fig. 5A). The first harsh note
was the longest, and increased in intensity
while rising in pitch. The second harsh note
was shorter and had a more uniform ampli-
tude. A typical loudsong could be transcribed
as “cchhhhhhh pipizhhhh”. On many occa-
sions birds sang songs with no discernible
“pip” notes between the harsh elements, al-
though spectrograms of such songs reveal a
distinct spike at the beginning of the second
harsh note (Fig. 5B). Less frequently birds
sang songs with thi'ee harsh elements instead
of two (Figs. 5C, D). The third harsh note in
such series was usually the shortest. Loud-
songs varied in duration, depending largely on
the number of harsh elements included. I
could find no consistent differences between
male and female songs, although males more
frequently inserted two “pip” notes between
harsh elements and females more frequently
sang songs with no “pip” notes.

The most commonly heard calls were long,
harsh, single notes, at a somewhat higher pitch
than the harsh elements of the song, and with
a peculiar, slightly nasal quality (Figs. 5E, E).
These “harsh calls” were given by both sexes,
although those of females were higher
pitched. “Harsh calls” seemed to be aggres-
sion calls given in response to tape playback
or by a bird disturbed by my presence. Birds
occasionally gave a soft rattle (Eig. 5G) in a
similar context. When Yapacana Antbirds of
either sex were suddenly startled or strongly
agitated, they gave one or more sharp and
somewhat squeaky “squip” notes (Fig. 6A).
These notes were similar in tonal quality to
the notes inserted between the primaiy harsh

Frequency (kHz)

Zimmer • CAURA AND YAPACANA ANTBIRDS

203

5 -
4 -
3

2 -
1 -

‘HiV Jk r ' u

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

Time (seconds)

FIG. 5. Spectrograms of Yapacana Antbird (Myrmecizo disjuncta) vocalizations: (A) male loudsong with
two “pip” notes inserted between the primary harsh elements, (B) loudsong variant with no discernible “pip"
notes between harsh elements (sex of singing bird unknown), (C) loudsong variant with three harsh elements
and single “pip” notes (sex of singing bird unknown), (D) loudsong variant with three harsh elements and no
“pip” notes (sex of singing bird unknown), (E) male harsh call, probably an aggression call, (F) female harsh
call, probably an aggression call, and (G) soft rattle call, given in an aggressive context.

204

THE WILSON BULLETIN • Vol. J 11, No. 2, June 1999

FIG. 6. Spectrograms of Yapacana Antbird (Myrmeciza disjuncta) calls: (A) “squip” notes given by a
startled or alarmed bird, (B) rarely heard complex call (sex of calling bird unknown), and (C) variant complex
call (sex of calling bird unknown).

elements of most loudsongs. Occasionally
birds of either sex gave a more complex call
that began with several “squip” notes and
ended with a descending series of soft, whis-
tled “wheee” or “whew” notes (Figs. 6B, C).
These complex calls had a distinct tailing off
quality, as in “squip squip squip wheee wheee
wheee whew whew.” The function of these
calls was not clear.

Behavior and sociality. — Yapacana Ant-
birds were encountered singly or in pairs and
did not associate with mixed-species flocks.
Given the relatively small size of their terri-
tories, members of pairs were rarely far from
their mates. An alarm call or song from one
bird almost invariably brought an immediate
vocal response from its mate. I did not witness
any confrontations between neighboring pairs
of antbirds, although on several occasions a
singing pair of birds stimulated an adjacent
pair to approach the apparent boundary and
countersing for several minutes.

Yapacana Antbirds typically maintained a
horizontal posture, with the head held higher
than the axis of the body. Singing birds usu-
ally maintained a more upright posture. The
tail was held within a few degrees above or
below horizontal, and often was kept slightly
fanned. Foraging birds often quickly flicked
the tail up and down in a shallow arc of less
than 10°, but Just as frequently dipped the tail
slowly downward at about a 30° angle before
flicking it back up more rapidly. Singing birds
frequently shivered the tail up and down more
rapidly throughout a song. Some individuals
wagged their tail sideways in a slow, some-

what jerky manner. Such motions involved the
entire tail being swung a few degrees away
from the axis of the body, held briefly in that
position, then swung still further in the same
direction before being swung back into align-
ment. This jerky motion is similar to some of
the tail movements employed by the Silvered
Antbird {Sclateria naevia; pers. obs.). Wheth-
er foraging or singing, Yapacana Antbirds
flicked their wings at least once during virtu-
ally every pause between hops. Wing-flicks
occurred both independent of and in synchro-
ny with tail movements. Singing birds rarely
sang consecutive songs from the same perch.
In response to tape playback males often ex-
posed a white interscapular patch.

Yapacana Antbirds foraged mostly on or
neai" the ground, always lower than 1.5 m.
They were restless, active foragers, moving by
short hops (often wing assisted) and seldom
pausing for more than 2 s in one spot. When
moving above the ground they clung laterally
to slender vertical saplings or perched across
horizontal limbs and vines, progressed in an
often enatic, zigzag course, and frequently
moved up and down. They were adept at
clinging to the thinnest stems, including slen-
der bamboo stalks and vines. On a few oc-
casions I saw birds hop headfirst down nesuiy
vertical stems or branches, almost in the man-
ner of a nuthatch (Sitta spp.). Birds frequently
took several hops on the ground before jump-
ing up to a low perch and then back down to
the ground. Small arthropod prey were
gleaned from stems and from tops and bot-
toms of live leaves by reaching out, up, or

Zimmer • CAURA AND YAPACANA ANTBIRDS

205

down on extended legs and with neck craned.
Prey were captured with a quick stabbing mo-
tion of the bill and swallowed entire. Larger
prey items were bashed against the perch and
mandibled one or more times before being
swallowed. Antbirds typically wiped their bill
on the perch after swallowing prey. Most prey
items that could be identified were small or-
thopterans (katydids and crickets), hemipter-
ans, and geometrid larvae.

I encountered Yapacana Antbirds foraging
in the open along the main trail on only three
occasions. Two of these encounters involved
pairs, and the other involved a lone female
plumaged bird. In each case the birds were
working the edge of the dense monte cenado
vegetation, as well as shrubs and clumps of
grass growing in parts of the trail. All foraged
mostly on the ground, progressing by a series
of short hops with minimal pauses in between,
and always with wings (and frequently the
tail) flicking. The most frequent attack ma-
neuvers were gleans from the surface of the
leaf litter or brief probes with the bill beneath
the leaf litter. The next most frequent tech-
nique was reaching up to glean from the un-
dersides of overhanging green leaves and
grass blades. On several occasions birds
jumped 6-15 cm upward to glean prey from
the undersides of leaves. The two pairs of ant-
birds encountered in the open were found in
the early morning before the sun had illumi-
nated the trail. The lone female plumaged bird
(possibly a subadult male) was found in mid-
moming, when the entire trail was sunlit and
temperatures were already above 30° C. I fol-
lowed this bird as it foraged steadily at the
edge of the woodland for more than 20 min-
utes and covered more than 50 m. It crossed
at least one known territorial boundary, but
remained silent and did not attract attention
from any other antbird.

On one occasion I found a female Yapacana
Antbird attending a foraging swarm of army
ants {Eciton sp.) within the monte ceirado. I
observed this bird over 20 minutes during
which it was the only bird attending the
swarm. The female antbird employed two
strategies in the vicinity of the ants. Pait of
the time she scanned the swarm from perches
within 0.3 m of the ground, dropped to the
ground to seize fleeing arthropods (orthopter-
ans, hemipterans, and spiders) and then re-

turned to a low perch to beat the prey on a
branch before swallowing. Slightly more time
was spent hopping on the ground between the
columns of ants and tossing dead leaves in the
manner of a leaftosser (Sclerurus spp.). Most
leaves were tossed by inserting the bill be-
neath the leaf and lifting it with a quick flak-
ing motion. Occasionally the bird picked up a
leaf in its bill and tossed it aside. David Wolf
(pers. comm.) observed another female ant-
bird (away from ants) that remained in one
spot tossing leaves in a similar manner for 1-
2 minutes.

I observed no other species of terrestrial or
semi-terrestrial antbirds in the monte cerrado.
The Black- throated Antbird (Myrmeciza atro-
thorax) and the Black-chinned Antbird {Hy-
pocnemoides melanopogon), both of which
routinely forage below 1.5 m (Hilty and
Brown 1986; pers. obs.), were locally com-
mon in nearby forest or edge habitats, but
were not found in the monte cerrado. The only
other passerine (besides M. disjuncta) in this
habitat that I found foraging below 1 .5 m was
the Buff-breasted Wren (Thryothorus leuco-
tis), which foraged everywhere from the
ground to the canopy.

DISCUSSION

Habitat and conservation. — My field work
indicates that both the Caura Antbird and the
Yapacana Antbird are habitat specialists, oc-
cuiTing in subtypes of more widely distributed
macrohabitats. Percnostola caurensis has
been recorded over a broad elevational range,
100-1300 m (Meyer de Schauensee and
Phelps 1978). In the Pantepui region this
range of elevations often spans the distance
between tall, seasonal humid forest and elfin
cloud forest. An antbird that occurs across
such a spectrum of habitats might normally be
considered an ecological generalist. However,
if the critical ecological factor determining its
distribution is the presence of lai'ge rocks
within forest regardless of elevation, then P.
caurensis is very much a specialist. This
could, in part, account for the absence of the
species from so many seemingly suitable low-
land sites in Bolivai' and Amazonas (Campa-
mento Junglaven: Zimmer and Hilty 1997;
Brazo Casiquiare: Paynter 1982).

Percnostola caurensis may have evolved as
something of a rock-specialist to occupy a

206

THE WILSON BULLETIN • Vol. HI, No. 2, June 1999

niche that is locally abundant in parts of the
highly eroded Guianan Shield. Mayr and
Phelps (1967:277) described the tepuis (table
top mountains) of this region: “many are ac-
tually strongly dissected and strewn with iso-
lated blocks, some more than 100 m high, and
with a large variety of other rock forms.”
Most Caura Antbirds have been collected
from the slopes of tepuis (Zimmer and Phelps
1947, Phelps and Phelps 1963, Meyer de
Schauensee and Phelps 1978). The most re-
cently collected specimen, from 1250 m at the
base of Pico Maguire (edo. Amazonas, Ven-
ezuela), was from a site described as “A
hanging valley in dense cloud forest with
moderately tall trees, and rocky forest floor
covered with thick moss” (Willard et al. 1991:
6). The apparent absence of P. caurensis from
much of the eastern portion of the tepui region
of Bolivar is a mystery.

Myrmeciza disjimcta appeal's to be restrict-
ed to woodlands growing on white sand soils.
White sand habitats are widely but patchily
distributed throughout Amazonia, with their
center of distribution in the upper Rfo Negro
region (Pires 1974, Stotz et al. 1996). White
sand soils support many different types of
vegetation, from scrub to tall forest (Anderson
1981). Several distinctly different types of
vegetation were found growing on white sand
soils in the Picua region but I found M. dis-
jimcta only in the monte cerrado. Similarly,
surveys of two nearby sites in Amazonas
(Campamento Junglaven and Pto. Ayacucho)
have failed to record M. disjimcta in spite of
the prevalence of white sand woodlands (Zim-
mer and Hilty 1997). The Yapacana Antbird
may therefore be restricted only to a particular
type of white sand woodland, the monte cer-
rado. This would indicate an even patchier
distribution than previously suspected and
would help explain how this species has es-
caped detection for so long.

Both the Caura Antbird and the Yapacana
Antbird appear to be locally common within
their prefeired habitats. The distributions of
both species are centered in the lowlands or
foothills of Amazonas and western Bolivar,
which are among the least populated regions
in Amazonia. Thus, neither species is under
immediate threat of extinction. However, their
patchy distributions and apparent restriction to
particular microhabitats make them more vul-

nerable than most other birds of the Guianan
lowlands. Stotz and co-workers (1996:4) have
noted that “. . . the first major waves of ex-
tinctions in the Neotropics are not occurring
in centers of diversity such as the Amazon.
Rather, extinctions are occuiring within cen-
ters of local endemism, especially among spe-
cies that have evolved ecological specializa-
tions that limit their ability to adapt to human
modifications of their habitats.”

Before the conservation threats facing P.
caurensis and M. disjimcta can be adequately
assessed, we must first confirm their depen-
dence on or preference for the microhabitats
in which I found them, then attempt to quan-
tify just how much appropriate habitat exists
within their ranges. More comprehensive sur-
vey work within the region is cleaidy needed.
Continued protection of existing parks or re-
serves, such as Yapacana National Paik, is vi-
tal, pai'ticularly with regaid to threats posed
by illegal gold-mining.

Intrafamilial relationships. — The Caura
Antbird was described by Hellmayr (1906)
and placed in the genus Sclateria. The Caura
Antbird was subsequently transfened to Schis-
tocichla (Zimmer and Phelps 1947), which
was later subsumed into Percnostola without
elaboration by Peters (1951). Subsequent au-
thors (e.g., Meyer de Schauensee 1966, 1970;
Sibley and Mom'oe 1990; Monroe and Sibley
1993) have continued this treatment, recog-
nizing five species: P. rufifrons, P. schistacea,
P. leiicostigma, P. caurensis, and P. lophotes.
Ridgely and Tudor (1994 unected the ge-
nus Schistocichla for schista a, leiicostigma,
and caurensis on the basis of their rounder,
uncrested heads (crested in P. rufifrons and P.
lophotes) and spotted rather than fringed wing
coverts.

On purely moiphological grounds, P. cau-
rensis, P. leiicostigma, and P. schistacea
would appear to comprise a natural grouping.
Plumage differences between P. caurensis and
P. leiicostigma are especially subtle, with size
and soft part coloration being the most im-
portant field characters for visually distin-
guishing the two species (pers. obs.). Vocal
similarities are less apparent, in part because
of pronounced geographic variation in the vo-
calizations of the various named subspecies of
P. leiicostigma (pers. obs.). Indeed, some vo-
cal differences within the P. leiicostigma com-

Zimmer • CAURA AND YAPACANA ANTBIRDS

207

plex are as great as the between species dif-
ferences in the Schistocichla group. Given
this, I feel that resolution of the intrageneric
relationships of the five species cunently in-
cluded in Percnostola should await molecular
comparisons, as well as a closer evaluation of
vocal and morphological differences as they
relate to the P. leucostigma group.

Almost since its description, there has been
speculation regarding the placement of the Ya-
pacana Antbird in Myrrneciza. Friedmann
(1948:478) offered that “the species is not too
distantly related to Myrrneciza atrothorax but
is clearly specifically distinct from that form.”
He went on to note that Zimmer had examined
the type and the paratype and had pointed out
that “the general plumage has about the tex-
ture of Cercornacra carbonaria” (Friedmann
1948:478). Friedmann (1948) further noted
Zimmer’s suggestion that a fully adult male
M. disjimcta might show a closer relationship
to Cercornacra than was suggested by the type
specimen. He also commented that “The pat-
tern of the markings of the upperwing coverts
is very like that of some forms of Cercornacra
(serva for example), but the bill is that of
Myrrneciza” (Friedmann 1948:478). Peters
(1951) alluded to the seemingly polyphyletic
nature of Myrrneciza as he defined it and made
several recommendations for the placement of
various species, but did not mention M. dis-
jimcta. Ridgely and Tudor (1994:333) also
noted the heterogeneous nature of Myrrneciza
and suggested specifically that M. disjuncta
may not belong in the genus and “perhaps is
more closely allied to Sclateria” . They also
seemed to suggest somewhat indirectly, that
M. disjuncta was or should be included in the
formerly recognized genus Myrmoderus.
However, I can find no evidence that M. dis-
juncta was included in the various shifts of
species between Myrrneciza and Myrmoderus
by Hellmayr (in Cory and Hellmayr 1924),
Todd (1927), or Peters (1951).

Morphological, vocal, and behavioral char-
acters offer contradictory clues to the possible
generic affiliations of the Yapacana Antbird.
Myrmoderus is not currently recognized, but
both Todd ( 1 927) and Peters (1951) advocated
that it be reserved for [Myrrneciza] loricata
and squamosa. The latter are clearly sibling
species that share several distinctive morpho-
logical, vocal, and behavioral characters, and

they are distant in all respects from M. dis-
juncta (pers. obs.). In some moiphological re-
spects, M. disjuncta is reminiscent of the
monotypic Sclateria, as suggested by Ridgely
and Tudor (1994), but it has fringed rather
than spotted wing coverts, white tail tips, an
interscapular patch, and differs greatly in both
vocal and behavioral characters. The song of
M. disjuncta is mildly reminiscent of that of
various members of the Cercornacra nigricans
group (as defined by Fitzpatrick and Willard
1990), and the plumage pattern of males,
lai'gely gray with white-fringed wingbars and
white tail-tips, and females ochraceous below,
fits several members of the C. tyrannina
group. However, Cercornacra antbirds tend to
be slender and proportionately long-tailed
(Ridgely and Tudor 1994), whereas M. dis-
juncta is relatively compact and short-tailed.
No Cercornacra approaches the white under-
parts of male M. disjuncta, the genus as a
whole has gray or blackish underparts. More
importantly, male-female antiphonal duets are
an important component of the vocal reper-
toires of virtually all species of Cercornacra
(Zimmer et al. 1997), but are not found in M.
disjuncta.

In size, proportions, and some aspects of
plumage Myrrneciza disjuncta is somewhat
suggestive of Hypocnemoides. However, nei-
ther species of Hypocnemoides is nearly as
sexually dimoiphic as is M. disjuncta, and
they lack any suggestion of the ochraceous
coloration found in female M. disjuncta. Myr-
meciza disjuncta lacks the black throat and
pale eye found in Hypocnemoides. Vocal dif-
ferences between M. disjuncta and Hypocne-
moides are much greater than the moipholog-
ical differences (pers. obs.). By themselves,
the two species of Hypocnemoides form a nat-
ural grouping, with great similarities in plum-
age, voice, foraging behavior, and habitat use.
Almost none of these characters are shared
with M. disjuncta.

I feel that Myrrneciza as cunently con-
structed is paraphyletic, with various sub-
groupings that do not appear to be closely al-
lied on the basis of morphological, vocal, or
behavioral characters (e.g., Peters 1951,
Ridgely and Tudor 1994). None of these sub-
groups is a good fit for M. disjuncta. Myr-
meciza atrothorax has been suggested as a
close relative of M. disjuncta by Friedmann

208

THE WILSON BULLETIN

VoL 111, No. 2, June 1999

(1948), but the males of the two species differ
dramatically in plumage, and the two species
share no vocal similarities that I can detect.
No compelling morphological similarities
clearly ally M. disjuncta with any other Myr-
meciza, nor with any other antbird. Similarly,
I have compared songs and calls of M. dis-
juncta to the other 1 8 species currently placed
in Myrmeciza (Isler and Whitney 1999; Zim-
mer, unpubl. data) and can find nothing to sug-
gest a close relationship between M. disjuncta
and any of the other species. Relationships
suggested by one or two morphological char-
acters in one sex are contradicted by morpho-
logical characters in the other sex, vocal char-
acters, behavioral characters, or by some com-
bination of the three. Although past descrip-
tions of genera have been based largely on
plumage characters which may or may not
have phylogenetic relevance, the addition of
vocal and behavioral data adds important ev-
idence in redefining these relationships (e.g,
comments in Remsen 1997, Remsen and
Schulenberg 1997). In the absence of a mo-
lecular based phylogeny the most conserva-
tive approach would be to leave M. disjuncta
where it is, as yet another poor fit in a genus
understood to be heterogeneous. However, in
my opinion, the sum of morphological, vocal,
and behavioral evidence would suggest that
the Yapacana Antbird is monotypic, deserving
of its own genus.

ACKNOWLEDGMENTS

Primary thanks go to M. and P. Isler for generously
lending their time and talents to the many figures in
this paper. Phyllis produced the many spectrograms af-
ter reviewing several hours of my tape recordings, and
Mort compiled the list of distributional records and
produced the map used in Eigure 1. The Islers made
many helpful comments on an early draft of this man-
uscript and have been an invaluable source of advice
and ideas with all aspects of my research on antbirds.
M. Lentino, C. Rodner, and R. Restall of the Colecion
Ornitoldgica Phelps generously supplied important
collection information. T. Schulenberg provided valu-
able advice as well as access to several references. J.
V. Remsen reviewed an early draft of the manuscript
and made numerous helpful comments that improved
it. 1 am most grateful to J. Coons for sharing infor-
mation on his experiences in the Picua region, partic-
ularly for bringing Alechiven Lodge to my attention.
D. Cooper was the first to report the probable presence
of Caura Antbirds in the Serrania dc la Cerbetana and
his advice was helpful in planning my trip to that re-
gion. Thanks also to the staffs of Hato Las Nieves and

Alechiven Lodge, my two bases of operation for this
research. L. Oleaga was of particular help in cutting
trails and facilitating my work at Hato Las Nieves. C.
Herrera, M. Ruiz and D. Eorbes of Caracas provided
invaluable assistance with trip logistics.

Special thanks must go to D. Wolf, who accompa-
nied me on my trip to Yapacana National Park. Dave
shared in the excitement of finding the Yapacana Ant-
birds and taping their vocalizations; his good humor
and knowledgeable insights made the heat and the bit-
ing jejenes more tolerable. Pinal thanks go to V. Eman-
uel and Victor Emanuel Nature Tours, Inc., for provid-
ing me with travel opportunities that made much of
my research possible.

LITERATURE CITED

Anderson, A. B. 1981. White-sand vegetation of Bra-
zilian Amazonia. Biotropica 13:199—210.

Cory, C. B. and C. E. Hellmayr. 1924. Catalogue of
birds of the Americas. Pteroptochidae, Conopo-
phagidae, Pormicariidae. Pield Mus. Nat. Hist.
(Zool. Series) 13(3): 1-369.

PiTZPATRiCK, J. W. AND D. E. WiLLARD. 1990. Ccrcom-
acra manu, a new species of antbird from south-
western Amazonia. Auk 107:239-245.
pRiEDMANN, H. 1945. A new ant-thrush from Venezue-
la. Proc. Biol. Soc. Wash. 58:83-84.
pRiEDMANN, H. 1948. Birds collected by the National
Geographic Society’s expeditions to northern Bra-
zil and southern Venezuela. Proc. U.S. Nat. Mus.
97:373-569.

Hellmayr, C. E. 1906. [Descriptions of new South
American birds.] Bull. Brit. Ornithol. Club. 19:
8-9.

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

Isler, M. L. 1997. A sector-based ornithological geo-
graphic information system for the Neotropics.
Ornithol. Monogr. 48:345-354.

Isler, M. L., P. R. Isler, and B. M. Whitney. 1997.
Biogeography and systematics of the Thamnophi-
lus punctatiis (Thamnophilidae) complex. Orni-
thol. Monogr. 48:355—382.

Isler, P. R. and B. M. Whitney. 1999. Voices of the
antbirds: Thamnophilidae, Pormicariidae, and
Conopophagidae. Library of Natural Sounds, Cor-
nell Laboratory of Ornithology, Ithaca, New York.
Mayr, E. and W. H. Phelps, Jr. 1967. The origin of
the bird fauna of the south Venezuelan highlands.
Bull. Am. Mus. Nat. Hist. 136:269-328.

Meyer de Schauen.see, R. 1966. The species of birds
of South America with their distribution. Living-
ston Press, Narbcth, Pennsylvania.

Meyer de Schauensee, R. 1970. A guide to the birds
of South America. Livingston Press, Narbeth,
Pennsylvania.

Meyer de Schauensee, R. and W. H. Phelps, Jr.
1978. A guide to the birds of Venezuela. Princeton
Univ. Press, Princeton, New Jersey.

Monroe, B. L., Jr. and C. G. Sibley. 1993. A world

Zimmer • CAURA AND YAPACANA ANTBIRDS

209

checklist of birds. Yale Univ. Press, New Haven,
Connecticut.

Parker, T. A., Ill, D. F. Stot/, and J. W. Fitzpatrick.
1996. Ecological and distributional databases. Pp.
1 13-436 in Neotropical birds: ecology and con-
servation (D. F Stotz, J. W. Fitzpatrick, T. A.
Parker, III, and D. K. Moscovits, Eds.). Univ. of
Chicago Press, Chicago, Illinois.

Paynter, R. a. 1982. Ornithological gazetteer of Ven-
ezuela. Museum of Comparative Zoology, Cam-
bridge, Massachussetts.

Peters, J. L. 1951. Checklist of birds of the world.
Vol. VII. Harvard Univ. Press, Cambridge, Mas-
sachussetts.

Phelps, W. H. and W. H. Phelps, Jr. 1963. Lista de
las aves de Venezuela con su distribucion. Parte
II. Passeriformes, second ed. Bol. Soc. Venez.
Cienc. Nat. 24:1—479.

PiRES, J. M. 1974. Tipos de vegetacao da Amazonia.
Publ. Avul. Mus. Goeldi. 20: 179-202.

Remsen, j. V., Jr. 1997. A new genus for the Yellow-
shouldered Grosbeak. Ornithol. Monogr. 48:89—
90.

Remsen, J. V., Jr. and S. K. Robinson. 1990. A clas-
sihcation scheme for foraging behavior of birds in
terrestrial habitats. Stud. Avian Biol. 13:144—160.

Remsen, J. V., Jr. and T. S. Schulenberg. 1997. The
pervasive influence of Ted Parker on Neotropical
field ornithology. Ornithol. Monogr. 48:7-20.

Ridgely, R. S. and G. Tudor. 1994. The birds of
South America. Vol. 2. Univ. of Texas Press. Aus-
tin.

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

Snow, D. 1982. The cotingas. Cornell Univ. Press, Ith-
aca, New York.

Stotz, D. E, J. W. Fitzpatrick, T. A. Parker, III, and
D. K. Moskovits. 1996. Neotropical birds: ecol-
ogy and conservation. Univ. of Chicago Press,
Chicago, Illinois.

Todd, W. E. C. 1927. New gnateaters and antbirds
from tropical America, with a revision of the ge-
nus Myrmeciza and its allies. Proc. Biol. Soc.
Wash. 40:149-178.

Willard, D. E., M. S. Foster, G. F. Barrowcloligh,
R. W. Dickerman, P. F. Cannell, S. L. Coats, J.
L. Cracraft, and j. P. O’Neill. 1991. The birds
of Cerro de la Neblina, Territorio Federal Ama-
zonas, Venezuela. Fieldiana Zool. 65:1-90.

Zimmer, J. T. and W. H. Phelps. 1947. Seven new
subspecies of birds from Venezuela and Brazil.
Am. Mus. Novit. 1338:1-7.

Zimmer, K. J. and S. L. Hilty. 1997. Avifauna of a
locality in the upper Orinoco drainage of Ama-
zonas, Venezuela. Ornithol. Monogr. 48:865-886.

Zimmer, K. J., A. Whittaker, and D. F. Stotz. 1997.
Vocalizations, behavior and distribution of the Rio
Branco Antbird. Wilson Bull. 109:663—678.

Wilson Bull., 111(2), 1999, pp. 210-215

HABITAT PATCH SIZE AND NESTING SUCCESS OF
YELLOW-BREASTED CHATS

DIRK E. BURHANS' - AND FRANK R, THOMPSON IIP

ABSTRACT. — We measured vegetation at shrub patches used for nesting by Yellow-breasted Chats {Icteria
virens) to evaluate the importance of nesting habitat patch features on nest predation, cowbird parasitism, and
nest site selection. Logistic regression models indicated that nests in small patches (average diameter <5.5 m)
that were parasitized by Brown-headed Cowbirds (Molothrus ater) experienced higher predation than unpara-
sitized nests in large patches. Nests in large patches were more likely to become parasitized by cowbirds, as
were nests with more large stems (>10 cm dbh) nearby. Patches used by chats for nesting had larger average
diameters than unused patches and tended to contain more small stems. Chats appeared to prefer large patches
and experienced lower nest predation there. Although they might experience higher brood parasitism frequencies
in large patches, losses to parasitism were balanced by higher nesting success because the mean number of chat
young that fledged did not differ between nests in small versus large patches. Received 12 Jan. 1998, accepted
28 Dec. 1998.

The nest “patch” has been defined as the
habitat patch immediately sunounding the
nest (Martin and Roper 1988). Characteristics
of the songbird nesting patch may differ from
the habitat available (Martin and Roper 1988;
Kelly 1993; Steele 1993; Kligo et al. 1996a,
b) and there may be differences between suc-
cessful and unsuccessful nests according to
nest patch characteristics (Martin and Roper
1988, Kelly 1993, Norment 1993, Tarvin and
Smith 1995). However, there is no consensus
on exactly what determines a nest patch. Pe-
tersen and Best (1985) and Martin and Roper
( 1988) defined the nest patch as the area with-
in 5 m of the nest, a criterion that other studies
since have adopted (Kligo et al. 1996a, b; Bar-
ber and Martin 1997). Other workers have
evaluated nest patches based upon other pre-
determined sizes (Conner et al. 1986, Kelly
1993, Norment 1993, Tarvin and Smith 1995),
multiple radius patch sizes (Petit et al. 1988,
Holway 1991, With 1994), or stem density
(Hoi way 1991, Knopf and Sedgewick 1992).
Knopf and Sedgwick ( 1992) based their patch
definition upon vegetation height and radius
descriptors rather than upon pre-determined
size, and concluded that individual plants
probably are functionally indistinguishable to
Yellow Warblers (Dendroicci petechia), which

' North Central Research .Station, USDA Forest Ser-
vice, 202 Natural Resources Building, Univ. of Mis-
souri, Columbia. MO 6521 I.

- Corresponding author;

E-mail: dburhan,s/nc_co@fs.fcd.us

select nests based on patch characteristics
rather than the nest plant.

We examined the relationship between nest
patch characteristics and nest predation, brood
parasitism, and nest site selection for the Yel-
low-breasted Chat {Icteria virens). Yellow-
breasted Chats are a common songbird of
shrub habitats (Nolan 1963, Thompson and
Nolan 1973) and at our sites often nested in
conspicuous dense thickets of shrubs. We
combined two approaches by measuring veg-
etation structure in a fixed-radius plot centered
on the nest and measuring dimensions of the
shrub patch in which the nest was located. Our
principle questions were: ( 1 ) ai'e chat nests in
large thickets, or patches, more likely to
fledge young than nests in small patches or
single shrubs and trees? and (2) are chat nests
that are placed further from the edge of the
nesting patch more likely to fledge young? We
predicted that chats nesting in larger patches
at greater distances from the patch edge would
be more likely to avoid predation because
large patches may impede the movements of
predators (Bowman and Hanis 1980, Holway
1991). Additionally, we predicted that nests
near greater numbers of trees would experi-
ence higher frequencies of cowbird paiasitism
because Brown-headed Cowbirds {Molothrus
ater) use trees to aid in finding nests (Ander-
son and Storer 1976, Romig and Crawford
1995, Clotfelter 1998). We also predicted that
size of nest patches would differ from the size
of patches selected at random. We tested these
predictions by monitoring nest success and

210

Bmiums and Thompson • YELLOW-BREASTED CHAT NESTING I’ATCH

21 1

cowbird parasitism of chats and by measuring
vegetation at nest sites and unused sites.

METHODS

We found Yellow-breasted Chat nests at Thomas
Baskett Wildlife Research and Education Center near
Ashland, (Boone County) Missouri, from 1992-1994
as part of a study of shrubland birds. Study sites were
six old fields ranging from 2.4 to 16.3 ha and surround-
ed by oak-hickory forest (see Burhans 1997 for de-
tailed site description). We monitored nests every 3-4
days and daily toward the end of the nestling period.
We considered nests that avoided predation and suc-
ceeded in Hedging either chat or cowbird young as
"fledged”. In most cases fledged nests were identified
by observing adults carrying food or scolding, or by
observing fledglings. Nests that were empty on the
fledging day (day 8, where day of hatching = day 0)
were classified as Hedged if they were active the day
before. We classified nests that were empty prior to
this time as depredated unless there were signs of pre-
mature Hedging, such as nearby fiedglings or adult
feeding activity. Parasitism status was determined for
all nests and only those nests that were initiated during
the period of cowbird parasitism (before the second
week of July) were considered in the parasitism anal-
ysis.

Vegetation samples were taken at nest sites and un-
used sites at the end of the nesting season. We mea-
sured nest height to the bottom of the nest cup. We
also measured nest “patch”, which was defined as in-
terlocking leafy shrub or tree vegetation at nest height
within which the nest plant was situated. Nest patches
varied in size from the single nest tree or shrub to an
entire fencerow. We measured length and width of
patches to the nearest 0. 1 m for distances within 3 m
and paced (calibrated at 1 m/pace) to the nearest m for
greater distances. “Average patch diameter” was the
sum of the length of the nest vegetation clump plus
the width of the clump divided by two. Nest patch
diameter varied greatly among patches (median = 5.5
m, range 0.3-65 m) so we grouped nests into “large”
(S:5.5 m) or “small” patches for analyses (see below).
“Patch-edge distance” was the distance (to the nearest
0.1 m) from the outside rim of the nest cup to the
nearest leafy edge of the nest patch. In order to further
characterize patches and evaluate potential cowbird
perches, we counted woody stems 1 1—20, 21—50, and
greater than 50 cm dbh (diameter at breast-height) in
an 11.3 m radius circle centered on each nest. We
counted shrub and sapling stems (^1 m high) in a 5
m radius circle around each nest in categories less than
2, 2-5, and at least 5 cm dbh. Many chat nests were
placed in large blackberry (Rnbus allegheniensis)
patches in which it was difficult to count stems. For
large blackberry patches (>10% of the circle) we es-
timated number of blackberry stems by counting the
number of stems in a square meter and extrapolating
to the proportion of the 5 m circle that was blackberry.

Unused sites were located by pacing in a randomly

determined compass direction to the first plant en-
countered of the same species and size category as the
nest plant (at least 40 m from the nest). As with nest
sites, we took patch diameter and stem count mea-
surements for unused patches. We did not sample veg-
etation for 10 nests destroyed by Hooding in 1993 and
storms in 1994 and did not include these nests in the
analysis. We also omitted 2 nests found immediately
before Hedging where it was not possible to inspect
chicks to determine parasitism status without forcing
Hedging.

Data analyses. — We evaluated Hedging success us-
ing both simple nesting success (number of successful
nests/total nests) and the Mayfield method (Mayfield
1961, 1975). For the Mayfield method half the number
of days between subsequent visits over which a nest
was empty were added to the number of previous days
the nest survived to obtain the total number of obser-
vation days for a nest. When calculating daily survival
probabilities we only included mortality caused by nest
predation. We calculated survival probabilities and
variances with standard errors according to Johnson
(1979). We compared survival probabilities using
CONTRAST (DOS; Sauer and Williams 1989). An-
other species that nested at this site (Indigo Bunting;
Passerina cyanea) suffered higher predation at para-
sitized nests (Dearborn in press), so we compared daily
survival probabilities between parasitized and unpar-
asitized nests. Simple nesting success was used for lo-
gistic regression models (below). Nests that Hedged at
least one chick (chat or cowbird) were considered
“Hedged”. When calculating mean number of chat
young Hedged, we assumed that the number success-
fully Hedged was equal to the number of chicks last
counted in the nest. We compared mean number of
chat chicks Hedged from nests in large versus small
patches with an independent sample two-tailed r-test.

We analyzed both nest predation and nest parasitsm
with logistic regression models. Nest height, patch dis-
tance. average patch diameter, stems defined as above,
total stems 10 cm dbh or smaller (“total small stems”),
and cowbird parasitism status (parasitized or not) were
evaluated in the nest predation model. Frequency of
parasitism has been related to nest height and nest veg-
etation (Hahn and Hatfield 1995, Britlingham and
Temple 1996), so we similarly used logistic regression
to analyze parasitism against nest height, average patch
diameter, patch distance, mean stems at least 10-20,
21-50, greater than 50 cm dbh. and combined stems
greater than 10 cm dbh (“total large stems"). Model
building for both nest predation and parasitism models
followed the method of Hosmer and Lemeshow ( 1989)
and consisted of running univariate logistic regression
models and retaining variables with P-values of 0.25
or less in a full model. The final reduced models in-
cluded those variables with P < 0.05. Decisions about
which variables should be left in final models were
based on probability values for individual variables
from a set of alternative multivariate models. We per-
formed Hosmer and Lemeshow ( 1989) goodness-of-fit
tests on the final models.

212

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 1. Parameter estimates, Wald X" statistics and probability levels for final nest predation (top) and
nest parasitism (bottom) logistic regression models.

ModelA'ariable

Parameter

X-

Nest predation

Parasitism

-2.23

5.44

0.02

Average patch diameter

-1.52

4.24

0.04

Nest parasitism

Average patch diameter

2.28

5.50

0.02

Total large stems (>10 cm dbh)

0.25

5.36

0.02

Nest sites and unused sites were compared with lo-
gistic regression rather than discriminant function be-
cause of the presence of binary explanatory variables
and non-normal variance of other variables (Press and
Wilson 1978, James and McCullogh 1990). Nest and
unused site variables were screened with univariate lo-
gistic regression models, and multivariate models were
developed similarly to the predation and brood para-
sitism models (above). Results for statistical tests are
reported as mean ± SE.

RESULTS

Nesting success and predation/parasitism
models. — Daily survival estimates of Yellow-
breasted Chat nests did not differ among years
from 1992-1994 (1992: 0.96 ± 0.01, 1993:
0.94 ± 0.02, 1994: 0.95 ± 0.02; x' = 0.7, df
= 2, P > 0.05) so data from all nests were
pooled for the predation analysis (0.95 ±

0. 01; n = 48 nests). Brood parasitism fre-

quency was 33% (n = 15 nests), 36% (n =
14 nests) and 23% (n = 13 nests) for 1992,
1993, and 1994 and did not differ between
years (Fisher exact test; P > 0.05). Cowbird
parasitism averaged 3 1 % over all years during
the seasonal period of parasitism (n = 42
nests). Parasitized nests did not have signifi-
cantly different survival rates than unparasit-
ized nests (parasitized nests 0.94 ± 0.02; un-
parasitized nests 0.96 ± 0.01; ~ 1-2, df =

1, P > 0.05).

Nest predation was best explained by a final
logistic regression model including parasitism
status and average patch diameter (Table 1;
Log likelihood for model = 51.8, x^ = 8.7, df
= 2, P = 0.01). Nests that were parasitized
and in small patches were more likely to suf-
fer predation. However, the nest parasitism
model indicated that nests in large patches
were more likely to become parasitized. The
nest parasitism model included the variables
average patch diameter and total large stems
(Table 1; Log likelihood for model = 40.9, x^

= 1 1.1, df = 2, P = 0.004). Nests with more
large stems were more likely to be parasitized,
but large patches did not have greater mean
values for total large stems than did small
patches (large patches 2.27 ± 0.67; small
patches 3.45 ± 1.01; / = 0.99, df = 40, P >
0.05). Distance from the nest to the edge of
the patch tended to be greater for paiasitized
nests (Table 2), but was eliminated from the
parasitism models because of the higher prob-
ability values associated with average patch
diameter, with which patch-edge distance was
positively conelated prior to transformation of
the former variable (r = 0.39, P = 0.009).
Mean number of chat young fledged did not
vary between nests in small versus large
patches (small patches: 1.04 ± 0.34 chat
young per nest; large patches 1.43 ± 0.36 chat
young per nest; t = —0.79, df = 44, P >
0.05).

Nest sites versus unused sites. — Univariate
logistic regression models indicated that nest
sites were situated in larger patches than un-
used sites (Table 3). When variables were
combined in the multivariate model only av-
erage patch diameter was significant (Log
likelihood for model = 121.07, X“ = 12.0, P
= 0.001).

DISCUSSION

As predicted, logistic regression models in-
dicated that Yellow-breasted Chats experi-
enced less predation in larger nest patches. As
with Indigo Buntings at these sites (Dearborn,
in press), predation was related to pai'asitism
status at Yellow-breasted Chat nests; nests that
were parasitized were more likely to experi-
ence predation. Chats tended to place nests in
larger patches with more small stems than
those in unused sites. Nests that were placed
faither from the patch edge were more sus-

Means (± standard error), parameter estimates, Wald statistics, and probability levels for individual variables from logistic regressions on predation

Burhciiis and Thompson • YELLOW-BREASTED CHAT NESTING PATCH

213

r--

CL.

r--

•o

'(A

d
+ 1

-t-l

r-'

Cl»

'(A

-1-1

-M

CfJ

'Ct

r--

—

Vw^

•a

11%

63%

013

-o

+ 1

'i-

1— •

—

»— .*

r--

•o

-a

+ 1

c/3

Variable

c/3

Lln

<L)

-C

■o

c/3

•o

c/3

C/3

-C

c/3

'-B

C/3

c/3

C/3

c/3

OX)

D-

*C/3

CdD

L-

C/3

-C

c/3

f—

c/3

r“

inj

c/3

—

•w

ceptible to parasitism (Table 2); however, we
were unable to separate the importance of
patch-edge distance from the size of the patch
itself (patch diameter). Although nests in large
patches were more likely to become parasit-
ized, higher nesting success in large patches
compensated for decrements in fitness caused
cowbird parasitism because the number of
host young that fledged was equal between
small and large patches.

Petersen and Best (1985), Knopf and Sedg-
wick (1992), and Holway (1991) found that
birds selected large shrubs or shrub stands for
nest placement. Holway (1991) and Knopf
and Sedgwick (1992) suggested that large
patches offer improved nest concealment;
Holway (1991) also believed that large patch-
es could impede the movements of mammals,
and could contain more potential nest sites for
predators to search (see also Martin and Roper
1988).

Several researchers have found that birds
place nests in denser cover than in unused
sites (Knopf and Sedwick 1992, Sedgwick and
Knopf 1992). Holway (1991) and Steele
(1993) found higher foliage and shrub density
at nests of Black-throated Blue Warblers
(Dendroica caerulescens) than at random
points. Wray and Whitmore (1979) and Nor-
ment (1993) found that successful Vesper
Sparrow (Pooecetes gramineus) and Harris
Sparrow (Zonotrichia querula) nests tended to
be placed in denser cover than unsuccessful
nests.

Chat nests parasitized by Brown-headed
Cowbirds were more likely to become dep-
redated. Dearborn (in press) found higher dai-
ly mortality at parasitized nests of Indigo Bun-
tings in a five year study from the same sites
we used. He suggested that louder vocaliza-
tion by cowbird nestlings was partly the
cause, although daily mortality was also high-
er at parasitized bunting nests during the in-
cubation stage. In our study, the sample size
of parasitized nests with cowbird chicks was
too small (/? = 4) to adequately compare daily
mortality between nests with cowbird chicks
and those without (Hensler and Nichols 1981).

Parasitized nests had more potential cow-
bird perches (laige stems) than unparasitized
nests. Recent studies have documented the im-
portance of perch proximity in brood parasit-
ism in both cuckoos (Cuculus canorus\ Al-

214

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 3. Means (± standard error), parameter estimates, Wald X‘ statistics, and probability levels for
individual variables from univariate logistic regressions comparing nest sites and unused sites.

Variable

Nest sites

Unused sites

Parameter

X-

Average patch diameter

52%

19%

1.55

10.91

0.001

(% in large patches)

Stems <2 cm dbh

313.23

± 108.47

60.19 ±

1 1.49

0.00

2.47

0.12

Stems 2-5 cm dbh

5.44

± 1.14

4.17 ±

1.01

0.02

0.68

0.41

Stems 5-10 cm dbh

1.48

± 0.29

1.10 ±

0.24

0.1 1

0.99

0.32

Stems 1 1-20 cm dbh

1.58

± 0.31

2.15 ±

0.42

-0.09

1.12

0.29

Stems 21-50 cm dbh

1.06

± 0.29

1.21 ±

0.32

-0.03

0.12

0.73

Stems >50 cm dbh

0.04

± 0.03

0.02 ±

0.02

0.71

0.33

0.57

Total small stems (<10 cm dbh)

320.14

± 108.19

65.45 ±

1 1.51

0.00

2.77

0.10

Total large stems (>10 cm dbh)

2.69

± 0.54

3.38 ±

0.65

-0.04

0.66

0.42

varez 1993, 0ien et al. 1996) and cowbirds
(Romig and Crawford 1995, Clotfelter 1998;
see also Anderson and Storer 1979). Previous
studies (Burhans 1997) on Field Sparrows
{Spizella pusilla) and Indigo Buntings nesting
at Thomas Baskett Wildlife Research and Ed-
ucation Center indicated no direct relationship
between perches near the nest and frequency
of parasitism. However, Yellow-breasted Chat
nest sites generally are situated in patches
with more trees and shrubs than old field nests
of Indigo Buntings and Field Sparrows. The
higher frequency of parasitism in large patch-
es was not an artifact of patch size, because
large patches did not necessarily contain more
total large stems.

Nest site selection, nesting success, and fre-
quency of cowbird parasitism at Yellow-
breasted Chat nests appear to be influenced by
patch size. However, relaxed predation in
large patches did not improve host fledging
success, because chats were more likely to be-
come parasitized in large patches and fledge
fewer of their own young. Although predation
and parasitism appeared to differ across patch
sizes, the effects of patch size on host fitness
appear to cancel each other out. Future studies
should look further at interactions between
site selection, brood parasitism, and predation,
and investigate tradeoffs in reproductive suc-
cess associated with these factors according to
different types of nest sites.

ACKNOWLEDGMENTS

Sugge.stions from D. Dearborn, and an anonymous
reviewer greatly improved the manuscript. We are
grateful to C. Lrceman, D. Dearborn, D. Martasian, M.
Alexander, and R. Kun/.a for help in the field and J.

Demand for help with data entry. We thank J. Laaborg

for his assistance. C. Lreiling kindly allowed us to

work on his property. This study was funded by the

USDA Eorest Service North Central Research Station.

LITERATURE CITED

Alvarez, F. 1993. Proximity of trees facilitates para-
sitism by Cuckoos Cuciilus canoni.s on Rufous
Warblers Cercotrichas galacfofes. Ibis 135:331.

Anderson, W. L. and R. W. Storer. 1976. Factors
influencing Kirtland’s Warbler nesting suceess.
Jaek-Pine Warbler 54:105-1 15.

Barber, D. R. and T. E. Martin. 1997. Influenee of
alternate host densities on Brown-headed Cowbird
parasitism rates in Black-eapped Vireos. Condor
99:595-604.

Bowman, G. B. and L. D. Harris. 1980. Effect of
spatial heterogeneity on ground-nest depredation.
J. Wildl. Manage. 44:806-813.

Brittingham, M. C. and S. A. Temple. 1996. Vege-
tation around parasitized and non-parasitized nests
within deeiduous forests. J. Field Ornith. 67:406-
413.

Burhans, D. E. 1997. Habitat and microhabitat fea-
tures associated with cowbird parasitism in two
forest edge cowbird hosts. Condor 99:866-872.

Clotfelter, E. D. 1998. What cues do Brown-headed
Cowbirds use to locate Red-winged Blackbird
host nests? Anim. Behav. 55:1181-1189.

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

Dearborn, D. C. In Press. Brown-headed Cowbird
nestling vocalizations and the risk of nest preda-
tion. Auk.

Hahn, D. C. and J. S. Hatfield. 1995. Parasitism at
the landscape scale: cowbirds prefer forests. Con-
■serv. Biol. 9:1415-1424.

Hensler, G. L. and j. D. Nichols. 1981. The Mayfield
method of estimating nesting success: a model,
estimators and simulation results. Wilson Bull. 93:
42-53.

limiuins anil Thompson • YELLOW-BREASTED CHAT NESTINCj PATCH

215

Holway, D. a. 1991. Nest-site seleetion and the im-
portance of nest concealment in the Black-throat-
ed Blue Warbler. Condor 93:575-581.

Hosmer, D. W., Jr. and S. Lemeshow. 1989. Applied
logistic regression. John Wiley & Sons, New
York.

J.AMES, E C. AND C. E. McCulloch. 1990. Multivariate
analysis in ecology and systematics: panacea or
pandora’s box? Annu. Rev. Ecol. Syst. 21:129-
166.

Johnson, D. H. 1979. Estimating nest success: the
Mayfield method and an alternative. Auk 96:651-
661.

Kelly, J. P. 1993. The effect of nest predation on hab-
itat selection by Dusky Flycatchers in limber pine-
juniper woodland. Condor 95:83-93.

Kligo, J. C., R. a. Sargent, B. R. Chapman, and K.
V. Miller. 1996a. Nest-site selection by Hooded
Warblers in bottomland hardwoods of South Car-
olina. Wilson Bull. 108:53-60.

Kligo, J. C., R. A. Sargent, K. V. Miller, and B. R.
Chapman. 1996b. Nest sites of Kentucky Warblers
in bottomland hardwoods of South Carolina. J.
Field Ornith. 67:300-306.

Knopf, E L. and J. A. Sedgwick. 1992. An experi-
mental study of nest-site selection by Yellow War-
blers. Condor 94:734-742.

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.

Mayheld, H. F. 1975. Suggestions for calculating nest
success. Wilson Bull. 87:456-466.

Nolan, V., Jr. 1963. Reproductive success of birds in
a deciduous scrub habitat. Ecology 44:305-313.

Norment, C. j. 1993. Nest-site characteristics and nest
predation in Harris’ Sparrows and White-Crowned
Sparrows in the northwest territories, Canada. Auk
1 10:769-777.

0IEN, 1. J., M. HoNZA, a. MoKSNES, and E. R0.SKAET.
1996. The risk of parasitism in relation to the dis-
tance from Reed Warbler nests to cuckoo perches.
J. Anim. Ecol. 65:147-153.

Petersen, K. L. and L. B. Best. 1985. Nest-site se-
lection by Sage Sparrows. Condor 87:217-221.

Petit, K. E., D. R. Petit, and L. J. Petit. 1988. On
measuring vegetation characteristics in bird terri-
tories: nest sites vs. perch sites and the effect of
plot size. Am. Midland Nat. 1 19:209-215.

Press, S. J. and S. Wilson. 1978. Choosing between
logistic regression and discriminant analysis. J.
Am. Stat. Assoc. 73:699-705.

Romig, G. P. and R. D. Crawford. 1995. Clay-colored
Sparrows in North Dakota parasitized by Brown-
headed Cowbirds. Prairie Nat. 27:193-203.

Sauer, J. R. and B. K. Williams. 1989. Generalized
procedures for testing hypotheses about survival
or recovery rates. J. Wildl. Manage. 53:137-142.

Sedgwick, J. A. and F. L. Knopf. 1992. Describing
Willow Flycatcher habitats: scale perspectives and
gender differences. Condor 94:720—733.

Steele, B. B. 1993. Selection of foraging and nesting
sites by Black-thoated Blue Warblers: their rela-
tive influence on habitat choice. Condor 95:568-
579.

Tarvin, K. a. and K. G. Smith. 1995. Microhabitat
factors influencing predation and success of sub-
urban Blue Jay Cyanocitta cristata nests. J. Avian.
Biol. 26:296-304.

Thompson, C. F. and V. Nolan, Jr. 1973. Population
biology of the Yellow-breasted Chat (Icteria vi-
rens L.) in southern Indiana. Ecol. Monog. 43:
145-171.

With, K. A. 1994. The hazards of nesting near shrubs
for a grassland bird, the McCown’s Longspun
Condor 96:1009-1019.

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

Wilson Bull., 1 1 1(2), 1999, pp. 216-228

AVIFAUNA OF A PARAGUAYAN CERRADO LOCALITY: PARQUE
NACIONAL SERRANIA SAN LUIS, DEPTO. CONCEPCION

MARK B. ROBBINS,' 3 ROB. C. FAUCETT,^ AND NATHAN H. RICE'^

ABSTRACT. — We recorded 181 avian species at the Paraguayan Cerrado site, Parque Nacional Serram'a San
Luis, depto. Concepcion, including the first record of Veery (Catharus fuscescens) for the country. We obtained
further evidence of hybridization between White-bellied (Basileiiterus hypoleucus) and Golden-crowned [B.
ciilicivorus) warblers. Our results combined with those of earlier workers document a total of 219 species for
this area. Pronounced differences in species composition exist between San Luis and a nearby Cerrado locality.
Three threatened and four near-threatened species were recorded at San Luis. Since our inventory the area that
we worked was traded for an adjacent, less human impacted sector of similar size. The new park boundary
contains relatively pristine campo (grassland), which is an important habitat for a number of resident and migrant
species suspected to be suffering serious population declines. Received 29 Jan. 1998, accepted 14 Nov. 1998.

RESUMEN. — Se registraron 181 especies de aves en el Parque Nacional Serrania San Luis, en el cerrado del
Paraguay, depto. Concepcion. La lista incluye el primer registro de Catharus fuscescens para el pafs, y mas
evidencia de hibridizacidn entre los parulidos Basileiiterus hypoleucus y B. culiciviorus. En combinacion con
los resultados de investigadores anteriores, han sido registrado 219 especies de aves para la zona. Existen
diferencias marcadas en la composicion avifaum'stica entre San Luis y otra localidad cercana en cerrado. Se
registraron tres especies amenazadas y cuatro casi-amenazadas en la zona. Desde que se hizo el presente estudio,
se cambio. el area de estudio por otro adyacente, menos impactado y de extenso similar; esta zona abarca mas
del campo (pastizal) en buen estado de conservacion, el cual es un refugio para varias especies residentes y
migratorias que se cuentran en declives poblacionales.

The Cerrado is the second largest ecologi-
cal region in South America (Ah’ Saber 1977),
but has only recently attracted the attention of
avian biogeographers (Haffer 1985; Silva
1995a, b, c). Silva (1995a) demonstrated that
about 70% of the Cerrado is inadequately
sampled and that the southern component es-
pecially is poorly known and most heavily im-
pacted by human activities. The Cerrado
reaches its southern terminus in northeastern
Paraguay, where it interdigitates with two oth-
er physiogeographic and biotic regions: the
Chaco and the Atlantic Forest (Hayes 1995a).

The lack of a quantitative inventory of this
region, coupled with the potential for exten-
sive biotic interchange among the contiguous
regions, make the Paraguayan Cerrado impor-
tant for its unique contribution to the biodi-
versity of the Cerrado. In this paper, we pre-
sent a preliminary avifaunal inventory for the

' Division of Ornithology, Natural History Museum,
Univ. of Kansas, Lawrence, KS 66045.

- Museum of Natural Science, Foster Hall, Louisiana
State Univ., Baton Rouge, LA 70803.

’ Corresponding author;

E-mail: mrobbins@falcon.cc.ukans.edu

•• Present address: Academy of Natural Sciences,
Dept, of Ornithology, 1900 Benjamin Franklin Park-
way, Philadelphia, PA 19103.

10,273 ha Parque Nacional Serrama San Luis,
created in 1991 to preserve a representative
sample of the Paraguayan Cerrado (Direccion
de Parques Nacionales y Vida Silvestre 1993).
We have incorporated results from earlier
work in the region to provide a more compre-
hensive list for this area. The San Luis list is
compared to another well-surveyed Paiaguay-
an Cerrado locality to examine avian species
turnover. Selected species accounts are pro-
vided where additional comment is merited to
clarify status, distribution, migration, food
habits, hybridization, and plumage characters.

STUDY AREA AND METHODS

We worked at the southern end of Parque Nacional
Serram'a San Luis (22° 40' S, 57° 21' W; taken with
Global Positioning System at park headquarters; Fig.
1) from 19-31 October 1996, and RCF revisited the
site from 29 November to 7 December 1996. The park
is located in the Serrania San Luis, an isolated, low-
lying (max. elevation 500 m) set of limestone hills
mostly covered with semihumid forest. Approximately
5,300 ha is cultivated grassland with about 70% of this
habitat now occupied by the aggressive African grass
Hyparrhenia rufa (Poaceae). The dominant savannah
tree, Tahehuia aurea (Bignoniaceae), was often asso-
ciated with Astronium urundeuva (Anacardiaceae), A/i-
adenanthera coluhrina (Leguminosae), and Rhainni-
dium elasocarpum (Rhamnaceae; Consorcio Parelc-
Foragro-Porto Real, unpubl. report). Hillsides were

216

Rohhins el al. • PARAGUAYAN CERRADO AVIFAUNA

217

FIG. 1 . Locations of Parque Nacional Serrania San
Luis and Parque Nacional Cerro Cora.

covered with shallow, rocky soil, resulting in shorter,
xeric forest; the more level areas along two arroyos
bordering the western and eastern sides of the park
held taller, more humid forest. Terrestrial bromeliads
and cacti were common in the dense, thorn-covered
understory on slopes and on isolated woodland in the
campo (open grassland). The largest area of campo,
about 4X2 km, had a few woodlots, some only a few
meters square. The upland forest and isolated wood-
lands had uneven canopies and were relatively short,
with few trees exceeding 20 m. Prominent trees in-
cluded; Amburana cearensis (Leguminosae), Aspido-
sperma pyrifolium (Apocynaceae), Calycophyllum
midtiflorum (Rubiaceae), and Astroniiim urundeuva
(Anacardiaceae). The understory was dominated by
trees of tbe genus Trichilia (Meliaceae).

Forests along the arroyos, especially east of Tagatlya
Guazu arroyo at the headquarters, were taller and more
humid, with some trees exceeding 40 m. These forests
were heavily disturbed by selective logging, and tree-
fall gaps with dense understory were common. Vine
tangles were also more common than in the upland
forest, especially about 2 km east of the park head-
quarters. Tall arborescent bamboo (Gucidua spp.) was
primarily restricted to disturbed areas along the ar-
royos.

This area is very seasonal with a cool, dry climate
from April through most of September when temper-
atures occasionally drop to near 0° C. Temperature and
rainfall increase in October and November with De-
cember through March being relatively hot and humid.
Average annual rainfall is 1300 mm with considerable
variation (A. Acosta, pers. comm.), and average annual
temperature is 24° C (Consorcio Parelc-Foragro-Porto
Real, unpubl. report). During our stay, weather was
highly variable. On the evening of 19 October a major
storm system from the southeast brought heavy rains
(70 mm) and strong winds. Precipitation also occurred
on 25 October, 12 mm; 26 October, 2 mm; and 27
October, 6 mm. Apparently it did not rain between our

October and December surveys (A. Acosta, pers.
comm.); however, it rained five of nine days during
the latter inventory period. Temperature highs ranged
from 24-38° C. The low was I8°C at dawn on 23
October. We did not record the temperature during No-
vember and December.

Four mist-nets were opened on the afternoon of 19
October, and 6 on 20 October; 15 nets were maintained
from 21-30 October, for a total of about 1 1,400 mist-
net-hr. All mist-nets were positioned in mesic forest
bordering Tagatlya Guazu arroyo; the most distant nets
were about 1.5 km east of the headquarters. Most nets
were moved every three to four days. Nets were typ-
ically opened at dawn and closed about noon, rarely
at sunset. Our inventory was confined to about 4 km
of forest trail east of the headquarters, about 5 km of
dirt road to the north of the headquarters, and about 2
km of dirt road from the headquarters to the southwest
entrance. On most mornings, RCF and NHR worked
the nets and adjacent forest, while MBR made inde-
pendent surveys of other areas from predawn until
11:00 or 12:00. Occasional forays were made in the
afternoon. Nocturnal fieldwork was limited to nights
with a full moon in October.

Specimens were deposited at the University of Kan-
sas Natural History Museum (KU), Lawrence, Kansas
and Museo Nacional de Historia Natural del Paraguay
(MNHNP), Asuncion, Paraguay. Tissue samples were
taken from every nonfluid preserved specimen and de-
posited at KU. Tape recordings will be deposited at the
Library of Natural Sounds, Cornell University, Ithaca,
New York.

To provide a more complete inventory of the Ser-
rania de San Luis region, we have integrated our spe-
cies list (Appendix) with results from earlier work in
San Luis de La Sierra that was conducted in Septem-
ber-October 1931 by Hans Krieg (referred to as “Apa-
Bergland”; summarized in Laubmann 1939, 1940).
Results from Kreig’s expeditions to Zanja Morotf,
Centurion, and Estrella, Concepci'on were not included
because these areas are farther removed from San Luis,
and the Estrella site was at the Ri'o Apa. We presume
that the riparian habitat and associated fauna along the
Rio Apa is somewhat distinct from that at San Luis.
We also excluded specimens taken by Emil Kaempfer
at La Fonciere, Concepcion (deposited at American
Museum of Natural History; Floyd Hayes and Paul
Sweet, pers. comm.). Except where noted otherwise,
we follow Hayes (1995a) for taxonomy and nomen-
clature.

RESULTS

We recorded a total of 181 species (Appen-
dix), of which three are designated as threat-
ened and four as near- threatened (Collar et al.
1992, 1994). Three of the 29 species identified
by Silva (1995c) as endemic to the Ceixado,
Planalto Foliage-gleaner (Philydor dimidia-
tus). Curl-crested Jay {Cyanocorax cristatel-
lus), and Black-throated Saltator {Saltator

218

THE WILSON BULLETIN • Voi 111, No. 2, June 1999

atricolUs), have been recorded at San Luis.
We recorded four species that Silva (1995c)
did not list for the Cerrado: South American
Painted-Snipe (Rostratula semicollaris),
White-naped Xenopsaris {Xenopsaris albimi-
cha). Red-crested Cardinal (Paroaria coron-
ata), and Golden-winged Cacique {Caciciis
chrysopterus). Hayes (1995a) did not list R.
semicollaris, X. albinucha, Golden-green
Woodpecker {Piculiis chrysochloros), nor
Dark-throated Seedeater (Sporophila ruficol-
lis) as occurring in his “Campos Cerrados”
category.

At least 64 of the 92 species we collected
had enlarged gonads indicating reproduction,
and nests or recently fledged young were ob-
served for 12 additional species that were not
collected (Appendix). Based on vocal activity
and behavior, breeding was suspected for sev-
eral species not collected, such as Bare-faced
Currasow {Crax fasciolata). Red-legged Ser-
iema (Cariama cristata). Short-tailed Night-
hawk {Lurocalis semitorquatus), and Crowned
Slaty Flycatcher {Griseotyrannus aurantioa-
trocristatus). Hence, at least 80 species were
actively nesting at the time of our studies
(mid-October-early December). One Neaictic,
one intratropical, and at least 32 Austral mi-
grants were recorded (Appendix; migrant ter-
minology follows Hayes 1995b).

Nocturnal avian vocal activity was most
pronounced shortly after dusk and in the hour
before dawn, with the greatest activity on
nights with the brightest moon light. The Fer-
ruginous Pymy-Owl (Glaucidium brasilian-
um) and both screech-owls (Otus) were heard
nightly, as were Pauraque (Nyctidromus albi-
collis). Rufous Nightjar (Caprimidgus rufus),
and Common Potoo (Nyctibius griseus). The
Little Nightjar (C. parvulus) was never heard,
even though individuals and a nest with an
egg and a day-old young, were collected at
the edge of camp.

Rodent eating hawks and owls were not
well represented in the San Luis avifauna. A
possible causal explanation is that rodent di-
versity and density were low. During October,
we observed no evidence of rodents. Squirrels
are unknown from this part of Paraguay; dur-
ing seven nights in late November— early De-
cember Texas Tech University mammalogists
captured few mammals in a variety of habitats
(M. Gorresen, pers. comm.). Raptor species

expected, but not detected include: Great
Black Hawk {Buteogallus urubitinga), Har-
ris’s Hawk {Parabuteo unicinctus), all Buteo
species. Spectacled Owl (Pulsatrix perspicil-
lata). Great Homed Owl {Bubo virginianus),
and Asio spp. Only the Savannah Hawk {Bu-
teogallus meridionalis) and Ornate Hawk-Ea-
gle {Spizaetus ornatus) were recorded (both
rarely) among the species of raptor whose di-
ets include a relatively high percent (>25%)
of rodents (Brown and Amadon 1968, del
Hoyo et al. 1994).

SPECIES ACCOUNTS

Snail Kite (Rostrhamus sociabilis). — On 22
and 23 October, we observed two flocks of
154 and 33 individuals (mostly adults), re-
spectively, pass through the campo. The birds
appeared from the north and landed on the
ground, in bushes, and in isolated palm trees,
flying only a few meters between each perch.
When the groups came to the forested Tagat-
lya Guazu aiToyo at the south end of the park,
they rose as a group into the air and disap-
peared to the south. Snail Kites were not ob-
served during late November-early December.
Our observations coincide with prior obser-
vations of migratory movements of this spe-
cies in Pai'aguay (Hayes et al. 1994).

Crowned Solitary Eagle { Harpy haliaetus
coronatus). — An adult with a tegu lizard (Tei-
idae; Tupinambis sp.) in its talons was flushed
from a large tree at the forest/campo edge on
19 October. It landed in the top of a neai’by
tree where it sat for about 3 min before it dis-
appeared with the lizard in its talons. This spe-
cies is considered vulnerable (“taxa believed
likely to move into the endangered category
in the near future if the causal factors continue
operating”; Collar et al. 1992) and has been
recorded few times from Paraguay with only
one prior record for depto. Concepcion (Hayes
1995a). Very little is known about this eagle’s
prey (Collar et al. 1992), and our observation
is apparently the first of it taking a. tegu. It is
known to take small mammals (Collar et al.
1992, del Hoyo et al. 1994), but we saw ho
signs of rodents. Del Hoyo and coworkers
(1994) noted that this eagle has short toes
characteristic of snake eaters.

Bare-faced Curassow (Crax fasciolata). —
Given that cracids are usually one of the first
avian species to disappear because of exces-

Robbins el at. • PARAGUAYAN CERRADO AVIFAUNA

219

sive hunting, we were encouraged to find at
least seven temtorial males calling along
about 4 km of trail on the east side of Tagatlya
Guazu aiToyo. Males called daily in October
from predawn until about 07:00, occasionally
until 10:00. We did not hear’ males in the drier
forest; however, we surveyed only a relatively
small area of this forest type at the optimal
time of day. No calling was heard in late No-
vember—early December.

Reddish-bellied Parrot (Pyrrhura fronta-
lis).— This species was common in both forest
types. Birds collected (KU 88346, MNHNP)
at San Luis were typical of the race P. f chi-
ripepe, and showed no sign of hybridization
with P. devillei as has been reported in areas
just northwest of San Luis (Short 1975).

Rufous Nightjar (Caprimulgus rufus). — The
single male (testes 11X6 mm; KU 88350)
obtained was not assignable to either the nom-
inate race or to southern C. r. rutilus. Like
birds from central Brazil (Goias, Bahia; Rob-
bins and Parker 1997), the San Luis speci-
men’s wing and tail measurements (173.6 and
119.7 mm, respectively) fall within the range
of variation of the nominate race, but plumage
color and pattern are closest to C. r. rutilus.

Blue-crowned Motmot (Momotus momo-
ta). — Our specimens (KU 88580, MNHNP)
from San Luis represent only the fifth locality
in Paraguay (Hayes 1995a, Ericson and Amar-
illa 1997), and the first for depto. Concepcion.
Both specimens, a probable pair, lack the blu-
ish-green throat of eastern Bolivian and Ar-
gentinian M. m. pilcomajensis; the greenish
underparts were washed with cinnamon, es-
pecially on the abdomen, which is more typ-
ical of southern Brazilian M. m. simplex.
Chapman (1923) was uncertain in assigning a
specimen from western Sao Paulo to either of
the above races and concluded that there was
considerable variability in this species south
of the Amazon. Ericson and Amarilla (1997)
believed that specimens they obtained from
Parque Nacional Defensores del Chaco, depto.
Chaco were intergrades between the above
forms.

Planalto Foliage-gleaner (Philydor dimi-
diatus). — Known in most of the literature as
the Russet-mantled Foliage-gleaner, we follow
Ridgely’s and Tudor’s (1994) appropriate sug-
gestion for an English name. Two pairs of this
poorly known foliage-gleaner were encoun-

tered about 2 km east of the headquarters (KU
88362, 88363, MNHNP). Both pairs were ob-
served foraging 2—4 m above the ground in
relatively dense understory. One pair was as-
sociated with a understory mixed-species
flock. Ridgely’s and Tudor’s (1994) descrip-
tion of the song and call accurately describe
what we recorded, with the San Luis birds
sounding very similar to birds from Patios de
Minas, Minas Gerais (tape recording by An-
drew Whittaker). Plumage differentiation and
taxonomy of this species will be presented
elsewhere.

Bare-throated Bellbird (Procnias nudicol-
lis). — San Luis lies at the western limit of this
species’ range (Snow 1982, Hayes 1995a). In
October, two birds were recorded. An adult
male (KU 88873) sang for more than 15 min
during mid-moming on 24 October from a
leafless, uppermost branch of one of the tallest
trees (ca 40 m) in the mesic forest east of the
headquarters. Only one or two other individ-
uals were heard in October, both gave the
bock call a few times, on two separate occa-
sions. In November and December, RCF heard
at least 12 displaying males daily along about
4 km of trail east of Tagatlya Guazu an'oyo.
An immature male (KU 88387: testes 8X4
mm) was one year old based on plumage
(Snow 1973) and bursa size (8X8 mm). The
plumage of this specimen resembles closely
that of an adult female (Snow 1982, Sick
1993), except that the head is entirely black
and the abdomen and crissum are primarily
white instead of pale yellow. It was in mod-
erate to heavy body molt, with #5 primai-y in
molt (primaries 1-4 fresh). Secondaiy 7 or 8
(one is missing on each side) was in molt.
Secondary 9 was very worn, and the others
appeared lightly worn. The central pair of rec-
trices was in molt with all other rectrices
heavily worn.

This bellbird has been documented as mi-
gratory in the southeastern part of its Brazilian
range (Belton 1985, Sick 1993); in Misiones,
Argentina it is apparently only transient (M.
Barnett in Lowen et al. 1996b). In Paraguay
it is thought to occur year round in depto. Ca-
nendiyii and probably depto. Amambay (Ma-
drono and Esquivel 1995). Its status elsewhere
in Paraguay is uncleai' (Lowen et al. 1996b).
Our limited fieldwork at San Luis suggests it
may occur there only seasonally. If this bell-

220

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

bird covers great distances in its seasonal
movements, then it may be especially vulner-
able to habitat fragmentation. Willis (1979) re-
garded large canopy frugivores to be vulner-
able because of their reliance on spatially and
temporally patchy resources. This bellbird and
the Red-ruffed Fruit Crow {Pyroderus scuta-
tus) completely disappeared from forests that
had been fragmented (largest fragment 198
ha) in the Lagoa Santa area of Brazil (Chris-
tiansen and Fitter 1997).

Veery (Catharus fuscescens). — We netted
an immature male (KU 88484) on 30 October.
This record is the first for Paraguay, and rep-
resents one of the southernmost records for
the continent (McFarlane 1974; Remsen and
Traylor 1983; Ridgely and Tudor 1989; Willis
and Oniki 1993; D. Stotz, pers. comm.). The
specimen is referable to the nominate subspe-
cies. All three Bolivian specimens originally
identified as C. f. salicicola by Remsen and
Traylor (1983) are now thought to be of either
the nominate race or C. f. fuliginosa (V. Rem-
sen, D. Stotz, pers. comm.). No subspecific
determinations have been made for southeast-
ern Brazilian specimens.

White-bellied Warbler (Basileuterus hypo-
leucus). — The presence of yellow in the lower
underparts of a few specimens from several
areas and observation of mixed-species pairs
document the occasional hybridization be-
tween B. hypoleucus and the Golden-crowned
Warbler (B. culicivorus; Hellmayr 1935, Wil-
lis 1986, Remsen and Traylor 1989, Silva
1991), although this literature has been over-
looked in recent paruline summations (Ridge-
ly and Tudor 1989, Curson et al. 1994). All
four specimens prepared as skins from San
Luis show signs of hybridization with varying
amounts of pale yellow on the center of the
abdomen and lower flanks (compared with
color plate and descriptions in Silva 1991).
We agree with prior assessments that this col-
or is indicative of gene flow between B. hy-
poleucus and B. culicivorus, and not with B.
flaveolus as implied in Hayes (1995a) for the
following reasons. Contrary to S. LaBar (in
Hayes 1995a), the vocalizations of B. hypo-
leucus are not more similar to those of B. flav-
eolus than to those of B. culicivorus. In fact,
the vocalizations of B. hypoleucus and B. cul-
icivorus are so similar that both species re-
spond to play-back recordings of each other’s

song (Silva 1991). Neither of these taxa’s song
resembles those of B. flaveolus, whose vocal-
izations and behavior clearly align it with the
Phaeothlypis assemblage (Ridgely and Tudor
1989; MBR, pers. obs.). Furthermore, the
plumages and behaviors of B. hypoleucus and
B. culicivorus are very similar (Silva 1991;
MBR, pers. obs.), unlike the phaeothlypine-
like B. flaveolus. We originally interpreted
Hayes’ (1995a) statement of “a mixed family
of B. hypoleucus and B. flaveolus (two of
each) at Estancia Fonciere [depto. Concep-
cion]” as meaning mixed-species pairs; his in-
tent, however, was to report “two of each spe-
cies intermingling together” (F. E. Hayes,
pers. comm.). Willis (1986) observed only a
single mixed pair of B. hypoleucus and B. cul-
icivorus-, all other pairs were conspecific. Nev-
ertheless, Silva (1991) documented more ex-
tensive hybridization in southeastern Brazil
and treated the two species as conspecific. To
our knowledge, all apparent hybrid morpho-
types are B. hypoleucus-\\ke,, i.e., ventrally
yellow B. culicivorus-XWe birds with some
white are unknown. Presumably the situation
in Paraguay is similar to that in Brazil (Willis
1986, Silva 1991), with hybridization appar-
ently occumng only at ecotones where the dry
forest inhabiting B. hypoleucus meets the
moist forest inhabiting B. culicivorus. Typical
B. culicivorus have been collected at Parque
Nacional Cerro Cora, depto. Amambay, only
about 130 km east of San Luis (Fig. 1), where
patches of moist forest are present (MNHNP
specimens, Hayes and Scharf 1995).

Screaming Cowbird (Molothrus rufoaxil-
laris). — Throughout most of its range, this
brood parasite has been documented to spe-
cialize on Bay-winged Cowbirds (Molothrus
badius', Friedmann 1963, Fraga 1986); how-
ever, Sick (1993) and Fraga (1996) demon-
strated that M. rufoaxillaris pai'asitizes the
Chopi Blackbird {Gnorimopsar chopi) in the
absence of M. badius. Our limited observa-
tions at San Luis also suggest that. M. rufoax-
illaris is parasiting a host other than M. bad-
ius. Male M. rufoaxillaris and Shining Cow-
birds {Molothrus bonariensis) were observed
displaying within 15 m of each other to black-
ish-appearing female cowbirds in areas where
the open forest and campo interdigitated. Giv-
en that no M. badius were observed at San
Luis, we presume that M. rufoaxillaris was

Rohhins et al. • PARAGUAYAN CERRADO AVIFAUNA

221

paiasitizing Gnoriniopsar chopi, which was
fairly common in the area. Unless male M.
nifoaxillahs and M. bonariensis are vocaliz-
ing, they ai'e extremely difficult to distinguish
under field conditions (Ridgely and Tudor
1989). However, Ridgely and Tudor (1989)
stated that pronounced plumage differences in
females (grayish-brown in M. bonahensis\
blackish in M. nifoaxillaris) could aid in iden-
tification in areas of sympatry. Our work at
San Luis, coupled with museum and literature
reviews, demonstrates that an unknown pro-
portion of female M. bonariensis in southern
South America also have blackish plumage.
Two adult female M. bonariensis collected at
San Luis (KU 88485, MNHNP; others ob-
served) were glossy, bluish-black on all but
the lower ventral surface and tail. Under field
conditions, these females closely resembled
accompanying males; only under excellent
light conditions were plumage differences de-
tected. This male-like plumage in female M.
bonariensis is not limited to Paraguay, as blu-
ish-black females that are very similar to the
San Luis birds have also been collected from
southeastern Brazil and extreme northeastern
Argentina (Friedmann 1927, Sick 1993; KU
65019). Observers should exercise caution in
identifying these cowbirds in areas of sym-
patry based on female plumage patterns. Data
are needed on what proportion of nominate
female M. bonariensis have male-like plum-
age.

DISCUSSION

Silva (1995a) used two criteria — a mini-
mum of 100 species recorded and at least 80
specimens collected — for defining a Cerrado
locality as minimally sampled. Based on his
criteria, our results (181 species, 220 speci-
mens) would be classified as a relatively thor-
ough Cerrado inventory; however, our results
are clearly preliminary. The inclusion of
Krieg’s results (Laubmann 1939, 1940) puts
the Serrania de la San Luis species list at 219.
With additional work, we predict that the
Parque Nacional Serrania San Luis avifauna
will surpass 250 species. Surveys are needed
at the drier northern end of the park, where
such species as Rusty-backed Antwren {For-
micivora rufa) and Rufous-sided Pygmy-Ty-
rant {Euscarthmus rufomarginatus) may oc-
cur. From our limited observations, it is clear

that San Luis is an important site for grassland
and forest Austral migrants, and we predict
that the majority of the species to be added to
the San Luis list will be from this component.
Year round surveys are needed to clarify the
status of many species.

Although San Luis is one of the Cerrado
sites closest to the Chaco, only one species
primarily restricted to the Chaco was found
there: the Great Rufous Woodcreeper (Xiph-
ocolaptes major). Considerable plumage var-
iation in the two specimens collected of this
species obfuscates subspecific determination.
For the Olivaceous Woodcreeper (Sittasomus
griseicapillus), the nominate form, principally
of the Chaco, was present; for the Narrow-
billed Woodcreeper (Lepidocolaptes angusti-
rostris) specimens collected were assignable
to the ventrally heavily streaked nominate
form than to the ventrally unmarked Cerrado
form, L. a. bivittatus.

The Atlantic Forest avifauna is also poorly
represented at San Luis, with only six species
recorded: Reddish-bellied Parakeet (Pyrrhura
frontalis). Variable Screech-Owl (Otus atri-
capillus), Rufous-capped Motmot (Bary-
phthengus ruficapillus). Ochre-breasted Fo-
liage-gleaner (Philydor lichtensteini). Eared
Pygmy-Tyrant {Myiornis auricularis), and the
Bare-throated Bellbird. The minimal Atlantic
Forest avifaunal contribution to San Luis con-
trasts with that of the only other Paiaguyan
Cenado site that has been adequately inven-
toried. A total of 201 species have been re-
corded at nearby Parque Nacional CeiTO Cora
(Hayes and Scharf 1995; R. Clay, pers.
comm.). Twenty-five Atlantic Forest species,
including all six known from San Luis, have
been reported from Cerro Cora. Eliminating
water inhabiting species (herons, waterfowl,
migratory shorebirds, kingfishers, marsh
dwelling passerines), 61 Ceno Cora species
have not been recorded at San Luis, whereas
62 species have been recorded at San Luis and
not Cerro Cora. Additional surveys at both
sites undoubtedly will reduce the uniqueness
of each site; nevertheless, the presence of hu-
mid forest patches at Ceno Cora and campo
at San Luis explain genuine species differenc-
es between these sites. The humid forest ele-
ment at Cerro Cora is composed not only of
species primarily restricted to Atlantic Forest,
but also of species more widely distributed in

222

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

humid forest across much of South America,
such as Plain Antvireo {Dysithamnus menta-
lis) and Red-rumped Cacique {Cacicus hae-
morrhous). Moreover, the difference in the
number of hummingbird species between the
two areas is striking, with ten species [Ru-
fous-throated Sapphire {Hylocharis sapphiri-
na) being added since Hayes and Scharf 1995;
R. Clay, pers. comm.] known from Cerro
Cora, but only two from San Luis. The pro-
nounced avifaunal differences between these
nearby localities support Silva’s (1995c) as-
sertion that additional Cerrado inventories are
needed.

Only 3 of the 29 Cenado endemic species
(Silva 1995c) have been found at San Luis.
We predict that additional work in the area
will document at least 6 more Ceirado endem-
ics: White-winged Nightjar {Caprimulgiis
candicans), White-lored Spinetail {Synallaxis
albilora). Chestnut-capped Foliage-gleaner
(Hylocryptus rectirostris). Collared Cres-
centchest {Melanopareia torquata). Reiser’s
Tyrannulet (Phyllomyias reiseri), and Helmet-
ed Manakin (Antilophia galeata), which have
been found at other Paraguayan Cerrado sites
(Laubmann 1939, 1940; Hayes 1995a; Lowen
et al. 1996a). Hence, maximum Cerrado en-
demic species diversity at San Luis is expect-
ed to be about nine species. Many of the re-
maining Cerrado endemic species have re-
stricted geographic ranges; six have extremely
small distributions (Silva 1995c). Indeed, one
generalization that can be made of the Cerrado
endemics is that most are not widespread in
the region. Even in the Cerrado’s geographic
center (eastern Mato Grosso/west-central
Goias), only about two-thirds of the endemics
have been found. Several factors may contrib-
ute to the somewhat reduced species diversity
in Paraguay, (1) less than an estimated 5% of
the total Cerrado habitat is found in Paraguay;
(2) northern Paraguay is at the southern limit
of this habitat’s cunent distribution; and (3)
Paraguay has not had the geographic isolating
mechanisms that have promoted speciation in
other parts of the CeiTado. However, compa-
rable numbers of endemics to Paraguay are
found in other peripheral Cenado areas, such
as western Mato Grosso and northern Goias;
hence, the Paraguayan Cenado is not espe-
cially depauperate when compared across the
range of the habitat.

The importance of Parque Nacional San
Luis cannot be overly stressed. Since our in-
ventory the area that we inventoried was trad-
ed for an adjacent, less human impacted sec-
tor, formerly known as Estancia Garay Kue
(R. Clay and A. Madrono, pers. comm.). The
new park boundary apparently encloses about
4500 ha of relatively pristine grassland. This
takes on special conservation significance giv-
en that virtually all grassland habitat in Pai'a-
guay has been moderately impacted by hu-
mans (Clay et al. 1998). Indeed, grasslands
are one of the most threatened habitats on the
continent (Goriup 1988, Bates et al. 1992,
Stotz et al. 1996, Silva et al. 1997). The grass-
lands at San Luis are critical to several avian
species, both resident and migrant. Three of
the neai-threatened avian species are grass-
land-inhabiting: Greater Rhea {Rhea ameri-
cana). Cock-tailed Tyrant {Alectnirus tricol-
or), and Dark-throated Seedeater {Sporophila
ruficollis). Moreover, we suspect that the
poorly-known and critically endangered
White-winged Nightjar {Caprimulgiis candi-
cans) occurs in the San Luis area. This spec-
tacular nightjar was only recently documented
in Paraguay (Lowen et al. 1996a), and quite
suiprisingly, in eastern Bolivia (Davis and
Flores 1994). Prior to these recent records it
was known only from a few sites in central
Brazil (Collar et al. 1992). Further surveys of
this pristine grassland may yield additional
grassland specialists, e.g.. Bearded Tachuri
{Polystictus pectoralis) and Sharp-tailed Ty-
rant {Cidicivora caiidacuta), that are declining
(Collar and Wege 1995, Stotz et al. 1996,
Parker and Willis 1997).

We commend the Paraguayan government
for having the foresight to establish San Luis
and Cerro Cora national parks. We hope that
our data will help underscore the importance
of these parks, and that this information may
be used in setting conservation priorities for
the rapidly disappearing Cenado.

ACKNOWLEDGMENTS

Director O. Romero of Departamento del Inventario
Biologico Nacional and MNHNP was indispensable in
too many ways to mention. C. Fox. director of Direc-
cion de Parques Nacionales y Vida Silvestre graciously
granted permits for our work al San Luis National
Park. A. L. Aquino, Director of CITES-Paraguay,
kindly provided logistical help in getting to San Luis.
We thank San Luis park guard, A. Acosta, for accom-

Rohhins et al. • PARAGUAYAN CERRADO AVIFAUNA

223

modating us. KU herpelologist J. Simmons provided a
litany of stories, recorded weather data, and kept us
from harm’s way of the Neotropical rattlesnake. D.
Stotz, T. Schulenberg, M. Adams, R. Clay, and P.
Sweet provided information from specimen material
not available to us. A. Whittaker kindly shared his re-
cordings of Philydor dimidiatiis. K. McVay translated
German literature and C. Freeman clarified vegetation
nomenclature. R. Clay, M. Foster, F. Hayes, A. Madro-
no, J. M. C. da Silva, and T. Peterson provided valu-
able comments on the manuscript. We thank the Field
Museum of Natural History (FMNH; J. Bates, D. Wil-
lard) and the Fos Angeles County Museum of Natural
History (FACMNH; K. GaiTett) for loan of specimens.
Funding in part was through the Kansas University
Ornithology Interest Group. We especially want to
thank the Burroughs Audubon Society for their gen-
erous support.

LITERATURE CITED

Ab’Saber, a. N. 1977. Os Dommios morfoclimaticos
da America do Sul. Primerira Aproxima9ao. Geo-
morfologia 52:1-21.

Bates, J. M., T. A. Parker, III, A. P. Capparella, and
T. J. Davis. 1992. Observations on the campo, cer-
rado, and forest avifaunas of eastern Dpto. Santa
Cruz, Bolivia, including 21 species new to the
country. Bull. Brit. Ornithol. Club 1 12:86-98.
Belton, W. 1985. Birds of Rio Grande do Sul, Brazil.
Part 2. Formicariidae through Corvidae. Bull. Am.
Mus. Nat. Hist. 180:1-242.

Brown, L. H. and D. Amadon. 1968. Eagles, hawks,
and falcons of the world. McGraw-Hill Book Co.,
New York.

Chapman, F. M. 1923. The distribution of the motmots
of the genus Momotus. Bull. Am. Mus. Nat. Hist.
48:27-59.

Chesser, R. T. 1997. Patterns of seasonal and geo-
graphical distribution of Austral migrant flycatch-
ers (Tyrannidae) in Bolivia. Ornithol. Monogr. 48:
171-204.

Christiansen, M. B. and E. Fitter. 1997. Species loss
in forest bird community near Lagoa Santa in
southeastern Brazil. Biol. Conserv. 80:23-32.
Clay, R. R, D. R. Capper, J. M. Barnett, I. J. Bur-
field, E. Z. Esquivel, R. Farina, C. P. Kennedy,
M. Perrens, and R. G. Pople. 1998. White-
winged Nightjars Caprimidgus ccindiccins and cer-
rado conservation: the key findings of Project
Aguara N. 1997. Cotinga 9:52-56.

Collar, N. J., L. P. Gonzaga, N. Krabbe, A. Mad-
rono Nieto, L. G. Naranjo, T. A. Parker, III,
AND D. C. Wege. 1992. Threatened birds of the
Americas. The ICBP/IUCN Red Data Book, third
ed. Part 2. Inter. Council for Bird Preservation,
Cambridge, U.K.

Collar, N. J. and D. C. Wege. 1995. The distribution
and conservation status of the Bearded Tachuri
Polystictus pectondis. BirdLife Int. 5:367-390.
Collar, N. J., M. J. Crosby, and A. J. Stattersfield.
1994. Birds to watch 2: the world list of threat-

ened birds. BirdLife International, Cambridge,
U.K.

CuRSON, J., D. Quinn, and D. Beadle. 1994. Warblers
of the Americas. Houghton Mifflin Co., Boston,
Massachusetts.

Davis, S. E. and E. Flores. 1994. Firsi record of
White-winged Nightjar Caprimulgus cundicans
for Bolivia. Bull. Brit. Ornithol. Club 114:127—
128.

Del Hoyo, J., A. Elliot, and J. Sargatal (Eds.).
1994. Handbook of birds of the world. Vol. 2.
New World vultures to guineafowl. Lynx Edi-
cions, Barcelona, Spain.

Direccion de Barques Nacionales y Vida Silvestre.
1993. Plan estrategico del Sistema Nacional de
Areas Silvestres Protegidas del Paraguay (SINA-
SIP). Ministerio de Agricultura y Ganaden'a, Sub-
secretan'a de Estado de Recursos Naturales y Me-
dio Ambiente, Direccion de Barques Nacionales y
Vida Silvestre, Asuncion, Paraguay.

Ericson, P. G. P. and L. a. Amarilla. 1997. First ob-
servations and new distributional data for birds in
Paraguay. Bull. Brit. Ornithol. Club 1 17:60—67.

Fraga, R. M. 1986. The Bay-winged Cowbird (Mol-
othrus badius) and its brood parasites: interac-
tions, coevolution and cooperative efficiency.
Ph.D. diss., Univ. of California, Santa Barbara.

Fraga, R. M. 1996. Further evidence of parasitism of
Chopi Blackbirds (Gnorimopsar chopi) by the
specialized Screaming Cowbird (Molothrus ru-
foaxillaris). Condor 98:866-867.

Friedmann, H. 1927. A revision of the classification
of the cowbirds. Auk 44:495-508.

Friedmann, H. 1963. Host relations of the parasitic
cowbirds. U.S. Natl. Mus. Bull. 223:1-276.

Goriup, P. D. (Ed.). 1988. Ecology and conservation
of grassland birds. Inter. Council for Bird Preser-
vation and Smithsonian Institution Press, Wash-
ington, D.C.

Haeter, j. 1985. Avian zoogeography of the Neotrop-
ical lowlands. Ornithol. Monogr. 36:113—146.

Hayes, F. E. 1995a. Status, distribution, and biogeog-
raphy of the birds of Paraguay. Monogr. Field Or-
nithol. 1:1-230.

Hayes, F. E. 1995b. Definitions for migrant birds: what
is a Neotropical migrant? Auk 112:521—523.

Hayes, F. E. and P. A. Scharf. 1995. The birds of
Parque Nacional Ceno Cora, Paraguay. Cotinga
4:20-24.

Hayes, F. E., Scharf, P. A., and R. S. Ridgely. 1994.
Austral bird migrants in Paraguay. Condor 96:83-
97.

Hellmayr, C. E. 1935. Catalogue of birds of the
Americas. Field Mus. Nat. Hist., Zool. Ser. 13(8):
1-541.

Laubmann, a. 1939. Die Vogel von Paraguay. Vol. 1.
Strecker und Schroder, Stuttgart, Germany.

Laubmann, A. 1940. Die Vogel von Paraguay. Vol. 2.
Strecker und Schroder, Stuttgart. Germany.

Lowen, j. C., L. Bartrina, T. M. Brooks, R. P. Clay,
AND J. A. Tobias. 1996a. Project YACUTINGA

224

THE WILSON BULLETIN • Vol. HI. No. 2. June 1999

'95; bird surveys and conservation priorities in
eastern Paraguay. Cotinga 5:14-19.

Lowen, J. C., L. Bartrina, R. P. Clay, and J. A. To-
bias. 1996b. Biological surveys and conservation
priorities in eastern Paraguay. CSB Conservation
Publications, Cambridge, U.K.

Madrono Nieto, A. and E. Z. Esquivel. 1995. Re-
serva Natural del Bosque Mbaracayu: su impor-
tancia en la conservacion de aves amenazadas,
cuasi-amenazadas y endemicas del bosque Atlan-
tico. Cotinga 4:52-57.

Marin, M. 1997. Species limits and distribution of
some New World spine-tailed swifts (Chaetura
spp.). Ornithol. Monogr. 48:431-444.

McEarlane, R. W. 1974. Unusual avian migrants in
Tarapaca. Idesia 3:181-184.

Parker, T. A., Ill and E. O. Willis. 1997. Notes on
three tiny grassland flycatchers, with comments on
the disappearance of South American fire-diver-
sified savannas. Ornithol. Monogr. 48:549—555.

Remsen, j. V., Jr. and M. A. Traylor, Jr. 1983. Ad-
ditions to the avifauna of Bolivia Part 2. Condor
85:95-98.

Remsen, J. V., Jr. and M. A. Traylor, Jr. 1989. An
annotated list of the birds of Bolivia. Buteo
Books, Vermillion, South Dakota.

Ridgely, R. S. and G. Tudor. 1989. The birds of
South America. Vol. 1. The oscine passerines.
Univ. of Texas Press, Austin.

Ridgely, R. S. and G. Tudor. 1994. The birds of
South America. Vol. 2. The suboscine passerines.
Univ. of Texas Press, Austin.

Robbins, M. B. and T. A. Parker, III. 1997. Voice
and taxonomy of Caprimulgus (rufus) otiosus
(Caprimulgidae), with a reevaluation of Capri-
miilgii.s rufus subspecies. Ornithol. Monogr. 48:
601-607.

Short, L. L. 1975. A zoogeographic analysis of the

South American Chaco avifauna. Bull. Am. Mus.
Nat. Hist. 154:167-352.

Sick, H. 1993. Birds in Brazil. Princeton Univ. Press,
Princeton, New Jersey.

Silva, J. M. C. da. 1995a. Avian inventory of the Cer-
rado Region, South America: implications for bi-
ological conservation. Bird Conserv. Int. 5:291 —
304.

Silva, J. M. C. da. 1995b. Biogeographic analysis of
the South American cerrado avifauna. Steenstru-
pia 21 :49-67.

Silva, J. M. C. da. 1995c. Birds of the cerrado region.
South America. Steenstrupia 21:69—92.

Silva, J. M. C. da, D. C. Oren, J. C. Roma, and L.
M. P. Henriques. 1997. Composition and distri-
bution of the avifauna of an Amazonian upland
savanna, Amapa, Brazil. Ornithol. Monogr. 48:
743-762.

Silva, W. R. 1991. Padroes ecologicos, bioacusticos,
biogeograficos e filogeneticos do complex Basi-
leuterus culicivorus (Aves, Parulidae) e demais
especies brasileiras do genero. Ph.D. diss., Univ.
Estadual de Campinas, Brazil.

Snow, D. W. 1973. Distribution, ecology, and evolu-
tion of the bellbirds (Procnias, Cotingidae). Bull.
Brit. Mus. (Nat. Hist.) Zool. 25:369-391.

Snow, D. 1982. The cotingas. Cornell Univ. Press, Ith-
aca, New York.

Stotz, D. E, j. W. Eitzpatrick, T. A. Parker, III, and
D. K. Moskovits. 1996. Neotropical birds: ecol-
ogy and conservation. Univ. of Chicago Press,
Chicago, Illinois.

Willis, E. O. 1979. The composition of avian com-
munities in remanescent woodlots in southern
Brazil. Pap. Avulsos Zool. (Sao Paulo) 33:1-25.

Willis, E. O. 1986. Vireos, wood warblers and war-
blers as ant followers. Gerfaut 76:177—186.

Willis, E. O. and Y. Oniki. 1993. New and confimied
birds from the state of Sao Paulo, Brazil, with
notes on disappearing species. Bull. Brit. Ornithol.
Club 113:23-34.

Rohhins el cil. • PARAGUAYAN CERRADO AVIFAUNA

225

APPENDIX. Avian species
cepcion, Paraguay.

and their relative abundance in

Parque Nacional Serrania San Luis, depto. Con-

Relative

abundance^

Status^

Habitat‘s

Documentation**

Rhea americana'^

Crypturellus imdulatiis

Cryptiirellus parvirostris

c,v

Rhynchotus rufescens

Nothura maculosa

Phalacrocorax brasilianus

Syrigma sibilatrix

Ardea alba

Bubulcus ibis

Butorides striatus

Nycticorax nycticorax

Phimosus infuscatus

Theristiciis caiidatus

Mycteria americana

Corcigyps atratus

Cathartes aura

Sarcoramphus papa

Cairina moschata

Leptodon cayanensis

Elanus leucurus

Rostrhamus sociabilis

Ictinia plumbea

Accipiter bicolor

Buteogallus meridionalis

Harpyhaliaetus coronatus

Buteo magnirostris

Spizaetus ornatus

Caracara plane us

Milvago chimachima

Herpetotheres cachinnans

Micrastur ruficollis

Falco sparverius

Crax fasciolata

Aramides cajanea

Porzana albicollis

Cariama cristata

Vanellus chilensis

Rostratula semicollaris

Tringa solitaria

Calidris fuscicollis

Gallinago paraguaiae

Columba cayennensis

Columbina squammata

Columbina talpacoti

Columbina picui

Claravis pretiosa

c,v

Leptotila verreauxi

Leptotila rufaxilla

Anodorhynchus hyacinthinus

Ara chloropterus

Ara maracana

Aratinga leucophthalmus

c,v

Aratinga aurea

Pyrrhura frontalis

c,v

Brotogeris chiriri

Pionus maximiliani

Amazona aestiva

226

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

APPENDIX. Continued

Relalive

abundance^

Status*’

Habitat*-'

Documentation**

Coccyziis melacof-yphus

Piayci cayana

Tupera naevia

Crotophaga major

Crotophaga ani

Guira guira

Otus atricapilliis

Otus choliba

Piilsatri.x per.spicillata

Bubo virginianu.s

Glaiicidium bra.'iilianiim

c,v

Speotylo cimicularia

Lurocalis seniitorquatu.s

Chordeile.s minor

Podager nacunda

Nyctidromii.s albicolli.s

c,v

Caprimulgits rufiis

c,v

Caprimulgiis parx iilas

P*ny

Nyctibius gri.seiis

c,v

Chaetura meridionali.s^

c,v

TIudurania furcata

Hylocluiris duy.stira

c,v

Trogon ciirucui

Momotus momota

c,v

Baryphthengii.s ruficapilliis

Chloroceryle americana

Chlorocerxle inda

Ny.stalus chacuru

Ny.slalus macular u.s

Pteroglossu.s castanotis

Ramphastos toco

Picumnu.s cirratus

MeUmerpe.s candid u.s-

Picoide.s mi.xtus

Veniliorni.s pa.s.serinu.s

c,v

Picul u.s ch ry.soch loros

Colapte.s campestris

Celeus lugubris

Dryocopu.s 1 ineatu.s

Campepliilus melanoleucos

Furnarius rufu.s

Sclioeniopbyla.x pluyganophila

c,v

S VI la lla.xi.s a ! be seen s

Pbacellodomu.s rufifrons

Philydor dimidiatiis

c,v

Philvdor Uchtensteini

c,v

Philydor rufu.s

c,v

Sirtasomus griseicapiUus

c,v

Xiphocolapte.s major

C,V

Dendrocolaptes platyro.stris

c,v

Lepidocolapte.s an gusli rostris

• c,v

ThamnophUus caerule.scen.s

c,v

HerpsHochmus atricapilliis

C,V '

Camptostoma ob.soletum

P?y

c,v

Pliaeomyia.s murina

Suiriri suiriri

Myiopagis caniceps

Mviopagis viridicata

p9*

c,v

Flaenia jlavogaster

Robbins el cil. • PARAGUAYAN CERRADO AVIF'AUNA

227

APPENDIX. Conlinued.

Kelalive

abundance*'

•Slams*’

Hahital^

Documentation^*

Elaenia albiceps

Elaenia pcin irostris

Inezici inonuita

P?*

c,v

Leptopogon amciiirocephaliis

c,v

Corythopis lielalandi

c,v

My iorn is atiriciilaris

Hemitriccus margaritacei venter

c,v

Tohnomvius sulphiirescens

c,v

Platvrinchiis mvstaceus

c,v

Myiophobus fasciatiis

Contopns cinereus

Lathrotriceus euleri

P?*

c,v

CnemotriccHS fiiscatus

P?*

c,v

Py roceph ulus rub inns

Xobnis cinerea

Xolmis velata

Hymenops perspicillatus

Eluvicola leucocephala

Alectrurus tricolor

P?*

Gubernetes yetcipa

c,v

Machelornis rixosus

c.v

Casiornis riifa

c,v

Sirvstes sibilator

c,v

Myiarchus swciinsoni

P?*

c,v

Myiarchiis tyrannulus

c,v

Pitangus sulphiiratus

Megarynchus pitangua

Myiodynastes macidatiis

C.V

Legatus leucophaius

Empidonornus variiis

Griseotyrcinniis aiirantioatrocristatus

C.V

Tyrannus melancholicus

Tyrannus scivana

Xenopsaris albinucha

Pachyramphus viridis

C.V

Pachyramphiis castaneiis

Pachyramphus polychopterus

Pachyramphus validus

P?n

Tityra cayana

Tityra inquisitor

c.v

Pvroderus sculatus

Procnias nudicollis

R,C

P?*

c.v

Pipra fasciicauda

Oxynmcus cristatus

Progne chalybea

Phaeoprocne tapera

Tachycineta leucorrhoa

Stelgidopteryx ruficollis

Cyanocorax cyanomelas

C.V

Cyanocorax cristatelhis

Cyanocorax chrysops

Troglodytes aedon

Polioptila dumicola

C.V

Turdus riifiventris

C.V

Turdus albicollis

C.V

Turdus amaurochalinus

Catharus fiiscescens

Mimus saturninus

228

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

APPENDIX. Continued.

Relative

abundance'

Status'’

Habitat'

Documentation‘*

Mimus triurus

Vireo olivaceus chivi

P?*

c,v

Cyclarhis gujanensis

Parula pitiayumi

Geothlypis aequinoctialis

Basileuterus ftaveolus

c,v

Basileuterus hypoleucus

P*y

c,v

Conirostrum speciosum

Euphonia chlorotica

Thraupis sayaca

Eucometis penicillata

c,v

Tachyphonus rufus

Piranga flava

Trichothraupis melanops

Cypsnagra hirundinacea

Nemosia pileata

Hemithraupis guira

c,v

Saltator similis

c,v

Saltator atricollis

c,v

Paroaria coronata

Coryphospingus cucullatus

c,v

Arremon flavirostris

c,v

Volatinia jacarina

Sporophila plumbea

c,v

Sporophila collaris

Sporophila caerulescens

Sporophila bouvreuil

Sporophila hypoxantha

Sporophila ruficollis

Sicalis flaveola

Embernagra platensis

Emberizoides herbicola

c,v

Ammodramus humeralis

c,v

Zonotrichia capensis

Gnorimopsar chopi

Pseudoleistes guirahuro

Leistes superciliaris

Molothrus bonarien.sis

Molothrus rufoaxillaris

c,v

Scaphidura oryzivora

Icterus cayanensis

c,v

Cacicus chrysopterus

c,v

Psarocolius decumanus

Carduelis magellanica

Passer domesticus

^ Relative abundance criteria based on our work. X = single observation; R = rare, not recorded daily when in appropriate habitat, and only in small
numbers when recorded, <5 individuals/day; U = uncommon, recorded in small numbers daily when in appropriate habitat, l-IO individual.s/day; C =
common, recorded daily in large numbers when in appropriate habitat, >10 individuals/day.

•’ Status: P = permanent resident; a breeder with at least a few individuals present throughout the year. Most species given this designation are not
known to make seasonal movements; virtually all with P? are presumed to have Austral migrant populations present at appropriate seasons; B = breeder
only. Austral migrant; * = at least one individual/species had enlarged gonads indicative of breeding, n = nest found, y = young observed; V = visitor;
nonbreeder, but may breed in adjacent areas to the park; IM = Intratropical migrant; AM = Austral migrant; NM = Nearctic migrant. Migrant terminology
follows Hayes (l99.Sb); ? = status uncertain. Hayes et al. (1994) and Chesser (1997) were consulted for Austral migrant status.

Habitat based on our work; I = open gra.ssland (campo); 2 = isolated woodlots in grassland, forest edge; 3 = forest; 4 = aerial.

‘'Documentation: C = collected; V = voice recorded; P = photographed; we use this designation only when a species was photographed but not "C”
or “V". .S = sight observation(s) only. L = based on specimens by the Kreig expedition to this region; referenced in Laubmann (1939, 1940); .see Methods
for explanation.

'.Species in boldface are threatened or near-threatened (Collar et al. 1992, 1994).

f Following Marin (1997).

Wilson Bull., 1 1 1(2), 1999, pp. 229-235

NOTES ON THE AVIFAUNA OF TABASCO

KEVIN WINKER,' STEFAN ARRIAGA WEISS, ^ JUANA LOURDES TREJO P.,^

AND PATRICIA ESCALANTE P'

ABSTRACT. — Tabasco, a Mexican state nearly half the size of the country of Costa Rica, lies juxtaposed
between the Isthmus of Tehuantepec and the Yucatan Peninsula. This state hosts a diverse Neotropical resident
avifauna, is a significant wintering area for Nearctic-Neotropic migrants, and has important biogeographic sig-
nificance. Surprisingly little recent ornithological study has occurred in Tabasco; the last major publication treats
data from 1939. Field work in March 1996 and recent specimens add nine species to the state list: Caprimulgus
vociferus, Chaetura vaiixi. Campy lopterus excellens, Enipidonax albigularis, Thryothorus modestus, Turdus in-
fuscatus. Myadestes unicolor, Limnothlypis swainsonii, and Vermivora ruficapilla. The status of 26 other species
is discussed. Further evidence of lowland forests being used as temporary refugia by birds from higher elevations
is also considered. Received 20 May 1998, accepted 19 Nov. 1998.

“Compared to many of the Mexican States, Tabasco has been slighted ornithologically. Anything collectors
could obtain in Tabasco could also be found in southern Veracruz with less effort. If the collector were to
exert the effort necessary to enter Tabasco, he might as well go the entire distance and travel on to the
Yucatan Peninsula, an area with many unique and peculiar forms.” Berrett (1962:4).

Tabasco, a state of approximately 24,600
km^, is situated on the northeastern side of the
Isthmus of Tehuantepec on the Gulf of Mex-
ico (Fig. 1). The ornithological neglect of Ta-
basco noted by Berrett (1962) has generally
continued for the past 35 years. Berrett (1962)
added 166 species to those previously known
from Tabasco and reported a total of 457 spe-
cies from the state. The bibliographic and mu-
seum database research of Centeno ( 1 994)
added 73 species to the total reported by Ber-
rett (1962), but the knowledge of avian dis-
tribution within Tabasco and specimen repre-
sentation of the birds of the state remain rel-
atively poor. Avian specimens apparently do
not even exist from 3 of the state’s 17 muni-
cipios (the equivalent of counties in the
U.S.A.; Centeno 1994).

Perhaps because the two major works on
Tabasco birds in the last 55 years are not read-
ily available (Berrett 1962, Centeno 1994) and

' Conservation & Research Center, NZP, Smithson-
ian Institution, 1500 Remount Rd., Front Royal, VA
22630.

^ Division Academica de Ciencias Biologicas, Univ.
Juarez Autdnoma de Tabasco, Km 0.5 Car. Villahcr-
mosa-Cardenas, Villahermosa, Tabasco 86000, Mexi-
co.

^ Instituto de Biologi'a, Depto. de Zoologfa, Univ.
Nacional Autonoma de Mexico, Apartado Postal 70-
153, Mexico, DF 04510.

■* Present address: Univ. of Alaska Museum, 907 Yu-
kon Dr., Fairbanks, AK 99775-6960.

■’Corresponding author; E-mail; ffksw@uaf.edu

because specimen representation is sparse, er-
rors and omissions regarding the distribution
and occurrence of birds in the state have been
fairly common (e.g., Peterson and Chalif
1973, Howell and Webb 1995). The last major
publication on the birds of Tabasco is the out-
dated report by Brodkorb (1943), which was
based on a collection made in the region by
the botanist Eizi Matuda and his assistants in
1939.

In this paper we report on our recent efforts
to increase existing knowledge of the birds of
Tabasco. In March 1996 we held a collabo-
rative field workshop in the municipios of
Centla and Huimanguillo. In Centla (the
northeastern most municipio in the state, bor-
dering Campeche), our field studies were con-
ducted from 4—13 March on the Rio Grijalva
in the Reserva de la Biosfera Pantanos de
Centla (Fig. 1). Our studies were focused
mainly on two sites in the area of 18° 29' N,
92° 38' W. The first site was a remnant tract
of mangrove {Rhizophora mangle) and puktal
(Bucida buceras) forest (a selva perennifolia,
or evergreen forest) along the banks of An oyo
Polo, a tributary of the Rio Grijalva about 4
km south of Frontera. The other area was a
partially flooded, shrubby pasture on the
banks of the Rio Grijalva and Rio San Pedrito,
about 14 km south of Frontera. Additional ob-
servations were made on water and land be-
tween these two sites.

In Huimanguillo, the southwestemmost

229

230

THE WILSON BULLETIN • Vol. HI, No. 2, June 1999

FIG. 1. The Mexican state of Tabasco, with its capital city of Villaheimosa and the location of our study
sites at the Biosphere Reserve of Pantanos de Centla and near the ejido of Malpasito.

municipio (bordering Veracruz and Chiapas),
our field studies were conducted from 14-22
March in the southwesternmost corner of the
state, below Cerro La Pava, west of Ejido
Malpasito, and approximately 5 km west of
the state’s border with Chiapas (ca 17° 20' N,
93° 36' W, 300-500 m elevation; Fig. 1). This
site consisted of open fields, pastures, small
fruit and coffee plantations, acahual (second
growth forest), and remnant rainforest patches
restricted to mountain crevices and aiToyos.

In addition to specimens and observations
obtained during our field work, we include
some additional specimen-based information
for the birds of Tabasco from the national or-
nithological collection of Mexico (Coleccion
Nacional dc las Aves — CNAV), which is
housed at the Instituto de Biologfa, Departa-

mento de Zoologia, Universidad Nacional Au-
tonoma de Mexico, in Mexico City. Our spec-
imens representing new records for the state
were also deposited in CNAV. A number of
records we report as new for the state were
predicted by Benett (1962). Several of our re-
cords expand the known distribution of par-
ticular species within the state; many of these
were also foreseen by Beirett (1962).

SPECIES ACCOUNTS

Broad-winged Hawk (Biiteo platyptenis). —
Berrett ( 1962) recorded only two observations
of this species in Tabasco, both of migratory
flocks (28 March 1960, 17 April 1961). What
appears to be the first state specimen, a female
in migration, was taken at Balneario Agua

Winker el al. • TABASCO AVIFAUNA

231

Blanca, Macuspana, on 13 March 1989
(CNAV 13,523).

Ruddy Crake (Laterallus ruber). — BeiTCtt
(1962) noted that the species had been re-
ported from only two areas in the state (near
Tenosique and Teapa), but that it was probably
more widespread. We found the species to be
very common at Pantanos de Centla, hearing
its distinctive vocalizations daily during our
stay in the aiea.

Common Snipe (Gallinago gallinago). —
Berrett ( 1 962) reported few Tabasco records
of this wintering migrant, including only three
sightings of his own. All but one of the re-
cords (from Sanchez Magallanes) were from
southern, inland areas of the state. We ob-
served at least one individual per day at Pan-
tanos de Centla on 8-10 March.

Sandwich Tern (Sterna sandvicensis). —
Berrett (1962) reported collecting the first
specimens for the state at Sanchez Magallanes
and Miramar; his observations of the species
did not extend beyond these areas. We found
this tern to be common on the Rio Grijalva
from Frontera to an area about 5 km to the
south. Two to approximately 40 individuals
were seen almost daily during our stay in this
area.

Least Tern (Sterna antillarum). — Berrett
(1962) reported only a single specimen from
the state taken by D. M. Lay on 2 May 1960
near Jonuta on the Rio Usumacinta. Apart
from the many individuals Lay observed near
Jonuta on 6 April and 2 May 1960, only two
other observations (both of single individuals)
were reported by Berrett: one from Chable on
the Rio Usumacinta and the other from near
Chontalpa (Huimanguillo) on the R(o Grijal-
va. These are all inland records. On 12 March
at Pantanos de Centla we had the opportunity
to closely observe approximately 40 individ-
uals feeding over the Rio Grijalva near Fron-
tera in loose aggregation with individuals of
the preceding species.

Mangrove Cuckoo (Coccyzus minor). —
Berrett ( 1 962) noted few Tabasco records and
felt that the species was restricted to coastal
areas. We note a male taken from a decidedly
inland locality 6 km S of Huimanguillo on 6
March 1984 by H. Munoz (CNAV 2,904).

Common Parauque (Nyctidromus albicol-
lis). — Berrett (1962) noted that his specimens
of this permanent resident taken between 1 1

March and 7 May were in breeding condition.
An individual found during the day in river-
bank mangrove/selva forest on 5 and 6 March
acted as though it was nesting when we
flushed it. We did not find a nest, but on 7
March found a single egg laid on the leaf litter
in one of our mist net lanes. We captured and
banded the bird when it returned to this nest.
A second egg was laid by 12 March (our next
visit to the site).

Whip-poor-will (Caprimulgus vociferus). —
New record. Berrett (1962) noted only a sin-
gle report of this species from Tabasco, that
of Rovirosa (1887). Rovirosa’s observations
are notoriously suspect, however; Brodkorb
(1943:8) stated that “ . . . scarcely any reli-
ance can be placed on his records which have
not been confirmed by other workers.” In fact,
Rovirosa (1887) simply pooled all of his ca-
primulgid observations under this single spe-
cific epithet, apparently not knowing what
species he had observed and perhaps liking
this particular name. Under “Antrostomus vo-
ciferus" he stated that various species of ca-
primulgid occurred in the state from January
to May. Thus, based on present evidence (i.e.,
Berrett 1962), it seems likely that Rovirosa
was discussing one or more of the four other
caprimulgids that have since been found to oc-
cur in the state. Given its commonness and
vocal habits, his observations probably con-
sisted mostly of Nyctidromus albicollis, but
this is speculative, and does not explain his
lack of records from the second half of the
year.

KW distinctly heard a Whip-poor-will sing-
ing at Pantanos de Centla on 9 Maixh. The
song was of the “Mexican” subspecific group
(C. V. arizonae/oaxacae/chiapensis), which
Howell and Webb (1995) suggested might oc-
cur in lowlands during the nonbreeding sea-
son. Subsequently, we discovered that a spec-
imen of this subspecies group had been taken
on 16 March 1984, 10 km S of Chontalpa,
Municipio de Huimanguillo, by F. Ornelas
(CNAV 2,917).

Vaux’s Swift (Chaetura vaiixi). — New Re-
cord; no specimen. A flock of approximately
15 individuals was seen and heard at close
range above Malpasito, Huimanguillo on 15
March, and a flock of 35-40 was seen each
day 16-18 March. The first migrant Chaetura
pelagica of the spring (a species that might be

232

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

confused with C. vaiixi) were seen and heard
on 21 March.

Violet Sabrewing (Campylopterus hemileu-
curus). — The four specimens recorded by
Berrett (1962) for the state all came from near
Teapa. The single additional sight record not-
ed was from near Chontalpa. We collected two
females on 17 and 19 March at Malpasito.
Neither had yet entered reproductive condi-
tion.

Long-tailed Sabrewing (Campylopterus ex-
cellens). — New Record. An adult female of
this species was taken at Malpasito in remnant
forest at the edge of a small arroyo. This is
the first record of this species for the state,
and also the northeastemmost occurrence of
the species, which until recently has been con-
sidered to have a very restiicted range (see
Winker et al. 1992a, Howell and Webb 1995).
Howell and Webb (1995) based their exten-
sion of the species’ range into Chiapas on a
previously misidentified specimen in the
American Museum of Natural History. Reex-
amination of this specimen (KW) suggests
that Howell and Webb (1995) were correct; it
is a Campylopterus excellens, and not a C.
curvipennis as originally identified. Together,
these two specimens suggest an approximate
doubling of the species’ total geographic
range.

Pygmy Kingfisher (Chloroceryle aenea). —
Hitherto, the only records for this species in
the state were from southern, inland localities
(Balancan and Ocuapan: Benett 1962; Chon-
talpa: CNAV 2,933). We collected a non-
breeding female at Pantanos de Centla on a
small, well-forested tributary (Arroyo Polo) of
the Rio Grijalva on 6 March.

Wedge-billed Woodcreeper (Glyphorhyn-
chus spirurus). — Recorded by Beirett (1962)
as an uncommon resident of the rainforest belt
of the state, records were lacking from the
westernmost rainforest in the state (Municipio
de Huimanguillo). We collected a nonbreeding
female in a patch of acahual (second growth
forest) at the base of the local mountains at
Malpasito on 17 March.

Ochre-bellied Flycatcher (Mionectes olea-
gineus). — Although Berrett (1962) recorded
this species near Chontalpa, he did not find it
to be particularly common. We found it to be
rather common at Malpasito, collecting seven
individuals 19-21 March. Their abundance at

the site was associated with the weather; they
were undetected before a relatively cold norte
(cold, wet weather system from the north) ar-
rived on 19 March.

Sepia-capped Flycatcher (Leptopogon
amaurocephalus). — A male taken on 20
March at Malpasito (testes moderately en-
larged) extends the range of this species with-
in the state to the westernmost occurrence of
the rainforest belt.

Sulphur-rumped Flycatcher (Myiobius sul-
phureipygius). — As with the previous species,
Berrett (1962) documented this species as a
permanent resident of the rainforest belt, but
lacked records for the westernmost rainforest
in the state. We collected a male with unen-
larged testes in acahual at Malpasito on 20
March.

White-throated Flycatcher (Empidonax al-
bigularis). — New Record. We collected what
appear to be the first specimens for the state
at Pantanos de Centla on 10 March in a shrub-
by pasture. Both birds were males with no fat
and a heavy molt that included body, wings,
and tail.

Great Crested Flycatcher (Myiarchus crin-
itus). — Berrett (1962) noted a single specimen
and a few sightings of this species in Tabasco,
all from May 1961 and neai' Balancan. We
note two other specimens, both males, taken
on 13 Mai'ch 1984 and 29 September 1965
(indicating a presence in fall migration as well
as spring), near Comalco and on the Rio San
Pablo, Municipio de San Pablo by F. Ornelas
and R. W. Dickerman, respectively (CNAV
3,491 and 17,215).

Gray-collared Becard (Pachyramphus ma-
jor).— An adult male, appai'ently the second
specimen for the state (Benett 1962), was tak-
en in 8-10 m acahual on 20 March at Mal-
pasito. The first specimen, also a male, was
taken in the eastern part of the state near Re-
forma, Balancan, on 28 May 1939 (Brodkorb
1943).

Mangrove Vireo (Vireo pallens). — BeiTett
(1962) collected the only previous specimen
for the state (male, 13 April 1961) near Villa-
hermosa and noted three additional sightings:
two near Villahermosa and another on the Rio
Usumacinta at Emiliano Zapata (all in April
1961 ). Peterson and Chalif ( 1973) and Howell
and Webb (1995) overlooked these records.
We collected two more individuals in low-

Winker et al. • TABASCO AVIFAUNA

233

lying, shrubby pasture on the Rio Grijalva at
Pantanos de Centla on 9 and 10 March. Both
were females with no fat and unenlarged ova-
ries. At least one more individual was later
seen in this same area.

Blue-headed Vireo (Vireo solitarius). —
Berrett (1962) reported only two observations
of this species in the state (March 1959,
March 1960), both near Teapa. We collected
what are apparently the first two state speci-
mens at Malpasito on 16 and 17 March, a fe-
male with no fat and a male with little fat.

Green Jay (Cyanocorax yncas). — Berrett
(1962) recorded specimens and sightings of
this species from only four localities in the
municipios of Balancan, Centla, and Teapa.
Our sightings of two individuals each on 15
and 21 March at Malpasito indicate that the
species also occurs in the western part of the
state.

Tree Swallow (Tachycineta bicolor). — Ber-
rett (1962) recorded only one specimen for the
state but noted a few additional sightings, in-
cluding several large flocks. Although it re-
mains unclear whether the species spends the
winter or is only a transient in migration, we
add the following records from Pantanos de
Centla: three specimens, all females with little
fat, taken on 8 March 1996, and sightings of
1-150 individuals almost daily from 4—12
March. In addition, a female was taken by R.
W. Dickerman 53 km W of Villahermosa on
3 April 1973 (CNAV 17,723).

Plain Wren (Thryothorus modestus). — New
record. An adult female was taken at Bal-
neario Agua Blanca, Macuspana, in selva me-
diana, by E. Diaz I. on 14 March 1989 (CNAV
13,389).

Slate-colored Solitaire (Myadestes unicoT
or). — New record. We collected a male on 20
March at Malpasito during the same norte as-
sociated with the occurrence of the Black
Robin (below) in the area. Movement of
Slate-colored Solitaires to lowland forest dur-
ing nortes is a frequent occurrence in the Si-
erra de Los Tuxtlas in southern Veracruz (Ra-
mos 1983; KW, pers. obs.).

Black Robin (Turdus infuscatus). — New re-
cord; no specimen. A female or immature
male of this highland species was seen feeding
in a fruiting tree on 20 March at Malpasito on
the last day of a relatively cold norte. It had
probably descended to this lower elevation (ca

450 m) to escape less suitable conditions in
the highlands, a common occurrence in some
areas of southern Mexico (see Winker et al.
1992b).

White-throated Robin (Turdus assimilis). —
Berrett ( 1 962) noted only three records of this
species in Tabasco, one specimen and two
sightings from near Teapa and near Chontalpa,
both southern localities. A male was taken at
Comalcalco (in the northern part of the state)
on 13 March 1984 by F. Ornelas (CNAV
6,776).

Nashville Warbler (Vermivora ruficapil-
la). — New record. Berrett (1962) reported a
brief glimpse of a bird that was probably this
species at Balancan. We collected the first
state specimen, a female with no fat, at Mal-
pasito on 20 March.

Yellow-rumped Warbler (Dendroica coron-
ata). — Berrett (1962) reported only a few rec-
ords for this species, all from December. We-
ber (1945) reported it at La Venta in spring
1943. We took three females with moderate to
heavy fat in shrubby pasture at Pantanos de
Centla on 9-11 March, and observed up to
three other individuals on these days.

Swainson’s Warbler (Limnothlypis swain-
sonii). — New record. The first state specimen,
a male with no fat and with tail characteristics
(uneven growth bars) of an after second year
(ASY) individual, was taken in a mixed forest
of mangrove and selva near the bank of Ar-
royo Polo (Rio Grijalva) near Frontera on 6
March. This bird was not in migratory con-
dition and the habitat it occupied was emi-
nently suitable for wintering (KW, pers. obs.;
see Graves 1998).

White-winged Tanager (Piranga leucop-
tera). — Berrett (1962) reported a few individ-
uals of this species in early and mid-Novem-
ber 1961 near Tenosique and surmised that
they may have been wintering birds from
higher elevations. We saw a single female at
Malpasito on 19 March, during the same norte
associated with the presence of the Black
Robin and Slate-colored Solitaire noted
above.

Orange-billed Sparrow (Arremon aurantii-
rostris). — Benett (1962:375) noted six speci-
mens from the state and referred to the species
as a “rarely recorded permanent resident of
the humid rain forest undergrowth.” Although
we worked intensively in the Malpasito ai'ea

234

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

for four days before encountering any individ-
uals of this species, during the relatively se-
vere norte of 19 and 20 March we collected
four individuals (three females, one male) in
our mist nets that had been in the same lo-
cations the entire time in a small patch of rem-
nant rainforest at the base of the local moun-
tains. This striking movement pattern of in-
creased captures during nortes is typical of
highland species at lowland sites, but not of
lowland species. We surmise that at this site
these typically lower-elevation birds might oc-
cupy less disturbed forest at slightly higher
elevations than the remnant forest available
where our field work was conducted.

Olive Sparrow (Arremonops riifivirga-
tus). — Benett (1962) reported only a single
specimen, but noted that the species was a
common permanent resident north of Balan-
can. These records were overlooked by How-
ell and Webb (1995). We found that the spe-
cies also occurs in the westernmost region of
the state. We collected a female with an un-
enlarged ovary on 18 March at Malpasito and,
in addition, found what appeared to be two
pairs (two singing males with associated in-
dividuals not singing) in a brushy area of
young second growth.

Savannah Sparrow (Passerculus sandwich-
ensis). — Four individuals were seen on 14
March beside scrubby pasture at Pantanos de
Centla on the Rio Grijalva, supplementing the
single specimen and several sightings reported
by Berrett (1962) from a southern part of the
state (Huastecas, Teapa, ca 29 km NE of Tea-
pa).

Lincoln’s Sparrow (Melospiza lincolnii). —
Berrett (1962) included this species as an un-
common wintering bird in the state, reporting
only a single specimen and scattered sight re-
cords from largely southern, inland sites. Our
records add northern, near-coastal records for
the state. We observed a single individual on
4 March at Pantanos de Centla in roadside
scrub, and collected a male with light fat on
1 1 March in shrubby pasture on the Rfo Gri-
jalva.

DISCUSSION

Our records consist of 10 migrant and 25
resident species. Of our nine new records for
the state, only two are of migrant species; sev-
en are resident. This is unusual for an area in

the northern Neotropics, where at the end of
the twentieth century one would expect our
knowledge of the avifauna to be sufficiently
strong that most new records would be gen-
erated by individuals of migrant species. We
consider the preponderance of resident species
in this report to be a reflection of Tabasco’s
status as a poorly known Neotropical region.
Further ornithological study of this region is
fully warranted. Tabasco is nearly half the size
of Costa Rica. It occupies a geographic posi-
tion making it important as a wintering ground
for Nearctic-Neotropic migrants and as a
breeding and nonbreeding area for a diverse
Neotropical resident avifauna. Further, it has
twofold biogeographic significance: one as a
transition zone between the Yucatan Peninsula
and the Isthmus of Tehuantepec (including a
significant portion of the rainforest belt at the
peninsula’s base) and another in its position at
the northern limits of Neotropical families
such as Heliomithidae, Eurypygidae, Galbu-
lidae, Bucconidae, and Pipridae.

Our data provide important new distribu-
tional information for both resident and mi-
grant species in this poorly known region. For
example, our records of Swainson’s Warbler
and the Long-tailed Sabrewing constitute sig-
nificant wintering and probable breeding
range extensions (respectively) for rare spe-
cies of conservation concern (see Howell and
Webb 1995, Winker et al. 1992a).

In addition, our data show inteimittent use
of lowland forest, indicating movements
among “resident” birds. The relatively low-
land sites we studied near Malpasito suddenly
became home to individuals of a number of
species that were not detected during the four
days of intensive field study prior to the arriv-
al of a norte on 19 March. At least five species
(including two new to the state) showed
movement patterns that were strongly tied to
the inclement weather: Mionectes oleagineus,
Turdus infuscatus, Myadestes unicolor, Pir-
anga leucoptera, and Arrenion aurantiirostris.
All of these species probably occupy forest at
higher elevations in this area. An increase in
overall capture rates and census detections
(unpubl. data) suggested increased numbers of
individuals of many species already present at
the lowland sites. These individuals may also
have come from the local highlands.

Temporary use of remnant lowland forest

Winker et al. • TABASCO AVIFAUNA

235

during nortes is a relatively common phenom-
enon in southern Veracruz (Ramos 1983,
Winker et al. 1997) and is an issue of grave
conservation concern; when lowland forests
have diminished to a degree of scarcity, they
are no longer available to individuals seeking
temporary refuge from inclement conditions
in the highlands.

ACKNOWLEDGMENTS

We thank the U.S. National Science Foundation
(NSF INT-9403053) and CONACyT (El 20) for sup-
porting our field studies, and SEMARNAP for issuing
the necessary permits. J. M. Arias R., C. A. Cordero
M., O. E. Escobar R, J. L. Gonzalez A., A. C. IbaiTa
M., G. Lopez S., M. de J. Mendez G., L. Montanez
G., M. Ramirez L., R. E. Sobrino R, and M. Suarez I.
proved to be excellent field companions and contrib-
uted to a stimulating workshop and field experience.
We also thank J. C. Romero, C. A. Jimenez B., P. del
Valle, R. del Valle Reyna, and the staffs of the Pan-
tanos de Centla Field Station and Aguaima for their
help and hospitality during our visits. F. Gonzalez G.
and J. H. Rappole provided helpful comments on an
earlier draft.

LITERATURE CITED

Berrett, D. G. 1962. The birds of the Mexican state
of Tabasco. Ph.D. diss., Louisiana State Univ., Ba-
ton Rouge.

Brodkorb, P. 1943. Birds from the lowlands of south-
ern Mexico. Misc. Publ. Mus. Zool. Univ. Mich.
55:1-88.

Centeno A., B. E. 1994. Estado actual del conoci-

miento de la avifauna de Tabasco: revision biblio-
grafica. Tesis Licenciado en Biologia, Univ. Jua-
rez Autonoma de Tabasco, Villahermosa.

Graves, G. R. 1998. Stereotyped foraging behavior of
the Swainson’s Warbler. J. Field Ornithol. 69: 121 —
127.

Howell, S. N. G. and S. Webb. 1995. A guide to the
birds of Mexico and northern Central America.
Oxford Univ. Press, New York.

Peterson, R. T. and E. L. Chalie. 1973. A field guide
to Mexican birds. Houghton Mifflin Co., Boston,
Massachusetts.

Ramos, M. A. 1983. Seasonal movements of bird pop-
ulations at a neotropical study site in southern Ve-
racruz, Mexico. Ph.D. diss., Univ. of Minnesota,
Minneapolis.

Rovirosa, j. N. 1887. Apuntes para la zoologia de
Tabasco: vertebrados observados en el Territorio
de Macuspana. Naturaleza 7:345-389.

Weber, W. A. 1945. Wildlife of Tabasco and Veracruz.
Nat. Geog. 87(2): 187-2 16.

Winker, K., M. A. Ramos, J. H. Rappole, and D. W.
Warner. 1992a. A note on Campylopterus excel-
lens in southern Veracruz, with a guide to sexing
captured individuals. J. Field Ornithol. 62:339-
343.

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

Winker, K., P. Escalante, J. H. Rappole, M. A. Ra-
mos, R. J. Oehlenschlager, and D. W. Warner.
1997. The evolution and conservation of Wet-
more’s Bush-Tanager: periodic migration and low-
land forest refugia in a “sedentary” neotropical
bird. Conserv. Biol. 11:692-697.

Wilson Bull., 111(2), 1999, pp. 236-242

PREDATION OF SMALL EGGS IN ARTIFICIAL NESTS: EFFECTS
OF NEST POSITION, EDGE, AND POTENTIAL PREDATOR
ABUNDANCE IN EXTENSIVE FOREST

RICHARD M. DEGRAAF,'^^ THOMAS J. MAIER,' AND TODD K. FULLERS

ABSTRACT. — Alter photographic observations in the field and laboratory tests indicated that small rodents
might be significant predators on small eggs, we conducted a field study in central Massachusetts to compare
predation ot House Sparrow (Passer domesticus) eggs in artificial nests near to (5-15 m) and far from (100-
120 m) forest edges and between ground and shrub nests. As in earlier studies in managed northeastern forest
landscapes that used larger quail eggs, predation rates on small eggs in nests at the forest edge did not differ
(P > 0.05) from those in the forest interior for either ground nests (edge = 0.80 vs interior = 0.90) or shrub
nests (edge = 0.38 vs interior = 0.28) after 12 days of exposure. However, predation rates on eggs in ground
nests were significantly higher (P < 0.001) than in shrub nests at both the edge and interior. There were no
significant (P > 0.05) differences in the frequency of capture of the 6 most common small mammal species
between forest edge and interior. Logistic regression analyses indicated a highly significant (P < 0.001) nest
placement effect but very little location or small mammal effect. Predation of small eggs by small-mouthed
ground predators such as white-footed mice (Peromyscus leiicopns) has not been documented as a major factor
in egg predation studies, but use of appropriately-sized eggs and quantification of predator species presence and
abundance seems essential to future studies. Received 31 March 1998, accepted 5 Jan. 1999.

Previously published evidence for elevated
nest predation rates at forest edges in the
northeastern U.S. is not consistent. For ex-
ample, in Maine, predation rates were higher
for artificial nests placed in shrubs at edges
than in forest interiors, but the distance to
edge had no effect on predation of ground
nests (Rudnicky and Hunter 1993). Also, nei-
ther the edge: area ratio of forest patches nor
the distance from edge affected artificial
ground nest predation rates (Small and Hunter
1988). Predation rates of artificial nests were
higher in extensive industrial forests than in
fragments, but within fragments, shrub nests
near edges were depredated at a higher rate
than those farther from edges. Furthermore,
the predation rate in clearcuts was lower than
that in forest fragments or plantations, and
within plantations, predation rates increased
with increasing distance from the edge (Van-
der Haegen and DeGraaf 1996). In Pennsyl-
vania, Yahner and Scott ( 1988) reported a di-
rect relationship between amount of forest

' USDA F'orest Service, Northeastern Research Sta-
tion, University of Massachusetts, Amherst, MA
01 003 USA.

’ Department of Forestry and Wildlife Management
and Graduate Program in Organismic and Evolution-
ary F3iology, University of Massachusetts, Amherst,
MA 01003-4210.

’ Corresponding author; E-mail:
rdegraaf@forwild.umass.edu

fragmentation caused by clearcutting and pre-
dation rates on artificial nests, yet Yahner and
CO workers (1993) did not find greater preda-
tion rates on such nests despite greater frag-
mentation resulting from additional cleai'cut-
ting on the same study area. In sum, the re-
sults of previous studies in the northeastern
U.S. are inconsistent, perhaps because the
large quail (Coturnix sp.) or chicken {Gallus
sp.) eggs used do not sample the entire pred-
ator community (Haskell 1995). Would the
use of eggs of approximately the same size as
most forest passerines shed light on patterns
of predation on artificial nests in relation to
forest edge?

White-footed mice {Peromyscus leucopus)
were frequently recorded by remotely-trig-
gered cameras at ground and shrub nests con-
taining eggs of Japanese Quail {Coturnix ja-
ponica)-, many of these same nests appeared
to be undisturbed at the end of the exposure
period and thus were not classified as visited
by predators (Danielson et al. 1997). Similar-
ly, Northern Bobwhite {Colinus 9irginianus)
eggs at ai'tificial nests in Minnesota that were
visited (as determined by photographs) by
red-backed voles {Clethrionomys gapperi) and
deer mice {Peromyscus maniculatus) were not
damaged, although those nests were classified
as depredated by Fenske-Crawford and Niemi
(1997). Small mammalian predators are clear-

236

DeGniafet at. • PREDATION ON SMAl.L EGGS IN ARTIFICIAE NESTS

237

ly able to locate artificial nests, but have lim-
ited ability to destroy quail eggs in these nests.
Quail eggs are not representative of the sizes
of eggs of most temperate forest passerines,
especially those of Neotropical migratory spe-
cies (Haskell 1995, DeGraaf and Maier 1996).
If appropriately-sized eggs were not available
to potentially common predators, then results
of previous studies to estimate nest predation
rates for forest songbirds may have been bi-
ased, contributing to the inconsistency of re-
sults in extensive northeastern forests.

Egg size is potentially important in nest
predation studies; even though small rodents
such as mice and eastern chipmunks (Tamias
striatus) may be egg predators (e.g., Maxon
and Oring 1978, Reitsma et al. 1990, respec-
tively), they apparently cannot readily open
and consume the larger eggs of quails and
chickens (Roper 1992, Haskell 1995; but see
Craig 1998). Roper (1992) showed that pred-
ators did not respond to quail eggs as they did
to native birds’ eggs in Panama because most
mammalian nest predators were too small to
eat quail eggs. Such eggs, however, are vir-
tually the only ones that have been used in
artificial nest predation studies (Major and
Kendal 1996).

These facts led us to conduct a laboratory
experiment of mouse predation on large (C.
japonica) and small (Zebra Finch, Taeniopy-
gia guttata) eggs (DeGraaf and Maier 1996).
Mouse predation on small eggs was immedi-
ate but did not occur on the large eggs. Sim-
ilar laboratory trials (Maier and DeGraaf, un-
publ. data) indicated that white-footed mice,
including juveniles, could open House Spar-
row (Passer domesticus) eggs; we conducted
a field study to evaluate egg predation in ar-
tificial nests containing such eggs.

We attempted to assess the effects of nest
location (edge vs interior), placement (shioib
vs ground), and the relative abundance of
small mammals on the predation of small
eggs. We hypothesized that small mammals
were equally abundant at edges and in forest
interiors (Heske 1995), that no edge-related
differences in nest predation would be found
for either ground or shrub nests (Major and
Kendall 1996), and that predation would be
greater on ground nests than on shmb nests
because small mammals such as mice and
chipmunks spend the majority of their forag-

ing time on the ground (Madison 1977, Elliot
1978, Graves et al. 1988).

METHODS

We placed artilicial nests near (5-15 ni) and far
( 100-120 m) from stand edges in 40 mature stands in
an extensive managed mixed-wood forest in central
Massachusetts during June to 15 July 1997. All stands
were at least 80 years old and of the red oak ( Quercus
n/^ra)-white pine (Finns strobus)-re.d maple (Acer
ritbrum) forest-cover type (Eyre 1980); edges were
formed by small (2-4 ha) clearcuts 1-6 years old. We
placed two ground nests and two shrub nests in each
stand, one of each type near and far from the edge and
at least 100 m from each other (Fig. 1). Nests (160
total) were wicker baskets 10 cm in diameter and 6
cm deep, weathered for 3 weeks before use, and con-
tained one fresh House Sparrow egg. To minimize hu-
man scent at nests, we wore rubber boots and clean
cotton gloves during nest placement (Whelan et al.
1994). Ground nests were set into the surface litter;
shrub nests were wired 1-1.5 m above the ground in
crotches or forks of branches of shrubs or small sap-
lings. All nests were checked after 12 days, approxi-
mately the mean incubation time for small forest pas-
serines. Eggs found out of the nest, destroyed in the
nest, or missing were classified as predations.

We analyzed the nest predation data as paired-sam-
ple nest types within stands (Zar 1996:163) and per-
formed statistical tests using SYSTAT 7.0 for Win-
dows. Sign tests were used to detect differences in the
number of nest predations among edge and interior
nests on the ground and in shrubs (Zar 1996:536).

We assessed the relative abundance of small mam-
mal species at edge and interior sites using 3-day re-
moval trapping (Miller and Getz 1977) at each site
immediately after the nest predation experiments. Two
circular trapping arrays (20 traps/20 m diameter array)
were set in each of the 40 stands, one midway between
edge nests and one between interior nests (Fig. 1 ). Dif-
ferent types of small mammal traps are more efficient
for trapping certain species under varying conditions,
e.g., weather (Williams and Braun 1983, Bury and
Corn 1987, Mengak and Guynn 1987); we used four
types of traps in an attempt to more completely sample
the small mammal community (Pelikan et al. 1977).
Small Victor snap traps with expanded pedals. Muse-
um Special snap traps with expanded pedals, large
(approx. 8 X 8 X 24 cm) Sherman traps, and modified
large Sherman live traps with circular glass windows
(5.5 cm diameter) in the rear door were used at each
array in equal numbers. All traps were baited with a
mixture of peanut butter, oatmeal, bacon, and black
sunflower seed and were checked daily. All small
mammals collected were deposited in the Vertebrate
Museum of the University of Massachusetts, Amherst,
Massachusetts. We followed the guidelines for the cap-
ture and handling of mammals approved by the Amer-
ican Society of Mammalogists (American Society of
Mammalogists 1998).

238

THE WILSON BULLETIN • Vol. Ill, No. 2. June 1999

Clearcut g

(2 - 4 ha) m

Edge ground nest

a c

a c
b d

a c

b d
a c
b d

10 m

Edge shrub nest

Forest

100- 120 m
\

Interior shrub nest

a c

bd bd

- bd

b d b d

a c

20 m dia

Interior ground nest

FIG. 1. Placement of small mammal trapping arrays and artificial nests in central Massachusetts, June-August.
1997. Four types of traps include: (a) small Victor snap, (b) Museum Special, (c) large Sherman, (d) large
Sherman with glass window. (Figure not to scale.)

Small mammal capture counts were compared by
species between edge and interior using the Wilcoxon
paired-sample test for species with sufficient n (Zar
1996:167). Logistic regression (PROC LOGISTIC,
SAS 1989) was used to assess the relationship between
small mammal counts and nest predation; we used a
model with nest placement (ground or shrub), location
(edge or interior), and small mammal abundance ef-
fects; the first two were treated as categorical variables
and the third as a continuous variable. The tests for
whether a coefficient is zero were carried out using Z
= (estimated coefficient/standard error) with the P-val-
ue obtained using the standard normal distribution
(Hosmerand Lemeshow 1989:17).

RESULTS

We did not detect any significant differenc-
es between the number of nest predations at
the forest edge and those in the forest interior
for either ground nests (Sign test: ties = 28,
4 “-b”, 8 critical value = 2, P > 0.05)
or shrub nests (Sign test: ties = 24, 10 “+”,
6 “ — critical value = 3, P > 0.05). How-
ever, the number of nest predations on ground
nests at both the edge (Sign test: ties = 15,
21 + 4 critical value = 7, P < 0.001)

and interior (Sign test: ties = 15, 25 “+”, 0
critical value = 1, P < 0.001) were sig-
nificantly higher than those on shrub nests.

Twelve species of small mammals were de-
tected; six species represented 99% of cap-
tures at both forest edge and interior. The dis-
tributions of the 6 most commonly detected
small mammal species did not differ signifi-
cantly (Wilcoxon paired-sample tests: P >
0.05) between stand edges and interiors (Table
1). White-footed mice were detected more
than all other species combined in both stand
edges and inteiiors and were the only small
mammal species detected in all 40 stands. Lo-
gistic regression analyses confirmed nest
placement (ground, shrub) effects but showed
no effect of small mammal abundance or lo-
cation (edge, interior) on nest predation rate
(Table 2).

DISCUSSION

Because this is the first study that we know
of to systematically evaluate artificial nest
predation in relation to forest edge using small
eggs, comparison with other studies where
larger eggs were used is difficult. In a recent
review of studies in both agricultural and set-
tled landscapes in North America and Europe,
Major and Kendal (1996) showed that egg
predation (on large eggs) was higher near the

DeGrcuifei al. • PREDATION ON SMALL EGGS IN ARTILICIAL NESTS

239

TABLE 1. Numbers of small mammals captured near edges and interiors of 40 stands in extensive forest
in central Massachusetts, July and August 1997. Wilcoxon paired-sample results for most commonly detected
species; N = stands species detected in, n = differences (N minus ties), T = smallest sum of ranks, Tuo-iij,,,, =
critical value (Zar 1996: table B.12),

Species^

Edge

Inlerior

Total

White-footed mouse (PeroiHyscus teiicopus)

251

235

486

262.5

182

>0.05

Red-backed vole (Clethriononiys gapperi)

137

90.5

>0.05

Northern short-tailed shrew (Blarina brevicaiida)

126

202.0

1 16

>0.05

Masked shrew (Sorex cinereus)

44.0

>0.05

Eastern chipmunk (Tamias striatiis)

28.5

>0.05

Smoky shrew (Sorex fiimeus)

13.5

>0.05

^ Species delected in ^3 siands: woodland jumping mouse {Napcteozcipus insifinis), flying squirrel {Glaucontys sp.). long-tailed weasel {Musiela frenata),
red squirrel {Tamiasciurus huJsonicus), pine vole {Microtus pineforum), meadow vole {Microius pennsylvcmicus).

forest edge in three studies, higher away from
the edge in one study, and equal in seven stud-
ies, Predation of artificial nests containing
Northern Bobwhite eggs in Wisconsin pine
barrens savannah patches was coirelated with
proximity to the edge (Niesmuth and Boyce
1997), Predation on artificial nests containing
small chicken, Japanese Quail, and plasticine
eggs in Alberta was highest in larger woodlots
and showed no edge effect (Hannon and Cot-
terill 1998). Two additional studies (which
used Japanese Quail eggs) in the northeastern
U.S. did not detect any difference in predation
rates between edges and interiors of exten-
sive-managed forests (Table 3; Vander Haegan
and DeGraaf 1996) or suburban/agricultural/
forest landscapes (Danielson et al. 1997).
Along with our current results, these vaiiable
findings suggest either that the “edge” effect
as related to egg predation (Andren and An-
gelstam 1988) is not a widespread phenome-

TABLE 2. Parameter estimates and statistics from
logistic regression analysis of placement (ground/
shrub), location (edge/interior), and small mammal
abundance in relation to predation of small eggs in
artificial nests in central Massachusetts, 1997. The tests
for whether a coefficient is zero were carried out using
Z = (est. coefficient/SE) with the P-value obtained us-
ing the standard normal distribution (equivalent to C
= Zr with the P-value based on X' distribution with I
df).

Variable

Estimated

coefficient

Standard

error

INTERCEPT

-1.7320

0.5080

0.0007

LOCATION

0.0002

0.3805

0.9996

W-F MICE

-0.0005

0.0602

0.9940

PLACEMENT

2.4655

0.3937

0.0001

non or that not all forest edges are the same;
i.e., forest-clearcut, forest-agriculture, and for-
est-suburb edges differ in the predators pre-
sent (Danielson et al. 1997).

Equally variable are the results of nest
placement studies (i.e., ground vs shrub/ele-
vated nests). Major and Kendal (1996) re-
ported higher predation at elevated nests in six
studies, higher predation at ground nests in
four studies, and equal predation rates in three
studies. Ground nests containing Japanese
Quail and plasticine eggs had increased pre-
dation along farm edge and interior in Sas-
katchewan, but there were no detectable dif-
ferences in predation rate between ground and
shrub nests at logged edge, logged interior, or
contiguous forest (Bayne and Hobson 1997).
Although two studies in the noitheastem U.S.
did not detect any difference in predation rates
between ground and shrub nests (Vander Hae-
gan and DeGraaf 1996, Danielson et al. 1997),
we found a strong placement effect (high pre-
dation on ground nests) using small eggs.

Where edge or nest placement effects oc-
cuiTed, generalist predators commonly were
presumed to depredate specific nest types dis-
proportionately. The variability in results
among studies may reflect differences in nest
predator guilds or the abundance of particular
species in study areas (e.g.. Pieman 1988). At-
tempts to identify individual egg predators in-
clude characterizations of predation remains
of real eggs (Gottfried and Thompson 1978,
but see Marini and Melo 1998), impressions
in plasticine (Bayne et al. 1997), and clay
eggs (Donovan et al. 1997), hair catchers
(Baker 1980), and remotely triggered cameras
(DeGraaf 1995). Nevertheless, egg predation

240

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 3. Predation rates (%) on eggs in artificial nests exposed for 12—14 days at forest edge and interior
(>50 m) in the northeastern U.S.

Nest

placement

Egg

size

Edge

Interior

Study

location

Reference

Rate

Ground

Sparrow

0.80

0.90

>0.05

Massachusetts

This study

Quail

0.45

0.41

>0.05

Massachusetts

DeGraaf, unpubl. data“

0.20

0.25

>0.05

Maine

Vander Haegan and DeGraaf 1996*’

0.29

0.21

>0.05

Maine

Rudnicky and Hunter 1993

Shrub

Sparrow

0.38

0.28

>0.05

Massachusetts

This study

Quail

0.60

0.51

>0.05

Massachusetts

DeGraaf, unpubl. data“

0.16

0.14

>0.05

Maine

Vander Haegan and DeGraaf 1996'’

0.55

0.29

0.015

Maine

Rudnicky and Hunter 1993

^ Recalculated from data used by Danielson et al. 1997.

’’ From Table 1 ; edge = 5 m and interior = 200 m from edge.

studies almost never include surveys of the
predator community in the study area or index
predator abundance (Yahner 1996). We in-
dexed small mammal abundance; the overall
abundance and distribution of small mammals
at forest edge and interior were similar, as
were the distributions of depredated ground
and shrub nests, but the abundance of small
mammals was not related to nest predation for
either nest type or location in one season. Ei-
ther the small mammals that we detected were
not major nest predators, or they did not vary
sufficiently in abundance in a homogeneous
landscape in one season to show a relationship
with nest predation. A relationship between
nest predation and small mammal abundance
may be detectable only over time; small mam-
mals vary greatly from year to year with food
abundance (Elkinton et al. 1996). Long-term
studies are needed to determine if this is the
case.

Small-mouthed nest predators such as Per-
omyscus were abundant in our study area ( 10—
40/ha; Elkinton et al. 1996) compared to larg-
er generalist predators such as fishers (Martes
pennanti; 21/100 km- in central Massachu-
setts; York 1996) that have been shown to
depredate artificial nests in northern New
England (DeGraaf 1995). If small eggs that
are susceptible to depredation by all potential
predators are used in artificial nests, then
ubiquitous, abundant predators (e.g., small
mammals) may swamp the effect of larger
generalist predators, even if the latter are more
abundant along forest edges (apparently not
the case in the northeastern U.S.). Moreover,
our data suggest that ground nests may be par-

ticularly vulnerable to predators such as mice
and chipmunks (Haskell 1995, Bayne et al.
1997), which spend more time foraging on the
ground than in shrubs or trees (Madison 1977,
Elliot 1978, Graves et al. 1988). Hence, the
hypothesis that egg predation rates are elevat-
ed at forest edges may, in large part, be an
artifact of egg size. Virtually all studies to date
have used quail eggs (see Paton 1994, Major
and Kendal 1996; but see George 1987) which
apparently cannot be opened by the most
abundant small-mouthed predators in temper-
ate forests.

Do natural nests containing small eggs
show edge related predation in the extensively
forested northeastern U.S.? In a 2-yeai' study
of ground nesting Ovenbird (Seiiirus aurocap-
illus) reproductive success in New Hampshire
(King et al. 1996), nests, tenitories, and ter-
ritorial males were equally distributed in edge
(0-200 m) and forest interior (201-400 m);
nest survival was higher in the forest interior
in year 1 , but not in year 2. The proportion of
pairs fledging at least 1 young, fledgling
weight, and fledgling wing chord did not dif-
fer between edge and interior over the course
of the study.

In extensive mixed- wood forests in New
England, edge related differences in artificial
nest predation rates have not been consistently
demonstrated. In our study predation rates
were substantially higher on artificial ground
nests that contained small eggs than those in
studies that used quail eggs (Table 3). All po-
tential predators can open small eggs, and
their use should result in higher predation
rates because small-mouthed predators are

DeGnulfci al. • PREDATION ON SMALL EGGS IN ARTILICIAL NESTS

241

more abundant than large nest predators. Pre-
dation rates of artificial nests often have been
assumed to track those of natural nests (Major
and Kendal 1996), but they may not unless
egg sizes closely approximate those of the
species of concern. For example, nest survival
of natural nests was lower than that of exper-
imental nests containing quail eggs in Panama
because of the abundance of small-mouthed
nest predators (Roper 1992). Predation rates
in quail egg experiments (e.g., Loiselle and
Hoppes 1983; Martin 1987, 1988) may be
useful to compaie local habitats, but may be
inappropriate for estimating natural predation
rates or for comparing ai’eas inhabited by dif-
ferent predators (Roper 1992). Nest predation
is a dominant factor in avian reproductive suc-
cess (Ricklefs 1969, Martin 1988); results of
experiments that exclude major sources of
mortality (i.e., small-mouthed predators) may
not be representative (Roper 1992).

Only if appropriate egg sizes are used can
predation rates in relation to habitat edge or
placement be generalized or approximated for
natural nests. Even then, effects such as nest
defense and appearance (Martin 1987) are dif-
ficult to address. Our data suggest that egg
predation rates may be strongly related to egg
size, other factors being equal, because the
most abundant predators can only open small
eggs.

ACKNOWLEDGMENTS

We thank M. Stoddard for assistance with field
work, J. Buonaccorsi for help with statistical analyses,
R. Askins, D. Haskell, D. King, R Sievert, and R. Yah-
ner for their critical reviews, and M. A. Sheremeta for
typing the manuscript.

LITERATURE CITED

American Society of Mammalogists. 1998. Guide-
lines for the capture, handling, and care of mam-
mals as approved by the American Society of
Mammalogists. J. Mamm. 79:1404-1409.
Andrew, H. and P. Angel.stam. 1988. Elevated pre-
dation rates as an edge effect in habitat islands:
experimental evidence. Ecology 69:.544-547.
Baker, B. W. 1980. Hair-catchers aid in identifying
mammalian predators of ground-nesting birds.
Wildl. Soc. Bull. 8:257-259.

Bayne, E. M. and K. A. Hob.son. 1997. Comparing
the effects of landscape fragmentation by forestry
and agriculture on predation of artificial nests.
Conserv. Biol. 11:1418—1429.

Bayne, E. M., K. A. Hobson, and P. Eargey. 1997.

Predation on artificial nests in relation to forest
type: contrasting the use of quail and plasticine
eggs. Ecography 20:233-239.

Bury, R. B. and P. S. Corn. 1987. Evaluation of pitfall
trapping in northwestern forests: trap arrays with
drift fences. J. Wildl. Manage. 51:112-119.

Craig, D. P. 1998. Chipmunks use leverage to eat
oversized eggs: support for the use of quail eggs
in artificial nest studies. Auk 115:486-489.

Danielson, W., R. M. DeGraae, and T. K. Fuller.
1997. Rural and suburban forest edges: effect on
egg predators and nest predation rates. Landscape
Urban Plann. 38:25-36.

DeGraae, R. M. 1995. Nest predation rates in man-
aged and reserved extensive northern hardwood
forests. For. Ecol. Manage. 79:227-234.

DeGraae, R. M. and T. J. Maier. 1996. Effect of egg
size on predation by white-footed mice. Wilson
Bull. 108:535-539.

Donovan, T. M., P. W. Jones, E. M. Annand, and F.
R. Thompson, III. 1997. Variation in local-scale
edge effects: mechanisms and landscape context.
Ecology 78:2064-2075.

Elkinton, j. S., W. M. Healy, J. P. Buonaccorsi, G.
H. Boettner, a. M. Hazzard, H. R. Smith, and
A. M. Liebhold. 1996. Interactions among gypsy
moths, white-footed mice, and acorns. Ecology
77:2332-2342.

Elliot, L. 1978. Social behavior and foraging ecology
of the eastern chipmunk (Tamis striatus) in the
Adirondack Mountains. Smithson. Contrib. Zool.
265:1-107.

Eyre, F. H. 1980. Forest cover types of the United
States and Canada. Society of American Foresters,
Washington, D.C.

Fenske-Crawford, T. J. and G. J. Niemi. 1997. Pre-
dation of artificial ground nests at two types of
edges in a forest-dominated landscape. Condor 99:
14-24.

George, T. L. 1987. Greater land bird densities on is-
land vs mainland: relation to nest predation level.
Ecology 68:1393-1400.

Gon'ERiED, B. M. AND C. F. Thompson. 1978. Exper-
imental analysis of nest predation in an old-field
habitat. Auk 95:304-312.

Graves, S., J. Maldonado, and J. O. Wolfe. 1988.
Use of ground and arboreal microhabitats by Per-
otnvscus leiicopiis and Peromyscits inaniculaliis.
Can. J. Zool. 66:277-288.

Hannon, S. J. and S. E. Cotthrill. 1998. Nest pre-
dation in aspen woodlots in an agricultural area in
Alberta: the enemy from within. Auk 1 15:16-25.

Haskei.l, D. G. 1995. Forest fragmentation and nest
predation: are experiments with Japanese Quail
eggs misleading? Auk 1 12:767-770.

Hhskk, E. j. 1995. Mammals in abundances on forest-
farm edges versus forest interiors in southern Il-
linois: is there an edge effect ? J. Mamm. 76:562-
568.

Hosmer, D. W., Jr. and S. Lemeshow. 1989. Applied

242

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

logistic regression. John Wiley and Sons, New
York.

King, D. I., C. R. Griffin, and R. M. DeGraaf. 1996.
Effects of clearcutting on habitat use and repro-
ductive success of the Ovenbird in forested land-
scapes. Conserv. Biol. 10:1380-1386.

Loiselle, B. a. and W. a. Hoppes. 1983. Nest pre-
dation in insular and mainland lowland rainforest
in Panama. Condor 85:93-95.

Madison, D. M. 1977. Movements and habitat use
among interacting Peromyscus leiicopiis as re-
vealed by radio telemetry. Can. Lield-Nat. 91:
273-281.

Major, R. E. and C. E. Kendal. 1996. The contri-
bution of artificial nest experiments to understand-
ing avian reproductive success: a review of meth-
ods and conclusions. Ibis 138:298-307.

Marini, M. A. and C. Melo. 1998. Predators of quail
eggs, and the evidence of the remains: implica-
tions for nest predation studies. Condor 100:395-
399.

Martin, T. E. 1987. Artificial nest experiments: effects
of nest appearance and types of predators. Condor
89:925-928.

Martin, T. E. 1988. Habitat and area effects on forest
bird assemblages: is nest predation an influence?
Ecology 69:74-84.

Maxon, S. J. and L. W. Oring. 1978. Mice as a source
of egg loss among ground-nesting birds. Auk 95:
582-584.

Mengak, M. T. and D. C. Guynn, Jr. 1987. Pitfalls
and snap traps for sampling small mammals and
herpetofauna. Am. Midi. Nat. 118:284—288.

Miller, D. H. and L. L. Getz. 1977. Factors influ-
encing local distribution and species diversity of
forest small mammals in New England. Can. J.
Zool. 55:806-814.

Niesmuth, N. D. and M. S. Boyce. 1997. Edge-related
nest losses in Wisconsin pine barrens. J. Wildl.
Manage. 61:1234-1239.

Paton, P. W. C. 1994. The effect of edge on avian
nesting success: how strong is the evidence? Con-
serv. Biol. 8: 1 7-26.

Pelikan, j., j. Zejda, and V. Holisova. 1977. Effi-
ciency of different traps in catching small mam-
mals. Folia Zool. 26:1-13.

PiCMAN, J. 1988. Experimental study of predation on

eggs of ground-nesting birds: effects of habitat
and nest distribution. Condor 90:124—131.

Reitsma, L. R., R. T. Holmes, and T. W. Sherry.
1990. Effects of removal of red squirrels, Tamias-
ciurus hudsonicus, and eastern chipmunk, Tamias
striatiis, on nest predation in a northern hardwood
forest: an artificial nest experiment. Oikos 57:
375-380.

Ricklefs, R. E. 1969. An analysis of nesting mortality
in birds. Smithson. Contrib. Zool. 9:1-48.

Roper, J. J. 1992. Nest predation experiments with
quail eggs: too much to swallow? Oikos 65:528-
530.

Rudnicky, T. C. and M. L. Hunter, Jr. 1993. Avian
nest predation in clearcuts, forests, and edges in a
forest-dominated landscape. J. Wildl. Manage. 57:
358-364.

SAS Institute Inc. 1989. SAS/STAT' user’s guide,
version 6, fourth ed. Cary, North Carolina.

Small, M. E and M. L. Hunter. 1988. Forest frag-
mentation and avian nest predation in forested
landscapes. Oecologia 76:62-64.

Vander Haegan, W. M. and R. M. DeGraaf. 1996.
Predation on artificial nests in forested riparian
buffer strips. J. Wildl. Manage. 60:542-550.

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.

Williams, D. E and S. E. Braun. 1983. Comparison
of pitfall and conventional traps for sampling
small animal populations. J. Wildl. Manage. 47:
841-845.

Yahner, R. H. 1996. Forest fragmentation, artificial
nest studies, and predator abundance. Conserv.
Biol. 10:672-673.

Yahner, R. H., C. G. Mahan, and C. A. DeLong.
1993. Dynamics of depredation on artificial
ground nests in habitat managed for Ruffed
Grouse. Wilson Bull. 105:172-179.

Yahner, R. H. and D. P. Scott. 1988. Effects of forest
fragmentation on depredation of artificial nests. J.
Wildl. Manage. 52:158-161.

York, E. C. 1996. Fisher population dynamics in
north-central Massachusetts. M.Sc. thesis, Univ.
of Massachusetts, Amherst.

Zar, j. H. 1996. Biostatistical Analysis, third ed. Pren-
tice-Hall, Englewood Cliffs, New Jersey.

Wilson Bull., 111(2), 1999, pp. 243-250

BIRD USE OF BURNED AND UNBURNED CONIFEROUS FORESTS

DURING WINTER

KAREN J. KREISEL' 2^ AND STEVEN J. STEIN'

ABSTRACT. — Cavity-nesting bird species have been shown to be associated with early post-fire habitat during
the breeding season but little study has been done of birds in the non-breeding season. We compared bird
composition and foraging behavior during the winter in burned and unburned forests. We conducted point counts
during four consecutive winters immediately following a stand replacement fire. Burned and unburned forests
had similar numbers of bird species, yet species composition was distinctly different. Trunk and branch foraging
species were 2.5 times more abundant in burned forest than in unburned forest. Within burned forests, trunk
and branch foraging species significantly decreased from the first winter post-fire to the fourth winter post-fire.
We conducted foraging observations of four woodpecker species within burned forests only. Woodpeckers used
western larch (Lari.x occidentalis), ponderosa pine {Pinus ponderosa) and Douglas-fir (Pseudotsuga menziesii)
snags that were greater than 23 cm in diameter. Stand replacement fires may play an important role in maintaining
populations of trunk and branch foraging species in mixed coniferous forests in northeastern Washington. Re-
ceived 24 August 1998, accepted 28 Dec. 1998.

Wildfire plays a major role in determining
landscape patterns by creating large mosaics
of burned habitat intermixed with unbumed
habitat. Prior to fire suppression policies in the
early 1900s, wildfires were more frequent and
widespread in the western United States
(Agee 1994, Hejl 1994). Mixed-conifer forests
of the Pacific Northwest have a history of
stand replacement fires, in which most trees
are killed, occurring every 140-340 years
(Agee 1994). Stand replacement fires create a
unique habitat of large patches of standing
dead trees that host great numbers of bark and
wood-boring beetles (Fumiss 1965, Amman
and Ryan 1991) which serve as food for birds
inhabiting recently burned forests (Spring
1965, Wickman 1965). This food resource is
thought to decrease dramatically 2-5 years
post-fire (Koplin 1972).

Cavity-nesting bird species are associated
with early post-fire forests (1-9 years post-
fire, Hutto 1995). Many researchers have
compared bird abundance in burned and un-
bumed forest during the breeding season and
found increased numbers of cavity-nesting
bird species in early post-fire forests (Bock
and Lynch 1970, Raphael and White 1984,
Raphael et al. 1987, Hutto 1995, Caton 1996,
Hitchcox 1996). Some species, including the

' Dept, of Biology, Eastern Washington Univ., Che-
ney, WA 99004.

- Present address: 2802 W. Depot Springs Rd., Span-
gle, WA 99031; E-mail: KKREISEL@hotmail.com
’ Corresponding author.

Black-backed Woodpecker {Picoides arcti-
cus). Three-toed Woodpecker (Picoides tri-
dactylus), and Mountain Bluebird (Sialia mex-
icana) have been shown to be more common
in the first few years after a fire than later
(Hutto 1995). How long this high abundance
of cavity-nesting bird species persists in post-
fire habitat is unclear.

There are few data on bird species com-
position of burned forests during the non-
breeding season. In Montana, Blackford
(1955) anecdotally noted numerous wood-
pecker species in burned forests during the
winter. Blake (1982) compared stand replace-
ment burned and unbumed ponderosa pine
(Pinus ponderosa) forests during the non-
breeding season and found bark insectivores
(including woodpeckers) to be more abundant
in burned forests. Non-breeding season habitat
may play a significant role in determining
overall survival and numbers of individuals
that breed (Conner 1979, Graber and Graber
1983, Klein 1988) and is probably as impor-
tant as breeding season habitat for the persis-
tence of avian populations. Many resident spe-
cies present in burned forest during the non-
breeding season may not remain in burned
forest year-round but seasonally migrate short
distances to different habitats, thereby maxi-
mizing their use of available resources.

Several studies have examined the chai'ac-
teristics of trees used for nesting by cavity-
nesting birds in burned forests. In general,
cavity-nesters seem to prefer broken top snags

243

244

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

and snags greater than 23 cm in diameter for
nesting (Raphael and White 1984, Hutto 1995,
Caton 1996, Hitchcox 1996). Fewer research-
ers have examined the snag characteristics
used for foraging by cavity-nesting birds in
burned forests (Hutto 1995, Caton 1996), even
though food availability may be as important
to excavating cavity-nesting species as the
availability of nest sites, especially in burned
forests where nest snags may be abundant
(Caton 1996). Because snags used for nesting
and foraging by cavity-nesting birds may dif-
fer, information on the characteristics of snags
used for nesting would be of particular interest
to managers to maintain fire-dependent bird
populations.

To investigate the use of burned forests dur-
ing the winter, we examined bird community
composition and foraging behaviors in burned
and unburned forests during four consecutive
winters in northeastern Washington.

STUDY AREA AND METHODS

Study urea. — The study sites were located in the
Kettle River Range on the Colville National Lorest in
the northeastern corner of Washington, 26 km north of
Republic (48° 65' N, 118° 73' W). This area is pre-
dominantly a mixed-conifer forest consisting of sub-
alpine fir (Abies lasiocarpa), Douglas-fir (Pseudotsuga
menziesii), Engelmann spruce (Picea engelmannii),
ponderosa pine, western larch (Lari.x occidentcdis), and
lodgepole pine (Pinus contorta), with a minor under-
story component of ninebark (Physocctrpus nuilva-
ceus). Both the burned and unburned study sites had
similar tree species composition and were located in
unlogged areas.

In August 1994 the Copper Butte fire was ignited
by lightning and burned 4000 ha, resulting in large
mosaics of high intensity burned areas mixed with low
intensity and unburned areas (U.S. Dept, of Agric.
1995). We chose two stand replacement burned sites
of greater than 80 ha within the Copper Butte fire area,
with elevations ranging from 1320-1650 m. The un-
burned site was located 20 km north of the burned
sites, 0.6 km from the burn boundary. The 80 ha un-
burned site had similar slope and aspect to the burned
sites but was only 1030-1320 m in elevation. Average
yearly rainfall for this area is 384 cm and average year-
ly temperature is 6.1°C. During November-Lebruary,
when all of the data were collected, the average month-
ly precipitation was 42 cm (predominantly snow) and
the average temperature was — 2.1°C.

Avian sampling. — We used the point count method
to quantify birds in burned and unburned forests
(Blondel ct al. 1981, Hutto et al. 1986). Point ctxmt
stations were systematically laid out at least 200 m
apart, 100 m from a change in habitat, and 50 m from

roads and creeks. We conducted ten-minute counts, re-
cording all birds seen and heard within 100 m (Hutto
1995). Unfortunately, we could not always distinguish
between Hairy (P. villosus). Three-toed (P. tridacty-
lus), or Black-backed (P. arcticus) woodpeckers so we
created a category of “unknown woodpecker” to ac-
count for these detections. Counts were conducted on
fair weather days when winds were less than 25 kph,
with little or no precipitation, and temperatures greater
than -9° C. Winter counts were conducted between
08:00 PST (I h after sunrise) and 16:00 from 8 No-
vember to 4 Lebruary 1994-1997.

Over the four years of the study, the number of point
count stations surveyed was as follows: first winter to
fourth winter post-fire, burned (8, 9, 13, 13) and un-
burned (0, 9, 9, 9). The number of stations on the
burned site varied over the four years because of log-
ging that occurred in 1996 at four of the stations.
These four stations were no longer used, and eight new
stations were added in a nearby similarly burned area.
The difference in the number of count stations in
burned and unburned forest affected bird diversity
very little. The four point count stations in addition to
nine on the burned study site added only one bird spe-
cies that was not detected in the first nine stations. On
both study sites there were two visits to each point
count station during a winter and these were averaged.
We used the mean number of birds per point in com-
paring burned and unburned forests and number of
points as sample size.

Bird species were assigned to three foraging guilds
based on Ehrlich and coworkers (1988) and Hutto
(1995): (1) trunk and branch foragers (timber drillers
and timber gleaners), (2) foliage foragers (including
aerial foragers), and (3) ground foragers.

Foraging sampling. — We recorded foraging obser-
vations of four woodpecker species in the burned for-
est only, during and after point count surveys. A bird
was considered foraging when it appeared to be ac-
tively searching for and/or obtaining food (i.e., digging
or pecking). Only the first foraging observation was
recorded for each individual (Hejl et al. 1990). The
following foraging data were recorded: bird species,
maneuver (pecking = tapping on the surface, flaking
= removing bark, drilling = excavating into wood),
zone of foraging (lower trunk, middle trunk, upper
trunk, branches), tree species, tree dbh class (8-22, 23-
37, 38-53, >53 cm), burn severity of trees (alive and
green, possibly alive with some green present, dead
with brown needles present, dead and severely
burned), and top condition (broken, crooked, double,
intact). We combined the foraging data for each spe-
cies over four winters iti order to increase sample size
(Morrison 1984). Consequently, we do not have infor-
mation on between year differences in foraging.

Tree sampling. — Several tree characteristics within
the burned forest were recorded at count stations and
between stations in ().()4-ha circular samples for a total
of 24 samples (Martin 1994): tree species, tree dbh,
burn severity, bark cover, branch condition, and top
condition.

Kreise! and Stein • BURNED FOREST BIRDS IN WINTER

245

TABLE 1. Species detecled in burned and imbiirned ear
ington during the winters, 1994-1997.

ly post-lire mixed-conifer

forests

in northeast Wash-

Species

Forage

guilcU

Nest

guild^

Burned^

Unburncd^

/xl

Year

Mean/pl

Year

Mcan/pt

Blue Grouse, Dendragapiis ohscnnts

0.083

—

>0.05

Downy Woodpecker, Picoides puhescens

1,2,4

0.175

—

>0.05

Hairy Woodpecker, Picoides viUosns

1, 2,3,4

0.410

2,4

0.056

0.035

Three-toed Woodpecker, Picoides tridactyins

1,2

0.042

—

0.047

Black-backed Woodpecker, Picoides arcticiis

1, 2,3,4

0.406

0.019

0.001

Pileated Woodpecker, Drycocopiis pi lea tits

—

0.019

>0.05

Unknown Woodpecker

1,2

0.144

0.019

0.05

Gray Jay, Perisorius canadensis

1,2

0.075

2,3

0.093

>0.05

Steller’s Jay, Cyanocitta stelleri

—

0.019

>0.05

Clark’s Nutcracker, Nucifraga coiumbiana

1,2

0.016

—

>0.05

Black-billed Magpie, Pica pica

0.014

—

>0.05

Common Raven, Coreas corax

0.016

0.037

>0.05

Black-capped Chickadee, Pants atricapiUiis

1,2

0.031

0.074

>0.05

Mountain Chickadee, Pants gambeli

1, 2,3,4

0.179

2,3,4

0.352

>0.05

White-breasted Nuthatch, Sitta carolinensis

0.048

—

>0.05

Red-breasted Nuthatch, Sitta canadensis

—

2,3,4

0.389

0.001

Brown Creeper, Certhis aniericana

0.019

—

>0.05

Winter Wren, Troglodytes troglodytes

—

0.019

>0.05

Golden-crowned Kinglet, Regiilits satrapa

—

2,3,4

0.260

0.001

Varied Thrush, Ixoreus naeviiis

—

0.056

>0.05

Red Crossbill, Loxia curvirostra

2,4

0.067

0.148

>0.05

^Forage guild: T = trunk and branch. F = foliage. G = ground.

^ Nest guild: O = open. C = cavity.

^ Burned and Unburned: year: year post-fire that bird species were present during the winter, mean/pl: mean number of birds per point averaged over 4
years.

^ /^-value: 2-way ANOVA, treatment effect (burned-unburned).

Statistical analyses. — The mean numbers of birds
per point in burned and unburned forests were com-
pared with a 2-way Analysis of Variance (ANOVA).
Bird abundance between years within a treatment was
compared using a Kruskal-Wallis nonparametric 1-way
ANOVA and pairwise comparisons were made using
Tukey’s BSD procedure. Bird abundances, grouped by
foraging guilds, were compared between burned and
unburned forests using a Mann-Whitney (/-test. To de-
termine trends in abundance over four years, a Spear-
man rank order correlation was performed (Zar 1996).
Comparisons of foraging observations and available
vegetation were made using the Goodness of Fit test
for categorical data. For example, if the distribution of
tree species used by Downy Woodpeckers was signif-
icantly different than the proportion of available tree
species the Goodness of Fit test would have a. P <
0.05. P values of less than 0.05 were considered sig-
nificant for all tests. Statistical analyses were per-
formed using SPSS (SPSS 1993).

RESULTS

Winter bird assemblage in burned and im-
biirned forests. — During the winter, 20 bird
species were detected in burned and unburned
forests combined (Table 1). Bird species com-
position differed between the two forests. In

burned forests 7 of the 14 species detected
wei'e restricted to burned forests and in un-
bumed forests, 6 of the 14 species detected
were restricted to unburned forests. Averaged
over four years tmnk and branch foraging spe-
cies were 2.5 times more abundant in burned
forest {U = 2.0, df = 1, P > 0.05), and foliage
foraging species were 3 times more abundant
in unburned forest ((/ = 2.0, df = 1, P > 0.05;
Fig. 1). Woodpecker species combined were
10 times more abundant in burned forests than
in unburned forest. The four most abundant
species detected in burned forest in descend-
ing order were the Hairy Woodpecker, Black-
backed Woodpecker, Mountain Chickadee
(Pants gambeli), and Downy Woodpecker
(Picoides piibescens), and in unburned forest
the Red-breasted Nuthatch (Sitta canadensis).
Mountain Chickadee, Golden-crowned King-
let (Regidus satrapa), and Red Crossbill (Lox-
ia curvirostra). Black-backed Woodpeckers
(2-way ANOVA: F = 11.26, df = 1, P =
0.001) and Hairy Woodpeckers (P = 4.62, df
= 1, P = 0.035) were significantly more abun-

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

246
2.0 -

1.8 -

1.6 -

1.4 -

1.2

1.0 -

-I--

0.6 ^

0.2 1
0.0 f

■ Burned
H Unburned

T T

y/A/j

TRUNK & FOLIAGE GROUND

BRANCH

LIG. 1. Mean (±SD) number of birds per point
over four years, by foraging guild in burned and un-
burned forests during winter (Mann-Whitney t/-test:
for each guild all P > 0.05).

LIG. 2. Trunk and branch foraging species abun-
dance over four years, in burned and unburned forest
during winter (Spearman rank: Burned s^ = -1.0, df
= 3, P = 0.001; Unburned s, = 0, df = 2, P > 0.05).

dant in burned forest regardless of year (Table
1). The Three-toed Woodpecker was signifi-
cantly more abundant (F = 4. 12, df = I, P =
0.047) in burned forest only during the second
winter post-fire (Kruskal- Wallis: = 11-9, df
= 3, P = 0.008; Tukey P = 0.042). The Red-
breasted Nuthatch (F = 13.10, df = 1, F =
0.001) and Golden-crowned Kinglet (F =
11.80, df = 1, F = 0.001) were significantly
more abundant in unbumed forest.

Change in bird abundance over four
years. — Bird species composition and abun-
dance in burned forests during the winter
changed from 1994-1997. Eleven species
were present the first or second winter post-
fire and were absent by the third winter post-
fire (Table 1). The Brown Creeper {Certhia
americana) was not present until the third
winter post-fire and the White-breasted Nut-
hatch (Sitta canadensis) was not present until
the fourth winter post-fire. The abundance of
trunk and branch foraging species decreased
3.8 times from winter 1994—1997 and showed
a significant negative trend (Spearman rank: r^
= — 1.0, F = 0.001; Fig. 2). Although all four
woodpecker species present in burned forest
during the winter had declined by the fourth
winter post-fire, none of the changes was sig-
nificant (all F > 0.05). Within burned forest
there was no significant trend in the abun-
dance of foliage and ground foraging species
from 1994—1997. Within unburned forest
there was a significant decreasing trend of fo-
liage foraging species (u = —1.0, F = 0.001)
during the four years.

Foraging. — In burned forest Downy, Hairy,
Three-toed, and Black-backed woodpeckers

foraged upon standing dead trees 99% of the
time and 1% of the time on logs (n = 145).
Woodpeckers used burned trees with brown
needles 51% of the time, significantly differ-
ent than the proportion available (20%; Good-
ness of Fit: x‘ = 14.98, df = 2, F = 0.001).
They foraged predominantly on Douglas-fir
(61%), western larch (38%), and ponderosa
pine (38%; Fig. 3). Western larch (13% avail-
able) and ponderosa pine (2% available) were
used for foraging significantly more than ex-
pected (x^ = 60.58, df = 4, F = 0.001). Trees
greater than 23 cm in diameter were used
(84%) significantly more than the proportion
available (36%; Fig. 4; X" = 85.86, df = 3, F
= 0.001). Broken top snags were used (14%)
in similar proportions to their availability
(12%; x^ = 0.28, df = 1, F > 0.05). Available
snags consisted mainly of severely burned
(80%), intact top (83%), Douglas-fir (78%),
western larch (13%), or sub-alpine fir (6%)
with a mean diameter of 22 cm.

The four woodpecker species foraged dif-
ferently. Downy Woodpeckers foraged pre-
dominantly by pecking while Hairy, Three-
toed, and Black-backed woodpeckers foraged
predominantly by flaking and drilling. Hairy
and Three-toed woodpeckers foraged on sim-
ilar tree species (Fig. 3) and on similai' parts
of trees (Fig. 5). Black-backed Woodpeckers
foraged on western larch and Douglas-fir (Fig.
3) and foraged predominantly on the middle
and lower trunks, of trees (Fig. 5). Downy
Woodpeckers foraged most frequently on
branches of ponderosa pine (Figs. 3, 5).

Kreise! anil Stein • BURNED FOREST BIRDS IN WINTER

247

FIG. 3. Proportion of tree species available and used for foraging by four woodpecker species during winter
in burned forests (Goodness of Fit, the tree species distribution used by all woodpeckers combined was signif-
icantly different than the available distribution, x" = 60.58, df = 4, P = 0.001; each bird species used a
significantly different tree species distribution than what was available (all P < 0.001) and each bird species
foraged on different tree species than each other (all P < 0.003) except Hairy and Three-toed woodpeckers
which foraged similarly (x‘ = 3.573, df = 4, P > 0.05)).

DISCUSSION

Trunk and branch foraging species were
more abundant in recently burned forests than
in unbumed forests. Other studies of burned
forests had similar results (Raphael and White
1984, Hutto 1995). This may be due to chang-
es in forest structure and related food resourc-
es as a result of stand replacement fire. Stand
replacement fire changes the structure of a for-
est from a dense canopy cover, with shrub un-
dergrowth and few standing dead trees to little
canopy cover, few shrubs, and numerous
standing dead trees. Food in unbumed forests
include a variety of seeds and insects on fo-
liage, bark, and shrubs. During the winter,
food in burned forests may be limited to seeds

8-22 23-37

TREE DIAME

FIG. 4. Proportion of tree diameters available (n —
in burned forests. Significantly different tree diameters
= 85.86, df = 3, P = 0.001).

from fire opened cones and bark and wood-
boring beetle larvae in fire-killed trees (Hutto
1995). Bird species that forage on the ground
and foliage probably have more food available
in unbumed forests, and species that forage on
tmnks and branches of trees probably have
more food available in burned forests. Bark
and wood-boring beetle larvae that are abun-
dant in fire-killed trees are the major food
source of woodpeckers, especially during win-
ter (Brawn et al. 1982). Therefore, fire-killed
trees may be cmcial in supplying year round
food for tmnk and branch foraging species.
Large diameter, thick barked snags are typical
locations for bark and wood-boring beetle lar-
vae (Otvos 1965). Woodpecker species in this

■ Available

H Foraged

38-52 > 53

R CATEGORIES (CM)

173) and used for foraging (/? = 145) by woodpeckers
were used than what was available (Goodness of Fit, x'

248

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

□ Downy □ Hairy ■ Three-toed 0 Black-backed

I I

Lower Trunk Middle Trunk Upper Trunk

Branches

FIG. 5. Foraging zones used by four woodpecker species during winter in burned forests [Goodness of Fit:
Downy (n = 26, x’ = 20.89, df = 3, P = 0.001) and Black-backed (n = 42, x" = 14.83, df = 3, P = 0.002)
woodpeckers used zones significantly different from an even distribution. Each woodpecker species foraged in
significantly different zones from each other (all P < 0.012) except Hairy (n = 55) and Three-toed (n = 12)
woodpeckers which foraged in similar zones (x“ = 6.01, df = 3, F > 0.05)].

Study used large snags between 23 and 37 cm
dbh disproportionately more often for forag-
ing in burned forest. In Montana, the average
foraging snag diameter was 10-30 cm (Caton
1996). Snag species used most often for for-
aging in this study included thick barked west-
ern larch, ponderosa pine, and Douglas-fir.
Hutto (1995) and Caton (1996) also found that
woodpeckers foraged on these same snag spe-
cies in burned forests in Montana.

Within burned forest there were several
changes in the bird community from the first
winter post-fire to the fourth. Some species
were present only the first or second winter
and were absent by the third winter. Wood-
peckers, however, decreased in abundance
from the first winter to levels equal to that
found in unburned forests by the fourth winter.
In the Sierra Nevada Mountains, Bock and
Lynch ( 1 970) found woodpeckers to be in low
numbers by six years post-fire. Hutto (1995)
found Three-toed and Black-backed wood-
peckers to be most common 1-9 years post-
fire. Post-fire forests appear to support a suc-
cession of bird species, with early post-fire
conditions being conducive to specific wood-
pecker species. The decrease in abundance of
the Downy, Hairy, Three-toed, and Black-
backed woodpeckers may have occurred as a
result of a decrease in food resources. Bark
and wood-boring beetle larvae dramatically
decrease 2-5 years post-fire (Koplin 1972).
Food resources may be as important to the
distribution of these species as nest-site avail-

ability. Caton (1996) suggested that the avail-
ability of suitable foraging snags played a
large role in the distribution of woodpecker
species in burned forests in Montana. Further
studies on food resource availability in burned
and unbumed forests may better help explain
their distribution.

Non-breeding season habitat quality is
probably as important to the persistence of
bird populations as breeding season habitat
(Conner 1979) because most bird mortality
occurs during the non-breeding season (Gra-
ber and Graber 1983). Favorable winter hab-
itat can increase winter survival and result in
more birds breeding the following season. Lo-
cal movements to burned habitats by resident
bird species may increase survival. The use of
burned forests during the winter by wood-
peckers may also increase their overwinter
survivorship by increasing food and decreas-
ing predation. Different foraging techniques
observed by woodpeckers in burned forest
may allow many species to coexist. All of
these factors may contribute to the increased
capability of recently burned forests to sup-
port bird species, including woodpeckers, dur-
ing the non-breeding season.

The 100 m radius point counts seemed, to
be an adequate way to compare the relative
abundance of birds in these two forests. Visual
detections of birds in the burned and unburned
forests were probably different, and audible
detections, which most of ours were, were
probably similar in the two forests. Differenc-

Kreisel and Stein • BURNED FOREST BIRDS IN WINTER

249

es in auditory detections were most likely
minimal during the winter when it was ex-
tremely quiet. Dellasala and coworkers (1996)
found little difference in detection rates of
birds in 100 m radius point counts in young
(20 yr) and old growth forests.

Fire suppression over the past 100 years
(Agee 1994, Hejl 1994) has probably had a
major effect on bird communities in mixed-
conifer forests. Stand replacement fires may
help to increase populations of cavity-nesting
and trunk and branch foraging species. Pop-
ulations of Hairy, Three-toed, and Black-
backed woodpeckers might be maintained by
periodic occurrences of stand replacement
fires throughout the landscape. The spatial and
temporal pattern of stand replacement fires
needed to maintain bird populations needs fur-
ther investigation.

Management implications. — Forest manag-
ers can increase cavity-nesting bird popula-
tions by relaxing fire suppression policies and/
or by initiating prescribed burning programs.
To manage for cavity-nesting birds in burned
forests snags of western larch, ponderosa pine
and Douglas-fir larger than 23 cm in diameter
should be present. If managers can delay log-
ging of burned forests three to four years, the
habitat will be less suitable for trunk and
branch foraging species and logging may have
less impact on these species. Future studies
investigating snag densities and patch sizes of
burned forests required to maintain popula-
tions of cavity-nesting species during summer
and winter are needed to help managers pro-
vide for these species.

ACKNOWLEDGMENTS

We thank D. J. Kinateder for invaluable field work
and helpful discussion. We give special thanks to S. J.
Hejl and M. A. O’Connell for suggestions and com-
ments on the manuscript. Funding was provided by
Kettle Range Conservation Group, Northwest Fund for
the Environment, Kongsgaard-Goldman Foundation,
Eastern Washington University, and the Barry Me-
morial Research Grant from the Turnbull Laboratory
for Ecological Studies.

LITERATURE CITED

Agee, J. A. 1994. Fire ecology of Pacific Northwest
forests. Island Press, Washington, D.C.

Amman, G. D. and K. C. Ryan. 1991. Insect infesta-
tions of fire-injured trees in the greater Yellow-
stone area. Intermountain Research Station, Re-

search Note lNT-398. U.S. Forest Service, Ogden,
Utah.

Bi.ackeord, j. L. 1955. Woodpecker concentrations in
burned forests. Condor 57:28-30.

Blake, J. G. 1982. Influence of fire and logging on
non-breeding bird communities of ponderosa pine
forests. J. Wildl. Manage. 46:404-415.

Blondel, j., C. Ferry, and B. Frochot. 1981. Point
counts with unlimited distance. Stud. Avian Biol.
6:414-420.

Bock, C. E. and J. F. Lynch. 1970. Breeding bird pop-
ulations of burned and unburned conifer forests in
the Sierra Nevada. Condor 72:182-189.

Brawn, J. D., W. H. Elder, and K. E. Evans. 1982.
Winter foraging by cavity nesting birds in an oak-
hickory forest. Wildl. Soc. Bull. 10:271-275.

Caton, E. L. 1996. Effects of fire and salvage-logging
on a cavity-nesting bird community in northwest-
ern Montana. Ph.D. diss., Univ. of Montana, Mis-
soula.

Conner, R. N. 1979. Seasonal changes in woodpecker
foraging methods: strategies for winter survival.
Pp. 95-105 in The role of insectivorous birds in
forest ecosystems (W. Lafeber, Ed.). Academic
Press, Orlando, Florida.

Dellasala, D. A., J. C. Hagar, K. A. Engel, W. C.
McComb, R. L. Fairbanks, and E. G. Campbell.
1996. Effects of silvicultural modifications of tem-
perate rainforest on breeding and wintering bird
communities. Prince of Wales Island, southeast
Alaska. Condor 98:706-721.

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988.
The birder’s handbook: field guide to the natural
history of North American birds. Simon and
Schuster, New York.

Furniss, M. M. 1965. Susceptibility of fire-injured
Douglas-fir to bark beetle attack in southern Ida-
ho. J. For. 63:8-1 1.

Graber, j. W. and R. R. Graber. 1983. Expectable
decline of forest bird populations in severe and
mild winters. Wilson Bull. 95:682—690.

Hejl, S. J. 1994. Human-induced changes in bird pop-
ulations in coniferous forests in western North
America during the past 100 years. Stud. Avian
Biol. 15:232-246.

Hejl, S. J., J. Verner, and G. W. Bell. 1990. Se-
quential versus initial observations in studies of
avian foraging. Stud. Avian Biol. 13:166-173.

Hitchcox, S. M. 1996. Abundance and nesting success
of cavity-nesting birds in unlogged and salvage-
logged burned forest in northwestern Montana.
M.S. thesis, Univ. of Montana, Missoula.

Hutto, R. L. 1995. Composition of bird communities
following stand replacement fires in northern
Rocky Mountain conifer forests. Conserv. Biol. 9:
1041-1058.

Huito, R. L., S. M. Pletschht, and P. Hendricks.
1986. A fixed-radius point count method for non-
breeding and breeding season use. Auk 103:593—
602.

250

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Klein, B. C. 1988. Weather-dependent mixed-species
flocking during the winter. Auk 105:583-584.

Koplin, J. R. 1972. Measuring predator impact of
woodpeckers on spruce beetles. J. Wildl. Manage.
36:308-320.

Martin, T. E. 1994. BBIRD field protocol. Montana
Cooperative Wildlife Research Unit, Univ. of
Montana, Missoula.

Morrison, M. L. 1984. Influence of sample size and
sampling design on analysis of avian foraging be-
havior. Condor 86:146-150.

Otvos, I. S. 1965. Studies on avian predators of Den-
droctonus breviconiis LeConte (Coleoptera: Scol-
ytidae) with special reference to Picidae. Can.
Ent. 97:1 184-1 199.

Raphael, M. G. and M. White. 1984. Use of snags
by cavity-nesting birds in the Sierra Nevada.
Wildl. Monogr. 86:1-66.

Raphael, M. G., M. L. Morrison, and M. R Yoder-
WiLLiAMS. 1987. Breeding bird populations during
25 years of postfire succession in the Sierra Ne-
vada. Condor 89:614-626.

Spring, L. W. 1965. Climbing and pecking adaptations
in some North American woodpeckers. Condor
67:457-488.

SPSS. 1993. SPSS for Windows 6.0. SPSS, Inc., Chi-
cago, Illinois.

U.S. Dept, op Agric. 1995. Pinal supplement to final
environmental impact statement for east Curlew
Creek area timber sales. Copper Butte fire salvage.
Republic Ranger District, Republic, Washington.

WiCKMAN, B. E. 1965. Black-backed Three-toed pre-
dation of Monochamus oregonensis (Coleoptera:
Cerambycidae). Pan-Pacific Ent. 41:162-164.

Zar, j. H. 1996. Biostatistical analysis, third ed. Simon
and Schuster, Upper Saddle River, New Jersey.

Wilson Bull., I 1 1(2), 1999, pp. 251-256

NEST PREDATORS OF OPEN AND CAVITY NESTING BIRDS IN

OAK WOODLANDS

KATHRYN L. PURCELL' ^ AND JARED VERNER'

ABSTRACT. — Camera setups revealed at least three species of rodents and seven species of birds as potential
predators at artificial open nests. Surprisingly, among avian predators identified at open nests, one third were
Bullock’s Orioles (Icterus hullockii). Two rodent species and three bird species were potential predators at
artificial cavity nests. This high predator diversity was consistent with previous studies, although the number of
avian predators at open nests was higher than expected. Received 31 March 1998, accepted 22 Nov. 1998.

As the primary source of nest failure among
birds (Lack 1968, Ricklefs 1969), predation is
a likely factor affecting species’ coexistence,
habitat selection, and conservation (Zimmer-
man 1984; Martin 1988a, b). When nest pre-
dation differs among species, habitats, and lo-
cations, it can influence life history traits such
as clutch size, nest placement, developmental
period, and number of broods (Ricklefs 1969;
Martin 1988c, 1995). Avian ecologists gen-
erally agree that predation rates differ among
species nesting in cavities and open (cup)
nests (Lack 1954, Nice 1957, Ricklefs 1969).
Predators may differ as well, but little is
known about predators of bird nests because
predation is rarely observed, and observations
are biased toward diurnal predators. Some re-
searchers have made assumptions about broad
classes of predators based on the appearance
of the depredated nest, but few data exist to
support those assumptions, and authors dis-
agree on evidence used to assign depredated
nests to predator groups and the reliability of
the evidence (Best 1978, Best and Stauffer
1980, Wray et al. 1982, Boag et al. 1984, Her-
nandez et al. 1998a, Marini and Melo 1998).
Here we report results of a camera study at
both artificial open and cavity nests. The pri-
mary objective of our study was to identify
nest predators as part of a larger study of re-
productive success among birds in oak-pine
woodlands in the west-central foothills of the
Sierra Nevada of California.

STUDY AREA AND METHODS

The study was done at the San Joaquin Experimen-
tal Range, approximately 40 km north of Fresno, Cal-

' USDA Forest Service, Pacific Southwest Research
Station, 2081 E. Sierra Ave., Fresno, CA 93710.

^ Corresponding author;

E-mail: kpurcell/psw_fresno@fs.fed.us

ifornia. The San Joaquin Experimental Range covers
about 1875 ha and ranges in elevation from 215 to 520
m. Climate is Mediterranean, with cool, wet winters
and hot, dry summers. A sparse woodland overstory
of blue oak (Querciis doiiglasii), interior live oak (Q.
wislizenii), and foothill pine (Pinus sabiniana) covers
most of the San Joaquin Experimental Range. A scat-
tered understory of shrubs includes mainly wedgeleaf
ceanothus (Ceanothus cuneatus), chaparral whitethorn
(C. leucodennis), redberry (Rhamnus crocea), and
mariposa manzanita (Arctostaphylos viscida maripo-
sa). The San Joaquin Experimental Range has been
lightly to moderately grazed since about 1900 and is
surrounded on all sides by similar habitat.

Using nests of California Towhees (Pipilo crissalis)
collected at the end of the previous field season, we
situated artificial open nests low in small trees or
shrubs in positions similar to those known to be used
by California Towhees (on a forked branch or sup-
ported by several twigs). At cavity setups, eggs were
placed with a “pick-up” tool, using cavities known to
be deep enough for cavity-nesting species at the San
Joaquin Experimental Range. Most cavities were ex-
cavated by primary cavity nesters, but some natural
cavities previously used for nesting were also used. A
fiberscope (Purcell 1997) was used to guide the place-
ment of eggs in cavities, to monitor eggs for possible
predation, and to measure cavity depth. To avoid leav-
ing olfactory cues at nests, field personnel washed their
hands before going into the field with a soap developed
to remove human scent and sprayed their boots with a
scent masker. We avoided dead-end trails and did not
create paths that might lead predators to nests.

One experimental egg was placed in open nests, and
one or two eggs were placed in cavity nests, the num-
ber and type depending on availability. Most eggs used
in open nests were from wild House Sparrows (Passer
domesticus) or captive Ringed Turtle-Doves (Strepto-
pelia risoria)-, most eggs in cavities were from captive
Zebra Finches (Poephila gidlata). We sometimes used
a Buttonquail (Turni.x sp.) egg as the second egg in a
cavity nest. House Sparrow eggs were slightly smaller
and Ringed Turtle-dove eggs slightly larger than those
of California Towhees (see Baicich and Harrison
1997). Although similar in size to eggs of the Plain
Titmouse (Baeolophus inornatus), the Zebra Finch

251

252

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

eggs were smaller than eggs of all cavity-nesting spe-
cies in our study area.

Predation at open nests was monitored mechanical-
ly, with an egg encircled by a loop of wire attached
within a nest of the California Towhee. Removal of
the egg activated an electrical signal to a solenoid, trip-
ping a camera mounted nearby. We used inexpensive,
autofocus, autoflash. Keystone 550D or 590AP cam-
eras, allowing identification of both diurnal and noc-
turnal predators. (Trade names and commercial prod-
ucts are mentioned for information only; no endorse-
ment by the U.S. Department of Agriculture is im-
plied.) Details on the mechanical system for open nests
are available from KLP. At cavity nests, we used Trail-
master Active Infrared trail monitors with weather-
proof, autoflash 35 mm cameras to monitor predation.
One box that transmitted (12.1 cm L X 8.3 cm W X
4.6 cm D) and one that received (19.1 cm L X 8.9 cm
W X 5.3 cm D) the infrared beam were placed on each
side of the cavity so that an animal entering it would
break the beam, triggering the camera positioned on a
nearby branch with a good view of the cavity. Because
Trailmaster units are designed to be set up horizontally
across trails, we modified the boxes so they could be
attached easily to the tree bole or limb with bungee
cords. Sensitivity was set at the minimum delay of 0.5
s (one pulse) before an event was recorded, and the
camera delay between photos was set at the minimum
of 6 s.

Based on the nesting seasons of cavity- and open-
nesting species, cameras were set up from March
through June 1995 (cavities) and April through June
1995 (open). We used 10 open-nest setups to monitor
70 open nests, and 7 Trailmaster monitors at 61 cavity
nests. Some data on cavity nests were also included
from the 1993 and 1994 field seasons (eight each year).
All setups were checked about every 4 days. If an egg
was taken, or not taken after 14 days, the setup was
dismantled and moved to another location and installed
using fresh eggs.

All artificial nests of the same nest type were sep-
arated by at least 200 m in an effort to reduce the
chance of visitation by the same animal at two or more
setups. This distance was thought to be enough to as-
sure independent samples of the small mammals iden-
tified as predators in this study. Based on spot mapping
at the San Joaquin Experimental Range (unpublished
data), territories of the Western Scrub-Jay (Apheloco-
ma californica), a common nest predator, were ap-
proximately 120-210 m in diameter. Mean territory
diameters of other common bird species ranged from
180 m (California Towhee) to 310 m (Western King-
bird, Tyranniis verticali.s). Some cavity setups were
closer than 200 m to open set-ups, but cameras and
eggs were not placed concurrently at the two nest
types.

At open nests, we measured nest height and the
height and diameter of the shrub or small tree contain-
ing the nest. Diameter was measured as the mean of
the maximum crown diameter and the widest diameter
perpendicular to the maximum diameter. At cavity

nests, we measured nest height, cavity depth, and hor-
izontal and vertical entrance diameters. We tested dif-
ferences in these attributes between predated and un-
predated nests using two-tailed /-tests (SAS version
6.12 for Windows, SAS Institute 1988), a = 0.05, and
Bonferroni adjustments for multiple tests. We calcu-
lated power according to Abramowitz and Stegun
(1964) based on specified effect sizes, an a of 0.05,
and two-tailed tests using an inhouse SAS program.

RESULTS

Open nests. — Eggs were removed from 39
of the 70 open nest setups, but we could iden-
tify the animals at only 29 of those. Rodents
were identified at four (14%); deer mouse
{Peromyscus spp., two cases), California
ground squirrel (Spermophilus beecheyi, one
case), and Merriam’s chipmunk {Eutamias
merriami, one case). At least five bird species
were photographed at the remaining 25 nests
(86%) from which an egg was taken: Western
Scrub-Jay (12 cases), Bullock’s Oriole {Icter-
us bullockii, 1 cases). Acorn Woodpecker
{Melanerpes fonnicivorus, 1 case). Western
Kingbird (1 case, a pair), and California To-
whee (1 case). We could not identify the bird
species at the three remaining setups.

In three additional cases, eggs were pecked,
chewed, or otherwise damaged but not re-
moved. A pair of Plain Titmice pecked a large
hole in the egg at one nest; a female Brown-
headed Cowbird {Molothrus ater) punctured
the egg in another nest; and either a dusky-
footed woodrat (Neotoma fuscipes) or a West-
ern Scmb-Jay chewed or pecked another egg
(both species were photographed).

We may have underestimated nocturnal pre-
dation. Only one photo of a deer mouse was
taken at night. In 2 of the 10 cases with no
identifiable predator, photos were taken at
night but were dark, perhaps because the cam-
era’s flash was too far from the nest or failed
to operate properly (see also Hernandez et al.
1998b).

No attribute measured at open nest setups
differed significantly between predated and
nonpredated nests (Table 1; P > 0.05 in all
cases, P < 0.017 required for Bonfenoni ad-
justment for multiple tests).

Cavity nests. — Photos were taken at 47 of
69 cavity nests where the egg was removed
or pecked open. Interpretation of the photos
was complicated, however. First, the mini-
mum camera delay did not allow a photo each

Purcell and Verner • NEST PREDATORS

253

TABLE 1. Nest site variables and results of r-tests for nonpredated (n = 18) and predated (n — 48) open
nests at the San Joaquin Experimental Range.

Nonpredated
open nests

Mean (SE)

Predated
open nests

Mean (SE)

Power

Nest height (in)

1.21 (0.11)

1.11 (0.06)

0.48

0.97'’

Substrate height (m)

4.61 (0.81)

4.25 (0.41)

0.67

0.9L

Substrate diameter (m)

6.03 (1.34)

4.90 (0.40)

0.43

0.82”

“ A F-value of 0.017 is needed for significance at a = 0.05 after Bonferroni adjustment for multiple comparisons.
’’ Based on an effect size of 0.5m.

Based on an effect size of 3 m.

time the infrared beam was broken after an
initial photo was taken. Consequently, photos
rarely showed animals leaving a cavity nest
and none showed one “caught in the act” of
leaving with an egg. We had to assume that
an animal in a photo consumed the missing
egg. Second, in seven cases the egg was gone
and no animal was evident in the photo (see
also Brooks 1996). The departure of some
predators from a cavity may have been too
rapid for it to be caught in the photo, or photos
with no animal may have resulted from direct
sunlight entering the receiver window (Kucera
and Barrett 1993). Third, in 15 instances more
than one species entered the cavity and tripped
the camera before we found that the egg was
gone. Fourth, in 16 cases nest material was
added to the cavity by birds or mammals so
we could not ascertain whether the eggs had
been eaten or simply buried. We did not in-
clude these cases in our comparisons of pre-
dated and nonpredated nests.

In nine cavity setups with missing eggs,
only one species appeared in the photos. The
assumed predators were European Starling
(Sturnus vulgaris; four cases). House Wren
{Troglodytes aedon; one case). Western Blue-

bird (Sialia mexicana', one case), deer mouse
(one case), and unidentified squirrels (proba-
bly California ground squirrels; two cases).
All photos were taken during daylight hours
except that of the deer mouse.

No attribute measured at cavity nests dif-
fered significantly between predated and non-
predated nests (Table 2, P > 0.05 in all cases,
P < 0.01 required after adjusting for multiple
tests).

DISCUSSION

Our results are consistent with other studies
using artificial nests in finding a high diversity
of nest predators, ranging from six to nine
species (Henry 1969, Wilcove 1985, Reitsma
et al. 1990, Leimgruber et al. 1994). We iden-
tified eight species at open nests, and two
(possibly three) other species pecked or
pierced eggs. Pieman and Schriml (1994)
found only one or two major predator species
in each of four vegetation types, although
predator diversity ranged from four (marsh) to
nine species (scrubland and forest). Lack of
independence of the setups in their study may
have overestimated the importance of some
predator species (see below). Interestingly, all

TABLE 2. Nest site variables and results of /-tests for predated and nonpredated cavity nests at the San
Joaquin Experimental Range.

Nonpredated cavities Predated cavities

Mean (SE)

Power

Nest height (m)

3.82 (0.24)

3.78 (0.23)

0.90

0.98'’

Depth (cm)

25.1 (1.94)

31.0 (2.42)

0.06

0.95”

Vertical entrance diameter (cm)

5.68 (0.67)

5.06 (0.16)

0.37

1.00*'

Horizontal entrance diameter (cm)

5.19 (0.19)

5.14 (0.13)

0.83

0.98“

Minimum entrance diameter (cm)

4.80 (0.15)

4.94 (0.14)

0.51

0.99“

“ A P- value of 0.010 is needed for significance at a = 0.05 after Bonferroni adjustment for multiple comparisons.
Based on an effect size of 1 m.

Based on an effect size of 0.5 cm.

Based on an effect size of 1.0 cm.

254

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

but one (Acom Woodpecker) of the avian
predators at open nests in our study were open
nesters and all avian predators at cavity nests
were cavity nesters.

We know of only two efforts to study pre-
dation at experimental cavity nests (Wilcove
1985, Sandstrdm 1991), neither of which
identified predators. Although some species
photographed at our cavity setups may have
been only reconnoitering potential nest sites,
they may also opportunistically eat eggs thus
encountered in a cavity. The importance of
birds as predators at active cavity nests, de-
fended by nesting birds, is unknown but prob-
ably significant in some instances. European
Starlings are known to usurp nest sites from
other bird species (Troetschler 1976, Ingold
1989) and, although the contents of the cavi-
ties were unknown, eggs were likely present
in some when usurpation occurred late enough
in the nesting cycle for egg-laying or incu-
bation to have begun. Our results suggest that
starlings probably consume eggs in the course
of usurping nest sites.

Although several species of corvids are be-
lieved to be important predators at open nests
(Yahner and Wright 1985, Yahner and Scott
1988, Andren 1992, Pieman and Schriml
1994, Hannon and Cotterill 1998), nest pre-
dation at open nests by noncorvids has not
often been documented. Gates and Gysel
(1978) reported anecdotal evidence of nest
predation by an Eastern Screech-Owl {Otus
asio). Pieman (1987) photographed Marsh
Wrens (Cistothorus palustnis). House Wrens,
and Gray Catbirds (Diimetella carolinensis)
predating nests. Pieman and Schriml (1994)
recorded predation events by Broad-winged
Hawks (Buteo platypterus). Eastern Mead-
owlarks {Sturnella magna), and Red-winged
Blackbirds (Agelaius phoeniceus). The Marsh
Wren was the only predator recorded at nests
of Yellow-headed Blackbirds (Xanthocephal-
us xanthocephalus'. Pieman and Isabelle
1995). Predation of open nests by woodpeck-
ers has been documented rarely (Bent 1939;
Watt 1980; Hernandez et al. 1998a, b; Robert
Cooper, unpubl. data), and never by the Acorn
Woodpecker. Egg removal by Bullock’s Ori-
oles was surprisingly common in this study.
Both sexes of this oriole are known to be ejec-
tors of Brown-headed Cowbird eggs, and they
sometimes consume the eggs before removing

the shells (Sealy and Neudorf 1995). Al-
though some of these “predators” may de-
stroy nests or eggs with no nutritional moti-
vation, predation of open nests by noncorvids
may be more frequent than previously
thought.

Leimgruber and coworkers (1994) and Pie-
man and Schriml (1994) also found that po-
tential predators visiting nests did not always
eat the eggs. Consistent with the findings of
Pieman and Schriml, mammals in our study
nearly always took the egg. Pieman and
Schriml (1994) classified as “accidental visi-
tors” several bird species that visited nests,
including Red-winged Blackbirds, that appar-
ently ate eggs at 6 of 29 nests visited. Re-
gardless of motivation, the effect of egg re-
moval on the nesting birds is the same.

Artificial nest studies are just that — artifi-
cial; some biases are certainly involved (Mar-
tin 1987, Reitsma et al. 1990, Whelan et al.
1994, Marini and Melo 1998, Wilson et al.
1998). As visual predators, birds may key in
on cameras or unrealistic nest placements, or
they may follow field workers and leai'n to
associate conspicuous mai'kers at nests with
food (Picozzi 1975, Gotmark 1992). At open
nests we were able to attain a realistic nest
placement or a good camera view, but usually
not both. At cavity nests, the transmitters and
receivers may have inhibited predators from
going to a cavity or may have attracted curi-
ous predators. Predation also may have been
more or less likely at artificial nests than real
nests defended by adult birds, but even real
nests are left unattended at regular intervals.
In spite of these potential biases, we feel that
useful data can be obtained from studies of
artificial nests because they establish baseline
data in an arena where so little is known.

We believe that the value of studies using
artificial nests can be substantially increased
if studies are designed more cai'efully to re-
duce potential biases. For example, most of
the studies we reviewed used distances be-
tween setups ranging from 20 to 60 m. Such
short intervals risk detection of the same in-
dividual predator at two or more setups, vio-
lating assumptions of independence. Ideally,
the distance between artificial nests should ex-
ceed the largest home-range diameter of the
suite of likely predators. One might argue that
shorter distances are appropriate for examin-

Purcell ami Verner • NEST PREDATORS

255

ing predation rates, although rates from arti-
ficial nest studies are of questionable value for
extrapolation to natural conditions (Martin
1987, Whelan et al. 1994, Wilson et al. 1998).
Replication over large areas is required to
characterize the suite of predators for a given
vegetation type, since predators are often un-
evenly distributed in space and time.

Further problems of independence may
have occurred in studies that replaced eggs in
nests that had been predated previously. Nour
and coworkers (1993) suggested that such egg
replacement may not be a problem in studies
using plasticine eggs or eggs made from mod-
eling clay because the eggs are not eaten and
provide the predator no incentive to return,
although predators could avoid nests with clay
eggs because of prior negative conditioning.

ACKNOWLEDGMENTS

We thank L. and S. Garner for instructions to con-
struct the camera apparatus to photograph nest preda-
tors at open nests. We were aided in the field by K.
Kalin, R. Miller, and, especially, by D. Cubanski and
J. Ohanesian, who improved the open-nest apparatus.
The manuscript benefitted from reviews by S. Hejl, C.
Maguire, C. Meslow, W. Laudenslayer, L. Reitsma, C.
Whelan, and three anonymous reviewers.

LITERATURE CITED

Abramowitz, M. and I. A. Stegun. 1964. Handbook
of mathematical functions. Dover Publications,
New York.

Andren, H. 1992. Corvid density and nest predation
in relation to forest fragmentation: a landscape
perspective. Ecology 73:794-804.

Baicich, P. J. and C. J. O. Harrison. 1997. A guide
to the nests, eggs, and nestlings of North Ameri-
can birds, second ed. Academic Press, San Diego,
California.

Bent, A. C. 1939. Life histories of North American
woodpeckers. U.S. Nat. Mus. Bull. 174:1—334.
Best, L. B. 1978. Field Sparrow reproductive success
and nesting ecology. Auk 95:9-22.

Best, L. B. and D. E Staubter. 1980. Factors affect-
ing nesting success in riparian bird communities.
Condor 82:149-158.

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.

Brooks, R. T. 1996. Assessment of two camera-based
systems for monitoring arboreal wildlife. Wildl.
Soc. Bull. 24:298-300.

Gates, J. E. and L. W. Gysel. 1978. Avian nest dis-
persion and fledging success in field-forest eco-
tones. Ecology 59:871-883.

Gotmark, F. 1992. The effects of investigator distur-
bance on nesting birds. Curr. Ornithol. 9:63-104.

Hannon, S. J. and S. E. Cotterii.l. 1998. Nest pre-
dation in aspen woodlols in an agricultural area in
Alberta: the enemy from within. Auk I 15:16-25.

Henry, V. G. 1969. Predation on dummy nests of
ground-nesting birds in the southern Appala-
chians. J. Wildl. Manage. 33:169-172.

Hernandez, E, D. Roluns, and R. Cantu. 1998a.
Evaluating evidence to identify ground-nest pred-
ators in west Texas. Wildl. Soc. Bull. 25:826-831.

Hernandez, E, D. Rollins, and R. Cantu. 1998b. An
evaluation of Trailmaster camera systems for iden-
tifying ground-nest predators. Wildl. Soc. Bull.
25:848-853.

Kucera, T. E. and R. H. Barrett. 1993. The Trail-
master camera system for detecting wildlife.
Wildl. Soc. Bull. 21:505-508.

Lack, D. 1954. The natural regulation of animal num-
bers. Oxford Univ. Press, London, U.K.

Lack, D. 1968. Ecological adaptations for breeding in
birds. Methuen, London, U.K.

Leimgruber, R, W. j. McShea, and J. H. Rappole.
1994. Predation on artificial nests in large forest
blocks. J. Wildl. Manage. 58:254-260.

Marini, M. A. and C. Melo. 1998. Predators of quail
eggs, and the evidence of the remains: implica-
tions for nest predation studies. Condor 100:395-
399.

Martin, T. E. 1987. Artificial nest experiments: effects
of nest appearance and type of predator. Condor
89:925-928.

Martin, T. E. 1988a. Processes organizing open-nest-
ing bird assemblages: competition or nest preda-
tion? Evol. Ecol. 2:37-50.

Martin, T. E. 1988b. On the advantage of being dif-
ferent: nest predation and the coexistence of bird
species. Proc. Nat. Acad. Sci., USA 85:2196-
2199.

Martin, T. E. 1988c. Nest placement: implications for
selected life-history traits, with special reference
to clutch size. Am. Nat. 132:900-910.

Martin, T. E. 1995. Avian life history evolution in
relation to nest sites, nest predation, and food.
Ecol. Monogr. 65:101-127.

Nice, M. M. 1957. Nesting success in altricial birds.
Auk 74:305-321.

Nour, N., E. Matthysen, and A. A. Dhondt. 1993.
Artificial nest predation and habitat fragmentation:
different trends in bird and mammal predators.
Ecography 16:111-116.

PiCMAN, J. 1987. An inexpensive camera setup for the
study of egg predation at artificial nests. J. Field
Ornithol. 58:372-382.

PiCMAN, J. AND A. Isabelle. 1995. Sources of mortality
and correlates of nesting success in Yellow-head-
ed Blackbirds. Auk 112:183-191.

PiCMAN, J. AND L. M. ScHRiML. 1994. A Camera study
of temporal patterns of nest predation in different
habitats. Wilson Bull. 106:456-465.

256

THE WILSON BULLETIN • Vol. HI, No. 2, June 1999

Picozzi, N. 1975. Crow predation on marked nests. J.
Wildl. Manage. 39:151-155.

Purcell, K. L. 1997. Use of a fiberscope for exam-
ining cavity nests. J. Lield Ornithol. 68:283-286.

Reitsma, L. R., R. T. Holmes, and T. W. Sherry.
1990. Effects of removal of red squirrels, Tamias-
ciurus hudsonicus, and eastern chipmunks, Tam-
ias striatus, on nest predation in a northern hard-
wood forest: an artificial nest experiment. Oikos
57:375-380.

Ricklefs, R. E. 1969. An analysis of nesting mortality
in birds. Smithson. Contrib. Zool. 9:1-48.

Sandstrom, U. 1991. Enhanced predation rates on
cavity bird nests at deciduous forest edges — an
experimental study. Ornis Fenn. 68:93-98.

SAS Institute. 1988. SAS/STAT user’s guide: statis-
tics. Release 6.03. SAS Institute, Inc., Cary, North
Carolinia.

Sealy, S. G. and D. L. Neudorf. 1995. Male Northern
Orioles eject cowbird eggs: implications for the
evolution of rejection behavior. Condor 97:369—
375.

Watt, D. J. 1980. Red-bellied Woodpecker predation

on nestling American Redstarts. Wilson Bull. 92:
249.

Whelan, C. J., M. L. Dilger, D. Robson, N. Hallyn,
AND S. Dilger. 1994. Effects of olfactory cues on
artificial-nest experiments. Auk 111:945-952.

Welcove, D. S. 1985. Nest predation in forest tracts
and the decline of migratory songbirds. Ecology
66:1211-1214.

Wilson, G. R., M. C. Brittingham, and L. J. Good-
rich. 1998. How well do artificial nests estimate
success of real nests? Condor 100:357—364.

Wray, T. II, K. A. Strait, and R. C. Whitmore. 1982.
Reproductive success of grassland sparrows on a
reclaimed surface mine in West Virginia. Auk 99:
157-164.

Yahner, R. H. and D. P. Scott. 1988. Effects of forest
fragmentation on depredation of artificial nests. J.
Wildl. Manage. 52:158-161.

Yahner, R. H. and A. L. Wright. 1985. Depredation
on artificial ground nests: effects of edge and plot
age. J. Wildl. Manage. 49:508-513.

Zimmerman, J. L. 1984. Nest predation and its rela-
tionship to habitat and nest density in Dickcissels.
Condor 86:68-72.

Short Communications

Wilson Bull.. 111(2). 1999, pp. 257-261

Juvenile Marbled Murrelet Nurseries and the Productivity Index

Katherine J. Kuletz'-^ and John F. Piatt-

ABSTRACT. — Late summer counts of juveniles at
sea are used as an index of Marbled Murrelet (Bra-
chyramplius mannoratu.s) reproductive success, but lit-
tle is known about juvenile dispersal or habitat use.
Further, it is not known whether these counts accu-
rately rellect absolute breeding success. To address
these questions we conducted five boat surveys for
Marbled Murrelets and Pigeon Guillemots (Cepphus
coliimba) in Kachemak Bay, Alaska between 7-24 Au-
gust 1996. Juvenile murrelet distribution in the bay
was patchy, and we identified a juvenile Marbled Mur-
relet ‘nursery’ area in the outer bay. Fifty-three of 61
juvenile murrelets were in this area, whereas after-
hatch-year (AHY) murrelets were dispersed throughout
the bay, as were juvenile and AHY Pigeon Guillemots.
The murrelet nursery was characterized by water in-
side of or at the edge of a 20 m deep contour, semi-
protected seas, productive waters, and a large bed of
Nereocystis kelp. Juveniles comprised 16.1% of all
murrelets and 24.8% of all guillemots observed at sea.
These data suggest a maximum reproductive success
of 0.32 chicks/pair if all AHY murrelets were breeding
and 0.46 chicks/pair if only 70% of AHY murrelets
were breeding. For guillemots, maximum productivity
estimated from at-sea counts was 0.50 chicks/pair if
all AHY were breeding and 0.71 chicks/pair if only
70% were breeding. The guillemot estimate was sim-
ilar to that obtained by concurrent studies at nine guil-
lemot colonies in the bay (0.56 chicks/pair). These re-
sults suggest that at sea surveys in late summer provide
a reasonable index of local productivity for nearshore
alcids. Further, if murrelet nursery areas can be found,
at sea counts may provide a valid measure of absolute
productivity. Received Jl June 1998, accepted 7 Jan.
1999.

Nests of the Marbled Murrelet (Brachyram-
phus marmoratus) are difficult to find or
study, and reproductive success is known only
from widely scattered nests studied over many
years. Because of the murrelet’s threatened

' U.S. Fish and Wildlife Service, 1011 E. Tudor Rd.,
Anchorage, AK 99503.

-Alaska Biological Sciences Center, U.S.G.S., 101 1
E. Tudor Rd., Anchorage, AK 99503.

^ Corresponding author;

E-mail; kathy_kuletz@fws.gov

status from British Columbia to California
(Ralph et al. 1995), considerable effort has
been devoted to finding alternate means of es-
timating murrelet reproductive success. The
most practical approach is to use a productiv-
ity index based on surveys at sea, which uses
the ratio of juveniles to adults or juvenile den-
sities during the fledging period as indices of
production (Ralph and Long 1995, Strong
1995, Kuletz and Kendall 1998). To be ac-
curate, surveys require some knowledge of
fledgling dispersal at sea, but little is known
about juvenile movements or habitat use.

Anecdotal evidence suggests that juvenile
Marbled Murrelets sometimes congregate in
“nursery areas”, often near shore or in exten-
sive kelp beds (Sealy 1975, Beissinger 1995,
Strachan et al. 1995, Strong et al. 1995). If
juveniles gather in specific habitats after
fledging, productivity surveys could be im-
proved by identifying their location and time
of use. Here, we report on a juvenile murrelet
nursery and describe associated habitat fea-
tures. We estimate muixelet productivity from
the ratio of juveniles to adults at sea, and com-
pare this with Pigeon Guillemot {Cepphus col-
umba) productivity estimates obtained by both
counts at sea and local colony studies.

STUDY AREA AND METHODS

We conducted surveys in Kachemak Bay. southcen-
tral Alaska on five days between 7-24 August 1996
(Eig. 1). We surveyed the south side of Kachemak Bay
because Marbled Munelet densities are highest on the
south side, which has deep water, many side bays, and
a predominantly rocky, convoluted shoreline (Agler et
al. 1998).

From a 10 m vessel we counted all Marbled Mur-
relets and Pigeon Guillemots within 100 m either side
of the boat. Two observers used 8 X 42 and 10 X 50
binoculars to identify species and plumages. Juvenile
murrelets, which resemble adults in basic plumage,
were identified using characteristics described in Cart-
er and Stein (1995) and Kuletz and Kendall (1998). A
third person entered observations into a laptop com-
puter using DLOG (Ecological Consulting Inc., Port-

257

258

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

FIG. 1. Survey routes (a-j) in Kachemak Bay, Alaska, surveyed by boat on five days on 7-24 August 1996.

land, Oregon). The DLOG data entry program was
linked with a Global Positioning System and every
observation had an associated latitude and longitude.
Survey routes followed a path parallel to shore. For
most of the survey we used radar to maintain a dis-
tance of 100 m from shore. In rocky or shallow sec-
tions we surveyed outside the 20 m depth contour.
From the head of the bay to Glacier Spit, and from
Kasitsna Bay to Seldovia Bay, we also surveyed 0.5-
1.0 km offshore (Fig. 1). The vessel traveled at speeds
of about 7 km/hr, but because this was a reconnais-
sance survey, we temporarily paused or left our path
to observe potential juvenile murrelets or guillemots
(birds in black and white plumages).

We surveyed a linear distance of 214 km on 10 dif-

ferent survey routes over five days for a total area sur-
veyed of 36.6 km- (Table 1). We refer to the area from
the head of Kachemak Bay to China Foot Bay as the
inner bay and the area west of China Foot to Seldovia
Bay as the outer bay. Our main objective was to de-
scribe the spatial distribution of murrelets during the
fledging period, but we obtained some temporal cov-
erage. Portions of the survey routes overlapped on dif-
ferent days and all regions of the bay were surveyed
both early and late in the fledging period (Table 1).
Survey dates (7-24 August) encompassed the main
and peak fledging period for mumelets, based on five
replicate surveys conducted independently between 7
August and 4 September 1996 near Kasitsna Bay by
KJK and J. Figurski. These dates correspond to the

TABLE 1. Numbers of adult (after-hatch-year) and juvenile Marbled Murrelets and Pigeon Guillemots ob-
served on survey routes in Kachemak Bay, Alaska, in August 1996. Area (km-) was calculated from the survey
route length X width.

Bay area

Survey

route

Date

Area

(kni2)

No. Marbled Murrelets

No. Pigeon Guillemots

Adults

Juveniles

Adults

Juveniles

Inner

8-13

2.86

Inner

8-24

3.44

Inner

8-13

3.64

Inner

8-24

1.52

Total

1 1.46

186

Outer

8-13

2.89

Outer

8-07

5.25

Outer

8-07

1.26

Outer

8-12

1.54

Outer

8-12

4.66

Outer

8-23

9.57

Total

25.17

131

139

SHORT COMMUNICATIONS

259

FIG. 2. Distribution of adult (after-hatch-year) and juvenile Marbled Murrelets on surveys conducted in
Kachemak Bay, Alaska, on 7-24 August 1996.

fledging period in nearby Prince William Sound, Alas-
ka (Kuletz and Kendall 1998). Pigeon Guillemot fledg-
ing dates were similar and were verified from local
colony studies (Piatt et al. 1997).

Because we wanted to describe the general distri-
bution of murrelets and our survey routes varied, we
pooled all bird counts for a single tally. For both Mar-
bled Murrelets and Pigeon Guillemots we determined
the ratio at sea of juveniles to adults and subadults
(after-hatching-year birds; AHY). We also calculated
an index of juveniles/pair based on counts of juveniles
and half the number of adults counted on the same
surveys. Piatt and coworkers (1997) obtained detailed
observations of Pigeon Guillemots on 60 nests in 9
colonies distributed along the south shore of Kache-
mak Bay from Glacier Spit to Seldovia Bay.

RESULTS

Fifty-nine of 61 juvenile Marbled Murrelets
were found in outer Kachemak Bay and two
were found in the inner bay (Table 1). Most
of the juveniles in the outer bay were concen-
trated 0. 5-1.0 km offshore, near the mouth of
Seldovia Bay (Fig. 2). This area has an exten-
sive kelp bed {Nereocystis sp.) and covers an
underwater shelf less than 20 m deep. Adult
Marbled Murrelets (n = 317; 5 in basic plum-
age) were distributed throughout Kachemak
Bay, with highest densities in the inner bay
between Glacier Spit and the bay head (Fig.
2). We found Pigeon Guillemots, both adults
(n = 249) and juveniles (n = 62), disti'ibuted
throughout Kachemak Bay (Table 1 ).

Juveniles represented 16.1% of all Marbled
Murrelets and 24.8% of all Pigeon Guillemots
counted. If all of the AHY Marbled Murrelets
were breeding, our counts suggest a maximum
reproductive success of 0.32 chicks/pair. A
more conservative estimate is that only 70%
of AHY birds were breeding (Piatt and Ford
1993), and therefore maximum productivity is
calculated as 0.46 chicks/pair. For Pigeon
Guillemots, the maximum productivity would
be 0.50 chicks/pair if all AHY birds were
breeding, and 0.71 chick/pair if only 70%
were breeding adults.

DISCUSSION

Juvenile Marbled Munelets in Kachemak
Bay showed a clear preference for the kelp
beds approximately 4 km on either side of the
mouth of Seldovia Bay. This distribution con-
trasts shaiply with the distribution of adult
mun-elets found throughout the bay. Adult
munelets forage on Pacific sand lance (Am-
modytes hexaptenis) in the inner bay (J. Piatt,
unpubl. data), suggesting that the distribution
of forage fish was not limiting the distribution
of juvenile munelets.

Why were juvenile munelets concentrated
along the shore in outer Kachemak Bay and
in extensive, dense beds of Nereocystis kelp?
Although exposed relative to the inner bay,
the orientation of the shoreline in this area

260

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

provided protection from prevailing south-
westerly winds. The southwest portion of Ka-
chemak Bay receives up welled waters from
the Alaska Coastal Current entering Cook In-
let from the southeast, and gyres in the outer
bay retain nutrients and promote high local
productivity (Trasky et al. 1977). The pres-
ence of Nereocystis, which attach to rocky
substrate and grow in water 20-40 m deep
where fast currents or upwelling occurs, is of-
ten associated with productive waters (Lalli
and Parsons 1993). Thus, shallow water, semi-
protected seas, the presence of kelp, and lo-
cally productive waters appear to combine
here to create a favorable nursery area for
newly-fledged murrelets. In addition, the kelp
made it difficult to see the juveniles, and so
may provide protection from avian predators
such as gulls and Bald Eagles {Haliaeetus leii-
cocephalus), which are common in this area.
Large Nereocystis kelp beds are not common
elsewhere in Kachemak Bay so this feature
may be the primary defining characteristic of
the nursery.

Juvenile murrelets may use the inner bay
temporarily after fledging, and if fledging
peaked early in August 1996, it is possible
that we missed seeing them before they emi-
grated to the outer bay. It is also possible that
juveniles were absent from the inner bay be-
cause few murrelets may breed there now as
the result of extensive damage to mature for-
ests from spruce beetle (Dendroctonus nifi-
pennis). However, the middle portion of the
bay (China Foot to Kasitsna Bay) still has
largely intact forests, and while the inner bay
is clearly an important foraging area for
adults, most juveniles were found in the outer
bay. The use of kelp beds in the outer bay by
juvenile murrelets appears to be a recurring
event; we have observed juvenile murrelets in
this area in previous years. Surveys of the en-
tire bay throughout the fledging period would
be necessary to determine whether, and if so
when, juveniles from throughout the bay
move to the kelp beds.

Estimates of Pigeon Guillemot productivity
obtained from juvenile surveys at sea com-
pared well to the productivity of guillemots
measured from colony-based reproductive
studies. Pigeon Guillemots at nine Kachemak
Bay colonies in 1996 produced 0.56 chicks/
pair, which falls within the range we estimated

from counts at sea. The estimate of production
we obtained for Marbled Murrelets also ap-
proximates that found for Marbled Murrelets
throughout their range (0.28 chicks/pair),
based on 32 nests followed to completion
(Nelson and Hamer 1995). It is noteworthy
that our estimate of murrelet production in
Kachemak Bay is much higher than those cal-
culated from surveys at sea in areas south of
Alaska (e.g., 0.001-0.1 1 chicks/pair), even af-
ter adjustments (0.01-0.17 chicks/pair) for the
timing of surveys (Beissinger 1995). This is
undoubtedly because we located the nursery
area near Seldovia Bay, which accounted for
53 of 61 juveniles we observed on surveys.

While the possibility of juvenile murrelet
nurseries has been suggested in some areas
(Sealy 1975, Strachan et al. 1995), they have
never been documented, and munelet distri-
bution may not always be as patchy as it ap-
pears to be in Kachemak Bay. In southeast
Alaska, VanVliet (pers. comm.) observed ju-
venile murrelets clustered near or in kelp beds
in late August in discrete areas of Port Al-
thorp, whereas adults were distributed from
Inian Pass to Icy Strait. In Prince William
Sound, Alaska, however, juvenile murrelets
were evenly dispersed in nearshore waters
(relative to local murrelet abundance), with
the exception of highly exposed shoreline
where they were absent (Kuletz et al. 1997).
The areas surveyed in Prince William Sound
did not have large kelp beds and were char-
acterized by convoluted, rocky shorelines with
numerous protected bays and coves. In addi-
tion, the juveniles in Prince William Sound
may not travel fai' in the first two weeks after
fledging (Kuletz and Marks 1997; Kuletz, un-
publ. data).

These results confirm that surveys at sea
provide a reasonable index of productivity for
nearshore seabirds such as Pigeon Guillemots
and Marbled Murrelets. However, it is impor-
tant to determine the post-fledging movements
of adults and juveniles for any given area of
study because adultrjuvenile ratios are sensi-
tive to late summer movements of adults and
subadults (Kuletz and Kendall 1998). Al-
though our temporal data were limited in this
study, we did not find obvious declines in
adult numbers in August, such as occurs in
Prince William Sound. If juvenile munelet
nurseries can be located, it would facilitate the

SHORT COMMUNICATIONS

261

use of juvenile densities to measure produc-
tivity and thus avoid problems associated with
using adult:juvenile ratios (Kuletz and Ken-
dall 1998). If adults remain in an area during
the main fledging period where muirelet nurs-
eries exist, surveys at sea may provide a valid
measure of absolute productivity, and not just
an index of production (e.g., Beissinger 1995).

ACKNOWLEDGMENTS

This work was conducted with funding from the
Exxon Valdez Oil Spill Trustee Council, the U.S. Geo-
logical Survey, and the U.S. Fish and Wildlife Service.
We are grateful to the Alaska Maritime National Wild-
life Refuge for logistic support, to B. Keitt and J. Fi-
gurski for assistance with surveys, and to M. Litzow
for use of unpublished data on Pigeon Guillemot pro-
ductivity.

LITERATURE CITED

Agler, B. a., S. J. Kendall, and D. B. Irons. 1998.
Abundance and distribution of Marbled and Kitt-
litz’s murrelets in southcentral and southeast Alas-
ka. Condor 100:254-265.

Beissinger, S. R. 1995. Population trends of the Mar-
bled Murrelet projected from demographic anal-
yses. USDA For. Serv. Gen. Tech. Rep. PSW-
GTR- 152:385-394.

Carter, H. R. and J. L. Stein. 1995. Molts and plum-
ages in the annual cycle of the Marbled Murrelet.
USDA For. Serv. Gen. Tech. Rep. PSW-GTR-152:
99-112.

Kuletz, K. J., S. Kendall, and D. Nigro. 1997. Rel-
ative abundance of adult and juvenile Marbled
Murrelets in Prince William Sound, Alaska: de-
veloping a productivity index. Exxon Valdez Oil
Spill Restoration Proj. Final Rep. 95031. U.S. Fish
and Wildl. Serv., Anchorage, Alaska.

Kuletz, K. J. and D. K. Marks. 1997. Post-fledging
behavior of a radio-tagged juvenile Marbled Mur-
relet. J. Field Ornithol. 68:421-425.

Kuletz, K. J. and S. J. Kendall. 1998. A productiv-

ity index for Marbled Murrelets in Alaska based
on surveys at sea. J. Wildl. Manage. 62:446-460.
Lalli, C. M. and T. R. Parsons. 1993. Biological
oceanography: an introduction. Pergamon Press,
New York.

Nelson, S. K. and T. E. Hamer. 1995. Nest success
and the effects of predation on Marbled Murrelets.
USDA For. Serv. Gen. Tech. Rep. PSW-GTR-152:
89-98.

Piatt, J. F. and R. G. Ford. 1993. Distribution and
abundance of Marbled Murrelets in Alaska. Con-
dor 95:662—669.

Piatt, J. F, M. Robards, S. Zador, M. Litzow, and
G. Drew. 1997. Cook Inlet seabirds and forage
fish studies. Exxon Valdez Oil Spill Restoration
Proj. Ann. Rep. 96163M, Biol. Resour. Div., U.S.
Geological Survey, Anchorage, Alaska.

Ralph, C. J., G. L. Hunt, Jr., M. G. Raphael, and J.
F. Piatt. 1995. Ecology and conservation of the
Marbled Murrelet. U.S. For. Serv. Gen. Tech.
Rep., PSW-GTR-152: 1-5.

Ralph, C. J. and L. L. Long. 1995. Productivity of
Marbled Murrelets in California from observa-
tions of young at sea. USDA For. Serv. Gen. Tech.
Rep. PSW-GTR-152:371-380.

Sealy, S. G. 1975. Aspects of the breeding biology
of the Marbled Murrelet in British Columbia.
Bird-Banding 46:141-154.

Strachan, G., M. McAllister, and C. J. Ralph.
1995. Marbled Murrelet at-sea and foraging be-
havior. USDA For. Serv. Gen. Tech. Rep. PSW-
GTR-152:247-253.

Strong, C. S. 1995. Distribution of Marbled Murre-
lets along the Oregon coast in 1992. Northwest.
Nat. 76:99-105.

Strong, C. S., B. S. Keitt, W. R. McIver, C. J. Palm-
er, AND I. Gaffney. 1995. Distribution and pop-
ulation estimates of Marbled Murrelets at sea in
Oregon during the summers of 1992 and 1993.
U.S. For. Serv. Gen. Tech. Rep., PSW-GTR-152:
339-352.

Trasky, L. L., L. B. Flagg, and D.C. Burbank. 1977.
Environmental studies of Kachemak Bay and
Lower Cook Inlet. Vol. 1. Impact of oil on the
Kachemak Bay environment. Alaska Dept, of Fish
and Game, Marine Coastal Habitat Management,
Anchorage, Alaska.

262

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Wilson Bull., 111(2), 1999, pp. 262-265

“Snorkeling” by the Chicks of the Wattled Jacana

Carlos Bosque'-^ and Emilio A. Herrera-

ABSTRACT — The chicks of the Wattled Jacana
(Jacana jacana) exhibited an unusual predator escape
behavior in the floodplains of Venezuela. When ap-
proached by a human, chicks dove and remained com-
pletely immobile while entirely submerged with just
the beak protruding vertically above the water line.
Since breathing should continue while hiding, we rea-
soned that it would be advantageous for the bird to
have the nostrils placed in a forward position along
the bill to facilitate breathing while submerged. To ex-
amine this expectation we compared the relative po-
sition of the nostrils of the Wattled Jacana with those
of species belonging to phylogenetically related Scol-
opacidae. In accordance with expectations. Wattled Ja-
canas have nostrils that are placed significantly more
forward along the bill than all species of Scolopacidae
measured. Nostril placement in species belonging to
other phylogenetically related families, Thinocoridae,
Pedionomidae, and Rostratulide is also basal as in
Scolopacidae. Lorward placement of nostrils seems to
be a derived character in Jacanas. The “snorkeling”
behavior of Wattled Jacana chicks is a behaviorally
elaborate predator escape mechanism, seemingly ac-
companied by anatomical adaptations. Received 16
June 1998, accepted 15 Dec. 1998.

Predation is often the major mortality factor
of young birds and is therefore an important
selective force shaping the behavior of young
birds and their parents (Rickleffs 1969, Martin
1992). Capabilities of precocial chicks to es-
cape predation or defend themselves from
predators are limited. Their chances of pre-
dation are largely dependent on actions taken
by their parents. When predators approach
their chicks, adults frequently perform “dis-
traction displays” that often divert the atten-
tion of the predator away from their offspring
(Skutch 1976). Safety of precocial chicks also
depends upon their small size, concealing col-
oration and immobility (Skutch 1976). In this
note we describe an unusual predator escape

' Dept. Biologia de Organismos, Univ. Simon Bo-
h'var. Aparlado 89. ()()(), Caracas 1080, Venezuela.

- Dept. Estudios Ambientales, Univ. Simdn Bohvar,
Apartado 89.000, Caracas 1080, Venezuela.

’ Corresponding author: E-mail; caiiosb@usb.ve

behavior exhibited by Wattled Jacana {Jacana
jacana) chicks and extend information on the
hiding behavior of Jacana chicks.

The Wattled Jacana inhabits freshwater
wetlands with floating and emergent vegeta-
tion throughout its range from northern South
America and Panama to central Argentina (del
Hoyo et al. 1996). Like several other Jacanas,
Wattled Jacanas show sex-role reversal and a
polyandrous mating system (Osborne 1982).
Although few details have been reported on
Wattled Jacanas (see Osborne and Bourne
1977, Osborne 1982), it is generally assumed
to have similai' behavioral traits to the North-
ern Jacana (7. spinosa; del Hoyo et al. 1996).
Chicks are highly precocious, leaving the nest
soon after hatching, but ai'e tended by the
male parent. Parental care includes brooding,
attending and defending, but not feeding the
chicks (Jenni and Collier 1972, Stephens
1984a, Betts and Jenni 1991). Females aid
males in defending offspring from potential
predators or conspecifics (Jenni and Collier
1972, Stephens 1984a). Heavy predation pres-
sure on eggs and chicks appears to be impor-
tant in both Wattled and Northern Jacanas
(Jenni 1974; Osborne and Bourne 1977; Ste-
phens 1984a, b).

Our observations were made on a savanna
flooded approximately 40 cm deep, covered
with floating and emergent vegetation (Eic-
chornia sp., Hymenachne amplexicaiilis) at
Hato El Frio, a cattle ranch and biological re-
serve in the southern Llanos (floodplains) of
Venezuela (7° 46' N, 68° 57' W). As E.H. ap-
proached a pair of Jacana adults with three
chicks, one of the adults, presumably the
male, performed a typical “broken wing” dis-
play, Jumping and apparently attempting un-
successfully to fly. As the observer continued
to approach, the parents flew away. When we
looked among the vegetation for the chicks
they were nowhere to be seen. While search-
ing, we found that what seemed to be an odd
looking slender yellowish flower was in fact

SHORT COMMUNICATIONS

263

the bill of one of the young jacanas. The chick
remained completely immobile while entirely
submerged with just the beak protruding ver-
tically above the water. By bringing a hand
from below we were able to pick up the bird,
which made no attempt to escape nor showed
any defense behavior. The chick was in the
downy stage.

This hiding behavior should reduce the
chances of detection by predators or aggres-
sive conspecifics. For it to be effective, the
chick should remain motionless underwater
for an unpredictable length of time. In order
to continue breathing while submerged it
would be advantageous to have the nostrils
placed toward the tip of the bill.

To examine this prediction, we compared
the relative position of the nostrils of the Wat-
tled Jacana with those of phylogenetically re-
lated species. Of the four other families in the
same parvorder (Sibley and Monroe 1990)
only Scolopacidae occur in Venezuela and
specimens were available in bird collections.
We selected at random 6 of the 12 genera of
Scolopacidae that occur in the country (Meyer
de Schauensee and Phelps 1978) and one spe-
cies from each of these, except for speciose
genus Calidris from which we chose two spe-
cies. In 10 individuals of each species we
measured the length of the exposed culmen
(EC) and the distance between the posterior
margin of the right nostril and the tip of the
bill (NT). Specimens, selected at random,
were measured at the Phelps Ornithological
Museum in Caracas with calipers to 0.1 mm.
For jacanas we measured separately adults, ju-
veniles, and the one downy chick in the col-
lection. The exposed culmen was measured
from the base of the frontal shield where it
rises more abruptly. From those measurements
we calculated an index (I) to describe the rel-
ative position of the nostrils along the bill:

I = (EC - NT)/EC

The value of this index should be zero in
those species in which the nostrils are placed
at the base of the bill and closer to one the
nearer the nostrils are to the bill tip. Statistical
analyses were done with SYSTAT 7.0 for
Windows (Wilkinson 1997).

Mean relative position of the nostrils dif-
fered between species (Fig. 1; Single Factor
ANOVA on the arcsine-transformed data: P

0.4

0.3

2 0.2
Q

■ f S ‘

0.0 — —

CH GG CF NP At CS BL JJa JJj JJc
SPECIES

FIG. 1. A comparison of the relative position of
the nostrils along the bill between Jacano jacana and
several species of Scolopacidae. Mean Index values (±
SD), see text for definition of the Index. Species are:
CH = Calidris himantopiis, GG = GaUinago gallin-
ago, CF = Calidris fuscicollis, NP = Numenius phaeo-
pus, AI = Arenaria interpres, CS = Catoptrophorus
semipalmatus, BL = Bartramia longicauda, JJa = Ja-
cana jacana (adults), JJj = J. jacana (juveniles), JJc
= J. jacana chick, n = 10 in all cases, except for the
single J. jacana chick. Horizontal lines above species
labels indicate similarity between the index mean of
those species (from Tukey HSD test). Jacana jacana
juveniles and the chick were not included in the test.

< 0.001, F = 143.632, df = 7); the single
jacana chick and juveniles were not included
in this comparison. A Tukey HSD a posteriori
test revealed a number of significant differ-
ences between species {P < 0.05; Fig. 1).
Wattled Jacana nostrils are placed significant-
ly more forward along the bill than all species
of Scolopacidae measured (Fig. 1 ). Nostrils
were also placed forward on juveniles and one
newly hatched downy chick.

Nostril placement in species belonging to
families phylogenetically related to the Jacan-
idae, but not available to us, can be seen in
photographs in del Hoyo and coworkers
(1996). Both species of Rostratulidae (belong-
ing in the same superfamily with Jacanidae)
have nostrils placed basally (del Hoyo et al.
1996:293-299). In the Plains-wanderer {Pe-
dionomus torquatus, Pedionomidae), the pos-
terior end of its longish naiina is clearly set

264

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

back (del Hoyo et al. 1996:535-536). The cor-
naceous flaps covering the nostrils of the
Thinocopridae {Thinocorus spp., Attagis gayi)
appear to be placed at the base of the bill also
(del Hoyo et al. 1996:539-543).

It appears that basal placement of the nos-
trils is ancestral in the group and that forward
displacement in jacanas is a derived character.
We cannot assert that breathing while sub-
merged was an important selective force in the
forward displacement of jacana nostrils, but
their current position should facilitate it.

Young of jacanas are known to avoid pred-
ators or aggressive conspecifics by taking to
water and snorkeling behavior is shared with
other species. The chicks of the Northern Ja-
cana, which is considered to form a superspe-
cies with the Wattled Jacana (del Hoyo et al.
1996), have been reported to swim (Gilliard
1967), to hide in the water (Stephens 1984a),
and to submerge with only their bills and the
tops of their heads showing above the surface
(Miller 1931). The hiding behavior of young
Pheasent-tailed Jacanas {Hydrophasianus chi-
rurgus) from India and Asia seems similar;
chicks “may freeze while hiding under a leaf
or even while completely submerged except
for the bill” (Johnsgard 1981:40). The chicks,
and sometimes even the adults, of the Lesser
Jacana {Microparra capensis) of Africa are
also known to submerge with only their bills
out of the water (Maclean 1972). Since snor-
keling seems to be accompanied by anatomi-
cal adaptations, it is likely that those species
have forward displaced nostrils also. In fact,
this seems to be the case in the Northern (del
Hoyo et al. 1996:277) and Pheasent-tailed Ja-
cana (drawing in del Hoyo et al. 1996:288).
For the Lesser Jacana it is difficult to judge
from available pictures. Other species of ja-
cana for which we do not have information on
snorkeling behavior also have nostrils placed
in forward position along the bill, most no-
tably the AWcan Jacana (Actophilornis afri-
canus-, del Hoyo et al. 1996:281). A broader
survey would be necessary to establish if
snorkeling and forward placement of the nos-
trils is shared by all species of the family.

In similar aquatic habitats in the central
Llanos of Venezuela, the young of at least
three other species swim or dive, although
adults do not: the altricial nestlings of Hoa-
tzins (Opisthocomus hoazin-, Strahl 1987) and

Greater Anis {Crotophaga major, Lau et al.,
in press) jump from their nest and swim or
dive in the water below; the precocial chicks
of Purple Gallinules {Porphirula martinica)
dive when threatened (Zaida Tarano, pers.
comm.). Interestingly, the latter and conge-
neric P. flavirostris have nostrils displaced
forward along the bill (pers. obs.).

ACKNOWLEDGMENTS

We gratefully acknowledge Dr. I. D. Maldonado for
his hospitality and permission to work at Hato El Erio.
M. Lentino, curator of the Phelps Ornithological Mu-
seum, allowed us to measure specimens and made
helpful suggestions. I. Carreno helped measure speci-
mens. Reviewers and editorial staff of The Wilson Bul-
letin improved the manuscript.

LITERATURE CITED

Betts, B. J. and D. A. Jenni. 1991. Time budgets and
the adaptaviness of polyandry in Northern Jaca-
nas. Wilson Bull. 103:578-597.
del Hoyo, J., A. Elliot, and J. Sargatal (Eds.).
1996. Handbook of the birds of the world. Vol.
3. Lynx Editions, Barcelona, Spain.

Gilliard, E. T. 1967. Living birds of the world. Dou-
bleday & Company, Inc., Garden City, New York.
Jenni, D. A. 1974. Evolution of polyandry in birds.
Am. Zool. 14:129-144.

Jenni, D. A. and G. Collier. 1972. Polyandry in the
American Jacana {Jacana spinosa). Auk 89:743-
765.

Johnsgard, P. A. 1981. The plovers, sandpipers and
snipes of the world. Univ. of Nebraska, Lincoln.
Lau, P. a., C. Bosque, and S. Strahl. In press. Nest
predation in relation to nest placement in the
Greater Ani {Crotophaga major). Ornitol. Neo-
trop. In press.

Maclean, G. 1972. Waders of waterside vegetation.

African Wildl. 26:163-167.

Martin, T. E. 1992. Interaction of nest predation and
food limitation in reproductive strategies. Curr.
Ornithol. 9:163—197.

Meyer De Schauensee, R. and W. H. Phelps, Jr.
1978. A guide to the birds of Venezuela. Prince-
ton Univ. Press, Princeton, New Jersey.

Miller, A. H. 1931. Observations on the incubation
and the care of the young in the jacana. Condor
33:32-33.

Osborne, D. R. 1982. Replacement nesting and poly-
andry in the Wattled Jacana. Wilson Bull. 94:206—
208.

Osborne, D. R. and G. R. Bourne. 1977. Breeding
behavior and food habits of the Wattled Jacana.
Condor 79:98-105.

Rickleffs, R. E. 1969. An analysis of nesting mor-
tality in birds. Smithson. Contrib. Zool. 9:1-48.

SHORT COMMUNICATIONS

265

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

Skutch, a. F. 1976. Parent birds and their young.
Univ. of Texas Press, Austin.

Stephens, M. 1984a. Interspecific aggresive behavior
of the polyandrous Northern Jacana (Jacana spi-
nosa). Auk 101:508-518.

Stephens, M. 1984b. Intraspecific distraction displays
of the polyandrous Northern Jacana, Jacana spi-
nosa. Ibis 126:70-72.

Strahl, S. D. 1987. The social organization and be-
haviour of the Hoatzin Opisihocomus hoazin in
central Venezuela. Ibis 130:483-502.

Wilkinson, L. 1997. SYSTAT 7.0: the system for sta-
tistics. SYSTAT Inc., Evanston, Illinois.

Wilson Bull., 1 1 1(2), 1999, pp. 265-268

Rapid Long-distance Colonization of Lake Gatun, Panama,

by Snail Kites

George R. Angehr'

ABSTRACT. — The distribution of the Snail Kite
(Rostrhamus sociabilis) is closely tied to that of apple
snails (Pomacea spp.), its nearly exclusive food. Be-
fore the early 1990s, the species occurred in Panama
primarily as a vagrant. Apple snails were introduced
to Lake Gatun in central Panama in the late 1980s, and
by 1994 Snail Kites had colonized the lake from pop-
ulation sources at least 350 km away and initiated
breeding. Since 1994 the population has increased rap-
idly and the species can now be found throughout the
lake. Received 7 Oct. 1998, accepted 6 Jan. 1999.

The Snail Kite {Rostrhamus sociabilis) is a
highly specialized raptor that ranges from
southern Florida and Mexico, through Central
America, to Bolivia, northern Argentina and
Uruguay (Beissinger 1988). Its distribution is
closely tied to that of apple snails {Pomacea
spp.), which form its diet almost exclusively
(Beissinger 1988). The kite uses its exception-
ally thin upper mandible to extract snails from
their shells (Snyder and Snyder 1969). Other
species of snails, turtles, crabs and other items
are taken on occasion (Beissinger 1990a, Sny-
der and Kale 1983, Sykes and Kale 1974).
The species is nomadic, moving in response
to changes in the availability of its favored
prey because of fluctuating water levels
(Sykes 1979, 1983; Beissinger and Takekawa
1983, Takekawa and Beissinger 1989).

Snail Kites are rare in southern Central

' Smithsonian Tropical Research Institute, Unit
0948, APO AA 34002-0948;

E-mail: angehrg@tivoli.si.edu

America, and there have been only seven pre-
vious reports from Panama (Ridgely and
Gwynne 1989). The sole Panama specimen,
an immature female, was collected near the
Colombian border at Perme, near Puerto Ob-
aldia, San Bias Province, in 1929 (Wetmore
1965). There were four reports of single birds,
all either females or immatures, at marshes
near Panama City, in 1971, 1973, 1977, and
1979. There were two reports from Chiriqui
province in western Panama; an adult male
and a female or immature near Gualaca in
1965, and several pairs and a nest in a marsh
near Remedies in 1973, the latter being the
only previous report of breeding activity in
Panama (Ridgely and Gwynne 1989). The
closest significant populations of Snail Kites
to Panama are in western Colombia, on the
west side of the Gulf of Uraba (Hilty and
Brown 1986), about 350 km from the Panama
Canal area, and the Tempisque Basin in north-
ern Costa Rica (Stiles and Skutch 1989), ap-
proximately 650 km away.

STUDY AREA

Lake Gatun (420 kmO is an artificial lake created
by the damming of the Chagres River to form the cen-
tral part of the Panama Canal in 1914. The lake level
is controlled by the Panama Canal Commission, and
may vary several meters between wet and dry seasons
as water is released when ships pass through the locks.
The Canal area experiences a strong four-month long
dry season from mid-December to mid-April.

The introduced aquatic plant Hydrilla verticillata is
a major problem in the lake. It apparently first became
established in the late 1920s or early 1930s, and had

266

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

become a significant problem by the 1960s (Maturell
and Salazar 1994). Rooting in shallow water, it now
occurs in huge beds in many parts of the lake.

The apple snail Pomacea lattrei was introduced to
Panama from Guatemala in the early 1980s by Pana-
ma’s National Directorate of Aquaculture. Snails were
established at a research station at Divisa, Herrera
Province, approximately 130 km west of Lake Gatun,
for use in rural aquaculture programs. Between 1986
and 1988, the owner of a sportfishing business in the
town of La Arenosa on the southwestern arm of Lake
Gatun obtained 15 snails from the Divisa station and
introduced them to a small artificial pond next to Lake
Gatun. After they had reproduced, she distributed
some to neighbors and introduced others into the lake.
A few months later it was noticed that the snails were
feeding on and reducing the abundant Hydrilla around
the community’s public dock and in surrounding areas.
Residents of other lakeside communities began delib-
erately introducing snails to their areas in an effort to
control Hydrilla (Maturell and Salazar 1994), a major
impediment to small boat travel. The snails are also
used for food by some local people.

Surveys by the Panama Canal Commission found
that the snails were distributed throughout the south-
eastern arm of the lake by 1991, were in the north-
western part by 1992, and had reached the town of
Gamboa at the east end of the lake by 1993. In some
areas Hydrilla cover was reduced by as much as 94%
in three years. Deliberate introduction by humans ev-
idently allowed the snail to disperse quickly around
the lake. The snails also spread by floating in currents
and by egg masses fixed to floating vegetation, logs,
and boats (Maturell and Salazar 1994).

A native species of apple snail, Pomacea cumingi,
occurs in the Canal area. It is a bottom-dwelling spe-
cies that typically occurs in low densities. It mostly
inhabits small streams, but is sometimes found in larg-
er rivers. The combination of habitat, bottom dwelling
habit, and low density evidently makes this species
unsuitable as prey for Snail Kites (L G. Thompson,
pers. comm.).

RESULTS

M. Santamaria (pers. comm.), a game war-
den at the Barro Colorado Nature Monument,
a reserve managed by the Smithsonian Trop-
ical Research Institute, first observed unusual
hawklike birds eating snails in Gigante Bay,
south of Barro Colorado Island, in May or
June 1994. On 3 February 1995 Santamaria
and I visited an area where he had recently
seen birds building nests in Guindilla Cove, a
narrow inlet about 1500 m long on the south
shore of Gigante Bay. We saw at least 14 Snail
Kites in the cove at that time, including at
least two adult males in black plumage. The
remaining birds were in brown plumage. Im-

matures and adult females have brown plum-
age and cannot reliably be distinguished in the
field (S. Beissinger, pers. comm.).

At four locations we observed nests con-
sisting of loose platforms of twigs at various
stages of construction. One small vine-cov-
ered tree standing in water (Site 1) had seven
nests while the other sites had one nest each.
No eggs or chicks were evident at that time,
although we could not see the contents of high
nests. Santamaria recently had seen birds
bringing twigs to add to the nests.

The area was visited again on 20 May 1994
by D. and L. Engleman. They did not see any
active nests at Site 1, but at least 17 kites were
soaring or perched in the area. They observed
five nests at another site (Site 2) about 300 m
north of Site 1 . These were located on a small
island composed of Annona glabra shrubs
overgrown with vines. Three nests were com-
plete, one with two eggs and another with at
least one egg and perhaps more. A third nest
appeared to have eggs but they could not be
counted. The remaining two nests were under
construction and were visited by birds in
brown plumage caiTying twigs. At least 20
kites were present at this site, yielding a total
of at least 37 in the cove (D. and L. Engleman,
pers. comm.).

I visited Guindilla Cove again on 15 July
1994. At that time seven nests were present at
Site 2. Two nests had three well-feathered
nestlings, while two more fledglings were
perched together on a branch near a third nest.
Three other birds which took flight from the
island had very short tails and appeared to
have recently fledged. On 27 August 1994 two
nests at Site 1 each had a single lai'ge downy
young. No active nests were present at Site 2,
but I saw two birds that apparently had re-
cently fledged. On 14 October no activity was
seen at Site 1, but three apparently recently
fledged birds were present at Site 2. On this
date I also surveyed several other ai’eas in Gi-
gante Bay that I had not previously investi-
gated. One nest with two small nestlings was
found near an island at the mouth of Guindilla
Cove, and a second nest with two fledglings
on the verge of flying was found on a large
Annona glabra island about 3 km west of
Guindilla. Approximately 15 adults and three
apparently recently fledged young were also
seen at this site.

SHORT COMMUNICATIONS

267

DISCUSSION

Apple snails probably reached the area of
Gigante Bay in 1991 (Matured and Salazar
1994). I am certain no kites were present in
the area before 1992, because I made three
surveys of the shoreline of parts of the bay by
canoe in 1991 and 1992, and surveyed Guin-
dilla Cove itself on 23 August 1992. Guindilla
Cove probably was colonized by kites in ei-
ther 1993 or 1994.

In 1995, in the central part of the lake neai'
the Panama Canal channel. Snail Kites were
restricted to Gigante Bay, based on informa-
tion from Smithsonian game wardens and re-
searchers who worked on the lake. Kites were
rarely if ever seen at the Smithsonian research
station on the north side of Barro Colorado
Island 4 km away.

Since 1995, kites have been seen more of-
ten in other parts of Lake Gatun and the Canal
area in general, ranging from the northern end
of the lake near the Gatun Locks to Miraflores
Locks near the Pacific entrance to the Panama
Canal. Although no comprehensive surveys of
kite distribution and numbers have been made.
Snail Kites appear to have spread throughout
the lake and occur wherever suitable habitat
is present.

In 1995 nest construction evidently began
in January, during the early dry season. A few
small nestlings were still present in mid-Oc-
tober, during the late rainy season. These latter
birds would probably have fledged in Novem-
ber, so that in 1995 the breeding season was
at least 10 months long. In Florida, the main
nesting season is January-August, peaking in
February-June, although in yeais of high wa-
ter breeding may begin in December and ex-
tend to September (Beissinger 1986, 1988;
Snyder et al. 1989). Nesting seasonality is
poorly known elsewhere in the tropics, but
may be tied to the rainy season. Nesting takes
place during the wettest period in Surinam,
Argentina and Venezuela (Beissinger 1988).

At Gigante, clutch or brood size ranged
from one to three. In Florida, clutch size pres-
ently ranges from one to three with the latter
much more frequent. However, clutch size ap-
pears to have declined historically in Florida,
with four-egg clutches previously having been
common (Beissinger 1986). The decline may
be attributable to habitat deterioration. Al-

though there are few data, clutch size else-
where in the kite’s range seems typically to be
two to three, although four-egg clutches often
occur in Argentina (Beissinger 1990b) and
clutch size may range up to six (Beissinger
1988). Despite a presumably superabundant
food supply at Gigante in 1995, there was no
evidence that kites increased their clutch size
in response.

The Snail Kite colony in Guindilla cove in-
creased from no birds in 1992 to at least 37
in 1995, an extraordinarily high rate of in-
crease if only a single founding pair had been
involved. Because Kites aie highly sociable,
initial colonization could have been by a small
flock of birds rather than a single pair. It is
possible that Guindilla Cove was not the ini-
tial site of colonization of Lake Gatun. Snails
were present in the southwestern arm of the
lake by the late 1980s. Kites could have col-
onized that area first, and then spread to Guin-
dilla. The southwestern arm of the lake is
large and remote from population centers, so
a colony there could easily have gone unde-
tected.

Nevertheless, Snail Kite populations in-
crease rapidly under favorable conditions
(Snyder et al. 1989, Beissinger 1995), such as
the essentially unlimited food supply Lake
Gatun would have offered the first colonists.
The age at first reproduction in Snail Kites is
very low, with some females nesting at 10
months. In favorable years in Florida, the
breeding season may last up to 10 months, and
some kites may re-nest and raise second
broods (Beissinger 1986, Snyder et al. 1989).
Parents of either sex may desert a nest leaving
the other member of the pair to continue to
raise the young alone (Beissinger and Snyder
1987). The remaining parent almost always is
able to rear the young to independence by it-
self, while the deserting parent has the oppor-
tunity to re-nest with another paitner.

The rapid colonization by Snail Kites of
Lake Gatun and subsequent population expan-
sion provides an interesting example of the
dispersal capabilities of this highly opportu-
nistic species. The most likely source of col-
onizing birds would have been the Colombian
population, approximately 350 km from Lake
Gatun. This population is much closer than
the one in Costa Rica, and there ai'e several
large areas of freshwater habitat, including

268

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

Lake Bayano and several large rivers, in the
intervening area. Dispersers discovered the
newly suitable habitat of Lake Gatun within
only a few years after the introduction of their
preferred food, and rapidly spread throughout
the available area.

ACKNOWLEDGMENTS

I would particularly like to thank M. Santamaria,
who first told me of the kites, D. and L. Engleman,
who provided additional information, and D. George,
who organized two of our visits to the colony. K.
Aparicio, S. Lollett, K. Kaufmann, and J. Nason also
assisted with field observations. I thank L G. Thomp-
son for identifying snail specimens and providing ad-
ditional information on Pomacea biology. S. R. Beis-
singer and R W. Sykes Jr. provided helpful reviews that
improved the manuscript. I would also like to thank
ANAM, Panama’s environmental authority, for grant-
ing the permits under which this research was con-
ducted.

LITERATURE CITED

Beissinger, S. R. 1986. Demography, environmental
uncertainty, and the evolution of mate desertion
in the Snail Kite. Ecology 67:1445-1459.
Beissinger, S. R. 1988. Snail Kite. Pp. 148-165 in
Handbook of North American birds. Vol. 4 (R. S.
Palmer, Ed.). Yale Univ. Press, New Haven, Con-
necticut.

Beissinger, S. R. 1990a. Alternative foods of a diet
specialist, the Snail Kite. Auk 107:327—333.
Beissinger, S. R. 1990b. Experimental brood manip-
ulations and the monoparental threshold in Snail
Kites. Am. Nat. 136:20-38.

Beissinger, S. R. 1995. Modeling extinction in peri-
odic environments: everglades water levels and
Snail Kite population viability. Ecol. Applications
5:618-631.

Beissinger, S. R. and N. E R. Snyder. 1987. Mate
desertion in the Snail Kite. Anim. Behav. 35:477-
487.

Beissinger, S. R. and J. E. Takekawa. 1983. Habitat
use and dispersal by Snail Kites in Florida during
drought conditions. Fla. Field Nat. 11:89-106.

Haverschmidt, F. 1970. Notes on the Snail Kite in
Surinam. Auk 87:580-584.

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

Maturell, j. C. and a. J. Salazar. 1994. Aspectos
de la introduccion y diseminacion del caracol gi-
gante Pomacea sp. en el Lago Gatun y sus efectos
sobre la abundancia de Hydrilla verticillata. Pan-
ama Canal Commission, Bureau of Engineering
and Construction, Balboa, Panama.

Ridgely, R. and j. Gwynne. 1989. A guide to the
birds of Panama. Princeton Univ. Press, Princeton,
New Jersey.

Snyder, N. E and H. A. Snyder. 1969. A compara-
tive study of mollusc predation by Limpkins, Ev-
erglade Kites, and Boat-tailed Grackles. Living
Bird 8:177-233.

Snyder, N. F. and H. W. Kale, II. 1983. Mollusk
predation by Snail Kites in Colombia. Auk 100:
93-97.

Snyder, N. E, S. R. Beissinger, and R. F. Chandler.
1989. Reproduction and demography of the Flor-
ida Everglade (Snail) Kite. Condor 91:300—316.

Stiles, F. G. and A. F. Skutch. 1989. A guide to the
birds of Costa Rica. Cornell Univ. Press, Ithaca,
New York.

Sykes, P. W., Jr. 1979. Status of the Everglade Kite
in Florida 1968-1978. Wilson Bull. 91:495-511.

Sykes, P. W., Jr. 1983. Recent population trends of
the Everglade Kite in Florida and its relationship
to water levels. J. Field Ornithol. 54:237-246.

Sykes, P. W., Jr. and H. W. Kale, II. 1974. Ever-
glades Kites feed on non-snail prey. Auk 91:818-
820.

Takekawa, J. E. and S. R. Beissinger. 1989. Cyclic
drought, dispersal, and the conservation of the
Snail Kite in Florida: lessons in critical habitat.
Conserv. Biol. 3:302—311.

Wetmore, a. 1965. The birds of the Republic of Pan-
ama. Vol. I. Tinamidae to Rynchopidae. Smith-
sonian Institution Press, Washington, D.C.

SHORT COMMUNICATIONS

269

Wilson Bull., 1 1 1(2), 1999, pp. 269-271

The “Significant Others” of American Kestrels:
Cohabitation with Arthropods

Jeffrey P Neubig' - and John A. Smallwood'’^

ABSTRACT. — We examined the arthropod fauna
that coexists in nest boxes with American Kestrel
chicks (Folco sparverius) in northwestern New Jersey.
Of the seven arthropod species present, five were scav-
enging beetles, including carrion beetles (Silpha in-
aeqiialis), hister beetles {Atholus arnericanus and Phel-
isier subrotundus), dermestid beetles (Dermestes ca-
ninus), and skin beetles (Trox foveicollis), which ap-
parently were attracted to prey remains that
accumulated in the nest boxes. Arthropod density and
species richness were significantly greater for nest box-
es in which kestrels bred than for unoccupied nest box-
es. Received 30 June 1998, accepted 16 Nov. 1998.

Studies of the association between Ameri-
can Kestrels {Falco sparverius) and arthro-
pods have focused primarily on ( 1 ) the occur-
rence of these invertebrates in kestrel diets
and the predatory behavior kestrels direct to-
ward them and (2) parasitic arthropods re-
ported to infest kestrels (for reviews see Sher-
rod 1978 and Smallwood and Bird in press,
respectively). However, arthropods interact
with kestrels in another functional role, that of
symbiotic scavengers. Kestrels are a cavity
nesting species; typical brood size is four or
five, and the nestling period lasts about 30
days (Johnsgard 1990). During this time prey
remains, regurgitated pellets, and other organ-
ic material accumulates in the nest cavity
(Balgooyen 1976; Smallwood, pers. obs.).
Balgooyen (1976) observed dermestid beetles
{Dermestes spp.) in each of approximately 40
kestrel nests in northern California and com-
mented on their role in nest sanitation. The
objectives of the present study were to ex-
amine the arthropod community that coexists
with kestrels breeding in nest boxes in New

' Dept, of Biology, Montclair State Univ., Upper
Montclair, NJ 07043.

- Current address: Dept, of Zoology, The Ohio State
Univ., Columbus, OH 43210.

^ Corresponding author;

E-mail: smallwood@saturn.montclair.edu

Jersey and to compare them to the invertebrate
fauna in nest boxes not occupied by kestrels.

STUDY AREA AND METHODS

The study area was located in rural northwestern
New Jersey, bordered to the north and west by the
Kittatinny Ridge and Delaware River, and to the east
and south by residential and commercial development.
This area is characterized by mixed agriculture, in-
cluding corn, hay, and cattle production, and forestland
in the ridge and valley physiographic region (Sauer et
al. 1997). Sixty wooden nest boxes (internal dimen-
sions: 20 X 23 cm floor, ca 34 cm in height) were
erected in open habitats in Sussex County (centered
approximately 41° 11' N, 74° 38' W) between 1 April
1995 and 25 April 1997, and 69 nest boxes in Warren
County (approximately 40° 47' N, 75° 04' W) between
5 August 1995 and 26 March 1997. Because kestrels
do not bring any nesting material into the nest cavity
(Bird and Palmer 1988), we covered the floor of each
nest box with approximately 6 cm of wood shavings
to provide a cushion and insulation for the eggs.

Each nest box was monitored at 4-week intervals
from 23 March through 13 July 1997 to determine oc-
cupancy status; a nest box was considered occupied by
breeding kestrels if at least one egg was observed.
Kestrels require 7-9 days to produce and 28-30 days
to incubate a clutch (Bird and Palmer 1988); thus, all
kestrel nesting attempts were discovered during the
laying or incubation stage. Nest boxes in which kes-
trels bred (herein “kestrel nest boxes”) were visited
3-7 times throughout the nesting attempt. During the
final visit chicks were banded and the bedding (in-
cluding cohabiting arthropods, prey remains, pellets,
and any other material) was collected and replaced
with fresh wood shavings. Pinal visits occurred be-
tween 21 June and 1 August 1997 when nestlings were
16-23 days old (75% were 20-22 days old).

In addition to kestrels, other vertebrates that bred in
the nest boxes included five avian species (Great Crest-
ed Flycatcher, Myiorchus crinitus; Tree Swallow,
Tachycineta hicolor. Eastern Bluebird, Siolia sialis\
European Starling, Sturnus vulgaris; and House Spar-
row, Baser domesticus) and four species of mammals
(eastern gray squirrel, Sciurus carolinensis; red squir-
rel, Tamiasciurus hudsonicus; southern flying squirrel,
Glaucomys volans; and white-footed mouse, Peromys-
cus leucopu.'i). Only four nest boxes remained unoc-
cupied by any vertebrate species, including kestrels,
during the breeding season (herein “unoccupied nest

270

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 1. Taxonomy and abundance of arthropods
in northwestern New Jersey.

co-inhabiting nest boxes with American Kestrel broods

Nest boxes

Individuals per occupied^ nest box

Taxonomy

occupied

{%)

Mean

Range

Arachnida

Araneida

Aglenidae

Tegenaria derhamii, funnel spider

6.3

1.0

0.0

1-1

Insecta

Coleoptera

Silphidae

Silpha inaequalis, carrion beetle

68.8

41.0

30.8

3-114

Histeridae

Atholus americanus, hister beetle

31.3

4.8

3.4

2-9

Phelister suhrotimdiis, hister beetle

37.5

8.8

11.1

1-24

Dermestidae

Denneste.s caninus, dermestid beetle

12.5

5.0

5.7

1-9

Trogidae

Trox foveicollis, skin beetle

6.3

3.0

0.0

3-3

Hymenoptera

Formicidae

Camponotus pennsylvanicus, black carpenter ant

6.3

1.0

0.0

1-1

“ Nest boxes in which a particular arthropod species was present.

boxes”). The bedding was collected from unoccupied
nest boxes on 11 and 13 July 1997.

All bedding samples were stored at — 22°C in air-
tight plastic bags to preserve any arthropod specimens
present. We subsequently extracted all arthropods vis-
ible at 1.75X magnification, preserved them in 70%
ethanol, and identified them with information from
Comstock and Gertsch (1948), Emerton and Lrost
(1961), Borrer and White (1970), Headstrom (1977),
Kaston (1978), Milne and Milne (1980), and Arnett
(1985).

Data on arthropod density and richness were tested
for normality. Because significant deviations were de-
tected, we used nonparametric statistical treatments ex-
clusively, including the Wilcoxon rank sum test,
Spearman’s correlation coefficient, and Eisher’s exact
test (Snedacor and Cochran 1980). Analyses were per-
formed using SAS 6.12 on a Sun Solaris 2.6 platform.

RESULTS

A total of 567 individual arthropods were
extracted from the bedding collected from 16
kestrel (Table 1) and four unoccupied nest
boxes. Arthropod density ranged from 0 (for
two kestrel and two unoccupied nest boxes)
to 115 individuals in one kestrel nest box.
Mean arthropod density for kestrel nest boxes
(35.3 ± 31.0 SD) was significantly greater
than that for unoccupied nest boxes (0.50 ±
0.58 SD; two-tailed Wilcoxon rank sums test:
Z = 2.421, P = 0.016). Maximum arthropod
density in unoccupied nest boxes was 1: a

crab spider (Thomsidae, Xysticus triguttus) in
one nest box and a European earwig (Forfi-
eulidae, Forficula aiiriciilaria) in one other.

Species richness per nest box ranged from
0-3 and mean richness for kestrel nest boxes
(1.69 ± 0.87 SD) was significantly greater
than that for unoccupied nest boxes (0.50 ±
0.58 SD; two-tailed Wilcoxon rank sum test:
Z = 2.258, P = 0.024). No significant corre-
lation was detected between arthropod species
richness and kestrel brood size (r^ = 0.17, P
> 0.05, n = 16), or between kestrel brood size
and arthropod density (r^ = -0.14, P > 0.05,
n = 16).

DISCUSSION

Of the seven arthropod species observed in
kestrel nest boxes in New Jersey, five (all the
beetles) are considered scavengers (Head-
strom 1977), and the caipenter ant is known
to forage primarily on insects (Palmer 1975).
The dermestid beetle (D. caninus), found in
two of the nest boxes, was the same species
commonly used in museums for cleaning flesh
from skeletal specimens (Headstrom 1977).

The mean and maximum densities of ar-
thropods in the kestrel nest boxes in New Jer-
sey were considerably less than those ob-
served in northcentral Florida (Smallwood,

SHORT COMMUNICATIONS

271

pers. obs.). In Florida, dermestid beetles and
other species typically occuned in large num-
bers, such that by the end of the nestling pe-
riod the substrate upon which the chicks stood
visibly pulsated as the result of the motion of
these arthropods. In an experiment to deter-
mine if high arthropod densities in nest cavi-
ties discourages second clutches, Smallwood
(unpubl. data) replaced the bedding for a ran-
domly selected group of nest boxes after kes-
trel chicks had fledged. Kestrels renested in
11.1% of the cleaned nest boxes (n = 27) and
14.0% of the control nest boxes {n = 50);
these percentages were not significantly dif-
ferent (one-tailed Fisher’s exact test, P =
0.76).

We saw no evidence of kestrel chicks bitten
or otherwise harmed by the arthropods living
in the New Jersey nest boxes. Rather, the re-
moval of much of the decaying organic ma-
terial (i.e., uneaten scraps of insects, small
birds, and rodents) may benefit kestrels by re-
ducing the risk of disease or infestation with
parasites. The significant difference in the
number and species of arthropods present be-
tween kestrel and unoccupied nest boxes, and
the fact that nearly all the arthropods were
scavengers, suggest that breeding kestrels are
producing the conditions that attract these in-
vertebrates.

ACKNOWLEDGMENTS

We thank GPU Energy and Sussex Rural Electric
Cooperative for allowing us to erect nest boxes on
their utility poles; A. Bonner, J. Bullis, D. Darlington,
C. Goodwin, S. Kamphausen, S. Kindler, A. Kinney,
J. Mershon, C. Natale, J. Trinca, P. Wargo, N. War-
hoftig, R. Wyman, and K. Zientko for monitoring nest
boxes. Release time for JAS was provided by the Fac-
ulty Scholarship Incentive Program, Montclair State
University.

LITERATURE CITED

Arnett, R. H., Jr. 1985. American insects. A hand-
book of the insects of America north of Mexico.
Van Nostrand Reinhold Co., New York.

Balgooyen, T. G. 1976. Behavior and ecology of the
American Kestrel {Falco sparveriits L.) in the Si-
erra Nevada of California. Univ. Calif. Publ. Zool.
103:1-83.

Bird, D. M. and R. S. Palmer. 1988. American Kes-
trel. Pp. 253-290 in Handbook of North American
birds. Vol. 5. Diurnal raptors. Part 2 (R. S. Palmer,
Ed.). Yale Univ. Press, New Haven, Connecticut.

Borrer, D. j. and R. E. White. 1970. A field guide
to the insects of America north of Mexico. Hough-
ton Mifflin Co., Boston, Massachusetts.

Comstock, J. H. and W. J. Gert.sch. 1948. The spider
book. Cornell Univ. Press, Ithaca, New York.

Emerton, j. H. and S. W. Frost. 1961. The common
spiders of the United States. Dover Publications,
Inc., New York.

Headstrom, R. 1977. The beetles of America. A. S.
Barnes and Co., Cranbury, New Jersey.

Johnsgard, P. A. 1990. Hawks, eagles, and falcons of
North America: biology and natural history.
Smithsonian Institution Press, Washington, D.C.

Kaston, B. j. 1978. How to know the spiders, third
ed. W. C. Brown and Co., Dubuque, Iowa.

Milne, L. and M. Milne. 1980. The Audubon Society
field guide to North American insects and spiders.
Alfred Knopf, New York.

Palmer, E. L. 1975. Fieldbook of natural history, sec-
ond ed. McGraw-Hill Book Co., New York.

Sauer, J. R., J. E. Hines, G. Gough, I. Thomas, and
B. G. Peterjohn. 1997. The breeding bird survey
results and analysis, version 96.3. Patuxent Wild-
life Research Center, Laurel, Maryland. URL =
http://www.mbr.nbs.gov/bbs/bbs.html.

Sherrod, S. K. 1978. Diets of North American Fal-
coniformes. Raptor Res. 12:49—121.

Smallwood, J. A. and D. M. Bird. In press. Ameri-
can Kestrel (Falco sparverius). In The birds of
North America (A. Poole and E Gill, Eds.). Acad-
emy of Natural Sciences, Philadelphia, Pennsyl-
vania; American Ornithologists' Union, Washing-
ton, D.C.

Snedacor, G. W. and W. G. Cochran. 1980. Statis-
tical methods. Seventh ed. Iowa State Univ. Press,
Ames.

Wilson, E. O. and W. H. Bossert. 1971. A primer
of population biology. Sinauer Associates. Inc.,
Sunderland, Massachusetts.

272

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Wilson Bull., 111(2), 1999, pp. 272-273

Barred Owl Nest in Attic of Shed

C. Stuart Houston'

ABSTRACT — Barred Owls (Strix varia) nested in
the attic of a shed during 12 out of 13 years at Llotten
Lake, Saskatchewan. Twelve nestlings were banded in
five seasons. Although two species of Strix are known
to nest in buildings in Europe, this appears to be the
first such instance in North America. Received 7 Oct.
1998, accepted 7 Jan. 1999.

For twelve out of thirteen years (1980—
1992), Barred Owls (Strix varia) nested in the
same attic of a shed along Flotten Lake, Sas-
katchewan. Prior to this. Barred Owls had
never been reported to use a building for nest-
ing purposes. Furthermore, this example rep-
resents the longest continuous use of one nest
site by this species, reported to date in Sas-
katchewan.

Barred Owls apparently moved into central
Saskatchewan from adjacent Manitoba in the
1950s. They occupied a relatively narrow
band of mixed forest, most often near lakes
and rivers. The first Saskatchewan nest was
located in 1961 (Houston 1959, 1961). In the
only Barred Owl nesting study conducted in
Saskatchewan, Mazur and coworkers (1997)
found nests only in mature forest: 6 in broken
tree snags, 4 in broken limbs, 2 on squirrel
platforms, 2 on stick nests, and 1 on a witch’s
broom platform. Of these nests, 10 were in
deciduous trees and 5 were in conifers.

On 15 May 1988, I first visited an aban-
doned shed (Fig. 1), built in 1946 in a 1.6-ha
man-made clearing located in mixed old-
growth forest near the shores of Flotten Lake,
Saskatchewan (54° 30' N, 108° 30' W). Dur-
ing the visit I banded three half-grown Barred
Owl nestlings. The property owners, Mr. and
Mrs. D. Mazuren, had observed Barred Owls
nesting on the attic floor in this building ever-y
year except one since 1980. Consequently,
1988 was the eighth year of use in nine years.
During other limes of year, the owls were seen

' 863 University Drive, Saskatoon, SK S7N ()J8,
Canada; E-mail: houstons@duke.u.sask.ca

in nearby trees but never perched on or in the
buildings. On 3 June 1989, we again banded
three young and took photographs. On 1 1
June 1990, we banded the single nestling
raised that year; on 28 May 1991, three nest-
lings; and on 30 May 1992, two nestlings
(only 2 eggs had been present on 19 April).
The shed attic has not been used since.

The 12 years of use is similar to the 10
consecutive years that Baned Owls used a
deep cavity in a dead oak in southeastern
Massachusetts (Bent 1938). I have found no
other records of Barred Owls using buildings
for nest sites, although they use artificial nest
boxes in Wisconsin and Minnesota (Johnson
and Follen 1984). In Europe, other species of

FIG. 1. Author banding young Barred Owls in the
attic of a shed. Photograph by D. G. Miller.

SHORT COMMUNICATIONS

273

Strix use buildings, especially bai'ns, as nest
sites. The Tawny Owl {Strix aluco) uses build-
ings as nest sites 15% of the time, similarly,
the Ural Owl {Strix iiralensis) nests in build-
ings 2-4% of the time (Mikkola 1983).

ACKNOWLEDGMENTS

I thank M. Belcher, J. S. Marks, J. F. Roy, and four
anonymous reviewers for constructive criticism.

LITERATURE CITED

Bent, A. C. 1938. Life histories of North American
birds of prey. U.S. Nat. Mus. Bull. 170:182—201.

Houston, C. S. 1959. First records of the Barred Owl
in Saskatchewan. Blue Jay 17:94.

Hou.ston, C. S. 1961. First Saskatchewan nest of
Barred Owl. Blue Jay 19:1 14-1 15.

Johnson, D. H. and D. G. Folehn. 1984. Barred owls
and nest boxes. Raptor Res. 18:34-35.

Ma/.ur, K. M., P. C. James, and S. D. Frith. 1997.
Barred Owl (Strix varia) nest site characteristics
in the boreal forest of Saskatchewan, Canada. U.S.
Forest Service General Technical Report NC-190:
267-27 1 .

Mikkola, H. 1983. Owls of Europe. Buteo Books,
Vermillion, South Dakota.

Wilson Bull., 111(2), 1999, pp. 273-216

Double Broo(iing in the Long-eareii Owl

Jeffrey S. Marks'-^ and Alison E. H. Perkins'

ABSTRACT. — Owls in the family Strigidae typi-
cally raise no more than one brood per year. We doc-
umented what apparently is the first unequivocal case
of double brooding in Long-eared Owls (Asia otiis). A
banded female raised 12 young in two nesting attempts
compared with a mean of 5.3 young for three single-
brooded females that nested in the same grove. Two
factors may have influenced the occurrence of double
brooding: the first nest was initiated unusually early in
the year (mid-February) and food availability (in the
form of voles) was high. The rare description of double
brooding in Long-eared Owls may be due to the dif-
ficulty of detecting it. Alternatively, double brooding
may be uncommon because it is seldom an economi-
cally viable strategy. Factors that would select against
double brooding include low probability of recruitment
of the first-brood young, and reduced survival and fe-
cundity of the adults. Received 17 Sept. 1998, accepted
29 Dec. 1998.

The number of young raised per year is an
important component of an individual’s life-
time reproductive success. Double brooding,
in which a second brood is attempted after a
successful first attempt, is a viable strategy if
the increase in fitness that results outweighs
the cost of any reduction in future survival or

' Montana Cooperative Wildlife Research Unit,
Univ. of Montana, Missoula, MT 59812.

'Corresponding author;

E-mail: jmarks(3> selway.umt.edu

fecundity of the adults and their young from
the first brood. The occunence of double
brooding may be influenced by factors such
as length of the breeding season, food avail-
ability, growth rates of the young, and the du-
ration and quality of parental care (e.g., Drent
and Daan 1980, Askenmo and Unger 1986,
Tinbergen and van Balen 1988).

Double brooding is relatively rare in rap-
tors, presumably because the length of the
breeding cycle and extended postfledging care
preclude its occunence (Newton 1979, Mor-
rison 1998). Among nocturnal raptors, double
brooding occurs regularly in Bani Owls {Tyto
alba-, Marti 1992, 1997) and occasionally in
Florida Bunowing Owls {Athene cimicularia
floridana-, Millsap and Bear 1990) and Boreal
Owls {Aegoliiis funereiis-, Kellomaki et al.
1977, Solheim 1983). During a study of
breeding Long-eared Owls {Asia otiis), we
documented a female that raised two broods
during the same nesting season. Here, we de-
scribe the event and discuss factors that may
have influenced its occuirence.

STUDY AREA AND METHODS

The study area is a small grove (ca 2 ha) of quaking
aspens (Populiis trenudoides) and black hawthorns
(Crataegus douglasii) located about 16 km west of
Poison. Lake County, Montana (47° 40' N, 114° 20'
W). The elevation at the site is 888 m, and the nesting

274

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

grove is surrounded by grasslands and agricultural
fields (mostly hay). The 1 1 Long-eared Owl nests that
occurred in the grove in 1997 and 1998 were in old
nests of Black-billed Magpies (Pica pica) and Amer-
ican Crows (Conns brachyrhynchos).

Adults were captured at night in mist nests placed
near the nest or at dusk with the aid of a plastic decoy
of a Great Horned Owl (Bubo virginianus). Captured
adults were classified as after hatching year (AHY) or
after second year (ASY) based on the absence or pres-
ence of two generations of secondaries, respectively
(see Pyle 1997). During the breeding season, the sex
of most adults can be determined in the field by dif-
ferences in plumage coloration and in the hand by
presence or absence of an incubation patch (Marks et
al. 1994).

RESULTS

On 16 February 1998, we observed a fe-
male Long-eared Owl incubating at a nest
(PSNII) about 25 m north of a nest that had
produced young the previous year. The male
was roosting nearby, but we could not deter-
mine whether he was banded. Three of the
five Long-eared Owl nests that occuired in the
grove in 1998 were initiated in February;
PSNII appeared to be the earliest of these
nests. On 2 April, JSM captured the adult
(AHY) female (band no. 951) by hand at
PSNII as she was brooding seven young that
ranged in age from about 1 to 3 weeks old.
Based on the estimated age of the chicks and
an incubation period of 28 days (Marks et al.
1994), female 951 would have initiated egg
laying on 12 February, and the oldest chick
would have hatched on 12 March. Other du-
ties prevented us from catching the mate of
female 951 during this nesting attempt.

During a visit to the nesting grove on 25
June, we found a new Long-eared Owl nest at
the northern edge of the grove 28 m from nest
PSNII. A female was brooding small chicks
that appeared to be about 2 weeks old, and a
male was flushed from the same roost site typ-
ically used by the PSNII male earlier that
spring. We returned to the nest on the evening
of 30 June and captured the female in a mist
net placed directly in front of the nest. She
proved to be 951, the same female that had
fledged seven chicks earlier in the spring. The
next morning we banded her five chicks,
which were about 2 to 3 weeks old. Female
951 was very aggressive as we handled her
chicks, diving and perching within 1 m of us
and enabling us to observe her band and to

note her pattern of flight-feather molt that we
had confirmed in the hand the previous night
(4 primaries and 2 secondaries growing on
each wing). Also at this time, we noticed that
her mate was banded. After several attempts,
we succeeded in capturing the male on the
evening of 13 July, at which time the oldest
chicks were capable of short flights from tree
to tree. The male proved to be no. 914, the
same male that had nested at this site in 1997.
Female 95 1 was still present, and both adults
presumably were provisioning their fledglings.
Interestingly, male 914 had not staited flight-
feather molt.

Female 951 fledged 12 young (defining
“fledging” as capable of sustained flight;
Marks 1986) in two nesting attempts com-
pared with a mean of 5.3 young (range 5 to
6) produced by the other three pairs that nest-
ed in the grove in 1998. The estimated time
between the initiation of 95 1 ’s two nesting at-
tempts was 90 days (i.e., 12 February and 12
May). At the time 951 initiated her second
clutch, the oldest offspring from her first
brood would have been about 6 weeks old.

DISCUSSION

Several records of double brooding by
Long-eared Owls have been reported in Eu-
rope (Reinsch and Warncke 1968, Rinne 1981,
Scott 1997), but in each case the evidence was
circumstantial. To our knowledge, ours is the
first report of double brooding in Long-eared
Owls based on a banded individual.

We suspect that weather and food avail-
ability played a major role in this case of dou-
ble brooding. The winter of 1998 was unusu-
ally mild in western Montana. The ground at
the study area was virtually free of snow from
January onward (pers. obs.), and the mean
temperature in February was 2.1° C above
normal at the Keir Dam weather station 1 3 km
from the study area (data obtained from the
National Climatic Data Center). In addition,
voles (Microtiis spp.) were abundant in winter
and spring; we saw many during the day, and
other vole-eating raptors [i.e.. Northern Har-
rier (Circus cyaneus). Rough-legged Hawk
(Buteo lagopus), and Short-eared Owl (Asia
flammeiis)] were numerous in the study area.
The mild weather and abundant food probably
induced Long-eared Owls to nest in Febmary,
which is very early for this species (see Marks

SHORT COMMUNICATIONS

275

et al. 1994). The continued high numbers of
voles in summer provided an opportunity for
double brooding, at least for one of the three
pairs that began nesting in February.

In general, the incidence of second nesting
attempts in facultatively double-brooded spe-
cies is negatively conelated with the laying
date of the first clutch (e.g.. Smith et al. 1987,
Geupel and DeSante 1990, Monison 1998).
Our case agrees with this finding, but it raises
the question of why the other two Long-eaied
Owl pairs that nested early did not raise a sec-
ond brood. One possibility is that the pheno-
typic quality of the double-brooded pair was
high relative to the other pairs (see Verboven
and Verhulst 1996). Although we have no ob-
jective measure of phenotypic quality in the
Long-eared Owls we studied, we note that the
male of the second nesting attempt had bred
successfully at the site in the previous year
(the other males did not breed there in 1997),
and the female that nested twice was in good
physical condition and was unusually aggres-
sive. Indeed, during her first attempt she at-
tacked JSM when he entered the nest. More-
over, the ratio of her body mass (g) to wing
length (mm) at first capture (1.33) was higher
than that of all but one of the other eight fe-
males captured late in the brooding-rearing
period in 1997 and 1998 (Jc = 1.11 ± 0.13
SD, range 0.98-1.34). The local experience of
the male and the physical condition and ag-
gressiveness of the female are consistent with
the notion that they were high-quality individ-
uals relative to the other early nesters in the
grove.

Double brooding in Long-eared Owls may
be more common than previously thought. Al-
ternatively, it may indeed be rare because it is
seldom an economically viable strategy. For
instance, the fitness gain from double brood-
ing would be marginal if the probability of
recruitment of first-brood young is low (i.e.,
because of reduced care from parents that di-
rect their efforts to a new brood), or if future
survival and fecundity of the adults are re-
duced. Female Long-eared Owls in Idaho de-
serted their broods when the young were 6.5
to 8 weeks old, and males continued to care
for the young until they were 8.5 to 1 1 weeks
old (Ulmschneider 1990). If parental care of
this duration is typical in Long-eared Owls,
then the first of the two broods would have

received a normal amount of parental care (the
second clutch was started when the oldest
chicks from the first nest were 6 weeks old)
only if the female changed mates between
nesting attempts (and the male continued to
care for the young), or if one or both parents
continued to provision the first brood while
starting the second (an unlikely occurrence
given that the male must provide food to the
incubating female). Moreover, the timing of
second broods could interfere with the molt
schedule of adults. Long-eared Owls generally
begin molting in early June soon after breed-
ing (Marks et al. 1994). The male attending
the second brood had not started flight-feather
molt in mid-July, suggesting that his molt was
delayed because of the late breeding effort.
Delayed molt potentially could influence sur-
vivorship and fecundity (see Pietiainen et al.
1984, Kjellen 1994).

In conclusion, double brooding appears to
be rare in Long-eared Owls, and it probably
occurs only when first nests are initiated early
and food availability is high. Nothing is
known about how double brooding affects re-
cruitment of young from first versus second
broods, or whether it affects the survivorship
and future fecundity of the pai'ents. Whether
double brooding is a viable strategy in Long-
eared Owls remains to be determined from ad-
ditional research.

ACKNOWLEDGMENTS

We thank R and D. Smith for granting access to their
land; C. Olson. T Dial, and R. Petty for help in catch-
ing owls; C. Marti for comments on the manuscript;
and H. Wright for translating two papers. This research
was supported by a grant from the University of Mon-
tana Research and Creativity Committee.

LITERATURE CITED

Askknmo, C. AND U. Ungkr. 1986. How to be double-
brooded: trends and liming of breeding perfor-
mance in the Rock Pipit. Ornis Scand. 17:237-
244.

Drknt, R. H. and S. Daan. 1980. The prudent parent:
energetic adjustments in avian breeding. Ardea
68:223-252.

Gkupkl, G. R. and D. F. DhSantk. 1990. Incidence
and determinants of double brooding in Wrentits.
Condor 92:67-75.

Khllomaki, E., E. Heinonen, and H. Tiainen. 1977.
Two successful nestings of Tengmalm’s Owl in
one summer. Ornis Fenn. 54:134-135.

276

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

Kjkllen, N. 1994. Moult in relation to migration in
birds: a review. Ornis Svecica 4:1-24.

Marks, J. S. 1986. Nest site characteristics and re-
productive success of Long-eared Owls in south-
western Idaho. Wilson Bull. 98:547-560.

Marks, J. S., D. L. Evans, and D. W. Holt. 1994.
Long-eared Owl {Asia otiis). In The birds of North
America. No. 133 (A. Poole and L. Gill, Eds.).
Academy of Natural Sciences, Philadelphia, Penn-
sylvania; American Ornithologists’ Union, Wash-
ington, D.C.

Marti, C. D. 1992. Barn Owl (Tyto alba). In The
birds of North America, no. 1 (A. Poole, P. Stet-
tenheim, and P. Gill, Eds.). Academy of Natural
Sciences, Philadelphia, Pennsylvania; American
Ornithologists’ Union, Washington, D.C.

Marti, C. D. 1997. Lifetime reproductive success in
Barn Owls near the limit of the species’ range.
Auk 1 14:581-592.

Millsap, B. a. and C. Bear. 1990. Double-brooding
by Plorida Burrowing Owls. Wilson Bull. 102:
313-317.

Morrison, J. L. 1998. Effects of double brooding on
productivity of Crested Caracaras. Auk 1 15:979-
987.

Newton, I. 1979. Population ecology of raptors. Bu-
teo Books, Vermillion, South Dakota.

PlETIAINEN, H., P. SaUROLA, AND H. KOLUNEN. 1984.
The reproductive constraints on moult in the Ural

Owl Stri.K uralensis. Ann. Zool. Penn. 21:277-
281.

Pyle, P. 1997. Plight-feather molt patterns and age in
North American owls. Am. Birding Assoc. Mon-
ogr. Pield Ornithol. 2:1-32.

Reinsch, a. and K. Warncke. 1968. Zweibruten bei
der Waldohreule. Anz. Ornithol. Ges. Bayern 8:
400-401.

Rinne, U. 1981. Beitrag zur Brutbiologie der Waldoh-
reule {Asia otus). Ornithol. Mitt. 33:62—65.

Scott, D. 1997. The Long-eared Owl. The Hawk and
Owl Trust, London, U.K.

Smith, H. G., H. Kallander, and J.-A. Nilsson.
1987. Effect of experimentally altered brood size
on frequency and timing of second clutches in the
Great Tit. Auk 104:700-706.

SoLHEiM, R. 1983. Bigyny and biandry in the Teng-
malm’s Owl Aegolius funereu.s. Ornis Scand. 14:
51-57.

Tinbergen, J. M. and J. H. van Balen. 1988. Pood
and multiple breeding. Acta Congr. Int. Ornithol.
19:380-391.

Ulm.schneider, H. 1990. Post-nesting ecology of the
Long-eared Owl (Asio otus) in southwestern Ida-
ho. M.S. thesis, Boise State Univ., Boise, Idaho.

Verboven, N. and S. Verhulst. 1996. Seasonal var-
iation in the incidence of double broods: the date
hypothesis fits better than the quality hypothesis.
J. Anim. Ecol. 65:264—273.

Wilson Bull, 1 1 1(2), 1999, pp. 276-278

Planning to Facilitate Caching:
Possible Suet Cutting by a Common Raven

Bernd Heinrich'

ABSTRACT — Many species of birds feed on suet
in winter. As far as is known, they all take bite-sized
chunks by pecking into this food randomly and/or they
tear off protruding pieces. I compared the peck-marks
left on suet by Blue Jays (Cyanocitta cristata) and
American Crows (Connis hrachyrhyncho.s) with those
left by Common Ravens (Corvus corax). Although
most ravens feed like jays and crows, at least one in-
dividual made distinct grooves, aligning dozens of
consecutive pecks, apparently to cut transportable
chunks off large suet blocks. Received 2H Aug. 1998,
accepted 7 Jan. 1999.

' Dept, of Biology, Univ. of Vermont, Burlington,
VT 05405.

The Common Raven, Corvus corax, is a
feeding generalist (Bent 1946, Ratcliffe 1997).
Ravens feed on carrion (Ewins et al. 1986),
fruit, grain, eggs, and “garbage” (Nelson
1934, Marquiss and Booth 1986, Engel and
Young 1989). Ravens also capture insects,
reptiles, amphibians, fish, small mammals,
and other birds (Man’ and Knight 1982, Camp
et al. 1993). I here describe a raveri removing
fat from a chunk of suet in an unusual or ab-
enant way that differs markedly from the
method used by jays, crows, and most other
ravens.

Chunks of beef suet that were either of suf-
ficient size so that they could not be earned
otl or that were nailed onto the frozen ground

SHORT COMMUNICATIONS

277

FIG. I. Two top photographs show grooves left by ravens in suet when they were interrupted while feeding
in the wild. A: Deep (1-2 cm) groove in a clear block of beef suet. B: Two grooves cut into suet adhering to
ribs. C: Typical pecking pattern in suet by American Crow. D: Typical pecking pattern in suet by a Blue Jay.
(Right two pictures show heads of spikes used to secure the suet onto frozen ground so that it could not be
carried off.)

were provided in the forest near Hinesburg,
Vermont. For many years the following spe-
cies fed on suet at this site: American Crows
{Corvus brachyrhynchos). Blue Jays {Cyano-
citta cristata). Hairy (Picoides villosus) and
Downy woodpeckers (Picoides pubescens).
Black-capped Chickadees (Poecile atricapil-
lus) and White-breasted Nuthatches (Sitta car-
olinensis). All of these birds fed on suet by
picking into it and/or taking the most promi-
nently protruding pieces.

On one occasion a raven flew away from
the feeding station as I approached, and the
pattern of pecks the raven had left on the suet
was distinctive (Fig. lA). The raven had
carved a 7.5 cm long (and 1—2 cm deep)
groove. I presume that I had inteiTupted the
bird just before it had finished the task of cut-
ting a smaller chunk of suet off the larger. The
same chunk of suet had what appeared to be
a pattern of a previous parallel cut (2-3 cm

anterior to the unfinished cut) where the bird
had already removed one slice of fat. The
readily available small pieces of frozen suet
chips (2-3 mm) that had been loosened were
next to the groove (Fig. lA).

When I inteiTupted two ravens as they were
feeding on the suet at the same site on three
later occasions, I found similar grooves and
only raven tracks around this suet. When I left
suet that was firmly attached to ribs, the ra-
vens cut grooves through the fat down to the
bone underlying the fat. No suet chunks could
be removed, but the bird(s) then cut another,
parallel groove (Fig. IB). Crows and Blue
Jays whose feeding patterns were observed
routinely at the same site always left only
peck-marks, never grooves (Fig. 1C, D). At
no time have I observed groove cutting in
caged ravens that I watched routinely. Neither
have I seen such behavior in hundreds of

278

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

hours of watching groups of ravens feeding
on frozen muscle meat in the wild.

What accounts for the ravens’ unique feed-
ing patterns on suet? The peck marks were
clearly aligned in rows. Each peck could have
provided only a small immediate reward, but
it made possible the removal of a laige chunk
of fat and hence a large reward later. Since
loose “crumbs” of fat were left (Fig. lA), the
delayed (greater) reward was apparently of
more importance to the birds than the small
proximate reward. Ravens exhibit similar ap-
parent foresight during some aspects of their
caching behavior (Heinrich and Pepper 1998).

Food access behavior is of interest because
it has traditionally provided a tool for exam-
ining cognition. Examples with corvids in-
clude studies of memory (Baida and Kamil
1989, Kamil and Baida 1985, Bednekoff et.
al. 1997), tool use (Hunt 1996), optimal for-
aging (Zach 1979, Waite and Ydenberg 1994),
and insight learning (Heinrich 1995).

The fat cutting behavior is probably very
rare in ravens and it is not likely an innate or
hard-wired response. Therefore, the raven
could have had insight of how to remove a
large chunk of food for storage and/or later
consumption.

LITERATURE CITED

Balda, R. P. and a. C. Kamil. 1989. A comparative
study of cache recovery by three corvid species.
Anim. Behav. 38:486-495.

Bednkkohh, P. a., a. C. Kamil, and R. P. Balda.
1997. Clark’s Nutcracker (Aves: Corvidae) spatial

memory: interference effects on cache recovery
performance? Ethology 103:554-565.

Bent, A. C. 1946. Life histories of North American
jays, crows, and titmice. Part 1. Dover, New York.

Camp, R. J., R. L. Knight, H. A. L. Knight, M. W.
Sherman, and T. Y. Kawashima. 1993. Food
habits of the nesting Common Ravens in the east-
ern Mojave Desert. Southwest. Nat. 38:163-165.

Engel, K. A. and L. S. Young. 1989. Spatial and
temporal patterns in the diet of Common Ravens
in southwestern Idaho. Condor 91:371—378.

Ewins, P. J., J. N. Dymond, and M. Marquiss. 1986.
The distribution, breeding and diet of ravens, Cor-
vus corax, in Shetlands. Bird Study 33:110-116.

Heinrich, B. 1995. An experimental investigation of
insight in Common Ravens (Corvus corax). Auk
1 12:994-1003.

Heinrich, B. and J. W. Pepper. 1998. Influence of
competitors on caching behavior in the Common
Raven Connis corax. Anim. Behav. 56:1083-
1090.

Hunt, G. 1996. Manufacture and use of hook-tools
by New Caledonia Crows. Nature 379:249-251.

Kamil, A. C. and R. P. Balda. 1985. Cache recovery
and spatial memory in Clark’s Nutcrackers (No-
cifraga Columbiana). J. Exp. Psychol. 1 1:95-1 1 1.

Marquiss, M. and C. J. Booth. 1986. The diet of
ravens, Corx’us corax, in Orkney. Bird Study 33:
190-195.

Marr, V. and R. L. Knight. 1982. Raven predation
of feral Rock Dove eggs. Murrelet 63:25.

Nelson, A. L. 1934. Some early summer food pref-
erences of the American Raven in southwestern
Oregon. Condor 36:10-15.

Ratcliffe, D. 1997. The Raven. Academic Press Inc.,
San Diego, California.

Waite, T. A. and R. C. Ydenberg. 1994. What cur-
rency do scatter-hoarding jays maximize? Behav.
Ecol. Sociobiol. 34:43—49.

Zach, R. 1979. Shell-dropping: decision making and
optimal foraging in Northwestern Crows. Behav-
iour 68: 106-1 17.

SHORT COMMUNICATIONS

279

Wilson Bull.. 1 1 1(2), 1999, pp. 279-281

Pairing Success of Wood Thrushes in a Fragmented
Agricultural Landscape

Lyle E. Friesen,' - Valerie E. Wyatt,' and Michael D. Cadman'

ABSTRACT — Habitat fragmentation has been as-
sociated with low pairing success of some Neotropical
migrant songbirds occupying forest fragments. From
1996 to 1998, we conducted a nest study of Wood
Thrushes (Hylocichia miistelina) in 21 woodlots rang-
ing in size from 3-12 ha in a highly fragmented ag-
ricultural landscape in southwestern Ontario. We found
active nests for 46 of 48 singing Wood Thrushes that
we detected in the forest fragments. Our results suggest
that in at least some highly fragmented agricultural
landscapes, most singing Wood Thrushes in small
woodlots are successfully paired. Received 17 Sept.
1998, accepted 18 Jan. 1999.

Habitat fragmentation has been associated
with low pairing success of some Neotropical
migrant songbirds residing in forest fragments
and along forest edges. For example, fewer
territorial male Ovenbirds (Seiurus aurocap-
illus) were paired in small forests than in large
ones in Missouri (Gibbs and Faaborg 1990,
Van Horn et al. 1995), New Jersey (Wander
1985), Ontario (Burke and Nol 1998), and
Quebec (Villard et al. 1993). Ziehmer (1993,
cited in Faaborg et al. 1995) documented low-
er pairing success for Red-eyed Vireos (Vireo
olivaceus) and Wood Thrushes {Hylocichia
mustelina) around clearings in large, selec-
tively logged forests in Missouri.

With respect to Wood Thrushes, diminished
pairing success may not apply generally
across all fragmented landscapes. We report
on high pairing success of Wood Thrushes in
forest fragments in Waterloo Region, an in-
tensively farmed landscape with 14% forest
cover and where the mean patch size of wood-
lots was 12.8 ha (± 18.3 SD). (See Friesen et
al., 1999 for a fuller description of the region-
al landscape.)

From 1996 to 1998, as part of a lai'ger re-

‘ Canadian Wildlife Service, 75 Farquhar Street,
Guelph, ON NIH 3N4, Canada.

- Corresponding author;

E-mail: lyle. friesen @sympatico.ca

gional study on nesting success and produc-
tivity of several species of forest birds (Frie-
sen et al. 1999), nest searches were conducted
in 21 woodlots known to hold Wood Thrush-
es; woodlot size ranged from 3-12 ha (x =
8.2 ± 3.3). The canopy at all sites, which av-
eraged 24 m in height, was dominated by sug-
ar maple {Acer sacchariim) and smaller
amounts of white ash {Fraxinus americana)
and American beech {Fagus grandifolia). Ma-
ple and ash saplings, alternate-leaved dog-
wood {Cornus alternifolia), and red-benied
elder {Sambucus pubens) predominated in the
openings created by ongoing and recent selec-
tive logging at all sites.

Searches for singing males involved four to
eight early morning visits to each site begin-
ning the last week of May and continuing to
20 June. We located and mapped the location
of all singing birds at each woodlot by walk-
ing parallel transects 100 m apart and using
taped song playbacks (Yahner and Ross
1995). We attempted to find nests for all sing-
ing birds using four- and five-person teams.
Nest searching, conducted within a 200 m ra-
dius of the singing bird, was discontinued if
an active nest (i.e., containing eggs or young)
was not found within 15 person hours. Sing-
ing birds were assumed to be paired if an ac-
tive nest was found in their vicinity.

Although many singing birds were detected
after 20 June, less effort was expended in find-
ing their nests because of the time constraints
involved in monitoring the nests found pre-
viously. In addition, first broods in southern
Ontario generally fledge aiound 20 June (Frie-
sen, Wyatt, and Cadman, unpubl. data) and
singing birds encountered thereafter could
have been unpaired males moving about be-
tween nests or they could have become un-
paired following the failure of eailier nests
(Roth and Johnson 1993). We were less con-
fident of locating all temtories later in the
breeding season because birds were then less

280

THE WILSON BULLETIN • Vol. HI. No. 2. June 1999

likely to respond to taped playbacks (Friesen,
Wyatt, and Cadman, pers. obs.)- Consequent-
ly, data collected after 20 June were not in-
cluded in this analysis.

One to three singing males were found in
each woodlot, with 48 territorial birds detect-
ed overall. Active nests were found for 46
(96%) of the singing Wood Thrushes, with 2.0
(± 1.7) person hours expended on average to
find each nest following the detection of a
singing bird. Nest height averaged 3.1 m (±
1.5 m, range 1.2-6. 4 m). Our estimate of pair-
ing success may be conservative because a
new but empty nest was found in the vicinity
of one of the two “unpaired” singing males
(suggesting recent predation) and it is possible
that we missed finding the nest of the other
“unpaired” male.

Although Neotropical migrants can experi-
ence poor pairing success in fragmented hab-
itats, sensitivity in this regard likely varies
among species. Based on their pairing success
in small fragments, Ovenbirds seem to be par-
ticularly sensitive to fragmentation effects
while Wood Thrushes are less so. Burke and
Nol (1998) speculated that the lower pairing
success of Ovenbirds in small forests might
be attributed to the absence of females who
avoid these areas of lower food abundance in
favor of larger forests. Wood Thrushes, but
not Ovenbirds, commonly inhabit small naral
woodlots in our study area (Friesen et al.
1995). As both species are ground foragers
dependent upon similar types of prey (Kauf-
man 1996), it may be that factors other than
food availability or perhaps subtle differences
in food preferences are limiting the distribu-
tion of Ovenbirds in our region.

It may be, too, that a species’ pairing suc-
cess varies across regions, perhaps in response
to factors such as differences in landscape
configuration, forest structure, disturbance re-
gimes, and population density. In Missouri,
where Wood Thrushes are “forest interior”
species that only occasionally occupy small
woodlots (Jacobs and Wilson 1997), lower
pairing success occurred as formerly contin-
uous forest became fragmented by logging ac-
tivities (Ziehmer 1993, cited by Faaborg et al.
1995). In our highly fragmented landscape.
Wood Thrushes exhibit a high level of pairing
success. Weinberg and Roth (1998) did not
explicitly comment on the pairing success of

Wood Thrushes in their study in Delaware;
however, based on the large number of active
nests (120 over two years) they found in 14
forest fragments ranging in size from 0.2— 2.1
ha, it appears that Wood Thrush experience a
high degree of pairing success in other frag-
mented landscapes as well.

Our findings on pairing success do not nec-
essarily imply that Wood Thrushes in south-
western Ontaiio are immune to fragmentation
pressures. For example, almost half of the
Wood Thrush nests found in Waterloo Region
in 1996 and 1997 were parasitized by Brown-
headed Cowbirds (Molothrus ater) resulting in
a significant decline in host productivity (Frie-
sen et al. 1999). With I'espect to mating status,
however, our data strongly suggest that most
singing males detected in forest fragments
early in the breeding season are likely to be
paired.

ACKNOWLEDGMENTS

Parts of this project were supported by the Ontario
Region of Environment Canada’s Canadian Wildlife
Service, Human Resources Development Canada, En-
vironmental Youth Corps-Ontario, Long Point Bird
Observatory, and the Regional Municipality of Water-
loo.

LITERATURE CITED

Burke, D. M. and E. Not. 1998. Influence of food
abundance, nest-site habitat, and forest fragmen-
tation on breeding Ovenbirds. Auk 1 15:96—104.
Faaborg, J., M. C. Brittingham, T. M. Donovan, and
J. Blake:. 1995. Habitat fragmentation in the tem-
perate zone. Pp. 357-380 in Ecology and man-
agement of neotropical migratory birds (T. E.
Martin and D. M. Finch, Eds.). Oxford Univ.
Press, New York.

Friesen, L. E., M. D. Cadman, and R. J. Mackay.
1999. Nesting success of neotropical migrant
songbirds in a highly fragmented landscape. Con-
serv. Biol. 13: 1-9.

Friesen, L. E., P. F. J. Eagles, and R. J. Mackay.
1995. Effects of residential development on for-
est-dwelling neotropical migrant songbirds. Con-
serv. Biol. 9:1408-1414.

Gibbs, J. P. and J. Faaborg. 1990. Estimating the
viability of Ovenbird and Kentucky Warbler pop-
ulations in forest fragments. Conserv. Biol. 4:193-
196.

Jacobs, B. and J. D. Wilson. 1997. Missouri breed-
ing bird atlas. Natural History Series. No. 6. Mis-
souri Dcpartmenl ol Conservation, Jefferson City.
Kaui-man, K. 1996. Lives of North American birds.
Houghton Mifflin Company, Boston, Massasschu-
.setts.

SHORT COMMUNICATIONS

281

Roth. R. R. and R. K. Johnson. 1993. Loiig-ierm
dynamics of a Wood Thrush population breeding
in a forest fragment. Auk 1 10:37-48.

Van Horn, M. A., R. M. Gentry, and J. Faaborg.
1995. Patterns of pairing success of the Ovenbird
{Seiiiriis aurocapillus) within Missouri forest
tracts. Auk 1 12:98-106.

ViLLARD, M., P. R. Martin, and C. G. Drummond.
1993. Habitat fragmentation and pairing success
in the Ovenbird. Auk 1 10:759-768.

Wander, S. A. 1985. Comparative breeding biology
of the Ovenbird in large vs. fragmented forests:

implications for the conservation of Neotropical
migrant birds. Pb.D. diss., Rutgers Univ., New
Brunswick, New Jersey.

Weinberg, H. J. and R. R. Roth. 1998. Forest area
and habitat quality for nesting Wood Thrushes.
Auk. 115:879-889.

Yahner, R. H. and B. D. Ross. 1995. Seasonal re-
sponse of Wood Thrushes to taped-playback
songs. Wilson Bull. 107:738-741.

Ziehmer, R. L. 1993. Effects of uneven-aged timber
management on forest bird communities. M.Sc.
thesis, Univ. of Missouri, Columbia.

Wilson Bull., 111(2), 1999, pp. 281-282

Connecticut Warbler, a North American Migrant New to Ecuador

Olaf Jahn,' - Maria Eugenia Jara Viteri,'* and Karl-L. Schuchmann- '*

ABSTRACT. — We present the first record of the
Connecticut Warbler (Oporornis agilis) for Ecuador.
The bird was mist-netted and photographed on 21 No-
vember 1996 at Playa de Oro, Rio Santiago, Esmer-
aldas Province, northwestern Ecuador. Received 15
Sept. 1998, accepted 14 Nov. 1998.

Between August 1995 and December 1996,
we studied understory bird communities near
the village Playa de Oro (00° 52' N, 78° 47'
W) situated at about 50 m above sea-level on
the Rio Santiago, Esmeraldas Province, north-
western Ecuador. Since February 1997 we
have carried out line transect censuses in dif-
ferent habitat types from 50 m to 400 m in
the community of Playa de Oro, close to the
border of the Cotocachi-Cayapas ecological
reserve. On the morning of 21 November
1996, a warbler of the genus Oporornis was
mist-netted in secondary vegetation near a
natural backwater pond on the outskirts of
Playa de Oro. The bird had a complete, pale.

' Fundacion Ecuatoriana de Estudios Ecologicos,
EcoCiencia, Isla San Cristobal 1523 e Isla Seymour,
P.O. Box 17-12-257, Quito, Ecuador.

- Alexander Koenig Research Institute and Museum
of Zoology, Research Group “Biology and Phylogeny
of Tropical Birds,” Adenauerallee 160, D-53 I 13 Bonn,
Germany.

^ Conocoto, Olmedo No 501 y Montarlo Valle de
los Chillos, Quito, Ecuador.

■' Corresponding author.

whitish-yellow eye ring, very long undertail
coverts, a dull brown breast band encircling a
paler throat, a grayish tinged hood and grayish
legs (Fig. 1). Referring to the description in
Curson and coworkers (1994), we determined
that the bird was a first winter Connecticut
Warbler {Oporornis agilis). Its measurements
were: total length (non-stretched) 125 mm,
wing 65 mm, tail 43 mm, wing — tail 22 mm,
bill length 11.95 mm, tai'sus 19.65 mm, body
mass 12.5 g. P7 and p8 were emarginated. Af-
ter the bird was photographed in different po-
sitions (uppeiparts, undeiparts, laterally, and
with opened wing) it was released. The pho-
tographs of the wing formula (primaiies)
show p9 longer than p6. The difference was
not measured exactly in the field. Because a
few female and immature O. Philadelphia also
have complete eye rings, it was necessary to
verify the identification using wing and tail
measurements. According to Lanyon and Bull
(1967), O. agilis can usually be separated
from O. Philadelphia by a wing - tail value
equal to 19 mm or more. The bird we captured
showed a wing — tail value of 22 mm, hence
its identification as O. agilis is virtually cer-
tain. Photos have been deposited at VIREO,
where the identity as O. agilis has been con-
firmed by L. Bevier and R. S. Ridgely (pers.
comm.). Academy of Natural Sciences, Phil-
adelphia.

Oporornis agilis is a rare to locally uncom-

282

THE WILSON BULLETIN • Vol. Ill, No. 2. June 1999

FIG. 1. Connecticut Warbler, Oporornis agilis, captured at Playa de Oro, Esmeraldas Province, northwestern
Ecuador.

mon boreal winter resident south and east of
the Andes, from eastern Colombia and Vene-
zuela south to eastern Peru, northern Bolivia
and west central Brazil (Curson et al. 1994,
Ridgely and Tudor 1989). Recent evidence (R.
S. Ridgley, pers. comm.) indicates that the
species occurs in the northern part of South
America mainly as a transient, but not during
the northern mid-winter months. The bird
trapped in Playa de Oro apparently represents
not only the first record of O. agilis west of
the Andes but also the first record from any-
where in Ecuador (R. S. Ridgely, pers.
comm.), although it had been expected in that
country.

ACKNOWLEDGMENTS

This work is a result of the project ‘Birds as Indi-
cators for Human-influenced Tropical Habitats’ and is
part of the ‘Tropical Ecology Support Program’ (TOB)
of the Deutsche Ge.sellschaft fiir Technische Zusam-
menarbeit (GTZ), Germany. The study would have
been impossible without the financial support of the
Brehm-Funds for International Bird Conservation,

Germany. The project was carried out in cooperation
with the Fundacion para el Estudio e Investigacion de
los Colibries Ecuatorianos (FEICE), Quito, as a na-
tional partner, with the Fundacion Ecuatoriana de Es-
tudios Ecologicos (EcoCiencia), Quito, as a logistical
partner, and with the logistic help, for which we are
most grateful, of the SUBIR Project, CARE/Ecuador,
and the Deutscher Entwicklungsdienst (DED), Ger-
many. We would like to thank the Institute Ecuatoriano
Forestal de Areas Naturales y Vida Silvestre (INE-
FAN), Quito, for permission to work in Ecuador, and
the community of Playa de Oro for permission to work
on their land. We thank L. Bevier and R. S. Ridgely
for the confirmation of the identity of our Oporornis
agilis photos and for critical comments on the manu-
script.

LITERATURE CITED

CuR.soN, J., D. Quinn, and D. Beadle. 1994. New
World warblers. Christopher Helm, London, U.K.
Lanyon, W. E. and J. Bull. 1967. Identification of
Connecticut, Mourning, and MacGillivray’s war-
blers. Bird-Banding 38:187-194.

Ridgely, R. S. and G. Tudor. 1989. The birds of
South America. Vol. 1. The oscine passerines. Ox-
ford Univ. Press. Oxford, U.K. . •

SHORT COMMUNICATIONS

283

Wilson Bull.. 1 I 1(2), 1999. pp. 283-286

Parental Behavior of a Bigamous Male Northern Cariiinal

Randall Breitwisch,'-^ Amy J. Schilling,' - and Joshua B. Banks'

ABSTRACT. — Parental behavior of a bigamous
male Northern Cardinal (Canlinalis cardintiUs) in
southwestern Ohio in 1997 is described. The male was
neither brighter in plumage nor larger than average.
Nesting periods of the two females overlapped. The
male provisioned the primary female during incubation
but not the secondary female. The male delayed pro-
visioning the secondary female’s nestlings until two
days after they hatched but then fed both sets of nest-
lings at rates typical of monogamous males. Despite
initially reduced paternal care, the brood of the sec-
ondary female fledged successfully. Received 12 Mar.
1998. accepted 15 Nov. 1998.

The majority of bird species are socially
monogamous (Lack 1968), the hypothesis be-
ing that ecological constraints explain the rel-
ative infrequency of polygyny in birds (Emlen
and Oring 1977). At the same time, males in
most species of socially monogamous birds
may have the behavioral capability of becom-
ing polygynous should ecological conditions
allow multiple mates (Smith et al. 1982,
Wingfield 1984). The description of infre-
quent cases of bigamy in socially monoga-
mous species is relevant to any discussion of
intraspecific variability in mating arrange-
ments. The behavior of bigamous males to-
ward two females and their offspring may pro-
vide information on the costs to females in-
volved in such mating arrangements.

Northei’n Cardinals {Cardinali.s cardinalis)
are socially monogamous and sexually dichro-
matic; the parental behavior of bigamous
males has not been previously described in de-
tail. Here, we document bigamy and paternal
care by a male Northern Cardinal observed
during 20 h over a 10 day period in early June
1997.

The three cardinals, all of unknown age.

' Dept, of Biology, Univ. of Dayton. Dayton, OH
45469-2320.

- Present address: Dept, of Anthropology. Univ. of
California, Davis, CA 95616-8522.

’’ Corresponding author;

E-mail: breit@neelix.udayton.edu

were members of a color-banded population
located at Aullwood Audubon Center, 15 km
northwest of Dayton, Ohio (39° 52' N, 84° 16'
W) and under continuous observation since
1991. The 80 ha property is a mixture of de-
ciduous woodlands, meadows, and prairies
where cardinals are abundant. The male that
became bigamous in 1997 was banded in the
spring of 1996. That year he was successful
over several others competing for a territory
that had been occupied for several years by a
male that disappeared over the 1995-1996
winter. The territory was one of the largest in
the study area and among those with the most
plant cover, a variable that might provide an
advantage to nesting success (Conner et al.
1986, Wolfenbarger 1996; however, see Fillia-
ter et al. 1994). The male enlarged this terri-
tory in 1997 and it became the site of the big-
amous mating.

One of the two females paired with this
male in 1997 (9 650) was banded in 1996 on
a tenitory adjacent to the one he occupied in

1996. This female’s mate disappeared in the
non-breeding season 1996-1997, as did the
mate of the [bigamous] male. Female 650 re-
mained on the same territory in 1997 and the
bigamous male expanded his 1996 territory to
include the area occupied by female 650. In

1997, this female was treated differently by
the male from the manner in which females
of monogamous males are treated by their
mates (described below), and for these reasons
we consider her the “secondary” female of
the bigamous male. The other female (9 555)
had been banded two years previously in an
area two territories away from the bigamous
male’s 1996-1997 territory. She mated in
1995 but was not found in 1996; her 1995
mate retained the same territory in 1996 and
mated with another female. Female 555 then
reappeared in 1997 on the territory from
which the bigamous male’s 1996 mate had
disappeared. The bigamous male treated fe-
male 555 similarly to how monogamous

284

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

males treat their mates, and we consider her
the “primary” mate of the bigamous male.
We do not know which female first associated
with the bigamous male in 1997.

Male cardinals provision their mates during
courtship, egg-laying, and incubation (Lemon
1968; Vemer and Willson 1969; Breitwisch,
Banks, Donahoo, LeClair, and Schilling, un-
publ. data); this is considered a form of indi-
rect parental care (Lyon and Montgomerie
1985). We sampled provisioning behavior of
the bigamous male toward both females dur-
ing six 1-h observations on six days of the
12-d incubation period. The nests, located ap-
proximately 60 m apart and out of sight of
each other, were monitored simultaneously
during these observation periods, which in-
cluded both mornings and afternoons. Neither
age of eggs nor time of day influences rate of
mate provisioning by male cardinals (Breit-
wisch, Banks, Donahoo, LeClair, and Schil-
ling, unpubl. data). Primary female 555 began
nesting a few days before secondary female
650, and the bigamous male provisioned fe-
male 555 at a similar rate (x = 1 .00 feeding/
h) to the mean value for the monogamous
population [x = 1.05 ± 0.48 (SD) feeding/h;
Breitwisch, Banks, Donahoo, LeClair, and
Schilling, unpubl. data]. However, female 650
received no food from the bigamous male.
This is in marked contrast to a sample of 18
monogamous males in the population, all of
which provisioned their mates during incu-
bation (Breitwisch, Banks, Donahoo, LeClair,
and Schilling, unpubl. data). Although the big-
amous male did not provision female 650, he
remained active throughout his tenitory and
did not appear to favor the area near the nest
of female 555.

Male cardinals feed nestlings at high rates,
frequently surpassing the rate at which fe-
males feed nestlings (Filliater and Breitwisch
1997, Linville et al. 1998). We sampled nest-
ling feedings by the bigamous male and the
two females during l-h samples on seven days
at female 555’s nest and six days at female
65()’s nest. Sampling periods on four of the
days were coincident (see below). Observa-
tion periods included mornings and after-
noons, although Filliater-Lee (1992) showed
that neither male nor female feeding rate is
related to time of day. The eggs of female 555
hatched three days before the eggs of female

650. The male fed the two nestlings of female
555 at a mean rate of 1.3 feeding/nestling/h,
similar to the feeding rate by monogamous
males (T = 1.1 ± 0.53 feeding/nestling/h; Fil-
liater and Breitwisch 1997). Female 555 fed
her nestlings at a mean rate of 0.57 feeding/
nestling/h, similar to the feeding rate by fe-
males mated to monogamous males (x = 0.87
± 0.38; Filliater and Breitwisch 1997). How-
ever, the male did not begin feeding the two
nestlings of female 650 until two days after
they hatched. Female 650 fed her nestlings at
a mean rate of 1.0 feeding/nestling/h, similar
to that of females mated to monogamous
males. When the male began to feed female
650’s nestlings, he fed them at a mean rate of
1.0 feeding/nestling/h, similar to the rate of
feeding by monogamous males. During four
days, nestlings were present in both nests and
the male fed nestlings at both, roughly alter-
nating his deliveries to the two sets of nest-
lings.

The fates of these two nests differed. The
nestlings of the primary female were preyed
upon a few days before they would have
fledged, but the nestlings of the secondary fe-
male fledged successfully. We do not know
whether the three adults maintained the biga-
mous relationship throughout the season.

We determined that the bigamous male was
neither exceptionally ornamented nor notably
large in body size. Using a technique de-
scribed by Linville and coworkers (1998), we
measured the brightness of the red breast
plumage of the male and of the red underwing
plumage of the two females. The bigamous
male and one other male were tied as the dull-
est in a sample of 14 males in 1997. The big-
amous male was also of average body size, as
measured by both tarsus and flattened wing
arc (R. Breitwisch and S.U. Linville, unpubl.
data). The females were both found to be at
least equal to the median plumage brightness
of 15 females in 1997. Primary female 555
was one score lower in brightness than sec-
ondary female 650. We lack size measure-
ments of the two females.

Our observations suggest that there are at
least potential costs for a secondaiy female
mated to a bigamous male cardinal. Most dra-
matically, the bigamous male failed to provi-
sion the secondary female during incubation.
A monogamous male typically provides ap-

SHORT COMMUNICATIONS

285

proxiniately 150 feedings at the nest during
the 12-d incubation period and probably sup-
plies the female with a significant amount of
food away from the nest (Breitwisch, Banks,
Donahoo, LeClair, and Schilling, unpubl.
data). The high rate of mate provisioning in-
dicates that the amount of food provided may
be important to the female’s nutritional state,
especially when considering the three or more
clutches of eggs laid by a typical female in
this population during a breeding season (Fil-
liater et al. 1994). The bigamous male’s be-
havior toward the secondary female’s nest-
lings was not typical of monogamous males
in this population. Although the male even-
tually began to feed the nestlings and did so
at a rate typical for monogamous males, he
delayed two days after these nestlings hatched
before beginning to feed them.

The primary contribution of male cardinals
to raising young appears to be pi'ovisioning
the female, nestlings, and fledglings. Guarding
and active defense against predators are of mi-
nor importance and effectiveness (Filliater et
al. 1994, Nealen and Breitwisch 1997). Thus,
we think it unlikely that any reduced level of
these components of paternal care were a sig-
nificant additional cost of bigamy to either fe-
male.

Bigamy in cardinals appears to be quite rare
(see Linville and Halkin, in press). Lemon
(1968) observed two cases in which he noted
that the females “less tended’’ by the biga-
mous males eventually left and were probably
unsuccessful (R. E. Lemon, pers. comm.). In
our own studies, the instance of bigamy de-
scribed here is the first witnessed in seven
years of monitoring mating relationships in
this population. Each of the last six years, we
have observed an average of about 20 terri-
tories, suggesting that the incidence of bigamy
is probably less than 5% [Verner and Will-
son’s (1969) criterion for monogamy] and
may be even less than 1%. Two other re-
searchers have not observed bigamy in multi-
year studies with a combined sample size of
more than 50 pairs (G. Ritchison, pers.
comm.; L. L. Wolfenbarger, pers. comm.). D.
M. Scott (pers. comm.) and R. E. Lemon
(1957, 1968, pers. comm.) have records of at
least three bigamous males in multi-year stud-
ies of more than 50 pairs of cardinals, al-

though Scott (pers. comm.) agrees with the
above estimate of less than a 5% incidence.

It has been hypothesized that staggered tim-
ing of nesting by two females mated to a big-
amous male may be critical to reducing the
cost of bigamy to the females (Verner 1964,
Breitwisch et al. 1986, Deirickson 1989). Big-
amous males should be able to apportion care
more easily when nests do not overlap in time.
Obviously, we cannot know if the bigamous
cardinal would have provisioned the second-
ary female if her incubation period had not
overlapped with that of the primary female.
Second, with staggered nesting, a female oc-
cupied with caring for eggs or nestlings might
display reduced aggression toward a second
female attempting to nest (Derrickson 1989).
Although female cardinals can be very ag-
gressive toward other females (R. Breitwisch,
pers. obs.), we did not witness aggression be-
tween the two females we observed.

In any case, the secondary female cardinal
was successful in producing fledglings despite
limited paternal care. Richmond (1978) re-
moved male cardinals from nesting pairs and
also found that females were able to raise
young by themselves. We speculate, as did
Richmond, that neglected females may still
pay a cost in future survival from such high
parental effort.

The question that remains is whether a sec-
ondary female in a bigamous relationship is
making the coirect decision at the time of
pairing with a male or committing an enor. It
seems likely that there is no single answer to
this question. In some monogamous species,
secondary females may be able to “predict”
that their young will receive paternal provi-
sioning [e.g.. Northern Shrikes, Lanius exciib-
itor, and Loggerhead Shrikes, L. ludovicianiis,
(Yosef 1992)]. In others, lack of paternal pro-
visioning may be equally predictable [e.g..
Song SpaiTOWS, Melospiza melodia (Smith et
al. 1982), Florida Scrub Jays, Aphelocoma c.
coendescens (Woolfenden 1976)], or paternal
provisioning may depend on degree of overlap
in nesting [e.g.. Northern Mockingbirds, Mi-
mas polyglottos (Logan and Rulli 1981, Breit-
wisch et al. 1986)]. Moreover, there are other
factors that may be involved in determining
level of paternal care at nests of secondary
females, both in species that are opportunis-
tically bigamous and those that ai'e typically

286

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

more polygynous. These include the degree to
which an aspect of paternal care is shareable,
the age and number of nestlings, and the
male’s confidence of paternity (Searcy and
Yasukawa 1995). Explanations of such varied
patterns will await additional reports on big-
amous relationships in socially monogamous
birds.

ACKNOWLEDGMENTS

We thank P. Donahoo and J. LeClair for assistance
in the field and D. Scott and R. Lemon for providing
unpublished data. C. Logan, J. Jawor, and an anony-
mous reviewer offered useful suggestions on the man-
uscript. C. Krueger, Director at Aullwood Audubon
Center and Farm, and J. Ritzenthaler, Head of Re-
search, granted permission to conduct this study on the
Aullwood property. This study was conducted under
USFWS Banding Permit No. 22351 and ODNR Band-
ing Permit No. 5-57-04, both issued to RB.

LITERATURE CITED

Brkitwisch, R., R. C. Ritter, and J. Zaias. 1986.
Parental behavior of a bigamous male Northern
Mockingbird. Auk 103:424—427.

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.

Derrickson, K. C. 1989. Bigamy in Northern Mock-
ingbirds: circumventing female-female aggres-
sion. Condor 91:728-732.

Emlen, S. T. and L. W. Oring. 1977. Ecology, sexual
selection, and the evolution of mating systems.
Science 197:215—223.

Filliater, T. S. and R. Breitwisch. 1997. Nestling
provisioning by the extremely dichromatic North-
ern Cardinal. Wilson Bull. 109:145-153.
Filliater, T S., R. Breitwlsch, and P. M. Nealen.
1994. Predation on Northern Cardinal nests: does
choice of nest site matter? Condor 96:761-768.
Filliater-Lee, T. S. 1992. Parental roles in feeding
nestlings, and nest sites and nest success in North-
ern Cardinals (Cardinalis cardinalis). M.Sc. the-
sis, Univ. of Dayton, Dayton, Ohio.

Lack. D. 1968. Ecological adaptations for breeding
in birds. Methuen, London, U.K.

Lemon, R. E. 1957. A study of nesting cardinals
(Richmondena cardinalis) at London, Canada.
M.Sc. thesis, Univ. of Western Ontario, London.

Lemon, R. E. 1968. The displays and call notes of
cardinals. Can. J. Zool. 46:141—151.

Linville, S. U., R. Breitwisch, and A. J. Schilling.
1998. Plumage brightness as an indicator of pa-
rental care in northern cardinals. Anim. Behav. 55:

1 19-127.

Linville, S. U. and S. L. Halkin. In press. Northern
Cardinal {Cardinalis cardinalis). In The birds of
North America (A. Poole and F. Gill, Eds.). The
Academy of Natural Sciences, Philadelphia, Penn-
sylvania; The American Ornithologists’ Union,
Washington, D.C.

Lyon, B. E. and R. D. Montgomerie. 1985. Incuba-
tion feeding in snow buntings: female manipula-
tion or indirect male parental care? Behav. Ecol.
Sociobiol. 17:279-284.

Nealen, P. M. and R. Breitwisch. 1997. Northern
Cardinal sexes defend nests equally. Wilson Bull.
109:269-278.

Richmond, A. W. 1978. An experimental study of ad-
vantages of monogamy in the cardinal. Ph.D.
diss., Indiana Univ., Bloomington.

Searcy, W. A. and K. Yasukawa. 1995. Polygyny
and sexual selection in Red-winged Blackbirds.
Princeton Univ. Press, Princeton, New Jersey.

Smith, J. N. M., Y. Yom-Tov, and R. Moses. 1982.
Polygyny, male parental care, and sex ratio in
Song Sparrows: an experimental study. Auk 99:
555-564.

Verner, j. 1964. Evolution of polygyny in the Long-
billed Marsh Wren. Evolution 18:252-261.

Verner, J. and M. E Willson. 1969. Mating systems,
sexual dimorphism, and the role of male North
American passerine birds in the nesting cycle. Or-
nithol. Monogr. 9:1-76.

Wingfield, J. C. 1984. Androgens and mating sys-
tems: testosterone-induced polygyny in normally
monogamous birds. Auk 101:665-671.

Wolfenbarger, L. L. 1996. Fitness effects associated
with red coloration of male Northern Cardinals
(Cardinalis cardinalis). Ph.D. diss., Cornell Univ.,
Ithaca, New York.

Woolfenden, G. E. 1976. A case of bigamy in the
Florida Scrub Jay. Auk 93:443-450.

Yosef, R. 1992. Behavior of polygynous and monog-
amous Loggerhead Shrikes and a comparison with
Northern Shrikes. Wilson Bull. 104:747—749.

Wilson Bull., 111(2), 1999, pp. 287-293

Special Report

A SURVEY OF UNDERGRADUATE ORNITHOLOGY COURSES IN

NORTH AMERICA

EDWARD H. BURTT, JR.'-^ AND W. HERBERT WILSON, JR.^

ABSTRACT. — The Committee on Undergraduate Education of the Wilson Ornithological Society conducted
a survey of ornithology courses in North America as a service for teachers of ornithology. Our survey of 26
responses uncovered 26 creative approaches to teaching ornithology. Nonetheless, a number of commonalities
exist. Courses at small colleges and large universities include both lecture and laboratory components and usually
extend into the spring. Most courses emphasize anatomy and physiology, nesting, evolution of birds, ecology,
and flight, with other topics receiving few or no lectures. Almost 60% of the courses include student dissection
or faculty demonstration. Some courses use preserved birds, others use birds that died accidentally, and one uses
roasted chickens that are eaten as part of the skeleto-muscular dissection. Laboratory sessions emphasize tax-
onomy and identification of local and, often, world birds. Most schools have at least a small collection of
specimens available for student use. Courses usually include an extensive project and written work. We hope
the results of the survey will stimulate discussion among teachers of ornithology as we seek to develop new
ideas for our courses. Received 29 Sept. 1997, accepted 8 Jan. 1999.

The Wilson Ornithological Society’s Com-
mittee on Undergraduate Education seeks to
increase the quality of teaching of ornithology
at the undergraduate level and to foster com-
munication among ornithology teachers about
successful and unsuccessful aspects of their
courses. With these goals in mind, we pre-
pared a questionnaire that was sent to all or-
nithology faculty who responded to a request
printed in the Ornithological Societies of
North America newsletter. The following is a
synthesis of the information provided by the
26 ornithologists who completed the question-
naire in 1993 and 1994. Some respondents left
one or more questions unanswered, thus our
analysis of some questions is based on fewer
than 26 responses.

The questionnaire included demographic
and course content questions. Copies of all
completed responses are available from the
Van Tyne Library at the University of Mich-
igan. We first describe the demographics of
our sample, then summarize the quantitative
data, and close with a discussion of successful
and unsuccessful aspects of the courses.

' Dept, of Zoology, Ohio Wesleyan Univ., Delaware,
OH 43015.

^ Dept, of Biology, Colby College, Waterville, ME
04901.

^ Corresponding author;

E-mail: ehburtt@cc.owu.edu

DEMOGRAPHICS OF THE
RESPONDENTS’ INSTITUTIONS

Our small sample is not amenable to mul-
tivariate analysis. Furthermore, because it is
based on only 26 respondents, our survey may
be biased. Our intent is to document the di-
versity of approaches and stimulate discus-
sion.

The 26 responses to the survey came from
ornithology teachers in 18 states and 1 Ca-
nadian province. Thirteen of the respondents
teach at schools in the Eastern Time zone, 1 1
at schools in the Central Time zone, and 2 in
schools in the Mountain Time zone. We re-
ceived no data from faculty teaching at
schools in the Western Time zone. Of the 24
schools in the Eastern and Central Time
zones, 5 are in southern states. Fifteen of the
colleges and universities aie located in small
towns, whereas 1 1 have suburban or urban
campuses. Sixteen of the respondents teach at
state-supported, public institutions; 7 teach at
privately supported, non-denominational col-
leges or universities; and 3 teach at church-
affiliated colleges. Twelve of the schools offer
the Ph.D., 4 the M.Sc. as their highest degree,
and the remaining 10 offer only bachelors’ de-
grees. Twelve schools have more than 10,000
undergraduates, hereafter referred to as large
schools, and 14 schools have fewer than
10,000 (small schools); 7 of these have 2,000
or fewer.

287

288

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

TABLE 1. Quantitative comparison of enrollment as affected by prerequisites tor ornithology courses taught
at large and small schools.

Prerequisites

Large schools
(>10,000)

Small schools
(£10,000)

No. of
schools

Mean class size {±SD)
(1988-1993)

No. of schools

Mean class size (±SD)
1988-1993)

None

50.8 ± 6.8

21.3 ± 2.7

1 semester biology

30.9 ± 20.0

12.6 ± 4.5

2+ semesters biology

26.6 ± 25.7

17.0 ± 12.5

Overall class size

31.4 ± 21.7

14.9 ± 8.1

THE COURSE

Goals. — The generally stated goal of the
classroom portion of the course was to pro-
vide students with a broad overview of orni-
thology, and to use birds as examples of fun-
damental concepts of biology and, to a lesser
extent, cognate disciplines. A second goal was
to use birds to illustrate the scientific process,
which includes hypothesis testing and stimu-
lation of new ideas through debate between
scientists with different interpretations of the
data. A third goal, emphasized by three re-
spondents, was to build a genuine admiration
of the many adaptations of birds, thereby pro-

TABLE 2. Quantitative comparison of some char-
acteristics of ornithology courses taught at large and
small schools (based on number of students enrolled).

Characleristics

Large

schools

(>10,000)

Small

schools

(£10,000)

Lrequency

Annually

Biannually

2-3 yr interval

Duration

14-16 wks

13 wks

10 wks

7-8 wks

Lecture h/wk

Use of primary literature

Assigned to students

Not assigned to students

Labs/wk/student

1 1

moting the life-long study and enjoyment of
birds.

Goals for the laboratory portion of the
course were more varied. Most faculty saw
field identification by sight and song as a pri-
mary goal. Family and order names were con-
sidered part of identification by most respon-
dents. Providing students with a working
knowledge of topography and anatomy, par-
ticularly of feathers, was another common
goal. Six respondents sought to instill appre-
ciation of behavior and ecology through field
experiences. Another six respondents indicat-
ed that a goal of the laboratory was to intro-
duce students to field (e.g., banding, census-
ing, recording of vocalizations, etc.) or mu-
seum (e.g., preparation and measurement of
study skins) techniques.

Structure. — One semester of introductory
biology or zoology was a common prerequi-
site although no prerequisite and two or more
prerequisites also occun'ed. Class size varied
significantly with the number of prerequisites
at both large (Fj 53 = 4.71, P < 0.05) and
small (Fj 49 = 4.05, P < 0.05) colleges and
universities (Table 1). Ornithology classes
with a single prerequisite had smaller enroll-
ments in schools of both sizes (large: t = 3.78,
df = 38, P < 0.001; small: t = 3.74, df = 35,
P < 0.001) than courses with no prerequisite.
Enrollment did not decline further with a sec-
ond or third prerequisite (Table 1). Overall,
class size at small schools was significantly
less it = -5.67, df = 106, P < 0.001) than
at large schools (Table 1).

Faculty at large schools were more likely
(X^ = 6.09, df = 1, P < 0.05) to offer orni-
thology annually than those at small schools,
but the duration of the courses varied similarly
among schools of different sizes (Table 2). All
but three of the courses were taught in the

liurtt anil Wilson • SURVEY OF ORNITHOLOCJY COURSES

289

Flight

Systematics

Nesting

Ecology

Migration,

Orientation

Conservation

Anatomy,

Physiology

Vocalization

Evolution
Biogeography

FIG. I . The proportion of the “consensus” ornithology course devoted to the subjects indicated.

spring semester beginning in January when
students learn to identify waterfowl, raptors,
and relatively few winter residents. Later in
the course, as their field skills improve, stu-
dents are exposed to an increasing diversity of
spring migrants and summer residents.

Twenty-four courses included both lecture-
discussion and laboratory. Lecture-discussion
sections met twice weekly for 75 min/meeting
in 12 courses, three times/week for 50 min/
meeting in 8 courses, and as 1 three-hour sem-
inar in 3 courses. Faculty at small schools pro-
vided more hours of lecture-discussion/week
(X" = 6.75, df = 1, P < 0.01, Table 2) than
those at large schools. Faculty expressed sat-
isfaction with the longer class period and with
occasions when lecture and laboratory could
be integrated.

The typical class period was what one re-
spondent characterized as a “loose lecture,” a
mix of lecture and discussion, illustrated with
specimens, slides and video tapes, and punc-
tuated with questions from the teacher. Sev-
eral faculty indicated plans to incorporate
software in the future, but none were using
computers in the classroom in 1993 and 1994
when the survey was completed.

Content. — Two courses used only the pri-
mary literature, 24 courses required texts, 9 of
these required two texts, and 1 required three.
Gill’s (1990; the survey was completed just
before the second edition) Ornithology was
the preferred text by a wide margin (Appen-
dix), but other texts were used. In addition to

a text, 15 respondents assigned their students
readings from the primary literature. No dif-
ference in use of the primary liteiature was
evident among schools of different sizes (Ta-
ble 2).

Each respondent was asked to provide a
syllabus of his or her course. We assigned the
lectures to 1 1 broadly defined topics. The
mean proportion devoted to each topic by all
respondents is shown in Fig. 1. Some eiTor
was unavoidable as we tried to categorize lec-
tures into the eleven topics. Nevertheless, this
figure represents the “consensus” course of
the surveyed teachers.

Most courses had one laboratory session/
student/week (Table 2) and in most courses it
was a mix of indoor and outdoor sessions.
Laboratory schedules were similai' at large
and small schools (Table 2).

Nineteen courses devoted one or more lab-
oratories to dissection of birds. Of these, two
courses had demonstration dissections by fac-
ulty only. Eight faculty provided their own
dissection guide (Appendix), but others relied
on Pettingill (1990) or Faaborg and Chaplin
(1988b). Preserved pigeons {Colwnba livia)
were used for dissection by 1 3 of 1 9 respon-
dents. One person used fresh pigeons. Chick-
ens (Gallus gallus), Japanese Quail {Coturnix
coturnix), European Starlings {Sturniis vulgar-
is), House SpaiTOWs (Passer domesticiis), and
birds killed in accidents were used in the re-
maining courses or for compainson with pi-
geons. One ornithologist brought a roasted

290

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

chicken to laboratory for dissection and sub-
sequent consumption.

Most instructors required students to own a
field guide. The appropriate Peterson guide
[eastern (1980) or western (1984) North
America] was the most popular choice (Ap-
pendix).

Twenty-three of 25 respondents who taught
a course with a laboratory component required
students to learn to identify species of birds
by sight, usually of the local avifauna. The
number of birds students had to learn varied
from fewer than 50 to over 200, with 101—150
being typical.

Seventeen of the 23 also required their stu-
dents to learn to identify some birds by song.
The number of species each student had to
learn ranged from 21 to 100 with 41-60 being
typical. To help students learn vocalizations,
respondents identified the Peterson tapes, the
Birding by Ear tapes, and the National Geo-
graphic Society tapes as particularly useful.
One respondent had prepared an audiotape
specific to the birds that students had to learn
in the course. A few respondents taped songs
with students and had the students analyze the
songs themselves. This not only taught stu-
dents recording and analytical techniques, but
also gave them a thorough knowledge of the
characteristics of the songs they recorded. To
test students’ abilities to identify birds by
sight and sound, 87% of the respondents gave
laboratory examinations and 35% gave ex-
aminations in the field.

The amount of taxonomy students had to
learn varied. Twenty of 25 instructors required
students to learn order names and know the
distinguishing features of each order. Family
names were required for students in 16 of the
courses. Few instructors required that genera
(two courses) and species (one course) names
be learned.

Most courses required a long written report,
and a few also required one or more short
written assignments. Long written assign-
ments included the following:

• detailed field journal based on 20 hours or
more of fieldwork in addition to the regular
laboratory field trips;

• term paper based on original field or labo-
ratory research or a literature review cov-
ering .some aspect of avian biology;

• Joint paper by several students working on

limited and local research topics. The teach-
er did the literature search. The students
added their own data and synthesized the
material;

• paper based on observations of a bluebird
box on campus. Students monitored the as-
signed box from late March until the young
fledged;

• paper based on the social behavior of a par-
ticular' species with monitoring of the spe-
cies over the course of the semester;

• research paper that usually involved field
research, data analysis, and literature re-
view in which the teacher and classmates
reviewed a rough draft before the final draft
was submitted;

• paper based on field research on the behav-
ior, ecology, or migration of a local bird
species. One insti-uctor disallowed referenc-
es to encourage the creativity and the ob-
servational and analytical skills of each stu-
dent;

• paper based on a census of the birds of a
site that has been censused annually since
1971;

• an account of a local bird in the style used
by the Birds of North America;

• paper based on the analysis of a large sam-
ple of banding data for Yellow Warblers
(Dendroica petechia). The students could
analyze site fidelity as a function of age and
sex, ai'rival dates as a function of age and
sex, etc.;

• analysis of the population dynamics of a
species based on Christmas Bird Count
data.

Short written assignments included the fol-
lowing:

• one page summary of a published article;

• computer spreadsheet assignments on en-
ergetics of flight and thermoregulation;

• three critiques of a set of three or four
papers with contradictory views on a pai-
ticular issue. Each student summarized each
paper and then offered critical comment on
each, taking a position on one side of the
controversy;

• weekly 5—10 minute essays written in class
on specific ornithological questions;

• two critiques of recent ornithological arti-
cles written in the format of the Recent Lit-
erature section of the Journal of Field Or-

Ihtnt and Wilson • SURVEY OF ORNITHOLOGY COURSES

291

nithology. The critiques had to be rewritten
until they reach “A” quality.

SUPPORT FOR UNDERGRADUATE
ORNTIHOLOGY

Avian specimens were equally available to
respondents at large and small schools. Fac-
ulty used collections to illustrate taxonomic
principles, avian systematics, and less often to
illustrate morphological, ecological, and be-
havioral adaptations. In five courses, students
were required to prepare one or more study
skins. Skin prepaiation was optional in six
other courses.

Collections available at small and large
schools were similar in size with 9 of 23 col-
lections having fewer than 1,000 specimens,
some with fewer than 300. Eighteen of the 23
collections had a regional focus, two had
broad North American representation, and
three had large collections representing birds
of the world. All collections included study
skins, most included mounts and skeletons,
and some included eggs, nests, and alcoholic
specimens.

Fifteen schools owned some of the natural
areas visited by the class. Use of these areas
varied from caiefully scheduled, multi-year
censusing of an arboretum managed by the
university, to intermittent visits to unmanaged
areas for “birding.” Here as with use of mu-
seum collections, faculty might benefit from
sharing ideas on how university-owned natu-
ral areas could be used in conjunction with an
ornithology course.

MOST SUCCESSFUL PARTS OF THE
COURSE

Respondents were asked to describe the
most successful parts of their courses. Fifteen
listed some aspect of field trips as the most
successful portion of the course. Interestingly,
one respondent found that some students
loved the laboratory/field portion of the
course while others hated it. Listed below are
the teaching aids and activities instructors
found most successful:

• audiotapes and CDs of bird songs to facil-
itate vocal identification of birds;

• breakfast with the class before or after
morning field trips;

• color slides, whether the instructor’s own or
supplemented from VIREO. One instructor

provided detailed notes on each slide so
that students could devote full attention to
the slides;

• demonstration or experiment that gets stu-
dents involved in active learning;

• field trips to build enthusiasm for learning
species identification and understanding the
biology of birds;

• laboratory and lecture sessions on the same
day to encourage integration of the mate-
rial;

• lectures on ecology and behavior;

• list of mnemonic devices generated by stu-
dents for learning vocalizations;

• lecture demonstrations, for example use of
parachutes, gliders, ornithopters, and
mounted wings in a wind tunnel to illustrate
principles of flight;

• mist-netting and bird-banding to excite stu-
dents’ interest, particularly early in the
course or in conjunction with ongoing re-
search in which the students could partici-
pate;

• morphological, ecological, and behavioral
adaptations of birds;

• study specimens before field trips;

• videotapes, especially those from the Na-
ture series on Public Broadcasting: for ex-
ample Marathon Bird, Rulers of the Wind,
Master Builders, the Bee Team (on social
behavior in White-fronted Bee-eaters), and
Jewels (hummingbirds);

• use of the Macintosh softwaie SoundEdit
(Mac Recorder). Annotated vocalizations of
40 species were provided on departmental
hard disks. Students could play vocaliza-
tions of species they found confusing. The
software also allowed students to make
sonograms and spectrograms of vocaliza-
tions;

• laminated color photographs of birds to
shaipen identification skills, most useful for
institutions with a limited teaching collec-
tion;

• students were provided with essay ques-
tions a week in advance of the test and
could return outlines of their answers at
least 48 hours before the test for comments
by the teacher;

• use of the “Gone Birding’’ game to intro-
duce students to identification. The game
helped promote the goals of enjoyment and
group learning as well as improving iden-

292

THE WILSON BULLETIN • Vol. Ill, No. 2, June 1999

tification skills and knowledge of species-
habitat associations.

LEAST SUCCESSFUL PARTS OF THE
COURSE

Each respondent was asked to identify the
least successful portions of his or her course.
Fellow ornithology teachers can offer little
help with three common complaints: insuffi-
cient time, cold weather, and conclusion of the
semester before the arrival of many spring mi-
grants. Lecture was most often listed as the
least successful part of the course. The least
successful lecture topics included systematics
and physiology and anatomy, despite the fact
that each occupied a substantial part of the
“consensus” course (Fig. 1 ). Additional areas
that some faculty listed as least successful in-
cluded using study skins to teach identifica-
tion and taxonomy, using tapes to leai'n bird
vocalizations, and teaching students field tech-
niques, especially how to quantify behavior in
the field. We hope that this list can stimulate
ideas that will improve these portions of the
courses.

TEACHING MATERIAL NOT PRESENTLY
AVAILABLE

When asked to identify teaching aids that
do not appear to be commercially available,
most respondents indicated a preference for
interactive software. Most would use such
software to help students learn identification
of birds. Specific needs are listed below. If
any readers know of such aids, please notify
us or a member of the Wilson Ornithological
Society Committee on Undergraduate Educa-
tion.

• good video, laser disk, or CD-ROM that
deals only with ordinal chai'acteristics;

• good video, laser disk, or CD-ROM that
deals only with familial characteristics;

• software that illustrates the principles of
taxonomy;

• video or computer disks illustrating field
characteristics of birds and their vocaliza-
tions accompanied by sonograms. Such ma-
terial would allow students to learn at their
own pace;

• software simulations of population dynam-
ics and evolution;

• simple, user-friendly manual on field tech-

niques in ornithology, directed toward stu-
dents.

CONCLUSION

Ornithology is part of the curriculum at
large and small, public and private, graduate
and undergraduate schools throughout North
America. The responses to our survey con-
firmed our sense that ornithology is an excit-
ing, interactive subject taught by men and
women with a genuine enjoyment of birds.
The desire to engender that same life-long en-
joyment in others was a common theme run-
ning throughout all the responses. Beyond this
common theme we were impressed with the
diversity of creative ideas contained within
the courses. We hope that our respondents’
ideas as we have presented them will stimu-
late others to introduce new ideas into their
ornithology courses and to bring those ideas
and their reception by students to ornitholog-
ical meetings for discussion by all who teach
the biology of birds. Finally the Wilson Or-
nithological Society’s Committee on Under-
graduate Education would be glad to work
with any ornithologist who has ideas or sug-
gestions related to the teaching of ornithology.
The Committee is eager to promote increased
public discussion of teaching philosophies and
ideas. We hope this paper marks the beginning
of that public discussion.

ACKNOWLEDGMENTS

We thank the 26 faculty who took the time to re-
spond to our questionnaire. We hope that we have rep-
resented their responses accurately. We thank the
members of the Wilson Ornithological Society’s Com-
mittee on Undergraduate Education for their sugges-
tions throughout the development of the questionnaire
and its analysis, most particularly E. J. Willoughby and
L. Moseley. The manuscript benefited from the many
helpful comments of K. L. Bildstein, A. J. Gatz, D. C.
Radabaugh, and an anonymous referee.

LITERATURE CITED

Faaborg, j. and S. B. Chaplin. 1988a. Ornithology:
an ecological approach. Prentice Half Englewood
Cliffs, New Jersey.

Faaborg, J. and S. B. CiiAPt.iN. 1988b. Ornithology:
an ecological approach. Laboratory manual and
field exercises. Prentice Hall, Englewood Clilfs,
New Jersey.

Gii.l., E B. 1990. Ornithology. W. H. Freeman and
Company, New York.

Mock, D. W. 1991. Behavior and evolution of birds.
W. H. Freeman and Company, New York.

Burn and Wilson • SURVEY OF ORNITHOLOGY COURSES

293

National Gkographic Socikty. 1983. Field guide to
the birds of North America. National Geographic
Society, Washington, D.C.

Pkrrins, C. a. and a. L. A. Middleton. 1985. The
encyclopedia of birds. Facts on File, New York.

Peterson, R. T. 1980. A field guide to the birds.
Houghton Mifflin Company, Boston, Massachu-
setts.

Peterson, R. T. 1984. A field guide to the western
birds. Houghton Mifflin Company, Boston, Mas-
sachusetts.

Pettingill, Jr., O. S. 1990. Ornithology in laboratory
and field. Academic Press, Inc., New York.

Proctor, N. S. and P. J. Lynch. 1993. Manual of
ornithology: avian structure and function. Yale
Univ. Press, New Haven, Connecticut.

Robbins, C. S., B. Brunn, and H. S. Zim. 1983. Birds
of North America. Golden Press, New York.

Welty, J. C. and L. F. Baptista. 1988. The life of

birds. Saunders College Publishing, Philadelphia,
Pennsylvania.

APPENDIX

Texts used for lecture, dissection, and field-
work are listed alphabetically by author (num-
ber of courses using the text are in parenthe-
ses).

Primary lecture text: Faaborg and Chaplin
1988a (2), Gill 1990 (18), Pettingill 1990 (5),
Welty and Baptista 1988 (1);

Supplementary lecture text: Mock 1991 (2),
Perrins and Middleton 1985 (1);

Dissection text: Faaborg and Chaplin 1988b
(1), instructor’s own (8), Pettingill 1990 (5);

Field guide: National Geographic 1983 (3),
Peterson 1980, 1984 (11), Pettingill 1990 (3),
Robbins et al. field guide 1983 (4).

Wilson Bull., 111(2), 1999, pp. 294-301

Ornithological Literature

Edited by William E. Davis, Jr.

SWALLOW SUMMER. By Charles R.
Brown. Univ. Nebraska Press, Lincoln, Ne-
braska. 1998: xiii + 371 pp., black-and-white
photographs. $16.95 (paper). — For the past 15
years, Charles R. Brown and his wife Mary
have studied Cliff Swallows {Petrochelidon
pyrrhonota) at Cedar Point Biological Station
in the Sand Hills of western Nebraska. The
Browns’ research was (and continues to be)
highly productive, partly because the swal-
lows proved to be excellent research material.
Cliff Swallows make mud nests in colonies
that may range from a few birds to thousands.
The swallows are relatively easy to capture
(but provide the basis for many adventures)
and are tolerant of extensive handling. Their
natural history is relevant to numerous issues
in the basic ecology of birds and to the un-
derstanding group behavior of animals.

This book is a non-scientific account of
how the Browns came to work on this bird
and the trials and tribulations of one season
of their studies. Each year they arrive in Ne-
braska in May in advance of the first swal-
lows. They capture, band, weigh, and measure
birds until the swallows stop breeding in late
July. With the help of numerous assistants,
they have obtained data from thousands of
swallows. In fact, their total sample size for
some measurements must be in six figures!

The writing is clear, funny, insightful, in-
teresting, and informative. If you have more
than a passing interest in swallows, group be-
havior, or basic avian ecology, you should also
read the scientific account, “Coloniality in the
Cliff Swallow: the effect of group size on so-
cial behavior” (co-authored with Mary
Brown; 1996, Univ. Chicago Press), if you
have not already done so.

Prairie thunderstorms, clouds of birds, the
mysteries of bird behavior, quotes from west-
ern movies (especially from “Lonesome
Dove”!), the deep pleasures of spending the
summer at a field station, the behavior of field
assistants, the dirt/frustration/exhaustion of
long days in the field, the satisfaction of dis-
covery— these are just a few of the rich

threads in the texture of Brown’s account. If
you have worked for a long time on one spe-
cies, spent a summer with students at a field
station, love western Nebraska, have struggled
to fund and run a field research project, or
simply would enjoy a good account of how
field ornithologists see the world, you will ap-
preciate this book. — CHARLES R. BLEM.

ATLAS OF BREEDING BIRDS OF IN-
DIANA. By John S. Castrale, Edward M.
Hopkins, and Charles E. Keller. Available
from: Indiana Department of Natural Resourc-
es, Customer Service Center, 402 W. Wash-
ington St., Rm. W160, Indianapolis, IN
46204. 1998: 388 pp., 14 numbered text figs.,
7 tables, 158 range maps. $20 plus $3.50 s&h,
$1 sales tax for Indiana residents (cloth). —
The Indiana Nongame and Endangered Wild-
life Program of the Indiana Department of
Natural Resources sponsored this atlas pro-
ject, coordinating the nearly 600 volunteers
and paid “block busters” during the fieldwork
conducted from 1985-1990. Workers targeted
647 “priority blocks,” west-central of the six
blocks of each U.S. Geological Survey IV2
topographic map of the state. A series of maps
depict counties, public lands, rivers, urban ar-
eas, natural regions, forested ai'eas (with sep-
arate maps for evergreen-deciduous and
shrubland-early successional woodlands), ag-
ricultural row-crops, pastureland, and marshes
and open water. In an attempt to provide some
indices of abundance, atlas accounts used
Breeding Bird Survey (BBS) and Summer
Bird Count (SBC) data for the 1985-1990 pe-
riod. The SBC uses the county as the sampling
area, and counts are conducted on multiple
days during June. A table lists in rank order
the percentage of blocks in which a species
was detected (e.g., American Robin, Turdus
rnigratorius, 100%, rank = 1), abundance val-
ues and rank for BBS routes, and birds/party
hour and rank for SBCs.

A biogeographic analysis by J. Dan Web-

294

ORNITHOLOGICAL LITERATURE

295

ster includes a table that lists species extinct
or extiipated by 1929, species extiipated be-
tween 1929 and 1979, possible or sporadic
nesters. Twentieth Century additions as breed-
ers, and confirmed breeding species since
1990. The analysis of bird distributions does
not correspond well with physiographic pat-
terns, vegetation, distribution of other groups
of organisms, or “natural regions.” The au-
thor suggests birds are poor indicators of bio-
geography in a small, flat state like Indiana.
Major changes this century that have had a
major impact on nesting species include the
virtual disappearance of the prairies, the drain-
age and pollution of wetlands, and the frag-
mentation of forests. The extirpation of prai-
rie, wetlands, and forest interior species, to-
gether with a uniform intrusion of alien spe-
cies has resulted in a more uniform avifauna
in the state.

The bulk of the book is devoted to species
accounts. Full species accounts accompany
the 158 species that were confirmed breeders,
supplemented by shorter accounts for 46 spe-
cies not confirmed as breeders during the at-
lasing period (including extirpated breeders
and those confirmed as breeding since 1990).
Each map occupies a full page, with the spe-
cies account on the facing page. The large size
of the maps makes them very easy to read.
The species accounts give very brief natural
histories synopses, and the bulk of the ac-
counts are concerned with historical distribu-
tion comparisons, analysis of the atlas results,
and comparisons with bird distributions in the
surrounding states of Ohio, Michigan, Illinois,
and Kentucky. Each species account is accom-
panied by a table summarizing the atlas, BBS,
and SBC data for north, central, and southern
regions of Indiana, as well as statewide.

This is a well-done atlas that is a bargain
at $20 — the more than 350 references, many
to local publications, alone are worth that. It
should be of interest to those concerned with
bird distribution.— WILLIAM E. DAVIS, JR.

A GUIDE TO THE NESTS, EGGS, AND
NESTLINGS OF NORTH AMERICAN
BIRDS, SECOND EDITION. By Paul J. Bai-
cich and Colin J. O. Harrison. Academic
Press. 1997: 347 pp., 64 color plates, and 103

black and white figures. $22.95 (paper). —
This is an updated version of the 1978 edition,
and this new edition is a must for the library
of anyone interested in the nesting period of
North American birds. The guide begins with
an introduction that describes a variety of as-
pects of breeding biology including how and
where nests are built; egg shape, color, and
size; and clutch size, incubation, hatching, and
the nestling period. Keys to nests, eggs, and
nestlings are also provided in the introductory
pages. The introductory section is followed by
a series of individual species accounts. Each
species account includes breeding habitat; lo-
cation, description, size, and materials of the
nests; number, shape, size, and color of eggs;
breeding season; length of incubation; de-
scription of nestlings; length and description
of nestling period; and roles of both sexes in
these activities.

For many users, the most valuable aspect of
this book is the color plates that provide pho-
tographs of the eggs of 597 species of North
American breeders as well as 147 color draw-
ings of nestlings. While most species’ eggs
are represented by a single photograph, sev-
eral with paiticularly variable eggs are repre-
sented by multiple photographs; for example,
the authors provide six different photographs
of Sandwich Tern {Sterna sandvicensis) eggs.
Additional information about variation within
a species is described in plate legends as well
as individual species accounts. Black-and-
white drawings ai'e scattered throughout the
text to illustrate additional nestlings and a va-
riety of nests.

The second edition of this guide includes
substantial information, photographs, and
drawings from the first edition; however, it
also includes updates and new information
that was not in the first edition. Species names
(both common and scientific) and taxonomic
affinities have been updated to reflect changes
since the first edition. Numerous species ac-
counts have been augmented with information
about breeding biology, numbers of eggs, and
incubation that was missing or not known in
the first edition. Species accounts have been
added for a number of species that were split
from existing species (e.g., Bicknell’s Thrush,
Catharus bicknellr, California Gnatcatcher,
Polioptila californica; and Island Scrub-Jay,
Aphelocoma insularis) or have begun to breed

296

THE WILSON BULLETIN

Vol. Ill, No. 2, June 1999

regularly in North America (e.g., Lesser
Black-backed Gull, Lams fuscus\ Buff-col-
lared Nightjar, Caprimulgus ridgwayr, and
Shiny Cowbird, Molothrus bonariensis) . The
plates have been updated to provide pictures
of the eggs of the new species covered in the
text and to reflect changes in taxonomic order,
species names, and family groups. Another
useful change in the revised edition is the
grouping of all the plates together in the cen-
ter of the guide rather than having the plates
of nestlings scattered throughout the text pag-
es. This change in format makes the guide
easier to use than the first edition. The guide
also now includes a selected bibliography of
important works used in the revision and in-
formation on how to contact the author to ob-
tain specific references for each of the ac-
counts.

Professional and amateur ornithologists will
find that this guide provides extensive infor-
mation about nests and eggs in an easy-to-use
format. I would highly recommend The Guide
to Nests, Eggs, and Nestlings of North Amer-
ican Birds to anyone interested in identifying
birds’ nests and eggs or learning more about
the nest and nestling stage of North American
birds.— SARA R. MORRIS.

BREEDING BIRDS OF WASHINGTON
STATE: LOCATION DATA AND PREDICT-
ED DISTRIBUTIONS. By Michael R. Smith,
Philip W. Mattocks, Jr., and Kelly M. Cassidy.
Available from Seattle Audubon Society,
8050-35th Avenue NE, Seattle, WA 98115.
1997: 538 pp., 10 numbered text figs., 6 ta-
bles, 244 range maps. $30 plus $3 s&h (pa-
per).— The authors attempt to fill two roles
with this book: compilation of the Seattle Au-
dubon Society’s breeding bird atlas project,
and a part of the final report of the Washing-
ton State Gap Analysis project (administered
by the National Biological Service). The goals
of this aspect of the Gap Project were to map
existing land cover, and model the breeding
distributions of birds. The book is divided into
an introductory chapter (22 pp.) and 258 spe-
cies accounts that include 244 maps (489 pp.).

The introductory chapter traces the history
of the Atlas project and summarizes the data
collected and analyzed. Over 600 volunteers

contributed data over the decade beginning in
1985. The sampling unit was a block of nine
square miles that constituted one quarter of a
township, for a total of 7912 blocks. Data pro-
vided by the Washington Department of Fish
and Wildlife, particularly from their shrub-
steppe bird study and National Heritage da-
tabase, were included. Coverage of blocks
was incomplete, with 56% containing at least
one “possible” species record, 51% a “prob-
able” record, and 44% a “confirmed” record.
More than 1 0 species were confirmed breeders
in 6% of the blocks (the highest number of
confirmed species in a block was 67). The au-
thors discuss the biases in the data, including
the concentration of records from more de-
veloped areas of the state. They also point out
that records do not reflect abundance, and give
the delightful example: “a confirmed breeding
record of a Ruby-crowned Kinglet in the Pon-
derosa Pine zone in the Blue Mountains
(where the species is uncommon, but can usu-
ally be found) occupies the same area on the
map as a confirmed record derived from the
Subalpine Fir zone, where Ruby-crowned
Kinglets are best measured by the ton.”

For most species models predicting breed-
ing distribution (extent of a species’ breeding
habitat) were prepared and appeared on the
species’ distribution map as “Habitats in core
zones” with, in some cases, an additional
“Habitats in peripheral zones” presented in a
lighter shade of gray. Habitats were selected
from a satellite image landcover map of the
state. Bird-habitat associations were devel-
oped from literature reviews, location of
breeding records, and consultation with ex-
perts. The introductory chapter detailed the
modeling process and mapping processes, al-
beit with a touch of jargonese (e.g., “Each
scene was spectrally clustered into 120 to 256
spectral classes using bands 1, 2, 3, 4, 5, and
7” or “the presumption that error in a cover
is equal to the product of the errors in each
layer comprising that cover is an over-simpli-
fication. The effect of an error in a source lay-
er upon a derived cover depends on what is
being derived.”). Ecoregions, vegetation, and
vegetation zones are detailed in tables, a
black-and-white map, and in two color plates.

Each species account is divided into three
parts, the first “Breeding Status and Distri-
bution,” the second “Model” gives the char-

ORNITHOLOGICAI, LITERATURK

297

acteristics of the model predicting the breed-
ing distribution, and the third “comments,”
provides aspects of the species’ biology. The
latter contains such information as maximum
abundance numbers for island nesting species,
comments on taxonomy, e.g., American (Cor-
vus brachyrhynochos) and Northwestern (C.
caurinus) crows, and an historical perspective
on the species. I found the “comments” sec-
tions pailicularly informative.

This is an interesting book. It is a hybrid
volume, and thus provides a great deal of in-
formation on habitat and computer modeling
not usually found in breeding bird atlases, but
is also somewhat jaixing because of the con-
trast of sophisticated jargon associated with
the Gap Analysis and the more usual distri-
bution descriptions associated with atlas pro-
jects. The atlas coverage seems sparse and I
wonder why Breeding Bird Survey data
weren’t used to provide a better measure of
bird abundance. As is usually the case, the
more than 200 references, many from local
journals and unpublished sources, are a gold
mine of regional information. Certainly this
book should be part of the library of every
serious student of bird distribution, and all ac-
ademic libraries.— WILLIAM E. DAVIS, JR.

MADE FOR EACH OTHER: A SYMBIOSIS
OF BIRDS AND PINES. By Ronald M. Fan-
ner, Oxford Univ. Press, New York, New
York. 1996: 160 pp., 14 chapters, 24 figures,
8 tables, and 15 color photographs on 4 plates.
$35.00 (cloth). — This slim book interweaves
a description of symbiosis on a “grand scale”
between some corvids and over twenty spe-
cies of pines across “vast tracts of North
American and Asian wildland” with the co-
evolutionary story of the Clark’s Nutcracker
(Nucifraga Columbiana) and whitebark pine
(Pinus albicaulis) of the western United
States. The author makes the case that the nut-
cracker-pine relationship is a strong mutual-
ism because nutcrackers are the most impor-
tant dispersal agents of the pine seeds and the
pine seeds are a nutrient-rich food essential
for nutcracker survival. The critical message,
delivered in the final chapter, “Is the Keystone
Slipping?,” is an alert that serious threats to
the whitebark pine may diminish populations

of the nutcracker and other animal species that
depend on the pine seeds.

This is a book for pine lovers. To support
his argument for the mutualism between cor-
vids and pines. Fanner examines the phytog-
eny, comparative morphology, and ecology of
pines, particularly the 35 species of “soft
pines” in the subgenus Strobus. This group of
species has a high proportion of species with
wingless seeds: a critical adaptation to dis-
persal by corvids. The five species of stone
pines — Eurasian and one North American —
that are close mutualists with the Eurasian
{Nucifraga caryocatactes) and Clark’s nut-
crackers receive the most attention. Charac-
teristics of stone pine cone fertilization and
development, and seed germination and nutri-
tional content, are described in detail. Some
of these descriptions are less relevant to the
main thesis of the book than others, but they
provide a rich natural history.

A review of the family Corvidae quickly
focuses on the “pine birds”: the nutcrackers
and the Pinyon Jay {Gymnorhinus cyanoce-
phalus). “Pine bird” adaptations to a diet of
pine seeds are long bills, the ability to carry
numbers of pine seeds, and well-developed
spatial memory for retrieving cached seeds.
The foundation for the pine-corvid mutualism,
including experimental evidence that Clark’s
Nutcrackers use fixed objects as visual cues
to find food caches, is provided primarily by
the research of Diana Tomback, Stephen Van-
der Wall, Russell Baida, and Fanner’s own
work on seed dispersal by birds. Fanner is
careful to point out that nutcrackers are not
completely dependent on a single species of
pine. He notes that nutcrackers use a variety
of other pine seeds and foods and migrate (ir-
rupt) when the pine seed crop fails. However,
nestlings are fed pine nuts that were cached
almost exclusively, and cached foods aie used
extensively during fall, winter, and spring
when other foods are scarce.

In Fanner’s view winged pine seeds are an-
cestral to wingless seeds, and corvids provide
the selection pressure to make this transition.
He presents a scenario for evolution of wing-
less seeds in two groups of closely related
pine species: the P. ayachuite-strobifomis-
flexilis complex in Mexico and the western
United States and the P. pannflora complex
in eastern Asia. Pine seeds cached by corvids

298

THE WILSON BULLETIN

Vol. Ill, No. 2. June 1999

are less likely to dessicate and more likely to
germinate than seeds dispersed by the wind.
Therefore, in drier climates, characteristics in
pines that enhance dispersal by corvids — ver-
tical fruiting branches, sessile cones, non-
opening “breakaway” scales, seed-retaining
cone cores, and large, wingless seeds — would
be advantageous.

Lanner’s two main theses — whitebark pine
depends on Clark’s Nutcracker for effective
seed dispersal, and the whitebark pine is a
keystone species — superficially seem difficult
to reconcile. Many animal species (birds,
squirrels, and bears) extract, move, and eat
whitebark pine seeds (the keystone concept),
but the nutcracker is essentially the only agent
of seed dispersal (a tight mutualism). The final
chapter (“Is the Keystone Slipping?”) pre-
sents the case that whitebark pine is a species
upon which Clark’s Nutcracker, red squirrel
(Tamiasciunis hudsonicus), and grizzly beai'
{Ursus arctos) depend to varying degrees. The
whitebark pine is seriously threatened by
“competition from more shade-tolerant trees
due to fire exclusion; heightened bark beetle
attacks, also engendered by fire exclusion;
loss of habitat through global warming; and
quick death from a parasitic fungus.” If these
threats severely diminish populations of
whitebark pine, over the long term and in a
diffuse ways, the nutcracker, and to a lesser
extent, the red squinel and grizzly bear will
suffer.

The writing is clear and the pace brisk. Fig-
ures and tables are used judiciously and the
plates are excellent. Some readers may object
to instances of anthropomorphism (“White-
bark pine takes its seeds very seriously indeed
when it comes to distributing the tree’s re-
sources.”) and unabashedly adaptationist in-
terpretations, but this is perhaps expected in a
book directed to a general readership. Details
of pine phylogeny and the uses of pine nuts
by people help to flesh out the story. I heartily
recommend this book to all interested in good
natural history writing and forest ecology in
particular. — R. TODD ENGSTROM.

SKUAS AND JAEGERS: A GUIDE TO
THE SKUAS AND JAEGERS OF THE
WORLD. Klaus Mailing Olsen and Hans

Larsson. Yale University Press, New Haven
and London. 1997: 190 p. 12 color and 1
black-and-white plate, 156 photographs, 20 of
these in color and 7 maps. $35.00 (cloth). —
This is an excellent field guide to the skuas
and jaegers, and the only current reference I
know that correctly illustrates and describes
the juvenal and winter plumages of all seven
species. For this reason, the book is an essen-
tial reference for all interested in seabird iden-
tification. The basic, alternate, and juvenal
plumages of three jaeger and four skua species
are all illustrated with attractive color paint-
ings by Larsson; most or all of these are de-
picted in clear photographs as well. A painting
of adults of the four species of skuas is oddly
reproduced in black-and-white; presumably
this was done to save costs — it slightly mars
the otherwise excellent graphic presentation.

The book is organized into an Introductory
section of 28 pages that includes such topics
as “Breeding Behavior”, “Skuas and Man”,
and “Observing Skuas in the Field”; followed
by the species accounts which form the main
body of the text. The species accounts contain
sections on identification, geographical vaiia-
tion, food, and range during migration and
winter (including maps). Compared to the ob-
vious precision and attention to detail char-
acteristic of the species accounts, I found the
Introductory sections to be somewhat cursory
and containing a number of questionable
statements. For example, on page 9 it is stated
that “Unlike gulls, skuas have supraorbital
salt glands . . . .” — gulls certainly have these
as well, as anyone who has watched “runny-
nosed” gulls at the seashore probably knows.
The dogma about clockwise migrations
around the North Pacific and North Atlantic
Oceans by South Polar Skuas is peipetuated,
despite the lack of evidence to support the no-
tion that any individual bird follows such a
path. My final quibble is about a point of
long-standing confusion about the occunence
of Brown Skuas (Catharacta skua) in the Ca-
ribbean. While the confusion is obviously no
fault of the authors, this book would have
been an ideal forum for the settling of this
issue. I leave it to the reader to puzzle out
whether the recovered banded skuas from
“Guadeloupe” and the “Lesser Antilles” are
one or two individuals, and how likely it is
that either were misidentified South Polar

ORNITHOLOGICAL LITERATURE

299

Skuas (C. macconnicki). Recent sightings
from the North Atlantic suggest that Brown
Skuas may be transequatorial migrants.

To me, the most novel and intei'esting in-
formation in this book pertains to recent ob-
servations of migrating skuas and jaegers,
both from land and from ships at sea. Not only
have some remarkable numbers of these birds
been observed, but compilations of such ob-
servations suggest they could be used as es-
timates of reproductive success the previous
season. This is due to conespondence within
year's of the proportions of juvenile birds seen.
In all, the authors are to be congratulated. This
is a valuable and long overdue work, skillfully
and artfully executed. — RICHARD R. VEIT.

JOHN ABBOT’S BIRDS OF GEORGIA:
SELECTED DRAWINGS FROM THE
HOUGHTON LIBRARY, HARVARD UNI-
VERSITY. Introduction and commentary by
Vivian Rogers-Price. Beehive Press, Savan-
nah, Georgia. 1997: vii-xlii, 25 color plates
and facing-page commentary, unnumbered
and unpaginated. $125 (cloth with linen slip-
case). — This publication presents the first col-
or reproductions, in book form, of bird paint-
ings of John Abbot (1751-1840), an English-
born artist-naturalist who spent most of his
long adult life in Georgia. He completed more
than 5000 watercolors of natural history sub-
jects including more than a thousand of birds,
most of which are extant, and another thou-
sand of insects, their life cycles and food
plants. The remainder includes everything
from mites and ticks to crabs and millipedes.
Abbot’s father had encouraged his early nat-
ural history illustration predilection, providing
him with an abundance of fine bird books and
professional instruction from teacher Jacob
Bonneau. At age 22 Abbot sailed to America
and landed in Virginia where he remained un-
til the rumblings preceding the American Rev-
olution prodded him into moving to the then
less militant Georgia. He remained there for
the rest of his life where he supported himself
largely through the sale of natural history col-
lections and watercolor paintings that were ea-
gerly sought by Europeans from his agent in
London, John Francillon. Abbot published lit-
tle himself, but in England several volumes

about butterflies, moths, and other insects
were illustrated by his work. John Latham
used Abbot’s drawings and specimens in his
book General History of Birds (1821-1824).
Abbot met and aided Alexander Wilson in the
compilation of his American Ornithology, and
aided George Ord in the completion of this
major work following Wilson’s untimely
death.

Probably at least partially because his work
was published by others and because of his
own parochialism, John Abbot’s bird work
was largely eclipsed by Wilson and Audubon.
However, Abbot’s style of presentation was
apparently modeled after George Edwards,
and similar to William Bartram’s, with styl-
ized foregrounds and birds perched on
dwarfed trees. Abbot did not progress artisti-
cally beyond the limitation of this approach,
which Wilson and Audubon did, and by the
time of his death in 1840, his bird portraits
appear' rather archaic. Nevertheless, I find it
remarkable that nearly two centuries elapsed
before a selection of this important Ar'nerican
bird artist’s watercolor paintings of birds were
published.

This large-format (27 X 31 cm) book be-
gins with a 17-page introduction that provides
a biographical sketch of Abbot’s life, high-
lighting his natural history collections and
paintings and his European colleagues and pa-
trons. It is a scholarly work, with 115 end-
notes that, in small type-face, at 19 pages
much exceed the length of the introduction.
The heart of the book is the 25 watercolors of
birds that are reproduced at the same size as
the originals. Each of the paintings is accoi-n-
panied on a facing page by a brief commen-
tary and any of Abbot’s notes that relate to
the species. The book concludes with the his-
tory and final disposition of 12 collections of
Abbot’s original paintings, including the col-
lection of 181 bird portraits, painted from
1801-1810, from which the paintings repro-
duced in this book were selected.

I compared the plates with the originals at
the Houghton Library at Harvard University.
The quality of reproduction was variable but
generally good (the American Oystercatcher,
Haematopus palliatus, was faded in appear-
ance but the Mourning Dove, Zenaida ma-
croura, was as crisp as the original). The
backgrounds in the reproductions were darker

300 THE WILSON BULLETIN • VoL III. No. 2, June 1999

and huffier than the originals and made the
reproductions warmer than the originals, but
muted the colors, especially the greens some-
what; and the contrast was not as crisp. This
was particularly a problem for the light-col-
ored egret and heron paintings. The selection
of paintings includes a hummingbird, which
was the only bird Abbot ever painted in flight,
a signed and dated Bald Eagle (Haliaeetus
leucocephalus), a vulture, an owl, three wood-
peckers, one sparrow, one wren, a nighthawk,
a dove, a crane, seven herons and ibises, an
oystercatcher, a tern, and four ducks. I found
the heavy emphasis on herons, ibises, and
ducks somewhat perplexing, although many,
like the preening Wood Duck (Aix sponsa),
illustrated interesting poses or behaviors.

I have a strong bias against footnotes and
endnotes, and found flipping back and forth
from the introductory text to the endnotes an-
noying. The commentary facing each plate
could have been expanded — there is a lot of
empty space. Minor problems aside, this is an
important work — thoroughly researched and
attractively presented. I commend the Beehive
Foundation for making Abott’s important
work accessible. I recommend the book to
those with particular interest in the history of
ornithology or bird art. The price may limit
sales, but every academic library should have
a copy.— WILLIAM E. DAVIS, JR.

OISEAUX DE LA REUNION. By Nicolas
Barre, Armand Barau, and Christian Jouanin,
illustrated by Nicolas Barre. Second edition,
revised and coirected by Nicolas Barre and
Christian Jouanin. Les Editions du Pacifique,
62 me du Couedic, 75014 Paris. 1996: 207
pp., 10 color plates, numerous color and
black-and-white text illustrations, bibliogra-
phy, indexes. ISBN 2-87868-027-8. Cloth. No
price given. — Although only ten pages longer,
the second edition of “Oiseaux de la Re-
union” is an improved and slightly larger-
sized volume (21 X 15 instead of 20 X 12.5
cm) than the first (which I reviewed in 1983,
Auk 100:541-543). The disadvantage of a
slightly larger size, of course, is that the new
book fits less easily in a pocket than the orig-
inal edition did. As the first edition had been
out of print since 1990, the need for a second

edition had been felt for quite some time. The
problem was that the two original authors
were no longer available for this job. Sadly,
one of them, Armand Barau, who was an
agronomist, had died in 1987, and the second
author and illustrator, Nicolas Barre, a veter-
inarian, had left Reunion in 1982, the year the
first edition was published. In the preface to
the second edition, Christian Jouanin, Asso-
ciate at the Museum national d’histoire natu-
relle in Paris, who brought the project of a
second edition to fmition, explains how Ar-
mand Barau’s widow and Nicolas Barre per-
suaded him to undertake the revision. We ai'e
fortunate that he accepted this task. This im-
portant guide (which, when still in print, was
difficult to get, as one had to write the authors
in Reunion in order to obtain copies) has now
been handsomely produced by the Editions du
Pacifique in Paris.

Although the text of the second edition is
quite similar- to that of the first, Jouanin has
brought the species accounts up to date and
has incorporated much new information, both
published and unpublished, that has been
gathered on the status and distribution of the
birds of Reunion by a number of workers
since the early 1980s. Eight of the ten color-
plates in the second edition are the same as
the eight of the first, except for their number-
ing: Plates VI, VIII, and X were numbered as
V, VII, and VIII, respectively, in the first edi-
tion. Two plates are new: numbers VII (in-
cluding seven species of kites, frigate-birds,
plovers, terns, and ducks; pages 148—149) and
IX (including 4 morphs of the endemic white-
eye Zosterops borbonicus borbonicus and
several plumage variations in the endemic
chat Saxicola tectes\ pages 178-179; a very
welcome addition). Interestingly, the eight
original plates are better reproduced in the
second than in the first edition (at least in my
copies). They are fresher, their colors are
crisper, and their slightly larger size mean that
the birds on each plate are slightly laiger, an
improvement in my opinion. As in the first
edition, the second has numerous black-and-
white text illustrations. Not all original draw-
ings aie included, however, and some have
been redrawn. For example, the attractive and
evocative drawings of several species of sea-
birds and a fishing boat (first edition, page 89)
and of shorebirds on a mudflat (first edition.

ORNITHOLOGICAL LITERATURFi

301

page 131) have been omitted from the second
edition, a pity. The attitudes of plovers, fresh
from a field sketchbook on page 138 of the
first edition, has been redrawn for the second
edition (page 132), but is now stiff and artis-
tically much less interesting. A novel feature
is the inclusion of a color portrait of each spe-
cies, taken from the plates, next to the name
of each species in the species accounts. As in
the first edition, the introductory sections are
illustrated with color photographs of habitats
of Reunion. Whereas there were 6 such pho-
tographs in the first edition, the second has 12,
thus giving the potential visitor a better over-
view of these landscapes.

In addition to the species accounts and the
plates, this volume includes an excellent de-
scription of Reunion (pages 14-23); a thor-
ough and fascinating (if sad) review of the
past avifauna (pages 26-52, illustrated with
nicely reproduced color plates of extinct spe-
cies, borrowed from older publications); a de-

tailed presentation of the modern avifauna
(pages 53-76, including hints about bird
watching and information about conserva-
tion); and suggestions about how to use the
guide. The book ends with a bibliography
(pages 194-199) and four indices (French,
Latin, English, and Creole names). One regret:
this book does not have a map showing the
position of Reunion in the Indian Ocean (this
was true of the first edition also).

Beautifully produced, full of carefully re-
searched information, easy to use thanks to its
user-friendly typography; I strongly recom-
mend this book to all students of insular avi-
faunas. No one visiting Reunion can be with-
out it. As I wrote in my review of the first
edition, this book “should be mandatory read-
ing for all school children of Reunion taking
courses on the geography of their magnificent
island.” My thanks go to Christian Jouanin
for having seen this second edition through
the press.— FRANCOIS VUILLEUMIER.

This issue of The Wilson Bulletin was published on 10 May 1999.

THE WILSON BULLETIN

Editor ROBERT C. REASON

Editorial Board KATHY G. BEAL

Department of Biology

State University of New York

1 College Circle

Geneseo, NY 14454

E-mail: WilsonBull@geneseo.edu

CLAIT E. BRAUN
RICHARD N. CONNER

Review Editor WILLIAM E. DAVIS, JR.

127 East Street

Foxboro, Massachusetts 02035

Editorial Assistants TARA BAIDEME

JOHN LAMAR
DANTE THOMAS
DORIS WATT

Index Editor KATHY G. BEAL
616 Xenia Avenue

Yellow Springs, Ohio 45387

SUGGESTIONS TO AUTHORS

See Wilson Bulletin, 110:152-154, 1998 for more detailed “Instructions to Authors.”
http://www.ummz.lsa.umich.edu/birds/wilsonbull.html

Submit four copies of manuscripts intended for publication in The Wilson Bulletin, 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 (Seventh Edition,
1998) insofar as scientific names of U.S., Canadian, Mexican, Central American, and West Indian birds
are concerned. Abstracts should be brief but quotable. 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 Scientific Style and Format Manual” (AIBS 1994). Photographs for illustra-
tions should have good contrast and be on glossy paper. Submit prints unmounted and provide a brief but
adequate legend for each figure with all captions on a single page. 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 Ornitho-
logical Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

The permanent mailing address of the Wilson Ornithological Society is: do 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.

MEMBERSHIP INQUIRIES

Membership inquiries should be sent to Laurie J. Goodrich, Route 2 Box 301 A, New Ringgold. PA
17960-9445; E-mail: goodrich@haukmountain.org.

CONTENTS

MAJOR PAPERS

A NEW SPECIES IN THE MYRMOTHERULA HAEMATONOTA SUPERSPECIES (AVES; THAM-
NOPHILIDAE) EROM THE WESTERN AMAZONIAN LOWLANDS OF ECUADOR AND PERU

Niels Krahhe, Morton L. Isler. Phyllis R. Isler, Bret M. Whitney,

Jose Alvarez A., and Paul J. Greenfield

COMPARATIVE SPRING HABITAT AND FOOD USE BY TWO ARCTIC NESTING GEESE

Suzanne Carriere, Robert G. Bromley, and Gilles Gauthier

A TEST OF THE CONDITION-BIAS HYPOTHESIS YIELDS DIFFERENT RESULTS FOR TWO

SPECIES OF SPARROWHAWKS (ACCIPITER) - -- - -

Edna Gorney, William S. Clark, and Yoram Yom-Tov

THE DEVELOPMENT OF A VOCAL THERMOREGULATORY RESPONSE TO TEMPERATURE IN
EMBRYOS OF THE DOMESTIC CHICKEN .. Shawn C. Bugden and Roger M. Evans

BEHAVIOR AND VOCALIZATIONS OF THE CAURA AND THE YAPACANA ANTBIRDS

Kevin J. Zimmer

HABITAT PATCH SIZE AND NESTING SUCCESS OF YELLOW- BREASTED CHATS

Dirk E. Burhans and Erank R. Thompson, III

AVIFAUNA OF A PARAGUAYAN CERRADO LOCALITY: PARQUE NACIONAL SERRANIA SAN
LUIS, DEPTO. CONCEPCION Mark B. Robbins, Rob. C. Faucett, and Nathan H. Rice

NOTES ON THE AVIFAUNA OF TABASCO — -

Kevin Winker, Stefan Arriaga Weiss, Juana Lourdes Trejo P., and Patricia Escalante P.

PREDATION OF SMALL EGGS IN ARTIFICIAL NESTS: EFFECTS OF NEST POSITION. EDGE.

AND POTENTIAL PREDATOR ABUNDANCE IN EXTENSIVE FOREST

Richard M. DeGraaf, Thomas J. Maier, and Todd K. Fuller

BIRD USE OF BURNED AND UNBURNED CONIFEROUS FORESTS DURING WINTER ^

Karen J. Kreisel and Steven J. Stein

NEST PREDATORS OF OPEN AND CAVITY NESTING BIRDS IN OAK WOODLANDS

Kathryn L. Purcell and Jared Venter

SHORT COMMUNICATIONS

JUVENILE MARBLED MURRELET NURSERIES AND THE PRODUCTIVITY INDEX

Katherine J. Kuletz and John F. Piatt

"SNORKELING" BY THE CHICKS OF THE WATTLED JACANA

Carlos Bosque and Emilio A. Herrera

RAPID LONG-DISTANCE COLONIZATION OF LAKE GATUN. PANAMA, BY SNAIL KITES
George R. Angehr

THE "SIGNIFICANT OTHERS” OF AMERICAN KESTRELS: COHABITATION WITH AR-
THROPODS - - -- Jeffrey P. Neubig and John A. Smallwood

BARRED OWL NEST IN ATTIC OF SHED C Stuart Hou.ston

DOUBLE BROODING IN THE LONG-EARED OWl

. Jeffrey S. Marks and Alison E. H. Perkins

PLANNING TO FACILITATE CACHING: POSSIBLE SUET CUTTING BY A COMMON RA-
VEN Bernd Heinrich

PAIRING SUCCESS OF WOOD THRUSHES IN A FRAGMENTED AGRICULTURAL LAND-
SCAPE Lyle E. Friesen, Valerie E. Wyatt, and Michael D. Cadman

CONNECTICUT WARBLER, A NORTH AMERICAN MIGRANT NEW TO ECUADOR

.. OlafJahn. Maria Eugenia Jara Viteri, and Karl-L. Schuchmann

PARENTAL BEHAVIOR OF A BIGAMOUS MALE NORTHERN CARDINAL

Randall Breitwisch, Amy J. Schilling, and Joshua B. Banks

SPECIAL REPORT

A SURVEY OF UNDERGRADUATE ORNITHOLOGY COURSES IN NORTH AMERICA

Edward H. Burtt, Jr. and W. Herbert Wilson, Jr.

ORNITHOLOGICAL LITERATURE

157

166

181

188

195

210

216

229

236

243

251

257

262

265

269

272

273
276
279
281
283

287

294

TKc Wilson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. Ill, NO. 3 SEPTEMBER 1999 PAGES 303-456

(ISSN (K)43-S643)

THE WILSON ORNITHOLOGICAL SOCIETY
FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President — John C. Kricher, Biology Department, Wheaton College, Norton, Massachusetts 02766;
E-mail: JKricher@wheatonma.edu.

First Vice-President — William E. Davis, Jr., College of General Studies, 871 Commonwealth Ave., Boston
University, Boston, Massachusetts 02215; E-mail: WEDavis@bu.edu.

Second Vice-President — Charles R. Blem, Dept, of Biology, 816 Park Ave., Virginia Commonwealth
Univ., Richmond, Virginia 23284; E-mail: cblem@saturn.vcu.edu.

Editor — Robert C. Reason, Department of Biology, State University of New York, 1 College Circle,
Geneseo, New York 14454; E-mail: WilsonBull@geneseo.edu.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper Montclair,
New Jersey 07043; E-mail: Smallwood@saturn.montclair.edu.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana 46556;
E-mail: DWatt@saintmarys.edu.

Elected Council Members — Charles E Thompson and Sara R. Morris (terms expire 2000), Jonathan L.
Atwood and James L. Ingold (terms expire 2001), Robert A. Askins and Jeffrey R. Walters (terms
expire 2002).

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

The Wilson Bulletin is sent to all members not in arrears for dues.

THE WILSON BULLETIN
(ISSN 0043-5643)

THE WIL.SON BULLETIN (ISSN 0043-5643) is published quarterly in March, June, September, and December by
the Wilson Ornithological Society, 810 East 10th Street, Lawrence, KS 66044-8897. The subscription price, both in the
United States and elsewhere, is S40.00 per year. Periodicals postage paid at Lawrence, KS. POSTMASTER: Send address
changes to THE WILSON BULLETIN, P.O. Box 1897, Lawrence, KS 66044-8897.

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.

.Subscriptions, changes of address and claims for undelivered copies should be sent to the OSNA, P.O. Box 1897,
Lawrence. K.S 66044-8897. Phone; (785) 843-1221; FAX: (785) 843-1274.

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

THE WILSON BULLETIN

A QUARTERLY JOURNAL OL ORNITHOLOGY
Published by the Wilson Ornithological Society

VOL. Ill, NO. 3 SEPTEMBER 1999 PAGES 303-456

Wilson Bull.. 111(3), 1999, pp. 303-313

ANTILLEAN SHORT-EARED OWLS INVADE SOUTHERN LLORIDA

WAYNE HOFFMAN,' GLEN E. WOOLFENDEN,^ AND P. WILLIAM SMITH^

ABSTRACT. — Recently, Short-eared Owls {Asio flammeus) have invaded extreme southern Florida during
spring and summer, most appear to be post-fledging dispersers. Morphological and plumage characteristics
identify the specimens as coming from the Antilles, most likely from Cuba, where numbers and range have
expanded greatly in recent years. This dispersal continues a trend that began in other bird species more than
half a century ago. Since 1932 about one landbird species per decade has colonized southern Florida from the
Antilles. Received 27 May 1998, accepted 30 Nov. 1998.

During the last two decades. Short-eared
Owls {Asio flammeus) have occurred with in-
creasing frequency in extreme southern Flor-
ida. especially from March through Septem-
ber. In the absence of specimens, these reports
were assumed to represent individuals of the
nominate race A. flammeus flammeus, which
was the only form of this polytypic species
known from the North American continent.
Asio f flammeus is Holarctic; in the western
hemisphere it breeds in northern North Amer-
ica and migrates south as far as southern Unit-
ed States, Mexico, and rarely the West Indies
(American Ornithologists’ Union 1957, 1998).
During the 1990s we obtained several speci-
mens of Short-eared Owls from extreme
southern Florida. Here we summarize all re-
cent records and reports of Short-eared Owls
from southern Florida, describe the character-

' 260 SE 97"’ Ct., South Beach, OR 97366.

^ Archbold Biological Station, Venus, FL 33960.
3p.O. Box 1992, Ocean Shores, WA 98569.

* Corresponding author;

E-mail: gwoolfenden@archbold-station.org

istics of Holarctic and Antillean Short-eared
Owls and conclude that most of the recent
spring-summer records are from an Antillean
population, and briefly review their nomencla-
ture. Finally, we discuss possible causes for
the dispersal of these owls into Florida and
their potential for colonizing the North Amer-
ican continent.

SPECIMENS, PHOTOGRAPHS AND
REPORTS

Specimens. — Between July 1990 and March
1998, we obtained eight dead Short-eared
Owls from the Florida Keys, Monroe County,
Florida. Three of the six males showed no
molt, while three showed light, scattered body
molt. All six males had small testes and all
had extensive black feathering around the
eyes, which in the nominate race typifies Ju-
venal plumage (Holt and Leisure 1993). One
female (GEW 5902, 25 May 1996, ovary 15
X 4 mm, largest ovum 1 mm, substantial body
molt) also appeared to be juvenile, but the oth-
er (GEW 5889, late April 1994, ovary 17 X

<—

FRONTISPIECE. Antillean Short-eared Owl photographed in Ft. Zachary Taylor State Park, Key West,
Monroe County, Florida on 4 April 1994 by Wayne Hoffman.

303

304

THE WILSON BULLETIN • Vol. IN, No. 3, September 1999

4, largest ovum 2 mm) had worn plumage and
possibly a regressing brood patch, and might
have been an adult.

Comparative material included five Short-
eared Owl specimens from the Antilles, four
from the southern Florida mainland, and 46
from elsewhere in North America. In June
1995 Orlando Garrido kindly loaned us an un-
sexed adult specimen (MNHN-1595) collect-
ed in Sancti Spiritus Province, Cuba. In 1996
Garrido donated to the Archbold Biological
Station collections a male specimen collected
near Havana (GEW 5925). We also examined
two nestlings collected in the Dominican Re-
public in November 1963 (LSUMZ 142354
and 142355; Schwartz and Klinikowski 1965).
The four mainland specimens, all from Dade
County, were two from Everglades National
Park (EVER 5035, collected 5 January 1971;
GEW 5890, collected 9 December 1990) and
two in the University of Miami Research Col-
lections (UMRC 948, collected 7 February
1956; UMRC 5387, collected 8 November
1966), which now are in the collections at
Archbold Biological Station.

In April 1997, Hoffman measured six spec-
imens of the Holarctic Asia flammeus flam-
meus in the Pennsylvania State University col-
lections and 40 specimens in the Carnegie
Museum collections. All were taken in North
America throughout the year. The few speci-
mens with missing or incompletely grown pri-
maries or central rectrices were excluded from
analysis. Hoffman also examined and mea-
sured the single Puerto Rican specimen in the
Carnegie Museum.

Measurements taken include lengths of
wing (flattened), tail, tarsus, and culmen from
the cere. Measurements were taken as de-
scribed by Palmer (1962) and Cramp and co-
workers (1977). Tarsi are difficult to measure
on the feathered feet of owls, so a dissecting
probe was used to assist in locating the mea-
suring points on the posterior of the intertarsal
joint and the anterior of the middle toe artic-
ulation with the tarsometatarsus.

Photographs and reports. — We examined
photographs of six individuals that did not be-
come specimens: three from the Dry Tortugas,
two from other of the Florida Keys, and one
from the Gulf of Mexico off Hernando Coun-
ty. We reviewed reports of Short-eared Owls
from Florida published in Audubon Field

Notes and American Birds (Loftin et al. 1991)
and evaluated other written and verbal reports
of these owls in southern Florida since 1978.

RESULTS

Current status of Short-eared Owls in
southern Florida. — Our review of citations in
Audubon Field Notes and American Birds
yielded about 68 reports of Short-eared Owls
in Florida before 1978. With one exception, a
bird seen 14 June 1963 at Lakeport, Glades
County, all pre-1978 reports were of occur-
rences between early October and late March.
We obtained information on 30 occurrences of
33-37 Short-eared Owls in Florida since
1978, including 16 records (specimens and
photographs) and 14 reports (no tangible ev-
idence). Twenty-three of the 30 appeared in
spring and summer, outside the early October-
late March dates dominating the earlier peri-
od. During spring— summer 1994 the influx
seemed particularly heavy. The first bird was
located at Ft. Taylor State Recreation Area in
Key West, 25 March (photographed on 4
April). It was joined by a second bird in mid-
April; both disappeared by mid-May. Mean-
while a birding tour located three birds at the
Dry Tortugas on 8 April (Woolfenden, pers.
obs.). One was picked up there in weakened
condition on 18 April and died while in transit
to Key West for treatment (GEW 5889). On 8
June, W. B. Robertson, Jr. flushed a group of
four owls on Long Key, Dry Tortugas. Thus,
a minimum of five, possibly as many as nine,
owls were found in southern Florida in spring
1994. Another record that summer was of an
individual plucked from the water offshore of
Hernando Co., in June (Table 1). The dates of
occurrence of these records and reports sug-
gest a source other than the Holarctic. The
specimens we have obtained allowed us to test
this hypothesis using geographic variation de-
scribed for Short-eared Owls.

Structural and plumage differences between
the Antillean and Holarctic Short-eared
Owls. — Based on Ridgway (1914) and Wet-
more (1928), and the specimens we examined,
Antillean Short-eared Owls differ from Hol-
arctic Short-eared Owls in size, proportions,
and plumage. These differences appear ade-
quate to distinguish all specimens in the hand,
and to allow identification of birds observed
closely in the field. They also appear sufficient

Hoffman el al. • SHORT-EARED OWLS IN SOUTHERN FLORIDA

r-*

■n

<3J

“O

c/3

C^*

c/3 C/3

c/3

c/5

c/3

D <U

c c

-3

1 £

-E

1 3 1

1 c

-E

-2

< <

d: <

1 <

< 1

1 <

OJ)

D-

OjO

-C

-O

c/3

Oi)

-C

D-

"o'

c/3

— O ri NO
NO ON NO 00
00 00 00 00
lo '/o lo wo

^ ^ ^ ^
U PJ UJ UJ
O O O O

00 00 oo" oo"
QQ CQ CQ 03
< < < <

a -

c/^ :

00 00 00 00

W^J

□

c/f

00'

c/3

(—

1“

. M

<1/

<1)

1 X 1

1 CL

1 Cl I

1 01

c/5

1 PL,

Cl-

0/J

o .

NJ 7: SJ

(U O 4J

a *5 c

c/3 Cl c/:

00 sO

OiJ

—

c/3

1/3

c/3

-o

■p

■a

■0

Cid

-a

< q; <

< < < <

CL (D
. CL

D (1)

^ -C

c/)

a s'

J s

o E

*5 ^3""
ctC —

c/5

CQ ^ W
tc d

oa u

cd ^

•—>

0 c/3

CJ u

c/3

C/3 O-

0 c

^ £

3 'u

u ^

L— I

00 Q

^ eg

^ CQ

“O

ojj

>»

-C

c/3 c/3 c/3 PU

eg eg eg
0/j 00 wj >;

<u cu

p >>

eg Uri
00

3 .E

u eg

>U-ClEQQQm5£QQS

CL 0-

C/5

c/3 C/3

13 “O

X) X

>> >. P

u u o

(U O

aj ^

^ s

a (N
Ofl ^

3 «

U (U

t2 ^

c/3 c/3 O

eg eg -rt
00 00 c

f^,

*— >

O/J

—

•u

C^J

•—i

r^i

—

u 1) u u ^

Q Q Q o

c
o
£

?3 P

>N eg

c/3

00 - .

*-r u. QJ — .

c/o Q O S

c/3 c/3

eg eg

00 n 50

i2 j f2

u 1) u

Q Q

C/3

•o

C ^ CL

3 ^ (D

^ u 00
P

00 <u —

— .

b 'nP
D. « a

< 5 <

_ - . , oo(Nr-

(N (N — n —

3 X
3

eg X

s ^ s

— vO IT)

r^t^oooooooo(X)ooooooooooovONONONONO'ONaNOsaNONONOsONONONO'ON

c^o^a^a^o^CT^^a\o^o^o^a^o^o^a^c^o^o^o^o^c^O'o^a^c^o^a^o^o^o^

305

306

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 2. Measurements of Short-eared Owl specimens. All measurements (in mm) taken by Hoffman. Sta-
tistical comparisons of means are by Student’s t-test (t), unpaired. Probability (P) of equal means is two-tailed.

/4. / fiammeus

West Indian

Measurement

Sex

Mean (Range)

Wing

313.8

(306-323)

289.8

(286-291)

13.77

<0.001

Wing

316.3

(308-325)

294.7

(292-299)

8.30

<0.001

Wing

both

315.0

(306-325)

291.0

(286-299)

16.05

<0.001

Tail

142.0

(133-148)

132.8

(127-141)

5.51

<0.001

Tail

146.1

(134-153)

140.7

(136-146)

1.68

>0.05

Tail

both

144.0

(133-153)

134.9

(127-146)

5.25

<0.001

Tarsus

42.7

(38-46)

52.2

(50-53.5)

10.28

<0.001

Tarsus

43.7

(39-46)

50.7

(50-54)

5.55

<0.001

Tarsus

both

43.2

(38-46)

1 1

51.8

(50-54)

11.67

<0.000

Culmen

26.3

(24.0-28.0)

28.9

(27.8-29.6)

5.21

<0.001

Culmen

26.2

(24.0-28.3)

29.8

(29.5-30.1)

4.51

<0.013

Culmen

both

26.3

(24.0-28.3)

29.1

(27.8-30.1)

6.94

<0.001

to distinguish Antillean specimens from all
other members of Asia flarnmeus.

Compared to Holarctic Asia f flarnmeus,
Antillean Short-eared Owls have shorter
wings and tails, longer tarsi, and slightly larg-
er bills (Table 2). Male and female A. f. flam-
meus overlapped broadly for all measure-
ments, and the Antillean male and female
specimens overlapped broadly in tail and tar-
sus measurements. No overlap existed be-
tween A. f. flarnmeus and the Antillean spec-
imens in length of the wing, tarsus, and cul-
men. The two populations overlapped sub-
stantially only in tail length. Because of the
broad overlap between the sexes for most
measurements, tests of significant mensural
differences between the groups were run for
both sexes combined as well as for males and
females separately. Differences between the
groups were highly significant for all compar-
isons except tail length among females (Table
2). Hoffman’s measurements of A./, flarnmeus
specimens are similar to those published in
Cramp and coworkers (1985), except that the
culmen measurements averaged 3 mm shorter,
suggesting a methodological difference. Ridg-
way (1914), Wetmore (1928), and Garrido
(1984) have published wing measurements for
Antillean owls, and Marshall (pers. comm.)
provided measurements on six specimens at
the US National Museum (Table 3). Together,
these sources provided measurements of 15
specimens (five males, three females, seven
unsexed); wing lengths ranged from 274 to

300 mm, which are similar to ours (286-299
mm).

The most notable plumage differences in-
volve coloration of the upper back and the un-
derparts (Fig. 1). On Holarctic birds the con-
tour feathers of the upper back, between and
anterior to the scapulars, are mostly tawny
with a dark brown central stripe. On Antillean
birds these back feathers are mostly dark
brown with tawny edgings. As a result the up-
per back of Holarctic birds appears distinctly
striped, whereas the backs of Antillean birds
appear overall dark brown, or dark brown
with obscure tawny mottling. The underparts
of A. f. flarnmeus are heavily streaked with
dark brown on a pale tawny to whitish back-
ground (Fig. 1). Streaks are broadest and most
dense on the upper breast, and gradually be-
come narrower and more sparse posteriorly.
The streaked feathers most posterior are lo-
cated on or near the knee joints. The under-
parts of the Antillean owls are much more
buffy overall. The streaking is similar on the
upper breast, but abruptly becomes much nar-
rower and more sparse at mid-breast. The low-
er breast and belly are mostly unstreaked with
a few narrow streaks (less than 2 mm wide)
on the flanks. The feathering on and around
the knee joints is unstreaked. The difference
in pattern and coloration of the underparts was
noted as early as 1770 (Buffon in Wetmore
and Swales 1931). Other plumage differences
include the color of the upper tail coverts
(fairly dark brown in Antillean birds versus

TABLE 3. Measurements (in mni) of Short-eared Owls, from the literature.

Hoffman el al. • SHORT-EARED OWLS IN SOUTHERN FLORIDA

307

(/5

C/5

r--

r-*

—

■a

*T^

k«

P 7 <5

P 7

u ^

2 ^
U ^

■o

—

Tj-

—

r--

in,

0^1

— H

— -

■D

C/5

c/5

'u — O') ’C — 'C 'u

r3 c3 c3 c3

> > > >

‘S: ^

^ c
o a
t-g

^ c

5 > X)
e e

S S

s $

c^U cx.c:l.

— ^5

“ o I

=: c

<u , .

1/5 P ^

3 £

ii: 3 ^

,o
o ,

s: v2

£ =

- o

E 2

C 00 -t

s I §

E >

£ >S — O

r<-, <v} ^

: II II II -3^

c s: s: 1:2

tawny yellow in A. f jiammeus). The sides of
the head just behind the lateral edges of the
facial disk have an unmarked brown patch not
found in A. f. flammeus. The dark bars on the
rectrices also tend to be more complete and
have straighter margins in the Antillean birds.

The feathering of the feet also differs be-
tween Holarctic and Antillean birds. In Asia f.
flammeus the tarsi and the upper surfaces and
sides of the toes are densely covered by fine
tawny feathers. On Antillean birds, the feath-
ering of the toes is restricted to the dorsal sur-
faces and terminates 3-6 mm from the talons.
The featherless skin of the under surfaces ex-
tends up the sides of the toes to near their
tops. Feathering on the toes is also more
sparse, with exposed skin showing between
individual feathers on museum specimens.
The feathers on the tarsi and toes are shorter
in the Antillean birds (ca 6 mm versus 16 mm
on the posterior surface midway along the tar-
sus and ca 3 mm versus 6 mm long on the
proximal phalanx of the middle toe).

Photographs of the Antillean owls appear to
show more prominent “ear” tufts than do A.
f. flammeus, but an attempt to measure the
tufts on the specimens at ABS failed to show
differences (tuft length 19-29 mm for 10 An-
tillean owls; 20—27 mm for 3 flammeus). The
surrounding head feathers appeared shorter on
the Antillean birds so the tufts may in fact
protrude farther, and possibly the Antillean
owls are more likely to erect the tufts when
confronted by photographers.

The Antillean birds appear to weigh sub-
stantially less than northern birds. Three
seemingly non-emaciated males of our Florida
specimens weighed 260 g, 280 g, and 299 g,
and the two non-emaciated females 274 g and
288 g. In comparison, mean weights of A. f
flammeus are 315 g for males, and 378 g and
411 g for two samples of females (Holt and
Leasure 1993).

Comparison of southern Florida specimens
with a series from the Greater Antilles. — We
sent the first two specimens we obtained
(GEW 5861 and GEW 5862) to Joe T. Mar-
shall at the National Museum of Natural His-
tory (NMNH) who compared them with a se-
ries of eight specimens from the Greater An-
tilles (six from Puerto Rico and two from
Cuba). He informed us (pers. comm.) that the
Florida specimens were similar to these An-

308

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

LIG. I. Specimens of Asia f. domin^ensis and Asia f. fiammeu.s. A. Dorsal view. Lett row from top; Asia,
f. dominf’ensis-, 1 unsexed adult from Cuba, 1 female from Llorida. and 4 males from Llorida. Right row from
top: Asia. f. flcimmeus', 3 females from mainland southern Llorida, 1 female from Grand Turk. B. Ventral view.
Left row from top: Asio. f. domin^en.sis', 1 unsexed adult from Cuba, 1 female from Llorida, and 4 males fiom
Llorida. Right row from top: A.sio. f flammeu.f, 3 females from mainland .southern Llorida, 1 female from Grand
Turk. Photograph by Reed Bowman.

Hoffmim el al. • SHORT-EARED OWLS IN SOUTHERN FH.ORIDA

309

tillean birds in wing length and back color and
pattern, but had paler underparts and smaller
bills. He did not assign them unequivocally to
any named population.

Comparison of southern Florida specimens
with Cuban specimens. — Five of the eight
Florida owls not referable to A. f flammeus
were compared with the unsexed Cuban spec-
imen (MNHN 1595), and all eight were com-
paied to the second Cuban specimen (GEW
5925). The first four males from Rorida had
somewhat paler underparts than the first Cuban
specimen (MNHN 1595), but the Florida fe-
male (GEW 5889) had shghtly darker under-
parts. As is true for A. f flammeus (Holt and
Leasure 1993), females of the Antillean form
may have darker underparts than males. The
second Cuban specimen (GEW 5925) has
some darker markings dorsally than most of the
Rorida specimens but otherwise appeared sim-
ilar. The Rorida specimens also agreed in de-
tail with the Cuban specimens in the foot-feath-
ering characteristics. We conclude that these
eight specimens from southern Rorida are va-
grants from the Greater Antilles, and probably
originate from Cuba. We suspect the slightly
paler underparts and smaller bill size noted by
Marshall are characteristics of age or sex.

Analysis of photographs. — The photo-
graphs of six owls from Florida agree in
plumage with the eight specimens described
above, so we consider them Antillean (Table
1). The first of these was discovered by Hoff-
man on Bush Key, Dry Tortugas, on 21 June
1978 and photographed in the hand by Bar-
bara Kittleson (Hoffman et al. 1979). When
the 1990 specimen was recognized as resem-
bling Antillean representatives of the Asia
flammeus species-group, we re-examined the
photographs and found them to show the dark
back and finely streaked buffy underparts of
Antillean birds. The second owl was photo-
graphed by Howard P. Langridge in April
1985 (Kale 1985). It shows extensive black
surrounding the eyes, and very fine streaking
on the breast, indicating it is likely a juvenile
and of Antillean origin. The third owl was
photographed by Paul Cavanagh on Bottle
Key, in northeast Florida Bay, on 24 August
1987. His photographs show the characteristic
unstreaked back and lightly streaked belly of
Antillean birds. The fourth owl, photographed
by Hoffman at Ft. Zachary Taylor, Key West

on 4 April 1994, also shows the unstreaked
back, lightly streaked belly, and the sparsely
feathered feet of Antillean birds, as well as the
black feathering around the eyes seen on the
five presumed juvenile male specimens. The
fifth owl is the first of seeming Antillean or-
igin found north of southernmost Florida. The
bird was rescued from the surface of the Gulf
of Mexico about 110 km west of Hernando
Co (northwest of St. Petersburg) in early June
1994. It was photographed, rehabilitated, and
eventually released by the Birds of Prey Cen-
ter of the Florida Audubon Society. The sixth
owl was photographed 29 April 1997 on Gar-
den Key, Dry Tortugas, by Darlene Friedman
and the photograph was forwarded to us by
Paul Lehman; it also appears to be Antillean.
We suspect that most of the other recent re-
ports from the Florida Keys and Dry Tortugas
also are of Antillean birds. Descriptions from
the observers generally support this conten-
tion. Table 1 includes one A. f. flammeus spec-
imen (collected 10 December 1992) and two
sightings that may belong to this race (24 No-
vember 1979, winter 1983-84).

DISCUSSION

We document the occurrence in southern
Florida of representatives of a population of
the Short-eared Owl previously unrecorded
from the North American continent. The doc-
umentation includes eight specimens and nu-
merous photographs of several birds. The only
taxon of this species previously known to oc-
cur on the continent is the Holarctic nominate
race Asia flammeus flammeus.

Based on the characteristics of the speci-
mens now available, we are confident that
these owls are from the Greater Antilles. On
geographical and historical grounds they most
likely come from Cuba. Seven of our speci-
mens, and perhaps most of the other southern
Florida birds, appear to be juveniles and to
have arrived during post-fledging dispersal.
Breeding by Short-eared Owls is known to oc-
cur in Hispaniola and Cuba during northern
hemisphere winter. Albert Schwartz collected
nestlings in the Dominican Republic in No-
vember (LSUMZ 142354 and 142355), and
Garrido (1984) reported a nest with eggs in
Cuba on 8 December. Based on seasonality of
reports, this pattern of dispersal to southern

310

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

Florida appears to be quite recent, beginning
in the late 1970s.

Nomenclature of Antillean Short-eared
Owls. — Assigning a scientific name to the
owls invading southern Florida is a problem
because the nomenclature of the Antillean
Short-eared Owl is confused and poorly doc-
umented. Briefly the history is as follows.
Muller (1776, reviewed in Wetmore 1928) de-
scribed an owl from Hispaniola, based on
Buffon (1770), as Strix domingensis. Subse-
quent authors ignored this taxon, or assumed
it was based on the Burrowing Owl {Athene
cunicularia) until Wetmore (1928) examined
one short-eared owl each from the Dominican
Republic and Haiti, and concluded these were
examples of the subject of Buffon’s illustra-
tion and account. However, prior to Wetmore’s
work, Ridgway (1882) described similar owls
from Puerto Rico as Asia portoricensis. Wet-
more (1928) compared his two specimens
from Hispaniola to five from Puerto Rico, and
concluded that those from Hispaniola were
only subspecifically distinguishable from the
Puerto Rican specimens. He used the names
Asia domingensis domingensis and A. d. por-
toricensis, respectively. We find no publica-
tion that proposes and explains the merger of
these taxa into Asio flammeus. Although Wet-
more and Lincoln (1933) treated domingensis
as a distinct species, only three years later
Bond (1936), without explanation, listed these
Antillean owls as subspecies of A. flammeus.
This treatment seems to have been followed
by most subsequent authors including Peters
(1940). The situation is further confused be-
cause some recent authors have grouped the
Hispaniolan {domingensis) and Puerto Rican
{portoricensis) populations together, also
without comment. Voous (1988) and Holt and
Leasure (1993), for example, referred to all
Antillean Short-eared Owls as the race por-
toricensis of A. flammeus despite the fact that
the name domingensis seemingly has priority
over portoricensis (Wetmore 1928). The ex-
istence of a Cuban breeding population has
been recognized only since 1981, and no for-
mal determinations of its taxonomic status
have been published. Pending further study,
we recommend using the single epithet dom-
ingensis for all the Antillean populations;
Wetmore’s (1928) justification for maintaining

portoricensis separate from domingensis
seems insufficient given his sample sizes.

We feel that the systematics and nomencla-
ture of the Short-eared Owl are in need of
revision. The validity of several races is in-
adequately established, in South America as
well as in the West Indies. We also suspect
that Asio flammeus may deserve splitting into
two or more species. Ideally, such a revision
would include detailed analyses of vocaliza-
tions, as well as studies of molecular genetic
differences. The West Indian birds are among
the most distinctive in plumage and structure,
but the Galapagos race and the South Amer-
ican populations could plausibly deserve spe-
cies status as well.

Biogeographic considerations. — The recent
occurrence and increasing frequency of these
owls in Florida raises the possibility of a new,
northward colonization from the Antilles.
Owls that reach mainland Florida may find
habitat suitable for nesting. Nesting habitat in
Cuba apparently includes pasturelands, rice
fields (Garrido 1984), and sugar cane planta-
tions (Garrido, pers. comm.). Habitats similar
to all of these occur extensively in southern
Florida.

If these Antillean owls colonize southern
Florida, they will be part of an ongoing wave
of colonizations from the West Indies. When
Howell’s Florida Bird Life was published
(Howell 1932), the breeding landbird fauna of
southern Florida contained only five species’
populations clearly of West Indian origin:
White-crowned Pigeon {Columba leucocepha-
la), (Cuban) Mourning Dove {Zenaida m. rna-
croura). Mangrove Cuckoo {Coccyzus minor).
Gray Kingbird {Tyrannus dominicensis), and
Black-whiskered Vireo {Vireo altiloquus).
Robertson and Kushlan (1984), in their in-
sightful analysis of the southern Florida avi-
fauna, considered all these to be quite recent
immigrants, in part because none showed geo-
graphic variation in Florida. A sixth and sev-
enth species, Zenaida Dove {Zenaida aurita)
and Key West Quail-Dove {Geotrygon chry-
sia), were reported breeding in the Florida
Keys prior to 1850, but both now occur only
as vagrants.

Since 1932 southern Florida has experi-
enced an average of about one natural land-
bird invasion per decade from the West Indies.
These recent immigrants are Smooth-billed

Hoffman el at. • SHORT-EARED OWLS IN SOUTHERN ET.ORIDA

Ani {Crotophaga ani; Sprunt, A. Jr. 1939,
1954), Cuban Yellow Warbler {Deudroica pe-
techia gundlachi; Greene 1942), Antillean
Nighthawk {Chordeiles gundlachii; Greene
1943), Fulvous Whistling-Duck {Dendrocyg-
na bicolor; reviewed by Palmer 1976, Turn-
bull et al. 1989), Cave Swallow [Hirundo fid-
va cavicola (= H. f. Jidva?); Smith et al.
1988j, and Shiny Cowbird {Molothrus bon-
airiensis; Smith and Sprunt 1987).

Most of these invading species had been
known as vagrants to southern Florida for
some time, and reports became increasingly
frequent before breeding in Florida was doc-
umented. This fits a general, but often over-
looked rule: range expansion tends to be driv-
en by population dynamics (often population
increases) in the source areas, rather than by
habitat changes or initial reproductive success
in the colonized areas.

Several West Indian birds, in addition to
these owls, seem poised to invade Florida
from the south or east (Bahamas). La Sagra’s
Flycatcher (Myiarchus sagrae; Smith and Ev-
ered 1992), Bahama Mockingbird (Mimus
gundlachii), and Thick-billed Vireo (Vireo
crassirostris; Smith et al. 1990) seem to be
increasing in frequency as vagrants. Pearly-
eyed Thrasher (Margarops fuscatus) has not
been documented in North America as of this
writing, but it has been extending its range
northward in the Bahamas in recent years and
could begin appearing in Florida in the near
future. The Cuban subspecies of American
Kestrel (Falco sparverius sparveroides) also
has been expanding its range in the Bahamas,
and recently was photographed by WH at Key
West. Robertson and Kushlan (1984:226)
speculated on potential immigrant West Indian
species, naming “the Masked Duck, a hum-
mingbird, Bahama [Tachycineta cyaneoviri-
dis), and Cave swallows, Bananaquit [Coereba
flaveola]. Stripe-headed Tanager [Spindalis
zena] and Black-faced Grassquit [Tiaris bi-
color]” most likely. Of these, the Cave Swal-
low and possibly the Masked Duck (Bowman
1995) already have colonized.

The rate of immigration in recent decades,
then, must be much higher than the overall
post-Pleistocene rate, unless prehistoric ex-
tinction rates for immigrant populations were
extremely high. The Zenaida Dove and Key
West Quail Dove, once reported to breed in

31 1

the Florida Keys, no longer do so. These pos-
sible extirpations of breeding populations
most likely resulted from hunting and habitat
destruction (Robertson 1978a, 1978b; Robert-
son and Woolfenden 1992).

The apparent increase in immigration rates
from the West Indies to Florida may have re-
sulted from anthropogenic changes in the en-
vironment, both in the West Indies and in
Florida. Flabitat changes, associated with for-
est clearing for grazing, cultivation, and urban
development, and global climate changes are
two nonexclusive anthropogenic changes to
the regional environments that could drive
these colonizations. All the known Cave
Swallow colonies in Florida, for example, are
located on concrete bridges and overpasses
along highways (Smith et al. 1988), so the in-
crease in numbers of vagrants prior to colony
establishment (Robertson and Woolfenden
1992) must have reflected population or hab-
itat changes in Cuba.

Explanations for northward colonization in-
volving global climate change potentially can
explain both the recent wave of colonizations
and the dearth of West Indian birds in the
southern Florida avifauna prior to this centu-
ry. A global warming trend has been under-
way since the end of the “Little Ice Age” in
about 1870 (Pielou 1991), a trend that has ac-
celerated because of increases in atmospheric
carbon dioxide, methane, and other green-
house gasses (Maul 1989). Some West Indian
birds may have colonized the Florida penin-
sula previously during the Holocene, but
stopped during the “Little Ice Age”. Al-
though records are inadequate or nonexistent,
the lower average temperatures of the “Little
Ice Age” likely were manifested in southern
Florida by more frequent and more severe ep-
isodes of cold winter weather rather than by
cooler summer weather. These hypothetical
cold episodes could have been particularly
damaging to West Indian bird populations at-
tempting to persist in southern Florida. The
current warming trend also may be fueling
population increases in the West Indies for the
species that have recently colonized, or are
appearing more frequently as vagrants. This
warming trend also may be making southern
Florida habitats subtly more suitable for these
birds (Robertson and Kushlan 1984).

312

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

ACKNOWLEDGMENTS

Much of the preparation of this paper took place at
Archbold Biological Station. We thank the staff for
providing us the facilities and opportunity to work
there. L. Quinn of the Llorida Keys Wild Bird Center
in Key Largo, B. Arnold of Wildlife Rescue of the
Llorida Keys in Key West, and K. Grinter of Marathon
Wild Bird Rescue have saved numerous bird speci-
mens for us, including all eight of the Antillean owls.
R. Collins provided information on the owl rehabili-
tated by Llorida Audubon’s Birds of Prey Center. S.
Cardiff and W. Meshaka assisted us in borrowing spec-
imens from the Museum of Natural Science, Louisiana
State University and Everglades National Park, respec-
tively. O. Garrido graciously transported the Cuban
specimens to us, and J. Marshall examined certain of
our recent specimens. H. Gonzalez Alonso, Director of
the Museo Nacional de Historia Natural, Havana, do-
nated to Archbold Biological Station specimen GEW
5925 through the request of O. Garrido. P. Lehman sent
to us the photograph documenting the 1997 record at
the Dry Tortugas. D. Steadman, R. Browning, and R.
Banks provided certain references. J. Litzpatrick, D.
Holt, J. Marshall, D. Steadman, and K. Voous re-
viewed various drafts of the manuscript and made nu-
merous helpful comments. We thank all of these per-
sons for their assistance.

LITERATURE CITED

American Ornithologists’ Union. 1957. Check-list
of North American birds, hfth ed. American Or-
nithologists’ Union, Washington, D.C.

American Ornithologists’ Union. 1998. Check-list
of North American birds, seventh ed. American
Ornithologists’ Union, Washington, D.C.

Bond, J. 1936. Birds of the West Indies. The Academy
of Natural Sciences of Philadelphia, Philadelphia,
Pennsylvania.

Bowman, M. 1995. Sighting of Masked Duck duck-
lings in Llorida. Lla. Lield Nat. 23:35.

Buffon, G. L. L. 1770. Histoire Naturelle des Oiseaux,
vol. 1. Paris, Prance.

Cramp, S., K. E. L. Simmons, I. J. Perguson-Lees, R.
Gillmor, P. a. D. Hollom, R. Hudson, E. M.
Nicholson, M. A. Ogilvie, P. J. S. Olney, K. H.
Voous, AND J. Wattel. 1977. Handbook of the
birds of Europe, the Middle East, and North Af-
rica: the birds of the Western Palearctic. Vol. 1.
Oxford University Press, Oxford, U.K.

Cramp, S., D. J. Brooks, E. Dunn, R. Gillmor, P. A.
D. Hollom, R. Hudson, E. M. Nicholson, M. A.
Ogilvie, P. J. S. Olney, C. S. Roselaar, K. E. L.
Simmons, K. H. Voous, D. 1. M. Wallace, J.
Wattel, and M. G. Wilson. 1985. Handbook of
the birds of Europe, the Middle East, and North
Africa: the birds of the western Palearctic. Vol. 4.
Oxford University Press, Oxford, U.K.

Garrido, O. H. 1984. A.sio fiammeu.s (Aves: Strigidae)
nesting in Cuba. Caribb. J. Sci. 20:67—68.

Greene, E. R. 1942. Golden Warbler nesting in lower
Florida Keys. Auk 59:114.

Greene, E. R. 1943. Cuban Nighthawk nesting in low-
er Florida Keys. Auk 60:105.

Hoffman, W., W. B. Robertson, Jr., and P. Patty.
1979. Short-eared Owl on Bush Key, Dry Tortu-
gas, Florida. Fla. Field Nat. 7:29—30.

Holt, D. W. and M. S. Leasure. 1993. Short-eared
Owl {Asia flammeus). In The birds of North
America, no. 62 (A. Poole and F. Gill, Eds.). The
Academy of Natural Sciences, Philadelphia, Penn-
sylvania; The American Ornithologists’ Union,
Washington, D.C.

Howell, A. H. 1932. Florida bird life. Coward-Mc-
Cann, Inc. New York.

Kale, H. W., II. 1985. Florida region. Am. Birds 39:
288-291.

Loftin, R. W., G. E. Woolfenden, and J. A Wool-
fenden. 1991. Florida bird records in American
Birds and Audubon Field Notes (1947—1989):
species index and county gazetter. Fla. Ornithol.
Soc. Sp. Pub. 4:1-99.

Maul, G. 1989. Implications of climatic changes in
the wider Caribbean region. CEP Tech. Rept. 3:
1-23.

Palmer, R. S. (Ed.). 1962. Handbook of North Amer-
ican birds. Vol. 1. Yale Univ. Press, New Haven,
Connecticut.

Palmer, R. S. (Ed.). 1976. Handbook of North Amer-
ican birds. Vol. 2. Yale Univ. Press, New Haven,
Connecticut.

Peters, J. L. 1940. Check-list of birds of the world.
Vol. 4. Harvard Univ. Press, Cambridge, Massa-
chusetts.

PiELOU, E. 1991. After the Ice Age: the return of life
to glaciated North America. Univ. of Chicago
Press, Chicago, Illinois.

Ridgway, R. 1882. Description of a new owl from
Porto Rico. Proc. U.S. Natl. Mus. 4:366-371.

Ridgway, R. 1914. Birds of North and Middle Amer-
ica. Bull. U.S. Natl. Mus. 50(6): 1-882.

Robertson, W. B., Jr. 1978a. Key West Quail-Dove.
P. 119 in Rare and endangered biota of Florida.
Vol. 2. Birds (H. W. Kale, Ed.). Univ. Presses of
Florida, Gainesville.

Robertson, W. B., Jr. 1978b. Zenaida Dove. P. 120 in
Rare and endangered biota of Florida. Vol. 2.
Birds (H. W. Kale, Ed.). Univ. Presses of Florida,
Gainesville.

Robertson, W. B., Jr. and J. A. Kushlan. 1984. The
southern Florida avifauna. Pp. 219-257 in Envi-
ronments of southern Florida: past and present II
(P. J. Gleason, Ed.). Miami Geological Society,
Coral Gable, Florida.

Robertson, W. B., Jr. and G. E. Woolfenden. 1992.
Florida bird species: an annotated list. Fla. Orni-
thol. Soc. Sp. Publ. 6:1-260.

Schwartz, A. and R. F. Klinikowski. 1965. Addition-
al observations on West Indian birds. Not. Nat.
376:1-16.

Hoffman et al. • SHORT-EARED OWLS IN SOUTHERN FLORIDA

313

Smith, R W. and D. S. Evered. 1992. Photo note — La
Sagra’s Flycatcher. Birding 24:294-297.

Smith, P. W., D. S. Evered, L. R. Messick, and M. C.
Wheeler. 1990. Eirst veritiable record of the
Thick-billed Vireo from the United States. Am.
Birds 44:372-376.

Smith, P. W., W. B. Robertson, Jr., and H. M. Ste-
venson. 1988. West Indian Cave Swallows nest-
ing in Florida, with comments on the taxonomy
of Hirundo fulva. Fla. Field Nat. 16:86—90.
Smith, P. W. and A. S. Sprunt, IV. 1987. The Shiny
Cowbird reaches the United States. Am. Birds 41:
370-371.

Sprunt, A. S. Jr. 1939. Smooth-billed Ani nesting in
Florida. Auk 56:335.

Sprunt, A. S. Jr. 1954. Florida bird life. Coward-
McCann, Inc., New York.

Turnbull, R. E., E A. John.son, and D. A. Brokhage.
1989. Status, distribution, and foods of Eulvous
Whistling-Ducks in southern Florida. J. Wildl.
Manage. 53:1046-1051.

Voous, K. H. 1988. Owls of the northern hemisphere.

The MIT Press, Cambridge, Massachusetts.
Wetmore, a. 1928. The Short-eared Owls of Porto
Rico and Hispaniola. Proc. Biol. Soc. Wash. 41:
165-166.

Wetmore, A. and F. C. Lincoln. 1933. Additional
notes on the birds of Haiti and the Dominican Re-
public. Proc. U.S. Natl. Mus. 82:1-68.

Wetmore, A. and B. H. Swales. 1931. The birds of
Haiti and the Dominican Republic. U.S. Natl.
Mus. Bull. 155:1-483.

Wilson Bull., 111(3), 1999, pp. 314-320

WITHIN- AND BETWEEN- YEAR DISPERSAL OF AMERICAN
AVOCETS AMONG MULTIPLE WESTERN GREAT

BASIN WETLANDS

JONATHAN H. PLISSNER,' SUSAN M. HAIG,' ^ AND LEWIS W. ORING^

ABSTRACT. — Connectivity of discrete habitat patches may be described in terms of the movements of
individual organisms among such patches. To examine connectivity of widely dispersed alkali lake systems, we
recorded post-breeding and subsequent breeding locations of color-banded American Avocets (Recun’irosira
americana) in the western U.S. Great Basin, from 1995—1997. Among individuals observed during the post-
breeding/premigratory season, over half of the 188 breeding adults were observed at lakes other than their
breeding locations, whereas 70% of 125 post-fledged young were observed only at their natal lake systems. Of
46 breeding adults observed in consecutive years, only eight (17%) dispersed between different lake systems.
Only 8% of chicks were observed after their first year, and only 1.3% returned to the natal area in subsequent
breeding sea.sons. Adult and recently fledged birds from the southernmost breeding site were regularly observed
in post-breeding aggregations at lakes several hundred kilometers to the north, suggesting seasonal differences
in habitat quality at the lake systems studied. These results indicate the importance of maintaining habitat for
post-breeding movements. Received 10 Dec. 1998, accepted 3 April 1999.

Concurrent with the recognition that habitat
fragmentation is a key threat to regional bio-
diversity, the role of dispersal in maintaining
connectivity between populations and subpop-
ulations has become a major focus in assess-
ing extinction risks and other dynamics of
populations and communities (Saunders et al.
1991, McPeek and Holt 1992, Taylor et al.
1993, Dunning et al. 1995, With et al. 1997,
Haig et al. 1998). For vertebrate populations,
dispersal studies generally have focused upon
movements within contiguous habitats or
among adjacent patches. Although appropriate
for some species, this approach fails to ac-
count for a significant proportion of dispersal
events among more mobile species, especially
those that inhabit discrete, widely-dispersed
patches of habitat. In addition, connectivity
measures require monitoring of multiple sites
as both potential sources and recipients of dis-
persers. Such interchange among patches has
been incorporated to varying degrees into nu-
merous spatial models of populations (see
Doebeli and Ruxton 1997, Ims and Yoccoz
1997, Wiens 1997), but empirical measures of
such rates of exchange are often difficult to
determine for vertebrate metapopulations.

' U.S. Geological Survey, Forest and Rangeland
Ecosystem Science Center, 3200 SW Jefferson Way,
Corvallis, OR 97331.

^ Dept, of Environmental and Resource Sciences,
Univ. of Nevada, 1000 Valley Rd., Reno, NV 89512.

’ Corresponding author; E-mail: haig.s@fsl.orst.edu

Specifically, studies of avian dispersal are
largely limited by the spatial scale that re-
searchers are able to effectively monitor dis-
persing individuals and by a traditional focus
on return rates rather than on broader dispersal
patterns. Typically, such studies focus on a
population inhabiting a single site or a few
neighboring areas, with the probabilities of
detection for dispersers decreasing geometri-
cally with distance from the point of origin
(Barrowclough 1978, Cunningham 1986).

To determine large-scale connectivity pat-
terns among discrete wetlands of the western
Great Basin, we examined movements of
banded American Avocets (Recurvirostra
americanci) among major lake systems of an
otherwise arid region. In addition to monitor-
ing dispersal in relation to breeding sites, we
also examined premigratory movements of in-
dividuals, an often neglected aspect of indi-
vidual life histories that may be a critical con-
nective element of patchy landscapes (Haig et
al. 1998).

METHODS

From 1995 to 1997, we color-banded and observed
American Avocets at four major alkali lake systems in
high desert regions of the western Great Basin: Sum-
mer Lake and Lake Abert in Lake County, Oregon;
Goose Lake in Lake County, Oregon and Modoc
County, California; and Honey Lake in Lassen County,
California (Fig. 1). Interlake distances range from 45-
315 km. Avocet breeding locations in the region, other
than our study areas, were scarce during the years of
the study; although additional local breeding popula-

314

Flissner et al. • AVOCET DISPERSAL

315

tions occur sporadically in response to suitable water
conditions (Neel and Henry 1996; Oring and Reed
1996; L. W. O., pers. observ.)- During breeding peri-
ods, our efforts at Summer Lake and Honey Lake were
focused on managed wetlands adjacent to the main
lake bodies: the Summer Lake Wildlife Area (SLWA)
and Jay Dow, Sr. Wetlands (JDW), respectively, al-
though we also conducted regular surveys along the
entire lake shorelines.

During the three breeding seasons, 339 incubating
adults were captured at nests and given unique com-
binations of color bands. Recently hatched chicks (n
= 457) and flightless fledglings {n = 19) were captured
and individually marked opportunistically near nesting
areas. In 1995, an additional 61 individuals in post-
breeding flocks were captured and banded at Summer
Lake. All birds were given U.S.D.I. numbered alumi-
num and celluloid or Darvic® plastic leg bands. In
1996 and 1997, radio transmitters were attached to alu-
minum leg bands and placed on 185 adult avocets,
distributed among the four lake systems (Plissner et al.
in press). Hatch-year birds were given brood-specific
band combinations, with individuals identified by col-
ored plastic tape wrapped around the numbered alu-
minum band. Sexes of adults were distinguished by
relative bill curvature (Hamilton 1975) whenever pos-
sible but were deferred to judgments during banding
in cases of conflict with observations of resighted in-
dividuals.

Banded avocets were resighted during the breeding
and post-breeding periods at each lake, from April
through September of each year. The Summer Lake
WA and Jay Dow, Sr. Wetlands were surveyed weekly
throughout the field season. Because of difficulties in
accessing most of the lakeshores, the four main lakes
were surveyed less frequently by foot and from vehi-
cles and hovercraft. In 1996 and 1997, complete sur-
veys of the three Oregon lakes were conducted weekly
during the breeding and post-breeding periods. Aerial
surveys and previous monitoring efforts at Honey
Lake indicated relatively little use of the main lake
body by avocets throughout the breeding and post-
breeding periods (L.W.O., unpubl. data). Nevertheless,
focal lake sites in proximity to Jay Dow. Sr. Wetlands

were regularly surveyed. Observers recorded band
combinations observed with 20-60X telescopes.
Breeding status was assigned to birds observed at nests
or in the presence of young chicks. As a conservative
measure, only birds observed after 1 August were as-
sumed to be post-breeding. We use the term “dispers-
al” to refer to movements between different lake sys-
tems; whereas we refer to birds returning to the same
lake system as “philopatric”. We also report anecdotal
observations of banded avocets from wintering areas
in California.

Resightings were subsequently screened against lists
of known band combinations. Because of occasional
band loss and the fact that some pairs of band colors
became difficult to distinguish as they faded over time
(Robinson and Oring 1997a), alternatives were consid-
ered for those observations that did not correspond to
known combinations. If only a single alternative com-
bination existed, the sighting was retained in the data-
set. Other observations were discarded. Observations
of individuals breeding in different years (bird-years)
were considered independent for all summaries and a-
nalyses. Frequency data were analyzed using G-tests
with Williams correction (Excel macro based on Sokal
and Rohlf 1981).

RESULTS

Within year movements. — During the study
period, we monitored postfledging/postbreed-
ing movements for 476 hatch-year and 339
breeding adult American Avocets (151 fe-
males, 170 males, 18 unknown gender) band-
ed in the western Great Basin. One bird, orig-
inally banded as a chick, was subsequently re-
captured and rebanded as a nesting adult.
Eight individuals were observed at nests dur-
ing two breeding seasons. Of the total moni-
tored, 125 chicks (26%) and 188 adults (55%)
observed during nesting periods were resight-
ed during the subsequent post-fledging/post-
breeding periods (i.e., after 1 August). Among

316

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 1. Post-breeding locations of adult American Avocets in the western Great Basin, 1995-1997.

Breeding location

Post-breeding location^

Adults monitored^

Honey Lake

Goose Lake

Lake Abert

Summer Lake

Honey Lake

105

Goose Lake

Lake Abert

Summer Lake

133

“ Includes seven individuals observed at nests during two breeding seasons.
Multiple post-breeding locations included for 30 individuals.

adults resighted during the post-breeding pe-
riod, 53% were observed at lakes other than
where they had bred (Table 1). There was no
significant difference in proportions of males
(53.5%, n = 99) and females (52.6%, n = 76)
that dispersed from breeding locations to other
lake systems (G^^, = 0.014, P > 0.05). Fur-
thermore, of observed post-fledged young,
30% (38 of 125) were resighted away from
their natal lake systems (Table 2), a significant
difference from the adult rates of post-breed-
ing dispersal (G^,, = 15.59, P < 0.05). Five
chicks and 16 adult breeders dispersed follow-
ing fledging or breeding from Honey Lake to
post-breeding locations among the northern
three lakes, while no birds from the northern
three lakes were observed postbreeding/post-
fledging at Honey Lake. Four chicks and three
adults were reported from wintering areas
along the northern California coast (Humboldt
Bay and San Francisco Bay) after leaving
Goose Lake and Honey Lake natal/breeding
areas. Prior to being reported on the wintering
grounds, all seven birds were observed on at
least one of the three Oregon lakes during the
post-breeding period.

Between-year movements. — We monitored
between-year dispersal locations for 454
chicks and 259 adult American Avocets (137
females, 122 males). Altogether, 197 birds
were observed in the region in multiple years,
including 51% of all banded adults and 8% of

chicks. Of 46 adults that were observed in the
region in consecutive breeding seasons, 38
(83%) returned to the same lake. Eight addi-
tional adults that bred at Honey Lake in 1995
were not observed the following year but re-
turned to breed in 1997. Overall, return rates
for chicks were 0.4% for birds observed in
their first post-natal breeding period (i.e., at
age 1) and 1.3% for birds observed during
their first two post-natal breeding seasons (i.e.,
ages 1 or 2). Of eleven banded chicks resight-
ed during subsequent breeding seasons, seven
(64%) returned to the natal lake system. The
other four chicks, all banded at Honey Lake,
dispersed to Oregon lakes in subsequent
breeding seasons. No adults dispersed from
Honey Lake to the northern lakes between
years, and no birds of any age class from
Goose, Abert, or Summer Lakes dispersed
south to Honey Lake during subsequent
breeding periods. Four birds (three males and
one female) were observed during the breed-
ing period of all three study years. Three of
the four were observed at Honey Lake during
all three years. The other individual bred at
Summer Lake in 1995 and 1996 and was ob-
served at Lake Abert during the breeding pe-
riod in 1997. Only four individuals that bred
in 1996 and 1997 were observed during the
post-breeding period of the previous year. One
of the four nested in 1997 at Summer Lake
after nesting and spending the post-breeding

TABLE 2. Post-Hedging locations of American Avocets in the we.stern Great Basin, 1995-1997.

Natal location

Po.st-fledging location

Chicks monitored

Honey Lake

Goose Lake

Lake Abert

Summer Lake

Honey Lake

201

Goose Lake

169

Lake Abert

Summer Lake

Flissiier el al. • AVOCET DISPERSAL

317

period at Lake Abert in 1996. The other three
bred and spent the post-breeding period at the
same lake in both years.

DISCUSSION

Our results demonstrate within and between
year patterns of dispersal at a regional scale.
Earlier studies of movement patterns of avo-
cets and other recurvirostrids focused on rates
of philopatry and dispersal of individuals
within a single wetland (Cadbury and Olney
1978, Watier and Fournier 1980, Sordahl
1984, Cadbury et al. 1989, James 1995, Rob-
inson and Oring 1997b) or relied upon occa-
sional surveys and anecdotal reports of
marked individuals from areas away from the
study area (Robinson and Oring 1996). By
regularly monitoring birds at multiple, distant
lake systems, we were better able to define the
movement patterns and extent of movements
at a scale more appropriate to the life history
of the species. Still, we recognize the fact that,
in order to gain an objective measure of dis-
persal, monitoring of more distant sites in
multiple directions from breeding areas would
be required.

Return rates of individuals we observed dif-
fered somewhat from those described for av-
ocets in an earlier Great Basin study at Honey
Lake, California. Robinson and Oring (1997b)
estimated that 21-25% of avocets chicks that
were known to have survived to breeding age
returned to the natal site to breed, whereas we
observed 64% of known survivors (i.e., indi-
viduals observed after their hatch-year) re-
turning to the natal lake system. In relation to
the total number of chicks banded, however,
our return rates were lower (4.2%, Robinson
and Oring 1997b; 1.3%, this study), suggest-
ing that survivorship of chicks was likely the
distinguishing variable. Return rates of breed-
ing adults were significantly different in the
two studies (24%, Robinson and Oring 1997b;
18%, this study; = 4.24, P < 0.05). Rel-
ative to other species of Charadriiformes (Or-
ing and Lank 1984), natal and breeding phil-
opatry rates of avocets in the western Great
Basin were low, perhaps resulting from the
extensive annual variability in breeding habi-
tat suitability.

The small number of interlake breeding and
natal dispersal events during the three years
of the study suggests limited, but adequate

(Wright 1951), gene fiow among Great Basin
breeding populations. Furthermore, the pattern
of both post-breeding and between year dis-
persal suggests an unbalanced pattern of con-
nectivity, perhaps indicative of source-sink
metapopulation dynamics (Pulliam 1988),
with Honey Lake/Jay Dow, Sr. Wetlands serv-
ing as a source population in relation to the
three Oregon lakes. This hypothesis is further
supported by observations that productivity at
Jay Dow, Sr. Wetlands has generally been
higher than at the other study areas (S.M.H.
and L.W.O., unpubl. data). Historically, in a
system with high interannual variability in
habitat patch quality resulting from fluctua-
tions in precipitation (Engilis and Reid 1997,
Robinson and Warnock 1997), sources and
sinks may have shifted periodically. Such
shifts would characterize a rescue effect me-
tapopulation (Brown and Kodric-Brown
1977), in which local populations, in danger
of extinction if isolated, nevertheless persist
as their numbers are buffered by immigration
from populations with more favorable breed-
ing conditions (Stacey et al. 1997). As a re-
cently developed, managed wetland, the Jay
Dow, Sr. Wetlands may in fact now be a stable
source for avocet populations at other wetland
systems throughout the western Great Basin.

Post-breeding movements also suggested
seasonal differences in the use of various wet-
lands within the region. Primary breeding ar-
eas such as the managed wetlands (Jay Dow,
Sr. Wetlands and Summer Lake WA) were rel-
atively less important as post-breeding habi-
tats, whereas the major lake bodies, particu-
larly the three northern waterbodies in our
study (Goose Lake, Lake Abert, and Summer
Lake), supported much higher densities of
birds in late summer and early fall than during
the breeding season (Warnock et al. 1998).
Our observations of post-breeding dispersal
from Honey Lake to northern lakes may re-
flect differential habitat suitability of the lake
systems during this time period. Avocets from
Honey Lake also have been reported in post-
breeding flocks at water bodies east and south
of the breeding/natal lake system (Robinson
and Oring 1996, 1997b; Plissner et al. in
press); indicating a multidirectional exodus
from this particular breeding area in late sum-
mer and early fall. As an alternative to hy-
potheses of seasonal resource tracking, other

318

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

studies have suggested that premigratory
movements of birds are exploratory or other-
wise associated with identification of future
breeding areas (Morton et al. 1991, Reed and
Oring 1992, Reed et al. in press). Our results,
however, do not provide evidence for associ-
ations between post-breeding locations and
subsequent nest sites, as would be expected
for support of this hypothesis.

The observed patterns of dispersal move-
ments were confirmed by radio telemetry data
(Plissner et al. in press), which also demon-
strated a northerly trend in post-breeding
movements of avocets between Carson Lake,
Churchill County, Nevada (Fig. 1) and the
three Oregon lakes of this study, without ev-
idence of southward post-breeding move-
ments by Oregon breeders. Telemetry data
further indicated that a very high proportion
of post-breeding adults (71%) frequented wet-
lands other than their breeding locations. In
addition, radio-tagged individuals frequently
were not detected at any of the primary lakes
during some telemetry surveys, suggesting
even greater rates of movements away from
breeding locations during the time period.
Therefore it is likely that visual resightings of
individuals represent a conservative estimate
of the extent of post-breeding movements in
the region. Nevertheless, these data provided
information on movements across multiple
years and for first-year birds, which can not
be obtained effectively using current teleme-
try methodologies.

Post-breeding locations may also be asso-
ciated with specific migration routes and/or
wintering areas for these populations. All
birds reported during the fall/winter along the
northern California coast, including those
originating from Honey Lake, were last ob-
served at one of the three northern lakes. Pre-
vious reports suggest that birds originating
from Honey Lake primarily migrate to win-
tering areas in California’s Central Valley and
coastal areas from San Francisco Bay south to
the central coast of Sinaloa, Mexico (Robin-
son and Oring 1996, Robinson et al. 1997).
Numbers of avocets at the northern extreme
of the species’ winter range (Humboldt Bay,
California), however, have been increasing
since I960 (Evans and Harris 1994), and it is
clear that many of these birds originated from
breeding areas throughout the western Great

Basin. Although further data are needed from
other wintering areas, it appears possible that
birds that aggregate at post-breeding areas in
the northwestern Great Basin may overwinter
along the northern California coast, whereas
others, even from the same breeding area, may
migrate directly south and west to more south-
erly post-breeding and wintering sites.

Our studies of American Avocet move-
ments in the western Great Basin indicate that
a dispersal-based evaluation of habitat con-
nectivity requires an understanding of move-
ments at multiple temporal and spatial scales.
Based solely upon movements of individuals
between breeding sites, connectivity among
the different wetlands of the region would ap-
pear to be weak. Rates of post-breeding move-
ments among different lake systems, however,
was substantially higher, providing evidence
of a strong link between the different systems.
An apparent northward trend in long-distance
post-breeding dispersal suggests that the
northern lake systems may provide better re-
sources for avocets during this time period,
while the same trend in between-year move-
ments may be simply a geographic artifact of
the location of a source population relative to
other study populations with lower productiv-
ity. This hypothesis also is supported by ob-
servations of avocets from Jay Dow, Sr. Wet-
land at stopover areas farther south (Robinson
and Oring 1996), suggesting a multi-direction-
al exodus of post-breeding birds from Honey
Lake breeding sites. Thus, avocets use a large
array of Great Basin wetlands within and
among years, suggesting that conservation ef-
forts should consider this complexity in defin-
ing appropriate habitat conservation strate-
gies.

ACKNOWLEDGMENTS

Funding for this project was provided by the Bio-
logical Resources Division of the U.S. Geological Sur-
vey, the USGS Forest and Rangeland Ecosystem Sci-
ence Center, a USDA Hatch grant to L.W.O. via the
University of Nevada-Reno, NSF grant DEB 9424375
to L.W.O., and USDA grant 95-37101-2026 to L.W.O.
and M. Rubega. Over 30 field assistants were indis-
pensible in gathering the data during the years of the
study. We also thank the 12 private landowners who
provided access to properties adjoining many of the
wetlands. M. Colwell and C. Gratto-Trevor provided
reports of banded birds observed along the California
coast during the non-breeding season. Additional sup-

Flissiier et al. • AVOCET DISPERSAL

319

port was provided by M. St. Louis, O. Tat't, Oregon

Department of Eish and Wildlife, the Jay Dow, Sr.

Wetlands, and the University of Nevada-Reno. We

thank C. Gratto-Trevor and two anonymous reviewers

for their helpful comments.

LITERATURE CITED

Barrowclough, G. E 1978. Sampling bias in dispersal
studies based on finite area. Bird-Banding 49:333-
341.

Brown, J. H. and A. Kodric-Brown. 1977. Turnover
rates in insular biogeography: effect of immigra-
tion on extinction. Ecology 58:445-449.

Cadbury, C. J., D. Hill, J. Partridge, and J. Sor-
ensen. 1989. The history of the Avocet population
and its management in England since recolonisa-
tion. R. Soc. Prot. Birds Conserv. Rev. 3:9-13.

C.xdbury, C. j. and P. j. S. Olney. 1978. Avocet pop-
ulation dynamics in England. British Birds 71:
102-121.

Cunningham, M. A. 1986. Dispersal in White-crowned
Sparrows: a computer simulation of study-area ef-
fects on estimates of local recruitment. Auk 103:
79-85.

Doebeli, M. and G. D. Ruxton. 1997. Evolution of
dispersal rates in metapopulation models: branch-
ing and cyclic dynamics in phenotype space. Evo-
lution 51:1730-1741.

Dunning, J. B., Jr., D. J. Stewart, B. J. Danielson,
B. R. Noon, T. L. Root, R. H. Lamberson, and
E. E. Stevens. 1998. Spatially explicit population
models: current forms and future uses. Ecol. Appl.
5:3-1 1.

Engilis, a., Jr. and E A. Reid. 1997. Challenges in
wetland restoration of the western Great Basin.
Int. Wader Stud. 9:71-79.

Evans, T. J. and S. W. Harris. 1994. Status and hab-
itat use by American Avocets wintering at Hum-
boldt Bay, California. Condor 96:178-189.

Haig, S. M., D. W. Mehlman, and L. W. Oring. 1998.
Avian movements and wetland connectivity in
landscape conservation. Conserv. Biol. 12:749-
758.

Hamilton, R. B. 1975. Comparative behavior of the
American Avocet and the Black-necked Stilt (Re-
curvirostridae). Ornithol. Monogr. 17:1-98.

IMS, R. A. AND N. G. Yoccoz. 1997. Studying transfer
processes in metapopulations: emigration, migra-
tion, and colonization. Pp. 247-265 in Metapopu-
lation biology: ecology, genetics, and evolution (I.
A. Hanski and M. E. Gilpin, Ed.) Academic Press,
San Diego, California.

James, R. A., Jr. 1995. Natal philopatry, site tenacity,
and age of first breeding of the Black-necked Stilt.
J. Field Ornithol. 66:107-1 1 I.

McPeek, M. a. and R. D. Holt. 1992. The evolution
of dispersal in spatially and temporally varying
environments. Am. Nat. 140:1010-1027.

Morton, M. L., M. W. Wakamatsu, M. E. Pereyra,
AND G. A. Morton. 1991. Postfledging dispersal.

habitat imprinting, and philopatry in a montane,
migratory sparrow. Ornis Scand. 22:98-106.

Neel, L. A. and W. G. Henry. 1997. Shorebirds of
the Lahontan Valley, Nevada, USA: a case history
of western Great Basin shorebirds. Int. Wader
Stud. 9:15-19.

Oring, L. W. and D. B. Lank. 1984. Breeding area
fidelity, natal philopatry, and the social systems of
sandpipers. Pp. 125-147 in Shorebirds: breeding
behavior and populations (J. Burger and B. L.
Olla, Ed.). Plenum Publishing, New York.

Oring, L. W. and J. M. Reed. 1997. Shorebirds of the
western Great Basin of North America: overview
and importance to continental populations. Int.
Wader Stud. 9:6-12.

Plissner, j. R, S. M. Haig, and L. W. Oring. In press.
Post-breeding movements of American Avocets
and wetland connectivity in the western great ba-
sin. Auk.

Pulliam, H. R. 1988. Sources, sinks, and population
regulation. Am. Nat. 132:652-661.

Reed, J. M., T. Boulinier, E. Danchin, and L. W.
Oring. In press. Informed dispersal: prospecting
by birds for breeding sites. Curr. Ornithol. 15: In
press.

Reed, J. M. and L. W. Oring. 1992. Reconnaissance
for future breeding sites by Spotted Sandpipers.
Behav. Ecol. 3:310-317.

Robinson, J. A. and L. W. Oring. 1996. Long-distance
movements by American Avocets and Black-
necked Stilts. J. Field Ornithol. 67:307-320.

Robinson, J. A. and L. W. Oring. 1997a. Fading of
UV-stable coloured bands on shorebirds. Wader
Stud. Group Bull. 84:45—46.

Robinson, J. A. and L. W. Oring. 1997b. Natal and
breeding dispersal in American Avocets. Auk 1 14:
416-430.

Robinson, J. A., L. W. Oring, J. P. Skorupa, and R.
Boettcher. 1997. American Avocet {Recun’iros-
tra americana). In The birds of North America,
no. 275 (A. Poole and E Gill, Ed.). The Academy
of Natural Sciences, Philadelphia, Pennsylvania;
The American Ornithologists’ Union, Washington,
D.C.

Robinson, J. A. and S. E. Warnock. 1997. The stag-
ing paradigm and wetland conservation in arid en-
vironments; shorebirds and wetlands of the North
American Great Basin. Int. Wader Stud. 9:37-44.

Saunders, D. A., R. J. Hobbs, and C. R. Margules.
1991. Biological con.sequcnces of ecosystem frag-
mentation: a review. Con.serv. Biol. 5:18-32.

SoKAL, R. R. AND E J. Rohlf. 1981. Biometry. W. H.
Freeman and Company. San Francisco, California.

SoRDAHL, T. A. 1984. Observations on breeding site
fidelity and pair formation in American Avocets
and Black-necked Stilts. N. Am. Bird Bander 9:
8-11

Stacey, P. B., V. A. John.son, and M. L. Taper. 1997.
Migration within metapopulations: the impact
upon local population dynamics. Pp. 267-291 in
Metapopulation biology: ecology, genetics, and

320

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

evolution (1. A. Hanski and M. E. Gilpin, Ed.).
Academic Press, San Diego, California.

Taylor, P. D., L. Fahrig, K. Henein, and G. Merri-
AM. 1993. Connectivity is a vital element of land-
scape structure. Oikos 68:571-573.

Warnock, N., S. M. Haig, and L. W. Oring. 1998.
Monitoring species richness and abundance of
shorebirds in the western Great Basin. Condor
100:589-600.

Waxier, J.-M. and O. Fournier. 1980. Elements de
demographie de la population d’Avocettes (Re-
curvirostra avosetta) de la cote atlantique fran-

(jaise. LOiseau Rev. Fran^aise Ornithol. 50:307-
321.

Wiens, J. A. 1997. Metapopulation dynamics and land-
scape ecology. Pp. 43—62 in Metapopulation bi-
ology: ecology, genetics, and evolution (I. A.
Hanski and M. E. Gilpin, Ed.). Academic Press,
San Diego, California.

With, K. A., R. H. Gardiner, and M. G. Turner.
1997. Landscape connectivity and population dis-
tributions in heterogeneous environments. Oikos
78:151-169.

Wright, S. 1951. The genetical structure of popula-
tions. Ann. Eugenics 15:323-354.

Wilson Bull., 111(3), 1999, pp. 321-329

HIGH MORTALITY OF PIPING PLOVERS ON BEACHES WITH
ABUNDANT GHOST CRABS: CORRELATION, NOT CAUSATION

DONNA L. WOLCOTT* 2 AND THOMAS G. WOLCOTT*

ABSTRACT. — Ghost crabs (Ocypode quad rata) have been implicated in mortality of eggs and chicks of the
beach-nesting Piping Plover (Charadrius melodus) whose Atlantic Coast populations are listed as threatened.
Through observation and experimentation, we investigated the interactions between ghost crahs and plovers on
Wild Beach, a Piping Plover nesting area on Assateague Island, Virginia. This site has a high abundance of
ghost crabs and historically low fledging success compared to adjacent areas with fewer crabs. We observed
encounters of crabs with plover eggs, chicks, and adults in the field, but never predation. In staged encounters
of crabs with eggs and chicks (using hatchery reared quail as plover surrogates), we were unable to elicit
predatory behavior either on the beach or in the lab. We conclude that although instances of ghost crab predation
on Piping Plover eggs and chicks occur, they are rare and cannot account for the high mortality frequently
reported on beaches where ghost crabs are abundant. Adult plovers behave toward crabs as if they were dan-
gerous to eggs and chicks, and their young hroods in the study area did not forage along the foreshore. Hence,
ghost crabs may increase mortality indirectly. Frequent responding to crabs by parents may attract more deadly
brood predators. Brood nutrition may suffer as adult plovers direct chicks away from areas where forage is
reportedly richer but crabs are abundant, such as the foreshore. Nutrient intake may be further reduced on more
southerly breeding grounds where high temperatures on backshores force chicks to stop foraging and take shelter
during mid-day. Although high mortality cannot be attributed directly to predation by crabs, it may be due to
factors that covary with crab abundance, such as high temperature, behavioral responses of adult birds, and poor

forage. Received 28 April 1998, accepted 7 Feb. 1999.

Anecdotal and published reports of ghost
crab predation on Piping Plovers {Charadrius
melodus; Loegering et al. 1995, Watts and
Bradshaw 1995) have led to concern that crab
predation may hamper recovery of plovers on
the Atlantic Coast, where the species is listed
as threatened (Loegering and Fraser 1995;
U.S. Fish and Wildlife Service 1993, 1996).
To assess the extent of crab-caused mortality,
we investigated interactions between ghost
crabs {Ocypode quadrata) and Piping Plovers
during incubation and chick rearing on Wild
Beach on Assateague Island, Virginia, within
the Chincoteague National Wildlife Refuge.
Compared to other portions of this barrier is-
land that are used for nesting areas by the Pip-
ing Plover, Wild Beach has higher abundances
of ghost crabs (Britton 1979) and lower rates
of fledging success (U.S. Fish and Wildlife
Service 1994).

Piping Plovers breed from eastern Canada
to North Carolina, as well as in the Great
Lakes region and the Great Plains of Canada
and the U.S. (Haig 1992). On the Atlantic
Coast, plovers typically lay four eggs in a

' Dept, of Marine, Earth and Atmospheric Sciences,
Box 8208, North Carolina State Univ., Raleigh, NC
27695-8208.

^ Corresponding author; E-mail: dwolcott@ncsu.edu

shallow scrape in the sand, usually well be-
yond the high-tide mark, or in shelly storm-
flattened areas (washouts) between and behind
the primary dunes. Chicks are precocial and
forage in moist backshore areas where avail-
able, or on the foreshore (the area between the
tides; Loegering and Fraser 1995).

Ghost crabs, named for their cryptic col-
oration, range along the Atlantic Coast from
Rhode Island to Brazil and throughout the Ca-
ribbean (Chace and Hobbs 1969). They are
among the fastest terrestrial invertebrates (Full
and Weinstein 1992) and formidable predators
with acute sensory receptors for vision, vibra-
tion, taste, and smell (Cowles 1908, Wellins
et al. 1989). They are most abundant on high
energy beaches, where they obtain over 90%
of their diet preying on intertidal invertebrates
(Wolcott 1978). They are extremely flexible
foragers, also scavenging, deposit feeding,
consuming seeds and insects, and are docu-
mented predators of turtle hatchlings (Arndt
1994, Robertson and Pfeiffer 1982). At dusk
crabs move from their burrows on the back-
shore and among the dunes to feed in the
swash zone. Crabs seek out and take refuge in
burrows as dawn approaches (Wolcott 1978).

Poor fledging success on beaches where
crabs are abundant, coupled with extensive

321

322

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

seasonal and spatial overlap between plovers
and the predatory ghost crabs, led to the hy-
pothesis that ghost crab predation is a com-
mon source of mortality for plover eggs and
chicks (U.S. Fish and Wildlife 1993, Loeger-
ing and Fraser 1995). To test this hypothesis,
we documented and quantified natural en-
counters between crabs and Piping Plover
eggs and chicks, and staged encounters be-
tween crabs and the eggs and chicks of non-
threatened species.

METHODS

We conducted our study on a 6.4 km stretch of the
Chincoteague National Wildlife Refuge’s (NWR) Wild
Beach, Virginia, in June and July, 1994. Wild Beach
has a steeply-sloped foreshore rising from a high en-
ergy surf zone to the berm and was characterized in
1994 by one or two wave-cut scarps less than 1 m
high. At the berm the slope decreased abruptly and
there was a narrow (20-50 m), flat backshore region.
Inland of the backshore are low vegetated foredunes,
fronting 2-4 m high artificially stabilized dunes veg-
etated with beach grass (Ammophila breviligulala) and
seaside goldenrod (Solidago semperviren.s).

We observed four natural plover nests through the
last days of incubation and past hatching using contin-
uous video monitoring to document crab predation on
eggs and newly hatched chicks. When the third egg
was laid in each nest, staff from the refuge surrounded
the nest with a 4 m diameter, 1.5 m high exclosure
constructed of 5 X 10 cm welded mesh wire supported
by reinforcing bar and covered with plastic netting. A
dummy camera was mounted at the top of an exclosure
support bar on the southwest side, facing toward the
nest and away from the prevailing wind. At the time
nest observation began, the dummy camera was re-
placed with a similar appearing video surveillance
camera fitted with infrared light-emitting diodes for
nocturnal illumination, and an infrared filter to prevent
saturation in strong sunlight. The camera imaged an
area approximately 2 X 3 m centered on the nest. In-
stalling the camera in place of the dummy took 3-12
min, and birds returned to the exclosure 1-9 min there-
after (mean = 4.25 min., n = 4). Coaxial and power
cables (150 m) led to a TV/VCR (Magnavox model
CCR095) powered by a deep-cycle lead-acid battery.
These were housed in a tent behind nearby dunes to
reduce disturbance to the plovers. Tapes were changed
at six-hour intervals and the battery every 18 hours.
According to the video record, the incubating adult
plovers typically did not flush from the nest at those
times. Video tapes were reviewed for the following
occurrences and the times logged: adult plover brood-
ing, alarming and flushing from the nest (rapid rising
off eggs, standing near nest with wing raises, rapid exit
from the field of view), parental exchanges of incu-
bating/brooding duty (second parent appears on screen
and changes places with the bird on the nest), eggs

hatching, activities of chicks (including walking, for-
aging and being brooded), and any activities of ghost
crabs.

To explore the role of crab predation after chicks
left the nest, we observed encounters between crabs
and adult plovers and chicks for eight days, concen-
trating on periods of peak crab activity. Observations
of hatched broods were made from a vehicle at least
30 m away, using binoculars during daylight hours,
and an image-intensifying scope (Varo Noctron IV) at
night fitted with an infrared diode laser to enhance
illumination. Plover chicks in two broods were marked
on each thigh with 10 mm diameter disks of Scotch-
lite® (3M) reflectorized tape glued to the surface of
their down to make them visible with night vision
equipment. Two of the marked chicks were the only
chicks to fledge in the study area; thus the treatment
did not appear to increase mortality. Data collected
during each observation period included the location
of the brood, the place and time spent foraging, and
the location of brooding.

To further assess the probability of ghost crabs prey-
ing on eggs, we presented crabs with surrogate eggs
that were similar in size and shape to those of Char-
adrius melodus (see Macivor et. al. 1990). Japanese
Quail (Coturni.x japonicu.s) eggs were obtained from a
local bird breeder, and Northern Bobwhite (Colinus
virginianu.s) eggs were obtained from Seven Oaks
Game Earm and Supply, Wilmington, North Carolina.
On three nights, we constructed four scrapes near the
berm and placed four Cotiimix eggs in each just prior
to the time that crabs emerge from their burrows. Dur-
ing the nocturnal peak in crab activity (19:30-22:00
EST), the artificial nests were observed for any ghost
crab encounters using the video camera unit fixed on
a tripod.

To further explore the vulnerability of eggs to ghost
crab predation, we placed four opaque gray plastic bins
(38L X 18W X 1 ID cm) in a rectangular array in the
laboratory and filled them to a depth of about 3 cm
with damp sand from the foreshore. In each we placed
a ghost crab [average weight 49.0 ± 5.4 (SD) g, with
a carapace width of 43.2 ± 1.6 mm, n = 8] freshly-
collected from the foreshore of Wild Beach, two Co-
turni.x eggs, and the bins were covered with chicken
wire. Crab activity was recorded using the infrared
camera and VCR from the initiation of the experiment
at 23:15 until 05:15, and condition of the eggs was
assessed at 10:00. The experiment was repeated using
dry beach sand and Colinu.s eggs on another night, and
condition of the eggs assessed after I 1 hours of ex-
posure to the crabs. From the video tapes we noted the
length of time crabs spent in contact with eggs, their
behavior, and the condition of the eggs after manipu-
lation by the crabs.

To elucidate potential predatory interactions be-
tween crabs and chicks, we used chicks of Northern
Bobwhite as surrogates for Piping Plover chicks. Bob-
white chicks are similar in size and behavior to Piping
Plover chicks, but darker and hence less cryptic on
beach sand. Two- and three-day-old chicks (u = 14)

Wolcotl ami Wolcott • PIPING PLOVERS AND GHOST CRABS

323

were released near the beach berm into areas of high
crab activity to maximize encounters with crabs and
were observed with binoculars as they wandered on
the beach. In four cases, chicks were deployed from
the parked vehicle (to which crabs showed little re-
sponse) into the immediate vicinity of active crabs. To
verify that crabs were motivated to forage, we placed
dead chicks, freshly cracked oysters, and pieces of
thawed chicken upwind near the mouths of occupied
burrows at the same time that live chicks were released
onto the beach. Encounters between crabs and chicks
and dead prey were documented.

To determine temporal overlap in activity between
crabs and plovers, nocturnal and diurnal observations
of crab activity and behavior were conducted from a
vehicle. The onset of migration of crabs to the surf
zone for nocturnal foraging was determined by noting
the time of appearance of the first crab moving down
the beach each night.

Statistics are reported as mean plus or minus one
standard deviation.

RESULTS

Predation by crabs on eggs. — Ghost crabs
did not prey on eggs in either natural nests or
in experimental trials, and crabs showed com-
parable behavior toward eggs in all settings.
In 147 hours of video observations of incu-
bation by Piping Plovers, there were seven ap-
pearances of ghost crabs in the video field.
These appearances generally were confined to
the periods of crepuscular movement by the
crabs (dawn, 04:16 ± 11.3 min, n = 2 and
dusk, 20:01 ± 4.6 min, n = 3). Two excep-
tions occurred on one rainy day when high
humidity and reduced insolation resulted in
diurnal crab activity. In two cases, crabs di-
rectly contacted and manipulated the eggs in
the nest cup after the incubating adult had
flushed from the nest. Crabs appeared to be
testing the eggs as potential prey, using chelae
and mouthparts which contain dense arrays of
chemoreceptors to “taste” the eggs. The du-
rations of manipulations were 13 s and 23 s.

Crabs that contacted Coturnix eggs in pseu-
do-nests in the field either continued slowly
toward the foreshore or stopped to manipulate
and taste the eggs. Crabs spent more time in
contact with eggs when manipulating them
(17.3 ± 3.06 s, n = 3) than when they were
not manipulating them (1 1.4 ± 8.4 s, n — 5).
In one instance, a crab spent 6 minutes and
12 seconds at a pseudo-nest of Coturnix eggs
on the beach, repeatedly tasting the eggs and
rolling one egg several cm prior to moving on
to the surf zone. We interpret this intense in-

terest as a response to the bird feces present
on that particular egg.

Crabs confined with Colinus eggs in the
laboratory explored the eggs as potential prey
at least once, not necessarily on the first con-
tact, but simply walked over the eggs in other
encounters. Four crabs spent an average of
12.2 s (± 8.2 s, n = 11), 10.6 s (± 11.2 s, n
= 8), 6.9 s (± 6.7 s, n = 10), and 7.3 s (±
5.9 s, n = 6) on eggs per encounter. Average
time of encounters in which manipulation oc-
curred (all crabs) was 13.7 s (± 8.4 s, n = 7).

Potential for predation by crabs on
chicks. — Three different Piping Plover broods
were observed for a total of 26.2 h. A chick
was seen to pass near a crab only once, and
in that instance, there was no response from
the crab.

After leaving the nest and exclosure, chicks
were almost always found foraging in vege-
tated areas, principally on the low foredunes
(ji = 18). Chicks from Nest Five were found
on high dunes because low foredunes were
uncommon in their territory. This brood was
seen once on the foreshore of Wild Beach on
the day after it left the exclosure. Surviving
chicks from all broods were taking 5-10 m
forays away from the attending parent by 1.5
days, often in divergent directions. The at-
tending adult was often seen following after a
rapidly moving chick. Typically the attending
adult stayed near the chicks, outside the veg-
etation, while the other adult stationed itself
near the berm, standing sentinel and frequent-
ly feeding on the foreshore. During midday
chicks alternated between foraging and resting
in the shade of vegetation. The parent typi-
cally moved the chicks back to the same gen-
eral area each evening to brood. Typical
brooding areas were in shell hash (broken
shell fragments) on the backshore near fore-
dunes, a few meters from dune vegetation.

Activity periods of birds and crabs showed
little overlap once the chicks left the nest. Par-
ent birds began brooding chicks as each even-
ing became cool, from about 18:30 to shortly
before 20:00 (19:05 ± 28 min, n = 8), until
chicks ceased making forays altogether, typi-
cally between 19:00 and 20:30 (20:00 ± 38
min, n = 8). First sightings of crabs moving
to the foreshore occurred about the same time
that chick activity fully ceased (19:51 ± 9
min, n = 7 nights). Crab activity increased

324

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

during the following hour, and chicks were
brooded during the peak of activity. In the
morning, crabs migrated from the foreshore to
their burrows before dawn (04:16 ± 1 1.3 min,
n = 2 mornings), and before broods were seen
to become active. Some crab activity persisted
throughout the day, generally restricted to the
vicinity of burrows whose densities increased
from the dunes toward the berm.

No instances of crab predation on bobwhite
chicks occurred, although more than 30 en-
counters were observed in 12 hours of diurnal
and nocturnal observation. In the one case in
which a crab seized a quail chick that was
precipitously deposited next to its burrow, the
crab promptly released the chick unharmed
and retreated to its burrow. Most commonly,
crabs showed no response to chicks that wan-
dered nearby within visual range. Crabs were
attracted to, and fed readily on, cracked oyster
and chicken during the same intervals in
which chicks were presented, demonstrating
the crabs’ willingness to forage and feed.
However, crabs routinely ignored dead quail
chicks, even when they physically contacted
them in the course of foraging.

Crab interactions with adult Piping Plo-
vers.— Interactions between incubating adult
Piping Plovers and approaching crabs were
variable. Plovers either remained on the nest
{n = 4) or flushed in = 4). Flushing typically
occurred while the crabs were further than 0.5
m from the nest cup (n = 3), but once not
until the crab approached within 10 cm. De-
fensive encounters initiated by the plovers
could involve both parents (n — 2). In the four
instances in which birds responded to crabs,
the minimum length of engagement in the vid-
eo field was 2 min (120 ± 49 s, n = 4). How-
ever, adults left the nest cup unattended for
about 5 min (307 ± 197 s, /? = 4), presumably
continuing the defense out of the camera’s
view.

Aggressive displays by incubating adult
plovers against approaching crabs were large-
ly ineffectual. Crabs generally remained mo-
tionless or maintained course (n = 3) when
confronted by adult plovers advancing slowly
with uplifted wings, but sometimes they ran
when charged by a displaying bird (n = 2).
In one instance, a crab whose course would
have bypassed the nest was deflected onto the
nest while veering from the displaying parent.

Thirteen encounters between crabs and
adult plovers with unfledged chicks were ob-
served, with variable behavior by the plovers.
Once, an adult passed within 10 cm of a crab
with no apparent response. Where interactions
occurred (n — 12), plovers always initiated
them, although the crab was 8-50 m from the
brood. Plovers would approach and display
within 10-20 cm of the crab. Birds were seen
feigning and leading crabs toward the fore-
shore (max distance = 10 m, max time = 4
min, n = 3). Three encounters involved both
parents.

Adult plovers appear to associate the pres-
ence of burrows with ghost crabs. A burrow
that was near a nest was ignored for the first
85 hours of video observation, but was closely
inspected by the adult plovers on 1 1 separate
occasions in the final 1 1 hours of observation
after a crab had approached the nest from that
direction. Burrows were also investigated on
the beach by parents with hatched broods (n
= 2).

Plover behavior and mortality. — Although
we observed no instances of Piping Plover
mortality directly attributable to predation by
ghost crabs, we documented other factors that
might directly or indirectly contribute to the
low fledging success on Wild Beach. Hatching
asynchrony (substantial time, e.g., >24 hrs
between hatching of the first and final eggs in
a given clutch) was responsible for the only
mortality for which a cause could be estab-
lished, and the only mortality to occur prior
to hatching. Hatching asynchrony showed a
strong seasonal correlation in this study. Six
clutches of eggs were laid on Wild Beach in
1994 (Table 1). For the two monitored broods
in June, hatching was highly synchronous.
The time between the hatching of the first egg
and the last egg in an entire clutch averaged
104 min. In the three July broods, duration of
hatching averaged at least 1680 min. The
asynchrony 'Contributed directly to the aban-
donment of an egg, which was determined to
be viable (Refuge Staff pers. comm.).

Of 23 chicks that hatched on Wild Beach
in 1994, two fledged. Half the mortality oc-
curred in the first two days, and 75% by day
five (Fig. 1 ). This pattern of chick loss soon
after hatching is typical for the species, but
even more pronounced than reported in pre-
vious studies (most mortality in the first 10

Wolcott and Wolcott • PIPING PLOVERS AND GHOST CRABS

325

TABLE 1. Data tor broods of Piping Plover {Cliaradrins melodns) hatching on Wild Beach, Assateague
Island, 1994. Based on six clutches laid on Wild Beach; nests 2, 3, 5 and 6 were monitored via video camera.

Nest 1

Nest 2

Nest 3

Nest 4

Nest .S

Nest 6

Hatch date

June 10

June 16

July 4-

6 July 6/7

July 8/9

Hatch duration

20 min

3 hr 08 min

>36 hr

>24 hr

Interval to walk

1 hr

1-2.5 hr

6 hr

5 hr

Time nest abandoned

1 ?

18:49

16:10

16:37

15:44-18:45

Survival, days (time

of death where known")

chick A

0.5 (n)

1.5

1-2 (d)

0.5 (n)

chick B

0.5 (n)

1.5

3-4 (d)

0.5-1. 5 (n)

chick C

1 (d)

fledged

4 (d)

1-2

chick D

1.5

fledged

never hatched

4-5

“ Where frequency of observation permits, the time of disappearance of chicks is given as day (d) or night (n).

days; Macivor 1990; Patterson 1988; U.S.
Fish and Wildlife Service 1993). Mortality of
chicks in all the June broods was highly con-
centrated in the first 48 hours (67%), with
25% occurring during the first night. In the
monitored July broods, mortality did not occur
during the first night, but was spread rather
uniformly over the first week.

We could not document the cause of mor-
tality and the fate of “disappeared” chicks.
All mortality on Wild Beach occurred after
broods had left the nest. Where confirmed
sightings allowed the time of disappearance of
chicks to be established with some certainty,
chicks were as apt to vanish during the night
as during the day (Table 1). Five chicks dis-
appeared between sunset and 09:00, and five
between 08:00 and 19:00. Signs of direct pre-
dation were never found.

Differences in diel activity patterns between
June and July broods were documented

through video observation, and indicated dif-
ferences in potential for prey acquisition and
for practicing locomotor and feeding behav-
iors during the first day of life. Hatching was
highly synchronous in June, but asynchronous
in July (Table 1). Chicks emerged from the
nest cup for brief excursions as soon as 1 hour
after hatching in the June broods, but after 5-
6 hours in the July broods (Table 1). For June
broods (2 and 3), chicks spent extended pe-
riods in the shade of vegetation near the nest,
or out of the camera’s field of view, before
returning to the brooding parent, and were ac-
tive throughout the day. July chicks spent little
time away from the nest and were continually
brooded during the hot midday.

Newly hatched chicks in all broods left the
exclosure between 16:00 and 19:00 (Table 1).
The chicks ranged from 4 to 12.5 hours in age,
and exhibited obvious age-related differences
in coordination for walking and running, both
within and between broods.

Once the nest was abandoned, it was no
longer possible to observe broods by using
video. However, in one of the asynchronous
broods, detailed observation of brooding and
foraging of day-old chicks was made by video
on three chicks that spent their first day post-
hatch (July 7) in the area of the nest while the
fourth egg was still being incubated. There
was an initial peak in activity of the chicks
(50% of each hour spent moving about) be-
tween 05:00 and 07:00 and another (30%) af-
ter 14:00. Activity remained below 10% be-
tween 09:00-13:00, and fell to zero at 12:00
when chicks were continuously brooded. For-
aging, as seconds per hour that a chick spent

326

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

in hunt and peck behaviors, was congruent
with the activity pattern (300 s/hr 05:(X)-07:00,
<10 s/hr 11:00-12:00).

DISCUSSION

Crab predation on Piping Plover chicks. —
Ghost crabs do not appear to be directly re-
sponsible for the poor breeding success of
Piping Plovers on Wild Beach on Assateague
Island. Ghost crabs showed no predatory re-
sponse to nearby chicks of Piping Plover or
Northern Bobwhite, and occasionally even ran
away from chicks. Since crabs that were ac-
tive during daylight hours readily fed on other
prey, we conclude that chicks (at least North-
ern Bobwhite) are not preferred prey.

Beyond the crabs’ apparent lack of interest
in chicks, there is temporal and spatial sepa-
ration in crab and chick activity that further
reduces, but does not eliminate the likelihood
of interactions. Ghost crabs are principally
nocturnal; they become active after sunset and
migrate to the foreshore to prey on macroin-
vertebrates. Crabs leave the foreshore and re-
turn to burrows by dawn. During the early
morning hours, they may be active near their
burrows, spending time on burrow repair, de-
fense, and intraspecific aggression, until they
are confined below the surface by microcli-
matic conditions. Crab activity is extended on
days with low desiccation risk (pers. obs.). Al-
though breeding adult plovers may forage ex-
tensively on the foreshore at night (Staine and
Burger 1994), unfledged chicks on Wild
Beach are almost entirely diurnal. During the
periods when crabs are most active, chicks are
being continuously brooded, at least through
the first week post-hatching when most chick
mortality occurs. In addition to temporal sep-
aration, spatial separation of chicks and crabs
was also evident. Only once was a brood seen
foraging in the intertidal prior to two weeks
post-hatching; all other foraging occurred
within 5 m of the dunes onto the backshore,
and hence was well inland of the berm and
the densest aggregation of crab burrows.

The possibility of occasional predation on
foraging chicks by crabs cannot be dismissed.
Only a single documented event (Loegering et
al. 1995) has occurred in over 1 16 h of direct
observation in this and other studies (Patter-
.son 1988, 44 h; Loegering 1992, 46 h; this
study, 26 h). Based on our observations, pre-

dation is most apt to occur when a chick star-
tles a crab, especially one that has been re-
cently defending its burrow. Aggression be-
tween crabs peaks during the early morning
hours as they compete for burrows. Land
crabs of several species will jump, either on
potential prey, or as part of aggressive en-
counters (Herreid 1963, Evans et al. 1976,
pers. obs.) On video we observed an hours-
old plover chick careening head first into a
crab burrow. Had the bunow been occupied,
the chick might have been killed. Neverthe-
less, our data indicate that stalking and killing
of chicks by crabs is highly unlikely.

Crab predation on eggs. — Video observa-
tions in our study show that ghost crabs that
make contact with Piping Plover eggs inves-
tigate the eggs as potential prey items, using
stereotypical tasting behaviors. Crabs showed
the same behavior toward surrogate eggs of
Coturnix japonica and Colinus virginianus on
the beach and in the lab. However, the claws
of the largest crabs on Wild Beach do not have
a gape large enough to directly crush a plover
egg, and no predation was observed.

Nevertheless, ghost crabs are confirmed,
though infrequent, predators of Piping Plover
eggs on barrier islands in Virginia (Watts and
Bradshaw 1995; Refuge staff, this study) and
North Carolina (S. Philhower, pers. comm.).
Viable eggs have been found in crab burrows
(S. Philhower, pers. comm.), but from known
crab behavior and observations during this
study, we conclude that ghost crabs are most
likely to attack or manipulate eggs that are
rotting, cracked, or dirty. Crabs use dactyls
and claws for contact chemoreception, and
distance olfaction to track odor plumes to the
source of a smell (Wellins et al. 1989). Ghost
crabs that encounter a large food parcel (e.g.,
a dead fish) typically dig a burrow immedi-
ately adjacent to it, which provides security
for extended scavenging. Crabs that burrow
next to a nest may do so because they identify
plover eggs as potential food. For instance, a
ghost crab burrow was found immediately ad-
jacent to a Piping Plover nest on a beach 10
km south of Wild Beach when refuge staff
inspected the nest after it was abandoned by
the adults after 35 days of continuous incu-
bation (normal development time is 27 d). The
eggs were missing. Excavation of the crab
burrow yielded three of the four eggs, one of

Wolcott ami Wolcott • PIPING PLOVERS AND GHOST CRABS

327

which was emitting a powerful smell. The
missing egg was assumed to have been con-
sumed by the crab. A crab that has experi-
enced eggs as prey may subsequently recog-
nize intact and odor-free eggs as food through
non-associative learning (Evans et al. 1976).
It is unclear whether broken eggs that have
been found in crab burrows were already
cracked, and hence emitting an attractive odor,
or were cracked from hitting each other in the
burrow. Whether instances of nest predation
by ghost crabs were initiated as scavenging or
as predation, the end result is that some viable
eggs are lost to crabs. Nevertheless, we con-
clude that egg predation by ghost crabs cannot
account for poor breeding success of Piping
Plovers on Wild Beach.

Piping Plover response to crabs. — What-
ever the actual threat from ghost crabs, adult
Piping Plovers treat them as potential preda-
tors. We observed 16 instances in which one
or both adults engaged in extensive displays
against crabs, and in 5 cases, the defense left
hatched broods unguarded. Further underscor-
ing the perceived threat from crabs, adult plo-
vers seem able to connect the presence of bur-
rows with ghost crabs, and invest time and
energy in investigating burrows.

The presence of abundant ghost crabs may
create indirect problems for plovers by several
mechanisms related to the adults’ perception
of crabs as potential predators. First, obvious
responses to ghost crabs may alert truly dan-
gerous predators, both avian and mammalian,
to the location of the brood at the very time
the parents are busy elsewhere and leave it
undefended. Second, more frequent alarm and
defense behaviors carry an energetic cost (re-
viewed in Walters 1984). Finally, it is possible
that the abundance of crabs on the backshore
induces the adult Piping Plovers to shepherd
their broods away from the foreshore, where
forage might be more abundant and have a
higher water content. Broods elsewhere on
Assateague Island are routinely taken to the
foreshore (refuge staff, pers. comm.).

Given the minimal direct threat posed by
ghost crabs, and the potential negative con-
sequences of frequent display and restricted
foraging, it seems maladaptive on the part of
the adult Piping Plovers to treat crabs as dan-
gerous predators. Natural selection acting on
adult defensive behavior should have elimi-

nated the behavior if ghost crabs are not sig-
nihcant predators and if engaging in defense
towards ghost crabs increases the likelihood
of predation by other predators. However, lack
of sufficient genetic isolation between birds
breeding in areas with and without ghost crabs
would preclude such selection. Instead, some
Charadriidae appear to recognize several ani-
mal categories, including avian, mammalian
(with a subset of ungulate), reptilian and “oth-
er”. They have evolved unique displays to
each group (reviewed in Gochfield 1984).
They generally distinguish potential predators
from non-predators, especially among birds,
thereby minimizing false alarms (Walters
1990). However, they seem less discriminat-
ing about other intruders, lumping disparate
taxonomic groups into a category of “poten-
tial threat” in an “urgency of response
scheme” (Walters 1990). Perhaps Piping Plo-
vers indiscriminately categorize anything ter-
restrial but “neither a large mammal nor a
snake”, and moving near a nest or near
chicks, as requiring immediate alarm. This
group might include dangerous predators such
as rats or mustelids to which immediate alarm
would be adaptive. If one postulates that or-
ganisms such as crabs and turtles, which do
not pose a significant threat but which do elic-
it alarm responses from parenting plovers (ref-
uge staff, pers. obs.), are lumped into the same
“dangerous predator” category in the alarm
response hierarchy, the apparently maladap-
tive alarming by Piping Plover parents could
be explained. Given the large geographic
range and variety in breeding habitat, with
concomitant and unpredictable variation in the
suite of predators, mounting a defensive dis-
play against anything novel in the area of the
brood might have at least neutral if not ben-
eficial effects on fitness.

Correlations betM’een low fledging success
and high ghost crab abundance. — We hypoth-
esize that high ghost crab abundance and low
fledging success of Piping Plovers have a cor-
relative, not causative, relationship. Three fac-
tors contribute to the correlation: beach and
dune morphology, climate, and parental be-
havior.

Ghost crabs are most abundant on high en-
ergy beaches backed by high dunes. The high
dunes provide overwintering habitat in which
crabs are able to burrow below their lethal iso-

328

THE WILSON BULLETIN • VoL III, No. 3. September 1999

therm (6—8° C) before being blocked by the
water table (T.G.W., unpubl. data), while high-
energy beaches provide habitat for the crabs’
preferred prey (Wolcott 1978). However,
beaches that are backed by dunes and that lack
low-lying moist habitat away from the surf,
may be poorer habitat for raising Piping Plo-
ver broods, even in the absence of crabs. On
Assateague Island, slower growth with con-
comitant reduced survival has been docu-
mented for chicks reared on an ocean beach,
compared with chicks from other areas with
low-lying moist habitat for foraging (Loeger-
ing and Fraser 1995). Prey abundance (mea-
sured in the wrack zone) and foraging rates
were lower on the ocean beach as well. Wild
Beach lacks the low-lying moist areas that are
the major foraging habitat for young plover
broods in the more productive breeding areas
on Assateague. Even the steep intertidal with
its sharp escarpments may pose a physical
barrier to young broods moving to the fore-
shore.

During the same time as our study, over
90% of chicks successfully fledged on a site
approximately 400 m inland from our site,
with extensive moist low-lying forage but no
access to the beach. Similarly 91% fledging
success occurred on a low energy beach at the
southern tip of Assateague Island, with exten-
sive backshore foraging areas. Meanwhile less
than 10% of chicks fledged on Wild Beach, a
percentage similar to its long-term average
(1988-1994, U.S. Fish and Wildlife 1994).
Poor forage is more likely to contribute to the
unusually high chick mortality on Wild Beach
than direct ghost crab predation.

Climate and latitude probably play a role in
the plover-crab relationship. From their north-
ern limit in New Jersey through the southern
limit of plover breeding in North Carolina,
ghost crabs increase in size and abundance
(U.S. Fish and Wildlife Service 1996). Pre-
sumably, at lower latitudes more adult crabs
are able to successfully overwinter, emerge
earlier in the spring, have a longer active sea-
son, and grow to a larger adult size. Higher
abundance of large crabs leads to more en-
counters between birds and crabs, with an in-
crease in adverse indirect effects on Piping
Plovers.

The hotter summer temperatures associated
with lower latitudes may directly affect brood

survival of Piping Plovers. High daytime tem-
peratures may speed the rate of embryo de-
velopment and lead to greater hatching asyn-
chrony (reviewed in Magrath 1990, Shields
1998). Asynchrony may contribute to mortal-
ity directly by causing abandonment of viable,
late eggs, and indirectly, by reducing the co-
operative attendance by adults of the hatched
young. In our study, severe asynchrony also
resulted in broods having chicks of very dif-
ferent locomotor capabilities, which could in-
crease the likelihood that chicks will become
separated from one another and lost.

Desiccation poses an even more immediate
danger to the chicks. Piping Plovers acquire
water from their food and thermoregulate by
panting (Haig 1992). During the critical first
day, broods hatching in hot weather during
our study showed a reduction in foraging
time and activity compared to broods hatch-
ing in cooler weather. Should heat cause a
persistent shift in activity, it implies that dur-
ing hot weather intake of food and its in-
cluded water is lowered at the same time that
evaporative losses become greater. Under-
standing the relationship between elevated
temperatures, asynchrony, and brood survival
is critical to informed management decisions
at the southern end of the Piping Plover’s
breeding range.

The behavior of the adult plovers to the per-
ceived threat of the crabs compounds the
problem of desiccation because the adults ap-
parently restrict their broods to the duneline,
where forage is sparser and drier. On longer
time scales, restricted feeding times and poor
forage will result in slow growth and in-
creased mortality (Loegering and Fraser
1995). The threat perceived by adult plovers
of ghost crabs on the Wild Beach may act as
a barrier between broods and the richer food
resources of the foreshore.

ACKNOWLEDGMENTS

Financial support for the study came from the U.S.
Biological Survey. Logistical support was provided by
Chincoteague National Wildlife Refuge and Virginia
Department of Game and Inland Fisheries. Special
thanks to A. Hecht, J. Schrocr and his staff, R. Cross,
K. Terwilliger for help with Scotch-liting® chicks, D.
Turner for donating quail eggs, and J. Walters for help-
ful comments on the manuscript.

Wolcoit and Wolcott • PIPING PLOVERS AND GHOST CRABS

329

LITERATURE CITED

Arndt, R. G. 1994. Predation on hatchling diamond-
back temipin, Mulaclemys terrapin (Schoepff), by
the ghost crab. Ocypode qiiadrata (Fabriciiis) II.
Fla. Sci. 57:1-5.

Britton, E. E. 1979. Evaluation of public use impacts
upon nesting shorebirds and the beach habitat on
Chincoteague National Wildlife Refuge. U.S. Fish
and Wildlife Service, Chincoteague National
Wildlife Refuge, Chincoteague, Virginia.

Chace, F. a. and H. H. Hobbs, Jr. 1969. The fresh-
water and terrestrial crabs of the West Indies with
special reference to Dominica. U.S. Natl. Mus.
Nat. Hist. Bull. 292:1-258.

Cowles, R. P. 1908. Habits, reactions, and associations
in Ocypode arenaria. Pap. Carnegie Inst. Wash.
2:3-41.

Evans, P. R. and M. W. Pienkowski. 1984. Population
dynamics of shorebirds. Behav. Mar. Anim. 5:91.

Evans, S. M., A. Cram, K. Eaton, R. Torrance, and
V. Wood. 1976. Foraging and agonistic behavior
in the ghost crab Ocypode kuhlii de Haan. Mar.
Behav. Physiol. 4:121-135.

Full, R. J. and R. B. Weinstein. 1992. Integrating the
physiology, mechanics and behavior of rapid run-
ning ghost crabs: slow and steady doesn’t always
win the race. Am. Zool. 32:382-395.

Gochfield, M. 1984. Antipredator behavior: aggres-
sive and distraction displays of shorebirds. Behav.
Mar. Anim. 5:289-377.

Haig, S. M. 1992. Piping Plover. In The birds of North
America, no. 2 (A. Poole, P. Stettenheim, and F.
Gill, Eds.). The Academy of Natural Sciences,
Philadelphia, Pennsylvania; The American Orni-
thologists’ Union, Washington, D.C.

Herreid, C. F. 1963. Observations on the feeding be-
havior of Cardisoma guanhumi (Latreille) in
southern Florida. Crustaceana 5:176-180.

Loegering, J. P. and J. D. Fraser. 1995. Factors af-
fecting Piping Plover chick survival in different
brood-rearing habitats. J. Wildl. Manage. 59:646-
655.

Loegering, J. R, J. D. Fraser, and L. L. Loegering.
1995. Ghost crab preys on a Piping Plover chick.
Wilson Bull. 10:768-769.

MacIvor, L. H. 1990. Population dynamics, breeding
ecology, and management of Piping Plovers on

outer Cape Cod, Massachu.setts. M.S. thesis, Univ.
of Massachusetts, Amher.st.

MacIvor, L. H., S. M. Melvin, and C. R. Griffin.
1990. Etlects of research activity on Piping Plover
nest predation. J. Wildl. Manage. 54:443-447.

Magrath, R. D. 1990. Hatching asynchrony in altri-
cial birds. Biol. Rev. 65:587-622.

Patterson, M. E. 1988. Piping Plover breeding biol-
ogy and reproductive success on A.s.sateague Is-
land. M.S. thesis, Virginia Polytechnic Institute
and State Univ., Blacksburg.

Robertson, J. R. and W. J. Pfeiffer. 1982. Deposit
feeding by the ghost crab Ocypode quad rata (Fa-
bricius). J. Exp. Mar. Biol. Ecol. 56:165-177.

Shields, M. A. 1998. Proximate and ultimate causes
of hatching asynchrony in the Brown Pelican.
Ph.D. diss.. North Carolina State Univ., Raleigh.

Staine, K. j. and j. Burger. 1994. Nocturnal foraging
behavior of breeding Piping Plovers (Charadrius
melodus) in New Jersey. Auk 111:579-587.

U.S. Fish and Wildlife Service. 1993. Piping Plover
monitoring and management, summer 1993. U.S.
Fish and Wildlife Service, Chincoteague NWR,
Chincoteague, Virginia.

U.S. Fish and Wildlife Service. 1994. Piping Plover
monitoring and management-summer 1994. U.S.
Fish and Wildlife Service, Chincoteague NWR,
Chincoteague, Virginia.

U.S. Fish and Wildlife Service. 1996. Piping Plover
{Charadrius melodus) Atlantic coast population
revised recovery plan. U.S. Fish and Wildlife Ser-
vice, Hadley, Virginia.

Walters, J. R. 1982. Parental behavior in Lapwings
(Charadriidae) and its relationships with clutch
sizes and mating systems. Evolution 36:1030-
1040.

Walters, J. R. 1984. The evolution of parental be-
havior and clutch size in shorebirds. Behav. Mar.
Anim. 5:234-287.

Walters, J. R. 1990. Antipredatory behavior of Lap-
wings: field evidence of discriminative abilities.
Wilson Bull. 102:49-70

Watts, B. D. and D. S. Bradshaw. 1995. Ghost crab
preys on Piping Plover eggs. Wilson Bull. 107:
767-768.

Wellins, C. a., D. Rittschof, and M. Wachowiak.
1989. Location of volatile odor sources by ghost
crab Ocypode quadrata (Fabriciiis). J. Chem.
Ecol. 15:1161-1169.

Wolcott, T. G. 1978. Ecological role of ghost crabs,
Ocypode quadrata (Fab.) on an ocean beach:
scavengers or predators? J. Exp. Mar. Biol. Ecol.
3 1 :67-82.

Wilson Bull., 111(3), 1999, pp. 330-339

A TAXONOMIC STUDY OF CRESTED CARACARAS

(FALCONIDAE)

CARLA J. DOVE' 2^ AND RICHARD C. BANKS^

ABSTRACT. — The taxonomic status of the crested caracaras (Caracara spp., Falconidae) has been unsettled
for many years. Current sources such as the AOU Check-list recognize a single species that includes three taxa
formerly considered distinct, citing observations by Hellmayr and Conover (1949) on two specimens considered
to be intermediate. We studied plumage characters and measurements of over 392 museum specimens and found
no evidence of clinal change between the northern and southern continental populations. Sixteen specimens from
localities near the Amazon River where these two populations sporadically meet exhibit a mosaic of plumage
elements from both forms. Measurements of wing chord, bill length, and bill depth indicate that size is positively
correlated with latitude north and south of the equator and that females are larger than males in the northern
population. These populations do not meet in western South America. We conclude that three biological species
can be identified in the crested caracaras; the insular Guadalupe Caracara {Caracara hitosus)-, and two continental
species. Northern (C. clieriway) and Southern caracara (C. plancus), neither of which shows subspecific variation.
Received 6 Oct. 1998, accepted 16 Feb. 1999.

The Florida population of the Crested Ca-
racara (Caracara plancus audubonii) is con-
sidered threatened by the U.S. Fish and Wild-
life Service (1987), a recognition that subjects
that population to strict permit regulations and
consideration for conservation efforts (Mor-
rison 1996). Although the generic name given
in the Fish and Wildlife Service listing is Po-
lyborus. Banks and Dove (1992) have shown
that the generic name should be Caracara.
The threatened status, together with the un-
certainty reflected by various taxonomic treat-
ments at the species and subspecies levels, has
led us to complement Morrison’s (1996) de-
tailed report by examining the taxonomy of
crested caracaras throughout their range.

Ridgway (1876) treated the crested caraca-
ras as three species in the genus Polyborus
(now Caracara)'. tharus Molina, 1782 of
southern South America; chertway Jacquin,
1784 from northern South America to south-
ern North America; and lutosus Ridgway,
1876 of Guadalupe Island, Mexico. Except for
the replacement of the specific name tharus
with the earlier plancus Miller, 1777 by Bra-

' Dept, of Biology, George Mason Univ., Fairfax,
VA 22030.

2 Present address: Dept, of Vertebrate Zoology, Na-
tional Museum of Natural History, Smithsonian Insti-
tution, Washington, DC 20560-01 16.

^ U.S. Geological Survey, Patuxent Wildlife Re-
search Center, National Museum of Natural History,
Washington. DC 20560-01 1 1.

* Corresponding author;

E-mail; dovec@nmnh.si.edu

bourne and Chubb (1912), an action followed
by Swann (1925) and all subsequent authors,
this treatment remained unchanged for three-
quarters of a century. Swann (1925) thought
that birds from the northern part of the range
of C. plancus in Brazil were “more or less
intermediate in appearance” between more
southerly C. plancus and C. clieriway, but
gave no details. Hellmayr and Conover
(1949), stating that C. clieriway “appears to
us nothing else but a well-marked race of the
Southern Caracara,” were the first to unite the
two continental populations into a single spe-
cies, C. plancus. The chief distinguishing
characters of blacker coloration and reduction
of white barring both on rump and chest were
considered by Hellmayr and Conover (1949)
as “merely differences of degree.” They also
noted that the apparent gap in measurements
between extreme southern C. plancus (Strait
of Magellan) and C. clieriway was bridged by
specimens from intermediate localities. Hell-
mayr and Conover (1949) admitted that they
had seen C. clieriway only from north of the
Amazon River but cited records of that form
from Santarem and Rhomes, south of the riv-
er. At the same time, they reported that spec-
imens from Marajo and Mexiana Islands, in
the mouth of the Amazon, belonged to C.
plancus. The only actual evidence of inter-
mediacy mentioned by Hellmayr and Conover
(1949:283-284) was based on two adult birds
from Obidos, Brazil, classified as C. clieriway,
“which, by more heavily barred lateral upper

330

Dove and Banks • TAXONOMY OF CRESTED CARACARAS

33]

TABLE

1. Plumage characters of Ca

racara cheriway and C. plancus.

Character — area

C. cheriway

C. plancus

1 Breast

dark spots or wedge-shaped
bars, heavier posteriorly

dark and light bars over entire breast area

2 Vent area

pale patch between thighs

dark feathers between thighs

3 Upper back/Scapulars

pale wedge-shaped patch
with broad black bars,
scapulars always black

finely barred, no wedge-shaped patch, scapu-
lars usually heavily barred, only lightly in
n. part of range

4 Lower back

black

barring continuous with upper back and tail

5 Upper tail coverts

white or faintly barred

barred

tail coverts, mark a decided step in the direc-
tion of plancus.”

Friedmann and coworkers (1950) and the
American Ornithologists’ Union (AOU 1957)
retained C. cheriway as a species. Wetmore
(1965) paraphrased Swann (1925) in men-
tioning intergradation south of the Amazon
and accepted “the present-day tendency to
unite” the two mainland forms. Brown and
Amadon (1968) considered C. plancus and C.
cheriway conspecific as did Blake (1977),
who briefly noted “evidence of intergrada-
tion in Brazil.” Vuilleumier (1970) was the
first to treat the insular C. lutosus as a well-
marked subspecies of the combined C. plan-
cus, apparently following the suggestion of
Brown and Amadon (1968:736) that if cher-
iway and plancus were combined, then “it
would be no great extension to include lu-
tosus” Stresemann and Amadon (1979) and
Sibley and Monroe (1990) merged the con-
tinental forms but retained C. lutosus as an
allospecies. The AOU (1983, 1998) merged
all the taxa, noting the report of intergrada-
tion “near the mouth of the Amazon.” Thus,
the merger of C. plancus and C. cheriway is
based mainly on Hellmayr’s and Conover’s
(1949:283) statement that color characters
were “merely differences of degree”, and on
the two specimens of C. cheriway from Ob-
idos, Brazil, that they stated showed a
“marked step” in the direction of C. plancus.
The degree of intergradation or variation has
never been thoroughly examined in the lim-
ited geographical area where the northern and
southern populations meet, and the distinc-
tiveness of C. lutosus has not been reviewed
since Friedmann (1950).

METHODS

To reevaluate the taxonomic relationships of the
crested caracaras at the species level, we compared five
plumage characters and three body measurements of
specimens from all continental geographic popula-
tions. All plumage color comparisons were made under
museum Examolites® or in daylight. We took mea-
surements on more than 392 specimens of wing chord,
bill depth (at the cere), bill length (from the base of
the bill to the tip), and tarsus of sexed adult birds from
all parts of the species’ range. Tarsal measurements
were extremely variable within populations and were
deemed not useful for comparison. Measurements of
males and females were analyzed separately to deter-
mine sexual variation.

We used an early version of Table 1 to categorize
plumage characters in specimens examined, and we
asked colleagues at some other museums (see ac-
knowledgments) to use that and xeroxed photographs
of Fig. 1 to evaluate specimens for us, thereby avoid-
ing the need of extensive loans of these large birds.
Voice recordings and tissue samples of this species are
insufficient for accurate analysis at this time.

Specimens from the northern populations were com-
pared with those from Bolivia and southern Brazil
south to Tierra del Fuego. Twenty-one specimens from
localities in Brazil along the Amazon River and the
northeastern coast of Brazil are from the contact zone
of the two populations and were examined separately
for plumage patterns. These were compared with
northern and southern populations to determine the ex-
tent of variation within this region. Because this spe-
cies is not sexually dichromatic in plumage color or
pattern, adult specimens (definitive plumage) were not
separated for plumage compari.sons.

At the subspecies level we examined the purported
characters of recently recognized taxa (Hellmayr and
Conover 1949, Peters 1931), all of which were de-
scribed long ago from the periphery of the range of
the species.

South American collecting localities were confirmed
using Stevens and Traylor (1983), Paynter (1988), and
Paynter and Traylor (1991). Descriptive statistics, AN-

332

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

FIG. 1 Typical adult plumage, ventral (above) and dorsal (below) views of (top) Cciracara cheriway (USNM
I 32101 . Sonora, Mexico), (middle) specimen from zone ol contact (USNM 276906, Maica, Brazil), and (bottom),
C. pUmcus (USNM 284790, Argentina).

Dove ami Banks • TAXONOMY OF CRESTED CARACARAS

333

OVAs, and /-tests were done with SYSTAT (1992)
version 5.0 for Windows.

RESULTS

Specific variation. — Typical cidult crested
caracaras from south (C. plancus) and north
(C. cheriway) of the Amazon have clearly dis-
tinct color patterns (Fig. 1 ) with major differ-
ences in all five plumage characters. The
plumage patterns of specimens from the
northern and southern portions of the range do
not overlap, and specimens from these popu-
lations can be identified and allocated un-
equivocally to either C. cheriway or C. plan-
cus on that basis. All adult birds from northern
Brazil (Roraima), Venezuela, Colombia, Ec-
uador, and Peru (Rio Chinchipe, Pacasmayo,
and Catacaos) north to the United States (ex-
cept Guadalupe Island) are of the C. cheriway
type. Birds from southern Brazil (Bahia) and
Bolivia to Tierra del Fuego are of the C plan-
cus type. In addition to the plumage differ-
ences evaluated (Table 1), the undertail co-
verts of C. cheriway are either all white, or
faintly or incompletely barred, giving the base
of the tail a whitish appearance, whereas the
undertail coverts of C. plancus are finely
barred with distinct wide white and narrow
dark bars that extend completely across the
feather. The base of the tail is more white than
barred on C. cheriway specimens, but this
character does not always hold true because
some C. cheriway specimens have barring
near the base of the tail.

Plumage patterns of specimens from the
zone of contact. — The apparent zone of sec-
ondary contact for the two populations ex-
tends from the mouth of the Amazon River
westward along the river and its southern trib-
utaries to the Rio Tapajos (Fig. 2). It may ex-
tend west as far as the Rio Purus (Canutama),
where two immature specimens (Field Muse-
um of Natural History; FMNH 100805,
100806) having mosaic plumage tendencies
were collected. A bird from farther west on
the Rio Jurua (Museum of Comparative Zo-
ology; MCZ 173161) is pure C. plancus and
was not considered to represent the contact
zone. The zone extends to the southeast from
the mouth of the Amazon to approximately
Morros de Mariana. The latest specimen col-
lected from the contact zone that we found
was dated 1937.

In two adult specimens (American Museum
of Natural History; AMNH 241501, 241502)
from well to the south at Remanso, on the Rio
Sao Francisco, the plumage on the upper
back/scapular area is similar to C. cheriway,
but the lower back area is intermediate in col-
or between C. cheriway and C. plancus. The
breast of a bird (Naturhistorisches Museum,
Vienna; NMWZ 39885) from Juazeiro, Bahia,
was scored as C. cheriway. Specimens from
other nearby Bahia localities (AMNH 163138,
Salvador; The Natural History Museum;
BMNH 73.3.19.4, Ilha de Itaparica; NMWZ
39884, Lago de Pamagua; NMWZ 39886,
Barra) all are pure C. plancus. The appearance
of these few c/zer/way- like characters well to
the south may indicate a much wider zone of
intergradation than we recognize, but the lack
of available specimens from the intervening
500 km makes this conclusion problematical.
This observation may merely be a reflection
of the generally more variable plumage in the
southern population. Gyldenstolpe (1951) also
has commented on the lack of comparative
material from parts of the range of this spe-
cies.

Hellmayr and Conover (1949) mentioned
birds from two other localities (Santarem and
Rhomes) in the zone and we accepted their
determination of those birds as cheriway (Ta-
ble 2). We found no gradual intergradation in
plumage characters but instead an abrupt shift
in plumage type. There was no consistent pat-
tern of intermediacy but rather a mosaic of
plumage combinations. A summary of the
plumage data shows that 54% of the 115 in-
dividual character states were like C. plancus,
40% were like C. cheriway, and 6% were in-
termediate. The lower back (character 4) re-
ceived 4 of the 6 intermediate ratings. Speci-
mens rated intermediate in this character have
dark backs with white-tipped, not barred,
feathers in the mid- to lower back region. No
specimen was intermediate in more than one
character. Seven of the birds from the contact
zone were rated as pure for one or the other
species.

The two specimens from Obidos that were
the basis for the original merger of the species
do not agree in plumage patterns. FMNH
101538 is typical C. cheriway, whereas
FMNH 101539 has the breast and heavily
barred tail pattern typical of C. plancus and

ATLANTIC OCEAN

334

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

FIG. 2. Map of localities in zone of contact with icons corresponding to character states listed in Table 2.

Dove ami Banks • TAXONOMY OF CRESTED CARACARAS

335

TABLE 2. Summary of plumage characters (numbers from Table I ) of specimens from the zone ol' contact
between northern anti southern continental populations of crested caracaras. Localities without specimen numbers
are included on the basis of records mentioned by Hellmayr and Conover (1949). C = C. cheriwav, P = C.
planciis. I = intermediate, ? = unscored.

Character

Locality

Museum tt

(Dbidos

FMNH 101538

FMNH 101539

BMNH 1908.8.21.17

Igarpe Arriba

FMNH 101424

FMNH 101425

FMNH 101427

Diamantina Parintins

USNM 121077

AMNH 276706

AMNH 277572

AMNH 277573

Maica Ituqui

USNM 276906

FMNH 101157

FMNH 101158

Agarape Brabo (Rio Tapajos)

AMNH 285747

Humberto do Campos (Maranho)

FMNH 100401

Santo Amaro Maranhao

MCZ 92682

Morros de Mariana

AMNH 241499

AMNH 241500

Ilha Mexiana

BMNH 73.3.19.4

Ilha de Marajo

MCZ 22996

Ilha Caviana

UM 7504

Santarem

Rhomes

the back pattern of C. cheriway with the
wedge-shaped barring on the upper back and
completely black mid- and lower back. An ad-
ditional specimen from Obidos (BMNH
1908.8.21.17) is typical of C. cheriway except
for the barring on the breast. Our efforts to
locate the two other specimens from Obidos
mentioned by Hellmayr and Conover (1949:
284) were unsuccessful.

Size variation. — We divided mensural data
from the continental populations into four
geographic sets as indicated in Table 3. A one-
way analysis of variance shows significant
differences among the four continental groups
in wing length (F = 27.8, P < 0.01, n = 336),
bill length (F = 3.4, P < 0.02, n = 376), and
bill depth (F = 4.2, P < 0.007, n = 391), but
body size, as estimated by wing chord mea-
surements, was positively correlated with lat-
itude (Fig. 3). Although individual variation
in body size is extensive throughout the range,
specimens from the extreme southern parts of
South America are the largest. Northern and

southern populations were analyzed separately
for sexual size dimorphism, f-tests of northern
and southern populations revealed a signifi-
cant sexual size difference in the northern
population, with females being larger (wing
length, t = —5.50, n = 243; bill length, t =
-5.96, n = 21 U bill depth, t = -6.93, n =
277; P < 0.001 in all measurements) but no
difference in the southern populations (wing
length, t = —1.37, n = 77; bill length, t =
-0.65, n = 84; bill depth, t = -0.84, n = 92;
P > 0.05 in all measurements). That the os-
tensible gap in measurements between C.
cheriway and extreme southern birds was
bridged by specimens from intermediate lo-
calities, was used by Hellmayr and Conover
(1949) as an argument to justify conspecific
treatment. Our data (Table 3, Fig. 3) show that
these caracaras vary clinally within each
hemisphere, and become smaller toward the
equator. This is true for wing chord, bill
length, and bill depth.

336

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 3. Descriptive statistics for four geographic
dalupe Island caracara.

groups of continental crested caracaras and the Gua-

Wing

(mm)

Bill length
(mm)

Bill depth
(mm)

US, Cuba, Mexico

d /?“ = 85

Rb = 343-409

28.6-36.7

16.2-19.9

x‘^ = 382.51

32.65

17.70

SD'* = 13.67

1.47

0.75

9 n = 19

R = 363-415

29.8-38.7

16.5-22.0

X = 391.27

33.60

18.3

SD = 1 1.35

1.44

0.88

Central Am. &

Northern South Am.

R = 353-409

29.5-35.0

14.8-19.2

X = 374.17

32.13

17.24

SD = 12.14

1.27

0.84

$ n = 37

R = 358-407

31.1-36.4

15.7-20.0

X = 383.57

33.28

18.05

SD = 11.64

1.31

0.81

Contact Zone

S n = 10

R = 351-401

30.2-34.7

16.3-18.9

X = 379.5

32.32

17.31

SD = 16.15

1.41

0.88

9/7 = 6

R = 359-406

29.2-33.8

17.3-18.8

X = 386.00

31.7

18.04

SD = 16.54

1.55

0.48

Southern South Am.

6 n = 35

Inch S. Brazil

R = 358-438

26.5-37.7

14.9-21.4

X = 397.34

32.49

17.92

SD = 22.46

2.37

1.47

R = 361-455

28.2-37.3

15.7-21.4

X = 404.45

32.82

18.18

SD = 22.84

2.28

1.44

Guadalupe Island

(5/7=1

406

32.2

17.8

9/7 = 2

401.5

31.0

18.2

“ n = number of specimens measured.
R = range.

X = mean.

SD = standard deviation.

DISCUSSION

Species limits. — We conclude that there are
3 species of crested caracaras: Caracara piem-
ens, C. cheriwciy, and C lutosus. Our exami-
nation of nearly 400 specimens from the con-
tinental range of the crested caracaras (Cara-
cara) revealed a mix in plumage characters of
the northern and southern populations only in
the limited zone of contact near the Amazon
River in Brazil. We consider this limited char-
acter sharing a result of secondary contact,
first suggested by Vuilleumier (1970). The

specimens indicate an abrupt shift in appear-
ance from the northern to the southern plum-
age pattern, and most of the intermediate
specimens exhibit a non-consistent mosaic of
characters (back pattern, breast). Juvenile
birds from the contact zone exhibit the same
tendencies, but were not studied as extensive-
ly as adults.

The essential reproductive isolation of these
populations is expressed by the low number
of intermediate characters in specimens from
the contact zone, and the relatively narrow

Dove am! Banks • TAXONOMY OF CRESTED CARACARAS

337

Wing Chord

FIG. 3. Linear regression model of size (indicated
by wing chord) and latitude of 196 specimen records.
Triangles represent specimens from south latitude and
circles represent specimens form north latitude.

area of overlap. Crested caracaras are not
widespread in Amazonia east of the Andes,
occurring only as wanderers in isolated patch-
es of savanna in Amazonian Colombia, Ec-
uador, and Bolivia. Individuals of the two
forms probably meet only infrequently when
they wander into areas of sufficiently open
habitat along the Amazon River. There is no
record of contact of the populations in western
South America. Localities of South American
specimens reveal a distributional gap from
northern Peru (excluding three cheriway spec-
imens taken from coastal Peru and the Peru-
Ecuador border) south to the middle of Chile.
Parker and coworkers (1982) indicate that
crested caracaras are uncommon or rare in
southern Peru, and Johnson (1965) reports
they are very scarce and confined to the sea
coast in the two northernmost provinces of
Chile (see map 77 in Brown and Amadon
1968).

Review of the Guadalupe Island species. —
Morphologically, adults of the extinct Gua-
dalupe Island population differ from those of
North and South American mainland popula-
tions more than the latter two differ from one
another. The crest of C. lutosus is brown rath-
er than black, and the crest feathers are longer
than those of mainland specimens. The throat

is buff to pale brown, not white. Ventral and
dorsal surfaces of the rest of the body are en-
tirely banded with brown and white or buff,
and there is no solid abdominal patch as in
the mainland populations. Remiges and some
scapulars are solid brown rather than black;
most upper-wing coverts are bordered narrow-
ly with darker brown. Sexed adult represen-
tatives of extinct C. lutosus are rare in collec-
tions, and only three specimens were mea-
sured (see Table 3). They do not differ in size
from mainland birds.

A few specimens from southern South
America (e.g., U.S. National Museum, USNM
13926, Patagonia) show some resemblance to
C. lutosus. These South American mainland
birds are primarily brown rather than black,
and most of the body, except for the thighs
and a small abdominal patch, is banded. How-
ever, the banding is finer on the mainland
birds than in lutosus. It was undoubtedly birds
like this that Ridgway (1876:460) alluded to
when he stated “This species resembles the P.
tharus [= C. plancus] much more than P.
cheriway. ...” An explanation for the mor-
phological similarity of birds at opposite ends
of the range, with birds of quite different ap-
pearance occupying the intervening continent,
would be speculative at this stage.

We agree with Ridgway (1876) that the
Guadalupe Island birds are specifically dis-
tinct. Recognition of that distinct and isolated
population follows logically from our sepa-
ration of C. cheriway from C. plancus, but we
suggest that the extinct C. lutosus should be
recognized at the species level regardless of
the treatment of the mainland populations.

Intraspecific variation. — Both southern (C.
plancus) and northern (C. cheriway) crested
caracaras have been variously divided at the
subspecific level by different authors. Popu-
lations of the Southern Caracara (C. plancus)
from northern Paraguay to the Amazon have
frequently been separated under the trinomial
C. p. brasiliensis (Gmelin, 1788) on the basis
of being darker and smaller (Swann 1925,
Wetmore 1926, Peters 1931). The name bras-
iliensis cannot be applied to any caracara
(Banks and Dove 1992), and Gyldenstolpe
(1951) has indicated that the subspecific name
caracara Spix, 1824 must be used if two
forms of birds south of the Amazon are rec-
ognized. However, we are unable to document

338

THE WILSON BULLETIN • Vol. 1 1 1, No. 3. September 1999

any consistent difference in plumage pattern,
color, or size among these birds, and so we
follow most recent authors (Hellmayr and
Conover 1949, Gyldenstolpe 1951, Blake
1977, Stresemann and Amadon 1979) in rec-
ognizing the populations south of the Amazon
as being monotypic Caracara plancus.

Populations of the northern mainland spe-
cies, C. cheriway, have been recognized by up
to four subspecific names (cheriway Jacquin,
1784; audubonii Cassin, 1865; pallidus Nel-
son, 1898; arnmophilus van Rossem, 1939).
Divisions in this species have been based on
the amount and intensity of black as opposed
to brown on the wings and back, and on size.
The amount of black is subject to individual
variation, and depends on stage of molt and
extensive fading related to the open habitat
occupied by the species. Contrary to Griscom
(1932), we do not believe that fading from
black to brown occurs post mortem. Size
varies less than some earlier writers have sug-
gested and increases clinally from the Ama-
zon to the north (Fig. 3), so that the birds in
the southern United States tend to be among
the largest. There are no major changes in
these characters to warrant the recognition of
van Rossem’s (1939) arnmophilus of Sonora,
Mexico with a supposedly smaller bill and
feet and more prominently barred tail, or, de-
spite the present disjunction of range, the tri-
nomial audubonii that Cassin (1865) based on
Florida’s larger birds. We agree with Strese-
mann and Amadon (1979) in considering
those names synonyms of C. cheriway.

Nelson (1898) separated birds from the Tres
Marias Islands off western Mexico under the
subspecific name pallidus, on the basis of pal-
er or lighter brown coloration and slightly
smaller size. Grant (1965) agreed that the is-
land birds are generally paler brown than
those on the mainland but attributed the dif-
ference to greater fading of the island birds.
We agree that the color is not diagnostic, and
note that new feathers or those generally hid-
den from solar radiation are no paler on the
island birds than on those from elsewhere in
the range. Furthermore, AMNH 471349, taken
by Nelson at the Tres Marias, is dark both
dorsally and ventrally. Grant (1965) consid-
ered a shorter terminal tail band and a shorter
wing and tarsus in males to be sufficient for
recognition of pallidus, even though compa-

rable differences could not be demonstrated in
females. However, tarsus length is too indi-
vidually variable to be a useful taxonomic
character. Wing and culmen lengths given by
Nelson (1898) and Grant (1965) fit well in the
range of variation of the much larger mainland
sample we measured (Table 3). We believe
that the minor difference in the length of the
terminal tail band may be related to wear of
the rectrices and that other differences of tail
band characters mentioned by Grant (1965)
are not sufficient to warrant separation of the
Tres Marias population. Some specimens from
the Tres Marias exhibit extensive barring on
the abdomen, but this is within the range of
variation of the mainland specimens. Thus, we
synonymize pallidus with C. cheriway.

ACKNOWLEDGMENTS

Part of this study was supported by a Collections
Study Grant to CJD from the American Museum of
Natural History (AMNH). The study was initiated as
a graduate course by CJD at George Mason University,
supervised by RCB and J. Shaffer. The work was con-
ducted at the U.S. National Museum of Natural His-
tory (USNM). We thank J. V. Remsen, S. W. Cardiff,
L. L. Kiff, K. C. Parkes, and D. Willard for scoring
specimens in their care and gratefully acknowledge
The Lield Museum of Natural History (LMNH) for the
loan of specimens. This study greatly benefited from
help by G. C. Banks in measuring and scoring many
specimens at the following collections; Lield Museum
of Natural History (LMNH), The Natural History Mu-
seum (BMNH), Louisiana State University Museum of
Natural Science, California Academy of Sciences, Cin-
cinnati Natural History Museum, Canadian Museum of
Nature, University of Michigan Museum of Zoology
(UM), Museum of Vertebrate Zoology (MVZ), Muse-
um of Comparative Zoology (MCZ), University of
Kansas Museum of Zoology, Yale Peabody Museum,
Royal Ontario Museum, Museum National d’Histoire
Naturelle, Paris, and Carnegie Museum of Natural His-
tory. CJD measured specimens at AMNH, USNM and
Naturhistorisches Museum, Vienna (NMWZ). A. R.
Phillips provided sub.specific comments; M. R. Brown-
ing and J. V. Remsen reviewed early drafts of this pa-
per. We thank M. Isler for assistance with the map and
V. Krantz (USNM) for the photographs.

LITERATURE CITED

Amkrican Ornithologists' Union. 1957. Check-list
of North American birds. Lifth ed. .Vmerican Or-
nithologi.sts’ Union, Baltimore, Maryland.
American Ornithologists’ Union. 1983. Check-li.st
of North American birds. Sixth ed. American Or-
nithologists’ Union, Washington, D.C.

American Ornithologlsts’ Union. 1998. Check-list

Dove and Banks • TAXONOMY OF CRESTED CARACARAS

339

of North American birds. Seventh ed. American
Ornithologists’ Union, Washington, D.C.

Banks, R. C. and C. J. Dove. 1992. The generic name
for Crested Caracaras (Aves: Falconidae). Proc.
Biol. Soc. Wash. 105:420-425.

Blake, E. R. 1977. Manual of Neotropical birds. Vol.

1. Univ. of Chicago Press, Chicago, Illinois.
Brabourne, L. and C. Chubb. 1912. The birds of

South America. Vol. 1. Taylor & Francis, London,
U.K.

Brown, L. and D. Amadon. 1968. Eagles, hawks and
falcons of the world. Vol. 2, Country Life Books,
London, U.K.

Cassin, J. 1865. Notes on some new and little known
rapacious birds. Proc. Acad. Nat. Sci. Phila. 17:
2-3.

Friedmann, H. 1950. The birds of North and Middle
America. Part XL U.S. Nat. Mus. Bull. 50:1-793.
Friedmann, H., L. Griscom, and R. T. Moore. 1950.
Distributional check-list of the birds of Mexico.
Part 1. Pac. Coast Avif. 29:1—202.

Grant, P. R. 1965. A systematic study of the terrestrial
birds of the Tres Marfas Islands, Mexico. Postilla
60:1-106.

Griscom, L. 1932. The distribution of bird-life in Gua-
temala. Bull. Amer. Mus. Nat. Hist. 64: 1-439.
Gyldenstolpe, N. 1951. The ornithology of the Rio
Purus region in western Brazil. Arkiv Zook, ser.

2, 2:1-320.

Hellmayr, C. E. and B. Conover. 1949. Catalogue
of birds of the Americas and adjacent islands.
Zook Sen, Field Mus. Nat. Hist. 13, Part 1 (4):1-
358.

Johnson, A. W. 1965. The birds of Chile and adjacent
regions of Argentina, Bolivia and Peru. Platt Es-
tablecimientos Graficos, S.A., Buenos Aires, Ar-
gentina.

Morrison, J. L. 1996. Crested Caracara (Caracara
plancus). In The birds of North America, no. 249
(A. Poole and F. Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Nelson, E. W. 1898. New birds from western Mexico.
Proc. Biol. Soc. Wash. 12:5-11.

Parker, T. A., Ill, S. A. Parker, and M. A. Plenge.
1982. An annotated checklist of Peruvian birds.
Buteo Books, Vermillion, South Dakota.

Paynter, R. a., Jr. 1988. Ornithological gazetteer of
Chile. Harvard Univ., Cambridge, Massachusetts.

Paynter, R. A., Jr. and M. A. Traylor, Jr. 1991.
Ornithological gazetteer of Brazil. Harvard Univ.,
Cambridge, Massachuetts.

Peters, J. L. 1931. Check-list of birds of the world.
Vol. 1. Harvard Univ. Press, Cambridge, Massa-
chusetts.

Ridgway, R. 1876. Studies of the American Falconi-
dae. Monograph of the Polybori. Bulk U.S. Geok
Geogr. Surv. Territories 1:451-473.

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

Stevens, L. and M. Traylor, Jr. 1983. Ornithological
gazetteer of Peru. Harvard Univ., Cambridge,
Massachusetts.

Stresemann, E. and D. Amadon. 1979. Order Ealcon-
iformes. Pp. 271—425 in Check-list of birds of the
world, Vol. 1, second ed. (E. Mayr and G. W.
Cottrell, Eds.). Museum of Comparative Zoology,
Cambridge, Massachusetts.

Swann, H. K. 1925. A monograph of the birds of prey
(order Accipitres). Part Ik Wheldon & Wesley,
London, U.K.

Systat. 1992. Version 5 ed. Systat, Inc. Evanston, Il-
linois.

U.S. Eish and Wildlife Service. 1987. Endangered
and threatened wildlife and plants; threatened sta-
tus for the Florida population of the Audubon’s
Crested Caracara. Federal Register 52:25229-
25231.

VAN Rossem, a. j. 1939. Some new races of birds from
Mexico. Ann. Mag. Nat. Hist. ser. 11, 4:439-443.

VuiLLEUMiER, F. 1970. Generic relations and speciation
patterns in the caracaras (Aves: Falconidae). Bre-
viora 355: 1-29.

Wetmore, a. W. 1926. Birds of Argentina, Paraguay,
Uruguay, and Chile. U.S. Nat. Mus. Bulk 133:95-
98.

Wetmore, A. W. 1965. Birds of Panama. Vol. 1.
Smithson. Misc. Coll. 150:1-483.

Wilson Bull., 111(3), 1999, pp. 340-345

VISUAL COMMUNICATION AND SEXUAL SELECTION IN A
NOCTURNAL BIRD SPECIES, CAPRIMULGUS RUFICOLLIS, A
BALANCE BETWEEN CRYPSIS AND CONSPICUOUSNESS

JUAN ARAGONES, ' LUIS ARIAS DE REYNA,' AND PILAR RECUERDA'

ABSTRACT. — Cryptic protective mechanisms and the conspicuousness required to communicate result in a
conflict of opposing selection. In the Red-necked Nightjar (Caprimulgus nificollis) a nocturnal bird, the use of
a restricted signaling strategy provides an appropriate balance between these two selection forces. Conspicuous
white wing and tail bands may have been favored by sexual selection in this species. We studied the variation
of visual signals and found conspicuousness to be closely related to sex and age, being much higher in males
and adults. This variation allows an individual to identify the reproductive status of conspecifics, providing
sexual selection a basis to select these visual signals in this and other nocturnal bird species. We believe that a
relationship between restricted signaling strategy and sexually selected visual signals may occur in nocturnal
species that use visual comunication. Received 24 April 1998, accepted 5 Feb. 1999.

Crypsis is one of the more effective anti-
predator mechanisms (Baker and Parker
1979). Because it relies on inconspicuousness,
its use can conflict with the conspicuousness
required for communication. Selection pres-
sure drives populations to address signals
among conspecifics (conspicuousness increas-
es at close range) and to conceal them from
predators (inconspicuouness or long range
crypsis; Butcher and Rohwer 1989). An ap-
propriate balance between crypsis and com-
munication is achieved by the use of restricted
signaling, for example, the presence of hidden
conspicuous signals in highly specific body
zones that are only exhibited in some situa-
tions (see Butcher and Rohwer 1989).

In strictly nocturnal birds, which account
for less than 3% of all bird species (Martin
1990), the visual channel is assumed to play
an insignificant role in communication com-
pared to diurnal birds. However, this is only
an assumption and nocturnal birds may use
visual signals to communicate more widely
than is thought. The Red-necked Nightjar {Ca-
primulgus ruficolUs) is a nightbird of open and
semi-open habitats that uses sight to capture
flying insects. The plumage of caprimulgids is
highly cryptic, but includes white spots on
wings, tail, and throat. These markings, be-
cause of their anatomical location, are only

' Dpto. Biologia Animal, Univ. de Cordoba, Avda.
San Alberto Magno, sn. 14004 Cordoba, Spain.

2 Present address: Avda. Cadiz n" 5, I4009-C6rdoba.
Spain.

’Corresponding author; E-mail: ba2biani@Lieo.es

visible in some contexts (agonistic, antipred-
ator, and sexual displays) when the sender in-
tentionally shows its markings to the receiver
(Bent 1940, Mengel 1972, Bruce 1973, Cramp
1985, Fry et al. 1988, Cleere 1998, Aragones
et al. in press). Some caprimulgid species
have been reported to show individual varia-
tion in spot size and color (Common Night-
hawk, Chordeiles minor, Selander 1954; Red-
necked Nightjar, Beven 1973, Soares 1973;
Blackish Nightjar, C. nigrescens, Ingels and
Ribot 1982). This variation in spots suggests
that signal conspicuousness might vary with
age and sex.

The objective of this paper is to describe
the variation of spots in signal conspicuous-
ness in the Red-necked Nightjar and to ex-
amine whether this variation is related to age
and/or sex. We hypothesize that if males, es-
specially adult males, show greater conspicu-
ousness in wing and tail bands, then these
plumage characters may be sexually selected.

METHODS

We used data from 170 specimens obtained from
bird collections at Estacion Biologica de Donana,
CSIC (/? = 22) and from road casualties (/t = 148) in
Cordoba (southern Spain) during the 1986 to 1994
breeding seasons. From this sample we measured the
following variables related to conspicuousness. (1)
Band size. The wing-band consisted of 3 to 4 spots
and the tail-band of 2 to 3 spots. Using a digital caliper,
the longer and shorter axis of each spot was measured
(± 0.1 mm) and the area for each was calculated. The
surface areas of the spots present in the outer primaries
of the right wing were summed and recorded as wing-
band size and those on right rectrices were summed
and recorded as tail-band size. The area of each spot

340

Arai>(mes et al. • CRYPSIS AND CONSPICUOUSNESS

341

FIG. 1. Wing-band (WBC) and tail-band contrast
(TBC), calculated from color and sharpness values of
spots forming the bands ( 1 = maximum contrast value,
white spot with sharp contrasting edges, 0 = minimum
contrast value, dark spots without shaip contrasting
edges).

was indexed by the product of the long and the short
axis. (2) Band contrast. Contrast was measured from
the color and sharpness of the spots that formed the
wing and tail bands. The contrast gradient was estab-
lished from a reference series formed by four types of
spots that were assigned the values 0, 0.33, 0.66 and
1; where 0 denoted the spot with the darkest colour
and least sharp contrasted edges, and 1 that with the
lightest color and sharpest edges. A given individual
can possess spots of different types (depending on col-
or and sharpness variations), so each spot was assigned
a value, and a mean index was used for the contrast
analysis in wing-band contrast and tail-band contrast
which included all spots of the wing and tail (Fig. 1).
(3) Uniformity of color spots within a band. Some in-
dividuals showed bands consisting of a single type of
spot, whereas others had one or two different types;
the former were designated “uniform” and the latter
“non-uniform”. (4) Spot number in wing and tail
bands. (5) Band conspicuousness. Band size and con-
trast values were first log transformed, normalizing the
data for the application of parametric statistics and
used to calculate a conspicuousness index (C) from the
following expression:

C = WBS X log(WBC+l) + TBS X log(TBC-M)/
WBS + TBS

where WBS = wing-band size, TBS = tail-band size,
WBC = wing-band contrast, and TBC = tail-band
contrast. This index varied from 0 to I, with 1 corre-
sponding to maximum conspicuousness. Sex was de-
termined by gonadal dissection. Differences in band
size and conspicuouness were analysed by ANOVA
and the Tukey test, and differences in the number of
spots and their uniformity by means of the G-test. Con-
trast differences were identified by multiple logistic re-
gression, which allows one or more categorical vari-
ables (contrast in our case) to be analysed and related
to a dependent variable. This method is especially suit-

able tor data that are not normally distributed and is
more eftective than other classifying methods such as
discriminant analysis (Press and Wifson 1978; see also
Hanell 1986, Schlinger 1990, Fancy et al. 1993). Con-
spicuousness data were transformed logarithmically
for parametric analysis. All analyses were performed
using JMP software (version 2, Macintosh computer).

RESULTS

There were significant differences in wing-
band size {F = 31.47, df = 3, 90, P < 0.001)
and tail-band size (F = 41.61, df = 3, 90, P
< 0.001) between sex-age categories. Wing
and tail bands were significantly larger in
males than in females and larger in adults rel-
ative to young. There was a clear trend for
band size to increase in an age-sex sequence,
both in wing-band (mean young females =
662.3 ± 37.87 SD, n = 29; mean adult fe-
males = 803.1 ± 52.66, A? = 15; mean young
males = 986.3 ± 37.24, n = 30; mean adult
males = 1210.5 ± 45.61, n = 20) and in tail
band (mean young females = 743.4 ± 67.06,
n = 29; mean adult females = 1046.4 ±
93.24, n = 15; mean young males = 1428.7
± 65.93, n = 30; mean adult males = 1852.6
± 80.75, n = 20). Both variables were found
to be positively correlated (r = 0.67, n = 93,
P < 0.001).

The multiple logistic regression discrimi-
nated between the four age-sex classes in
terms of wing-band or tail-band contrast (Ta-
ble 1). There were significant differences in
sex and age for wing-band contrast (x^3 =
86.33, P < 0.001, n = 94) and tail-band con-
trast X"3 = 94.21, P < 0.001, n = 94). Males
and adults always had the largest contrast in-
dex both for wing-band and tail-band (Fig. 2).
Sharp edges were detected only in males with
wing-band or tail-band contrast index close to
one because only such spots had sharp edge
contrasts. Males showed greater uniformity in
wing-band contrast than females, and adults
were more uniform than young in this respect.
All groups showed uniform tail bands (Table
2).

The presence of four spots on the wing is
an exclusive feature of adult males (Table 2);
over 45% (/? = 16) of adult males have the
fourth spot. About 80% of adult males also
tended to possess three caudal spots (/? = 28;
Table 2). Additional spots on the tail was not
significantly different among the other groups
(Table 2). The additional wing spot increased

342

THE WILSON BULLETIN • Vol. HI, No. 3. September 1999

TABLE 1. Parameter estimates of logistic functions derived from multiple logistic regression.

Logistic function parameters

Estimate

Wald test

(Young female vs adult female)

Intercept I

9.47

17.62

<0.001

Wing-band contrast

-13.22

21.22

<0.001

(Adult female vs young male)

Intercept 2

7.52

1 1.51

<0.001

Wing-band contrast

-10.17

14.43

<0.001

(Young male vs adult male)

Intercept 3

1.63

0.56

>0.05

Wing-band contrast

-1.34

0.32

>0.05

(Young female vs adult female)

Intercept 1

10.05

23.21

<0.001

Tail-band contrast

-16.03

31.09

<0.001

(Adult female vs young female)

Intercept 2

8.45

16.36

<0.001

Tail-band contrast

-12.29

20.59

<0.001

(Young male vs adult male)

Intercept 3

6.09

10.54

<0.001

Tail-band contrast

-6.90

10.20

<0.001

the wing-band size of adult males an average
of 6% and the tail-band size of adult males an
average of 12%. The presence of additional
spots on wing and tail increased the overall
band area by 9.6% and presumably increased
signal conspicuousness. There were signifi-
cant differences in conspicuousness (Table 2)
with regard to age-sex classes, with greater
conspicuousness in males and adults.

DISCUSSION

We detected a marked trend for signal con-
spicuousness to increase with age and in
males which was reflected in an increase of
contrast, number of spots, wing- and tail-
bands area, and uniformity of contrast. The
differences in conspicuousness in wing- and
tail-bands were related to the age and sex class
of the individual. Potentially, these signals let
conspecifics evaluate one another in aggres-
sive or reproductive contexts. Signal conspic-
uousness seemingly involves little energy ex-
penditure but is combined with expensive dis-
plays; the combination results in an efficient
signal that conveys a much greater amount of
information (Hasson 1991). We found that
spot number was related to sex and age, but
Forero and coworkers (1995) found no statis-
tically significant differences in the average

FIG. 2. Heavy line indicates probability of belong-
ing to a sex cla.ss as a function of wing-band contra.st
(WBC) or tail-band contra.st (TBC), thin lines indicates
probability of belonging to an age class.

Aragones cl al. • CRYPSIS AND C(3NSPICUOUSNESS

343

TABLE 2. Frequency distribution of
index for age-sex groups.

contrast uniformity.

additional spots in

bands and

conspicuousness

Female

Male

Young

Adult

Young

Adult

Contrast uniformity (n = 94)

n = 29

/7 = 15

n = 30

/; = 20

WBC (G, = 27.29)

37%

66%

95%

<0.001

TBO' (Gj = 1 1.09)

88%

87%

91%

89%

>0.05

Additional spots (n = 170)

n = 50

n = 31

n = 54

n = 35

In wing (G, = 40.33)

55%

<0.001

In tail (G, = 39.85)

80%

<0.001

Conspicuousness (/; = 94)

n = 29

/; = 15

n = 30

n = 20

(F = 71.76, n = 3, 90)

0.23

0.43

0.91

<0.001

“ WBC = wing-band contra.si, TBC =

tail-band contrast

number of spots between groups of the same
species. We believe the addition of new spots
has a strong biological significance because it
occurs only in adult males as reported by For-
ero and co workers (1995), but their method,
failed to demonstrate a significant difference.

In the Red-necked Nightjar increased signal
conspicuousness may increase the likelihood
of mating (Andersson 1982, Andersson 1992,
Saetre et al. 1994), and conspicuous males
may have a higher mating success (Payne
1982, Price 1984, Jarvi et al. 1987, Harvey

FIG. 3. For restricted signaling strategy to be effective, conspicuous plumage zones (black and white areas
in the drawing) must be concealed by cryptic plumage zones (gray areas in the drawing) while at rest. A
hypothetical model based on the genus Caprimulf’us (I) represents the maximum possible conspicuousness for
restricted signaling strategy; a larger conspicuous area would defeat the defensive system based on crypsis and
hence cancel restricted signaling strategy. The Nacunda Nighthawk (2), the Sand-Colored Nighthawk (3) and
the White-Winged Nigthjar (4) use three restricted signaling strategies that are very similar to that of the model,
esspecially the Sand-Colored Nighthawk.

344

THE WILSON BULLETIN • VoL III, No. 3, September 1999

and Bradbury 1991, Sundberg and Larsson
1994). The fact that increased signal conspic-
uousness was associated with adult males sug-
gests that the signal could be modified by sex-
ual selection (Hbglund 1993). For adult males,
this signal indicates that those individuals
have survived at least two reproductive sea-
sons and, therefore, may reflect a high repro-
ductive quality. In another nocturnal bird spe-
cies, the Great Snipe {Gallinago media) males
have white spots on their tails, and females
choose the males with the whitest signals
(Hbglund et al. 1992). This suggests that sig-
nal conspicuousness in nocturnal birds could
be an effective way of distinguishing between
potential mates.

Our results show that in Red-necked Night-
jar both sexes possess ornaments, and that in
both sexes spots tend to increase in conspic-
uousness with age. Forero and co workers
(1995) claimed that visual signals increased
only in males. We believe that their results are
due to a less precise method of measuring spot
size.

The restricted signaling strategy allows two
scarcely compatible mechanisms to be com-
bined and might occur in many cryptic species
of the genera Burhinus, Charadrius and Gal-
linago, although this has not yet been inves-
tigated. For a nocturnal signal, contrast is
more important than color, as in the capri-
mulgids and other species that use restricted
signaling strategy with white spots (i.e., Gal-
linago media, Hbglund et al. 1992; Burhinus
spp., Martin 1990, Hayman et al. 1986).
Therefore, increasing the amount of white col-
or increases contrast and presumably maxi-
mizes the signaling ability. Similar trends of
increasing conspicuousness with age have
been detected in other caprimulgids (Little
Nightjar, C. parvulus heterurus, Schwartz
1968; Chuck-WiU’s-Widow, C. vociferus,
Rohwer 1971; Blackish Nightjar, Ingels and
Ribot 1982; Scrub Nightjar, C. anthonyi, Rob-
bins and Ridgeley 1994; Nacunda Nighthawk,
Podager nacunda, Aragones unpubl. data;
Pauraque, Nyctidromus albicollis, Aragones
1997a, b).

From the distribution of visual signals in
Red-necked Nightjar (throat, wings, and tail-
bands), we developed a model plumage pat-
tern for a nightjar that represents the optimum/
maximum distribution of conspicuous areas in

the body (Fig. 3). We subsequently found that
the Nacunda Nighthawk and the White-
Winged Nightjar {Caprimulgus candicans)
have patterns that closely resemble our model,
and the Sand-colored Nighthawk’s (Chordei-
les rupestris) visual signals patterns are iden-
tical with that of the model. All three are gre-
garious South American species of open or
semiopen habitats with sexual variations in
wing- and tail-bands (Ffrench 1986, Hilty and
Brown 1986, Sick 1993). It is interesting to
note that the Sand-colored Nighthawk and the
Yellow-billed Tern {Sterna superciliaris), two
Amazonic species that use the same fluvial
habitat, show almost identical visual signal
patterns and general appearance (see Sick
1993). This convergent evolution stresses the
significance of open and semiopen habitats for
restricted signaling strategy evolution. Such
habitats are also characterized by other species
that employ this strategy, Burhinus spp. and
the Great Snipe (Hayman et al. 1986). We be-
lieve that a relationship between restricted sig-
naling strategy and sexually selected visual
signals may occur in nocturnal species that
uses visual comunication.

ACKNOWLEDGMENTS

Valuable assistance was provided at various stages
of this study by many friends. We thank M. C. Casaut.
A. Leiva, M. A. Nunez and R. Pulido for help with
the collection of traffic casualties, and E M. Coco for
help with statistical analysis. We are grateful to M. A.
Nunez, E M. Coco, R. Reques, S. Carpintero and J.
Marin for practical support, discussion, and comments.
Two anonymous referees provided valuable sugges-
tions on an early version of this manuscript.

LITERATURE CITED

Andersson, M. 1982. Female choice selects for ex-
treme tail lengh in a widowbird. Nature 299:818-
820.

Andersson, S. 1992. Female preference for long tails
in lekking Jackson's Widowbirds: experimental
evidence. Anim. Behav. 43:379-388.

Aragones, J. 1997a. Influencia de la cripsis sobre el
comportamiento del Chotacabras pardo, Capri-
mulf>u.s ruficoUis. Tesis Doctoral. Univ. de Cor-
doba. Cordoba, Spain.

Aragones, J. 1997b. Risk taking and flushing distance:
a way of parental investment in the Pauraque
(Nyctidromu.s albicollis). Etologia 5:83—90.
Aragones, J., L. Arias de Reyna, and P. Recuerda.
1998. Antipredator strategies and mimicry in Red-
necked Nightjar Caprimulgus ruficollis. Etologia
6:53-59.

Aragones el al. • CRYPSIS AND CONSPICUOUSNESS

345

Baker, R. E. and G. A. Parker. 1979. The evolution
of bird coloration. Phil. Tran.s. R. Soc. Lond. B
287:63-130.

Bent, A. C. 1940. Life histories of North American
cuckoos, goatsuckers, hummingbirds and their al-
lies. Bull. U.S. Nat. Mus. 176:147-254.

Beven, G. 1973. Studies of less familiar birds. Red-
necked Nightjar. Brit. Birds 66:390-396.

Bruce, A. 1973. Tail flashing display in the Whip-
poor-will. Auk 90:682.

Butcher, G. S. and S. Rohwer. 1989. The evolution
of conspicuous and distinctive coloration for com-
munication in birds. Curr. Ornithol. 6:51-108.

Cleere, N. 1998. Nightjars. A guide to nightjars and
related nightbirds. Pica Press. Sussex, U.K.

Cramp, S. 1985. The birds of the western Palearctic.
Vol 4. Oxford Univ. Press, Oxford, U.K.

Fancy, S. G., T. K. Pratt, G. D. Lindsey, C. K. Har-
ada, a. H. Parent, and J. D. Jacobi. 1993. Iden-
tifyng sex and age of Apapane and liwi on Ha-
waii. J. Field Ornithol. 64:262—269.

Ffrench, R. 1986. A guide to the birds of Trinidad and
Tobago. Harrowood Books, Valley Forge, Penn-
sylvania.

Forero, M. G., j. L. Tella, and L. Garcia. 1995. Age
related evolution of sexual dimorphism in the
Red-necked Nightjar Caprirnulgus ruficolUs. J.
Ornithol. 136:447-451.

Fry, C. H., S. Keith, and E. K. Urban. 1988. The
birds of Africa. Vol 3. Academic Press, London,
U.K.

Harvey, P. H. and J. W. Bradbury. 1991. Sexual se-
lection. Pp. 203—233 in Behavioural ecology. An
evolutionary approach (J. R. Krebs and N. B. Da-
vies, Eds.). Blackwell Scientific Publications, Ox-
ford, U.K.

Harrell, F. E. 1986. The logistic procedure. Pp. 269—
293 in SUGI supplemental library user’s guide,
version 5 (R. P. Hastings, Ed.). SAS Institute Inc.,
Cary, North Carolina.

Hasson, O. 1991. Sexual displays as amplifiers: prac-
tical examples with emphasis on feather decora-
tions. Behav. Ecol. 2:189-197.

Hayman, R, P. Marchant, and T. Prater. 1986.
Shorebirds: an identification guide to the waders
of the world. Croom Helm, London, U.K.

Hilty, S. L. and W. L. Brown. 1986. The birds of
Colombia. Princeton Univ. Press, Princeton, New
Jersey.

Hoglund, j. 1993. Cost of mate advertisement. Eto-
logia 3:143-150.

Hoglund, J., M. Eriksson, and L. Lindell. 1992. Fe-
males of the lek breeding Great Snipe, GoUinago

media, prefer males with white tails. Anim. Be-
hav. 40:23-32.

Ingels, J. and J. H. Ribot. 1982. Variations in the
white markings of the Blackish Nightjar Capri-
iniilgus nigrescens. Bull. Brit. Ornithol. Club 102:
1 19-122.

Jarvi, T, E. Roskaft, M. Bakken, and B. Zum.steg.
1987. Evolution of variation in male .secondary
sexual characteristics: a test of eight hypothesis
applied to Pied Flycatchers. Behav. Ecol. Socio-
biol. 20:161-169.

Martin, G. 1990. Birds by night. T. & A. D. Poyser,
London, U.K.

Mengel, R. M. 1972. Wing clapping in territorial and
courtship behavior of the Chuck-will’s-widow and
Poor- Will (Caprimulgidae). Auk 89:441-444.

Payne, R. B. 1982. Ecological consequences of song
matching: breeding success and intraspecific song
mimicry in Indigo Buntings. Ecology 63:401-
411.

Press, S. J. and S. Wilson. 1978. Choosing between
logistic regression and discriminant analysis. J.
Am. Stat. Assoc. 73:699-705.

Price, T. D. 1984. Sexual selection on body size, ter-
ritory and plumage variation in a population of
Darwin’s finches. Evolution 38:327-341.

Robbins, M. B. and R. S. Ridgeley. 1994. Voice,
plumage and natural history of Anthony’s Nightjar
Caprirnulgus anthonyi. Condor 96:224-228.

Rohwer, S. 1971. Molt and the annual cycle of the
Chuck-will’s-widow, Caprirnulgus carolinensis.
Auk 88:485-519.

Saetre, G. R, S. Dale, and T. Slagsvold. 1994. Fe-
male Ried Flycatchers prefer brightly coloured
males. Anim. Behav. 48:1407-1416.

ScHLiNGER, B. A. 1990. A non-parametric aid in iden-
tifying sex of cryptically dimorphic birds. Wilson
Bull. 102:545-550.

Schwartz, R. 1968. Notes on two Neotropical night-
jars, Caprirnulgus anthonyi and C. pan’ulus. Con-
dor 70:223-227.

Selander, R. K. 1954. A systematic review of the
booming nighthawks of we.stern North America.
Condor 56:57-82.

Sick, H. 1993. Birds in Brazil. A natural history.
Rrinceton Univ. Rress, Rrinceton, New Jersey.

Soares, A. A. 1973. Sobre Caprirnulgus europaeus e
C. ruficollis em Rortugal (Aves, Caprimulgifor-
mes). Arquivos Museu Bocage 4:215-233.

SuNDBERG, J. AND C. Larsson. 1994. Male coloration
as an indicator of parental quality in the Yellow-
hammer, Emheriza citrinella. Anim. Behav. 48:
885-892.

Wilson Bull., 111(3), 1999, pp. 346—353

INTERSPECIFIC INTERACTIONS WITH FORAGING
RED-COCKADED WOODPECKERS IN SOUTH-CENTRAL FLORIDA

REED BOWMAN,' 5 DAVID L. LEONARD, JR.,' ^ LESLIE K. BACKUS,' ^ AND

ALLISON R. MAINS'^

ABSTRACT. — Inter.specific competition for Red-cockaded Woodpecker (Picoicles borealis) cavities has been
well documented and may be one factor contributing to the species’ decline. Other forms of interspecific inter-
actions have rarely been documented over most of the species’ range and have received little attention. During
806 hours of Red-cockaded Woodpecker foraging observations in south-central Florida we documented 306
interspecific interactions with 19 species. We observed fewer non-foraging interactions (98) than foraging inter-
actions (208). Red-cockaded Woodpeckers lost 70 (71%) of the non-foraging interactions and 177 (85%) of the
foraging interactions. Most non-foraging interactions (64%) were with non-woodpecker species, several of which
frequently and consistently dominated Red-cockaded Woodpeckers. Together, Eastern Kingbirds (Tyrannus tyr-
annies), Great Crested Flycatchers (Myiarclnis crinitus). Eastern Bluebirds (Sialia sialis), and Pine Warblers
(Dendroica pinus) won 45 of their 48 (94%) non-foraging interactions with Red-cockaded Woodpeckers. Most
foraging interactions (97%) were with other woodpecker species. Red-bellied Woodpeckers (Melanerpes caro-
linits) were involved in 172 (85%) of these interactions, of which they won 168 (98%). We found no relationship
between the rate of interactions and the habitats or the local landscape in which these interactions occuiTed.
Red-cockaded Woodpeckers did not appear to move to different and possibly less productive foraging sites after
being usurped. In south-central Florida, where hardwood basal areas are relatively low in Red-cockaded Wood-

pecker habitat, the foraging niche of these two species
range. Received 20 July 1998, accepted 5 Feb. 1999.

The Red-cockaded Woodpecker (Picoides
borealis) is a cooperative breeder restricted to
the old growth pine forests of the southeastern
United States (Jackson 1971). Despite nearly
30 years of Federal protection, Red-cockaded
Woodpecker populations have continued to
decline (James 1991). Habitat loss and frag-
mentation have ultimately been responsible
for the species’ decline (Lennartz et al. 1983,
Conner and Rudolph 1991). Interspecific com-
petition for Red-cockaded Woodpecker nest
and roost cavities has been well documented
(Jackson 1978, Harlow and Lennartz 1983,
Kappes and Harris 1995) and may be one
proximate factor contributing to the species’
decline (U.S. Fish and Wildlife Service 1985).

Interspecific interactions, other than those
involving cavities, have rarely been reported
over most of the Red-cockaded Woodpecker’s

‘ Archbokl Biological Station, PO Box 2057, Lake
Placid, FL 33862.

j.'^ Dept, of Wildlife Ecology and Conservation, Univ.
of Florida, PO Box 1 1043, Gainesville, FL 3261 1.

’James San Jacinto Mountains Re.serve, Univ. of
California, PO Box 1775, Idyllwild, CA 92549.

’ Dept, of Forestry, Wildlife and Fisheries, Univ. of
Tennessee, Knoxville, TN 37996.

’ Corresponding author;

E-mail: rbowman@archbold-station.org

may overlap to a greater extent than elsewhere in then-

range (Morse 1970, Nesbitt et al. 1978).
Hooper and Lennartz (1981) observed forag-
ing Red-cockaded Woodpeckers from May to
March in South Carolina and documented 21
interspecific interactions between Red-cock-
aded Woodpeckers and one of four wood-
pecker species or the Brown-headed Nuthatch
{Sitta pusilla). Only three interactions were re-
lated to foraging. Ligon (1970) reported six
interactions between Red-cockaded Wood-
peckers and Downy {Picoides pubescens) and
Hairy woodpeckers {P. villosus) during 240
hours of observations from May to December
in north-central Florida. In contrast, Nesbitt
and coworkers (1981) documented 149 inter-
specific interactions between Red-cockaded
Woodpeckers and five woodpecker species
during 221 hours of observations from July to
October in southwestern Florida. Most inter-
actions involved Red-bellied Woodpeckers
{Melanerpes carolinus) that often usurped
Red-cockaded Woodpeckers from foraging
sites. These interactions may have reduced the
caloric intake of foraging Red-cockaded
Woodpeckers (Nesbitt et al. 1981).

Geographic variation in interactions be-
tween species is common (Travis 1996). Ex-
plaining this variation may lead to a better un-
derstanding of geographical differences in be-

346

Bowman et al. • RED-COCKADED WOODPECKER EORAGING INTERACTIONS

347

havior, demography, and habitat selection of
potentially interacting species. In this paper
we report on interspecific interactions with
Red-cockaded Woodpeckers in a small popu-
lation in south-central Florida.

METHODS

The Avon Park Air Force Range (APR) i.s a 42,900
ha, multiple-use, active military training installation in
Polk and Highlands counties, Florida. Dominant native
pine communities consist of longleaf (Finns paliistris)
and south Florida slash pine (P. elliotlii var. densa)
and approximately 9,000 ha planted in north Florida
slash pine (P. elliottii var. elliottii). The pine habitats
are interspersed with other communities typical of this
region such as oak scrub and fresh water marshes. The
natural pine habitats support the characteristic bird
community for this region (Engstrom 1993), including
21 groups of Red-cockaded Woodpeckers.

To determine the foraging preferences of Red-cock-
aded Woodpeckers at Avon Park AFR, we observed
individuals from 12 groups once a month from April
1995 to March 1996. Red-cockaded Woodpeckers
were observed from dawn to dusk whenever possible;
observations that ended prior to 13:00 EST were re-
peated. During a foraging observation period, we re-
corded the location of the focal individual, its foraging
maneuver, and substrate use at 10 minute intervals.
Locations were entered into a Geographical Informa-
tion System (ArcView GIS Version 3.0). We deter-
mined home range boundaries and overlaid these
boundaries with existing habitat type coverages. From
these maps, we calculated the area of each habitat type
(13 categories) in each home range and linked indi-
vidual foraging locations to specific habitat types.
These habitat types included pine flatwoods, scrubby
flatwoods, oak scrub, sand pine scrub, pine plantation,
mixed natural pine and plantation, pine swamp, oak
hammock, hardwood forests, cypress, marsh, lake, and
human disturbed.

During a subset (806 hours) of the total observation
time (1168 hours), we documented all interspecific in-
teractions. We recorded the species, sex (if determin-
able), type of aggression (aerial chase, tree chase,
lunge, usurp, etc.), and the outcome (winner/loser). In-
dividuals that retreated without retaliation were clas-
sified as losing. We categorized interactions as forag-
ing or non-foraging interactions. Interactions where the
winner examined, or foraged at, the usurped site were
categorized as foraging related. All other interactions
were categorized as non-foraging. Monthly observa-
tion periods varied as did the number of individuals in
each group. To avoid observation time and group size
biases, we used only those interactions that involved
the breeding pair in each cluster and converted those
interactions to a rate per hour for all analyses. We u.sed
the number of interactions between all individuals to
describe the species involved in interspecific interac-
tions with Red-cockaded Woodpeckers, and the num-
ber, type and outcome of those interactions.

Because most interactions were instantaneous or no
longer than 15-60 s (aerial cha.ses), we assumed inter-
actions between the same individuals were indepen-
dent if they occuired more than 1 5 min apart. For in-
teractions that occuiTed less than 15 min apart, we ex-
cluded all but the first interaction as long as the type
of interaction (foraging or non-foraging) and habitat
were the same. When the type of interaction or habitat
differed, we excluded all the interactions, since they
could not be aggregated into a single type of interac-
tion. However, if we suspected two different individ-
uals of the other species (e.g., one male and one fe-
male) were involved in sequential interactions less
than 15 min apart then both observations were consid-
ered independent.

To examine whether the frequency of interspecific
interactions was habitat specific, we compared the fre-
quency of interactions per habitat type to the expected
frequency based on the proportion of time Red-cock-
aded Woodpeckers foraged in each habitat type. To
determine if the frequency of interspecific interactions
was related to the local landscape, we compared the
frequency of interactions per group to the mean basal
area of pines and hardwoods in each Red-cockaded
Woodpecker’s home range. Given the frequency of for-
aging interactions between Red-cockaded and Red-bel-
lied woodpeckers, we repeated the above analyses for
those interactions.

To determine whether usurpations had a measurable
effect on Red-cockaded Woodpecker foraging patterns
we performed two analyses. First, we compared the
habitats used by Red-cockaded Woodpeckers before
and after interactions with Red-bellied Woodpeckers
to determine whether the former moved to a different
and potentially less productive habitat after an inter-
action. Second, we compared foraging tree character-
istics [dbh (diameter at breast height) and height] and
Red-cockaded Woodpecker foraging height before and
after usurpations by Red-bellied Woodpeckers to de-
termine whether they moved to different micro-sites
after interactions. Male and female Red-cockaded
Woodpeckers forage at different locations (Ligon
1970), therefore we analyzed each sex separately. All
statistical tests were nonparametric and were per-
formed in the Microsoft Windows 95 operating system
using SPSS (version 8.0).

RESULTS

We observed 306 independent interspecific
interactions between 45 color-banded Red-
cockaded Woodpeckers and 19 other bird spe-
cies (Table 1 ). Interactions involved 26 breed-
ing adult Red-cockaded Woodpeckers (13 <3,
13 $), 10 hatch-year birds (6 d, 4 9), 6 older
helpers (all c3), and 2 floaters (both 6). Of the
306 interactions observed, 203 occurred with
the breeding Red-cockaded Woodpeckers, 50
with hatch-year birds, 16 with older helpers,
and 6 with floaters. Red-cockaded Woodpeck-

348

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

TABLE 1. Species observed interacting with Red-cockaded Woodpeckers, the outcome (loser or winner),
and the type (foraging or non-foraging) of interaction during 806 hours of foraging observations at the Avon
Park Air Force Range, 1995-1996.

Red-cockaded Woodpecker

Loser

Red-cockaded Woodpecker

Winner

Species

Foraging

Non-foraging

Foraging

Non-foraging

Red-shouldered Hawk

Red-bellied Woodpecker

168

Red-headed Woodpecker

—

Northern Flicker

—

Yellow-bellied Sapsucker

—

Downy Woodpecker

—

Hairy Woodpecker

—

Eastern Kingbird

—

Great Crested Flycatcher

—

1 1

—

Florida Scrub-Jay

—

Blue Jay

—

Brown-headed Nuthatch

—

Eastern Bluebird

—

Loggerhead Shrike

—

Northern Mockingbird

—

Pine Warbler

—

Eastern Towhee

—

Red-winged Blackbird

—

Summer Tanager

—

Total # of Interactions

177

Total # of Species

ers lost 247 (81%) interactions to 18 species
and won 59 (19%) interactions with 9 species
(Table 1). Windoss ratios for Red-cockaded
Woodpeckers did not differ between life his-
tory stage (x^ = 1.23, df = 3, P > 0.05) or
sex (x^ = 1.49, df = 1, P > 0.05).

Interactions between Red-cockaded Wood-
peckers and other woodpecker species were
most frequent, accounting for 237 (77%) of
all interactions. Excluding species with fewer
than five observed interactions, five species
won more than 85% of their interactions with
Red-cockaded Woodpeckers [Eastern King-
bird (Tyrannus tyrannus), 100%; Great Crest-
ed Flycatcher {Myiarchus crinitus), 100%;
Red-bellied Woodpecker, 94%; Eastern Blue-
bird {Sicilia sialis), 89%; and Pine Warbler
(Dendroica pinus), 86%; Table 1]. The
Downy Woodpecker was the only species
Red-cockaded Woodpeckers consistently
dominated (34 of 36 encounters).

The rate of interspecific interactions with
breeding Red-cockaded Woodpeckers was
greatest in June and July. Although interac-
tions varied between months from 0.10 (±
0.03 SE) to 0.40 (± 0.31) interactions per

hour (Fig. la), these differences were not sig-
nificant (Kruskal-Wallis One Way ANOVA:
X“ = 16.7, df = 11, P > 0.05).

Non-foraging interactions. — Red-cockaded
Woodpeckers had 98 non-foraging interac-
tions with 18 species; however, they had only
non-foraging interactions with 12 of those
species (Table 1). Interactions with Red-head-
ed Woodpeckers {Melanerpes erythocephal-
us). Eastern Kingbirds, Great Crested Fly-
catchers, Northern Mockingbirds {Mirnus po-
lyglottos), and Pine Warblers often involved
aerial chases that lasted from 15-60 s. Few of
these interactions were initiated by Red-cock-
aded Woodpeckers. However, in one instance
a group of Red-cockaded Woodpeckers
mobbed and successfully evicted a Red-shoul-
dered Hawk {Buteo lineatus). Of the 98 inter-
actions, Red-cockaded Woodpeckers won 28
and lost 70.

Non-foraging interactions were most fre-
quent during June and July (Fig. lb) and
monthly differences were statistically different
(Kruskal-Wallis One Way ANOVA: x^ ~
23.9, df = 11, P = 0.013).

Foraging interactions. — Red-cockaded

Bowman el at. • RED-COCKADED WOODPECKER FORAGING INTERACTIONS

349

Month

FIG. 1. Interspecific interactions (T ± 1 SE for each month) observed per hour with breeding adult Red-
cockaded Woodpeckers at Avon Park Air Force Range, in south-central Florida, 1995-1996; (a) all interactions,
(b) non-foraging interactions, (c) foraging interactions, and (d) all interactions with Red-bellied Woodpeckers.

Woodpeckers had 208 foraging interactions
with eight species. With six of these species
they had both foraging and non-foraging in-
teractions, but foraging interactions were
more frequent than non-foraging interactions
(Table 1 ). Most foraging interactions were be-
tween Red-cockaded Woodpeckers and other
woodpeckers (202 of 208, 97%), but interac-
tions with Blue Jays (Cyanocitta cristata).
Eastern Kingbirds, Brown-headed Nuthatches,
and Pine Warblers also were observed. Red-
cockaded Woodpeckers lost most (177 of 208,
85.1%) foraging interactions; however, 74%
of the 31 interactions they won were with
Downy Woodpeckers. Red-cockaded Wood-
peckers lost a greater percentage of foraging
interactions (85.1%) than they did non-forag-
ing interactions (71.4%; = 7.14, df = 1, P

= 0.008). None of the foraging interactions
were initiated by Red-cockaded Woodpeckers,
except for those with Downy Woodpeckers.

Downy Woodpeckers frequently foraged near
Red-cockaded Woodpeckers and often were
aggressively chased and their foraging loca-
tions usurped. The rate of foraging interac-
tions with breeding Red-cockaded Woodpeck-
ers did not vary monthly (Kruskal-Wallis One
Way ANOVA: x" = 14.2, df = 1 1, P = 0.22;
Fig. Ic).

Red-bellied Woodpeckers. — Most interspe-
cific interactions occurred between Red-cock-
aded and Red-bellied woodpeckers (186 of
306, 61%). Of the 186 interactions between
Red-cockaded and Red-bellied woodpeckers,
the latter won 174 (94%). Red-bellied Wood-
peckers successfully usurped foraging Red-
cockaded Woodpeckers in all but 4 of 168 for-
aging interactions. Red-bellied Woodpeckers
frequently foraged within sight of Red-cock-
aded Woodpeckers but usurped them only af-
ter the Red-cockaded Woodpecker had found
food. We also observed Red-bellied Wood-

350

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 2. Variation in habitat characteristics and the rate (x ± 1 SE) of all interspecific interactions (per
hour of observation), foraging and non-foraging interactions, and interactions with Red-bellied Woodpeckers,
with breeding Red-cockaded Woodpeckers from different groups at the Avon Park Air Eorce Range, 1995-1996.

Hardwood

RCW group

Pine basal area
(m-/ha)

basal area
( m-/ha)

All interactions

Foraging

interactions

Non-foraging

interactions

Red-bellied Woodpecker
interactions

8.13

0.0

0.16

-h

0.09

0.11

0.07

0.05 ±

0.04

0.10

0.06

12.44

0.0

0.19

-h

0.05

0.15

0.05

0.04 ±

0.04

0.11

0.04

8.97

0.0

0.20

-F

0.06

0.17

0.01

0.04 ±

0.02

0.14

0.06

6.6

0.28

0.20

0.06

0.15

0.01

0.06 ±

0.03

0.14

0.06

7.49

0.07

0.10

0.05

0.06

0.03

0.04 ±

0.04

0.02

9.28

0.04

0.10

0.06

0.04

0.02

0.06 ±

0.06

0.04

0.02

1 1.34

0.03

0.42

0.09

0.27

0.09

0.15 ±

0.05

0.22

0.09

9.18

0.46

0.22

0.07

0.18

0.07

0.04 ±

0.02

0.15

0.05

10.25

0.44

0.17

0.11

0.07

0.05 ±

0.04

0.11

0.07

6.55

0.01

0.42

0.15

0.34

0.13

0.08 ±

0.03

0.35

0.14

12.15

0.86

0.22

0.09

0.16

0.08

0.07 ±

0.03

0.13

0.07

12.33

0.0

0.34

0.09

0.24

0.07

0.09 ±

0.04

0.22

0.02

peckers following Red-cockaded Woodpeck-
ers as they foraged between different pine
stands. Red-cockaded Woodpeckers won only
12 interactions with Red-bellied Woodpeck-
ers: 4 foraging interactions and 10 non-for-
aging interactions.

The rate of interactions between Red-bel-
lied Woodpeckers and breeding Red-cockaded
Woodpeckers varied monthly (Kruskal-Wallis
One Way ANOVA: x' = 19.3, df = 11, P =
0.055; Fig. Id); however, no consistent pattern
was evident. The rate of interactions did not
differ between the breeding and non-breeding
season (Mann-Whitney [/-test: Z = -0.59, P
> 0.05). No sex-related difference existed in
the rate of interactions between Red-cockaded
and Red-bellied woodpeckers (x^ = 5.13, df
= 1, P > 0.05).

Habitat relationships. — Red-cockaded
Woodpeckers foraged predominately (93.8%
of observation time) in pine flatwood, scrubby
flatwood, and pine plantation habitats (Bow-
man et al. 1998, unpubl. data). The frequency
of foraging and non-foraging interactions did
not differ from the relative frequency of hab-
itats used by foraging Red-cockaded Wood-
peckers (x^ — 4.74 and 2.18, df = 4 and 2,
respectively, P > 0.05). We also found no sig-
nificant correlations between the rate of inter-
actions and the area of any of the 13 habitat
types in Red-cockaded Woodpecker home
ranges (Pearson correlations: all P > 0.05).
Pine basal area in home ranges varied from

6.6 to 12.4 m^ per ha, and hardwood basal
area varied from 0.0 to 0.44 m^ per ha (Table
2); however, neither the total number of inter-
actions nor foraging or non-foraging interac-
tions were correlated with the basal area of
pines or hardwoods within each home range.

Interactions with Red-bellied Woodpeckers
in different habitat types did not differ from
the relative frequency of habitats used by Red-
cockaded Woodpeckers (x^ = 0.18, df = 3, P
> 0.05). However, the rate of these interac-
tions was positively correlated with the per-
centage of each home range comprised of pine
plantation (Spearman rank correlation: r =
0.62, P < 0.05). No significant correlations
existed between the rate of interactions and
the area of any of the other 12 habitat types
or the pine or hardwood basal area in Red-
cockaded Woodpecker home ranges (Spear-
man rank correlation: all P > 0.05).

Red-cockaded Woodpeckers moved to a
new habitat type following only 5 of 120 (4%)
usurpations by Red-bellied Woodpeckers for
which we had data. No significant difference
existed in the dbh or height of the trees used
by foraging Red-cockaded Woodpeckers
(male or female) before and after usurpation
(Kruskal-Wallis One Way ANOVAs: all P >
0.05), nor did any differences exist in the
height at which Red-cockaded Woodpeckers
(male or female) foraged before and after
usurpation (Kruskal-Wallis One Way ANO-
VAs: all P > 0.05).

liowimin et al. • RED-COCKADED WOODPECKER FORAGING INTERACTIONS

351

DISCUSSION

Thirty-two percent of interspecific interac-
tions were not related to foraging. Non-for-
aging interactions were highly seasonal, oc-
cuiTing during the breeding season for most
species. Many of these interactions may have
been related to nest and/or fledgling defense
as many occurred near nests or young of the
species interacting with Red-cockaded Wood-
peckers. Although these interactions were sea-
sonal and relatively infrequent in our popu-
lation, other forms of non-foraging interac-
tions (e.g., cavity competition) could play an
important role in the dynamics of Red-cock-
aded Woodpecker populations (Kappes and
Harris 1995).

Most interspecific interactions were related
to foraging and occurred between Red-cock-
aded and Red-bellied woodpeckers. Red-cock-
aded Woodpeckers lost virtually all foraging
interactions with Red-bellied Woodpeckers.
Red-cockaded Woodpeckers interacted fre-
quently with Downy Woodpeckers, winning
most encounters. Therefore, the latter inter-
actions likely had no deleterious impacts on
Red-cockaded Woodpeckers.

Habitat use, foraging behavior, and diet of
Red-cockaded and Red-bellied woodpeckers
appear to be dissimilar. Red-bellied Wood-
peckers use most habitats occurring within
their range (Sprunt 1954, Breitwisch 1977 and
references within) but may prefer hardwood
habitats (Short 1982, Root 1988). In Florida,
their use of tree species for foraging is diverse
and varies by habitat type (Breitwisch 1977).
Red-bellied Woodpeckers spend 20-69% of
their foraging time on dead trees (Williams
1975, Breitwisch 1977, Williams and Batzli
1979). In contrast, Red-cockaded Woodpeck-
ers forage almost exclusively on living pines
(Hooper and Lennartz 1981 ) in relatively open
pine forests.

In south Florida pine habitat, Breitwisch
(1977) observed foraging Red-bellied Wood-
peckers gleaning and probing (80%) but rarely
excavating (10%). At Avon Park AFR, for-
aging Red-cockaded Woodpeckers used sur-
face probes (54%) most frequently, excavated
frequently (40%), and rarely gleaned (4%;
Bowman et al.l998, unpubl. data).

Little dietary overlap appears to exist be-
tween Red-cockaded and Red-bellied wood-

peckers (Beal 191 1). Red-bellied Woodpecker
stomachs (n — 271) contained 3 1 % animal
matter, of which 6% was ants; Red-cockaded
Woodpecker stomachs (n = 76) contained
81% animal matter, of which 56% was ants.
Both species consumed a similar percentage
of beetles (~10%); however, little overlap ex-
isted in the remaining fraction of animal mat-
ter.

Niche overlap between these two species
appears to be low, even in south Florida, yet
interactions between Red-bellied and Red-
cockaded woodpeckers appear to be higher
here than elsewhere in their ranges. It is pos-
sible that these interactions are simply over-
looked elsewhere, especially if they are more
frequent outside of the breeding season. If so,
and these interactions have deleterious im-
pacts on Red-cockaded Woodpeckers, then
they should be examined more closely else-
where. However, geographical variation in in-
terspecific competition may be real and be
caused by variation in population densities of
the species (Thompson 1988), indirect effects
as species assemblages change, the productiv-
ity or vegetation composition of habitats
(Travis 1 996) or some interaction of these fac-
tors.

Data on the density of Red-bellied Wood-
peckers across their range are not available; in
general they appear as abundant in Florida as
elsewhere in the southeastern coastal plain
(Bock and Lepthien 1975, Root 1988, Price et
al. 1995). At Avon Park AFR, the density of
Red-cockaded Woodpeckers is low compared
to populations outside of peninsular Florida
(Bowman et al. 1998, unpubl. data). Data on
the regional variation in density of both Red-
cockaded and Red-bellied woodpeckers are
needed to determine whether differences in
density contribute to variations in interspecific
interactions.

Indirect effects related to the presence of
other species may have contributed to the high
rate of observed interactions. At Avon Park
AFR, five species of woodpeckers and the
Brown-headed Nuthatch are sympatric with
Red-cockaded Woodpeckers; however, many
of these species are sympatric in pine habitats
outside of peninsular Florida. The abundance
and diversity of species utilizing similar re-
sources in different habitats may contribute to

352

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

variation in the rate of interspecific interac-
tions.

Differences in pine forests between south-
central Florida and more temperate forests
may have contributed to the relatively high
rates of interactions with Red-bellied Wood-
peckers. In southern Florida, most Red-cock-
aded Woodpeckers occur in mesic and hydric
flatwoods. These habitats have lower hard-
wood basal area than do more temperate pine
communities (Beever and Dryden 1992, pers.
obs.). Elsewhere, Red-bellied Woodpeckers’
preference for hardwoods may minimize their
foraging overlap with Red-cockaded Wood-
peckers, but we know little about habitat-spe-
cific foraging strategies of either species. Al-
though hardwood basal area varied among the
12 Red-cockaded Woodpecker home ranges,
overall, basal area was low and was not cor-
related with the frequency of Red-bellied
Woodpecker interactions.

All Red-cockaded Woodpecker populations
in peninsular Florida support fewer than 50
groups (Cox et al. 1995). In peninsular Flor-
ida, Red-cockaded Woodpeckers have larger
home ranges (Nesbitt et al. 1981; DeLotelle
et al. 1983; Bowman et al. 1998, unpubl. data)
and produce fewer fledglings (Jansen and Pat-
terson 1983; DeLotelle and Epting 1992;
Bowman et al. 1998, unpubl. data) than other
populations. These characteristics suggest that
these populations may occupy relatively poor
quality habitat; however, few correlations ex-
ist between various measures of Red-cockad-
ed Woodpecker demography and habitat char-
acteristics (Beyer et al. 1996; Bowman et al.
1998, unpubl. data). Although these results do
not suggest a deleterious effect of interspecific
competition, the relatively high rates we doc-
umented bear further investigation, especially
where these interactions have not been re-
ported. Aggressive interaction between spe-
cies is not sufficient to demonstrate competi-
tion, but interspecific competition may con-
tribute to variation in the abundance and re-
productive potential of species. It is possible
that some synergistic interaction of habitat
and community structure, such as competition,
may be related to regional differences in Red-
cockaded Woodpecker demography.

ACKNOWLEDGMENTS

Financial support was provided by the United States
Department of Defense. We thank P. Ebersbach, B.

Progulske, P Walsh, and K. Olson for cooperation
throughout the course of this study. T. Dean, D. Barber,
P. Barber, and D. Swan provided very helpful com-
ments on early drafts of this manuscript and R. Conner,
E James, C. Rudolph and J. Thomlinson provided con-
structive reviews of a later version.

LITERATURE CITED

Beal, F. E. L. 1911. Food of the woodpeckers of the
United States. U.S. Dept. Agric. Biol. Surv. Bull.,
37:1-64.

Beever, J. W. and K. A. Dryden. 1992. Red-cockaded
Woodpeckers and hydric slash pine flatwoods.
Trans. N. Am. Wildl. Nat. Res. Conf. 57:693-700.
Beyer, D. E., R. Costa, R. G. Hooper, and C. A.
Hess. 1996. Habitat quality and reproduction of
Red-cockaded Woodpecker groups in Florida. J.
Wildl. Manage. 60:826—835.

Bock, C. E. and L. W. Lepthien. 1975. A Christmas
count analysis of woodpecker abundance in the
United States. Wilson Bull. 87:355-366.

Bowman, R., D. L. Leonard, Jr., L. K. Backus, P. M.
Barber, A. R. Mains, L. M. Richman, and D.
Swan. 1998. Demography and habitat character-
istics of the Red-cockaded Woodpecker (Picoides
borealis) at the Avon Park Air Force Range: Final
Report, 1994-1998. Department of Defense,
Avon Park, Florida.

Breitwisch, R. j. 1977. The ecology and behavior of
the Red-bellied Woodpecker, Centurus carolinus
(Linnaeus) (Aves: Picidae), in south Florida.
M.Sc. thesis, Univ. of Miami, Miami, Florida.
Conner, R. N. and D. C. Rudolph. 1991. Forest hab-
itat loss, fragmentation, and Red-cockaded Wood-
pecker populations. Wilson Bull. 103:446—457.
Cox, J., W. W. Baker, and D. Wood. 1995. Status,
distribution, and conservation of the Red-cock-
aded Woodpecker in Florida: a 1992 update. Pp.
475-464 in Red-cockaded Woodpecker: recovery,
ecology, and management (D. L. Kulhavey, R. G.
Hooper, and R. Costa, Eds.). College of Forestry,
Stephen F. Austin State Univ., Nacogdoches, Tex-
as.

DeLotelle, R. S. and R. J. Epting. 1992. Reproduc-
tion of the Red-cockaded Woodpecker in central
Florida. Wil.son Bull. 104:285—294.

DeLotelle, R. S., J. R. Newman, and A. E. Jerauld.
1983. Habitat use by Red-cockaded Woodpeckers
in central Florida. Pp. 59-67 in Proceedings of
the Red-cockaded Woodpecker symposium 11
(D.A. Wood, Ed.). Florida Game and Fresh Water
Fish Comm., Tallahassee.

Engstrom, R. T. 1993. Characteristic mammals and
birds of longleaf pine forests. Proc. Tall Timbers
Fire Ecol. Conf. 18:127-138.

Harlow, R. F. and M. R. Lennartz. 1983. Interspe-
cific competition for Red-cockaded Woodpecker
cavities during the nesting season in South Caro-
lina. Pp. 41-43 in Red-cockaded Woodpecker
symposium II (D. A. Wood, Ed.). Florida Game
and Fresh Water Fish Comm., Tallahassee.

Bowman et al. • RED-COCKADED WOODPECKER EORAGING INTERACTIONS

353

Hooper, R. G. and M. R. Lennartz. 1981. Eoraging
behavior of the Red-eockaded Woodpecker in
South Carolina. Auk 98:321-334.

Jackson, J. A. 1971. The evolution, taxonomy, distri-
bution, past populations and current status of the
Red-cockaded Woodpecker. Pp. 4-29 in The ecol-
ogy and management of the Red-cockaded Wood-
pecker (R. L. Thompson, Ed.). Bur. Sport Eish.
Wildl., U.S. Dept, of Interior, and Tall Timbers
Research Station, Tallahassee, Florida.

Jackson, J. A. 1978. Competition for cavities and Red-
cockaded Woodpecker management. Pp. 103-112
in Endangered birds: management techniques for
threatened species (S. A. Temple, Ed.). Univ. of
Wisconsin Press, Madison.

James, E C. 1991. Signs of trouble in the largest re-
maining population of Red-cockaded Woodpeck-
ers. Auk 108:419-423.

Jansen, D. K. and G. A. Patterson. 1983. Late nest-
ing and low reproductive success of Red-cockad-
ed Woodpeckers in the Big Cypress National Pre-
serve, Florida, in 1983. Pp. 99 in Red-cockaded
Woodpecker symposium II (D. A. Wood, Ed.).
Florida Game and Fresh Water Fish Comm., Tal-
lahassee.

Kappes, j., Jr. and L. D. Harris. 1995. Interspecific
competition for Red-cockaded Woodpecker cavi-
ties in Apalachicola National Forest. Pp. 389-393
in Red-cockaded Woodpecker: recovery, ecology,
and management (D. L. Kulhavey, R. G. Hooper,
and R. Costa, Eds.). College of Forestry, Stephen
F. Austin State Univ., Nacogdoches, Texas.

Lennartz, M. R., P. H. Geissler, R. F. Harlow, R. C.
Long, K. M. Chitwood, and J. A. Jackson. 1983.
Status of the Red-cockaded Woodpecker on fed-
eral lands in the south. Pp. 7—12 in Red-cockaded
Woodpecker Symposium II (D. A. Wood, Ed.).

Florida Game Fresh Water Eish Comm., Tallahas-
see.

Ligon, j. D. 1970. Behavior and breeding biology of
the Red-cockaded Woodpecker. Auk 87:255-278.

Morse, D. H. 1970. Ecological aspects ol'some mixed-
species foraging flocks of birds. Ecology 40: 1 19-
168.

Nesbitt, S. A., D. T. Gilbert, and D. B. Barbour.
1978. Red-cockaded Woodpecker fall movements
in a Florida flatwood community. Auk 95:145-
151.

Nesbitt, S. A., B. A. Harris, A. E. Jerauld, and
C. B. Brownsmith. 1981. Report of the investi-
gation of Red-cockaded Woodpeckers in Charlotte
County. Florida Game and Fresh Water Fish
Comm., Tallahassee.

Price. J., S. Droege, and A. Price. 1995. The summer
atlas of North American birds. Academic Press.
San Diego, California.

Root, T. 1988. Atlas of wintering North American
birds. Univ. of Chicago Press, Chicago, Illinois.

Short, L. L. 1982. Woodpeckers of the world. Dela-
ware Museum of Natural History, Greenville.

Sprunt, a., Jr. 1954. Florida bird life. Coward-Mc-
Cann, Inc., New York.

Thompson, J. N. 1988. Variation in interspecific inter-
actions. Annu. Rev. Ecol. Syst. 19:65-87.

Travis, J. 1996. The significance of geographic vari-
ation in species interactions. Am. Nat. 148:S1-S8.

U.S. Fish and Wildlife Service. 1985. Red-cockaded
Woodpecker recovery plan. U.S. Fish and Wildlife
Service, Atlanta, Georgia.

Williams, J. B. 1975. Habitat utilization by four spe-
cies of woodpeckers in a central Illinois wood-
land. Am. Midi. Nat. 93:354-367.

Williams, J. B. and G. O. Batzli. 1979. Interference
competition and niche shifts in the bark-foraging
guild in central Illinois. Wilson Bull. 91 :400-41 1.

Wilson Bull.. 111(3), 1999, pp. 354-362

SPATIAL AND TEMPORAL DYNAMICS OF A PURPLE MARTIN

PRE-MIGRATORY ROOST

KEVIN R. RUSSELL' 23 AND SIDNEY A. GAUTHREAUX, JR.‘

ABSTRACT. — We used simultaneous WSR-88D radar (NEXRAD) and direct visual observations to investi-
gate the spatial and temporal dynamics of a Purple Martin (Prague suhis) pre-migratory roost in South Carolina.
The timing of mass flights of martins from and to the roost was related to levels of ambient light. Each morning,
the birds first departed approximately 40 min before sunrise independent of date, with peak departures occurring
about 10 min before sunrise. The time of evening flights was more variable, but peak movement of birds into
the roost consistently occurred at sunset. Purple Martins exited the roost in organized, annular departures (360°)
that were visible on radar up to 100 km away from the roost, but returned to the roost over an extended period
in scattered flocks. During morning departures we recorded flight speeds up to 13.4 m/s. Radar echoes corre-
sponding to martin flights were recorded farther from the roost, and flights from and to the roost occuiTed later
and earlier, respectively, in response to increased cloud cover. The departures of birds from the roost appeared
to be displaced by winds aloft. At the peak of the roosting season in late July, the total roost population was
estimated to be at least 700,000 birds. Received 18 Aiig. 1998, accepted 5 Feb. 1999.

Purple Martins (Progne siibis) are neotrop-
ical migratory swallows that breed across
North America (Brown 1997, AOU 1998). In
eastern North America, Purple Martins are
conspicuous colonial nesters that almost ex-
clusively are dependent on man-made nesting
houses. As a result, martin breeding biology
and behavior have been the focus of consid-
erable study (see Brown 1997). After the
fledging period eastern populations of Purple
Martins often congregate in distinctive noc-
turnal roosts that may reach concentrations of
100,000 or more as a prelude to fall migration
(Brown 1997). From late June through August
or early September these assemblages engage
in two mass movements daily; a morning ex-
odus from the roost for aerial foraging and an
evening return (Allen and Nice 1952). Al-
though the existence of these large pre-migra-
tory roosts is well documented (e.g.. Bent
1942, Allen and Nice 1952, Anderson 1965,
Brown and Wolfe 1978, Rogillio 1989, Rus-
sell et al. 1998), little quantitative data are
available concerning their spatial and tempo-
ral dynamics.

Studies of communal roosts may be limited
by the inability to collect data at appropriate
spatial and temporal scales (Caccamise et al.

' Dept, of Biological Science.s. Clemson Univ.,
Clem.son. SC 29634.

2 Current addre.ss: Willamette Industries, Inc., P.O.
Box 488. Dallas. OR 97338-0488.

’ Corresponding author;

E-mail; krussell@wii.com

1983, Russell et al. 1998). Locating roosts of-
ten is labor-intensive (see Caccamise and
Fischl 1985, Komar 1997). Even when roost
locations are known, visual surveys alone pro-
vide only limited data on the spatial extent
and direction of roosting flights (Brown and
Wolfe 1978), and the time of movements (e.g.,
pre-dawn flights) may limit visual observa-
tions (Russell and Gauthreaux 1998). Recent-
ly, we developed methods for using the Na-
tional Weather Service’s new doppler weather
surveillance radar, the WSR-88D or NEX-
RAD to locate and study communal roosting
assemblages (Russell and Gauthreaux 1998,
Russell et al. 1998). In this study, we provide
quantitative data on the spatial and temporal
dynamics of a pre-migratory roost of Purple
Martins in South Carolina using simultaneous
WSR-88D radar and direct visual observa-
tions. Our specific objectives were to docu-
ment the daily timing and spatial pattern of
roosting flights, flight directions and speeds,
the influence of weather conditions on roost-
ing flights, and seasonal changes in roost pop-
ulation size.

METHODS

Study area. — Our study site was a Purple Martin
roost on Lunch fsland (34° 03' N. 81° 18' W), a 5-ha
island located in Lake Murray, near Columbia, South
Carolina (Russell and Gauthreaux 1998: fig. 2). We
conducted visual surveys from a peninsula on the
shoreline of the lake, approximately 3 km south of the
roost. Radar data were collected at the National Weath-
er Service Office, Columbia Metropolitan Airport, 28
km southeast of the roost site. A detailed de.scription

354

Russell and Gauthreaux • PURPLE MARTIN ROOSTING DYNAMICS

355

of the study area is provided in Russell (1996) and
Russell and Gauthreaux (1998).

Sun’ey methods. — We conducted timed visual sur-
veys of roosting flights 5 days per week from 30 June
to 27 August 1995. We originally conducted morning
surveys from 05:50 to 06:50 EST, but extended them
to 07:30 when the duration of departures lengthened
in late July. Evening surveys were conducted from 19:
15 to 21:15 throughout the study. During each visual
survey, we recorded the numbers and flight directions
of individual martins as they passed over a natural cir-
cular opening (24 m diameter; 48° angle of observa-
tion) in the forest canopy, using methods described by
Lowery and Newman ( 1963) and applied to WSR-88D
radar by Russell and Gauthreaux ( 1998). We identified
martins in flight by their distinctive profile, behavior,
and vocalizations (Brown 1997). The presence of other
roosting species was negligible during the study (Rus-
sell 1996). When both numbers and flight directions
could not be accurately recorded, we ( I ) assigned di-
rections to flocks, or (2) only made counts for many
of the birds. During surveys we also made incidental
observations of the roost site with a 30X spotting
scope and recorded general weather conditions. We lat-
er obtained detailed local climatological data (LCD)
from the National Weather Service Office at the Co-
lumbia Metropolitan Airport.

Immediately after each visual survey we visited the
National Weather Service Office and acquired WSR-
88D radar images that coincided with our surveys, ex-
cept for eight mornings when weather conditions ob-
scured martin flights on radar. Evening flights of Pur-
ple Martins typically failed to produce recognizable
patterns on the WSR-88D (Russell and Gauthreaux
1998); thus we present no radar data on evening arriv-
als. We collected base reflectivity images to monitor
flights of martins in the proximity of the roost and
composite reflectivity images to track flights over a
wider geographic area (Russell and Gauthreaux 1998).
We also collected several radial velocity images to de-
termine directions and speeds of each mass flight. We
recorded radar images by taking a 35-mm color slide
exposure of each image as it was displayed on a mon-
itor at the radar site. Detailed descriptions of the WSR-
88D, radar images, and our survey methods are pro-
vided in Russell and Gauthreaux (1998).

Analyses. — We examined the timing of roosting
flights by recording the beginning, end, duration, and
peak of each mass flight from both the radar and visual
data (Russell and Gauthreaux 1998). We also exam-
ined temporal changes in the numbers of departing and
arriving birds by pooling our visual counts into 15-min
totals ba.sed on our timed surveys (Russell 1996). We
characterized the spatial patterns of roosting flights by
examining the origin, timing, intensity, number, and
spatial distribution of echoes from each radar image of
the roost area (Russell and Gauthreaux 1998). We mea-
sured the maximum spatial extent of each mass flight
by relating echoes created from the leading edge of the
flight to landmarks depicted on the radar (e.g., county
boundaries). We excluded mornings when regional

precipitation or spurious ground echoes obscured the
maximum extent of flights.

We pooled visual flight tracks of individual birds so
that peak movement from or to the roost was not di-
vided artificially (Russell 1996); then examined tem-
poral changes in flight directions using circular statis-
tics (Zar 1984). We scored directions and speeds of
each mass flight directly from the radar images (Ru.s-
sell and Gauthreaux 1998).

We used univariate correlation and stepwise multi-
ple regression to examine the influence of weather con-
ditions on roosting flights. Daily climatological data
(mean, 04;00, 07;00, 19:00) included temperature, dew
point, total precipitation, atmospheric pressure, surface
wind speed, visibility, and cloud cover. Variables ini-
tially entered or left the stepwise models at P = 0. 1 .
We examined the influence of winds aloft on morning
departures by scoring the modal azimuth of annular
displacement (relative to the roost site; Eastwood et al.
1962) and regressing these values against the corre-
sponding direction of geostrophic winds aloft (1000 m
above the ground or 900 mb; Riehl 1972) at 07:00.

The total roost population (TRP) represents the es-
timated daily sum of all birds at the roost (Caccamise
et al. 1983). Based on a preliminary examination of
the radar data, we assumed that departures and arrivals
of Purple Martins were omnidirectional at approxi-
mately equal densities. Our visual sample area was
0.47° or 1/766 of the circumference around the roost
at a distance of 3 km. Thus, we calculated separate
morning and evening TRPs by multiplying the total
count for each movement by 766 (Russell 1996) and
examined changes in roost size. We used univariate
correlation to investigate relationships between our
morning and evening counts, and patterns of flight
with time and date. We used JMP 3.1 (SAS Institute
1995) on an IBM computer for all analyses with prob-
ability values P < 0.05 recognized as significant and
means reported ± 1 standard deviation (SD) unless
otherwise noted. We standardized all survey times by
converting to minutes before or after sunrise or sunset
(as given by the National Weather Service Office in
Columbia, South Carolina).

RESULTS

Timing of flights. — We detected the initia-
tion of mass departures as a single radar echo
corresponding to the location of the roost.
Time of initial departure from the roost aver-
aged 41.4 ± 4.0 (SD) min before sunrise
(range 31—48 min) and was independent of
date {r- = 0.02, P > 0.05, n = 32; Russell
and Gauthreaux 1998). Time of peak depar-
ture also was independent of date (r-' = 0.06,
P > 0.05, n = 32). Mean duration of morning
departures (interval from first radar echo to
last bird observed visually) was 67.9 ± 12.2
min (range 49-101 min) and was positively

356

THE WILSON BULLETIN • Vol. ///, No. 3, September 1999
A B

TIME BEFORE (-)/AFI ER (+) SUNRISE (MIN) TIME BEFORE (+)/AFI'ER (-) SUNSET (MIN)

FIG I . Changing numbers of roosting Purple Martins during visual counts of (A) morning departures and
(B) evening arrivals, by time before and after sunrise or sunset.

related to the number of birds counted exiting
the roost (r^ = 0.60, P < 0.001, n = 32).

Changes in the numbers of Purple Martins
observed departing the roost were related to
the time of sunrise. Because some birds first
departed in darkness, changes in the percent-
age of birds we observed 40—30 min before
sunrise reflected their increased visibility in
the morning sky rather than actual numbers
aloft (Fig. lA). Corresponding radar data in-
dicated that, on average, the movement of
most birds from the roost began 20.4 ± 4.6
min before sunrise, when sufficient light was
available for accurate visual counts (Russell
and Gauthreaux 1998). The number of birds
departing from the roost consistently peaked
about 10 min before sunrise, then steadily de-
clined over the next 40 min (Fig. lA). We
occasionally observed a small, secondary in-
crease in the number of birds departing from
the roost as late as 30-40 min after sunrise
(Fig. lA).

Purple Martins first returned to the roost an
average of 59.8 ± 13.5 min before sunset
(range 35-84 min), and the last birds entered
the roost 2-24 min after sunset (T = 15.5 ±
5.9 min). We found a weak negative relation-
ship between the time that the last birds were
observed entering the roost and date (r =
0.23, P = 0.0033, n = 36), but not for the
time of peak movement into the roost (r-’ =
0.07, P > 0.05, n = 36). Mean duration of
observed evening arrivals at the roost (71.4 ±

15.2 min, range 43-103 min) was not related
to our visual counts (r-’ = 0.05, P > 0.05, n
= 36), but we detected a negative relationship
between the duration of evening flights and
date = 0.71, P < 0.001, n = 36).

Numbers of Purple Martins returning to the
roost increased 50-30 min before sunset (Fig.
IB). At about 30 min before sunset, the num-
ber of arrivals increased rapidly until peaking
at sunset (Fig. IB). After sunset, numbers
aloft declined rapidly and ceased by 24 min
after sunset, and always before complete dark-
ness.

Spatial pattern of flights. — Each morning
until the peak of the roosting season. Purple
Martins departed en masse from the roost in
all directions (360°). These omnidirectional
departures first were visible on radar as a
roughly circular mass of echoes that extended
across Lake Murray (Russell and Gauthreaux
1998: fig. 3). As departures continued, an ex-
panding ring or annulus was seen on radar as
the departing birds extended across the land-
scape directly away from the roost (Russell
and Gauthreaux 1998: fig. 4). The maximum
daily extent of flights we could detect on radar
averaged 77.9 ± 14.1 km (range 50-100 km,
n = 20; Russell and Gauthreaux 1998); We
found no relationship between the maximum
distance of flights and date (r^ = 0.02, P >
0.05, n = 20) or our morning counts =
0.01, P > 0.05, n = 20). Beginning 12 August
we observed on radar that more birds departed

Russell (uul Gauthreaux • PURPLE MARTIN ROOSTING DYNAMICS

357

south than other directions; by 18 August
birds only departed to the south and after 26
August we no longer observed departures on
radar.

In the evening. Purple Martins arrived at
the roost in sporadic, loosely organized flocks.
The birds flew over the observation point at
tree-top height (ca 18-25 m) and then flew
down to just above the lake’s surface, usually
below the radar beam (Russell and Gauth-
reaux 1998). After reaching the roost the birds
remained aloft in a growing mass that circled
high over the island counter-clockwise until a
final, spiraling entry into the roost at sunset.

Flight directions and speeds. — During
morning departures, the Purple Martins we
observed visually were consistently and sig-
nificantly {P < 0.001) oriented to the south.
The mean direction (±95% confidence inter-
val, vector length, angular deviation) of mar-
tins at 40-21 min before sunrise was 180° (2°,
0.953, 18°); at 20-1 min before sunrise was
184° (3°, 0.922, 23°); at 0-20 min after sunrise
was 183° (4°, 0.884, 28°); at 21-40 min after
sunrise was 182° (4°, 0.885, 28°); and at 41-
60 min after sunrise was 190° (3°, 0.931, 21°).
The radar images also showed the birds flying
in a southerly direction over the observation
point, consistent with an omnidirectional de-
parture from the roost. Radar images of ve-
locity that we acquired on 32 mornings
showed that flight speeds of the departing
birds ranged from 10.8-13.4 m/s (see figure 5
in Russell and Gauthreaux 1998).

We observed distinct changes in the orien-
tation of Purple Martins during evening flights
to the roost. The mean direction of martins at
90-71 min before sunset was 141° (90°, 0.364,
65°) and not significant (P > 0.05). During
this time we often observed birds feeding in
small flocks and flying away from the roost.
As the time of sunset approached, however,
the mean direction of martins at the observa-
tion point became significant (P < 0.001) and
increasingly oriented towards the northeast,
but still east of the roost (azimuth = 10° from
observation point). The mean direction of
martins at 70-51 min before sunset was 93°
(18°, 0.318, 67°); at 50-31 min before sunset
was 73° (14°, 0.399, 63°); at 30-1 1 min before
sunset was 57° (9°, 0.590, 52°); at 10 min be-
fore to 9 min after sunset was 54° (5°, 0.836,

33°); and at 10-29 min after sunset was 53°
(4°, 0.843, 32°).

Weather conditions and roosting flights. —
Purple Martins first departed the roost earlier
relative to sunrise with higher atmospheric
pressure, surface wind speed, and relative hu-
midity, but later relative to sunrise with in-
creased cloud cover (stepwise multiple regres-
sion: = 0.66, P < 0.001, n = 32). Of these

variables, atmospheric pressure was the most
important and the only significant univariate
relationship {r- = 0.20, P = 0.0068, n = 32).
We also found a significant relationship be-
tween the duration of morning departures and
atmospheric pressure (stepwise multiple re-
gression: R~ = 0.29, P < 0.001, n = 32).

Purple Martins were detected on radar far-
ther from the roost on days with increased
cloud cover or decreased visibility at 07:00
(stepwise multiple regression: R^ = 0.61, P <
0.001, n = 20), although univariate analysis
indicated a significant relationship only for
cloud cover (ri = 0.35, P = 0.0063, n = 20).
Mass departures from the roost also appeared
to be displaced by winds aloft. Azimuthal
wind direction was strongly related to the
modal direction of annulus displacement (r- =
0.76, P < 0.001, n = 20).

Arrival of the first birds over the observa-
tion point was not related to any of the weath-
er variables, but arrival of the last martins at
the roost occurred earlier relative to sunset on
days with more cloud cover and later in the
season (stepwise multiple regression: R~ =
0.40, P < 0.001, n = 36). Univariate analysis
also indicated a significant relationship for
cloud cover: (ri = 0. 13, P = 0.0275, n = 36).
Maximum temperature, atmospheric pressure,
and surface wind speed at 19:00 were identi-
fied by stepwise regression as significant pre-
dictors of the duration of evening flights {R~
= 0.89, P < 0.001, n = 36), although no var-
iable was significant when subjected to uni-
variate analysis.

Seasonal changes in total roost popula-
tion.— Despite fluctuations in the daily number
of departing (x = 368.3 ± 201.9 SD, range 29-
916, n — 14,732 birds) and returning (x =
755.1 ± 548.9 SD, range 108-2,531, n =
27,182 birds) martins, the roosting population
exhibited a seasonal pattern of growth and de-
cline (Fig. 2). On the first morning census of
3 July we counted only 172 birds over the ob-

358

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

MORNING EXODUS

EVENING RETURN

FIG 2. Seasonal changes in visual counts of roost-
ing Purple Martins during (A) morning departures and
(B) evening arrivals.

servation point yielding an estimated total roost
population (TRP) of 131,752 birds, but on 31
July we counted 916 birds, yielding a peak
morning TRP of 701,656 birds. On the last
morning census (27 August), we counted only
29 birds (TRP = 22,214). Although evening
counts usually were higher than morning
counts, numbers of birds returning to the roost
also followed a pattern of increase and decline
(Fig. 2). The first (30 June), peak (17 July),
and last (27 August) evening counts yielded
TRPs of 301,804 birds, 1,938,746 birds, and 0
birds, respectively. However, the peak morning
count occurred two weeks after the peak even-
ing count and morning counts were not related
to evening counts from the previous night (r"
= 0.001, P > 0.05, n = 32) or the same night
(r = 0.01, P > 0.05, n = 32).

DISCUSSION

Timing of flights. — Our observations indi-
cate that flights of Purple Martins from and to
pre-migratory roosts are related to levels of

ambient light. Purple Martins (Brown and
Wolfe 1978, Oren 1980, Hill 1988), Crag
Martins (Ptyonoprogne rupestris', Elkins and
Etheridge 1974), and other species of swal-
lows (Rudebeck 1955, Loske 1984, Skutch
1989, Komar 1997) have been shown previ-
ously to enter and exit roosts in response to
changing intensities of light. Although endog-
enous factors likely are a significant influence
on the timing of departures and arrivals of
birds at roosts (Aschoff 1967), cueing on
changing light levels has been suggested as a
selective advantage for maximizing time
available for feeding (Eastwood et al. 1962,
Summers and Eeare 1995).

The timing of evening flights was less pre-
dictable than morning departures and ap-
peared to be related to date. Across a season
the departure or arrival times of the first and
last birds at roosts may show greater variation
than peaks of departure or arrival (Summers
and Eeare 1995). In our study, more variation
was associated with the times martins first ar-
rived in the vicinity of the roost than when
flights were terminated, as was reported pre-
viously for other roosting species (Eastwood
et al. 1962, Meanley 1965, Bunning 1973,
Krantz and Gauthreaux 1975). We suggest
that relationships between date and the timing
of evening arrivals of martins are related
mostly to the large variation associated with
these flights.

Spatial pattern of flights. — As with other
roosting species (Harper 1959, Eastwood et al.
1962), the omnidirectional departures of Pur-
ple Martins that we observed on radar were
associated with daily feeding flights (Russell
1996). In contrast, directionally biased pat-
terns of flight on radar are often associated
with migration departures (Harper 1959, Rich-
ardson and Haight 1970). The strong southerly
bias in the annulus and its eventual disap-
pearance during the latter part of August, in
conjunction with our declining visual counts,
marked increasing fall migration departures
and eventual abandonment of the roost.

Among our more surprising results was the
long distances Purple Martins flew from the
roost. Prior to our study, it was believed that
the birds foraged within 10-15 km of the
Lake Murray roost (J. Cely, pers. comm.).
Brown and Wolfe (1978) suggested that Pur-
ple Martins may travel as far as 48 km from

Russell and Gauthreaux • PURPLE MARTIN ROOSTING DYNAMICS

359

pre-niigratory roosts during the day while
feeding. Our radar data clearly showed that
birds regularly flew almost 80 km and occa-
sionally as far as 100 km from the roost. If
the large population size of the roost and thus
foraging competition (Caccamise et al. 1983,
Sunmiers and Feare 1995) was responsible for
the flight distances we observed, a significant
relationship should have emerged between the
extent of departures and changes in TRP, but
it did not. In fact, some of the longest flights
occurred early in the season, when TRP was
relatively small. Because we did not track in-
dividuals, we do not know the ultimate des-
tination of the birds or whether all departing
individuals returned the same evening.

Flight directions and speeds. — In marked
contrast to the uniform departures of Purple
Martins, we observed high variability in the
flight tracks of birds arriving at the roost.
Even during the peak of arrivals, mean flight
direction of the birds was east of the roost.
This directional bias and our own incidental
observations indicate that Purple Martins re-
turn to the roost via specific flight corridors.
Use of preferred flight lines often is associated
with evening movements to roosts (Eastwood
et al. 1962, Meanley 1965, Skutch 1989). Af-
ter arriving at the roost the birds assembled in
a high, circling mass before a final entry at
sunset. This pattern is very similar to the
phases of “staging” or pre-settling behavior
(initial aimless flight, formation of a tight
flock near the roost, final descent; Loske
1984) previously described for Prague spp.
(Brown and Wolfe 1978, Oren 1980, Hill
1988) and other genera of roosting swallows
(Bent 1942; Loske 1984, 1986; Skutch 1989,
Komar 1997). Flight speeds of departing Pur-
ple Martins we recorded on radar (10.8-13.4
m/s) were similar to those reported by Evans
and Drickamer (8.45-11.09 m/s; 1994) and
Southern (12.1 m/s; 1959).

Weather conditions and roosting flights. —
Because Purple Martins apparently respond to
some threshold level of ambient light for the
cueing of flights from and to the roost, the
timing of these movements should vary with
daily differences in light levels caused by
changing cloud cover (Richardson 1978, Elk-
ins 1983). In our study, cloud cover and at-
mospheric pressure were the most important
weather conditions explaining variation in the

timing of roosting flights. Increasing or de-
creasing light levels reach a given intensity
later or earlier, respectively, on days with
overcast conditions compared to clear days.
Atmospheric pressure is often inversely cor-
related with the amount of cloud cover (Riehl
1972). The inclusion of other variables into
the stepwise models also may have resulted
from intercorrelations with cloud cover, at-
mospheric pressure, or date (Richardson
1978). Because these weather variables were
highly intercorrelated, the most that can be as-
sumed is that the birds responded to some as-
pect of the weather that was interrelated with
the significant variables (Richardson 1978).
Thus, we view our results as a preliminary
explanation of variation in the flight behaviors
of roosting Purple Martins, rather than repre-
senting causal mechanisms.

Finlay (1976) observed the influence of
cloud cover on Purple Martins during the nest-
ing season; the birds departed from nests ear-
lier on clear days than on cloudy days. Also,
Elkins and Etheridge (1974) found that Crag
Martins returned to roosts up to 2.5 h earlier
on overcast or cloudy days. The influence of
cloud cover on the extent of martin roosting
flights may in part be related to the aerial for-
aging habits of the birds (Elkins 1983). In two
studies of Purple Martin food habits (Spice
1972, Walsh 1978), the intake of airborne in-
sects was negatively correlated with cloud
cover. Cloudy weather also was shown to re-
duce the amount and type of airborne insects
taken by Brown-chested Martins {Phaeoprog-
ne tapera) in Venezuela (Turner 1984). Ad-
ditionally, Finlay (1976) reported that during
overcast conditions martins appeared to spend
more time away from nests, presumably
searching for food. It is possible that cloudy
conditions forced roosting Purple Martins to
travel longer distances in search of prey (Elk-
ins 1983).

Our radar observations that winds aloft dis-
placed the annular departures of Purple Mar-
tins indicate that the birds continued to fly
with constant headings and speeds, apparently
making no correction for drift in spite of their
presumed knowledge of the local topography.
Similar radar observations were made by
Eastwood et al. (1962) of wind displaced de-
partures of European Starlings {Sturnus vul-
garis) from roosts.

360

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

Seasonal changes in total roost popula-
tion.— The overall increase in TRP through
the end of July likely reflects recruitment of
local and regional populations after the fledg-
ing period (Brown 1997). Although Purple
Martins are colonial breeders they nest asyn-
chronously (Brown 1997), and in South Car-
olina clutches have been observed as early as
1 1 April and as late as 19 June (Post and
Gauthreaux 1989). Thus, it is not surprising
that increases in TRP during July occurred
gradually. The influx of birds at the roost also
occurred too early in the season to be signif-
icantly influenced by migration. However,
subsequent declines in TRP and changes in
flight patterns evident on radar throughout
August corresponded with fall migration de-
partures (Hamel 1992, Brown 1997). In the
southeastern United States, Purple Martins of-
ten depart for South America in early August
(Hamel 1992).

Although the roost exhibited a seasonal pat-
tern of growth and decline consistent with pre-
migrator y assembly, large daily fluctuations in
TRP were evident. The accuracy of our TRP
estimates depended on the following assump-
tions: (1) departures and arrivals were equally
distributed around the roost, (2) each bird was
counted only once during a survey, and (3)
birds did not switch among roosts. The an-
nular patterns from radar and uniform flight
tracks over the observation point indicate that
the first two assumptions probably were met
for most morning departures, although some
daily variation was likely. In contrast, the
large variability and directional bias of even-
ing flights and the lack of relationships be-
tween morning and evening counts indicate
that the birds did not return to the roost in a
uniform manner, and at least early in the even-
ing some birds were counted multiple times.
Thus, our evening counts likely did not pro-
vide an accurate estimate of the number of
martins at the roost on a daily basis.

Another potential source of variation in
TRP is movement of birds among roosts. Dur-
ing the present and subsequent (Russell et al.
1998) studies, we discovered another pre-mi-
gratory roost of martins 100 km west in Geor-
gia and others within 230 km. Fluctuations in
TRP through late July may reflect some
switching of birds among roosts on a regional
basis, while temporary increases in TRP late

in the season also could result from birds en-
countering the roost during migration from
more northerly breeding or roosting sites.
Likewise, birds migrating from Lake Murray
probably encounter roosts along the Gulf
coast (Russell et al. 1998). Although we be-
lieve our morning counts provide a good ini-
tial assessment of daily changes in TRP and
maximum roost size, more accurate estimates
will require monitoring roost populations on a
regional basis. In part this may require rigor-
ous, labor intensive ground surveys at multi-
ple roost sites (e.g., Caccamise et al. 1983).
However, with further refinements in meth-
odology, WSR-88D radar may provide the po-
tential to remotely monitor seasonal and an-
nual changes in roost populations over large
geographical areas (Russell and Gauthreaux
1998, Russell et al. 1998).

Why do Purple Martins assemble in pre-
migratory roosts? — Large pre-migratory
roosts of Purple Martins are neither a recent
nor an isolated phenomenon. Wayne (1910),
Stone (1937), and Anderson (1965) observed
roosts reaching concentrations of at least
100,000, and a well-established roost at Lake
Pontchartrain, Louisiana may support 200,000
birds (Rogillio 1989). Some roosts, including
Lake Murray and a site in southern Oklahoma
(Brown and Wolfe 1978) have been used by
martins for 20 years or more (Russell et al.
1998). At least 30 additional major pre-migra-
tory roosts are known to exist in the eastern
United States (Russell et al. 1998). Enormous
concentrations of roosting Purple Martins also
have been documented on their wintering
grounds (Oren 1980; Hill 1988, 1993).

Several selective advantages have been pro-
posed for communal roosting behavior: pre-
migratory assembly (Allen and Nice 1952,
Michael and Chao 1973, Skutch 1989), re-
duced risk of predation (Lack 1968), more ef-
ficient thermoregulation (Williams et al.
1991), enhanced foraging ability through in-
formation exchange (Ward and Zahavi 1973,
Brown and Brown 1996), and association with
super-abundant food supplies or other diurnal
activity centers (Caccamise and Morfison
1986, Caccamise 1993). Although the late-
summer roosting habits of Purple Martins of-
ten are attributed to the need for pre-migratory
assembly (Allen and Nice 1952), this seasonal
pattern of roosting does not explain the ben-

Russell and Gaullireau.x • PURPLE MARTIN ROOSTING DYNAMICS

361

efits gained by gathering in a large communal
assemblage for several weeks prior to migra-
tion (Caccamise et al. 1983, Brown 1997). Be-
cause the Lake Murray roost is on an island
and at least 3 km from the lakeshore the birds
likely have little threat from most predators or
other disturbances; other large pre-migratory
roosts of Purple Martins also are associated
with bodies of water (Russell et al. 1998). Ad-
ditionally, the anti-predator benefits of roost-
ing should reach their maximum value at rel-
atively small population sizes (e.g., < 1000;
Pulliam 1973). Although roost advertising is
evident in the pre-settling flight behavior of
Purple Martins, the omnidirectional departures
of the birds and their absence until evening
make it unlikely that Lunch Island functions
as an information center about local changes
in food supply (Skutch 1989, Brown and
Brown 1996, Brown 1997).

Our study has quantified aspects of the pre-
migratory roosting behavior olF Purple Martins
that previously were known only from anec-
dotal accounts (see Brown 1997) or brief com-
ments made during studies of nesting or im-
mediate post-fledging activities (Allen and
Nice 1952, Finlay 1971, Brown 1978). Future
studies should focus on the costs and benefits
of large pre-migratory roosts to Purple Mar-
tins and be conducted at scales sufficiently
large to monitor spatial and temporal patterns
of roosting on a regional basis (Caccamise et
al. 1983, Russell et al. 1998). Studies employ-
ing radiotelemetry or other tracking methods
also are needed to determine the daily and
seasonal movements of individual martins, in-
cluding whether individuals switch among
roosts, or commute between roosts and diurnal
activity centers (e.g., Caccamise and Morrison
1986, Caccamise 1993).

ACKNOWLEDGMENTS

Funding was provided by the South Carolina Elec-
tric & Gas Company (SCE&G) and the Riverbanks
Zoological Park and Botanical Garden (Columbia,
South Carolina). The Department of Defense (DoD),
Legacy Resource Management Program provided ad-
ditional funding. We thank B. Palmer and the staff of
the National Weather Service Office (Columbia, South
Carolina) for use of the WSR-88D and assistance with
its operation. Logistical support from SCE&G and P.
Krantz of Riverbanks was greatly appreciated. We also
gratefully acknowledge the helpful contributions of M.
A. Russell and C. G. Belser. J. A. Waldvogel, J. D.

Lanham, C. R. Brown, and anonymous reviewers pro-
vided critical ct)mments that significantly improved the

manuscript.

LITERATURE CITED

Allen, R. W. and M. M. Nice. 1952. A study of the
breeding biology of the Purple Martin (Progne
suhis). Am. Midi. Nat. 47:606-665.

American Ornithologists’ Union. 1998. Checklist of
North American birds, seventh ed. American Or-
nithologists’ Union, Washington, D.C.

Anderson, D. 1965. A preliminary study of a Purple
Martin roost. Bluebird 32:3-4.

Aschoff, J. 1967. Circadian rhythms in birds. Proc.
Int. Ornithol. Congr. 14:81-105.

Bent, A. C. 1942. Life histories of North American
flycatchers, larks, swallows, and their allies. U.S.
Natl. Mus. Bull. 179:1-555.

Brown, C. R. 1978. Post-fledging behavior of Purple
Martins. Wilson Bull. 90:376-385.

Brown, C. R. 1997. Puiple Martin (Progne suhis). In
The birds of North America, no. 287 (A. Poole
and E Gill, Eds.). The Academy of Natural Sci-
enees, Philadelphia, Pennsylvania; The American
Ornithologists’ Union, Washington, D.C.

Brown, C. R. and M. B. Brown. 1996. Coloniality in
the Cliff Swallow: the effect of group size on so-
cial behavior. Univ. of Chicago Press, Chicago,
Illinois.

Brown, C. R. and S. D. Wolfe. 1978. Post-breeding
movements of Purple Martins in the Lake Texoma
area. Bull. Tex. Ornithol. Soc. 11:22-23.

Bunning, E. 1973. The physiological clock. Springer-
Verlag, New York.

Caccamise, D. F. 1993. The “patch-sitting” hypothe-
sis: a parsimonious view of communal roosting
behavior. Wilson Bull. 105:372-378.

Caccamise, D. F and J. Fischl. 1985. Patterns of as-
sociation of secondary species in roosts of Euro-
pean Starlings and Common Graekles. Wilson
Bull. 97:173-182.

Caccamise, D. F and D. W. Morrison. 1986. Avian
communal roosting: implications of diurnal activ-
ity centers. Am. Nat. 128:191-198.

Caccamise, D. F, L. A. Lyon, and J. Fischl. 1983.
Seasonal patterns in roosting flocks of starlings
and Common Crackles. Condor 85:474-481.

EA.STWOOD, E., G. A. ISTED, AND G. C. Rider. 1962.
Radar ring angels and the roosting movements of
starlings. Philos. Trans. R. Soc. Lond. B 156:242-
267.

Elkins, N. 1983. Weather and bird behaviour. T & A
D Poyser, Staffordshire, U.K.

Elkins, N. and B. Etheridge. 1974. The Crag Martin
in winter quarters at Gibraltar. Brit. Birds 72:376-
387.

Evans, T. R. and L. C. Drickamer. 1994. Flight
speeds of birds determined using Doppler radar.
Wilson Bull. 106:154-156.

Finlay, J. C. 1971. Post-breeding nest cavity defense
in Pui-ple Martins. Condor 73:381-382.

362

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

Finlay. J. C. 1976. Some effects of weather on Puiple
Martin activity. Auk 93:231-244.

Hamel, P. B. 1992. Land manager's guide to the birds
of the South. The Nature Conservancy, Chapel
Hill, North Carolina.

Harper, W. G. 1959. Roosting movements of birds and
migration departures as seen by radar. Ibis 101:
201-208.

Hill, J. R., III. 1988. A tale of two cities: the Purple
Martin on its Brazilian wintering grounds. Puiple
Martin Update 1(4): 1-5.

Hill, J. R.. III. 1993. A giant martin roost on the Am-
azon River. Purple Martin Update 4(4):5— 7.

Komar, O. 1997. Communal roosting behavior of the
Cave Swallow in El Salvador. Wilson Bull. 109:
332-337.

Krantz, P. E. and S. a. Gauthreaux, Jr. 1975. Solar
radiation, light intensity, and roosting behavior in
birds. Wilson Bull. 87:91-95.

Lack, D. 1968. Ecological adaptations for breeding in
birds. Methuen, London, U.K.

Loske, K. H. 1984. Observations on the swallow Hi-
riiiulo riistica at mass roosting places in the mid-
dle of North Rhine, Westphalia, Germany. Vogel-
welt 105:51-60.

Loske, K. H. 1986. On the behavior of swallows Hi-
rundo riistica at mass roosting places in Namibia,
Southwest Africa. Beitr. Vogelkd. 32:273-280.

Lowery, G. H., Jr. and R. J. Newman. 1963. Studying
bird migration with a telescope. Museum of Zo-
ology, Louisiana State Univ., Baton Rouge.

Meanley, B. 1965. The roosting behavior of the Red-
winged Blackbird in the southeastern United
States. Wilson Bull. 77:217-228.

Michael, E. D. and W. H. Chao. 1973. Migration and
roosting of Chimney Swifts in east Texas. Auk 90:
100-105.

Oren, D. C. 1980. Enormous concentration of martins
(Prague spp.) in Iquitos, Peru. Condor 82:344-
345.

Post, W. and S. A. Gauthreaux, Jr. 1989. Status and
distribution ot South Carolina birds. Charleston
Museum, Charleston, South Carolina.

Pulliam, H. R. 1973. On the advantages of flocking.
J. Theor. Biol. 38:419-422.

Richardson, W. J. 1978. Timing and amount ot bird
migration in relation to weather: a review. Oikos
30:224-272.

Richardson, W. J. and M. E. Haight. 1970. Migration
departures from starling roosts. Can. J. Zool. 48:
31-39.

Riehl, H. 1972. Introduction to the atmosphere, second
ed. McGraw-Hill, New York.

Rogillio, C. 1989. Plans for a Purple Martin sanctuary
take wing. Purple Martin Update 2(l):6-7.

Rudebeck, G. 1955. Some observations at a roost of
European swallows and other birds in south-east-
ern Transvaal. Ibis 97:572—580.

Russell, K. R. 1996. Spatial and temporal patterns of
a Purple Martin {Prague siihis) roost: a radar and
direct visual study. M.S. thesis, Clemson Univ.,
Clemson, South Carolina.

Russell, K. R. and S. A. Gauthreaux, Jr. 1998. Use
of weather radar to characterize movements of
roosting Purple Martins. Wildl. Soc. Bull. 26:5—
16.

Russell, K. R., D. S. Mizrahi, and S. A. Gauth-
reaux, Jr. 1998. Large-scale mapping of Purple
Martin pre-migratory roosts using WSR-88D
weather surveillance radar. J. Field Ornithol. 69:
316-325.

SAS Institute. 1995. JMP statistics and graphics
guide. Vers. 3.1. SAS Inst. Inc., Cary, North Car-
olina.

Skutch, a. F. 1989. Birds asleep. Univ. of Texas Press,
Austin.

Southern, W. E. 1959. Homing of Puiple Martins.
Wilson Bull. 71:254—261.

Spice, H. 1972. Food habits of nestling Purple Martins.
M.S. thesis, Univ. of Alberta, Edmonton.

Stone, W. 1937. Bird studies at Old Cape May, vol.
2. Delaware Valley Ornithol. Club, Academy of
Natural Sciences of Philadelphia, Philadelphia,
Pennsylvania.

Summers, R. W. and C. J. Feare. 1995. Roost depar-
ture by European Starlings Stiinnis vulgaris: ef-
fects of competition and choice of feeding site. J.
Avian Biol. 26:193-202.

Turner, A. K. 1984. Nesting and feeding habits of
Brown-chested Martins in relation to weather con-
ditions. Condor 86:30-35.

Walsh, H. 1978. Food of nestling Puiple Martins. Wil-
son Bull. 90:248-260.

Ward, P. and A. Zahavi. 1973. The importance of
certain assemblages of birds as “information cen-
tres” for food-finding. Ibis 115:517-534.

Wayne, A. T. 1910. Birds of South Carolina. Charles-
ton Museum, Charleston, South Carolina.

Williams, J. B., M. A. du Plessis, and W. R. Sieg-
fried. 1991. Green Woodhoopoes (Phaenicidits
piirpureus) and obligate cavity roosting provide a
test of the thermoregulatory insufficiency hypoth-
esis. Auk 108:285-293.

Zar, j. H. 1984. Biostatistical analysis, second ed.
Prentice-Hall, Englewood Cliffs, New Jersey.

Wilson Bull., 111(3), 1999, pp. 363-367

AGGRESSIVE RESPONSE OE CHICKADEES TOWARDS
BLACK-CAPPED AND CAROLINA CHICKADEE CALLS IN

CENTRAL ILLINOIS

ERIC L. KERSHNER' 23 AND ERIC K. BOLLINGER*

ABSTRACT. — Aggressive responses of Black-capped (Poecile atricapillus) and Carolina chickadees (Poecile
carolinensis) to heterospecitic and conspecific vocalization playbacks were measured across a historic contact
zone in east-central Illinois to determine the magnitude of interspecific aggression. Within the traditional Carolina
Chickadee range, chickadees responded more aggressively towards Carolina Chickadee calls than Black-capped
Chickadee calls. Within the traditional Black-capped Chickadee range, chickadees did not respond to either
vocalization significantly more than the other. The aggressive response towards presumed heterospecific vocal-
izations for all chickadees was marginally more aggressive closer to the contact zone. Thus, we conclude that
interspecific aggression may not act as a gap producing mechanism between chickadee ranges. Received 24 Nov.
1998, accepted 3! March 1999.

Black-capped {Poecile atricapillus) and
Carolina {Poecile carolinensis) chickadees are
extremely similar in appearance, behavior, and
ecology (Brewer 1963, Johnston 1971, Merritt
1978). They occupy largely parapatric breed-
ing ranges, although there are a few narrow
zones of overlap. Interspecific encounters and
recognition of heterospecifics are common in
these overlap areas (Ward and Ward 1974).
Interspecific territoriality in chickadees is rare
except with other chickadees in contact zones
(Brewer 1963, Smith 1993). Interspecific ter-
ritoriality may arise through the competition
for limited resources, which may contirbute to
a competitively-induced gap between Black-
capped and Carolina chickadee ranges (Tanner
1952, Slade and Robertson 1977).

We measured the level of aggression exhib-
ited by chickadees across a historic contact
zone in Illinois defined by Brewer (1963) to
test the role of interspecific aggression as a
range segregating mechanism. We hypothe-
sized that responses towards the conspecific
vocalization would be more aggressive than
those to heterospecific vocalizations. We also
hypothesized that levels of aggression to het-
erospecific vocalizations would be greatest
closer to the contact zone and weaker away
from the contact zone. We expected this re-

' Dept, of Zoology, Eastern Illinois Univ., Charles-
ton, IL 61920.

^ Present address: Dept, of Natural Resources and
Environmental Sciences, 350 Burnsides Research Lab-
oratory, 1208 W. Pennsylvania Ave., Univ. of Illinois,
Urbana, IL 61801.

^ Corresponding author; E-mail: kershner@uiuc.edu

sponse because closer to the zone of overlap
chickadees would have more encounters with
congeners and therefore should exhibit more
aggressive territorial defense if interspecific
aggression is used to maintain range bound-
aries.

METHODS

Study area. — Our 5 study sites formed a transect
across the historical contact zone for chickadees in
east-central Illinois (Fig. 1). The distance from the
midline of the contact zone varied for each study site
(Shelby ville 10 km, Douglas Hart 22 km. Fox Ridge
40 km, Lincoln Trail 58 km, Sangchris 65 km).
Sangchris State Park and Shelbyville State Park were
located in traditional Black-capped Chickadee range.
Fox Ridge State Park, Douglas Hart Nature Center, and
Lincoln Trail State Park were located in traditional
Carolina Chickadee range. Study sites were visited
weekly.

Playback e.xperiments. — Two chickadee calls and
one White-breasted Nuthatch (Sitta carolinensis) vo-
calization were used in this experiment. A playback
tape (Maxell UDII 60 minutes) was made for each
vocalization on a Magnavox FA9403 dual recording
stereo system. Each tape had a call rate of 18 calls/
minute. Vocalizations of both chickadee species and
the White-breasted Nuthatches were taken from the
Peterson Guide to Bird Songs® (Peterson 1983). One
call of each species was used for all trials. This seemed
reasonable given that differences between species calls
are much greater than variation of calls within a spe-
cies. Vocalizations were played to subjects on a Pan-
asonic FW18 dual speaker cassette recorder.

Playbacks were used to test the abilities of chicka-
dees to discriminate between conspecific and hetero-
specific vocalizations (Emlen et al. 1975). Several re-
searchers have shown that chickadees respond to con-
specific songs more than to heterospecific songs (Hill
and Lein 1989, Menitt 1978, Robbins et al. 1986b,

363

364

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

EIG. I. Map of the study sites. Dark line indicates
the historic contact zone across east-central Illinois
(from Brewer, 1963). North of the contact zone rep-
resents Black-capped Chickadee range and south of the
contact zone represents Carolina Chickadee range.
Study sites are (1) Sangchris State Park, (2) Shelby-
ville State Park, (3) Douglas Hart Nature Center, (4)
Pox Ridge State Park, and (5) Lincoln Trail State Park.

Ward and Ward 1974). The chick-a-dee call was used
in this study because it contains sufficient information
that can potentially be used by chickadees for individ-
ual recognition (Mamman and Nowicki 1981, Smith
1991). These calls tend to be short, less musical, less
variable, and relatively specialized for particular lunc-
tions such as alarm or changing the spacing between
individuals (Smith 1991). The playback of chick-a-dee
calls may elicit elevated calling rates (Nowicki 1983)
by simulating intruding males (Hill and Lein 1989,
Shackleton et al. 1992), and therefore would measure
any differences in aggression levels between vocali-
zation type.

Playback experiments were conducted 08:00-1 2:00
CST, 1 May-31 July 1995. Two different paired play-
back trials were conducted. One trial type consisted of
broadcasting Black-capped and Carolina chick-a-dee
call to chickadees in all study sites. The .second trial

type consisted of playing both a randomly-selected
chickadee call (Black-capped or Carolina) and a
White-breasted Nuthatch vocalization. This second tri-
al type was played every third trial at each study site.
Nuthatches were used because chickadees should be
familiar with this vocalization through winter flock as-
sociation. Therefore, the responses to the nuthatch vo-
calizations gave us a baseline level of aggression to
compare to the responses to chickadee vocalizations.

Playback trials were conducted by slowly walking
around the study site until chickadees were detected.
Trials were conducted by approaching a single bird as
close as possible without visibly agitating it. Only data
that fit the following criteria were used: (1) the focal
bird could be approached within 15 m, and (2) weather
conditions matched those required for the Breeding
Bird Survey (Robbins et al. 1986a). Each bird was
exposed to only one trial (either a two-chickadee trial
or a chickadee-nuthatch trial), and specific areas within
each study site were used for only one trial to avoid
influencing neighboring chickadees. The minimum
distance between trial locations was 500 m and was
usually over 1000 m. Vocalizations were played for 2
minutes with a 5 minute silent period between the two
sets of vocalizations. The silent period allowed the fo-
cal bird to return to normal activity after being exposed
to the first vocalization. The order of the two vocali-
zations were alternated to reduce bias (Kroodsma
1989, Lampe and Baker 1994, Ward and Ward 1974).

Statistical analyses. — The degree of aggression by
the focal bird was quantified based on its behavior dur-
ing the two-minute trial period (Table 1). The identity
of species at each study site was assumed to be that
of historic record, although the possibility of hybrid-
ization may render this assumption invalid. However,
all analyses were conducted without regard to species
identity. Response scores were analyzed by paired t-
tests. To compare all chickadee calls combined to nut-
hatch vocalizations, one trial per site was used and all
sites were pooled together. The relationships between
mean aggressive response and both date and distance
to the contact zone were analyzed by Pearson Cone-
lation Analysis (SAS Institute 1994).

RESULTS

Chickadees responded more aggressively to
Black-capped and Carolina chickadee calls
combined than to songs of White-breasted
Nuthatches (t = 4.6, df = 9, P < 0.001; Fig.
2). Chickadees also responded more aggres-
sively to each chickadee species call separate-
ly than to White-breasted Nuthatch vocaliza-
tions (Black-capped: t = 8.9, df = 4, P <
0.001; Carolina: t = 5.65, df = 4, P < 0.005;
Fig. 2).

There was no significant difference in ag-
gressive response to Carolina calls versus
Black-capped chickadee calls (t = 1.06, df =
33, P > 0.05; Fig. 3). Chickadees within tra-

Kershner ami Bollinf>er • PLAYBACK RESPONSES OF CHICKADEES

365

TABLE 1 . Categories of behavioral responses for
Black-capped Chickadees (Poecile ciiricapillus) and
Carolina Chickadees (Poecile carolinensis) to play-
backs of conspecihc and heterospecific calls. Catego-
ries were derived by combining information from
Brindley (1991), Censky and Ficken (1982), Ficken
and Wiese (1984), Popp and coworkers (1990),
Schroeder and Wiley (1983), and Shackleton and co-
workers (1992). The categories run on a 0—10 scale
with 10 being the most aggressive.

Category

Behavior

Flights <2 m from tape player, wing twit-
tering

Flights 2-5 m from tape player

Flights <5 m from tape player, >2 gargles

Flights <5 m from tape player, <2 gargles

Flights 5-10 m from tape player, >4 calls
made

Flights 5-10 m from tape player, <4 calls
made

Flights 5—10 m from tape player, songs
elicited

Flights >10 m from tape player, calls elic-
ited, some gargling

Flights >10 m from tape player, songs
elicited

minimal interest shown in recording, mov-
ing away from area

no interest, left area during trial

ditional Black-capped Chickadee range did
not respond significantly more aggressively to
Black-capped calls than to Carolina calls (/ =
1.35, df = 12, F > 0.05; Fig. 3). However,
chickadees within traditional Carolina Chick-
adee range were more aggressive towards Car-
olina calls than Black-capped calls (r = 2.75,
df = 20, F < 0.01; Fig. 3).

When sites were analyzed separately, only
chickadees at Fox Ridge State Park responded
significantly more aggressively towards a giv-
en chickadee vocalization. Chickadees at Fox
Ridge responded more aggresssively to Car-
olina calls than Black-capped chickadee calls
{t = -2.74, df = 8, F < 0.03; Fig. 3).

Aggressive responses towards both Black-
capped (r = —0.71, n = 5) and Carolina
chickadee vocalizations (r = —0.29, n = 5)
increased at decreasing distances to the con-
tact zone, although neither was statistically
significant (F > 0.05). With both species com-
bined, the relationship (r = —0.49, F = 0.08,
n = 10) was still not significant (F > 0.05).
Futhermore, chickadees did not show a sea-

7-,

Combined Black-capped Carolina

Vocalization

FIG. 2. Mean aggressive response of chickadees
in paired Black-capped Chickadee-nuthatch (n = 5),
Carolina Chickadee-nuthatch (n = 5), and for all trials
combined (n = 10). Dark bars represent the response
to chickadee calls and light bars represent the response
to nuthatch vocalizations. ** p < 0.01, *** P < 0.001.

Study Site

FIG. 3. Mean aggressive response of chickadees
to Black-capped and Carolina chick-a-dee vocaliza-
tions at all study sites. Dark bars represent the response
to Black-capped Chickadee calls and light bars repre-
sent the response to Carolina Chickadee calls. The
study sites are arranged from West to East and the
contact zone is noted between Shelbyville and Douglas
Hart. (Sample sizes were 6, 7, 5, 9, 7 respectively; *
* Z’ < 0.05.)

366

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

sonal difference in response, when all trials
were combined (r = —0.12, P > 0.05, n =
36). Separately, there was no relationship be-
tween mean aggressive response to either
Black-capped or Carolina Chickadee vocali-
zations and date (r = —0.28, P > 0.05, n —
18 and r = 0.05, P > 0.05, n = 18, respec-
tively).

DISCUSSION

The fact that chickadees responded more
aggressively to chickadee calls than nuthatch
vocalizations indicated that chickadees were
able to discriminate between congeneric and
heterogeneric vocalizations (Fig. 2). These re-
sults are consistent with other studies that sug-
gest that interspecific territoriality in chicka-
dees is rare except with other chickadees in
contact zones (Brewer 1963, Smith 1993).

We found that within the traditional Caro-
lina Chickadee range, chickadees responded
more aggressively to presumed conspecific
calls. This is consistent with the results of
studies on buntings in allopatric populations
(Emlen et al. 1975) and in tropical birds in
Peru (Robinson and Terborgh 1995). Several
researchers found that chickadees responded
more aggressively to their own song type than
to songs of other chickadee species except in
the contact zone where they responded ag-
gressively to both conspecific and heterospe-
cific song types (Ratcliffe and Weisman 1986,
Robbins et al. 1986b). This differs from our
results at one site. Chickadees at Fox Ridge
State Park showed significantly more aggres-
sion to presumed conspecific vocalizations
than towards heterospecifics. Overall, chicka-
dees clearly responded more aggressively to
chickadee calls than nuthatch calls with little
difference between presumed hetero- and con-
specific chickadee calls. This suggests that
chickadees may either not perceive nuthatches
as a competitive threat or that chickadees near
contact zones may not distinguish between
chickadee species calls.

We did not get the predicted increase in ag-
gression towards heterospecifics closer to the
contact zone. It is possibile that the maximum
distance from the contact zone used in this
study was not far enough to detect any sig-
nificant differences in aggression. This would
suggest that chickadees across the area are fa-
miliar with congeners. Merritt (1981) sug-

gested that individual chickadees expand and
contract their ranges seasonally. As the ranges
of these species approach each other, cogni-
zance of the heterospecific vocalization should
increase (Ward and Ward 1974). At increasing
distances from the contact zone, there could
be a point where the mean aggressive respons-
es would be significantly lower than closer to
the contact zone.

Two other possible explanations exist for
the aggression shown towards a presumed het-
erospecific call within species’ ranges: mis-
directed aggression and the presence of hy-
brids. Misdirected aggression could arise from
mistaken identity (Murray 1971, 1981). This
is possible because the vocalizations of both
species are similar and variable between in-
dividuals, and we used only one example of
each vocalization type in this experiment
(Mammen and Nowicki 1981, Smith 1991).
However, this aggression could be intentional,
because response to heterospecific calls may
promote recognition and facilitate heterospe-
cific spacing (Emlen et al. 1975, Merritt 1981,
Robinson and Terborgh 1995). Hybridization
may be more common in contact zones than
previously thought (Brewer 1963, Johnston
1971, Rising 1968, Robbins et al. 1986b,
Ward and Ward 1974). Thus, the lack of spe-
cies specific aggressive responses could be the
result of the presence of hybrids that are fa-
miliar with and respond similarly to calls of
both species. If hybridization is the cause for
the observed interspecific aggression, it is
likely that interspecific territoriality may not
act as a gap producing mechanism. Gaps be-
tween Black-capped and Carolina chickadee
ranges may occur if hybrids within these gaps
had severely reduced fitness (Brewer 1963).
However, other factors may cause gaps. One
such example is the lack of suitable habitat in
gaps areas (Grubb et al. 1994).

ACKNOWLEDGMENTS

We would like to thank Eastern Illinois University
Council for Faculty Research for partially funding this
project. Thanks to T. Grubb Jr., M. Morrison, and 3
anonymous reviwers for comments on this manuscript.

LITERATURE CITED

Baker, M. C. 1991. Response of male Indigo and Laz-
uli buntings and their hybrids to song playback in
allopatric and sympatric populations. Behaviour
119:225-242.

Kershner am! Bolliiif’er • PLAYBACK RESPONSES OE CHICKADEES

367

Brewer, R. 1963. Ecological and reproductive rela-
tionships of Black-capped and Carolina chicka-
dees. Auk 80:9-47.

Brindley, E. L. 1991. Response of European Robins
to playback of song: neighbor recognition and
overlapping. Anim. Behav. 41:503-512.

Censky, E. J. and M. S. Ficken. 1982. Responses of
Black-capped Chickadees to miiTors. Wilson Bull.
94:590-593.

Emlen, S. T, J. D. Rising, and W. L. Thompson. 1975.
A behavioral and morphological study of sympat-
ry in the Indigo and Lazuli buntings of the Great
Plains. Wilson Bull. 87:145-177.

Ficken, M. S. and C. M. Weise. 1984. A complex call
of the Black-capped Chickadee (Pants atricapil-
lus). 1. Microgeographic variation. Auk 101:349-
360.

Grubb, T. C., Jr., R. A. Mauck, and S. L. Earnst.
1994. On no-chickadee zones in midwestern
North America: evidence from the Ohio Breeding
Bird Atlas and the North American Breeding Bird
Survey. Auk 111:191-197.

Hill, B. G. and M. R. Lein. 1989. Natural and sim-
ulated encounters between sympatric Black-
capped Chickadees and Mountain Chickadees.
Auk 106:645-652.

Johnston, D. W. 1971. Ecological aspects of hybrid-
izing chickadees in Virginia. Am. Midi. Nat. 85:
124-134.

Kroodsma, D. E. 1989. Suggested experimental de-
sign for song playbacks. Anim. Behav. 37:600-
609.

Lampe, H. M. and M. C. Baker. 1994. Behavioral
response to song playback by male and female
White-crowned Sparrows of two subspecies. Bio-
acoustics 5:171—185.

Mammen, D. L. and S. Nowicki. 1981. Individual dif-
ferences and within-flock convergence in chicka-
dee calls. Behav. Ecol. Sociobiol. 9:179-186.

Merritt, P. G. 1978. Characteristics of Black-capped
and Carolina chickadees at the range interface in
Northern Indiana. Jack-Pine Warbler 56:170—179.

Merritt, P. G. 1981. Narrowly disjunct allopatry be-
tween Black-capped and Carolina chickadees in
Northern Indiana. Wilson Bull. 93:54-66.

Murray, B. G. 1971. The ecological consequences of
interspecific territorial behavior in birds. Ecology
52:414-423.

Murray, B. G. 1981. The origins of adaptive inter-
specific tenitorialism. Biol. Rev. 56:1-22.

Nowk ki, S. 1983. Flock-specific recognition of chick-
adee calls. Behav. Ecol. Sociobiol. 12:317-320.

Peterson, R. T. 1983. A held guide to bird songs of
eastern and central North America. Houghton
Mifflin Co. Boston, Massachusetts.

Popp, J. W., M. S. Ficken, and C. M. Weise. 1990.
How are agonistic encounters among Black-
capped Chickadees resolved? Anim. Behav. 39:
980-986.

Ratcliffe, L. M. and R. G. Weisman. 1986. Song
sequence discrimination in the Black-capped
Chickadee. J. Comp. Psychol. 100:361-367.

Rising, J. D. 1968. A multivariate assessment of inter-
breeding between the chickadees. Pants atricap-
illtis and P. carolittensis. Syst. Zool. 17:160-169.

Robbins, C. S., D. Bystrak, and P. H. Geissler.
1986a. The breeding bird survey: its hrst hfteen
years, 1965-1979. U.S. Fish Wildl. Serv. Resour.
Publ. 157:1-196.

Robbins, M. B., M. J. Braun, and E. A. Tobey.
1986b. Moiphological and vocal variation across
a contact zone between the chickadees Pants atri-
capillus and P. carolittensis. Auk 103:655-666.

Robinson, S. K. and J. Terborgh. 1995. Interspecihc
aggression and habitat selection by Amazonian
birds. J. Anim. Ecol. 64:1-1 1.

SAS Institute. 1994. SAS/STAT user’s guide. SAS
Institute, Inc., Cary, North Carolina.

Schroeder,.D. j. and R. H. Wiley. 1983. Communi-
cation with shared song themes in Tufted Titmice.
Auk 100:414-424.

Shackleton, S. a., L. Ratcliffe, and D. M. Weary.
1992. Relative frequency parameters and song
recognition in Black-capped Chickadees. Condor
94:782-785.

Slade, N. A. and P. B. Robertson. 1977. Comments
on competitively-induced disjunct allopatry. Occ.
Pap. Mus. Nat. His. 65:1-8.

Smith, S. M. 1991. The Black-capped Chickadee: be-
havioral ecology and natural history. Cornell
Univ. Press, Ithaca, New York.

Smith, S. M. 1993. Black-capped Chickadee in The
birds of North America, no. 39. (A. Poole. P. Stet-
tenheim, and E Gill, Eds). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithological Union, Washington D.C.

Tanner, J. T. 1952. Black-capped and Carolina chick-
adees in the southern Appalachian Mountains.
Auk 69:407-424.

Ward, R. and D. A. Ward. 1974. Songs in contiguous
populations of Black-cappcd and Carolina chick-
adees in Pennsylvania. Wilson Bull. 86:344-356.

Wilson Bull., 111(3), 1999, pp. 368-375

USE OF SONG TYPES BY MOUNTAIN CHICKADEES

{POECILE GAMBELI)

MYRA O. WIEBE'- AND M. ROSS LEIN' ^

ABSTRACT. — We investigated the composition and function of individual song repertoires in Mountain
Chickadees (PoecUe gambeli) in Alberta, Canada. Individual males had repertoires of 4—7 song types, but three
types made up 90% of all songs. We tested and rejected the hypothesis that all song types convey the same
behavioral messages. Different song types were associated with different behavioral situations. Males used 3-
note songs predominantly during undisturbed singing and 2-note songs predominantly during non-aggressive
activity. Three-note songs with each successive note lower pitched were associated with male— male interactions.
We suggest that different song types convey messages indicating different levels of aggression by the singer.
The function of individual repertoires in Mountain Chickadees appears to be similar to that of other North
American chickadees and titmice, with different song types having different communicative functions. Received
7 Aug. 1998, accepted II Feb. 1999.

Although individuals of some species of
songbirds sing only one type of song (Searcy
1983), males of many species possess reper-
toires comprising a number of discrete cate-
gories of songs, or song types (Dodson and
Lemon 1975). The significance of song rep-
ertoires has been the subject of much research
and speculation (reviewed by Krebs and
Kroodsma 1980, Kroodsma 1982, Kroodsma
and Byers 1991, MacDougall-Shackleton
1998).

Most species of the genera Parus, Poecile,
and Baeolophus have individual song reper-
toires (Hailman 1989), although the role of
repertoires appears to differ among species
[these closely-related genera were formerly
lumped as Pcirus (American Ornithologists’
Union 1997)]. Individual male Great Tits {Pa-
rus major) may use different song types in
sequence with no apparent change in the ex-
ternal situation (Hinde 1952), suggesting that
all song types convey the same messages.
Consequently, Great Tits have been used to
test many of the hypotheses that suggest that
overall repertoire size is important (e.g., Krebs
1976, 1977; McGregor et al. 1981; Baker et
al. 1986; Lambrechts and Dhondt 1988). In

' Divi.sion of Ecology (Behavioral Ecology Group),
Dept, of Biological Sciences, Univ. of Calgary, Cal-
gary, Alberta T2N 1N4, Canada;

E-mail; mrlein@ucalgary.ca

2 Present address: Canadian Wildlife Service, Suite
301, 5204 50th Avenue, Yellowknife, Northwest Ter-
ritories XI A 1E2, Canada;

E-mail; myra.robertson@ec.gc.ca.

’ Corresponding author.

contrast, other species, such as the Blue Tit
{Parus caeruleus), the North American tit-
mice {Baeolophus), and some of the North
American chickadees {Poecile), seem to pos-
sess song types that convey different messag-
es (Smith 1972, Dixon and Martin 1979, Gad-
dis 1983, Schroeder and Wiley 1983, Bijnens
and Dhondt 1984, Johnson 1987).

We examine the role of individual reper-
toires of Mountain Chickadees {Poecile gam-
beli). Mountain Chickadees sing relatively
simple songs consisting of 2-6 whistled notes,
with any number of these notes shifted to fre-
quencies lower than the others (Hill 1987).
Song types are defined easily by variation in
number and pitch of notes; individual reper-
toires consist of 3-5 song types (Hill and Lein
1989). Although no studies have investigated
whether Mountain Chickadees can differenti-
ate among song types, this seems probable be-
cause closely related Black-capped Chicka-
dees {Poecile atricapillus) distinguish among
songs varying in note number and pitch (Rat-
cliffe and Weisman 1986; Weisman and Rat-
cliffe 1989).

We documented song variation and singing
behavior of male Mountain Chickadees during
the breeding season. Our null hypothesis was
that all song types of the Mountain Chickadee
convey the same behavioral messages, and
thus the song type that a male sings would be
independent of the situation in which it is
used. The alternate hypothesis was that song
types convey different messages and thus cer-
tain song types would have a higher proba-
bility of being sung in specific situations.

368

Wiehe and Lean • SONG TYPES OF MOUNTAIN CHICKADEES

369

METHODS

The study was conducted at the Barrier Lake site
(51° 00' N, I 15° 00' W) of the University of Calgary’s
Kananaskis Field Stations in the Kananaskis Valley of
the Rocky Mountains of southwestern Alberta. The
forest inhabited by Mountain Chickadees is dominated
by trembling aspen (Populus tremuloides), white
spruce (Picea glaiica), and lodgepole pine (Pinus con-
torta). Mountain Chickadees are secondary cavity-
nesters, using pre-existing cavities such as natural
crevices or deserted nests of other cavity-nesting birds
(Hill 1987, pers. obs.). In southwestern Alberta, Moun-
tain Chickadees start to search for suitable nesting cav-
ities during April (Hill 1987, pers. obs.). Nest-building
occurs in early May. Incubation starts near the end of
May and lasts about 14 days. The nestling period gen-
erally lasts 18-21 days (Dahlsten and Copper 1979)
but may be as short as 14 days (pers. obs.). Nestlings
fledge during the last week in June and early July.

Mountain Chickadee songs, and information on the
situations of song use, were recorded from April to
July 1993. Data were collected from 1 1 males, of
which 7 were marked with a unique combination of
colored plastic leg bands. We could identify unbanded
males because males in adjacent territories were band-
ed; we also used territorial location, use of favorite
singing sites, or association with a particular nest cav-
ity to identify unbanded males. All males except one
foraged frequently with one other Mountain Chickadee
and therefore were presumed to be mated. Because
Mountain Chickadees defend their entire home range
(Hill 1987), we determined the extent of a male’s ter-
ritory by noting areas in which he was found regularly.
Observations of aggressive interactions between males
helped to confirm the locations of territorial boundar-
ies.

It was impossible to sample songs of all individuals
in one day, but usually each individual was observed
at least once in a three-day period. Observations oc-
curred between 04:00 and 12:00 (MST), the period
when chickadees are most active. An observation pe-
riod began when a male started to sing and lasted from
a few minutes to over an hour, depending on how long
the individual sang. Songs and verbal descriptions of
the different elements of the situation during singing
were recorded onto Sony C-90HF cassette tapes using
AKG D190E or Sony ECM-33P microphones, Sony
PBR-330 parabolic reflectors, and Sony TCM-5000 or
Sony TCM-5000EV tape recorders.

All recordings were made by either MOW or a field
assistant. Before working independently, the field as-
sistant accompanied MOW during 15 observation pe-
riods to ensure that both observers were making com-
parable observations. We did not notice chickadees en-
gaging in behaviors directed at the observer during any
of the observation periods, so we believe that the effect
of our pre.sence was minimal.

We recorded three elements of the situation while
males were singing. (1) Behavior of the singer: a. un-
disturbed (i.e., singing while stationary and not en-

gaged in other activities); b. engaged in non-aggressive
activities (e.g., foraging, preening, feeding nestlings,
etc.); c. engaged in aggre.ssive activities (i.e. counter-
singing with or chasing territorial neighbors). (2) Po-
sition on territory: a. within 50 m of the nest site; b.
away from the ne.st site. (3) Location of mate: a. pres-
ent in the vicinity of the singer; b. absent from the
vicinity. Each song was assigned to a particular situ-
ation. If the situation changed while the male was sing-
ing, then songs recorded before the change were as-
signed to the first situation, and songs recorded after
the change assigned to the second situation. Nests of
six pairs were found, allowing us to also assign songs
for these males to particular breeding stages (nest
search/build, egg-laying, incubation, or nestling).

Mountain Chickadees often were out of sight when
in the tops of coniferous trees. We assumed that songs
given during such intervals were in the same situation
as the last song given prior to disappearance. However,
if a subject was out of sight for more than 5 minutes,
we recorded its behavior for that interval as unknown.

We categorized different song types by variation in
number of notes and relative pitch of notes within a
song. These two features showed the most obvious
variation among songs and were relatively easy to dis-
tinguish by ear. We confirmed these classifications by
examining audiospectrographs of many songs using
SIGNAL bioacoustical analysis software (Engineering
Design, Belmont, Massachusetts). Transcriptions of re-
cordings were made using OBSER'VER software (Nol-
dus Information Technology, Wageningen, The Neth-
erlands).

We examined separately the relative importance of
number of notes in a song and the pitch of notes within
a song. To determine the influence of number of notes,
we combined all song types with the same number of
notes regardless of the pitch of the notes within the
song. Statistical tests were performed only for 2-note
and 3-note songs because sample sizes for songs with
1 note and 4 notes or more were too small. To examine
the influence of pitch, statistical tests were performed
only for “common 3-note songs’’ and “descending 3-
note songs” (see descriptions in Results). Sample sizes
of other song types with variations in pitch were too
small for statistical testing.

If different song types are not used in different sit-
uations, one would predict that a particular song type
should occur in a specific situation at the frequency
expected if song types are used at random. To examine
the influence of note number in songs, we calculated
X' values for 2X2 contingency tables of either the
relative number of 2-note songs and all other song
types compared between specific situations of use, or
the relative number of 3-note songs and all other song
types compared between specific situations of use. To
examine the influence of pitch in songs, we calculated

values for 2 X 2 contingency tables of the number
of common 3-note songs and descending 3-note songs
compared between specific situations of use. The in-
dependent event in all contingency tables was a single
song.

370

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

We conducted a separate contingency analysis for
each individual for every situation of use. We were not
able to record all individuals in different situations.
Consequently, the number of individuals used in each
comparison varied from four to seven. We combined
the results of individual contingency analyses using a
test described by Cochran (1971:151). This test ac-
counts for differences in direction of response among
individuals and can be used even if there is a wide
range of sample sizes and probabilities among individ-
uals. The test calculates a z-value that can be compared
to the normal distribution to determine significance,
with non-significant results indicating that the song
types are independent of the situation of use. The sign
of the z-value indicates the direction of deviation from
the expected.

Most statistical tests were performed using STATIS-
TIX 3.5 for Windows (Analytical Software, St. Paul,
Minnesota). Differences at an a < 0.05 were consid-
ered to be significant.

RESULTS

Composition of song repertoires. — We re-
corded an average of 1,501 songs (± 435 SE;
range = 222-4,372) from each of the 1 1 focal
males, with an individual male being recorded
an average of 15.1 days (± 2.2; range = 6-
28). Average total repertoire size was 7.4 song
types (± 0.3; range = 6-9). However, song
types used rarely by an individual may have
been “accidental” productions rather than
regular elements in the repertoire. Excluding
song types that represented less than 1% of
the total songs for an individual, the average
repertoire size was 5.1 song types (± 0.3;
range = 4-7). There was no relationship be-
tween the number of songs recorded from an
individual and the estimated size of his rep-
ertoire (all song types: Spearman r = 0.36, P
> 0.05, n = 11; song types > 1%: Spearman
r = -0.27, P > 0.05, n = 11), indicating that
all repertoires were sampled adequately.

“Common 3-note songs” (3-note songs
with the last two notes lower in pitch than the
first note. Fig. lA), “common 2-note songs”
(2-note songs with the second note lower-
pitched, Fig. IB), and “descending 3-note
songs” (3-note songs with each successive
note at least 200 Hz lower in pitch than the
previous note. Fig. 1C) were the prevalent
song types. The most-frequent song type was
common 2-note song for 5 of the 1 1 focal
males, common 3-note song for 4 males, and
descending 3-note song for 2 males. All males
sang these three song types, but for three in-

. 2-1

^ 6-

■

1 -T 1 1 1 H

C 4-
0)

O’ O.

k 2-

U-

1 1 1 1 j

m m m

1 1 ’"'I 1 1

Time (seconds)

PIG. 1. Audiospectrograms of song types most
commonly used by Mountain Chickadees on the study
area. A. Common 3-note song. B. Common 2-note
song. C. Descending 3-note song. D. 2-note song with
both notes the same pitch. E. Three-note song with all
notes the same pitch. F. Typical 4-note song. All high-
quality recordings of songs had the short, upswept note
at the beginning of the song that is present in the com-
mon 3-note, the common 2-note song, and the de-
scending 3-note song shown in this figure.

dividuals < 4% of total songs recorded were
descending 3-note songs. All other variants
had frequencies of < 12% of an individual’s
repertoire, including 2-note and 3-note songs
with all notes of the same pitch (i.e., within
200 Hz of one another; Figs. ID, IE) and
songs with four or more notes (Fig. IF). Wie-
be (1995) gives details of frequencies of use
of song types by each focal male and the use
of 2-note songs, common 3-note songs and
descending 3-note songs, in different situa-
tions by individual males. These data are sum-
marized in the following sections.

Variation in song use among breeding stag-
es.— Two-note and 3-note songs were not
used at random during many breeding stages,
although there was much individual variation
in the relative use of these song types within
all stages (Tables 1, 2). Likewise, common
and descending 3-note songs were not used at
random during different stages and there was
also much individual variation (Table 3).

Wiehe and Lean • SONG TYPES OF MOUNTAIN CHICKADEES

371

TABLE I. The use of 2-note songs by male Mountain Chickadees in various situations, compared to all
other song types. Columns headed “AM males” refer to combined analysis of contingency tables for individual
males. Columns headed “Individual males” indicate numbers of contingency tables for individual males with
significant departures from expectations.

Situation

All males

Individual males

Preferred song type

Preferred song^

c-value

2-note

Other

None

Nest searching/building stage

Other

-6.77

<0.01

Egg-laying stage

+ 1.86

0.06

Incubation stage

2-note

+ 5.99

<0.01

Nestling stage

Other

-2.30

0.02

Near nest

2-note

+3.94

<0.01

Mate present

Other

-2.45

0.01

Undisturbed singing

Other

-9.1 1

<0.01

Male-male interaction

-0.22

0.83

“ Indicate.s that 2-note songs were used significantly more frequently (2-note) or less frequently (Other) than expected in that situation. A blank indicates
no significant deviation from expected frequencies.

Variation in song use among behavioral sit-
uations.— When singing near the nest site,
males used 2-note songs significantly more
frequently (Table 1) and 3-note songs signifi-
cantly less frequently (Table 2) than expected.
This pattern also is reflected in the song types
used preferentially by individuals. Common
3-note songs were less frequent than expected
(and descending 3-note songs more frequent)
when males were singing near the nest, al-
though there was much individual variation
(Table 3).

Males sang 2-note songs significantly less
frequently than expected in the presence of
their mates, as indicated by the combined
analysis and by the preferred song types of
individuals (Table 1). Common 3-note songs
were used significantly more frequently than
expected in this situation, with five of seven

individuals using common 3-note songs pref-
erentially (Table 3).

When engaged in undisturbed singing, in
comparison to singing while engaged in an-
other non-aggressive activity such as foraging,
2-note songs were used significantly less fre-
quently (Table 1), and 3-note songs signifi-
cantly more frequently (Table 2) than expect-
ed. Song type preferences of individuals are
entirely concordant with this overall pattern.
Frequencies of common and descending 3-
note songs did not differ from expectations
during undisturbed singing (Table 3).

During male— male interactions, common 3-
note songs were used significantly less fre-
quently, and descending 3-note songs signifi-
cantly more frequently, than expected (Table
3). Three of five males showed significant
preferential use of descending 3-note songs

TABLE 2. The use of 3-note songs by male Mountain Chickadees in various situations, compared to all
other song types. See Table I for an explanation of the format.

Situation

All males

Individual males

Preferred song type

Preferred song“

z-value

3-note

Other

None

Nest searching/building stage

3-note

+ 5.71

<0.01

Egg-laying stage

-0.76

0.45

Incubation stage

Other

-5.55

<0.01

Nestling stage

+ 1.55

0.12

Near nest

Other

-4.82

<0.01

Mate present

+ 1.48

0.14

Undisturbed singing

3-note

+ 10.34

<0.01

Male-male interaction

+0.59

0.56

^ Indicates that 3-note songs were u.sed significantly more frequently (3-note) or less frequently (Other) than expected in that situation. A blank indicates
no significant deviation from expected frequencies.

372

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 3. The use of common 3-note songs by male Mountain Chickadees compared to the use of descend-
ing 3-note songs in various situations. See Table 1 for an explanation of the format.

Individual males
Preferred song type

All males

Descend-

- Common

ing

Situation

Preferred song“

z-value

3-note

None

Nest searching/building stage

Common

+ 8.73

<0.01

EgR-laying stage

Common

+ 8.36

<0.01

Incubation stage

Descending

-10.73

<0.01

Nestling stage

Descending

-5.65

<0.01

Near nest

Descending

-4.35

<0.01

Mate present

Common

-6.56

<0.01

Undisturbed singing

+0.94

0.35

Male-male interaction

Descending

-6.72

<0.01

^ Indicates the 3-note song type that was used significantly more frequently than expected in that situation. A blank indicates no significant deviation
from expected frequencies.

during this situation whereas none used com-
mon 3-note songs preferentially.

DISCUSSION

Individual song repertoires. — Excluding in-
frequent song types, males had individual rep-
ertoires of 4-7 song types. Hill and Lein
(1989) estimated a slightly smaller repertoire
size of 3-5 song types. However, they sam-
pled songs less intensively and may not have
recorded enough songs to obtain complete
song repertoires for all individuals.

Other researchers imply that the most com-
mon song type of Mountain Chickadees is a
3-note song with all notes of the same pitch
(Gaddis 1985, Hailman 1989, Hill and Lein
1989). In contrast, we found that males in our
study rarely sang songs of this type. We are
unable to explain this difference, but it may
support Gaddis’ (1985) suggestion of geo-
graphical variation in Mountain Chickadee
songs.

Use of .song types in different situations. —
Our analyses of song length grouped different
song types with the same numbers of notes
into a single category, possibly obscuring
some patterns in the use of different song
types. However, because almost all 2-note
songs recorded were common 2-note songs,
the results from analyses of all 2-note songs
would be almost identical to results using only
common 2-note songs. Almost all 3-note
songs recorded were either common or de-
scending 3-note songs. Any differences be-
tween these song types that were obscured in

the song length analyses should be revealed
in the song pitch analyses.

Different song types had significant asso-
ciations with different breeding stages. How-
ever, there was also much individual variation
in all stages, with different males using dif-
ferent songs preferentially during the same
breeding stage. This suggests that although the
z-values were significant statistically, the re-
sults may not be meaningful biologically. Dif-
ferent song types are probably not signaling
messages about the breeding stage of the sing-
er. Instead, different breeding stages may be
associated with other situations that are more
relevant biologically to the messages of the
song type.

The number of notes in songs of Mountain
Chickadees was associated significantly with
location relative to the singer’s nest. All in-
dividuals showing significant preferences ei-
ther used more 2-note songs near the nest than
away from it, or more 3-notes away from the
nest than near it. Some other studies on song-
birds also have found that males sang different
song types depending on their territorial lo-
cation (e.g., Lein 1978, Weary et al. 1994).
Although pitch of the last note in 3-note songs
was associated with location relative to the
nest, this association is weak because of the
large degree of individual variation.

In Mountain Chickadees, the tendency to
sing 2-note songs near the nest and 3-note
songs away from the nest may be influenced
by factors other than just the singer’s location.
Males were more likely to engage in non-ag-

Wiehe am! Lein • SONG TYPES OF MOUNTAIN CHICKADEES

373

gressive activities while singing near the nest
than when singing away from the nest, sug-
gesting that location relative to the nest and
male behavior were related. Although note
number may communicate some information
about the singer’s location, these factors may
be related only indirectly and the association
between male behavior and note number may
be more important biologically.

Different song types were associated with
the presence or absence of the singer’s mate.
Other researchers (e.g., Temrin 1986, Staicer
1989) have claimed that song types used pref-
erentially in the presence of females have a
greater intersexual function. One methodolog-
ical problem in our study is that a female may
have been recorded as absent when she actu-
ally was nearby in the nest hole, but not vis-
ible to the observer. Two-note songs and de-
scending 3-note songs, which were positively
associated with the absence of the female,
were also positively associated with the nest
site. Furthermore, in most cases singers did
not seem to be directing song specifically at
females. Males sometimes sang low-volume,
“quiet” songs, usually of 1 or 2 notes, when
approaching nests to feed incubating mates.
Although such songs may be directed specif-
ically at females, we never observed normal
volume songs used in this manner. Therefore,
we are hesitant, without further experimental
study, to suggest that common 3-note songs
have a greater intersexual function than do 2-
note songs and descending 3-note songs.

The behavior of singing males was associ-
ated with the number of notes in the song, but
not with pitch. Two-note songs were positive-
ly associated with singing while engaged in
non-aggressive activity and 3-note songs were
positively associated with undisturbed sing-
ing. Although not all males had individual re-
sults that were significant, all males showed
this trend.

We suggest that 3-note songs signal a high-
er motivation level of the singer to sing than
do 2-note songs. If so, 3-note songs might in-
dicate that the singer is more willing to en-
gage in some of the agonistic actions associ-
ated with singing in Mountain Chickadees,
such as countersinging bouts or interacting
with other males at the edge of the territory
to confirm boundaries. Thus, 3-note songs

could convey more aggressive messages than
2-note songs.

Male-male interactions in Mountain Chick-
adees were associated with changes in pitch
of notes in the song. Because most males were
more likely to sing common 3-note songs
when apparently unprovoked by another bird’s
activities than during interactions with rival
males, this song type may function in spon-
taneous advertisement of the territory. De-
scending 3-note songs were associated with
male-male interactions and it is probable that
lowering the pitch of the last note in 3-note
songs may convey some message to the rival.
Other researchers have suggested that song
types associated with male-male interactions
probably convey more aggressive messages
than do other song types (e.g.. Nelson and
Croner 1991). Interactions between males are
situations of high levels of agonistic stimula-
tion for Mountain Chickadees and so they
may be more likely to use songs that convey
stronger aggressive tendencies at this time.
Thus, descending 3-note songs may convey
more aggressive messages than common 3-
note songs.

Comparison to closely-related species. —
Our findings suggest that the function of in-
dividual repertoires in Mountain Chickadees
is similar to that of other North American
chickadees and titmice, with different song
types used in different situations and appear-
ing to have different communicative func-
tions. Three species of North American tit-
mice have certain song types that are used
predominantly in male-male interactions
(Gaddis 1983, Schroeder and Wiley 1983,
Johnson 1987). These are probably similar in
function to the descending 3-note song of the
Mountain Chickadee, which is also used in
male-male interactions. The Bridled Titmouse
(Baeolophus wollweheri) has one song type
used predominantly in spontaneous advertise-
ment of territory (Gaddis 1983). and we found
that the common 3-note song of the Mountain
Chickadee is used predominantly in undis-
turbed singing. Schroeder and Wiley (1983)
suggested that different song types of the Tuft-
ed Titmouse (B. bicolor) convey different lev-
els of aggression by the singer. This corre-
sponds to our suggestion that, in Mountain
Chickadees, descending 3-note songs, com-
mon 3-note songs, and 2-note songs indicate

374

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

high, intermediate, and low levels of aggres-
sion, respectively.

There are also some differences in the man-
ner in which titmice and Mountain Chicka-
dees use songs. For instance, the Bridled Tit-
mouse has a song type used predominantly in
long-distance countersinging (Gaddis 1983).
We did not find any song type in Mountain
Chickadees that was used in this manner, al-
though it is possible that some songs that we
categorized as “undisturbed” singing may
have actually been in response to far away
song that was inaudible to the observer. John-
son (1987) noted that the Plain Titmouse
{Baeolophus inornatus) was more likely to
use some song types in situations related to
nesting activities. Although 2-note songs were
associated with close proximity to the nest,
there was no indication that Mountain Chick-
adees were using these songs in any way that
was related specifically to nesting activity.

Carolina Chickadees (Poecile carolinensis)
have one song type associated with counter-
singing that is thought to be a more aggressive
song type (Smith 1972). This song type could
be similar in function to descending 3-note
songs given by Mountain Chickadees during
male— male interactions. However, Smith
(1972) also observed that Carolina Chicka-
dees were more likely to use this aggressive
song type while patrolling territorial bound-
aries whereas we did not note any strong as-
sociation between descending 3-note songs
and territorial boundaries.

Pitch may be an important cue in coding
information in the songs of both Mountain
and Black-capped chickadees, but the two
species differ in how they vary the pitch of
their songs. Unlike Black-capped Chickadees,
which shift the entire song downward in pitch.
Mountain Chickadees shift individual notes in
a song to a lower pitch. Black-capped Chick-
adee songs shifted downward in pitch were
observed during countersinging between
males in the field and in response to playback
in both wild and captive birds (Ratcliffe and
Weisman 1985, Hill and Lein 1987). Moun-
tain Chickadees 3-note songs, with the last
note lower in pitch (descending 3-note songs),
were associated with male-male interactions
whereas 3-note songs with the last two notes
of the same pitch (common 3-note songs)
were associated with the less-aggressive situ-

ation of territorial advertisement. In both spe-
cies, pitch seems to be lowered during more
aggressive situations. Morton (1977) suggest-
ed that calls of low pitch indicate higher ag-
gressiveness by the signaler than do calls of
higher pitch. This idea seems to be applicable
to song in Mountain Chickadees and Black-
capped chickadees.

ACKNOWLEDGMENTS

S. R. M. Shima provided able assistance in collec-
tion of the data. R Rodriguez de la Vega, M. R. Evans,

J. Bolstad, J. Goddard, and K. Olson also assisted with
field work. We thank G. Chilton for his help and ad-
vice during various stages of the project. The staff at
the Barrier Lake site of the University of Calgary’s
Kananaskis Field Stations provided support and ac-
commodation during the field season. R. M. R. Bar-
clay, A. P. Russell, P. A.squith, and J. P. Hailman gave
constructive comments on earlier drafts of the manu-
script. This research was supported by a graduate
scholarship from Ellis Bird Farm Limited and research
assistantships from the University of Calgary to MOW,
and a research grant from the Natural Sciences and
Engineering Research Council of Canada to MRL.

LITERATURE CITED

American Ornithologists’ Union. 1997. Forty-first
supplement to the American Ornithologists’
Union Check-list of North American Birds. Auk
1 14:542-552.

Baker, M. C., T H. Bjerke, H. Lampe, and Y. Esp-
MARK. 1986. Sexual response of female Great Tits
to variation in size of males' song repertoires. Am.
Nat. 128:491-498.

Bijnens, L. and a. a. Dhondt. 1984. Vocalizations in
a Belgian Blue Tit, Pants c. caeruleus, popula-
tion. Le Gerfaut 74:243-269.

Cochran, W. G. 1971. Some methods for strengthen-
ing the common x’ test. Pp. 132—156 in Readings
in statistics for the behavioral scientist (J. A. Ste-
ger, Ed.). Holt, Rinehart, and Winston, New York.
Dahlsten, D. L. and W. A. Copper. 1979. The use of
nesting boxes to study the biology of the Moun-
tain Chickadee (Pants gamheli) and its impact on
selected forest insects. Pp. 217-260 in The role of
insectivorous birds in forest ecosystems (J. G.
Dickson, R. N. Conner, R. R. Fleet, J. C. Kroll,
and J. A. Jackson, Eds.). Academic Press, New
York.

Dixon, K. L. and D. J. Martin. 1979. Notes on the
vocalization of the Mexican Chickadee. Condor
81:421-423. . -

Dodson, D. W. and R. E. Lemon. 1975. Re-examina-
tion of monotony threshold hypothesis in bird
song. Nature 257:126-128.

Gaddis, P. K. 1983. Differential usage of song types
by Plain, Bridled and Tufted titmice. Ornis Scand.
14:16-23.

Wiehe ami Lein • SONG TYPES OF MOUNTAIN CHICKADEES

375

Gaddis, P. K, 1985. Structure and variability in the
vocal repertoire of the Mountain Chickadee. Wil-
son Bull. 97:30-46.

H.ailman, J. P. 1989. The organization of major vocal-
izations in the Paridae. Wilson Bull. 101:305-343.

Hill, B. G. 1987. Tenitorial behaviour and ecological
relations of sympatric Black-capped (Pams alri-
capiUus) and Mountain chickadees (Pams gam-
beli) in southwestern Alberta. M.Sc. thesis, Univ.
of Calgary, Calgary, Alberta.

Hill, B. G. and M. R. Lein. 1987. Function of fre-
quency-shifted songs of Black-capped Chicka-
dees. Condor 89:914-915.

Hill, B. G. and M. R. Lein. 1989. Natural and sim-
ulated encounters between sympatric Black-
capped Chickadees and Mountain chickadees.
Auk 106:645-652.

Hinde, R. a. 1952. The behaviour of the Great Tit
(Pams major) and some related species. Behav-
iour, Suppl. 2:1-201.

Johnson, L. S. 1987. Pattern of song types use for
territorial defense in the Plain Titmouse Parus in-
ornatus. Ornis Scand. 18:24—32.

Krebs, J. R. 1976. Habituation and song repertoires in
the Great Tit. Behav. Ecol. Sociobiol. 1:297-303.

Krebs, J. R. 1977. The significance of song reper-
toires: the Beau Geste hypothesis. Anim. Behav.
25:475-478.

Krebs, J. R. and D. E. Kroodsma. 1980. Repertoires
and geographical variation in bird song. Adv.
Study Behav. 11:143-177.

Kroodsma, D. E. 1982. Song repertoires: problems in
their definition and use. Pp. 125-146 in Acoustic
communication in birds. Vol. 2. Song learning and
its consequences (D. E. Kroodsma and E. H. Mill-
er, Eds.). Academic Press, New York.

Kroodsma, D. E. and B. E. Byers. 1991. The func-
tionfs) of bird song. Am. Zool. 31:318-328.

Lambrechts, M. and a. A. Dhondt. 1988. The anti-
exhaustion hypothesis: a new hypothesis to ex-
plain .song performance and song switching in the
Great Tit. Anim. Behav. 36:327-334.

Lein, M. R. 1978. Song variation in a population of
Chestnut-sided Warblers (Dendroica pensylvani-
ca): its nature and suggested significance. Can. J.
Zool. 56:1266-1283.

MacDougali.-Shackleton, S. a. 1998. Sexual selec-
tion and the evolution of song repertoires. Curr.
Ornithol. 14:81-124.

McGregor, P. K., J. R. Krebs, and C. M. Perrins.
1981. Song repertoires and lifetime reproductive
success in the Great Tit (Pams major). Am. Nat.
188:149-159.

Morton, E. S. 1977. On the occurrence and signifi-
cance of motivation-structural rules in some bird
and mammal .sounds. Am. Nat. 11 1:855-869.

Nelson, D. A. and L. J. Croner. 1991. Song catego-
ries and their functions in the Field Sparrow (Spi-
zella piisUla). Auk 108:42-52.

Ratcliffe, L. and R. G. Weisman. 1985. Frequency
shift in the fee bee song of the Black-capped
Chickadee (Pams alricapillus). Condor 87:555—
556.

Ratcliffe, L. and R. G. Weisman. 1986. Song se-
quence discrimination in the Black-capped Chick-
adee (Pams atricapillus). J. Comp. Psychol. 100:
361-367.

Schroeder, D. j. and R. H. Wiley. 1983. Communi-
cation with repertoires of song themes in Tufted
Titmice. Anim. Behav. 31:1128-1138.

Searcy, W. A. 1983. Response to multiple song types
in male Song Sparrows and Field Sparrows.
Anim. Behav. 31:948-949.

Smith, S. T. 1972. Communication and other social
behavior in Pams carolinensis. Publ. Nuttall Or-
nithol. Club 11:1-125.

Staicer, C. a. 1989. Characteristics, use, and signifi-
cance of two singing behaviors in Grace’s Warbler
(Dendroica graciae). Auk 106:49-63.

Temrin, H. 1986. Singing behaviour in relation to po-
lyterritorial polygyny in the Wood Warbler (Phyl-
loscopus sihilatrix). Anim. Behav. 34:146-152.

Weary, D. M., R. E. Lemon, and S. Perreault. 1994.
Male Yellow Warblers vary use of song types de-
pending on pairing status and distance from the
nest. Auk 1 1 1 :727-729.

Weisman, R. G. and L. Ratcliffe. 1989. Absolute and
relative pitch processing in Black-capped Chick-
adees, Pams atricapillus. Anim. Behav. 38:685-
692.

WiEBE, M. O. 1995. The function of song types in the
Mountain Chickadee (Pams gamheli). M.Sc. the-
sis, Univ. of Calgary, Calgary, Alberta.

Wilson Bull.. 111(3). 1999, pp. 376-380

SURVIVAL AND LONGEVITY OF THE PUERTO RICAN VIREO

BETHANY L. WOODWORTH,' JOHN FAABORG,^ AND WAYNE J. ARENDU

ABSTRACT. — The Puerto Rican Vireo {Vireo latimeri), a Puerto Rican endemic, is declining in at least one
forest reserve as the result of pressures from introduced nest predators and an introduced brood parasite. We
collected data on adult survival, adult longevity, and juvenile survival from a long-term mist netting study
(1973—1999) and a demographic study of color-marked birds (1990—1993) in Guanica Forest, Puerto Rico. Of
the adult birds banded in the first three years of the demographic study, 24 of 32 males (75%) and 6 of 7 females
(86%) were known to survive until June of the year following their banding. Model-based estimates of adult
survival rate from capture/resighting of 65 color-marked birds was 0.74 (± 0.05 SE); for 51 adult males analyzed
separately, survival rate was 0.74 (±0.06; data were insufficient to estimate survival rate of females). We
recorded a new longevity record for the Puerto Rican Vireo of 13 years, 2 months. Juvenile survival was
estimated by enumeration to be 0.40 (±0.15). Juveniles spent prolonged periods on their natal teiTitory, which
might increase their probability of surviving to first breeding. Puerto Rican Vireos have lelatively high survival
rates despite the presence of numerous introduced predators in their habitat, a highly seasonal environment, and
the stress of renesting as many as 6 times in a season. Received 19 Oct. 1998, accepted 25 Feb. 1999.

Survival rate is an important component of
life history models, and it is a central parameter
examined in comparative demographic studies
(e.g., Ricklefs 1982, Martin 1995). From a con-
servation perspective, precise estimation of
adult and Juvenile survival rates is critical be-
cause population dynamics often show great
sensitivity to variation in these parameters
(e.g., Lande 1988, Ryan et al. 1993). Despite
their importance, relatively few data are avail-
able regarding the survival rates and longevity
of tropical birds in general, and insular species
in particular (but see Karr et al. 1990, Faaborg
and Arendt 1995, Johnston et al. 1997).

The Puerto Rican Vireo (Vireo latimeri), is
a small (11-12 g) passerine restricted to the
island of Puerto Rico (Wetmore 1916). The
population of vireos in Guanica Forest, Puerto
Rico’s largest dry forest reserve, has declined
steadily over the past 20 years as a result of
parasitism by an exotic avian brood parasite,
the Shiny Cowbird (Molothrus bonarien.sis),
and nest predation by introduced mammals

' Dept, of Ecology, Evolution and Behavior, Univ.
of Minnesota, 1987 Upper Buford Circle. St. Paul, MN
55108.

- Present address: Pacific Island Ecosystems Re-
search Center, Biological Resources Division,
U.S.G.S., Kilauea Field Station, PO. Box 44, Hawaii
National Park, HI 96718.

•’ Division of Biological Sciences, Univ. of Missouri,
Columbia, MO 6521 1.

■' USDA Forest Service, International In.stitute of
Tropical Forestry, Sabana Field Research Station, P.O.
Box 490, Palmer. PR 00721.

^ Corresponding author;

E-mail: Bethany-WoodworthC® usgs.gov

[rats (Rattus spp.), mongoose (Herpestes au-
ropunctatus), and feral cats (Felis catus)-, Faa-
borg et al. 1997; Woodworth 1997, 1999]. A
population dynamics analysis indicated that es-
timates of the Puerto Rican Vireo’s population
growth rate were very sensitive to adult sur-
vival rate; consequently precise estimates of
this parameter are crucial for useful population
modehng (Woodworth 1999). In this paper we
present data on adult and juvenile recoveries,
from which we estimate survival rates and lon-
gevity of the Puerto Rican Vireo, compare
them to temperate mainland congeners, and
comment on their implications for persistence
of this single-island endemic.

METHODS

We .studied the Puerto Rican Vireo population in
Guanica Forest Reserve (17° 58' N, 66° 52' W) along
the southwestern coast of Puerto Rico. The reserve
comprises 4,0 1 5 ha of mature dry subtropical forest
over shallow limestone soils. Rainfall averages 860
mm annually, almost all of which tails between April
and November (Murphy and Lugo 1986), and Puerto
Rican Vireos generally breed from April through July
(Woodworth 1997). Guanica is the site of a long-term
constant effort mist netting study of wintering and res-
ident landbirds (see Faaborg and Arendt 1989a), dur-
ing which 135 Puerto Rican Vireos were marked with
aluminum bands from 1973-1996.

From 1990-1993. B.L.W. conducted a demographic
study of color-marked vireos in four 50-ha study areas
(Woodworth 1997). Resident vireos were captufed by
playing recorded vireo songs to lure territorial males
or pairs into mist nets. Because males were more ag-
gressive than females toward intruders, most known
sex birds we captured were males (78%, n = 65; 1
bird was of unknown sex). Individuals were resighted
by revisiting all territories within the study areas and

376

Woodworth el at. • PUERTO RICAN VIREO

377

areas within 300 ni (about two territory widths) of their
borders, and by broadcasting Puerto Rican Vireo song.
Color-marked birds were relocated every few days
(range: 1-9) throughout the breeding season as part of
a study of their seasonal reproductive success (Wood-
worth 1997). In order to approximate the general as-
sumption of capture-recapture models that all sampling
is instantaneous, "capture periods” were defined so as
to be short in relation to interval length (breeding sea-
son; Smith and Anderson 1987). Thus we defined two
sample (capture) periods each breeding season (1990,
1991, and 1993), one consisting of the first two weeks
after arrival on the study area, and the second includ-
ing the two weeks immediately preceding the end of
the field season. Birds banded at other times were in-
cluded only if they were resighted during one of these
sampling periods and were treated as if they were orig-
inally banded during that sampling period. The June
1992 and January 1993 capture periods consisted of
15 and 11 day visits to the study area, respectively.

We estimated adult survival rate from capture/resight-
ing data on 66 color-marked, territorial adults over 4
years and 7 capture intervals [average interval length =
0.45 ± 0.31 (SE) years]. Five of the birds used in this
analysis were originally banded by J.F. and W.J.A. prior
to 1990, and so were included in a survival analysis by
Faaborg and Arendt (1995), but time periods of the two
survival datasets did not overlap. We used the program
JOLLY (Pollock et al. 1990) to produce estimates of
survival rate under five different capture-recapture mod-
els which vary in their assumptions about capture and
survival probabilities. These models and their assump-
tions have been presented in detail elsewhere (Pollock
et al. 1990 and references therein). In general, the cap-
ture/resighting field methods used here and the more
widely u.sed constant effort mist netting methods meet
(or not) the assumptions of the Jolly-Seber models to
similar degrees, with a few exceptions: (1) although
fixed placement of nets in relation to territory bound-
aries may result in heterogeneous capture probabilities
in mistnetting .studies (Pollock et al. 1990), we were able
to search entire territories for marked individuals; (2)
trap response (net shyness) was not a concern in this
study because we did not need to catch a bird in a net
in order to resight it; (3) we were able to exclude tran-
sients from the study (the presence of transients in a
sample may bias survival rate estimates if special mod-
els are not employed; Pradel et al. 1997); (4) we were
able to rule out temporary emigration. Becau.se the cap-
ture probability in this study was very high (0.92), we
expect the model-based estimators to provide reasonably
unbiased estimates of survival rate despite the relatively
short time span of the study and moderate sample size
(Gilbert 1973).

The program JOLLY provides goodness-of-fit te.sts
to assess the fit of a model to a given data set. Where
several models fit the data, likelihood ratio tests were
used to test among models, with the simplest adequate
model preferred. For statistical comparisons among
survival rates we used the X“ statistic proposed by
Sauer and Williams (1989).

Model-based estimators account for the possibility
that a bird is alive and in the study area, but is not
resighted in a particular sample period. To facilitate
comparison with other studies, we also present surviv-
al rate as the number of birds banded in the first three
years of the study that were known to be alive in the
June following their banding (i.e., enumeration).

Because of the small sample size of fledglings, we
could not use model-based estimators of juvenile sur-
vival. Thus, juvenile survival was calculated as the pro-
portion of birds originally banded as fledglings that were
recaptured or resighted in any subsequent year, and var-
iance was estimated assuming binomial sampling.

As is true in all capture-recapture studies of open
populations, dispersal outside of the study area could
not be distinguished from mortality. However, typical
dispersal rates and distances for the Puerto Rican Vireo
are small (Woodworth et al. 1998) so the effect should
be relatively minor in this study.

During the demographic study, B.L.W. recaptured 7
birds that had been originally banded prior to 1990.
Estimated maximum longevity of the recaptured birds
was calculated as the time from initial banding to the
last recapture, plus the time from initial banding to the
previous June 1, assuming that all birds were hatched
on that date (following Klimkiewicz et al. 1983).

RESULTS

We color-banded 51 males, 14 females, 13
fledglings, and 10 birds of unknown sex (win-
ter captures). Of the adult birds banded in the
first three years of the study, 24 of 32 males
(75%) and 6 of 7 females (86%) were known
to survive (i.e., were alive and present on the
study area) until the June of the year follow-
ing their banding (Table 1). Over the four
years, there were 59 opportunities for males
to survive between breeding seasons, and the
males survived in at least 43 of these cases
(73%). Females were documented to survive
in 11 of 13 opportunities (85%).

Territorial adult Puerto Rican Vireos during
this study had an estimated annual survival rate
of 0.74 (± 0.05). JOLLY model D, which is
based on constant survival and capture proba-
bilites throughout the study, provided the best
fit to the data (overall ~ 9.7, df = 8, P =
0.29). Our capture/resighting methodology re-
sulted in a very high annual capture probability
(0.92 ± 0.03). Male annual survival rate esti-
mated for 51 males was 0.74 (± 0.06; Model
D, capture probability = 0.95 ± 0.02). Data
for 14 females were insufficient to fit a model
describing female survival because most were
banded in the last year of the study. Juvenile

378

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 1. Bandings and resightings of male Puerto Rican Vireos known to be at least one year old when
banded (demographic study only). The 8 sample periods, each 2 weeks long, took place in April 1990, July
1990, March 1991, August 1991, June 1992, January 1993, March 1993, and August 1993, resulting in 7
intervals. No new' birds were banded during sample periods in July 1990, June 1992, or August 1993.

Sample period
when banded

Returns in

sample period following banding

Number

banded

One

Two

Three

Four

Five

Six

Seven

Apr 1990

Mar 1991

—

Aug 1991

1 1

—

Jan 1993

—

Mar 1993

—

Total

survival rate from fledging to first breeding
was estimated at 0.40 (± 0.15; n = 10).

The oldest Puerto Rican Vireo we recap-
tured was at least 13 years, 2 months (13-02)
old, exceeding the previous longevity record
for the Puerto Rican Vireo by nearly 4 years
(09-04; Faaborg and Arendt 1989b). Three
other birds that nearly matched the previous
record were also recaptured (09-01, 09-02,
and 09-02). All were color-banded territorial
males that we observed over 1-2 complete
breeding seasons before they disappeared or
the study ended. None had dispersed more
than 500 m in the decade since they were orig-
inally banded (see Woodworth et al. 1998 for
details of dispersal behavior).

DISCUSSION

Faaborg and Arendt (1995) estimated an
adult annual survival rate of 0.68 (± 0.08) for
the Puerto Rican Vireo population in Guanica,
based on their long-term mist netting study
(Jolly-Seber model D, 19 individuals over 18
years and 15 capture intervals). Although their
mean survival value is slightly less than that
presented here (probably because of the inclu-
sion of a higher proportion of female birds,
along with non-territorial individuals), the two
estimates are not significantly different (y^ =
0.40, df = 1, P > 0.05).

Our Puerto Rican Vireo survival rate esti-
mate is high relative to survival and recovery
rates reported for temperate vireos. Recovery
percentages of White-eyed Vireos {Vireo gri-
.seus) studied over 9 years on their breeding
grounds were 48% for males, and 50% for
females (Hopp et al. 1999). Return rates of
adult male Black-capped Vireos {V. atricap-
illiis) to breeding territories in central Texas

were 55-75% (Grzybowski 1991). The sur-
vival rate of Red-eyed Vireos {V. olivaceous)
based on returns to breeding grounds in Mar-
yland, was estimated at 59% (Jolly-Seber
model A; Nichols et al. 1981). Return rates of
adult Gray Vireos (V. vicinior) to wintering
territories in Mexico were 46-71% (Bates
1992), and for Bell’s Vireo {V. bellii) return-
ing to breeding territories in California, 47%
(Salata 1983). Interestingly, compared to sur-
vival rate estimates for other tropical island
passerines studied to date, the survival rate of
the Puerto Rican Vireo is not unusually high
[e.g., average 68% (51—79%) for 7 Puerto Ri-
can species, Faaborg and Arendt 1995; 65.3%
(45-85%) for 17 Trinidadian species, John-
ston et al. 1997; and 76% (55-88%) for 5 Ha-
waiian species, van Riper 1987, Lepson and
Freed 1995, Ralph and Fancy 1995, Wood-
worth et al. in press].

Likewise, the longevity record for the
Puerto Rican Vireo of 13 years, 2 months is
long relative to most of its temperate conge-
ners. Records for six other temperate Vireo
species range from 6 years, 1 month to 10
years (Davis 1995, Kennard 1975, Klimkiew-
icz et al. 1983, Rodewald and James 1996).
The record for the Warbling Vireo (Vireo gil-
vis), a neotropical migrant, is very similar (13-
01, Klimkiewicz et al. 1983) to the Puerto Ri-
can Vireo. Such records are complicated by
many factors (Krementz et al. 1989). Al-
though it is surprising that such a seemingly
long longevity record from a nonmigratory
species would be equaled by a long distance
migrant, it is worth noting that over 15,000
Warbling Vireos have been banded (Kli-
mkiewicz et al. 1983), but only a few hundred

Woodworth et cil. • PUERTO RICAN VIREO

379

Puerto Rican Vireos have been bandeci and
many of these have been long lived.

The relatively high survival and longevity
of this insular species is remarkable in light
of the presence of numerous introduced pred-
ators in its habitat, high rates of nest failure
causing females to renest up to 6 times in a
season (Woodworth 1997), and the stresses of
a highly seasonal environment (almost no rain
falls from December to March, and Guanica
loses up to 50% of its leaf area in winter;
Murphy and Lugo 1986). Puerto Rican Vireos
have a small clutch size relative to temperate
vireos (Woodworth 1995), which, when cou-
pled with the generally observed trade-off be-
tween fecundity and survival (Martin 1995),
might allow birds to survive through more
breeding seasons (Cody 1966). In addition, a
non-migratory insular species might outlive
its migratory counterparts because it does not
pay the price of annual migration.

Juvenile survival rates of passerines are
poorly known, especially for tropical birds.
Return rates of juvenile Bell’s Vireos and
Black-capped Vireos were measured as 24%,
although actual survival rate is likely to be
higher (Salata 1983, Grzybowski 1991; re-
spectively). Survival of juvenile Wood Tlirush
{Hylocichla mustelina) in their first 12 weeks
is only 0.42 (Anders et al. 1997). Juvenile sur-
vival may be enhanced if young are allowed
to remain in their natal territory for an ex-
tended period (discussed in Karr et al. 1990)
as Puerto Rican Vireo fledglings have been
observed to do (at least 80-98 days post-
fledging; Woodworth 1995).

A population dynamics model of this pop-
ulation showed that, as is common in many
population models, the vireo ’s predicted pop-
ulation growth rate was greatly dependent upon
the value of adult survivorship used in the
model (Woodworth 1999). Therefore, a precise
and accurate estimate of adult survivorship is
critical to evaluating the long-term prospects
for survival of this population. The close agree-
ment between two independent estimates (Faa-
borg and Arendt 1995 and this study) of adult
survival rate for this population improves con-
fidence in the predictions of a model using
these estimates, although additional data on fe-
male and juvenile survivorship is needed.

The relatively high adult and juvenile sur-
vival rates we documented would seem to bode

well for the persistence of Puerto Rican Vireos
in Guanica Forest. However, other work on this
population has shown that the vireos suffer ex-
tremely high nest losses to native and intro-
duced predators, and to parasitism by the ex-
otic Shiny Cowbird (Woodworth 1997). De-
spite as many as 6 nest attempts in a single
season, females succeed in fledging young
from only 0.41-0.67 nests per year (Wood-
worth 1997). These factors result in an overall
negative population growth rate for the vireo
over the range of “reasonable” survival rate
values (the 95% confidence limits of the esti-
mates; Woodworth 1999). Thus, the declines
observed over the previous decade (Faaborg et
al. 1997) are likely to continue unless active
management is undertaken to reduce predation
and/or brood parasitism in the forest.

ACKNOWLEDGMENTS

J. Colon kindly provided information on two birds
he originally banded. We thank the dozens of volun-
teers who have participated in these projects over the
years. The Puerto Rico Departamento de Recursos Na-
turales Ambiente and M. Canals kindly gave permis-
sion to work in Guanica. B.L.W. was supported by the
International Council for Bird Preservation-U.S. Sec-
tion; Frank M. Chapman Fund of the American Mu-
seum of Natural History; Sigma Xi Grant-in-Aid of
Research; Dayton Natural History Fund and Wilkie
Fund for Natural History Research, Bell Museum of
Natural History; Eastern Bird Banding Association;
Paul A. Stewart Award of the Wilson Ornithological
Society; fellowships and assistantships from the De-
partment of Ecology, Evolution and Behavior, Univer-
sity of Minnesota; and Grants for Research Abroad and
a Doctoral Dissertation Fellowship from the Graduate
School of the University of Minnesota. Support for J.F
and W.J.A. was provided by the Frank M. Chapman
Fund of the American Museum of Natural Hi.story, Na-
tional Sciences Foundation Doctoral Dissertation Im-
provement Fund, University of Missouri-Columbia
(Research Council of the Graduate School), USD A
Forest Service (International Institute of Tropical For-
estry), Biological Resources Division of the U.S. Geo-
logical Survey, and the U.S. Fish and Wildlife Service.
K. Dugger, S. Hopp, J. Bates, and two anonymous re-
viewers provided helpful comments on earlier drafts of
this manuscript.

LITERATURE CITED

Anders, A. D., D. C. Dearborn, J. Faaborg, and F.
R. Thompson, III. 1997. Juvenile survival in a
population of Neotropical migrant birds. Conserv.
Biol. 11:698-707.

Bates, J. M. 1992. Winter tenitorial behavior of Gray
Vireos. Wilson Bull. 104:425-433.

380

THE WILSON BULLETIN • VoL III. No. 3. September 1999

Cody, M. L. 1966. A general theory of clutch-size.
Evolution 20:174-184.

Davis, J. N. 1995. Hutton's Vireo (Vireo hiittoni). In
The birds of North America, no. 189 (A. Poole
and L Gill, Eds.). The Academy of Natural Sci-
ences, Philadelphia, Pennsylvania; The American
Ornithologists' Union, Washington, D.C.

Faaborg, J. and W. J. Arendt. 1989a. Long-term de-
clines in winter resident warblers in a Puerto Ri-
can dry forest. Am. Birds 43:1226-1230.

Faaborg, J. and W. J. Arendt. 1989b. Longevity es-
timates of Puerto Rican birds. N. Am. Bird Bander
14:1 1-13.

Faaborg, J. and W. J. Arendt. 1995. Survival rates
of Puerto Rican birds: are islands really that dif-
ferent? Auk 1 12:503-507.

Faaborg, J., K. M. Dugger, W. J. Arendt, B. L.
Woodworth, and M. E. Baltz. 1997. Population
declines of the Puerto Rican Vireo {Vireo lalimeri)
in Guanica Forest. Wilson Bull. 109:195-202.

Gilbert, R. O. 1973. Approximations of the bias in
the Jolly-Seber capture-recapture model. Biomet-
rics 29:501-526.

Grzybowski, j. a. 1991. Survivorship, dispersal and
population structure of Black-capped Vireos at the
Kerr Wildlife Management Area, Texas. Resour.
Protection Div., Texas Parks Wildl. Dept., Austin.

Hope, S. L., A. Kirby, and C. A. Boone. 1999. Band-
ing returns, arrival pattern, and site-hdelity of
White-eyed Vireos. Wilson Bull. 111:46—55.

Johnston, J. P, W. J. Peach, R. D. Gregory, and S.
A. White. 1997. Survival rates of tropical and
temperate passerines: a Trinidadian perspective.
Am. Nat. 150:771-789.

Karr, J. R., J. D. Nichols, M. K. Klimkiewicz, and
j. D. Brawn. 1990. Survival rates of birds of trop-
ical and temperate forests: will the dogma sur-
vive? Am. Nat. 136:277—291.

Kennard, j. H. 1975. Longevity records of North
American birds. Bird-Banding 46:55—57.

Klimkiewicz, M. K., R. B. Clapp, and A. G. Futcher.
1983. Longevity records of North American birds:
Remizidae through Parulinae. J. Field Ornithol.
54:287-294.

Krementz, D. G., j. R. Sauer, and J. D. Nichols.
1989. Model-based estimates of annual survival
rate are preferable to observed maximum lifespan
statistics for use in comparative life-history stud-
ies. Oikos 56:20.3-208.

Lande, R. 1988. Demographic models of the NoHhern
Spotted Owl (Sfri.x occidenlali.s caiirina). Oecol-
ogia (Berlin) 75:601-607.

Lep.son, j. K. and L. a. Freed. 1995. Variation in male
plumage and behavior ot the Hawaii Akepa. Auk
1 12:402-414.

Martin, T. E. 1995. Avian life history evolution in
relation to nest sites, nest predation, and food.
Ecol. Monogr. 65:101-127.

Murphy, P. G. and A. E. Liigo. 1986. .Structure and
biomass of a subtropical dry fore.st in Puerto Rico.
Biotropica 18:89-96.

Nichols, J. D., B. R. Noon, S. L. Stokes, and J. E.
Hines. 1981. Remarks on the ii.se of mark-recap-
ture methodology in estimating avian population
size. Stud. Avian Biol. 6:121-136.

Pollock, K. H., J. D. Nichols, C. Brownie, and J. E.
Hines. 1990. Statistical inference for capture-re-
capture experiments. Wildl. Monogr. 107:1—97.

Pradel, R., j. E. Hines, J. D. Lebreton, and J. D.
Nichols. 1997. Capture-recapture survival models
taking account of transients. Biometrics 53:60—72.

Ralph, C. J. and S. G. Fancy. 1995. Demography and
movements of 'Apapane and 'I'iwi in Hawai'i.
Condor 97:729-742.

Ricklefs, R. E. 1982. Comparative avian demography.
Cliit. Ornithol. 1:1-32.

Rodewald, P. G. and R. D. James. 1996. Yellow-
throated Vireo ( Vireo jiavifrons). In The birds of
North America, no. 247 (A. Poole and F. Gill,
Eds.). The Academy of Natural Sciences, Phila-
delphia. Pennsylvania; The American Ornitholo-
gists' Union, Washington, D.C.

Ryan, M. R., B. G. Root, and P. M. Mayer. 1993.
Status of Piping Plovers in the Great Plains ot
North America: a demographic simulation model.
Conserv. Biol. 7:581—585.

Salata, L. R. 1983. Status of the Least Bell's Vireo
on Camp Pendleton, California: research done in
1983. U.S. Fish and Wildl. Serv., Laguna Niguel,
California.

Sauer, J. R. and B. K. Williams. 1989. Generalized
procedures for testing hypotheses about survival
or recovery rates. J. Wildl. Manage. 53:137—142.

Smith, D. R. and D. R. Anderson. 1987. Effects of
lengthy ringing periods on estimates ot annual
survival. Acta Ornithol. 23:69—76.

van Riper, C.. III. 1987. Breeding ecology of the Ha-
waii Common Amakihi. Condor 89:85—102.

Wetmore, a. 1916. Birds of Porto Rico. U.S. Dept.
Agric. Bull. 326:1-140.

Woodworth, B. L. 1995. Ecology of the Puerto Rican
Vireo and the Shiny Cowbird in Guanica Forest,
Puerto Rico. Ph.D. diss., Univ. of Minnesota, St.
Paul.

Woodworth, B. L. 1997. Brood parasitism, nest pre-
dation. and season-long reproductive success of a
tropical island endemic. Condor 99:605-612.

Woodworth, B. L. 1999. Modeling the population dy-
namics of a songbird exposed to parasitism and
predation and evaluating management options.
Conserv. Biol. 13:67—76.

Woodworth, B. L., J. Faaborg, and W. J. Arendt.
1998. Breeding and natal dispersal in the Puerto
Rican Vireo. J. Field Ornithol. 69:1-7.

Woodworth. B. L., J. T. Nel.son, E. J. TVeed, S. G.
Fancy, M. P. Moore, E. B. Cohen, and M. S.
Collins. In press. Breeding productivity and sur-
vival of the endangered Hawai'i Creeper in a wet
forest refuge on Manna Kea. Hawai'i. In The sta-
tus, ecology, and conservation of the Hawaiian
avifauna (.1. M. Scott, S. Conant, and L. A. Freed.
Eds.). Stud. Avian Biol.

Wilson Bull.. 111(3), 1999, pp. 381-388

EFFECTS OF PRIOR RESIDENCE AND AGE ON BREEDING
PERFORMANCE IN YELLOW WARBLERS

G. A. LOZANO' - ' AND R. E. LEMON'

ABSTRACT. — Age-related increases in reproductive success could be the result of better survival by suc-
cessful breeders (survival hypothesis), greater dispersal by unsuccessful breeders (dispersal hypothesis), and/or
age-related differences in the ability to compete for breeding opportunities (constraint hypothesis). We used
banding and nesting data horn four consecutive breeding seasons to examine the effects of prior residency on
several indices of breeding performance in Yellow Warblers {Dendroica petechia). We compared the breeding
performance of returning birds with that of new arrivals, and of individuals between successive breeding seasons.
There were no differences in clutch size between new anivals and returning individuals, nor within individuals
between successive breeding seasons. Among males, prior residence had no effect on whether a clutch was
started, but among females the number of prior residents that initiated a clutch was higher than expected, and
the number of new arrivals that did not was lower than expected. In contrast, there were no differences in laying
or hatching date between new arrivals and returning individuals, but within-individual comparisons showed that
males bred earlier in successive breeding seasons. Previous reproduction increased subsequent return rates only
1 out of 3 years in both sexes. Returning males were larger than new arrivals, but there were no differences in
females. Within-individual size increases between successive breeding seasons occurred in both sexes. These
results are consistent with the constraint hypothesis, but the proximate mechanisms by which these differences
arise remain to be determined. Received 1 Oct. 1998, accepted 20 Feb. 1999.

Age-related increases in reproductive suc-
cess have been documented in many bird spe-
cies (reviewed by Saether 1990, Forslund and
Part 1995). Several hypotheses have been pro-
posed to explain this phenomenon. Forslund
and Part (1995) divided these hypotheses into
three groups depending on whether they are
based on (1) the gradual appearance or dis-
appearance of certain phenotypes, (2) the life-
time optimization of reproductive effort, or (3)
age-related improvements in competence.
These three groups of hypotheses are not mu-
tually exclusive.

In the first group of hypotheses, differences
in reproductive success among age classes are
not viewed as the result of individuals increas-
ing their reproductive success with age, but
rather as a result of the elimination of some
phenotypes from the population. The proba-
bility of breeding may be positively correlated
with the likelihood of survival (survival hy-
pothesis) or negatively correlated with the
likelihood of dispersal (dispersal hypothesis;
e.g.. Smith 1981, Nol and Smith 1987, Wheel-
wright and Schultz 1994); either mechanism

' Dept, of Biology, McGill Univ., 120.3 Dr. Pentield
Ave., Montreal, Quebec, Canada H3A IBI.

^ Present address: Dept, of Biological Sciences, Si-
mon Fraser Univ., Burnaby, British Columbia, Canada,
V5A IS6; E-mail: LOZANO (®SFU.CA
^ Corresponding author.

would yield a positive correlation of age with
reproductive success. These hypotheses are
clearly not applicable to species that show
within-individual increases in reproductive
success between successive reproductive
bouts (Pyle et al. 1991, Smith 1993).

The second group of hypotheses asserts that
older individuals allocate more effort to re-
production. The restraint hypothesis states that
this is because residual reproductive value de-
creases with age, so as individuals become
older the value of current reproduction in-
creases relative to the value of future repro-
duction. This idea is based on the theoretical
trade-off between current and future reproduc-
tion (Williams 1966, Pianka and Parker 1975)
and only applies if the probability of survival
decreases with age. Reproductive effort may
also increase with age if age-specific improve-
ments in breeding experience make each suc-
cessive reproductive bout relatively less ex-
pensive (Reid 1988). Accordingly, individuals
may allocate the same relative effort into re-
production as they age, but increase it in ab-
solute terms. These hypotheses are based on
life history, in which age is the explanation.

In contrast, hypotheses in the third group,
collectively known as the constraint hypoth-
esis (Curio 1983), are based on ecological
considerations. These hypotheses posit that
younger individuals are unable to reproduce

381

382

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

as well as older ones because the former are
less capable of competing for breeding op-
portunities. Age-related variance in reproduc-
tive success is considered to be a result of
differences in the ability to obtain territories,
forage efficiently, resist competitors, avoid
predators, attract prospective mates, and/or
raise offspring. These factors may, in turn, be
affected by the familiarity of individuals with
the breeding area. This effect, known as “lo-
cal experience” or “local familiarity”, may
allow prior residents to exploit the resources
of an area more efficiently than new arrivals
(Hinde 1956, Greenwood 1980).

Here we used banding and nesting data
from four consecutive breeding seasons to ex-
amine the effects of prior residency on several
indices of breeding performance in Yellow
Warblers (Dendroica petechia). The Yellow
Warbler is a socially monogamous 10 g pas-
serine found from Alaska and northern Can-
ada to the coasts of Peru and Venezuela, in-
cluding the Caribbean and Galapagos islands.
Formerly, several species, or subspecies were
recognized (Aldrich 1942, Bent 1963), but
currently these are considered “groups” of
one species (Sibley and Monroe 1990). South-
ern, non-migratory subspecies are easily dis-
tinguished from North American migratory
populations by morphological differences (Al-
drich 1942, Wiedenfeld 1991, Curson et al.
1994) and by phylogenetic analyses based on
mitochondrial DNA (Klein and Brown 1994).
The migratory group {Dendroica petechia
aestiva) used in this study, breeds in the Unit-
ed States and Canada and winters in Central
America, northern South America, and the
Caribbean.

Using morphological and nesting data from
four consecutive breeding seasons, we first de-
termined whether prior residence increased
the likelihood or timing of breeding in Yellow
Warblers. We then tested the survival and dis-
persal hypotheses by examining whether in-
dividuals were more likely to breed or to do
so earlier in successive seasons, and we dealt
with the possible disappearance of phenotypes
from the population by relating morphology
and breeding performance to subsequent re-
turn rates. Finally, we explored morphological
differences between new arrivals and return-
ing birds, and individual morphological
changes between successive breeding seasons.

METHODS

Data were collected from 1992 to 1995 at a 5 ha
(approximately 500 m X 100 m) area between Pointe
a Lourneau and Pointe dii Moulin on lie Perrot, Que-
bec, Canada (45° 22' N, 73°51' W). Yellow Warblers
began arriving to our study area during the second
week of May in all four years. Typically one or two
males arrived first; two or three days later large num-
bers of males arrived, along with the first females.
Nests are built exclusively by females, and nest-build-
ing takes about four days. Eggs are usually laid on
consecutive days, at a rate of one per day. Clutches
normally contain four or five eggs. Incubation begins
when the last egg is laid, and the first egg hatches 10
days later. During the nestling period only females
brood, but both parents feed their nestlings. Pledging
normally occurrs approximately 10 days after hatch-
ing.

Adult birds were captured using mist nets; each bird
captured was banded with a U.S. Fish and Wildlife
Service (USFWS) aluminum band and three colored
plastic bands. Mist nets were operated throughout the
study area on a daily basis from May to early June,
starting at dawn and ending at 11:00-13:00, weather-
permitting. Three to five nets were used simultaneous-
ly in close proximity to one another, and these were
moved such that the entire study area was sampled
every 7-10 days. After 3 or 4 such cycles, we focused
our effort for 1-2 weeks on unbanded birds whose
nests had already been found. Nestlings were banded
at the nest when they were six days old, but only with
a USFWS band. In any given year, each adult bird
captured was classified as a returning bird if it had
been banded as an adult in previous years; otherwise
it was classified as a new arrival. It is impossible to
know with certainty whether every bird was captured
every year, so a few returning birds may have been
misclassified as being new to the site; all tests must
therefore be considered conservative.

Several standard morphological measurements were
taken from each bird: the lengths of the flattened wing
chord, the ninth primary feather, the outermost rectrix,
and the tarsus. The weight of each bird was recorded
to the nearest 0.1 g using a calibrated Pesola® spring
balance.

From early May to the middle of June the area was
thoroughly searched daily for nests, concentrating in
areas where birds had been banded and using behav-
ioral cues from females. Except during days of contin-
uous rain, every nest was visited daily to document the
chronology of nest building and egg laying. Many
nests failed as a result of harsh weather or predators;
renesting attempts usually followed. Two- categorical
measures were used to determine the breeding stage
reached: whether at least one egg was laid at a given
nest (clutch initiation) and whether at least one egg
hatched. Clutch initiation and hatching dates were used
to indicate the timing of reproduction; replacement
nests were not used in these comparisons. Laying date
was defined as the date the first egg was laid, and

Lozano and Lemon • BREEDING SUCCESS IN YELLOW WARBLERS

383

TABLE

were new

test.s; male

1. Frequency of Yellow Warblers that reached the "first egg” stage depending on
to the area or returning birds, by year and sex (expected frequencies in parentheses),
s: X- = 0.006, df = 2, P > 0.05; females: X' = 0.706, df = 2, P > 0.05.

whether they
Heterogeneity

Did not reach the

■•first egg" stage

Reached the "first

egg" stage

New

Return

New

Return

Males

1993

13 (13.3)

13 (12.7)

1 1 (10.7)

10 (10.3)

0.02

»0.()5

1994

30 (31.1)

15 ( 13.9)

19 (17.9)

7 (8.1)

0.09

»0.05

1995

19 (19.3)

12 (11.7)

14 (13.7)

8 (8.3)

0.01

»0.05

Total

171

62 (63.2)

40 (38.8)

44 (42.8)

25 (26.4)

0.06

»0.05

Females

1993

21 (16)

2 (7)

1 1 (16)

12 (7)

8.32

0.004

1994

23 (21.1)

8 (9.9)

22 (23.9)

13 (11.1)

0.52

»0.05

1995

15 ( 1 1.4)

4 (7.6)

12 (15.6)

14 (10.4)

3.65

0.056

Total

157

59 (48.3)

14 (24.7)

45 (55.6)

39 (28.3)

1 1.8

0.001

similarly, hatching date as the date the first egg
hatched. As used here, the terms “clutch initiation”,
"first-egg stage", “laying date”, and “hatching date”
refer to features of the nest, and therefore apply to both
the male and the female associated with the nest. In
several other species clutch size often decreases as the
breeding sea.son progresses (Erikstad et al. 1985, Mur-
phy 1986, Perrins and McCleery 1989), so only initial
nesting attempts were used in comparisons of clutch
size. Other estimates of breeding performance were not
used in these analyses because intrusive experiments,
beginning at the time of hatching but random with re-
spect to the variables examined here, were carried out
every year (Lozano and Lemon 1995, 1996, 1998).

Although the date on which birds began to arrive
during the four years of the study varied by up to two
weeks, the first day on which an egg was laid in the
population was fairly consistent. Erom 1992 to 1995,
these first egg dates were May 26 and 24, June 2, and
May 25 respectively. These yearly first egg dates were
used to account for variation in laying date among
years; relative laying and hatching dates of individual
nests are defined as the number of days after the re-
spective yearly first egg dates, added to the mean first
egg date from all four years.

Relationships between categorical variables were
examined using Fisher’s exact tests for n < 20, tests
for 40 > n > 20, and x“ with Yates’ adjustment for
continuity for n > 40 (Cochran 1954, Fienberg 1980,
Everitt 1992). Because the mobility of the two sexes
may have differed, e.specially early in the breeding .sea-
son when the birds were banded, the data were ana-
lyzed separately for males and females. Within-indi-
vidual increases in size and breeding performace be-
tween con.secutive seasons were te.sted with one-tailed
paired /-tests or Wilcoxon matched-pairs signed ranks
te.sts; if an individual was sampled repeatedly in sev-
eral years, data from only the first 2 years were used
in these analyses. Size differences between groups
were also as.sessed using the four morphological vari-
ables in a MANOVA, with group and year as the in-

dependent variables. All statistical analyses were per-
formed or verified using Statistica (5.1 and 98, under
Windows 3.1 1 and 95). Statistical significance was ac-
cepted at P < 0.05.

RESULTS

Among males there was no relation be-
tween prior residence and whether a clutch
was initiated in their nests (Table 1). In con-
trast, among females the number of prior res-
idents that initiated a clutch was higher than
expected, as was the number of birds new to
the area that failed to initiate a clutch. The
pattern was the same in all three years, sig-
nificant in 2 of the 3 years and in the analysis
with data pooled from all years (Table 1). Fi-
nally, there were no significant differences in
first egg date (F, ,5, = 0.65, P > 0.05), clutch
size (F, ,47 = 0.17, P > 0.05), or hatching date
~ 0.21, P > 0.05) between birds new
to the area and returning individuals.

Differences between groups do not neces-
sarily imply changes within individuals. Com-
parisons within individuals in successive
breeding seasons show no increases in clutch
size in either sex (Wilcoxon matched-pairs
signed-ranks tests: males Z = 0.36, // = 18, F
> 0.05; females Z = 0.82, n = 21, F > 0.05).
However, for males laying date was signifi-
cantly earlier in successive breeding seasons,
on average 2.8 days earlier (Wilcoxon
matched-pairs signed-ranks test: Z = 1.78, n
~ 18, F = 0.04; Fig. la). This was largely
due to late-nesting males nesting markedly
earlier in the following year, as there was little
difference among males that nested in late

Relative Clutch Initiaion Date, Year (n+1 ) Relative Clutch Initiation Date, Year (n+1 )

384

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

25-May 30-May 4-Jun 9-Jun 14-Jun 19-Jun 24-Jun 29-Jun
Relative Clutch Initiation Date, Year n

29- Jun

24-Jun

19-Jun

14-Jun

9-Jun

4-Jun

30- May

25-Ma'

•

Females

• * O 0

. o o

o o

May 30-May 4-Jun 9-Jun 14-Jun 19-Jun 24-Jun 29-Jun
Relative Clutch Initiation Date, Year n

FIG. 1. Clutch initiation dates by individuals in
consecutive breeding seasons, corrected for differences
between years (see methods). The dashed lines have a
slope of 1, which would result if clutch initiation oc-
curred on the same relative date in successive breeding
seasons. The two solid circles in the lower graph were
categorized as outliers (see results).

May— early June (Fig. la). In contrast, the dif-
ference in females was only 1.3 days earlier,
and was not significant (Wilcoxon matched-
pairs signed-ranks test Z = 0.93, n = 2\, P =
0.18; Fig. lb). However, the relationship is
heavily influenced by 2 nests that may have
been re-nesting attempts. Without these out-
liers the margin becomes 4.0 days earlier and
statistically significant (Wilcoxon matched-
pairs signed-ranks test: Z = 1.85, n = 19, P
= 0.032; Fig. lb). The removal of these out-
liers did not lead to categorically different
conclusions in all other analyses.

A significant relationship between breeding
performance and subsequent return occurred
in only one of the three years. Both males and
females that had at least started a clutch in
1992 were more likely to return the following
year (Table 2), but there was no evidence of
this relationship in the other 2 years. Although
the results were also significant when using
pooled data, this was due solely to the 1992
results (Heterogeneity t^sts: males P =
0.014; females P = 0.065; Table 2). Similar
results were obtained if return frequency was
compared to having hatched at least one off-
spring. Among birds that nested, there were
no significant differences in first egg date
(F,,i66 = 0-60, P > 0.05), clutch size (F,.i66 =
0.21, P > 0.05), or hatching date (F, =

1.10, P > 0.05) between birds that subse-
quently returned and those that did not. There-
fore, the effect of breeding performance on
subsequent return, as measured here, was cer-
tainly not consistent, if present at all.

TABLE ^ Relationship between having reached the “first egg" stage in one breeding season and returning

in the following year. Continuity adjusted or Fisher's exael tests (•) were used dependin_g on the sample siee

and its distribution (expected frequencies in parentheses). Heterogeneity tests; males, x 8.51. dt

0.014; females: X' = 5.461, df = 2, F = 0.065.

Did not reach the

■•first egg" .stage

Reached the

“Hrst egg" stage

Did not return

Returned

Did not return

Returned

Males

1992

29 (20.5)

3 (11.7)

14 (22.5)

21 (12.5)

16.5

<0.001

1995

15 (14.9)

1 1 (11.1)

12 (12.1)

9 (8.9)

0.07

»0.05

1994

53 (32.3)

12 (12.7)

18 (18.7)

8 (7.3)

0.01

»0.05

Total

185

77 (67.4)

26 (35.6)

44 (53.6)

38 (28.4)

8.07

0.005

Females

1 99'>

8 (4.1)

1 (4.9)

9 (12.9)

19 (15.1)

0.005

1995

17 (14.5)

6 (8.5)

12 (14.5)

1 1 (8.5)

1.49

»0.05

1 994

25 (26.3)

6 (4.7)

31 (29.7)

4 (5.3)

0.31

»0.05

Total

149

50 (43.1 )

13 (19.9)

52 (58.9)

34 (27.1)

5.17

0.023

Lozano and Lemon • BREEDING SUCCESS IN YELLOW WARBLERS

385

Wing Chord 9th Primary Tail Tarsus

FIG. 2. Mean size ( + SE) of Yellow Warblers ver-
sus residency history. MANOVAs were carried out us-
ing all variables (Males: Wilks’ \ = 0.901, /^4i5i =
4.14, P = 0.003; Females; Wilks’ X = 0.976, C4 ,33 =
0.803, P ^ 0.05). Asterisks indicate significant differ-
ences {P < 0.05) resulting from univariate ANOVAs.

Returning males were significantly larger
than new arrivals (Wilks’ \ = 0.901, ^4 ,5, =
4.14, P = 0.003; Fig. 2) but returning females
were not larger than new arrivals (Wilks’ X. =
0.976, ^4 ,33 = 0.803, P > 0.05; Fig. 2). In

contrast, there were significant within-individ-
ual increases in size between successive
breeding seasons in both sexes (Table 3).

Finally, there were no significant differenc-
es in morphology between nesting birds that
subsequently returned and nesting birds that
did not return, in either sex (MANOVA:
Males Wilks’ \ = 0.989, F4 ,67 = 0.448, P >
0.05; Females Wilks’ X = 0.9477, F4 ,,5 =
1.724, P > 0.05).

DISCUSSION

The effects of prior residency on breeding
performance differed substantially between
the sexes. Comparisons of new arrivals with
returning birds showed that the likelihood of
breeding increased with prior residency in fe-
males, but not in males. Among birds that
nested, there were no significant differences
in the timing of breeding and clutch size be-
tween new arrivals and returning birds. How-
ever, within-individual comparisons showed
that males began breeding significantly earlier
in subsequent breeding seasons. Therefore, the
effect of prior residency was greater for fe-
males than for males; in females local expe-
rience affected the prospect of breeding, and
in males it only affected the timing of breed-
ing.

There can be several advantages to breed-
ing earlier. Early breeders often have larger
clutches (Erikstad et al. 1985, Murphy 1986,
Perrins and McCleery 1989). In the popula-
tion we studied the median clutch size de-
creased from five eggs in nests started before
June 15, to four eggs in nests initiated there-
after. Fledglings from earlier nests also have

TABLE 3. Within-individual increa.se.s
results from one-tailed paired /-tests.

in size between

successive

breeding seasons

in Yellow Warblers and

Variable

Difference (mm)

Year^ , i - Year,

Males

Tail

0.50

0.32

1 .55

0.06

Wing chord

0.46

0.21

2.14

0.02

Ninth primary

0.97

0.28

3.53

<0.001

Tarsus

0.20

0.14

1.49

0.07

Females

Tail

0.27

0.31

0.89

>0.05

Wing chord

0.74

0.34

2.18

0.02

iNinth primary

0.74

0.32

2.32

0.01

Tarsus

0.31

0.14

2.28

0.02

386

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

more time to develop before migration, which
in our area begins in the latter half of July.
Perhaps most important, nest losses, whether
they are due to inclement weather or preda-
tion, are very common in open nesting birds
(Nice 1957, Martin and Li 1992). An earlier
start provides a longer available breeding sea-
son, which enhances the probability of re-
nesting following nest losses (Lozano et al.
1996).

Birds that arrive first may also benefit by
obtaining preferred territories (Wooller and
Coulson 1977, Newton 1988, Morris and
Lemon 1989, Lemon et al. 1996). On the other
hand, early arrivals also risk death from ex-
posure to cold weather and lack of food early
in the spring (Anderson 1965, Whitmore et al.
1977). Several other researchers have shown
that earlier arrivals are larger and in better
condition than later arrivals (e.g., Arvidsson
and Neergaard 1991, Lozano 1994, Stolt and
Fransson 1995). Size is important because ar-
rival time depends partially on the ability to
withstand adverse conditions early in the
breeding season. In our population, however,
it is unlikely that size affected arrival time,
because there was little variance in arrival
dates, but even given similar arrival times,
large size may still be advantageous in intra-
sexual competition during territory establish-
ment and defense (Arcese 1987, Hogstad
1989).

Individual males were larger in subsequent
breeding seasons, and, as a group, returning
males were also larger than males new to the
area. In contrast, returning females were not
significantly larger than females new to the
area, but individual females recaptured in con-
secutive breeding seasons were larger from
one breeding season to the next. At first
glance, the latter results may appear contra-
dictory, but the two analyses are not equiva-
lent. Females may increase in size between
consecutive breeding seasons without neces-
sarily leading to returning females being larg-
er than new arrivals. The differences between
the sexes may result, for instance, if the breed-
ing dispersal or size variance of females is
greater than that of males.

Size may also play a role in mate choice
later in the season. In Yellow Warblers extra-
pair paternity is widespread (Yezerinac et al.
1995), and within-pair paternity increases

with male size (Yezerinac and Weatherhead
1997). Female preference for larger males
may occur if size is an honest indicator of
phenotypic or genotypic quality, or if size is
being used as an indicator of age. If there is
a genetic component of survival ability, age,
in itself, would be a measure of genetic qual-
ity, and females would benefit by mating with
older males (Manning 1985).

We do not know the exact ages of all adult
birds studied, but we assume the mean age of
new arrivals is less than that of returning
birds. This is because in Yellow Warblers
breeding dispersal is limited (Yezerinac and
Weatherhead 1997), but natal dispersal is
common, as it is in birds in general (Green-
wood 1980, Greenwood and Harvey 1982,
Clarke et al. 1997). We have no measures of
dispersal exclusive of mortality, but only 4.5%
of all nestlings returned, compared to an over-
all yearly return of 36% for adults. Although
we cannot be certain that every individual
classified as a new arrival was only one year
old, returning birds were, by necessity, at least
two years old. Therefore, when compared as
two separate groups, not as individuals, it is
safe to conclude that returning birds were old-
er than new arrivals (see also Yezerinac and
Weatherhead 1997).

Age-dependent increases in reproductive
success have been well documented in birds
(reviewed by Ssether 1990, Forslund and Part
1995), but the proximate mechanisms respon-
sible have been difficult to determine because
potential factors are often correlated. Apparent
age-dependent increases in reproductive suc-
cess can be a statistical consequence of the
gradual disappearance of poor breeders from
a population, which would result if the phe-
notypic or genotypic quality of an individual
affects both its chances of breeding and sur-
viving (Curio 1983). Alternatively, individuals
may be more likely to disperse to other areas
in subsequent breeding seasons after a failed
breeding attempt (Harvey et al. 1979, Bensch
and Hasselquist 1991). Either mechanism
would yield a difference in reproductive suc-
cess between age classes, but this did not oc-
cur in our study. Except for one year, breeding
performance did not affect subsequent return
for either sex. Furthermore, there were no
morphological differences between birds that
subsequently returned and those that did not.

Lozano and Lemon • BREEDING SUCCESS IN YELLOW WARBLERS

387

These results also confirm that non-breeders
were not actually transients, captured while on
route to their final destinations. Had this been
the case they would have been less likely to
return than breeders.

The restraint hypothesis proposes that
young individuals deliberately withhold repro-
ductive effort. Life-history theory suggests
that age of first reproduction is an important
component of lifetime reproductive success
(Charlesworth 1980, Clutton-Brock 1988).
Our records show that 64% of all adult Yellow
Warblers fail to return, which makes it very
unlikely that individuals would purposely
forego the opportunity to reproduce, especial-
ly after having already migrated to the breed-
ing grounds. Moreover, the decision not to re-
produce would be optimal only if there is a
large cost to reproduction, but we found that
breeding was not related to subsequent return
rates. Therefore, the restraint hypothesis is
probably not appropriate to explain age-de-
pendent increases in reproductive success in
Yellow Warblers or other short-lived migra-
tory species (Wheelwright and Schultz 1994).

Our results are consistent with the con-
straint hypothesis, which predicts that younger
birds are disadvantaged when competing for
breeding opportunities. However, it is difficult
to know to what extent these results are
caused by differences in local experience or
by age in itself. Experimental work will be
required to determine the ecological and prox-
imate mechanisms responsible for these dif-
ferences (Martin 1995).

ACKNOWLEDGMENTS

We thank C. Daniel, G. Goggin-Michaud, D. Lafleiir
and C. Riley for their help with the field work. Finan-
cial support was provided by NSERC, FCAR, and the
John K. Cooper Foundation. We appreciate the com-
ments of J. Grant. D. Kramer, E. Nol, G. Pollack, R.
Titman, H. Reiswig, N. Wheelwright, and D. A. Wie-
denfeld.

LITERATURE CITED

Aldrich, J. W. 1942. Specific relationships of the
Golden and Yellow warblers. Auk 59:47-449.
Anderson, D. W. 1965. Spring mortality in insectiv-
orous birds. Loon 37:134-135.

Arcese, P. 1987. Age, intrusion pressure and defense
against floaters by temtorial male Song Sparrows.
Anim. Behav. 35:773-784.

Arvidsson, B. L. and R. Neergaard. 1991. Mate

choice in the Willow Warbler — a field experiment.
Behav. Ecol. Sociobiol. 29:225-229.

Bensch, a. and D. Hasselquist. 1991. Territory infi-
delity in the polygynous Great Reed Warbler Ac-
rocephaliis ariindinacetis: the effect of variation
in tenitory attractiveness. J. Anim. Ecol. 60:875-
871.

Bent, A. C. 1963. Life histories of North American
wood warblers, part one. Dover Publications Inc.,
New York.

Charlesworth, B. 1980. Evolution in age-structured
populations. Cambridge Univ. Press, Cambridge,
U.K.

Clarke, A. L., B-E. SasTHER, and E. Rpskaft. 1997.
Sex bia.ses in avian dispersal: a reappraisal. Oikos
79:429-438.

Clutton-Brock, T. H. 1988. Reproductive success.
Univ. of Chicago Press, Chicago, Illinois.

Cochran, W. G. 1954. Some methods for strengthen-
ing the common tests. Biometrics 10:417-451.

Curio, E. 1983. Why do young birds reproduce less
well? Ibis 125:400-404.

CuRSON, J., D. Quinn, and D. Beadle. 1994. Warblers
of the Americas. Houghton Miffin Company, Bos-
ton, Massachusetts.

Erikstad, K. E., H. C. Pedersen, and J. B. Steen.
1985. Clutch size and egg size variation in Willow
Grouse Lagopus /. lagopus. Ornis Scand. 16:88-
94.

Everitt, B. S. 1992. The analysis of contingency ta-
bles, second edition. Chapman & Hall, London,
U.K.

Fienberg, S. E. 1980. The analysis of cross-classified
categorical data, second ed. MIT Press, Cam-
bridge, Massachusetts.

Forslund, P. and T. Part. 1995. Age and reproduction
in birds-hypotheses and tests. Trends Ecol. Evol.
10:374-378.

Greenwood, P. J. 1980. Mating systems, philopatry
and dispersal in birds and mammals. Anim. Be-
hav. 28:1140-1 162.

Greenwood, P. J. and P. H. Harvey. 1982. The natal
and breeding dispersal of birds. Annu. Rev. Ecol.
Syst. 13:1-21.

Harvey, P. H., P. J. Greenwood, and C. M. Perrins.
1979. Breeding area fidelity of the Great Tit (Pa-
ms major). J. Anim. Ecol. 48:305-313.

Hinde, R. a. 1956. The biological significance of the
ten'itories of birds. Ibis 98:340-369.

Hogstad, O. 1989. The presence of non-territorial
males in Willow Warbler Phyllo.uopiis trochilus
populations — a removal .study. Ibis 131:263-267.

Klein, N. K. and W. M. Brown. 1994. Intraspecific
molecular phylogeny in the Yellow Warbler (Den-
droica petechia) and implications for avian bio-
geography of the We.st Indies. Evolution 48: 1914-
1932.

Lemon, R. E., S. Perreault, and G. A. Lozano. 1996.
Breeding di.spersions and site fidelity of American
Redstarts (Setophaga mticilla). Can. J. Zool. 74:
2238-2247.

388

THE WILSON BULLETIN • Vol. Ill, No. 3, Sepleinher 1999

Lozano, G. A. 1994. Size, condition, and territory
ownership in male Tree Swallows (Tachycineta
bicolor). Can. J. Zool. 72:330-333.

Lozano. G. A. and R. E. Lemon. 1995. Food abun-
dance and parental care in Yellow Warblers (Den-
droica petechia). Behav. Ecol. Sociobiol. 37:45—
50.

Lozano, G. A. and R. E. Lemon. 1996. Male plumage,
paternal care and reproductive success in Yellow
Warblers, Dendroica petechia. Anim. Behav. 51:
265-272.

Lozano, G. A. and R. E. Lemon. 1998. Parental care
responses by Yellow Warblers (Dendroica pete-
chia) to simultaneous manipulations of food abun-
dance and brood size. Can. J. Zool. 76:916—924.

Lozano, G. A., S. Perreault, and R. E. Lemon. 1996.
Age, arrival date and reproductive success of male
American Redstarts Setophaga nilicilla. J. Avian
Biol. 27:164-170.

Manning, J. T. 1985. Choosy females and correlates
of male age. J. Theor. Biol. 1 16:349-354.

Martin, K. 1995. Patterns and mechanisms for age-
dependent reproduction and survival in birds. Am.
Zool. 35:340-348.

Martin, T. E. and P. Li. 1992. Life history traits of
open- vs. cavity-nesting birds. Ecology 73:579-
592.

Morris, M. M. J. and R. E. Lemon. 1989. Mate choice
in American Redstarts: by territory quality? Can.
J. Zool. 66:2255-2261.

Murphy, T. M. 1986. Temporal components of repro-
ductive variability in Eastern Kingbirds (Tyranini.s
tyrannu.s). Ecology 67:1483—1492.

Newton, I. 1988. Age and reproduction in the Spar-
rowhawk. Pp. 201-219 in Reproductive success
(T. H. Clutton-Brock, Ed.). Univ. of Chicago
Press, Chicago, Illinois.

Nice, M. M. 1957. Nesting success in altricial birds.
Auk 74:305-321.

Nol, E. and j. N. M. Smith. 1987. Effects of age and
breeding experience on seasonal reproductive suc-
cess in the Song Sparrow. J. Anim. Ecol. 56:301-
313.

Perrins, C. M. and R. H. McCleery. 1989. Laying

date and clutch size in Great Tits. Wilson Bull.
101:236-253.

PiANKA, E. R. and W. S. Parker. 1975. Age-specific
reproductive tactics. Am. Nat. 109:453-464.

Pyle, R, L. B. Spear, W. J. Sydeman, and D. G. Ain-
LEY. 1991. The effects of experience and age on
the breeding performance of Western Gulls. Auk
108:25-33.

Reid, W. V. 1988. Age-specific patterns of reproduc-
tion in the Glaucous-winged Gull: increased effort
with age? Ecology 69:1454-1465.

S/ETHER, B-E. 1990. Age-specific variation in repro-
ductive performance of birds. Curr. Ornithol. 7:
251-283.

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

Smith, J. N. M. 1981. Does high fecundity reduce sur-
vival in Song SpaiTOWS? Evolution 35: 1 142-1 148.

Smith, H. G. 1993. Parental age and reproduction in
the Marsh Tit Parit.s pahistris. Ibis 135:196-201.

Stolt, B. O. and T. Fransson. 1995. Body mass, wing
length and spring arrival of the Ortolan Bunting
Emheriza hortidana. Ornis Fennica 72:14-18.

Wheelwright, N. T. and C. B. Schultz. 1994. Age
and reproduction in Savannah Span'ows and Tree
Swallows. J. Anim. Ecol. 63:686-702.

Whitmore, R. C., J. A. Mosher, and H. H. Frost.
1977. Spring migrant mortality during unseason-
able weather. Auk 94:778-781.

WiEDENFELD, D. A. 1991. Geographical moiphology of
male Yellow Warblers. Condor 93:712—723.

Williams, G. C. 1966. Natural selection, the cost of
reproduction and a refinement of Lack's principle.
Am. Nat. 100:687—690.

WOOLER, R. D. AND J. C. CouLSON. 1977. Factors af-
fecting the age of first breeding of the Kittiwake.
Ris.sa tridactyki. Ibis 119:339-349.

Yezerinac, S. M. and P. J. Weatherhead. 1997. Ex-
tra-pair mating, male plumage coloration, and sex-
ual selection in Yellow Warblers (Dendroica pe-
techia). Proc. R. Soc. Lond. B 264:527-532.

Yezerinac, S. M., P. J. Weatherhead, and P. T. Boag.
1995. Extra-pair paternity and the opportunity for
sexual selection in a socially monogamous bird
(Dendroica petechia). Behav. Ecol. Sociobiol. 37:
179-188.

Wilson Bull.. 1 I 1(3), 1999, pp. 389-396

DISTRIBUTION AND HABITAT ASSOCIATIONS OF THREE
ENDEMIC GRASSLAND SONGBIRDS IN
SOUTHERN SASKATCHEWAN

S. K. DAVIS, D. C. DUNCAN,'^ AND M. SKEEL'^

ABSTRACT. — We conducted 1675 point counts on 93 survey routes to determine the distribution and habitat
associations of three endemic grassland songbirds across the four prairie ecoregions of southern Saskatchewan,
Canada. Within the four habitat types surveyed, Sprague’s Pipits (Anthiis spragueii) and Chestnut-collared Long-
spurs (Calcariits ornatu.s) occuned more frequently in native and seeded pastures than in hayland and cropland,
whereas Baird's Sparrows (Ammodramus hairdii) occurred as frequently in hayland as in native and seeded
pastures. The occuiTcnce of Baird's Spanows and Chestnut-collared Longspurs did not differ significantly be-
tween lightly, moderately, and heavily grazed native pastures, whereas Sprague’s Pipits occurred less frequently
in heavily grazed pastures. Sprague’s Pipits and Chestnut-collared Longspurs occurred more often in the drier
prairies of the southern portion of the province, but Chestnut-collared Longspurs were virtually absent from the
cypress upland ecoregion. In contrast, Baird’s Sparrows occurred most frequently in the .semi-arid grasslands of
the moist-mixed grassland ecoregion. In native pastures. Chestnut-collared Longspurs were associated with a
lower density of short grasses and lesser amounts of litter whereas Baird’s Sparrows were associated with a
higher density of taller grasses and sparse shrub cover. Our results suggest that conservation programs that
convert annually tilled cropland to perennial forage could provide additional habitat for endemic grassland birds.
Received 2 Nov. 1998, accepted 2 April 1999.

Land settlement and agriculture have great-
ly altered the landscape of southern Saskatch-
ewan. Only 17% of the province’s original na-
tive prairie remains (Samson and Knopf 1994)
and is currently threatened by cultivation,
over-grazing by livestock, invasion by exotic
plant species, and urban development. In areas
of Saskatchewan where soils and landscapes
are particularly suited for crop production,
less than 0.1% of the original native prairie
remains (Riemer et al. 1997).

Despite the loss of native grassland, Sas-
katchewan supports a diverse grassland avi-
fauna (Smith 1996), including 1 1 of the 12
primary endemic, and 17 of the 25 secondary
endemic grassland bird species as outlined by
Knopf (1994). Primary endemic species gen-
erally have more restricted breeding ranges
(Sauer et al. 1997) and are less flexible in their
habitat requirements than more broadly dis-
tributed species (Owens and Myres 1973,

Knopf 1996, Davis and Duncan in press).

Consequently, continental populations of 7 of

' Saskatchewan Wetland Conservation Corporation.
202-2050 Cornwall St., Regina SK, Canada, S4P 2K5.

-Present address: Canadian Wildlife Service, 200-
4999-98"- Ave., Edmonton AB, Canada, T6B 2X3.

^Present address: Nature Saskatchewan, 206-1860
Lome St., Regina SK, Canada, S4P 2L7.

^ Corresponding author; E-mail: .sdavis@wetland.sk. ca

the 1 1 primary endemic species are currently
in decline (Sauer et al. 1997), possibly as a
result of loss and degradation of native prairie
habitat.

Information on the habitat associations of
endemic songbirds can provide insight into
population declines by identifying habitat fea-
tures that correlate with their occurrence. This
information may be used in making manage-
ment decisions and formulating land-use pol-
icies. Although many researchers have ex-
amined habitat selection by grassland song-
birds (Cody 1968, Owens and Myres 1973,
Whitmore 1979, Rotenberry and Wiens 1980,
Johnson and Temple 1986, Mahon 1995), lit-
tle research has been focused on birds of the
northern mixed-grass prairie. Furthermore,
most studies in the mixed-grass prairie region
have been conducted on intensively managed
sites (Dale 1983, Winter 1994, Madden 1996,
Dale et al. 1997) or were located within a
small geographic area (Arnold and Higgins
1986, Sutter 1996; but see Johnson and
Schwartz 1993, Davis and Duncan in press).

We examined habitat associations and the
distribution of three endemic songbird species
of the northern mixed-grass prairie across the
entire Prairie Ecozone of southern Saskatch-
ewan. The objectives of the study were (1) to
determine whether the frequencies of occur-

389

390

THE WILSON BULLETIN • Vol. III. No. 3. September 1999

rence of Sprague’s Pipit {Anthus spragueii),
Baird’s Sparrow {Ammodramus bairdii), and
Chestnut-collared Longspur (Calcarius orna-
tus) differ among native pasture, seeded pas-
ture, hayland, and cropland; (2) to determine
whether various levels of grazing intensity in-
fluence the occurrence of these species; (3) to
determine the distribution of the three species
relative to the prairie ecoregions of Saskatch-
ewan; and (4) to identify structural compo-
nents of native prairie vegetation important in
predicting songbird occurrence.

STUDY AREA AND METHODS

Study area. — We conducted grassland bird surveys
throughout the Prairie Ecozone of southern Saskatch-
ewan. The Prairie Ecozone covers 24,103,000 ha and
comprises four ecoregions; cypress upland, mixed
grassland, moist-mixed grassland, and aspen parkland
(Fig. 1; Ecological Stratification Working Group
1995). The cypress upland in the extreme southwestern
portion of the province rises 400—500 m above the
prairie landscape. This region is characterized by slop-
ing escarpments, valleys, and coulees. Wheatgrass
(Agropyron spp.) and speargrass {Stipa spp.) dominate
the dark brown soils of the lower elevations, whereas
fescue (Festuca spp.) prairie predominates on the
slopes and at higher elevations. The mixed grassland
is the driest region of Saskatchewan and is character-
ized by wheatgrass, speargrass, and blue grama grass
(Bouteloua gracilis). Because of the lack of moisture,
trees and wetlands are scarce; shrubs are restricted to
mesic areas. The moist-mixed grassland represents the
northern extent of the open grasslands in Saskatche-
wan. This region is characterized by semiarid condi-
tions and dark brown soils. Speargrass, wheatgrass,
and deciduous shrubs predominate. The aspen park-
land ecoregion is characterized by trembling aspen
{Populus tremuloide.s) groves and fescue grasslands,
although the latter habitat is now rare (Sask. Wetland
Conserv. Corp., unpubl. data).

Bird surveys were conducted in native pasture, seed-
ed pasture, hayland, and cropland. Survey routes were
designed to sample mostly grassland habitat, although
each of the four habitat types was sampled on most
routes. Native prairie was characterized by Stipa spp.,
June grass (Koeleria cristata). northern wheatgrass
(Agropyron dasystachyum), western wheatgrass (A.
smithii). blue grama grass, Care.x spp., club moss (Se-
laginella densa). pasture sage (Artemisia frigida), and
various forbs. The most common shrubs were western
snowberry (Symphoricarpos occidentalis), rose (Rosa
spp.), and wolf willow (Eleagmis commutata). Seeded
pasture was defined as land that had been broken and
seeded with exotic perennial gras.ses tor grazing, most
commonly crested wheatgrass (Agropyron cri.statum)
or brome grass (Bronnts spp.), with altalta (Medi( ago
spp.) or sweet clover (Melilotus spp.) sometimes pre-
sent. Hayland was defined as cultivated land that had

been seeded to perennial crops for haying. Vegetation
on hayland ranged from 100% alfalfa, to mixes of al-
falfa, sweet clover and introduced grasses such as
brome grass, crested wheatgrass, or bluegrass (Poa
spp.). Cropland was cultivated land that was seeded to
annual crops, most commonly wheat (Triticum aesti-
vum) or canola (Brassica spp.).

Route selection. — We numbered (1-42) townships
from the U.S. border to the northern extent of the as-
pen parkland, and renumbered (1-64) range locations
from the Manitoba border west to the Alberta border.
The starting points of 76 routes were then located by
selecting township-range numbers from a random
number table. Because the target species (Baird’s Spar-
row) was believed to be a grassland specialist (Cart-
wright et al. 1937, Owens and Myres 1973), we only
included townships where most point counts (>80%)
could be located in grassland habitat. Thus more routes
were established in the mixed grassland ecoregion than
the other ecoregions because most of the grassland in
Saskatchewan exists in this region (Saskatchewan Dig-
ital Land Cover Project, unpubl. data). Because Baird’s
Sparrow was a threatened species in 1994 (Goossen et
al. 1993), we assumed they were uncommon and es-
tablished an additional 19 non-random routes in areas
where the species was thought to occur regularly. The
proportion of point counts in grassland habitat was
nearly identical for random and non-random routes (79
and 80%, re.spectively).

Bird occurrence. — Six surveyors recorded the num-
ber of singing males using five-minute, 100-m fixed
radius point counts (Ralph et al. 1993) conducted from
roads and trails at approximately 0.8 km intervals
along each route. Each point count was subdivided into
half-circle counts by recording birds detected on the
left or right side of the road or trail because habitat
types and grazing intensity often differed on each side.
Although roadside sampling may confound species-
habitat as.sociations because of the influence of road-
side vegetation, the three species in this study are not
attracted to roadside habitat (Sutter et al., in press).
Thus the occurrence of these species in the four habitat
types is likely not confounded by roadside vegetation.
Surveyors attended a training session prior to the study
to standardize protocol and reduce surveyor bias. Sur-
veys were conducted between 4 June and 2 July, 1994,
commenced 30 minutes before sunrise, on days with
no precipitation, and winds less than 20 km/hr. Each
route had 20-25 stops.

Habitat as.sociations. — Bird surveyors quantified
eight measures of vegetation structure at every second
native pasture point count location on the same day as
the survey. Vegetation was assessed on only native
pastures because of logistic constraints in sampling all
habitats over a large area and because of our increased
sampling of native pasture. Each surveyor laid a meter
stick on the ground at least 35 m from the road/trail
at a random location within each portion of the half-
circle. A 6 mm diameter metal rod was passed verti-
cally through the vegetation at each end of the meter
stick and the number of contacts by different vegeta-

Davis el al. • ENDEMIC GRASSLAND SONGBIRDS

391

FIG. 1. Location of random (closed circles) and non-random (open circles) survey routes and ecoregions
within the Prairie Ecozone of southern Saskatchewan.

tive life forms (e.g., standing dead vegetation, narrow-
leaf grass, broad-leaf grass, forb, shrubs >15 cm, and
dwarf shrubs <15 cm high) counted in each successive
1 decimeter (dm) height interval (Rotcnbeny and
Wiens 1980). The occuiTence of each life form was
lumped into two height categories; 1 dm (number of
contacts <1 dm) and 2+ dm (number of contacts >1
dm) height intervals because vegetation contacts in na-
tive pastures rarely occurred in the higher levels. Litter
depth was measured from the surface of the ground to

the top of the litter, and distance to the nearest shrub
was visually e.stimated.

Grazing intensity was estimated in June and early
July by a range ecologist as idle, lightly, moderately,
or heavily grazed. Idle prairie was defined as native
vegetation that had not been grazed for at least two
years. Heavily grazed pa.stures were characterized by
virtually all plant material and litter removed, greater
than 20% bare .soil, greater than 40% club moss, and
small plants with poor vigor. Lightly grazed pasture

392

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 1 . Influence of habitat type,
singing males in southern Saskatchewan.

ecoregion, and grazing intensity on the occurrence (% half circles) of
For each species, values followed by the same letter do not differ (P

> 0.05) from one another.

Sprague's Pipit

Baird's Sparrow

Chestnut-collared

Longspur

Habitat

Native pasture (n = 1158)

Seeded pasture (/; = 192)

Hayland (/? = 1 16)

Cropland {n = 209)

18.5 A

14.1 A

2.6 B

0.5 B

30.5 A
31.2 A
25.0 A

5.3 B

21.4 A

17.7 A

1.7 B

0.5 B

Ecoregion

Aspen parkland (;? = 238)

Cypress upland (n = 149)

Mixed grassland {n = 805)

Moist-mixed grassland (n = 483)

8.4 A

24.2 B

16.9 B

6.2 A

15.9 A

16.8 A

25.8 B
37.7 C

6.7 A

2.0 B

24.4 C

14.5 D

Lightly grazed (/? = 112)

Moderately grazed (n = 137)

Heavily grazed (/? = 81 )

Grazing intensity

19.6 A

22.6 A

9.9 B

34.8

37.9

25.9

23.2

24.1

22.2

was defined as having little evidence of grazing, abun-
dant litter and plant material, less than 10% bare soil,
less than 10% club moss, and robust and vigorous
plants. Moderately grazed pastures exhibited charac-
teristics intermediate between the previous two types.

Statistical analyses. — We used songbird occurrence
(presence/absence) within half-circle point counts for
all analyses because more than one individual was re-
corded in only 9%, 8%, and 2% of the half-circle
counts for Baird's Sparrow, Chestnut-collared Long-
spur and Sprague’s Pipit, respectively. Because the two
half-circle counts at any given stop do not represent
independent observations, we randomly chose only
one half-circle count on either side of the road or trail
for inclusion in all subsequent analyses. x“ contingency
analyses were used to determine whether the frequency
of occurrence of songbirds inside halt-circle point
counts was influenced by land-use, grazing intensity,
and ecoregion. Pair-wise comparisons were pertormed
only for those species where the overall significance
level was P < 0.05. Although most of the native and
seeded pastures were grazed by cattle, 5% of the sam-
ple points had been idle for at least two years and were
thus omitted from all analyses. A multivariate assess-
ment of songbird occurrence in native pasture was con-
ducted using step-wise logistic regression on half-cir-
cle point counts using vegetative structure, grazing in-
tensity, and ecoregions. None of these variables were
highly correlated with each other (all comparisons r’
< 0.42, P > 0.001 ) except for the number of contacts
of broad-leaf grass in the first, and 2+ decimeter cat-
egories (/-’ = 0.71. P < 0.001); thus only broad-leaf
grass contacts in the first decimeter were used. Level
of significance for variable inclusion in the models was
set at 0.05. All analyses were performed using SAS
statistical software v. 6.12 (.SAS Institute Inc. 1989).

RESULTS

Land use. — The occurrence of each of the
three endemic species differed significantly
among habitat types (x^ = 59.2-75.6, df = 3,

P < 0.001). Sprague’s Pipit and Chestnut-col-
lared Longspur occurred more frequently in
native and seeded pasture than in hay land or
cropland, whereas Baird’s Sparrows occurred
as frequently in native and seeded pastures as
in hayland, but occurred least frequently in
cropland (Table 1).

Distribution. — The occurrence of each spe-
cies differed among ecoregions (Table 1 ; x" =
51.1-74.4, df = 3, P < 0.001). Furthermore,
ecoregion type was a significant predictor of
occurrence in each of the three logistic re-
gression models (Table 2). Sprague s Pipits
were recorded in relatively low abundance
throughout the study area, occurring most fre-
quently in the cypress upland and mixed
grassland ecoregions (Table 1). Chestnut-col-
lared Longspurs were primarily restricted to
the extreme southern portion of the province,
particularly within the mixed grassland ecore-
gion (Tables 1, 2). Baird’s Sparrows Were
abundant throughout much of the study area,
but were recorded most frequently in the
moist-mixed grassland (Tables 1, 2).

Habitat associations. — Grazing intensity
had little influence on the occurrence of

Davis el al. • ENDEMIC GRASSLAND SONGBIRDS

393

TABLE 2. Results summary ot step-wise logistic regression analyses of grassland songbird occurrence in
native pasture. Variables are presented in the order they were entered into the model.

Purunieter

Species

Variable

estimate

Wald

Sprague’s Pipit

Intercept

-0.894

19.922

<0.001

Moist-mixed grassland

-1.920

12.377

<0.001

0.092

Heavily grazed

-0.981

4.261

0.039

Baird's Sparrow

Intercept

- 1 .346

26.653

<0.00 1

Narrow-leaf grass 2' dm

0.315

9.480

0.002

0.075

Moist-mixed grassland

0.825

8.209

0.004

Shrub distance

0.008

3.941

0.002

Chestnut-collared Longspur

Intercept

-0.605

3.664

0.055

Litter depth

-0.035

4.525

0.033

0.088

Narrow-leaf grass 1 dm

-0.278

6.053

0.014

Mixed grassland

0.675

4.535

0.033

Baird’s Sparrow and Chestnut-collared Long-
spur in native pasture (Table 1; x“ = 3.4, df
= 2, P > 0.05 and = 0.1, df = 2, P >
0.05, respectively). The occurrence of
Sprague’s Pipits in native pasture differed sig-
nificantly among grazing intensity levels (x^
= 6.2, df = 2, P = 0.045). Sprague’s Pipits
were negatively associated with heavy grazing
(Table 2), occurring twice as often in lightly
and moderately grazed pastures as in heavily
grazed pastures (Table 1). Although grazing
intensity did not significantly influence the oc-
currence of Baird’s Sparrows, the species was
associated with pastures having greater cov-
erage of grasses over 10 cm and were attracted
to pastures with sparse shrub cover (Table 2).
In contrast. Chestnut-collared Longspurs ap-
peared to be associated with pastures that
were characterized by less dense vegetative
cover. The species was negatively associated
with depth of the litter and the density of nar-
row-leaf grasses in the first decimeter (Table
2). However, the amount of variation ex-
plained by each model was extremely poor
(Table 2).

DISCUSSION

Land-use. — Sprague’s Pipits and Chestnut-
collared Longspurs were mostly restricted to
grassland habitat in southern Saskatchewan.
Although we detected no differences in the
frequency of occurrence of these species in
native and seeded pastures, others have re-
ported Sprague’s Pipits (Owens and Myres
1973, Hartley 1994, Madden 1996, Dale et al.
1997) and Chestnut-collared Longspurs
(Stewart and Kantrud 1972, Owens and Myres

1973) to prefer native prairie over a number
of other habitat types. The attractiveness of
seeded pastures in this study may have been
influenced by the age and structural compo-
sition of the seeded pastures such that a num-
ber of these pastures may have been similar
to native pastures. Indeed, surveyors consult-
ed with the range ecologist on several occa-
sions to confirm whether pastures were native
or had been cultivated in the past. More re-
cently seeded pastures, or those dominated by
certain exotic plant species may be less suit-
able for Sprague’s Pipits and Chestnut-col-
lared Longspurs. In Saskatchewan, both spe-
cies occur more frequently and in higher
abundance in native pastures than in pastures
dominated by crested wheatgrass (Sutter
1996, Davis et al. 1996, Davis and Duncan in
press). Siinilarly, fields comprised predomi-
nantly of smooth brome grass {Bromus iner-
mis) are unsuitable for these species (Wilson
and Belcher 1989).

In our study, Baird’s Sparrows exhibited
more flexibility in their habitat use than either
Chestnut-collared Longspurs or Sprague’s
Pipits. The occurrence of Baird’s Sparrows in
habitats other than native prairie has been well
documented (reviewed in Davis et al. 1996)
despite earlier studies that suggested the spar-
row was a native prairie specialist (Cartwright
et al. 1937, Owens and Myres 1973). Al-
though we frequently recorded Baird’s Spar-
rows in hayfields, this habitat may act as a
population sink (Pulliam 1988). Using a pro-
ductivity index. Dale and coworkers (1997)
found significantly fewer signs of productive

394

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

behavior following mowing of hayfields in
southern Saskatchewan. Mowing destroys ap-
proximately 50% of ground nests and the pro-
ductivity of breeding birds in hayfields is of-
ten below that required to maintain a stable
population (Frawley 1989, Bollinger et al.
1990).

Distribution and habitat associations. — We
found that the distribution of Sprague’s Pipit,
Baird’s Sparrow, and Chestnut-collared Long-
spur closely resembled that outlined by the
Saskatchewan breeding bird atlas (Smith
1996) and the Breeding Bird Survey (BBS;
Sauer et al. 1997). Additionally, our results
suggest that these species are associated with
certain regions within the Prairie Ecozone.
Sprague’s Pipits and Chestnut-collared Long-
spurs occurred mostly in the drier prairies of
the southern portion of the province, but
Chestnut-collared Longspurs were virtually
absent from the cypress upland ecoregion de-
spite the presence of large tracts of native
grassland. Grassland habitat in this region is
taller and denser than the surrounding grass-
lands (W. Harris, pers. comm.) and thus may
not provide Chestnut-collared Longspurs with
the short, sparse vegetative cover they require
(Harris 1944, Owens and Myres 1973, Dale
1983). In contrast to Sprague’s Pipits and
Chestnut-collared Longspurs, Baird’s Spar-
rows occurred most frequently in the more
mesic grasslands of the moist-mixed grassland
ecoregion, consistent with their overall pref-
erence for taller and denser vegetation (Dale
1983, Winter 1994, Madden 1996, Sutter and
Brigham 1998, Davis and Duncan in press).
Furthermore, our study indicates that the Mis-
souri Coteau, which borders the mixed and
moist-mixed grassland ecoregions, is an im-
portant landform for Baird’s Sparrows in Sas-
katchewan (McMaster and Davis, upubl.
data). This landform has also been identified
as a significant region for Baird s Sparrows in
North Dakota (Stewart 1975).

Grazing by livestock can have a profound
influence on the structure of rangeland vege-
tation (Wiens and Dyer 1975, Ryder 1980).
Because habitat selection by grassland birds is
likely influenced by habitat structure (Wiens
1969), grazing intensity should affect the oc-
currence of the three songbird species exam-
ined in this study. Sprague’s Pipits, and to a
lesser degree, Baird’s Sparrows, were both in-

fluenced by grazing intensity. Light to mod-
erately grazed pasture has been described as
preferred habitat for both Sprague s Pipits and
Baird’s Sparrows (Kantrud 1981; Kantrud and
Kologiski 1982, 1983; Knopf 1996) although
others have suggested that these species prefer
idle prairie (Maher 1973, Owens and Myres
1973, Dale 1984). The latter studies, however,
examined only a small number of grazed and
ungrazed sites and did not discriminate be-
tween lightly or heavily grazed pastures. We
could not assess the suitability of ungrazed
prairie to either species in our study because
this habitat type is uncommon in Saskatche-
wan. Despite the lack of information on the
response of Sprague’s Pipits and Baird s Spar-
rows to low intensity grazing relative to un-
grazed native prairie, our results suggest that
low levels of grazing intensity are tolerated by
these species.

We found no difference in the response of
Chestnut-collared Longspurs to grazing inten-
sity in native pasture. Numerous studies have
reported that Chestnut-collared Longspurs re-
spond positively to grazing (Maher 1973, Dale
1983, Kantrud and Kologiski 1983, Renken
and Dinsmore 1987, Bock et al. 1993). Graz-
ing may not have influenced Chestnut-collared
Longspurs in our study because the structure
of the vegetation under all grazing intensities
may have fallen within an acceptable range.
Descriptions of Chestnut-collared Longspur
breeding habitats have ranged from over-
grazed pastures with sparse vegetation to sit-
uations where the “. . .thicker and taller grass-
es afford adequate concealment” (DuBois
1935:70). Chestnut-collared Longspurs in our
study were generally associated with sparsely
vegetated native pastures with low plant litter
depths (see also Harris 1944, Owens and My-
res 1973, Dale 1983, Johnson and Schwartz
1993).

Conservation. — While continental popula-
tions of Baird’s Sparrows and Chestnut-col-
lared Longspurs appear to be relatively stable
(Sauer et al. 1997), Sprague’s Pipits are cur-
rently undergoing significant population de-
clines of 4.7% per year, one of the steepest
declines recorded for grassland songbirds in
North America. The conversion of native
grassland to annually cropped land, and the
pattern of habitat loss (i.e., habitat fragmen-
tation) have likely played significant roles in

Dcivis ei al. • ENDEMIC GRASSLAND SONGBIRDS

395

these declines (Davis, unpubl. data). For ex-
ample, Sprague’s Pipits reach their highest
densities in southeastern Alberta and south-
western Saskatchewan (Sauer et al. 1997), ar-
eas characterized by large tracts of contiguous
native grassland (South Digital Land Cover
Project, unpubl. data). While land-use pro-
grams that convert annually tilled cropland to
perennial cover will likely provide additional
habitat for endemic grassland birds (Johnson
and Schwartz 1993, Reynolds et al. 1994, Sut-
ter and Brigham 1998, Davis and Duncan in
press), it is imperative that the reproductive
consequences of selecting alternative nesting
habitats be determined to accurately assess
habitat quality (Johnson and Temple 1986,
Van Horne 1983, Vickery et al. 1992).

ACKNOWLEDGMENTS

We thank G. Butcher, M. Hartley, B. Dale, D. Hjer-
taas, D. H. Johnson, and D. McKinnon for their helpful
advice on the initial design of the study. We are grate-
ful to our held crew, L. Banman, C. Bjorklund, R Hjer-
taas, R. Kreba, J. Pollock, and T. Troupe. Thanks to J.
Keith (Saskatchewan Conservation Data Center) for
assistance in the production of maps and T. Hairison
(Saskatchewan Wetland Conservation Corporation) for
his expertise in rangeland ecology. Comments by S. L.
Jones, D. G. McMaster, G. C. Sutter, and an anony-
mous reviewer greatly improved the manuscript. This
study would not have been possible without the co-
operation of private landowners and the staff and pa-
trons of the Prairie Farm Rehabilitation Administration
pastures. Provincial Community pastures, and Provin-
cial Co-op pa.sture.s. Financial support for this study
was provided by the National Fish and Wildlife Foun-
dation (U.S.), Endangered Species Recovery Fund
(World Wildlife Fund Canada and Canadian Wildlife
Service of Environment Canada), and Saskatchewan
Endangered Species Fund (Saskatchewan Environment
and Resource Management).

LITERATURE CITED

Arnold, T. W. and K. F. Higgins. 1986. Effects of
shrub coverages on birds of North Dakota mixed-
grass prairies. Can. Field-Nat. 100:10-14.

Bock, C. E., V. A. Saab, T, D. Rich, and D. S. Dob-
kin. 1993. Effects of livestock grazing on Neo-
tropical migratory landbirds in western North
America. Pp. 296—309 in Status and management
of Neotropical migratory birds (D. M. Finch and
P. W. Stangel, Eds.). U.S. For. Serv. Gen. Tech.
Rep. RM-229, Fort Collins, Colorado.

Bollinger, E. K., P. B. Bollinger, and T. A. Gavin.
1990. Effects of hay-cropping on ea.stern popula-
tions of the Bobolink. Wildl. Soc. Bull. 18:142-
150.

Cartwrigh'i, B. W., T. M. Siiorit, and R. D. Harris.
1937. Baird’s Sparrow. Trans. R. Can. Inst. 46:
153-198.

Cody, M. L. 1968, On the methods of resource divi-
sion in grassland bird communities. Am. Nat. 102:
107-147.

Dale, B. C. 1983. Habitat relationships of seven spe-
cies of passerine birds at Last Mountain Lake,
Saskatchewan. M.Sc. thesis, Univ. of Regina, Re-
gina, Saskatchewan.

Dale, B. C. 1984. Birds of grazed and ungrazed grass-
lands in Saskatchewan. Blue Jay 42:102-105.

Dale, B. C., P. A. Martin, and P. S. Taylor. 1997.
Effects of hay management on grassland song-
birds in Saskatchewan. Wildl. Soc. Bull. 25:616-
626.

Davis, S. K., D. C. Duncan, and M. Skeel. 1996. The
Baird’s SpaiTow: status re.solved. Blue Jay 54:
185-191.

Davis, S. K. and D. C. Duncan. In press. Grassland
songbird occurrence in native and crested wheat-
grass pastures of southern Saskatchewan. Stud.
Avian Biol.

Dubois, A. D. 1935. Nests of Horned Larks and long-
spurs on a Montana prairie. Condor 37:56-72.

Ecological Stratification Working Group. 1995. A
national ecological framework for Canada. Agri-
culture and Agri-Food Canada and Environment
Canada, Ottawa.

Fairfield, G. M. 1968. Chestnut-collared Longspun
U.S. Nat. Mus. Bull. 237:1635-1652.

Frawley, B. j. 1989. The dynamics of nongame bird
breeding ecology in Iowa alfalfa fields. M.S. the-
sis, Iowa State Univ., Ames.

Goossen, j. R, S. Brechtel, K. D. De Smet, D. Hjer-
TAAS, AND C. Wershler. 1993. Canadian Baird’s
Sparrow recovery plan. Recovery of Nationally
Endangered Wildlife Rep. No. 3. Canadian Wild-
life Federation, Ottawa.

Harris, R. D. 1944. The Chestnut-collared Longspur
in Manitoba. Wilson Bull. 56:105-115.

Hartley, M. J. 1994. Passerine abundance and pro-
ductivity indices in grasslands managed for wa-
terfowl nesting cover. Trans. Am. Wildl. Nat. Re-
sour. Conf. 59:322-327.

Johnson, D. H. and M. D. Schwartz. 1993. The con-
servation reserve program: habitat for grassland
birds. Great Plains Res. 3:273-295.

Johnson, R. G. and S. A. Temple. 1986. Assessing
habitat quality for birds nesting in fragmented tail-
grass prairie. Pp. 245-249 in Wildlife 2000. Mod-
elling habitat relationships of tenestrial verte-
brates (J. 'Verner, M. L. Morrison, and C. J. Ralph,
Eds.). Univ. of Wisconsin Press, Madison.

Kantrud, H. a. 1981. Grazing intensity effects on the
breeding avifauna of North Dakota native grass-
lands. Can. Field-Nat. 95:404-417.

Kantrud, H. A. and R. L. Kologiski. 1982. Effects
of soils and grazing on breeding birds of uncul-
tivated upland grasslands of the northern great

396

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

plains. U.S. Eish and Wildlife Service. Washing-
ton, D.C.

Kantrud, H. a. and R. L. Kologiski. 1983. Avian
associations of the northern Great Plains grass-
lands. J. Biogeogr. 10:331—350.

Knopf, E L. 1994. Avian assemblages on altered grass-
lands. Stud. Avian Biol. 15:247-257.

Knopf, E L. 1996. Prairie legacies-birds. Pp. 135—148
in Prairie conservation. Preserving North Ameri-
ca's most endangered ecosystem (E B. Samson
and E L. Knopf, Eds.). Island Press, Washington,
D.C.

Madden, E. M. 1996. Passerine communities and bird-
habitat relationships on prescribed-burned, mixed-
grass prairie in North Dakota. M.S. thesis, Mon-
tana State Univ., Bozeman.

Maher, W. J. 1973. Birds: I. Population dynamics.
Matador project technical report No. 34. Canadian
Committee for the International Biological Pro-
gramme, Saskatoon, Saskatchewan.

Mahon, C. L. 1995. Habitat selection and detectability
of Baird's Sparrows in southwestern Alberta.
M.Sc. thesis, Univ. of Alberta, Edmonton.

Owens, R. A. and M. T. Myres. 1973. Effects of ag-
riculture upon populations of native passerine
birds of Alberta fescue grassland. Can. J. Zool.
51:697-713.

Pulliam, H. R. 1988. Sources, sinks, and population
regulation. Am. Nat. 132:652—661.

Ralph, C. J., G. R. Geupel, P. Pyle, T. E. Martin,
and D. E Desante. 1993. Handbook of field
methods for monitoring landbirds. Gen. Tech.
Rep. PSW-GTR-144. Pacific Southwest Research
Station, Albany, California.

Renken, R. B. and j. j. Dinsmore. 1987. Nongame
bird communities on managed grasslands in North
Dakota. Can. Field-Nat. 101:551—557.

Reynolds, R. E., T. L. Shaffer, J. R. Sauer, and B.
G. Peterjohn. 1994. Conservation Reserve Pro-
gram: benefit for grassland birds in the noithern
plains. Trans. Am. Wildl. Nat. Resour. Conf. 59:
328-336.

Riemer, G., T. Harrison, L. H.all, and N. Lynn. 1997.
The native prairie stewardship program. Pp. 111-
1 16 in Caring for the home place: protected areas
and landscape ecology (P. Jonker, J. Vandall, L.
Ba.schak, and D. Gauthier. Eds.). Univ. Extension
Press. Univ. of Saskatchewan, Saskatoon.

Rotenberry, j. T. and j. A. Wiens. 1980. Habitat
structure, patchiness, and avian communities in
North American steppe vegetation: a multivariate
analysis. Ecology 61:1228-1250.

Ryder, R. A. 1980. Effects of grazing on bird habitat.
Pp 51-66 in Management of western forests and
grasslands for nongame birds (R. M. DeGrafl and
N. G. Tilghman. Eds.). U.S. For. Serv. Gen. Tech.
Rep., Ogden, Utah.

Samson, E B. and E L. Knopf. 1994. Prairie conser-
vation in North America. Bioscience 44:418-421.

SAS Institute Inc. 1989. SAS/STAT user's guide. Re-
lease 6.12 ed. SAS Inst., Inc., Cary, North Caro-
lina.

Sauer, J. R., J. E. Hines, G. Gough, I. Thomas, and
B. G. Peterjohn. 1997. The North American
breeding bird survey results and analysis. Version
96.4. Patuxent Wildlife Research Center, Laurel,
Maryland. URL = http://www.mbr.nbs.gov/bbs/
bbs.html

Smith, A. R. 1996. Atlas of Saskatchewan birds. Sas-
katchewan Natural History Society, Regina, Sas-
katchewan.

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

Stewart, R. E. and H. A. Kantrud. 1972. Population
estimates of breeding birds in North Dakota. Auk
89:766-788.

Sutter, G. C. 1996. Habitat selection and prairie
drought in relation to grassland bird community
structure and the nesting ecology of Sprague s
Pipit, Anthii.s spragueii. Ph.D. diss., Univ. of Re-
gina, Regina, Saskatchewan.

Sutter, G. C. and M. Brigham. 1998. Avifaunal and
habitat changes resulting from conversion of na-
tive prairie to crested wheat grass: patterns at
songbird community and species levels. Can. J.
Zool. 76:869-875.

Sutter, G. C., S. K. Davis, and D. C. Duncan. In
press. Grassland songbird abundance along roads
and trails in southern Saskatchewan. J. Field Or-
nithol.

Van Horne, B. 1983. Density as a misleading indi-
cator of habitat quality. J. Wildl. Manage. 47:893-
901.

Vickery, P. D., M. L. Hunter, Jr., and J. V. Wells.
1992. Use of a new reproductive index to evaluate
relationship between habitat quality and breeding
success. Auk 109:697—705.

Whitmore. R. C. 1979. Short-term changes in vege-
tation structure and its effects on Grasshopper
SpaiTows in West Virginia. Auk 96:621—625.

Wiens, J. A. 1969. An approach to the study of eco-
logical relationships among grassland birds. Or-
nithol. Monogr. 8:1-93.

Wiens, J. A. and M. I. Dyer. 1975. Rangeland avi-
faunas: their composition, energetics, and role in
the ecosystem. Pp. 146—182 in Symposium on
management of forest and range habitats for non-
game birds (D. R. Smith, Ed.). U.S. For. Serv.
Gen. Tech. Rep. WO-L, Tucson, Arizona.

Wilson. S. D. and J. W. Belcher. 1989. Plant and
bird communities of native prairie and introduced
Eurasian vegetation in Manitoba, Canada. Con-
serv. Biol. 3:39—44.

Winter, M. 1994. Habitat selection of Baird's Spar-
rows in northern mixed-grass prairie. Diplomar-
beit, Univ. of Tubingen, Tubingen, Germany.

Wilson Bull.. 1 1 1(3), 1999, pp. 397-414

BIRD COMMUNITIES IN NATURAL FOREST PATCHES IN

SOUTHERN BRAZIL

LUIZ DOS ANJOS' 3 AND ROBERTO BO^ON^

ABSTRACT. — Avifaimal composition was evaluated for natural (not artificial) patches of mixed temperate
rain forest in the Campos Gerais region, Parana State, southern Brazil. A large patch (840 ha) and 1 1 smaller
patches (0.5-40 ha) were censused from September to December of 1995 (five hours per month, each site). The
total species number was strongly conelated with patch size (r = 0.92, P < 0.001). However, the number of
edge species increased with decreasing patch area; the opposite happened with forest species. Thus, the ratio of
edge to forest species increased with decreasing patch area. The number of leaf insectivore species decreased
the most with a decrease in area. The mean Simpson similarity index was 73.8% among forest patches of similar
size. Smaller forest patches linked to the 840 ha patch were more similar to this larger patch than isolated
patches. Point counts from January to December 1991 in four patches (72 points each area) showed that several
species, specially trunk (and twig) insectivores and omnivores, increa.sed in relative abundance with decreases
in area of the patch (density compensation). The “habitat appropriation” hypothesis, the expansion of niches to
include slightly different habitats, could explain the increased relative abundance of two trunk (and twig) insec-
tivores: Cranioleuca ohsoleta and Cranioleuco pallida. Received 15 Oct. 1998, accepted 19 Feb. 1999.

Forest fragmentation in the Neotropical re-
gion has been considered an important force
in the loss of biodiversity (Bierregaard and
Lovejoy 1989). Decreases in the number of
bird species and changing avifaunal compo-
sition have been documented by many work-
ers (Willis 1974, 1979; Karr 1982; Bierre-
gaard 1990; Anjos 1992; Aleixo and Vielliard
1995).

Density compensation (increased relative
abundance of species in biologically isolated
habitats) is another frequent feature in habitat
fragments (Mac Arthur et al. 1972, Wright
1980). Reduced competition, habitat differ-
ences, and differences in colonization (Rick-
lefs and Cox 1978, Blondel 1991) also have
been suggested as possible reasons for density
compensation in habitat fragments. Blondel
and coworkers (1988) presented the “habitat
appropriation” hypothesis in which popula-
tions increase on islands because of expansion
into additional habitats.

Studies of birds in Neotropical forest frag-
ments have been carried out after fragmenta-
tion took place through logging. The state of
Parana, southern Brazil, has a grassland re-

‘ Univ. Estadual de Londrina, Depto. de Biol. Ani-
mal e Vegetal, Caixa Postal 6001, Londrina 86051-
970, Parana, Brazil; E-mail: llanjo.s@sercomtel.com.br
^ Socied. de Pesquisa em Vida Selvagem e Educagao
Ambiental, Rua Gutemberg 345, Curitiba 80420-030,
Parana, Brazil.

Corresponding author.

gion called “Campos Gerais” with isolated
forest patches of 0.1-100 ha (Fig. 1). This
landscape is natural; it was not deforested by
humans. The forest patches appear in areas of
suitable soil conditions surrounded by grass-
land and are slowly increasing in area through
a natural ecological succession (Klein 1960,
1972). The forest structure among different
patch series is similar (Klein 1960, Maack
1981, Klein and Hatschbach 1971). Because
fragmentation has been natural, the biological
processes may not have been affected by hu-
man activity. Our goal in this study was to
analyze the differences in the composition of
the bird community among various-sized for-
est patches in Campos Gerais region.

STUDY AREA AND METHODS

Study sites. — The study areas are the Fazenda Santa
Rita (25° 15' S, 49° 48' W) and Vila Velha State Park
(25° 15' S, 49° 55' W), Parana State, southern Brazil,
in a region called Campos Gerais (Fig. 2). The patches
of forest are mixed temperate rainforest. The dominant
trees are: Araucaria angustifolia ( Araucariaceae), Po-
docarpus lainhertii (Podocaipaceae), Sehastiana coni-
tnersoniana (Euphorbiaceae), Ocotea porosa, and Nec-
tandra grandijiora (both Lauraceae; Klein and Hatsch-
bach 1971). The average annual temperature is
17.3° C, varying from 20.9° C during summer (Decem-
ber-February) and 14.I°C during winter (June-Au-
gust). The range of average annual precipitation is
l5()0-200() mm. Elevations range from 950-1100 m
a.s.l. (Maack 1981).

Eleven forest patches (called B, C, D, E, F, G, H, I,
J, L and M) from 0.5-40 ha in size and one large forest
patch of 840 ha (called A) were censused. All these

397

398

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

LIG. 1. Landscape in the “Campos Gerais” region, southern Brazil; grassland with naturally isolated forest
patches.

sites are covered by similar forest structure. Some
patches are linked (patches B, C, D, and E) to the
largest forest patch (A) by forest corridors while others
are completely isolated by grassland (Table 1 ).

Species composition. — Bird species in each patch,
micro-habitat, and feeding habit were recorded during
five-hour monthly visits in the morning (20 h total)
September-December, 1995. Each species was classi-
fied as to its typical habitat and feeding habit through
field observations and references (Eitzpatrick 1980;
Belton 1984, 1985; Sick 1997; Ridgely and Tudor
1989, 1994). The habitats were categorized as open
area, forest border (up to 5 m from the edge), and
forest interior. We divided the forest into three levels;
(1) understory (below 2 m), (2) mid-levels (2-7 m)
and (3) canopy (above 7 m). Feeding habits were cat-
egorized as; omnivore, insectivore, Irugivore, carni-
vore or nectarivore. For insectivores, the area where
the insect or its larvae was most often captured was
also recorded. Three classes of capture sites were de-
fined: (1) trunk (and twigs), (2) leaves, and (3) gen-
eralized insectivores.

Relative abundance. — Relative abundance was de-
termined by monthly point counts of unlimited dis-
tance (Blondel et al. 1970) January-December in
1991. Fifty-three points were used to calculate relative
abundance: 24 in A, 12 in B, 10 in C. and 7 in G.
Each point was 100 m from another and at least 50 m
from the edge of the forest. Each month six points in
each site were chosen at random to be sampled giving

a total of 72 counts per site during the year (288 point
counts total in the four sites). The relative abundance
of each species was determined by dividing the total
number of species contacts by the total number of
points (72) sampled in each patch (Blondel et al.
1970). Sampling began at dawn at the first randomly
selected point and finished about 3.5 h later at the sixth
point. The time for sampling at each point was 20 min.
Snecies were identified primarily by sound (99%). The
same observer (LdA) performed all counts. Each pair
or flock of each species was counted once (one con-
tact) while vocalizing. Precautions were taken not to
count the same individual or group more than once (a
form was used in order to locate the counted individ-
uals), especially highly mobile species. Bird recordings
(1601 recordings of 414 species, deposited in the Bio-
acoustic Laboratory of the Universidade Estadual de
Londrina) were used to aid identification. The observer
had six years (1984-1990) of field experience in spe-
cies identification in the region (Anjos 1992, Anjos
and Graf 1993, Anjos et al. 1997).

Atudxses. — The birds were identified mostly using
the taxonomy of Meyer de Schauensee (1982) and
Sick (1997). Similarities in bird species composition
between forest patches were determined using the
Simpson Index. 3 diversity (Whittaker 1960) was used
to measure the degree of turnover in species compo-
sition along the six points sampled monthly (in A and
G). This index is a measure of how different the sam-
ples were, in terms of the variety of species. Differ-

Anjos ami Bo<^on • BIRD COMMUNITIES IN FOREST PATCHES

399

FIG. 2. Study area in “Campos Gerais” region, Parana State, southern Brazil, indicated by hatching.

ences in numbers of species or numbers of contacts of
species between the sites were tested with X" analysis
at a < 0.05. Analysis of Variance (one way ANOVA)
and the Tukey multiple range test (P < 0.05) were used
to evaluate the average number of species and contacts
recorded per sampled point in A, B, C and G. The
relationship between number of bird species and size
of fragment was expressed through a transformation
following Preston (1962): log^ = log^ + zlogA where

TABLE 1. Size (ha) and on the distance (m) from
a continuous forest of each study site.

Site

Size

Type-Di.stance

840 ha

continuous forest

40 ha

linked by forest corridor- 100 m

20 ha

linked by forest corridor-200 m

12 ha

linked by forest corridor-600 m

10.5 ha

linked by forest corridor-4()0 m

10 ha

isolated-3000 m

9 ha

isolated-2000 m

8.5 ha

isolated-800 m

6.5 ha

isolated-2500 m

4 ha

isolated- 1500 m

1.5 ha

isolated-500 m

0.5 ha

isolated- 1000 m

S is the number of species and A is the size of the area.
SAS/STAT (version 6.11, IBM, mainframe) was used
to calculate most of the above values.

RESULTS

Habitat distribution. — Birds in all patches
totaled 189 species, including 13 open area
species, 51 edge species, and 125 forest spe-
cies (Appendix). The open area birds used the
forest only for roosting and/or for nesting but
not for foraging.

There were 13 edge species in the largest
patch (A) and 32 species in the 6.5 ha frag-
ment (1). By contrast, there were 93 forest spe-
cies in the largest patch and only 13 species
in the smallest fragment (M; Table 2). These
results indicate a general decrease in the ratio
of forest to edge species with a decrease in
area (Fig. 3). The exceptions are the patches
E (34/21 species), G (41/28 species), and L
(32/19 species). The number of understory
forest species decreased faster {y} = 15.37, df
= 2, P < 0.05) with decrease in area of forest
(33 species in A, 20 species in B, and 8 spe-

400

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

TABLE 2. Number of bird species in the study sites accordin;

g their habitats and feeding

habits.

Sites

Habitats

Open area

Edae

Forest/understory

Forest/mid-levels

Forest/canopy

Feeding Habits

Omnivores

Insectivores/trunk

Insectivores/leaves

1 1

Insectivores/general

1 1

Frugivores

Carnivores

Nectarivores

Total Species

108

cies in C) than mid-level forest species (32
species in A, 29 species in B, and 27 species
in C; X- = 0-43, df = 2, F > 0.05); canopy
forest species were intermediate in the ratio of
decreasing (28 species in A, 19 species in B,
and 16 species in C; “ 3.71, df = 2, P >
0.05; Table 2).

Similarity between sites. — The similarity
index (Simpson) between the largest site (A:
108 species) and the smallest (M; 45 species)
was 40% (17 species common to both sites);
however, the mean was 73.8 ± 2.7% (SE)
among forest patches of similar size.

The number of species was correlated with
area (Fig. 4; r = 0.92, df = 10, P < 0.001).
Species living in open areas were not consid-

FIG. 3. Ratio of forest (solid) and edge (hatched)
species in the study sites. Sites B-E linked to A by
forest corridors.

ered in this analysis; edge species were in-
cluded with forest species because the major-
ity of them were commonly found within the
forest. The number of species can be predicted
according to the Preston (1962) model; \ogS
= 3.81 + 0.17 logA, where S is the number
of species and A is the fragmented area in ha.

Feeding habits. — Of the 189 species re-
corded, 85 were insectivores, 58 omnivores,
28 frugivores, 12 carnivores, and 6 nectari-
vores. Among the insectivores, 33 were leaf
insectivores, 3 1 generalized insectivores (cap-
ture insects in various ways), and 21 trunk
insectivores.

Insectivores were less abundant in B (40 ha;

FIG. 4. Correlation (log/log) between number of
species (S) and area (A) of fragment. Dotted lines in-
dicate 95% confidence interval.

Anjos am! lio^on • BIRD COMMUNITIES IN FOREST PATCHES

401

42 species) than in A (840 ha; 59 species)
although not significantly so (x" = 2.86, df =
I, P > 0.05). In patches smaller than 40 ha
the rate of decrease slowed as patch size di-
minished (36 species/20 ha, 33 species/12 ha,
28 species/10.5 ha); in patches smaller than
10.5 ha, there was no clear pattern (Table 2).
The greatest reduction in species occurred in
leaf insectivores, from 30 species in the forest
(A) to 16 species in the largest patch (B; x“
= 4.26, df = 1, P < 0.05); the loss of trunk
insectivores was not so great (18 species in A
and 14 species in B; x‘ = 0.5, df = 1, P >
0.05) while generalized insectivores gained
one species (11 species in A and 12 species
in B).

Omnivores were represented by two more
species in B (32 species) than in A (30 spe-
cies). Their numbers remained relatively con-
stant down through the 12 ha patch (D; 31
species) but decreased in smaller patches (Ta-
ble 2).

Frugivores decreased slightly in number (x^
= 0.46, df = 2, P < 0.05) from A (15 species)
to B and C (both with 12 species). The num-
ber of frugivores stayed between 5 and 8 spe-
cies in patches smaller than 12 ha (Table 2).

Carnivores and nectarivores were repre-
sented by few species in all sites (1-4 species;
Table 2).

Number of species and contacts. — In point
counts conducted in 1991 we recorded 138
species in A, 125 in B, 103 in C, and 91 in
G. The average numbers of species per sam-
pled point were; 23.6 in A, 16.5 in B, 18.5 in
C, and 21.4 in G. The number of species per
sampled point was not significantly different
(Tukey test: P > 0.05) between A and G al-
though they are different in the size and total
number of species recorded. The intermediate-
sized patches B (40 ha) and C (20 ha) did not
show significant differences (Tukey test: P >
0.05) among the numbers of species per sam-
pled point, but both were significantly differ-
ent from A and G (Tukey test: P < 0.05).

The average numbers of contacts per sam-
pled point were: 27.0 in A, 19.1 in B, 22.1 in
C, and 25.8 in G. The numbers of contacts per
point were not significantly different between
A and G nor between B and C; but they were
significantly different between A-B, A-C, G-
B and G-C (Tukey test: P < 0.05).

There were variations in the average num-

15
10

FIG. 5. Average number of contacts in the studied
sites during the year. Square = A, triangle = B, circle
= C, diamond = G.

ber of contacts during the year in the four
study sites. The number of contacts was high-
est in September/October— January than Feb-
ruary-August (Fig. 5).

Relative abundance. — The highest relative
abundance (number of contacts divided by the
number of sampled points) in this study was
Basileuterus leucoblepharus (1.2) followed by
Turdus rufiventris (0.85), Lepidocolaptes
squamatus (0.68), Basileuterus culicivorus
(0.68), Cyclarhis gujanensis (0.67), and Cran-
ioleuca obsoleta (0.61).

The majority of species recorded in the
point counts presented significant differences
in the values of relative abundance when oc-
curring in three or four sampled sites (indi-
cated with “s” in Appendix; df = 2 or 3, P

< 0.05). Some of these species increased in
abundance with the decrease in area (density
compensation) such as Veniliornis spilogaster,
Cranioleuca obsoleta, Turdus amaurochali-
nus, Cyclarhis gujanensis. Panda pitiayumi,
Thraupis sayaca, and Stephanophorus diade-
matus. The distribution of the abundances in
a rank order showed that the 12 species with
highest abundances (indicated with an * in
Appendix) in the smallest patch analyzed (G)
comprised 27% of total contacts but they com-
prised only 20% in the continuous forest (A).

Relative abundance increased with decreas-
ing area for 12 omnivorous species (Appen-
dix); Penelope obscura (x“ = 9.12, df = 3, P

< 0.05), Pachyramphus polychopterus (x^ =
73.87, df = 3, P < 0.001), Pitangus sulphur-
atus (x' = 106.39, df = 3, P < 0.001), Elaen-

402

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

ia mesoleuca (x“ = 23.38, df = 3, P < 0.001),
Cyanocorax caeruleus (x" = 21.08, df = 3, P

< 0.001), Turdus rufiventris (x‘ = 8.52, df =

3, P < 0.05), Turdus amaurochalinus (x' =
23.9, df = 3, P < 0.001), Cyclarhis gujanen-
sis (x- = 36.19, df = 3, P < 0.001), Stephan-
ophorus diadematus (x“ — 12.92, df = 3, P <
0.01), Thraupis sayaca (x“ = 19.8, df = 3, P

< 0.001), Zonotrichia capensis (x^ = 33.64,

df = 3, P < 0.001), and Carduelis magellan-
icus (x" = 14.07, df = 3, P < 0.01). The sum
of relative abundances of the omnivorous spe-
cies increased (x“ 28.41, df = 2, P < 0.001)

from A (6.56) to the patches C (7.52) and G
(9.0).

The sum of relative abundances of leaf in-
sectivorous species was highest (x^ = 77.15,
df = 3, P < 0.001) in A (8.42), followed by
G (6.18), C (5.56), and B (5.14). Parula pi-
tiayumi (x“ = 12.86, df = 3, P < 0.01) in-
creased in relative abundance with decreases
in area (Appendix).

The sum of relative abundances of species
of trunk insectivores increased (x“ = 10.8, df
= 1, P < 0.01) from A (4.06) to G (5.24) as
the result of a substantial increase in relative
abundance of two species: Cranioleuca ob-
soleta (0.21 in A and 1.54 in G) and C. pallida
(0.14 in A and 0.93 in G).

Relative abundances of Furnarius rufus (x^
= 49.89, df = 2, P < 0.001), Tyrannus rne-
lancholicus (x^ = 23, df = 3, P < O.OOL), and
Phyllomyias fasciatus (x^ = 39.72, df = 3, P
< 0.001) generalized insectivores, increased
with decreases in area (Appendix). The sum
of the relative abundances of this group was
higher (x^ = 8.43, df = 1, P < 0.01) in G
(3.82) than A (2.93).

The sum of the relative abundances of fru-
givores decreased quickly (x' “ 176.93, df —
3^ p < 0.001) with decreases in area: A, 4.18;
B, 2.89; C, 1.76, and G, 0.92. Only Leptotila
verreauxi (x^ = 20.43, df = 3, P < 0.001 ) and
L. rufaxilla (x^ ~ 11.96, df = 3, P < 0.01)
increased in relative abundance.

DISCUSSION

Number of species. — Based on censuses
from a few visits to many islands in the spe-
cies/area studies, Haila and Jiirvinen (1981:
561) suggested that the “gain by increasing
the number of visits to an island is the im-
proved accuracy in the recording of rare spe-

cies.” Almost 90% of the species in a site are
recorded during a single visit (Haila and Ku-
usela 1982). Taking into account the point
counts of one year carried out in the present
study, the combined samplings in A, B, C, and
G from September to December (four visits)
revealed 73—78% of the total species in these
sites; this is less than the 90% predicted by
Haila and Kuusela (1982). The fewer species
we recorded in four visits is problably due to
a higher proportion of rare species in our
study sites. Therefore researchers in the Neo-
tropical region, especially in forests, should
make several visits to a site instead of a single
visit.

There was a strong correlation (r = 0.92,
df = 10, P < 0.001) between the number of
bird species and the size of the forest patches
in the Campos Gerais region. Other effects,
such as habitat heterogeneity, might be more
important than area in predicting the number
of species found on an island (Martin et al.
1995). In more homogeneous sites it is ex-
pected that the effect of area would increase
(Connor and McCoy 1979). Galli and co-
workers (1976) studied woodlots in New Jer-
sey with similar foliage height diversity and
found a strong correlation between the num-
ber of species and size of the area. Martin
(1981) studied homogeneous shelterbelts in
South Dakota and concluded that area was the
most important factor in determining the num-
ber of species. The similarity in vegetation
structure between Campos Gerais forest
patches is due to their being at approximately
the same stage of ecological succession (Klein
1960, 1972; Klein and Hatschbach 1971;
Maack 1981), and probably explains the high
species/area correlation.

Based on the study by Tomialojc and co-
workers (1984) in Poland, Blondel (1986)
suggested that the habitats in large forests ap-
pear to be broken into a mosaic of “sink” and
“source” because a large forest is represented
by a mosaic of habitat patches. This results in
a potential heterogeneity of the distribution of
birds in large forests. In order to test the
source/sink (Blondel 1986) role of larger for-
ests, we calculated the ^-diversity (Whittaker
1960) among the six points sampled monthly,
in the 9 ha patch (G) and in the largest patch
(A, 840 ha). The mean (3-diversity was higher
in A (1.68) than in G (1.27; Mann-Whitney

Anjos am! lioi-on • BIRD COMMUNI TIES IN FOREST PATCHES

403

t/-test: U = 12.5, P < 0.05), which suggests
a more homogeneous bird distribution in the
smaller area. The average number of species
per point sampled was not significantly dif-
ferent (Tukey test: P > 0.05) between G and
A, which suggests that the higher number of
total species in the largest patch is mainly due
to the heterogeneity of species’ distributions
compared to smaller patch.

It was expected that there would be more
species in A (138 species, 840 ha) than in G
(91 species, 9 ha) because of the difference in
size of the fragments. The small difference
can be explained by the heterogeneity of spe-
cies distribution in A; the area sampled in A
(24 sampled points, which represents around
80 ha) corresponds to about 10% of the total
site, but the total area of G (9 ha) was sam-
pled. Two hundred and seven species were re-
corded in A in a parallel study (Anjos 1992).
Thus, the heterogeneity in bird distribution re-
sulted in Just 67% of the total avifauna in A
being recorded in this study.

Composition of the bird communities. — In
the Neotropics, as in the rest of the world,
losses of species and changes in avian com-
munity composition occur with fragmentation
of forest areas. In the Campos Gerais, the
scarcity of forest species are, in part, balanced
by gains in edge species (Fig. 3). Willis
(1979) pointed out that the woodlots of Sao
Paulo, southeastern Brazil, have become more
like temperate zone forests with an emphasis
on oscine birds characteristic of canopy and
edge. Bierregaard and Lovejoy (1988) noted
some edge species {Troglodytes aedon and
Ramphocelus carbo) wandering into small
forest patches (1 ha) after experimental defor-
estation in Amazonia. In the Campos Gerais
region a gradual increase in the ratio of edge
species to forest species was obvious with a
decrease in fragment area. The majority of
these edge species are oscines, some of them
living in the canopy.

The proportion of forest/edge species does
not seem to be related only to the size of the
area. MacArthur and Wilson (1963, 1967)
demonstrated that isolation is also an impor-
tant factor. Among the isolated patches, L (1 .5
ha) possessed a higher proportion of forest
species than F (10 ha), H (8.5 ha), I (6.5 ha),
or J (4 ha) probably because it was closer to
the largest patch (Fig. 3, Table 1).

Among lorest birds, the principal decrea.se
in number ol species in the Campos Gerais
patches was associated with the loss of un-
derstory species. Canopy species probably fly
to a nearby patch when there is a shortage of
resources. Bierregaard and Lovejoy (1989)
verified a decrease in the number of species
also occurring among understory birds in
Amazonia.

Aleixo and Vielliard (1995) pointed out that
leaf insectivores of the understory were those
most likely to be absent in a woodlot of 251
ha in Sao Paulo. We also found that most of
the understory species that were absent from
the small patches in the Campos Gerais region
were leaf insectivores. Trunk insectivores
were rarely found in small woodlots by Willis
(1979); these were mostly small birds, such as
Picumnus spp. (12 g) and Veniliornis spilo-
gaster (43 g). Large trunk insectivores, such
as Phloeoceastes robustus (263 g) and Dry-
ocopus lineatus (246 g), were recorded only
in the forest of our study. Only small or mid-
size species such as Cranio I euca obsoleta (14
g), C. pallida (12 g), Lepidocolaptes squa-
matus (28 g), and Veniliornis spilogaster were
recorded in the smallest patches (Appendix).

Frugivores were rare in small patches of
forest of Amazonia (Bierregaard and Lovejoy
1989) and Sao Paulo (Willis 1979, Aleixo and
Vielliard 1995). Willis (1979) thought that be-
cause frugivores can easily travel to other
woodlots to look for available resources they
probably disappear from the small woodlots;
they depend on scattered trees of different
species at different seasons or years, and only
large woodlots have enough tree diversity to
support them. Frugivores occurred in all the
patches we studied in the Campos Gerais re-
gion. This ability to easily move between
patches may have prevented their numbers
from declining (although the relative abun-
dance of frugivores in smaller patches was
low, as shown above).

Omnivores, which can switch from fruit to
insects or vice versa, may benefit from small
woodlots of Sao Paulo (Willis 1979). This
group was represented by the greatest number
of species in the Campos Gerais and tended
to increase its contribution to the avifaunal
composition with decreases in area; the sum
of relative abundance increased from A (6.56)
to G (9.0).

404

THE WILSON BULLETIN • Vol. 1 1 1, No. 3. September 1999

The present composition of the avifauna in
the patches at the Campos Gerais region is the
result of a long period during which, presum-
ably, many factors have operated. Forest
patches resulting from human disturbance of
a continuous forest are isolated more rapidly.
The effects of isolation on natural and artifi-
cial forest fragments may be different. Ac-
cording to Willis (1979), frugivores seem to
be more sensitive to isolation in woodlots in
Sao Paulo than in those of the Campos Gerais.
But the woodlots studied by Willis (1979) are
larger and more isolated than the patches stud-
ied in Campos Gerais. Thus, a study carried
out under similar conditions is necessary for
a more precise comparative analysis.

Relative abundance. — Vielliard and Silva
(1990) and Aleixo and Vielliard (1995), using
the same census method of point counts in
Sao Paulo, Brazil, obtained a similar pattern
of monthly variation in number of contacts as
we did (Fig. 5). This is probably because Sep-
tember-December is the main period when all
the species are more obvious when they breed
and vocalize (Vielliard and Silva 1990, Aleixo
and Vielliard 1995). There are also summer
transients of some species in the study area
from September to March (Anjos and Graf
1993). Therefore, counts of relative abun-
dance are influenced by vocalization and mi-
gratory behavior. Because censuses were per-
formed monthly in each patch, seasonal vari-
ation was assumed to be the same for all sites,
making comparative analysis possible be-
tween the patches. In addition, because habitat
physiognomy was similar for all of our patch-
es, bias between the sites resulting from dif-
ferences in detection of vocalizations (Schieck
1997) was unlikely.

Density compensation was detected in the
present study for several species. For edge
species the greater relative abundance is prob-
ably due to the relatively greater habitat area
in the smaller patches. Edge species may also
be better colonizers of isolated patches than
forest species.

Decreased competition in smaller patches
could explain density compensation (Ricklefs
and Cox 1978). This seems to be the case for
two leaf insectivores {Thamnophilus caerules-
cen.s and Parula pitiayurni) and some trunk
insectivores {Veniliornis spUoga.ster, Lepido-
colaptes .squarnatu.s, Cranioleuca ohsoleta.

and C. pallida) in the present study. These
groups decreased in number of species as area
decreased, which could mean fewer competi-
tors. However, documenting that competition
occurs among species is not easy (Wiens
1989).

The “habitat appropriation” hypothesis of
Blondel and coworkers (1988; a population
size increases in an island because of an ex-
pansion of habitat occupation) was examined
with two trunk and twig insectivores. Crani-
oleuca obsoleta (x" “ 162.72, df = 3, F <
0.05) and C. pallida (x^ = 91.41, df = 3, F
< 0.05) increased in relative abundance with
decreases in area (Appendix). Points were se-
lected in A and G with either dense or open
understory. The relative abundance of the two
species was calculated in these two habitats
(dense and open understory). Cranioleuca ob-
soleta had a relative abundance of 0.33 in
open understory and 0.08 in dense understory
areas of the 840 ha patch (A) but 2.07 and
1.06, respectively, in the 9 ha patch (G). Cran-
ioleuca pallida was not observed in dense un-
derstory and had a relative abundance of 0.67
in open understory areas of A, and 0.43 and
1.29 respectively in G. This suggests that both
species increase the habitats they use (they
tended to be more abundant in areas of dense
understory in G). These results were consis-
tent with what could be interpreted as the
“habitat appropriation” hypothesis. This
should be considered as a complementary ex-
planation for density compensation. Crani-
oleuca obsoleta also was common with high
relative abundance in another natural patch of
forest in Curitiba city, Parana (Anjos and Lar-
oca 1989).

The causes of density compensation are not
clear but they may be different for each spe-
cies.

ACKNOWLEDGMENTS

We acknowledge Conselho Nacional de Pesquisa
(CNPq-Brasilia) for research grants (LdA-350054/95-
9). Financial support for the field study was provided
by the Universidade Estadual de Londrina, Londrina,
the Consorcio Intermunicipal para Prote9ao Antbiental
do rio Tibagi. Londrina, and the Klabin Fabricadora
de Papel e Celulose, Tclemaco Borba. We thank T.
Matsuo for statistical analysis assistance and M. M. F.
Costa and A. H. F. de Toledo for English revision. M.
C. Dias and M. Torrezan helped us with the vegetation
description. We appreciated the assistance of E. A.

Anjos ami Boi^on • BIRD COMMUNITIES IN FOREST PATCHES

405

Lima, H. D. Garcia, and V. L. Ogassawara using SAS/

STAT J. C. V. Lopes kindly permitted us to work at

the Santa Rita farm. Finally we sincerely thank W.

Belton, J. Bates, E. O. Willis and Y. Oniki for com-
ments and criticisms that improved the manuscript and

J. M. E. Vielliard for suggestions during the held work.

LITERATURE CITED

Aleixo, a. and j. M. E. Vielliard. 1995. Composi^ao
e dinamica da avifauna da mata de Santa Genebra,
Campinas, Sao Paulo, Brasil. Rev. Bras. Zool. 12;
493-5 1 1 .

Anjos, L. dos. 1992. Riqueza e abundancia de aves
em “ilhas” de floresta de araucaria. Ph.D. diss.,
Univ. Federal do Parana, Curitiba, Brazil.

Anjos, L. dos and S. Laroca. 1989. Abundancia re-
lativa e diversidade especifica em duas comuni-
dades urbanas de aves de Curitiba (sul do Brasil).
Arq. Biol. Tecnol. (Curitiba) 32:637-643.

Anjos, L. dos and V. Graf. 1993. Riqueza de aves da
fazenda Santa Rita, regiao dos Campos Gerais,
Palmeira, Parana, Brasil. Rev. Bras. Zool. 10:673—
693.

Anjos, L. dos, K.-L. Schuchmann, and R. Berndt.
1997. Avifaunal composition, species richness,
and status in the Tibagi River basin, Parana State,
southern Brazil. Omitol. Neotrop. 8:145—173.

Belton, W. 1984. Birds of Rio Grande do Sul, Brazil.
Part 1. Bull. Amer. Mus. Nat. Hist. 178:369-636.

Belton, W. 1985. Birds of Rio Grande do Sul, Brazil.
Part 2. Bull. Amer. Mus. Nat. Hist. 180:1-241.

Bierregaard, R. O., Jr. 1990. Avian communities in
the understory of the Amazonian forest fragments.
Pp. 333-343 in Biogeography and ecology of for-
est bird communities (A. Keast, Ed.). SPB Aca-
demic. The Hague, Netherlands.

Bierregaard, R. O., Jr. and T. E. Lovejoy. 1988.
Birds in Amazonian forest fragments: effects of
insularization. Proc. Ornithol. Congr. 19:1564-
1579.

Bierregaard, R. O., Jr. and T. E. Lovejoy. 1989. Ef-
fects of forest fragmentation on Amazonian un-
derstory bird communities. Acta Amazonica 19:
215-241.

Blondel, j. 1986. Biogeographie evolutive. Masson,
Paris, France.

Blondel, J. 1991. Birds in biological isolates. Pp. 45-
72 in Bird population studies (C. M. Perrins, J.-
D. Lebreton, and G. J. M. Hirons, Eds.). Oxford
Univ. Press, Oxford, U.K.

Blondel, J., C. Ferry, and B. Frochot. 1970. La
methode des indices ponctuels d’abondance
(I. PA.) ou des releves d’avifaune par “stations
d'ecoute”. Alauda 38:55-71.

Blondel, J., D. Chessel, and B. Frochot. 1988. Birds
impoverishment, niche expansion, and density in-
flation in mediterranean island habitats. Ecology
69:1899-1917.

Connor, E. F. and E. D. McCoy. 1979. The statistics
and biology of the species-area relationship. Am.
Nat. 113:791-833.

Fitzpatrick, J. W. 1980. Foraging behavior of Neo-
tropical tyrant flycatchers. Condor 82:43—57.

Galli, a. E., C. F. Lecj, and R. T. Forman. 1976.
Avian distribution patterns within sized forest is-
lands in central New Jersey. Auk 93:356-365.

Haila, Y. and O. Jarvinen. 1981. The underexploited
potential of bird censuses in insular ecology. Stud.
Avian Biol. 6:559-565.

Haila, Y. and S. Kuusela. 1982. Efficiency of one
visit censuses of birds breeding on small islands.
Ornis Scand. 13:17-24.

Karr, J. R. 1982. Avian extiction on Barro Colorado
Island, Panama; a reassessment. Am. Nat. 119:
220-239.

Klein, R. M. 1960. O aspecto dinamico do pinheiro
brasileiro. Sellowia 12:17-44.

Klein, R. M. 1972. As florestas da America do Sul.
Univ. de Brasilia, Brasilia, Brazil.

Klein, R. M. and G. Hatschback. 1971. Fitofision-
omia e notas complementares sobre a mapa fito-
geogrifico de Quero-quero (Parana). Bol. Par.
Geoc. 28: 159-188.

Leck, C. F. 1979. Avian extinctions in an isolated trop-
ical wet-forest preserve, Ecuador. Auk 96:343-
352.

Maack, R. 1981. Geografia ffsica do Estado do Para-
na. Livraria Jose Olympio, Rio de Janeiro, Brasil.

MacArthur, R. and E. O. Wilson. 1963. An equilib-
rium theory of insular zoogeography. Evolution
17:373-387.

MacArthur, R. and E. O. Wilson. 1967. The theory
of island bigeography. Princeton Univ. Press,
Princeton, New Jersey.

MacArthur, R., J. R. Kaar, and J. M. Diamond.
1972. Density compensation in island faunas.
Ecology 53:330-342.

Martin, J.-L., A. J. Gaston, and S. Hitier. 1995. The
effect of island size and isolation and old growth
forest habitat and bird diversity in Gwaii Haanas
(Queen Charlotte Islands, Canada). Oikos 72:1 15-
131.

Martin, T. E. 1981. Limitation in small habitat islands;
chance or competition. Auk 98:715-734.

Meyer de Schauensee, R. 1982. A guide to the birds
of South America. Academy Natural Science,
Philadelphia, Pennsylvania.

Preston, E W. 1962. The canonical distribution of
commoness and rarity. Ecology 43:185-215.

Ricklefs, R. E. and G. W. Cox. 1978. Stage of taxon
cycle, habitat distribution and population density
in the avifauna of the West Indies. Am. Nat. 1 12:
875-895.

Ridgely, R. S. and G. Tudor. 1989. The birds of
South America. Vol. 1. Oxford Univ. Press, Ox-
ford, U.K.

Ridgely, R. S. and G. Tudor. 1994. The birds of
South America. Vol. 2. Oxford Univ. Press, Ox-
ford, U.K.

SAS Institute, Inc. 1989. SAS/STAT® user's guide,
version 6.1 1, fourth ed. SAS Institute, Inc., Carey,
North Carolina.

406

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

SCHIECK, J. 1997. Biased detection of bird vocaliza-
tions affects comparisons of bird abundance
among forested habitats. Condor 99:179-190.

Sick, H. 1997. Omitologia brasileira. Nova Lronteira
ed., Rio de Janeiro, Brazil.

Tomialojc, L., T. Wesolowski, and W. Walankiew-
icz. 1984. Breeding bird community of a primae-
val temperate forest (Bialowieza National Park,
Poland). Acta Omitol. 20:241-310.

ViELLiARD, J. E. M. AND W. R. SiLVA. 1990. Nova me-
todologia de levantamento quantitativo da avifau-
na e primeiros resultados no interior do Estado de
Sao Paulo, Brasil. Pp. 117-151 in Anais do IV
Encontro de Anilhadores de Aves (S. Mendes,

Ed.). Univ. Federal Rural de Pernambuco Press,
Recife, Brazil.

Whittaker, R. H. 1960. Vegetation of the Siskiyou
Mountains, Oregon and California. Ecol. Monogr.
30:279-338.

Wiens, J. A. 1989. The ecology of bird communities.

Vol. 2. Cambridge Univ. Press, Cambridge, U.K.
Willis, E. O. 1974. Populations and local extinctions
of birds on Barro Colorado Island, Panama. Ecol.
Monogr. 44:153-169.

Willis, E. O. 1979. The composition of avian com-
munities in remanescent woodlots in southern
Brazil. Papeis Avulsos Zoologia 33:1-25.
Wright, S. J. 1980. Density compensation in island
avifaunas. Oecologia 45:385—389.

Aujos and Bo^on • BIRD COMMUNITIES IN FOREST PATCHES

(U w

3
U
CJ
O

-C

(u •£

-O o

X V

w CJ

■o
c

,(U -O

c O

OJD

•— c/2

c/2 (U

c/:

O 3
3 >■
O ^
L3 c/2
^ (D
C -3

c/2 .3

O'^

U D.

c/2

03 ^

c/2 CC

.3 O

C4 ^

^ c/2

■£ ^

C O

CJ .

■o

<72

“O

C .3
o c/2

H >-

- •£
Z o

“ o
c 1/
c a.

-c

S Cj ftC CO

9 S
£ ^
^ 3

7 C

g S

2 =

in -2

QC r3

X ^
H

<< <<< <<S2<- S

uu uuu uuoouo o

<< <Q< UUQQ<< S

OO OUO uih-WWOO if

o o
ci. ^
^ C
2f 3

U U

■f

^ 5

Rc- S

^ 5

•3 o

"S ^

c I

'3 "S.

—

00 (N ro

VO 00

CN CN 00

>0 (O

O (N O

o —

d d d

d d

r^j

^T)

ON CD

o —

d d

VD VD (N

— . — c ro

O O O

d d d

— CO

— ■ 00

— o

d d

c C

5 i: i; <<

w o

9 2

z S-

< 3

CO 5

< ^

407

APPENDIX. CONTINUED.

408

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

UQUQUQQPDD

mutLLija,tLiwti.u-p-

-1- n

— — r)-

o o o

un 00

o o
d d

Qi Di Oi Qi Oi Qi Qi
u- U, U- U- U- tin u-

u Q u u u u u

tu UJ tU Uh pH tl. tt,

—

'd-

0^ 00 ^
m <N —
— o o
d d d

OO ^ 00

— cN m — ' n

o o o p p

d d d d d

E CL

-c

-j

'S.

ii 3 ~

.S- § ^

s S: ^

^ S'

s ^ ^

^ s

UPO'oONO-JhjO

C/1

b d 5 c

2i -H 3 3

S- .2 -i; ^ d

-C d Cl. .5

C: - „ i- ~

~ .to U 3i

1-^

i; d -C: C 2

^ ^ ^ 5
s c ^ 2

^ ^ S c 5

^ k. -S -S s

k, CQ a, d. rt;

Cl,

S S S S
o o o o

S S q q

Ph tu PJ PJ

< < <
u u u

u Q Q

P- PJ w

r<'i

o o

r*^

d d

^ —

■tt

o —

—

d d

5 t

U ^ ^

c <

I s

= 2

^ H

-c

■ s:
c

-c r;;

H. -s:

<■ .c

5 s

5 'C
O

APPENDIX. CONTINUED.

Anjos and Bo^on • BIRD COMMUNITIES IN FOREST PATCHES

409

PQUJWUPJW SS (V

Z2ZZZZ OO £

QQQpS SS U

upjwUhE Euh [i,

QuSuuQSu

WPuEPhPuPJUhPh

H H H H H H H

S S S
u. u.

.— (

r-H

'Cf

IT)

.— •

C^J

— -

— H

Tj-

C^l

>ri

C^l

1-H

rz o

01
H

':3 ^

-C5

-s: o
0. U

a
oo ^

•§l

^ -S S

Co G F. ^

s: c
Oc Cic

c c

ti; f;

Ct'

i*.

•5

-2 ■§

"S S

-O K

S: *>M

>- ^
3 3

S E

3 3

;j ;j

^-2

>1
3

3
>)

S3 >.

3 I

i2 2

3 :
3

So So

3 3 -E
U U O.

3^3
3 3 0
3 3 0
O ?

Q 5 ^

s: c

H
Co
<

J
O
U
O
DC
Q

W Q
Q

<4)

c\ C t.“

a -ii 'c
■2 :§ a.

•C ^ 2

^ 2 ia,

S--S

s ^ a

C C Cj

2 i g

a S

.s:^G

CO >< Q

3
Oc s

S' cS

CO c*-.

Cn Co to
^ ^

S- S' ^
2 -2 “2
c "2 "5

2i ?= 2s

^ ^ ^

APPENDIX. CONTINUED.

410

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

PL,[UP-U,t)HpHU,U.P-UpHtl-P-U-pHtUP-pH

s s

O O J J J

j J o o o o

pdSSSdddppp

uuP-Puu-ti-P-.P-P-u,ti-u.

ON NO
O fN

d d

* *
— ON
— ^

\0 — 00
d d d

O'! NO 00

o o o
d d d

* *
r) — '

■Nt ro NO >£N
iri ON (N

o o o

— 00 Nt NO
00 ol ^ ro

— IT) (N

d d d d

■*■

— ri

00 cn

(N in (N 00

— (N On) (N

^ O — (N P

d d d d d

oominincNcSNOoo

m m
O 00
^ o
d d

(N
NO ^
NO O

d d

in NO

o —
d d

(n tn

o o

o o c

'll-

— 't:}- — csr^-^oocn
t^t^Nomr'NO'^rNi
;pp

dddddddd

O On
00 m

NO —

d d

m 00 On
mom
mm—'
d d d

— -minoomON^m
minmoooNOONNO

m (N
d d d

d o
d d d

^ 00 (N in —

m ON m

NO in m m
m NO — ■ m

o o o o o o o

c; No ^

^ tjs ->
^ ^ i;
d 2

F 5 5
*»■ —*

:S ^ t.

C d

£3.

5 c

c c

!3.

-J

^ S lo
2
i;

= c
-cs >■.

^ ^c.0

^ :::

5 §

PS-

^ s ^

'nJ 1

2 i

'nJ ^

5 Nc 3;
a ? £3

C 33

C
No j:;

£3 £3
K? li

^ £u k;
53 3:^

K C Cn.

NnJ ^ t*"* •>* -.C

^ £3 S d S .2f

•<S!;5;i<l<£n>5 QCQ^iSf^QQft-GCdO

APPENDIX. CONTINUED.

Atijos and Bo(;(m • BIRD COMMUNITIES IN FOREST PATCHES

411

S S S 2 QJ 05

o o o o £ £

c «
c o.

>• ^
CL, C

^ Ei.

5 S
-c:

&• &•

I I
^ &■
"5 "C

a <3

Q, Q,

^ I

a £:■

2 ^

2 2

;>>

3! ^
EJ Q

o: oi

U, Ph

oaooaooaooojoojaooaoj

^U^<<QQQQQQuuQQSppQ<p

OiiOOOtuwujWpQpjupppwpjEEEwOEI-

'tt

t»

— H

c:^

•o

1-^

>-H

1—

ID m

so O

— H

d d

■Nt

—

-H

<L>

. ^ s

Q- Q,

■S- g

W § I

I -~

Z C
< .5 -S

^ ? r2

>H >< tJ

to S
-3

C V
-C tp

cx Oc

^ '5

to to
3 3
oc Oc

-2i -Si
"c "S
A .£.

Xo 5^

.to

■S a
=: "s
"3 ^

po. J)
to 3

=3 b
S>c EX

tJJ

.£.

■ I- ^

^ §

Cp -2

s ::

^ .'o

■S S
^ i

IX. tp

5 "5

-s;

3 c .:i

5 5 ^

'S.

6C ■

5 5

Co •:2
2 §
3 to

b .*3

-5 I

A ^

to to

C 3
2 -p

£ 5

3 to
S! 2i t3
EU 5 P
.C Eu

3 P; 3

» eS -2
2 § -

2 I- S
,0 S c
U UJ u

3 3
tL> ^
3 'J

■5b S

g 5

35 S

3 -C
C 'J
;3 c

Cp c E-
.J 3
.b 3!

^ £

APPENDIX. CONTINUED.

412

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

J J J

JJhJJJJOOOOhJOOJJPh

UDDDDSQSQQQQQUUUSPD

5 o-

-c s:

•2 2

s: ^

•S- o
'c

5 =s

^ C3

-Ci

Ci. o

o -3

K K

0, O,

C/5

■Si

-3

-2 -2
kj kj

•Si

-S

k»

-3

•S

■S

S. "s
■§

3 3

kO kj

5
o

oo ^

<3 2

•2 s

3 5

_3 2

C -Cj

^ 5

S §■
■1,'^
5 ^

^ k
•3 3

a, <

O O

U 2

kn P-

S 2 S S S
o o o o o

S 2 S S P

Uh Ph Ph Ph P-.

CNl ^

ON —

»n

(N 0

d d

■O

os —

10 »n

0 0

0 — '

d d

r' (N ON ko
On ■■t — ' 00

O o m -r]-

d P d d

-H Tt

O 00

d d

ON NO

— (N

d d

■■t

r' ro

00 — ' U1

00 — ' ro

d d d

-H Tf NO

^ — I 00

NO O ■■t

d d d

td
^ 2
U 2*

3 §

3 ?

2 S

'2 3

c -s

3 FX

-Si

3 -3

3 3

■3

- 3
3. 2

P a.

U O

> s

Q S.

3 I

8 I

Di K

at.

•S-

■— .j

-3

•2

a. L ^

^33

•2^3

.2?^ 2 :2

3 k 3 ■

to >1 t<3

3 3 3-

^ s

APPENDIX. CONTINUED.

Anjos and Bo^on • BIRD COMMUNITIES IN FOREST PATCHES

413

>>.2
rz o

I ^

c/5

2 « S
o o o

C ^
^ 3

^ Cj

Ci-

S S 2 S
o o o o

Q U U Q
PJ tin Uh W

J O J J

EEEotSoooooooo

u u u

P> PP Ph

U Q U Q P

pH U P-, U Pu

.-H

«-H

r--

<— 1

1-H

— H

1-H

'tt

*— (

w—(

t-H

»— (

— i

— a\
o o
d d

W
<
Q

.to

•g

■5

§ ^

-3

to t/5

3 3

3 .

I ^

^ Cj

^ O
^ C

§■

w
<

P 5 G O P

s -=
Q. a

^ t^

.a -2 S

Cp £X ^
r* •^‘ S

-S ^

3 ■

O, c:) cq

-3

-C5

'HH

5u t3

a „
03 W

Cei .^u

-Ci

'■**.*

-a

■a

a a a

■a a;
Op a,
a a
kj kj

a ^
a Si

a •a-
kj a.

.2 S

^ a
a a;

ct ^

s S

05 2
ac -j-
a Ea
,a Si

Ki Co

d s

a '3

I § .2

tr: -a

to a
a. 5.

5 S 5

-a a: ^

p p 5:

a a
a -2
a tp

^ -2
to S

2 ^
a Q

a c
a a

^ 2
■to
eS tt;

to Ck

a. a
ST a
a a
a a
^ a

a iH

a a
a h
a -£
a -2
a a

-S

a, a:

414

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

ooooooooo

qsqdqqdqqq

[UtuUPuUtOU-WUU

CL.

w-H

1-^

•—

— H

r--

Tj-

£ o-

i2"’

§ 'C

■ij

t:?

^ 5

3 'c:
to ^ ^

C ^3 ^

■a N

ji ^ a

^ ^ -ii

to LO

Cs t

I 2

a -2
-c

&• •-

z E

"S. I

1? -2

w s;

•a ^

5 I,

5u 3

3 ^a
•2 ^
3 3
t3
C >-
C 3
0, O

Habitat (OA = open area; ED = edge; FU = forest understory; FM = forest mid-levels; FC = forest canopy).

Feeding habit (OM = omnivore; FR = frugivore; NE = nectarivore; CA = carnivore; GI = generalized insectivore; LI = leaf insectivore; T1 = trunk insectivore) is shown for each species.

Wilson Bull., 111(3), 1999, pp. 415-420

DO MAMMALIAN NEST PREDATORS FOLLOW HUMAN SCENT
TRAILS IN THE SHORTGRASS PRAIRIE?

SUSAN K. SKAGEN,' 2 THOMAS R. STANLEY,' AND M. BETH DILLON'

ABSTRACT. — Nest predation, the major cause of nest failure in passerines, has exerted a strong influence on
the evolution of life history traits of birds. Because human disturbance during nest monitoring may alter predation
rates, we investigated whether human scent affected the survival of artificial ground nests in shortgrass prairie.
Our experiment consisted of two treatments, one in which there was no attempt to mask human scent along
travel routes between artificial nests, and one in which we masked human scent with cow manure, a scent
familiar to mammalian predators in the study area. We found no evidence that human scent influenced predation
rates, nor that mammalian predators followed human trails between nests. We conclude that scent trails made
by investigators do not result in lower nesting success of passerines of the shortgrass prairie where vegetation
trampling is minimal, mammalian predators predominate, and avian predators are rare. Received 9 Nov 1998,
accepted 10 Feb. 1999.

Predation has exerted a strong influence on
the evolution of habitat selection and life his-
tory traits for many avian species (Martin
1993b). Research on a broad array of ecolog-
ical topics requires estimates of avian fecun-
dity. Because nest predation is the major cause
of nest failure in passerines (Ricklefs 1969;
Martin 1992, 1993a, b), researchers have fre-
quently expressed concerns that monitoring
might artificially increase predation rates
(Mayfield 1975, Major 1990, Gotmark 1992).

Predators might be attracted to nests by vi-
sual cues, such as the presence of researchers,
trampling of vegetation, increased activity of
parent birds, and by olfactory cues. Mamma-
lian predators are thought to follow tracks in
the vegetation and to respond to human scent
along the trails or at the nests (Creighton
1971, Wilson 1976, Nol and Brooks 1982,
Gotmark 1992, Whelan et al. 1994). In a re-
view paper on investigator bias, Gotmark
(1992) concluded there was little or no evi-
dence that researcher disturbance increased
mammalian predation rates. Of three studies
that have directly addressed whether human
scent increases mammalian predation rates
(Keith 1961, Macivor et al. 1990, Whelan et
al. 1994), one (Whelan et al. 1994) supported
the hypothesis. Even though evidence is scant,
the use of rubber boots and gloves is widely

' U.S. Geological Survey, Biological Resources Di-
vision, Midcontinent Ecological Science Center, Fort
Collins, CO 80525-3400.

^ Corresponding author;

E-mail; susan_skagen(g> usgs.gov

recommended to alleviate the potential prob-
lem of human scent leading to bird nests (Nol
and Brooks 1982, Yahner et al.l993. Major
and Kendal 1996).

Artificial bird nests have been widely used
in predation studies (e.g., Gottfried and
Thompson 1978, Yahner and Wright 1985,
Yahner et al. 1993). Despite problems with in-
terpretation of results (Major and Kendal
1996), they remain a useful tool for testing
predation theories. We conducted an experi-
ment using artificial ground nests in a short-
grass prairie where the primary nest predators
are mammals and human presence is rare. Our
objective was to test if human scent increased
the rates of predation on shortgrass prairie
ground nesting birds by comparing two meth-
ods of experimenter travel between nests.

The purpose of our study was to determine
the most expedient technique for ongoing
breeding bird studies in the shortgrass prairie.
We do not intend to make inferences from this
study to other ecosystems and predator com-
munities. Because breeding systems vary in
predator communities, predator behavior, ex-
posure to human presence, vegetation struc-
ture, and nest position, many systems need to
be evaluated before we can fully understand
the effect of human scent on predation rates.

METHODS

We conducted this experiment in July 1997 on Paw-
nee National Grassland. 7 km northwest of Briggsdale,
Weld County, Colorado (40° 41' N, 104° 24' W). The
259 ha tract of grazed shortgrass prairie is character-
ized by short and mid-grasses, cacti {Opunlia sp.).

415

416

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

forbs, and patchy areas of yucca {Yucca glauca). Com-
mon ground nesting passerines in the vicinity include
Homed Lark {Eremophila alpestris). Lark Bunting
(Calamospiza melancorys), McCown’s Longspur (Cal-
cariu.s mccownii). Chestnut-collared Longspur {Cal-
carius omatus), and Western Meadowlark {Sturnella
neglecta). Potential predators of ground-nesting birds
include thirteen-lined ground squirrel {Spermophilus
tridecemlineatu.s), deer mouse (Peromyscus manicu-
latu.s), northern grasshopper mouse {Onychomys leu-
cogaster), coyote (Canis latrans), swift fox (Vulpes ve-
lox), raccoon (Procyon lotor), long-tailed weasel {Mus-
tela frenata), badger {Taxidea taxus), striped skunk
{Mephitis mephitis), and several snake species.

We placed 100 artificial nests along two transects
established 800 m apart in similar habitat. Each tran-
sect contained 25 lines perpendicular to the transect
and alternating in opposite directions at 50 m intervals
so that adjacent lines were 100 m apart. Each line con-
tained two nests; one at 100 m (Nest A) and the second
at 200 m (Nest B) from the transect. From Nest B we
walked an additional 100 m so that both nests on the
line were treated equally. To aid in relocating nests we
noted any distinguishing features around the nest and
tied surveyor’s tape to low-growing vegetation 10 m
from each nest, a distance not associated with in-
creased predation rates (Major and Kendal 1996).

To test if human scent trails influenced predation
rates, we subjected artificial nests to two treatments.
In Treatment 1 (human scent), we wore leather boots
and made no attempts to mask human scent while
walking between nests. In Treatment 2 (masked scent),
we masked human scent with a scent familiar to po-
tential predators in the study area by wearing rubber
boots that were sprayed with a cow manure tea (fresh
cow manure steeped in water, in a 1:3 mixture for at
least 12 hours) at the beginning of each line. Treatment
types were randomly assigned to the 50 lines (25 lines
per transect); both nests on a line received the same
treatment. Because we specifically wanted to deter-
mine an effect of scent trails, we wore vinyl craft
gloves (standard field practice) while handling eggs in
both treatments to minimize human scent on the eggs.

Nests consisted of a scrape on the ground and con-
tained two fresh Japanese quail (Cnturnix japonica)
eggs (mean length X width, 3.3 X 2.6 cm, n = 20)
and one clay egg (2.2 X 1.5 cm, n = 20). Scrapes
were created using the broad end of a large wooden
tongue depressor. While wearing rubber gloves we
constructed clay eggs out of soft modeling compound
(Sculpey III brand) to approximate the size of Lark
Bunting eggs. Clay eggs aided in the identification of
nest predators and enabled us to record predation by
predators too small to handle quail eggs (i.e., small
rodents; Major and Kendal 1996).

Nests were set out on 9 July 1997 and checked three
days later, a time interval during which we expected
50% of the ne.sts to survive ba.sed on preliminary re-
sults of trials using artificial nests constructed in the
same manner. Although .several studies used longer tri-
al intervals, we expected that our ability to detect dif-

TABLE 1. Predation outcomes for 25 lines receiv-
ing the human scent treatment and 25 lines receiving
the masked scent treatment.

Predation outcome

( I = depredated. Number of lines

0 = survived) with outcome

Nest A

Nest B

Human scent
(«,)

Masked scent
("!/)

ferences would be diminished if nearly all nests were
depredated. Nests were classified as intact or disturbed
based on signs of disturbance to either quail or clay
eggs. Nests were considered disturbed if quail eggs
were missing, broken, or moved, or if clay eggs were
missing, moved, or had tooth impressions. We collect-
ed extant clay eggs for examination and identification
of any diagnostic marks. We classified markings on the
clay eggs as rodent, non-rodent, insect, or unknown by
comparing them with known tooth impressions made
from skulls in the zoology collection at Colorado State
University, Fort Collins, Colorado. In the absence of
other signs of disturbance, nests containing clay eggs
with only insect marks were considered intact.

The data from this experiment are counts and can
be arranged into an / X j contingency table (Table 1),
where i denotes the predation outcome and j denotes
the treatment (i.e., human or masked scent). While the
cell probabilities for such tables are commonly mod-
eled and estimated using standard loglinear models
(e.g., Agresti 1990), reparameterization of the under-
lying multinomial model can lead to loglinear models
that are difficult to construct or difficult to interpret.
In this study we reparameterize the underlying multi-
nomial model to address the following specific ques-
tions: (1) do predation probabilities differ for nests on
a line because of differences in their proximity to the
transect, (2) do predation probabilities for nests differ
because of differences in human and masked scent
treatments, and (3) is there evidence that predators fol-
lowed the investigator’s trail between nests on a line.
Hence, instead of using a loglinear modeling approach,
we derived parameter estimates and constructed hy-
pothesis tests using classical maximum likelihood
methods (e.g., Lar.sen and Marx 1986:261). The gen-
eral procedure was to ( 1 ) construct the appropriate
likelihood function for the data, (2) derive estimators
and compute estimates for parameters under the model,
(3) evaluate the likelihood function at the maximum
likelihood parameter estimates to obtain the deviance
(here we omit the term for the saturated model and
define deviance as -2 X (log-likelihood), and then (4)
test specific hypotheses using likelihood ratio tests for
nested models (Agresti 1990:21 1). The models used in
this study are presented in the Appendix. SAS statis-

Skcificn el al. • MAMMALIAN NEST PREDATORS

417

TABLE 2. Three candidate models for estimating
predation probabilities constructed using (A.l).

Model

Pariinielers and Con.straint.s

Deviance

Pa> Pb,

137.182

138.549

Pa Pb Q Cg

138.589

tical software (version 6.12 on an IBM-eompatible mi-
crocomputer; SAS Institute Inc. 1990) was used for all
computations. Values reported are means (±SE).

RESULTS

During the trial, 49 of 100 nests were dis-
turbed, 24 from the human scent treatment
and 25 from the masked scent treatment. Dis-
turbance to quail eggs was apparent in 45 of
49 (92%) nests; eggs were missing from 12
nests, broken in an additional 7 nests, and
moved in an additional 26 nests. In four nests,
the quail eggs were undisturbed, yet clay eggs
were either moved or had tooth impressions.
Clay eggs were undisturbed in only two nests
with disturbed (broken) quail eggs. Rodent
tooth impressions were identified on 22 clay
eggs, and non-rodent impressions on one clay
egg. Clay eggs were missing from 20 nests.
Quail eggs and clay eggs that were moved
were displaced an average of 31.3 cm (± 9.10,
median 1 cm, range 0.5-330 cm, n = 41) and
43.2 cm (± 15.07, median 20 cm, range 0.5-
250 cm, n = 20) from their original positions,
respectively.

Predation outcomes for the two treatments
are summarized in Table 1. In general, few
differences between the treatments were evi-
dent. In Table 2, the three candidate models
constructed under A.l are presented, along
with their deviance. The likelihood ratio test
between model 2 and model 1, which tests for
differences in predation rates between nest A
and nest B caused by their proximity to the
transect, had a P-value of 0.505 (y^ = 1.37,
df = 2). Hence, there appears to be no effect
as a result of proximity to the transect. The
likelihood ratio test between model 3 and
model 2, which tests for differences in pre-
dation rates caused by differences in the hu-
man and masked scent treatments, had a P-
value of 0.841 (x^ = 0.04, df = 1). Hence, we
conclude there was no treatment effect.

For the two-parameter model in A. 2 (i.e., p,

p'), which allows unconditional and condi-
tional predation probabilities for nests on a
line to differ, we get a deviance of 97.094.
When we impose the constraint p = p' (giving
us a one-parameter model that is equivalent to
model 3 in Table 2), we get a deviance of
98.387. The likelihood ratio test for these
models, which tests whether predators were
following the investigator’s trail between nests
on a line, had a P-value of 0.256 (x^ = 1.29,
df = 1). Consequently, we conclude predators
did not follow the investigator’s trail between
nests on a line.

The one parameter models from A.l and
A. 2 are mathematically equivalent and, based
on the likelihood ratio tests, are the appropri-
ate models to use for parameter estimation.
Hence, the estimated three-day predation
probability for nests in this study was 0.49
(±0.050), which gives an estimated daily sur-
vival probability of 0.80 (±0.026).

DISCUSSION

We found no evidence that human scent
trails to nests altered predation rates on arti-
ficial nests in grasslands where the main pred-
ators are small mammals, nor did we find ev-
idence that predators were more likely to dep-
redate nests on the same trail. A learned as-
sociation of human scent with food is unlikely
because human presence is rare throughout
much of our study area. Rather, the scent
would be novel to small mammals of the re-
gion. We found no evidence that novel scent
was an attractant to predators in our region.

Whether human scent is an attractant or de-
terrent to predators has been a topic of spec-
ulation (Creighton 1971, Mayfield 1975, Wil-
son 1976, Gotmark 1992) that has been di-
rectly tested in only three other studies. Re-
sults differ between studies. Keith (1961)
reported no effect of human scent on survival
of artificial duck nests in wetlands with pre-
dominately mammalian predators. Macivor
and coworkers (1990) found that red fox {Vid-
pes vulpes) avoided human scent associated
with experimental plover nests along a beach.
In contrast, Whelan and coworkers (1994) re-
ported raccoons in a forested system preying
on nests with human scent and novel scent
more frequently than nests with no scent or
familiar scent. In another study evaluating the
influence of familiar and novel scents, Clark

418

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

and Wobeser (1997) determined that a novel
odor (lemon juice and ground ginger root) did
not affect survival of artificial waterfowl
nests. Collectively, these studies represent a
broad variability in predator community com-
position, predator behavior, exposure to hu-
man presence, vegetation structure, and nest
placement.

Evidence of other aspects of investigator
bias on predation rates is also equivocal. Fre-
quency of nest visits had no effect on preda-
tion rates of artificial nests in several studies
(Bowen et al. 1976, Gottfried and Thompson
1978, Erikstad et al. 1982, Maclvor et al.
1990, Gotmark 1992, Mankin and Warner
1992), but did in two studies (Major 1990,
Esler and Grand 1993), presumably because
of vegetation trampling. In our study, we did
not evaluate the effects of frequency of visi-
tation nor vegetation trampling. In the short-
grass prairie, the one visit to artificial nests
during construction resulted in minimal veg-
etation trampling.

One criticism of artificial nests is that they
often contain only quail eggs and that small
predators unable to handle the quail eggs may
be under-represented (Major and Kendal
1996). We addressed this problem by consid-
ering nests disturbed when eggs were moved
as well as broken or removed, and by using
smaller clay eggs in addition to quail eggs. We
found, however, that only two nests would
have been misclassified as undisturbed if only
quail eggs had been used.

We conclude that the procedures we used
while visiting nests are unlikely to contribute
to reduced nesting success of passerines of the
shortgrass prairie where vegetation trampling
is minimal, mammalian predators predomi-
nate, and avian predators are rare. Our con-
clusion is consistent with Gotmark (1992)
who surmises that passerines are less sensitive
to investigator disturbance than other groups
of birds, scent having less effect than vege-
tation trampling, and increases in predation in
response to human cues more common for
avian than mammalian or reptilian predators.
We recommend that investigators continue to
evaluate whether human scent alters predation
rates in avian breeding systems and not make
inappropriate inferences across systems. Hu-
man scent studies that identify and describe
the predator communities, habitat structure.

and human influence will ultimately contrib-
ute to better understanding of observer bias in
research.

ACKNOWLEDGMENTS

We thank A. Yackel Adams and R. Adams for field
assistance, J. Miller for suggestions on experimental
design, B. A. Wunder for access to the zoology col-
lection at Colorado State University, and D. T. Arm-
strong for information on mammals of the study area.
J. Bradley and M. Laubhan reviewed a draft of the
manuscript. This study was funded by the Biological
Resources Division of the U. S. Geological Survey.

LITERATURE CITED

Agresti, a. 1990. Categorical data analysis. John Wi-
ley and Sons, New York.

Bowen, D. E., R. J. Robel, and R G. Watt. 1976.
Habitat and investigators influence artificial
ground nest losses. Trans. Kans. Acad. Sci. 79;
141-148.

Clark, R. G. and B. K. Wobeser. 1997. Making sense
of scents: effects of odour on survival of simulat-
ed duck nests. J. Avian Biol. 28:31—37.
Creighton, P. D. 1971. Nesting of the Lark Bunting
in northcentral Colorado. U.S.I.B.P. Grassland Bi-
ome Tech. Rep. 68. Colorado State Univ., Port
Collins.

Erikstad, K. E., R. Blom, and S. Myrberget. 1982.
Territorial Hooded Crows as predators on Willow
Ptarmigan nests. J. Wildl. Manage. 46:109-114.
Esler, D. and J. B. Grand. 1993. Factors influencing
depredation of artificial duck nests. J. Wildl. Man-
age. 57:244-248.

Gotmark, E 1992. The effects of investigator distur-
bance on nesting birds. Cuix Ornithol. 9:63-104.
Gottfried, B. M. and C. F. Thompson. 1978. Exper-
imental analysis of nest predation in an old-field
habitat. Auk 95:304-312.

Keith, L. B. 1961. A study of waterfowl ecology on
small impoundments in southeastern Alberta.
Wildl. Monogr. 6:1-86.

Larsen, R. J. and M. L. Marx. 1986. An introduction
to mathematical statistics and its applications, sec-
ond ed. Prentice-Hall, Englewood Cliffs, New Jer-
sey.

MacIvor, L. H., S. M. Melvin, and C. R. Griffin.
1990. Effects of research activity on Piping Plover
nest predation. J. Wildl. Manage. 54:443-447.
Major, R. E. 1990. The effect of human observers on
the intensity of nest predation. Ibis 132:608-612.
Major, R. E. and C. E. Kendal. 1996. The contri-
bution of artificial nest experiments to understand-
ing avian reproductive success: a review of meth-
ods and conclusions. Ibis 138:298-307.

Mankin, R C. and R. W. Warner. 1992. Vulnerability
of ground nests to predation on an agricultural
habitat island in east-central Illinois. Am. Midi.
Nat. 128:281-291.

Martin, T. E. 1992. Breeding productivity consider-

Skai>en el al. • MAMMALIAN NEST PREDATORS

419

ations: what are tlie appropriate habitat features
for management? Pp. 455-473 in Ecology and
conservation of Neotropical migrant land birds (J.
M. Hagan and D. W. Johnston, Eds.). Smithsonian
Institution Press, Washington, D.C.

Martin, T. E. 1993a. Nest predation among vegetation
layers and habitats: revising the dogmas. Am. Nat.
141:897-913.

Martin, T. E. 1993b. Nest predation and nest sites:
new perspectives and old patterns. BioScience 43:
523-532.

Mayfield, H. E 1975. Suggestions for calculating nest
success. Wilson Bull. 87:456-466.

Nol, E. and R. J. Brooks. 1982. Effects of predator
exclosures on nesting outcome of Killdeer. J. Field
Ornithol. 53:263—268.

Ricklefs, R. E. 1969. An analysis of nesting mortality
in birds. Smithson. Contrib. Zool. 9:1—48.

SAS Institute Inc. 1990. SAS language reference,
version 6, first ed. SAS Institute, Inc., Cary, North
Carolina.

Whelan, C. J., M. L. Kilger, D. Robson, N. Hallyn,
AND S. Dilger. 1994. Effects of olfactory cues on
artificial-nest experiments. Auk 111:945-952.
Wilson, J. K. 1976. Nesting success of the Lark Bun-
ting near the periphery of its breeding range.
Kans. Ornithol. Soc. Bull. 27:13-22.

Yahner, R. H. and a. W. Wright. 1985. Depredation
on artificial ground nests: effects of edge and plot
age. J. Wildl. Manage. 49:158-161.

Yahner, R. H., C. G. Mahan, and C. A. Delong.
1993. Dynamics of depredation on artificial
ground nests in habitat managed for Ruffed
Grouse. Wilson Bull. 105:172-179.

APPENDIX

For a particular line receiving either the hu-
man scent or the masked scent treatment, four
outcomes are possible (Table 1). If we denote
these outcomes by / (/ = 1, . . . , 4), let the
probability of the /-th outcome be tt, and y,
(respectively) for human and masked scent
treatments, and let n, and m, (respectively) be
the number of lines for which the /-th outcome
was observed for human and masked scent
treatments, then the probability of the ob-
served data is the product of two multinomi-
als:

4 4

;=1 1=1

where C, and C2 are multinomial coefficients,
2,77, = 2,y,- = 1, and 2,/2, = 2,m, = 25. Under
the assumption that lines and nests on a line
are independent (the latter assumption is test-
ed below using A. 2), we can reparameterize
this model in terms of the probability nest A

and nest B were depredated for human scent
treatments and p/d and the probability nest
A and nest B were depredated for masked
scent treatments (c^ and Cn), to obtain a model
with likelihood function proportional to:

L(1 - Pa){\ -Pb)]'''[Pa(1 -Pfi)''4(l - Pa)Pb]''^

X - Q)(1 - Cb)Y"[ca{\ - Cb)]"'2

X [(1 - Q)cB]"'3[c^Ca]'"L (A.l)

We derived estimators for p^, pg, c^, and Cg
using standard maximum likelihood methods
(Larsen and Marx 1986:261). Differences in
predation probabilities between nests A and B
due to differences in proximity to the transect,
and differences in predation probabilities due
to differences in treatments, were tested by
constraining parameters in A.l to obtain the
appropriate submodels, and then performing
likelihood ratio tests. In the first submodel pa-
rameters were constrained so that, within a
treatment, predation probabilities for nests A
and B were constant (i.e., p^ = pg and =
Cg). In the second submodel parameters were
constrained so that predation probabilities
were constant between nests A and B and
across treatments (i.e.. Pa — Pb ~ = Cg, see

Table 1).

In an effort to determine whether predators
were following the human trail between nests,
one additional model was constructed. This
model assumed that predation probabilities
among nests on a line and among treatments
did not differ, but allowed the unconditional
and conditional predation probabilities of
nests on a line to differ. Here, the conditional
predation probability is the probability nest A
would be depredated given nest B had already
been depredated, or the converse. If we denote
the unconditional predation probability by p
and the conditional predation probability by
p' , then the probability neither nest on a line
is depredated is given by (1 — p)(l — p) and
the probability both nests on a line are dep-
redated is given by pp' . To obtain the proba-
bility that only one nest on a line is depre-
dated, we exploit the fact that the cell proba-
bilities for a multinomial must sum to one.
Hence, the probability that only one nest on a
line is depredated is given by 1 — (1 — p)(l
— p) ~ pp' which, after some algebraic ma-
nipulation, yields the intuitively reasonable

420

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

p{\ — p') + (1 — p)p. Thus, the resulting
model has likelihood function proportional to:

[(1 - p)ii - p)Y'’r\w - p') +

(1 - (A.2)

where Uj and ntj (i = 1, . . . , 4) are as defined
above. A test for Hq: p — p' versus H^: p ^
p' was constructed using a likelihood ratio
test.

Short Communications

Wilson Bull., 111(3), 1999, pp. 421-422

Christmas Shearwater Egg Dimensions and Shell Characteristics on
Laysan Island, Northwestern Hawaiian Islands

G. C. Whittow^'^ and M. B. Naughton^

ABSTRACT. — The mean fresh egg mass of Christ-
mas Shearwaters (Puffinus nativitatis) on Laysan Is-
land, in the Northwestern Hawaiian Islands, was 44.9
± 3.4 (SD) g, and the mean egg volume was 42.3 ±
2.9 cm-’. The measured length and breadth of the eggs,
the shell mass, shell thickness, and number of pores in
the shell were within 10% of predictions for procel-
lariiform birds, based on fresh egg mass or on both
fresh egg mass and incubation period. These data con-
form with evidence that there are few allometric dif-
ferences between the eggs of tropical Procellariiformes
and those of Procellariiformes from higher latimdes.
Received 19 Oct. 1998, accepted 25 Feh. 1999.

The Christmas Shearwater (Puffinus nativ-
itatis) is a tropical procellariiform seabird that
breeds on islands in the central North and
South Pacific oceans (Warham 1990). It has
been little studied, perhaps because the spe-
cies is not abundant anywhere (Shallenberger
1984). The purpose of the present note is to
report data for Christmas Shearwater eggs col-
lected on Laysan Island (28° 12' N; 177° 20'
W) in the northwestern Hawaiian Islands.

We measured egg volume by weighing the
egg in air and again submerged in water. From
the difference between the mass in air and in
water, and the density of water, egg volume can
be determined in accordance with Archimedes’
Principle (Rahn et al. 1976). We determined
fresh egg mass by weighing the egg after fiUing
the airceU with distilled water (the mass loss of
the egg during incubation being entirely the re-
sult of the loss of water vapor; Grant et al.
1982), and we measured both the length and
breadth of the egg with a dial caliper. We ob-
tained the shell mass, its thickness, and we

■ Dept, of Physiology, John A. Bums School of Med-
icine, Univ. of Hawaii, Honolulu, HI 96822; E-mail:
whittowg@jabsom.biomed.hawaii.edu
^ Corresponding author.

’ U.S. Fish & Wildlife Service, Prince Kuhio Fed-
eral Building, Honolulu, HI 96850.

counted the number of pores in the shell of ran-
domly-selected sub-samples of eggs after drying
the shells in a desiccator. We measured pore
density using the procedure described by Tyler
(1953) and Roudybush and coworkers (1980).

The mean fresh egg mass of 18 Christmas
Shearwater eggs was 44.9 ± 3.4 (SD) g.
Knowledge of the fresh egg mass provides an
opportunity to compare some of the other mea-
sured values (Table 1) with predictions for pro-
cellariiform eggs, based on the mass of their
freshly laid eggs (Rahn and Whittow 1988).
There are no predictive equations for the vol-
ume of the eggs of Procellariiformes, but mea-
sured egg lengths and breadths were similar
(100.3% and 96.0%, respectively) to predicted
values (Table 1). Measured shell mass and shell
thickness were also similar (94.7% and
105.6%, respectively) to predictions (Table 1).
Rahn and Whittow (1988) presented two pre-
dictive equations for the total number of pores
in the eggshell. Both require the calculation of
the surface area of the egg from its mass (Tul-
lett and Board 1977), which is then multiplied
by the measured pore density (Table 1). The
resulting estimated total number of pores in the
shell of a Christmas Shearwater’s egg is 3103
pores. This value falls between the two pre-
dicted values (2963 and 3584). The predicted
values were both based on the incubation pe-
riod as well as on the fresh egg mass; for this
purpose, we used an incubation period of 53
days (Byrd et al. 1983, Naughton 1983) and a
fresh egg mass of 44.9 g.

The measured values for the eggs and egg-
shells of the tropical Christmas Shearwater are
close to predictions for Procellariiformes in
general. This finding supports evidence that
there are few differences in the allometric re-
lationships of eggs and eggshells of Procel-
lariiformes between tropical and non-tropical
species. In contrast, there are substantial dif-
ferences between Procellariiformes and other

421

422

THE WILSON BULLETIN • Vol. HI. No. 3, September 1999

TABLE 1. Measured and predicted values for the eggs and eggshells of Christmas Shearwaters on Laysan
Island. The mean measured values (±1 SD) are shown; n = the numbers of observations. Predicted values are
calculated following Rahn and Whittow (1988).

Measured {x ± SD)

Predicted

Volume (cm^)

42.3 ± 2.9

Length (cm)

55.6 ± 2.0

55.5

Breadth (cm)

38.3 ± 1.3

39.9

Eggshell

Mass (g)

2.9 ± 0.2

3.0

Thickness (mm)

0.3 ± 0.0

0.3

Pore density [pores (cm0“']

59.8 ± 4.7

orders of seabirds in this regard (Whittow
1984, Ar and Rahn 1985).

ACKNOWLEDGMENTS

We thank the U.S. Fish and Wildlife Service for
granting permission to collect eggs on Laysan Island.

LITERATURE CITED

Ar, a. and H. Rahn. 1985. Pores in avian eggshells:
gas conductance, gas exchange and embryonic
growth rate. Resp. Physiol. 61:1-20.

Byrd, G. V., D. I. Moriarty, and B. G. Brady. 1983.
Breeding biolgy of Wedge-tailed Shearwaters at
Kilauea Point, Hawaii. Condor 85:292—296.
Grant, G. S., C. V. Paganelli, T. N. Pettit, and G.
C. Whittow. 1982. Determination of fresh egg
mass during incubation. Condor 84:121-122.
Naughton, M. 1982. Breeding biology of the Christ-
mas Shearwater (Puffinus nativitatis) on Laysan
Island, Hawaii. Pac. Seabird Group Bull. 9:71-72.
Rahn, H. and G. C. Whittow. 1988. Adaptations to a

pelagic life: eggs of the albatross, shearwater and
petrel. Comp. Biochem. Physiol. 91A:415— 423.
Rahn, H., C. V. Paganelli, I. C. T. Nisbet, and G. C.
Whittow. 1976. Regulation of incubation water
loss in eggs of seven species of terns. Physiol.
Zool. 49:245-259.

Roudybush, T, L. Hoffman, and H. Rahn. 1980.
Conductance, pore geometry, and water loss of
eggs of Cassin’s Auklet. Condor 82:105-106.
Shallenberger, R. J. 1984. Fulmars, shearwaters and
gadfly petrels. Pp. 42—56 in Seabirds of eastern
North Pacific and Arctic waters (D. Haley, Ed.).
Pacific Search Press, Seattle, Washington.
Tullett, S. G. and R. G. Board. 1977. Determina-
tions of avian eggshell porosity. J. Zool. (Lond.)
183:203-21 1.

Tyler, C. 1953. Studies on egg shells. II: method for
marking and counting pores. J. Sci. Food Agric.
4:266-272.

Warham, j. 1990. The petrels: their ecology and
breeding systems. Academic Press, London, U.K.
Whittow, G. C. 1984. Physiological ecology of incuba-
tion in tropical seabirds. Stud. Avian Biol. 8:47-72.

Wilson Bull., 111(3), 1999, pp. 422-424

The Paint-billed Crake Breeding in Costa Rica

David M. Watson'-^ and Brett W. Benz'

ABSTRACT. — We report a recent observation from
southern Costa Rica of the Paint-billed Crake {Neocre.x
erythrops), a little known species from eastern and
northern South America. An adult and recently
hatched chick were observed at close range in wet

‘ Natural History Museum & Biodiversity Research
Center, and Dept, of Ecology and Evolutionary Biol-
ogy, The Univ. of Kansas, Lawrence, KS 66045.

2 Corresponding author; E-mail: vergil@ukans.edu

grassy second-growth. This observation constitutes the
first record of the young of this species and represents
the only breeding record for Central America. Re-
ceived 12 Nov. 1998, accepted 12 Feb. 1999.

On 5 June 1998, at 16:30 an adult Paint-
billed Crake (Neocrex erythrops) was ob-
served, accompanied by a chick, near the

SHORT COMMUNICATIONS

423

town of Golfito on the Pacific coast of Costa
Rica, close to the Panama border (8° 37' N,
83° 1 r W). Observations were made by
both authors and K. Cohoon, while walking
slowly beside the Golfito airstrip amidst
grassy second growth interspersed with a
row of large Ficus trees. Behind this vege-
tation was a slow moving stream with thick-
ets of tall grass along its banks. We saw the
birds from approximately 15 m and watched
them for 25 seconds using binoculars. The
adult paused in the middle of the path, even-
tually returning to the wet grassy second
growth from which it had walked. The
bright yellow bill with a scarlet base, and
black and white barred flanks were clearly
visible, clearly distinguishing it from the
congeneric Colombian Crake (Neocrex col-
ombianus). The chick was covered uniform-
ly in black natal down and the tarsi were
dark grey or horn. Further soft-part colors
were not noted because it quickly ran away
from the adult, across the path into thick un-
dergrowth beside the airstrip. Despite sub-
sequent visits to this locality for several
weeks, we made no further observations of
this species.

Neocrex erythrops is a little known species
that ranges widely in eastern and central South
America; N. e. olivascens is known east of the
Andes from Colombia and Venezuela south to
Paraguay and Argentina, and N. e. erythrops
from west of the Andes in coastal Peru and
the Galapagos Islands (Ripley 1977). There
have been reports of vagrants within South
America (Osborne and Beissinger 1979), from
suburban areas and up to 3375 m elevation
(Remsen and Traylor 1983), and some recent
reports (Tostain et al 1992, Haverschmidt and
Mees 1994) suggest that the species may be
resident in Surinam and French Guyana. As
with many other species of rail, the chicks of
this species are undescribed (Ripley 1977, del
Hoyo et al 1996). The uniform black down
and dark tarsi are similar to the young of other
neotropical rails in the genera Loterallus and
Porzana.

The status of this species in Central Amer-
ica is unclear. There is only one definite record
from Costa Rica, from the Sarapiqui lowlands
in the northeast by Stiles and Rosselli on 22
August 1987 (Stiles and Skutch 1989). There
is an additional record of either this species or

the similar N. colombianus from southern
Costa Rica, near Hitoy Cerere in March 1985
(Pratt et al., reported in Stiles and Skutch
1989). Two specimens collected in the coastal
lowlands of Bocas del Toro, Panama on 10
November 1981, were later identified by Rip-
ley as the wide-ranging N. erythrops olivas-
cens-, several individuals were seen at Tocu-
men Marsh in eastern Panama by Behrstock
(1983). All of these records are from the Ca-
ribbean lowlands, thus the record reported
herein constitutes the first for the Pacific slope
of Central America.

There are two records from North Ameri-
ca: from east central Texas on 17 February
1972 (Arnold 1978), and Virginia on 15 De-
cember 1978 (Blem 1980). Both these re-
cords were probably wandering individuals,
a pattern seen in many other species of rail
(Remsen and Parker 1990). In contrast to
these winter records, the two records from
Costa Rica are from August and March.
Based on the June 15 record we report, and
the clear evidence of breeding, we suggest
that N. erythrops has a breeding population
in southern Costa Rica.

ACKNOWLEDGMENTS

We thank D. Levey, J. Eberhard, and two anony-
mous reviewers for helpful and constuctive comments
on the manuscript.

LITERATURE CITED

Arnold, K. A. 1978. First United States record of
Paint-billed Crake (Neocrex etythrops). Auk 95:
745-746.

Behrstock, R. A. 1983. Colombian Crake (Neocrex
colombianus) and Paint-billed Crake (N. eiy-
throps): hrst breeding records for Central Ameri-
ca. Am. Birds 37:956-957.

Blem, C. R. 1980. A Paint-billed Crake in Virginia.
Wilson Bull. 92:393-394.

DEL Hoyo, J., A. Elliot, and J. Sargatal. 1996.
Handbook of the birds of the world. Vol. 3: Hoa-
tzin to auks. Lynx Edicions, Barcelona, Spain.
Haverschmidt, E and G. E Mees. 1994. Birds of Su-
riname. Vaco Press, Paramaribo, Suriname.
Osborne, D. R. and S. R. Beissinger. 1979. The
Paint-billed Crake in Guyana. Auk 96:425.
Remsen, J. V., Jr. and M. A. Traylor, Jr. 1983. Ad-
ditions to the avifauna of Bolivia, part 2. Condor
85:95-98.

Remsen, J. V., Jr. and T. A. Parker, III. 1990. Sea-
sonal distribution of the Azure Gallinule (Porphy-

424

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

nila flavirostris), with comments on vagrancy in
rails and gallinules. Wilson Bull. 102:380-399.
Ripley, S. D. 1977. Rails of the world. David R. God-
ine. Pub., Boston, Massachusetts.

Stiles, E G. and A. E Skutch. 1989. A guide to the

birds of Costa Rica. Cornell Univ. Press, Ithaca,
New York.

Tostain, O., J. L. Dujardin, C. Erard, and J. M.
Thiollay. 1992. Oiseaux de Guyana. Societe
d’Etudes Ornithologiques, Brunoy, Guyana.

Wilson Bull., 11 1(3), 1999, pp. 424-426

Additional Records of Fall and Winter Nesting by Killdeer in Southern

United States

Kimberly G. Smith,*’^ W. Marvin Davis,- Thomas E. Kienzle,^ William Post,^ and

Robert W. Chinn^

ABSTRACT. — We report on successful nesting at-
tempts in fall by Killdeer (Cliaradrius vociferii.s) in
southern Mississippi in November, 1987 and in central
Arkansas in October, 1998, and a winter nesting at-
tempt in South Carolina in December, 1998. The first
nest was found 1 year before previously reported fall
nestings in the Southeast and 1 month earlier in the
season. The second is the most northern and western
fall nesting site in the South, and the third is the latest
reported nesting attempt in the southern United States.
Taken together with 3 other reported successful fall
nests in Mississippi and South Carolina, Killdeer
would appear to be the only fall breeding shorebird in
North America and, based on those 6 widely-scattered
observations over the last 1 1 years, should now be
considered a rare fall and winter breeder across the
southern United States. Received 24 Nov. 1998, ac-
cepted 31 March 1999.

Although an anomalous report of breeding
in November exists from Michigan in 1982
(Tessen 1983), Jackson and coworkers (1995)
were the first to document fall and winter
breeding by Killdeer {Charcidrius vociferus)

' Dept, of Biological Sciences, Univ. of Arkansas,
Fayetteville, AR 72701; E-mail:
kgsmith@comp.uark.edu

2 The School of Pharmacy, Univ. of Mississippi,
University, MS 38677.

' Dept, of Microbiology & Immunology, Univ. of
Arkan.sas for Medical Sciences, Little Rock, AR
72205.

■’Charleston Museum, 360 Meeting St., Charleston,
SC 29403.

■’ 7666 Chippendale Rd., North Charleston, SC
29420.

’’ Corresponding author.

in the southeastern United States, reporting 1
set of chicks and adults on 16 November and
another set on 1 1 December 1988 in Okibbeha
Co., Mississippi. Subsequently, Post (1996)
reported 3 downy young, apparently 1-2 days
old, taken to a veterinarian in Berkeley Co.,
South Carolina on 13 November 1995. Here
we report on two more successful fall nesting
attempts by Killdeer in the south: one from
Mississippi that is earlier than observations by
Jackson and coworkers (1995), and one from
central Arkansas, the most northern and west-
ern fall nesting site yet reported in the South.
We also document a mid-December winter
nesting attempt in South Carolina, which is
the latest nesting activity yet reported.

On 7 November 1987, W.M.D. found and
photographed a pair of adults with one chick,
which appeared to be several days old, at the
wastewater treatment plant lagoon in Wave-
land, Hancock Co., coastal Mississippi. The
race-track shaped lagoon was surrounded by
a 4. 5-9. 2 m raised strip of excavated soil,
which varied from well-grassed to almost bare
areas, one of which was evidently chosen for
nesting by the Killdeer.

On 5 October 1998, T.E.K. and his wife dis-
covered a nest with four eggs located in a
stone area on the barrier of the parking lot at
the Veterans Administration Hospital in Little
Rock, Pulaski Co., Arkansas. During daily ob-
servations, two birds were usually present and
the female was observed incubating during
the day. They found two chicks on 26 October
and a third on 28 October. The nest was aban-

SHORT COMMUNICATIONS

425

doned with one egg remaining on 29 October.
Subsequent analysis determined that the egg
was fertile, but did not hatch.

On 5 December 1998, R.W.C. found a Kill-
deer nest containing four eggs in North
Charleston, Charleston Co., South Carolina.
The nest was a shallow depression located on
an approximately 0.2 ha lawn covered with
short (2-3 cm) grass. The nest was about 10
m from a frequently traveled road in a U.S.
Post Office complex. He made repeated visits
to the site, and found adults incubating daily
during 5-15 December. On 16 December, af-
ter arrival of a cold front on the coast, no Kill-
deer were seen in the area, and Chinn con-
cluded that the adults had abandoned the nest.
On 17 December at 15:00 EST, W.P. checked
the area, and finding no Killdeer, collected the
four eggs (ChM # 1998.11.50a-d), which
were intact but cold. The heaviest egg
weighed 11.7 g; the lightest, 10.8 g. Dimen-
sions of the longest egg were 40.2 X 25.6
mm; the shortest, 37.9 X 26.5 mm. No egg
had a discernible embryo, although the con-
tents of all appeared to be fresh.

Killdeer would appear to be the only fall
nesting shorebird in North America. There are
apparently no records of Killdeer breeding in
Mexico (P. Escalante, A. T. Peterson, pers.
comm.), and no evidence of breeding later
than July in southern California based on
clutches in the collection at the Western Eoun-
dation of Vertebrate Zoology (M. Marin, pers.
comm.). However, based on breeding records
from October, Eebruary, May, and August,
Robertson (1962) concluded that Killdeer
breed throughout the year in the Caribbean.
Schardien (1981) determined that Killdeer
have year-round territories in Mississippi and
copulations have been observed during winter
months (Jackson and Jackson, in press). Thus,
those six widely scattered records of breeding
over the last 1 1 years would suggest that the
Killdeer should now be considered a rare fall
and winter breeder in southern United States.

Incidental fall breeding in temperate re-
gions has been documented for a wide vari-
ety of birds (e.g., Orians 1960), but repeated
fall (or winter) breeding seems to be trig-
gered by either an appropriate stimulus ap-
pearing naturally, e.g., green cones stimulat-
ing breeding in Pinyon Jays (Gymnorhinus
cyanocephalus; Ligon 1978), or an appropri-

ate stimulus occurring during the wrong sea-
son, e.g.. Tricolored Blackbirds (Agelaius tri-
color) breeding in fall in response to flooding
of rice fields (Orians 1960). Those recent ob-
servations of fall breeding by Killdeer could
be due to a combination of events, including
more birdwatchers being active during fall
and early winter in areas of the South than
in previous times. However, the most likely
explanation is the unusually warm years that
have occurred over the last decade or so
(Mann et al. 1998). Jackson and coworkers
(1995) noted that their observations followed
a summer drought and mild fall weather,
which is similar to the situation in Arkansas
during 1998. The period from May through
September of 1998 was the hottest on record
at Little Rock, with 18 days above 37.8° C,
59 days above 35° C, and 111 days above
32.2° C and below average rainfall (National
Weather Service, North Little Rock, Arkan-
sas). Warm fall weather may be extending the
breeding season into fall, and even winter,
months, but there are apparently few, if any,
breeding attempts reported for late August
and September. More likely, extended warm
fall weather stimulates Killdeer to resume
breeding, as they typically are very early
spring breeders in the south [e.g., nests found
in Eebruary of 1999 in Louisiana (W. M.
D.)].

ACKNOWLEDGMENTS

J. and B. J. Jackson supplied unpublished informa-
tion on Killdeer; A. Jobes, H. Parker. R. Payne, and
R. E. Ricklefs supplied important references; and C. E
Bailey examined the Arkansas egg. E. Nol and P. W.
Bergstrom made helpful suggestions as reviewers. This
is Contribution No. 2 from the Arkansas Breeding Bird
Atlas.

LITERATURE CITED

Jackson, B. J. S. and J. A. Jackson. In press. Killdeer
(Cluiradrius vociferus). In The birds of North
America (A. Poole and E Gill. Eds.). The Acad-
emy of Natural Sciences. Philadelphia, Pennsyl-
vania; The American Ornithologists' Union,
Washington, D.C.

Jackson, J. A., M. E Hodoes, D. J. Ingold, and B. J.
S. Jackson. 1995. Fall nesting of Killdeers in Mis-
sissippi. Miss. Kite 25:16—17.

Eicon, J. D. 1978. Reproductive interdependence of
Pifion Jays and pinon pines. Ecol. Monogr. 48:
I I 1-126.

Mann, M. E., R. S. Bradley, and M. K. Hughes.

426

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

1998. Global-scale temperature patterns and cli-
mate forcing over the past six centuries. Nature
392:779-787.

Orians, G. H. 1960. Autumnal breeding in the Tricol-
ored Blackbird. Auk 77:379-398.

Post, W. 1996. Late autumnal breeding by Killdeer.
Ela. Eield Nat. 24:109.

Robertson, W. B., Jr. 1962. Observations of the birds
of St. John, Virgin Islands. Auk 79:44-76.
ScHARDiEN, B. J. 1981. Behavioral ecology of a south-
ern population of Killdeer. Ph.D. diss., Mississippi
State Univ., Mississippi State.

Tessen, D. D. 1983. Western Great Lakes region. Am.
Birds 37:182-185.

Wilson Bull.. 1 1 1(3), 1999, pp. 426-427

Wild Turkeys {Meleagris gallopavo) Renest After Successful Hatch

Craig A. Harper' --^ and Jay H. Exurn^"*

ABSTRACT. — Wild Turkey [Meleagris gallopavo)
hens frequently renest after disturbance on the nest,
especially while laying or early during the incubation
period. However, no record exists of Wild Turkey hens
renesting after a successful hatch. We document three
Wild Turkey hens that renested after having hatched a
brood. None of the renests were successful. Received
2 Dec. 1998, accepted 6 March 1999.

Nesting success of Wild Turkeys {Melea-
gris gallopavo) varies widely across their
range and is influenced by many factors (Van-
glider 1992). A successful nest generally is
defined as one in which at least one poult
hatches. Researchers have documented many
instances of Wild Turkey hens renesting after
their initial nest was disturbed or depredated
(Everette et al. 1980, Williams et al. 1980,
Vangilder et al. 1987), most often while laying
or during early incubation (Williams et al.
1976). Causes for nest failure and subsequent
renesting include nest destruction by predators
(Speake 1980, Vander-Haegen et al. 1988), se-
vere weather (Roberts and Porter 1998, Kim-
mel and Zwank 1985), and disturbance by re-
searchers (Still and Baumann 1990). Renest-
ing after a successful nest was not thought to

' Dept, of Eorcst Resources, Clcmson Univ., Clem-
son. SC 29634; E-mail: charpei@utk.edu

- Pre.sent address: Dept, of Forestry, Wildlife and
Fisheries. P.O. Box 1071, Univ. of Tennessee, Knox-
ville. TN 37901.

' Dept, of Zoology-Entomology, Auburn Univ., Au-
burn, AL 36849.

■' Present address: Glatting, Jackson, Kcrcher, An-
glin. Lopez and Rinehart, Orlando, EL 32801.

■' Corresponding author.

occur. In his Book of the Wild Turkey, Wil-
liams (1981:53) stated, “No example is
known of a hen nesting again in the same year
after her brood hatched, and there has been no
reported case of a turkey hatching two broods
in one year.” Below, we document three cases
of hens renesting after having successfully
nested but with early loss of broods.

In 1983, while working in southern Ala-
bama, J.H.E. found a Wild Turkey hen that
renested three times after hatching a brood
that did not survive more than two days. This
particular hen hatched all 1 1 eggs in her initial
clutch after a normal incubation period (28
days). On the day after hatching, five of the
poults were found dead in the nest from un-
known causes. The hen had no poults with her
two days after hatching, and the fate of the
remaining six poults was never determined.
Twenty-five days after hatching, the hen re-
nested. This nest was disturbed by investiga-
tors, which prompted the hen to abandon that
nest and eventually renest two more times.
None of the renests was successful.

In the southern Appalachians of North Car-
olina, C.A.H. monitored two wild turkey hens
that renested after hatching clutches of 1 1 and
14 eggs. These nests were initiated in early
April of 1996, and incubated 29 and 27 days.
Both broods were killed within five days. One
hen initiated a second nest 17 days after her
initial clutch hatched. This renest contained
nine eggs that were incubated 65 days, 37
days beyond the normal 28-day incubation pe-
riod. Subsequently, this nest was abandoned,
and the eggs were determined to be infertile.

SHORT COMMUNICATIONS

427

The second hen began incubating another nest
27 days after her initial clutch hatched. This
renest was incubated for 1 1 days and aban-
doned for unknown reasons.

None of the renests we documented was
successful, either because of infertility or nest
abandonment. Ultimately, infertility could be
a primary factor limiting success of late re-
nests. Sperm may be contained in a hen’s ovi-
duct up to 56 days after copulation (Blanken-
ship 1992) and remain viable for a ‘normal’
renesting attempt. However, the renests we
documented were initiated considerably later
than most renests because they occurred after
an entire incubation period plus some addi-
tional days. Thus, a hen might need to copu-
late again in order to lay fertile eggs two
months after the primary mating season (i.e.,
April-May). However, copulation in July
would be exceptional because the urge to
breed, which is associated with the rise in tes-
tosterone (for males) and prolactin (for fe-
males) levels, is regressing (Blankenship
1992).

The ability of Wild Turkey hens to renest
after having hatched a brood did not contrib-
ute to productivity in these cases. However,
we now know it is possible for hens to renest
after a successful nesting attempt and it is
conceivable that such nesting could be suc-
cessful.

LITERATURE CITED

Blankenship, L. H. 1992. Physiology. Pp. 84-100 in
The Wild Turkey: biology and management (J. G.

Dickson, Ed.). Stackpole Books, Harrisburg,
Pennsylvania.

Everetth, D. D., D. W. Speake, and W. K. Maddox.
1980. Natality and mortality of a north Alabama
Wild Turkey population. Proc. Natl. Wild Turkey
Symp. 4:1 17-126.

Kimmel, E G. and P. J. Zwank. 1985. Habitat selection
and nesting responses to spring flooding by east-
ern Wild Turkey hens in Louisiana. Proc. Natl.
Wild Turkey Symp. 5:155-171.

Roberts, S. D. and W. E Porter. 1998. Relation be-
tween weather and survival of Wild Turkey nests.
J. Wildl. Manage. 62:1492-1498.

Speake, D. W. 1980. Predation on Wild Turkeys in
Alabama. Proc. Natl. Wild Turkey Symp. 4:86-
101.

Still, H. R., Jr. and D. P. Baumann, Jr. 1990. Wild
Turkey nesting ecology on the Erancis Marion Na-
tional Forest. Proc. Natl. Wild Turkey Symp. 6:
13-17.

Vander-Haegen, W. M., W. E. Dodge, and M. W.
Sayre. 1988. Factors affecting productivity in a
northern Wild Turkey population. J. Wildl. Man-
age. 52:127-133.

Vangilder, L. D. 1992. Population dynamics. Pp.
144-164 in The Wild Turkey: biology and man-
agement. (J. G. Dickson, Ed.). Stackpole Books,
Harrisburg, Pennsylvania.

Vangilder, L. D., E. W. Kurzejeski, V. L. Kimmel-
Truitt, and j. B. Lewis. 1987. Reproductive pa-
rameters of Wild Turkey hens in north Missouri.
J. Wildl. Manage. 51:535-540.

Williams, L. E., Jr. 1981. The book of the Wild Tur-
key. Winchester Press, Tulsa, Oklahoma.

Williams, L. E., Jr., D. H. Austin, and T. E. Peoples.
1976. The breeding potential of the Wild Turkey
hen. Proc. Annu. Conf. Southeast. Assoc. Fish
Wildl. Agenc. 30:371-376.

Williams, L. E., Jr., D. H. Austin, and T. E. Peoples.
1980. Turkey nesting success on a Florida study
area. Proc. Natl. Wild Turkey Symp. 4:102-107.

428

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

Wilson Bull., 111(3), 1999, pp. 428-432

Post-migration Weight Gain of Swainson’s Hawks in Argentina

Michael I. Goldstein,' Peter H. Bloom,- Jose H. Sarasola,^ and Thomas E. Lacher'

ABSTRACT. — Swainson’s Hawks (Biiteo swain-
soni\ Aguilucho Langostero) were captured and band-
ed in La Pampa, Argentina from 28 November 1996
through 25 January 1997. We collected morphometric
measurements to determine if hawks gained weight on
the austral (non-breeding) grounds. Hawks in appar-
ently healthy condition weighed 819.7 ± 1 1.1 g (mean
± SE, n = 127, range 540-1090 g). Weight increased
significantly over the length of the non-breeding sea-
son in Argentina (P = 0.0059), but wingspan (124.9
± 0.5 cm, n = 127, range 105.7-137.1 cm) and tail
(19.8 ± 0.1 cm, n = 127, range 16.5-23.1 cm) did
not. When separated by age, weight increased signifi-
cantly for juveniles (P = 0.0083) but was marginally
non-significant for adults (P = 0.0555). Received 15
Oct. 1998, accepted 15 Feb. 1999.

Swainson’s Hawks (Buteo swainsoni) are
long-distance migrants that travel between the
plains, shrublands, and pampas (grasslands) of
North and South America. During the south-
ward migration, some hawks temporarily stop
in Central America or in the agricultural zones
of Mexico (Ridgely and Gwynne 1989, Stiles
and Skutch 1989, Howell and Webb 1995).
Occasionally some hawks are found in Florida
(Robertson and Woolfenden 1992) and others
winter in California (Herzog 1996). The pam-
pas of Argentina are a major non-breeding
(austral) destination, supporting much of this
species’ population from mid-November
through early March. Two satellite radio-
tagged hawks trapped by Brian Woodbridge
in California in 1994 were tracked to La Pam-
pa province, Argentina (Woodbridge et al.

1995). Large numbers of hawks were subse-
quently found in La Pampa, Buenos Aires,

' Texas A&M Univ.. Dept, of Wildlife and Fisheries
Scienees, 210 Nagle Hall, College Station, TX 77843-
2258.

2 Western Foundation of Vertebrate Zoology, 439
Calle San Pablo, Camarillo, CA 93010.

^ Univ. Naeional de La Pampa, Facultad de Cieneias
Exactos y Naturales, Uruguay 151, (6300) Santa Rosa,
La Pampa, Argentina.

■’ Corresponding author;

E-mail; mgold.stein@tamu.edu

and Cordoba provinces (Goldstein et al.

1996) .

Adults feed primarily vertebrates to nes-
tlings (England et al. 1997). Summer flocks of
non-breeding birds eat a more varied diet, in-
cluding insects (Johnson et al. 1987). On the
non-breeding grounds, hawks are generally
seen in large aggregations: roosting, foraging,
and traveling together. In the pampas, the
hawks primarily eat invertebrates (White et al.
1989, Jaramillo 1993, Woodbridge et al. 1995,
Goldstein et al. 1996). The dietary shift allows
for more birds to be supported per unit area.
In La Pampa, flocks of birds following grass-
hopper (Orthoptera) outbreaks have been re-
ported (Rudolph and Fisher 1993). We found
flocks as large as 12,000 birds (Goldstein

1997) .

Long-distance migration using only stored
fat has been suspected but not documented
(Smith et al. 1986). Whether hawks forage or
fast en route to Argentina, or whether specific
stopover habitats are regularly used is not
known (Goldstein and Smith 1991, Kirkley
1991). The extent of predation on airbom
dragonflies and other flying insects during mi-
gration is also unknown. Nevertheless, if mi-
grating hawks used only stored fat they might
arrive in the pampas in poor condition (Smith
et al. 1986). Hawks have been reported arriv-
ing in Argentina in such weak condition that
they were picked up by hand (C. C. Olrog in
Smith 1980). None of these birds, however,
were checked for contaminants.

Swainson’s Hawks are found in Argentina
from early November through mid-February,
although later arrivals and earlier departures
have been documented (England et al. 1997).
Substantial periods of fasting during migration
would result in a substantial loss of weight,
weakened condition, and a subsequent in-
crease in weight when on the wintering
grounds. From November 1996 through Jan-
uary 1997, we captured and measured Swain-
son’s Hawks on the non-breeding grounds. We

SHORT COMMUNICATIONS

429

present these data and observations regarding
Swainson’s Hawk austral weight gain.

METHODS

The research area in northern La Pampa (35° 14' S,
63° 57' W, 149 m ASL) is a flat grassland dominated
by row crop agriculture. One hundred and twenty-eight
Swainson's Hawks were captured and sampled in La
Pampa from 28 November 1996 through 25 January
1997. Most hawks were trapped by bal chatri traps in
open fields or on dirt roads between foraging fields
(Bloom 1987). Nine of the 128 hawks were captured
by hand after they were grounded from their nighttime
Eucalyptus viminalis roost by a thunderstorm. All
hawks were captured in the morning between 05:15
and I 1 :00 local time.

Hawks were banded and weighed to the nearest
gram with a 1500 g Pesola scale. Wingspan and tail
length were recorded to the nearest mm. Hawks were
classified as juveniles or adults based on plumage, with
immatures grouped as juveniles (Wheeler and Clark
1995). A regression analysis was used to analyze
changes in weight, wingspan, and tail length for the
population and for each age category. Time was ex-
pressed in number of trap days from Day 1 (28 No-
vember 1996). Because sexes could not be distin-
guished morphologically, we did not analyze the data
by sex (Wheeler and Clark 1995). Statistical analyses
were performed using SPSS 8.0 for Windows (SPSS
Inc., Chicago, Illinois).

RESULTS

We collected complete data sets for 127
healthy hawks. One individual that might have
been exposed to pesticides was excluded
(Goldstein 1997). Weight averaged 819.7 g [±
11.1 (SE), range 540-1090 gj. Wingspan
measured 124.9 cm (± 0.5, range 105.7-137.1
cm), and tail length measured 19.8 cm (± 0.1,
range 16.5-23.1 cm). Regression analysis in-
dicated that weight increased significantly
over time (R^ = 0.0705, P = 0.0059), but nei-
ther wingspan (R^ = 0.0001, P > 0.05) nor
tail length (R^ = 0.0078, P > 0.05) increased
during the season. This pattern suggested an
increase in mass from late November until late
January without structural growth.

There were 65 adult, 58 juvenile, and 4 un-
known aged birds. Adults weighed 836.6 g (±
16.5, range 560-1090 g). Juveniles weighed
794.2 g (± 15.9, range 540-1080 g). Weight
significantly increased over time for Juveniles
{P = 0.0083, df = 57, R2 = 0.1 179; Fig. 1 A),
but was marginally non-significant for adults
(P = 0.0555, df = 64, R^ = 0.0570; Fig. IB).
Although both age categories showed weight

gains, the larger slope for juveniles indicated
they gained weight at a faster rate than adults.
When examined by age, neither wingspan nor
tail length changed over time (P > 0.05 for
all).

Wingspan was significantly and positively
correlated with tail length (r = 0.509; P <
0.05) and with weight (r = 0.342; P < 0.05).
Weight and tail length showed no correlation
(r = -0.048; P > 0.05).

DISCUSSION

Swainson’s Hawks remain in Argentina
from mid-November through mid-February
(Houston and Schmutz 1995, England et al.
1997). When we began following large flocks
on 20 November 1996 we saw no signs of
birds in poor condition. Whether birds arrived
in Argentina in poor condition is uncertain,
but neither our observations nor conversations
with farmers supported that conjecture. Al-
though hawks were not in poor condition
when captured in late November, they were
somewhat lighter than those measured on the
breeding grounds. Breeding males weighed
808 g (range 693-936 g, n = 69) and females
weighed 1109 g (range 937-1367 g, n = 50),
for an overall mean of 934 g (J. K. Schmutz
in England et al. 1997). Although hawks cap-
tured in Argentina averaged 819.7 g, birds
captured on 28 November- 17 December 1996
were notably lighter (779.1 g) than the mean.
Hawks captured from Days 21-59 weighed

860.5 g. Both adults and juveniles, when ex-
amined separately, show similar patterns of
weight for Days 1-20 (adults 814.8 g, juve-
niles 737.0 g) and Days 21-59 (adults 872.5
g, juveniles 847.7 g). This pattern suggests
that both adult and juvenile hawks lost weight
during migration and then regained it during
their stay in Argentina, although the trend for
juveniles was stronger. However, if the hawks
foraged in the northern pampas for several
days or weeks prior to their arrival near the
La Pampa field sites, our measure of weight
loss is underestimated.

Rectrix lengths were consistent with those
reported from North America (England et al.
1997). Alberta males averaged 18.4 cm (range
17.0-19.8 cm, /? = 61) and females averaged

20.5 cm (range 19.3-22.1 cm, n = 43) for an
overall length of 19.3 cm (J. K. Schmutz in
England et al. 1997). We found no wingspan

430

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

1200

1000

800

% 600
O
§

400

200

0 10 20 30 40 50 60 70

Time (trapday)

FIG. 1. Swainson’s Hawk juvenile (A) and adult (B) austral weights from 28 November 1996 through 25
January 1997 as measured in northern La Pampa, Argentina. Juvenile weight gain was significant (Y = 2.3972x
+ 736.65, = 0.1 179, n = 58, P = 0.0083). Adult weight gain was marginally non-significant (Y = L8196x
+ 806.89, R2 = 0.0570, n = 65, P = 0.0555).

data for North American hawks in the litera-
ture. In addition, hawks were molting
throughout the non-breeding season in Argen-
tina, a phenomenon that has not been studied.

We captured only one hawk that was light
weight and appeared weak. The hawk
weighed 560 g when trapped on 25 November
1996. It gained 100 g in 5 days of captivity
and was subsequently released. Whether this
bird was weak from migration or had been

exposed to toxins is unknown. Although no
chemical residues were found on either foot-
wash or feather residue samples, cholinester-
ase measurements from blood plasma samples
taken at the time of capture were 30% below
those taken at the time of release (Goldstein
1997). Exposure to organophosphate insecti-
cides decreases plasma cholinesterase activity
levels (Hill and Fleming 1982).

Large flocks foraging on the ground during

SHORT COMMUNICATIONS

431

migration were thought unlikely in the past
(Smith et al. 1986), but such behavior may not
be uncommon (see England et al. 1997, Gold-
stein 1997). Insect outbreaks in agricultural
grasslands may provide terrestrial foraging
opportunities along the migration corridor. Re-
cently discovered populations wintering in
southwestern Mexico confirm the need for
further study of stopover habitats. In addition,
hawks are known to forage aerially for grass-
hoppers and dragonflies across the pampas of
Argentina (White et al. 1989, Jaramillo 1993,
Rudolph and Fisher 1993, Woodbridge et al.
1995, Goldstein 1997). We have no idea what
altitudes are reached while traveling on ther-
mal air currents across the insect-rich tropical
and sub-tropical rainforests, nor whether any
appreciable distance is ever achieved without
aid of thermals. We do not know whether ae-
rial insects are in adequate supply at the
heights of travel. The question of the duration
and extent of fasting of Swainson’s Hawks
during migration needs further study although
lengthy periods of fasting seem unlikely given
the weights of November birds captured in
Argentina.

ACKNOWLEDGMENTS

M. I. Goldstein wishes to express thanks to the
Graduate Program Enhancement Fund of the Depart-
ment of Wildlife and Fisheries Sciences of Texas
A&M University for financial support, enabling pre-
sentation of this work at the annual meeting of the
Raptor Research Foundation in Ogden, Utah in Sep-
tember 1998. We wish also to thank: A. Lanusse and
S. Salva who hosted our stay at Estancia Chanilao in
La Pampa; B. Woodbridge and M. Bechard who pro-
vided financial assistance and technical guidance; M.
Parker, M. Bechard, M. Kochert, S. Canavelli, S. Bak-
er, S. Weidensaul, and the Lanusse family for trapping
hawks; and M. E. Zaccagnini and J. L. Panigatti for
logistical support. This project was funded in part by
the Hawk Mountain-Zeiss Raptor Research Award,
the Frank M. Chapman Memorial Fund, U.S. Fish and
Wildlife Service (Office of International Affairs), No-
vartis Crop Protection, and the Archbold Tropical Re-
search Center. P. H. Bloom thanks B. Woodbridge and
the U.S. Forest Service for facilitating travel to Ar-
gentina. A. Palka, C. S. Houston, A. Jaramillo, and an
anonymous reviewer provided critical reviews of ear-
lier versions of the manuscript.

LITERATURE CITED

Bloom, P. H. 1987. Capturing and handling raptors.
Pp. 99—123 in Raptor management techniques
manual (B. A. G. Pendleton, B. A. Milsap, K. W.

Cline, and D. M. Bird, Eds.). National Wildlife
Federation. Wa.shington, D.C.

England, A. S., M. J. Bechard, and C. S. Hou,ston.
1997. Swain.son's Hawk (liiileo swainsoni). In The
birds of North America, no. 265 (A. Poole and F.
Gill Eds.). The Academy of Natural Sciences,
Philadelphia, Pennsylvania; The American Orni-
thologists’ Union, Washington, D.C.

Goldstein, D. L. and N. G. Smith. 1991. Response to
Kirkley. J. Raptor Res. 25:87-88.

Goldstein, M. I. 1997. Toxicological assessment of a
neotropical migrant on its non-breeding grounds:
case study of the Swainson’s Hawk in Argentina.
M.Sc. thesis, Clemson Univ., Clemson, South
Carolina.

Goldstein, M. I., B. Woodbridge, M. E. Zaccagnini,
S. B. Canavelli, and A. Lanusse. 1996. An as-
sessment of mortality of Swainson’s Hawks on
wintering grounds in Argentina. J. Raptor Res. 30:
106-107.

Herzog, S. K. 1996. Wintering Swainson’s Hawks in
California’s Sacramento-San Joaquin River Delta.
Condor 98:876-879.

Hill, E. F. and W. J. Fleming. 1982. Anticholinester-
ase poisoning of birds: field monitoring and di-
agnosis of acute poisoning. Fnv. Toxicol. Chem.
1:27-38.

Houston, C. S. and J. K. Schmutz. 1995. Swainson’s
Hawk banding in North America to 1992. N. Am.
Bird Bander 20:120-127.

Howell. S. N. G. and S. Webb. 1995. A guide to the
birds of Mexico and northern Central America.
Oxford Univ. Press, New York.

Jaramillo, A. P. 1993. Wintering Swainson’s Hawks
in Argentina: food and age segregation. Condor
95:475-479.

Johnson, C. G., L. A. Nickerson, and M. J. Bechard.
1987. Grasshopper consumption and summer
flocks of non-breeding Swainson’s Hawks. Con-
dor 89:676-678.

Kirkley, J. S. 1991. Do migrant Swainson’s Hawks
fast en roitte to Argentina? J. Raptor Res. 25:82-
86.

Ridgely, R. S. and j. a. Gwynne. Jr. 1989. A guide
to the birds of Panama with Costa Rica, Nicara-
gua, and Honduras, second ed. Princeton Univ.
Press, Princeton, New Jersey.

Robertson. W. D., Jr. and G. E. Woolfenden. 1992.
Florida bird species: an annotated list. Florida Or-
nithological Society, Cocoa. Florida.

Rlidolph, D. C. and C. D. Fisher. 1993. Swainson’s
Hawk predation on dragonflies in Argentina. Wil-
son Bull. 105:365-366.

Smith. N. G. 1980. Hawk and vulture migrations in
the Neotropics. Pp. 51-65 in Migrant birds in the
Neotropics: ecology, behavior, distribution, and
conservation (A. Keast and F. S. Morton. Eds.).
Smithsonian Institution Press, Washington, D.C.

Smith, N. G., D. L. Goldstein, and G. A. Barthol-
omew. 1986. Is long-distance migration possible
using only stored fat? Auk 103:607-61 1.

432

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

Stiles, E G. and A. E Skutch. 1989. A guide to the
birds of Costa Rica. Cornell Univ. Press, Ithaca,
New York.

Wheeler, B. K. and W. S. Clark. 1995. A photo-
graphic guide to North American raptors. Aca-
demic Press, San Diego, California.

White, C. M., D. A. Boyce, and R. Straneck. 1989.
Observations on Buteo .swainsoni in Argentina,

1984, with comments on food, habitat alteration
and agricultural chemicals. Pp. 79—87 in Raptors
in the modern world (B. U. Meyburg, and R. D.
Chancellor, Eds.). World Wildlife Group on Birds
of Prey, London, U.K.

WOODBRIDGE, B., K. K. PiNLEY, AND S. T. SEAGER.
1995. An investigation of the Swainson’s Hawk
in Argentina. J. Raptor Res. 29:202-204.

Wihon Bull., 111(3), 1999, pp. 432-436

Siblicide at Northern Goshawk Nests: Does Food Play a Role?

Wendy A, Estes,' - Sarah R, Dewey, and Patricia L. Kennedy'

ABSTRACT. — Siblicide as a mechanism for brood
reduction has been reported in a number of asynchro-
nously hatching bird species. Although researchers
have documented the occurrence of facultative sibli-
cide in several raptor species, its cause is still debated.
Most hypotheses relate incidences of siblicide to food
availability. The food-amount hypothesis predicts a
negative relationship between the amount of food
available and nestling aggression. While the food-
amount hypothesis has received much attention, few
studies show more than correlational support for this
activity in raptors. Our observation of a siblicide event
at a Northern Goshawk (Accipiter gentili.s) nest used
as a control in a supplemental feeding experiment, and
a similar incident where a nestling goshawk's death
can be attributed to siblicide provide support for the
negative correlation between food amount and sibling
aggression. These observations and the lack of any re-
ported sibling aggression at seven supplementally fed
nests showing extreme hatching asynchrony also in-
dicate a relationship between food resources and brood
reduction. Our observations are consistent with the
idea that goshawks exhibit facultative siblicide, and
that resource levels as predicted by the food-amount
hypothesis directly influence it. Received 7 Oct. 1998,
accepted 16 Feb. 1999.

Hatching asynchrony in birds facilitates
brood reduction because the last hatched nest-

' Dept, of Eishery and Wildlife Biology, Colorado
State Univ., Eort Collins, CO 80523.

2 Pre.scnt address: 104 Biological Sciences East,
Univ. of Arizona, Tuc.son, AZ 85721.

’ Present address: USDA Forest .Service, Ashley Na-
tional Forest, Vernal Ranger District, 355 N. Vernal
Ave., Vernal, UT 84078.

•' Corresponding author;

E-mail: patk@cnr.colostate.edu

ling is at a competitive disadvantage if re-
sources provided by the parents prove inade-
quate. In asynchronously hatched broods, the
youngest nestling occasionally dies from ag-
gressive sibling behavior including pecking,
exclusion during feeding bouts, or eviction
from the nest (Lack 1954, Mock et al. 1990,
Creighton and Schnell 1996). Asynchronous
hatching results in adapting a brood size to an
unpredictable food supply by allowing all
young to survive when food is plentiful, but
ensuring brood reduction to match parental
provisioning capabilities when prey levels are
meager (Lack 1954, Newton 1979, Bryant and
Tatner 1990, Heeb 1994). Species in which
the frequency of siblicide events are variable
are termed facultative, while those in which
siblicide occurs in nearly all nest attempts are
called obligate (Edwards and Collopy 1983).
Although the occurrence of obligate siblicide
appears to be largely innate (Mock et al. 1990,
Gerhardt et al. 1997), the causes of facultative
siblicide are still debated (Forbes and Mock
1994).

Fatal sibling aggression has been docu-
mented in a range of avian species (Stinson
1979, Braun and Hunt 1983, Anderson 1989,
Drummond and Garcia Chavelas 1989, Bryant
and Tatner 1990, Mock et al. 1990; Mock and
Lamey 1991, Heinsohn 1995, Reynolds
1996). However, an understanding of the
proximate factors that influence the occur-
rence of facultative siblicide remains elusive
because such events are rare and unpredict-
able. Most similar hypotheses attempt to ex-
plain facultative siblicide in relation to food.

SHORT COMMUNICATIONS

433

The food-amount hypothesis predicts a nega-
tive relationship between the amount of food
available and nestling aggression (Mock et al.
1987, Creighton and Schnell 1996). By killing
its sibling when food is scarce, a nestling may
increase its chance of survival by increasing
its share of food delivered to the nest. Mock
and coworkers (1990) found that smaller food
morsels can be monopolized through combat
and, therefore, reward sibling aggression.
Higher rates of aggression were observed in
larger broods of Cattle Egrets (Bubulcus ibis)
where individual food portions are expected
to be smaller (Mock et al. 1987).

Because siblicide events are uncommon,
and tend to go unwitnessed unless nests are
under constant watch, few studies have estab-
lished a causal link between food resources
and fatal sibling aggression (but see Mock et
al. 1987). Facultative siblicide has been doc-
umented in several raptor species, but its
cause has not been fully investigated (Schnell
1958, Pilz and Seibert 1978, Newton 1979,
Bechard 1983, Zachel 1985, Bortolotti et al.
1991, Boal and Bacorn 1994). Although the
hypothesis that food supplies influence sibling
aggression is intuitively appealing, few stud-
ies, with the exception of Wiebe and Borto-
lotti (1995) and Wellicome (1997), have pro-
vided more than correlational support for this
activity in raptors.

Northern Goshawks (Accipiter gentilis)
hatch asynchronously and exhibit siblicide
(Newton 1979, Stinson 1979). Observational
accounts of siblicide in goshawks are rare
(Schnell 1958, Zachel 1985, Boal and Bacorn
1994), and its occurrence is thought to be lim-
ited to times when food is in very low supply
(Newton 1979). Experimental data linking
food resources and incidence of siblicide in
this species are nonexistent. In this paper, we
report the occurrence of a siblicide event at a
goshawk nest in northeastern Utah. We also
describe another incident in which a nestling’s
death was likely the result of siblicide and we
provide experimental evidence that the inci-
dence of sibling aggression may be related to
food supplies. Our observations are consistent
with the hypothesis that goshawks exhibit fac-
ultative siblicide directly influenced by food
resources.

Our siblicide observations occurred at nests
that were part of a study on the influence of

food provisioning on female nest attendance
and nestling begging vocalizations. This in-
vestigation was part of a larger experiment ex-
amining the influence of supplemental food on
parental care strategies and juvenile survival
(Dewey 1999). In 1997, 14 nests were includ-
ed in the food supplementation experiment
(experimental design similar to that of Ward
and Kennedy 1996). Seven of these nests were
randomly assigned as treatments and were
provided Japanese Quail (Coturnix coturnix)
from hatching through the fledgling depen-
dency period. We visited treatment nests every
two to three days and provided sufficient food
to meet the energy requirements of the female
and young until the next scheduled visit (see
Dewey 1999 for details). Control nests were
visited at the same interval and for the same
amount of time, but were not given food. The
nest attendance/vocalization study was con-
ducted from mid-June to mid-July 1997 and
consisted of a subset of the nests used in the
food supplementation experiment. Each nest
was observed for a 3 h period on three dif-
ferent occasions from a portable blind located
approximately 30 m from the nest. Observa-
tion times were rotated to include both morn-
ings and afternoons.

The first event was witnessed during a 3 h
observation period on the afternoon of 10 July
at control nest LGD. The adult female had not
been observed in the nest stand since 8 July,
and likely had deserted the nest or died. Al-
though we were unable to trap the female to
verify her age prior to her disappearance, her
unusually dark maroon colored eyes, behavior,
and degree of scarring above and around her
right eye indicated old age. The role of the
male in caring for the nestlings after the fe-
male’s disappearance is unknown because he
was never observed visiting the nest.

Shortly after observations had begun, two
nestlings (21 and 22 days old) were begging
periodically. The older nestling (N 1 ) then be-
gan flapping its wings and pecking at the head
of nestling 2 (N2). Nestling 2 initially retali-
ated by flapping its wings and pecking at the
head of N 1 , but soon turned its back to N 1
and assumed the defense stance described by
Schnell (1958), with its head lowered and its
rump elevated. Nestling 1 responded by in-
creasing its intensity of pecking and then be-
gan pulling down out of N2’s thighs and

434

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

rump. Nestling 2 uttered a high-pitched call
and moved toward the edge of the nest. This
behavior continued as N2 was forced out of
the west side of the nest and onto the nest
branch. Nestling 1 perched on the edge of the
nest and continued pulling down from N2’s
rump until N2 moved out of reach, at which
point N 1 walked to the center of the nest and
rested.

After approximately 15 min, during which
several strong gusts of wind nearly blew N2
off the branch, N2 moved back into the nest;
N 1 resumed aggressive attacks within 5 min.
Nestling 2 again assumed a defense stance,
and N1 began tearing down from N2’s back-
side, forcing N2 onto the south edge of the
nest. Nestling 1 then began rushing at N2 and
colliding with N2’s hind end. This behavior
continued while N2 called and flapped its
wings in an attempt to maintain balance as it
clung to the rim of the nest. Nestling 2 then
turned quickly and climbed over N1 and into
the center of the nest. Nestling 1 pursued N2
to the east edge of the nest next to the tree
trunk, where N2 again took a defense stance.
Nestling 1 resumed ramming and tearing
down from N2. Nestling 2 was knocked out
of the nest but caught its wing on a branch.
Nestling 1 leaned out of the nest and contin-
ued to rip down from N2 while N2 screamed.
Nestling 1 then backed off, uttered an adult-
like alarm call, and returned to the center of
the nest. Nestling 2 climbed back into the nest
and remained in the nest for approximately 10
min until N1 again chased N2 out of the nest
and onto the nest branch. Nestling 2 was not
allowed back onto the nest for the rest of the
observation period.

Two days later we found N2 dead on the
ground under the nest. Nestling 1 directed
loud alarm and begging calls at us throughout
the visit to the nest stand. Nestling 1 was
found dead in the nest on 14 July 1997. Bod-
ies of both nestlings were sent to the Colorado
Veterinary Diagnostic Laboratory at the Col-
lege of Veterinary Medicine, Colorado State
University, where necropsies were performed.
Nestling 2 was mildly emaciated, had two
fractured ribs, and pulmonary hemorrhaging,
presumably incurred during its fall from the
nest tree. The exact cause of death for N1 was
unknown; however, the necropsy showed this
bird suffered from advanced emaciation re-

sulting in pectoral muscle atrophy, which
strongly suggests starvation.

The other probable case of siblicide oc-
curred at the control nest SNK on 2 July 1997.
When we entered the nest stand the female
was not in the immediate vicinity. Two nest-
lings (20 and 22 days old) were in the nest;
one was obviously dead with blood around its
head. The adult female returned shortly there-
after, poked at the dead nestling briefly, and
then carried the body away from the nest. She
returned within several minutes without the
dead nestling. Although we did not witness
aggression between the siblings, the fact that
the dead nestling was still in the nest suggests
that a predator did not kill it. Because of the
obvious head injury, we believe the nestling’s
death resulted from siblicide and not merely
starvation. Although, it is possible that the re-
maining nestling attempted cannibalism after
its sibling had died, we did not observe the
nestling trying to feed on its dead sibling. The
SNK nestling’s death could also have been the
result of filial infanticide if parental behavior
(e.g., nest desertion, favoritism, or aggression)
contributed to its death (Mock and Parker
1997). However, O’Connor’s (1978) brood re-
duction model predicts that conditions favor-
able to siblicide will occur more often than
those favorable to filial infanticide. To our
knowledge, filial infanticide resulting from fa-
tal parental aggression has not been docu-
mented in goshawks; nevertheless, we cannot
rule it out as a possibility.

Several details of these observations dif-
fered from those of similar events observed in
goshawks and other raptors. Cannibalism was
documented to have followed siblicide in a
Swainson’s Hawk (Buteo swainsoni) nest (Pilz
and Seibert 1978), three Burrowing Owl
{Athene cuniciilaria) nests (Wellicome 1997),
four American Kestrel {Falco spar\>erius)
nests (Bortolotti et al. 1991), and three gos-
hawk nests (Schnell 1958, Zachel 1985, Boal
and Bacorn 1994). However, cannibalism was
not observed at either nest in this study. At
the LCD nest cannibalism might have oc-
curred if N2 had not fallen to its death; but
NTs behavior gave no indication that it was
attacking N2 for the purpose of consumption.
Nestling 1 seemed intent on expelling N2
from the nest and Nl’s aggression stopped
once N2 was out of the nest. If N1 was at-

SHORT COMMUNICATIONS

435

tempting to kill N2 for consumption we would
have expected the aggression to continue until
N2 was dead. Cannibalism was also not ob-
served at the SNK nest but this may have been
due to the presence of the adults that were
providing food to the remaining nestling, or
to the removal of the dead nestling before it
could be cannibalized.

A second disparity between our observa-
tions and those in the literature is the potential
function of the submissive posture of the de-
fense stance. According to Schnell’s (1958)
observations of nestling aggression, the ag-
gressor terminated attacks when its sibling as-
sumed the defense stance. In our observations,
N1 continued aggressive attacks after N2 as-
sumed the defense stance. Nestling I’s behav-
ior also differed from aggressor behavior in
other documented siblicide events in that N I’s
attacks were aimed primarily at N2’s rump
and thighs instead of at its sibling’s head
(Schnell 1958, Pilz and Seibert 1978, Boal
and Bacom 1994).

Although our LGD siblicide observation
differs in the aforementioned ways from those
previously reported by Schnell (1958), Zachel
(1985), and Boal and Bacom (1994), our
event is similar in that it occurred during a
period of apparent low food supply. We did
not measure food availability in our study
area, but provided half of our experimental
goshawks with supplemental food. Including
the LGD nest failure and the SNK mortality,
we documented brood reductions at four of
the seven control nests in 1997 and no nest-
ling deaths at any of the treatment nests. Three
of the seven (43%) control nests failed (i.e.,
fledged no young). In addition, the youngest
nestling at one supplemented nest hatched 10
days after its closest sibling (mean age differ-
ence between oldest sibling and each of the
younger siblings = 2.12 days) and was no-
ticeably smaller than its two nest mates, yet
survived to fledging age with little aggression
between siblings. Ward and Kennedy (1996,
unpubk data) documented similar results in
their experiment, where a nest with supple-
mental food successfully fledged four young
including a nestling 7-10 days younger than
its closest sibling. Because nestlings that hatch
significantly later than their siblings in asyn-
chronous broods often die unless enough food
is provided (Bryant and Tatner 1990, Wiebe

and Bortolotti 1995), we attribute the higher
survival of these treatment nestlings to the
high food abundance.

Although our study was not designed to in-
vestigate the role of food in sibling aggression
in goshawks, our finding of higher survival
for supplementally fed nestlings, coupled with
the siblicide observations provided us with the
opportunity to consider this relationship. Lack
(1954) hypothesized that asynchronous hatch-
ing in avian species occurs to facilitate brood
size reduction to match available levels of re-
sources provided by parents. If Lack’s hy-
pothesis is correct, occurrences of siblicide
should be influenced by levels of prey abun-
dance. Forbes and Mock (1994) differentiate
two types of facultative siblicide: one in
which aggression is triggered by food short-
age and the other where it is not. Mock and
coworkers (1987) observed that the occur-
rence of fatal sibling aggression in some spe-
cies was only indirectly influenced by food.
They observed aggressive behavior between
siblings regardless of food levels, but mortal-
ity from aggression was lower if food was
abundant because the younger siblings were
sufficiently strong to withstand the attacks.
Our observations are consistent with the form
of facultative siblicide directly influenced by
resource levels and provides evidence for the
hypothesis that low food supplies trigger sib-
ling aggression in goshawks. Additional em-
pirical research, coupled with measurement of
background resource levels is needed to fur-
ther substantiate this assertion and clarify the
nature of the relationship.

ACKNOWLEDGMENTS

These observations were made while conducting re-
search funded by the USDA Forest .Service, Ashley
National Forest. We thank C. W. Boal. R. R Gerhardt.
J. S. Marks, D. W. Mock and two anonymous referees
for their helpful reviews of the manuscript. We would
also like to thank J. Cropley, S. Lewis, C. Neilson, M.
Painter, and S. Rayroux for assistance in the Held, K.
Paulin for her support of this research, and C. Sigurd-
son for generously volunteering her time and expertise
to conduct necropsies.

LITERATURE CITED

Anderson, D. J. 1989. The role of hatching asynchro-
ny in siblicidal brood reduction of two booby spe-
cies. Behav. Ecol. Sociobiol. 25:363—368.
Bechard, M. j. 1983. Food supply and the occurrence

436

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

of brood reduction in Swainson’s Hawk. Wilson
Bull. 95:233-242.

Boal, C. W. and J. E. B acorn. 1994. Siblicide and
cannibalism at Northern Goshawk nests. Auk 111:
748-750.

BoRTOLorn, G. R., K. L. Wiebe, and W. M. Iko. 1991.
Cannibalism of nestling American Kestrels by their
parents and sibhngs. Can. J. Zool. 69:1447-1453.

Braun, B. M. and G. L. Hunt, Jr. 1983. Brood re-
duction in Black-legged Kitti wakes. Auk 100:
469-476.

Bryant, D. M. and R Tatner. 1990. Hatching asyn-
chrony, sibling competition and siblicide in nest-
ling birds: studies of swiftlets and bee-eaters.
Anim. Behav. 39:657—671.

Creighton, J. C. and G. D. Schnell. 1996. Proximate
control of siblicide in Cattle Egrets: a test of the
food-amount hypothesis. Behav. Ecol. Sociobiol.
38:371-377.

Dewey, S. R, 1999. Effects of supplemental food on
parental care strategies and juvenile survival in
Northern Goshawks. M.Sc. thesis, Colorado State
University, Fort Collins.

Drummond, H. and C. Garcia Chavelas. 1989. Food
shortage influences sibling aggression in the Blue-
footed Booby. Anim. Behav. 37:806—819.

Edwards, T. C., Jr. and M. W. Collopy. 1983. Ob-
ligate and facultative brood reduction in eagles:
an examination of factors that influence fratricide.
Auk 100:630-635.

Forbes, L. S. and D. W. Mock. 1994. Proximate and
ultimate determinants of avian brood reduction.
Pp. 237-256 in Infanticide and parental care (S.
Parmigiani and F. S. vom Saal, Eds.). Hardwood
Academic Publ., Chur, Switzerland.

Gerhardt, R. P, D. M. Gerhardt, and M. A. Vas-
QUEZ. 1997. Siblicide in Swallow-tailed Kites.
Wilson Bull. 109:1 12-120.

Heeb, P. 1994. Intraclutch egg-mass variation and
hatching asynchrony in the Jackdaw Con'us mo-
nedula. Ardea 82:287—297.

Heinsohn, R. G. 1995. Hatching a.synchrony and brood
reduction in cooperatively breeding White-winged
Choughs Corcora.x melanorhamphos. Emu 95:
252-258.

Lack, D. 1954. The natural regulation of animal num-
bers. Oxford Univ. Press, Oxford, U.K.

Mock, D. W. and T. C. Lamey. 1991. The role of
brood size in regulating egret sibling aggression.
Am. Nat. 138:1015-1026.

Mock, D. W. and G. A. Parker. 1997. The evolution
of sibling rivalry. Oxford Univ. Press, Oxford,
U.K.

Mock, D. W., T. C. Lamey, and B. J. Ploger. 1987.
Proximate and ultimate roles of food amount in
regulating egret sibling aggression. Ecology 68:
1760-1772.

Mock, D. W., H. Drummond, and C. H. Stinson.
1990. Avian siblicide. Am. Sci. 78:438-449.

Newton, I. 1979. Population ecology of raptors. Buteo
Books, Vermillion, South Dakota.

O’Connor, R. J. 1978. Brood reduction in birds: se-
lection for fratricide, infanticide and suicide.
Anim. Behav. 26:79-96.

PiLZ, W. R. AND L. K. Seibert. 1978. Fratricide and
cannibalism in Swainson’s Hawk. Auk 95:584-
585.

Reynolds, P. S. 1996. Brood reduction and siblicide
in Black-billed Magpies (Pica pica). Auk 113:
189-199.

Schnell, J. H. 1958. Nesting behavior and food habits
of goshawks in the Sierra Nevada of California.
Condor 60:377-403.

Stinson, C. H. 1979. On the selective advantage of
fratricide in raptors. Evolution 33:1219-1225.

Ward, J. M. and P. L. Kennedy. 1996. Effects of sup-
plemental food on growth and survival of juvenile
Northern Goshawks. Auk 113:200-208.

Wellicome, T. I. 1997. Reproductive performance of
Burrowing Owls (Speotyto cimicularia): effects of
supplemental food. J. Raptor Res. 9:68-73.

Wiebe, K. L. and G. R. Bortolotti. 1995. Food-de-
pendent benefits of hatching asynchrony in Amer-
ican Kestrels Falco .xpan’eriu.K. Behav. Ecol. So-
ciobiol. 36:49-57.

Zachel, C. R. 1985. Food habits, hunting activity, and
post-fledging behavior of Northern Goshawks
(Accipiter gentili.x) in interior Alaska. M.Sc. the-
sis, Univ. of Alaska, Fairbanks.

SHORT COMMUNICATIONS

437

Wilson Bull., I 1 1(3), 1999, pp. 437-439

Cooperative Foraging in the Mountain Caracara in Peru

Jason Jones' -

ABSTRACT. — Cooperative foraging behavior is
rarely observed in ground-walking birds. I report on
observations of cooperative foraging behavior by
Mountain Caracaras (Phalcoboenits megaloptenis) in
the puna region of Peru in September of 1995. On
several occasions, three individuals (two adults and
one immature) were observed working together to turn
over large rocks to obtain prey from beneath. These
cooperative foraging events are notable in that, unlike
cooperative foraging behavior observed in other
ground-walking birds, only one individual obtained
prey from a given cooperative effort. The presence of
the immature individual may be indicative of delayed
dispersal, a behavior not previously described for this
poorly known species. Received 14 Dec. 1998, ac-
cepted 28 March 1999.

Cooperative foraging involves two or more
individual organisms assisting one another in
obtaining a food item. Among vertebrates,
this behavior is well described in many social
mammals (Macdonald 1983, Serfass 1995)
and some fish (Dugatkin and Mesterton-Gib-
bons 1996) but is relatively uncommon in
birds (Sullivan 1984). Among birds, cooper-
ative hunting appears most frequently in sea-
birds (Parasitic Jaegers, Stercorarius parasi-
ticus, Pruett- Jones 1980; Brown and Ameri-
can White pelicans, Pelecanus occidentalis
and P. erythrorhynchos, J. Jones, unpub.
data). Examples from land birds include the
cooperative hunting behavior exhibited by
Harris’ Hawks (Parabuteo unicinctus; Mader
1979, Bednarz 1988), Golden Eagles (Aqiiila
chrysaetos; Collopy 1983) and Crested Ca-
racaras {Caracara plancus\ Morrison 1996).
Most of these instances of cooperative hunt-
ing involve highly mobile prey items; coop-
erative foraging for less mobile organisms is
uncommon (Sullivan 1984). In this report, I

' 4974 Lakeshore Rd., Kelowna, British Columbia,
Canada, VI Y 7R3.

- Present address: Dept, of Biology, Queen’s Univ.,
Kingston, Ontario, Canada, K7L 3N6;

E-mail: jonesja® biology. queensu.ca

detail observations of cooperative foraging
behavior in the Mountain Caracara (Phalco-
boenus megalopterus). This species is adept
at ground foraging and in non-urban areas
feeds on large arthropods, rodents, and birds
(Brown and Amadon 1968). Breeding usually
occurs between October and December with
two, rarely three, eggs laid (Brown and Ama-
don 1968).

STUDY AREA

The observations were made in the Peruvian puna
zone on the road between the towns of Quillabamba
and Ollantaytambo, Department of Cuzco (1 3° 9' S,
72° 14' W; 3750 m elevation). This region is charac-
terized by dry grasslands, dominated by genera such
as Calamagrostis and Festuca with interspersed shrubs
of the genera Astragalus, Berberis, and Lupiniis (Park-
er et al. 1982). Mountain Caracaras are common in this
region (Parker et al. 1982) and are often found near
towns where they feed on refuse and carrion (White
and Boyce 1987).

RESULTS

I observed cooperative rock-turning on four
occasions from 1-6 September 1995, as I
watched three individuals (two adults and one
immature) foraging together on the puna
grassland. The immature bird was easily dis-
tinguished by its plumage. On each occasion,
one of the adults approached a large rock,
walked around it, uttered a high-pitched
kieeer, and then stood by the rock. Apparently
responding to the vocalization, the other in-
dividuals joined the first at the rock and pro-
ceeded to work together to flip the rock from
its resting place, with each bird using one of
its talons. The bird that made the call partic-
ipated in the the turning but also appeared to
act as a “watcher” and was the individual re-
sponsible for prey capture. On one occasion,
the item was captured by an adult which then
gave it to the immature bird. No begging vo-
calizations were uttered by the younger bird
nor did it adopt any unusual posture. Each lift-
ing event took approximately 30 min from call

438

THE WILSON BULLETIN • Vol. Ill, No. 3. September 1999

to prey capture. This species, although a
ground-foraging specialist, has weak legs
(Brown and Amadon 1968). After examining
the rocks (approximate dimensons 30 X 20 X
10 cm), 1 do not believe that one individual
could have turned over any of the rocks by
itself.

At no time during the days of observation
was the immature bird more than 100 m from
one or the other of the adults, although the
adults were often separated by distances up to
500 m. In a series of 12 one-hour watches {n
= 4 for each individual), I determined that the
two adults appeared to spend more time for-
aging than did the immature bird (adult =
68.6%, imm. = 52.3%) although the differ-
ence was not statistically significant (x“ =
3.334, df = 1, P = 0.067).

Subsequent investigation revealed that
worms or arthropods could be found under
most (19 of 24) of the rocks in the vicinity;
several of the rocks (5 of 24) also hid rodent
runways. Examination of the surrounding
grassland showed that prey items of a similar
size but different taxa (e.g., grasshoppers rath-
er than millipedes) were available without
rock lifting.

DISCUSSION

Unlike most land bird species. Mountain
Caracaras were not using cooperative foraging
behavior to pursue and subdue large, highly
mobile prey items. Rather, they required co-
operation to obtain access to otherwise inac-
cessible prey items. While the turning over of
small rocks was one of the main foraging be-
haviours exhibited by these individuals, the
cooperative rock-turning events did not occur
within the set watches and did not seem to
represent a major foraging strategy for these
individuals. The main difference between the
cooperative behavior observed in Mountain
Caracaras and that observed in other cooper-
atively foraging birds is that only one individ-
ual obtained food from a given foraging event;
Harris’ Hawks, for example, share large prey
that are cooperatively caught (Bednarz 1988).
This disparity in obtaining a food reward may
even out over time (e.g., one of the four prey
items was given to the immature bird) but
there apparently is often no immediate reward
for some of the individuals participating. That
individuals are willing to help without a re-

ward is perhaps indicative of the length of
time these birds remain together as a foraging
unit; that is, an individual is willing to help
today because its turn will come eventually
(see Trivers 1971 for discussion of reciprocal
altruism).

Congeners of the Mountain Caracara hatch
their eggs in December and fledglings are usu-
ally independent by March (Brown and Ama-
don 1968). As my observations took place in
September, the immature member of the trio
was probably a chick from a previous breed-
ing effort and its presence, therefore, may rep-
resent delayed dispersal. Delayed dispersal is
fairly common in Neotropical raptors (Mader
1981). How common delayed dispersal is in
Mountain Caracaras and how it may affect the
incidence of cooperative foraging, is uncer-
tain.

ACKNOWLEDGMENTS

I thank J. Barg, E. Canuthers, F. Chavez-Ramirez,
D. Jone.s, W. Rendell and four anonymous reviewers
for providing helpful comments on earlier drafts of this
manuscript.

LITERATURE CITED

Bednarz, J. C. 1988. Cooperative hunting in Hanis’
Hawks {Parahuteo unicinctus). Science 239:
1525-1527.

Brown, L. and D. Amadon. 1968. Eagles, hawks and
falcons of the world. Vol. 2. McGraw-Hill, New
York.

COLLOPY, M. W. 1983. Foraging behavior and success
of Golden Eagles. Auk 100:747-749.

Dugatkin, L. a. and M. Mesterton-Gibbons. 1996.
Cooperation among unrelated individuals: recip-
rocal altruism, by-product mutualism and group
selection in tishes. Biosystems 37:19-30.
Macdonald, D. W. 1983. The ecology of carnivore
social behaviour. Nature 301:379-384.

Mader. W. J. 1979. Breeding behavior of a polyan-
drous trio of Harris' Hawks in southern Arizona.
Auk 96:776-788.

Mader, W. J. 1981. Notes on nesting raptors in the
llanos of Venezuela. Condor 83:48-51.

Morri.son, j. L. 1996. Crested Caracara (Ccinieara
plancus). In The birds of North America, no. 249
(A. Poole and E Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; .The
American Ornithologists' Union, Washington, D.C.
Parker, T. A., Ill, S. A. Parker, and M. A. Plenge,
1982. An annotated checklist of Peruvian birds.
Buteo Books, Vermillion, South Dakota.

Pruett -Jones, S. G. 1980. Team-hunting and food
sharing in Parasitic Jaegers. Wilson Bull. 92:524-
526.

SHORT COMMUNICATIONS

439

Serfass, T. L. 1995. Cooperative toragiiig by North
American river ottens. Lutra caiuuk'iisis. Can.
Field-Nat. 109:458-459.

Sullivan, K. A. 1984. Cooperative foraging and court-
ship feeding in the Laughing Gull. Wilson Bull.
96:710-71 1.

Trivf:r.s, R. L. 1971. The evolution of rceiprtteal altru-
ism. Q. Rev. Biol. 46:35-57.

White, C. M. and D. A. Boyce. 1987. Notes on the
Mountain Caracara ( kluilcohoenu.s inef>cilople-
rus) in the Argentine puna. Wilson Bull. 99:
283-284.

Wilson Bull.. 1 1 1(3), 1999, pp. 439-440

Predation by Rufous Motmot on Black-and-Green Poison Dart Frog

Terry L. Master'

ABSTRACT. — I observed a Rufous Motmot (Bar-
yphthengus martii) feeding a black-and-green poison
dart frog (Denclrobates aurotiis) to another motmot in
the Caribbean Slope lowland rainforest of northeastern
Costa Rica. Neither individual appeared to suffer any
ill effects from what was probably courtship feeding.
Small vertebrates are typical prey for the larger species
of motmots. Blue-crowned Motmots (Momotus moni-
ota) have been observed consuming several species of
poison dart frogs raised in captivity but captive reared
frogs either do not contain, or have reduced levels of,
the toxins that native frogs produce. Relatively little is
known about the effects of poison dart frog toxins on
predators. Presumably, the digestive system of the Ru-
fous Motmot is capable of neutralizing the potentially
toxic effects of such prey. Received 15 Sept. 1998,
accepted 15 Feb. 1998.

Poison dart frogs have long been known to
possess toxic skin secretions, and, because of
their bright coloration, are thought to be apo-
sematic to visually hunting predators such as
Rufous Motmots {Baryphthengus martii)
which presumably have excellent color vision
(Brodie and Tumbarello 1977). Smith (1975)
demonstrated that hand-reared Torquoise-
browed Motmots (Eumomota superciliosa)
showed an innate avoidance of snake-shaped
models with patterns simulating those of coral
snakes. All other snake models were readily
attacked implying that aposematic coloration
is a deterrent to this species. Observations in-
dicate that Blue-crowned Motmots {Momotu.s
momota) at the National Aquarium consume
several species of poison dart frogs including

' Dept, of Biological Sciences, East Stroudsburg
Univ., East Stroudsburg, PA 18301;

E-mail: tmaster@esu.edu

the black-and-green poison dart frog (Dendro-
bates auratus) and phantasmal poison dart
frog (Dendrobates tricolor). However, these
frogs were raised in captivity and either do not
produce or have relatively low levels of the
characteristic skin toxins (Kricher 1997; C.
Rowsom, pers. comm.).

At approximately 9:30 CST on 26 March
1995, an adult Rufous Motmot was observed
in secondary lowland tropical forest from a
hiking trail located at Estacion Biologica La
Suerte, near Cariari, Limon Province, north-
eastern Costa Rica (10° 26' N, 83° 46' W).
The bird landed 25 m from the trail on an
exposed perch 3 m above the ground and was
easily observed for approximately 4 min. Af-
ter 4 min another individual landed on the
same branch next to the first individual. The
newly arrived motmot was carrying a black-
and-green poison dart frog in its beak which
it fed immediately to the first individual. It is
not possible to distinguish between sexes in
Rufous Motmots; however, this behavior was
interpreted as a male who was feeding the fe-
male as a courtship gesture. Both individuals
had diagnostic black breast marks and raquet-
tails indicative of adult birds, suggesting that
this was probably not a fledgling being fed.
The pair continued sitting on the branch for
approximately 30 min after which they flew
off together into the forest. Neither individual
appeared to suffer any ill effects from either
grasping or consuming the poison dart frog.

The typical diet of motmots varies somewhat
in conjunction with body size. Smaller species
prefer insects while larger species consume in-
sects along with other invertebrates, small ver-

440

THE WILSON BULLETIN • Vol. HI, No. 3, September 1999

tebrates, and fmit (Orejuela 1980, Remsen et
al. 1993). The Rufous Motmot consumes ar-
thropods, other invertebrates including crabs,
small vertebrates including fish, lizards and
birds, as well as fruit (Remsen et al. 1993).
Frogs have been reported as a dietary compo-
nent of the Rufous, Broad-billed {Electron pla-
tyrhynchum) and Torquoise-browed motmots
(Remsen et al. 1993), and Blue-crowned Mot-
mots in captivity (C. Rowsom, pers. comm.).

The effect of poison dait frog toxins on var-
ious potential predators has received relatively
little attention. Brodie and Tumbarello (1977)
tested the response of garter snakes {Thamno-
phis sirtalis) to D. auratus offered as prey.
Snakes readily mouthed, or in some cases con-
sumed the frogs but all exhibited head shaking,
mouth opening, convulsions, and loss of equi-
librium. Only one snake actually died and that
was after consuming its third frog. These
snakes do not possess color vision and might
not be influenced by the aposematic coloration
to the extent that an organism with color vision
would be (Brodie and Tumbarello 1977).

While motmots in general may be warned
by aposematic coloration, the Rufous Motmot
at least is capable of handling and consuming
this particular species of poison dart frog.
Dendrobates auratus reaches densities of 1 in-
dividual/180 m^ in one locality at La Suerte

known to be frequented by Rufous Motmots
(B. Graves, pers. comm.) One pair was ob-
served on the ground rummaging through leaf
litter where they would undoubtedly encoun-
ter D. auratus (B. Graves, pers. comm.). The
level of toxins in the frogs of this area, how
the motmots physiologically handle the tox-
ins, and the frequency with which they con-
sume D. auratus remain unknown.

ACKNOWLEDGMENTS

I would like to thank Dr. T LaDuke of East Strouds-
burg University, Dr. B. Graves of Northern Michigan
University, and C. Row.som of the National Aquarium
in Baltimore for encouragement and assistance with
this manuscript.

LITERATURE CITED

Brodie, E. D. and M. S. Tumbarello. 1977. The an-
tipredator functions of Dendrobates auratus (Am-
phibia, Anura, Dendrobatidae) skin secretion in
regard to a snake predator. J. Herp. 12:264—265.
Kricher, J. 1977. A Neotropical companion. Princeton
Univ. Press., Princeton, New Jersey.

Orejuela, J. E. 1980. Niche relationships between
Torquoise-browed and Blue-crowned motmots.
Wilson Bull. 92:229-244.

Remsen, J. V., M. A. Hyde, and A. Chapman. 1993.
The diets of Neotropical trogons, motmots, bar-
bets and toucans. Condor 95:178-192.

Smith, S. M. 1975. Innate recognition of coral snake
pattern by a possible avian predator. Nature
(Lond.) 187:759-760.

Wil.son Bull.. 111(3), 1999, pp. 440-442

Evidence Of Egg Ejection In Mountain Bluebirds

Percy N. Hebert'

ABSTRACT. — When the last two eggs of Mountain
Bluebird {Sialia currucoides) clutches were replaced
with another bluebird egg and one House Sparrow
{Passer domesticus) egg, 20% (3/15) of the spaiTOW
eggs were removed within 24 hr. None of the surrogate
bluebird eggs was removed. This is the first recorded
instance of interspecific egg ejection in a bluebird spe-
cies, and hole-nesters in general. Received 2 Nov.
I99S, accepted IH Feb. 1999.

' Dept, of Zoology, Univ. of Manitoba. Winnipeg,
Manitoba. Canada, R3T 2N2;

E-mail: phcberKScc.umanitoba.ca

Of the approximately 140 biological hosts
of the Brown-headed Cowbird (Molothrus
ater), fewer than 7% have been classified as
rejectors (Friedmann and Kiff 1985, Ortega
1998). Rejectors typically remove cowbird
eggs from the nest within 24 hr of introduc-
tion (Rothstein 1982). Ejection is accom.-
plished either by grasping the cowbird egg. be-
tween the mandibles or by puncturing the egg
with the beak and then lifting the egg out of
the nest (Sealy 1996). Acceptors, by contrast,
do not remove cowbird eggs and in most cases
provision the cowbird nestling(s) (see Petit
1991, Sealy 1996).

SHORT COMMUNICATIONS

441

Unlike the Shiny Cowbird {M. honarensis),
the Brown-headed Cowbird infrequently par-
asitizes hole-nesters (Ortega 1998; but see
Petit 1991). Bluebirds (Sicilia spp.) are para-
sitized infrequently by Brown-headed Cow-
birds (Friedmann and Kiff 1985). Cowbird
eggs have been found in 0.2-2. 6% of Eastern
Bluebird (S. sialia) nests, but there are only 4
records of parasitism on Mountain Bluebirds
(S. currucoides) and none for the Western
Bluebird (S. mexicanus; Friedmann and Kiff
1985). These low frequencies of parasitism
may be due to aggression by adult bluebirds
towards female cowbirds (Gowaty and Wag-
ner 1988). Furthermore, the cowbird parasit-
izes smaller hosts than itself (Friedmann et al.
1977), thus female cowbirds may be too large
to squeeze through bluebird cavity entrances
(Friedmann et al. 1977, Pribil and Pieman
1997).

Given such low frequencies of parasitism
by Brown-headed Cowbirds on hole-nesters in
general (Friedmann and Kiff 1985), apparent-
ly there has been little selection pressure fa-
voring the evolution of rejection behavior
(Davies and Brooke 1989). In fact, there is
only one published record of interspecific egg
ejection in hole nesting species (Moksnes et
al. 1990). Here I present data that indicate that
Mountain Bluebirds apparently cannot distin-
guish between conspecific eggs, whereas they
can recognize interspecific eggs as different
from their own, and that these eggs are some-
times removed from the nest.

METHODS

I collected the data between May and July, in 1995
and 1996, on a population of Mountain Bluebirds nest-
ing in boxes near Virden, Manitoba (49° 51' N,
100° 55' W). Nest-boxes were visited every 2-3 days
during ne.st-building and daily during laying. Eggs
were measured and weighed within 24 hr of laying,
and numbered on the blunt end using a non-toxic felt
marker. Once the clutch was complete, the penultimate
and ultimate eggs were removed for 24 and 48 hr, re-
spectively. To minimize the risk of abandonment, some
of these clutches received one bluebird egg from failed
clutches and one House Sparrow [Passer clomeslicus)
egg. The presence or absence of these replacement
eggs was then recorded 24 and 48 hours later when
the original eggs were returned to their clutches.

Because House Sparrow eggs are very similar to
Brown-headed Cowbird eggs (see Lowther 1993,
Lowther and Cink 1992), I expected bluebirds to re-
spond to a sparrow egg the same way they would re-
spond to a cowbird egg (see also Rothstein 1977).

RESULTS

Fifteen nests received a bluebird egg and a
sparrow egg, and none of these nests was
abandoned. None of the replacement bluebird
eggs was removed from the nest within 48 hr.
By contrast, 3/15 (20%) of the sparrow eggs
were removed from the nest, all within 24 hr
of introduction. In 2 of the 3 ejections, the
sparrow egg was removed from the nest-box,
whereas in the third instance the undamaged
egg ended up on the rim of the nest. For both
years combined, 12 bluebird eggs were known
to have been cracked or dented during mea-
suring. Of these, one was found on the rim of
the nest cup the following day, 7 were gone
the following day, and 4 remained in the nest.

DISCUSSION

The results of this study indicate that
Mountain Bluebirds are capable of egg ejec-
tion. Mountain Bluebirds possess several traits
that Rothstein (1975) identified as pre-adap-
tations for the evolution of ejection behavior.
For instance, Rothstein (1975) suggested that
the evolution of ejection behavior would be
facilitated if the hosts’ eggs differed from
those of the cowbird in at least two respects:
base color, maculation, and size. Mountain
Bluebird eggs differ from sparrow and cow-
bird eggs in color and maculation (see Lowth-
er 1993, Lowther and Cink 1992, Power and
Lombardo 1996). Mountain Bluebirds can re-
move their damaged eggs from the nest; their
eggs are similar in size to those of the cow-
bird. Thus it can be assumed that bluebirds
would be capable of removing cowbird eggs
from their nests.

Given that Mountain Bluebirds are sympat-
ric with Brown-headed Cowbirds (see Lowth-
er 1993, Power and Lombardo 1996), and that
their eggs are sufficiently different to facilitate
recognition and ejection of a cowbird egg, a
rejection rate of only 20% would appear to be
low. However, the low rate of ejection I ob-
served is likely an underestimate of the fre-
quency of ejection behavior in Mountain
Bluebirds. For example, Rothstein (1982) ob-
served that American Robins (Turdus migra-
torius) are less likely to eject cowbird eggs
that are introduced into the nest after laying.
As I introduced sparrow eggs at clutch com-
pletion, it is thus possible that bluebirds were

442

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

less likely to eject them. Furthermore, the
ejected sparrow eggs were removed within 24
hr of placement in bluebird nests. Such a
quick response is typical of most rejector spe-
cies (Sealy 1996) and suggests at least a mod-
erate level of intolerance (sensu Rothstein
1982) to cowbird parasitism in Mountain
Bluebirds.

ACKNOWLEDGMENTS

I thank E Asselin, H. Khan, and H. McElmoyle for
their assistance in the field. J. Hare, A. Smith and the
Canadian Wildlife Service provided valuable logistical
support. I also thank S. Rondeau for providing some
of the House Sparrow eggs. S. G. Sealy, and 3 anon-
ymous reveiwers made helpful comments on a previ-
ous draft of the manuscript. This study was supported
by a research grant from the Natural Sciences and En-
gineering Research Council of Canada.

LITERATURE CITED

Davies, N. B. and M. de L. Brooke. 1989. An ex-
perimental study of co-evolution between the
Cuckoo, Cuculus canoriis, and its hosts. II. Host
egg markings, chick discrimination and general
discussion. J. Anim. Ecol. 58:225-236.
Friedmann, H., L. E Kief, and S. I. Rothstein. 1977.
A further contribution to knowledge of the host
relations of the parasitic cowbirds. Smithson.
Contr. Zool. 235:1-75.

Friedmann, H. and L. E Kiff. 1985. The parasitic
cowbirds and their hosts. Proc. West. Found. Vert.
Zool. 2:226-302.

Gowaty, P. a. and S. j. Wagner. 1988. Breeding sea-
son aggression of female and male Eastern Blue-
birds (Sialia .sialis) to models of potential conspe-
cific and interspecihc egg dumpers. Ethology 78:
238-250.

Lowther, P. E. 1993. Brown-headed Cowbird (Mol-
othrii.s ater). In The birds of North America, no.

47 (A. Poole and F. Gill, Eds.). The Academy of
Natural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Lowther, P. E. and C. L. Cink. 1992. House Sparrow.
In The birds of North America, no. 12 (A. Poole,
P. Stettenheim, and E Gill, Eds.). The Academy
of Natural Sciences, Philadelphia, Pennsylvania;
The American Ornithologists’ Union, Washington,
DC.

MOKSNES, a., E. R0SKAFT, A. T. Braa, L. Korsnes, H.
M. Lampe, and H. C. Pedersen. 1990. Behav-
ioural responses of potential hosts towards artih-
cial cuckoo eggs and dummies. Behaviour 116:
64-89.

Ortega, C. P. 1998. Cowbirds and other brood para-
sites. Univ. of Arizona Press, Tucson.

Petit, L. J. 1991. Adaptive tolerance of cowbird par-
asitism by Prothonotary Warblers: a consequence
of nest-site limitation. Anim. Behav. 41:425-432.

Power, H. W. and M. P. Lombardo. 1996. Mountain
Bluebird. In The birds of North America, no. 222
(A. Poole and E Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
DC.

Pribil, S. and j. Picman. 1997. Parasitism of House
Wren nests by Brown-headed Cowbirds: why is it
so rare? Can. J. Zool. 75:302-307.

Rothstein, S. I. 1975. An experimental and teleonom-
ic investigation of avian brood parasitism. Condor
77:250-271.

Rothstein, S. I. 1977. Cowbird parasitism and egg
recognition of the Northern Oriole. Wilson Bull.
89:21-32.

Rothstein, S. I. 1982. Mechanisms of avian egg rec-
ognition: which egg parameters elicit responses by
rejecter species? Behav. Ecol. Sociobiol. 11:229-
239.

Sealy, S. G. 1996. Evolution of host defenses against
brood parasitism: implications of puncture-ejec-
tion by a small passerine. Auk 1 13:346-355.

SHORT COMMUNICATIONS

443

Wilson Bull.. 111(3), 1999, pp. 443-444

Foraging Ovenbird Follows Armadillo

Douglas J. Levey'

ABSTRACT — I report an observation of a foraging
Ovenbird (Seiurus aurocapillus) following a nine-
banded armadillo (Dasypits novemcinctus), near
Gainesville, Florida. Close attendance only while the
armadillo was moving and disturbing leaf litter sug-
gests the Ovenbird was taking advantage of increased
prey availability caused by the armadillo's flushing of
insects. Received 27 May 1998, accepted / Oct. 1998.

Many species of birds forage in association
with groups of other species. The two most
common explanations of such interspecific
groups relate to decreased risk of predation
and increased foraging efficiency (Bertram
1978). Foraging efficiency can be increased
via several mechanisms (Morse 1970). Per-
haps the least-well studied mechanism occurs
when one species follows another species and
captures prey incidentally flushed by the sec-
ond species. Such a relationship between “fol-
lowers” and “beaters” has been reported for
groups of birds and groups of cattle (Scot
1984), dolphins (Evans 1987), primates (Ter-
borgh 1983, Boinski and Scott 1988), wolves
(Silveira et al. 1997), ants (Willis and Oniki
1978), and other birds (Bennetts and Dreitz
1997). Here I report an observation of an Ov-
enbird {Seiurus aurocapillus) following a
nine-banded armadillo (Dasypus novemcinc-
tus).

On 6 March 1998 at 10:35 EST along the
rim of Paynes Prairie (Alachua County, Flor-
ida), I flushed an Ovenbird from the ground.
After perching for 1-2 min, it flew to the
ground within 3 m of an armadillo. As the
armadillo started to move forward, the Ov-
enbird flew directly to it, landing approxi-
mately 30 cm from its tail and maintaining
that distance of separation as the armadillo
walked. I followed them for 17 min at a dis-
tance of 10-12 m. The armadillo often

' Dept, of Zoology, PO Box 1 18525, Univ. of Flor-
ida, Gainesville, FL 3261 1-8525;

E-mail: DLEVEY(§)zoo. ufl.edu

Stopped briefly (< 10 s) to dig or push its
snout into the leaf litter. The Ovenbird did not
approach the armadillo’s head during these
times but rather remained by its tail. On two
occasions the armadillo stopped for 1-2 min
and the Ovenbird walked 2-3 m away from
it. When the armadillo started to move again,
the Ovenbird immediately resumed following
it, once returning by flight. It frequently
pecked at the leaf litter. I was unable to de-
termine if these presumed foraging attempts
were successful. They did not appear more
frequent when the Ovenbird was following
close behind the armadillo than when it tem-
porarily foraged by itself. I did not notice any
insects being flushed by the armadillo, but
there was heavy shade and I lacked binocu-
lars. I stopped my observation when the ar-
madillo walked into a clearing and the Ov-
enbird did not follow.

The behavior of the Ovenbird suggests its
association with the armadillo was not due to
both animals being attracted to an area of high
prey abundance. In particular, its close prox-
imity to the armadillo only when the armadillo
was moving suggests it was using the arma-
dillo as a beater. A similar pattern of atten-
dance has been noted for Double-toothed
Kites {Harpagus didenlatus), Gray-headed
Tanagers {Eucornetis pencillata), and Tawny-
winged Woodcreepers (Dendrocincla anaba-
tina) following Squirrel Monkey {Sainiiri oer-
stedi) troops; attendance frequencies of these
species were higher when the Squirrel Mon-
key troops were moving and foraging than
when they were immobile (Boinski and Scott
1988).

Despite high levels of disturbance created
by armadillos while foraging, I am unaware
of any other published accounts of birds fol-
lowing them. Further observations are re-
quired before it can be concluded that the bird
I watched was not idiosyncratic and that Ov-
enbirds benefit from following armadillos.

444

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

ACKNOWLEDGMENTS

Thanks to E. Bollinger, S. Duncan, G. Pryor, K.

Smith and an anonymous reviewer for constructive

comments on the manuscript.

LITERATURE CITED

Bennetts, R. E. and V. Dreitz. 1997. Possible use of
wading birds as beaters by Snail Kites, Boat-tailed
Crackles, and Limpkins. Wilson Bull. 109:169-
173.

Bertram, B. C. R. 1978. Living in groups: predators
and prey. Pp. 64-96 in Behavioural ecology (J. R.
Krebs and N. B. Davies, Eds.). Sinauer Press,
Sunderland, Massachusetts.

Boinski, S. and P. E. Scott. 1988. Association of birds
with monkeys in Costa Rica. Biotropica 20:136-
143.

Evans, D. L. 1987. Dolphins as beaters for gulls? Bird
Behav. 7:47-48.

Morse, D. H. 1970. Ecological aspects of some mixed-
species foraging flocks of birds. Ecol. Monogr. 40:
119-168.

Scot, D. 1984. The feeding success of Cattle Egrets
in flocks. Anim. Behav. 32:1089-1100.

Silveira, L., a. T. a. Jacomo, E H. G. Rodrigues,
and P. G. Crawshaw. 1997. Hunting association
between the Aplomado Falcon (Falco femoralis)
and the Maned Wolf {Chry.socyon bracbyurus) in
Emas National Park, central Brazil. Condor 99:
201-202.

Terborgh, J. 1983. Five New World primates. Prince-
ton Univ. Press, Princeton, New Jersey.

Willis, E. O. and Y. Oniki. 1978. Birds and army ants.
Annu. Rev. Ecol. Syst. 9:243-263.

Wilson Bull.. 1 1 1(3), 1999, pp. 445-456

Ornithological Literature

Edited by William E. Davis, Jr.

THE GREAT BLUE HERON: A NATU-
RAL HISTORY AND ECOLOGY OF A
SEASHORE SENTINEL. By Robert W. But-
ler. UBC Press, Vancouver, British Columbia.
1997: 167 pp., 30 black-and-white photos, 24
color plates, 17 numbered text figs., 18 tables.
$39.95 (Canadian) (cloth). — This is a very
nice book about a very interesting bird — the
Great Blue Heron {Ardea herodias) — with a
particular emphasis on the subspecies that fre-
quents the northwest coast of North America,
A. h. fannini. However, frequent reference to
herons elsewhere broadens the scope and per-
spective of the book. It is more than just the
study of a heron, however, it is the story of
the ecology and conservation of a region.
Robert Butler is certainly well qualified to tell
this story — 24 of the more than 1 50 references
cited bear his name as an author. In the brief
introductory chapter Butler outlines the aims
of the book that include detailing the natural
history of the Great Blue Heron, “a worthy
symbol of the conservation of coastal habi-
tats,” along the 27,000 kms of fragmented
British Columbia coastline. Chapters that fol-
low consider in detail the habitat of the heron
(the shores of temperate rainforest), the her-
on’s food web, the sites where the heron has
been studied year around, all with an histori-
cal perspective on interactions with man wo-
ven through the narrative. There are chapters
on foraging, food and diet, social and territo-
rial behavior, colonial nesting, habitat selec-
tion, population dynamics, and finally a chap-
ter on the conservation of Great Blue Herons
and the Strait of Georgia ecosystem. An epi-
logue concludes with the optimistic thought
that many people are beginning to recognize
that the environment provides more than just
resources, and that restoring environments
will yield great future dividends. Appendices
report records of Great Blue Heron colonies,
length-mass regression equations for fish, ef-
fects of increased disturbance on heron pop-
ulations, and lists of scientific names of plants
and animals.

The book is very well written. It is clear

that the author thoroughly enjoys his research
on herons (including night-vision telescopic
sleuthing), and his descriptions have an almost
poetic touch: “1 look back on the long hours
spent watching herons catch fish as an enjoy-
able period of my life. Perched in the shade
on a prominent location overlooking spectac-
ular scenery, the smell of the sea and arbutus
leaves carried on a warm breeze, gulls and
shorebirds busily feeding along the mudflat —
it was hard to beat,” or “On calm nights I
often slept on the beach beneath the stars. I
welcomed the silence of the night after a day
in the colony, though the quiet was periodi-
cally disturbed by landing calls ringing from
the forest.” Even his descriptions of natural
history phenomena make for pleasant reading:
“The delicate choreographed displays of a
threatening heron are exquisite.”

I thoroughly enjoyed reading this book. It
contained a great deal of interesting and im-
portant scientific information about Great
Blue Herons, and a well articulated conser-
vation perspective. I recommend it to anyone
interested in avian biology or conservation. —
WILLIAM E. DAVIS, JR.

A GUIDE TO THE IDENTIFICATION
AND NATURAL HISTORY OF THE SPAR-
ROWS OF THE UNITED STATES. By
James D. Rising. Illustrated by David D. Bea-
dle. Academic Press, New York. 1996: 365
pp., 27 color plates with captions, 53 color
range maps, $42.00 (cloth). — Until the last
three years, those of us fascinated by details
of the natural history of sparrows in North
America had only Bent’s life histories for a
reference. With the publication of Sparrows
and Buntings — A Guide to the Sparrows and
Buntings of North American and the World in
1995, and now this guide, we have substan-
tially more information of a recent vintage at
our finger tips. This is a good book. Because
of its less ambitious geographic coverage.

445

446

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

there is more information and greater detail
for each species in this book.

Sixty-two species, from Olive Sparrow
(Embernagra rufivirgata) through McKay’s
Bunting (Plectrophenax hyperboreus) are
treated. Each species account includes infor-
mation about identification, similar species,
details of plumage color, voice, habits, habitat,
breeding behavior, range, history, geographic
variation, measurements, and technical refer-
ences (for further reading). A clearly drawn,
easily interpreted, color range map shows
breeding, winter, and “all year” ranges for
each species. Including state and provincial
boundaries on the range maps facilitates their
interpretation substantially. Where sufficient
data exist, an additional color map, showing
relative abundance, based on Breeding Bird
Survey information, also is included as part of
the species account. Though somewhat useful,
the shading of the relative abundance maps is
such that it is difficult to distinguish the
shades chosen for the two middle classes of
the four relative abundance categories
mapped. In addition, the two middle catego-
ries of 5-20 and 20-50 individuals appear to
overlap. One wonders how the computer map-
ping software dealt with abundances of 20 in-
dividuals and how one really should interpret
the relative abundance maps.

The book is comprehensive in its coverage.
Among the 62 “species” descriptions in the
book are a few forms not yet elevated to full
species status by the AOU Committee on
Classification and Nomenclature in its 1998
check-list. These include Bell’s Sparrow [con-
sidered a “group” of Sage Sparrow (Amphis-
pizcibelli) by AOUJ and Red, Sooty, and Slate-
colored Fox sparrows [considered “groups”
of Fox Sparrow (Passerella iliaca) by AOUJ.
There also is substantial treatment of a num-
ber of “Alaskan rarities,” including Pine {Em-
berizci leucocephaliis). Little (E. pu.silla). Rus-
tic (E. ru.stica). Yellow-breasted {E. aureola).
Gray (E. variablis), Pallas’s (E. pallasi), and
Reed buntings {E. schoeniclus). Two other
rare North American species. Yellow-faced
(Tiaris oUvacea) and Black-faced {T. bicolor)
grassquits, reported occasionally from the
southern U.S., also are described. I also ap-
preciated the concise comments and descrip-
tions of subspecies for Savannah Sparrow
( Passe rculu.s sandwichensis). Song Sparrow

{Melospiza melodia), and Dark-eyed Junco
(Junco hyemalis), including tables of mea-
surements illustrating geographic variation in
size. Special introductory sections for most of
the genera described in the book provide help-
ful additional information about intergeneric
relationships among the sparrows and identify
areas where more research is needed to clarify
relationships among species and genera.

The illustrations by David Beadle add sig-
nificantly to the book’s appeal. In addition to
excellent color plates, each species account is
illustrated by a black-and-white drawing of
the species. The artist has done an excellent
job of capturing the subtle beauty of plumage
colors for this generally somber-colored as-
semblage of species.

In general this is a good, solid, user-friendly
reference book for a challenging group of
North American species. It is small enough to
be carried in the field, though not necessarily
a “field guide” in the traditional sense. I rec-
ommend it to anyone as an exceptionally well
illustrated, very readable introduction to North
American sparrows. — CHARLES R. SMITH.

NATURAL HISTORY OF THE WATER-
FOWL. By Frank S. Todd. Ibis Publishing
Co., Vista, California. 1996: 490 pp., more
than 750 color photographs with captions, 164
range maps, appendix. $80.00 (cloth). — This
visually stunning and informative book effec-
tively portrays the diversity of behavior, ecol-
ogy, and plumage of the more than 160 spe-
cies of waterfowl. Its large format {XQVi by
13") allows the author to display to best ad-
vantage the many fine photos that he has taken
in pursuit of waterfowl and other birds, al-
though even the heaviest binoculars will seem
like featherweights after lifting this hefty
tome! The book begins with introductions to
natural history (Chapter 1) and taxonomy
(Chapter 2), then proceeds to cover primarily
typical waterfowl groups in the next 15 chap-
ters (including chapters on “Waterfowl Odd-
ities” and “Whitewater Ducks”). The chapter
on waterfowl natural history includes a nice
overview of this group and its adaptations,
and many photos are included to illustrate typ-
ical comfort movements and postures. The
photo legends throughout the book add infor-

ORNITHOLOGICAL LITERATURE

447

mation that is sometimes not present in the
text, including general information on birds.
The level of detail varies for examples given
in this chapter, creating some redundancy with
future chapters, and many of the examples are
from work on geese and swans. The inclusion
of the term “gang rape” (p. 41) is inappro-
priate and unnecessary since the preferable
terminology (forced copulation) is also used
in the text, and references to “hyperactive”
and “hot-blooded” males in photo legends de-
picting forced copulation attempts are regret-
table and misleading. The taxonomy chapter
includes a very brief coverage of past and cur-
rent thoughts on waterfowl taxonomy. The au-
thor recognizes the need for further corrobo-
ration of more recent taxonomic treatments of
this group (some of which have not been sup-
ported by recent findings), and, for consisten-
cy with other sources, he uses more traditional
groups for subsequent chapters with a few ex-
ceptions.

Chapters 3-17 each begin with an overview
of the similarities within the group being cov-
ered, including movements on land and in the
air, basic calls, feeding habits, nesting and
egg-laying, flocking, mating and family be-
havior, and migration. Individual species ac-
counts follow, including descriptions of ex-
tinct species and separate accounts for each
race. The description of each species is ac-
companied by a small color range map (on
which race distributions are not delineated),
and usually several excellent photos of the
species in the wild. Valuable information on
conservation, captive propagation, and human
uses/conflicts is presented in addition to more
detailed information on topics outlined in the
chapter overview. Incubation period, time to
fledging, conditions under which dump nest-
ing data were collected, and nesting density
are presented inconsistently within the species
accounts, although further details on weight,
egg and clutch size, incubation period, and
fledging period are presented for each species
(and race if appropriate) in the Appendix. In-
sights from the close observation of captives
complement information from extensive field
experience in many species accounts.
Throughout these accounts, the author has in-
serted personal experiences and origins of
some scientific names that add extra interest
for the reader. I especially found interesting

the accounts of endangered species and races
within various waterfowl groups.

The accounts of the northern geese (Chap-
ter 5), swans (Chapter 6), and eiders (Chapter
14) are some of the most complete of the vol-
ume, demonstrating the author’s experience
with these groups and also information avail-
able from the scientific studies of The Wild-
fowl and Wetlands Trust in England. These
chapters are full of photos, including all of the
Canada Goose {Branta canadensis) races, and
details on the Nene (Hawaiian Goose; Branta
sandvicensis) reintroduction project and suc-
cessful Barnacle Goose {Branta leucopsis)
conservation. Nesting densities for the eiders
are well-documented, as is the solving of the
mystery of the location of Spectacled Eider
{Somateria fischeri) wintering areas. Atypical
waterfowl, such as the Magpie Goose {Anser-
anas semipalmata\ Chapter 3); Cape Barren
Goose {Cereopsis novaehollandiae). Freckled
Duck (Stictonetta naevosa). Spur-winged
Goose (Plectropterus gambensis). Comb
Duck {Sarkidiornis melanotos), and Pink-
eared Duck {Malacorhynchus membranaceus)
(grouped as “waterfowl oddities” in Chapter
7); and screamers (Chapter 17) are covered
quite extensively in accounts that highlight the
unusual traits of these species and their con-
tinued puzzling taxonomy. The author groups
together and covers well the white-water
ducks (Blue Duck, Hyrnenolairnus malacor-
hynchos'. Torrent Duck, Merganetta armata;
and Salvadori’s Duck, Salvadorina waigiuen-
sis) in Chapter 10, describing adaptations to
this demanding environment such as their sed-
entary lifestyle, territorial behavior, and diet
of benthic invertebrates.

The sheldgeese and shelducks, sharing
strong pairbonds and an aggressive disposi-
tion, are described in Chapter 8, including es-
pecially good information on the ongoing con-
flicts with humans when sheldgeese use crop
lands. Presented in Chapter 9 is another very
pugnacious group of waterfowl that includes
three flightless species, the steamerducks. The
chapters on the pochards (Chapter 13), sea-
ducks (Chapter 15), and stiff-tailed ducks
(Chapter 16) present many useful photos and
information on feeding and diving adapta-
tions. Also included in these chapters are fas-
cinating accounts of the (probably) extinct
Pink-headed Duck {Rhodonessa caryophylla-

448

THE WILSON BULLETIN • Vol. HI. No. 3, September 1999

cea), the extinct Labrador Duck {Camptorhyn-
chus labradorius) and Auckland Islands Mer-
ganser (Mergus australis), and the only wa-
terfowl obligate brood parasite, the extant
Black-headed Duck (Heteronetta atricapilla).
Accounts of the whistling ducks (Chapter 4),
including the White-backed Duck (Thalassor-
nis leuconotus), are informative but generally
short because of a lack of detailed information
for many species. This general lack of infor-
mation available for southern hemisphere
ducks is especially evident in short accounts
for these species in Chapters 1 1 (perching
ducks) and 12 (dabbling ducks). Chapter 11
covers many of the traditional perching duck
group members while recognizing that the
taxonomic organization of these ducks is still
changing. The account of a well-studied spe-
cies, the (North American) Wood Duck {Aix
sponsa), was disappointing in its omission of
available information on dump-nesting and
other aspects of its natural history. In general,
members of the large dabbling duck group re-
ceive more complete coverage, including
more personal observations and photos by the
author, although I found some aspects of the
Andean Teal (Anas flavirostris andiurn) and
Puna Teal (Anas puna) accounts at odds with
my own observations.

The Epilogue presents a balanced treatment
of threats to waterfowl populations and ben-
efits greatly from the author’s own experience
with the continued conservation challenges
that face this group. Overall, the book is writ-
ten in a style that is easy to read; a glossary
is included to aid the lay reader and the Index
at the end of the book facilitates finding in-
formation on particular species. Scientific
names are not presented in the book except
for those of waterfowl species. There are few
citations in the text and the bibliography in-
cludes mostly books, limiting the use of this
volume as a scientific or research reference.
Despite a few reoccurring grammatical prob-
lems, the writing style conveys well the thrill
of viewing waterfowl and the author’s enthu-
siasm about his experiences with this group of
birds. This book is a treat to the eyes for any-
one that appreciates birds, and in addition, is
at the least a good overview of waterfowl nat-
ural history. Highly recommended as a visual,
general reference book for anyone interested
in waterfowl. — GWENDA L. BREWER.

FAIRY-WRENS AND GRASSWRENS
MALURIDAE. By Ian Rowley and Eleanor
Russell, illus. by Peter Marsack. Oxford Uni-
versity Press, Oxford. 1997: 274 pp., 9 color
plates, numerous maps, tables and black-and-
white line drawings. $75 (cloth). — This fas-
cinating family of birds is found only in Aus-
tralia and New Guinea and consists of five
genera: emu-wrens (Stipiturus) and grass-
wrens (Amytornis) found only in Australia,
tree- wrens (Sipodotus) and russet- wrens (C/v-
tomyias) of New Guinea, and the largest ge-
nus, fairy-wrens (Malurus) found in both.
Several species of fairy-wrens have been in-
tensively studied using color-banded popula-
tions and biochemical analyses and the results,
particularly those relating to breeding biology,
are intriguing and make for fascinating read-
ing.

This fourth volume in Oxford University
Press’ series on bird families of the world pro-
vides a thorough review of this interesting
family of birds by authors who have done
much of the primary research on several of
the species considered. The monograph is di-
vided into two parts: the first consists of eight
chapters dealing with various aspects of the
biology, behavior, ecology, evolution, and
conservation of the Maluridae; and the second
consists of accounts of the 5 genera and 25
species that constitute the family. Chapter 2
discusses two centuries of the rather confused
taxonomy of the Maluridae, including more
recent biochemical studies. Chapter 3 deals
with the environment, biogeography, and evo-
lution, including plate tectonics and Gond-
wana breakup, past and present climates and
vegetation, refugia and speciation during the
past two million years, changes since human
settlement, and the evolution of the Maluridae
including the five main lineages of fairy-
wrens and the grass wrens. Chapter 4 deals
with morphology, locomotion, and feeding be-
havior of these largely insectivorous, ground,
and shrub-dwelling birds. Chapter 5 discusses
vocal communication and social organization,
and includes a number of sonagrams, and a
thorough analysis of courtship displays. The
chapter also details the remarkable findings
from electrophoresis and DNA fingerprinting
studies focused on reproductive biology in
fairy-wrens. Although monogamous and mat-
ed for life, extra-pair copulations outside of

ORNITHOLOGICAL LITERATURE

449

the territorial family group in one study, ac-
counted for more than three-quarters of the
young! Chapter 6 is devoted to co-operative
breeding and an analysis of helpers at the nest
(mostly surviving young from earlier years).
Chapter 7 contains a generalized life history
study of the Maluridae, based mostly on long-
term studies of fairy-wrens, and includes sec-
tions on nests and nest building, eggs, clutch
size, number of broods, reproductive success,
parasitism and predation, dispersal, and sur-
vival rates. Chapter 8, on conservation, traces
the clearance and fragmentation of vegetation
for agriculture, forestry, and grazing, and ac-
companying habitat degradation, introduction
of alien plants, draining of wetlands, and
changes introduced into fire regimes by Eu-
ropeans. The author’s analysis suggests that 5
species and 5 additional subspecies are cur-
rently threatened.

In Part II, brief accounts of each genus are
followed by detailed accounts of each species.
These accounts begin with descriptions of
adult males and females in breeding and
eclipsed plumages, of immatures and moult
sequences, and typically continue with history
of taxonomy, weights and measurements, field
characters, voice, range and status, habitat,
displays, breeding behavior, and life cycle.
Each account is accompanied by a range map
and sonagram(s). Additional brief accounts of
groupings of species within genera are present
where needed, e.g., a section on the four
chestnut-shouldered fairy-wren species. The
eight color plates are excellent. They include,
where appropriate, depictions of adult males
and females, immatures, eclipsed males, and
subspecies. The color of the plates is excel-
lent, although the breast of the Blue-breasted
Fairy- Wren (Maluras piilcherrimus) appears
bluer than I remember it in the field. A color
figure, grouped with the plates, contains six
photographs of typical Australian habitats. A
glossary helps with terms like “samphire,”
“spinifex,” or “billabong” that might not be
familiar to everyone. The bibliography in-
cludes more than 500 references.

This is a well-written, thoroughly re-
searched, monograph. I looked through the
Acknowledgments for people who I know
have done work on malurids — they were all
there. This is a comprehensive book, easy to
read, and loaded with interesting information

about a fascinating and lovely family of birds.
Anyone working on Australasian birds or with
interest in avian breeding biology should have
this book.— WILLIAM E. DAVIS, JR.

AVIAN CONSERVATION. By John M.
Marziuff and Rex Sallabanks (Eds.). Island
Press, Washington, D.C. 1998; 563pp. (no
price given) — Avian Conservation is a collec-
tion of chapters of which about one half were
presented in a symposium of the American
Ornithologist’s Union and Raptor Research
Foundation meeting in 1996. The remaining
chapters were solicited by the editors to fill
gaps and to provide a land manager’s per-
spective on relevant research for avian con-
servation. The book is divided into 7 parts and
31 chapters. Part 1 introduces the reader to the
past and present approaches taken in conser-
vation. Part 2 reviews a variety of techniques
applied in conservation research including ge-
netics, spatial modeling, indicator species con-
cept, and monitoring landbirds. Part 3 pro-
vides examples of approaches used to con-
serve endangered and sensitive species. Part 4
deals with conservation of forested landscapes
and Part 5 covers non-forested and urban
landscapes. Part 6 examines conservation of
birds outside North America and includes the
European agricultural environment, research
needs and applications for Neotropical birds,
and conservation in Israel, Russia, the Mari-
ana Islands and Australia. Part 7 is a discus-
sion on making conservation research relevant
to land managers. The aims of the authors are
to review current research and identify infor-
mation.

Marziuff and Sallabanks should be com-
mended for assembling concise reviews of
many important topics for conservationists
and applications for land managers. Anyone
interested in the latest information and status
of projects aimed at conserving the Northern
Spotted Owl (Strix occidentalis). Northern
Goshawk (Accipter gentiiis), Hawaiian Goose
(Anser sandvicensis), Red-cockaded Wood-
pecker {Picoides borialis) and others will find
this book very useful. Also anyone wishing to
apply techniques such as GIS modeling of
populations, understanding threats to seabirds,
invasions by exotics or affects of urban en-

450

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

vironments on birds should read this book. Is-
land Press has published an attractive book in
an easy-to-read format.

This book contains a wealth of information
and is an excellent review of the topic al-
though it is biased in favor of North American
issues — 6 of the 31 chapters were devoted to
areas of the world outside the USA, 2 of those
were written by Americans, and 50 of the 58
contributors were from American institutions.
Consequently, the conclusions were strongly
biased towards problems perceived by Amer-
icans. The topic of bird conservation is im-
mense and the book would have been more
successful if it had dealt only with birds that
use the USA. A minor annoyance is the as-
sumption by many American authors that all
readers know that place names such as Pacific
Northwest, Midwestern States, the Snake Riv-
er, and issues such as the enactment of the
Endangered Species Act are in the USA when
they are writing for a world-wide audience.
Once again, if the book had been focused on
birds that use the USA, these terms would
have been appropriate. A more important
oversight is the small amount of attention de-
voted to existing conservation programs in the
USA and abroad. For example, the North
American Waterfowl Management Plan
(NAWMP) is the largest and most ambitious
avian conservation program undertaken in
North America. It has overcome political bar-
riers, raised billions of dollars and set aside
1000s of hectares of wetlands in Canada, the
USA, and Mexico. The success of NAWMP
has prompted other programs such as the
Western Hemisphere Shorebird Reserve Net-
work (WHSRN), Partners in Flight, and the
Seaduck Joint Venture. There is no or very
little mention of these and many other con-
servation programs in Avian Conservation.
Both NAWMP and WHSRN have been in
place for many years, with well established
newsletters and web sites but only one paper
addressed concerns about waterfowl and it fo-
cused on the Hawaiian Goose; there were no
papers on shorebirds. A few chapters on the
status and trends of these and other birds
would have strengthened the book consider-
ably. There was no mention of Birdlife Inter-
national’s Important Bird Areas program in
Europe, the Middle East, Canada, and Mexi-
co, and a similar program by National Au-

dubon Society and the American Bird Con-
servancy in the US was also overlooked. A
review of the many approaches, their success-
es and failures, and the research questions
they require would have greatly strengthened
Avian Conservation. These oversights suggest
that a wide gap remains between the two sol-
itudes of research and conservation manage-
ment, at least in the USA. The five chapters
on land management were written by Ameri-
cans and for Americans. The superb reviews
in Avian Conservation will appeal to conser-
vationists world-wide and should be on their
shelves as an up-to-date summary of field and
a reference source. However, its applicability
is limited largely to a North American audi-
ence.—ROBERT W. BUTLER.

RUDDY DUCKS AND OTHER STIFF-
TAILS: THEIR BEHAVIOR AND BIOLO-
GY. By Paul A. Johnsgard and Montserrat
Carbonell. University of Oklahoma Press,
Norman, Oklahoma. 1996: 291 pp., 16 color
photos with captions, 33 numbered text fig-
ures including line drawings and range maps,
19 tables, 13 black-and-white illustrations.
$49.95 (cloth). — The collaboration of these
two authors brings together a wealth of ex-
perience with wild and captive stifftails, and
with studies of waterfowl natural history and
behavior in general. The result is a fine book
on a fascinating group of ducks that includes
excellent illustrations and much detailed in-
formation. Although there are few color plates
(photos of both sexes of all species covered
would be helpful), there are nice illustrations
of each species and excellent drawings of dis-
play behavior by the first author, many of
which are tracings from films or photos.

The first section of the book covers general
characteristics of the stifftails, effectively in-
troducing a number of interesting features of
this group. The summary chapters that follow
make especially good use of the second au-
thor’s thesis work on a variety of captive stiff-
tails at The Wildfowl and Wetlands Trust in
England. Covered as the stifftail group are
members of the genera Heteronetta, Nomonyx,
Oxyura, and Bizinra. Although the White-
backed Duck (Thassalornis leuconotus) and
Freckled Duck {Stictonetta naevosa) are no

ORNI FHOLOGICAL LITERATURE

451

longer considered to be stifftails, some com-
ments are still included where relevant (es-
pecially display behavior). Chapter 1 begins
with a detailed historical treatment of the tax-
onomy of the stifftails, and concludes with a
synthesis of available information and a dis-
cussion of remaining questions (some of
which have recently been tackled through ge-
netic analyses). Chapter 2 presents a detailed
summary of stifftail morphology and anato-
my, including a discussion of the adaptations
for diving that these species possess and com-
parisons between them. Molts and plumages
are also treated in this chapter, but in a very
general way, with detailed accounts appearing
in species chapters when possible. I believe
that it would have been helpful to include a
clear summary of at least the most common
molt patterns of stifftails here as an overview.
Chapter 3, General Behavior and Ecology, at-
tempts to summarize the postures and loco-
motion, comfort movements, time budgets,
feeding behavior, habitats, dispersal, migra-
tion, and important interspecific interactions
of the stifftails. Accounts of comfort move-
ments are unusually complete due to a com-
bination of information obtained by the au-
thors and previous work by Frank McKinney.
Time budget data and dive durations are pri-
marily available only from captives, and diet
data are presented from previous studies on
wild birds. Especially rare for waterfowl are
the data on dive durations and activity budgets
for ducklings.

Sex ratios, pairbonds, ritualized display be-
havior, aggression and territoriality, and con-
tributions of display behavior to taxonomic re-
lationships are presented in Chapter 4 (Com-
parative Social and Sexual Behavior). This
chapter also is based on a mixture of data
from captive and wild birds. Unfortunately,
some of the sample sizes from captives are
small, and there is a general lack of data on
number of individuals and variability, making
it difficult for the reader to interpret the level
of support for the authors’ statements. Also,
the display information focuses almost com-
pletely on male displays in this chapter. A ta-
ble of male stifftail structures and displays
makes for an easy comparison of behavioral
similarities, and hints at some of the recent
findings on the taxonomy of this group using
genetic characters. Reproductive and popula-

tion biology are summarized in Chapter 5, in-
cluding a plea for more studies on species oth-
er than the well-known North American Rud-
dy Duck {Oxyura jamaicensis). To underscore
this point, the timing of pairbonding, breed-
ing, and nesting, and hatching success, brood
behavior, renesting, and annual recruitment in
this chapter are only available in any detail
for the North American Ruddy Duck. Much
of this information was drawn from the same
three studies. Information from more species,
including data from captives, is summarized
for nest site characteristics, eggs and laying
behavior, clutch sizes, and duckling weights,
often in tables that allow comparisons to be
made between the stifftails. Nest parasitism is
discussed in some detail, including an indi-
cation of reproductive success relative to host
nests for North American Ruddy Duck and
Black-headed Duck (Heteronetta atricapilla),
an obligate nest parasite.

The second section of the book. Chapters
6-13, presents species accounts that include
vernacular names, range of species and races,
measurements and plumage descriptions,
identification cues, ecology, annual cycle, so-
cial and sexual behavior, nesting and parental
behavior, and reproductive success and status.
Range maps have been updated and improved
from Johnsgard’s Ducks, Geese, and Swans of
the World (1978). In general, the species ac-
counts present a good degree of specific data,
and sample sizes are given for measurements
and some other data. Plumage and soft tissue
descriptions are supplemented by references
to the location of photographs or illustrations
in the literature, and for each species there are
detailed drawings or tracings from film of
courtship display postures or sequences in ad-
dition to written accounts. Although phonetic
descriptions of calls are given, sonagrams
would have been useful. Identification of birds
in the hand and in the field is noted, and an
Appendix features a dichotomous key to in-
hand identification with drawings of adults
and ducklings.

The section on ecology for each species in-
cludes habitat, density, foods, foraging, and a
short coverage of competitors, predators, and
symbionts. Data on densities, foods con-
sumed, and predators are unavailable for
many species, and the discussions of compet-
itors are mostly speculations about diet over-

452

THE WILSON BULLETIN • Vol. Ill, No. 3, September 1999

laps with various waterbirds and other stiff-
tails. Symbionts apparently refers to parasitic
egg-laying interactions, and these data would
appear to have been more appropriate in the
section on nest choice. Movements and mi-
grations, molts and plumages, and the breed-
ing cycle are discussed as parts of the annual
cycle, with limited information on wild birds
for the lesser-known species. Descriptions of
the mating system, territoriality, courtship and
pairbonding, and copulatory behavior are in-
cluded under social and sexual behavior.
Again, information is limited for wild birds
for a number of species, and captive studies
have provided the majority of the detailed de-
scriptions of courtship displays and copula-
tion. The use of “rape behavior” is outdated
and even in the book’s Glossary the reader is
instructed to “See forced copulation”. The
nesting and parental behavior subsection in-
cludes nest choice and egg laying, and hatch-
ing and brood-related behavior. A mixture of
data from wild and captive birds expands the
coverage of these topics appreciably, although
in general, little information on duckling be-
havior is known for any of the ducks. Aspects
of reproductive success and status for each
species were summarized in tables in the first
section of the book, so in the species accounts,
the focus is primarily on population estimates
and conservation challenges (although cover-
age is a little disappointing on this topic).
Brought to light here is the especially alarm-
ing predicament of the White-headed Duck
iOxyura leucocephala), which is threatened
by overhunting, habitat destruction, and hy-
bridization and interactions with the intro-
duced North American Ruddy Duck. The sta-
tus of most of the .southern hemisphere species
has been difficult to determine because of a
lack of focused studies, but as human activi-
ties continue to increase pressure on wetland
habitats, we cannot afford our ignorance if
these species, and many others, are to survive.

The book is peppered with a number of ty-
pographical errors, but it is quite readable and
a glossary is included to aid the lay reader. In
general, the references throughout both sec-
tions of the book do not seem to be as up-to-
date as they could have been, probably as a
result of delays between preparation and pub-
lishing. Some unpublished material supple-
ments the text, and the references section in-

cludes sources not cited in the text. This book
goes a long way towards identifying where
our gaps in knowledge lie and what future
studies are needed on this interesting group of
birds. 1 recommend it as a valuable and quite
easy to use resource for researchers and stu-
dents of waterfowl behavior and ecology or
others particularly interested in the stiff-
tails.— GWENDA L. BREWER.

THE NUTHATCHES. By Erik Matthysen,
illus. by David Quinn. T & A D Poyser, Lon-
don, U.K. U.S. edition published by Academic
Press, San Diego, CA. 1998: xx-l-315 pp., one
color plate, many black & white drawings, 17
black & white photos, 103 figs., 29 tables
$39.95. — Books devoted to single bird fami-
lies are in vogue nowadays and several series
are being published. The series coming from
the Poyser company of England differs from
most in that the books do not attempt to be a
field guide and they have no extensive color
plates. Instead most of them are thorough
studies of the natural history of the species
involved.

The nuthatch volume is of that nature.
While treating all 24 species of the genus Sit-
ta, approximately 60% of the space is devoted
to the Eurasian Nuthatch {S. europaea). The
author has spent a major portion of his pro-
fessional career studying this species starting
with an undergraduate thesis in 1982. The
species is treated in 9 chapters: Taxonomy,
Morphology and Moult; Habitat and Popula-
tion Density; Foraging, Food and Hoarding;
The Pair and Its Territory; Breeding Biology;
Finding a Territory; Dispersal and Migration;
Population Dynamics; and Nuthatches in For-
est Fragments. Each chapter bristles with data,
often of a sort not usually found in avian life
history studies. Tables and graphs abound. Be-
sides his own data from Belgium, the author
gathered data from the literature from
throughout the range of the species.' The result
is an exemplary life history account.

The other 23 species are treated in four
chapters. In so far as possible the species are
treated by the same topics as above, but un-
derstandably very little is known about some
aspects of the biology of some species.

The Mediterranean Nuthatches are three

ORNITHOLOGICAL LITERATURE

453

Red-breasted Nuthatch (5. canadensis) look-
alikes: Algerian (5. ledanti), Corsican (S. whi-
teheadi), and Krueper’s (5. krueperi) nut-
hatches. All these have very limited distribu-
tion and the Algerian Nuthatch was discov-
ered as late as 1975. The story of the
discovery of this species is told in detail. The
original population estimate was 12 pairs in a
very restricted area but more recently the es-
timate is somewhere between 500 and a few
thousand pairs as the result of the discovery
of additional populations. All three of these
species are susceptible to threats due to loss
of habitat.

The two Rock Nuthatches (5. tephronota
and S. neumayer) have abandoned the forest
habitat of the rest of the family and are found
on rocky slopes and cliffs in dry regions from
Yugoslavia to Pakistan. The two overlap in
Iran and eastern Turkey, and it was not real-
ized until 1911 that they were two species.

Fourteen species are covered under the col-
lective heading. Oriental Nuthatches. These
range from the Himalayas to the Philippines,
and as many as 7 species can be found to-
gether in some parts of southeastern Asia. As
might be expected many of these are not well
known and the treatment is less detailed than
for other species. Included in this group are
the 195 mm Giant Nuthatch (5. magna), the
Blue Nuthatch (5. azurea), and the Beautiful
Nuthatch (S. formosa) both of which depart
from the drab coloration of most of the family,
as well as two more Red-breasted look-alikes.

The final chapter discusses the 4 New
World Nuthatches: White-breasted {S. caroli-
nensis). Red-breasted (5. canadensis), Brown-
headed (S. pusilla), and Pygmy (5. pygmaea).
Despite the abundance of these species none
of them has been studied to the extent that S.
europaea has. The Red-breasted is the only
migrating nuthatch. The “Dwarf” nuthatches,
S. pusilla and S. pygmaea, are unique in the
family in having small, often overlapping, ter-
ritories, communal roosts, and extra male
helpers at the nest.

The evolutionary history of the S. canaden-
sis superspecies which consists of one species
in northern North America, three species in
the Mediterranean region, and two geograph-
ically separated species in China presents an
interesting puzzle.

David Quinn’s black-and-white sketches.

many illustrating behaviors, enhance the book
and the single color plate illustrates four spe-
cies that will be unfamiliar to British and
American readers. As of present knowledge
this appears to be the definitive work on the
family Sittidae. — GEORGE A. HALL.

WORKING EOR WILDLIFE: THE BE-
GINNING OF PRESERVATION IN CANA-
DA. By Janet Foster, with a foreword and an
afterword by Lome Hammond. University of
Toronto Press, Toronto. Second ed. 1998: 297
pp., 38 black and white photographs, 5 maps.
$21.95 (paper). — In 1904, Howard Douglas,
Superintendent of Canada’s Rocky Mountains
Park, learned that Michel Pablo wanted to sell
a large herd of bison he kept in western Mon-
tana. Although as many as 20 to 30 million
of the animals had once roamed the North
American continent, by the end of the Nine-
teenth Century their numbers had been
thinned to fewer than a thousand. Pablo’s
herd, which he estimated at around 360 indi-
viduals, represented the largest surviving ag-
gregation of a species that seemed to be rush-
ing headlong into oblivion. Although Ameri-
can conservationists hoped to keep Pablo’s bi-
son in the United States, they failed to secure
the necessary funds. At Douglas’s urging, the
Canadian government purchased the herd and
shipped it by rail to Rocky Mountains Park.

This story is one of many fascinating epi-
sodes that Janet Foster recounts in her study
of the origins of Canadian wildlife conserva-
tion. According to Foster, the first significant
efforts to address wildlife decline in Canada
came at the end of the nineteenth century,
when a handful of senior federal civil servants
began using their position and influence to
push a protectionist agenda. Howard Douglas
and his dedicated colleagues — Robert Camp-
bell, Director of the Forestry Branch, Depart-
ment of Interior; James Harkin, Commissioner
of Dominion Parks; Maxwell Graham, Chief
of the Parks Branch Animal Division; and
Gordon Hewitt, Division Entomologist with
the Department of Agriculture — are the main
protagonists in Eoster’s account, which begins
in the mid 1880s and ends in the early 1920s.

None of this small group was a particularly
prominent public figure, and only one, Harkin,

454

THE WILSON BULLETIN • VoL III, No. 3, September 1999

had any formal training in the biological sci-
ences. Yet, working together with provincial
officials, other interested citizens, and their
counterparts in the United States, they were
remarkably successful in transforming their
personal commitment to wildlife into federal
policy. Fighting a pervasive belief in the su-
perabundance of nature, a national agenda that
emphasized settlement and development, a
public that seemed largely indifferent to the
desperate plight of wildlife, and a tradition
that left resource management in the hands of
provincial governments, these federal officials
established national parks, created wildlife
preserves, rallied public support for native
species, and pushed through protective legis-
lation and regulations. Aiding this quintet of
federal civil servants was a larger cast of char-
acters who receive much less attention in this
book, including the ornithologists Hoyes
Lloyd, Percy Tavener, James Fleming, and
others.

Foster’s account of the Migratory Bird
Treaty, negotiated in the years around World
War I, will be of particular interest to readers
of this journal. After more than a decade of
lobbying, in 1913 wildlife advocates in the
United States finally secured a federal law
protecting migratory birds. Fearing that the
new legislation might be struck down on con-
stitutional grounds, the bill’s supporters then
moved to have its provisions introduced into
a treaty with Great Britain. Negotiations soon
bogged down, however, when officials from
the Maritime Provinces balked at the idea of
eliminating spring shooting. Foster demon-
strates Hewitt’s central role in garnering Ca-
nadian support for this landmark treaty, which
remains in effect to this day.

This book is a second edition of a work first
published two decades ago. The environmen-
tal historian Lome Hammond has contributed
the only significant additions: a new foreword,
which briefly describes the larger context of
Foster’s book, and a new afterword, which re-
views the literature on Canadian wildlife con-
servation published since the first edition. If
Foster were to write her book today, undoubt-
edly she would pity more attention to the con-
tributions of sportsmen, naturalists, humani-
tarians, and provincial wildlife officials — all
of whom have received much scholarly atten-
tion since the first edition. Yet, because her

book is so well written and based on solid
archival research, it remains a useful starting
point for anyone interested in the early history
of wildlife conservation in Canada. — MARK
V. BARROW, JR.

HABITATS FOR BIRDS IN EUROPE: A
CONSERVATION STRATEGY FOR THE
WIDER ENVIRONMENT. Compiled by Gra-
ham M. Tucker and Michael I. Evans.
BirdLife Conservation Series No. 6, BirdLife
International, Cambridge, U.K. 1997: 464 pp.,
6 appendices. $45.00 (paper) (in North Amer-
ica, contact via email: BTUCKER®
SIPRESS.SI.EDU). — This ambitious compi-
lation from 8 habitat working groups marks
the third and final leg of a decadal marathon
sponsored by BirdLife to promote the conser-
vation of Europe’s birds (the first two culmi-
nated in Grimmet and Jones’ 1989 Important
Bird Areas in Europe and number 3 in the
BirdLife series. Tucker and Heath’s 1994
Birds in Europe: their conserx’ation status).
Thirteen workshops were held across Europe
bringing experts together to prepare conser-
vation strategies for each of the following ma-
jor habitat types: marine habitats; coastal hab-
itats; inland wetlands; tundra, mires (bogs),
and moorlands; lowland Atlantic heathland;
boreal and temperate forests; Mediterranean
forest, shrubland, and rocky habitats; and ag-
ricultural and grassland habitats. For each
habitat type, information is provided on cur-
rent distribution (with maps) and trends, its
history, physical and biological processes, and
its dominant flora and fauna. In addition, the
chief values of habitat to humans is given, and
the major threats to the habitat quality and
quantity of priority bird species are identified
in both text and tables. The last section of
each habitat chapter then lists conservation
opportunities such as legislation, financial in-
centives, and policy initiatives, then broad
conservation recommendations are given.

The rationale for priority bird rankings is
explained early in the book, providing a very
useful model for other large-scale bird (or oth-
er fauna) conservation efforts around the
globe. In this scheme, 5 classes are established
of “Species of European Conservation Con-
cern” (or SPECs); SPEC 1 species are of

ORNITHOLOGICAL LITERATURE

455

global concern (rare, endangered, or declining
populations), while SPEC 4 and 5 species
have favorable conservation status in Europe.
Next, priority categories (from A to D) for
bird species in each habitat are established us-
ing a matrix of SPEC category X habitat im-
portance (percent of European population us-
ing that habitat). Thus, Priority A species in
any particular habitat are those most vulner-
able to further losses, while Priority D species
are those that are more stable and widespread,
with less dependence upon that particular hab-
itat.

Next, the principles and strategies for broad
conservation initiatives in Europe are out-
lined, spanning across international treaties
and conventions, economic instruments, and
policy doctrine. A dazzling litany of some 25
legislative instruments are reviewed, from
global to more local European perspectives,
ranging from specifics (agricultural nitrate
control policy) to broad measures of biodi-
versity (so-called Rio Convention of 1992).
Then, numerous economic instruments are re-
viewed (e.g., various European Union, EU,
and World Bank funds) followed by other
broad initiatives (e.g.. Birds and Habitats Di-
rectives in EU). It would seem that with this
bewildering array of conventions and plans
cutting across landscapes and political bound-
aries, that Europe’s conservation needs would
all be well taken care of! But alas, as with
most large Plans, “the devil is in the de-
tails”— developing consensus for habitat pro-
tection, harvest criteria, or emission standards
among an array of nations with vastly differ-
ent ideologies and histories in human-nature
interactions is difficult.

A survey across the habitat chapters reveals
the following order (from most to least) from
the perspective of priority species: (1) Agri-
cultural and grassland species — 173 species
(ca 70% unfavorable conservation status),
with 6 Priority A species (4 are SPEC 1); (2)
Boreal and temperate forests — 114 species
(40% unfavorable status), with 2 Priority A
species; (3) Inland wetlands — 102 species
(55% unfavorable), with all 8 Priority A spe-
cies being SPEC 1; (4) Mediterranean forest,
shrubland, and rocky habitats — 100 species
(65% unfavorable), with 10 Priority A species
(1 SPEC 1); (5) Coastal habitats — 75 species
(70% unfavorable), with 13 Priority A species

(5 are SPEC 1); (6) Tundra, mires, and moor-
land— 73 priority species (37% unfavorable),
with only 2 Priority A species (1 SPEC 1); (7)
Marine habitats — 62 species (45% unfavor-
able), with 6 Priority A species (1 SPEC 1);
(8) Lowland Atlantic heathland — 16 species
(all small populations), no Priority A or B spe-
cies.

Some recurrent themes are the need to in-
tegrate habitat conservation planning with
other sectors and programs using mechanisms
such as Environmental Impact Assessment
and Strategic Environmental Assessment, to
modify established policies within the EU
such as Common Agricultural Policy and
Common Eisheries Policy to work toward sus-
taining biodiversity rather using the more my-
opic traditional focus, to remove economic in-
centives that destroy habitat and reduce di-
versity (e.g., non-indigenous tree plantations),
to better educate the public about the benefits
and ecological services of such natural habi-
tats, and to work to develop standards and cri-
teria for sustainability of habitats across na-
tional boundaries. The fact that nearly 40% of
Europe’s bird species show an unfavorable
conservation status is largely due to intensi-
fication. That is, intensification of farming and
silvicultural practices on land, aquaculture and
open sea fishing in coastal and marine habi-
tats, and of coastal development, recreation,
and tourism, much of the latter especially in
Mediterranean Europe.

This book represents an enormous effort by
many experts. The details are displayed in the
extensive appendices, the figures and tables
are very useful in summarizing the vast
amounts of species, habitat, and threat infor-
mation, and the chapters are neatly and co-
herently packaged so that they can nearly
stand alone for those with more specific in-
terests. The approaches taken in prioritizing
the species and their threats, and the conser-
vation recommendations made by each habitat
working group provide an extremely valuable
reference for bird conservationists in any con-
tinent, not simply Europe. The challenges now
lie in translating all of those directives, con-
ventions, and biodiversity initiatives into real
Action Plans that can survive the turbulent po-
litical and economic seas, especially as the
new Euro currency takes hold. This book
should be a library requisite for conservation-

456

THE WILSON BULLETIN • VoL HI, No. 3, September 1999

ists, land managers, and environmental policy
professionals because it transcends issues of
bird conservation and prompts us to consider
true integration not just of land- and ocean-

scapes and their intersections, but also of ecol-
ogy, agriculture, forestry, fisheries, econom-
ics, and of course, politics. — R. MICHAEL
ERWIN.

Wilson Bull., 1 1 1(3), 1999, pp. 456

Announcement

The Lincoln Park Zoo Scott Neotropic and
Africa/Asia Funds support field research in
conservation biology around the world. The
Scott Neotropic fund focuses on projects un-
dertaken in Latin America and the Caribbean.
The fund emphasizes the support of graduate
students and other young researchers, partic-
ularly those from Latin America. Since 1986,
the fund has awarded over 126 grants in 19
countries. The Africa/Asia fund, lauched in
1997, focuses on projects throughout Africa,
Asia, and the Pacific. Each fund supports
projects of young conservation biologists and
between 5 and 15 projects for each fund are

supported each year. The fund awards are sel-
dom greater than US$7500, and most awards
fall in the range of $3000-$6000. Initial sup-
port is for up to 12 months from the date of
award, and the maximum duration of support
is two years. The current deadline for receipt
of Scott Neotropic proposals is 1 September,
and Africa/Asia proposals have no deadline
for 1999. For additional information and ap-
plication procedures go to www.lpzoo.com,
email steveedC® ix.netcom.com, or write to:
LINCOLN PARK ZOO SNF/AA FUNDS,
% Director of Conservation and Science,
Lincoln Park Zoo, Chicago, IL 60614.

This issue of The Wilson Bulletin was published on 10 August 1999.

THE WILSON BULLETIN

Editor ROBERT C. REASON

Editorial Board KATHY G. BEAL

Department of Biology

State University of New York

1 College Circle

Geneseo, NY 14454

E-mail: WilsonBull@geneseo.edu

Review Editor WILLIAM E. DAVIS, JR.

127 East Street

Foxboro, Massachusetts 02035

CLAIT E. BRAUN
RICHARD N. CONNER

Editorial Assistants TARA BAIDEME

Index Editor KATHY G. BEAL

JOHN LAMAR
DANTE THOMAS
DORIS WATT

616 Xenia Avenue
Yellow Springs, Ohio 45387

SUGGESTIONS TO AUTHORS

NOTICE OF CHANGE OF ADDRESS

If your address changes, notify the Society immediately. Send your complete new address to Ornitho-
logical Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

MEMBERSHIP INQUIRIES

Membership inquiries should be sent to Laurie J. Goodrich, Route 2 Box 301 A, New Ringgold, PA
17960-9445; E-mail: goodrich@haukmountain.org.

CONTENTS

ANTILLEAN SHORT-EARED OWLS INVADE SOUTHERN FLORIDA

- Wayne Hoffman, Glen E. Woolfencien, and P. William Smith 303

WITHIN- AND BETWEEN-YEAR DISPERSAL OF AMERICAN AVOCETS AMONG MULTIPLE

WESTERN GREAT BASIN WETLANDS

Jonathan H. Plissner, Susan M. Hai^, and Lewis W. Orin^ 3 1 4

HIGH MORTALITY OF PIPING PLOVERS ON BEACHES WITH ABUNDANT GHOST CRABS:

CORRELATION, NOT CAUSATION Donna L. Wolcott and Thomas G. Wolcott 321

A TAXONOMIC STUDY OF CRESTED CARACARAS (FALCONIDAE)

- - — Carla J. Dove and Richard C. Banks 330

VISUAL COMMUNICATION AND SEXUAL SELECTION IN A NOCTURNAL BIRD SPECIES, CA-
PRIMULGUS RUF/COLUS, A BALANCE BETWEEN CRYPSIS AND CONSPICUOUSNESS ...

Juan Aragones, Luis Arias De Reyna, and Pilar Recuerda 340

INTERSPECIFIC INTERACTIONS WITH FORAGING RED-COCKADED WOODPECKERS IN

SOUTH-CENTRAL FLORIDA

Reed Bowman, David L. Leonard, Jr., Leslie K. Backus, and Allison R. Mains 346

SPATIAL AND TEMPORAL DYNAMICS OF A PURPLE MARTIN PRE-MIGRATORY ROOST ......

- - Kevin R. Russell and Sidney A. Gauthreau.x, Jr. 354

AGGRESSIVE RESPONSE OF CHICKADEES TOWARDS BLACK-CAPPED AND CAROLINA

CHICKADEE CALLS IN CENTRAL ILLINOIS Eric L. Kershner and Eric K. Bollinger 363

USE OF SONG TYPES BY MOUNTAIN CHICKADEES (POECILE GAMBELL)

Myra O. Wiehe and M. Ross Lein 368

SURVIVAL AND LONGEVITY OF THE PUERTO RICAN VIREO

- - Bethany L. Woodworth, John Faahorg. and Wayne J. Arendt 376

EFFECTS OF PRIOR RESIDENCE AND AGE ON BREEDING PERFORMANCE IN YELLOW WAR-
BLERS G. A. Lozano and R. E. Lemon 381

DISTRIBUTION AND HABITAT ASSOCIATIONS OF THREE ENDEMIC GRASSLAND SONG-
BIRDS IN SOUTHERN SASKATCHEWAN S. K. Davis, D. C. Duncan, and M. Skeel 389

BIRD COMMUNITIES IN NATURAL FOREST PATCHES IN SOUTHERN BRAZIL

Luiz Dos Anjos and Roberto Bo^ on 397

DO MAMMALIAN NEST PREDATORS FOLLOW HUMAN SCENT TRAILS IN THE SHORTGRASS

PRAIRIE? Susan K. Skagen, Thomas R. Stanley, and M. Beth Dillon 415

SHORT COMMUNICATIONS

CHRISTMAS SHEARWATER EGG DIMENSIONS AND SHELL CHARACTERISTICS ON LAY-

SAN ISLAND, NORTHWESTERN HAWAIIAN ISLANDS

- G. C. Whittow and M. B. Naughton 42 1

THE PAINT-BILLED CRAKE BREEDING IN COSTA RICA

David M. Watson and Brett W. Benz 422

ADDITIONAL RECORDS OF FALL AND WINTER NESTING BY KILLDEER IN SOUTHERN

UNITED STATES Kimberly G. Smith, W. ManJn Davis, Thomas E. Kienzle,

William Post, and Robert W. Chinn 424

WILD TURKEYS (MELEAGRIS GALLOPAVO) RENEST AFTER SUCCESSFUL HATCH

— Craig A. Harper and Jay H. E.xum 426

POST-MIGRATION WEIGHT GAIN OF SWAINSON’S HAWKS IN ARGENTINA

Michael I. Goldstein, Peter H. Bloom. Jose H. Sarasola, and Thomas E. Lacher 428

SIBLICIDE AT NORTHERN GOSHAWK NESTS: DOES FOOD PLAY A ROLE?

. Wendy A. Estes, Sarah R. Dewey, and Patricia L. Kennedy 432

COOPERATIVE FORAGING IN THE MOUNTAIN CARACARA IN PERU Ja.son Jones 437

PREDATION BY RUFOUS MOTMOT ON BLACK-AND-GREEN POISON DART FROG

Terry L. Master 439

EVIDENCE OF EGG EJECTION IN MOUNTAIN BLUEBIRDS Percy N. Hebert 440

FORAGING OVENBIRD FOLLOWS ARMADILLO .. Douglas J. Uvey ,443

ORNITHOLOGICAL LITERATURE ...._ 445

The Wilson Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. Ill, NO. 4 DECEMBER 1999 PAGES 457-630

(ISSN 0043-5643)

THE WILSON ORNITHOLOGICAL SOCIETY
FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President John C. Kricher, Biology Department, Wheaton College, Norton, Massachusetts 02766;
E-mail; JKricher@wheatonma.edu.

First Vice-President — William E. Davis, Jr., College of General Studies, 871 Commonwealth Ave., Boston
University, Boston, Massachusetts 02215; E-mail; WEDavis@bu.edu.

Second Vice-President — Charles R. Blem, Dept, of Biology, 816 Park Ave., Virginia Commonwealth
Univ., Richmond, Virginia 23284; E-mail; cblem@saturn.vcu.edu.

Editor — Robert C. Beason, Department of Biology, State University of New York, 1 College Circle,
Geneseo, New York 14454; E-mail; WilsonBull@geneseo.edu.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper Montclair,
New Jersey 07043; E-mail; Smallwood@mail.montclair.edu.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame, Indiana 46556;
E-mail; DWatt@saintmarys.edu.

Elected Council Members — Charles F. Thompson and Sara R. Morris (terms expire 2000), Jonathan L.
Atwood and James L. Ingold (terms expire 2001), Robert A. Askins and Jeffrey R. Walters (terms
expire 2002).

Membership dues per calendar year are; Active, $21.00; Student, $15.00; Family, $25.00; Sustaining,
$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 of books, reprints, and orni-
thological magazines from members and friends of the Society. Two members have generously established
a fund for the purchase of new books; members and friends are invited to maintain the fund by regular
contribution. The fund will be administered by the Library Committee. William A. Lunk, University of
Michigan is Chairman of the Committee. The Library currently receives over 200 periodicals as gifts and
in exchange for The Wilson Bulletin. For information on the library and our holdings, see the Society’s
web page at http;//www.ummz. Isa. umich.edu/birds/wos.html. 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; Josselyn Van Tyne Memorial Library, Museum of Zoology, The
University of Michigan, 1 109 Geddes Ave., Ann Arbor, MI 48109-1079 USA. Contributions to the New
Book Fund should be sent to the Treasurer.

THE WILSON BULLETIN
(ISSN 0043-5643)

THE WILSON BULLETIN (ISSN 0043-5643) is published quarterly in March, June, September, and December by
the Wilson Ornithological Society, 810 East 1 0th Street, Lawrence, KS 66044-8897. The subscription price, both in the
United States and elsewhere, is $40.00 per year. Periodicals postage paid at Lawrence, KS. POSTMASTER: Send address
changes to THE WILSON BULLETIN, P.O. Box 1897, Lawrence, KS 66044-8897.

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.

Subscriptions, changes of address and claims for undelivered copies should be sent to the OSNA, P.O. Box 1897,
Lawrence, KS 66044-8897. Phone: (785) 843-1221; FAX: (785) 843-1274.

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

MCZ

LIDRA.^Y

btU 22 1^99

HA'^V'AR'O

UlMlVZi^JirY

THE WILSON BULLETIN

A QUARTERLY JOURNAL OF ORNITHOLOGY
Published by the Wilson Ornithological Society

VOL. Ill, NO. 4 DECEMBER 1999 PAGES 457-630

Wilson Bull., 111(4), 1999, pp. 457-464

A NEW SPECIES OF HAWK-OWL NINOX FROM NORTH

SULAWESI, INDONESIA

PAMELA C. RASMUSSEN*

ABSTRACT. — A distinctive new species of hawk-owl, Ninox ios, is described from a specimen collected in
1985 in forest at 1120 m in Bogani Nani Wartabone (then Dumoga-Bone) National Park, North Sulawesi,
Indonesia. It was previously identified as a rufous morph of the Ochre-bellied Hawk-Owl, N. ochracea. Ninox
ios is small, predominantly bright chestnut, and lacks facial patterning; it has pink orbital skin, yellow irides,
triangular whitish scapular spots, a finely banded and relatively long tail, unusually short, slender tarsi that are
feathered for most of their length, and weak claws. Its relationships within the genus Ninox are unclear; it differs
in several morphological characters from all other species. Because Ninox ios is only known from one specimen,
its distribution and conservation status are unknown; nothing is known of its ecology, but it probably occurs
primarily at higher elevations than N. ochracea. Received 14 Dec. 1998, accepted 5 May 1999.

For many years two endemic species of the
genus Ninox were thought to occur on the cen-
tral Indonesian island of Sulawesi. Of these,
the Speckled Hawk-Owl (Ninox punctulata)
primarily inhabits disturbed lowland habitats
throughout the island (White and Bruce 1986),
and is morphologically quite different from
other endemic Indonesian Ninox. The poorly
known Ochre-bellied Hawk-Owl [A. ochracea
(= perversa)] of the lowland rainforests in
North and Central Sulawesi (White and Bruce
1986) is a small, fairly typical member of its
genus (Frontispiece). Because there had been
no indication that a third species might occur,
it was a surprise when in 1985 F. G. Rozen-

' NHB 336 MRC 1 14, Smithsonian Institution,
Washington, DC 20560, and Michigan State University
Museum, East Lansing, MI 48824-0590;

E-mail: rasmussen.pamela@nmnh.si.edu

daal netted an almost entirely bright rufous
Ninox (Frontispiece) in Bogani Nani Warta-
bone (then Dumoga-Bone) National Park,
North Sulawesi, Indonesia (Fig. 1). He con-
cluded that this individual represented “a pre-
viously undescribed rufous phase” of N.
ochracea (Rozendaal and Dekker 1989), and
this treatment was followed by Coates and
Bishop (1997).

While working on small owls at the Na-
tional Museum of Natural History/Naturalis,
Leiden (NNM, formerly Rijksmuseum van
Natuurlijke Historie, RMNH) in June and Oc-
tober 1998, I chanced to see the rufous Sula-
wesi specimen, which had been registered as
RMNH 84701 but had not yet been incorpo-
rated into the main collection following its
purchase by NNM. On the second occasion I
noted that it differed in several morphological
features from Ninox ochracea, in addition to

FRONTISPIECE. Cinnabar Hawk-Owl (Ninox ios, upper two) compared with Oehre-bellied Hawk-Owl (TV.
ochracea, lower left), and Bum race of Moluccan Hawk-Owl (TV. squamipila hantu, lower right). Original
watercolor painting by Ian Lewington.

457

458

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

FIG. 1. Map of Wallacea showing collection locality of the holotype of the Cinnabar Hawk-Owl (Ninox
ios), other locations mentioned in the text, and approximate ranges within the region of other species of Ninox
and subspecies of W. scutulata, which occurs throughout the region.

the obvious color differences. Subsequent
mensural analyses of series of all species of
Ninox have confirmed the distinctness of the
rufous Sulawesi specimen (an adult in good
condition) in many characters. Although sev-
eral Ninox species from other areas are typi-
cally rufous, morphism (and thus true rufous
morphs) appears to be unknown for any Ni-
nox, and in any case most of the differences
are structural and thus would not be related to
morph. Despite the fact that only one rufous
specimen is known thus far from Sulawesi,
there is no reason to believe that any of its
several novel character states are aberrant, and
there can be no reasonable doubt that it rep-
resents a new species.

CINNABAR HAWK-OWL Ninox ios sp. nov.

Holotype. — RMNH 84701, adult male
(Frontispiece), according to the label collected
in a forested valley at 1 120 m at Clark’s camp
(Hill 1440), east-central Bogani Nani Warta-
bone National Park, North Sulawesi, Indone-
sia (ca 0°40' N, 123° 0' E) by F. G. and C.
M. Rozendaal the night of 5-6 April 1985 (the
label date of 7 April presumably indicates date
of death). Label data: “Completely ossified
skull”, “weight 78 g”.

Diagnosis. — A small, lightly built, nearly
uniformly rich chestnut hawk-owl with a rel-
atively long tail and narrow pointed wings, lax
feathering, no facial pattern, mostly feathered

Rasmussen • A NEW SPECIES OF HAWK-OWL

459

short slender tarsi, and rufous, narrowly dark-
barred wings and tail.

Compared with all flying states of Ninox
ochracea [n = 20 (three of which are fully
grown juveniles); 6 males, 4 females, 10 un-
sexed], N. ios is much smaller in most dimen-
sions (Table 1, Fig. 2), but has a relatively
longer tail and rictal bristles. Its wing, while
shorter than that of N. ochracea, is narrower
and more pointed (Fig. 3). Ninox ios has a
much shorter, shallower bill and smaller nares
than N. ochracea. It has short, slender tarsi
that are mostly feathered on both surfaces,
whereas N. ochracea has longer, stout tarsi
that are largely unfeathered on the anterior
(acrotarsal) side and are virtually unfeathered
on the posterior (plantar) side, with numerous
stiff bristles over the unfeathered areas. The
new species has relatively sparse, fine rufous
bristles on the extreme lower tarsi and on its
slender toes (although the bristles are heavier
and longer on the hallux), while N. ochracea
has more profuse, heavier, mostly pale bristles
(which are usually longer but sometimes worn
down to stubs) on the tops and sides of its
stouter toes. Ninox ios has much smaller, more
slender claws that are dark for most of their
length (vs large and mostly pale in N. ochra-
cea). The holotype of TV. ios had pink orbital
skin (vs blackish in TV. ochracea) and yellow
eyes, as does TV. ochracea according to Strese-
mann (1940), who based this statement on G.
Heinrich’s specimens [although Meyer and
Wiglesworth (1898) mentioned a brown-eyed
TV. ochracea]. The base of the bill and the cere
of TV. ios appear entirely pale (vs the basal
two-thirds conspicuously dark in specimens of
TV. ochracea).

In plumage, TV. ios differs conspicuously
from both adults and juveniles of TV. ochracea
in its overall bright rufous coloration (vs dark
brown and yellow-ocher). Unlike all flying
stages of TV. ochracea, it lacks facial pattern-
ing, including the whitish supercilia typical of
most of its relatives, and also lacks white
markings in the wing coverts and flight feath-
ers. Less obvious distinctions from TV. ochra-
cea include its more triangular (vs squarer
tipped) whitish scapular spots, its mainly ru-
fescent rictal bristles (vs blackish with white
bases), its more narrowly barred rectrices, its
vaguely dark-scalloped lower underparts (vs
plain ocher or somewhat brown- streaked), and

the patterning of its breast feathers, which
have a light rufous (vs dark brown) area sur-
rounding the whitish shafts.

The Philippine Hawk-Owl {Ninox philip-
pensis) superspecies (sensu Dickinson et al.
1991, but see Collar and Rasmussen 1998) is
composed of several dark brown to brown-
and-ocher forms that are either barred or
streaked below. None of the taxa included in
TV. philippensis can be described as warmer-
toned than rufescent brown. All have much
heavier claws and relatively shorter tails (Fig.
2A) than TV. ios, from which they also differ
in wing shape (Fig. 3). One form, TV. [philip-
pensis] mindorensis (see Frontispiece), is
somewhat similar in overall size and tarsal
feathering to TV. ios than is any other taxon,
including TV. ochracea (Fig. 2C), but not in
plumage or the above-mentioned shape char-
acters.

All taxa of the paraphyletic Moluccan
Hawk-Owl {Ninox squamipila\ split provi-
sionally into at least three species by Norman
et al. 1998) are considerably larger and heavi-
er-legged than TV. ios, and all differ from it
additionally in having whitish-barred under-
parts and scapulars. Despite the above differ-
ences, Ninox s. hantu (Frontispiece) of Bum
superficially resembles the much smaller TV.
ios because of its overall mfescence and re-
duced barring below, as well as its obscure
facial pattern and finely barred tail. The Sum-
ba Hawk-Owl (TV. rudolfi) is large and strik-
ingly different, with a heavily spotted crown,
barred underparts, and broadly banded and
speckled upperparts. The widespread and var-
iable Brown Hawk-Owl {Ninox scutulata) is
also a much larger species, with a broadly
banded tail and large, heavily feathered tarsi.
It is dark brown above with the underparts
heavily streaked, or nearly solid dark brown
in TV. s. obscura of the Andamans. The nom-
inate race of the Andaman Hawk-Owl (TV. a.
affinis) is smaller than TV. scutulata, to which
it is otherwise quite similar, while the larger
Nicobar race (TV. a. isolata) is even more like
some races of TV. scutulata.

The highly varied subspecies (including a
new one described from Roti Island, south-
west of Timor, Lesser Sundas; Johnstone and
Darnell 1997) usually grouped in the Southern
Boobook {Ninox novae seelandiae) as well as
the Manus Hawk-Owl (TV. meeki) are also

460

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

f \ r"

(j V
CL ^

c/5

C !

o •

s '5

> *5

- ^

-* ^'.SJ

I d

^ s

<£ ° M

_ c

- "5 ^

(/)

C -O
o u

C/5

S3 ^
1) ^

+1-^2
+1 -

__ SJ

ULl V

J i
CQ .S

U«

c/5

<D .ti

c/5

2 ofl

O c^

ro fN St O

D. 0

fO ro fO fO

C^5

" (N

d) 1

-C —

*§

;:;

3.9

3.5

sd O; ;

04 d d 04

Ci.

■

O CJ

<9j

+1 +1 +1 +1

OX)

CL C

tN NO St

NO d d 1/5

—H

I*-*

— -

^ -=

c/5

E ji

OJJ

to ^

04 — OS O'

•o

SO so m »n

“O

3 ^

ON On —

o 2

— d d 04

^ 2

s/3

<<1

+1 +1 +1 +1

p o

O 3- St NO

—

§;

,-H

sd d d in

,(D

(U Nm

£ 3

2 o

■<— *

C/5 lU

^ 3

y s

-5 3

in so so so

04 04 04 04

OX) ^

2 13

5.2

1.7

1.2

0.8

2.1

2.6

2.7

2.8

3.9

— ■£

+1 +1 +1 +1

+ 1

c-^

2 «

■l

— ' -t ro NO

•d

O'!

— d 2

»— 1

C/5 d)

_ x:

&- d:;

- o

c a

o -

/-'V

r-s

c/5

S',

M x;

IT)

Wi -w

> E

ON ro — .

1 1 1 1

04 O O loi

r- “O

w w w w

3 QJ

r- m

0 E

u ^

—

— '

^ ^ ^

M s

§ 2

ir)'r'--00O000N00^O'O'Ot^0N— I'sTroiriO
— r-iOu-istr<S— 'O— -(NCNricNCNVO— "OOrO
+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
m — ^cNO(Nmr<^staNOt^t^— 'f^'Ooor^oo

d —

ro — O' lO O
fs n — 00 lo (N

0\ — ■ On 00 On m

'sj' On OJ

On NO O;

r'CNror'imvocNO
— 04 (N r' in —

NO O 00 NO
04— ' — (SOONl^t^fNldON— '
— <04rOrON^ONr4 —

s: c

w d)
&0 —
C d)

c/5

E «
^ 2
< Qi

— 3:

222222222

tcct^dctcdtt:

ooooooooc

^C/5C/5C/5C/5C/5C/5C/5C/5

-^rl^^i^»r)'Or-ooc^

cucuCua<(xa<cu(XCL

-C

c/5

Clij

-E

c/5

■S

c/5

£ S3
00
c

s s D

Rasmussen • A NEW SPECIES OF HAWK-OWL

461

Culmen from skull (mm)

Tarsus length (mm)

FIG. 2. Bivariate scatter plots (measurements in mm) for Ninox ios (filled circles), N. ochracea, N. “squam-
ipila” (sensu White and Bruce 1986), and N. philippensis (sensu Dickinson et al. 1991): A. culmen vs tail
length; B. auricular vs rictal bristle length; and C. tarsus length vs extent of unfeathered tarsus. For A, the main
Philippine taxon groups are treated separately, while for B and C they are combined.

462

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

60-

FIG. 3. Shortfalls from wingpoint of each primary
(PI = outer primary; shortfalls are distance of tip of
each primary from longest primary in folded wing) for
Ninox ios, N. ochracea, N. [philippensis] spilocephala,
and N. [/?.] spilonota. Ninox p. philippensis and similar
races are virtually identical in pattern of primary short-
falls to N. [p.] spilocephala and thus are not shown
separately, while all bar-bellied populations in the N.
philippensis superspecies are similar to race N. p. spi-
lonota.

larger than N. ios and are streaked or heavily
blotched below. None of the remaining Aus-
tralasian taxa (Papuan Boobook Owl, N. theo-
macha; Rufous Owl, N. rufa'. Powerful Owl,
N. strenua-, or Barking Owl, N. connivens) ap-
proach N. ios more closely than the above.
The other Sulawesi endemic. Speckled Hawk-
Owl (N. punctulata), and some Melanesian
taxa (Bismarck Hawk-Owl, N. variegata-. New
Britain Hawk-Owl, N. odiosa; and Solomons
Hawk-Owl, N. jacquinoti) are strikingly dif-
ferent in plumage and morphology, with short
tails, very heavy tarsi, and Athene plum-
age pattern and toe bristles; in fact some had
been placed in that genus (among others) in
the past. The White-browed Owl (Ninox su-
perciliaris) of Madagascar is very different
from other Ninox (H. F. James, pers. comm.)
as would be predicted by its distribution.

Distribution. — To date Ninox ios is known
only from the type locality in North Sulawesi,
Indonesia. It might occur at similar elevations
elsewhere in the Minahasa Peninsula of North
Sulawesi.

Description of the holotype. — Color match-
ing was done under natural light using Mun-

sell (1977) notation, in which the first number
and letters refer to the hue, the number pre-
ceding the slash is the value or lightness, and
the last number is the chroma or saturation.
The holotype was directly compared with 11
specimens of N. ochracea (including the ho-
lotype) at NNM, and a series of photographs
of it was compared to specimens from other
museums.

Front of head from base of bill through cen-
ter of forecrown and including supercilia, uni-
form rich chestnut (5YR 5/8); center of crown
through mantle slightly darker (close to 5YR
4/6); rictal bristles fairly long (maximum 24
mm), profuse, and dark chestnut, somewhat
blackish near tips; auriculars with fairly long
distally extended barbs (total length of longest
feather 23 mm) that are paler basally (5YR 5/
10) and grade to black near the tips; chin and
throat paler chestnut (5YR 6/8) than forehead.

Sides of neck and breast, back, rump, and
uppertail coverts are all approximately the
same rich dark chestnut (5YR 4/8). The un-
derparts appear very lightly dappled, slightly
paler chestnut (5YR 5/8) than upperparts.
Most breast feathers have pale shaft streaks
(5YR 7/8) and pale rufous surrounding areas,
some with darker dappling at sides, and feath-
ers of lower underparts are mostly pale rufous
with vague darker scalloping (2.5YR 5/8); un-
dertail coverts rufescent whitish with the tips
scalloped rufous (5YR 6/8).

The scapulars have large mostly triangular
whitish spots with broad dark chevron-shaped
tips (5YR 4/4). The upper secondary coverts
are almost uniformly rufous (5YR 6/8) and the
upper primary coverts are darker (5YR 4/2).
The remiges are faded, pale, and worn, in
striking contrast to the fresh, richly colored
scapulars. The inner webs of the primaries and
narrow vague dark bands of the outer webs
are dark grayish brown (5YR 4/4); only the
outer webs have broader light bands (5YR 7/
6). The base color of the secondaries is dull
rufescent ochraceous (7.5YR 6/8), with fine
dark dusky brown bars (7.5YR 4/4). The inner
webs of the undersurfaces of the inner pri-
maries and secondaries are basally pale rufous
(7.5 YR 8/6), as are the uppersurfaces of the
inner webs of the inner secondaries, which
contrast strongly with the dark bands. The un-
derwing coverts are solid pale rufous (7.5YR
7/8). The uppertail surface has pale bands of

Rasinussen • A NEW SPECIES OF HAWK-OWL

463

dull rufous (SYR 5/6) that are narrow basally
and wider distally, and about 12 narrow very
dark brown bands (SYR 3/2) that fade out to-
ward the tip. There are no definite bands for
the terminal 20 mm. The rectrices are heavily
worn and faded.

The short, slender tarsi are completely
feathered with short pale cinnamon (7. SYR 7/
6) pennaceous feathers to about 12 mm ante-
riorly (measured from joint of digits 1-2 of
middle toe) and posteriorly to about 6 mm
(measured from base of hallux). The toes ap-
pear to have been slender, with sparse, short
rufous bristles on the tops and sides of each
toe. The claws are small, delicate, and mostly
blackish but with pale bases.

The soft part colors recorded on the original
label are: eyes “bright yellow; pink orbital
skin”, bill “ivory”, feet “pale whitish-yel-
low”.

Measurements of the holotype (by au-
thor).— Culmen (from skull) 17.9 mm; cul-
men (from distal edge of cere) 10.7 mm; tar-
sus 22.6 mm; wing 172 mm; tail 97 mm. Total
length of prepared specimen 220 mm. See Ta-
ble 1 for measurements of other characters of
the holotype and those of other species.

Etymology. — This new species is named
Ninox ios (Greek for rust) for its striking over-
all coloration. The specific epithet is here used
as a noun in apposition to Ninox, which, al-
though usually treated as feminine, is a port-
manteau combining Nisus and Noctua. The
common name “cinnabar” also refers to its
predominant color, which is similar to that of
mercuric sulfide before prolonged exposure to
light.

DISCUSSION

Voice. — Not definitely known. Rozendaal
(Rozendaal and Dekker 1989:97) mentioned
“disyllabic calls ascribed to [N. ochracea] re-
corded at Clark’s camp and on the summit of
G.[unung] Muajat during April 1985.” Ek-
strom and co workers (1998:39) reported “an
unknown owl Ninox sp.” giving a series of
dry hoots rising and falling in pitch in dense
evergreen valley forest near the eastern
boundary of Lore Lindu National Park, at
about 1300 m, in the northern part of central
Sulawesi (J. Tobias, pers. comm.). Either of
these reports might refer to Ninox ios but con-
firming field data are required.

Habitat and elevation. — Most researchers
have considered N. ochracea to be restricted
to the lowlands below 800 m (Stresemann
1939, White and Bruce 1986, Stattersfield et
al. 1998). More recently, Coates and Bishop
(1997) gave the elevational range of N. ochra-
cea as up to 1780 m, but this was probably
based on the questionable vocal records men-
tioned in Rozendaal and Dekker (1989) and
the collection of the type of N. ios at 1120 m.
All montane records of N. ochracea therefore
require review in light of this new species.
Ninox ios clearly occurs in sympatry with, al-
though very likely at higher elevations than,
N. ochracea.

Molt, breeding, and ecology. — The holo-
type of N. ios clearly had recently molted its
scapulars, which were bright and fresh and
contrasted strikingly with the relatively dull
tertials and other flight feathers. The feathers
on the crown appeared to be worn, while those
of the back appeared fresh. Only 10 rectrices
were present. Active molt of the flight feathers
was not detected, but avoid damaging the
unique specimen a thorough examination was
not attempted. The size of the label drawing
of the largest testis (which measures 6X4
mm) suggests a bird not completely reproduc-
tively quiescent. Because nothing is known of
the habits of N. ios, it is possible only to spec-
ulate that its morphology (which recalls that
of owlet-nightjars Aegothelidae) suggests the
likelihood of its preying largely upon soft-
bodied invertebrates caught in flight.

Systematics. — The affinities of Ninox ios
are unclear; it shows many morphological dif-
ferences from all other species, particularly in
its small size, relatively long tail, narrow
pointed wing, and weak tarsi and claws. Al-
though membership in the polytypic N. phi-
lippensis superspecies might ‘^eem geograph-
ically plausible, the pattern ^f primary feather
lengths shown by TV. ios is closer to that of TV.
ochracea than to any form of TV. philippensis.
Phylogenetic analyses will be required to un-
derstand the relationships of TV. ios.

Conservation. — As only one specimen is
known, it appears likely that Ninox ios has a
limited range and/or is rare. However, noctur-
nal birds are frequently overlooked. Also,
most scientific bird collecting in Sulawesi
took place before mist-nets were widely avail-
able, and at lower elevations. Ascertaining its

464

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

vocalizations and calling periods will be a pre-
requisite to carrying out effective surveys,
which will be essential to establish the degree
of risk faced by this unique new species.

The only other bird species thought to be
restricted to North Sulawesi is the poorly
known Matinan Flycatcher (Muscicapa san-
fordi), which has been found only in the Du-
moga-Bone and Tentolo-Matinan mountains
between 1400 and 1780 m. The fact that a
species as distinctive as Ninox ios could have
escaped description until now clearly under-
scores the fact that our knowledge of the avi-
fauna of Sulawesi is still in a rudimentary
state.

ACKNOWLEDGMENTS

Special thanks are due R. W. R. J. Dekker, M. S.
Hoogmoed, and H. van Grouw, NNM/Naturalis, Lei-
den, and to E G. Rozendaal. Specimens of comparative
taxa were examined in museums too numerous to enu-
merate individually, but chiefly the following, for
which I thank G. E Barrowclough and R Sweet, Amer-
ican Museum of Natural History, New York; G. Hess,
Delaware Museum of Natural History, Wilmington; D.
Willard, Field Museum of Natural History, Chicago;
R. S. Kennedy, Museum of Natural History, Cincin-
nati; S. Prijono and Darjono, Museum Zoologicum Bo-
goriense, Cibinong, West Java; S. L. Olson, G. R.
Graves, and B. M. McPhelim, National Museum of
Natural History, Smithsonian Institution, Washington,
D.C.; M. P. Walters and R. P. Prys-Jones, The Natural
History Museum, Tring, U.K.; G. Boenigk, Staatliches
Naturhistorische Museum, Braunschweig, Germany;
and E C. Sibley, Yale Peabody Museum, New Haven.
The manuscript was improved by the comments of R.
W. R. J. Dekker, N. J. Collar, B. M. Beehler, and R.
Hill; J. Tobias provided information; and the frontis-
piece was painted by 1. Lewington.

LITERATURE CITED

Coates, B. J. and K. D. Bishop. 1997. A guide to the
birds of Wallacea. Alderley, Queensland, Austra-
lia.

Collar, N. J. and P. C. Rasmussen. 1998. Species
limits in the Ninox philippensis complex. Ostrich
69(3-4);398.

Dickinson, E. C., R. S. Kennedy, and K. C. Parkes.
1991. The birds of the Philippines. British Orni-
thologists’ Union, Tring, U.K.

Ekstrom, j., j. Tobias, and J. Robinson-Dean. 1998.
Forests at the edge of Lore Lindu National Park,
central Sulawesi. Oriental Bird Club Bull. 28:36—
39.

Johnstone, R. and J. Darnell. 1997. Description of
a new subspecies of Boobook Owl Ninox novae-
seelandiae (Gmelin) from Roti Island, Indonesia.
West. Aust. Nat. 21:161—173.

Meyer, A. B. and L. W. Wiglesworth. 1898. The
birds of Celebes and the neighbouring islands.
Vol. 1. R. Friedlander & Sohn, Berlin, Germany.

Munsell. 1977. Munsell color charts for plant tissues.
Gretagmacbeth, New Windsor, New York.

Norman, J. A., L. Christidis, M. Westerman, and E
A. R. Hill. 1998. Molecular data confirms the
specific status of the Christmas Island Hawk-Owl
Ninox natalis. Emu 98:197—208.

Rozendaal, E G. and R. W. R. J. Dekker. 1989. An-
notated checklist of the birds of the Dumoga-Bone
National Park, North Sulawesi. Kukila 4:85-109.

Stattersfield, a. j., M. J. Crosby, A. J. Long, and
D. C. Wege. 1998. Endemic bird areas of the
world. BirdLife International, Cambridge, U.K.

Stresemann, E. 1939. Die Vogel von Celebes. Part 1.
J. Ornithol. 87:299-425.

Stresemann, E. 1940. Die Vogel von Celebes. Part 2.
J. Ornithol. 88:389-487.

White, C. M. N. and M. D. Bruce. 1986. The birds
of Wallacea. British Ornithologists’ Union, Lon-
don, U.K.

Wilson Hull., 1 1 1(4), 1999. pp. 465-471

PATTERNS OF VARIATION IN SIZE AND COMPOSITION OF
GREATER SCAUP EGGS: ARE THEY RELATED?

PAUL L. FLINT' 3 AND J. BARRY GRAND' ^

ABSTRACT. — We studied egg size variation of Greater Scaup (Aythya mcirila) nesting on the Yukon-Kus-
kokwini Delta, Alaska from 1991 — 1996. Mean egg size was 64.36 ± 0.03 (SE) ml. Egg size did not vary with
clutch size or serve as an index of body size. There was less than 2% overlap in total clutch volumes for clutches
of different sizes indicating that phenotypic clutch size-egg size trade-offs are not occurring among individuals.
At the population level. Greater Scaup have less variation in egg size than other species of waterfowl. The
proportion of variation in egg size caused by differences among females was 0.20, caused by differences within
females among years was 0.25, and caused by differences within females and years (i.e., clutches) was 0.56.
The proportion of egg lipid decreased with increasing egg size while the proportion of egg protein increased
with egg size. Thus, Greater Scaup appear to trade-off lipid for protein as egg size increases. The proportion of
variation that was due to differences among females in total egg protein was 0.79 and in total egg lipid was
0.49. We conclude that in the absence of a fitness trade-off between clutch size and egg size, selection has
reduced among-individual variation in egg size. Received 16 April 1999, accepted 4 August 1999.

Lack (1967) suggested that trade-offs may
occur between the number and size of eggs
produced by waterfowl species. Rohwer
(1988) argued that the same trade-off between
number and size of eggs should occur within
species as well. The high repeatability of egg
size (i.e., volume or weight) generally found
in waterfowl suggests that such trade-offs
likely occur among rather than within individ-
uals (Lessells et al. 1989, Larsson and Fors-
lund 1992). Thus, the concept of a clutch
size-egg size trade-off implies that females
laying small clutches of large eggs and fe-
males laying large clutches of small eggs have
equal fitness. Accordingly, for a trade-off to
exist both clutch size and egg size must be
positively related to fitness.

Waterfowl laying larger clutches may be
more fit because they tend to fledge more
young (Lessells 1986, Rockwell et al. 1987,
Flint 1993). For example, Rockwell and co-
workers (1987) demonstrated that female
Lesser Snow Geese (Chen caerulescens cae-
rulescens) laying larger clutches recruited
more young than females laying smaller
clutches. Fitness may also be related to egg
size in .some species of waterfowl because

' Alaska Biological Science Center, U..S. Geolog-
ical Survey, 1011 East Tudor Rd., Anchorage, AK
99503.

- Present address: Alabama Cooperative Fisheries
and Wildlife Research Unit, 331 Funchess Hall, Au-
burn University, Auburn, AL 36849.

^ Corresponding author; E-mail; pauLflint(§>usgs.gov

young from larger eggs are better able to sur-
vive extreme conditions (Ankney 1980,
Rhymer 1988, Thomas and Brown 1988). For
example, Dawson and Clark (1996) found that
Lesser Scaup {Aythya affinis) ducklings from
large eggs survived better than those from
small eggs under natural conditions (but see
Williams et al. 1993). The mechanism by
which larger eggs yield higher juvenile sur-
vival may be related to egg composition
(Dawson and Clark 1996). Egg composition,
in terms of lipid and protein, typically varies
isometrically with egg weight for waterfowl
species (Ankney 1980, Hepp et al. 1987,
Owen and West 1988, Hill 1995, Slattery and
Alisauskas 1995); however, some species
show a proportional increase in lipid with in-
creasing egg size (Birkhead 1984, 1985; Ali-
sauskas 1986; Rohwer 1986; Williams 1994).
In either case, young hatching from large eggs
tend to be larger at hatching and have abso-
lutely larger reserves than young hatching
from smaller eggs (Ankney 1980, Slattery and
Alisauskas 1995, Erikstad et al. 1998).

Fitness trade-offs are potential mechanisms
maintaining heritable variation in both egg
size and composition at the population level
(Falconer 1989). Egg size has been shown to
be, at least partially, under genetic control,
and heritable genetic variation has been found
in several species of waterfowl (Batt and
Prince 1978, Lessells et al. 1989, Larsson and
Forslund 1992). Thus, in the absence of fitness
trade-offs, mean egg size for a population

465

466

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

would be expected to respond to directional
selection. Therefore, we predict that patterns
of egg size variation within and among pop-
ulations are linked to clutch size-egg size
trade-offs. Additionally, if clutch size-egg
size trade-offs exist, then the corresponding
variation in egg composition can be used to
infer the required relationship between egg
size and fitness. It follows then, that patterns
of variation in egg composition within popu-
lations are also influenced by clutch size-egg
size trade-offs.

Our goal in this study was to describe egg
size variation in Greater Scaup {Aythya mar-
ilci) with regard to female body size and clutch
size. Further, we looked for evidence of clutch
size-egg size trade-offs among females. We
examined variation in egg composition within
and among females and in relation to egg size
and compared these results to other species.

METHODS

This study was conducted along the lower Kashunuk
River drainage (61° 20' N, 165° 35' W) on the outer
coastal fringe of the Yukon Delta National Wildlife
Refuge, Alaska. This study was conducted under ap-
propriate Federal (PRT-692350), state (93-69), and ref-
uge special use collection permits. The study area con-
sists of relatively flat sedge meadows and numerous
ponds (Flint and Grand 1996a, Grand et al. 1997). Nest
searches were conducted from mid-May through mid-
July (see Flint and Grand 1996a). When nests were
discovered, the number of eggs, nest location, and date
were recorded. Eggs were individually numbered and
candled to determine the stage of embryonic devel-
opment (Flint and Grand 1996a). Nests were revisited
at 7 day intervals and the number of eggs and stage
of development of embryos were recorded. Maximum
lengths and breadths of all eggs were measured to the
nearest 0. 1 mm. Clutch size was defined as the number
of eggs known to have been laid into a nest for nests
found during egg laying and as the number of eggs in
the nest at time of discovery for nests found during
incubation.

In 1994-1996 a sample of hens was captured on the
nest at hatching using bow traps (Flint and Grand
1996b). Hens were weighed to the nearest 10 grams
and culmen and total tarsus lengths were measured to
the nearest 0.1 mm. Females were marked with alu-
minum U.S. Fi.sh and Wildlife Service tarsal bands.

In 1996, we collected a sample of 30 complete
clutches at 4 days of incubation in conjunction with a
separate study of renesting ecology. These eggs were
weighed, measured (length and breadth) with calipers,
and we measured the external volume by submerging
the egg and measuring the displacement ol water to
the nearest 0.5 ml. Variables in the relationship be-
tween volume and linear measures were estimated us-

ing analysis of covariance with volume as the depen-
dent variable, females as a factor, and length X
breadth^ as the single covariate (Hoyt 1979, Flint and
Sedinger 1992, Flint and Grand 1996b). We included
the interaction between females and length X breadth^
to compare the slope of this relationship among fe-
males. The fitted relationship was used to predict egg
volume from linear measures for eggs where volume
was not measured directly (i.e., eggs measured in the
field).

In 1993, a sample of nests was visited every other
day after 7 eggs had been laid to determine the date
of clutch completion. We collected a sample of 88 eggs
from 1 1 clutches as soon as laying was complete.
These eggs were boiled for 15 min and frozen. Each
egg was later thawed and weighed whole. Because en-
tire clutches were collected at the end of laying and
some incubation occurred during laying, vasculariza-
tion or embryonic development was noted as being
present during dissection. The egg was separated into
shell, albumen, and yolk and these components were
dried to a constant weight at approximately 80° C. Fat
was extracted from the yolk using petroleum ether
(Dobush et al. 1985) in a Soxhlet apparatus. Total yolk
protein was estimated as whole yolk dry weight minus
yolk fat. We assumed that dried albumen weight was
composed entirely of protein (Montevecchi et al. 1983)
and estimated total egg protein as total yolk protein
plus albumen dry weight.

We examined variation in egg size with clutch size
using a nested ANOVA with clutches nested within
clutch sizes. We used the mean square error among
clutches with clutch sizes as the denominator in the F-
test of clutch size effects. Total clutch volume was
calculated as the sum of the individual egg volumes
within a clutch for the sample of nests that survived
to incubation. We used the sum of the log transformed
measurements of culmen and tarsus as an index of
structural size and examined variation in mean egg size
in relation to this index for the sample of nests from
which we captured females using linear regression.
The proportion of variation in egg size caused by dif-
ferences within clutches, within females among years,
and among females (i.e., repeatability) were calculated
using a nested ANOVA and modifying the methods of
Lessells and Boag (1987) for a nested design (Sokal
and Rohlf 1981).

We examined variation in both total egg lipid and
protein (separately) using ANCOVA with female (i.e.,
clutch) and egg development as factors and egg size
as a covariate. We also included an interaction between
female and egg size. To examine allometric relation-
ships of egg components we used log|„-logio regression
of egg components against egg weight; isometry was
concluded if the slope of these relationships was not
different from 1. Repeatability of egg components was
e.stimated using a nested design with eggs nested with-
in females. All analyses were conducted using SAS
version 6.12 (SAS Institute 1990).

Flint iind Grand • GREATER SCAUP EGG SIZE AND COMPOSITION

467

(f)

250 300 350 400 450 500 550 600 650 700 750

Total dutch volume (ml)

FIG. 1. Total clutch volume for Greater Scaup clutches of different sizes from the Yukon-Kuskokwim Delta
1991-1996. Clutch volumes were rounded to the nearest 5 ml. Overall only 3 of 235 (1.3%) clutch volumes
(i.e., two 9 egg clutches and one 10 egg clutch) occurred within the volume distribution of another clutch size.
Other stacked bars represent tied values resulting from rounding of total clutch volumes where actual estimates
of total clutch volume did not overlap. For example, a 7 egg clutch with a volume of 467 ml is stacked with
an 8 egg clutch of 468 ml because rounding is required to develop the distributions.

RESULTS

We measured length, breadth, weight, and
volume of 271 eggs. The equation:

Volume = 31.84 + 0.2729 X length

X breadth^, ( 1 )

described the relationship between displace-
ment (ml) and linear egg measurements (cm)
(r2 = 0.44, F, 269 = 209.10, P < 0.001). To
compare variation in predicted egg volumes
with variation in measured volumes we cal-
culated the repeatability of measured egg vol-
umes (33.4%) and found it similar to the re-
peatability of estimated egg volumes (36.4%).

We measured a total of 3937 eggs in the
field. Mean egg length was 6.352 ± 0.003
(SE) cm and egg breadth was 4.328 ± 0.002
cm. Mean estimated egg size was 64.36 ±
0.03 ml (CV = 0.03). We had no data on var-
iation in egg size with laying sequence, but
the average range of egg size within clutches
was 3.3 ml.

Egg size did not vary with clutch size
(^12.470 “ 0.74, P > 0.05). For the analyses of
the relationship between egg size and body
size we used 93 observations of individual fe-
males captured on nests. Average egg size per
clutch was not related to our index of body
size (F| 9o = 0.35, P > 0.05). Overall only 3

of 235 (1.3%) clutch volumes (i.e., two 9-egg
clutches and one 10-egg clutch) occurred
within the volume distribution of another
clutch size (Fig. 1).

For estimates of repeatability of egg size we
captured 20 females 41 times (only 1 female
was captured 3 times). The proportion of var-
iation in egg size attributed to differences
among females (i.e., repeatability) was 0.20,
proportion of variation attributed to differenc-
es within females among years was 0.25, and
proportion of variation attributed to differenc-
es within females and years (i.e., clutches)
was 0.56. Using a standard approach of cal-
culating the mean egg size per clutch and
treating clutches as individual observations
within females (e.g., Flint and Grand 1996b)
yields a repeatability of mean egg size of 0.36.

Total egg protein varied with egg size m
= 67.96, P < 0.001), but not among females
(^9,64 ~ 1.55, P > 0.05). The relationship be-
tween egg protein and size also did not vary
among females (i.e, no interaction, ^9^,4 =
1.59, P > 0.05). Total egg protein was not
significantly influenced by early embryonic
development ^ = 3.23, P > 0.05). Simi-
larly, total egg lipid varied with egg size (F, (,4
= 65.80, P < 0.001), but not among females
(^9.64 ~ 1.98, P = 0.055). However, the rela-

468

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 1. Allometric relationships between total eg

;g lipid, protein and water

in relation to egg weight.

Allometric relationship

Slope

ECU'

UCL-’

r''

Weight and egg lipid'-'

0.62

0.85

0.71

0.99

2.09

0.04

Weight and egg protein‘s

0.84

1.11

1.01

1.21

2.16

0.03

Weight and egg water

0.95

1.01

0.96

1.06

0.52

>0.05

“ Upper and lower 95% confidence limits for slope.
'’Test of null hypothesis that slope = 1.0.

Egg lipid and protein are in dry weight.

tionship between total egg lipid and egg size
varied among females (i.e., interaction; =
2.04, P = 0.049). Total egg lipid was not in-
fluenced by early embryonic development
(F, ,4 = 0.17, P > 0.05).

The proportion of lipid in an egg declined
with increasing egg weight, whereas the pro-
portion of protein increased with increasing
egg weight (Table 1). Among-female differ-
ences explained most of the variation in total
egg protein; whereas within-female (i.e., with-
in clutch) differences explained most of the
variation in egg lipid components (Table 2).
Even with the variation in egg composition
within and among females, total clutch vol-
ume still predicted most of the variation in
total protein {F — 0.86, F, ^ = 48.71, P <
0.001) and lipid (r- = 0.93, F, g = 100.05, P
< 0.001) investment in a clutch.

DISCUSSION

Egg size. — The relationship between length
X width^ and measured volumes fit poorly (F
= 0.44) for Greater Scaup. For other species
of waterfowl these relationships have shown
a high F (>90%; Flint and Sedinger 1992,
Flint and Grand 1996b, Slattery and Alisaus-

TABLE 2. Average egg components dry weight
and proportion of variation due to differences among
females.

Among

female

variance

component

Variable

(%)

Albumen (g)

4.97

0.046

87.9

Shell (g)

5.23

0.032

73.2

Yolk (g)

14.84

0. 1 09

52.4

Total egg lipid (g)

9.97

0.069

48.8

Total egg protein (g)

9.89

0.077

79.1

Lipid per mi

0.15

0.001

36.2

Protein per ml

0. 1 5

0.001

79.9

Density (g/ml)

0.91

0.003

67.9

kas 1995). We believe this relatively poor fit
is the result of the low coefficient of variation
in egg size (i.e., lack of variation in the y-
axis). Additionally, we likely detected varia-
tion in slope among females because we sam-
pled entire clutches and thus had relatively
large samples within females. Other studies
have failed to find variation in slope of pre-
dictive relationships between egg measures
and egg size among females but only sampled
a few eggs from each clutch (Flint and Sedin-
ger 1992, Flint and Grand 1996b). We found
little difference between estimates of repeat-
ability of predicted volumes and measured
volumes. The comparability of these two re-
sults indicates that error associated with pre-
dicting egg volumes is random within and
among females.

We found less variation in egg size at the
population level than has been reported for
other species of waterfowl. The coefficient of
variation of average egg size for Greater
Scaup (3%) was substantially less than for
Northern Pintails {Anas acuta, 11%; Flint and
Grand 1996b), Blue- winged Teal {Anas dis-
cors, 9%; Rohwer 1986), Canada Geese
{Branta canadensis, 7—9%; Cooper 1978,
Leblanc 1989), Black Brant {Branta bernicla
nigricans, 8%; Flint and Sedinger 1992), or
Lesser Snow Geese (8%; Cooke et al. 1995).
Thus, it appears selection has reduced the var-
iation in egg size at the population level com-
pared to other species.

The repeatability of egg size that we mea-
sured among females is substantially lower
than reported for other waterfowl (Batt and
Prince 1978, Duncan 1987, Lessells et al.
1989, Flint and Sedinger 1992, Flint and
Grand 1996b). When we partitioned variance
using a nested ANOVA design, including var-
iation within clutches, we found the among
female component of the variance was even
lower. Using an approach similar to ours.

Flint and Grand • GREATER SCAUP EGG SIZE AND COMPOSITION

469

Leblanc (1989) reported that the proportions
of variation among female Canada Geese,
within females-among years, and within
clutches were 0.62, 0.05, and 0.33, respec-
tively. Thus, the patterns of egg size variation
within and among female Greater Scaup are
substantially different than those reported for
other waterfowl. The average range of egg
sizes within clutches was similar to the range
of egg sizes reported for other waterfowl
(Leblanc 1987, Owen and West 1988, Flint
and Sedinger 1992, Robertson and Cooke
1993, Cooke et al. 1995). Thus, the low re-
peatability we report is not the result of rela-
tively high variation within females (i.e.,
clutches), but is strongly influenced by the
lack of variation in egg size among females.
Relative to other waterfowl species, it appears
selection has resulted in reduced egg size var-
iation among female Greater Scaup and main-
tained variation within clutches that may be
related to sequence (Flint and Sedinger 1992,
Cooke et al. 1995).

Mean egg size did not vary with clutch size
in this study. This is consistent with findings
for other waterfowl (Duncan 1987, Rohwer
1988, Rohwer and Eisenhauer 1989, Flint and
Sedinger 1992, Flint and Grand 1996b).
Therefore, we did not detect a trade-off be-
tween clutch size and egg size predicted under
the nutrient limitation hypothesis and a simple
model of nutrient allocation to eggs (Flint et
al. 1996). Additionally, mean egg size was not
related to body size and the alternative allo-
cation model described by Flint and cowork-
ers (1996) cannot explain the failure to detect
a negative relationship between clutch size
and egg size. Finally, the lack of significant
overlap in total clutch volumes for clutches of
different size, indicated phenotypic trade-offs
between clutch size and egg size among fe-
males with equal investments in their clutches
did not occur (Flint and Sedinger 1992, Flint
and Grand 1996b). Thus, we find no evidence
of phenotypic clutch size-egg size trade-offs
among individuals for Greater Scaup.

Ankney and Bissett (1976) proposed that
egg size variation in a population was main-
tained by annual variation in environmental
conditions that caused annual variation in op-
timal egg size. While not explicitly stated, the
concept of a clutch size-egg size trade-off is
inherent in this hypothesis. In their example.

Ankney and Bissett (1976) state that the ad-
vantage to females laying small eggs is that
they can produce more eggs from finite re-
serves than females laying large eggs. We
found considerably less egg size variation at
the population level than has been reported for
other species. Correspondingly, we also found
no evidence of a clutch size— egg size trade-
off whereas studies of other species have
shown some evidence of a trade-off among
individuals with equal investments in their
clutches (Ankney and Bissett 1976, Flint and
Sedinger 1992, Flint and Grand 1996b). Fur-
ther, repeatability sets the upper limit to her-
itability (Falconer 1989); thus, the low re-
peatability we measured implies little genetic
variability for egg size in our study popula-
tion. Traits influenced by selection are ex-
pected to approach fixation and therefore have
low heritabilities (Falconer 1989). We suggest
for Greater Scaup that clutch size-egg size
trade-offs do not occur and selection has re-
duced variability in egg size.

Egg composition. — Because female water-
fowl commonly begin incubation before egg
laying is complete (Flint et al. 1994), and we
did not collect eggs until egg laying was ter-
minated, some eggs had slight embryonic de-
velopment at the time of collection. The pres-
ence of egg development was not related to
either total egg lipid or protein after control-
ling for variation among females, egg volume,
and an interaction between females and egg
volume. Thus, we do not believe that the early
development observed in some of our col-
lected eggs influenced our results.

The relationship between egg lipid and egg
size for Greater Scaup varied among females.
This is similar to what Rohwer (1986) found
for Blue-winged Teal. Birkhead (1985) re-
ported greater variation in egg composition
among females than within clutches for Mal-
lards {Anas platyrhynchos). Further, Alisaus-
kas (1986) found that egg lipid was highly
variable among female American Coots {Fu-
Uca americana) and varied within clutches in
relation to egg sequence. Hepp and coworkers
(1987) found that about half of the variation
in egg lipid in Wood Ducks {Aix sponsa) was
due to differences within females, similar to
our results. Varying egg composition may al-
low females to slightly adjust the total invest-
ment in the clutch without altering clutch size

470

THE WILSON BULLETIN • Vol. HI, No. 4, December 1999

or egg size (Owen and West 1988). The con-
sistent finding that egg lipid varied within fe-
males suggests that there may be some adap-
tive partitioning of resources within clutches,
perhaps related to egg laying sequence. How-
ever, even with the variation in egg compo-
sition described above, total clutch volume ex-
plained more than 85% of the variation in to-
tal lipid and protein investment in a clutch.
Thus, contrary to the results of Flint and
Grand (1996b) for Northern Pintails, clutch
size is a good predictor of nutrient investment
in Greater Scaup clutches.

Heavier eggs had proportionally more pro-
tein and less lipid than lighter eggs. Thus, it
appears that Greater Scaup trade off lipid for
protein as egg weight increases. These results
differ from studies of waterfowl that show a
proportional increase in lipid with egg size
(Birkhead 1984, 1985; Alisauskas 1986; Roh-
wer 1986; Williams 1994), and differ from
studies of other precocial species where egg
protein and lipid both increase proportionately
with egg weight (Ankney 1980, Hepp et al.
1987, Owen and West 1988, Hill 1995, Slat-
tery and Alisauskas 1995). Ankney and Bis-
sett (1976) argued that because egg yolk, and
hence egg lipid, increased with egg size,
young hatching from larger eggs survived bet-
ter because they had absolutely larger lipid re-
serves. However, young from larger eggs tend
to be larger at hatch (Ankney 1980, Slattery
and Alisauskas 1995, Erikstad et al. 1998) and
thus will also have absolutely higher energy
requirements (Rhymer 1988). For Greater
Scaup, the proportion of egg lipid decreased
with increasing egg size. Therefore, we would
not expect large egg size to confer a survival
advantage to offspring in our study popula-
tion.

Given our conclusion that clutch size-egg
size trade-offs do not occur, and egg size has
been optimized by selection for our popula-
tion, we would expect no relationship between
egg size and fitness for our study population.
The proportional decline in egg lipid with in-
creasing egg size fits this prediction. There-
fore, we conclude that clutch size-egg size
trade-offs likely influence both egg size vari-
ation and patterns of egg composition within
species. If this hypothesis is correct, we would
predict that species showing evidence ol phe-
notypic trade-offs between clutch size and egg

size (e.g.. Northern Pintails; Flint and Grand
1996b) would also have patterns of variation
in egg composition consistent with the expec-
tation that offspring from larger eggs have a
survival advantage.

ACKNOWLEDGMENTS

This project was funded by U.S. Fish and Wildlife
Service, Office of Migratory Bird Management, Re-
gion 7, and the Alaska Biological Science Center. We
thank D. Derksen, R. Oates, and R. Leedy for their
support, and J. Morgart and the staff of the Yukon
Delta National Wildlife Refuge for logistical support.
We acknowledge the many people who helped search
for nests and measure eggs over the years. We thank
S. Lee for conducting nutrient composition analysis of
eggs and R. Rockwell and C. Babcock for helping us
understand repeatability analysis. We thank T. Arnold,
D. Derksen, K. Lessells, M. MacCluskie, and J. Se-
dinger for reviewing earlier drafts of this manuscript.

LITERATURE CITED

Alisauskas, R. T 1986. Variation in the composition
of the eggs and chicks of American Coots. Condor
88:84-90.

Ankney, C. D. 1980. Egg weight, survival, and growth
of Lesser Snow Goose goslings. J. Wildl. Manage.
44:174-182.

Ankney, C. D. and A. R. Bissett. 1976. An expla-
nation of egg-weight variation in the Lesser Snow
Goose. J. Wildl. Manage. 40:729-734.

Batt, B. D. j. and H. H. Prince. 1978. Some repro-
ductive parameters of Mallards in relation to age,
captivity, and geographic origin. J. Wildl. Man-
age. 42:834-842.

Birkhead, M. 1984. Variation in the weight and com-
position of Mute Swan {Cygnus olor) eggs. Con-
dor 86:489-490.

Birkhead, M. 1985. Variation in egg quality and com-
position in the Mallard Anas platyrhynchos. Ibis
127:467-475.

Cooke, E, R. E Rockwell, and D. B. Lank. 1995.
The Snow Geese of La Perouse Bay: natural se-
lection in the wild. Oxford Univ. Press, New York.
Cooper, J. A. 1978. The history and breeding biology
of the Canada Goose of Marshy Point, Manitoba.
Wildl. Monogr. 61:1-87.

DAW.SON, R. D. AND R. G. Clark. 1996. Effects of
variation in egg size and hatching date on survival
of Lesser Scaup Aythya qffinis ducklings. Ibis 1 38:
693-699.

Dobush, G. R., C. D. Ankney, and D. G. Krementz.
1985. The effect of apparatus, extraction time, and
.solvent type on lipid extractions of Snow Geese.
Can. J. Zool. 63:1917-1920.

Duncan, D. C. 1987. Variation and heritability in egg
size of Northern Pintails. Can. J. Zool. 65:992-
996.

Erikstad, K. E., T. Tveraa, and J. O. Bustnes. 1998.

Flint and Grand • GREATER SCAUP EGG SIZE AND COMPOSITION

471

Significance ot intraclutch egg-size variation in
Common Eider: the role of egg size and quality
of ducklings. J. Avian Biol. 29:3-9.

Falconer, D. S. 1989. Introduction to quantitative ge-
netics. Third ed. John Wiley and Sons, Inc., New
York.

Flint, P. L. and J. S. Sedinger. 1992. Reproductive
implications of egg-size variation in the Black
Brant. Auk 109:896-903.

Flint, P. L. 1993. Pre fledging survival and reproduc-
tive survival in Black Brant. Ph.D. diss., Univ. of
Alaska, Fairbanks.

Flint, P. L., M. S. Lindberg, M. C. MacCluskie, and
J. S. Sedinger. 1994. The adaptive significance of
hatching synchrony of waterfowl eggs. Wildfowl
45:248-254.

Flint, P. L. and J. B. Grand. 1996a. Nesting success
of Northern Pintails on the Coastal Yukon-Kus-
kokwim Delta. Condor 98:54-60.

Flint, P. L. and J. B. Grand. 1996b. Variation in egg
size of the Northern Pintail. Condor 98:162-165.

Flint, P. L., J. B. Grand, and J. S. Sedinger. 1996.
Allocation of limited reserves to a clutch: a model
explaining the lack of a relationship between
clutch size and egg size. Auk 113:939-942.

Grand, J. B., P. L. Flint, and P. J. Heglund. 1997.
Habitat use by breeding Northern Pintails on the
Yukon-Kuskokwim Delta, Alaska. J. Wildl. Man-
age. 61:1 199-1207.

Hepp, G. R., D. j. Strangohr, L. A. Baker, and R.
A. Kennamer. 1987. Factors affecting variation in
the egg and duckling components of Wood Ducks.
Auk 104:435-443.

Hill, W. L. 1995. Intraspecific variation in egg com-
position. Wilson Bull. 107:382-387.

Hoyt, D. F. 1979. Practical methods of estimating vol-
ume and fresh weight of bird eggs. Auk 96:73-
77.

Lack, D. 1967. The significance of clutch-size in wa-
terfowl. Wildfowl 18:125-128.

Larsson, K. and P. Forslund. 1992. Genetic and so-
cial inheritance of body and egg size in the Bar-
nacle Goose (Branta leucopsis). Evolution 46:
235-244.

Leblanc, Y. 1987. Intraclutch variation in egg size of
Canada Geese. Can. J. Zool. 65:3044-3047.

Leblanc, Y. 1989. Variation in size of egg of captive
and wild Canada Geese. Ornis Scand. 20:93-98.

Lessells, C. M. 1986. Brood size in Canada Geese: a
manipulation experiment. J. Anim. Ecol. 55:669-
689.

Lessells, C. M. and P. T. Boag. 1987. Unrepeatable

repeatabilities: a common mistake. Auk 104:1 16-
121.

Lessells, C. M., F. Cooke, and R. F Rockwell. 1989.
Is there a trade-off between egg weight and clutch
size in Lesser Snow Geese (Anser c. caerides-
cens)l J. Evol. Biol. 2:457-472.

Montevecchi, W. a., 1. R. Kirkman, D. D. Roby, and
K. L. Brink. 1983. Size, organic composition, and
energy content of Leach’s Storm Petrel (Ocean-
odronia leuohoa) eggs with reference to position
in the precocial-altricial spectrum and breeding
ecology. Can. J. Zool. 61:1457-1463.

Owen, M. and J. West. 1988. Variation in egg com-
position in semi-captive Barnacle Gee.se. Ornis
Scand. 19:58-62.

Rhymer, J. M. 1988. The effect of egg size variability
on thermoregulation of Mallard (Anas platyrhyn-
chos) offspring and its implications for survival.
Oecologia 75:20-24.

Robertson, G. J. and F. Cooke. 1993. Intraclutch egg
size variation and hatching success in the Com-
mon Eider. Can. J. Zool. 71:544-549.

Rockwell, R. F, C. S. Findlay, and F Cooke. 1987.
Is there an optimal clutch size in Snow Geese?
Am. Nat. 130:839-863.

Rohwer, F. C. 1986. Composition of Blue-winged Teal
eggs in relation to egg size, clutch size, and the
timing of laying. Condor 88:513-519.

Rohwer, F. C. 1988. Inter- and intraspecific relation-
ships between egg size and clutch size in water-
fowl. Auk 105:161-176.

Rohwer, F. C. and D. I. Eisenhauer. 1989. Egg mass
and clutch size relationships in geese, eiders, and
swans. Ornis Scand. 20:43-48.

SAS Institute. 1990. SAS/STAT user’s guide. Version
6. SAS Institute, Inc. Cary, North Carolina.

Slattery, S. M. and R. T. Alisauskas. 1995. Egg
characteristics and body reserves of neonate Ross’
and Lesser Snow geese. Condor 97:970-984.

SoKAL, R. R. AND F J. Rohlf. 1981. Biometry, second
ed. W. H. Freeman and Co., New York.

Thomas, V. G. and H. C. P. Brown. 1988. Relation-
ships among egg size, energy reserves, growth
rate, and fasting resistance of Canada Goose gos-
lings from .southern Ontario. Can. J. Zool. 66:957-
964.

Williams, T. D. 1994. Intraspecific variation in egg
size and egg composition in birds: effects on off-
spring fitness. Biol. Rev. 68:35-59.

Williams, T. D., D. B. Lank, F Cooke, and R. F.
Rockwell. 1993. Fitness consequences of egg-
size variation in the Lesser Snow Goose. Oecol-
ogia 96:331-338.

Wilson Bull., 111(4), 1999, pp. 472-477

POSTFLEDGING BEHAVIOR OF GOLDEN EAGLES

LAURA T. O’TOOLE,' 2 PATRICIA L. KENNEDY,' ^ RICHARD L. KNIGHT,' AND

LOWELL C. MCEWEN'

ABSTRACT. — We predicted that extended parental care, asynchronous hatching, and incidences of siblicide
in Golden Eagles (Aquila chrysaeto.s) could increase the chances for conflict between siblings, and between
parents and offspring as juveniles aged. This conflict could motivate independence and dispersal in this species.
To test our predictions, during the 1993 and 1994 breeding seasons we examined post-fledging behavior in
Golden Eagles from the Little Missouri National Grassland and contiguous areas of western North Dakota. We
collected observations of 28 radio-tagged juveniles to determine whether predispersal movements were correlated
with age and with the presence of a sibling or parent during the first 6—10 weeks after fledging. We also recorded
juvenile vocalization rates to determine if they changed with age or the presence of a parent. We found that
distance from the natal nest increased with time since fledging. This was attributed to an increased proficiency
in flight and gradual development of independence from parental care. We found that calling rate and distance
between individuals of sibling pairs did not change with time after fledging but was highly variable. Calling
rates of fledglings in the presence of parents were higher than when parents were absent. Increased calling may
facilitate juvenile location or inform the parents of offspring nutritional status. Parents were not visible for most
observation periods and we did not observe any aggression by parents directed toward offspring. Siblings
engaged in “pl^y” activity and we did not detect any signs of aggression between siblings. Our data do not
support the predictions that an increase in parental and/or sibling aggression is associated with independence m
this population of Golden Eagles. Received 15 Oct. 1998, accepted 20 April 1999.

Factors influencing the timing and duration
of dispersal may be extrinsic (environmental),
endogenous, or some combination of the two
(Howard 1960, Ritchison et al. 1992, Belthoff
and Duffy 1995). Extrinsic factors include pa-
rental aggression toward young (Alonso et al.
1987, Hiraldo et al. 1989, Wiggett and Boag
1993), sibling aggression (Holleback 1974, De
Laet 1985, Strickland 1991), ectoparasitism
(Brown and Brown 1992), increased predation
risk (Harfenist and Ydenberg 1995), and de-
clining food availability within the natal area
(Messier 1985, Kenward et al. 1993, Busta-
mante 1994b). In raptors, the role of sibling
and parental aggression in family break-up is
disputed. In some species, dispersal may oc-
cur after a period of parent-offspring conflict,
the parents being the ones who promote the
independence of juveniles by gradually reduc-
ing the food supply or increasing aggression
towards the juveniles (Alonso et al. 1987, De-
lannoy and Cruz 1988, Hiraldo et al. 1989).
Family break-up may also take place without
apparent conflict (Bustamante and Hiraldo

' Dept, of Fishery and Wildlife Biology, Colorado
State Univ., Ft. Collins. CO 80523.

2 Current address: 1815 W. Grange Avenue, Mil-
waukee, WI 53221 .

’ Ct^rresponding author;

E-mail : patk@cnr.colostate.edu

1990, Bustamante 1994b). Although Watson
(1997) thought that there was little evidence
of aggression between adult Golden Eagles
and their young. Walker (1987) observed
some aggression by parents toward offspring
during the fledgling dependency period. There
is little information on the factors that affect
timing of dispersal in Golden Eagles because
few studies have been conducted on the be-
havior of juvenile eagles during this stage
(Watson 1997). We predicted that the extend-
ed parental care, asynchronous hatching (Wat-
son 1997), and high incidence of siblicide
(Edwards and Collopy 1983, Edwards et al.
1988) in this species would increase the prob-
ability of predispersal conflict.

METHODS

Study area. — We conducted this study from May to
November 1993-1994 in the Little Missouri National
Grassland (46° 00-48° 07' N, 102° 50'-104° 00' W)
and contiguous areas in western North Dakota. The
dominant habitat was mixed prairie with patches of
shrubs managed primarily for livestock grazing (Hop-
kins et al. 1986, Fowler et al. 1991). Woodlands (1.8%
of total vegetative cover) were found in areas of higher
soil moisture, such as valley bottoms, lower valley
slopes, and along stream banks and floodplains of the
Little Missouri River (Girard et al. 1989). Large buttes,
easily eroded sandstone, and clay badlands character-
ized the topography. The climate is semi-arid conti-
nental with wide daily fluctuations in temperature and

472

O'Toole ei al. • FLEDGLING GOLDEN EAGLE BEHAVIOR

473

variable precipitation. Annual precipitation averages
33-41 cm. and the average temperature in July is
21°C (Jensen 1972). We found Golden Eagle nests on
siltstonc and clay buttes, and in trees along rivers and
streams.

Data collection. — We located active Golden Eagle
nests by aerial and ground surveys. We aged chicks
based on behavior and plumage (Ellis 1979), and sexed
them based on bill depth and head, hallux claw, and
tail length (A. Harmata, pers. comm.). We entered each
of the 20 eagle nests (/; = 10 in 1993; /? = 10 in 1994)
when nestlings were 8-10 weeks of age (near fledg-
ing). We selected this age because harassment can
cause prolonged ab.sence of eagle adults which could
result in nestlings being exposed to direct sunlight or
missed feedings — both fatal to younger chicks. Be-
cause young eagles will fledge prematurely if dis-
turbed, we waited until nestlings were fully feathered
and could fly well enough to avoid injury if they
fledged in response to our presence (Fyfe and Olen-
dorft 1976). Twenty-eight (12 male, 16 female) chicks
(/! = 12 in 1993; n = 16 in 1994) were weighed, mea-
sured, fitted with a 25-g backpack-style radio trans-
mitter (L. L. Electronics, Mahomet, Illinois), and band-
ed with an aluminum U.S. Fish and Wildlife Service
leg band. After we instrumented and measured each
eagle, we placed it back in the nest where it remained
until it fledged. We spent an average of 51.3 (± 3.9
SE) min from eagle capture to replacement in the nest.

We considered a juvenile as fledged after its first
flight from the nest. After fledging, we located birds
using ground searches every 2-4 days for behavioral
observations. We recorded over 416 hr of direct be-
havioral observations from 14 July to 20 August 1993,
and 15 July to 22 August 1994. We recorded move-
ments of radio-telemetered birds until they dispersed
from their natal territories, or we found them dead.
During the first five weeks, we were usually able to
locate eagles visually (94.1% of attempts). We pri-
marily u.sed triangulation (Kenward 1987) by four ob-
servers to find juveniles in the later stages of the post-
fledging period.

We observed eagles during daylight hours (06:00-
21:00 MST) using I4-45X zoom spotting scopes from
vehicles or at locations at lea.st 300 m from the bird.
We determined target observation times by randomly
selecting a 4-hr block of daylight. We covered the oth-
er time periods during later observations using the
same method. During each observation period, wc
continually recorded data until the predetermined time
elapsed, or until we could no longer observe the bird.
We collected a mean of 15.4 (± 1.2) hr of observation
per bird over 4.3 (± 0.3) observation periods. We re-
corded eagle locations and activities on a data check-
sheet. We did not record sibling behaviors simulta-
neously. We recorded the following data on adults: lo-
cation in relation to focal bird, incidences of food de-
liveries by an adult, stooping and physical aggression
directed toward the focal bird. We observed few prey
deliveries (/? = 9), so we could not analyze temporal
changes in food provisioning rates.

We attempted to sample behavior equally among
birds tor diltcrent ages and times of day. We calculated
distance from local bird to the nest and the distance
between siblings using simple geometry and UTM co-
ordinates.

Statistical analy.si.s. — Using Repeated Measures
Analysis of Variance (ANOVA; Proc GLM; SAS Inst.
Inc. 1987; PC ver. 6.10) we tested our predictions on
the pooled 1993—1994 data because there were too few
data to examine a year effect using repeated measures
analysis. To determine the proper sampling unit for
analysis (individual bird vs nest) we used the ratio of
Type III Sum of Squares (Proc GLM). A relatively
large ratio (>2) would indicate that individual birds
should be combined into one “nest” unit because sep-
arate consideration results in unexplained error (Sum
of Squared Errors; P. Chapman, pers. comm.). There
was no evidence of a strong sibling effect (average
ratio = 1.3), and because all behavioral observations
were made after fledging, we used individuals as the
unit of analysis.

We compared hourly calling rate between three age
groups: 0-14, 15-28, and 29-42 days post-fledging;
and a fledgling’s distance from the natal nest over six
age periods (0-15, 16-28, 29-52, 53-74, 75-97, and
98-137 days post-fledging). We tested whether the dis-
tance between siblings increased over three post-fledg-
ing age periods (0-37, 38-74, and 75-121 days). For
analysis, we grouped fledgling ages differently because
the data collected for calling, distance from the nest,
and distance between siblings varied in quantity. We
were able to collect data on distance from nest longer
than any other measure. Also, calling rates were re-
corded hourly, unlike both distance measurements. The
selected age groups fulfilled minimum sample size re-
quirements for the most suitable statistical tests and to
achieve a stratified distribution of data points (P. Chap-
man, pers. comm.). All data sets used for these anal-
yses were log-transformed to equalize the variance of
errors (SAS Inst. Inc. 1987; PC ver. 6.10) but the raw
data are reported in this paper. The a level for all sta-
tistical tests was 0.05.

To analyze data on parent-offspring interactions, we
used ANOVA (Proc GLM; SAS Inst. Inc. 1987; PC
ver. 6.10) on log-transformed data to test whether call-
ing rates or fledgling distance from the nest changed
when parents were present. We also examined whether
fledgling distance from the nest depended upon an in-
teraction between parents' presence and fledging age.

RESULTS

Hourly calling rate. — The number of calls
per hour did not change with time after fledg-
ing (Wilks’ Lambda ^24,2 = 0.06, P > 0.05).
The mean values for each of the three groups
were similar, with an overall mean of 1 1.4 (±
0.01) calls hr ' (Fig. 1). Calling rate did not
differ between the sexes (F, 4,3 = 0.50, P >
0.05).

474

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

JC.

20 -

15 -

0-14

15-28 29-42

Days postlledging

FIG. 1. Mean number of calls ISEI per hour for
Golden Eagles in three post-hedging time periods (0-
14, 15-28, 29-42 d) in the Little Missouri National
Grassland of western North Dakota, 1993 and 1994.

Distance from nest. — The distance of fledg-
lings from their nests increased significantly
with time (Wilks’ Lambda F4 ,o = 11.53, P =
0.001; Fig. 2) and eventually resulted in loss
of radio contact within the study area (27,500
km^) as the birds dispersed. There was no dif-
ference between sexes in the distance individ-
uals dispersed (F, ,3 = 0.06, P > 0.05).

Sibling interactions and juvenile behav-
ior.— We collected data from six sibling pairs
to determine the mean distance between sib-
lings with time since fledging. There was no
significant (Wilks’ Lambda F2.3 = 0.34, P >
0.05) change over the time period of our study
(Fig. 3).

Fledgling pairs frequently exhibited non-
aggressive social behavior. Siblings often fol-
lowed one another or flew together. We ob-
served mutual preening or “nibbling (Ellis
1979) when the fledglings perched together (n
= 288), as they frequently did. We observed
three different sibling pairs display “play
catching and plucking of prey together. We
recorded no overtly aggressive social behavior
between sibling pairs, and this social activity
did not appear to change over time.

The mean age at first flight was 10.1 (±
0.08) weeks. We observed 1 1 attempts of prey
capture by juveniles. Two of the observed at-

FIG. 2. Mean distance from the nest (m) ISEI for
Golden Eagles in six post-fledging time periods (0-15,
16-28, 29-52, 53-74, 75-97, 98-137 d) in the Little
Missouri National Grassland of western North Dakota,
1993 and 1994.

500

450

400

350

300

s 250

e 200

150

100

0-37 38-74

75-121

Days postfledging

FIG. 3. Mean distance ISEI between Golden Eagle
siblings (6 pairs) in three post-fledging time periods
(0-37, 38-74, 75-121 d) in the Little Missouri Na-
tional Grassland of western North Dakota, 1993 and
1994.

O'Toole el al. • FLEDGLING GOLDEN EAGLE BEHAVIOR

475

tempts were successful, with prey items in-
cluding an unidentified snake and a rodent.

Parental interactions. — We observed par-
ents near the focal bird relatively infrequently
(29.6%) during observation periods. Calling
rates of the young were higher in the presence
of parents than when parents were not visible
(^1.41.1 ~ 14.39, P — 0.001). The rate of calling
nearly doubled for juveniles with a parent
present (x ± SE = 21.8 ± 3.5 calls hr ‘) ver-
sus parents absent (11.5 ± 1.7 calls hr”‘).
Fledglings were closer to the nest when par-
ents were present (F, ,3 = 10.81, P = 0.001),
but distance from the nest was not signifi-
cantly correlated with the interaction of fledg-
ling age and parental presence (F4 10 = 0.96,
P > 0.05). We did not observe any aggressive
behavior by the parent toward the offspring.

DISCUSSION

Calling rate did not change as the birds
aged after fledging, although we observed a
high degree of individual variation. A similar
finding was reported in a Japanese population
of Black Kites (Milvus migrans; Koga and
Shiraishi 1994). Increased calling with age
was reported in Spanish populations of Black
Kites (Bustamante and Hiraldo 1990), the
Egyptian Vulture {Neophron percnopterus;
Ceballos and Donazar 1990), and the Spanish
Imperial Eagle (Aquila adalberti; Alonso et al.
1987). The increased calling rates in these
studies were attributed to a decrease in pro-
visioning by parents as the young aged. Our
results do not corroborate findings of in-
creased calling with age. It is possible that
food provisioning did not decrease over time
in our study, resulting in no change in calling
rates. It is also possible that food provisioning
decreased, but calling did not increase because
the young were becoming independent and
beginning to hunt on their own. We cannot
evaluate these hypotheses because we did not
have enough data to analyze food provision-
ing rates.

Throughout our study, the appearance of a
parent resulted in increased juvenile calling
compared with when parents were absent. In-
creased calling in the presence of a parent is
commonly observed in raptors (Alonso et al.
1987, Ikeda 1987, Hiraldo et al. 1989, Bus-
tamante 1994a). Calling by juveniles may fa-
cilitate juvenile location by parents (Ikeda

1987). Calling also informs the parents of the
nutritional status of their offspring (Trivers
1974).

Distance from the nest increased as the ju-
veniles aged, probably in part due to increased
flying proficiency. As independence nears,
movements may not represent a linear dis-
persal, but may resemble “wanderings” out-
side the natal area. Similar observations have
been reported for other species (Boeker and
Ray 1971, Beecham and Kochert 1975, Steen-
hof et al. 1984, Walker 1987, Bahat 1992).

Siblings tended to move together after leav-
ing the nest, and the distance between them
remained relatively constant. This contradicts
Newton’s (1979) idea that fledgling raptors
tend to perch apart. Other studies also re-
vealed increased distance between siblings
with age (Hiraldo et al. 1989, Bustamante and
Hiraldo 1990, Ceballos and Donazar 1990).
Increased sibling distance has been attributed
to “tension” between juveniles or increased
flight proficiency. Bustamante (1994a) report-
ed that Kestrels {Falco tinnunculus) perched
close together and engaged in social behavior
similar to the young Golden Eagles we ob-
served. Bustamante (1993) also found siblings
of Black-shouldered Kites often perched to-
gether. Kenward and coworkers (1993) saw
the same tendency with Northern Goshawk
{Accipiter gentilis) siblings.

If parents and offspring were in conflict, re-
sulting in the adults “driving-off” the juve-
niles, or if parents used aggression to evaluate
the young’s flight proficiency (Ferrer 1992),
we should have observed agonistic interac-
tions during the times when the parents were
seen with the offspring, but we did not. Al-
though parental aggression has been observed
in raptor species (Robertson 1985, Alonso et
al. 1987, Walker 1987, Hiraldo et al. 1989), it
is often the case that the parents feed their
offspring long after they have fledged (Ikeda
1987, Walker 1987, Bustamante and Hiraldo
1990, Ceballos and Donazar 1990, Bahat
1992).

We did not find an increase in aggression
between siblings, as we predicted. Although
juveniles moved away from their parental
home range, they apparently remained togeth-
er, indicative of sibling attraction rather than
aversion. Our data do not support the predic-
tions that an increase in parental and/or sibling

476

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

agression is associated with independence in
this population.

ACKNOWLEDGMENTS

We would like to acknowledge the following for
field assistance; M. Melendez, L. Masten, P. O’Brien,
R. Brunotte, C. Stubblefield, and B. O'Toole. Many
thanks go to C. Althouse and B. Peterson for technical
assistance; T. Tiara and M. Hooper for blood chemical
analysis; B. Tokaruk for recovery of one eagle in Sas-
katchewan; and D. Getzy for necropsy. We thank A.
Harmata, G. Fob, R. Kreil, K. Sanchez, L. Brand, K.
Winks, D. Hirsch, W. Cushing, and J. Jeske for other
help. We thank J. Smith and two anonymous reviewers
for their helpful reviews of the manuscript. Support for
this research project was provided by the United States
Department of Agriculture, Animal and Plant Health
Inspection Service, PPQ Grasshopper IPM Project; the
U.S. Fish and Wildlife Service, U.S. Forest Service
LMNG; and the North Dakota Game and Fish De-
partment. The Animal Care and Use Committee of
Colorado State University approved the methodologies
used in this project.

LITERATURE CITED

Alonso, J. C., L. M. Gonzalez, B. Heredia, and J.
L. Gonzalez. 1987. Parental care and the transi-
tion to independence of Spanish Imperial Eagle
Aqiiila heliaca in Donana National Park, South-
west Spain. Ibis 129:212-224.

Bahat, O. 1992. Post-fledging movements of Golden
Eagles (Aquila chry.saetos homeryeri) in the Ne-
gev Desert, Isreal, as determined by radio-telem-
etry. Pp. 612-621 in Wildlife telemetry: remote
monitoring and tracking of animals (LG. Preide
and S.M. Swift, Eds.). Ellis Horwood Ltd, West
Sussex, U.K.

Beecham, j. j. and M. N. Kochert. 1975. Breeding
biology of the Golden Eagle in .southwestern Ida-
ho. Wilson Bull. 87:506—513.

Belthoff, j. R. and a. M. Duffy, Jr. 1995. Loco-
motor activity levels and the dispersal of Western
Screech-Owl, Otus kennicottii. Anim. Behav. 50:
558-561.

Boeker, E. L. and T. D. Ray. 1971. Golden Eagle
population studies in the Southwest. Condor 73:
463-467.

Brown, C. R. and M. B. Brown. 1992. Ectoparasitism
as a cause of natal dispersal in Cliff Swallows.
Ecology 73:1718—1723.

Bu,stamante, j. 1993. The post-fledging dependence
period of the Black-Shouldered Kite {Elanns cae-
ruleiis). J. Raptor Res. 27:185-190.

BU.STAMANTE, J. 1994a. Behavior of colonial Common
Kestrels (Falco tinminculus) during the post-
fledging dependence period in southwestern
Spain. J. Raptor Res. 28:79-83.

Bustamante, J. 1994b. Family break-up in Black and
Red kites Milvu.s mif>ran.s and M. milvii.s: is time

of independence an offspring decision? Ibis 136:
176-184.

Bustamante, J. and F. Hiraldo. 1990. Factors influ-
encing family rupture and parent-offspring con-
flict in the Black Kite (Milvu.s mii’rans). Ibis 132;
58-67.

Ceballos, O. and j. a. Donazar. 1990. Parent-off-
spring conflict during the post-fledging period in
the Egyptian Vulture Neophron perenopterus
(Aves, Accipitridae). Ethology 85:225-235.

DeLaet, J.V. 1985. Dominance and aggression in ju-
venile Great Tits, Panes' major major, in relation
to dispersal. Pp. 375-380 in Behavioral ecology:
ecological consequences of adaptive behavior
(Sibley, R. M. and R. H. Smith, Eds.). Blackwell
Scientific Publ., London, U.K.

Delannoy, C. a. and a. Cruz. 1988. Breeding biol-
ogy of the Puerto Rican Sharp-shinned Hawk (Ac-
cipiter striatus Venator). Auk 105:649—662.

Edwards, T. C., Jr. and M. W. Collopy. 1983. Ob-
ligate and facultative brood reduction in eagles;
an examination of factors that influence fratricide.
Auk 100:630-635.

Edwards, T. C., Jr., M. W. Collopy, K. Steenhof,
AND M. N. Kochert. 1988. Sex ratios of fledgling
Golden Eagles. Auk l(J5;793-796.

Ellis, D. H. 1979. Development of behavior in the
Golden Eagle. Wildl. Monogr. 43:1-94.

Ferrer, M. 1992. Regulation of the period of post-
fledging dependence in the Spanish Imperial Ea-
gle/\r/n//u Ibis 134:128—133.

Fowler, A. C., R. L. Knight, T. L. George, and L.
C. McEwen. 1991. Effects of avian predation on
grasshopper populations in North Dakota grass-
lands. Ecology 72:1775—1781.

Fyfe, R. W. and R. R. Olendorff. 1976. Minimizing
the dangers of nesting studies to raptors and other
sensitive species. Can. Wildl. Serv. Occas. Pap.
23:1-17.

Girard, M. M., H. Goetz, and A. J. Bjugstad. 1989.
Native woodland habitat types of southwestern
North Dakota. USDA For. Serv. Res. Pap. RM-
281:1-36.

Harfenist, a. and R. C. Ydenberg. 1995. Parental
provisioning and predation risk in Rhinoceros
Auklets (Cerorhinca monocerala): effects on nest-
ling growth and fledging. Behav. Ecol. 6:82-86.

Hiraldo, E, M. Delibes, and R. R. E.strella. 1989.
Observations of a Zone-tailed Hawk family during
the post-fledging period. J. Raptor Res. 23:103—

1 06.

Holleback, M. 1974. Behavioral interactions and the
dispersal of the family in Black-capped Chicka-
dees. Wilson Bull. 76:28-36.

Hopkins. R. B., J. F. Cassel, and A. J. Bjugstad.
1986. Relationships between breeding birds and
vegetation in four woodland types of the Little
Missouri National Grasslands. USDA For. Serv.
Res. Pap. MR-27():I-I2.

Howard, W. E. 1960. Innate and environmental dis-

O'Toole et al. • FLEDGLING GOLDEN EAGLE BEHAVIOR

477

persal of individual vertebrates. Am. Midi. Nat.
63:152-161.

Ikeda, Y. 1987. Ecological research on the Japanese
Golden Eagle Aquila chrysaetos japonica during
the post-fledging period in the Hakusan Range. J.
Raptor Res. 21:79-80.

Jensen, R. R. 1972. Climate of North Dakota. North
Dakota State Univ., Eargo.

Kenward, R. E. 1987. Wildlife radio tagging: equip-
ment, field techniques and data analysis. Academ-
ic Press, London, U.K.

Kenward, R. E., V. Marcstrom, and M. Karlblom.
1993. Post-nestling behavior of Goshawks, Accip-
iter genfilis. 1. The causes of dispersal. Anim. Be-
hav. 46:65-70.

Koga, K. and S. Shiraishi. 1994. Parent-offspring re-
lations during the post-fledging dependency peri-
od in the Black Kite {Milvus migrans) in Japan.
J. Raptor Res. 28:171-177.

Meissier, E 1985. Solitary living and extraterritorial
movements of wolves in relation to social status
and prey abundance. Can. J. Zool. 63:239-245.

Newton, I. 1979. Population ecology of raptors. Buteo
Books, Vermillion, South Dakota.

Ritchison, a. j., j. R. Belthoff, and E. J. Sparks.

1992. Di.spersal restlessness: evidence for innate
dispersal by juvenile Eastern Screech Owls?
Anim. Behav. 43:57-65.

Robertson, A. S. 1985. Observations on the post-
fledging dependence period of Cape Vultures. Os-
trich 56:58-66.

SAS Institute, Inc. 1987. SAS stati.stical guide for
personal computers, version 6.10. SAS Inst. Inc.,
Cary, North Carolina.

Steenhof, K., M. N. Kochert, and M. Q. Moritsch.
1984. Dispersal and migration of southwestern
Idaho raptors. J. Field Ornithol. 55:357-368.

Strickland, D. 1991. Juvenile dispersal in Gray Jays:
dominant brood member expels siblings from na-
tal teiTitory. Can. J. Zool. 69:2935-2945.

Trivers, R. L. 1974. Parent-offspring conflict. Am.
Zool. 14:249-264.

Walker, D. G. 1987. Observations on the post-fledg-
ing period of the Golden Eagle Aquila chrysaetos
in England. Ibis 129:92-96.

Watson, J. 1997. The Golden Eagle. T & A D Poyser,
London, U.K.

WiGGETT, D. R. and D. a. Boag. 1993. The proximate
cause of male-biased natal emigration in Colum-
bian ground squirrels. Can. J. Zool. 71:201-218.

Wilson Bull., 111(4), 1999, pp. 478-487

GROWTH PATTERNS OF HAWAIIAN STILT CHICKS

J. MICHAEL REED,' ELIZABETH M. GRAY,^^ DIANNE LEWIS,^
LEWIS W. ORING,3 RICHARD COLEMAN, ^ TIMOTHY BURR,^ AND

PETER LUSCOMB^

ABSTRACT. — We studied chick growth and plumage patterns in the endangered Hawaiian Stilt (Himantopus
mexictmus knudseni). Body mass of captive chicks closely fit a Gompertz growth curve, revealing a growth
coefficient {K) of 0.065 day ' and point of inflection (T) of 17 days. When chicks fledged about 28 days after
hatching, they weighed only 60% of adult body mass; at 42 d, birds still were only 75% of adult mass; culmen,
tarsus, and wing chord at fledging also were less than adult size. This trend of continued growth to adult size
after fledging is typical for most shorebirds. After hatching, captive chicks grew more rapidly than wild chicks,
probably because of an unlimited food supply. We found no evidence for adverse effects of weather on the
growth of wild chicks. As with other shorebirds, the tarsus started relatively long, with culmen and then wing
chord growing more rapidly in later development. Tarsal and wing chord growth were sigmoidal, whereas culmen
growth was linear. We describe plumage characteristics of weekly age classes of chicks to help researchers age
birds in the wild. Received 28 Dec. 1998, accepted 20 April 1999.

Avian growth patterns have been studied
primarily because of their relationships to the
ecology and evolutionary history of different
species (Ricklefs 1968, 1973, 1983; O’Connor
1984; Anthony et al. 1991), and to maximize
food yields of domestic animals (e.g., An-
thony et al. 1991). Although there is selection
for rapid independence of chicks, which
should reduce variance in growth rates, intra-
specific growth patterns can be variable and
flexible because of environmental variability
and competing selective pressures (Cooch et
al. 1991, Emlen et al. 1991). In studies of wild
birds, altricial species have been studied more
often than precocial species, at least in part
because the former remain in the nest from
hatching until fledging.

In this paper we present information on
chick growth patterns of the Hawaiian Stilt

' Biological Resources Research Center, Univ. of
Nevada, Reno, NV 89557.

2 Current address; Dept, of Biology, Tufts Univ.,
Medford, MA 02155; E-mail; mreed@tufts.edu

’Environmental and Resource Sciences/ 1 86, 1000
Valley Rd., Univ. of Nevada, Reno, NV 89512-0013.

Current address; USG.S-BRD-PIERC, RO. Box 44,
Hawaii National Park, HI 96718.

’ U.S. Fi.sh and Wildlife Service, 91 1 NE 1 1th Ave.,
Eastside Federal Complex, Portland. OR 97232.

^ Southwest Division, Naval Facilities Engineering
Command, 1220 Pacific Hwy., San Diego, CA 92132-
5190.

' 151 Kapahulu Ave., Honolulu Zoo, Honolulu, HI
96815.

** Corresponding author.

{Himantopus mexicanus knudseni), a precocial
bird that is an endangered subspecies of the
Black-necked Stilt. Like all shorebirds, stilts
are precocial and nidifugeous. Hawaiian Stilts
are significantly larger than the nominate race
(Coleman 1981) and differ somewhat in adult
plumage characteristics (Wilson and Evans
1893, Coleman 1981). Stilts are found on all
five major islands in Hawaii, breed exclusive-
ly in shallow, lowland wetlands (USFWS
1985), and statewide population counts indi-
cate a steady increase in population size (Reed
and Oring 1993). Our specific objectives were
to ( 1 ) describe patterns of Hawaiian Stilt chick
growth from captive and wild birds and com-
pare them to other shorebirds, and (2) provide
a method for aging chicks in the field. The
last objective was designed for studying pre-
adult mortality patterns by providing aging
criteria that do not requiring capturing the
bird.

METHODS

Captive birds. — Growth data for captive birds came
from 15 individuals raised from eggs in 1980 in the
Honolulu Zoo. Because chicks were kept in a common
enclosure, some competition for food might have oc-
curred, although food was provided ad libitum. Be-
cause all birds were subject to the same feeding and
environmental conditions, inter-individual variability
in growth should be minimized. All birds were
weighed daily for 42 days to the nearest 0.1 g. Ha-
waiian Stilts fledge approximately 28 d after hatching
(Coleman 1981).

One of the 15 birds was used only for the first 13 d
because a bill deformity developed at this time, caus-

478

Reed et al. • HAWAIIAN STILT CHICK GROWTH

479

ing the individual to lose mass quickly. A sixteenth
bird was not included in the analysis because of ab-
eiTant fluctuations in growth. Its mass at hatch was
over 5 standard deviations above the mean, and it
gained mass rapidly for 1 1 days. Between days 12-17,
however, it lost 25% of its body mass, dropping well
below the mean (ca 2 standard deviations); on day 18
it began to grow rapidly again, reaching mean mass
for the group 24 d later.

Other variables (culmen, tarsus, and wing chord)
were measured less regularly. Measurements were
made every 2-4 d after hatching and became less fre-
quent (every 4—10 d) after fledging. Some individuals
were measured more often than others. Despite this
variation, we were able to derive useful growth pat-
terns for these body measurements. Mass was mea-
sured by one person and lengths by another.

A growth curve for body mass was fit to a Gompertz
equation (r^ = 0.99; SPSS, Inc. 1995, NONLIN pro-
cedure) because it is used most often for shorebirds
(e.g., Beintema and Visser 1989a) and we wanted to
allow interspecific comparisons to be made (O’Connor
1984). The fit was made on average values for each
day from 12—15 individuals. The Gompertz equation
has the form

W = A X

where W is body mass (g), A is asymptotic (adult) mass
(g), K is the growth coefficient (day ‘), t is age (d),
and e is the base for natural logarithms. Adult mass
came from 43 adult males and 42 adult females (Cole-
man 1981). Although adult females weigh slightly
more than males (mean difference = 7.0 g), the dif-
ference is a small percentage (<4%) of total body
mass, consequently A was averaged across sexes
(202.5 g).

Wild birds. — Wild chicks were captured by hand on
the islands of Oahu, Maui, and Kauai in 1978-1980
and 1993. During 1978-1980, we captured chicks with
known hatching dates 142 times. Because chicks from
the same clutch were not considered to be independent,
they were averaged within each clutch (maximum of
four chicks averaged per clutch). This resulted in 33
measurements of chicks less than 24 h old (designated
day 0; « = 64 chicks). Chicks were remeasured every
time they were encountered and captured. This resulted
in 43 measurements of birds from 2-32 d old in = 78
chicks). We measured mass to the nearest 1 .0 g, cul-
men and tarsus lengths to the nearest 0.1 mm, and
wing chord to the nearest 1.0 mm. In 1993, we took
measurements on 55 birds ranging in age from hatch-
ing to fledging using the above methods. During 1993
we rarely knew the exact age of each chick, so these
measurements were used only to determine the rela-
tionships among body measurements. Tarsus and wing
chord measurements were made on the right side of
the chick and the same person made all measurements
in 1993. We also noted the presence or absence of an
egg tooth. Field measurements from 1978-1980 were
made by one person, and in 1993 by another, so values
were not compared.

Fluinage. — We considered only tho.se plumage char-
acteristics that were visible in the field: fuzzy appear-
ance associated with down, brown versus black cast,
presence of an eye ring, etc. We used the above char-
acteristics to describe plumage of weekly age classes.
In several cases, plumage descriptions for weekly age
classes were incomplete (e.g., lacking description of
wing coloration for week 3). Because plumage is es-
sentially the same for chicks of both Hawaiian and
Black-necked stilts (Coleman 1981), we supplemented
our descriptions of Hawaiian Stilts with plumage ob-
servations of wild, known-aged Black-necked Stilt
chicks at Honey Lake, California in 1997. Plumage of
adult Hawaiian Stilts is different from fledglings (Rob-
inson et al., in press).

Analyses. — Statistical analyses were conducted us-
ing version 7.0 of SPSS (SPSS, Inc. 1995). One as-
sumption in comparing body measurements between
captive and wild birds is that initial body sizes are
equal. To test this, we used multiple analysis of vari-
ance (MANOVA) to compare mass and culmen length,
tarsus and wing chord measurements between known-
aged captive and wild hatch day (day 0) birds. For ages
after day 0, we determined whether or not mean values
for wild birds fell within 95% confidence intervals for
mean values of captive birds. All statistical tests were
two-tailed. Values presented are means ± SD.

RESULTS

Growth in captivity. — Growth parameters
for the Gompertz equation indicated a growth
coefficient {K) of 0.065 and time to inflection
point (T) of 17 days. Although chick mass
varied little among the 1 1 individuals on day
of hatch (15.7 ± 0.6 g), variability in mass
among individuals increased greatly over the
first two weeks (60.4 ± 9.2 g), and remained
high up to fledging at day 28 (122.5 ± 10.6
g). In general, differences among chick mass
at day 14 are consistent until fledging, indi-
cating that chicks that gain relatively more
mass in the first two weeks after hatching tend
to fledge at a heavier mass than chicks that
gain less mass their first two weeks. Captive
individuals did not experience a significant
mass loss between day 0 (hatch day) and day
1 (paired t-test: t = —0.432, df = 10, P >
0.05).

At fledging, chicks had not attained adult
body mass or body measurements. Mass at
fledging was 60% of adult mass, culmen
length was 67% of adult length, tarsus length
was at 66%, and wing chord length was at
55% (adult measurements from Coleman
1981).

Growth in the wild. — There was no differ-

480

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

LIG. 1. Captive and wild Hawaiian Stilt chick mass and percentage of adult body mass (202.5 g) as a
function of age. Values are means ± SE.

ence between mean mass of captive (x = 15.7
± 0.6 g) and wild (x = 15.6 ± 1.1 g) chicks
at hatch {t = 0.551, df = 40, P > 0.05). The
apparent decrease in mass between day 0
(hatch day) and day 1 for wild chicks was not
significant (paired f-test: t = 0.585, df = 12,
P > 0.05). From days 1 to 17, masses of
same-aged wild birds typically fell within the
95% confidence interval of captive birds,
though below the mean. In three comparisons
(day 9, 14, 15), the mass of wild birds fell
below the 95% confidence interval for captive
mass. Mass gain with age generally followed
a sigmoidal pattern, with individuals not
reaching an asymptote until after 42 days of
age (Fig. 1). Similarly, from days 1 to 17,
mean wing chord of same-aged wild birds fell
within the 95% confidence interval of captive
birds, with the exception of days 9, 14 and 15,
when mean wing chord measurements for
wild birds fell below the 95% confidence in-
terval. Growth of the wing chord also fol-
lowed a sigmoidal pattern, although the slope
of the curve was less steep for wing chord
growth than it was for mass gain (Fig. 2).
Mean culmen length and mean tarsus length
did not differ between wild and captive birds
from days 1 to 17. Mean culmen growth for
both wild and captive chicks was relatively
linear with increasing age (Fig. 2).

Relative growth rates. — Relative growth
rates among different parts of the body can be
assessed without reference to age. We found
tarsus length to be long in early development
relative to culmen and wing chord, and it con-
tinued to grow at a faster rate than the culmen
throughout development. Culmen and wing
chord grew at approximately the same rate in
early development until wing chord reached
about 40 mm; as wing chord continued to
grow, culmen length growth rate slowed con-
siderably. Changes in wing chord and body
mass were similar throughout the growth pe-
riod observed (Fig. 3). Changes in tarsus
length and body mass also were similar until
individuals reached approximately 80 g, when
tarsus growth slowed.

Plumage. — Using field data from known-
aged chicks, we constructed a table of weekly
plumage characteristics for Hawaiian Stilt
chicks (Table 1). The presence or loss of
down, as well as overall body color, appear to
be the two best indicators of chick age in the
wild for weeks 1-3. Aging during this time is
more precise if one can determine the pres-
ence and condition of primary sheaths; this
cannot be done, however, without chicks in
hand. Specifically, in week 1 chicks are en-
tirely covered with down, and primary sheaths
are absent. The dorsal surface of the body in-

Reed el al. • HAWAIIAN STILT CHICK GROWTH

481

FIG. 2. Growth patterns for wing chord, culmen
length, and tarsus length of captive and wild Hawaiian
Stilt chicks.

eluding head, neck, back, and wing is mottled
black, golden brown, and white; the ventral
surface is creamy white. In week 2, the head
begins to turn brown and is distinctly lighter
than the rest of the body. Mottling on the neck
changes to a more solid pattern of gray and
tan. Most importantly, primary sheaths
emerge on day 12. During week 3, down be-
gins to disappear, giving chicks a sleeker ap-
pearance. Overall body coloration changes
from mottled black, golden brown, and white
to plain gray and white, and primary sheaths
are broken about day 16. In all cases where
we had information on both subspecies, plum-
age descriptions of known-aged Black-necked
Stilt chicks matched exactly the plumage de-
scriptions of known-aged Hawaiian Stilt
chicks up to and including six weeks of age.

160i

120

40-

a □

□ □ oa°

^>13 °

Q f □

a„Eb a

160-

120-

80-

4 0

8 0

1 20

1 60

f □ %

4 0

8 0

120 160

160 -

■a

120 -

80 -

40 -

□ B

40 80 120

Mass (g)

160

FIG. 3. Relative growth rates of three body mea-
surements of wild Hawaiian Stilt chicks compared to
body mass.

From 4-6 weeks, age classes can be differ-
entiated by the presence of tail feathers, the
ability to fly, and the presence of an eye patch
and eye ring. In week 4, tail feathers emerge
and the eye patch and eye ring become visible.
During week 5, all down is lost, wing feathers
are fully developed enabling short distance
flight (up to 1.5 m), and the eye patch is dis-
tinct. Finally in week 6, chicks are capable of
prolonged flight.

As with other shorebirds (Clark 1961), the
egg tooth typically was lost after the first day
and always was gone after 48 h.

DISCUSSION

Because shorebird chicks feed themselves,
they hatch with well developed legs and a

482

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

-C

■o ^

I ^

E o

* <u

^ 2

Oi)

o •-

C« £

s ^

2 15

M §

^ (U

cd • ''

.ti

j= ^
^ 2

^ u

S ^

o i-

^ >

cd • »•

1 ”1
s

- i

'5 o £1

“o

■o

o ^
B c

2 o

c/3 U

2 13
o h
wj c

^ (U

>
u ^
cd •

— (U

n w
■= U
"O ■§ .ti
u 5 x:

s ^ *

ffi"

c3 o

c
o

i:; Ui

o ^
2

- TD

^ X>

S ^

2 o

C to

3 w

!l! c

O cd
4=
W) ^
2

c -2
.3 Wj
13

2 13

o h

00 C

r (U

cd • ^
— (u

t-i

e 2

o ^

•o

i:: X) W)

•o

zz u-

O ^

^ a

c/3

X
c/3 *“

0 ^

c/3

4^ O
O ^
Cd X

■X

3
O

X '-

O x:

c/3

-u

§ ^

V <€.

o o

>»

CO <u

c ^

2 Un

. ^

c/3 ^

o ^

3 ^

0) X

o 3

c/3 cd

4— >

c/3

00 T3
c ^
.3 (U
3 X

3 ^

X x

•3 C

w w

W) o

3 3

»rt

Week 6 None Black & white Mostly black w/brown flecks Dorsal gray; Ventral white Mottled black & golden

brown

Reed el al. • HAWAIIAN STILT CHICK GROWTH

483

00
>> C

w-c

(U O

c ca

<l>

C 2

c/3

C/)

c/3

j_,

^ .

^ o

—

Oi)

6
*U
fU

T3 >

:= O

O ^

(/)

>>
ca
•O

CJ c/3

cd y

^ 5::^
iH

= P

<U C
T3 'C
<u
Cl

^ -5

^ B
^ ■
CJ

•o

ji: ^
w o
cd cii
o

c/3

• f

‘w 2

CJ W

> X

.S X

to ^

c/3

13 O

w ^

O X)

c/3

—

>»

’O

-o

C/)

c/3

s/5

i~>

c/3

CZI

"O c/3

^3 c/3

C dj
dJ J-
C/3 C

-S?

CZl

c/3

.9-

o .B-

3 .&

.9-

■*— '

CL w

■*— '

Tai

^ I

c/3

C/3 »H

L- CX

(1^ ^

« O

^3; Lh o j-i

CQ X) X3

U "u

C ..

c ^

03 ^

(U T3
00

^s--

o i3
u Z 5
> -a ^
"' „ o

o -c

o -C) <u

^ N- SJ

o -D lU

— i.. o =$
■a Z' x; ^

C/3

r<i

»o

*>.

484

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

partly developed bill; wing development be-
gins later and is rapid once started (Galbraith
1988, Thompson et al. 1990). Growth patterns
are variable among species (Holland and Yal-
den 1991; Starck and Ricklefs 1998a, b). For
example, body mass at fledging as a percent-
age of asymptotic adult body mass varies in
shorebirds (Scolopacidae and Charadriidae)
from 53% to 91% (Beintema and Visser
1989a). In addition, it has been suggested that
shorebirds have a higher growth coefficient
{K) than other terrestrial, precocial birds
(Beintema and Visser 1989b). Of 15 species
reviewed by Beintema and Visser (1989a), K
ranged from 0.051 to 0.158, and the inflection
point {T) ranged from 5.5 to 23.8 d after
hatch. Not surprisingly, body size is positively
correlated with the inflection point and nega-
tively correlated with the growth coefficient.
That is, larger species reach the half-way point
in growth at a relatively larger size, and grow
at a slower rate in proportion to their adult
body size, than do smaller species. Hawaiian
Stilts conform to these patterns.

Shape of growth curves. — Captive Hawai-
ian Stilt chicks grew from approximately 15
g at hatching to 125 g at fledging, attaining
only 60% of adult body mass when they
fledged. Culmen, tarsus, and wing chord also
were still growing at fledging, well below
adult sizes, and did not reach adult values un-
til after 42 days after hatching. Culmen and
tarsus sizes increased rapidly between hatch-
ing and fledging, with culmen growth gener-
ally following a linear trajectory and tarsus
following a slightly sigmoidal pattern. Wing
chord growth was sigmoidal, with slow
growth from hatch day to day 12 followed by
a substantial increase in growth rate when
chicks reached 13—15 days old.

Mass loss in the first 24-48 h after hatch
has been reported in some shorebird species
(e.g.. Lapwing, Vcinellus vanellus; Galbraith
1988) and is attributed to movement away
from the nest cup soon after hatching. Al-
though Hawaiian Stilts also leave their nest
cup within a day of hatching, we found no
significant mass loss for captive or wild chicks
from day of hatch to day 1. Differences in the
distance traveled and the amount of food
available in the first 24 h may explain inter-
specific and intraspecific differences in shore-
bird mass loss immediately after hatching.

Reasons for variation in shorebird post-hatch-
ing mass loss require further investigation.

Comparison of captive and wild chick
growth. — Captive and wild chick masses did
not differ significantly for most ages; when
they differed, wild birds were lighter than cap-
tive birds. By the end of week 1 captive
chicks generally were growing at a faster rate
than wild chicks for all growth parameters
measured. This trend mirrors results from oth-
er studies of precocial birds (Beintema and
Visser 1989a). In most cases, captive and wild
chicks have similar growth curves, with more
variation in the growth of wild chicks (Visser
and Ricklefs 1993). Faster growth in captivity
could be due to an unlimited food supply,
while slower growth in the wild could be at-
tributed to colder weather, which increases the
costs of thermoregulation and reduces the
amount of time that chicks can spend forag-
ing. A study of time budgets in the field of
three precocial charadriiform species revealed
that during adverse weather, young chicks
were brooded for 75% of the daytime, and as
a result, they could not obtain enough food to
satisfy their energy requirements (Beintema
and Visser 1989a). In contrast, during good
weather conditions, chicks foraged almost
continuously once they were able to thermo-
regulate.

Beintema and Visser (1989a, b) hypothe-
sized that for shorebird species, cold temper-
atures and cold with rain are the main causes
of slower chick growth in the wild. Specifi-
cally, temperatures dropping below 15° C
slowed chick growth. In Hawaii, temperatures
in coastal wetlands where Hawaiian Stilts
breed rarely fall below 21° C, and there are no
records of temperatures as low as 15° C. In
addition, rains at coastal areas typically are
short-lived. The fact that growth was slower
in wild chicks despite temperatures above
15° C suggests that temperature itself is not
the main factor affecting slower Hawaiian
Stilt chick growth in the field. At warmer tem-
peratures, Pierce (1986) observed faster
growth in other stilt species. Either a different
threshold applies to Hawaiian Stilts or differ-
ences were due to food availability (Beintema
1994).

Comparison to other species. — Hawaiian
Stilts grow slowly in comparison to other
shorebirds. Of the 42 growth coefficients

Reed et al. • HAWAIIAN STILT CHICK GROWTH

485

Starck and Ricklefs (1998a) reported for 27
species of shorebird, only 5 were lower than
what we calculated for Hawaiian Stilts, and
all came from heavier species. The only pub-
lished estimates of Himantopus growth coef-
ficients are Starck’s and Ricklefs’ (1998a) cal-
culations from Pierce’s (1986) data on Pied
{Himantopus himantopus leucocephalus) and
Black {H. novaezealandiae) stilts. These spe-
cies have lower adult masses (129 g and 130
g, respectively) than do Hawaiian Stilts, but
do not fledge until a later age. Hawaiian Stilts
fledge approximately 28 days after hatching;
Pied Stilt chicks do not fledge until they are
34 d, and the Black Stilt fledges even later (at
46 d; Pierce 1986). Similar to the Hawaiian
Stilt, both species continue to grow after
fledging. However, based on data presented by
Pierce (1986: fig. 6), Pied and Black stilts
fledge at a higher percent of their adult body
mass. Consequently, despite the longer time
to fledging. Pied and Black stilt growth co-
efficients are consistent with expectations
based on their adult size {K = 0.175 and 0.129
respectively; Starck and Ricklefs 1998a). A K
of 0.074 would be expected for the 202.5 g
Hawaiian Stilt (Beintema and Visser 1989a),
but we observed K = 0.065 for Hawaiian
Stilts in captivity (and possibly lower in the
field; Fig. 1). Starck and Ricklefs (1998a) also
reported faster growth coefficients for the Eu-
ropean Avocet (Recurvirostra avosetta\ K =
0.213 and 0.171 from two different studies),
which is similar in mass to Hawaiian stilts
(168 g and 250 g, respectively). Although the
relationship between body m.ass and K in
Charadriiformes, is poor {H = 8%, n = 75
species; Starck and Ricklefs 1998b), these
data demonstrate that the slow growth rate ob-
served in Hawaiian Stilts is not a character-
istic of the Recurvirostridae.

We do not know why Hawaiian Stilts have
slow growth. The two obvious hypotheses do
not provide satisfactory explanations. First,
growth rate could be correlated with latitude.
Tropical environments provide a longer breed-
ing season, and growth rates of tropical altri-
cial species are lower than are those of taxo-
nomically related temperate species (Ricklefs
1976, Oniki and Ricklefs 1981). The Hawai-
ian Stilt breeding season lasts six months
(Coleman 1981). Despite this, neither the in-
cubation nor fledging period is prolonged.

Worldwide, stilts average 22-26 days of in-
cubation (Johnsgard 1981), which incorpo-
rates the Hawaiian Stilt’s incubation length of
25 days (Colemen 1981). As noted above, the
fledging time is shorter in this species than in
others of its genus (Johnsgard 1981, Pierce
1986) so there is no extended time as a chick.
There are no studies of which we are aware
comparing growth rates of precocial species
across a latitudinal gradient, but it would be
an interesting assessment.

Second, the lower growth rate could be a
consequence of evolving in an island environ-
ment where predation rates might have been
relatively low before human occupancy, and
selection for rapid growth might have been
relaxed. Most recorded mortality of adult Ha-
waiian Stilts is attributed to introduced species
(Woodside 1979). However, one would expect
slower growth to be associated with an older
age at fledging, which does not occur. In con-
trast, the Hawaiian Stilt fledges at a smaller
percent of adult body mass than do other stilts,
resulting in an extended post-fledging growth
period.

Estimating age. — Ideally, estimates of
chick age would be based on a trait that
changes rapidly and monotonically throughout
growth. One problem with this method is that
often no one trait is ideal throughout the entire
growth period. Rather, traits differ in their ac-
curacy for aging as chicks become older. For
example, measurements of tarsus and wing
chord for Hawaiian Stilts are not useful for
aging chicks at early and late ages because of
their sigmoidal growth patterns. Using mass
as an indicator of chick age is problematic be-
cause it fluctuates rapidly, depending on en-
vironmental conditions and when chicks are
weighed in relation to their last feeding. For
Hawaiian Stilts, culmen length may be the
most useful parameter for aging chicks be-
cause its growth trajectory is fairly linear. Be-
cause it typically has a constant growth rate
throughout the chick stage, culmen length has
been used to age chicks of other shorebird
species in the wild (Beintema and Visser
1989a). However, even for traits that tend to
vary linearly and monotonically throughout
development, there is a tremendous amount of
individual variation in daily growth. Unfor-
tunately, this individual variation is magnified
by measurement error when all measurements

486

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

are not made by the same person. Thus, de-
termination of chick age using body measure-
ments and mass, regardless of the species,
should be viewed as approximate at best.

As a result, we decided to describe general
plumage patterns for Hawaiian Stilt chicks of
known age in the field to set up criteria for
establishing weekly age classes for chicks, de-
fined by the presence or absence of specific
plumage characteristics. The ability to identify
approximate chick age in the field without
capturing young of this endangered species
could aid in management by helping to iden-
tify the age at which chicks disappear. To this
end, we found definitive differences between
plumage characteristics of specific age classes
of Hawaiian Stilts. This should increase abil-
ities to assess survival, the least understood
life-history component of this species (Reed
et al. 1998).

We found plumage characteristics to be use-
ful for identifying weekly age classes of Ha-
waiian Stilts. Because culmen length is similar
for captive and wild chicks and has a linear
positive relationship with age throughout de-
velopment, a combination of culmen length
and plumage description may be the most ac-
curate way to age wild Hawaiian Stilt chicks.
Relying primarily on plumage characteristics,
specifically because they are non-invasive,
and supplementing these observations with
culmen lengths if chicks are captured, will
help minimize interference in this endangered
species while providing managers with a tool
for monitoring reproductive success and pop-
ulation numbers. Because adults and fledg-
lings differ in plumage patterns, it also will
allow accurate monitoring of reproductive
success before molt.

ACKNOWLEDGMENTS

We thank the United States Fish and Wildlife Ser-
vice’s Pacific Refuges Office, the Hawaii State Divi-
sion of Forestry and Wildlife, and the Marine Corps
Base Hawaii, Kaneohe Bay for support and use of their
facilities, and M. Silbernagle and C. Terry for logistic
support. This manuscript benefited from reviews by C.
Elphick, N. Warnock, R. Ricklefs, and two anonymous
reviewers. Research was supported by a grant from the
North DakoUi/National Science Foundation EPSCoR
to LWO, by National Science Foundation grants DEB
9322733 fto J.M.R. and L.W.O.) and DEB 9424375
(to L.W.O.), and by the U.S. Dept, of Agriculture.

LITERATURE CITED

Anthony, N. B., D. A. Emmerson, K. E. Nestor, W.
L. Bacon, P. B. StECEL, and E. A. Dunnington.
1991. Comparison of growth curves of weight se-
lected populations of turkey, quail, and chickens.
Poultry Sci. 70:13-19.

Beintema, a. J. 1994. Condition indices for wader
chicks derived from body-weight and bill-length.
Bird Study 41:68-75.

Beintema, A. J. and G. H. Visser. 1989a. Growth pa-
rameters on chicks of charadriiform birds. Ardea
77:169-180.

Beintema, A. J. and G. H. Visser. 1989b. The effect
of weather on time budgets and development of
chicks of meadow birds. Ardea 77:181-192.

Clark, G. A., Jr. 1961. Occurrence and timing of egg
teeth in birds. Wilson Bull. 73:268—278.

Coleman, R. A. 1981. The reproductive biology of the
Hawaiian subspecies of the Black-necked Stilt,
Himantopus mexicanu.s knudseni. Ph.D. diss.,
Pennsylvania State Univ., State College.

CoocH, E. G., D. B. Lank, A. Dzubin, R. E Rockwell,
AND E Cooke. 1991. Body size variation in Lesser
Snow Geese: environmental plasticity in gosling
growth rates. Ecology 72:503-512.

Emlen, S. T, P. H. Wrege, N. J. Demong, and R. E.
Hegner. 1991. Flexible growth rates in nestling
White-fronted Bee-eaters: a possible adaptation to
short-term food shortage. Condor 93:591-597.

Galbraith, H. 1988. Adaptation and constraint in the
growth pattern of Lapwing Vanellus vanellu.s
chicks. J. Zool. Lond. 215:537—548.

Holland, P. K. and D. W. Yalden. 1991. Growth of
Common Sandpiper chicks. Wader Study Group
Bull. 62:13-15.

JoHNSGARD, P. A. 1981. The plovers, sandpipers, and
snipes of the world. Univ. of Nebraska Press, Lin-
coln.

O’Connor, R. J. 1984. Growth and development of
birds. Wiley, New York.

Oniki, Y. and R. E. Ricklefs. 1981. More growth rates
of birds in the humid New World tropics. Ibis 123:
349-354.

Pierce, R. J. 1986. Differences in susceptibility to pre-
dation during nesting between Pied and Black
stilts (Himantopus spp.). Auk 103:273-280.

Reed, J. M., C. S. Elphick, and L. W. Oring. 1998.
Life-history and viability analysis of the endan-
gered Hawaiian Stilt. Biol. Conserv. 84:34—45.

Reed, J. M. and L. W. Oring. 1993. Long-term pop-
ulation trends of the endangered Ae’o (Hawaiian
Stilt, Himantopus mexicanus knudseni). Trans.
West. Wild. Soc. 29:54-60.

Ricklefs, R. E. 1968. Patterns of growth in birds. Ibis
1 10:419-451.

Ricklefs, R. E. 1973. Patterns of growth in birds. 11.
Growth rate and mode of development. Ibis 1 15:
177-200.

Ricklefs, R. E. 1976. Growth rate of birds in the hu-
mid New World tropics. Ibis 118:176-207.

Reed el al. • HAWAIIAN STILT CHICK GROWTH

487

Ricklefs, R. E. 1983. Avian po.stnatal development.
Avian Biol. 7: 1-83.

Robinson, J. A., J. M. Reed, J. R Skorupa, and L. W.
Oring. In press. Black-necked Stilt (Himantopus
mexicanus). In The birds of North America (A.
Poole and E Gill, Eds.). The Academy of Natural
Sciences, Philadelphia, Pennsylvania; The Amer-
ican Ornithologists’ Union, Washington, D.C.

SPSS, Inc. 1995. SPSS for Windows. Base system
user’s guide. Release 7.0 ed. SPSS Inc., Chicago,
Illinois.

Starck, j. M. and R. E. Ricklefs. 1998a. Avian
growth rate data set. Pp. 381-423 in Avian growth
and development: evolution within the altricial-
precocial spectrum (J. M. Starck and R. E. Rick-
lefs, Eds.). Oxford Univ. Press, New York.

Starck, J. M. and R. E. Ricklefs. 1998b. Variation,
constraint, and phylogeny: comparative analysis
of variation in growth. Pp. 247-265 in Avian

growth and development: evolution within the al-
tricial-precocial spectrum (J. M. Starck and R. E.
Ricklefs, Eds.). Oxford Univ. Press, New York.

Thompson, P. S., C. McCarty, and W. G. Hale. 1990.
Growth and development of Redshank Trinf^a lo-
taniis chicks on the Ribble saltmarshes, N.W. Eng-
land. Ringing & Migr. 11:57-64.

U.S.F.W.S. 1985. Hawaiian waterbirds recovery plan.
U.S. Fish and Wild!. Serv., Portland, Oregon.

VissER, G. H. AND R. E. Ricklefs. 1993. Development
of temperature regulation in shorebirds. Physiol.
Zool. 66:771-782.

Wilson, S. B. and A. H. Evans. 1893. Aves Ha-
waiienses: birds of the Sandwich Islands. Arno
Press (1974), New York.

WooDSiDE, D. H. 1979. Limited study of nesting by
stilt on the islands of Maui, Oahu and Kauai. Pro-
gress Report, Job No. R-III-C, Project No. W-18-
R-4. Division of Forestry and Wildlife, Honolulu,
Hawaii.

Wilson Bull., 111(4), 1999, pp. 488-493

NESTING BEHAVIOR OF THE LILAC-CROWNED PARROT

KATHERINE RENTON' ^ AND ALEJANDRO SALINAS -MELGOZA^

ABSTRACT. — Nesting behavior of the Lilac-crowned Parrot (Amazona finschi) was observed over a three
year period at 24 nests in the tropical dry forest of the Chamela-Cuixmala Biosphere Reserve, western Mexico.
Nest site characteristics and the pattern of parental care throughout the nesting cycle are described for this
mainland Amazon parrot and compared with that reported for other Amazon parrot species. Nest sites were
located in natural cavities of large mature trees characteristic of semi-deciduous forest. Nest sites were similar
to one another in tree species, tree size, cavity height, and entrance width, indicating that Lilac-crowned Parrots
may select nest sites based on these characteristics. Unlike most parrot species. Lilac-crowned Parrots showed
low nest site reuse and high synchrony of nest initiation. Throughout the nesting cycle, females and nestlings
were fed only twice a day on average. Nest attendance during feeding visits was short. The infrequent feeding
visits and short nest attendance exhibited by Lilac-crowned Parrots corresponds with that found for other main-
land Amazon parrots in northeastern Mexico, but contrasts with the multiple feedings and longer nest attendance
observed for island Amazon species. The distinct aspects of Lilac-crowned Parrot nesting behavior may be
related to predation rate and food resource availability during the extreme dry season. Received 10 March 1999,
accepted 15 July 1999.

Most studies on the ecology of Neotropical
Amazon parrots have been conducted on spe-
cies inhabiting the Caribbean Islands (Snyder
et al. 1987, Gnam 1991, Gnam and Rockwell
1991, Wilson et al. 1995). A comparative
study of three species of mainland Amazon
parrot in north-eastern Mexico found distinct
differences compared with island species in
some aspects of reproductive behavior and
productivity (Enkerlin-Hoeflich 1995). Fe-
males and nestlings of three mainland Ama-
zon parrot species were fed only twice a day
(Enkerlin-Hoeflich 1995), in comparison with
the multiple feedings observed for Caribbean
Amazons (Snyder et al. 1987, Gnam 1991,
Wilson et al. 1995) and other Neotropical par-
rots (Fanning 1991, Waltman and Beissinger
1992). Additional data on mainland Amazon
parrots are needed to determine whether con-
clusions from studies on island species are ap-
plicable to mainland species. The Amazon
parrot species of Mexico are particularly suit-
ed for comparison with the Amazona species
of the Greater Antilles in the Caribbean be-
cause of their close evolutionary relationship
(Snyder et al. 1987, Forshaw 1989). The Li-
lac-crowned Parrot (Amazona finschi) is en-

' Durrell Institute of Con.servation & Ecology, The
Univ. of Kent at Canterbury, Kent, CT2 7NJ, U.K.

^ Fundacion Ecologica de Cuixmala, Apartado Post-
al 161 , .San Patricio-Melaque, Jalisco, CP 48980, Mex-
ico.

’ Corresponding author;

E-mail; fundacion_ecologica@gdl.icanet.net.mx

demic to western Mexico and has a restricted
distribution from southeastern Sonora to Oa-
xaca (Forshaw 1989). There have been no
studies on the ecology of the Lilac-crowned
Parrot, and little is known of its breeding bi-
ology (Forshaw 1989). Anecdotal reports
from captive breeding give an incubation pe-
riod of 28 days, with the young chick leaving
the nest after 60 days (Mann and Mann 1978).
In this paper we present observations on the
nest site requirements and nesting behavior of
the Lilac-crowned Parrot in the wild, and
compare them to observations for other island
and mainland Amazon parrots.

STUDY AREA AND METHODS

Studies on the breeding biology of the Lilac-
crowned Parrot were conducted at the 13,142 ha Cha-
mela-Cuixmala Biosphere Reserve (19° 22' N,
104° 56' W to 19° 35' N, 105° 03' W) on the Pacific
coast of Mexico. The study site has a dry tropical
climate exhibiting a marked seasonality in precipita-
tion, with 80% of the 748 mm average annual rainfall
occurring June to November, and a prolonged drought
from mid-February to late May (Bullock 1986). The
reserve has a hilly topography varying in elevation
from 20-520 m above sea level. The dominant veg-
etation type on the slopes is tropical dry deciduous
forest, with semi-deciduous forest in the larger drain-
ages and more humid valleys (Lott et al. 1987, Lott
1993). Monospecific forests of Celaenodendron mex-
icaniun also occur as discontinuous patches within
the tropical deciduous forest mosaic (Martijena and
Bullock 1994).

Observations on the nesting behavior of Lilac-
crowned Parrots were conducted from January to June
in 1996-1998. Nest searches were earned out in Feb-

488

Renum and Salinas-Melf>oza • LILAC-CROWNED PARROT NESTING BEHAVIOR

489

TABLE 1.

Cavity dimensions for 26 Lilac-crowned Parrot nests.

Nest character

Mean

Standard deviation

Range

Coefficient of variation
(s/m) X 100

Tree diameter at brea.st height (cm)

43.1

1 1.4

27.7-66.3

26.3%

Cavity height from ground (m)

9.7

1.7

7.4-14.7

17.8%

Entrance width (cm)

10.0

2.3

6.4-14.0

22.5%

Entrance length (cm)

21.0

16.0

7.5-71.6

76.2%

Internal diameter (cm)

19.9

6.8

10.5-35.0

34.3%

Cavity depth (cm)

66.2

51.7

24-260

78.1%

Circumference at entrance (cm)

86.7

30.0

26-135

34.5%

Neare.st active nest (m)

948.9

707.7

25-2419

74.6%

ruary during the nest prospecting and early incubation
phases of the parrot breeding cycle. No additional
nests were located later in the nesting cycle because
the behavior of breeding pairs made detection of nest
sites difficult. A cavity was considered a potential nest
site if one or both of the adult panots were observed
entering it. The cavity was considered an active nest
site if one of the adult parrots remained within the
cavity for longer than 30 min. Nest site reuse was de-
termined from the frequency of cavity occupancy be-
tween years.

Access to nest cavities was achieved using both sin-
gle-rope ascending (Perry 1978, Perry and Williams
1981) and a tree bole climbing technique (Donahue
and Wood 1995). Nesting requirements of the Lilac-
crowned Parrot were determined by measurement of
nest cavity dimensions: tree species, diameter at breast
height (DBH) of the tree, height above ground of the
entrance, width and length of entrance, cavity depth,
internal diameter, and circumference of the tree at en-
trance (Saunders 1979, Saunders et al. 1982). The lo-
cation of each nest site and where possible the tree
used by the nesting pair for the transfer of food from
the male to the female were obtained using a geo-
graphic positioning system. The coefficient of varia-
tion was determined for the mean cavity dimensions
to evaluate the variability of characteristics between
nest sites.

Behavior of breeding pairs was determined by ob-
servations of parrot nests from covered blinds using
10 X 40 binoculars. Continuous dawn to dusk obser-
vations were conducted on 30 man-days (360 hours)
at 8 nests. No activity was observed at nests during
mid-day; therefore additional observations were re-
stricted to the first four hours after sunrise and the last
three hours prior to sunset giving an additional 299
hours of observation at 16 nests. Parental care and in-
vestment was evaluated from the number of feeding
visits to the nest, arrival time, duration of feeding visit,
time spent in the nest cavity, and time spent in the nest
area (defined as within 100 m of the nest). Descriptive
statistics are presented with means, ranges, and stan-
dard deviations.

RESULTS

Nest site characteristics. — A total of 29
nest sites were located in 1995-1998, all of
which occurred in natural cavities. Nest cav-
ities were located in live trees of Celaenoden-
dron rnexicanum, local name Guayabillo
(51.7%, n = 15), and Astronium graveolens,
local name Culebro (31.0%, n = 9). Of the
remaining 5 cavities, 2 were located in a Ta-
bebua species, 1 was located in a dead tree,
and 2 were located in unidentified trees. Nest
site reuse was low, with only 3 (10.3%) of the
29 nests sites located between 1995 and 1998
being used by nesting pairs over more than
one breeding season. One cavity was used in
three of the four years; 1995, 1997, and 1998.
Two other cavities were reused once after a
vacancy of one year.

Mean cavity dimensions for 26 active nest
sites are presented in Table 1; three cavities
could not be accessed for safety reasons. The
cavity dimensions with the least variation
were height of entrance from the ground and
width of entrance. Diameter of tree at breast
height was relatively consistent between nest
sites and reflects the fact that parrot nests were
located in large, mature trees characteristic of
semi-deciduous forest. The greatest variability
was found in depth of cavity and length of
entrance.

Egg-laying and incubation. — Timing of
egg-laying was highly synchronized between
nests with most pairs commencing incubation
within 14 days of the first nest being initiat-
ed. Mean nest initiation date was 6 February
± 4.6 (SD) days in 1996 (range: 30 January-
13 February, n — 8), and 15 February ±5.3

490

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

days in 1997 (range: 10-23 February, n =
6).

Nest attendance by the female was high
during incubation, with the female leaving the
nest only once in the morning and once in the
afternoon to be fed by the male. The female
spent a mean of 39.4 ± 26.5 min per day out
of the nest over both morning and afternoon
feeding sessions (range: 15-95 min per day, n
= 20). Conversely, the male was rarely ob-
served entering the nest or perching on the
nest rim. Daily activity periods were consis-
tent between nests with the male making an
average 2.1 ± 0.3 nest visits per day (range
2-3 visits, n — 35) to feed the female. Mean
arrival times for the morning and afternoon
activity periods were respectively 08:24 EST
± 46 min (range: 06:07-10:08, n — 45) and
18:14 EST ± 26 min (range: 17:15—19:03, n
= 40). Each feeding visit by a male lasted an
average of 33.9 ± 25.6 min (range: 5-113
min, n = 80).

The male usually vocalized loudly on his
approach to the nest area and perched in a tree
adjacent to the nest cavity making low contact
vocalizations until the female emerged. The
nesting pair gave a characteristic take-off
squawk, or bugle, as the female flew from the
nest cavity to join the male. Food transfer
from the male to the female took place in a
regular perch tree located an average 423 ±
228 m (range: 149-983 m, n = 11) from the
nest cavity. The food transfer session was the
only time during the incubation phase when
both adults were away from the nest area and
was short in duration (average 11.5 ± 10.4
min, range: 2-48 min, n = 78).

Parental care. — Eggs hatched asynchro-
nously, and females continued to brood nest-
lings during the day until the oldest nestling
was 19.6 ± 2.7 days old (range: 15-23 days,
n = 9). Females ceased roosting in nests over-
night when the youngest chick was in its third
week. During this early nestling phase, the
male continued to feed the female twice a day
and was occasionally observed to enter the
nest for a mean 1.7 ± 2.8 min (range: 0—8
min, n — 9) per feeding visit. Later in the
nesting cycle when the chicks were larger,
both parents entered the nest to feed the
young.

The behavior of nesting pairs altered once
the female began to forage with the male.

Nesting pairs became more secretive around
nests, arriving and departing silently. Pairs
used low, almost inaudible vocalizations when
in the nest area. Pairs were cautious about ap-
proaching the actual nest, and would not do
so if they detected an observer or another dis-
turbance. The nesting pair made an average
2.6 ± 0.9 visits per day (range: 2-4 visits, n
= 25) to feed the nestlings. Average duration
of feeding visits during the nestling phase was
72.3 ± 42.3 min (range: 12-171 min, n = 31).
However, the nesting pair spent the majority
of this time perched in trees around the nest
area. Attendance at the actual nest cavity was
short, lasting an average of 10.6 ± 11.2 min
[range: 1.0-27.2 min, n = 30 (total time either
adult in nest cavity or at entrance)], with a
mean of 5.4 ± 4.5 min per visit (range: 0-17
min, n — 30) spent within the nest cavity, and
a mean of 6.0 ± 12 min (range: 0-20.6 min,
n = 30) perched at the nest rim. Each adult
spent a mean of only 4.0 ± 3.5 min (range:
0.42-17 min, n = 41) inside the nest cavity
per feeding visit. This was sufficient time to
feed the young; however, there was no indi-
cation that parent birds spent any other time
in the nest with the young except when feed-
ing them.

Prior to fledging, nestlings began to climb
to the nest entrance, and were fed at the nest
rim. During this stage, nesting pairs spent
more time perched near the nest entrance
making low contact vocalizations to the
young. Mean age at fledging was 63.7 ± 3.2
days (range: 56-68 days, n = 22). Nestlings
fledged asynchronously, and all nests fledged
young within a 2-3 week period. All nestlings
fledged within 12 days in 1996 (mean fledge
date = 10 May ± 4.34 days, range: 6-18 May,
« = 8), 17 days in 1997 (mean fledge date =
18 May ± 6.68 days, range: 11-28 May, n =
7), and 13 days in 1998 (mean fledge date =
8 May ± 5.16 days, range: 2-15 May, n =
7).

DISCUSSION

The low variability between nest sites in
tree species, size, cavity height, and entrance
width suggests that Lilac-crowned Parrots
may select nest sites based on these charac-
teristics. Predation rates decrease with increas-
ing height of nest sites from the ground (Nils-
son 1984, Wilcove 1985), while the increased

Renton and Salinas-Mel}>oza • LILAC-CROWNED PARROT NESTING BEHAVIOR

491

size of nest entrance required by large birds
may pose greater risks from predation, leading
to specific requirements for entrance dimen-
sions (Christman and Dhondt 1997). Amazon
parrots in northeastern Mexico appear to se-
lect cavities based on tree species, cavity
height, and entrance length (Enkerlin-Hoeflich
1995). Australian cockatoos also demonstrate
species specific requirements related to body
size for entrance dimensions and internal di-
ameter of nest hollows (Saunders et al. 1982).
Enkerlin-Hoeflich (1995) suggested that vari-
ability in several cavity characteristics com-
bined with narrow criteria for a few key char-
acters may provide parrots with the flexibility
to exploit a wide range of available cavities
while limiting predation and competition
threats. In addition, low nest site reuse by Li-
lac-crowned Parrots is contrary to the 30-
40% cavity reuse observed for most other par-
rot species (Saunders 1982, Snyder et al.
1987, Rowley and Chapman 1991, Smith
1991, Enkerlin-Hoeflich 1995). Natural pre-
dation is the main cause of nest failure for
Lilac-crowned Parrots (Renton 1998), hence
infrequent cavity reuse may help to prevent
predators from learning nest site locations
(Sonerud 1985, 1989).

The Lilac-crowned Parrot is notably differ-
ent from other parrot species in its high syn-
chrony in nest initiation, with all nests com-
mencing within two weeks in each season,
and the general nest initiation period compris-
ing the first three weeks in February. There is
no evidence that breeding pairs of the Lilac-
crowned Parrot relay after a nest failure,
which would also lengthen the nesting period.
Most Amazona species have a 3-5 week nest
initiation period each breeding season (Snyder
et al. 1987, Gnam 1991, Enkerlin-Hoeflich
1995). The Monk Parakeet (Myiopsitta mon-
achus) in Argentina extends egg-laying over
a nine week period (Navarro et al. 1992). Aus-
tralian cockatoos have a similar broad egg-
laying period of 5-8 weeks (Saunders 1982,
Smith and Saunders 1986, Rowley and Chap-
man 1991, Smith 1991). The nesting season
of the Lilac-crowned Parrot may be so sharply
defined by the extreme climatic seasonality in
tropical deciduous forest and food resource
availability. Nesting pairs may need to fledge
young before the end of the long dry season
in late May-June when food abundance de-

clines (Renton 1998). Delaying nest initiation
may result in breeding pairs having to conduct
energetically demanding activities of raising
young during this environmentally difficult
period.

The infrequent feeding visits to the nest by
breeding pairs of the Lilac-crowned Parrot
contrasts with the multiple daily feedings not-
ed for island Amazona species (Snyder et al.
1987, Gnam 1991) and other Neotropical par-
rots (Lanning and Shiflett 1983; Lanning
1991; Waltman and Beissinger 1992; K.R.,
pers. obs.), but is consistent with the two nest
visits per day observed for three mainland
Amazon parrots in northeastern Mexico (En-
kerlin-Hoeflich 1995). Morning and afternoon
arrival times, approximately one hour after
sunrise and one hour before sunset, for nesting
pairs of the Lilac-crowned Parrot were similar
to the three Amazona species in northeastern
Mexico (Enkerlin-Hoeflich 1995). Large
cockatoos in dry areas of Australia also re-
strict nest visitation activity to the early morn-
ing and late afternoon, spending the hot, mid-
day periods resting under the shade of leafy
trees (Saunders 1982). Lilac-crowned Parrots
at the study site have been noted to demon-
strate signs of heat stress during the mid-day
hours of 12:00—14:00 by holding wings away
from their bodies and panting with beaks open
(K.R., pers. obs.). Therefore, restricting feed-
ing activity to the early morning and late af-
ternoon may enable parrots to conserve en-
ergy during high mid-day temperatures, par-
ticularly in dry habitats.

In addition to being infrequent, nest atten-
dance by Lilac-crowned Parrots during feed-
ing visits was brief. Most Lilac-crowned Par-
rot activity was conducted away from the nest
area. Nesting pairs were never observed to
forage near the nest, and food transfers from
the male to the female took place an average
423 m from the nest. Island Amazon parrots,
by comparison, may spend longer periods in
the nest cavity brooding and preening young
(Snyder et al. 1987), as well as conducting
food transfers and foraging activities near the
nest (Snyder et al. 1987, Gnam 1991). Infre-
quent visits, short nest attendance, and feeding
away from the nest by Lilac-crowned Parrots
may serve to limit the amount of activity in
the nest area, and reduce the risk of attracting
predators to the nest.

492

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Breeding birds are restricted by competing
demands to maintain their own energetic re-
quirements and to protect and nourish the
young (Martin 1987). While infrequent, short
nest visits may reduce both mid-day energetic
expenditures by foraging adults and the risks
of nest predation, there is a cost in having to
meet the energy demands of the young in a
shorter time. Hence, variations in nesting be-
havior suggest that island and mainland Am-
azon parrots may be employing differing strat-
egies to meet time constraints in caring for the
young.

ACKNOWLEDGMENTS

We are grateful to the Cuixmala Ecological Eoun-
dation for financial and logistical support during the
study, and to the Mexican government department of
SEMARNAP who granted permits for the research.
K.R. received a grant from the Durrell Institute of Con-
servation and Ecology and A.S.M. was supported by
the Cuixmala Ecological Foundation and the Denver
Zoological Foundation. S. Berman, R. A. Griffiths, D.
V. Fanning, B. Miller, N.ER. Snyder, and an anony-
mous reviewer provided constructive comments on the
manuscript.

LITERATURE CITED

Bullock, S. H. 1986. Climate of Chamela, Jalisco, and
trends in the south coastal region of Mexico. Arch.
Met. Geoph. BiocL, Sen B 36:297-316.
Chrlstman, B. J. and A. A. Dhondt. 1997. Nest pre-
dation in Black-capped Chickadees: how safe are
cavity nests? Auk 1 14:769-773.

Donahue, P. K. and T. M. Wood. 1995. A safe, flex-
ible, and non-injurious technique for climbing tall
trees. Selbyana 16:196—200.

Enkerlin-Hoeflich, E. C. 1995. Comparative ecology
and reproductive biology of three species of Ama-
zonci parrots in northeastern Mexico. Ph.D. diss.,
Texas A & M Univ., Kingsville.

Forshaw, J. M. 1989. Parrots of the world, third (re-
vised) ed. Lansdowne Editions, Melbourne, Aus-
tralia.

Gnam, R. S. 1991. Nesting behaviour of the Bahama
Parrot Atnazona leucocephala bahamensis on
Abaco Island, Bahamas. Acta Congr. Int. Ornithol.
20:673-680.

Gnam, R. S. and R. F. Rockwell. 1991. Reproductive
potential and output of the Bahama Parrot Ama-
zona leucocephala bahamen.sLs. Ibis 133:400-
405.

Fanning, D.V. 1991. Distribution and breeding biol-
ogy of the Red-fronted Macaw. Wilson Bull. 103:
357-365.

Fanning, D. V. and J. T. Shielett. 1983. Nesting ecol-
ogy of Thick-hilled Parrots. Condor 85:66-73.
Lorr, E. J. 1993. Annotated checklist of the vascular

flora of the Chamela Bay region, Jalisco, Mexico.
Occas. Pap. Calif. Acad. Sci. 148:1-60.

Lott, E. J., S. H. Bullock, and J. A. Solis-Magal-
LANES. 1987. Floristic diversity and structure of
upland and arroyo forests of coastal Jalisco. Bio-
tropica 19:228-235.

Mann, R. E. H. and P. D. Mann. 1978. Breeding
Finch’s Amazon Parrot (Atnazona finschi). Avicul.
Mag. 84:187-189.

Martijena, N. E. and S. H. Bullock. 1994. Mono-
specific dominance of a tropical deciduous forest
in Mexico. J. Biogeogr. 21:63-74.

Martin, T. E. 1987. Food as a limit on breeding birds:
a life-history perspective. Annu. Rev. Ecol. Syst.
18:453-487.

Navarro, J. L., M. B. Martella, and E. H. Bucher.
1992. Breeding season and productivity of Monk
Parakeets in Cordoba, Argentina. Wilson Bull.
104:413-424.

Nilsson, S. G. 1984. The evolution of nest-site selec-
tion among hole-nesting birds: the importance of
nest predation and competition. Ornis Scand. 15:
167-175.

Perry, D. R. 1978. A method of access into the
crowns of emergent and canopy trees. Biotropica
10:155-157.

Perry, D. R. and J. Williams. 1981. The tropical rain
forest canopy: a method providing total access.
Biotropica 13:283-285.

Renton, K. 1998. Reproductive ecology and conser-
vation of the Lilac-crowned Parrot (Atnazona fin-
schi) in Jalisco, Mexico. Ph.D. diss., Univ. of
Kent, Canterbury, U.K.

Rowley, I. and G. Chapman. 1991. The breeding bi-
ology, food, social organization, demography and
conservation of the Major Mitchell or Pink Cock-
atoo, Cacatua leadbeateri, on the margin of the
Western Australian wheatbelt. Aust. J. Zool. 39:
21 1-261.

Saunders, D. A. 1979. The availability of tree hollows
for use as nest sites by White-tailed Black Cock-
atoos. Aust. Wild. Res. 6:205—216.

Saunders, D. A. 1982. The breeding behaviour and
biology the short-billed form of the White-tailed
Black Cockatoo Calyptorhynchus fiinereits. Ibis
124:422-455.

Saunders, D. A, G. T. Smith, and I. Rowley. 1982.
The availability and dimensions of tree hollows
that provide nest sites for cockatoos (Psittacifonnes)
in Western Australia. Aust. Wild. Res. 9:541—556.

Smith, G. T 1991. Breeding ecology of the Western
Long-billed Corella, Cacatua pastihator paslina-
/or. Wild. Res. 18:91-1 10.

Smith, G. T. and D. A. Saunders. 1986. Clutch size
and productivity in three sympatric species of
cockatoo (Psittacifonnes) in the south-west of
Western Australia. Aust. Wild. Res. 13:275-285.

Snyder, N. E R., J. W. Wiley, and C. B. Kepler.
1987. The parrots of Luquillo: natural history and
conservation of the Puerto Rican Parrot. Western

Renton and Salinas-Melgoza • LILAC-CROWNED PARROT NESTING BEHAVIOR

493

Foundation of Vertebrate Zoology, Los Angeles,
California.

SoNERUD, G. A. 1985. Nest hole shift in Tengmalm’s
Owl Aeo}>olius funereiis as defence against nest
predation involving long-term memory in the
predator. J. Anim. Ecol. 54:179-192.

SoNERUD, G. A. 1989. Reduced predation by pine mar-
tins on nests of Tengmalm’s Owl in relocated box-
es. Anim. Behav. 37:332-334.

Waltman, J. R. and S. R. Beissinger. 1992. Breeding
behavior of the Green-rumped Parrotlet. Wilson
Bull. 104:65-84.

WiLCOVE, D. S. 1985. Nest predation in forest tracts
and the decline of migratory songbirds. Ecology
66:121 1-1214.

Wilson, K. A., R. Field, and M. H. Wilson. 1995.
Successful nesting behavior of Puerto Rican Par-
rots. Wilson Bull. 107:518-529.

Wilson Bull., 111(4), 1999, pp. 494-498

RELATIONSHIPS AMONG RED-COCKADED WOODPECKER
GROUP DENSITY, NESTLING PROVISIONING
RATES, AND HABITAT

RICHARD N. CONNER,' 3 D. CRAIG RUDOLPH,' RICHARD R. SCHAEFER,'
DANIEL SAENZ,' AND CLIFFORD E. SHACKELFORD^

ABSTRACT. — We examined Red-cockaded Woodpecker (Picoides borealis) food provisioning rates of nest-
lings during the 1992 and 1993 breeding seasons on the Vernon Ranger District of the Kisatchie National Forest
in Louisiana. Provisioning rates were monitored at nest trees in moderate (9.8 groups/2 km radius, n = 10) and
low (5.9 groups/2 km radius, n = 10) density populations. Habitat around each cluster was measured within
three radii (100 m, 400 m, and 800 m) to evaluate the possible influence of habitat quality on group density
and nestling provisioning rates. We tested the null hypothesis that habitat quality and provisioning rates would
be similar in areas with different densities of woodpecker groups. We failed to detect differences in nestling
provisioning rates between woodpecker groups in moderate versus low group densities. Woodpecker groups
from areas where group densities were moderate attempted to nest significantly more often than woodpecker
groups occurring in low densities. Hardwood midstory vegetation was more abundant in areas with low wood-
pecker group density. Old-growth pines, which are known to be important for cavity excavation, were present
in habitat around cavity-tree clusters of moderate-density groups, but generally absent in areas where group
density was low. Woodpecker group density may be related to hardwood midstory conditions and the abundance
and spatial distribution of remnant old pines. Received 18 August 1998, accepted 9 August 1999.

The Red-cockaded Woodpecker {Picoides
borealis) is a cooperatively breeding species
closely associated with older-growth pine for-
ests of the southeastern United States (U.S.
Fish and Wildlife Service 1985, Walters et al.
1988). A single tree, or aggregation of cavity
trees, termed the cluster, is inhabited by a
group of woodpeckers that includes a single
breeding pair. Other adults present in the
group are typically male offspring from pre-
vious breeding seasons (Ligon 1970, Lennartz
et al. 1987).

Considerable information is known about
the woodpecker’s cavity tree requirements.
The Red-cockaded Woodpecker requires old
living pines for its cavities (Conner and
O’Halloran 1987). The presence of fungal de-
cay within the heartwood of pines increases
in frequency as pines age and significantly
shortens the time required for woodpeckers to
excavate nest and roost cavities (Hooper et al.

' Wildlife Habitat and Silviculture Laboratory
(maintained in cooperation with the College of For-
estry, Stephen F. Austin State Univ.), Southern Re-
.search Station, U.S.D.A. Forest Service, Nacogdoches,
TX 75962.

2 Texas Partners in Flight, Texas Parks and Wildlife
Department, Austin, TX 78704.

’ Corresponding author;

E-mail: c_connerrn@ titan.sfasu.edu

1991, Conner et al. 1994, Rudolph et al.
1995). A sufficient diameter of heartwood,
which increases with pine age, also is required
to provide adequate space for cavity excava-
tion (Conner et al. 1994). Pines with suitable
cavities are known to be a critical resource for
Red-cockaded Woodpeckers (Walters et al.
1992a, Conner and Rudolph 1995). Thus, lo-
cation of existing cavities and the number and
distribution of old-growth pines that can be
excavated for cavities likely has a strong in-
fluence on the density and spatial distribution
of woodpecker groups.

The quality of foraging habitat across the
forest landscape also may influence nesting
success and the density of Red-cockaded
Woodpecker populations. Male and female
Red-cockaded Woodpeckers forage at differ-
ent locations on trees and use different meth-
ods to exploit arthropod prey (Ligon 1968,
Hooper and Lennartz 1981). Hooper and Har-
low (1986) observed that foraging Red-cock-
aded Woodpeckers showed little preference
among pine stands that were more than 30
years old, and concluded that once pine stands
reach 30 years old their quality as a foraging
substrate does not improve with further aging.
More recent studies suggest that old-growth
pines provide an increased foraging benefit
(Zwicker 1995, Jones and Hunt 1996, Eng-

494

Conner el al. • RED-COCKADED NESTLING PROVISIONING

495

Strom and Sanders 1997). In this study we
evaluate relationships among forest habitat,
adult woodpecker provisioning rates of nes-
tlings, and woodpecker fledging success of
woodpecker groups where group density is
moderate and where group density is low. We
ask if foraging habitat characteristics are re-
lated to woodpecker group density, nestling
provisioning rates, and nesting success. If
woodpecker group density is a function of for-
aging habitat quality, provisioning rates of
nestlings where woodpecker group density is
moderate might be expected to exceed those
of groups living in lower densities.

STUDY AREA AND METHODS

The Vernon Ranger District of the Kisatchie Na-
tional Eorest (31° 01' N, 93° 02' W) is located in west-
central Louisiana. Longleaf pines (Finns palustris)
compose the bulk of the overstory with grasses and
forbs as the primary ground cover. Hardwood midstory
vegetation is typically minimal on the Vernon Ranger
District but does occur in some areas where the effec-
tiveness of prescribed fire has been reduced. We se-
lected 10 Red-cockaded Woodpecker groups from ar-
eas of the national forest where group density was
moderate (x = 9.8 active cavity-tree clusters per 2-km
radius) and 10 groups from portions of the Ranger Dis-
trict where group density was low [x = 5.9 active cav-
ity-tree clusters per 2-km radius; moderate vs low
group density: = 5.01, P < 0.001; see Hooper and

Lennartz (1995) for classification (moderate vs low) of
woodpecker group densities]. Prior to the 1992 breed-
ing season, all adult woodpeckers roosting within these
20 clusters were captured at their roost cavities,
weighed, and banded with U.S. Fish and Wildlife Ser-
vice metal bands and combinations of color bands to
facilitate individual recognition.

At the onset of the nesting seasons during 1992 and
1993 we climbed cavity trees using Swedish climbing
ladders to determine the location of the nest tree, the
number of eggs in the clutch, and the initial number
of nestlings. During July and August of both years we
determined the number of post-fledging survivors for
each woodpecker group. We quantified adult provi-
sioning rates of nestlings for each woodpecker group
when nestlings were 8, 20, and 23 days old (see Schae-
fer 1996). The total number of provisioning trips made
by adults to feed nestlings was counted during the 3-
hour period following the breeding male’s initial de-
parture from the nest cavity in the morning. We also
identified which adult brought food to nestlings during
each provisioning trip and made an estimate of prey
size (cm) using the adult woodpecker’s bill as a size
scale. Nest trees were climbed after each 3-h provi-
sioning sampling period to verify the number of nest-
lings present in each nest cavity. An adjusted provi-
sioning rate was calculated for each nest by dividing
the total number of feeding trips by the number of

adult woodpeckers in the group, yielding the number
of provisioning trips per adult. Provisioning rates were
also adjusted to simultaneously account for different
numbers of nestlings in cavities and group size.

Red-cockaded Woodpecker use of forest stands for
toraging depends in part on the distance ol' the stand
from cavity trees (DeLotelle et al. 1987). We measured
vegetational characteristics around the geometric cen-
ter of each woodpecker cavity-tree cluster within three
radii: 0-100 m, 101—400 m, and 401-800 m. Forest
compartment stand maps were obtained from the Kis-
atchie National Forest supervisor’s office for those
compartments falling within 800 m of each cluster
studied. Each compartment is comprised of forest
stands of varying size and tree age. Five dominant or
codominant pine trees were randomly selected as cen-
tral points in each forest stand, and habitat character-
istics for each stand were gathered around each of
these five trees. For each stand, means were calculated
for each habitat measurement taken around the five
central trees.

Stand age was determined by coring each central
tree at breast height (1.3 m) with an increment borer
and counting the growth rings of the cores. Five years
were added to the growth ring counts for longleaf pine
to adjust for the minimum years spent as a seedling
(Conner and O’Halloran 1987). Stands were divided
into five age classes: 0-29, 30-49, 50-69, 70-89, and
>90 years old. The diameters of each central tree and
of all live stems larger than 2 cm within 1 1.3 m of the
central tree were measured at breast height with cali-
pers. Stands were divided into two diameter classes
based on average diameters of pines: 30-40 cm and
40.1-50 cm diameter at breast height (dbh). Pines in
smaller diameter classes were excluded because they
were rarely encountered. A one-factor metric basal
area prism was used to measure pine overstory basal
area (m°/ha). Stands were divided into three basal area
classes based on average basal areas of overstory
pines: 0—3, 3.1-12, and 12.1-21 m^/ha.

Hardwood midstory density was visually estimated
and placed into one of five categories: none, sparse,
moderate, dense, and very dense. The effects of mid-
story height (measured with a clinometer) and mids-
tory density may not be obvious when considered in-
dependently. For example, tall, dense midstory con-
ditions may have a different impact on the woodpeck-
ers than would tall, sparse midstory conditions.
Therefore, both midstory height and midstory density
were considered together to obtain measures of suit-
able and unsuitable midstory conditions. Midstory
conditions were considered suitable if the height was
less than 3 m regardless of the density, or if the density
was none to sparse regardless of the height. Midstory
conditions were considered unsuitable if the height
was more than 3 m and the density was moderate to
very dense.

The area (ha) of each forest stand within distance
zones of 100 m, 400 m. and 800 m from each nest tree
was measured with a digitizer using Sigma-Scan*. The
area of each stand was summed for each habitat vari-

496

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 1. Comparisons of mean (± SD) nestling provisioning (feeding) and nest productivity rates between
Red-cockaded Woodpecker groups in areas of moderate (n = 10) and low {n = 10) woodpecker group density
on the Vernon Ranger District, Kisatchie National Lorest, Louisiana, during the 1992 and 1993 breeding seasons.

Group density

Variable

Moderate

Low

No. feeding trips to 8-day-old nestlings

37.0 (17.3)

37.6 (15.5)

0.61

>0.05

No. feeding trips to 20-day-old nestlings

42.1 (17.5)

43.9 (29.2)

0.19

>0.05

No. feeding trips to 23-day-old nestlings

42.6 (18.4)

46.1 (19.6)

0.06

>0.05

Adjusted 8-day feeding rate

12.9 (7.4)

13.3 (4.0)

0.13

>0.05

Adjusted 20-day feeding rate

14.3 (5.9)

15.5 (9.5)

0.38

>0.05

Adjusted 23 -day feeding rate

14.8 (6.3)

16.2 (5.1)

0.57

>0.05

No. adult woodpeckers in group

3.1 (0.9)

2.9 (0.8)

0.48

>0.05

No. of eggs in clutch from successful groups

3.1 (0.9)

2.8 (1.2)

0.80

>0.05

No. of nestlings fledged from successful groups

1.9 (0.9)

1.7 (0.9)

0.67

>0.05

able by age, dbh, and basal area classes as described
above for the 100 m, 400 m, and 800 m concentric
distance zones around each woodpecker cluster ex-
amined and converted to a percentage for each distance
zone.

Data were analyzed using SAS (version 6.12; SAS
Inst. Inc. 1988) on an IBM compatible computer. Per-
centage data for stand area in various habitat classes
were transformed with an arcsin transformation, and
count data for nesting variables were transformed with
a square root function to approximate normality in all
parametric statistical tests. A stepwise discriminant
analysis was used as a data reduction technique to
compare habitat variables between groups occurring in
moderate and low densities. In order to test the hy-
pothesis that habitat quality influences provisioning
rates and woodpecker group density, we calculated
two-tailed r-tests (adjusted for unequal variances) to
test for differences in nest productivity and provision-
ing effort. Lrequencies of nest success, nest attempts,
and nest failures for moderate and low group densities
were compared with analysis using 2X2 contin-
gency tables. Logistic regression (stepwise) was used
to evaluate nest success (successful or not) and nest
attempts (attempted to nest or not) as a function of
habitat characteristics.

RESULTS

Of all habitat variables measured only the
percentage area with suitable midstory con-
ditions within 100 m of the nest tree (Wilks’
Lambda 0.78, P = 0.0356, 1, 18 df) entered
the stepwise discriminant function (75% clas-
sification accuracy) comparing groups in mod-
erate [96.3% ± 6.5 (SE) area with suitable
midstory] and low (67.8% ± 38.5) densities,
suggesting considerable homogeneity of hab-
itat throughout the Vernon Ranger District.
Stands with old-growth pines (pines >90
years old) within 100 m and 400 m of clusters
were present around groups occurring in mod-

erate densities (1.7% ± 5.2 and 2.4% ±5.7
of the area, respectively) but were totally ab-
sent in areas with low woodpecker group den-
sities.

Differences in woodpecker group density
did not appear to be a function of foraging
habitat quality. We detected no significant dif-
ferences between moderate and low wood-
pecker group density in the unadjusted and ad-
justed rates that adults fed nestlings at 8, 20,
or 23 days post hatching, woodpecker group
size, clutch size, and the number of young
successfully fledged (Table 1). However, the
power of our ability to detect a difference is
low (5-9%) because of the relatively small
sample size (n = 20). As with provisioning
rates, we failed to detect a difference between
moderate and low woodpecker group density
in the size of prey that adults fed to nestlings
(t = 0.45, P > 0.05, df = 24). The average
weight of breeding males was identical in ar-
eas of moderate group density (48.3 g) and
low group density (48.3 g). Breeding female
woodpeckers differed by only 0.1 g, 46.6 g
and 46.7 g, respectively.

A significantly higher proportion of wood-
pecker groups attempted to nest in moderate
group densities (19 of 20 nest years) than in
low group densities (10 of 20 nest years;

= 10.2, P < 0.001, df = 1). Also, a signifi-
cantly higher proportion of woodpecker
groups nested successfully in moderate group
densities (17 of 20 nest years) than in low
group densities (9 of 20 nest years; = 7.03,
P - 0.008, df = 1). The three groups that
failed to produce fledglings in the areas with
moderate group density were the result of two

Conner el al. • RED-COCK ADED NESTLING PROVISIONING

497

nesting attempts that failed and one instance
where the woodpeckers did not attempt to nest
because the breeding female abandoned the
cluster (or died) immediately prior to the
breeding season. The 1 1 groups that failed to
produce fledglings in the low-density groups
were the result of one attempt (eggs laid) that
failed, five instances where the breeding fe-
male disappeared immediately prior to the
breeding season, and five instances where a
pair was present but did not attempt to nest.
Two (10.5%) of 19 nesting attempts in areas
with moderate group density failed, whereas
1 (10.0%) of 10 nesting attempts in areas with
low group density failed (x“ = 0.002, P >
0.05, df = 1). Cluster abandonment by fe-
males prior to the breeding season was mar-
ginally higher in low group densities (5 of 20
nest years) than in moderate group densities
(1 of 20 nest years; = 3.14, P = 0.08, df
= 1).

Attempts to evaluate nest success (success-
ful or not) and nest attempts (attempted to nest
or not) as a function of habitat characteristics
through logistic regression failed as no mea-
sured habitat characteristic had a sufficient re-
lationship to enter the analyses.

DISCUSSION

The observed lower rate of nesting attempts
in the low-density groups relative to moder-
ate-density groups could be related to at least
several factors singly or in combination: (1)
failure to nest because of a foraging habitat
insufficiency, (2) demographic dysfunction re-
sulting from increased isolation of low-density
groups relative to moderate-density groups,
and (3) an inadequate number of older- growth
pines suitable for nest trees. Inadequacy of
foraging habitat appears to be an unlikely ex-
planation. The lack of older-growth pines in
the low group density area may have reduced
the number of sites available for cavity-tree
clusters, and because of increased group iso-
lation, may have had a negative effect on pop-
ulation demographics. Unfortunately, the his-
toric demographics of groups we studied was
not known. It is also possible that prey avail-
ability, as provided by these older-growth
pines, had an influence on cluster abandon-
ment by females and whether groups attempt-
ed to nest or not. Limiting factors appear to

prevent nest initiation rather than decreasing
the success rate of nesting attempts.

Recent research suggests that Red-cockad-
ed Woodpeckers have a preference for older
pines. Zwicker (1995), Engstrom and Sanders
(1997), and Jones and Hunt (1996) observed
that Red-cockaded Woodpeckers used larger,
older-growth pines at much higher rates than
would be expected based on availability.
However, using logistic regression we failed
to detect a relationship between availability of
old pines and nesting attempts.

The age and experience of the breeders oc-
cupying moderate and low density group areas
also may have influenced the observed differ-
ences in nesting productivity. Older and more
experienced woodpeckers might preferentially
capture and occupy habitat with higher den-
sities of older-growth pines because such hab-
itat is viewed as better quality than habitat
lacking older-growth pines. The number of
young fledged by Red-cockaded Woodpeckers
is known to increase with the age and expe-
rience of breeders (Walters et al. 1992b). We
did not know the ages of the woodpeckers in
the groups we studied.

The suitability of hardwood midstory con-
ditions within 100 m of the center of cavity-
tree clusters was significantly greater for mod-
erate-density groups, but it was not related to
nest productivity or the propensity of groups
to nest. The greater presence of unsuitable
midstory conditions in areas of low group
density than areas of moderate group density
suggests that cluster abandonment could also
have influenced the observed differences in
group density. The Red-cockaded Woodpeck-
er’s requirement for open pine stands relative-
ly devoid of hardwood midstory is well
known (Conner and Rudolph 1989, Loeb et
al. 1992).

ACKNOWLEDGMENTS

We thank U.S. Forest Service personnel from the
Vernon Ranger District of the Kisatchie National For-
est for logistical assistance throughout the study and
the Southern Region Office R-8 for financial support.
We thank K. G. Beal. N. R. Canie, R. T. Engstrom. S.
Forbes, E C. James, R. G. Hooper, N. E. Koerth, J. D.
Ligon, and J. R. Walters for constructive comments
leading to the improvement of the manuscript.

LITERATURE CITED

Conner, R. N. and K. A. O'Halloran. 1987. Cavity-
tree selection by Red-cockaded Woodpeckers as

498

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

related to growth dynamics of southern pines.
Wilson Bull. 99:398-412.

Conner, R. N. and D. C. Rudolph. 1989. Red-cock-
aded Woodpecker colony status and trends on the
Angelina, Davy Crockett and Sabine National
Forests. U.S.D.A. For. Serv., Res. Pap. SO-250:1-
15.

Conner, R. N. and D. C. Rudolph. 1995. Excavation
dynamics and use patterns of Red-cockaded
Woodpecker cavities: relationships with coopera-
tive 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 F. Austin State
Univ., Nacogdoches, Texas.

Conner, R. N., D. C. Rudolph, D. Saenz, and R. R.
Schaefer. 1994. Heartwood, sap wood, and fungal
decay associated with Red-cockaded Woodpecker
cavity trees. J. Wildl. Manage. 58:728-734.

DeLotelle, R. S., R. J. Epting, and J. R. Newman.
1987. Habitat use and territory characteristics of
Red-cockaded Woodpeckers in central Florida.
Wilson Bull. 99:202-217.

Engstrom, R. T. and F J. Sanders. 1997. Red-cock-
aded Woodpecker foraging ecology in an old-
growth longleaf pine forest. Wilson Bull. 109:
203-217.

Hooper, R. G. and R. F. Harlow. 1986. Foraging
stands selected by foraging Red-cockaded Wood-
peckers. U.S.D.A. For. Serv. Res. Pap. SE-259: 1-
10.

Hooper, R. G. and M. R. Lennartz. 1981. Foraging
behavior of the Red-cockaded Woodpecker in
South Carolina. Auk 98:321—334.

Hooper, R. G. and M. R. Lennartz. 1995. Short-term
response of a high density Red-cockaded Wood-
pecker population to loss of foraging habitat. Pp.
283-289 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, Tex-
as.

Hooper, R. G., M. R. Lennartz, and H. D. Muse.
1991. Heart rot and cavity tree selection by Red-
cockaded Woodpeckers. J. Wildl. Manage. 55:
323-327.

Jones, C. M. and H. E. Hunt. 1996. Foraging habitat
of the Red-cockaded Woodpecker on the

D’Arbonne National Wildlife Refuge, Louisiana.
J. Field Ornithol. 67:511-518.

Lennartz, M. R., R. G. Hooper, and R. F. Harlow.

1987. Sociality and cooperative breeding of Red-
cockaded Woodpeckers, {Picoides borealis). Be-
hav. Fcol. Sociobiol. 20:77-88.

Ligon, j. D. 1968. Sexual differences in foraging be-
havior in two species of Dendrocopos woodpeck-
ers. Auk 85:203-215.

Ligon, J. D. 1970. Behavior and breeding biology of
the Red-cockaded Woodpecker. Auk 87:255-278.
Loeb, S. C., W. D. Pepper, and A. T. Doyle. 1992.
Habitat characteristics of active and abandoned
Red-cockaded Woodpecker colonies. So. J. Appl.
For. 16:120-125.

Rudolph, D. C., R. N. Conner, and R. R. Schaefer.
1995. Red-cockaded Woodpecker detection of red
heart infection. Pp. 338-342 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.

SAS Inst., Inc. 1988. SAS/STAT user’s guide. Release
6.03. SAS Institute, Inc., Cary, North Carolina.
Schaefer, R. R. 1996. Red-cockaded Woodpecker re-
production and provisioning of nestlings in rela-
tion to habitat. M.Sc. thesis, Stephen F. Austin
State Univ., Nacogdoches, Texas.

U.S. Fish and Wildlife Service. 1985. Red-cockaded
Woodpecker recovery plan. U.S. Fish and Wildlife
Service, Atlanta, Georgia.

Walters, J. R., C. K. Copeyon, and J. H. Carter, III.
1992a. Test of the ecological basis of cooperative
breeding in Red-cockaded Woodpeckers. Auk
109:90-97.

Walters, J. R., P. D. Doerr, and J. H. Carter, III.
1992b. Delayed dispersal and reproduction as a
life-history tactic in cooperative breeders: fitness
calculations from Red-cockaded Woodpeckers.
Am. Nat. 139:623-643.

Walters, J. R., P. D. Doerr, and J. H. Carter, III.

1988. The cooperative breeding system of the
Red-cockaded Woodpecker. Ethology 78:275-
305.

ZwiCKER, S. M. 1995. Selection of pines for foraging
and cavity excavation by Red-cockaded Wood-
peckers. M.Sc. thesis. North Carolina State Univ.,
Raleigh.

Wilson Bull., 111(4), 1999, pp. 499-504

RESPONSES OF BELL’S VIREOS TO BROOD PARASITISM BY THE
BROWN-HEADED COWBIRD IN KANSAS

TIMOTHY H. PARKER' 2

ABSTRACT. — I studied patterns of cowbird parasitism and responses to this parasitism by Bell’s Vireos (Vireo
hellii) in Kansas. Bell’s Vireos abandoned parasitized nests at a significantly higher rate than unparasitized nests.
Lower probability of brood parasitism later in the season may help make abandonment followed by renesting
beneficial. Burial of cowbird eggs by vireos was also observed in several cases. I did not detect a strong
relationship between nest site vegetation characteristics and the probability of brood parasitism. Received 9 Nov.
1998. accepted 27 May 1999.

Bell’s Vireo (Vireo bellii) is a well known
host of the brood parasitic Brown-headed
Cowbird (Molothrus ater\ Barlow 1962,
Mayfield 1965, Franzreb 1987, Brown 1993).
The arrival of the cowbird in California in
this century (Laymon 1987) has been cited
as a major factor causing the severe range
restriction and endangerment of the Least
Bell’s Vireo (V. b. pusillus; Franzreb 1987,
1989; Laymon 1987, Brown 1993). In his re-
view Brown (1993) reported that between
one third to over one half of all Bell’s Vireo
nests monitored in California were parasit-
ized by cowbirds. High rates of parasitism
were also reported in the Great Plains race
(V. b. bellii', Barlow 1962, Brown 1993). Al-
though declines in Bell’s Vireo population
have been detected in some areas of the Great
Plains by the Breeding Bird Survey (Brown
1993), this species is still at least locally
common. The long-term data set (1981-
1997) from my study site shows no decline
[Konza Prairie Long Term Ecological Re-
search (LTER) Site, data set CBPOl]. Sur-
veys elsewhere in the region also have de-
tected higher Bell’s Vireo densities than near-
by Breeding Bird Survey routes (Robbins et
al. 1992, 1993; M. B. Robbins, pers. comm.).
This suggests that cowbird parasitism, de-
spite its frequency, may not be causing a rap-
id decline in Bell’s Vireo on the Great Plains.

It is important to study nest success be-
cause local population numbers may not re-
flect local reproduction (because of source-

' Division of Biology, Kansas State Univ., Manhat-
tan, KS 66506.

^ Present address: Dept, of Biology, Univ. of New
Mexico, Albuquerque, NM 87131;

E-mail: tparker@unm.edu

sink dynamics; Brawn and Robinson 1996).
If the Bell’s Vireo is not declining rapidly on
the Great Plains, we might expect this pop-
ulation to possess traits that would allow its
persistence in the face of cowbird parasitism.
Vireos could try to avoid parasitism altogeth-
er, they could attempt to salvage nesting at-
tempts after parasitism has occurred, or they
could simply abandon parasitized nests and
renest (Clark and Robertson 1981, Hill and
Sealy 1994). Avoidance measures could in-
clude cryptic nest placement, secretive be-
havior around the nest (Uyehara and Narins
1995), and/or aggressive nest defense (Neu-
dorf and Sealy 1994, Robertson and Norman
1977). Two means of salvaging a parasitized
nest include removal of cowbird eggs (Roth-
stein 1975) or burial of cowbird eggs with
nesting material (Clark and Robertson 1981,
Sealy 1996).

In this paper, I consider the potential roles
for avoidance of cowbirds and salvaging or
abandoning parasitized nests by Bell’s Vireos
in Kansas. Analysis of nest site vegetation
coupled with observations of nest contents al-
lowed exploration of cryptic nest placement,
burial of cowbird eggs, and nest abandonment
followed by renesting.

METHODS

From May through Augu.st of 1996 I inve.stigated
cowbird parasitism and nest success of Bell's Vireos
in the Flint Hills of northeastern Kansas. My study site
was located on a portion of the Nature Con.servancy's
Konza Prairie Research Natural Area (in Riley and
Geary counties). The site consisted of tallgrass prairie
interspersed with deciduous shrub vegetation concen-
trated around ephemeral streams and limestone out-
croppings. Vireos arrive at this site beginning in mid-
May and initiate nest building in late May, but re-
nesting attempts continue into early July. Nests are

499

500

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

placed in low deciduous shrubs usually within 1.5 m
of the ground (unpubl. data).

Nest building by Bell’s Vireos usually takes 3 to 4
days and egg laying follows 1 or 2 days after nest
completion. Males aid in nest site selection, nest con-
struction, incubation, and feeding of young. Typically
4 eggs are laid (Brown 1993). On the Great Plains,
two broods are sometimes reared in one season (Bar-
low 1962, Brown 1993).

I searched for and monitored nests throughout the
season using the locations of singing males to narrow
my search. I located most nests before the onset of
incubation. The entire study area (110 ha) was
searched every 3-4 days. I also visited active nests
once every 3—4 days to record the number of eggs and/
or nestlings present (including cowbird) and to look
for any buried cowbird eggs. If no adult vireo was
active in the vicinity of the nest, I felt the eggs for
warmth to determine if they were being incubated. I
ceased visiting a nest and concluded it had failed after
two successive visits where I observed cold eggs and
no parental activity. If the entire contents of a nest
were removed, or if during incubation or the nestling
period most of the contents were removed and the par-
ents ceased attending the nest, I concluded that the nest
was depredated. I did not visit nests if cowbirds or
predators were in the vicinity (Martin and Geupel
1993).

To minimize nest disturbance, I waited until the nest
was no longer in use before assessing vegetation near
the nest. I measured height of nest, height of the nest
shrub, and depth of leaf litter. I counted the number of
woody stems within a 50 cm horizontal radius of the
nest shrub and estimated nest concealment (percentage
of nest hidden from the observer by vegetation at dis-
tance of 1 m, BBIRD protocol) above, below, and in
the four cardinal directions around the nest. I estimated
ground cover (a proxy for vegetation density; defined
as the percent of ground covered by a given vegetation
type within 5 m of the nest) for nest substrate (the plant
species in which nest placed), large shrubs (the size
class used by vireos for nesting), all woody vegetation,
woody clumps (closed canopy continuous woody veg-
etation), sparse woody clumps (open canopy continu-
ous woody vegetation), grass, and the three most com-
mon woody species within 5 m. Also within the 5 m,
1 estimated the median height of the woody canopy,
measured the height of the tallest woody stem, and
counted the number of dead woody stems and the
numbers of live woody stems under 2.5 cm diameter
and over 2.5 cm diameter. To assess the area of ground
covered (both within and outside of the 5 m radius)
by the woody clump in which the nest was placed, 1
measured the maximum width of the clump and the
width perpendicular to the maximum. The distance
from the nest to the nearest corridor of woody vege-
tation along a stream bed was recorded as well.

By noting changes in nest contents and whether or
not nests were active, I assessed vireo re.sponses to
cowbird parasitism: nest material placed over cowbird
eggs or nest abandonment subsequent to the laying of

TABLE 1. Fates of Bell’s Vireo nests parasitized
and not parasitized by Brown-headed Cowbirds. All

nests included were completed and
found during building or laying.

active

and were

Para.sitized

Unparasitized

nests

Abandoned

32“

8*’

Depredated

Fledged (cowbird if parasitized.

8^-

vireos if not)

8“=

TotaE

“ Includes two nests in which cowbird eggs were buried and other cowbird
eggs were later laid.

^ Includes one nest in which a cowbird egg was buried.

Total offspring fledged = 24.

One nest counted as parasitized here contained a buried cowbird egg.
Without further cowbird eggs being laid, the nest was later abandoned and
so was considered unparasitized when abandoned.

cowbird eggs. A nest was considered abandoned (as
opposed to depredated) if parental activity ceased and
either the number of vireo eggs had not declined be-
tween visits or the number of vireo eggs had decreased
but the number of cowbird eggs had inereased between
visits. Using a test (Sokal and Rohlf 1987), I com-
pared the proportion of parasitized nests that were
abandoned to the proportion of unparasitized nests
which were abandoned. All completed (nest lining
complete), active (adults defending nest) nests located
before or during the laying stage were included in the
analysis (/? = 63).

To identify the factors associated with nest aban-
donment, I compared parasitized nests (n = 43; does
not include 1 nest abandoned after vireos had buried
a cowbird egg; Table 1) that were abandoned to those
that were not abandoned based on the numbers of vireo
eggs and cowbird eggs in the nests. Numbers of vireo
and cowbird eggs were considered separately in two t-
te.sts (using f-test assuming equal variances, Microsoft
Excel 7.0).

Although the vireos were not banded, I conserva-
tively estimated renesting attempts by comparing nest
locations with dates of nest use for all nests (n = 63)
included in this study. I considered a nesting attempt
to be a renesting event if it occurred within 7 days of
the cessation of use of a nearby nest. If a nesting at-
tempt was begun after a longer period, I considered it
a possible renesting attempt (presumably in some of
these cases I may have missed an intervening nesting
attempt). Furthermore, a nest could be considered part
of a given .series of renestings (or possible renestings)
only if the location of the nest did not overlap with
the locations of a different series of apparent rene.st-
ings. Becau.se Bell’s Vireos are territorial (Brown
1993), I made the conservative as.sumption that terri-
tories (i.c., series of nesting attempts) did not overlap
and were consistent throughout the season to avoid
overestimating renesting. 1 located a number of i.solat-
ed nests which, based on their late dates of initiation.

Parker • BROOD PARASITISM OF BELL’S VIREOS

501

were probably renesting attempts; however, I did not
count these as renestings because I could not identify
any previous nests.

A reasonable estimate of the proportion of pairs pro-
ducing offspring was not possible because for 25 of
the estimated 33 vireo pairs, only one or two nesting
attempts were observed for each pair. Therefore I
could not rule out the possibility that other, possibly
successful, nesting attempts were not detected.

To assess the timing of nest initiation on nest suc-
cess, I conducted the following analyses. I compared
vireo nest initiation dates (Julian dates) for both par-
asitized and unparasitized nests {n = 56) using a
Mann- Whitney f/-test. This analysis included all com-
pleted, active nests found during building or laying
except for those unparasitized nests that were aban-
doned early in the nesting cycle (/; = 7). For the nests
1 excluded from the analyses, I could not rule out the
possibility that parasitism might have occurred had the
nest remained in use. I also compared the nest initia-
tion dates for both depredated and fledged (fledged ei-
ther cowbird or vireo young) nests {n = 23) using a
Mann- Whitney t/-test. Included in this analysis were
completed, active nests found during building or lay-
ing that were either depredated or fledged. Finally, us-
ing a Mann-Whitney t/-test, I compared vireo nest ini-
tiation dates for successful (fledged vireos) and unsuc-
cessful (all other fates) nests. All 63 complete, active
nests found during building or laying were included.
t/-test F-values were obtained from Sokal and Rohlf
(1987).

I included 28 variables describing vegetation sur-
rounding nests in a step-wise discriminant function
analysis (using PROC STEPDISC, SAS 6.12, for a
UNIX operating system) to compare parasitized to un-
parasitized nests. I set the critical P-value for entering
and remaining in the model at 0.05. I included nests
found at all stages of the nesting cycle for which I had
measured vegetation (unparasitized n = 15, parasitized
n = 50) except for those unparasitized nests that were
abandoned early in the nesting cycle. For these aban-
doned nests, I could not rule out the possibility that
parasitism might have occurred had the nest remained
in use.

RESULTS

Of the 63 completed and active Bell’s Vireo
nests found during nest building and laying,
44 (70%) were parasitized by at least one
cowbird egg but only 3 of these fledged a
cowbird young (Table 1). None of the para-
sitized nests fledged any vireo young. A mean
of 1.5 cowbird eggs were laid in each para-
sitized nest, and a mean of 1.5 vireo eggs were
present in each such nest after cowbird activ-
ity (possibly egg removal; Brown 1993).

Of the 44 parasitized Bell’s Vireo nests in-
cluded in this analysis, in only 4 (9%) did the
vireo parents use additional nest material to

cover one or more cowbird eggs laid in their
nests. All nests with buried eggs subsequently
failed for a variety of reasons (Table 1). In
none of the nests with buried eggs could 1 rule
out the possibility that cowbird eggs had been
buried during the process of nest building be-
cause they were laid in nests under construc-
tion.

Nest abandonment following cowbird par-
asitism in my study was frequent. Of the 43
parasitized nests (does not include 1 nest
abandoned after vireos had buried cowbird
egg, see Table 1), 32 were abandoned. This is
a significantly higher proportion of abandon-
ment than that expected based on the frequen-
cy of abandonment for unparasitized nests (8
of 20; = 21.22, P < 0.001).

Abandoned nests had significantly fewer
host eggs than non-abandoned nests [aban-
doned: X = 0.9 ±0.1 (SE); non-abandoned: x
= 3.3 ± 0.1; t = -7.04, P < 0.001]. Aban-
donment was not significantly related to the
number of cowbird eggs laid (abandoned: x =
1.6 ± 0.1; non-abandoned: x = 1.2 ± 0.1; f
= 1.45, P > 0.05).

Of 63 nests, 1 estimated 20 (32%) were re-
nesting attempts and 10 (16%) were probable
renesting attempts. Of the 8 nests that fledged
vireo young, 6 appeared to have been renest-
ing attempts.

Unparasitized nests (« =12) were initiated
significantly later {U — 378, P < 0.05) than
parasitized nests {n = 44; Fig. 1). No differ-
ence in initiation date was found between dep-
redated {n = 12) and fledged (/? = 11) nests
in initiation date ((/ = 91, P > 0.05; Fig. 1).
Successful nests (/? = 8) did not differ from
failed nests (n = 55) in date of initiation {U
= 235.5, P > 0.5; Fig. 1).

The nest substrate species was selected by
step-wise discriminant analysis as a significant
predictor of cowbird parasitism (F = 5.29, P
= 0.0248, F = 0.08). Unparasitized nests
were surrounded within 5 m by more of the
plant species in which the nest was placed
than were parasitized nests. No other vegeta-
tion variables distinguished parasitized from
unparasitized nests.

DISCUSSION

During my one season of study, abandon-
ment (and apparent renesting) was the most
common response of Bell’s Vireos to brood

number of nests number of nests number of nests

502

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

nest initiation dates

failed nests

\///A successful nests

J7A ^

”5

“D

LIG. 1. Dates of Bell’s Vireo nest initiation (in 10
day intervals; May-July 1996) for (1 A) parasitized and
unparasitized nests, (IB) depredated and fledged (ei-
ther cowbird or vireo young) nests, and ( 1C) failed and
successful (Bell’s Vireo fledged) nests.

parasitism by cowbirds. Abandoned nests had
fewer host eggs than non-abandoned nests.
This result is consistent with other findings
that egg removal by cowbirds, rather than the
presence of cowbird eggs in the nest, is the
stimulus that leads to nest abandonment (Bar-
low 1962, Hill and Sealy 1994, Woodworth
1997). It is also consistent with the hypothesis
that nest abandonment is a generalized re-
sponse to egg loss as opposed to a specific
response to parasitism (Rothstein 1975).

The seasonal activity of Brown-headed
Cowbirds could be a factor favoring nest
abandonment by the Bell’s Vireo. Unparasit-
ized vireo nests were initiated significantly
later in the season. Egg laying by cowbirds

declined more quickly than vireo nest initia-
tion over the breeding season at my study site;
a finding similar to other studies (Scott and
Ankney 1980, Hill and Sealy 1994). Those
Bell’s Vireos that nest later therefore are less
likely to be parasitized. This suggests that nest
abandonment followed by renesting is bene-
ficial for the vireos. No costs to later nesting
were detected; neither depredated nests nor
nests that failed from all causes differed in
initiation date from successful nests. However,
post-fledging success was not followed.

Nest abandonment may be a complemen-
tary tactic to egg burial. Abandonment may
be effective at high rates of parasitism while
egg burying may be effective at lower rates.
In this study, burying of cowbird eggs was
rare and was not a successful tactic, partially
because of subsequent cowbird parasitism.
However, when rates of parasitism are lower
(i.e., with a lower probability of subsequent
cowbird eggs being laid) this behavior might
be beneficial. Burial is probably less expen-
sive energetically than constructing a new nest
(Clark and Robertson 1981). Frequency of
parasitism on my study site may be unusually
high in comparison to the Great Plains as a
whole (29% of nests parasitized, Friedmann et
al. 1977; 13-69% of nests parasitized. Brown
1993). If this is so, then this study may un-
derestimate the importance of egg burying in
allowing Great Plains Bell’s Vireos to persist
in the presence of cowbirds.

Egg burying has not been reported for
Bell’s Vireos in California (Salata 1983, Gray
and Greaves 1984, Franzreb 1989). Cowbirds
have occupied most of California only in the
past century (Laymon 1987), so their hosts
there may not have had enough time to evolve
adaptive responses to brood parasitism (May-
field 1965).

I did not attempt to compare cowbird in-
duced nest abandonment rates in my study to
those reported from California because of two
potentially confounding factors. Unlike my
study, in California cowbird eggs were re-
moved by researchers (Salata 1983, Gray and
Greaves 1984, Franzreb 1989). Therefore, the
observed rate of abandonment in California
may be reduced because not all vireos that
abandon parasitized nests do so immediately
upon receiving a cowbird egg (pers. obs.).
Secondly, usually only one cowbird egg was

Parker • BROOD PARASIl'ISM OF BELL’S VIREOS

503

laid per nest in California (Salata 1983, Gray
and Greaves 1984, Franzreb 1989), possibly
coinciding with removal of only one host egg
by the cowbirds (Lowther 1993). The lower
intensity of parasitism in California than in
Kansas could mean a less intense proximate
cue for vireos to abandon in California. This
could lead to the observation of different
abandonment tendencies in these two vireo
populations regardless of the presence or ab-
sence of any evolved differences between
them.

To better understand cowbird behavior and
the possibility for cryptic nest placement by
vireos, I considered the relationship between
nest-site vegetation and parasitism. I found
that unparasitized nests had more ground cov-
ered (within 5 m of the nest) by the plant spe-
cies in which a given nest was placed (nest
substrate). However, this finding does not nec-
essarily support the idea that an increased den-
sity of vegetation generally hinders searching
by cowbirds because no other measures of
vegetation density were associated with brood
parasitism. Although cowbird parasitism
seems to be affected by vegetation structure
in forests (Brittingham and Temple 1996),
such effects were not apparent in this study.
The predictive value of the variable ‘nest sub-
strate’ was low (r^ = 0.08). With such a weak
relationship between nest placement and
brood parasitism, cowbirds may be a negligi-
ble selective pressure further refining nest
placement in the Bell’s Vireo.

ACKNOWLEDGMENTS

I would like to thank J. L. Zimmerman for his ideas
as I conducted this research. I also wish to thank J. E
Cavitt for his help during data collection. R. Kimball
and C. Fellows provided useful suggestions during
manuscript preparation. Anonymous reviewers made
helpful comments on earlier versions of this manu-
script. I conducted this study at the Konza Prairie Re-
search Natural Area Long Term Ecological Research
Site and I utilized resources provided by Konza LTER.
Nest site measurement protocol and data entry pro-
gram were provided by BBIRD.

LITERATURE CITED

Barlow, J. C. 1962. Natural history of the Bell Vireo,
Vireo hellii Audubon. Univ. Kans. Publ. Mus.
Nat. Hist. 12:241-296.

Brawn, J. D. and S. K. Robinson. 1996. Source-sink
population dynamics may complicate the interpre-
tation of long term census data. Ecology 77:3-12.

Brittingham, M. C. and S. A. Templk. 1996. Vege-
tation around parasitized and non-parasitized nests
within deciduous forest. J. Field Ornithol. 67:406-
413.

Brown, B. T. 1993. Bell’s Vireo (Vireo hellii). In The
birds of North America, no. 35 (A. Poole, P. Stet-
tenheim, and F Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
D.C.

Clark, K. L. and R. J. Robertson. 1981. Cowbird
parasitism and evolution of anti-parasite strategies
in the Yellow Warbler. Wilson Bull. 93:249-258.

Franzreb, K. E. 1987. Endangered status and strate-
gies for conservation of the Least Bell’s Vireo
(Vireo hellii pusillu.s) in California. West. Birds
18:43-49.

Franzreb, K. E. 1989. Ecology and conservation of
the endangered Least Bell’s Vireo. Biological Re-
port 89(1). U.S. Fish and Wildlife Service, Wash-
ington, D.C.

Friedmann, H., L. E Kief, and S. I. Rothstein. 1977.
A further contribution to knowledge of the host
relations of the parasitic cowbirds. Smithson. Con-
trib. Zool. 235:1—75.

Gray, M. V. and J. M. Greaves. 1984. Riparian for-
ests as habitat for the Least Bell’s Vireo. In Cal-
ifornia riparian systems: ecology, conservation,
and productive management (R. E. Warner and K.
M. Hendrix, Eds.). Univ. of California Press,
Berkeley.

Hill, D. P. and S. G. Sealy. 1994. Desertion of nests
parasitized by cowbirds: have Clay-colored Spar-
rows evolved an anti-parasite defense? Anini. Be-
hav. 48:1063-1070.

Laymon, S. a. 1987. Brown-headed Cowbirds in Cal-
ifornia: historical perspectives and management
opportunities in riparian habitats. West. Birds 18:
63-70.

Lowther, P. E. 1993. Brown-headed Cowbird (Mol-
othru.s ater). In The birds of North America, no.
47 (A. Poole, P. Stettenheim, and F. Gill, Eds.).
The Academy of Natural Sciences, Philadelphia,
Pennsylvania; The American Ornithologists’
Union, Washington, D.C.

Martin, T. E. and G. R. Geupel. 1993. Nest-monitor-
ing plots: methods for locating nests and moni-
toring success. J. Field Ornithol. 64:507-519.

Mayfield, H. 1965. The Brown-headed Cowbird, with
old and new hosts. Living Bird 4:12-28.

Neudorf. D. L. and S. G. Sealy. 1994. Sunrise ne.st
attentiveness in cowbird hosts. Condor 96:162—
169.

Robbins, M. B., D. A. Easterla, and D. Mead. 1992.
Avian census of the Nodaway River, northwestern
Missouri. Bluebird. 59:105-107.

Robbins, M. B., D. A. Ea.sterla, and D. Mead. 1993.
1993 avian census of the Nodaway River, north-
western Missouri. Bluebird. 60:110-1 1 1.

Robertson, R. J. and R. E Norman. 1977. The func-
tion and evolution of aggressive host behavior to-

504

THE WILSON BULLETIN ♦ Vol. Ill, No. 4, December 1999

wards the Brown-headed Cowbird (Molothrus
ater). Can. J. Zool. 55:508-518.

Rothstein, S. I. 1975. An experimental and teleonom-
ic investigation of avian brood parasitism. Condor
77:250-271.

Salata, L. R. 1983. Status of the Least Bell’s Vireo
on Camp Pendleton, California. U.S. Fish and
Wildlife Service, Laguna Niguel, California.

Scott, D. M. and C. D. Ankney. 1980. Fecundity of
the Brown-headed Cowbird in southern Ontario.
Auk 97:677-687.

Sealy, S. G. 1996. Evolution of host defenses against
brood parasitism: implication of puncture-ejection
by a small passerine. Auk 1 13:346-355.

SOKAL, R. R. AND E J. Rohle. 1987. Introduction to
biostatistics, second ed. W. H. Freeman and Com-
pany, New York.

Uyehara, j. C. and P. M. Narins. 1995. Nest defense
by Willow Flycatchers to brood-parasitic intrud-
ers. Condor 97:361—368.

Woodworth, B. L. 1997. Brood parasitism, nest pre-
dation, and season-long reproductive success of a
tropical island endemic. Condor 99:606-621.

Wilson Bull., I 1 1(4), 1999, pp. 505-514

THE TYPE B SONG OF THE NORTHERN PARULA: STRUCTURE
AND GEOGRAPHIC VARIATION ALONG PROPOSED
SUB-SPECIES BOUNDARIES

MICHAEL D. BAY' 2

ABSTRACT. — The type B song of the Northern Parula (Parula aniericana) was described from 120 males
recorded throughout much of the species range in North America. Most songs were structured with a series of
complex syllables, followed by simple syllables, trill syllables, and a terminal simple syllable. Some birds sang
songs that contained 2 phrases per song with syllables that varied in structure and number between individuals.
Analysis ot song variables revealed variation at the macrogeographic level with songs from western populations
differing significantly from eastern populations in song duration, frequency, number of trill syllables, and simple
syllables. In addition, variation was evident between eastern and western populations in the structuring of phrase
patterns. Received 8 March J999. accepted 3 August 1999.

Several investigations into wood-warbler
(Parulidae) song behavior have shown that
some species sing two song types (Ficken and
Ficken 1967, Morse 1967, Highsmith 1989)
while other species have a repertoire of sev-
eral songs classified as first and second cate-
gory songs (Lein 1978, Staicer 1989, Byers
1995). The two song types have been referred
to as types A and B (Morse 1967, Nolan
1978), types I and II (Lanyon and Gill 1964,
Gill and Murray 1972, Morrisson and Hardy
1983), or accented and unaccented ending
songs (Morse 1966, Lein 1978). Studies on
the function of song suggest that type A, I, or
accented ending songs [or the type B in
Black-throated Green Warblers (Dendroica vi-
rens) and Blackburnian Warblers (D. fusca)]
are used as intersexual signals and are more
stereotyped; while B, II, or unaccented ending
songs (or the type A in Black- throated Green
and Blackburnian warblers) are used intrasex-
ually and tend to be more variable (e.g.,
Kroodsma 1981, reviewed by Spector 1992).

Although detailed descriptions of parulid
song types can be found for several species, a
few, like the Northern Parula {Parula ameri-
cana), are less well studied. Moldenhauer
(1992) presented a detailed account of the
type A song, but the type B song has yet to
be described in detail spectrographically. In
this study, I present a description of B songs

' Dept, of Biological Sciences. Sam Houston State
Univ., Huntsville, TX 77341.

^ Present address: Dept, of Biology, East Central
Univ., Ada, OK 74820;

E-mail : mbay @ mailclerk.ecok.edu

recorded from several males located through-
out the species’ breeding range (eastern Unit-
ed States and southeastern Canada).

METHODS

Type B songs of 29 male Parula Warblers were re-
corded from 20 localities in Texas, Alabama, Missis-
sippi, Louisiana, Georgia, and Tennessee (Appendix)
from 15 May to 10 June 1986, during the morning
hours of 07:00-10:00 (CST). Although no birds were
color marked, only 1—3 individuals were recorded per
locality, with two investigators recording different in-
dividuals, often at the same time. Type B songs were,
in most instances, elicited by song playback. These
recordings (46 songs) were made using a Uher 4000
Report IC recorder at a tape speed of 19 cm/sec and a
Dan Gibson P650 parabolic microphone. Spectrograms
of recorded songs were produced with a Kay Eleme-
trics 606 IB sonagraph with a wideband filter. Addi-
tional song recordings were obtained from the Texas
Bird Sound Library (Department of Biological Scienc-
es, Sam Houston State University; 33 birds, 50 songs),
Cornell Library of Natural Sounds (15 birds, 27
songs), and the Borror Laboratory of Bioacoustics of
The Ohio State University (43 birds, 58 songs). These
songs were recorded using Nagra III (38 cm/sec) or
Magnemite (38 cm/sec) recorders. An AKG micro-
phone (or an unknown type) was used with a parabola
to record.

The following variables were measured for each
song: (1) duration of song (sec), (2) total number of
syllables, (3) number of syllable types, (4) minimum
frequency, and (5) maximum frequency. For individ-
uals with multiple recordings, 1 computed a within-bird
mean. For the sake of comparison with variation as
reported in the type A song (Moldenhauer 1992), I
used analysis of variance (ANOVA; SAS Institute
1985; a = 0.05) to test for significant differences be-
tween song populations from three geographic areas.
Moldenhauer (1992) primarily reported on the differ-
ences between songs in the western versus eastern re-
gions of the species’ breeding range. I followed this

505

506

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

M if i

^ 0 30 see ^

FIG. 1. Syllable classes of Northern Parula type B
songs. (A) complex syllable (B) simple syllable (C)
trill syllable.

division with the exception that I divide the eastern
population into northeast and southeast (see Appen-
dix). These geographical ranges correspond to the sub-
species (western race, P. a. hidoviciancc, southeastern
race, P. a. americaner, northeastern race, P. a. pusillci)
proposed by Oberholser (1974).

1 classified song elements based on morphological
structure (Baptista 1974, Weins 1982, Bradley and
Bradley 1983) using terminology partly adopted from
Baptista (1974) and Staicer (1989). A note was any
short tracing on the spectrogram, and syllables were
represented by repeated notes or series of notes form-
ing a coherent unit. Three classes of syllables were
identified: (1) simple syllables (SS), those containing
1 or 2 simple notes; (2) complex syllables (CS), syl-
lables with more than 2 notes forming a coherent unit,
or in rare instances 1-2 wavy and continuous notes;
(3) trill syllables (TS), high frequency slurs near the
end of a song (Fig. 1). To distinguish between the
number of different variations or types within each syl-
lable class, a subscript number was added to a sylla-
ble’s abbreviation (e.g., CS14).

A phrase was defined as a series of repeated sylla-
bles forming the following 4 phrase classes of a B
song: (1) complex (C) phra.se, composed of complex
syllables of one type; (2) mixed (M) phra.se, composed
of a mixture of repeated complex and simple syllables
of one type each; (3) trill (T) phrase, composed of trill
syllables of one type; (4) repeated trill (R) phrase,
composed of trill syllables of 2 types (Fig. 2). A phrase
pattern was the entire sequence of phrases in a song
and was symbolized by the letter codes for each phrase
class (e.g., C-T-SS = a song composed of a C phrase,
T phrase, and ending with a simple syllable; Fig. 3).

RESULTS

Song structure. — Thirty-nine syllable types
from 3 syllable classes were discovered in the
type B song, including 21 complex syllables
(Fig. 4), 1 1 simple syllables, and 7 trill syl-
lables (Fig. 5). It appeared that no one indi-
vidual contained more than one complex syl-
lable type and no more than 2 types of simple
and trill syllables.

Each of the syllable classes (complex, sim-

81 A

8 D

\mm

6 05 L5

Time (sec)

FIG. 2. Phrase classes of Northern Parula type B
songs. (A) C phrase (B) M phrase (C) T phrase (D) R
phrase.

pie, and trill) was used to construct song
phrases that typically began with an introduc-
tory series of complex syllables {x = 3.6 syl-
lables per song, SD = 1.4, n = 181 songs)
followed by a high frequency trill (T phrase)
(x = 4.6 syllables per song, SD = 2.7, n =
181 songs) and ending in one to several sim-
ple syllables (SS; x = 0.88 syllables per song,
SD = 2.6, n = 181 songs). Some birds sang
songs that lacked a few of these syllable class-
es (e.g., only the introductory complex sylla-
bles were present) and therefore were shorter
and usually more difficult to hear (low ampli-
tude; e.g., somewhat like muted songs; Morse
1967). The number of syllables within a song
ranged from 2 to 19.5 (x = 10.7), while most
(54%) were composed of 10 to 15 syllables.
Few songs contained more than 15 syllables
(12.6%). Some birds began their song with
chip notes (see Morse 1967) but this was rare-
ly recorded and was not considered in this
analysis.

The most frequent phrase pattern, C-T-SS
(complex-trill-simple syllable), accounted for
most of the songs (67%). Nine different pat-
terns were found, the five most common are
illustrated in Fig. 6. It is likely that some ex-
tremely rare patterns may represent scrambled

Bay • NORTHERN PARULA TYPE B SONG

507

i- ^ f 1 ^ ^ ^ 1 ^ ] ^ 1 hHf--U-Hh-]hHfH h-H

CSi3 syllables TSs syllables SS9

[ - - Song Duration - - -I

EIG. 3. Measurements and coding of a type B song. MxF = maximum frequency, MnF = minimum fre-
quency, CS = complex syllable, TS = trill syllable, SS = simple syllable.

(e.g., T-M-T-SS occurred in 0.4% of songs) or
incomplete song phrases (e.g., M occurred in
1.4% of songs and M-T occurred in 1.5% of
songs).

Geographic variation. — Eight phrase pat-
terns (except for the phrase C-T-SS) occurred
in populations that occupied the western half
of the species’ breeding range and correspond
to those birds that sang the western type A
song (see Moldenhauer 1992). Birds occupy-
ing the eastern range and corresponding to
those that sing the eastern type A song, sang
B songs that were constructed of only three
phrase patterns (C-T-SS, C-T, C; Fig. 6). The
additional variation in phrase patterns ob-
served for the western population is attributed
to the addition of simple and complex sylla-
bles to the introductory portion of the song to
form M phrases. This phrase difference ap-
pears to change or overlap between the Mis-
sissippi and Alabama boundary much as oc-
curs with the type A song (Moldenhauer
1992). However, more data are needed to de-
termine the distribution of type B song phrase
patterns within the east to west type A song
overlap zone.

In addition to phrase variation, songs varied
significantly between geographic areas in
mean duration, mean maximum frequency,
and in the mean number of simple and trill
syllables (Table 1). Songs of western birds av-
eraged 0.10 second longer than songs of
southeastern birds and 0.20 second longer
than songs of northeastern birds {F = 10.90,
df = 2, 134, P < 0.001). Birds from the latter
region averaged 0.5 kHz less than the average

maximum frequency of western and south-
eastern populations (F = 4.31, df = 2, 134, P

< 0.025).

Comparing the usage of syllable types with-
in the 3 syllable classes (complex, simple, and
trill) among birds in the three geographic ar-
eas, the most frequent syllable for western
birds was CSjj (56.6%; Fig. 4, no. 13) fol-
lowed by Tj (43.9%; Fig. 5, no. 8) and SS^
(29.2%; Fig. 5, no. 8). For birds in the south-
east, songs most frequently contained SS^
(47.3%; Fig. 5, no. 8) followed by T, and T,
(both at 39.4%; Fig. 5, no. 1 and 2, respec-
tively). The most frequently used complex
syllables for southeastern songs were CS,3
(18.4%; Fig. 4, no. 13) and CS4 (15.7%; Fig.

4, no. 4). Birds from the northeastern region
primarily used T, (51.2%; Fig. 5, no. 3), CS,9
(36.5%; Fig. 4, no. 19), and SS9 (34.1%; Fig.

5, no. 9) to construct their songs. Of the three
syllable types used to construct the type B
song, significant differences were evident be-
tween the three geographical areas in the
mean number of syllables used per song (Ta-
ble 1). For instance, western singers used sig-
nificantly more simple syllables to construct
the type B song (F = 16.3, df = 2, 134, P <
0.005) but fewer trill syllables in comparison
to eastern singers (F = 23.9, df = 2, 134, P

< 0.005).

Only within specific localities did some
birds sing identical songs (i.e., same phrase
patterns and/or syllables), while others sang
songs using a different type of one or more
syllables within the three syllable classes (e.g.,
one individual might use 08,3 and T^,, while

Frequency (IcHz)

508

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Complex Syllables

^ 0.30 sec ^

FIG. 4. Complex syllables of Northern Parula type B songs. Syllable types 1, 3, 5, 10, 11, and 20 from
eastern birds only, while variants 14, 16, 17, and 18 from we.stern birds only.

another used CS4 and TSO- It is unknown
whether individuals with the same song (i.e.,
same phra.se and syllables) within a specific
locality were neighbors because most of these
songs were not recorded by the author. It is
likely that some of these individuals were
members of the same local population because

many of these songs were recorded within the
same year. Some morphological differences
were evident in specific syllable types be-
tween individuals (e.g., CS,, might differ
slightly between two birds; Fig. 7). Morpho-
logical differences in simple syllables were
evident between those used in the middle of

Frequency (IdHz)

Bay • NORTHERN PARULA TYPE B SONG

509

Simple Syllables

\/ 7 X! y

Trill Syllables

I-

0.30 sec

FIG. 5. Simple and trill syllable.s of Northern Parula type B songs. Simple syllable types 1, 3, and 5 and
trill syllable type 6 from western birds only.

Frequency (IcHz)

510

THE WILSON BULLETIN • Vol. HI. No. 4. December 1999

I-

• I

-C ph rase I I--T phrase--I simple syllable

I M phrase I I — T phrase — I simple syllable

' / / A,/ f A' / W

. 1 ' 1 t 1 Ml'

I C phrase I I--T phrase— I

Ml ^

fi ‘M

I C phrase I

6 0.5 EO E6 ?0

Time (sec)

FIG. 6. The five most common song or phrase patterns observed in Northern Panda type B songs. (A)
pattern C-T-SS [67% of all .songs (west = 0.0%, southeast = 92%, northeast = 100%)], (B) pattern M-T-SS
1 13.1% of all songs (west = 45.0%, southeast = 0.0%, northeast = 0.0%)], (C) pattern C-T [6.2% of all songs
(west = 17%. .southeast = 2.6%, northeast = 0.0%)], (D) pattern M-R-SS [3.6% of all songs (west = 12.1%,
southeast = 0.0%, northeast = ().()%)], (E) pattern C [2.1% of all .songs (west = 2.4%, southeast = 4.1%,
northeast = 0.0%)].

Hay • NORTHERN PARULA TYPE B SONG

511

TABLE 1. A comparison of song variables from western, southeastern, and northeastern type B songs of
the Northern Panda (values are mean ± SD). Number of individuals in which songs were analyzed: West = 41,
Southeast = 38, Northeast = 41.

Song population

ANOVA

Variable

West

Southeast

Northeast

/■'-value'-'

Song duration (s)“

1.5

-h

0.2

1.4

-h

0.2

1.3

-i-

0.1

10.90**

Number of syllables

11.9

2.9

11.3

-h

3.1

11.4

-h

2.7

0.53

Syllable types in a song

4.0

-h

1.3

3.7

-h

1.0

3.6

-h

0.5

1.26

Maximum frequency (k.Hz)'“

7.0

-h

0.4

7.2

0.4

6.6

-h

0.4

4.31*

Minimum frequency (kHz)

4.4

-h

0.5

4.4

-h

0.5

4.2

0.4

2.10

Complex syllables''^’

3.7

-h

1.6

3.5

-h

1.3

3.4

H-

0.9

0.71

Simple syllables"*’

4.6

-h

3.0

1.1

-h

1.0

-F

0.3

16.3***

Trill syllables"*’

2.9

1.5

5.1

-+■

2.6

6.5

H-

2.7

23 9***

^ Multiple Comparison (Bonferroni Correction; a = 0.05): Song Duration (W # Se = Ne), Maximum Frequency (W = Se Ne). Simple Syllables (W
# Se = Ne). Trill Syllables (W # Se Ne).

^ Number per song.

>=* P < 0.025. ** P < 0.001. *** P < 0.0005.

the song (to form M phrases) versus the single
simple syllable terminating the song (as usu-
ally occurs in most parula B songs).

DISCUSSION

Song structure. — Northern Parula B songs
are complex and show much intraspecific var-
iation. Song complexity occurs because most
songs contain multiple phrases (usually two)
that vary in syllable types between individu-
als. A similar arrangement occurs in B songs
of the Grace’s Warbler (Dendroica graciae;
Staicer 1989) and the type II song of the Blue
and Golden-winged warblers (Gill and Murray
1972, Highsmith 1989). Some evidence sug-
gests that complexity in B songs may be a
result of intrasexual usage (e.g., territorial
clashes between males), while A songs, used
for mate attraction, are more stereotyped
(Kroodsma 1981).

Geographic variation. — Northern Parula
type A songs are very similar in most vari-
ables (except trill rate, frequency, and song
length), but differ significantly between east-
ern and western populations in the type of ter-
minal syllable (Moldenhauer 1992). Males
from each population recognize and respond
differentially to the two types of A songs (Re-
gelski and Moldenhauer 1996). Similarly, the
type B song exhibited differences between
eastern and western populations in song
length and maximum frequency, but primarily
differed in the number of specific syllable
types and phrase patterns composing songs.
For instance, western birds frequently used

more simple syllables per song and often used
these syllables to construct the M phrase pat-
tern, which was absent from eastern singers.

Moldenhauer (1992) argued for subspecific
recognition of P. americana (P. a. americana
for the east and P. a. ludoviciana for the west)
based upon the terminal note difference in the
type A songs. Results from my study are con-
sistent with this division, based upon B song
phrase patterning between eastern and western
populations. The terminal note in the type A
song is readily identifiable both audibly and
visually by sonogram. Likewise, B songs with
M phrases (western population) may be au-
dibly distinguished from songs without M
phrases (eastern population); they are longer
(composed of more syllables) and more buz-
zy. These differences are easily viewed by
comparing sonograms (Fig. 6B, D vs Fig. 6 A,
C). Whether these structural differences in
eastern and western type A and B songs are
influenced by learning and/or have a genetic
basis has yet to be determined.

Populations from each geographic region
contained a repertoire of unshared syllables
(15.3% confined to the eastern population,
20.5% confined to the western population);
however, many (57%) were shared. Syllable
confinement within specific macrogeographi-
cal areas reported in my study might be
viewed with some speculation. For instance,
some individual songs contained more than
one syllable type within a syllable class, par-
ticularly trill and simple syllables. No individ-

Frequency (IcHz)

512

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

6-
5-
4-
3-
2-
I-
0-

' 0 30 see *

FIG. 7. Examples of morphological variations of specific syllable types between individuals (A) CSivIcom-
plex syllable), (B) SS, (simple syllable), (OTSj (trill syllable).

Bay • NORTHERN PARULA TYPE B SONG

513

uals for which multiple songs were examined
sang more than 1 complex syllable type. How-
ever, because of the small sample size (only 5
individuals recorded with at least 5 songs
from each), these results should be viewed
with caution. Within specific microgeographic
areas, syllable sharing could be variable even
though a few individuals sang some or all of
the same syllables.

In other warbler species that have a song
similar to the type B of the Northern Parula,
local dialect patterns are usually evident,
which suggests that young males learn B
songs from neighboring conspecifics (Kroods-
ma 1981). Thus an individual from an area
that contains a song similar to his neighbors,
will be more effective during countersinging
bouts (Kroodsma 1981). Playback experi-
ments show that Blue-winged Warbler males
respond more intensely to their local type II
(B) songs but do not differentiate among type
I (A) songs of different localities (Kroodsma
et al. 1984). Although Northern Parula B
songs tend to be complex and differ between
locations (in either phrase pattern or syllable
types used), Bay (1987) found no discemable
patterns or dialects in songs from the two best
sampled areas (representing 10 and 15 indi-
viduals) in Texas.

Future studies should concentrate on gath-
ering information concerning site fidelity in
successive breeding years to allow a better un-
derstanding of type B song structure at the
microgeographic level. Such data might also
reveal whether young males learn the same
types of syllables and phrase patterns in suc-
cessive years or different ones as a result of
the exposure to new individuals. In addition,
researchers should determine what role, if any,
the disjunct winter distribution of the North-
ern Parula has on song learning and the geo-
graphical differences in the type A and B
songs on the breeding grounds

ACKNOWLEDGMENTS

Many thanks to R. Moldenhauer for suggesting this
research and for his advice throughout the study. D.
Spector, C. Staicer, and anonymous reviewers made
valuable suggestions for improving the manuscript. A.
Dewees and M. Duggan gave helpful advice on statis-
tics. Part of this study was caixied out to fulfill require-
ments for the M.A. degree at the Dept, of Biological
Sciences, Sam Houston State University.

LITERATURE CITED

Baptista, L. E 1974. The effects of songs of wintering
White-crowned Spanows on .song development in
.sedentary populations of the .species. Z. Tierpsy-
chol. 34:147-171.

Bay, M. D. 1987. Singing behavior and geographic
variation in the type B song of the Northern Parula
(Parula americana). M.A. thesis, Sam Houston
State Univ., Huntsville, Texas.

Bradley, D. and R. Bradley. 1983. Application of
sequence comparison to the study of bird songs.
In Time warps, string edits and macromolecules:
the theory and practice of sequence comparison
(D. Sankoff and J. Kruskal, Eds.). Proceedings of
the Conference on Sequence Comparison, Mon-
treal, Canada.

Byers, B. E. 1995. Song types, repertoires and song
variability in a population of Chestnut-sided War-
blers. Condor 97:390-401.

Eicken, M. S. and R. W. Eicken. 1967. Singing be-
havior of the Blue-winged Warblers and Golden-
winged Warblers and their hybrids. Behaviour 28:
149-181.

Gill, E B. and B. G. Murray. 1972. Song variation
in sympatric Blue-winged and Golden-winged
warblers. Auk 89:625-643.

Highsmith, R. T. 1989. The singing behavior of Gold-
en-winged Warblers. Wilson Bull. 101:36-50.

Kroodsma, D. E. 1981. Geographical variation and
functions of song types in warblers (Parulidae).
Auk 98:743-751.

Kroodsma, D. E., W. R. Meservey, A. L. Whitlock,
AND W. M. Vanderhaegen. 1984. Blue-winged
Warblers (Vermivora pinus) “recognize” dialects
in type II but not type 1 songs. Behav. Ecol. So-
ciobiol. 15:127-131.

Lanyon, W. E. and E B. Gill. 1964. Spectrographic
analysis of variation in the songs of a population
of Blue-winged Warblers {Vermivora pinus). Am.
Mus. Novitates, No. 2176:1 — 18.

Lein, M. R. 1978. Song variation in a population of
Chestnut-sided Warblers (Dendroica pensylvani-
ca): its nature and suggested significance. Can. J.
Zool. 56:1266-1283.

Moldenhauer, R. R. 1992. Two song populations of
the Northern Parula. Auk 109:215-222.

Morrison, M. M. and J. W. Hardy. 1983. Vocaliza-
tions of the Black-throated Gray Warbler. Wilson
Bull. 95:640-643.

Morse, D. H. 1966. The context of .songs in the Yel-
low Warbler. Wilson Bull. 78:444—455.

Morse, D. H. 1967. The context of songs in the Black-
throated Green and Blackburnian warblers. Wil-
.son Bull. 79:64-74.

Nolan, V., Jr. 1978. The ecology and behavior of the
Prairie Warbler, Dendroica discolor. Ornithol.
Monogr. 26:1-595.

Oberholser, H. C. 1974. The bird life of Texas, 2
vols. Univ. of Texas Press, Austin.

Regelski, D. j. and R. R. Moldenhauer. 1996. Dis-

514

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

crimination between regional song forms in the
Northern Parula. Wilson Bull. 108:335-341.

SAS Institute. 1985. U.ser's guide: statistics. SAS In-
stitute Inc., Cary, North Carolina.

Spector, D. a. 1992. Wood warbler song systems: a
review of paruline singing behaviors. Curr. Orni-
thol. 9:199-238.

Staicer, C. a. 1989. Characteristics, use, and signifi-
cance of two singing behaviors in Grace’s Warbler
(Dendroica graciae). Auk 106:49-63.

Weins, J. a. 1982. Song pattern variation in the Sage
Sparrow (Amphispiza belli): dialects or epiphe-
nomena? Auk 99:208-229.

APPENDIX

Recording localities for Northern Parula
type B songs. Localities are listed by geo-
graphic region and then alphabetically by state
(or country) and county, parish, or province.

Number of individuals per locale are indicated
in parenthesis.

Southeast region. — Alabama: Dale (1),
Dallas (1), Clark (1), Covington (2), Monroe

(2) , Wilcox (4). Florida: Alatchua (1), Dade

(3) , Flager (1), Jefferson (2), Polk (1), Wak-
ulla (1). Georgia: Bulloch (3), Chatham (4),
Jenkins (6), Richmond (1). North Carolina:
New Hanover (3). South Carolina: Richland
(8). Tennessee: Dickson (1), Sevier (1).

Northeast region. — Canada: Gaspe (1).
Maine: Lincoln (44). Ohio: Franklin (2), Sen-
eca (1). West Virginia: Monogalia (1).

West region. — Lousiana: St. Tammany (2).
Mississippi: George (3), Hancock (4). Texas:
Hardin (2), Montgomery (2), Trinity (2),
Walker (25).

Wilson Bull., 1 I 1(4), 1999, pp. 515-527

NESTING BIOLOGY OF DICKClSSELS AND HENSLOW’S
SPARROWS IN SOUTHWESTERN MISSOURI
PRAIRIE FRAGMENTS

MAIKEN WINTER' 2

ABSTRACT. — According to data from the North American Breeding Bird Survey, populations of Dickcissel
(Spizci americana) and Henslow’s SpaiTow {Ammodrannis henslowii) have declined severely during the last 30
years. The reasons for their population declines seem to differ; habitat fragmentation on the breeding grounds
has been suggested to have little negative impact on Dickcissels, but appears to be a major reason for Henslow’s
Sparrow declines. Previous reports on the status of Dickcissels and Henslow’s Sparrows largely were based on
density estimates without considering the nesting biology of the two species. My comparison of the nesting
biology of Dickcissel and Henslow’s Spanow provides some insight into potential factors that might contribute
to their population declines. During 1995-1997, I studied the nesting biology of Dickcissels and Henslow’s
Sparrows in fragments of native tallgrass prairie in southwestern Missouri. Both species had similar clutch sizes,
rates of hatching success, and numbers of young fledged per successful nest. Dickcissels tended to have lower
rates of nesting success and higher rates of brood parasitism by Brown-headed Cowbirds (Molothrus ater) than
Henslow’s Sparrows. Although several vegetation characteristics at the nest differed between successful and
depredated nests in Dickcissels, no differences were found between successful and depredated Henslow’s Spar-
row nests or between parasitized and unparasitized Dickcissel nests. My results indicate that Dickcissels might
reproduce less successfully than Henslow’s Sparrows in southwestern Missouri, and might therefore be of higher
conservation concern on the breeding ground than previously thought. Received 27 January 1999, accepted 3
June 1999.

Data from the North American Breeding
Bird Survey indicated that populations of
Dickcissel (Spiza americana) and Henslow’s
Sparrows (Ammodramus henslowii) have de-
clined by about 39% and 91%, respectively,
during the last 30 years (Peterjohn et al.
1994). The reasons for the declines are
thought to differ between the two species:
Dickcissels are assumed to have declined
mainly because of poisoning on their South
American wintering grounds (Basili and Tem-
ple 1995, Basili 1997), and are thought to be
little affected by breeding habitat loss or frag-
mentation (Herkert et al. 1993). In contrast,
the population decline of Henslow’s Sparrows
seems to be mainly caused by loss and frag-
mentation of suitable grassland habitat on
their breeding grounds (Herkert 1994). How-
ever, status assessments of Dickcissels and
Henslow’s Sparrows are based largely on es-
timates of density or relative abundance, with-
out considering the breeding ecology of the
two species. A comparison of the breeding
ecology of Dickcissels and Henslow’s Spar-

' Division of Biological Sciences, Univ. of Missouri,
Columbia, MO 65211.

^Current address; 611 Winston Ct, Apt 4, Ithaca,
NY 14850; E-mail; mwinte02@syr.edu

rows might provide information on factors
that could cause differential reproductive suc-
cess in the two species. Such factors might
include clutch sizes, rates of brood parasitism
by Brown-headed Cowbirds {Molothrus ater),
rates of nest predation, and hatching and
fledging rates. Vegetation characteristics at the
nest site might differ between the species and
cause one species to be more susceptible to
nest predation or cowbird parasitism.

In southwestern Missouri, little information
has been collected on the nesting success of
passerines breeding in tallgrass prairie frag-
ments. In this study I describe and compare
nesting characteristics of the Dickcissel and
the Henslow’s Sparrow in fragments of native
tallgrass prairie in southwestern Missouri be-
tween 1995 and 1997. Detailed analyses on
the effect of fragment size, proximity to hab-
itat edge, management practices, and land-
scape structure on density and nesting success
of these species are described elsewhere (Win-
ter 1998, Winter and Faaborg in press).

Dickcissel. — Dickcissels are grassland hab-
itat generalists; they can be found breeding in
a wide variety of grassland vegetation (Bent
1968). Because males often sing from elevat-
ed perches and females often place their nests
above the ground, they tolerate a relatively

515

516

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

large number of shrubs and trees. As with
most polygynous species, the female generally
tends both eggs and young alone.

In spite of their ability to use secondary
habitats such as non-native grasslands or road
right-of-way, Dickcissel populations have de-
clined by about 39% since 1966 (Peterjohn et
al. 1994). Population trends in Dickcissels are
difficult to estimate because their abundance
and distribution fluctuate considerably among
years (Fretwell 1986), but habitat fragmenta-
tion did not seem to be responsible for their
population declines (Herkert et al. 1993). In-
stead, poisoning of tens of thousands of birds
on the wintering grounds has been suggested
as cause for its population decline (Basili and
Temple 1995, Basili 1997). Recent evidence
indicates that habitat fragmentation on the
breeding grounds might also have a negative
impact on this species (Winter 1998, Winter
and Faaborg in press).

Henslow's Sparrow. — Little is known
about the nesting behavior of this inconspic-
uous species because of its furtive behavior
and its tendency to spend most of its time on
the ground (Bent 1968). Its highest densities
occur in grasslands with tall, dense vegetation
and a well-developed layer of litter (Wiens
1969, Robins 1971, Zimmerman 1988, Her-
kert 1994, Mazur 1996, Winter 1998). Based
on the few existing nesting studies (Hyde
1939, Bent 1968, Robins 1971, Schulenberg
et al. 1994, Rohrbaugh et al. in press), we
know that this monogamous species generally
nests close to the ground in tall dense vege-
tation, preferably within large clumps of litter.

With the destruction of tallgrass prairie and
similar grassland habitats, the breeding range
of Henslow’s Sparrows has contracted consid-
erably during the last 30 years, mainly in the
northeastern, eastern, and northwestern parts
of its range (Pruitt 1996). Although Henslow’s
Sparrows also nest in secondary habitats such
as hayfields and reclaimed surface mines (see
review in Swanson 1996), it has shown a con-
sistent population decline (Peterjohn et al.
1994, Herkert 1997). Analysis of Christmas
Bird Count data in the southeastern United
States also indicates population declines on
the wintering grounds (Butcher and Lowe
1990). The major reason for the large popu-
lation decline of Henslow’s Spanows has been
suggested to be loss and fragmentation of hab-

itat on the breeding grounds (Herkert 1994).
However, studies in Missouri, Kansas, and
Ohio indicated that Henslow’s Sparrows can
occurr in even small fragments (see Winter
1998), and since 1988 its populations have
been steadily increasing in Illinois (J. R. Her-
kert, pers. comm.).

STUDY SITES AND METHODS

Study area. — Between 1995 and 1997, I studied the
nesting biology of Dickcissels and Henslow’s Spar-
rows in 13 fragments of native tallgrass prairie in
southwestern Missouri (approx. 37° 30' N, 93° 30' W;
Winter 1998). Dominant grasses in the study area in-
cluded big bluestem (Andropogon gerardii), little blue-
stem (Schizocliyrium scoparium), and Indian grass
{Sorghastrum nutans). Dominant forbs included sun-
flower (Heliantlius spp.), milkweed (Asclepias spp.),
blazing star (Liatris spicata), and sensitive briar
(Schrankia nuttallii). Prairies were owned by the Mis-
souri Department of Conservation, the Missouri Prairie
Foundation, The Nature Conservancy, and the Missou-
ri Department of Natural Resources and were actively
managed by prescribed burning and haying (see Winter
1998).

Nest searching and monitoring. — Throughout each
field season (early May to end of July) my field assis-
tants and I located and monitored nests of all grassland
species that we found, but focused our nest searching
efforts on finding nests of Henslow's Sparrows and
Dickcissels. Nests were found by walking across the
study sites and adjacent areas of similar vegetation,
while paying close attention to behavior and vocali-
zations of nearby adult birds. Most nests were found
by observing adults (Dickcissels: 80%; Henslow's
Sparrows: 56%). Behavioral patterns of adults that we
used as clues that nests might be nearby were chipping,
flying short distances away or around the observer,
flushing close to the observer followed by a short
flight, and canying nest material, fecal sacs, or food.
The location of a potential nest site was marked with
a short length of flagging tape at three locations within
I m of a potential nest site forming a triangle. We then
retreated 10-30 m and tried to locate the nest when
the bird returned. Nests also were located by flushing
birds while randomly walking across the prairie (Dick-
cissel: 10%; Henslow's Sparrow: 30%). The remaining
10% and 14% of all nests, respectively, were found
fortuitously; flushing birds while doing other research
activities such as vegetation measurements or census-
ing. Because rope-dragging and systematic search
were ineffective methods for nest finding in 1995, 1
did not use those methods in the following years.

We did not search for or monitor nests when vege-
tation was wet (after rain or heavy dew immediately
after sunrise) to minimize disturbance of vegetation
surrounding nests. Each nest was marked with a flag
5 m to the north and a small ribbon was placed about
30 cm south of those nests that were hard to find. Ev-
ery 3-4 days nest fate was checked by walking past

Winter - HENSLOW’S SPARROW AND DICKCISSEL NESTING

517

the nest to avoid creating ‘‘dead ends” that might lead
nest predators to the nest. During each nest check we
recorded the number of host and cowbird eggs and
young, presence or absence of adults, and the state of
the nest if the nest was found empty. An empty nest
was considered successful if one or more of the fol-
lowing cues were observed; feces in the nest, feather
sheaths in the nest, nest rim flattened, adults carrying
food or chipping, or fledgling close to nest.

Nest vegetation. — Nest vegetation was characterized
within one week after activity at a nest had ceased.
Vegetation was measured at five locations around the
nest site: directly at the nest and 0.5 m from the nest
in each cardinal direction. At each of the five points I
measured vegetation cover (Daubenmire 1959), the
number of woody stems within each Daubenmire
frame, vegetation height, litter depth, and visual ob-
struction (Robel et al. 1970; for a more detailed de-
scription see Winter 1998). For each nest I calculated
the mean for each of the five measuring points, and
used the mean of those five data points for further
analysis.

Estimates of nesting success. — When calculating
rates of nesting success, I excluded nests for which it
was not possible to determine if predation happened
before or after a nest was abandoned. This was true
for nests that had small clutch sizes ( 1-2 eggs) and
were depredated the next time the nest was checked.
Those nests, however, were included for estimating
rates of cowbird parasitism. For each year I estimated
species specific probabilities of daily nest survival
(Mayfield 1975) separately for incubation and nestling
stages, and for the total nesting period. The total prob-
ability of nest survival was defined as the probability
that a nest successfully survived incubation and nest-
ling periods and fledged at least one young of the pa-
rental species. In the two species that I investigated,
incubation begins with the laying of the last egg (Bent
1968). I used the following exponents to estimate the
probability of nesting success over the entire nesting
period; Dickcissel: 21 days (12 incubation days plus 9
nestling days), and Henslow’s Sparrow: 20 days (11
incubation days plus 9 nestling days; Ehrlich et al.
1988). Standard errors for daily nest survival rates
were calculated by using the formula for binomial dis-
tributions (Zar 1996). 1 used means and confidence in-
tervals (Johnson 1979) to compare rates of nesting suc-
cess among years and between incubation and nestling
stages in each year. To allow comparison with other
studies that did not use Mayfield estimates, 1 also pre-
sent the apparent proportion of successful nests.

Statistical analyses. — Logistic regression was used
to investigate if nesting success was related to the date
in the breeding season. I calculated mean clutch size
for each week in the breeding season and used linear
regression analysis to investigate if clutch size varied
during the nesting season. Because the number of nests
found varied among weeks, I weighted the mean week-
ly clutch size by its standard error. For this analysis I
used all unparasitized nests of the three years of the
study.

Vegetation characteri.stics of depredated and suc-
cessful Dickcissel and Henslow’s Sparrow nests were
compared with a two-tailed /-test. T’-values were com-
pared to the T’-values obtained from a sequential Bon-
ferroni adjustment (Rice 1989). The same analysis was
used to compare characteristics between Dickcissel
and Henslow’s Sparrow nests. Dickcissel was the only
species with enough parasitized nests to allow for sta-
tistical analysis. For this species, host clutch size, num-
ber of host fledglings, and nest characteristics of par-
asitized and unparasitized nests were compared with
/-tests. Nesting success of unparasitized and parasitized
Dickcissel nests were compared by using means and
confidence intervals (Johnson 1979). Logistic regres-
sion was used to investigate if cowbird parasitism was
related to the date in the breeding season. All data
were analyzed with SAS (SAS 6.03 for PC; SAS In-
stitute, Inc. 1988) and are presented as means and stan-
dard errors; the level of significance was set at 0.05.

RESULTS

Nesting biology. — Henslow’s Sparrows typ-
ically arrived in the study area in early May,
about 1-2 weeks earlier than most Dickcis-
sels, and stopped nesting by the end of July,
also about 1-2 weeks earlier than most Dick-
cissels (Fig. 1). The latest observed initiation
of incubation in Henslow’s Sparrows was 16
July. In contrast to Henslow’s Sparrows, some
Dickcissels were observed carrying nesting
material in early August. Although most Dick-
cissels seem to have completed their nesting
activity by the end of July, some might nest
until the end of August. The peak of Dickcis-
sel nest initiation did not occur until early
June, and they continued to nest throughout
June and early July (Fig. lA). In contrast to
Dickcissels, Henslow’s Sparrows had two
peaks of nest initiation, one in the second and
third week of May, and one in the middle of
June (Fig. IB).

Dickcissels and Henslow’s Sparrows had
almost identical clutch sizes, hatching and
fledging rates, and lengths of nestling stages
(Table 1). Clutch size of Dickcissels and Hen-
slow’s Sparrows tended to decline with date
in the breeding season but not significantly
(Dickcissel: F = 4.8, F = 0.35, df = 9,10, P
= 0.06, slope = —0.05 ± 0.02; Henslow’s
Sparrow: F = 5.0, F = 0.38, df = 8,9, P =
0.06, slope = -0.05 ± 0.02).

Nesting success. — The main cause of nest
failure was nest predation; 86% of all failed
Henslow’s Sparrow nests and 84% of all
failed Dickcissel nests were depredated (Table

518

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

a) Dickcissel

0.20 -

1-7 8-14 15-21 22-28 29-4 5-11 12-18 19-25 26-2 3-9 10-16 17-23

May June July

b) Henslow's Sparrow

FIG. 1. Frequency distribution of clutch initiation dates in (A) Dickcissels and (B) Henslow’s SpaiTOWS in
southwestern Missouri, 1995-1997.

1 ). Nest failure from unknown causes (prob-
ably weather) or nest abandonment were min-
imal (Table 1). None of the nests of either
species failed as a result of cowbird parasit-

ism. Dickcissel nesting success was lower
during incubation than during the nestling
stage in 1996, whereas nesting success during
incubation and nestling stages did not differ

Winter • HENSLOW’S SPARROW AND DICKCISSEL NESTING

519

TABLE 1. General nesting data (.v ± SE) of Dickcissel
prairie fragments, 1995-1997.

and Henslow’s Sparrow

in southwestern Missouri

Dickcissel
(II = 242 f

Henslow’s Sparrow
(n = 59)

General nesting data:

Successful nests {n)

112

Depredated nests (n)

128

Unknown loss (n)

Abandoned nests («)*’

a) during nest building

b) with eggs

c) parasitized

Mowed nests (nY

Nesting success:

Mayfield nesting success (%)‘*

29.7

39.5

Apparent nesting success^

46.3

57.6

Nesting biology:

Clutch size (n)^

3.9 ± 0.05 (227)

3.8

± 0.10 (56)

Incubation days («)«

11.45 ± 0.08 (11)

12.0

± 0 (1)

Nestling days (n)*’

8.7 ± 0.02 (52)

9.1

± 0.08 (9)

Hatching success («)'

92.9 (69)

93.2 (12)

Young fledged/nest

1.7

2.0

Young fledged/successful nest

3.6

3.5

Broad parasitism:

Parasitized nests (%>'

8.8 ± 0.005 (21)

5.3

± 0.006 (3)

Young fledged from successful unparasitized nests

3.7 ± 0.13 (105)

3.6

± 0.25 (33)

Young fledged from successful parasitized nests

2.3 ± 0.26 (6)

2.0

± 0 (1)

^ Total number of nests found excluding those that were abandoned and mowed,

^ Not included in the total number of nests.

Not included in the total number of nests.

After Mayfield (1975).

' Percent of successful nests from all nests found.
fQnly unparasitized nests were used.

8 n = number of nests that could be followed from nest building until hatching.

n = number of nests that could be followed from hatching until fledging.

‘ Percent hatched eggs from all eggs for which the clutch size was known with certainty (see Methods).
J Percent of parasitized nests out of all nests found.

significantly in any other year or for Hen-
slow’s Sparrows (Table 2). Nesting success
did not vary significantly with the date in the
breeding season for either Dickcissels (Wald-
= 0.22, P > 0.05, n — 240) or Henslow’s
Sparrows (Wald-x^ = 2.66, P > 0.05, n = 59).
Dickcissel nesting success was higher in 1997
than in 1996, whereas nesting success of Hen-
slow’s Sparrows did not vary significantly
among years (Table 2). Mayfield nesting suc-
cess tended to be higher in Henslow’s Spar-
rows (40%) than in Dickcissels (30%; Tables
1, 2); however, the 95% confidence intervals
for the estimates of nesting success in these
two species overlapped.

Cowbird parasitism. — The rate of brood
parasitism by Brown-headed Cowbirds was

low, but slightly higher in Dickcissels (9.6%)
than in Henslow’s Sparrows (5.3%; Table 1).
Dickcissel nests were parasitized throughout
the nesting season except for the first and third
week of May and the last week of July (Fig.
2).

Parasitized nests generally had smaller
clutches, fewer fledglings, and lower nesting
success than unparasitized nests (Table 3). On
average, cowbirds laid 1.4 eggs per parasit-
ized Dickcissel nest. None of the three para-
sitized Henslow’s Sparrow nests had more
than one cowbird egg. Host clutch size in both
species was reduced by about 0.9 eggs per
parasitized nest (Table 3). The reduction in
clutch size was significant in Dickcissels {t =
4.07, df = 23, P < 0.001), with fewer host

520

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 2.
1995-1997.

Nesting success of Dickcissel and Henslow’s Sparrow in southwestern Missouri prairie fragments.

Year

Nest cycle
interval

Depredated

(«)•>

Exposure

days'^

Survival ± SE‘*

Success

(%)'

Dickcissel

1995

Incubation

69.0

0.94 ± 0.03

21.6 < 48.8 < 100

Nestling

74.5

0.95 ± 0.03

35.0 < 60.8 < 100

Totab

143.5

0.94 ± 0.02

10.9 < 30.0 < 65.0

1996

Incubation

511.5

0.91 ± 0.01

24.7 < 34.0 < 44.58

Nestling

388.5

0.95 ± 0.01

52.0 < 63.7 < 76.08

Total

113

900.0

0.92 ± 0.01

12.0 < 19.7 < 24.98

1997

Incubation

517.0

0.94 ± 0.01

36.8 < 46.4 < 61.3

Nestling

618.0

0.95 ± 0.01

53.0 < 63.9 < 74.6

Total

117

1235.0

0.96 ± 0.01

32.6 < 39.8 < 55.18

Alb

Incubation

181

1097.5

0.93 ± 0.01

34.0 < 40.3 <51.4

Nestling

171

1081.0

0.95 ± 0.01

46.4 < 63.6 < 70.6

Total

248

128

2278.5

0.94 ± 0.00

21.8 < 29.7 < 34.0

Henslow’s Sparrow

1995

Incubation

34.5

1.0 ± 0.00

100.0

Nestling

59.0

0.97 ± 0.02

52.0 < 73.3 < 100

Total

83.5

0.98 ± 0.02

29.0 < 66.8 < 100

1996

Incubation

1 1

103.5

0.95 ± 0.02

35.4 < 58.0 < 89.5

Nestling

146.0

0.94 ± 0.02

38.7 < 60.2 < 83.4

Total

249.5

0.95 ± 0.01

23.4 < 34.3 < 54.4

1997

Incubation

84.5

0.95 ± 0.02

35.4 < 58.7 < 89.5

Nestling

133.0

0.95 ± 0.02

42.8 < 66.0 <91.0

Total

217.5

0.95 ± 0.01

23.4 < 39.0 < 54.0

All

Incubation

222.5

0.96 ± 0.01

50.6 < 63.5 < 80.1

Nestling

334.0

0.95 ± 0.01

52.0 < 64.3 < 76.0

Total

550.5

0.95 ± 0.01

24.4 < 39.5 < 52.2

“ Total number of nests monitored in a specific nesting interval during incubation and nestling stages. Because nests were mostly monitored during parts
of both nesting stages, the sum of nests in each interval is higher than the total number of nests found.

Total number of depredated nests.

Total number of exposure days (Mayfield 1975).

Probability of daily Mayfield nesting succe.ss (Day) = - (# depredated nest.s/# Mayfield days) + 1 SE = sqrt ((Day » (# depredated nests/# Mayfield
days))/# Mayfield days).

' Probability of nesting succe.ss over the entire interval = Day‘"'"''“' shown are means and lower and upper 95% confidence intervals (Johnson

1979).

f Both nesting stages combined.

? Intervals with the same letter do not overlap.

All years combined.

young fledged from successful parasitized
nests (t = 4.10, df = 34, P < 0.001). The
reduction in nesting success was not caused
by competition with cowbird young, but rather
by a higher predation rate on parasitized nests;
alt successful parasitized nests fledged young
of both host and cowbird.

Nest characteristics. — Compared to Hen-
slow’s Sparrows, Dickcissels chose a variety
of nest sites. Most (45%) nests were placed in
forbs, especially leadplant (Amorpha canes-
cens) and ashy sunflower {Helianthus mollis),
but shrubs (29%), grass (16%), and litter
(10%) also were used as nesting substrates.
Nests were typically woven in the stems of
forbs or woody plants. Because nest searches

were restricted to grassland habitat, nests were
not found within shrubby edge habitats. How-
ever, many Dickcissels were observed breed-
ing in such edge habitats (Winter 1998). Suc-
cessful nests had taller vegetation, greater vi-
sual obstruction values, greater coverage by
grass, and smaller areas of bare soil than un-
successful nests (Table 4). None of the vege-
tative characteristics that 1 measured at Dick-
cissel nest sites differed between parasitized
and unparasitized nests (Table 4).

Henslow’s Sparrows typically placed their
nests among layers of thick litter (82% of all
nests). Compared to Dickcissel nests, the veg-
etation surrounding Henslow’s Sparrow nests
had deeper litter (3.5 ± 0.27 vs 1.9 ± 0.23

Winler • HENSLOW’S SPARROW AND DICKCISSEL NESTING

521

May June July

FIG. 2. Frequency distribution of rates of cowbird parasitism on Dickcissel nests in southwestern Missouri,
1995-1997.

cm; t = 3.3, df = 296, P = 0.001), lower
vegetation (42.0 ± 1.12 vs 46.1 ± 0.67 cm; t
= —2.8, df = 296, P = 0.005), greater cover
by litter (29.6 ± 2.0 vs 11.6 ± 0.71%; t =
10.4, df = 296, P < 0.001), and less cover by
forbs (17.9 ± 1.03 vs 26.0 ± 0.83%; t =
—4.6, df = 296, P < 0.001), woody plants
(1.5 ± 0.12 vs 5.2 ± 0.62%; t = -2.9, df =
296, P = 0.004), and soil (0.4 ± 0.12 vs 4.8
± 0.40%; t = -5.5, df = 296, P < 0.001).
Henslow’s Sparrow nests also had a higher
percentage of nest cover (90.3 ± 2.68 vs 67.7

± 1.79%; t = 5.9, df = 296, P < 0.001), and
were located closer to the ground (7.2 ± 0.53
vs 18.4 ± 1.04 cm; t = —5.4, df = 134, P <
0.001). All significant P-values remained sig-
nificant after a sequential Bonferroni adjust-
ment. Henslow’s Sparrows were never ob-
served to place their nest within or in imme-
diate proximity to woody vegetation. In con-
trast to Dickcissels, Henslow’s Sparrows did
not weave their nests into the surrounding
vegetation, but placed them loosely among the
surrounding stems of grass and dead vegeta-

TABLE 3. Clutch size (.v ± SE) and nesting success of unparasitized and parasitized nests in Dickcissel and
Henslow’s Sparrow in southwestern Missouri, 1995-1997.

Unparasitized ne.sts

Parasitized nests

Year

Clutch size

Exposure

days

Failed

nests

Success”

(%)

Host clutch

Cowbird clutch

Exposure

days

Failed

nests

Succe.ss

(%)

Dickcissel

1995

3.9

-+*

0.13

143.5

35.0

3.0 ± 0

2.0 ± 0

12.0

16.1

1996

104

3.9

0.07

931.0

23.5

3.4 ± 0.40

2.0 ± 0.55

26.0

1.1

1997

3.9

-+-

0.07

1014.5

39.4

3.0 ± 0.26

1.2 ± 0.14

152.0

27.7

All

220

3.9

H-

0.05

2089.0

1 13

31.2

3.0 ± 0.22

1.4 ± 0.18

190.0

17.8

Henslow’s

Sparrow

1995

4.3

-h

0.21

79.0

59.9

3.0 ± 0

1.0 ± 0

15.0

100

1996

3.7

-h

0.17

138.0

13.8

N/A

1997

3.7

-h

0.15

201.5

44.5

3.0 ± 0

1.0 ± 0

16.5

7.5

All

3.8

0.10

418.5

32.3

3.0 ± 0

1.0 ± 0

31.5

26.9

“Probability that a nest survived both incubation and nesting periods, estimated after Mayfield (1975).

522

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Urn

— m — ootNoo

ddddddddddd

C/D

C/5

fNr^l fN(N(N(N (N(N

c/5

00 — re) oOinsOOOCN— 'fN

-C

d — ddd — -^-^d-^d

1 III

c/5

■y.

c/5

C/5

— vori^ —

•eb — p On ■'t ^ Os (^j rn VD

^ac

doi4d— 'oireirN — dcN

7. II

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

cz ^

p p p 00 (N ^ lo p

Q. ^

— (ddddoOONU'rtrritC

CS — -eb (N —

•o

i/^r^lONt^rNONOsrriTtt^m

— t^NOOt^OOOONO-'tON —

•o

.3 ^

ddddddddd —

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

b 11

00 00 00 p p p oj p b' (N (p

— \or~-d — ^re)d'rt4Ndod

ri- rN — (^1 no —

o F

~o S

— 00 Oi ON

C X

— OOOO^OON — Oro —

TZ D

— OO — ooO'rt-u-iOrNp

X ^

a,.

ddddddddddd

w a

V V

V -p

= o

? fc

r3 ''P

re)0N>/-)rr)ini/^>/^O’rrt''00

— OOnO — — — — 'P — On

r-i n — M ri ri oi (N — (N

^ CQ

1) a

"O 3

e3 f3 n rt a

C '**-

— ^rnr-i— -dreiddr-i — d

3 c

^ C3

■yi

•j'.

III 1 II

X (J

(U ^

H 'c

3 .SJJ
X 'x

ONce, re, — onOnOnonO^On

— .O — — ^OnO — ^ONNDi/^'e)-

c 73

•— u

o — — dd— ■— -ddrN —

^ E

3 O
73 —

+1 +1 +1 +1 +1 4-1 4-1 +\ 4-1 4-1 4-1
>/',^0'ND'e)-r'' — — poop

Q —

— ^et->6d— •dUiEiri4U

^ OJ

X 3

Tj- ri — in (N no —

X ^

m m o 00 oi no NO 00 m On O'

T a.

oi 00 00 O On rN| m p p p

d d d d d — — d d ONi —

O c
tZ rt

y, —

4-1 4-1 4-1 4-1 -H -H 4-1 -H 4-1 4-1 4-1

u 3
i:

U r-

a II

3 -
X 3.

ooNNOitoiopppoip
oiodoid— ■NOin'rfmONOo

Urn

OXJ
X • —

-rf ri — m O'! NO —

"S <!

y— s

!-■

E S

d 3

n 1

^ r:

1— 1 iTi

. O'
^ —

UJ X
J c
CC y

Variable

3'Ci

r«px

r“ U- <— *

•< E

Litt

Veg

Visi

No.

Litt

Gra

For

Soi

Nc!

Ne:

•I)

tion. In late June and throughout July, 18% of
all nests were found in areas that had been
burned the same spring and therefore lacked
any litter. In these areas, Henslow’s Sparrows
placed their nests within large clumps of grass
(mostly big bluestem and Indian grass) close
to the ground. Successful and depredated Hen-
slow’s Sparrow nests did not differ in any nest
characteristic (Table 5).

DISCUSSION

In southwestern Missouri, Dickcissels and
Henslow’s Sparrows had nearly identical
clutch size, hatching success, length of incu-
bation and nestling stages, and number of
young fledged per successful nest. These var-
iables were similar to previous reports for
Dickcissels (Bent 1968, Harmeson 1974, Zim-
merman 1982, Fretwell 1986, Patterson and
Best 1996). Fewer studies have monitored
Henslow’s Sparrow nests, because their nests
are difficult to locate (Bent 1968; Robins
1971; Schulenberg et al. 1994; D. Reinking,
pers. comm.). Clutch size of Dickcissels and
Henslow’s Sparrows tended to decrease with
date in the breeding season in southwestern
Missouri. For Dickcissels, Harmeson (1974)
described a peak in clutch size in the middle
of the nesting season, whereas changes of
Henslow’s Sparrow clutch size over time had
not yet been described.

Although Dickcissels and Henslow’s Spar-
rows had almost identical nesting variables,
their nesting phenologies seemed to differ.
Generally, nesting success is relatively low in
most grassland nesting birds, varying from
25-50% (Wiens 1969, Vickery et al. 1992,
Martin 1995). Grassland birds often compen-
sate for low nesting success by several re-
nesting attempts throughout the breeding sea-
son. Consequently, most grassland nesting
species raise an average of 1.5-2 broods per
female per year (Wiens 1969, Martin 1995).
The single nesting peak of Dickcissels in my
study seemed to indicate that Dickcissels
raised only one brood in my study area, as has
been described by Zimmerman (1982, 1984).
However, E. Bollinger (pers. comm.) observed
a second brood in one color-banded Dickcissel
female, indicating that Dickcissels can be dou-
ble-brooded. Because Dickcissels appear to
frequently move within one breeding season
(Fretwell 1986), possibly because of displace-

Winler • HENSLOW’S SPARROW AND DICKCISSEL NESTING

523

TABLE 5. Henslow’s Sparrow nest characteristics (.f ± SE) at
western Missouri prairie fragments, 1993-1997.

successful and depredated ne.sts

in south-

Variable

Successful
(n = 35)

Depredated
(« = 25)

Litter depth (cm)

3.3 ± 0.37

3.7 ± 0.41

0.72

OAl

Vegetation height (cm)

43.0 ± 1.48

40.7 ±1.71

- 1 .00

0.32

Visual obstruction (dm)

25.4 ± 1.89

24.9 ± 2.09

-0.20

57.7

0.84

No. woody stems

0.43 ± 0.14

0.26 ±0.11

-0.94

57.7

0.35

Litter cover (%)

27.1 ± 2.70

33.1 ± 2.85

1.50

0.14

Grass cover (%)

51.4 ± 2.50

49.4 ± 2.43

-0.57

0.57

Eorb cover (%)

19.1 ± 1.35

16.2 ± 1.58

-1.42

0.16

Woody cover (%)

1.82 ± 0.68

1.14 ± 0.45

-0.98

55.2

0.40

Soil cover (%)

0.55 ± 0.12

0.22 ±0.11

-1.47

50.4

0.15

Nest cover (%)

89.6 ± 4.02

91.7 ± 3.03

0.41

57.7

0.68

Nest height (cm)

7.0 ± 0.63

7.9 ± 1.02

0.74

26.8

0.47

ment from hayfields after mowing (Igl 1991,
Frawley and Best 1991), females might renest
or raise a second brood in another area. None
of the Dickcissels in my study area were col-
or-banded; therefore, I could not determine if
late nesting females (as also described by Har-
meson 1974) had arrived from other areas, or
if they had started a second brood or a re-
nesting attempt in the same area.

Henslow’s Sparrows seemed to be more
likely to be double-brooded in southwestern
Missouri than Dickcissels because they clearly
exhibited two peaks of nest initiation. How-
ever, as with Dickcissels, individual birds
were not color-banded, making it impossible
to determine if the second nesting peak was
caused by females on their second brood, by
renesting attempts, by newly arriving females,
or if it was an artifact of small sample size.
This lack of adequate information is also true
for all other studies that describe this species
as double-brooded (Hyde 1939, Bent 1968,
Robins 1971).

Nest predation was the main reason for nest
failure, as has been described for many other
bird species (Martin 1993, Patterson and Best
1996). Mean Mayfield nesting success of
Dickcissels was similar to that reported from
Kansas (Zimmerman 1984) and Missouri
Crop Reserve Program fields (McCoy 1996),
but lower rates of nesting success were re-
ported from Iowa (Bryan and Best 1994, Pat-
terson and Best 1996), Kansas (Hill 1976),
and Oklahoma (Rohrbaugh et al. in press).
Robins (1971) reported that 6 of 1 1 Henslow’s
Sparrow nests found in Michigan successfully
fledged young. This apparent success rate

(54.5%) is comparable to the apparent success
rate in my study (57.6%). However, the num-
ber of young fledged per nest in Michigan
(0.37) and the number of young fledged per
successful nest (2.8) were lower than in Mis-
souri. In Oklahoma, 40.9% of 22 Henslow’s
Sparrow nests were successful (D. Reinking,
pers. comm.), which was about 17% lower
than in Missouri. The number of young
fledged per unparasitized Henslow’s Sparrow
nest was also slightly lower in Oklahoma than
in Missouri (3.3 vs 3.6 young fledged per nest;
Reinking, pers. comm.). Southwestern Mis-
souri thus seems to be a relatively productive
breeding area for Henslow’s Sparrows.

Daily Mayfield nesting success in Dickcis-
sels was lower during incubation than during
nestling stages in 1996, and tended to be low-
er in 1995 and 1997. Higher nesting success
during the nestling stage was also reported by
Bryan and Best (1994) and by Harmeson
(1974), and generally is the most frequently
observed pattern of nest survival (Nice 1957;
but see Patterson and Best 1996). Nesting suc-
cess could be lower during incubation because
poorly concealed nests are the first to be found
by nest predators, or because visually hunting
nest predators find nests with eggs more eas-
ily. Shorter and sparser vegetation at depre-
dated Dickcissel nests indicated that these
nests were in fact less well concealed than
successful nests. High incidence of nest pre-
dation by mammals (see Winter 1998), which
hunt based on visual and olfactory cues, might
explain the tendency for slightly lower nesting
success during incubation in southwestern
Missouri. However, the only nest predators

524

THE WILSON BULLETIN • Vo/. Ill, No. 4. December 1999

that I observed at Dickcissel nests were two
snakes, one eastern yellowbellied racer {Col-
uber constrictor flaviventris) and one prairie
kingsnake {Lximpropeltis calligaster calligas-
ter). Because rates of nesting success in Hen-
slow’s Sparrows were nearly identical during
incubation and nestling stages, and because
their nests were extremely well concealed, it
seems that visually hunting nest predators
rarely destroy its nests. Instead of visually
hunting predators, snakes are possibly the
main nest predators of Henslow’s Sparrow
nests. This could also be the reason why nest
vegetation did not differ between successful
and depredated nests.

Rates of brood parasitism by Brown-headed
Cowbirds in southwestern Missouri were rel-
atively low compared to parasitism rates de-
scribed in other studies on grassland-nesting
birds (Hergenrader 1962; Zimmerman 1966,
1983; Hill 1976; Elliott 1978; Patterson and
Best 1996; Koford et al. in press). Because
cowbirds did not cause direct mortality to any
Dickcissel young in my study, brood parasit-
ism by itself did not directly decrease nesting
success. However, the reduction in clutch size
decreased the number of host fledglings by
about one young per parasitized nest, as also
was reported by Hill (1976).

In Henslow’s Sparrows, brood parasitism
by Brown-headed Cowbirds was slightly low-
er than in Dickcissels; only 5% of all Hen-
slow’s Sparrow nests were parasitized in
southwestern Missouri. The only other records
of parasitized Henslow’s Sparrow nests are
from Oklahoma (Reinking, pers. comm.) and
Kansas (Schulenberg et al. 1994). In
Oklahoma, 2 out of 22 Henslow’s Sparrow
nests were parasitized. Only 1 of the parasit-
ized nests successfully fledged both host and
cowbird young, whereas the other nest was
depredated. The one Henslow’s Sparrow nest
that was found by Schulenberg and coauthors
(1994) in Kansas contained two cowbird eggs
and was abandoned during incubation. Cow-
bird parasitism is probably low in Henslow’s
Sparrows because their nests are well con-
cealed. Low parasitism rates in Henslow’s
Sparrows were previously noted by Bent
(1968:786), who mentioned that this species
“appears to escape heavy parasitism, possibly
because the nests are so well hidden.’’

Nest placement differed significantly be-

tween Dickcissels and Henslow’s Sparrows.
Dickcissels chose a variety of nesting habitats
(Bent 1968); they preferred forbs and shrubs,
and did not avoid edge habitats. Henslow’s
Sparrow nests, on the other hand, were never
found in either of the nest substrates preferred
by Dickcissels or within shrubby edge habitat.
Instead, this species built its nest lower to the
ground, mainly within large clumps of litter
where it was almost 100% covered by vege-
tation (Hyde 1939, Robin 1971, Schulenberg
et al. 1994). Several researchers that described
the relationship between Henslow’s Sparrow
breeding densities and vegetation parameters
also noted the species’ preference for tall
grass and litter cover (Wiens 1969, Skinner et
al. 1984, Herkert 1994, Swanson 1996). 1
found that Henslow’s Sparrows were able to
build nests in recently burned areas that
lacked litter, as did Zimmerman (pers. comm,
in Schulenberg et al. 1994) in Kansas. Be-
cause Dickcissel nests were more conspicuous
than Henslow’s Sparrow nests, they were
probably more easily detected by visually
hunting nest predators and Brown-headed
Cowbirds, resulting in slightly higher rates of
nest predation and nest parasitism in this spe-
cies.

General nesting data indicated that Dickcis-
sels tended to be less productive in south-
western Missouri than Henslow’s Sparrows.
These findings are in contrast to the general
notion that Dickcissels are of little conserva-
tion concern on the breeding grounds (Herkert
et al. 1993, Swanson 1996). The discrepancy
may be because previous reports on Dickcis-
sels and Henslow’s Sparrows were based only
on breeding density estimates. This study
showed that basic data on the general nesting
ecology of a species are necessary for a better
understanding of the factors that might influ-
ence a species in a given area.

ACKNOWLEDGMENTS

I thank J. Faaborg for hi.s support throughout my
study, and J. Faaborg, E. Fink, W. Hochachkil, L. Igl,
D. Johnson, R. Koford, M. Ryan, and R Vickery for
many comments that greatly improved the manuscript.
I also thank my field assi.stants J. Bernier, M. Harry,
P. Kohn, S. Panken, D. Rutka, and K. Warren. This
research was supported by the U.S. Geological Survey
(Northern Prairie Wildlife Research Center), the Mis-
souri Department of Conservation, the U.S. Fish and

Winter • HENSLOW’S SPARROW AND DICKCISSEL NESTING

525

Wildlife Service, the Missouri Prairie Foundation, and

Sigma Xi.

LITERATURE CITED

Basili, G. D. 1997. Continental-scale ecology and con-
servation of Dickcissels. Ph.D. diss., Univ. of
Wisconsin, Madison.

Basili, G. and S. A. Temple. 1995. A perilous migra-
tion. Nat. Hist. 104:40-46.

Bent, A. C. 1968. Life histories of North American
cardinals, grosbeaks, buntings, towhees, finches,
sparrows, and allies. Bull. U.S. Natl. Mus. 237:1-
1889.

Bryan, G. G. and L. B. Best. 1994. Avian nest den-
sity and success in grassed waterways in Iowa
rowcrop fields. Wildl. Soc. Bull. 22:583-592.

Butcher, G. and Lowe. 1990. Population trends of
twenty species of migratory birds as revealed by
Christmas Bird Counts, 1963-87. Final report.
Cooperative Agreement No. 14-16-0009-88-941,
U.S. Fish and Wildl. Serv. Office of Migratory
Bird Manage., Washington, D.C.

Daubenmire, R. 1959. A canopy-coverage method of
vegetational analysis. Northwest Sci. 33:43—64.

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988.
The birder’s handbook. Simon and Schuster Inc.,
New York.

Elliott, P. E 1978. Cowbird parasitism in the Kansas
tallgrass prairie. Auk 95:161-167.

Frawley, B. J. and L. B. Best. 1991. Effects of mow-
ing on breeding bird abundance and species com-
position in alfalfa fields. Wildl. Soc. Bull. 19:135-
142.

Fretwell, S. 1986. Distribution and abundance of the
Dickcissel. Current Ornithol. 4:211-242.

Harmeson, J. P. 1974. Breeding ecology of the Dick-
cissel. Auk 91:348-359.

Hergenrader, G. L. 1962. The incidence of nest par-
asitism by the Brown-headed Cowbird (Molothrus
ater) on roadside nesting birds in Nebraska. Auk
79:85-88.

Herkert, j. R. 1994. Status and habitat selection of
the Henslow’s Sparrow in Illinois. Wilson Bull.
106:35-45.

Herkert, J. R. 1997. Population trends of the Hen-
slow’s Sparrow in relation to the Con.servation Re-
serve Program in Illinois, 1975-1995. J. Field Or-
nithol. 68:235-244.

Herkert, J. R., R. S. Szafone, V. M. Kleen, and J.
E. ScHWEGMAN. 1993. Habitat establishment, en-
hancement, and management for forest and grass-
land birds in Illinois. Division of Natural Heritage,
Illinois Department of Conservation, Springfield.

Hill, R. A. 1976. Host-parasite relation.ships of the
Brown-headed Cowbird in a prairie habitat of
west-central Kansas. Wilson Bull. 33:555-565.

Hyde, A. S. 1939. The life history of Henslow’s Spar-
row, Passerherbulus henslowii (Audubon). Univ.
Mich. Misc. Publ. 41:1-72.

Igl, L. D. 1991. The role of climate and mowing on
Dickcissel (Spiza americana) movements, distri-

bution, and abundance. M..Sc. thesis, Iowa State
Univ., Ames.

Johnson, D. H. 1979. Estimating nest success: the
Mayfield method and an alternative. Auk 96:651-
661.

Koford R. R., B. S. Bowen, J. T. Lokemoen, and A.
D. Kruse. In press. Cowbird parasitism in grass-
land and cropland in North Dakota. In The ecol-
ogy and management of cowbirds (J. N. M. Smith,

T. Cook, S. K. Robinson, S. I. Rothstein, and S.
G. Sealy, Eds.). University of Texas Press, Austin.

Martin, T. E. 1993. Nest predation among vegetation
layers and habitat types: revising the dogmas. Am.
Nat. 141:897-913.

Martin, T. E. 1995. Avian life history evolution in
relation to nest sites, nest predation, and food.
Ecol. Monogr. 65:101-127.

Mayfield, H. F. 1975. Suggestions for calculating nest
success. Wilson Bull. 87:456—466.

Mazur, R. 1996. Implications of field management for
Henslow’s Sparrow habitat at Saratoga National
Historical Park, New York. M.Sc. thesis. State
Univ. of New York, Syracuse.

McCoy, T. D. 1996. Avian abundance, composition,
and reproductive success on Conservation Re-
serve Program fields in northern Missouri. M.Sc.
thesis, Univ. of Missouri, Columbia.

Nice, M. M. 1957. Nest success in altricial birds. Auk
74:305-321.

Patterson, M. P. and L. B. Best. 1996. Bird abun-
dance and nest success in Iowa CRP fields: the
importance of vegetation structure and composi-
tion. Am. Midi. Nat. 135:153-167.

Peterjohn, B. G., j. R. Sauer, and W. A. Link. 1994.
The 1992 and 1993 summary of the North Amer-
ican breeding bird survey. Bird Populations 2:46-
61.

Pruitt, L. 1996. Henslow's Sparrow status assessment.

U. S. Fish and Wildlife Service, Minneapolis. Min-
nesota.

Rice, W. R. 1989. Analyzing tables of statistical tests.
Evolution 43:223—225.

Robel, R. j., j. N. Briggs, A. D. Dayton, and L. C.
Hulbert. 1970. Relationship between visual ob-
struction measurements and weight of grassland
vegetation. J. Range Manage. 23:295-297.

Robins, J. D. 1971. A study on Henslow's SpaiTows
in Michigan. Wilson Bull. 83:39-48.

Rohrbaugh, R. W., D. L. Reinking, D. H. Wolfe, S.
K. Sherrod, and M. A. Jenkins. In pre.ss. Effects
of prescribed burning and grazing on nesting and
reproductive success of three grassland passerine
species in tallgrass prairie. Stud. Avian Biol.

SAS Institute, Inc. 1988. SAS/STAT user's guide, re-
lease 6.03 ed. SAS Inst., Cary, North Carolina.

ScHULENBERG, J. H., G. L. Horak, M. D. Schwilling,
and E. j. Finck. 1994. Nesting of Henslow's Spar-
rows in Osage County, Kansas. Kans. Ornith. Soc.
Bull. 44:1-4.

Skinner, R. M., T. M. Baskett, and M. D. Blenden.

526

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

1984. Bird habitat on Missouri prairies. Missouri
Department of Conservation, Jefferson City.
Swanson, D. A. 1996. Nesting ecology and nesting
habitat requirements of Ohio's grassland-nesting
birds; a literature review. Ohio Fish Wildl. Serv.
Rep. 13:1-60.

Vickery, P. D., M. L. Hunter, and J. V. Wells. 1992.
Evidence of incidental nest predation and its effect
on nests of threatened grassland birds. Oikos 63;
281-288.

Wiens, J. A. 1969. An approach to the study of eco-
logical relationships among grassland birds. Or-
nithol. Monogr. 8; 1-93.

Winter, M. 1998. Effects of habitat fragmentation on
grassland-nesting birds in southwestern Missouri.
Ph.D. diss., Univ. of Missouri, Columbia.
Winter, M. and J. Faaborg. In press. Patterns of area-

sensitivity in grassland-nesting birds. Conserv.
Biol.

Zar, j. H. 1996. Biostatistical analysis, third ed. Pren-
tice-Hall, Englewood Cliffs, New Jersey.

Zimmerman, J. L. 1966. Polygyny in the Dickcissel.
Auk 83:534-546.

Zimmerman, J. L. 1982. Nesting success of Dickcissels
{Spiza americana) in preferred and less preferred
habitats. Auk 99:292-298.

Zimmerman, J. L. 1983. Cowbird parasitism of Dick-
cissels in different habitats and at different nest
densities. Wilson Bull. 95:7—22.

Zimmerman, J. L. 1984. Nest predation and its rela-
tionship to habitat and nest density in Dickcissels.
Condor 86:68-72.

Zimmerman, J. L. 1988. Breeding season habitat se-
lection by the Henslow’s Sparrow (Ammodramus
henslowii) in Kansas. Wilson Bull. 100:17—24.

Wilson Bull., 1 1 1(4), 1999, pp. 527-535

RESPONSE OF A BIRD ASSEMBLAGE IN SEMIARID CHILE TO

THE 1997-1998 EL NINO

FABIAN M. JAKSIC 2 AND IVAN LAZO'

ABSTRACT. — The semiarid region of Chile is influenced by El Nino Southern Oscillation. Its absence causes
droughts and its presence causes wet years, which in turn result in variations in resource levels for avian
assemblages. We show that bird species richness and density follow some of these pulses closely. Sixty-one bird
species, 32 of which were Passeriformes, were sighted during five years in Las Chinchillas National Reserve
(300 km N of Santiago). Overall, 30 species (49%) were residents and 31 (51%) were migratory. The most
speciose trophic groups were insectivores (34%), carnivores (28%), and granivores (25%). Bird species richness
and density declined from 43 species and 45-50 individuals/ha in spring 1993, to 29 species and 15-20 indi-
viduals/ha in autumn 1996. Increases were observed with the onset of El Nino, reaching totals of 42 species (a
45% increase from 29) and densities of 55—60 bird.s/ha in summer 1997. Similar trends were observed in one
of two major food resources measured: small mammals. Positive correlations were found between raptor species
richness and density and small mammal density, but not between insectivorous bird species richness or density
and terrestrial arthropod abundance. Because the climate was very dry during most of the time of our study, we
may have witnessed the lowest boundary for species richness and bird density. Whether the 1997-1998 El Nino
brought the maximum bird species richness and density for the site is yet to be seen. Received 20 Jan. 1999,
accepted 2 June 1999.

Bird assemblages vary through time in spe-
cies composition and absolute and relative
abundances, in the short term (seasonal, Av-
ery and van Riper 1989), medium term (be-
tween years, Wiens 1990a, b), or both. Most
authors agree that this variation reflects
changes in the resource base, mainly food
(Feinsinger et al. 1985, Wiens 1993). Three
factors are generally proposed to account for
how birds use resources (Pearson 1991): com-
petition (Martin 1987, Pulliam and Dunning
1987), predation (Lima 1987), and physical or
abiotic stresses (Karr and Freemark 1983).
The degree to which these factors determine
bird assemblages is controversial (Loiselle
and Blake 1991). Although they are not mu-
tually exclusive, these three factors are usually
assessed individually, rarely two simulta-
neously (Martin 1985, Kotler and Holt 1989).
Most attention has concentrated on the role of
species interactions rather than on physical
factors (Pearson 1991), although arguably the
latter set the stage for biotic interactions.

The periodic intrusion of El Nino Southern
Oscillation along the western coasts of the
Americas constitutes a major physical distur-
bance that brings warm water to the shores

‘ Dept, de Ecologi'a, Pontificia Univ. Catolica de
Chile, Casilla 1 14-D, Santiago, Chile;

E-mail: fjak.sic @ genes. bio. puc. cl
^ Corresponding author.

and increased precipitation to the adjacent
land masses. Although initially studied in its
oceanographic and climatic aspects, increased
awareness of the multiple effects of El Nino
is shifting the focus to the effects of this phe-
nomenon on birds (Barber and Chavez 1983,
Schreiber and Schreiber 1984, Gibbs and
Grant 1987, Grant and Grant 1987, Hall et al.
1988, Miskelly 1990, Massey et al. 1992,
Lindsey et al. 1997). Previous studies dealt
with the putative El Nino effects on seabird
colonies or on terrestrial island birds. Effects
on inland birds have been little studied.

The semiarid areas of northcentral Chile
(27-32° S), apart from seasonal fluctuations in
weather and food resources, are characterized
by medium term fluctuations in rainfall (Pu-
entes et al. 1988). Accordingly, plant cover,
amount of herbage production, and size of
seed bank vary markedly among years (Gu-
tierrez et al. 1993). Small mammals track
these food resources closely (Jimenez et al.
1992) and perhaps arthropods do also (Puen-
tes and Campusano 1985). It has become in-
creasingly clear that the unusual rainfall
brought by El Nino events to semiarid Chile
are responsible for increased primary produc-
tivity, which in turn leads to population out-
breaks of small mammals and to local increas-
es in the populations of carnivorous birds that
prey on them (Meserve et al. 1995; Jaksic et
al. 1996, 1997).

527

528

THE WILSON BULLETIN • Vol. Ill, No. 4. December 1999

Interestingly, no multi-year studies have
been conducted on the response of an entire
bird assemblage in such semiarid areas to
changes in precipitation, or in primary (veg-
etation) or secondary production (mammals,
arthropods). Here we report upon the effect of
variation in such secondary production on
bird species richness and density.

STUDY AREA AND METHODS

Study area. — The study site is located inside Las
Chinchillas National Reserve (simply called the re-
serve hereafter), at 31° 30' S, 71° 06' W, about 300 km
N of Santiago, Chile, and 60 km E of the Pacific coast.
The reserve spans 400-1700 m elevation, has a rugged
topography, and is dominated physiographically by an
alternation of ridges and ravines, with few flat areas
between. The climate is classified as semiarid by di
Castri and Hajek (1976), with sporadic precipitation
concentrated during Austral winter months (June
through August). Mean annual rainfall 1986-1997 was
170 mm, but with marked increases in 1987 (513 mm),
1992 (307 mm), and 1997 (367 mm) associated with
the respective El Nino events ( 1986-1987, 1991-1992,
1997-1998). Two dominant landscape types character-
ize the study site; ravines and slopes. Ravines are more
vegetated (70.3% shrub cover) than slopes (52.7%
cover).

Vegetation. — The vegetation is a thornscrub com-
posed mainly of spiny dicots, bromeliads, and cacti
(details in Gajardo 1978). Dominant species in ravines
are Stevia sp. (15.5% cover), Colliguaya odorifera
(12.5%), Pleocarphu.s revolutu.s (6.6%), Bacchari.s
paniculata (5.4%), Proustia cimeifolia (5.4%), and
Maytenus hoaria (5.2%). Dominant species in slopes
are Bahia amhrosioide.s (16.9%), Proustia cuneifolia
(19.3%), Baccharis paniculata (8.6%), Porlieria chi-
lensis (8.0%), Lobelia polyphylla (4.0%), and Puya
berteroniana (3.0%).

Census techniques. — During the four calendar sea-
sons of every year from July (winter) 1993 to January
1998 (summer), we conducted fixed band transects to
census birds (Burnham et al. 1980, Conner and Dick-
.son 1980, Bibby et al. 1993). One transect was in a
ravine (1500 X 20 m = 3 ha. time spent = 45-55 min/
transect) and another in a southern exposure mid-slope
(500 X 40 m = 2 ha. 20-30 min/transect) of each El
Cobre and El Grillo creek beds (sampling effort = 4
transects/season, the two creek beds combined). The
two ravine tran.sects were surveyed throughout the
study period, but the two on slopes were terminated in
the summer of 1995. The two creek beds were dry
except for the wet winter of 1997, but are hereafter
called creeks nonetheless. Transects were started 1 h
after the sun's rising above the top of ridges east of
the two creeks. Diurnal raptors (Falconiformes) were
censused opportunistically in a 2000 ha area centered
around the two creeks. The abundance of the four noc-
turnal raptor species (Strigiformes) at the site was as-
sessed based on responses to playbacks of their calls

(Johnson et al. 1981, Haug and Didiuk 1993) 0.5 h
after the first star was spotted in the sky. Every play-
back was broadcast for 1 min for each species sequen-
tially, at four fixed stations located in the ravine of the
El Cobre creek, at two stations in the ravine of El
Grillo creek, and at two stations in the bottom of the
cliffs that border the Auco stream. For the Strigiformes,
we considered the area sampled to be about 2000 ha,
and all estimates of abundance refer to the minimum
number of individuals detected. All densities are stan-
dardized as the number of individuals per species per
hectare.

Bird categorizations. — Bird nomenclature follows
Meyer de Schauensee (1982). We categorized species
according to their reproductive status, diet, residence
status, and habitat. Reproduction: birds were classified
as either nesting or non-nesting at the site, depending
on whether they were observed actually nesting. Some
cryptic nesters may have escaped our detection and
thus the number of nesting species may be underesti-
mated. Diet; we followed Jaksic and Feinsinger (1991)
in establi.shing the following primary diets: carnivores,
insectivores, granivores, nectarivores, frugivores, foli-
vores, and omnivores. These categorizations come
from the literature or from direct observation. Resi-
dence: we considered a species to be resident at the
study site if it was observed on 13 (ca 70%) of the 19
visits to the site. A species was categorized as migrant,
either if it stayed at the site only during spring and
summer (summer visitor) or if it stayed only during
autumn and winter (winter visitor). Habitat; the follow-
ing landscape units were considered as “habitats”; ra-
vines (bottom of creeks), slopes (sides of creeks), flat
areas, and water edges (around permanent streams).

Food availability. — -Small mammal density was as-
sessed during five days/four nights trapping bouts in
each season on opposite slopes of El Cobre creek.
Each of the two grids consisted of a 7 X 7 arrangement
of stations at 15 m intervals equipped with one Sher-
man live trap at each station. Total trapping area was
2.2 ha. Jimenez and coworkers ( 1992) provided details
of this trapping scheme. Teirestrial arthropods were
sampled with Barber pitfall traps (Southwood 1978)
consisting of a 200 ml plastic vial with its rim at
ground level, inside of which a 100 ml plastic vial was
tightly fitted and filled with water and biodegradable
detergent. One hundred such traps were placed at El
Cobre creek along four tran.sects with stations at 10 m
intervals. Two transects were on opposite slopes and
had 20 traps each; two transects (with 30 traps per
transect) were in the ravine. Specimens were collected
at 24 h intervals during two consecutive days and their
total abundance expressed as individuals per trap per
24 h. This total abundance was weighted by the pro-
portion of traps placed in each habitat type (2:3,
slopes : ravine). Unlike the case for mammals, this
method provides information only on relative densities
of large terrestrial arthropods through time.

Statistical analyses. — To determine whether there
were as.sociations between habitat and bird densities
through time, we compared data obtained in the two

Jaksic ami Lazo • BIRD RESPONSES TO EL NINO IN CHILE

529

slopes with those from the two ravines because these
were the two physiognomically most salient landscape
features in the reserve. The degree of association be-
tween bird densities and two food resources (mammals
and arthropods) was determined separately for carniv-
orous (Falconiformes and Strigiformes) and insectiv-
orous birds by Spearman correlation coefficients (So-
kal and Rohlf 1981). Because raptor sightings could
not be assigned unequivocally to either slopes or ra-
vines (as they apparently hunted in both habitat types),
these data were pooled. In the case of insectivorous
birds, we could assign sightings unequivocally to one
or the other of the two habitats, and thus we were able
to make separate comparisons with arthropod abun-
dance in either slopes or ravines. Bird and arthropod
densities were compared by means of Wilcoxon
matched-pairs tests (Sokal and Rohlf 1981). The re-
sults from Barber traps may be considered only as ap-
proximations of terrestrial arthropod density. We did
not sample aerial or foliage arthropods. All statistical
analyses were performed with software Systat version
7.0 for Windows 95.

RESULTS

Characterization of the bird assemblage. —
We sighted 61 species during our 5 year study
(Table 1), 32 (53%) of which were Passeri-
formes. Thirty species (49%) were residents
(but only 13 species were sighted in all 19
visits). Transient species were equally divided
between winter visitors (10%) and summer
visitors (10%). Another 31% of the species
were sighted too few times to enable us to
categorize them, save as accidentals or occa-
sional. Two species were not observed but
known to be present. The White-tailed Kite
(Elanus leucurus) had been sighted previously
(Jaksic et al. 1996) and thus we categorized it
as a migrant. We considered the Great Homed
Owl Bubo virginianus to be resident, despite
having detected it during only five visits be-
cause we collected freshly regurgitated pellets
at each visit (Jaksic et al. 1996).

In terms of species numbers, the best rep-
resented trophic groups were insectivores
(34%), carnivores (28%), and granivores
(25%), accounting for 87% (53 species) of the
local assemblage (Table 1). Thirty-eight
(62%) of the 61 species nested in the reserve,
including three summer visitors: Aplomado
Falcon (Falco femoralis). Giant Hummingbird
{Patagona gigas), and White-tufted Tyrant
{Elaenia albiceps; Table 1). Seventy-seven
percent of the species were sighted in ravines,
54% on slopes, 11% on flat areas, and 7%
near small streams and ponds. (These per-

centages add to more than 100% because
some species visit more than one habitat type;
Table 1.) Bird densities were not different, ei-
ther between the two ravines (Wilcoxon
matched-pairs test; Z = 0.283; n = 19, P >
0.05) or between the two slopes (Z = 0.891;
n = 11; P > 0.05). Thus, we felt justified to
analyze our data from the two creeks by hab-
itat type only.

Multi-year trends. — Drought conditions
prevailed during the first four years of our
study (1993-1996); rainfall ranged 40—106
mm compared to a mean of 170 mm for 1986-
1997 (Fig. 1). By contrast, 1997 had over
twice the mean annual precipitation recorded
for the study site (Fig. 1). Concomitantly,
there was a declining trend in bird species
richness from 43 species in winter 1993 to 29
at the end of the drought in autumn 1997 (a
33% decrease; Fig. 1). Twenty-nine species
may well represent the minimum number of
bird species present in the reserve at any time.
As soon as El Nino driven precipitation re-
appeared at the site (the previous two occur-
rences were 1987 and 1992), there was an in-
crease from 29 species in autumn 1997 to 42
species in summer 1997 (a 45% increase; Fig.
1). Whether this is close to the maximum bird
species richness that the site can accommo-
date has not been determined.

Bird density also tracked precipitation pat-
terns (Fig. 2). Bird numbers in ravines de-
clined from about 45-50 individuals/ha
(1993) to about 10/ha in summer 1994 (the
driest year in the series; Fig. 1), and started a
slow recovery through 1995 and 1996, reach-
ing densities of 55-60 birds/ha during sum-
mer 1997. Bird densities on slopes were ap-
parently less than those in ravines (Fig. 2) but
paralleled the same trends until summer 1995,
when we terminated censuses on slopes.

Similar trends were observed among small
mammals but not among terrestrial arthropods
(Fig. 3). Small mammals declined markedly
through 1993 and 1994, remained at very low
levels 1995-1996, and recovered after the
spring 1997 (Fig. 3). Terrestrial arthropods did
not display such marked fluctuation (Fig. 3).
There were significant positive correlations
between small mammal density and both rap-
tor species richness {r^ = 0.77, df = 17, P <
0.001) and raptor density (r^ = 0.76, df = 17,
P < 0.001). Although positive, there were no

530

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 1. Birds of Las Chinchillas National Reserve (northcentral Chile) and their categorization by repro-

diictive status, dietary category, residence status (in parentheses,
sighted), and habitat type (landscape aspect).

, number of calendar

seasons the species was

Specie.s

Reproduction

Diet

Residence

Habitat

Nothoprocta perdicaria

Nesting

Granivore

Resident (18)

Ravine/Slope

Ca.s me radius albus

Non-nesting

Carnivore

Migrant (6)“

Water edge

Egretta thula

Nesting

Carnivore

Migrant (8)

Water edge

Nycticorax nycticorax

Non-nesting

Carnivore

Migrant (1)

Water edge

Vultur gryphus

Non-nesting

Carnivore

Migrant (5)

Slope

Elanus leucurus

Non-nesting

Carnivore

Migrant (0)'’

Elat areas

Geranoaetus melanoleucus

Nesting

Carnivore

Resident (16)

Ravine/Slope

Buteo polyosoma

Nesting

Carnivore

Resident (14)

Ravine

Parabuteo unicinctus

Non-nesting

Carnivore

Migrant (8)

Flat areas

Milvago chimango

Nesting

Insectivore

Resident (14)

Ravine/Flat areas

Ealco peregrinus

Non-nesting

Carnivore

Summer visitor (5)

Ravine/Slope

Falco femoralis

Nesting

Carnivore

Summer visitor (4)

Slope

Ealco sparverius

Nesting

Insectivore

Resident (13)

Ravine/Flat areas

Callipepla californica

Nesting

Granivore

Resident (19)

Ravine/Slope

Rallus sanguinolentus

Nesting

Carnivore

Migrant (8)

Water edge

Vanellus chilensis

Nesting

Carnivore

Resident (14)

Flat areas

Columba araucaria

Non-nesting

Granivore

Winter visitor (3)

Ravine

Zenaida auriculata

Nesting

Granivore

Resident (14)

Ravine/Slope

Metriopelia melanoptera

Unknown

Granivore

Summer visitor (11)

Ravine/Slope

Cyanoliseus patagonus

Unknown

Erugivore

Resident (15)

Ravine

Tyto alba

Nesting

Carnivore

Migrant (11)

Ravine/Slope

Bubo virginianus

Nesting

Carnivore

Resident (6)'’

Ravine/Slope

Glaucidium nanum

Nesting

Carnivore

Resident (13)

Ravine/Slope

Speotvto cunicularia

Nesting

Carnivore

Resident (16)

Ravine/Slope

Caprirnulgus longirostris

Nesting

Insectivore

Resident (19)

Ravine

Patagona gigas

Nesting

Nectarivore

Summer visitor ( 12)

Ravine/Slope

Seplianoides galeritus

Non-nesting

Nectarivore

Winter visitor (10)

Ravine/Slope

Colaptes pitius

Non-nesting

Insectivore

Migrant (8)

Slope

Picoides lignarius

Nesting

Insectivore

Resident (19)

Ravine/Slope

Geositta rufipennis

Non-nesting

Granivore

Winter visitor (6)

Ravine/Slope

Upucerthia dumetaria

Non-nesting

Insectivore

Migrant (10)

Ravine

Upucerthia ruficauda

Non-nesting

Insectivore

Summer visitor (5)

Ravine

Chilia melanura

Nesting

Insectivore

Resident (19)

Ravine/Slope

Leptasthenura aegithaloides

Nesting

Insectivore

Resident (19)

Ravine/Slope

Asthenes modes ta

Nesting

Insectivore

Resident (16)

Ravine/Slope

Pteroptochos megapodius

Nesting

Insectivore

Resident (19)

Ravine/Slope

Scelorchilus albicollis

Nesting

Insectivore

Resident (19)

Ravine/Slope

Scytalopus magellanicus

Nesting

Insectivore

Resident (16)

Ravine

Agriornis livida

Non-nesting

Carnivore

Migrant (7)

Slope

Pyrope pyrope

Nesting

Insectivore

Resident (18)

Ravine/Slope

Muscisaxicola macloviana

Non-nesting

Insectivore

Winter visitor (2)

Ravine/Slope

Elaenia albiceps

Nesting

Insectivore

Summer visitor (7)

Ravine

Anairetes parulus

Nesting

Insectivore

Resident (18)

Ravine/Slope

Colorhamphus parvirostris

Non-nesting

Insectivore

Winter visitor (8)

Ravine

Tachycineta leucopyga

Nesting

Insectivore

Resident (19)

Flat areas

Pygochelidon cyanoleuca

Non-nesting

Insectivore

Migrant (3)

Flat areas

Troglodytes aedon

Nesting

Insectivore

Resident ( 19)

Ravine/Slope

Phytotoma rara

Non-nesting

Eolivore

Migrant ( 1 )

Ravine

Turdus falcklandii

Nesting

Omnivore

Migrant (9)

Ravine

Mimus thenca

Ne.sting

Omnivore

Resident ( 19)

Ravine/Slope

Sicalis sp.

Non-nesting

Granivore

Migrant (3)

Ravine

Zonotrichia capensis

Nesting

Granivore

Resident ( 19)

Ravine/Slope

Sturnella loyca

Nesting

Omnivore

Resident (19)

Ravine

Curaeus curaeus

Nesting

Omnivore

Resident ( 16)

Ravine/Slope

Phrygilus gayi

Non-nesting

Granivore

Winter visitor (8)

Ravine/Slope

Jciksic and Uizo • BIRD RESPONSES TO EL NINO IN CHILE

531

TABLE 1. CONTINUED.

Species

Reproduction

Diet

Residence

Habiial

Phiygilus fniticeti

Nesting

Granivore

Resident (18)

Ravine/Slope

Phry’gilus aluudinus

Nesting

Granivore

Migrant (8)

Ravine

Diuca diuca

Nesting

Granivore

Resident ( 19)

Ravine/Slope

Carduelis uropygialis

Non-nesting

Granivore

Migrant (2)

Ravine

Corduelis harhatus

Non-nesting

Granivore

Migrant (2)

Slope

Passer domesticus

Nesting

Granivore

Migrant (12)

Ravine

^ Species categorized simply as migrants did not yield enough data to determine whether they are summer or winter visitors, occasionals, or accidentals.
^ Not sighted during our work, but present in the area. See text for details.

significant correlations between terrestrial ar-
thropod abundance and insectivorous bird
species richness (r^ = 0.06, df = 17, P >
0.05), density on slopes (r^ = 0.49, df = 9, P
> 0.05) nor density on ravines (r^ = 0.37, df
= 17, P > 0.05).

During the transition from drought (1993-
1996) to wet year (1997), two previously un-
recorded species arrived at the reserve, the fo-
livorous Rufous-tailed Plantcutter {Phytotoma

rara) and the fish and amphibian eating
Brown Heron (Nycticorax nycticorax).

DISCUSSION

The percentage (51%) of species at our
study site that were migrants, is similar (48%
of 88 species) to that reported by Marone
(1992a,b) in the Monte scrubland across the
Andes at similar latitudes in Argentina. As in
Chile, the high percentage of migrants in the

Calendar seasons

EIG. 1. Seasonal variation in rainfall and number of bird species at Las Chinchillas National Reserve
throughout the study period (19 calendar .seasons = 5 years). W = Winter (June-August), Sp = Spring (Sep-
tember-November), Su = Summer (December-Eebruary), A = Autumn (March-May). The segmented line
represents the mean annual rainfall of the last 12 years.

532

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Calendar seasons

FIG. 2. Bird density (mean ± SD) at the reserve throughout the study period, by habitat type: ravines
and slopes. W = Winter, Sp = Spring, Su = Summer, A = Autumn. Censuses on the slopes were discontinued
after summer 1995.

Calendar seasons

FIG. 3. Abundance of terrestrial arthropods (mean ± SD) and of small mammals (mean ± SD) at the
reserve throughout the study period. Abundance of arthropods is the mean of four samples (two from opposite
facing slopes and two from a ravine), that of mammals is an average for the north and south facing slopes. W
= Winter, Sp = Spring, Su = Summer, A = Autumn.

Jciksic and Uizo • BIRD RESPONSES TO EL NINO IN CHILE

533

Argentine Monte may be associated with the
extreme fluctuations in precipitation that char-
acterize arid and semiarid regions of South
America. Varying precipitation levels in Chile
are associated with changes in primary and
secondary production (Fuentes and Campu-
sano 1985; Jaksic et al. 1996, 1997). Migrant
birds may exploit these unpredictable resourc-
es by moving north from the more mesic
Mediterranean region of central Chile or down
from the Coastal Range to the west or the An-
des to the east of our study area.

Marone (1992a,b) observed that insecti-
vores (52%), carnivores (19%), and grani-
vores (15%) dominated the Argentine Monte,
accounting for 86% of the local assemblage,
similar to the 87% we recorded in Chile. Nev-
ertheless, there were more insectivores and
fewer carnivores and granivores in the Argen-
tine site than in our Chilean site. This suggests
that there may be differences in resource lev-
els between Chilean and Argentine semiarid
sites. More data are needed from the Monte.

We detected positive correlations between
number and/or density of bird species and spe-
cific resource levels (i.e., with regard to small
mammals but not to terrestrial arthropods).
Gutierrez and coworkers (pers. comm.) mea-
sured the seed bank of the reserve before
(1996) and during El Nino (1997). We found
a positive association between this resource
and richness/density of granivorous birds dur-
ing the respective years. In central Chile, Lo-
pez-Calleja (1995) found that the granivorous
Diuca Finch (Diuca diuca) and Rufous-col-
lared Sparrow (Zonotrichia capensis), two
species that were also present at our site,
changed their seed preferences in response to
short-term fluctuations in seed abundance. A
detailed study of the abundance of these two
species and their respective diets in relation to
the seed bank in the reserve is needed.

Why were birds in our study site more
abundant in ravines than in slopes? Perhaps
food levels are higher in ravines than on
slopes. Also, there is more vegetated ground
in ravines (70%) than on slopes (53%). Struc-
tural aspects of the former habitat may render
it more attractive to birds. The presence of a
tree layer (e.g., Maytenus boaria, Quillaja sa-
ponaria) in ravines, which is absent from
slopes, may favor higher bird densities be-
cause it provides more shelter, roosting and

nesting sites, and perhaps greater food diver-
sity (e.g., foliage insects).

The bird assemblage in the reserve showed
both short (seasonal) and multi-year variation
in its composition and density. Although bird
species richness and density in our study site
were lower during drought years and higher
during wet years, it should be noted that
droughts are more frequent and last longer
than El Nino events (eight dry years versus
four wet ones in 1986-1997). In a semiarid
climate, perhaps dry years set the baseline in
species richness and density for the bird as-
semblage, and both increase during wet years
because of the immigration of opportunistic
species from elsewhere. Thus, El Nino driven
rains impose a strong abiotic influence, which
cascades from increased primary and second-
ary productivity to bird species richness and
density.

ACKNOWLEDGMENTS

B. Saavedra, E. Silva, R. Soto, and S. Tellier helped
us with their respective expertises. R Feinsinger helped
us in many ways. C. F. Estades. F. Hertel, M. Marini,
F Vuilleumier and two anonymous reviewers made co-
gent criticisms that helped us improve our presentation
of data. The study was funded by grants Fondecyt 193-
0639 and 196-0319, NSF-INT 92-14085, the Mellon
Fund. Lazo acknowledges the support of the Mellon
Foundation toward postgraduate training. Jaksic is sup-
ported by a Presidential Chair in Science, Republic of
Chile.

LITERATURE CITED

Avery, M. L. and C. van Riper, III. 1989. Seasonal
changes in bird communities of the chapanal and
blue-oak woodland in central California. Condor
91:288-295.

Barber, R. T. and F P. Chavez. 1983. Biological con-
sequences of El Nino. Science 222:1203-1210.
Bibby, C. J.. N. D. Burgess, and D. A. Hill. 1993.
Bird census techniques. Academic Press, London,
U.K.

Burnham, K. R, D. R. Anderson, and J. L. Laake.
1980. Estimation of density from line transect
sampling of biological populations. Wildl. Mon-
ogr. 72:1-202.

Conner, R. N. and J. G. Dickson. 1980. Strip transect
sampling and analysis for avian habitat studies.
Wildl. ,Soc. Bull. 8:4-10.

di Castri, F. and E. Hajek. 1976. Bioclimatologi'a de
Chile. Ediciones Univ. Catolica de Chile, Santia-
go, Chile.

Feinsinger R, L. A. Swarm, and J. A. Wolfe. 1985.
Nectar-feeding birds on Trinidad and Tobago:

534

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

comparison of diverse and depauperate guilds.
Ecol. Monogr. 55:1-28.

Fuentes, E. R. and C. Campusano. 1985. Pest out-
breaks and rainfall in the semi-arid region of
Chile. J. Arid Envir. 8:67-72.

Fuentes, E. R., E. R. Hajek, and G. Espinoza. 1988.
Some consequences of rainfall variability for
Mediterranean type ecosystems in Chile. Pp. 347-
360 in Time scales and water stress (F. di Castri,
C. Floret, S. Rambal, and J. Roy, Eds.). Interna-
tional Union of Biological Sciences, Paris, France.

Gajardo, R. 1978. Antecedentes para la determinacion
de las comunidades vegetales de Auco. (Illapel,
IV Region). Cs. For. Chile 1:19-27.

Gibbs, H. L. and P. R. Grant. 1987. Ecological con-
sequences of an exceptionally strong El Nino
event on Darwin’s finches. Ecology 68:1735-
1746.

Grant, P. R. and B. R. Grant. 1987. The extraordi-
nary El Nino event of 1982-83: effects on Dar-
win’s finches on Isla Genovesa, Galapagos. Oikos
49:55-66.

Gutierrez, J. R., P. L. Meserve, E M. Jaksic, L. C.
Contreras, S. Herrera, and H. Vasquez. 1993.
Structure and dynamics of vegetation in a Chilean
arid thornscrub community. Acta Oecol. 14:271 —
285.

Hall, G. A., H. L. Gibbs, P. R. Grant, L. W. Bots-
FORD, AND G. S. Butcher. 1988. Effects of El
Nino-Southem Oscillation (ENSO) on terrestrial
birds. Proc. Int. Ornithol. Congr. 19:1759—1775.

Haug, E. a. and a. B. Didiuk. 1993. Use of recorded
calls to detect burrowing owls. J. Field Ornithol.
64:188-194.

Jaksic, E M. and P. Feinsinger. 1991. Bird assem-
blages in temperate forests of North and South
America: a compari.son of diversity, dynamics,
guild structure, and resource use. Rev. Chilena
Hist. Nat. 64:491-510.

Jaksic, E M., P. Feinsinger, and J. E. Jimenez. 1996.
Ecological redundancy and long-term dynamics of
vertebrate predators in semiarid Chile. Conserv.
Biol. 10:252-262.

Jaksic, F. M., S. I. Silva, P. L. Meserve, and J. R.
Gutierrez. 1997. A long-term study of vertebrate
predator responses to an El Nino (ENSO) distur-
bance in western South America. Oikos 78:341-
354.

Jimenez, J. E., P. Feinsinger, and E M. Jaksic. 1992.
Spatiotemporal patterns of an imiption and de-
cline of small mammals in north-central Chile. J.
Mamm. 73:356-364.

Johnson, R., R. Brown, L. T. Haight, and J. M. Simp-
son. 1981. Playback recording as a special avian
censusing technique. Stud. Avian Biol. 6:68-75.

Karr, J. R. and K. E. Freemark. 1983. Habitat selec-
tion and environmental gradients: dynamics in the
“stable” tropics. Ecology 64:1481-1494.

Koti.er, B. P. and R. D. Holt. 1989. Predation and
competition: the interaction of two types ot spe-
cies interactions. Oikos 54:256-260.

Lima, S. L. 1987. Clutch size in birds: a predation
perspective. Ecology 68:1062-1070.

Lindsey, G. D., T. K. Pratt, M. H. Reynolds, and J.
D. Jacobi. 1997. Response of six species of Ha-
waiian forest birds to a 1991—1992 El Nino
drought. Wilson Bull. 109:339—343.

Loiselle, B. a. and j. G. Blake. 1991. Temporal var-
iation in birds and fruits along an elevational gra-
dient in Costa Rica. Ecology 72:180-193.

Lopez-Calleja, M. V. 1995. Dieta de Zonotrichia ca-
pensis (Emberizidae) y Diuca cliuca (Fringillidae):
efecto de la variacion estacional de los recursos
troficos y la riqueza de aves granivoras en Chile
central. Rev. Chil. Hist. Nat. 68:321—331.

Marone, L. 1992a. Seasonal and year-to-year fluctu-
ations of bird populations and guilds in the Monte
Desert, Argentina. J. Field Ornithol. 63:264—308.

Marone, L. 1992b. Estatus de residencia y categori-
zacion trofica de las especies de aves en la Re-
serva de la Biosfera de Nacunan, Mendoza. Hor-
nero 13:207-210.

Martin, T. E. 1985. Resource selection by tropical fru-
givorous birds: integrating multiple interactions.
Oecologia 66:563-573.

Martin, T. E. 1987. Food as a limit on breeding birds:
a life history perspective. Annu. Rev. Ecol. Syst.
18:453-487.

Massey, B. W., D. W. Bradley, and J. L. Atwood.
1992. Demography of a California Least Tern col-
ony including effects of the 1982—83 El Nino.
Condor 94:976-983.

Meserve, P. L., J. A. Yunger, J. R. Gutierrez, L. C.
Contreras, W. B. Milstead, B. K. Lang, K. L.
Cramer, S. Herrera, V. O. Lagos, S. I. Silva, E.
L. Tabilo, M. a. Torrealba, and E M. Jaksic.
1995. Heterogeneous responses of small mammals
to an El Nino Southern Oscillation event in north-
central semiarid Chile and the importance of eco-
logical scale. J. Mamm. 76:580-595.

Meyer de Schauensee, R. 1982. A guide to the birds
of South America. International Council for Bird
Preservation, Philadelphia, Pennsylvania.

Miskelly, C. M. 1990. Effects of the 1982—83 El Nino
event on two endemic landbirds on the Snares Is-
lands, New Zealand. Emu 90:24—27.

Pearson, D. L. 1991. A basis for developing broader
generalizations of bird community structure and
species co-occurrence. Proc. Int. Ornithol. Congr.
20:1462-1469.

Pulliam, H. R. and J. B. Dunning. 1987. The influ-
ence of food supply on local density and diversity
of sparrows. Ecology 68:1009-1014.

ScHREiBER, R. W. AND E. A. ScHREiBER. 1984. Central
Pacific seabirds and the El Nino Southern Oscil-
lation: 1982 to 1983 perspectives. Science 225:
713-716.

SoKAL, R. R. AND E J. Rohlf. 1981. Biometry: the
principles and practice of statistics in biological
re.search. Second ed. W. H. Freeman & Co., San
Francisco, California.

SouTHWOOD, T. R. E. 1978. Ecological methods with

Jciksic and Lazo • BIRD RESPONSES TO EL NINO IN CHILE

535

particular reference to the study of insect popu-
lations, second ed. Chapman and Hall, London,
U.K.

Wiens, J. A. 1990a. The ecology of bird communities.
Vol. 1. Cambridge Univ. Press, Cambridge, U.K.

Wiens, J. A. 1990b. The ecology of bird communities.

Vol. 2. Cambridge Univ. Pre.ss, Cambridge, U.K.
Wiens, J. A. 1993. Pat times, lean times and compe-
tition among predators. Trends Ecol. Evol. 8:348-
349.

Wilson Bull., 111(4), 1999, pp. 536-540

POTENTIAL FOR PREDATOR LEARNING OF ARTIFICIAL
ARBOREAL NEST LOCATIONS

RICHARD H. YAHNER" AND CAROLYN G. MAHAN' -

ABSTRACT. — We examined the potential for predators to learn the location of artificial arboreal (1.5 m above
ground) nests in a managed forested landscape of central Pennsylvania from June— July 1995. We tested the
hypothesis that predators do not learn the location of artificial arboreal nests placed repeatedly at the same sites
(fixed nests) versus those placed at random sites in three habitats created by clearcutting (forested patches,
forested corridors, contiguous forest). Sixty-nine (23%) of 299 total nests in five combined trials were disturbed
by predators; 11 (16%) of these disturbances were attributed to corvids. Predation rates were greater on nests
placed at random (28%) compared to fixed sites (18%, P < 0.05), indicating predators did not learn or return
to the location of arboreal nests during our study. Predation rates varied significantly (P < 0.001) among habitats,
with 49% of the nests disturbed in the forested-patch habitat versus only 7% and 13% in forested-corridor and
contiguous-forest habitats, respectively. We propose that predation was higher in forested patches than in the
other two habitats because the former had greater amounts of edge. Received 12 Nov. 1998, accepted 10 May
1999.

Artificial nest studies have been useful in
examining the relationships between avian
nesting success and landscape patterns (e.g.,
Paton 1994, Bayne and Hobson 1997). Sev-
eral investigators have indicated that depre-
dation of artificial and natural avian nests in
managed forests varies with landscape pat-
terns created by clearcutting (Yahner and Ross
1995, Vander Haegen and DeGraaf 1996,
Yahner and Mahan 1996a). However, if pre-
dation rates on artificial nests are used as an
indicator of temporal or spatial trends in avian
nesting success (Yahner 1996, Sargent et al.
1998, Wilson et al. 1998), then the potential
effect of the ability of predators to learn the
locations of artificial nests needs to be deter-
mined. For example, as a consequence of
clearcutting in a localized area, the availability
of suitable nest sites may decline, thereby en-
abling predators to find nests located in the
remaining uncut forested tracts (patches or
conidors).

Forest clearcutting for Ruffed Grouse {Bon-
asa umbellus) at the Barrens Grouse Habitat
Management Area (GHMA) in central Penn-
sylvania provided us with an ideal opportunity
to test the hypothesis that predation rates did
not vary between artificial arboreal (1.5 m

' .School of Forest Resources, The Pennsylvania
.State Univ., University Park, PA 16802-4300; E-mail:
rhy@psu.edu

2 Current address: Dept, of Biology, Penn State Al-
toona, Altoona, PA 16601.

’ Corresponding author.

above ground) nests placed at sites used re-
peatedly (fixed nests) versus random sites in
a managed forested landscape. To our knowl-
edge, predation rates on artificial nests at fixed
vs random sites has been examined only with
ground nests (Yahner and Mahan 1996a).

STUDY AREA AND METHODS

Our study was conducted on a 1166-ha Barrens
GHMA, State Game Lands 176, Centre County, Penn-
sylvania, where a series of experimental studies deal-
ing with depredation of artificial and actual nests have
been conducted (e.g., Yahner and Wright 1985, Yahner
1991, Yahner and Ross 1995, Yahner and Mahan
1996a). The Barrens GHMA includes reference (con-
tiguous forest habitat) and treated (forested-patch and
forested-corridor habitats) sectors of similar size (Fig.
1). The treated sector is divided into 136 contiguous
4-ha blocks, and each block is partitioned into four 1
ha ( 100 X 100 m) plots arranged in a elockwise pattern
(plots A-D). At the first cutting cycle (winter 1976-
1977), plot A was clearcut in each block. At the second
cycle (winter 1980-1981), plot B was clearcut in each
block of the forested-patch habitat. At the third and
last cycle (winters 1985—1986 and 1986—1987), plot B
in each block of the forested-conidor habitat and plot
C in each block of the forested-patch habitat were
clearcut. The remaining uncut plots in the treated sec-
tor and forest in the reference sector have not been
clearcut for 75-80 years. As a result of these three
cutting cycles, a mosaic of uncut plots (plot D) entirely
surrounded by clearcut plots of three age classes (plots
A-C) occuned in the forested-patch habitat, whereas
100 m wide corridors of uncut plots (plots C— D) re-
mained in the forested-corridor habitat (Fig. 1).

We placed artificial arboreal ( 1 .5 m above ground)
nests during five time periods (trials) from early June
through July 1995 (Yahner and Mahan 1996a). A trial
was 6 days in length, with 8 days between trials. At

536

Yahner and Mahan • PREDATOR LEARNING

537

REFERENCE TREATED

SECTOR

Tt 976-77 r>

’777777Z

31985-87T

1 Uncut 1

1 Uncut \

UNIMPROVED ROAD
I I CONTIGUOUS-FOREST HABITAT
V777\ FORESTED-CORRIDOR HABITAT
FORESTED-PATCH HABITAT

FIG. I. Schematic of reference and treated sectors at the Barrens GHMA, Centre County, Pennsylvania.
Dates of cutting cycles are given in plots A and B of the 76 blocks in the forested-corridor habitat of forest
clearcutting and in plots A— C of the 60 blocks in the forested-patch habitat of clearcutting. Forest in the
contiguous-forested habitat of clearcutting (reference sector), in plots C and D of the forested-corridor habitat,
and in plot D of the forested-patch habitat.

the beginning of the study, 10 uncut plots (plot D)
were chosen randomly in both forested-patch and for-
ested-corridor habitats and 10 sites were randomly se-
lected in the contiguous forest. These 30 sites were
designated as fixed nests and were used in all trials (1—
5) for ne.st placement. For each trial, we randomly
chose 10 additional uncut plots (plot D) each in both
forested-patch and forested-corridor habitats and 10
sites in the contiguous forest; these additional 30 sites
were termed random nests. This resulted in 60 nests/
trial, with 20 nests/habitat (forested patch, forested cor-
ridor, and contiguous forest) and 30 nests/nest-site type
(fixed and random).

Artificial nests (10 cm diam and 10 cm deep) were
constructed of chicken wire painted flat black to reduce
glare and lined with leaf litter; nests were attached to
the nearest woody stem (1-5 cm dbh) with green wire
(Yahner and Scott 1988). Two fresh, brown chicken
eggs were placed in each nest and sunk slightly below
the rim of the nest to minimize detection. We chose
large brown chicken eggs in this study because they
allowed us to directly compare our results with those
obtained in other studies at the study site, including

artificial ground and arboreal nest studies conducted
before the third cutting cycle (e.g., Yahner and Wright
1985, Yahner and Scott 1988), an artificial ground nest
study conducted after the third cutting cycle (Yahner
and Mahan 1996a), and a study of Wood Thmsh nest-
ing success after the third cutting cycle (Yahner and
Ross 1995). One nest was established at each site. In
forested-patch and forested-corridor habitats, nests
were located 50 m from an edge in the center of plot
D; in the contiguous forest, nests were placed at least
50 m from an edge (e.g., logging road). Rubber gloves
and boots were worn when placing nests to reduce
human scent (Nol and Brooks 1982).

We determined the fates of nests (e.g.. undisturbed,
di.sturbed by an avian predator, disturbed by a nona-
vian predator) at the end of each trial (Yahner and
Mahan 1996a). Nest predators were classified by mode
of disturbance and general nest appearance; eggs with
peck holes were categorized as preyed upon by birds,
and nests without eggs or with crushed eggs were clas-
sified as preyed upon by nonavian predators (Rearden
1951, Yahner and Scott 1988, Hernandez et al. 1997).
Eggs and egg fragments were removed from nests at

538

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 1. Pate of artificial arboreal nests in relation to type of nest site, habitat, and trial in a
forested landscape at the Barrens GHMA, Centre County, Pennsylvania, June— July 1995.

managed

Nest fate

Type of
nest-site

Trial

Habitat

Total

Undisturbed

Pixed

Porested patch

Porested corridor

Contiguous forest

122

Random

Porested patch

Forested corridor

Contiguous forest

108

Disturbed

Pixed

Forested patch

Forested corridor

Contiguous forest

Random

Forested patch

Forested corridor

Contiguous forest

the end of each trial. The location of one nest in the
forested-corridor habitat was not found after nest
placement during trial 1 .

Common bird species nesting in uncut forest within
2 m of ground level at the Barrens Grouse HMA were
Wood Thrush (Hylocichla mustelina) and Eastern
Towhee (Pipilo erythrophlhaImu.s\ Yahner 1991). Po-
tential predators on artificial arboreal nests were Amer-
ican Crow (Corvus hrachyrhynchos). Blue Jay (Cyan-
ocitta cristata), and raccoon (Procyon lotor, Yahner
and Scott 1988, Yahner and Morrell 1991). Smaller
mammalian predators, e.g., eastern chipmunk (Tamia.s
striatu.s) and white-footed mice (Peromy.scu.s leuco-
pu.s), probably had minimal effect on our artificial
nests because of the relatively large egg size (see Rop-
er 1992, Haskell 1995, DeGraaf and Maier 1996, Yah-
ner and Mahan 1996b).

We examined dependency of nest fate (undisturbed
and disturbed) on nest-site type (fixed versus random),
habitat (forested patch, forested corridor, and contigu-
ous forest), and trial (1-5) using a four-way test-of-
independence (BMDP4L Log-Linear Model; Dixon
1990). Likelihood ratios (G’) were used to determine
interactions of nest fate with the three other variables
using log-linear models (Dixon 1990. Sokal and Rohlf
1995). If nest fate was dependent on a variable with
more than two levels, we used 2X2 G-tests-of-inde-
pendence about the cell(s) of interest.

RESULTS

Sixty-nine (23%) of the 299 artificial ar-
boreal nests were disturbed during the five tri-
als combined (Table 1); one nest location was

not found in trial 1. We attributed 11 (16%)
of the disturbed nests to avian predators. Nest
fate was dependent on nest type, with fewer
arboreal nests disturbed at fixed than at ran-
dom sites (18% vs 28%, respectively; G =
4.0, df = I, P < 0.05).

Nest fate varied with habitat (G = 55.8, df
= 2, P < 0.001). Rate of nest disturbance was
higher in the forested-patch habitat (49%)
compared to either forested-corridor (7%) or
contiguous-forested habitats (13%; G ^ 22.3,
df = 1, P < 0.001). The number of disturbed
nests in the forested-corridor habitat, however,
was similar to that in the contiguous-forest
habitat (P > 0.05). In contrast, nest fate was
not associated with trial or with interactions
of two or more variables (P > 0.05).

DISCUSSION

We believe that predators did not learn the
location of arboreal nests in our study (Eibl-
Eibesfeldt 1970, Krebs 1978, Yahner and
Wright 1985) because disturbance rates were
higher at random than at fixed sites and be-
cause rates did not vary among trials. In an-
other study of artificial nests, both avian and
mammalian predators preyed upon nests ran-
domly and did not learn the location of ex-
perimental nests (Angelstam 1986). In con-

Yalwer ami Mahan - PREDATOR LEARNING

539

trast, previous work at the Barrens GHMA
showed that predators probably learned the lo-
cation of ground nests at fixed nests in the
forested-patch sector, particularly as the study
progressed (trials 4 and 5; Yahner and Mahan
1996a).

Because artificial nests pose potential biases
and the debate on their usefulness in assessing
success of natural nests continues, caution
should be used in interpreting the results ob-
tained from artificial nest studies in making
management decisions (e.g., Yahner 1996, Or-
tega et al. 1998, Wilson et al. 1998). Care
should be used when extrapolating results ob-
tained from artificial nest studies compared to
naturally occurring nests because predation
rates on the two types of nests may vary and
predation rates may differ among years (Sto-
raas 1988). For example, predators may use
behavioral cues from nesting birds to locate
naturally occurring nests. Well designed stud-
ies using artificial nests remain a useful ap-
proach to making inferences about factors af-
fecting avian nesting success, especially when
comparisons are made between local habitats,
among nests in a given locality, at the same
locality over several years, or in detecting
trends in rates of predation (Roper 1992, Yah-
ner and Mahan 1996a, Wilson et al. 1998).

Our study and others provided evidence
that uncut wooded corridors, which are at
least 100 m wide in a forested landscape af-
fected by clearcutting, may provide consider-
ably more secure nesting habitat for breeding
birds than small uncut forest stands. For ex-
ample, Yahner and Ross (1995) found lower
predation on Wood Thrush nests in the for-
ested-corridor habitat (50%) than in the con-
tiguous forest (61%) or forested-patch habitats
(100%). Based on a study of nest predation
along uncut buffer strips retained after clear-
cutting near streams in Maine, Vander Haegen
and DeGraaf (1996) provided evidence that
relatively wide (>150 m) strips enhanced
nesting success of forest birds. Their study in-
cluded artificial ground and arboreal nests
containing Japanese Quail (Coturnix coturnix)
eggs. Despite conflicting evidence for preda-
tor learning of the location of artificial arbo-
real versus ground nests, we recommend that
investigators using artificial nests in frag-
mented forested landscapes carefully random-
ize nest placement in order to mitigate detec-

tion of nests by predators (see Yahner and Ma-
han 1996a).

ACKNOWLEDGMENTS

Our study was funded by the Pennsylvania Agricul-
tural Experiment Station and the Max McGraw Wild-
life Eoundation. We thank B. Niccolai, B. Ross, and
C. Stem for field assistance.

LITERATURE CITED

Angelstam, P. 1986. Predation on ground-nesting
birds’ nests in relation to predator densities and
habitat edge. Oikos 47:365-373.

Bayne, E. M. and K. A. Hobson. 1997. Comparing
the effects of landscape fragmentation by forestry
and agriculture on predation of artificial nests.
Conserv. Biol. 11:1418-1429.

DeGraaf, R. M. and T. J. Maier. 1996. Effect of egg
size on predation by white-footed mice. Wilson
Bull. 108:535-539.

Dixon, W. J. (Chief Ed.). 1990. BMDP statistical soft-
ware manual. Univ. of California Press, Berkeley.
Eibl-Eibesfeldt, I. 1970. Ethology: the biology of be-
havior. Holt, Rinhart, and Winston, New York.
Haskell, D. G. 1995. Reevaluation of the effects of
forest fragmentation on rates of bird-nest preda-
tion. Conserv. Biol. 9:1316—1318.

Hernandez, E, D. Rollins, and R. Cantu. 1997. An
evaluation of Trialmaster* camera systems for
identifying ground-nest predators. Wildl. Soc.
Bull. 25:848-853.

Krebs, J. R. 1978. Optimal foraging: decision rules for
predators. Pp. 23-63 in Behavioural ecology (J.
R. Krebs and N. B. Davies, Eds.). Sinauer Asso-
ciates, Inc., Sunderland, Massachusetts.

Nol, E. and R. j. Brooks. 1982. Effects of predator
exclosures on nesting success of Killdeer. J. Field
Ornithol. 53:263-268.

Ortega, C. P, J. C. Ortega, C. A. Rapp, and S. A.
Backensto. 1998. Validating the use of artificial
nests in predation experiments. J. Wildl. Manage.
62:925-932.

Baton, W. C. 1994. The effect of edge on avian nest-
ing success: how strong is the evidence? Conserv.
Biol. 8:17-26.

PiCMAN, J. 1988. Experimental study of predation on
eggs of ground-nesting birds: effects of habitat
and nest distribution. Condor 90:124-131.
Rearden, j. D. 1951. Identification of waterfowl nest
predators. J. Wildl. Manage. 15:386-395.

Roper, J. J. 1992. Nest predation experiments with
quail eggs: too much to swallow? Oikos 65:528-
530.

Sargeant, R. a., j. C. Kilgo, B. R. Chapman, and K.
V. Miller. 1998. Predation of artificial nests in
hardwood fragments enclosed by pine and agri-
cultural habitats. J. Wildl. Manage. 62: 1438-1442.
SoKAL, R. R. AND E J. Rohlf. 1995. Biometry, third
ed. W. H. Freeman and Company, San Francisco,
California.

540

THE WILSON BULLETIN • Vol. HI, No. 4, December 1999

Storaas, T 1988. A comparison of losses in artificial
and naturally occurring Capercaillie nests. J.
Wildl. Manage. 52:123-126.

Vander Haegen, W. M. and R. M. DeGraaf. 1996.
Predation on artificial nests in forested riparian
buffer strips. J. Wildl. Manage. 60:542-550.
Wilson, G. R., M. C. Brittingham, and L. J. Good-
rich. 1998. How well do artificial nests estimate
success of real nests? Condor 100:357-364.
Yahner, R. H. 1991. Avian nesting ecology in small
even-aged aspen stands. J. Wildl. Manage. 55:
155-159.

Yahner, R. H. 1996. Forest fragmentation, artificial
nest studies, and predator abundance. Conserv.
Biol. 10:672-673.

Yahner, R. H. and C. G. Mahan. 1996a. Depredation

of artificial ground nests in a managed forested
landscape. Conserv. Biol. 10:285-288.

Yahner, R. H. and C. G. Mahan. 1996b. Effects of
egg type on depredation of artificial ground nests.
Wilson Bull. 108:129-136.

Yahner, R. H. and T. E. Morrell. 1991. Depredation
of artificial nests in irrigated forests. Wilson Bull.
103:113-117.

Yahner, R. H. and B. D. Ross. 1995. Distribution and
success of Wood Thrush nests in a managed for-
ested landscape. Northeast Wildl. 52:1-9.

Yahner, R. H. and D. P. Scott. 1988. Effects of forest
fragmentation on depredation of artificial avian
nests. J. Wildl. Manage. 52:158-161.

Yahner, R. H. and A. L. Wright. 1985. Depredation
on artificial nests: effects of edge and plot age. J.
Wildl. Manage. 49:508-513.

Wilson Bull., 111(4), 1999, pp. 541-549

PREDATION ON ARTIFICIAL NESTS ALONG THREE EDGE TYPES
IN A NORTH CAROLINA BOTTOMLAND HARDWOOD FOREST

JAMES F. SARACCO'^ AND JAIME A. COLLAZO'

ABSTRACT. — Many researchers have reported high rates of nest predation near forest edges. However, edges
may be of various types (e.g., interior or exterior, abrupt or gradual), which may not always result in elevated
predation. We compared predation rates on artificial arboreal nests along three types of edges in a bottomland
forest in North Carolina during the 1996 breeding season. Edge types were forest— farm, forest— river, and the
transition zone between the two dominant forest types in the floodplain (cypress-gum swamps-natural levees).
We tested for differences in predation rates using two egg types: Northern Bobwhite (Colinus virginianus) and
clay eggs. Predation rates were higher (P < 0.05) along forest-farm edges than along the other two edges.
Predation rates did not differ between forest-river and transition zone edges. Patterns of predation on the two
egg types and higher avian predator abundance on forest-farm edges suggested that avian predators may have
exerted more predation pressure along these edges. These results are consistent with other studies, which suggest
that encroachment by agriculture into forested landscapes may negatively affect breeding birds. Our findings
also suggest that not all edge types are equivalent in terms of predation rates. This is important in assessing the
conservation value of bottomland forests, which may contain various edge types resulting from natural processes
(e.g., hydrodynamics). Received 19 Feh. 1999, accepted 6 July 1999.

Predation is the primary cause of nest loss
for a wide range of passerine birds (Martin
1992) and may be the most important factor
affecting their population dynamics (Temple
and Cary 1988). Forest birds nesting in highly
fragmented landscapes or near edges may ex-
perience higher rates of nest predation than
birds nesting in contiguous forests (Paton
1994, Andren 1995, Robinson et al. 1995).
However, forest edges occur in a variety of
contexts which may not always lead to in-
creased predation levels. For example, edges
may be in the interior (e.g., clearcuts within
contiguous forest) or along the exterior (e.g.,
agricultural encroachment from outside) of
forests and they exhibit varying degrees of
contrast from subtle to abrupt (Ratti and Reese
1988, Yahner et al. 1989, Hawrot and Niemi
1996, Fenske-Crawford and Niemi 1997,
Suarez et al. 1997). Most researchers reporting
high predation rates near edges have exam-
ined abrupt exterior edges (reviewed by An-
dren 1995). Those that have considered inte-
rior and more subtle edges have reported less
consistent results (e.g., Ratti and Reese 1988,
Yahner et al. 1989, Fenske-Crawford and Nie-

‘ North Carolina Cooperative Fish and Wildlife Re-
search Unit, Biological Resources Division, U.S. Geo-
logical Survey, Dept, of Zoology, North Carolina State
Univ., Raleigh, NC 27695-7617.

^Corresponding author; E-mail: jfsaracc(§>unity.ncsu.
edu

mi 1997, Suarez et al. 1997). Further inves-
tigation into the characteristics of edges that
influence levels of predation is clearly needed.
Such information could be used to assess the
conservation value of complex landscapes,
such as bottomland hardwood forests that sup-
port diverse breeding bird communities (e.g.,
Wharton et al. 1981, Mitchell and Lancia
1990, Mitchell et al. 1991, Pashley and Bar-
row 1992). These forested wetlands may con-
tain a variety of edge types that result from
the patchwork of plant communities whose ar-
rangement is influenced by site-specific hy-
drodynamics and sediment deposition rates
along floodplains (Wharton et al. 1982).

We compared predation rates on artificial
arboreal nests among three edge types in bot-
tomland hardwood forests along the Roanoke
River in North Carolina. The three edge types
were; (1) forest-farm edge (an abrupt exterior
edge), (2) forest-river edge (an abrupt interior
edge), and (3) levee-swamp edge (a gradual
interior edge where the two dominant plant
communities in the floodplain meet). We used
artificial nests primarily because of the logis-
tic and experimental advantages afforded by
their use. We do not claim that predation rates
on artificial nests represent those experienced
by natural nests, only that the pattern of pre-
dation among edge types are likely to be sim-
ilar for the two. For example, the few studies
that have compared patterns of predation
among habitats using both artificial and nat-

541

542

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

ural nests as well as studies comparing similar
habitats using either of these methods have
typically found a close match in predation pat-
terns for the two nest types (see Andren
1995). To our knowledge, this is the first study
to examine predation rates at an edge between
two relatively undisturbed forested plant com-
munities and only the third to examine a forest
edge abutting water (Bollinger and Peak 1995,
Vander Hagen and DeGraff 1996). Differenc-
es in predator communities among the three
edge types were assessed by comparing pat-
terns of predation on two egg types and the
abundance of likely avian nest predators.

STUDY AREA AND METHODS

This study was conducted within a contiguous forest
corridor along the lower Roanoke River between the
towns of Palmyra and Jamesville, North Carolina
(36° 9' N to 35° 50' N, 77° 20' W to 76° 53' W). The
forested areas we studied have been undisturbed for
more than 60 years. Loss and alteration of forests in
the floodplain have come primarily from crop (e.g.,
peanuts, cotton, wheat) and timber production. The
lower Roanoke ecosystem is comprised of 20 vegeta-
tive community types (Schafale and Weakley 1990), 2
of which are clearly dominant: cypress-gum swamp
and coastal plain levee forests (hereafter swamps and
levees, respectively). Swamps are flooded for extended
periods throughout the year. The dominant canopy spe-
cies are water tupelo (Nyssa aquatica) and bald cy-
press (Taxodium distichum)-, Carolina ash (Fraxinu.s
caroliniana) is common in the understory (Lynch et
al. 1994). Levees occur at slightly higher elevations
and are comprised of a diverse mixture of canopy spe-
cies including American elm (Ulmu.s americana),
green ash {Frcixinu.s penn.sylvanica), hackberry (Celtis
laevigata), boxelder (Acer negundo), water hickory
(Carya aquatica), and sweetgum (Liquidamhar .styra-
ciflua). The understory of levees is characterized by
pawpaw (A.simina triloba), ironwood (Carpinu.s caro-
liniana), and various vines (Lynch et al. 1994). Al-
though the sizes and shapes of patches of the two for-
est types are variable, levees generally occur as linear
patches close to the river channel formed by the de-
position of sediment following flooding events. Farther
from the river channel, these forests grade into
swamps. Levee-swamp edges are comprised of a mix-
ture of species typical of the two forest types. Forest-
river edges are compri.sed of species typical of levees.
Forest-farm edges are dominated by swamp trees, with
red maple (Acer rubrum) also a dominant species.

Artificial nests were placed along two 1.5 km tran-
sects established within each edge type. Survey tape
was u.sed to mark 59 25-m intervals (nest site locations
1-60) along each transect. Nests were placed at 50 m
intervals beginning at the first survey flag (nest site 1)
during trials one and three and beginning at the second
survey flag (nest site 2) during trial two. Thus, 30 nests

were placed along each transect during each trial. All
nests were placed on a suitable substrate within 15 m
of the survey flag. Transects along forest-river and for-
est-farm edges ran parallel to the river and fields, re-
spectively, and were approximately 15 m inside the
forest. Levee-swamp edge transects ran along the es-
timated center of the levee-swamp transition zone.
Transition zones were characterized by the presence of
bald cypress and water tupelo, wetter soils (often with
standing water), and a noticeable opening of the un-
derstory. This design resulted in all nests being within
30 m of a habitat boundary. Paton (1994) found that
edge effects on nest predation are typically found with-
in 50 m of a habitat boundary; the 30 m distance cutoff
we used was well within this range. All transects were
separated by at least 2 km and were at least 100 m
from any other edge type.

Because predators may respond to artificial nests
differently than to natural nests (Major and Kendal
1996), we attempted to mimic as closely as possible
the size, color, and locations of nests of Acadian Fly-
catchers (Empidonax virescens), a common breeding
species in the floodplain. Several other common breed-
ing species place their nests in similar locations (Lynch
et al. 1994). Artificial nests were constructed from
commercially available miniature grape vine wreaths
(approximately 8 cm outside and 5 cm inside diame-
ters) with bottoms of dried grass or leaves lining wire
mesh frames (approximately 4 cm deep). Nests were
attached with wire to the fork of a low hanging tree
branch, sapling, or shrub at a height of approximately
2.5 m.

Three 15 day trials were run over the course of the
1996 nesting season (30 May-24 July). Fifteen days
approximates a typical incubation period for open-
nesting passerines in the area. Two egg types were
placed in each nest: one Northern Bobwhite egg (Col-
inu.s virginianus) and one smaller white clay egg
(“Plastalina”, Van Aken International; approximately
20 X 10 mm) to account for potential biases associated
with egg type (Roper 1992; Haskell 1995a, b; Major
and Kendal 1996). Eggs were placed in each nest 3-5
days after nests were placed in the field. This was in-
tended to mimic the interval between nest building and
egg laying (Marini et al. 1995). We minimized human
scent at nest sites by wearing rubber boots and gloves
while placing nests and eggs, and while checking nests
(Nol and Brooks 1982). Nests were checked for signs
of predation on three occasions during each trial (day
5, 10, and 15). We considered a nest to be depredated
if either egg was damaged or missing. Predation was
attributed to a bird if the clay egg was found with bill
imprints and/or the bobwhite egg was found with
punctures suggestive of a bill (e.g., as described for
crows by Rearden 1951). We considered a nest to be
depredated by a large mouthed mammal if bobwhite
eggs were found half eaten from one end [suggesting
raccoon, Procyon lotor (Rearden 1951), or gray squir-
rel, Sciurus carolinen.sis (C. J. Whelan, pers. comm.)],
if chewed up clay eggs were found, or if nests were
destroyed (e.g., nest ring gone or pulled apart; Best

Saracco and Collazo • ARTIFICIAL NEST PREDATION AT EDGES

543

and Stauffer 1980). Nests for which tooth imprints or
scratches were found on clay eggs, or for which both
eggs were found still in the nest or in the immediate
vicinity and the clay egg was scratched, were consid-
ered to have been depredated by small-mouthed mam-
mals (e.g., Peromysciis mice; Major 1991, Haskell
1995b). Although snakes may have also contributed to
predation, we were unable to attribute predation events
to snakes based on evidence at nest sites.

Birds were censu.sed at 10 count stations located at
150 m intervals along each transect. All birds seen or
heard within a 50 m radius and more than 50 m but
within the area of interest, over a 10 min interval were
recorded (Hutto et al. 1986). One census was con-
ducted along each transect during the morning hours
(06:30-09:45 EST) between 20 May and 6 June.
Abundance of species likely to depredate nests [Amer-
ican Crow (Con>us hrachyrhynchos). Fish Crow (Cor-
vus ossfragus). Blue Jay (Cyanocilta chstata) and
Common Crackle (Quiscalus quiscula)] was expressed
as total detections per point (i.e., all detections, un-
bounded radius). Although some independence among
sampling stations may have been sacrificed by using
detections at all distances, each of these were “high-
detection-ratio” species (i.e., each had a high propor-
tion of the total detections recorded outside of the 50
m radius count circle), suggesting that detections at all
distances within the edge and immediately adjacent
habitats were more appropriate for comparisons (Hutto
et al. 1986).

Univariate repeated measures ANOVA was used to
test for differences in predation rates among edge types
(Proc GEM, SAS Institute 1990; Winer et al. 1991).
The response variable was the proportion of the 30
nests depredated on each transect. The independent
variable was edge type (forest— farm, levee-swamp,
and forest-river); trial (1, 2, and 3) and days of ex-
posure (5, 10, and 15) were repeated measures. Prior
to analyses the response variable was square root-arc-
sine transformed to meet homogeneity of variance as-
sumption (Levene’s test: P > 0.05; JMP, SAS Institute
1994). In order to test for potential biases associated
with egg type, we used McNemar’s tests conducted
separately for each edge type (Proc FREQ, SAS Insti-
tute 1990). We tested for differences in selected avian
predator abundance among edge types using nested-
ANOVA (Proc NESTED, SAS Institute 1990). The re-
sponse variable was the number of detections of se-
lected avian predators per point. Model terms were
edge type and transect [edge type]. Data met homo-
geneity of variance assumption (Levene’s test; P >
0.05; JMP, SAS Institute 1994). Differences in abun-
dance of individual species of avian predators among
edge types were assessed using Kruskal-Wallis tests
(Proc NPARIWAY, SAS Institute 1990). For species
where a significant edge effect was found, a posteriori
contrasts were computed using the nonparametric all-
treatments multiple contrast test described in Hollander
and Wolfe (1999). An a < 0.05 was used for all an-
alyses and values presented are means ± SE. Statistical

analyses were performed with JMP (version 3.2.2) and
SAS (version 7.0) for Windows.

RESULTS

Predation rates differed significantly among
edge types (F = 11.33, df = 2, 3; P = 0.04)
and were higher along the agricultural field-
forest edges than along the other two edge
types (F — 22.31, df = 1, 3; P = 0.01; Fig.
1). Predation rates did not differ between for-
est-river and levee-swamp edges (F = 0.35,
df = 1, 3; P > 0.05). There was no difference
in predation rate among trials (P = 0.05, df
= 2, 6; P > 0.05). Within trials, predation rate
increased with day of exposure (F = 94.54,
df = 2, 6; P < 0.001). Interaction between
day of exposure and edge type was nearly sig-
nificant (P = 3.55, df = 4, 6; P = 0.08). This
nearly significant interaction was likely
caused by differences in response pattern
(slope) between levee-swamp and forest-river
edges from 5-10 days of exposure (Fig. 1).
The difference among these two edges at 5
days of exposure was not significant (P =
5.23, df = 1,3; P > 0.05). Predation rates
were highest along forest-farm edges regard-
less of exposure time.

The number of nests for which the bob-
white egg was damaged or missing was high-
est on forest-farm edges, while the number of
nests for which only the clay egg was depre-
dated was similar among edge types (Fig. 2A).
Bobwhite eggs were preyed upon more fre-
quently on forest-farm edges than on the other
two edge types (Fig. 2B). Conversely, the per-
centage of depredated nests in which only the
clay egg was preyed upon was lowest on for-
est-farm edges and highest on levee-swamp
edges. For each edge type, the clay egg was
depredated significantly more often than the
bobwhite egg in nests where only one egg was
depredated (Forest-farm: = 9.49, df = 1;

P < 0.01; Forest-river: G,,jj = 19.15, df = 1;
P < 0.001; Levee-swamp: G,,jjj = 45.83, df =
1; P < 0.001). Despite this egg type bias, the
pattern of predation, higher on forest-farm
edges than on the other two edge types, was
the same regardless of whether predation was
on bobwhite or clay eggs (Fig. 2A).

We identified nest predators for 30% of
depredated nests (114/368). Of these, 69%
(79) were birds, 22% (25) were smaller
mouthed mammals, and 9% (10) were larger

544

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

100
90
80
70

c 60

"S 50

^ 40

30
20
10
0

LIG. 1. Predation rates on artificial nests at three edge types in bottomland hardwood forest along the

Roanoke River floodplain. North Carolina during the 1996 breeding season. Predation rates were significantly

higher along the forest-farm edges than on the other two edges {P < 0.05).

5 10 15

Days of exposure

mouthed mammals. Although measurements
of bill imprints in clay eggs were not taken,
the size and shape of these imprints suggested
that crows, Blue Jays, and Common Crackles
were among the avian predators. The abun-
dance of these nest predators differed by edge
type (F = 36.84, df = 2, 3; F < 0.01) and
was higher on forest-farm edges than along
forest-river and levee-swamp edges (F =
65.79, df = 1, 3; F < 0.01; Fig. 3). Avian
predator abundance did not differ significantly
between forest-river and levee-swamp edges
but tended to be higher along the forest-river
edge (F = 7.89, df = 1, 3; F = 0.07; Fig. 3).
Considered individually, the four predator
species were not consistent in their responses
to edge type (Fig. 3). The numbers of Amer-
ican and Fish Crows detected differed signif-
icantly among edge types (American Crow;

= 1 1.21, df = 2, F < 0.01; Fish Crow: =

12.27, df = 2; F < 0.01), and both of these
species were significantly more abundant
along forest-farm edges than along levee-
swamp edges (American Crow q = 4.28, Fish

Crow q = 4.42; F < 0.01). American Crows
were also significantly more abundant on for-
est-farm edges than along forest-river edges
{q — 3.65; F < 0.05). Fish Crows tended to
be more abundant along forest-farm edges
than along forest-river edges, although this
difference was not significant {q = 3.06; 0.05
< F < 0.10). In contrast. Blue Jays, which
also showed a significant edge effect (x^ =
6.05, df = 2; F < 0.05), were more abundant
along forest-river edges than on levee-swamp
edges {q = 3.33; F < 0.05). Contrasts be-
tween forest-farm edges and the other two
edge types for this species were not significant
(F > 0.05). Common Crackle abundance did
not differ significantly among edge types (x^
= 3.74, df = 2; F > 0.05).

DISCUSSION

Our findings are consistent with most pre-
vious studies that have reported high rates of
nest predation along abrupt exterior edges
(our forest-farm edge type; see Andren 1995).
The presence of more avian predators along

% of depreciated nests Number depredated

Saracco and Collazo • ARTIFICIAL NEST PREDATION AT EDGES

545

100 T

90
80
70 -
60
50
40
30
20
10
0

Forest-farm Forest-river Swamp-levee

FIG. 2. (A) Absolute number of depredated nests and (B) the percentage of depredated nests for which the

bobwhite egg, clay egg, or only the clay egg was preyed upon. Re.sults of McNemar's tests conducted for each
edge type (/? = 180 nests) .suggested a significant egg type bias (P < 0.01 for each edge type); however, the
same pattern of predation (highest on forest-farm edges) was apparent regardless of whether bobwhite or clay
eggs were considered.

Forest-farm Forest-river Swamp-levee

546

THE W ILSON BULLETIN • Vol. III. No. 4. December 1994

2.5

o '•■5
O

0.5

FIG. 3. Mean number of detections per point for selected avian nest predators at three edge types in a
bottomland hardwood forest along the Roanoke River floodplain. North Carolina. The pooled abundance for all
species was significantly higher on forest-farm edges than on the other two edge types (F < 0.01); the responses
of individual species were not consistent among edge types.

IIIIIIIIP

■ All Species
e American Crow
ID Fish Crow
^ Blue Jay
0 Common Crackle

Forest-farm

Forest-river

Swamp-levee

forest-farm edges may have contributed to the
higher predation rates we observed along
these edges. Marini and coworkers (1995)
found a positive correlation between avian
predator abundance (American Crows. Blue
Jays, and Common Crackles) and predation
levels on artificial nests in forest saplings, as
well as significantly higher predation rates on
these nests at forest-fimn edges. Other re-
searchers have also related avian nest predator
abundance (e.g.. corvids) to forest-farm edges
or shown predation on artificial nests by these
predators to be higher near such edges (e.g..
Whitcomb et al. 1981. Angelstam 1986. An-
dren 1992. Nour et al. 1993).

Our finding that nests at forest-river edges
experienced lower predation rates than forest-
farm edges is in accordance w ith Vander Hae-
gen and DeGraff ( 1 986) w ho found no effect
of distance from a river edge on predation
rate. In contrast. Bollinger and Peak (1995)
found predation rates to be uniformly high on
artificial ground nests along a forest edge bor-

dering water and a forest-fami edge in one
forest fragment in an agricultural setting.
Small forest fragments in agricultural land-
scapes such as this may become inundated
w ith certain mammalian predator species (e.g..
raccoons; gray squirrels. Sciiirus carolinensis:
and opossums. Didelphis marsupiali.s', Bider
1968. Matthiae and Steams 1981). The rela-
tively w ide and heavily forested river corridor
in our study may have alleviated any such
packing effects by predators.

The levee-swamp edges we studied are
unique in that they ivre naturally occurring
boundaries between plant communities rather
than edges resulting from human activities
(e.g.. agriculture, forestry practices). As such,
they may not be perceived as edges by some
predators that may move freely between le-
vees and swamps rather than concentrating ac-
tivities along the edge or using it as a travel
lane (Bider 1968. Chasko and Gates 1982).
This could explain the relatively low preda-
tion rates we observed at these edges.

Sarcuco ami Collazo • AR'I Il-ICIAL NEtST RKHDA I'ION A I lilKiliS

547

It is difficult to determine the relative im-
pact of different predators at the three edge
types because predators were only identified
for 30% of depredated nests. The greater
abundance of avian predators at forest-farm
edges may have contributed to the higher pre-
dation rates there; however, some mammalian
predators and snakes might also be abundant
and concentrate their activities or travel along
abrupt edges (Bider 1968, Chasko and Gates
1982, Durner and Gates 1993; but see Heske
1995). Unfortunately, we were unable to as-
sess the relative abundance of non-avian pred-
ators or their relative contribution to predation
on artificial nests. Nonetheless, there was
some indication that the predators responsible
for depredating nests may have differed
among edge types. For example, because
small mouthed predators may have been un-
able to damage the Northern Bobwhite eggs
[as has been reported for Japanese Quail {Co-
turnix coturnix) eggs; Roper 1992, Haskell
1995aJ, our finding that the proportion of
nests for which only the clay egg was depre-
dated was higher at the forest interior edges
suggests that small mouthed predators (e.g.,
mice) may have been more important at these
edges. In contrast, both the proportion and ab-
solute number of depredated nests for which
the bobwhite egg was preyed upon was high-
est at forest-farm edges. This supports the
contention advanced by Haskell (1995b) and
Nour and coworkers (1993) that avian and
larger mammalian predators increase in im-
portance in small forest patches or at the edg-
es of forests. Smaller mouthed predators, al-
though possibly more frequent at the interior
edges, appear to have depredated similar pro-
portions of nests along the three edge types.
Finally, differential predation rates at different
types of edges could also be influenced by
factors other than the types of predators in-
volved and their abundance. Future studies
should be designed to consider factors influ-
encing nest site selection (e.g., number of po-
tential nest sites) and nest densities of avian
community members (Martin 1993).

Bottomland hardwood forests of the south-
eastern U.S. are being destroyed and frag-
mented at high rates (Turner et al. 1981, Ab-
ernathy and Turner 1987). These areas provide
important breeding habitats for many migra-
tory and resident birds (Wharton et al. 1981,

Mitchell and Lancia 1990, Mitchell et al.
1991, Pashley and Barrow 1992). Understand-
ing how edges resulting from natural process-
es (e.g., hydrodynamics), as well as from an-
thropogenic modifications, affect breeding
bird communities is important to their conser-
vation and management. Our results suggest
that encroachment by agriculture may nega-
tively affect breeding birds through higher
predation rates along forest-farm edges. Nat-
ural edges between adjacent plant communi-
ties and at the forest-river interface may not
affect breeding birds in the same way.

ACKNOWLEDGMENTS

We thank L. Peoples and E. Sandlin for assistance
in the field. We also thank The Nature Conservancy,
the North Carolina Wildlife Resources Commission,
U.S. Fi.sh and Wildlife Service, and private land own-
ers for granting permission to work on and facilitating
access to their land. C. Parkhurst provided Northern
Bobwhite eggs. K. Pollock provided advise regarding
the design of the experiment, and G. Brown provided
assistance with statistical analyses. This project was
supported by the North Carolina Chapter of The Na-
ture Conservancy, North Carolina State University,
and Biological Resources Division, U.S. Geological
Survey. We thank M. Groom, D. Haskell, J. Lyons, T.
Simons, J. Walters, C. Whelan, and two anonymous
reviewers for comments on earlier versions of this
manuscript.

LITERATURE CITED

Abernathy, Y. and R. Turner. 1987. U.S. forested
wetlands: status and changes 1940-1980. Bio-
Science 37:721—727.

Andren, H. 1992. Corvid density and nest predation
in relation to forest fragmentation: a landscape
perspective. Ecology 73:794—804.

Andren, H. 1995. Effects of landscape composition on
predation rates at habitat edges. Pp. 225-255 in
Mosaic landscapes and ecological processes (L.
Hansson, L. Fahrig, and G. Merriam. Eds.). Chap-
man and Hall, London, U.K.

Angelstam, P. 1986. Predation on ground-nesting
birds nests in relation to predator densities and
habitat edge. Oikos 47:365-373.

Be.st, L. B. and D. E Stauffer. 1980. Factors affect-
ing nesting success in riparian bird communities.
Condor 82:149-158.

Bider, J. R. 1968. Animal activity in uncontrolled ter-
restrial communities as determined by a sand tran-
sect technique. Ecol. Monogr. 38:269-308.
Bollinger, E. K. and R. G. Peak. 1995. Depredation
of artificial avian nests: a comparison of forest-
field and forest-lake edges. Am. Midi. Nat. 134:
200-203.

Chasko, G. G. and J. E. Gates. 1982. Avian habitat

548

THE WILSON BULLETIN • Vol. HI, No. 4, December 1999

suitability along a transmission-line corridor in an
oak-hickory forest region. Wildl. Monogr. 82:1-
41.

Durner, G. M. and J. E. Gates. 1993. Spatial ecology
of black rat snakes on Remington Farms, Mary-
land. J. Wildl. Manage. 57:812-826.

Fenske-Crawford, T. J. and G. J. Niemi. 1997. Pre-
dation of artificial ground nests at two types of
edges in a forest-dominated landscape. Condor 99;
14-24.

Haskell, D. G. 1995a. Forest fragmentation and nest
predation; are experiments with Japanese Quail
eggs misleading? Auk 112:767-770.

Haskell, D. G. 1995b. A reevaluation of the effects
of forest fragmentation on rates of bird-nest pre-
dation. Conserv. Biol. 9:1316-1318.

Hawrot, R. Y. and G. j. Niemi. 1996. Effects of edge
type and patch shape on avian communities in a
mixed conifer-hardwood forest. Auk 113:586-
598.

Heske, E. j. 1995. Mammalian abundances on forest-
farm edges versus forest interiors in southern Il-
linois; is there an edge effect? J. Mamm. 76:562-
568.

Hollander, M. and D. A. Wolfe. 1999. Nonpara-
metric statistical methods. Second ed. John Wiley
and Sons, Inc., New York.

Hutto, R. L., S. M. Pletschet, and P. Hendricks.
1986. A fixed-radius point count method for non-
breeding and breeding season use. Auk 103:593-
602.

Lynch, J. M., J. A. Collazo, and J. R. Walters.
1994. Breeding birds of the lower Roanoke River
floodplain — A land managers guide. The Nature
Conservancy, Durham, North Carolina.

Major, R. E. 1991. Identification of nest predators by
photography, dummy eggs, and adhesive tape.
Auk 108:190-195.

Major, R. E. and C. E. Kendal. 1996. The contri-
bution of artificial nest experiments to understand-
ing avian reproductive success: a review of meth-
ods and conclusions. Ibis 138:298-307.

Marini, M. A., S. K. Robinson, and E. J. Heske. 1995.
Edge effects on nest predation in the Shawnee Na-
tional Forest, Southern Illinois. Biol. Conserv. 74:
203-213.

Martin, T. E. 1992. Breeding productivity consider-
ations; 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
Institution Press, Washington, D.C.

Martin, T. E. 1993. Nest predation and nest sites: new
perspectives on old patterns. BioScience 43:523-
532.

Matthiae, P. E. and E Stearns. 1981. Mammals in
forest islands in .southeastern Wisconsin. Pp. 55-
66 in Forest island dynamics in man-dominated
landscapes (R. L. Burgess and D. M. Sharpe.
Eds.). Springer- Verlag, New York.

Mitchell, L. J. and R. A. Lancia. 1990. Breeding bird

community changes in bald-cypress-tupelo wet-
land following timber harvesting. Proc. Annu.
Conf. Southeast. Assoc. Fish Wildl. Agencies 44:
189-201.

Mitchell, L. J., R. A. Lancia, R. Lea, and S. A.
Gauthreaux, Jr. 1991. Effects of clearcutting and
natural regeneration of breeding bird communities
of bald-cypress-tupelo wetlands. Pp. 155—161 in
Proc. Symp. Wetlands River Corridor Manage. (J.
Kusler and S. Daly, Eds.). Association of Wetland
Managers, Inc., Berne, New York.

Nol, E. and R. j. Brooks. 1982. Effects of predator
exclosures on nesting success of Killdeer. J. Field
Ornithol. 53:263-268.

Nour, N., E. Matthysen, and A. A. Dhondt. 1993.
Artificial nest predation and habitat fragmentation:
different trends in bird and mammal predators.
Ecography 16:111-116.

Pashley, D. N. and W. C. Barrow. 1992. Effects of
land use practices on Neotropical migratory birds
in bottomland hardwood forests. Pp. 315-320 in
Status and management of Neotropical migratory
birds (D. M. Finch and P. W. Stangel, Eds.).
USDA Forest Service, Rocky Mountain Forest
and Range Experimental Station. Gen. Tech. Rep.
RM-229. Fort Collins, Colorado.

Paton, P. W. C. 1994. The effect of edge on avian nest
success: how strong is the evidence? Conserv.
Biol. 8:17-26.

Ratti, j. T. and Reese, K. P. 1988. Preliminary test of
the ecological trap hypothesis. J. Wildl. Manage.
52:484-491.

Rearden, j. D. 1951. Identification of waterfowl nest
predators. J. Wildl. Manage. 15:386-395.

Robinson, S. K., F. R. Thompson, III, T. M. Donovan,
D. R. Whitehead, and J. Faaborg. 1995. Re-
gional forest fragmentation and the nesting suc-
cess of migratory birds. Science 267:1987-1990.

Roper, J. J. 1992. Nest predation experiments with
quail eggs; too much to swallow? Oikos 65:528-
530.

SAS Institute. 1990. SAS/STAT users guide. Version
6, fourth ed. SAS Institute Inc., Cary, North Car-
olina.

SAS Institute. 1994. JMP statistics and graphics
guide. Version 3.0. SAS Institute Inc., Cary, North
Carolina.

SCHAFALE, M. P. AND A. S. WEAKLEY. 1990. Classifi-
cation of the natural communities of North Caro-
lina: third approximation. North Carolina Natural
Heritage Program, Raleigh, North Carolina.

Small, M. E and M. L. Hunter. 1988.. Forest frag-
mentation and avian nest predation in forested
landscapes. Oecologia 76:62-64.

Suarez, A. V., K. S. Pfennig, and S. K. Robinson.
1997. Nesting success of a disturbance-dependent
songbird on different kinds of edges. Conserv.
Biol. 11:928-935.

Temple, S. A. and J. R. Cary. 1988. Modeling dy-
namics of habitat-interior bird populations in frag-
mented landscapes. Conserv. Biol. 2:340-347.

Saracco and Collazo • ARTIFICIAL NEST PREDATION AT EDGES

549

Turner, R. E., S. Forsythe, and N. Craig. 1981. Bot-
tomland hardwood forest land resources of the
southeastern U.S. Pp. 13-18 in Wetlands of bot-
tomland hardwood forest (J. R. Clark and J. Ben-
forado, Eds.). Elsevier Scientific Publishing Co.,
New York.

Vander Haegen, W. M. and R. M. DeGraaf. 1996.
Predation on artificial nests in forested riparian
buffer strips. J. Wildl. Manage. 60:542-550.

Wharton, C. H., W. M. Kitchens, F. C. Pendleton,
AND T. W. SiPE. 1982. The ecology of bottomland
hardwood swamps of the Southeast: a community
profile. FWS/OBS-81/37. U.S. Fish and Wildlife
Service, Biological Services Program, Washing-
ton, D.C.

Wharton, C. H., V. W. Lambour, J. Newson, P. V.
Winger, L. L. Gaddy, and R. Mancke. 1981. The
fauna of bottomland hardwoods of the southeast-

ern United States. Pp. 87-160 in Wetlands of bot-
tomland hardwood forest (J. R. Clark and J. Ben-
forado, Eds.). Elsevier Scientific Publishing Co.,
New York.

Whitcomb, R. E, C. S. Robbins, J. F. Lynch, B. L.
Whitcomb, M. K. Klimkiewicz, and D. Bystrak.
1981. Effects of forest fragmentation on avifauna
of the eastern deciduous forest. Pp. 125-205 in
Forest island dynamics in man-dominated land-
scapes (R. L. Burgess and D. M. Sharpe, Eds.).
Springer- Verlag, New York.

Winer, B. J., D. R. Brown, and K. M. Michaels.
1991. Statistical principles in experimental design.
McGraw-Hill, New York.

Yahner, R. H., T. E. Morrell, and E. S. Rachael.
1989. Effects of edge contrast on depredation of
artificial avian nests. J. Wildl. Manage. 49:508-
513.

Wilson Bull., 111(4), 1999, pp. 550-558

THE RESPONSE OF A KANSAS WINTER BIRD COMMUNITY TO
WEATHER, PHOTOPERIOD, AND YEAR

MARTIN A. STAPANIAN,' CHRISTOPHER C. SMITH, AND ELMER J. FINCK^

ABSTRACT. — We conducted a bird census along the same route nearly each week for 14 winters (194
censuses), and compared the mean number of species per station and the total number of species recorded on
the census with the length of photoperiod and weather variables. We found significant differences among winters
for both indicators of species richness. This result is consistent with previous studies in which abundance of
food was measured in the same general area. Both indicators of species richness were negatively associated with
the number of days after 1 November. This result is consistent with the hypothesis that wintering species
dependent on nonrenewed food resources lose individuals to mortality or emigration. Further, there was a positive
relationship between photoperiod and both indicators of species richness. This result is consistent with the
hypothesis that the detection of individuals in the early morning hours increases with the amount of daylight
they have available for foraging and social behaviors. Wind speed and temperature had negative and positive
relationships, respectively, to species richness. The number of species per station was greatest on days when the
ground was covered with dew and least on days when snow depth was more than 15 cm. When the “winters”
were divided into four 30-day “quarters”, most of the 61 species were recorded with equal frequency in each
quarter. Eight species were detected less frequently at the end of winter than in the beginning. Four species
exhibited the reverse pattern. Two species were recorded more frequently at the beginning and at the end of the
winter than during the middle. Temperature, wind, photoperiod, successive winter day, year, and species-specific
evolutionary history all affect winter bird species richness. Received 1 Oct. 1998, accepted 5 August 1999.

Winter is a stressful season of the year for
endotherms at mid- and high latitudes. Severe
cold, short photoperiod, and a mostly nonre-
newed food supply make it a challenge to
maintain a constant body temperature. Many
bird species migrate to more hospitable cli-
mates. For those species that overwinter at
higher latitudes, weather conditions have been
shown to affect the amount of body fat stored
(White and West 1977, Dawson and Marsh
1986, Peach et al. 1992, Waite 1992, Houston
and McNamara 1993, Rogers et al. 1994, Pi-
lastro et al. 1995). Collins (1989) provided a
short review on some of the major physiolog-
ical adaptations in birds for surviving the win-
ter. Robbins (1972, 1981a) and Altman (1983)
discussed the importance of weather condi-
tions on winter bird populations.

Although detailed, long-term winter studies
exist for specific species (Loery and Nichols

' Ohio Cooperative Fish and Wildlife Re.search Unit,
The Ohio State Univ., 1735 Neil Ave., Columbus, OF!
43210.

2 Division of Biology, Ackert Hall, Kansas State
Univ.. Manhattan, KS 66506.

’ Division of Biological Sciences, Box 4050, Em-
poria State Univ.. Emporia, KS 66801.

■* Present address: USGS/BRD. Lake Erie Biological
Station, 6100 Columbus Ave., Sandusky, OH 44870.

' Corresponding author; E-mail: cccsmith@ksu.edu

1985), data for studies of overwintering bird
communities often are collected only for a few
days per year (cf., Erskine 1992). For exam-
ple, three counts per year are made for the
Finnish winter bird census routes (Hilden
1987), and the Christmas Bird Count is an an-
nual one-day count of an area. The daily ef-
fects of weather components and the annual
effects of available food resources (e.g., mast
crop failure) on the number of species in an
area are often difficult to determine or are sta-
tistically confounded. Instructions for the
Winter Bird-Population Study (Robbins
1981b) call for a minimum of six visits per
site per year. However, daily weather data for
the study sites and analyses of the effects of
weather components on species richness are
typically lacking (Robbins 1981a). Further,
depending on species-specific responses to
abiotic factors and food abundance, bird spe-
cies may differ in their detectability during the
course of winter.

We analyzed data from bird censuses con-
ducted at nearly weekly intervals for 14 years
(194 censuses) along the same route. We in-
clude in our analyses weather data collected
from a permanent station approximately 10
km from the route. Our objectives were to (1)
quantify the effects of weather components,
photoperiod, and the cumulative number of

550

Suipcinian el al. • WEATHER EFFECTS ON WINTER BIRDS

551

winter days on species richness; and (2) to
determine if the frequency of detection of in-
dividual species changed over the course of
the winter. To accomplish the first objective,
we used a statistical procedure that accounts
for the correlation structure (i.e., time-depen-
dency) between censuses taken within each
winter. We accomplished the second objective
by testing the null hypothesis that individual
species were recorded with equal frequency
within each of four 30-day intervals during
the winter.

METHODS

Census route and field method. — Our study, like
others based on seeing and/or hearing birds to count
their presence, measured the visual and auditory de-
tectability of birds. Birds were counted with a modified
Breeding Bird Survey procedure (Robbins et al. 1986)
along a regular census route across the border between
Riley and Pottawatomie counties, Kansas (Stapanian
1982, Stapanian et al. 1994). The route consisted of
16 stations; unlike the Breeding Bird Survey routes,
the stations were not separated by regular 0.81 km in-
tervals. Instead, stations were selected to represent typ-
ical upland and riparian forest habitats with some tree
species bearing fleshy, bird-dispersed fruit in propor-
tion to their presence in the Kansas Flint Hills. Nine
stations were along one road and seven were along
another. There were eight convenient sequences in
which the 16 stations could be visited. Each of the
eight sequences of stations was used during eight con-
secutive censuses. Therefore, there was no consistent
pattern in the time after official sunrise that each sta-
tion was visited. Distances between stations on the
same road ranged from 0.3 to 1.6 km (mean = 1.0).
The nearest stations on the two roads were separated
by 13 km. Because our goal was to quantify the effects
of weather, photoperiod, and cumulative number of
winter days on species richness in the entire area, data
were pooled for all stations. Birds were identified to
species, and the number of individuals was counted
for 3 min at each station. Birds flying overhead were
included in the analysis. Censuses were conducted at
approximately weekly intervals November-February,
1982-1996. Each census began within 1 h after sunrise
and required approximately 2 h to complete. In accor-
dance with instructions for Breeding Bird Surveys
(Robbins 1981b), no censuses were conducted in fog,
steady drizzle, prolonged rain, or winds stronger than
Beaufort 3 (13-19 km/h).

In selecting stations for the census, the original cri-
terion was a wooded area with concentrations of trees
of Juniperus virginiana, Monts rubra, or Celtis occi-
dentalis that would attract frugivorous birds (Stapanian
1982). The two roads along which the stations were
spaced held a variety of habitats (Table 1 ), which af-
fected our bird censuses. At each station we visualized
a line perpendicular to the road and classified each of

TABLE 1 . Habitat type for census stops by the
number of stops at which the habitat was represented
and by the number of 90° arcs at the 16 stops that were
predominantly composed of that habitat.

Habitat

Number of
census
slops

Number of
90^" arcs

Native prairie

Cj grass pastures

Row crops

Residential and farm buildings

Dense shrub

Juniper forest

Young mixed forest

Mature mixed forest with oak

Mature mixed forest without oak

Riparian margin forest

Forest beyond crops

—

the four 90° sections thus formed as being predomi-
nantly in one category for Table 1. Thus, there are a
total of 64 sections for the 16 stations that form Table
1 . The Flint Hills area of Kansas held almost no forests
before European settlement (Axelrod 1985). Only
about 16 species of native trees have spread into the
area from the eastern deciduous forests after the con-
trol of prairie fires. Two stations were completely sur-
rounded by forest, but 14 stations had at least one 90°
section of forest holding one of the three tree species
producing fleshy fruit and the other two stations had
fence rows with M. rubra. The mature forests are sep-
arated into those with and without bur oaks {Quercus
macrocarpa) because this tree species must have a
large acorn crop in order for Red-headed Woodpeckers
(Melanerpes erynhrocephalus) to winter in the area.
Some of the residences near stations on the census
were homes with lawns while others had corrals for
livestock. At five stations birds could be heard calling
from mature forests beyond extensive fields of row
crops (Table 1 ).

Although our survey has been conducted along the
same route nearly every week since 1978 (Stapanian
et al. 1994), because weather data are not available
before 1982, we only analyzed data from November
1982 through February 1996. We .selected the period
between 1 November through 28 February because it
represents a time interval in the study area during
which (1) food sources are not renewed and (2) Neo-
tropical migrants are rarely present. We divided this
interval into four 3()-day periods (quarters) for analysis
of the presence of individual species. Our censuses
were designed to monitor populations of upland birds
(Stapanian 1982, Stapanian et al. 1994). Aquatic and
nocturnal species were eliminated from the present
analysis. For each census, we calculated the mean
number of species recorded per station and the total
number of species recorded from all stations.

Our procedures differed from Breeding Bird Sur-
veys in three ways. First, when no birds were evident

552

THE WILSON BULLETIN • Vol. 111. No. 4. December 1999

at a station, we spished to attract them. Second, when
we were unable to find new birds where the car was
parked, we walked along the road in search of birds.
Third, we had more than one observer on 58% of the
censuses. Neither of the first two differences biased the
data. Making noise and walking along the road when
no birds were evident would tend to overestimate the
number of bird species and individuals when they were
lowest. Thus, any conclusions we would make about
which factors decreased bird activity and the number
of species would be conservative. The number of ob-
servers ranged from one (82 censuses, 42. 1 %) to four
(3 censuses, 1.5%). The number of censuses in which
there were two and three observers were 81 (41.5%)
and 29 (14.9%), respectively. In exploratory analyses,
we found that the number of species recorded was
greater when more than one observer participated in
the census. Therefore, we adjusted mean species per
station and total species per census for the number of
observers. In controlled experiments performed during
winter on this route (C. C. Smith, unpubl. data), we
found that the mean number of species per station and
total species per census increased on average by fac-
tors of 1.32 and 1.08, respectively, for multiple ob-
servers over those values found by one observer. Thus,
when the number of observers was greater than 1, we
divided mean species per station and total species per
census by 1.32 and 1.08, respectively. Further, C.C.S.
participated in all censuses and his hearing still allows
him to detect a Brown Creeper (Certhia americana) at
30 m. E.J.E participated in almost all censuses from
1982 through February 1989. J. Cavitt, S. Hansen, S.
Hull, C. Pacey, G. Radke, and C. Rebar participated
in at least four censuses each.

Weather data. — Weather data for each census were
collected automatically from a permanent station at the
Konza Prairie Research Natural Area, located within
18 km from all our census stations. The weather station
measured wind speed at hourly intervals on the hour.
We, therefore, selected weather data recorded at 07:00
on each census day. Because each cen.sus began within
1 h after official sunrise, 07:00 does not represent a
standard time relative to sunrise for all censuses. How-
ever, we were confident that the data were represen-
tative of the weather during our censuses.

We use a standard weather service formula to con-
vert temperature and wind speed to a wind chill tem-
perature. Wind chill temperature exceeded air temper-
ature only for wind speeds greater than 6.7 km/h,
which occurred on only six cen.suses. In exploratory
analyses of variance, we found that of the weather var-
iables recorded, only temperature and wind speed ac-
counted for a significant proportion of the variance in
our .statistical models.

We ranked ground conditions from 1 though 6 ac-
cording to what we perceived as increased difficulty
for birds in finding food on the ground: (1) dry, (2)
dew, (3) frost, (4) wet from rain or melting snow, (5)
snow 15 cm or le.ss deep, and (6) snow more than 15
cm deep. Ground condition was recorded at the first
station we visited on all but six censuses. All stations

were then assigned the same weather data and ground
condition class as the first station for the census.

Statistical analyses. — “Winter day” was designated
as the number of days after 31 October for each cen-
sus. Photoperiod was calculated from published tables
(U.S. Naval Observatory 1945) as the number of min-
utes between official sunrise and official sunset on
each census date, and ranged from 565 to 679 min.

We tested for the effects of winter day, photoperiod,
and weather components on our two indicators of spe-
cies richness. In exploratory analyses and previous
studies (Stapanian et al. 1994) we found considerable
variance in the species richness among winters. Fur-
ther, we found significant time dependency among the
successive censuses within winters. Therefore, we used
a mixed models procedure (Crowder and Hand 1990,
SAS Institute 1992, Littell et al. 1996) in which each
winter was treated as a random effect (i.e., whole plot),
and the remaining variables were treated as fixed co-
variates (i.e., subplots) to account for the correlation
structure among the censuses within winters. The de-
grees of freedom and mean squares were adjusted for
time dependency based on the covariance structure and
inference space. Ecologically, this meant that we re-
moved winters as a random effect from the statistical
model tests for the effects of the fixed covariates based
on an average set of conditions at the beginning of
winter. The resulting model was general, not winter-
specific.

We evaluated the covariance structure in three ways:
(1) uniform correlation (compound symmetry), (2) ex-
ponentially decaying, and (3) Markov chain. In ex-
ploratory analyses, we found that the uniform corre-
lation method best represented the covariance structure
of the data set. Further, we found that none of the two-
way interactions between the fixed covariates contrib-
uted significantly to the models {P > 0.05 in all cases).
Thus, we performed the mixed models analyses only
on the main effects of the fixed covariates. We per-
formed Tukey’s tests for a posteriori testing on the
effects of specific ground condition classes on diver-
sity.

We defined ordinal year as the ordinal number of a
census year (i.e., year 1 = 1982-1983, year 2 = 1983-
1984, . . . , year 14 = 1995-1996). We performed stan-
dard Pearson correlations between ordinal year, species
richness, and our weather variables. In this manner, we
were able to test for overall temporal trends in weather
and diversity on census days on our census route.

For each species, we calculated the proportion of the
censuses conducted in each quarter (30-day interval)
of each year in which that species was recorded (Ap-
pendix). These quarterly proportions were then pooled
across all 14 winters for each species. Using analysis
of variance, we then tested the null hypothesis that
each species was recorded in equal proportions in all
four quarters. Tukey’s pair-wi.se comparisons were
used for all a posteriori testing. SAS for Personal
Computers, version 6.12 for Windows was used for
statistical computations.

Stapauian el cil. • WEATHER EFFECTS ON WINTER BIRDS

553

TABLE 2. Summary statistics for each winter for the total number of species recorded on each census and
the mean number of species per station on each census. These variables were adjusted for number of observers.
Years that share a grouping letter were not significantly different (Tukey’s pair-wise comparisons, P > 0.05) for
that indicator of diversity.

Species per census

Species per station

Year

IV'

Mean

SD*’

Mean

SD*’

1982-1983

24.18

2.29 B, C, D

5.00

0.74 A, B, C

1983-1984

18.21

4.97 E

3.52

1.54 D, E

1984-1985

19.48

2.65 D, E

3.99

1.02 C, D

1985-1986

23.40

3.32 B, C

5.04

1.44 A, B, C

1986-1987

1 1

22.09

3.67 B, C, D, E

4.89

1.02 A, B, C, D

1987-1988

23.73

2.59 B, C

5.50

1.18 A, B

1988-1989

24.45

2.07 B, C

5.37

0.91 A, B

1989-1990

23.75

4.16 B, C

4.98

0.59 A, B, C

1990-1991

25.93

2.76 B

5.86

1.09 A

1991-1992

18.39

3.26 E

3.65

1.25 D

1992-1993

23.71

2.99 B, C

4.97

1.07 A, B, C

1993-1994

29.81

3.67 A

5.36

1.12 A, B

1994-1995

24.00

2.16 B, C

5.34

1.36 A, B

1995-1996

22.32

1.37 C, D

4.63

0.95 B, C, D

^ n = number of censuses.
^ SD = standard deviation.

RESULTS

There were significant differences among
winters for the annual means of both total spe-
cies per census and species per station (F,3 ,go
= 13.87 and 5.46, respectively, P < 0.001 in
both cases; Table 2). Consequently, we treated
winters as random effects in our mixed model
analysis. The results from the mixed model
procedure (Table 3) suggested significant ef-
fects from winter day, photoperiod, tempera-
ture, and wind speed for the number of species
per census and species per station. Tempera-
ture and photoperiod were positively related
to both indicators of species richness when the
covariance structure was taken into account
(slopes in Table 3). On the other hand, wind
speed and winter day had negative effects on
both indicators of diversity (slopes in Table 3).

On average, a change of 1° C in temperature
or 1 km/h in wind speed had a greater effect
on species richness than did either a change
of 1 min in photoperiod or 1 day further into
winter.

Ground condition had a significant effect on
species per station, but not on species per cen-
sus in the mixed model analyses (Table 3).
Values of species per station were lowest
when there was more than 15 cm of snow on
the ground and greatest when the ground was
covered with dew (Table 4).

Both of our indicators of species richness
increased over the course of our study. There
was weak but positive correlation between
species per census and ordinal year (r = 0.27,
df = 12, P < 0.001) and between species per
station and ordinal year (r — 0. 15, df = 12,

TABLE 3. Results of the mixed models ANOVA procedure. Two-way interactions were not found to be
significant in exploratory analyses (P > 0.05). Slopes and standard errors (SE) of the slopes are not reported
for ground condition because it was not a continuous variable.

Specie.s per census Species per station

Source tlf P > F Slope SE /•' P > F Slope SE

Winter day (days)

1 20.35

0.001

Photoperiod (minutes)

1 11.51

0.0009

Temperature (°C)

1 7.65

0.0063

Wind speed (km/h)

1 26.36

0.0001

Ground condition

5 1.10

0.3636

0.033

0.007

21.89

0.0001

-0.01 1

0.002

0.029

0.008

40.03

0.0001

0.017

0.003

0.1 1 1

0.040

30.53

0.0001

0.070

0.013

0.634

0.123

45.86

2.54

0.0001

0.0301

-0.264

0.039

554

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

TABLE 4. A posteriori tests (Tukey’s pair-wise
comparisons) on the effects of ground condition on the
mean values of the mean number of species per station
on a census. Means of ground condition classes with
at least one letter in common are not signihcantly dif-
ferent (P > 0.05).

Ground

condition

Class

Mean of mean
species per
station

Clear, dry

4.93 B

Dew

6.12 A

Erost

5.14 B

Wet

4.75 B

Snow < 15 cm

4.47 B

Snow > 15 cm

3.56 C

^ n = number of censuses.

P = 0.035).

Further,

there was

a weak but

negative correlation between wind speed and
ordinal year (r = —0.16, df = 12, P = 0.022),
which suggested that wind speed on census
trips decreased over the course of this study.
We made no conscious change in our policy
of when to hold censuses during our study that
would have resulted in lower wind speeds dur-
ing censuses. Neither temperature nor ground
condition class was significantly correlated
with ordinal year (P > 0.05 in both cases).

Sixty-one species were recorded for our
study (Appendix). Only two species, Ameri-
can Crow (Corvus brachyrhynchos) and
Black-capped Chickadee (Poecile atricapil-
lus), were recorded on all 194 censuses.
Twenty-two species (36.1% of the total spe-
cies) were recorded on at least 50% of the
censuses. Thirteen species (21.3% of the total
species) were seen on less than 5% of the cen-
suses. There was no consistent pattern to the
temporal occurrences of individual species
(Appendix). For most species, the proportion
of the censuses in which they were recorded
was the same for each quarter. Eight species.
Northern Flicker (Colaptes ciurcitus). Golden-
crowned Kinglet (Regulus satrapa). Northern
Bobwhite (Colinus virginianus). White-
crowned Sparrow (Zonotrichio leucophrys).
Common Crackle (Quiscalus quiscula). Field
Sparrow (Spizella piisiUa), White-throated
Sparrow (Zonotrichia alhicollis), and Lin-
coln’s Sparrow (Melospiz.o lincolnii), occurred
more frequently in early winter than in late
winter. The reverse trend was exhibited by
four species, American Tree Sparrow (Spizella

arborea). Tufted Titmouse (Baeolophus bicol-
or), Western Meadowlark (Sturnella neglec-
ta), and Eastern Meadowlark (Sturnella mag-
na). Two species. Red-winged Blackbird
(Agelaius phoeniceus), and Ring-necked
Pheasant (Phasianus colchicus), were record-
ed less frequently in mid-winter (i.e., in the
second and third quarters) than at the begin-
ning or end. Species observed on fewer than
5% of the censuses were not considered com-
mon enough to test for patterns of occurrence
by winter quarter (Appendix).

DISCUSSION

Our study is the first of which we are aware
that demonstrated significant effects of pho-
toperiod, cumulative number of winter days,
and weather components on bird species rich-
ness of upland and riparian forest birds in
winter. The results appear to differ consider-
ably from those of Robbins (1981a). He de-
tected no effects of weather conditions on the
numbers of selected species or families of
birds from repeated coverage of a Winter Bird
Survey route. Robbins’ 8-km route in Mary-
land was covered at least three times per year
for five consecutive years in late December or
early January. He also analyzed data from
eight years of Audubon Winter Bird-Popula-
tion Studies on two forest plots in Maryland.
There were no significant effects of tempera-
ture on the number of species he recorded.
The differences between our results and those
from Robbins may be due to (1) our larger
sample size, (2) a longer season (i.e., Novem-
ber through February) in our study, (3) the
fact that Robbins’ (1981a) analyses were re-
stricted to selected species and families, or (4)
differences in location and climate. Most im-
portantly, Robbins (1981a) selected for calm,
dry mornings in both studies. Therefore, it is
not surprising that he reported no weather ef-
fects on the number of species recorded.

Although ours is a long-term study, there
were too few censuses to analyze, the effects
of number of winter days, photoperiod, and
weather on species richness for specific win-
ters. The data strongly suggest that differences
in the detection of bird species occurred
among winters. Previously, we (Stapanian et
al. 1994) estimated extremely low seed crops
for weeds, herbs, grasses, and bur oak for the
winters of 1983-1984 and 1984-1985. Simi-

Stapanum et cil. • WEATHER EEFECTS ON WINTER BIRDS

555

lady, we estimated extremely low weed seed
and wild fleshy fruit crops for the winter of
1991—1992. These low food supplies may
partly explain why the fewest species were re-
corded in those years. Large crops of herb and
grass seeds were estimated for the winter of
1982-1983, which had relatively high values
for species per station. Similarly, there were
large crops of acorns and fleshy fruits for
1988-1989 and 1990-1991. In both winters,
species richness was relatively high. These
trends support the importance of the size of
unrenewed food supplies in determining the
detection of winter bird populations.

By treating the large winter differences as
random effects, the statistical analysis dem-
onstrated that photoperiod affects the morning
activity of birds. The influences of photope-
riodism on the physiology and activity of
birds are well documented (Bissonette 1932,
1937; Bartholomew 1949; Welty and Baptista
1988 and references therein; Ball 1993; Hau
et al. 1998). Perhaps when less time is avail-
able for feeding, as in mid-winter, birds spend
less time in easily detected behaviors. The be-
havioral effect of reduced feeding time rela-
tive to energy needs is likely to be larger flock
size (Caraco 1979, Sullivan 1988) and a lower
probability of seeing birds at the average sta-
tion.

Temperature and ground condition classes
were significantly and negatively correlated (r
—0.37, P < 0.001, n = 187). Thus, what
we perceived as difficult foraging conditions
might have been simply a consequence of low
temperature. Dew (ground condition class 2)
and ground wet from rain or snowmelt (class
4) required that the air temperature exceed
0° C, while the temperature may be below
0° C for dry ground (class 1) and will be for
frost (class 3). When we switched the number
class of frost to 4 and rain or snowmelt to 3,
the correlation coefficient between tempera-
ture and ground condition class increased in
absolute magnitude (r = —0.508, P < 0.001).
Ground condition may have little effect in-
dependent of temperature.

We are not sure how to interpret the nega-
tive effects of wind speed on bird species per
census or bird species per station. Wind speed
had a negative effect on both indicators of
species richness even when we considered
only those censuses in which wind speeds

were less than 6.7 km/h, the speed above
which wind chill temperature is less than air
temperature. The effects of wind speed on
species richness appear to be due to neither a
decrease in our ability to hear birds, nor ap-
parent additional thermoregulatory stress for
the birds. However, wind speed typically in-
creases after sunrise, and the wind speed at
the end of a census may be greater than at the
beginning. Small differences in wind speed at
07:00 may be magnified later in the census.
There is evidence that some species, particu-
larly those with small body sizes, can reduce
metabolic demands in winter by selecting mi-
crohabitats that are sheltered from the wind
and exposed to solar radiation (Wolf and
Walsberg 1996).

Similarly, we are unsure why both indica-
tors of species richness increased in later
years. The same principal observer (C.C.S.)
was present for all censuses in our study. Eye-
sight and hearing typically deteriorate over
time (Cyr 1981), but these effects can be
countered by individual experience with a
specific route. The increases in species rich-
ness were not due to changes in mean annual
temperature, because temperature and ordinal
census year were not significantly correlated
(r = 0.109, df = 12, P > 0.05). Wind speed
on the census trips was negatively correlated
with year and with both indicators of species
richness. Thus, a decrease in wind speed on
census days may partially explain the increas-
es in species per station and species per cen-
sus over the census years. Species composi-
tion on the census route changed over time.
For example, the population of Carolina Wren
(Thryothorus ludovicicvms), a sedentary bird
species, increased steadily during the 14 years
of the census after a time when it was at low
levels in the Manhattan Christmas bird counts.
Wild Turkeys (Meleagris gallopavo) were re-
introduced in the area in the early 1980s and
their populations have increased since. These
changes may be due to milder winters during
the study period. There may have been an in-
crease in canopy closure or structural diversity
of the habitat on the route over the 14 years
of the study, but that was not measured.

White-crowned Sparrows and White-throat-
ed Sparrows feed in large mixed flocks of
sparrows in late winter and in smaller groups
in late fall. These species were recorded on a

556

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

greater proportion of the censuses in the first
quarter than in any other. November (first
quarter) typically has the mildest average
weather of any quarter and food abundance
should be greatest at this time. Therefore, the
selective pressure on birds for being in large
flocks should be least at this time (Caraco
1979, Sullivan 1988). If members of a species
were more widely distributed geographically,
an observer would be more likely to record
that bird species at least once on a census.
Field Sparrows, Lincoln’s Sparrows, and
Common Crackles were also seen with greater
frequency in November than in late winter.
The Kansas Flint Hills are along the northern-
most edge of the winter range of these three
species (Thompson and Ely 1992). Although
most members of these species migrate south
out of the census area, a few overwinter (Zim-
merman 1993). The decrease in numbers for
these species during the winter could be a re-
sult of continued migration south or mortality
in a marginal range.

The proportions of censuses on which
Northern Bobwhite (a resident species). Gold-
en-crowned Kinglet (a winter migrant), and
Northern Flicker (winter resident and winter
migrant; Thompson and Ely 1989, 1992; Zim-
merman 1993) were recorded dropped steadily
from the second through the fourth quarter.
This decline may have been due to mortality
to the wintering populations of those species.
The results from our study for Golden-
crowned Kinglet agreed with those of Zim-
merman (1993) who hypothesized that the
variation in departure of Golden-crowned
Kinglets from the area was related to avail-
ability of food.

Eastern Meadowlark, Western Meadowlark,
and Tufted Titmouse were recorded most fre-
quently on censuses in the last quarter. These
are resident species (Thompson and Ely 1992,
Zimmerman 1993) that breed early in spring
and begin establishing territories and/or ob-
taining mates in late winter. The American
Tree Sparrow, a winter migrant to the area
(Zimmerman 1993), was recorded least fre-
quently in the first quarter. This agrees with
Finck (1986), who found this species to be
most numerous from December through Feb-
ruary, suggesting a late migratory arrival.
Red-winged Blackbird and Ring-necked
Pheasant were most often recorded in Novem-

ber and February. These are resident or partly
resident species that flock in severe weather
in mid-winter. However, they begin prepara-
tion for breeding in late winter (Zimmerman
1993). Zimmerman (1993) found Red- winged
Blackbird to be “occasional” during the win-
ter months of most years in upland habitats
until the migrants returned in late winter.

Our results are consistent with at least five
hypotheses: (1) species are lost by mortality
resulting from nonrenewed resources over the
course of winter, (2) resident species move in
and out of detection distance in the census
area, (3) selective pressures for flock sizes
change with weather conditions and food
abundance, (4) some species are more easily
detected in late winter because of early court-
ship behavior, and (5) species richness in the
census area changes as a result of the arrival
and departure of seasonally migrant species.
The results suggest a complex relationship
among weather components, photoperiod,
abundance of resources, and species-specific
evolutionary histories on winter bird species
richness. We suggest further studies to analyze
responses by individual species to resource
abundance and abiotic factors in winter.

ACKNOWLEDGMENTS

R. Leighty provided assistance with statistical ana-
lyses. C. Robbins, T Grubb, and J. Harder commented
on earlier drafts of the paper. Weather data were col-
lected as part of the NSF Long Term Ecological Re-
search Program at Konza Prairie Research Natural
Area. The USGS/BRD Ohio Cooperative Fish and
Wildlife Research Unit provided support to work on
this manuscript.

LITERATURE CITED

Altman, R. 1983. The effects of time of day on a
winter bird survey. Ky. Warbler 59:23-26.
Axelrod, D. I. 1985. Rise of the grassland biome.
Central North America. Bot. Rev. 51:163-201.
Ball, G. E 1993. The neural integration of environ-
mental information by seasonally breeding birds.
Am. Zool. 33:185-199.

Bartholomew, G. A. 1949. The effect of light inten-
sity and day length on reproduction in the English
Sparrow. Bull. Mus. Comp. Zool. 101:433-476.
Blssonnette, T. H. 1932. Light and diet as factors in
relation to sexual photoperiodicity. Nature (Lond.)
129:612.

Bissonnette, T. H. 1937. Photoperiodicity in birds.
Wilson Bull. 49:241-270.

Caraco, T. 1979. Time budgeting and group size: a
test of theory. Ecology 60:618-627.

StapanUm et al. • WEATHER EFFECTS ON WINTER BIRDS

557

Collins, R T. 1989. Surviving the winter; the physi-
ology of thermoregulation in birds. Passenger Pi-
geon 51:315-320.

Crowder, M. J. and D. J. Hand. 1990. Analysis of
repeated measures. Monogr. Stat. Appl. Probabil-
ity 41:1-257.

Cyr, a. 1981. Limitation and variability in hearing
ability in censusing birds. Stud. Avian Biol. 6:
327-333.

Dawson, W. R. and R. L. Marsh. 1986. Winter fat-
tening in the American Goldfinch and the possible
role of temperature in its regulation. Physiol. Zool.
59:357-368.

Erskine, a. j. 1992. A ten-year urban winter bird
count in Sackville, New Brunswick. Can. Field-
Nat. 106:499-506.

Finck, E. j. 1986. Birds wintering on the Konza Prairie
Research Natural Area. Proc. N. Am. Prairie Conf.
9:91-94.

HaU, M., M. WlKELSKI, AND J. C. WiNGFlELD. 1998. A
Neotropical bird can measure the slight changes
in tropical photoperiod. Proc. R. Soc. London B
265:89-95.

Hidden, O. 1987. Finnish winter bird censuses: long-
term trends in 1956-1984. Acta Oecol. 8:157-
168.

Houston, A. I. and J. M. McNamara. 1993. A theo-
retical investigation of the fat reserves and mor-
tality levels of small birds in winter. Ornis Scand.
24:205-219.

Littell, R. C., G. a. Milliken, W. W. Stroup, and
R. D. WoLFiNGER. 1996. SAS systems for mixed
models. SAS Institute, Inc., Cary, North Carolina.

Loery, G. AND J. D. Nichols. 1985. Dynamics of a
Black-capped Chickadee population, 1958-1983.
Ecology 66:1195-1203.

Peach, W. J., D. P. Hodson, and J. A. Fowler. 1992.
Variation in the winter body mass of starlings
Sturnus vulgaris. Bird Study 39:89-95.

PiLASTRO, A., G. BERTORELLE, AND G. Marin. 1995.
Winter fattening strategies of two passerine spe-
cies: environmental and social influences. J. Avian
Biol. 26:25-32.

Robbins, C. S. 1972. An appraisal of the winter bird-
population study technique. Am. Birds 26:1-4.

Robbins, C. S. 1981a. Bird activity levels related to
weather. Stud. Avian Biol. 6:301-310.

Robbins, C. S. 1981b. Reappraisal of the winter bird-
population study technique. Stud. Avian Biol. 6:
52-57.

Robbins, C. S., D. Bystrak, and P. H. Geissler. 1986.
The breeding bird survey; its first fifteen years,
1965-1979. Spec. Sci. Rep.-Wildl. 157:1-196.

Rogers, C. M., V. Nolan, Jr., and E. D. Ketterson.
1994. Winter fattening in the Dark-eyed Junco:
plasticity and possible interaction with migration
trade-offs. Oecologia 97:526-532.

SAS Institute, Inc. 1992. SAS/STAT software:
changes and enhancements. Release 6.07. SAS In-
stitute, Inc., Cary, North Carolina.

Stapanian, M. a. 1982. Evolution of fruiting strate-
gies among fleshy-fruited plant species in eastern
Kansas. Ecology 63:1422-1431.

Stapanian, M. A., C. C. Smith, and E. J. Finck. 1994.
Population variabilities of bird guilds in Kansas
during fall and winter: weekly censuses versus
Christmas bird counts. Condor 96:58-69.

Sullivan, K. A. 1988. Ontogeny of time budgets of
Yellow-eyed Juncos: adaptation to ecological con-
straints. Ecology 69:118-124.

Thompson, M. C. and C. Ely. 1989. Birds in Kansas;
vol. I. Univ. of Kansas, Univ. Press of Kansas,
Lawrence.

Thompson, M. C. and C. Ely. 1992. Birds in Kansas:
vol. II. Univ. of Kansas, Univ. Press of Kansas,
Lawrence.

U.S. Naval Observatory. 1945. Supplement to the
American Ephemeris, 1946. Tables of sunrise,
sunset, and twilight. U.S. Government Printing
Office, Washington, D.C.

Waite, T. A. 1992. Winter fattening in Gray Jays: sea-
sonal, diurnal, and climatic correlates. Ornis
Scand. 23:499-503.

Welty, j. C. and L. Baptista. 1988. The life of birds,
fourth ed. Saunders College Publishing, New
York.

White, C. M. and G. C. West. 1977. The annual lipid
cycle and feeding behavior of Alaskan Redpolls.
Oecologia 27:227-238.

Wolf, B. O. and G. E. Walsberg. 1996. Thermal ef-
fects of radiation and wind on a small bird and
implications for microsite selection. Ecology 77:
2228-2236.

Zimmerman, J. L. 1993. The birds of Konza. Univ.
Press of Kansas, Lawrence.

558

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

APPENDIX

The number of all censuses (All) and the mean proportion of censuses in 30-day intervals over 14 winters in
which species were observed. For each species, we tested the null hypothesis that it was recorded in equal
proportions in all four quarters. A posteriori tests (Tukey’s pair-wise comparisons) were performed on those
species for which the null hypothesis was rejected. Quarterly means for a species having the same letter were
not significantly (P < 0.05) different. Species listed at the bottom lacked significant quarterly differences.
Abbreviations: All = all censuses combined, n = number of censuses.

Common name

Scientific name

All

n = 194

1-30 Nov
n = 52

1-30 Dec
n = 42

31 Dec-29 Jan
n = 46

30 Jan-28 Feb
n = 54

Northern Flicker

Colaptes aiiratiis

171

0.962 A

0.986 A

0.873 AB

0.768 B

American Tree Sparrow

Spizella arborea

157

0.562 B

0.958 A

0.962 A

0.857 A

Tufted Titmouse

Baeolophus bicolor

153

0.673 B

0.642 B

0.902 A

0.929 A

Red-winged Blackbird

Agelaiiis phoeniceus

143

0.926 A

0.626 BC

0.579 C

0.831 AB

Western Meadowlark

Sturnella neglecta

0.576 AB

0.318 C

0.396 BC

0.737 A

Golden-crowned Kinglet

Regulus satrapa

0.604 A

0.669 A

0.321 B

0.127 B

Northern Bobwhite

Colinus virginianus

0.382 AB

0.461 A

0.364 AB

0.211 B

White-crowned Sparrow

Zonotrichia leucophrys

0.385 A

0.095 B

0.143 B

0.157 B

Ring-necked Pheasant

Phasianus colchicus

0.240 A

0.183 AB

0.014 B

0.294 A

Common Crackle

Quiscalus quiscula

0.308 A

0.060 B

0.024 B

0.071 B

Eastern Meadowlark

Sturnella magna

0.070 B

0.018 B

0.089 B

0.205 A

Field Sparrow

Spizella pusilla

0.270 A

0.071 B

0.056 B

0.032 B

White-throated Sparrow

Zonotrichia albicollis

0.173 A

0.060 AB

0.018 B

Lincoln’s Sparrow

Melospiza lincolnii

0.130 A

0.065 AB

0.000 B

0.050 AB

American Crow (Corvus brachyrhynchos) 194, Black-capped Chickadee (Poeciie alricapillus) 194, Dark-eyed Junco (Junco hyemalis) 193, Northern
Cardinal (Cardinalis cardinalis) 192, House Sparrow (Passer dome sticus) 192, Blue Jay (Cyanocilla cristala) 190, American Goldfinch (Carduelis iristis)
187, White-breasted Nuthatch (Silla carolinensis) 187, Red-bellied Woodpecker (Melanerpes carolinus) 184, Downy Woodpecker (Picoides pubescens)
170, European Starling (Slurnus vulgaris) 169, American Robin (Tiirdus migralorius) 167, Red-tailed Hawk (Buteo jamaicensis) 157, Hairy Woodpecker
(Picoides villosus) 143, Harris’s Sparrow (Zonotrichia querula) 136, Eastern Bluebird (Sialia sialis) 128, Carolina Wren (Thryothorus ludovicianus) 97,
Red-headed Woodpecker (Melanerpes erythrocephalus) 92, Rock Dove (Columba livia) 81, Song Sparrow (Melospiza melodia) 68, American Kestrel
(Falco sparverius) 49, Brown Creeper (Cerlhia americana) 47, Northern Harrier (Circus cyaneus) 41, Cedar Waxwing (Bombycilla cedrorum) 40, Bewick’s
Wren (Thrvomanes bewickii) 36, Spotted Towhee (Pipilo maculalus) 33, Wild Turkey (Meleagris gallopavo) 29, Mourning Dove (Zenaida macroura) 27,
Pine Siskin (Carduelis pinus) 23, Brown-headed Cowbird (Molothrus ater) 20, Yellow-bellied Sapsucker (Sphyrapicus varius) 16, Winter Wren (Troglodytes
iroglodvles) 16, Sharp-shinned Hawk (Accipiter sirialus) 14. Yellow-rumped Warbler (Dendroica coronata) 1 1, Loggerhead Shrike (Lanius ludovicianus)
7. Rough-legged Hawk (Buteo lagoputs) 6, Horned Lark (Eremophila alpestris) 6, Rusty Blackbird (Euphagus carolinu.s) 6, Cooper’s Hawk (Accipiter
cooperii) 4. Purple Finch (Carpodacus purpureus) 3, Prairie Falcon (Falco mexicanus) 3, Fox Sparrow (Passerella iliaca) 3, Sedge Wren (Cistothorus
platensis) 2, Red-breasted Nuthatch (Sitta canadensis) 2, Brown Thrasher (Toxostoma rufum) 2. Hermit Thrush (Catharus guttatus) 1. Yellow-headed
Blackbird (Xanthocephalus .xanthocephalus) 1.

Short Communications

Wilson Bull., 1 1 1(4), 1999, pp. 559-560

Possible Winter Quarters of the Aleutian Tern?

Norman P. Hill' and K. David Bishop-^

ABSTRACT. — Recent observations of the Aleutian
Tern {Sterna aleutica) in the coastal waters around
Hong Kong in spring and fall, and Singapore and the
Indonesian islands of Karimun and Bintan between
October and April indicate that at least part of the pop-
ulation of this species migrates through and winters in
these areas. Our observations during December 1997,
suggest that the coastal waters of Java, Bali and Su-
lawesi may form an additional part of the winter range
of this species. Received 3 August 1998, accepted 17
June 1999.

The Aleutian Tern {Sterna aleutica) breeds
along the western coast of Alaska (USA) and
in Asia on the east coast of Kamchatka and
Sakhalin (American Ornithologists’ Union
1998). In Alaska, birds return to colonies dur-
ing early May and then disperse during Au-
gust and September after breeding, (Harrison
1983). Both the AOU Checklist (1998) and
Harrison (1983) state that this species’ winter
range is “unknown”.

Daring the last decade or so there has been
a steady accumulation of records for this spe-
cies outside its breeding range from Southeast
Asia. Lee (1992) reported six specimens col-
lected in May 1984 in the Mindanao Sea off
Bohol, Philippines. Brazil (1991) noted ap-
proximately ten records from Honshu and
Hokkaido, Japan including one instance of
probable breeding. One exceptional record in-
volves the occurrence of a single vagrant bird
on the Fame Islands off the coast of northeast
England during May 1979 (Dixey et al. 1981).
More recently, during August and September
1992 as many as 190 birds were observed off
the southern and southeastern coast of Hong
Kong (Kennerley et al. 1993).

Initially birds were observed in breeding

' 38 North Main Street, Assonet, MA 02702.

2 ‘Semioptera’, RO. Box 6068, Kincumbcr, NSW
2251, Australia.

3 Corresponding author; E-mail: kdbishop@ozemail.
com.au

(alternate) plumage in late August but most
then molted into non-breeding (basic) plum-
age during September. Details of the latter,
poorly known plumage, can be found in Lee
(1992), Kennerley and coworkers (1993) and
Kennerley and Ollington (1998). Subsequent-
ly, small numbers have been observed annu-
ally in Hong Kong waters including individ-
uals in breeding plumage during April-June.
It is now established there as a regular and
fairly common migrant in varying numbers.
Several hundred migrants have also been re-
corded annually in August and September,
with occasional birds recorded in October
(Leven et al. 1994; Carey et al. 1995, 1996).

From September to October 1994 this spe-
cies was common within the Riau Archipela-
go, Indonesia, and a single bird was also re-
corded there in March 1996 (Rajathurai 1996).
Kennerley and Ollington (1998) observed
small numbers between 18 September and 25
April in the Straits of Malacca and in the seas
around the island of Grand Karimun with a
maximum number of 15 positively identified
birds on 19 January 1996, with more than 100
distant terns also present.

Small numbers of Aleutian Terns were also
observed off the east coast of Singapore in
September and October 1994 (Kennerley and
Ollington 1998). No records were reported
during 1995 but a concerted effort to locate
this species in 1996 resulted in at least 15
birds sighted 13 October in the Straits of Sin-
gapore between Jurong and the Horsburgh
Light (Kennerley and Ollington 1998).

We report here the first probable records
from Java, Bali, and Sulawesi in the Republic
of Indonesia and identify a likely proportion
of this species’ wintering range.

On 30 November 1997 between Labuan
and the north coast of Ujung Kulon National
Park, at the western tip of Java, from a boat
we observed, about 20 probable adult Aleu-
tian Terns in groups of 3-6, all in non-breed-

559

560

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

ing plumage. We obtained good views of these
birds in flight and perched as they fished to-
gether with other seabirds among and from the
numerous fish-trap platforms. Similarly, on 4
December 1997 we also observed, from a boat
a flock of approximately 35 probable Aleutian
Terns during the late afternoon. They were
feeding just off the west coast of Bali, within
the maritime boundary of Bali Barat National
Park. On 13 December 1997 we observed a
group of 15 probable Aleutian Terns and
Crested Terns {Sterna bergii) fishing at the
mouth of the Bone River, Gorontalo, Sula-
wesi. At the same time, about 200 unidentified
terns were feeding offshore along the interface
between the muddy river water and the clear
sea.

We observed the flocks closely for 10-30
minutes, as two of the flocks of the birds often
fed close to our boat. N.P.H. is familiar with
the species on its breeding grounds and has
observed it twice in Alaska and once in Si-
beria. Both of us are familiar with the black-
billed race of Common Tern {Sterna hirundo
longipennis), which the Aleutian Tern most
resembles. All the birds we saw appeared to
be adults in winter or near winter plumage
with a sharply defined pattern of extensive
white forehead, with white extending onto the
crown of some individuals creating a bald-
headed appearance. Some individuals had a
black nape with whitish streaking in the upper
edge of the black nape (for an example see
photographs 6 and 10 in Kennerley and Ol-
lington 1998). The wings and back were dark
grey with a contrasting pale rump and tail.
The bill was black on all the birds as were the
legs of perched birds off the west coast of
Java. Flight appeared to be slower than that
of a Common Tern and feeding was accom-
plished more by dipping, similar to Chlidon-
ias terns or the tropical Sooty and Bridled
terns {S. fuscata and S. anaethetus). Unfortu-
nately we did not take note of the underwing
pattern which Kennerley and Ollington (1998)
and Kennerley (pers. comm.) demonstrate to
be a diagnostic field identification character
for the Aleutian Tern in winter plumage.

While our records of the Aleutian Tern in
Java, Bali and Sulawesi are not conclusive, we
have presented our records here in order to
draw attention to a possible wintering area for
this species and to encourage field workers
visiting Southeast Asia and Wallacea to be
more diligent in their examination of tern
flocks, especially from September to April.
Our records of a probable Aleutian Tern in
Indonesia together with previously published
observations of this species elsewhere in
Southeast Asia suggest that a significant pro-
portion of the western Pacific population of
Aleutian Tern migrates along the southern
coast of China and Southeast Asia to winter
in the islands of Indonesia and possibly the
Philippines.

ACKNOWLEDGMENTS

We are most grateful to P. Kennerley for his invalu-
able comments on our manuscript.

LITERATURE CITED

American Ornithologists’ Union. 1998. Checklist of
North American birds, seventh ed. American Or-
nithologists’ Union, Washington, D.C.

Brazil, M. A. 1991. The birds of Japan. Christopher
Helm. London, U.K.

Carey, G. J., D. A. Diskin, P. J. Leader, M. R. Leven,
R. W. Lewthwaite, M. L. Chalmers, P. R. Ken-
nerley, AND V. B. PiCKEN. 1996. Systematic list.
Hong Kong Bird Report 1995:45.

Carey, G. J., D. A. Diskin, V. B. Picken, and P. J.
Leader. 1995. Systematic list. Hong Kong Bird
Report 1994:53.

Dixey, a. E., a. Ferguson, R. Heywood, and A. R.
Taylor. 1981. Aleutian Tern on the Fame Islands.
Brit. Birds 74: 41 1-416.

Harrison, P. 1983. Seabirds an identification guide.

Croom Helm Ltd., Beckenham, Kent, U.K.
Kennerley, P. R., P. J. Leader, and M. R. Leven.
1993. Aleutian Tern: the first records for Hong
Kong. Hong Kong Bird Report 1992:107-113.
Kennerley, P. and R. Ollington. 1998. Aleutian
Terns in South-East Asia. Oriental Bird Club Bull.
27:34-41.

Lee, D. S. 1992. Specimen records of Aleutian Terns
from the Philippines. Condor 94:276-279.

Leven, M. R., G. J. Carey, and V. B. Picken. 1994.

Systematic list. Hong Kong Bird Report 1993:50.
Rajathurai, S. 1996. The birds of Batam and Bintan
Islands, Riau Archipelago. Kukila 8:86-113.

SHORT COMMUNICATIONS

561

Wilson Bull., 111(4), 1999, pp. 561-564

Arthropods and Predation of Artificial Nests in the Bahamas:
Implications for Subtropical Avifauna

Nancy L. Staus' --^ and Paul M. Mayer^'^

ABSTRACT. — Little is known of nest predation
patterns in the dry subtropics. We used artificial nests
to examine patterns of nest predation and to identify
possible nest predators in the Bahamas. Unlike pre-
dation patterns in temperate areas, we found no rela-
tionship between predation rates and nest cover or dis-
tance to the road. Instead, the rate of nest predation
depended on distance to ocean. This result and a pho-
tograph taken at a disturbed nest implicated the giant
white land crab {Cardisoma guanhumi) as a possible
nest predator. Because land crabs are prevalent
throughout the subtropics and could potentially influ-
ence nesting behavior, we advise researchers to con-
sider variables associated with land crabs when ex-
amining nest predation in the subtropics. Received 14
July 1998, accepted 15 April 1999.

Nest predation studies are abundant in the
literature; most have been conducted in north-
ern, temperate areas (reviewed by Paton 1994,
Major and Kendal 1996, Hartley and Hunter
1998). Although a few similar studies have
taken place in the tropics (e.g., Gibbs 1991,
Laurance et al. 1993) and wet subtropics (Lat-
ta et al. 1995), no such study has been con-
ducted in dry, subtropical habitat where pred-
ator species assemblages may be quite differ-
ent. Patterns of nest predation might differ in
the dry subtropics as a result of differences in
numbers and species of egg predators.

Long Island, an outer island in the southern
Bahamas archipelago, is characterized by dry.

' James Ford Bell Museum of Natural History and
Graduate Program in Conservation Biology, 100 Ecol-
ogy Building, Univ. of Minnesota, St. Paul, MN
55108.

^ Conservation Biology Institute, 800 NW Starker
Ave., Suite 3 1C, Corvallis, OR 97330.

^Graduate Program in Conservation Biology, 100
Ecology Building, Univ. of Minnesota, St. Paul, MN
55108.

U.S. Environmental Protection Agency, Kerr Re-
search Center, P.O. Box 1 198, Ada, OK 74820.

^ Corresponding author; E-mail;
nstaus@earthdesign.com

scrubby vegetation and a relatively depauper-
ate fauna. There has been no prior study to
examine avian nest predation on any of the
Bahama islands or to determine which egg
predators are present. We used artificial nests
to determine factors influencing nest survival
of ground-nesting birds and to identify im-
portant nest predators on Long Island and Hog
Cay, Bahamas. Artificial nests are frequently
used in predation experiments where it is as-
sumed that they provide a reasonable assess-
ment of the impact of predators on real nests
(Burger et al. 1994, but see Major and Kendal
1996). In temperate zone studies, nest preda-
tion rates often varied with nest visibility (Ma-
jor and Kendal 1996) and distance from edge
(Paton 1994). We conducted an experiment to
determine whether patterns of nest predation
in the Bahamas were similar to those observed
elsewhere and to identify possible nest pred-
ators.

STUDY AREA AND METHODS

We conducted our study on the northern 20 km of
Long Island and on Hog Cay, Bahamas. Long Island,
one of the outer islands of the Bahamas archipelago,
is 128 km long and 6.4 km wide at its widest point.
Hog Cay is a small (100 ha), privately owned island
located off the northern tip of Long Island. Both is-
lands are covered with dry, scrubby vegetation. Man-
groves (Rhn.ophora mangle and Avicennia genninans)
grow along the coasts of both islands.

We observed 7 ground nesting bird species on Long
Island and Hog Cay, Bahamas. The largest included
the West Indian Whistling-duck (Dendrocygna arho-
rea) and White-cheeked Pintail (Anas hahatnensis ha-
hamensis). Smaller species included Antillean Night-
hawks (Chordeiles minor). Common Ground-Doves
(Calumhigallina passerina). Snowy and Wilson's plo-
vers (Cliaradrius ale.xandrinus and C. wilsonia), and
Willets (Catoptrophorus semipalmatu.s).

Potential terrestrial nest predators included intro-
duced rats (Rattus spp.), domestic dogs (Canis fami-
liaris), and native giant white land crabs (Cardisoma
guanhumi). Possible avian egg predators included
Laughing Gulls (Larus atricilla). Yellow-crowned

562

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Night-Herons {Nyctanassa violacea), and Smooth-
billed Anis (Crotophaga ani).

We utilized artificial ground nests in two experi-
mental trials lasting from mid-May through July 1995.
Each trial was 30 days, the approximate incubation
time for several ground nesting species in the Bahamas
(Paterson 1972). For each trial, two plots were selected
on Long Island and one plot was selected in similar
habitat on Hog Cay. Trial 1 took place May 17-June
19 and consisted of 25 and 16 nests on Long Island,
and 30 nests on Hog Cay. Trial 2 took place June 20-
July 21 and consisted of 24 and 16 nests on Long
Island, and 30 nests on Hog Cay. Artificial nests were
placed in a grid pattern within each of the six sites
such that one edge of the grid was located parallel to
and 50 m from a road. All nests within the grid were
placed 25 m apart and were randomly assigned as
“hidden” or “open.” Hidden nests were completely
covered by vegetation, whereas open nests could be
seen from within 1 m.

Nests consisted of a shallow scrape containing five
domestic chicken eggs located under thatch palm
(Thrinax microcarpa) or a dense bush. Although some
studies have detected effects of egg size because small
predators (e.g., mice) were unable to break larger eggs
(Pieman 1988, Haskell 1995), we believe that all po-
tential predators in our study sites were large enough
to handle chicken eggs. Nests were examined for sur-
vivorship at days 6, 12, 18, and 24, and were consid-
ered depredated if one or more eggs was missing or
damaged.

To identify specific nest predators, we placed three
automatic cameras with flash capability at one nest on
each study site during each trial. Cameras were trig-
gered by a motion-sensitive mercury switch glued to
the bottom of the eggs. Because cameras were con-
spicuous, they were placed at previously depredated
nests, which were then rebaited with chicken eggs.
Cameras were rotated among nests within study sites
on a weekly basis.

We also noted the remnants of eggs at the first pre-
dation event for each nest. Predators can sometimes be
identified by the type of egg remains they leave behind
(Reardon 1951; but see Trevor et al. 1991 ). Depredated
eggs were classified as missing or broken (portion of
an egg remaining in nest), and appearance of broken
eggs was also noted (e.g., many small fragments, half
shell remaining).

We developed a logistic regression model to exam-
ine the dependency of ne.st fate on nest type (i.e., hid-
den or open), distance to road (to examine edge ef-
fects), and distance to the ocean (a variable as.sociated
with land crab presence). Logistic regression models
have been used to analyze factors affecting the success
of both natural (Thomas ct al. 1996) and artificial nests
(Burger et al. 1994, Vander Haegen and DeGraaf
1996) and are appropriate when response variables are
binary (e.g., nest succc.ss or failure) and factors are
continuous (e.g., distance to road/occan; Hosmer and
Lemeshow 1989). We determined the suitability of the
model by using the Hosmer and Lemeshow Goodness-

of-Fit Test and associated statistic with a significance
level of P < 0.05. Individual variables within the mod-
el were tested with the Wald statistic. Analyses were
conducted with SAS (Windows version 6.12; SAS In-
stitute Inc., Cary, North Carolina).

RESULTS

Of 141 artificial nests, 99 (70%) were dep-
redated during the two trials combined. Our
overall regression model fit our data (Hosmer
and Lemeshow Goodness-of-Fit statistic =
8.57, df = 8, P = 0.38) and was significant
(score statistic = 21.73, df = 3, P < 0.001).
Nest fate depended primarily on distance to
the ocean (score statistic = 16.4, df = 1, P <
0.001). Nests ranged from 100-1500 m from
the ocean; the probability of nest success in-
creased with distance from the ocean. Nests
located farthest from the ocean (1500 m) had
the greatest success (71%), whereas those lo-
cated 100-325 m from the ocean and had an
average success rate of 23% (range 7-37%).

Nest fate was independent of nest type (hid-
den or open; x“ = 2.56, df = 1, P > 0.05);
44 (63%) and 54 (76%) of the nests were dis-
turbed at hidden and open nests, respectively.
In addition, nest fate was not associated with
distance to roads (x^ = 0.18, df = 1, P >
0.05).

One camera successfully captured activity
near a nest. A photograph was taken of a giant
white land crab near two damaged eggs in a
nest on Hog Cay. It was not clear whether the
crab broke the eggs, or found them after they
had been broken.

Of the 190 eggs from 92 nests on which
data were collected, 73 (38%) were missing
and 117 (62%) were broken at the first nest
check after predation. Thirteen (12.5%) of the
broken eggs were attributed to rat predation
(Flack and Lloyd 1976, pers. obs.).

DISCUSSION

Although the results of artificial nest ex-
periments conducted in the temperate zone are
often inconsistent, a few common patterns
have emerged. In general, predation rates -are
higher in nests that are more visible and in
habitats with little understory cover (Major
and Kendal 1996, Hartley and Hunter 1998).
In addition, Paton (1994) found a negative re-
lationship between nest predation rates and
distance from habitat edge in most of the 14

SHORT COMMUNICATIONS

563

artificial nest studies he re-analyzed. We
found no such patterns in the Bahamas.

In our study, the only environmental vari-
able that successfully predicted nest fate was
distance to the ocean, suggesting that nests
were depredated by a species residing in or
near the water. Our photograph of a giant
white land crab at a nest suggests that land
crabs were depredating artificial nests.

Although land crabs are terrestrial and do
not rely on the ocean directly on a daily basis,
they do need some source of water nearby to
survive and the females migrate to the ocean
for reproduction to release larvae (Wolcott

1988) . Given their reliance on water for re-
production and oxygen exchange, land crabs
are generally limited to low-lying areas near
mangroves, swamps, and streams, and are
rarely more than a few kilometers from the
sea (Wolcott 1988).

Giant white land crabs were abundant on
both Hog Cay and Long Island but did not
appear to be associated with edge habitats.
Their mostly vegetarian diets and ground for-
aging habits could bring them into contact
with ground nests regardless of whether the
nests were hidden or open. Experiments in-
volving captive land crabs revealed that crabs
were able to crack and consume eggs of var-
ious sizes corresponding to the egg sizes of
ground nesting birds in the Bahamas (Staus
and Barnwell 1996). This study also indicated
that crabs could be responsible for both bro-
ken and missing eggs.

Although some studies document chick pre-
dation by giant white land crabs (Gnam 1991),
ours is the first study to implicate Cardisoma
crabs as egg predators. Egg eating behavior
has been documented in several Gecarcinus
spp., land crabs in the same family as C.
guanhiimi (Rockwell 1932, Atkinson 1985,
Burger and Gochfeld 1988, Burger et al.

1989) . Other egg eating species include hermit
crabs (Coenobita rugoscr, Atkinson 1985, Bur-
ger et al. 1989), coconut crabs (Birgus latro\
Atkinson 1985), and ghost crabs (Ocypode
quadrata; Watts and Bradshaw 1995).

It has been suggested that land crabs play
an ecological role similar to that of rats, and
that crabs may have exerted considerable in-
fluence on tropical island avifuanas (Atkinson
1985). For example, after examining the fossil
record, Olson (1981) hypothesized that Ge-

carcinus land crabs in the South Atlantic may
have prevented the colonization of some is-
lands by burrowing and ground-nesting pe-
trels. Burger and Gochfeld (1988) provided
evidence that Roseate Terns {Sterna dougallii)
in Puerto Rico chose nest sites far from suit-
able land crab (G. ruricola) habitat. The wide-
ly distributed giant white land crab might have
a similar effect on bird populations within its
range.

Our results suggest that egg predation pat-
terns and predator species assemblages in the
dry subtropics may be different than those in
northern temperate areas. Specifically, land
crabs may play a significant role as egg pred-
ators. In the future, we urge researchers to
consider environmental variables associated
with the presence of land crabs (e.g., density
of crab burrows, altitude, distance from ocean)
when examining nest predation in the sub-
tropics.

ACKNOWLEDGMENTS

This research was funded by grants from Ducks Un-
limited (Bahamas), Birdlife International, Dayton and
Wilkie Funds for Natural History and Behavior, As-
sociation of Field Ornithologists, and World Nature
Association. Permission to work in the Bahamas was
granted by M. Isaacs of the Ministry of Agriculture.
Mr. and Mrs. P. Graham kindly allowed us to work on
Hog Cay. We thank N. Graham for excellent field as-
sistance during this project. Thanks also to L. Gape of
the Bahamas National Trust, P. Maillis, M. Lightbourn,
C. Crissey, and E. Wilson for their friendship and lo-
gistical support throughout this project. We gratefully
acknowledge K. Beal for her assistance with the sta-
tistical analysis. A special thanks to D. F. McKinney
for guidance and support throughout the project, and
helpful comments on earlier drafts of this manuscript.

LITERATURE CITED

Atkinson, I. A. E. 1985. The spread of commensal
species of Rattus to oceanic islands and their ef-
fects on island avifaunas. Pp. 35-81 in Conser-
vation of island birds (P J. Moors, Ed.). ICBP
Technical Publication No. 3, Page Bros. Ltd., Nor-
wich, England.

Burger, J. and M. Gochfeld. 1988. Nest-site selec-
tion by Roseate Terns in two tropical colonies on
Culebra, Puerto Rico. Condor 90:843—851.
Burger, J., M. Gochfeld. J. E. Saliva, D. Gochfeld,
AND H. Morales. 1989. Antipredator behaviour in
nesting Zenaida Doves (Zenaidci auhta): Parental
investment or offspring vulnerability. Behaviour
111:129-143.

Burger, L. D., L. W. Burger, Jr., and J. Faaborg.
1994. Effects of prairie fragmentation on preda-

564

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

tion of artificial nests. J. Wild!. Manage. 58:249-
254.

Flack, J. A. D. and B. D. Lloyd. 1976. The effect of
rodents on the breeding success of the South Is-
land Robin. Pp. 59—66 in The ecology and control
of rodents in New Zealand nature reserves (P. R.
Dingwall, I. A. E. Atkinson, and C. Hay, Eds.).
Departments of Land and Survey, Wellington,
New Zealand.

Gibbs, J. P. 1991. Avian nest predation in tropical wet
forest; an experimental study. Oikos 60:155-161.

Gnam, R. S. 1991. Natural history of the Bahama Par-
rot Amazona leucocephalo hahamensis on Abaco
Island. Proc. Symp. Nat. Hist. Bahamas 4:67-74.

Hartley, M. J. and M. L. Hunter, Jr. 1998. A meta-
analysis of forest cover, edge effects, and artificial
nest predation rates. Conserv. Biol. 12:465-469.

Haskell, D. G. 1995. A reevaluation of the effects of
forest fragmentation on rates of bird-nest preda-
tion. Conserv. Biol. 5:1316-1318.

Hosmer, D. W. and S. Lemeshow. 1989. Applied lo-
gistic regression. John Wiley and Sons, New York.

Latta, S. C., j. M. Wunderle, Jr., E. Terranova,
and M. Pagan. 1995. An experimental study of
nest predation in a subtropical wet forest follow-
ing hurricane disturbance. Wilson Bull. 107:590—
602.

Laurance, W. E, j. Garesche, and C. W. Payne.
1993. Avian nest predation in modified and nat-
ural habitats in tropical Queensland: an experi-
mental study. Wildl. Res. 20:711—723.

Major, R. E. and C. E. Kendal. 1996. The contri-
bution of artificial nest experiments to understand-
ing avian reproductive success: a review of meth-
ods and conclusions. Ibis 138:298-307.

Olson, S. L. 1981. Natural history of vertebrates on
the Brazilian islands of the mid south Atlantic.
Nat. Geog. Soc. Res. Rep. 13:481-492.

Paterson, A. 1972. Birds of the Bahamas. Durell Pub-
lications, Kennebunkport, Maine.

Paton, P. W. C. 1994. The effect of edge on avian nest
success: how strong is the evidence? Conserv.
Biol. 8:17-26.

PiCMAN, J. 1988. Experimental study of predation on
eggs of ground-nesting birds: effects of habitat
and nest distribution. Condor 90:124-131.

Reardon, J. D. 1951. Identification of waterfowl nest
predators. J. Wildl. Manage. 15:386—395.

Rockwell, R. H. 1932. Southward through the dol-
drums. Nat. Hist. 32:424—436.

Staus, N. L. and E H. Barnwell. 1996. The tropical
Atlantic land crab Cardi.soma guanhumi feeds on
bird eggs. Am. Zool. 36:80A.

Thomas, B. G., E. P. Wiggers, and R. L. Clawson.
1996. Habitat selection and breeding status of
Swainson’s Warblers in southern Missouri. J.
Wildl. Manage. 60:611-616.

Trevor, J. T, R. W. Seabloom, and R. D. Sayler.
1991. Identification of mammalian predators at ar-
tificial waterfowl nests. Prairie Nat. 23:93-99.

Vander Haegen, W. M. and R. M. DeGraaf. 1996.
Predation on artificial nests in forested riparian
buffer strips. J. Wildl. Manage. 60:542-550.

Watts, B. D. and D. S. Bradshaw. 1995. Ghost crab
preys on Piping Plover eggs. Wilson Bull. 107:
767-768.

Wolcott, T. G. 1988. Ecology. Pp. 55-96 in Biology
of the land crabs (W. W. Burggren and B. R. Mc-
Mahon, Eds.). Cambridge Univ. Press, Cam-
bridge, U.K.

Wilson Bull: 111(4), 1999, pp. 564-569

Notes About the Distribution of Pauxi pauxi and
Aburria aburri in Venezuela

Jose L. Silva' -

ABSTRACT. — In this paper I review the current
distribution of the Northern Helmeted Curassow
(Baicxi pciuxi) and the Wattled Guan {Aburria aburri)
in Venezuela. The historical range of P. pau.xi was re-
duced as a result of human population growth and hab-
itat perturbations. The current distribution corresponds

' NYZS The Wildlife Con.servation Society, 185th
& Southern Blvd., Bronx, New York 10460-1099.

2 Current address: Univ. of Florida, Wildlife Ecolo-
gy and Con.servation Dept., RO. Box I 10430, 303
Newins — Ziegler Hall, Gainesville, FL 32611-0430;
E-mail: jo.selugo@ufl.edu

principally with 18 national parks located from the
northern coastal mountains of central Venezuela to the
Andes Cordillera and Sierra de Perija. Pau.xi pau.xi was
recorded only in three localities outside national parks
and may have expanded from its historical distribution
in the eastern part of the country. Aburria aburri was
recorded in Sierra de Perija and western Merida to
southern Tachira, including four new localities; three
in national parks. Both species are endangered in Ven-
ezuela and their survival will depend on environmental
education programs and enforcement of the law. Re-
ceived 9 Feb. 199H, accepted 20 .July 1999.

SHORT COMMUNICATIONS

565

The Wildlife Conservation Society of the
New York Zoological Society funded a study
on human impacts on game species in pro-
tected areas of Venezuela from 1985 to 1990
(Silva and Strahl 1991, 1994, 1996, 1997).
During 1985-1996, censuses and interviews
were conducted and new data about the dis-
tribution of Pauxi pauxi (Northern Helmeted
Curassow) and Aburria aburri (Wattled Guan)
were collected. My objective in this paper is
to present these data and review the status of
P. pauxi and A. aburri in Venezuela.

In Venezuela Pauxi pauxi ranges from the
northern coastal mountains of central Vene-
zuela to the Andes Cordillera and Sierra de
Perija in rain forest and cloud forest. Most au-
thors (Wetmore and Phelps 1943; Phelps and
Phelps 1958, 1962; Delacour and Amadon
1973; Meyer de Schauensee and Phelps 1978;
Collar et al. 1992; Rodriguez and Rojas-Suar-
ez 1995) cited the states and localities of the
historical range as follows: P. p. pauxi: south-
ern Miranda state in Cerro Negro (Guatopo
National Park); north Caracas in El Calvario;
Distrito Federal; coastal mountains in Aragua
state (Henri Pittier National Park); Carabobo
state in Valencia, San Esteban, and Montal-
ban; east Falcon to Yaracuy state in Tucacas,
Nirgua, mountains inland from Aroa, and La-
gunita de Aroa; Lara state in Cubiro, and Ya-
cambu National Park; from northern Merida
to southern Tachira state in Montana de Li-
mones. La Azulita, and Burgua. Pauxi p. gil-
liardi: Zulia state from southern Sierra de Per-
ija (Sierra de los Motilones) to southern Rio
Tucuco, in Fila Macoita, Campamento Avispa,
Cerro Yin-taina, upper Rio Negro, and La Sa-
bana. A continuous distribution in the histor-
ical range was assumed (Fig. lA) because of
historical records, and Central Cordillera and
Los Andes Cordillera were almost a continu-
ous forest in the past.

The habitat available for P. pauxi has been
greatly reduced as a result of deforestation,
fragmentation, and habitat alteration. Almost
all the remnant forest available in northern
Venezuela was decreed as national parks by
the Venezuelan government. Consequently,
these national parks are isolated. 1 found that
the current distribution of P. pauxi mainly co-
incided with the distribution of national parks
situated in its historical range (Fig. IB), as
well as new localities such as Sierra de San

Luis, Cueva Quebrada del Toro, and Tirgua
National Parks. Although it was reported in
Morrocoy National Park (Collar et al. 1992,
Wege and Long 1993), according to the rang-
ers, it was no longer present in the park in
1996.

Pauxi pauxi is rare in national parks and
almost extinct outside national parks because
hunting pressure is highest outside the parks
(Silva and Strahl 1991, 1996, 1997). The few
locations where P. pauxi was found outside
national parks included the Sanchon River
Hydraulic Reserve (10° 24' N, 68° 09' W), the
Cojedes River Protectoral High Basin (10° 24'
N, 68° 15' W) and Finca El Jaguar (10° 26'
N, 68° 59' W).

From interviews with hunters I found that
P. pauxi probably existed or may still live in
eastern Venezuela. A hunter in Teresen (Mon-
agas State) narrated the size, color pattern, and
helmeted color of this species, and imitated its
booming song. He recognized the bird from a
set of cracid pictures. Pauxi pauxi was seen
in La Hormiga (9° 54' N, 62° 58' W), Cano
Payanuco, Guarapiche Forest Reserve (Sucre
and Monagas States) between 1968 and 1973.
Because only 1 of 25 interviewed hunters in
Teresen saw a P. pauxi, and saw it only once,
this should not be interpreted as range exten-
sion. The nearest locality of the historical dis-
tribution (Guatopo National Park) is approxi-
mately 405 km from Guarapiche Forest Re-
serve, and this separation is settled with towns
and cities. More likely the former distribution
record was incomplete. Pauxi pauxi is very
likely to be extinct in Guarapiche because of
high hunting pressure.

An interesting characteristic of P. pauxi is
the brown phenotype that sometimes occurs
in females. Males and females are typically
black with a white belly. Hunters call the
brown morph “Canaguey” or “Pauji Amaril-
lo.” It was reported in the Sierra de Perija,
where two specimens were collected between
1941 and 1957 (Delacour and Amadon 1973).
Here I report 26 new localities of the brown
morph seen between 1949 and 1993 (Table 1).
Of the 34 birds sighted, a single brown phe-
notype was seen with one black phenotype at
Fila Real (1975), one with two black pheno-
types at Casa de Tejas (1980), and one with
seven black phenotypes at El Corazon (1988).
Two brown phenotypes were seen with two

566

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

FIG. I . A. Historical range of Pait.xi pau.xi according to Delacour and Amadon { 1973), Meyer de Schauensee
and Phelps (1978), Collar and coworkers (1992), and Rodriguez and Rojas-Suarez (1995). B. Cunent range of
PaiLxi pau.xi in Venezuela. Boundaries of national parks are shown, but the species may not be distributed
through the entire park. Numbers are national parks, Roman numerals are other localities, and letters are states
with the exception of Caracas (Ca), the capital of Venezuela. Code; 1 = Guatopo (122,464 ha). 2 = El Avila
(81,800 ha). 3 = Macarao (15,000 ha). 4 = Henri Pittier (107,800 ha). 5 = San Esteban (43,500 ha). 6 =
Morrocoy (32,090 ha). 7 = Siena de San Luis (2(),()()() ha). 8 = Cueva Quebrada del Toro (4,885 ha). 9 =
Yurubi (23,670 ha). 10 = Tirgua. 1 1 = Terepaima (18,650 ha). 12 = Yacambii (14,580 ha). 13 = Dinira (42,000
ha). 14 = Guaramacal (21.000 ha). 15 = Sierra de La Culata (200,400 ha). 16 = Siena Nevada (276,446 ha).
17 = Tapo — Caparo. 18 = El Tama (109,000 ha). 19 = Siena de Perija (295,288 ha). I = Sanchon River
Hidraulic Re.serve (8,100 ha). 11 = Cojedes River Protectoral High Basin (276,000 ha). Ill = Finca El Jaguar
(I6,()()() ha). IV = Guarapiche Forest Reserve (576,500 ha). Mo = Monagas. S = Sucre. G = Guarico. Mi =
Miranda. Ca = Caracas. A = Aragua. C = Carabobo. Y = Yaracuy. F = Falcon. L = Lara. Tr = Trujillo. Me
= Merida. T = Tachira. Z = Zulia.

SHORT COMMUNICATIONS

567

TABLE 1.

Brown phenotype of Pauxi pauxi recorded

in Venezuela.

Locality

Date

Location

Coordinates

Henri Pittier

1975

La Glorieta

10° 28'N 67°45'W

Henri Pittier

03/1984

La Regresiva

I0°22'N 67°44'W

Henri Pittier

1993

El Saltico

Henri Pittier

1993

Los Riitos

—

San Esteban

1955

Burro Sin Cabezas

—

San Esteban

1960

Buito Sin Cabezas

—

San Esteban

1986-87

Flor Amarillo

—

San Esteban

1987

Flor Amarillo

—

San Esteban

07/1989

El Tanque

—

San Esteban

1991

San Felipe

—

San Esteban

El Dique

10° 18'N 67°59'W

San Esteban

La Panta (Qda. Yaguas)

—

San Esteban

Ranchitos

—

San Esteban

La Manguera

—

Terepaima

1949

Los Portones

9° 52'N 69° 20'W

Terepaima

1975

Fila Real

9° 55'N 69° 16'W

Terepaima

04/1992

Fila Real

9°53'N 69° 17'W

Yacambu

1979

El Blanquito, Qda. La Toma

9°42'N 69° 34'W

Yacambu

1983

Barro Amarillo

—

Yacambu

1988

El Corazdn

—

Yacambu

1988

La Canada

—

Yacambu

06/1992

El Blanquito

9° 42'N 69° 34'W

Yacambu

La Escalera

9° 42'N 69° 30'W

Yacambu

La Postora

9°41'N 69°37'W

Yacambu

El Blanquito

9° 42'N 69° 34'W

Yacambu

Cerro Blanco

9° 37'N 69° 30'W

Sierra Nevada

Alto de la Aguada

8° 37'N 70°40'W

Sierra Nevada

San Benito

8° 40'N 70° 37'W

Rio Sanchon

1980

La Cumbre del Cacho

—

Rio Sanchon

11/1980

Casa de Tejas

—

“ Name of localities do not appear on maps because they are local names used by hunters and the exact locations are unknown.

black phenotypes at El Saltico and Los Riitos
in 1993.

Aburria aburri was mainly recorded in the
western part of Venezuela (Fig. 2). The his-
torical range was reported to be in the Sierra
de Perija and west Merida to southern Tachira
in rain and cloud forest (Delacour and Ama-
don 1973, Meyer de Schauensee and Phelps
1978, Rodriguez and Rojas-Suarez 1995). The
current distribution of A. aburri indicates that
the record of the historical range may have
been incomplete. Aburria aburri was recorded
in Sierra Nevada National Park (54 inter-
viewed hunters), in Terepaima National Park
(observed), and in Yacambu National Park
and the basin of Yacambu River (S. Boher,
pers. comm., and 338 interviewed hunters).
These were new distribution records, but they
did not suggest an extension of the historical
range because hunters over 60 years old hunt-

ed A. aburri since they were young. Perhaps,
the historical range was continuous. Although
Rodriguez and Rojas-Suarez (1995) stated that
A. aburri probably was found in the eastern
part of Costa Cordillera, they did not mention
the source of their information.

The present status of P. pauxi and A. aburri
is worrisome. According to the population
censuses (Silva and Strahl 1991, 1997) and
the interviews (Silva and Strahl 1996), P.
pauxi and A. aburri have very low densities
with A. aburri being more rare than P. pauxi.
Both species were considered Endangered by
the Cracid Specialist Group (Strahl et al.

1994) and by researchers of a recent study in
Venezuela (Rodriguez and Rojas-Suarez

1995) . Habitat destruction and illegal hunting
are the principal causes of the decline in pop-
ulation of both species, and their conservation
will rely on hunter education (Silva and Pel-

568

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

(B)

FIG. 2. Historical (A) and current distribution (B) of Ahurria aburri in Venezuela. Abbreviations are the
same as Fig. 1.

legrini 1996, Silva 1997) and enforcement of
the law.

ACKNOWLEDGMENTS

NYZS The Wildlife Conservation Society funded
the main projects. I thank hunters who collaborated
with the interviews. A. Grigss reviewed the manu-
.script.

LITERATURE CITED

Collar. N. J., L. P. Gonzaga, N. Krabbe, A. Mad-
rono Nieto, L. G. Naranjo, T. A. Parker, III,
AND D. C. Wege. 1992. Threatened birds of the
Americas: the ICBP/IUCN red data book 2. Inter-
national Council for Bird Preservation, Cam-
bridge, U.K.

Delacour, J. and D. Amadon. 1973. Curassows and
related birds. American Museum of Natural His-
tory, New York.

Meyer de Schauensee, R. and W. H. Phelps, Jr.
1978. A guide to the birds of Venezuela. Princeton
Univ. Press, Princeton, New Jersey.

Phelps, W. H. and W. H. Phelps, Jr. 1958. Lista de
Aves de Venezuela con su dislribucion I: no pas-
ser! formes. Bol. Soc. Venezolana Cienc. Nat. 19:
1-317.

Phelps, W. H. and W. H. Phelps, Jr. 1962. Two new

subspecies of birds from Venezuela, the rufous
phase of Pauxi paiixi and other notes. Proc. Biol.
Soc. Wash. 75:199-204.

Rodriguez. J. P. and F. Rojas-Suarez. 1995. Libro
Rojo de la Fauna Venezolana. ExLibris, Caracas,
Venezuela.

Silva, J. L. 1997. Proyecto de Educacion para la Con-
servacion de la Familia Cracidae en los Parques
Nacionales al Norte de Venezuela. Pp. 424-436
in The Cracidae: their biology and conservation
(S. Strahl, S. Beaujon, D. M. Brooks, A. J. Be-
gazo, G. Sedaghatkish, and F. Olmos, Eds.). Han-
cock House, Hong Kong, China.

Silva, J. L. and N. Pellegrini. 1996. Proyecto de Ed-
ucacion para Cazadores Furtivos en Parques Na-
cionales. Natura 106:59-62.

Silva, J. L. and S. Strahl. 1991. Human impact on
populations of guan. curassow and chachalacas in
Venezuela. Pp. 36-52 in Neotropical wildlife use
and conservation (J. G. Robinson and K. H. Red-
ford, Eds.). Chicago Univ. Press, Chicago, Illinois.

Silva, J. L. and S. Strahl. 1994. Usos Folcloricos de
la Fauna Silvestre en Nueve Parques Nacionales
al Norte de Venezuela. Vida Silvestre Neotropical
3:100-107.

Silva, J. L. and S. Strahl. 1996. La Caza Furtiva en
los Parques Nacionales al Norte de Venezuela.
Vida Silvestre Neotropical 5:126-139.

SHORT COMMUNICATIONS

569

Silva, J. L. and S. Strahl. 1997. Presion de Caza
sobre Poblaciones de Cracidos en los Parques Na-
cionales al Norte de Venezuela. Pp. 437—448 in
The Cracidae: their biology and conservation (S.
Strahl, S. Beaujon, D. M. Brooks, A. J. Begazo,
G. Sedaghatkish, and F. Olmos, Eds.). Hancock
House, Hong Kong, China.

Strahl, S., S. Ellis, O. Byers, and C. Plasse. 1994.
Conservation assessment and management plan

for Neotropical guans, curassows, and chachala-
cas. lUCN/SSC Conservation Breeding Specialist
Group, Apple Valley, Minnesota.

Wege, D. C. and a. Long. 1993. Bird conservation
priorities; country profile, Venezuela. BirdLife
Conserv. Series 3:1-85.

Wetmore, a. and W. H. Phelps. 1943. Description of
a third form of curassow of the Genus Pauxi. J.
Washington Acad. Sci. 33:142-146.

Wilson Bull., 111(4), 1999, pp. 569-571

Western Burrowing Owls in California Produce Second Broods of Chicks

Jennifer A. Gervais' - and Daniel K. Rosenberg'

ABSTRACT. — We present the first evidence that
western Burrowing Owls are capable of raising a sec-
ond brood of chicks within a nesting season once their
first brood successfully fledges. Two pairs of owls in
central California known to have successfully fledged
chicks from a first brood renested in 1998, with one
pair producing five additional fledglings. Received 29
March 1999, accepted 15 July 1999.

Western Burrowing Owls {Athene cunicu-
laria) are thought to be declining throughout
much of their range (DeSante et al. 1997,
James and Espie 1997). The potential causes
of these declines vary with location, but likely
include large-scale habitat destruction from
farming or development, reductions in species
such as ground squirrels that create the bur-
rows that the owls use, and agricultural chem-
icals (James and Espie 1997, Gervais et al. in
press). Because of the perceived threat to the
viability of Burrowing Owl populations, the
species has been listed as endangered, threat-
ened, or of special management concern in a
number of North American states and prov-
inces (Haug et al. 1993).

Effective conservation at the species level
requires understanding the population dynam-
ics of the species in question, which in turn
means accurate estimation of demographic pa-

' Oregon Cooperative Fish & Wildlife Research
Unit, Dept, of Fisheries and Wildlife, Oregon State
Univ., Corvallis, OR 97331.

- Corre.sponding author; E-mail:
gervaisj @ ucs.orst.edii

rameters such as survival and reproductive
rates. These can be used in simplified models
that allow the examination of the effects of
possible management actions or environmen-
tal perturbations on population persistence.
Such an approach has recently been used for
the northern Spotted Owl (Strix occidentalis\
Noon and Biles 1990), and for predicting the
effects of pesticide exposure on wildlife pop-
ulations (Caswell 1996, Calow et al. 1997).

Simulations of generalized life history strat-
egies have shown that for a species with rel-
atively low adult survivorship and a short life
span, reproductive success may be most influ-
ential in maintaining population viability (Em-
len and Pikitch 1989). This is likely to be gen-
erally true for Burrowing Owls. They are ca-
pable of producing up to 12 eggs in a clutch
(Haug et al. 1993), and we have observed up
to 10 young fledged per nest in good repro-
ductive years. In addition. Burrowing Owl an-
nual adult survivorship appears to be quite
low, with between-year return rates ranging
from 33-58% (Haug et al. 1993), and a lon-
gevity record for a wild banded owl of 8 years
and 8 months (Kennard 1975). If sensitivity
analyses prove that the Burrowing Owl fits the
predictions of the Emlen-Pikitch model (Em-
len and Pikitch 1989) for a small, relatively
short-lived species, then accurate assessment
of reproductive potential of Burrowing Owls
is essential to evaluating population processes.

Only Florida Burrowing Owls have been
known to produce second broods within a sea-
son (Millsap and Bear 1990). We report two

570

THE WILSON BULLETIN • Vol. 1 1 1, No. 4, December 1999

instances of western Burrowing Owl pairs at-
tempting second broods after the first brood
had successfully been fledged during the 1998
breeding season. To our knowledge, this is the
first time the production of more than a single
brood per season has been verified in western
Burrowing Owls.

METHODS AND RESULTS

We have conducted demographic research
since 1997 on a population of Burrowing
Owls at Naval Air Station Lemoore (36° 20'
N, 119° 57' W), 50 km southwest of Fresno,
Califoma. The naval air station supports ap-
proximately 65 breeding pairs of owls, which
appear to be winter residents (Gervais and Ro-
senberg, unpubl. data). Nesting habitat on the
station is primarily small patches of exotic an-
nual grasses along runway easements and in
wildlife areas surrounded by agricultural
fields. Wildlife areas are fallow fields that are
composed of exotic annual grasses and weeds,
although they are burned annually. Approxi-
mately 75% of the adult resident population
of owls is now banded with U.S. Fish and
Wildlife Service bands and unique alpha-nu-
meric rivet bands (Acraft Bird Bands, Edmon-
ton, Alberta, Canada).

Early in the season, we collected eggs for
use in an ongoing toxicology study (Gervais
et al. in press). One sampled burrow contained
at least four eggs and the incubating female
on 19 April 1998. At that time, we identified
the female from her bands and we removed
one egg. Her mate also was previously banded
and was identified by resighting his bands ear-
ly in the nesting season. This nest successfully
fledged two chicks in early June. We recap-
tured the female owl on 14 June using a
mouse baited spring net. At capture, she
weighed 198 g, well above the 150 g average
for this species (Haug et al. 1993), and was
noticeably swollen in the lower abdomen. Her
brood patch was well developed and vascu-
larized, suggesting nesting activity.

To verify that this female was indeed relay-
ing, we used an infrared burrow probe (Chris-
tensen Designs, Manteca, California) to ex-
amine the burrow on 16 June. We observed
the two fledged chicks in the entrance to the
nesting chamber, but were unable to see be-
yond them. The burrow entrance had fresh
decorations of coyote dung and the nest tunnel

was lined with similar debris. The adults were
observed at the burrow entrance throughout
the next few weeks; individual identity was
confirmed using their color bands.

We examined the nest again on 27 June,
and observed four eggs in the nest chamber
after the female flushed from the burrow en-
trance. We removed two eggs through an ac-
cess hole originally dug for the toxicological
study egg sampling (Gervais et al., in press).
The eggs were cool, but the shells were very
clean and candling revealed clear egg contents
with no visible development. The eggs were
returned to the nest after inspection and the
access holes covered again with dirt and
boards. We do not believe these eggs were
from the previous nesting attempt because of
their clear contents and clean shells. Eggs that
sit in the burrow for eight weeks would have
dark contents as they began to rot and shells
would be covered with dirt and fecal matter
from the chicks.

When we examined the burrow on 14 July,
the eggs were gone. No owls were present at
the burrow during that visit, although both
adults continued to be sighted in the area
through July.

A second double nesting attempt also oc-
curred in 1998. We observed with the infrared
burrow probe a banded female owl in her bur-
row with nine eggs on 16 April; she raised one
chick to fledging after the disappearance of her
mate. We observed this same female at the
same burrow entrance in early September with
five buffy breasted chicks. These chicks clearly
had recently emerged from their burrows, be-
cause juvenile owls fledged during the main
breeding attempt at this site have typically un-
dergone a body molt by this time and their
breasts are heavily streaked in the manner of
adult birds. No other nests within the area still
contained chicks at this time. The five chicks
were frequently seen at the burrow entrance
through the middle of September when field-
work was discontinued. This is. typical of
young owls still fully dependent on their par-
ents for food; owls fledged earlier in the season
had dispersed from their natal burrows by early
August as indicated by radio telemetry (Ger-
vais and Rosenberg, unpubl. data). The owl’s
mate for this second attempt was also banded,
but his bands were consistently too muddy to
read and he was never identified.

SHORT COMMUNICATIONS

571

Our fieldwork did not include detailed ob-
servations of all nests in our study area
throughout the summer and early fall, but we
did not find any other evidence in support of
double brooding attempts. These attempts
may be quite rare and only occur in excep-
tional years. The 1998 breeding season was
marked by very late rains, resulting in a high
proportion of renesting efforts by the owls
(Rosenberg and Gervais, unpubl. data). The
prolonged growing season that followed the
wet spring may have led to conditions con-
ducive to late-season breeding attempts, such
as greater food availability.

Nevertheless, it is clear that at least in some
conditions western Burrowing Owls can raise
two broods of chicks, thus increasing their re-
productive output. This may be important for
individuals whose first broods were small be-
cause of predation or the loss of a mate, or
for populations recovering from environmen-
tal damage such as pesticides or burrowing
rodent control. This information is also im-
portant for use in sensitivity modeling such as
that done by Emlen and Pikitch (1989) or
Noon and Biles (1990), because accurate de-
mographic parameter estimation is essential to
determining life history strategies and evalu-
ating demographic risks to populations.

ACKNOWLEDGMENTS

This research was funded by the U.S. Navy Engi-
neering Field Activity West, the Bureau of Land Man-
agement, Bakersfield, California Field Office, the U.S.
Fish and Wildlife Service Partners in Wildlife Act
Fund, California Department of Fish and Game, and
the National Fish and Wildlife Foundation. We thank
T. Lanman and J. Podulka for assistance with field ob-
servations; E.S. Botelho, G.A. Green, and an anony-
mous reviewer for comments on the manuscript; and
J. Crane for logistical support at NAS Lemoore that
made this research possible. Cooperators ot the Oregon
Cooperative Fish and Wildlife Research Unit include
the U.S. Fish and Wildlife Service, Oregon State Uni-
versity, Oregon Department of Fish and Wildlife, the

Wildlife Management Institute, and the Biological Re-
sources Division of the U.S. Geological Survey. This
work was conducted in conjunction with The Institute
for Bird Populations, and is The Institute for Bird Pop-
ulations Publication No. 175.

LITERATURE CITED

Calow, P, M. Sibley, and V. Forbes. 1997. Risk as-
sessment on the basis of simplified life-history
scenarios. Environ. Toxicol. Chem. 16:1983-
1989.

Caswell, H. 1996. Demography meets ecotoxicology:
untangling the population level effects of toxic
substances. Pp. 255-288 in Ecotoxicology: a hi-
erarchical approach (M. C. Newman and C. H.
Jagoe, Eds.). Lewis Publishers, Boca Raton, Flor-
ida.

DeSante, D. E, E. D. Ruhlen, S. L. Adamany, K. M.
Burton, and S. Amin. 1997. A census of Burrow-
ing Owls in central California in 1991. Pp. 38-48
in The Burrowing Owl: its biology and manage-
ment (J. L. Lincer and K. Steenhof, Eds.). Raptor
Research Foundation, Inc., Boise, Idaho.

Emlen, J. M. and E. K. Pikitch. 1989. Animal popu-
lation dynamics: identification of critical compo-
nents. Ecol. Model. 44:253—273.

Gervais, J. A., D. K. Rosenberg, D. M. Fry, L. Tru-
Lio, AND K. K. Sturm. In press. Burrowing Owls
and agricultural pesticides: evaluation of residues
and risks for three populations in California. En-
viron. Toxicol. Chem.

Haug, E. a., B. a. Millsap, and M. S. Martell.
1993. Burrowing Owl {Speotyto cunicularia). In
The birds of North America, no. 61 (A. Poole and
F Gill, Eds.). The Academy of Natural Sciences,
Philadelphia, Pennsylvania; The American Orni-
thologists’ Union, Washington, D.C.

James, P. C. and R. H. M. Espie. 1997. Cunent status
of the Burrowing Owl in North America: an agen-
cy survey. Pp. 3—5 in The Burrowing Owl: its bi-
ology and management (J. L. Lincer and K. Steen-
hof, Eds.). Raptor Research Foundation, Inc., Boi-
se, Idaho.

Kennard, j. H. 1975. Longevity records of North
American birds. Bird-Banding 46:55—73.

Millsap, B. A. and C. Bear. 1990. Double brooding
by Florida Burrowing Owls. Wilson Bull. 102:
313-317.

Noon, B. R. and C. Biles. 1990. Mathematical demog-
raphy of Spotted Owls in the Pacific Northwest. J.
Wildl. Manage. 54:18-27.

572

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Wilson Bull., 111(4), 1999, pp. 572-573

Continuous Nesting of Barn Owls in Illinois

Jeffery W. Walk,'-^ Terry L. Esker,- and Scott A. Simpson-

ABSTRACT. — Bam Owls (Tyto alba) typically be-
gin nesting in temperate zones in early spring. The
species has high reproductive output (large clutch size,
occasionally double-brooded) and high mortality for a
member of the Strigiformes. We report on Barn Owls
in one nest box that hatched hve clutches and fledged
young from four clutches within 23 months. Laying,
incubation, or brood-rearing was attempted in every
month of the year. In the only clutch mortality we wit-
nessed, three young apparently died of exposure dur-
ing a period of cold weather (temperatures as low as
— 15°C). Received 24 February 1999, accepted 27
May 1999.

Bam Owls {Tyto alba) have a high repro-
ductive rate and relatively short life span, fit-
ting an r-selected life history strategy (Colvin
et al. 1984, Marti 1997). Average clutch sizes
in North America range from 4.2— 7.1 eggs
(mostly 5-6 eggs; Hands et al. 1989, Marti
1992) and the average Bam Owl life span is
less than 2 years (1.4 years, Stewart 1952; 1.7
years, Keran 1981). Most nest initiations in
temperate areas occur from February to June,
with the peak probably occurring March-May
(Hands et al. 1989, Marti 1992). Nestlings
have been banded in every month except Feb-
ruary in the northern United States (Stewart
1952), suggesting that nesting may occur in
any season. Second broods are uncommon in
temperate climates. For example, only about
5% of pairs in Utah produced a second brood
(Marti 1992). In contrast, nesting in tropical
areas occurs year around with double broods
being common (e.g., Lenton 1984). In captiv-
ity, Maestrelli (1973) reported a pair fledging
six broods in 22 months.

A nest box (enclosed design from Colvin
1983) was erected 4.6 m above ground in an
empty barn in 1986. This location was within

' Univ. of Illinois, Dept, of Natural Resources and
Environmental Sciences, Urbana, IL 61801.

^ Illinois Dept, of Natural Resources, Prairie Ridge
State Natural Area, Newton, IL 62448.

' Corresponding author; E-mail: j-walk@uiuc.edu

a 64 ha grassland tract of Prairie Ridge State
Natural Area, Marion County, Illinois (38° 45'
N, 88° 51' W). Prairie Ridge State Natural
Area grasslands are a combination of restored
native grasses and introduced cool-season
grasses managed by the Illinois Department of
Natural Resources for grassland wildlife.
Within 2 km of the nest site, land use was
about 70% rowerop agriculture, 15% grass-
lands (96 ha of Prairie Ridge State Natural
Area, 98 ha of Conservation Reserve Program
grasslands), 8% small grains, and small
amounts of woodland, pasture and farmsteads.

The nest box was checked periodically and
Rock Dove {Columba livia) nests were re-
moved. We first observed one adult Barn Owl
perched upon the nest box on 20 September
1993. On 1 November 1993, two adults and a
clutch of six eggs were noted. We saw three
chicks on 21 December 1993 and found them
dead in the nest box on 6 January 1994. The
young apparently died of exposure to harsh
weather between 23 and 31 December. Mean
temperature during this period was — 5.5°C
(—15° to 3° C) with 7 cm snowfall on 25 De-
cember and northeasterly winds 20-25 km/hr
blowing into the nest box on 28 and 29 De-
cember (weather data from Midwestern Cli-
mate Center, Champaign, Illinois). The second
nest attempt began about 1 1 March 1994 (one
egg in nest box), with five young seen through
May 1994. Fledging occurred in early July
(four grown birds observed 14 July). An adult
and the third clutch of five eggs were noted
23 August 1994. Four chicks from this clutch
fledged between 25 and 30 October. The
fourth nest attempt was apparently initiated in
early to mid-February 1995. A clutch of seven
eggs was observed 13 March, and five yolmg
about to fledge were seen 9 May 1995. A fifth
clutch consisted of three eggs recorded on 7
July 1995, three young (estimated two weeks
old) on 23 August and two fledglings on 19
October 1995. Single Barn Owls were ob-
served only sporadically after this date.

SHORT COMMUNICATIONS

573

The five clutches averaged 5.2 eggs per
clutch (range 3-7). Of the four successful
clutches, 3.75 young fledged per clutch (range
2-5). We estimate an average of 160 days
(range 130-190 days) for each successful nest
cycle (estimated time between the start of one
attempt and the start of the next). While these
birds were not banded or marked, their con-
stant presence at the nesting site from Septem-
ber 1993 to October of 1995 suggests the
same adults were involved in all nest attempts.

Excluding the first nest attempt, the tem-
poral pattern of the four successful nests better
fits the typical early spring/late summer pat-
tern of double-brooded Barn Owls (Taylor
1994), although the 1995 nests were about one
month earlier than their 1994 counterparts.
Taken as a whole, this nest site was used near-
ly continuously for two years with an addi-
tional delay of about 30 days between nest
attempts during the coldest time of year.

This is only the fourth report of Bam Owls
using a nest box in Illinois (Illinois Biological
and Conservation Data System, unpubl. data).
Once this nest site was discovered, breeding
activity was brief, but quite productive.

ACKNOWLEDGMENTS

We thank R. Day, R. Edgin, B. Griffith and R. Jan-
sen for their assistance with monitoring the Barn Owl
nest box. We gratefully acknowledge C. Becker, D.
Cooper, J. Herkert, E. Kershner, G. Kruse, C. Marti,
D. Olson, D. Smith, G. Therres, and R. Warner for
reviewing this manuscript. We would also like to thank
the Illinois Endangered Species Protection Board, the

Illinois Natural History Survey, the Illinois Nature Pre-
serves Commission and The Nature Conservancy for

their support of Prairie Ridge State Natural Area.

LITERATURE CITED

Colvin, B. A. 1983. Nest boxes for Barn Owls. Publ.
346(183). Ohio Dept, of Nat. Res., Columbus.

Colvin, B. A., P. L. Hegdal, and W. B. Jackson.
1984. A comprehensive approach to research and
management of Common Barn-Owl populations.
Pp. 270-282 in Proceedings of workshop man-
agement of nongame species and ecological com-
munities (W. Comb, Ed.). Univ. of Kentucky, Lex-
ington.

Hands, H. M., R. D. Drobney, and M. R. Ryan. 1989.
Status of the Common Barn-Owl in the northcen-
tral United States. Missouri Cooperative Fish and
Wildlife Research Unit, Columbia.

Reran, D. 1981. The incidence of man-caused and
natural mortalities to raptors. Raptor Res. 15:108—
112.

Lenton, G. M. 1984. The feeding and breeding ecol-
ogy of Barn Owls Tyto alba in peninsular Malay-
sia. Ibis 126:551-575.

Maestrelli, J. R. 1973. Propagation of Barn Owls in
captivity. Auk 90:426-428.

Marti, C. D. 1992. Barn Owl. In The birds of North
America, no. 1 (A. Poole, P. Stettenheim, and E
Gill, Eds.). The Academy of National Sciences,
Philadelphia, Pennsylvania; The American Orni-
thologists’ Union, Washington, D.C.

Marti, C. D. 1997. Lifetime reproductive success in
Barn Owls near the limit of the species’ range.
Auk 114:581-592.

Stewart, P. A. 1952. Dispersal, breeding behavior and
longevity of banded Barn Owls in North America.
Auk 69:227-245.

Taylor, I. 1994. Barn Owls: predator-prey relation-
ships and conservation. Cambridge Univ. Press,
Cambridge, U.K.

574

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Wilson Bull., 111(4), 1999, pp. 574-575

Unusual Nest Sites for Southwestern Willow Flycatchers

Scott H. Stoleson''^ and Deborah M. Finch'

ABSTRACT — The endangered southwestern Wil-
low Elycatcher (Empidona.x traillii extimus) is an ob-
ligate riparian species that typically nests in willow
(Salix spp.) thickets or other dense, shrubby habitats.
We report on the first nests in Arizona sycamore (Pla-
tanus wrightii) and in a climbing rose vine (Rosa mul-
tiflora). Although these nests were located in novel
substrates, they were typical for the species in being
supported by multiple small stems and in having a
dense canopy cover. We suggest that nest substrate
preferences of Willow Elycatchers in the Southwest
may be broader than generally considered. Received
17 Nov. 1998, accepted 6 March 1999.

The southwestern Willow Flycatcher (Empi-
donax trcdllii extimus), a federally-listed endan-
gered species, is an obligate riparian specialist
that breeds in dense vegetation associated with
watercourses (U.S. Fish and Wildlife Service
1995). Most studies of habitat preferences in
Willow Flycatchers have shown a strong asso-
ciation with willow (Salix spp.) thickets or other
shrubby habitats (McCabe 1991, Sedgwick and
Knopf 1992). In the Southwest, nests of this
subspecies have been found most commonly in
willows, salt cedar (Tamarix spp.), and locally
in forbs such as stinging nettles (Urtica dioica
holosericea) or trees such as boxelder (Acer ne-
gundo), alder (Alnus spp.), Russian olive
(Eleagnus angustifolia), and young Fremont
cottonwoods (Populus fremontii; Sferra et al.
1997; Sogge et al. 1997). In this paper we report
on the first recorded incidence of Willow Fly-
catchers nesting in Arizona sycamore (Platanus
wrightii) and in a nonnative climbing rose (Rosa
multiflora).

These observations were made as part of a
study of southwestern Willow Flycatchers in
the Gila River valley near the towns of Cliff
and Gila. Grant County, New Mexico (32° 57'
N, 108° 35' W). The study area consists of

' USDA Forest Service, Rocky Mountain Re.search
Station, 2205 Columbia SE, Albuquerque, NM 87106.

^Corresponding author; E-mail:
sstolcso/rmrs_albq@fs.fcd.us

patches of riparian woodland along the river
and earthen irrigation ditches at elevations av-
eraging 1400 m. Most of the valley bottom is
used for ranching and farming. Woodland
patches are composed primarily of Fremont
cottonwood, Goodding’s willow (Salix good-
dingii), boxelder, Arizona sycamore and Ari-
zona walnut (Juglans major), with an under-
story of shrubs, forbs, and grasses. This valley
supports the largest known population of Wil-
low Flycatchers in the Southwest, with an es-
timated 230 pairs in 1998 (Stoleson and
Finch, unpubl. data; P. Boucher, pers. comm.).

Nesting habits at this site differed from
what has been reported elsewhere in the
Southwest. Of 257 nests located in 1997-
1998, 76.5% were placed in boxelder, 8.6% in
willows, 6.3% in Russian olive, and the re-
mainder (<5% each) in Arizona alder (Alnus
oblongifolia), seepwillow (Baccharis glutino-
sa), Fremont cottonwood, salt cedar, Arizona
sycamore, and rose.

One pair of Willow Flycatchers was found
building a nest in a sycamore on 8 June at the
Fort West Ditch site on the Gila National For-
est (FS-4). This nest was too high for its con-
tents to be visible (nest characteristics in Table
1). Parents were observed carrying food to the
nest on 1 and 6 July. On 13 July, the parents
were observed feeding at least two fledglings
in the surrounding trees. This nest was in a
cluster of five vertical twigs on a small
branch. Although the nest tree was very open,
the nest itself was immediately beneath a
dense layer of foliage.

A second sycamore nest (GRP-7) was lo-
cated in The Nature Conservancy of New
Mexico’s Gila Riparian Preserve on 23 July.
The female was on the nest incubating or
brooding. On 31 July the nest was found errip-
ty and disheveled, presumably the result of
predation. This nest was located in a cluster
of about 1 2 vertical twigs at the end of a short,
broken branch. Like the previous nest, the nest
was visible from the sides but covered from
above by dense foliage.

SHORT COMMUNICATIONS

575

TABLE 1. Characteristics and outcomes of southwestern Willow Flycatcher nests in rose and sycamore in
the Gila River valley, southwestern New Mexico, 1998.

Nesl characierisiics

Nesl plant

Ne.st

Sub.striue

Nest ht.
(m)

Canopy ht.
at nest (m)

Canopy
cover (%)^

diameter

(cm)*'

Distance to
water (cm)*^

Distance to
edge (m)‘*

Outcome

SEl-19

rose

3.5

13.9

0.7

fledged 2

FS-4

sycamore

13.6

17.9

33.0

fledged ^2

GRP-7

sycamore

8.0

12.2

33.2

depredated

^ Average percent canopy cover measured at base of nest plant and at points 4 and 8 m from base in tour cardinal directions, measured using densiomeiers.
^Measured at 1.7 m above ground.

^ Horizontal distance between base of nest plant and nearest perennial water.

^ Defined here as horizontal distance between nest site and nearest area with no tree cover (i.e., shrubs, sand, or pasture).

On 18 June, an incubating bird was flushed
from a nest in a rose vine climbing a large
boxelder tree on the U-Bar Ranch (SEl-19).
The nest contained two eggs at that time. On
13 July, two almost fully feathered fledglings
were observed being fed in the undergrowth
near the nest. The nest was placed at the junc-
tion of four stems of the nonnative Rosa mul-
tiflora, hanging from and about a meter below
a leaning trunk of boxelder.

Willow Flycatcher nests have been found
only rarely in native shrubby Rosa species in
the Southwest, in California (W. Haas, pers.
comm.) and at high elevations in Arizona
(McCarthey et al. 1998). In the Palouse Hills
of Washington, where Willow Flycatchers are
not restricted to riparian habitats, rose was the
most frequent nest substrate (King 1955).
Similarly, 56% of nests in the interior of Brit-
ish Columbia were in rose (Campbell et al.
1997). Nests have been reported in rose else-
where as well (Walkinshaw 1966; McCabe
1991; J. Sedgwick, pers. comm.).

Our observations emphasize that Willow Fly-
catchers are opportunistic in their choice of nest-
ing substrates, apparently requiring only dense
fohage and a suitable twig structure to support
their nests (McCabe 1991, Sogge et al. 1997).
Although the three nests reported here were un-
usual in terms of substrate species, they were
very typical of flycatcher nests with respect to
fohage density and twig structure (Table 1).

ACKNOWLEDGMENTS

We thank G. Bodner, K. Brodhead, P. Chan. J. Gar-
cia, B. Gibbons, D. Hawk.sworth, and H. Walker for
field assistance; P. Boucher, J. Monzingo, and R. Pope
of the Gila National Forest and T. Bays, T. Shelley,
and C. Rose of Phelps Dodge for logistical support; D.

Parker for sharing his expertise; and T and D. Ogilvie
for their hospitality. Funding was provided by the Gila
National Forest, Phelps Dodge Corporation, and The
Nature Conservancy. Comments by T. Bays, D. Mei-
dinger, C. Rose, J. Sedgwick. M. Whitfield and two
anonymous reviewers improved the manuscript.

LITERATURE CITED

Campbell, R. W, N. K. Da we, I. McTaggart-Cowan,
G. E. Smith, and J. M. Cooper. 1997. The birds of
Briti.sh Columbia; passerines: flycatchers through
vireos. Univ. of British Columbia Press, Vancouver.
King, J. R. 1955. Notes on the life history of Traill’s
Flycatcher {Empidonax traillii) in southeastern
Washington. Auk 72:148-173.

McCabe, R. A. 1991. The little green bird. Palmer
Publications, Inc., Amherst, Wisconsin.
McCarthey, T. D., C. E. Paradzick, J. W. Rourke,
M. W. Sumner, and R. F. Davidson. 1998. Ari-
zona Partners in Flight southwestern Willow Fly-
catcher 1997 survey and nest monitoring report.
Nongame and Endangered Wildlife Program
Technical Report 130. Arizona Game and Fish
Department, Phoenix.

Sedgwick, J. A. and F. L. Knopf. 1992. Describing
Willow Flycatcher habitats: scale perspectives and
gender differences. Condor 94:720-733.

Sferra, S. j., T. E. Corman, C. E. Paradzick, J. W.
Rourke, J. A. Spencer, and M. W. Sumner. 1997.
Arizona Partners in Flight southwestern Willow
Flycatcher survey; 1993-1996 summary report.
Nongame and Endangered Wildlife Program
Technical Report 113. Arizona Game and Fish
Department, Phoenix.

Sogge, M. K., R. M. Marshall, S. J. Sferra, and T.
J. Tibbitts. 1997. A southwestern Willow Fly-
catcher natural history summary and survey pro-
tocol. National Park Service Technical Report
NPS/NAUCPRS/NRTR-97/12. Flagstaff, Arizona.
U.S. Fish and Wildlife Service. 1995. Final rule deter-
mining endangered status for the southwestern Wil-
low Flycatcher. Federal Register 60:10694-10715.
Walkinshaw, L. H. 1966. Summer biology of Traill’s
Flycatcher. Wilson Bull. 78:31—46.

576

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Wilson Bull.. 1 1 1(4), 1999, pp. 576-577

Mortality of an Adult Veery Incurred During the Defense of Nestlings

David I. King'

ABSTRACT — Cost-benefit analyses of the adaptive
significance of nest defense in birds are based on the
assumption that parent birds attacking nest predators
risk serious injury or death. However, there are few
published records of adult birds dying during nest de-
fense. I found an adult male Veery (Catharus fiisce-
sens) dead in circumstances indicating that the bird
died while defending his nest. This observation sup-
ports speculation that adult birds risk injury or death
in the course of nest defense, and lends support to
explanations of variability in nest-defense behavior
that are presented in terms of cost-benefit analysis and
the optimization of fitness. Received 11 Jan. 1999, ac-
cepted 27 May 1999.

On 27 June 1997 I located a Veery {Catha-
rus fusee sens) nest containing 3 eggs 1.6 m
from the ground in a 9 year-old clearcut in the
White Mountain National Forest (43° 58' N-
70° 97' W) in north-central New Hampshire.
On 1 1 July I approached the nest and ob-
served an adult on the rim of the nest with its
tail tilted up in an unusual posture. The adult
was dead, the body cold. Autopsy revealed
that the adult was a male and that the body
had numerous puncture wounds, one on the
central-posterior region of the right pectoralis
muscle and four more on the back in the vi-
cinity of the synsacrum. The four puncture
wounds on the back were symmetrically ori-
ented on either side of the spine, 1. 5-2.0 cm
apart and up to 7 mm deep, and probably
caused the death of the bird. The depth and
spacing of the wounds suggest that the pred-
ator may have been a Sharp-shinned Hawk
(Accipiter striatus), a species known to attack
Veery nests (Day 1953).

The adult was missing all of its secondary
feathers and all but three rectrices. The miss-
ing feathers were scattered over the nest and
the immediate vicinity of the nest bush. An-
other adult, presumably the female, was heard

‘ Dept, of Forestry and Wildlife Management, Univ.
of Massachusetts Amherst, Amherst, MA 01003;
E-mail: daveking@forwild.umass.edu

alarm calling nearby. The nest cup was tom
down and had come to rest on some branches
of the nest bush below the original level of
the nest. Two nestlings were found prone on
ground beneath the nest. The nestlings were
cold, and gaped weakly. Based on previous
experience with the nesting phenology of this
species, I estimated the nestlings to be at least
three days short of fledging. The original
clutch size was three eggs. It is unknown
whether the third egg hatched. No sign of a
third nestling or eggshells was found in the
vicinity.

On the basis of several lines of evidence, I
conclude the male Veery was killed in the
course of nest defense. First, based on the
spacing of the puncture marks, the predator
was evidently large enough to consume a Vee-
ry (K. Doyle, pers. comm.). Under these cir-
cumstances, it is difficult to conceive of a sce-
nario in which a predator would kill the adult,
leave it on the nest, attempt to depredate the
nest, and subsequently leave. In contrast, the
disposition of the male, and of the nest and its
contents are all consistent with the hypothesis
that the adult was killed in association with
nest defense. The adult was killed yet not con-
sumed, indicating that it was not the original
target of the predator. Furthermore, the nes-
tlings were not consumed, suggesting that the
predator was interrupted during the predation
event. Veerys have been observed successful-
ly defending nests by striking the predator
with their wings (Nice 1962; Pettingill 1976,
pers. obs.), which would account for the sym-
metrical loss of wing feathers I observed. It
could be argued that the male was the original
target of the predator, was wounded else-
where, and returned to the nest seeking a se-
cure hiding place, were it not for the fact that
the nest had been attacked.

Cost-benefit analyses of the adaptive sig-
nificance of nest defense in birds are based on
the assumption that birds attacking predators
during the course of nest defense are at some

SHORT COMMUNICATIONS

577

risk of injury or death (Montgomerie and
Weatherhead 1988). However, observations of
nest predation events under natural conditions
are rare (Pettingill 1976) and observations of
attacks on parent birds by predators during the
course of nest predation are even more scarce.
Brunton (1986) observed a Killdeer {Char-
adrius vociferous) killed by a red fox (Vulpes
vulpes) while performing a distraction display.
This observation of a Veery confronting a
predator at the cost of its own life during ac-
tive defense of the nest is to my knowledge,
unprecedented. This observation supports
speculation that adult birds assume risk of in-
jury or death in the course of nest defense
(Curio and Regelman 1985), and lends sup-
port to explanations of variability in nest-de-
fense behavior that are couched in terms of
cost-benefit analysis and the optimization of
fitness (Montgomerie and Weatherhead 1988).

ACKNOWLEDGMENTS

I thank K. Doyle of the Vertebrate Museum at the
University of Massachusetts Amherst for assistance in
conducting the autopsy, and D. Albano, R. DeGraaf,
and C. Griffin for commenting on the manuscript.

LITERATURE CITED

Brunton, D. H. 1986. Fatal antipredator behavior of
a Killdeer. Wilson Bull. 98:605-607.

Curio, E. and K. Regelman. 1985. The behavior and
dynamics of Great Tits (Purus major) approaching
a predator. Z. Tierpsychol. 69:3-18.

Day, K. C. 1953. The home life of the Veery. Bird-
Banding 24:100-106.

Montgomerie, R. D. and R J. Weatherhead. 1988.
Risks and rewards of nest defense by parent birds.
Q. Rev. Biol. 63:167-187.

Nice, M. M. 1962. Observations on breeding behavior
of Veeries in Michigan. Bird-Banding 33:114.
Pettingill, O. S., Jr. 1976. Observed acts of predation
on birds in northern lower Michigan. Living Bird
15:33-41.

Wilson Bull., 111(4), 1999, pp. 577—581

Relationships of Clutch Size and Hatching Success to Age of Female

Prothonotary Warblers

Charles R. Blem,'- Leann B. Blem,' and Claudia I. Barrientos'

ABSTRACT. — We obtained 1033 clutch sizes from
281 known-age female Prothonotary Warblers (Pro-
tonotaria citrea) nesting in nest boxes at Pre.squile Na-
tional Wildlife Refuge in eastern Virginia from 1987
through 1998. Prothonotary Warblers typically nested
twice during each breeding season; first clutches of all
birds averaged 1.01 eggs greater than second clutches
[4.96 ± 0.72 (SD) vs 3.94 ± 0.55]. Clutch size was
significantly smaller in first nests of one-year-old war-
blers (4.64 ± 0.48) than in first clutches of females
two to eight years old (5.05 ± 0.62). First clutches did
not differ among age classes of birds older than one
year. The mean size of second clutches was not sig-
nificantly different among any of the age classes. One
year old birds initiated laying significantly later than
older birds (125.0 ± 6.4 vs 121.5 ± 7.7; Julian dates).
The average number of infertile eggs in first clutches
was larger in one year old females and differed sig-
nificantly from that of older females (1.01 ± 0.90 vs

' Dept, of Biology, Virginia Commonwealth Univ.,
816 Park Ave., Richmond, VA 23284-2012.

- Corresponding author; E-mail:
cblem@saturn.vcu.edu

0.63 ± 0.87). The number of infertile eggs in second
clutches did not differ significantly with female age.
Significantly fewer eggs hatched in first nests of one
year old birds than in those of older birds (3.75 ± 0.89
vs 4.33 ± 1.09). Received 2 Dec. 1998, accepted 2
May 1999.

The Prothonotary Warbler {Protonotaria ci-
trea) is unusual among wood warblers (Pa-
rulidae) because it nests in secondary cavities.
It shares this trait with only one other member
of the 116 members of the subfamily Paruli-
nae [Lucy’s Warbler {Vermivora luciae): Cur-
son et al. 1994], It is likewise noteworthy
among the birds of the eastern United States
in that it migrates farther than the other small
passerines nesting in secondary cavities. De-
terminants of clutch size of Prothonotary War-
blers therefore may be of interest for compar-
ison with other cavity-nesting passerines and
with other Neotropical migrants. Several re-

578

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

searchers have documented changes in repro-
ductive success associated with age in cavity-
nesting, passerine birds (Bryant 1988, Mc-
Cleery and Perrins 1989, Sternberg 1989,
Seether 1990), but to date there have been few
such data for a warbler.

Prothonotary Warblers are declining in the
United States (Sauer et al. 1997), thus any
knowledge of their demography may be useful
in conservation of the species. In this paper
we provide an analysis of a large set of mea-
surements of reproductive performance of the
Prothonotary Warbler that we have accumu-
lated over the past 12 years. Specifically, we
examine clutch size and infertility in this spe-
cies and ask the question: Does clutch size
and/or infertility of eggs change with age of
females?

STUDY AREA AND METHODS

Our data were obtained from 1987-1998 in swamp
forest along the James River near and on Presquile
National Wildlife Refuge, Hopewell, Virginia (37° 20'
N, 77° 15' W). The habitat of the study area is tidal
swamp in which the dominant tree species are black
gum (Vv.v.vfl .'iylvatica), red maple (Acer ruhrum), and
ash {Fraxinu.s sp.). These swamps have a relatively
harsh environment where tree-surface temperatures
regularly exceed 45° C and tidal amplitude in the
swamp during spring tides often exceeds 1 m. Begin-
ning in March 1987, we placed nest boxes made of
salt-treated pine or red cedar at 100 m intervals along
the creek banks. Box dimensions were 28 L X 9 W X
6 D cm and the entrance hole was 3.8 cm in diameter
(see Blem and Blem 1991, 1992, 1994, for details).
We gradually increased the number of nest boxes in
the study from 141 in 1987 to 300 in 1993—1998.

We checked the contents of boxes 8-15 times during
each breeding season. Old nest material was removed
from the boxes in late winter. Eggs that failed to hatch
were opened to determine fertility and degree of de-
velopment. The present paper includes only those
clutches that were incubated by females and only those
eggs that failed to hatch because of infertility. Clutches
that failed because nests were abandoned were not in-
cluded in the analyses. Parasitism by Brown-headed
Cowbirds (Molotlirus ater) is relatively uncommon at
our .study site (<5% of all clutches), but nests con-
taining cowbird eggs were excluded from our analyses.
We recorded dates of hrst eggs and clutch sizes only
for those nests visited often enough that we could be
certain of laying dates. In some instances, we deter-
mined clutch size but ne.sts were subsequently taken
by predators and we were unable to determine fertility
of the eggs. Sample sizes therefore vary among various
subsets of the data. Because Prothonotary Warblers
typically produce two clutches ctich sea.son (Petit 1989,
Blem and Blem 1992), we divided nests with eggs into

two groups: first clutches in which first eggs were laid
from 25 April through 20 May and second clutches in
which first eggs were laid after 20 May. Recaptures of
banded birds indicated that this division was accurate
for this data set. Nest boxes were originally attached
to trees. We moved them to metal poles in 1995, al-
most completely eliminating predation on nests. Since
then many females have been recaptured during sec-
ond broods in the same nest box. We captured adults
by hand-netting them as they emerged from boxes and
banded all birds with aluminum USFWS bands. In
1998, we used the criteria in Pyle (1997) to age adults,
but older adults and many birds captured before 1998
could only be aged relative to previous captures. We
designated such birds with a + (i.e., 3 + ) to indicate
minimal age, and analyzed age classes accordingly.

All data are reported as means ± SD. Differences
among groups were analyzed using nonparametric
Kruskal-Wallis tests (x' approximation; Zar 1984, SAS
Institute Inc. 1990, Proc NPARIWAY). In all statisti-
cal tests, a probability of 0.05 or less was accepted as
significant (P < 0.05). All analyses were performed
using SAS (Ver. 6; SAS Institute, Inc. 1990) on an
IBM mainframe computer (VM operating system).

RESULTS

Sample size. — Over the 12 years we banded
2968 nestlings and 482 adult females. Birds
were recaptured opportunistically, therefore
sample sizes varied from year to year and in-
dividual age classes came from various years.
We recovered 487 adults and 103 birds banded
as nestlings. Recaptures during the same clutch
were counted only once. Of all females banded
as adults, 47.9% were recaptured at least once
in subsequent years {n = 231). Only 1.7% of
all nestlings were recaptured (n = 50). Some
females {n = 112) were captured over several
years and, therefore, are represented in several
age classes in Tables 1 and 2.

Clutch size.- — We obtained 1033 clutch siz-
es from 281 female Prothonotary Warblers of
known age (Table 1). First clutch sizes dif-
fered significantly among age classes (x^ =
22.4, P = 0.002, df = 7), but there was no
difference in second clutches (x^ = 5.0, P >
0.05, df = 7). One year old female Protho-
notary Warblers laid an average of 0.4 fewer
eggs in first clutches (4.64 ± 0.48; n = 42)
than did older females (5.03 ± 0.73; x^'~
17.7, P < 0.001, df - 1; Table 1). First clutch-
es did not differ among age classes of birds
older than one year (x^ = 4.6, P > 0.05, df =
6). First clutches of all birds averaged 1.01
eggs more than second clutches (4.96 ± 0.72
vs 3.94 ± 0.55; x^ = 356.8, P < 0.001, df =

SHORT COMMUNICATIONS

579

TABLE 1. Clutch
Refuge, Virginia.

size of known-age

female Prothonotary Warblers captured

at Presquile National Wildlife

Age (years)'*

First clutchesh

Second clutches^

Julian date of first egg*’

4.64 ± 0.48 (42)

3.81 ± 0.60 (11)

125.0 ± 6.4 (40)

1 +

4.91 ± 0.72 (395)

3.86 ± 0.62 (128)

124.1 ± 5.5 (381)

4.94 ± 0.79 (33)

3.90 ± 0.32 (10)

123.2 ± 5.6 (32)

5.04 ± 0.79 (211)

4.06 ± 0.46 (65)

121.4 ± 5.8 (205)

5.20 ± 0.42 (10)

4.25 ± 0.50 (4)

120.3 ± 2.5 (10)

3 +

4.97 ± 0.59 (58)

4.12 ± 0.33 (17)

121.3 ± 6.2 (55)

4.50 ± 0.71 (2)

— (0)

121.0 ± 4.2 (2)

5-8

5.18 ± 0.53 (33)

4.00 ± 0.39 (14)

121.0 ± 6.1 (32)

Totals

4.96 ± 0.72 (784)

3.94 ± 0.55 (249)

123.0 ± 5.8 (757)

^ Plus signs indicate that females were that age or older.

^ Numbers in parentheses are sample sizes. Values are means ± SD.

1; Table 1). The mean of second clutches of
one year old birds was 0.24 eggs fewer than
that of females 2-8 years old (3.81 ± 0.60 vs
4.05 ± 0.54), but did not differ significantly
among the age classes (x^ = 9.1, P > 0.05,
df = 5; Table 1). We found no change in
clutch size over consecutive years for 82 of
122 individuals; in 24 cases clutch size in-
creased by one egg, in 14 cases clutch size
decreased by one egg, and in two cases clutch
size decreased by two eggs.

Nest initiation dates. — Nest initiation dates
(Julian) differed significantly among age clas-
ses in the whole data set (x^ = 54.7, P <
0.001, df = 7), but not among age classes of
females 2 years old or older (x^ = 6.2, P >
0.05, df = 5). One year old birds initiated lay-
ing of first clutches significantly later than
older birds (125.0 ± 6.4 vs 121.5 ± 5.8; (x^
= 13.9, P < 0.001, df = 1). The mean date
of nest initiation was remarkably stable

among females greater than two years old,
varying very little from 1 May (Julian date =
121).

Infertility rate. — The number of infertile
eggs in first clutches was significantly larger
in one year old females than in older birds (x^
= 3.9, P < 0.05, df = 1; Table 2), but the
number of infertile eggs in second clutches
did not differ significantly with female age (x“
= 8.7, P > 0.05, df = 7). Infertile eggs were
more frequent in first clutches than in second
(X^ = 8.8, P < 0.01, df = 1). We found no
effect of age on frequency of clutches in
which all eggs hatched, regardless of clutch
size (x' = 0.04, P > 0.05, df = 1). One year
old birds hatched all eggs in 36.0% (18/50) of
their clutches. Older birds hatched all eggs in
35.6% (235/660) of their clutches and there
was no significant difference between the two
groups (x^ = 0.005, P > 0.05, df = 1).

TABLE 2. Number of infertile eggs and nestlings per clutch of known-age Prothonotary Warblers captured
1987-1998 at Presquile National Wildlife Refuge, Virginia.

Number of infertile eggs

Number of nestlings

Age (years)”

First clutches*’

Second clutches*’

First

clutches*’

Second clutches*’

1.00 ± 0.90 (28)

0.43 ± 0.53 (7)

3.70 ±

0.89 (28)

3.43 ± 0.79 (7)

1 +

0.55 ± 0.75 (295)

0.42 ± 0.61 (67)

4.41 ±

0.92 (295)

3.58 ± 0.69 (67)

0.92 ±1.41 (26)

0.50 ± 0.68 (10)

4.04 ±

1.64 (26)

3.40 ± 0.71 (10)

2-h

0.66 ± 0.88 (179)

0.47 ± 0.79 (49)

4.36 ±

0.94 (179)

3.61 ± 0.86 (49)

0.90 ± 1.20 (10)

0.50 ± 0.71 (2)

4.30 ±

1.25 (10)

3.50 ± 0.71 (2)

3 +

0.71 ± 1.01 (45)

0.21 ± 0.43 (14)

4.31 ±

1.18 (45)

3.86 ± 0.36 (14)

0.50 ± 0.71 (2)

— (0)

4.00 ±

0.00 (2)

— (0)

5-8

0.73 ± 0.87 (26)

0.33 ± 0.71 (9)

4.46 ±

1.17 (26)

3.55 ±1.01 (9)

Totals

0.64 ± 0.87 (611)

0.42 ± 0.69 (158)

4.34 ±

1.01 (611)

3.59 ± 0.77 (158)

“ Plus signs indicate that females were that age or older,
h Numbers in parentheses are sample sizes. Values are means ± SD.

580

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

DISCUSSION

In eastern Virginia, Prothonotary Warblers
lay 2-8 eggs per clutch and clutch size varies
significantly among years (Blem and Blem
1992). In 1986, we initiated a study to identify
those factors that were responsible for those
yearly differences. To lend explanatory
strength to our analyses, we attempted to iden-
tify potential sources of variation. For exam-
ple, date of clutch initiation is significantly re-
lated to clutch size (Blem and Blem 1992),
while nest cavity volume, position of the nest,
presence of old nest materials, and female
mass are not (Blem and Blem 1991; Blem et
al. 1999, unpubl. data). Female age is related
significantly to Prothonotary Warbler clutch
size in a manner similar to that found in sev-
eral other cavity-nesting species (Klomp
1970, Saether 1990), particularly the European
Starling {Sturnus vulgaris; Kluijver 1935),
House Martin {Delichon urbica; Bryant
1988), Pied Flycatcher (Ficedula hypoleuca;
Sternberg 1989) and Great Tit (Parus major;
McCleery and Perrins 1989). In all of these
species, older females laid 0.4— 1.0 more eggs
per clutch than birds producing their first
clutch. The increase in mean size of first
clutches with age may be due to several fac-
tors including proximate factors such as in-
creased development of reproductive tracts or
enhanced ability to collect and process energy,
thus allowing females to produce more eggs.

The effects of senility on clutch size are less
well known. There seems to be no document-
ed decreases of clutch size accompanying lon-
gevity in passerine birds. This might be due
to the difficulty in obtaining clutch sizes from
the rare individuals that reach more than a few
years of age. In the present study we obtained
clutch measurements from several birds more
than four years old, including one each from
six, seven, and eight year old birds. One fe-
male originally caught as an adult in 1990 was
recaptured a total of 8 times from 1990—1997.
Her first clutch sizes were 4 (1990), 5 (1992),
5 (1993), 5 (1995), 5 (1996), and 5 (1997). In
1997 the bird was at least eight years of age.
The previous longevity record for Prothono-
tary Warblers was 5 years, 1 1 months (Ken-
nard 1975).

Conservation measures, including intensive
use of predator-proof nest boxes, have been

successful in increasing local abundance of
Prothonotary Warblers (Blem and Blem
1992). However, elimination of predation at
nest boxes could skew age structures of war-
bler populations either by increasing produc-
tion of young or by decreasing mortality of
adult females nesting in boxes. Skewed age
structures could then affect clutch size and in-
fertility rates, this factor must be taken into
account in any analysis of annual variations
in clutch size. For conservation of the species,
reduced clutch size and greater infertility of
young birds seem to have only a modest im-
pact on reproductive performance of Protho-
notary Warblers. Furthermore, effects of se-
nility were not obvious even in relatively old
warblers.

ACKNOWLEDGMENTS

We thank D. Brehmer, C. Cosgrove, S. Horne, B.
Monroe, J. Reilly, R. Reilly, A. Seidenberg, K. Sei-
denberg, and T. Thorp for help in monitoring nest box-
es. We are grateful to B. Brady, refuge manager of
Presquile National Wildlife Refuge, for his continued
cooperation in this research. The North American
Bluebird Society provided funds to place nest boxes
on metal pipes to reduce predation of warblers. The
comments of two anonymous reviewers significantly
improved an earlier version of this manuscript.

LITERATURE CITED

Blem, C. R. and L. B. Blem. 1991. Nest box selection
by Prothonotary Warblers. J. Field Ornithol. 62:
299-307.

Blem, C. R. and L. B. Blem. 1992. Prothonotary War-
blers nesting in nest boxes; clutch size and timing
in Virginia. Raven 63:15-20.

Blem, C. R. and L. B. Blem. 1994. Composition and
microclimate of Prothonotary Warbler nests. Auk
1 1 1:197-200.

Blem, C. R., L. B. Blem, and L. S. Berlinghoff.
1999. Old nests in Prothonotary Warbler nest box-
es: effects on reproductive performance. J. Field
Ornithol. 70:95-100.

Bryant, D. M. 1988. Lifetime reproductive success of
House Martins. Pp. 173-188 in Reproductive suc-
cess (T. H. Clutton-Brock, Ed.). Univ. of Chicago
Press, Chicago, Illinois.

CuRSON, J., D. Quinn, and D. Bendle. 1994. Warblers
of the Americas. Houghton-Mifflin Co., Boston,
Massachusetts.

Kennard, j. H. 1975. Longevity records of North
American birds. Bird-Banding 46:55—73.

Klomp, H. 1970. The determination of clutch-size in
birds: a review. Ardea 58:1-124.

Kluijver, H. N. 1935. Waarnemingen over de leven-
swijze van den Spreeuw (Sturnus v. vulf^ciris L.) met
behulp van geringde individuen. Ardea 24:133-166.

SHORT COMMUNICATIONS

581

McCleery, R. H. and C. M. Perrins. 1989. Great Tit.
Pp. 35-53 in Lifetime reproduction in birds (I.
Newton, Ed.). Academic Press, New York.

Petit, L. S. 1989. Breeding biology of Protlionotary
Warblers in riverine habitat in Tennessee. Wilson
Bull. 101:51-61.

Pyle, P. 1997. Identification guide to North American
birds. Part I. Columbidae to Ploceidae. Slate
Creek Press, Bolinas, California.

S^THER, B. 1990. Age-specific variation in reproduc-
tive performance of birds. Current Ornithol. 7;
251-283.

SAS Institute. 1990. SA.SASTAT® user's guide, ver-
sion 6, fourth ed. Vol. 2. SAS Institute, Inc., Cary,
North Carolina.

Sauer, J. R., J. E. Hines, G. Gough, 1. Thomas, and
B. G. Peterjohn. 1997. The North American
breeding bird survey results and analysis. Patux-
ent Wildlife Research Center, Laurel, Maryland.

Sternberg, H. 1989. Pied Flycatcher. Pp. 55-74 in
Lifetime reproduction in birds (1. Newton, Ed.).
Academic Press, New York.

Zar, j. H. 1984. Biostatistical analysis. Second ed.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

Wilson Bull., 111(4), 1999, pp. 581-584

Hybridization Between Clay-colored Sparrow and Field Sparrow in

Northern Vermont

David J. Hoag'

ABSTRACT. — A male sparrow showing hybrid
characteristics between Clay-colored {Spizella pallida)
and Field sparrows (Spizella pusilla) was first observed
in Grand Isle, Vermont, in 1997. In 1998, the same
hybrid defended a territory and mated with a female
Field Sparrow. The pair produced one fledgling. The
hybrid's signature song was composed of the buzzy
notes of a Clay-colored Sparrow rising to a final trill
as if copying a Field Sparrow's accelerating clear whis-
tles. Received 18 Dec. 1998, accepted 9 May 1999.

I have found only two previous records of
Clay-colored Sparrows (Spizella pallida) and
Field Sparrows (Spizella pusilla) cooperating
at a nest. Finch and Smart (1974) mention,
without further details, a Clay-colored Spar-
row found breeding with a Field Sparrow at
Rockefeller Institute, Dutchess County, New
York; “young were taken for study.” The one
example of hybridization presented by Knap-
ton (1994) is the account by Brooks (1980) of
a trio of adults, a male Clay-colored Sparrow
and a pair of Field Sparrows, at a nest near
Millbrook, Dutchess County, New York; how-
ever, “the fledged young appeared identical to
young Field Sparrows.” Carey and coworkers
(1994) refer to the same report as “possible”
hybridization. Hybridization between these

' 173 West Shore Rd., Grand Isle, VT 05458;
E-mail: sr71blbrd@aol.com

two species is not unexpected because of their
close phylogenetic relationship (Patten and
Fugate 1998). Examples exist of apparent
crossbreeding between Clay-colored Sparrows
and other Spizella species, and between Chip-
ping Sparrow (Spizella passerina) and Brew-
er’s Sparrow (Spizella breweri; Knapton 1994,
Pyle and Howell 1996).

Clay-colored Sparrows are rarely reported
in Vermont (Faccio et al. 1997, 1998). In con-
trast, Field Sparrows may be abundant in
proper habitat such as the abandoned over-
grown fields and pastures of Grand Isle, a
town on Lake Champlain in northwestern Ver-
mont. There, on 29 May 1997, I identified a
Clay-colored Sparrow by its song which con-
sisted of two long buzzes. At 09:30, 11:00,
and 16:30 EST, for a total of 30 minutes, I
listened to the sparrow sing from elevated
perches in a grassy clearing surrounded by red
cedar (Juniperus virginiana), staghorn sumac
(Rhus typhina), and common barberry (Ber-
heris vulgaris). This sparrow was relocated
650 m north on 2 June, and last heard on 5
June.

From 23 July through 14 August, 1 ob-
served and recorded a second Clay-colored
Sparrow in a similar clearing 300 m southwest
of the original location. I recorded the songs
on a microcassette recorder and transferred the
songs to a computer using either Creative

582

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

3 kHz

2 kHz

Time 1 s

5 kHz

4 kHz -

3 kHz

2 kHz

5 kHz

4 kHz -

3 kHz

2 kHz

Time 1 s

FIG. 1. Three sonograms illustrate the buzzy songs of the probable Clay-colored/Field Sparrow hybrid. A.
Trilled version of the hybrid Clay-colored Sparrow’s song recorded July 1997. B. Basic song of the same hybrid
Clay-colored Sparrow recorded May 1998. C. Trilled version of the same hybrid sparrow’s song recorded May
1998. The two trilled songs have a very similar pattern to a Field Sparrow’s clear whistles as illustrated by D.
D. A Field Sparrow’s basic song recorded May 1998.

Technology’s Sound Blaster Pro Voice Editor
version 2.08, or Microsoft Windows Sound
Recorder version 3.1. The sonograms were
created in Windows 3.1 with Spectrogram
(version 2.3; Horne 1995).

I assumed this second bird was a different
Clay-colored Sparrow because its song, 5-8
short buzzes, was extraordinarily different
from the first bird’s song of two long buzzes.
The second sparrow’s song frequently ended
with a buzzy trill that varied in duration (Fig.
lA). With the added trill, the song resembled
a Field Sparrow’s basic song (Fig. ID). The
sonograms show the similarity of the song
pattern to a Field Sparrow song, and show the
dissimilarity of the buzzy notes to the clear
whistles of a Field Sparrow. The mimicry ei-
ther was learned from Field Sparrows in the
surrounding area (Knapton 1994) or perhaps
was a product of hybridization.

Although the face markings of the second

sparrow seemed somewhat indistinct, the
plumage pattern was generally compatible
with Clay-colored Sparrow. The wide central
stripe and the streaking on the crown were
typical of Clay-colored Sparrow. The shapes
of the bill and the tail were also representative
of Clay-colored Sparrow. However, the red-
dish tint of its plumage and the pink color of
its entire bill caused me and other observers
to accept it as a probable hybrid between a
Clay-colored Sparrow and a Field Sparrow.
Two observers thought the flank color at the
bend of the wing was indicative of Field Spar-
row parentage. The Vermont Bird Records
Committee agreed with the hybrid designation
at its November 1998 meeting (Nicholson,
pers. comm.).

From 7—14 April 1998, Field Sparrows re-
turned to the area. On 28 April, on the same
territory that had been occupied by the hybrid
in July and August of 1997, I found a bird

SHORT COMMUNICATIONS

583

FIG. 2. A Clay-colored Sparrow’s mimicry of a Field Sparrow’s complex song as illustrated by two sono-
grams. A. The new Clay-colored Sparrow’s four-part song recorded 8 July 1998. B. A Field Spanow’s three-
part complex song recorded July 1998.

whose coloration matched that of the hybrid I
had seen in 1997. Also, the bird’s song was
the same distinct vocalization of about eight
buzzes (Fig. IB) with trilled notes occasion-
ally added (Fig. 1C), as heard and recorded in
1997 (Fig. lA). According to Knapton (1994),
Clay-colored Sparrows retain their song type
from one year to the next.

On 23-24 July 1997, the hybrid sang rapid
eight buzz songs interspersed with seven buzz
songs having the added trill. The length of the
trill varied; twice it was very short with only
three notes. The repertoire from 25 July to 3
August consisted of five buzzes heard 14
times, seven buzzes heard 9 times, seven
buzzes with a trill heard 17 times, eight buzz-
es heard over 50 times, and nine buzzes heard
twice. Errors in judging sounds may have af-
fected the true syllable count. The sonograms
verify the softness of the initial buzz. In two
variant seven buzz songs, the fifth syllable
was abbreviated. One long song of faster
buzzing was heard on 7 August. In 1998, the

hybrid’s songs were less variable. I heard few
five buzz songs, and seven buzz songs without
a trill were very scarce. Approximately 25%
of the songs were seven buzzes with a trill;
75% were eight buzzes.

By 3 May 1998, the hybrid had moved
about 100 m north. Singing occurred less fre-
quently during the second half of May when
the hybrid acquired a Field Sparrow as a mate.

I made frequent observations throughout May
and June and found no extra Field Sparrows
within the territory. I did not observe any ex-
tra-pair copulations.

Nest inspections at 09:00—10:30 revealed
no egg on 4 June, one egg on 5 June, two
eggs on 7 June, and three eggs on 8 June, the
first day that the Field Sparrow was brooding.
On 19-23 June, the female spaiTow sat on the
nest, blocking viewing of the nest contents.
Both adult sparrows carried food to the nest
21-26 June. On 28 June, the abandoned nest
contained two infertile eggs. Both adults chap-
eroned me and chipped continuously as I at-

584

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

tempted to view the single fledgling at 07:45
and 13:35. The hybrid carried food on 2 and
5 July and continued to scold me throughout
July. On 26 July, the hybrid guarded the ju-
venile. The two adults and the juvenile were
still in close association.

The female Field Sparrow’s reaction during
my unsuccessful search for a second nest on
26 July indicated that a second brood existed.
I saw the hybrid carrying food again on 5 Au-
gust, a late date for assisting the first brood
juvenile hatched approximately 40 days ear-
lier. In December 1998 I found the second
nest which was obscured by grass within a
thicket of barberry centrally located in the 0.3
ha territory. This second nest contained one
Field Sparrow egg. Since all 1997 Field Spar-
row nests in the surrounding area were de-
stroyed by ice accumulation during three days
of freezing rain in January 1998, this undam-
aged nest provided additional evidence that
the hybrid and its mate raised a second 1998
brood.

I did not hear the hybird singing from mid-
June until 5 July, the day a new Clay-colored
Sparrow began a four day encroachment upon
the territory. The new Clay-colored Sparrow’s
song (Fig. 2A) closely resembled a Field
Sparrow song (Fig. 2B). All songs of this new
Clay-colored Sparrow were identical except
for minor variations in length. The hybrid re-
sponded with its own songs (Fig. lA-C)
through 12 July.

The ability of some emberizid sparrows to
learn other species’ songs (Tasker 1955, Bap-
tista et al. 1981), possibly useful in defense of
territory, may also attract mates from closely
related species. Albrecht and Oring (1995) in-
dicated that the primary function of song for
Chipping Sparrows is mate attraction rather
than territorial defense. The song mimesis
may result in occasional interspecific pairing
of Spizella species. Unpublished accounts of
Spizella mimesis include Chipping Sparrows
and Clay-colored Sparrows singing each oth-
er’s .song (Bailey, pers. comm.) and a Field
Sparrow singing a Chipping Sparrow song
(unpubl. data).

The probable hybrid’s unusual plumage, its
atypical song, its pairing with a female Field
Sparrow, and its intensive parental activity at
the nest imply that crossbreeding between

Clay-colored Sparrows and Field Sparrows
occurs. A DNA study of these individuals
might verify hybridization.

Note added in proof: The hybrid sparrow
returned 5 May 1999 and eventually paired
with a female Field Sparrow 9 July to 2 Au-
gust. His songs in 1999 initially were identical
to those in 1997 and 1998, but stopped in-
cluding the trill during the last half of the
breeding season.

ACKNOWLEDGMENTS

I thank D. Capen for reviewing and assisting with
the manuscript. I thank the referees for their assistance
and suggestions for improving the manuscript.

LITERATURE CITED

Albrecht, D. J. and L. W. Oring. 1995. Song in Chip-
ping Sparrows, Spizella pa.sserina: structure and
function. Anim. Behav. 50:1233-1241.

Baptista, L. E, M. L. Morton, and M. E. Pereyra.
1981. Interspecific song mimesis by a Lincoln
Sparrow. Wilson Bull. 93:265-267.

Brooks, E. W. 1980. Interspecific nesting of Clay-col-
ored and Field sparrows. Wilson Bull. 92:264-
265.

Carey, M., D. E. Burhans, and D. A. Nelson. 1994.
Field Sparrow (Spizella pu.silla). In The birds of
North America, no. 103 (A. Poole and F. Gill,
Eds.). The Academy of Natural Sciences, Phila-
delphia, Pennsylvania; The American Ornitholo-
gists' Union, Washington, D.C.

Faccio, S., J. Nicholson, J. Peterson, and C. Rimmer.
1997. Vermont bird records committee report
1995. Vermont Institute of Natural Science,
Woodstock.

Faccio, S., J. Nicholson, J. Peterson, B. Pfeiffer,
AND C. Rimmer. 1998. Vermont bird records com-
mittee report 1996. Vermont Institute of Natural
Science, Woodstock.

Finch, D. W. and R. Smart. 1974. Clay-colored Spar-
row. Kingbird 24:211.

Horne, R. S. 1995. Spectrogram, version 2.3. URL =
www.intranet.csupomona.edu/~biology/
gram.html

Knapton, R. W. 1994. Clay-colored Sparrow (Spizella
pallida). In The birds of North America, no. 120
(A. Poole and F. Gill, Eds.). The Academy of Nat-
ural Sciences, Philadelphia, Pennsylvania; The
American Ornithologists’ Union, Washington,
D.C.

Patten, M. A. and M. Fugate. 1998. Systematic, re-
lationships among the emberizid sparrows. Auk
1 15:412-424.

Pyle, P. and S. N. G. Howell. 1996. Spizella spar-
rows: intraspecific variation and identification.
Birding 28:374-387.

Tasker, R. R. 1955. Chipping Sparrow with song of
Clay-colored Sparrow at Toronto. Auk 72:303.

Wilson Bull., I 1 1(4), 1999. pp. 585-588

Commentary

A CRITIQUE OF WANG YONG AND FINCH’S
FIEFD-IDENTIFICATIONS OF WIFFOW FFYCATCHER
SUBSPECIES IN NEW MEXICO

John P. Hubbard'

In a recent paper in the Wilson Bulletin,
Wang Yong and Finch (1997; henceforth
Y&F) reported that they subspecifically iden-
tified 83 of 84 Willow Flycatchers {Ernpidon-
ax traillii) captured, banded, and released in
central New Mexico in spring and autumn
1994 and 1995. Given the nature of these sub-
species and the means by which Y&F appar-
ently identified them, I am extremely doubtful
about the reliability of their determinations
and thus the validity of these as scientific data.
The fact is that identifying these taxa is quite
difficult, even for trained taxonomists working
in the laboratory under the best protocols and
conditions. This difficulty stems from a num-
ber of factors, the major one being the per-
vasive subtlety of the plumage-color charac-
ters by which these subspecies mainly differ.
Not surprisingly, these differences are difficult
to describe in words, which is exacerbated by
the fact that none of the available classifica-
tion systems accurately portrays the range of
plumage coloration observed in this flycatcher
(e.g.. Browning 1993). This means that this
species’ plumage-color characters are best ob-
served in specimens (i.e., study or flat skins),
which also provide the best avenue for iden-
tifying subspecies. To do this, one must first
assemble series of skins representing all rel-
evant taxa, as well as such important subcat-
egories as age classes (e.g., adult vs immature)
and seasonal groupings (e.g., spring vs au-
tumn). Then one sorts “unknowns” (which
could include live birds) into subcategories
and compares them to the taxa therein, which
should produce at least tentative subspecific
identifications. In fact, this is the standard lab-
oratory approach for identifying color-based
subspecies, and it is the only means proven

' Route 5, Box 431, Espanola, New Mexico 87532.

reliable for this purpose in the Willow Fly-
catcher.

As my earlier comments suggest, I do not
believe Y&F used the approach described
above in their attempts to identify subspecies
in the Willow Flycatcher. In other words, they
did not take synoptic series of study skins into
the field, against which the birds they captured
were compared to determine subspecific iden-
tities. However, I cannot be 100% certain
about this because the methods section in their
paper is so incomplete and otherwise deficient
one can only guess at many aspects of their
approach. Nonetheless, it seems logical that if
they had used skins as the basis for their iden-
tifications, they would have said so. Given this
assumption, if they did not use skins, how did
they go about identifying their birds to sub-
species? On this matter Y&F are at best
vague, providing a few clues but no definitive
explanations of their identification methodol-
ogy. For example, we are told that they
“. . . adopted the four-subspecies classification
system of Hubbard (1987) and Unitt (1987),
in which “subspecies identity ... is based [in
part] on . . . coloration of the head [= crown]
and neck [= forenape] and its contrast with
the back, and the contrast between the breast-
band and the throat (see Phillips 1948, Hub-
bard 1987, Unitt 1987, Browning 1993).”
Based on this, I assume that Y&F chose lit-
erature descriptions (as opposed to specimen
comparisons) as the basis for their identifica-
tion of Willow Flycatcher subspecies. In ad-
dition, 1 also suspect they converted these de-
scriptions into the color values of Smithe
(1975), as this is the system they used to clas-
sify coloration in birds captured in the field.
Beyond this, one could also speculate on such
matters as (a) how converted values were ac-
tually used to identify birds, e.g., whether in
a dichotomous key, probability table, or other

585

586

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

framework; or (b) what Y&F’s perceptions
were of color characters in various races, giv-
en that no such descriptions were offered by
them. However, I see no purpose in further
speculation concerning these or other aspects
of their methodology. This is because if they
did base their identifications on the literature
rather than specimens, 1 believe the process
became so flawed that the details are irrele-
vant— like rearranging deck chairs on the
sinking Titantic!

The message here is that the literature is no
substitute for specimen comparisons for any-
one attempting to identify Willow Flycatcher
subspecies, at least if attaining the most reli-
able scientific data is the goal. Furthermore,
given logistical and other problems, 1 doubt
even specimen comparisons would consistent-
ly yield reliable identifications of live birds
under field conditions. Not only would it be
unwieldy to take and use museum skins in the
field, but setting up and maintaining constant
conditions (e.g., lighting) would also be dif-
ficult. In addition, except for recaptures, only
one opportunity would be available to identify
each live bird in the field. This means that one
could not reassess identifications at a later
time, which is both frequent and necessary
when studying specimens in the laboratory. In
this regard, photographs and certainly color
readings (e.g., from Smithe 1975) would not
be adequate for such reexaminations because
these do not exactly duplicate colors observed
in the birds or specimens themselves. Given
these considerations, I believe that identifying
subspecies in the Willow Flycatcher is best
done in the laboratory, using study skins ex-
amined under proper protocols and procedures
by people trained in the process. In other
words, this is a task that should be left to an
alpha-taxonomic approach, which is appropri-
ate when one considers that subspecies arose
and largely remain as products of that realm.

Even when approached as outlined above,
the reality is that not every specimen or even
population of this flycatcher can be reliably
assigned to subspecies. Intergradation and
overlap occur in all characters that distinguish
these taxa, so birds exhibiting such character-
istics may be un- or misidentified as a result.
In addition, characteristics in some popula-
tions remain poorly known, mainly because of
the paucity of specimens from these areas. For

example, in the latest revision of the species.
Browning (1993) could only assemble 270
specimens of breeding season adults — includ-
ing fewer than 20 of the endangered subspe-
cies E. t. extirnus of the Southwest. As a con-
sequence, it is not surprising that he ques-
tioned boundaries between four of the five
subspecies recognized in his paper. Even
when populational characteristics are better
known, opinions may differ as regards their
taxonomic treatment. Thus, Browning (1993)
recognized two subspecies (i.e., E. t. trail Hi
and E. t. campestris) as breeding in the region
east of the Rocky Mountains, whereas Unitt
(1987) merged the latter with the nominate
form. Differences in opinion also exist on a
broader scale, such as concerning the overall
number of subspecies recognizable in the Wil-
low Flycatcher. For example, some taxono-
mists maintain that none should be recognized
(e.g., Mayr and Short 1970, Traylor 1979),
while others accept four to six as valid (e.g.,
Phillips 1948, Aldrich 1951, Wetmore 1972,
Oberholser 1974, Unitt 1987, Browning
1993). Thus, although specimen comparisons
provide our only reliable means for identify-
ing subspecies in this flycatcher, this approach
must be used with the clear recognition that it
is just the first step in this very difficult en-
deavor.

Incidentally, the above differences in taxo-
nomic opinion present a problem for those
that rely largely or entirely on the literature
for their knowledge of geographic variation in
this species. That is, how does one choose
which authorities to follow and thus which
viewpoints to accept on this subject? Among
others, one way around this would be to ad-
here strictly to a single point of view, such as
the recent revision of this flycatcher by
Browning (1993). However, Y&F chose not
do this, instead electing to cobble their con-
cept of variation from a variety of sources
(e.g., Phillips 1948, Hubbard 1987, Unitt
1987, Browning 1993). Given the lack of con-
sensus among these sources, this was a ques-
tionable decision. In fact, it would be a chal-
lenge even for people with firsthand experi-
ence with geographic variation in this species,
as seen from the variety of opinions cited
above. As a consequence, it is not surprising
that I would quibble with Y&F’s choices, in-
cluding that of which authorities to follow.

Huhhard • COMMENTARY

587

For example, as indicated earlier, they cited
my unpublished paper (Hubbard 1987) as a
basis for the “four-subspecies classification
system” adopted in their study. However, that
so-called system was actually a cobbling job
itself, my aim being to summarize color char-
acters of various subspecies from the treat-
ments of Phillips (1948), Aldrich (1951), Wet-
more (1972), and Oberholser (1974). As such,
it was not meant either to provide definitive
descriptions of these subspecies or to recom-
mend which should be recognized as valid.
For it to have been otherwise used by Y&F
may seem flattering, but it certainly was not a
sound decision from a taxonomic viewpoint.

Given the flawed nature of their approach,
it is no surprise that Y&F’s findings on Wil-
low Flycatcher subspecies would also be open
to question. For example, when compared
with what is known from specimens (e.g.,
Hubbard 1987), significant differences emerge
on the New Mexico status of three of the four
taxa recognized in that study. (In light of the
relative scientific standing of the two sources,
I would obviously accept the specimen ver-
sion over that of Y&F in every case.) The
most significant difference occurs in the sub-
species E. t. brewsteri (sensu stricto), which
breeds along the Pacific slope of North Amer-
ica. Although occurring regularly in migration
eastward to Arizona (Monson and Phillips
1981), this form has rarely been collected east
and north of that state, e.g., in Utah (Behle
1985), Colorado (Bailey and Niedrach 1965),
Oklahoma (Sutton 1967), and Texas (Ober-
holser 1967). Hard data from New Mexico
clearly conform to this pattern, with only two
(4.7%) of the 43 specimens so attributed in
Hubbard (1987) and even these were some-
what equivocal. By contrast, Y&F identified
33 (39.8%) of their 83 birds as E. t. brewsteri,
which is about 8.5 times more frequent than
reported by Hubbard. Another notable depar-
ture involves the subspecies E. t. traillii (in
which Y&F include E. t. campestris), which
breeds from the Great Plains to the northeast-
ern Atlantic Coast. In the Southwest, E. t.
traillii/campestris occurs regularly in the
plains of eastern Colorado (Bailey and Nied-
rach 1965) and New Mexico (Hubbard 1987),
but it has not been collected as far west as
Arizona (Monson and Phillips 1981). Yet
Y&F reported that 8.4% of their birds were

this form, even though the the middle Rio
Grande Valley lies some 200 miles west of the
nearest specimen localities in New Mexico.
Finally is the race E. t. adastus, which breeds
widely in the interior U.S. north of the .south-
western states, through which it passes in both
spring and autumn. In New Mexico, it com-
prised 25.6% of the specimens reported by
Hubbard (1987), compared to 10.8% in Y&F’s
sample.

As for the fourth subspecies {E. t. extimus),
Y&F identified 34 (41.1%) of their birds as
this form, compared to the 48.8% from
throughout New Mexico by Hubbard (1987).
Thus, on the face of it, their findings would
seem not to differ significantly from what is
known from specimens of this taxon. How-
ever, the number of questionable literature re-
cords of this subspecies suggests it may be
more subject to misidentification than certain
other forms, such E. t. brewsteri and E. t.
traillii (both sensu lato). Birds that might be
mistaken for E. t. extimus could include sun-
bleached or worn individuals of other races,
as well as pale variants of E. t. adastus, in-
tergrades between the latter and E. t. extimus,
and carelessly-examined E. t. campestris. If so
misidentified, such instances could help ex-
plain records of E. t. extimus from areas out-
side its known breeding range, such as the
northern two-thirds of Colorado (Bailey and
Niedrach 1965) and Texas east of the Trans-
Pecos region (Oberholser 1974). As for New
Mexico, 1 am dubious of E. t. extimus records
from the eastern plains, such as two speci-
mens reported in Hubbard (1987) from Roo-
sevelt County. In addition, 1 have definitely
reidentified two of the purported E. t. extimus
from that report, one from San Juan County
(= E. t. adastus > extimus) and another from
Socorro County (= E. t. extimus > adastus).
Of course, as mentioned earlier, we do not
have the luxury of reexamining E. t. e.xtimus
(or other subspecies) reported by Y&F, so
their identifications cannot be reassessed in
light of potential sources of misidentification.
Given this and their flawed methodology, 1 see
no reason to regard their findings on this form
as any more acceptable than those on the other
races reported in their paper. As a final point,
Y&F make no mention of the differences be-
tween their findings on the various subspecies
and the specimen record as discussed above.

588

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

While the need for this would not have been
obvious as regards E. t. extimus and perhaps
even E. t. adastus, this could hardly have been
the case with E. t. traillii and especially E. t.
brewsteri.

To summarize, geographic variation in the
Willow Flycatcher mainly involves subtle dif-
ferences in plumage coloration, concerning
which taxonomists disagree in terms of the
number of subspecies that should be recog-
nized. Anyone contemplating identifying
these subspecies should do so with these ca-
veats in mind, as well as by approaching the
process through the use of specimen compar-
isons— preferably in the laboratory under con-
trolled conditions and with proper training in
alpha-taxonomic procedures. Given that
Y&F’s approach appears to have been other-
wise, I submit that their field identification of
these subspecies cannot be regarded as a bona
fide assessment of this parameter in the birds
they processed in New Mexico in 1994 and
1995. Furthermore, for those that would use
their subspecific findings, I urge them to do
so with extreme caution to say the least. Be-
yond this, I would like to state that as an al-
pha-taxonomist, I am dismayed that a study
with such a flawed approach to subspecies
identification could make its way into print in
a major ornithological journal. To wit, orni-
thology has come to rely almost entirely on
non-specimen data for monitoring the distri-
bution and status of birds on this planet. While
not necessarily a bad thing, sometimes we
may fail to recognize the very real limitations
of such data. No better example of this exists
than as regards the identification of difficult
taxa, of which subspecies in Empidonax trail-
lii provide a perfect case in point.

LITERATURE CITED

Aldrich, J. W. 1951. A review of the races of Traill’s
Elycatcher. Wilson Bull. 63:192-197.

Bailey, A. M. and R. J. Niedrach. 1965. Birds of
Colorado, vol. 2. Denver Museum of Natural His-
tory, Denver, Colorado.

Behle, W. C. 1985. Utah birds: geographic variation
and systematics. Utah Mus. Nat. Hist. Occ. Pub.
5:1-147.

Browning, M. R. 1993. Comments on the taxonomy
of Empidonax traillii West. Birds 24:241-257.

Hubbard, J. P. 1987. The status of the Willow Fly-
catcher in New Mexico. Unpublished report. New
Mexico Dept, of Game and Fish, Santa Fe.

Mayr, E. and L. L. Short, Jr. 1970. Species taxa of
North American birds. Nuttall Ornithithological
Club, Cambridge, Massachusetts.

Monson, G. and a. R. Phillips. 1981. Annotated
checklist of the birds of Arizona. Univ. of Arizona
Press, Tucson.

Oberholser, H. C. 1974. The birdlife of Texas, vol.
2. Univ. of Texas Press, Austin.

Phillips, A. R. 1948. Geographic variation in Etnpi-
donax traillii. Auk 65:507-514.

Phillips, A. R., J. Marshall, and G. Monson. 1964.
The birds of Arizona. Univ. Arizona Press, Tuc-
son.

Smithe, F. B. 1975. Naturalist’s color guide. American
Museum of Natural History, New York.

Sutton, G. M. 1967. Oklahoma birds. Univ. of
Oklahoma Press, Norman.

Traylor, M. A., JR. 1979. Check-list of birds of the
world, vol. 8. Harvard Univ., Cambridge, Massa-
chusetts.

Unitt, P. 1987. Empidonax traillii extimus'. an endan-
gered subspecies. West. Birds 18:137-162.

Yong, W. and D. M. Finch. 1997. Migration of the
Willow Flycatcher along the Middle Rio Grande.
Wilson Bull. 109:253-268.

Wetmore, a. 1972. The birds of the Republic of Pan-
ama, part 3. Smithsonian Misc. Coll. 150(3): 1-
631.

Vonf; and Finch • RESPONSE

589

Wilson Bull., 111(4), 1999, pp. 589-592

RESPONSE

Wang Yong' and Deborah M. Finch* -

Hubbard (1999) criticizes our paper Migra-
tion of the Willow Flycatcher along the middle
Rio Grande (Yong and Finch 1997), where we
reported aspects of stopover ecology of the
species including timing, abundance, fat
stores, stopover length, and habitat use. Hub-
bard questions our identification of subspecies
of the Willow Flycatcher {Empidonax traillii)
and the methods we used to identify them. He
also attempts to evaluate the accuracy of our
results of subspecies composition by compar-
ing them with data from other researchers. We
welcome and applaud this scrutiny in the hope
that this interchange will stimulate greater in-
terest, research, and capability to distinguish
the phenotypic characteristics of subspecies of
the Willow Flycatcher. Given that the south-
western race {E. t. extimus) of the Willow Fly-
catcher is federally listed as Endangered, re-
liable methods for identifying this subspecies
need to be developed to more effectively con-
serve and recover its populations.

We are aware that the subspecific taxonomy
of the Willow Flycatcher is inconsistent
among taxonomists as are the techniques to
identify subspecies. Consequently, reliable
identification of subspecies is difficult, espe-
cially in field situations. We acknowledge that
issues of taxonomic status, population distri-
butions, and identification methods of subspe-
cies of the Willow Flycatcher should be ex-
plored further. However, Hubbard’s criticisms
of our paper are generally based on erroneous
information as well as incorrect assumptions
about our methods, and they do not alter our
conclusions about Willow Flycatcher stopover
ecology at the species level.

Hubbard’s first criticism focuses on the
methods we used for identifying the subspe-
cies. Rather than using an assemblage of sub-

' USDA Forest Service, Rocky Mountain Forest Re-
search Station, 2205 Columbia, SE, Albuquerque, NM
87106.

^ Corresponding author; E-mail: Finch_Deborah_M/
rmrs_albq@fs.fed.us

species skins as advocated by Hubbard to
identify Willow Flycatcher subspecies in the
field, we relied on descriptions and records of
coloration and morphology published in the
available literature by taxonomists. Contrary
to what Hubbard speculates, we did not con-
vert color descriptions into Smithe’s (1975)
color code values. We based our identification
of back plumage color on the most recent re-
search by Unitt (1987) and Browning (1993).
Using Smithe’s color codes to describe back
plumage, Unitt (1987) writes: “In brewsteri
the green is in the direction of olive green
(color 48), in adastus in the direction of
greenish olive (color 49), and in extimus and
traillii in the direction of grayish olive (color
43). That is, brewsteri is a dark brownish ol-
ive, adastus a dark grayish green, and extimus
and traillii a pale grayish green. . .’’ Browning
(1993) suggested that Smithe’s color system
is problematical because the color swatches
generally are not identical matches for actual
colors. Hence, he used Munsell Color Charts
(1990) to describe the crown and back con-
trast for his specimens. During our fieldwork,
we consulted both Unitt’s (1987) color codes
for subspecies’ back color and Browning’s
color contrast scores between crown and back.
Although Hubbard suggests that live speci-
mens have some disadvantages, we counter
that the plumage coloration of live birds is
more likely to be true to type than skin spec-
imen plumage that may have faded. If our hy-
pothesis that the coloration of fresh plumage
differs from that of faded plumage is correct,
then data collected from live specimens may
be more reliable, or at least not less reliable,
than results obtained from study skins. Birds
occasionally called or sang in our study after
being released. Information about song and
call characteristics were also recorded when
possible. Such data are available from living
flycatchers but not from skins. Sedgewick’s
(pers. comm.) preliminary analyses of Willow
Flycatcher song and call signatures collected

590

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

in different regions suggest that E. t. extimus
song structure can be distinguished from that
of its northern conspecifics and we used this
kind of data to aid identification also.

We did not rely solely on coloration for
subspecies identification, contrary to Hub-
bard’s second assumption. Unitt (1987) sug-
gested that wing formula (relative length of
primary feather length) can be used to assist
subspecies identification. Of the 305 speci-
mens that Unitt (1987) examined, wing for-
mula distinguished 93% of the E. t. extimus
and E. t. traillii, 88% of the E. t. adastus and
E. t. traillii, and 89% of the E. t. brewsteri
and E. t. traillii. Browning (1993) also applied
wing formula to assess variation in subspecific
characteristics, and his results also demon-
strated that wing formula may be useful for
distinguishing some subspecies although his
sample size was smaller than Unitt’s. Hubbard
himself (1987) noted that E. t. brewsteri was
smaller than other described forms. In the
field, we relied partly on non-overlapping ex-
treme wing measurements to assist in the
identification of this subspecies. In addition,
we measured and recorded more than 30 var-
iables from each individual. Following Unitt
(1987), we used wing formula to aid in iden-
tifying subspecies.

Thirdly, Hubbard (1999) comments that
“even when characteristics of populations are
better known, opinions may differ as regards
their taxonomic treatment” because of limited
sample sizes, interbreeding among populations,
and differences in taxonomists’ methods,
views, and findings. Although we agree that
taxonomists have been inconsistent in their
treatment of subspecific taxonomy, we consid-
er this to be an incentive for finding areas of
common ground among researchers, rather
than a justification for concluding that reliable
identification of subspecies is impossible.
Hubbard states that we should have strictly
adhered to a single view of subspecies tax-
onomy. We followed a single view of subspe-
cies treatment, but we did not credit this single
view to a single researcher. We made it clear
that we adopted the “four subspecies classi-
fication system of Hubbard (1987) and Unitt
(1987).” We warned readers in our Methods
section that: “Given morphological overlap
and hybridization among subspecies, complete
accuracy in identifying subspecies is not

achievable.” Although taxonomists disagree
in their interpretations of within-species vari-
ation and subspecies recognition, there is un-
mistakable agreement about use of a four sub-
species classification among recent research
papers (Hubbard 1987, Unitt 1987, Browning
1993). Hubbard (1987) clearly advocates ac-
ceptance of the four subspecies classification
in his report by stating that: “Given the degree
of agreement among recent workers, I believe
the most prudent course is to accept all of the
above subspecies [i.e., E. t. extimus, brewsteri,
and adastus^ and traillii as valid — at least un-
til more definitive studies are available.” Al-
though in his commentary Hubbard declares
his own report to be a “cobbling job”, its
quality is deemed sound by other authorities.
Indeed, it has been widely distributed and cit-
ed both unofficially and officially by the En-
dangered Species Programs of U.S. Fish and
Wildlife Service regions, by state Game and
Fish Departments, and by other agencies and
ornithologists in the western United States, es-
pecially in the Southwest. Given Hubbard’s
background as a competent taxonomist in
New Mexico and as an officer of the state en-
dangered species branch, his paper is judged
as an authoritative source on the species. For
example, in the process use.d for listing the
southwestern Willow Flycatcher as a federally
endangered subspecies, Hubbard’s paper was
one of the most heavily cited reports by the
U.S. Fish and Wildlife Service (1995).

Unitt (1987) also states that the four races
of E. traillii are valid and may be distin-
guished from each other by “color, wing for-
mula, or both”. Browning (1993) further sep-
arated subspecies E. t. traillii into two popu-
lations: E. t. campestris of the Great Plains
and Great Lakes regions, and E. t. traillii to
the southeast of E. t. campestris. We recently
became aware, that Unitt has conducted fur-
ther research on the same specimens and may
soon be updating his taxonomic treatment (P.
Unitt, pers. com. through J. E. Cartron). These
different authors describe subspecies distri-
butions that are very similar although popu-
lation boundaries are not exactly the same.
U.S. Fish and Wildlife Service relied partly
on these studies to conclude that listing the
southwestern Willow Flycatcher as an endan-
gered subspecies was appropriate.

Fourthly, Hubbard evaluates our results by

Y(>nf> and Finch • RESPONSE

591

comparing our subspecies composition data
with subspecies data from his own and other
reports and sources. While such comparisons
may be valid for the purpose of exploring po-
tential sources of viuriation, the conclusions
that Hubbard draws are incorrect because of
spatial and temporal differences among stud-
ies. Species, subspecies, and population com-
position of migratory birds captured at spe-
cific stopover sites in fall or spring can dra-
matically differ from what is observed at the
same location during the breeding season at
the same location or from other locations dur-
ing migration. For example, the overall spe-
cies composition we detected indicated that
the majority of individuals captured were not
local breeders and many did not even breed
in New Mexico (Finch and Yong 1999). While
we used a standardized, systematic procedure
to sample throughout the entire migration sea-
sons of spring and fall, 1994 and 1995, other
studies that Hubbard (1999) cites and com-
pares to ours were not conducted during mi-
gration seasons and/or did not use standard-
ized procedures. In addition, source studies
cited by Hubbard are heterogeneous in rela-
tion to study goals, year of study, number of
years, geographical location, sampling design,
sampling season, and quality of data, leading
to uncontrolled and unknown factors that in-
validate comparisons with our data set. Our
data are restricted to two sites during two
years in the middle Rio Grande valley of New
Mexico, and thus are only truly comparable to
other data from the same vicinity, year, and
sampling design. Given that different studies,
especially earlier ones, used controversial cri-
teria for classifying and counting their speci-
mens, Hubbard’s argument that our results are
inaccurate because they are not completely
consistent with other studies that, when com-
pared, also yielded dissimilar results is circu-
lar. In our manuscript, we did not make such
comparisons for at least two reasons: (1) our
research focus was on the stopover biology of
the species, not on the taxonomic status of the
subspecies, and (2) other data sources were
not homogeneous or similar enough to draw
comparisons.

Our data and conclusions about the fly-
catcher’s stopover ecology are not dependent
on the validity or accuracy of its subspecies
status or on the methods used to identify sub-

species. Because E. t. extimus is endangered,
U.S. Fish and Wildlife permits for collecting
voucher specimens during migration are not
issued in the Southwest, eliminating the pos-
sibility of having an alpha-taxonomist identify
locally caught specimens to subspecies for the
purpose of setting standards. Because most
current research studies and conservation ef-
forts pertaining to the Willow Flycatcher have
focused on its breeding grounds, the impor-
tance of our research centers on when, where,
and how migration stopover sites in riparian
woodlands along the middle Rio Grande are
used for resting and fat depositions by the spe-
cies. Without understanding the migration
strategy of the species and without justifying
efforts to conserve the stopover habitat that
the species uses, the Willow Flycatcher’s fate
in the Southwest will be jeopardized regard-
less of how perfect or imperfect our ability in
identifying subspecies is.

Throughout ornithological history, subspe-
cies classification and identification have tra-
ditionally been a “problematic” area, partic-
ularly within the genus Empidonax. Uncer-
tainties about subspecies or even species sta-
tus do not negate the value of our migration
research or refute our results about Willow
Flycatcher stopover ecology or intraspecific
variation in migration patterns. We assert that
increased knowledge of the stopover behavior
and energetic condition of the Willow Fly-
catcher is important for understanding the bi-
ology of the species as a whole and that in-
formation about within-species variation is
valuable in conserving the endangered south-
western subspecies.

Our paper and Hubbard’s (1999) critique
have opened up the opportunity to develop
and expand discussion and evaluation of the
different subspecies, the subspecies concept as
a whole, and whether subspecies should be
recognized for the Willow Flycatcher given
the disagreement about their identification and
the difficulty in identifying birds in hand. We
invite and challenge others to contribute ideas
and knowledge to this controversy in the hope
that new or better techniques for identifying
willow flycatcher subspecies may result. Such
discussion or results would certify beyond a
doubt the worthwhile contribution of our pa-
per. Subjecting any paper to a critical com-
mentary, however, automatically attracts the

592

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

notice of additional readers. We are pleased
with the extra attention in the hope that further
research, understanding, and conservation ef-
forts will be directed toward the endangered
southwestern Willow Flycatcher and its dis-
appearing habitat.

ACKNOWLEDGMENT

We thank R. Banks for his review of the manuscript.

LITERATURE CITED

Browning, M. R. 1993. Comments on the taxonomy
of Empidonax traillii (Willow Elycatcher). West.
Birds 24:241-257.

Einch, D. M. and W. Yong. 1999. Landbird migration
in riparian habitats of the middle Rio Grande.
Stud. Avian Biol. In press.

Hubbard, J. R 1987. The status of the Willow Fly-

catcher in New Mexico. New Mexico Dept, of
Game and Fish, Endangered Species Program,
Santa Fe.

Hubbard, J. P. 1999. A critique of Wang Yong and
Finch’s Field-identification of Willow Flycatcher
subspecies in New Mexico. Wilson Bull. Ill:
585-588.

Munsell Color Charts. 1990. Munsell Color Com-
pany, Inc, Baltimore, Maryland.

Smithe, F. B. 1975. Naturalist’s color guide. American
Museum of Natural History, New York.

Unitt, P. 1987. Empidonax traillii extiinu.s: an endan-
gered subspecies. West. Birds 18:137-162.

U.S. Fish and Wildlife Service. 1995. Final mle de-
termining endangered status for the Southwestern
Willow Flycatcher. Federal Register 60:10694—
10715.

Yong, W. and D. M. Finch. 1997. Migration of the
Willow Flycatcher along the middle Rio Grande.
Wilson Bulletin 109:253-268.

Wilson Bull., 1 1 1(4), 19W, pp. 593-599

Ornithological Literature

Edited by William E. Davis, Jr.

THE AUKS. By Anthony J. Gaston and
Ian Jones, illustrated by Ian Lewington, line
drawings by Ian Lewinton and Ian Jones. Ox-
ford University Press, New York, New York.
1998: XX plus 349 pp., 8 color plates with
caption figs., 41 figures, 32 maps, 43 tables
29 photographs. ISBN 0-19-8540320-9. $75
(Cloth). — This book summarizes the biology
of these northern seabirds. The authors have
spent much time studying many species of
auks and are well prepared to summarize the
biology of these northern seabirds. As the au-
thors point out, these birds are primarily ma-
rine organisms but most studies have been
conducted at their breeding sites on land.

The book includes three sections: Plan of
the book. General chapters, and Species ac-
counts. The Plan of the book is comparable to
the introduction in most books: it explains the
intent of the authors and the layout of the
book. This section includes an important table
presenting common nomenclature between
Europe and North America. The plates of the
species are excellent. Literature coverage is
extensive. The book is written in British En-
glish rather than American English.

The General chapters include 7 chapters:
Auks and their world, Systematics and evo-
lution, Distribution and biogeography. Auks
and ecosystems. Social behavior. Chick de-
velopment and the transition from land to
sea, and Populations and conservation.
These chapters include the authors’ under-
standing of the family, comparative analyses,
and syntheses. These are worth the price of
the book. These chapters are strong in that
they deal with auks at sea as well as the
breeding on land. Sections in these chapters
are usually one to a few pages in length. As
summaries, these cover the material ade-
quately but each could be expanded into a
more thorough monograph. Our understand-
ing of these birds has developed slowly be-
cause these are marine birds breeding in re-
mote areas. These summary chapters include
an important historical perspective when

covery of auks and systematics. These sec-
tions allow the reader to put the literature
into an historical perspective. We are only
now increasingly understanding their marine
biology, while often information in the lit-
erature is based on early articles based on
early assumptions.

The authors discuss the biology of these
birds at sea and during breeding at land.
Some comparisons are among most auks and
related species; some comparisons are among
select auk species. These comparisons are
important in keeping these birds in perspec-
tive with other seabirds, and for understand-
ing how they are adapted to their marine en-
vironment. The chapter on populations and
conservation is very timely. It includes sec-
tions on changes in populations, species of
concern, as well as fisheries impacts. While
this is a good review, it should have included
more timely information on immediate is-
sues. The Species accounts are well written.
They include sections on Description, Range
and status. Habitat, food and feeding behav-
ior, Displays and breeding behaviour. Breed-
ing and life cycle, and Population dynamics.
It would have been nice to include a sum-
mary of the conservation status.

This is a fine book with a great deal of in-
formation. However, confusion may result
from inconsistencies among, and in some cas-
es poor organization of, tables, figures, and
maps. Caption formats are different between
the General chapters and the Species accounts.
Timing of activity at colonies is dealt with
inconsistently among the species accounts but
is summarized in a table in the chapter on So-
cial behaviour under the confusing heading of
Activity timing. Furthermore, in the tables
that include measurements in the Species ac-
counts, data are summarized according to ref-
erence numbers with no indication as to where
to find the references, leaving it to the reader
to guess where in the text these reference
numbers are indicated. Maps in the first seven
chapters have a figure number (without men-
tion of Fig.); among the species accounts.

dealing with auks and people, scientific dis-

593

594

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

maps do not contain a figure number or a title,
leaving it to the reader to assume that these
are for the species accounts in which they are
included. Finally, among range maps, the
summer range coloration is often difficult to
distinguish from the background map colora-
tion. These problems will lead to confusion
and detract from an otherwise fine and infor-
mative book.

The seven summary chapters present the
author’s family analyses and strongly based
understanding of these birds. This information
is well worth the book. The species accounts
and difficulty in extracting information are
disappointing. — MALCOLM C. COULTER.

THE ECOTRAVELLERS’ WILDLIFE
GUIDE TO COSTA RICA. By Les Beletsky.
Academic Press, San Diego, California. 1998:
426 pp., 80 color plates, 19 habitat photos.
$27.95.

THE ECOTRAVELLERS’ WILDLIFE
GUIDE TO BELIZE AND NORTHERN
GUATEMALA. By Les Beletsky. Academic
Press, San Diego, California, 1998: 488 pp.
104 color plates, 20 habitat photos. $27.95.

THE ECOTRAVELLERS’ WILDLIFE
GUIDE TO TROPICAL MEXICO. By Les
Beletsky, Academic Press, San Diego, Cali-
fornia, 1999: 497 pp. 104 color plates, 21
habitat photos. $27.95. — Each of these vol-
umes is part of a series recently released un-
der the sponsorship of the Wildlife Conser-
vation Society and authored by Les Beletsky.
As all of the volumes are quite similar they
are reviewed together. The books are intend-
ed as ecological introductions. There is a
brief opening section about ecotourism that
is essentially identical in each of the books,
as is the structure of each book. There is an
ecological overview of the country followed
by chapters on how to use the book and on
conservation issues. The major parts of each
book are chapters on amphibians, reptiles,
birds, and mammals, and, in the Belize book,
on marine life. Unfortunately terrestrial in-

vertebrates are totally ignored. There is noth-
ing about butterflies or other prominent in-
verts in either book. It is disappointing to
look in the index and see reference to ant-
birds, anteaters, antshrikes, ant-tanagers,
antthrushes, but not to ants!

Birds are featured prominently in these
books, the bird chapter being 89 pages long
in Costa Rica, 80 pages long in Belize, 83
pages long in Tropical Mexico (with 50 color
plates on birds in each book). Much of the text
and plates is duplicated among the three vol-
umes. Chapters on taxa discuss natural histo-
ry, breeding, ecological interactions, lore and
notes, and conservation issues. “Profiled”
species are illustrated on the plates. Illustra-
tions are large format and are confined to (ar-
guably) the most common species. For most
readers of The Wilson Bulletm these guides
will not be useful in field identification of
birds because too many species are omitted.
However, these books will help bird watchers
and ornithologists to identify amphibians and
reptiles, many of which are not illustrated in
convenient field-sized books. As with birds,
better and more complete guides exist for
mammals and fish. It is regretable that Belet-
sky selected tropical Mexico, much of which
duplicates what he includes in his Belize
book, rather than western Mexico, where there
are major avifaunal as well as other taxonomic
distinctions.

Beletsky’s text draws heavily from the
published literature and does a credible job
of presenting sound overviews of the subject
matter. The text ranges widely, from sum-
maries of why some birds have evolved to
become frugivores to Mayan legends about
how hummingbirds became so bright. Much
information, though charming in a way, is su-
perfluous to a utilitarian field guide. The lore
and notes sections abound with such quaint
insights as the belief that cows belonging to
a farmer who has destroyed a swallow’s nest
will give bloody milk. We are told (in each
volume) that the Common Raven (Corvus
corax) is the largest passerine, though the Au-
thor readily acknowledges that they do not
occur in Costa Rica, Belize, northern Gua-
temala, or the Mexican Yucatan. I always
thought the Superb Lyrebird {Menura novae-
hollandiae) was actually the largest passerine
(neither does it occur in the Neotropics).

ORINTHOLOGICAL LITERATURE

595

Nonetheless, visitors to the countries profiled
should find the appropriate book useful in
gaining a better understanding (and a fair
dose of trivia) about the local vertebrate
wildlife.— JOHN KRICHER.

A BIRD-FINDING GUIDE TO MEXICO.
By Steve N. G. Howell, illus. by Sophie
Webb. Cornell University Press, Ithaca, New
York. 1999; 365 pp., 54 locality maps. $20
softcover. — This is an exciting addition to the
libraries of couch-birders and those eager to
bird or investigate new corners of Mexico.
This compact and well-organized guide is 6"
X 9" and appears to be solidly bound. Steve,
Sophie, and their occasional fellow travelers
have spent hundreds of days and nights in fine
resorts, dives, and camping sites to gather lo-
cality lists for III sites from Baja to the Yu-
catan. Every locality map is extremely useful
to any visitor, although indicated habitat may
be altered in time.

This is the first detailed information on
many important sites in Mexico, many of
which scream for recognition and protection.
Very few sites are protected in Mexico. Lo-
cality descriptions are well written, to the
point, and include attention to directions with-
in a tenth of a kilometer. Comments on avail-
ability of second-class buses and camping
sites are included for those without their own
(or rental) vehicles and those on limited bud-
gets. The species lists are English common
names only, and wisely run in 4 columns per
page (very economical). Relative abundances
are not given, except for those rarely seen.
Mexican endemics are bolded in the text and
lists.

A variety of taxonomic decisions and En-
glish name selections differ from Seventh
Edition of the A.O.U. Check-List of North
American Birds (1998). Most involve
“splits” as both the world bird species list
and the Dow Jones Average struggle to pass
10,000. Well-differentiated subspecies clus-
ters are given distinct English names in an-
ticipation of these forms being given full spe-
cies status in the future. Numbers gathering
bird listers pressure for all splits, boo all
lumps. One would hope to see less attention
by birders focused on one unit of taxonomy

and more attention to higher (genus) and
lower (subspecies) levels in the future. As for
the names selected, it’s great to see the use
of whitestart replacing redstart for members
of the genus Myioborus. Redstart was created
for Old World thrushes with red on the tail,
and is erroneously used for an American pa-
rulid which could easily be renamed Oran-
gestart, Setophaga ruticilla. There is no red
in the tail of any Myioborus, most have no
red anywhere, and they are no longer placed
adjacent to Setophagal Guy Tudor and 1
came up with whitestart back in the late
1970s as a solution. This British author has
sold his publisher on using grey in place of
the American gray throughout in contradic-
tion of the 1998 A.O.U. Check-List. I am
concerned with using subspecific modifiers in
front of species names; it can create much
confusion. Perhaps it should be Sooty race of
Fox Sparrow or Fox Sparrow (Sooty race) for
Passerella iliaca unaleschensis, not Sooty
Fox Sparrow.

This book is not designed for “lite birders”
doing cruises of the Mexican Riviera, the
whale lagoons, or the islands of the Gulf of
California, nor for those doing single desti-
nation beach resort vacations. It will greatly
aid self sufficient, street-smart birders and or-
nithologists with a taste for adventure and
great birding. Great job Steve! — PETER AL-
DEN.

THE BIRDS OF SONORA. By Stephen M.
Russell and Gale Monson, illus. By Ray
Harm. The University of Arizona Press, Tuc-
son, Arizona. 1998: 360 pp., 2 color plates, 34
b+w figures. ISBN 0-8165-1635-9. $75 hard-
cover.— A long awaited book authored by two
excellent fieldmen who have lived just across
the border in or near Tucson for years. This
is a fairly heavy book {^Vi" X 11") focused on
the ranges, habitats, seasonal abundances, his-
torical records, and current status of over 500
species of birds. It accomplishes its tasks well,
especially with the well-researched range
maps for most species. This work is the first
update in many decades for a province that
should attract many more birders from the
southwestern states. Sonora has tropical de-
ciduous forest around the colonial hilltown of

596

THE WILSON BULLETIN • Vol. HI, No. 4. December 1999

Alamos, great pine-oak woodland in the Sierra
Madre Occidental, and cactus-scapes of the
Sonoran Desert.

The dust jacket features a colorful White-
fronted Parrot {Arnazona albifrons), which is
not reproduced within and lost to users of li-
braries that routinely toss dust jackets. Nice to
see the separate large maps of mountain rang-
es, cities, rivers and reservoirs, and the full
color vegetation map in the introduction. I
question the wisdom of using only the metric
system to indicate elevations and distances.
Outside of scientific circles the metric system
is dying in the U.S., why obfuscate the ma-
jority of the book’s users? Add American
equivalents in parentheses.

The geographical coverage excludes the
Sonoran Islands in the Gulf of California. Isla
Tiburon and other islets have no endemic
birds and this book should have included a
summary of known residents and visitors.
While habitat loss is discussed and lamented,
a rundown of any protected areas and a focus
on areas most in need of protection would
have been welcome.

The appendices cover plants named in the
text, an exhaustive gazetteer useful to any bi-
ologist, and literature cited. Adding a Spanish
common name throughout the text and includ-
ing them in the index is an outstanding step.
One hopes this will be made available in
Spanish for Sonoran citizens.

Wouldn’t it be nice to have similar books
state by state throughout Mexico being re-
searched and published in an orderly fashion?
This would be a good model. — PETER AL-
DEN.

A FIELD GUIDE TO THE BIRDS OF
MEXICO AND ADJACENT AREAS (BE-
LIZE, GUATEMALA, AND EL SALVA-
DOR), third edition. By Ernest Preston Ed-
wards, principal illustrator Edward Murrell
Butler. University of Texas Press, Austin, Tex-
as. 1998: 284 pp. incl 51 color plates. $35
hardcover, $17.95 softcover. — Ernest Edwards
has been a pioneer in producing a series of
bird-finding guides and compact bird field
guides to Mexico. This third edition updates
names and taxonomy, adds a few plates, and
competes with the Peterson’s and Chalif’s

Field Guide to Mexican Birds (Houghton Mif-
flin, Boston, 1973). As a portable pocket
guide it is significantly wider than the Peter-
son’s (fitting fewer pockets), the artwork a bit
stiff and stylized. The text is concise and the
book well indexed between text and plates and
back. The sequence of families and species in
the color plates is jumbled and confusing.
Neither book has range maps. Gray is spelled
gray not grey in Edwards.

At four times the size and weight of the
Edwards or Peterson/Chalif, Steve Howell’s
and Sophie Webb’s A Guide to the Birds of
Mexico and Northern Central America (New
York, Oxford Univ. Press, 1995) is the clear
choice for serious students of Mexican birdlife
with its exhaustive text, superior plates, and
excellent range maps. However, its bulk and
weight will force many to consider leaving it
at home or in the car when deciding which of
the two portable quick reference guides to
take in the field. The Edwards book gives
much less thought to taxonomic changes and
English name modification than does Steve
Howell’s works, a mixed blessing.

The time has come to stop redoing at-
tempts to cover close to 1100 species of birds
from such disparate places in 4-6 countries.
What’s needed is field guides to Pacific slope
birdlife. Gulf and Caribbean slope birdlife,
highland birdlife, etc. No matter where you
are, one’s book has over 50% of its species
totally inapplicable to wherever you are. Far
more useful would be a guide to, say, just
Yucatan, Belize, and Caribbean Guatemala’s
birds. That’s one area that can support such
a smaller geographical focus book. — PETER
ALDEN.

CHECKLIST OF THE BIRDS OF EUR-
ASIA. By Ben E King. Ibis Publishing, Vista,
California. 1997: 105 pp. $19.95 (paper). —
Ben King has produced a concise, and func-
tional birder’s checklist to the contiguous con-
tinents of Europe and Asia, plus their atten-
dant islands, ranging all the way from Iceland
to Novaya Zemlya to Japan, the Philippines,
the Greater Sundas, and Wallacea, two-thirds
of the Old World, in fact. For purposes of dis-
tributional coding. King has partitioned this
vast expanse into ten regions: Europe, the

ORINTHOLOGICAL LITERATURE

597

Middle East, the former Soviet Union, Japan,
the Indian Subcontinent, China, Southeast
Asia, the Greater Sundas, the Philippines, and
Wallacea. The list follows “generally” that of
Peters’ world list, and includes 3062 species.
Taxonomy is reported to be conservative, al-
though the author notes that he has adopted
“some new ideas from the literature as well
as unpublished field studies, especially where
they appear to corroborate my own experi-
ence.” The checklist indicates regional pres-
ence or absence, but provides no information
on status (e.g., breeding, migrant, vagrant) or
abundance. The author notes that introduced
species are, in many cases omitted. Species
endemic to a single one of King’s regions are
indicated by boldface.

The main focus of King’s six page intro-
duction is the construction and clarity of En-
glish names. From King’s strongly worded
statement, it is clear that a main purpose of
this checklist was to provide a standardized
and revised set of English names for the birds
of Eurasia. King notes that names that are elit-
ist or that are difficult to pronounce need to
be changed, as do patronyms memorializing
westerners. Conducting such a nomenclatural
“cleansing” must be a difficult task, indeed.
And it is interesting to compare King’s bird
names against those used in the array of other
checklists and field guides for the region. For
reasons of space I will focus only on a few
widespread Asian passerine species that hap-
pen to also inhabit Wallacea, an area familiar
to me. For this I refer to the English names
used in King and Dickinson’s (1975) Field
Guide to the Birds of South-East Asia, White
and Bruce’s (1986) Birds of Wallacea, Inskipp
et al.’s (1996) Annotated Checklist of the
Birds of the Oriental Region, and Coates and
Bishop’s (1997) Guide to the Birds of Wal-
lacea. At least for these widespread Asian
species, the names King today chooses to use
are all quite reasonable, and, in fact, the no-
menclature across the various publications
produced over a 24 year span is surprisingly
uniform.

It is rare that one can obtain consensus on
English names, even when working in com-
mittee. I am happy to report, however, that, at
least by comparison of 16 widespread song-
bird species that served as my sample exhibit
vary little variation in name in the six sampled

texts. Of these 16, only 4 were represented by
more than one name (Red Avadavat/Straw-
berry Waxbill, Chestnut/Black-headed Munia,
Hair-crested/Spangled Drongo, and Eurasian
Tree-Sparrow/Tree Sparrow). One of the birds
with two names, the drongo, is in fact a geo-
graphically variable taxon that may constitute
more than a single species. Thus it is evident
that even for Southeast Asian birds there is
considerable stability and uniformity of En-
glish nomenclature. Still, given that King
deals with 3000 species, there will be plenty
of species with multiple English names in cur-
rent use. Thus, one of the disappointments of
the King checklist is that alternative names are
not listed, probably because of space limita-
tions.

The checklist is completed by a compre-
hensive index of English and scientific names,
which lists all scientific names by species
(e.g., ‘deucophaeus, Dicrurus'") and all En-
glish names by group-name (e.g., “Drongo,
Ashy”). At the very least I would have pre-
ferred to see listing by genus, as well (e.g.,
""Dicrurus leucophaeus"). That complaint
aside, this is a compact, well-produced, and
useful checklist that covers a huge avifauna.
This will be a must buy for many world bird-
ers and ornithogeographers. — BRUCE M.
BEEHLER.

THE HANDBOOK OF BIRD IDENTIFI-
CATION FOR EUROPE AND THE WEST-
ERN PALEARCTIC. By Mark Beaman and
Steve Madge, illus. by Hilary Burn, Martin
Elliott, Alan Harris, Peter Hayman, Laurel
Tucker, and Dan Zetterstrom. Princeton Uni-
versity Press, Princeton, New Jersey. 1998:
868 pp., 291 color plates with captions, 77
other color illustrations dropped in the text
elsewhere, 625 color-coded range maps,
$99.50 (cloth). — This book, originally pub-
lished in Great Britain by Christopher Helm,
is a monumental achievement. It covers al-
most 900 species known to have occurred in
the Western Palearctic, the area defined for the
9-volume Oxford University Press series ed-
ited by S. Cramp et al.. The Birds of the West-
ern Palearctic (1977-1994). This area extends
from Franz Joseph Land and Novaya Zemlya
south to Kuwait, west to the southwest comer

598

THE WILSON BULLETIN • VoL III, No. 4. December 1999

of Morocco, and north to Iceland, Jan Mayen,
and Spitsbergen, including the Azores, Ma-
deira, and Canary Island groups.

All species are illustrated in color, but that
is only the beginning. Many are shown in
flight, at rest, as adults and juveniles, from
above and below, and in several racial forms
as appropriate. About 600 of the species cov-
ered breed in the area and the rest occur only
seasonally or are vagrants. Although a vagrant
may have occurred only once or twice, its
plumages are covered thoroughly, usually in
an illustration dropped in the text, and typi-
cally with several images. Vagrants that occur
with some frequency are included on the
plates with the local species. To provide a per-
spective on the breadth of coverage, the 33
plates depicting 60 species of jaegers, skuas,
gulls, terns, and alcids contain no fewer than
480 images and include at least 12 species that
are vagrants from North America or Asia.
Two plates and 29 images are devoted to the
recently-split Yellow-legged (Lams cachin-
nans), Heuglin’s (L. heuglini), and Armenian
gulls (L. arrnenicus) alone. Swainson’s Hawk
(Buteo swainsoni) has only occurred two or
three times, but it rates eight images including
both pale and dark morphs, perched and in
flight, and even the rare rufous variant. The
plates are supported by detailed discussion of
identification criteria and comparisons be-
tween similar species in the text, including
mention of racial populations where current
taxonomy is in doubt or where geographical
variation is significant.

The same lavish treatment is given every
group, including sandpipers and plovers. A
high percentage of the shorebird species
known for eastern North America is depicted
on the plates right next to the most similar
Western Palearctic species. Even Eskimo Cur-
lew is there (Niimenius horeals). Indeed, spe-
cies such as Semipalmated (Colidns pusilla).
Least (C. minytilla). Western (C. mouri),
Baird’s (C. hairdd), White-rumped (C. fusci-
collis). Stilt (C. himantopu.s) and Pectoral (C.
mekmotos) sandpipers and the two dowitchers
are treated more fully than in all but special-
ized shorebird guides. If anything. Western
Palearctic species are even more fully treated.
For those with a virtual field-identification
death-wish, there are two plates with 32 im-
ages of Phylloscopus warblers that give new

meaning to the notion of confusing (and vir-
tually inseparable) species, including five
forms of Chiffchaff (P. collybita) that may be
made separate species someday.

Though the greatest wealth of detail is re-
served, appropriately, for local species. North
American birders will be impressed with the
description details provided for North Amer-
ican vagrants. Six species of thrushes, 22
wood warblers, 2 tanagers, 15 emberizids, and
5 icterids are covered exhaustively, for ex-
ample. In addition to critical field marks, var-
iations associated with age or sex, voice, and
preferred habitat in each species’ natural range
are covered in detail, just the same as for local
birds.

The taxonomy is relatively conservative
and current to about 1995, a significant
achievement itself when dealing with so many
species. However, the authors have carefully
called attention to many races that may be el-
evated to full species rank in the future: Taiga
and Tundra Bean Goose (Anser fabalis ssp.),
two or more forms of Brant (Branta bernicla).
Common and Black scoters (Melanitta nigra
ssp.). Pharaoh Eagle Owl (Bubo bubo asca-
laphus), two or more Yellow Wagtails (Mo-
tacilla flava), Moroccan Wagtail (Motacilla
alba subpersonata), Sykes’ Warbler (Hippo-
lais caligata rama), the Chiffchaffs mentioned
above, and others.

There are brief but excellent introductory
sections outlining the content of the species
accounts, defining terms, and commenting on
the techniques and pitfalls of field identifica-
tion. Full indices of English and scientific
names are provided. There are also appendices
of 14 recent additions with full descriptions,
another of 50 species intentionally omitted,
and an appendix listing other important dis-
tribution and identification references for the
area covered. As is inevitable in a publication
of such size, there are occasional typographi-
cal errors, but I noted only three or four in
studying the book for over eight hours. Each
of the illustrators has done superlative work,
and though every plate can be called “good”,
in this reviewer’s opinion there are a few
plates that are less successful than others at
capturing a vibrant, lifelike quality to the im-
ages.

It is easy to lapse into superlatives after
only a short acquaintance with this book. It

ORINTHOLOGICAL LITERATURE

599

weighs about five pounds, and at two inches
thick it is too large and heavy to fit conve-
niently in the pocket of a field jacket. But
most active North American birders will want
to have it, despite the high price, even if they

do not always carry it with them. It sets a new
high standard for the part of the world it cov-
ers, a standard not yet met or even closely
approached elsewhere. Highly recommend-
ed.—ALLAN R. KEITH.

Wilson Bull., 111(4), 1999, pp. 600-607

PROCEEDINGS OF THE EIGHTIETH
ANNUAL MEETING

JOHN A. SMALLWOOD, SECRETARY

The eightieth annual meeting of the Wilson Orni-
thological Society was held Thursday, 10 June,
through Sunday, 13 June, 1999, at Colby College in
Waterville, Maine. W. Herbert Wilson chaired the local
committee; support for the conference was provided
by the Special Programs office of Colby College.

The Council met from 13:07 to 18:27 on Thursday,
10 June, and again from 15:35 to 16:30 on Saturday,

12 June, in Room 335, Olin Hall, Colby College. On
Thursday evening there was an informal reception in
Cotter Union for conferees and their guests.

The opening session on Friday convened in Room
101 Keys Hall, being called to order at 08:37 by WOS
President Edward H. Burtt, Jr. After several announce-
ments were made by Local Chair Herb Wilson and by
Scientific Program Chair Ted Davis, President Burtt
introduced Professor Ed Yeterian, Dean of Faculty at
Colby College, who welcomed those in attendance.
Welcoming remarks were also offered by Dr. Miriam
Bennett, Professor Emeritus, Department of Biology,
Colby College. Eollowing these opening remarks.
President Burtt introduced the third annual Margaret
Morse Nice Plenary Lecture, “Intraspecific variation
in the sizes and shapes of birds,” presented by Frances
C. James. The Nice Lecture was followed by the first
business meeting of the Wilson Ornithological Society.

The scientific program included 45 presentations or-
ganized into six paper sessions and one session for the

13 poster presentations. Approximately half of all pre-
sentations were by students. In addition, there were
two workshops, one on post-baccalaureate careers in
wildlife and conservation, presented jointly by the Wil-
son Ornithological Society and the Ornithological
Council, and the other on bird skinning, conducted by
WOS Librarian Janet Hinshaw.

On Friday evening the conferees enjoyed a “down
East” lobster bake at Johnson Pond on the Colby Col-
lege campus. At the conclusion of this meal, the sat-
isfied decapodivores reconvened for a social at the
Cotter Union. Field trips on Friday and Saturday morn-
ings included excursions to Sidney Bog and Colby Ar-
boretum. Several longer field trips were scheduled for
Sunday, including birding along coastal sites in south-
ern Maine, white-water rafting on the Kennebec River,
birding in the greater Waterville area, and a pelagic
trip to Matinicus rock, where those participating
viewed Atlantic Puffins (Fratercula arctica). Razor-
bills (Alca torcla). Black Guillemots {Cepphus grylle),
and several species of procellariiformids.

A social hour preceded the annual banquet, which
was held in Dana Hall. At the conclusion of a fine
dinner President Burtt briefly addressed the conferees,
commended Herb Wilson and the Local Committee for
a pleasant conference venue and Ted Davis and the

Scientific Program Committee for a successful meet-
ing, thanked retiring Members of Council Peter Fred-
erick and Danny Ingold for their service to the Society,
and recognized all student presenters for their contri-
butions. The following awards were presented:

MARGARET MORSE NICE MEDAL

(for the WOS plenary lecture)

Frances C. James, “Intraspecific variation in the siz-
es and shapes of birds.”

EDWARDS PRIZE

(for the best major article in volume 110 of
The Wilson Bulletin)

Sheila Conant, H. Douglas Pratt, and Robert J. Shal-
lenberger, 1998, “Reflections on a 1975 expedi-
tion to the lost world of the Alaka’i and other
notes on the natural history, .semantics, and con-
servation of Kaua’i birds.” Wilson Bull. 110:1-
22.

LOUIS AGASSIZ FUERTES AWARD

Kazuya Naoki, “Community evolution in the An-
dean tanagers of the genus Tangara.'"

PAUL A. STEWART AWARDS

Paul M. Brandy, “Olive-sided Flycatcher habitat use
in managed landscapes.”

Thomas V. Dietsch, “Ecology and conservation of
Neotropical birds in coffee agro-ecosystems of
southern Mexico.”

Amanda D. Rodewald, “Disturbance in forested
landscape: influence of type and magnitude on
forest birds.”

WILSON ORNITHOLOGICAL SOCIETY
TRAVEL AWARDS

Thomas V. Dietsch, “Relating Neotropical birds and
vegetative structure to certification criteria for
coffee agrosystems in Chiappas, Mexico.”

Falk Huettmann, “Wintering Razorbills, Alca torcla,
and auk assemblages in the lower Bay of Fundy,
Canada. Results from two winter surveys 1997/
98 and 1998/99.”

Rachael Z. Jennings, “Spatial and temporal Varia-
tion in the distributions of Calypte hummingbirds
along an elevational tran.sect. Riverside County,
California.”

Karl E. Miller, “Nesting .success of the Great Crest-
ed Flycatcher in natural nests and in nest boxes:
predation rates increase with nest box age.”

Kimberly A. Peters, “Swain.son’s Warbler habitat

600

Annum, rhport

601

selection in a managed bottomland hardwood for-
est in South Carolina.”

ALEXANDER WILSON PRIZE

(for best student paper)

Christopher M. Somers, "Bird depredation of grapes
in a Niagara Vineyard; do predictable trends ex-
ist?”

Selection committee for the Nice Medal-Edward H.
Biirtt., Jr. (chair), John Kricher, William E. Davis, Jr.,
and W. Herbert Wilson; for the Edwards Prize-Robert
Beason (chair), Charles Blem, and Doris Watt; for the
Fuertes and Stewart Awards— Richard Stiel (chair),
James Sedgwick, and Kelli Stone; for the Wilson Or-
nithological Society Travel Award— William E. Davis,
Jr. (chair), and John C. Kricher; and for the Wilson
Prize-William E. Davis, Jr. (chair), Jon Barlow, Robert
Beason, Richard Conner, and Sara Morris.

FIRST BUSINESS MEETING

The first business meeting was called to order by
President Burtt at 10:02 on Friday, 1 1 June, in Room
105 Keyes Hall. Secretary Smallwood then presented
to those who had gathered a synopsis of Thursday’s
council meeting, commenting on the successful efforts
of Herb Wilson and the local committee and of Second
Vice-president Ted Davis and the scientific program
committee. He reviewed the awards offered by the So-
ciety, the winners to be announced at the annual ban-
quet, and reported that Council had enthusiastically in-
creased their monetary values: the prestigious Louis
Agassiz Fuertes Award increased to $2500, the Mar-
garet Morse Nice Award increased to $1000, and the
annual funding for up to four Paul A. Stewart Awards
was increased to $2000. Secretary Smallwood in-
formed the membership that to alleviate any confusion
over two separate WOS awards made in honor of Mar-
garet Mor.se Nice, namely the MMN Award for ama-
teur ornithologists and the MMN Medal for ornithol-
ogists presenting plenary lectures on a lifetime of or-
nithological research, the MMN Award would, in the
spirit of the contribution of amateurs to the field of
ornithology, henceforth be known as the George A.
Hall and Harold E Mayfield Award. Council had also
increased the amount of funding available for student
travel awards to $5000 annually. Secretary Smallwood
announced that Bob Beason had been elected to a third
year as Editor of The Wilson Bulletin, noting that the
journal was not only published under budget, due
mostly to format changes and electronic submissions,
but also published slightly ahead of schedule. The sec-
retary then reviewed the Conservation Committee re-
port on reauthorization of the Endangered Species Act.
The 2000 meeting will be held 27—30 April at Hotel
Galvez in Galveston, at the invitation of the Houston
Audubon Society and the Gulf Coast Bird Observa-
tory; Dwight Peake will be the local host. The 2001
meeting will be held jointly with the Arkansas Audu-
bon Society, 3-6 May, at the University of Arkansas

Continuing Education Center and Fayetteville Hilton;
Doug James will chair the local committee.

Although this information was not available at the
time of the first business meeting, here the Wilson Or-
nithological Society honors the memory of WOS
members who passed away since the 1998 meeting:
Robert E. Ball (North Canton, OH), Roger M. Evans
(Winnipeg, MB), Richard R. Graber (Golconda, IL),
Frances Hamerstrom (Plainfield, WI), Robert R.
Knickmeyer (Hazelwood, MO), H. Elliot McClure
(Camarillo, CA), Henri Ouellet (Hull, QC), Edward F.
Rivinus (Upper Marlboro, MD), and Charles G. Sibley
(Santa Rosa, CA).

The treasurer’s report was then presented by Doris
Watt.

Bob Beason offered the editor’s report.

President Burtt delivered the report of the Nominat-
ing Committee for Peter Stettenheim, Chair, who was
not able to attend; other committee members included
Kenneth Able, Patricia Gowaty, and Ellen Ketterson.
The following slate of candidates was offered: Presi-
dent, John C. Kricher; First Vice-President, William E.
Davis, Jr.; Second Vice-President, Charles R. Blem;
Secretary, John A. Smallwood; Treasurer, Doris J.
Watt, and Members of Council for 1999—2002, Robert
A. Askins, Gary Ritchison, Charles E Thompson, and
Jeffrey R. Walters.

The first business meeting was adjourned at approx-
imately 10:20

SECOND BUSINESS MEETING

The second business meeting was called to order by
President Burtt at 13:31, Saturday, 12 June, in Room
105 Keys Hall, at which time he recalled the report of
the nominating committee to the floor. Calling for ad-
ditional nominations and hearing none. President Burtt
accepted a motion from Dick Banks, seconded by Sara
Morris, that the nominations be closed. That motion
passed unanimously. Jerry Jackson moved that the
slate of officers (but not councilors) be passed unani-
mously by acclimation. After a second from Sara Mor-
ris, it was. Because there were four candidates for three
seats on the Council (two Members of Council for
1999-2002, and one to complete Charles Blem's
1997-2000 Council term) election was by paper ballot,
with the following result: Members of Council for
1999-2002, Robert A. Askins and Jeffrey R. Walters;
Member of Council for 1999-2000. Charles E Thomp-
son.

Dwight Peake, Local Chair, updated those assem-
bled on the 2000 meeting in Galveston. After a brief
discussion about the proposed name change for The
Wilson Bulletin, and remarks on the electronic publi-
cation of the Bulletin and the Ornithological Newslet-
ter, Sara Morris moved for adjournment. Ted Davis
and Peter Frederick seconded the motion, and the
membership indicated with the appropriate body lan-
guage that the motion had indeed passed. This oc-
curred shortly before 14:00.

602

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

REPORT OF THE TREASURER

1 July 1998 through 30 June 1999

GENERAL EUNDS

Balance Eorward ^

Receipts

Regular and Sustaining Memberships $ 29,688.00

Student Memberships $ 4 278.00

Family Memberships $ 112.00

Total Dues j

Subscriptions $ 24,005.00

Newsletter $ 181.00

Back Issues $ 406.25

Page Charges $ 16,911.25

Total Income from Publications $

Contributions: Research Fund $ 291.00

Van Tyne Library $ 560.00

General Endowment $ 30.00

Life, Patrons (Gen Endowment) $ 1,000.00

Unrestricted $ 937.00

Total Contributions $

Royalties $ 1,127.70

Interest from Checking Account $ 490.52

Interest from Endowments $ 24,573.16

Dividends from Dreyfus Acct (reinvested) $ 999.02

Subtotal $

Miscellaneous $

Net OSNA Adjustment $

TOTAL RECEIPTS $

Disbursements
Bulletin Publication

June 1998 $ 17,191.67

September 1998 $ 16,280.53

December 1998 $ 18,239.61

March 1999 $ 16,570.33

Editor’s expenses $ 4,315.07

Total Publication Costs $

OSNA Expenses $ 5,921.00

Officer’s expenses $ 253.95

CPA (tax filing) $ 465.00

Incorporation fee $ 5.00

Editor’s honorarium $ 2,000.00

Editor’s travel $ 500.00

Library $ 1,000.00

Miscellaneous $ 81.80

Total Operating Expenses $

Organizational Awards $ 1,700.00

Ornithological Council Contribution $ 1,000.00

ABC Dues $ 150.00

Total Philanthropies $

TOTAL DISBURSEMENTS $'

Endinf’ Balance $

CASH ACCOUNTS

First Source Bank Checking 1 July 99 $ 61,832.05

Dreyfus Liquid Assets I July 99 $ 22,662. 16

Total Cash on Hand $

Van Tyne Library Accounts

Starting Balance $ 2,531.63

64,264.33

34,078.00

41,503.50

2,818.00

27,190.40

994.35

(680.41)

105,903.84

72,597.21

10,226.75

2,850.00

85,673.96

84,494.21

Annual rkport

603

Receipts . . .
Expenses . .
Ending Balance

SUTTON DESIGNATED ACCOUNT

$ 8,788.79
$ (3,046.13)
$

Endowment Principal as of 1 July 98 $ 66,956.97

Increase in Principal Value 98-99* $ 68,898.72

1998 Earnings* $ 1,998.06

Funds disbursed for Color Plates (2) $ 1,935.00

Balance $

8,274.29

63.06

TOTAE ENDOWMENT FUNDS

1996 Market Value $ 676,029.00

1997 Market Value $ 814,315.00

1998 Market Value (1 July 98) $ 968,838.00

1999 Market Value (1 July 99) $ 1,096,733.81

*Based on 2.9% yield on Mellon account

EDITOR’S REPORT— 1998

During 1998, the editorial office of The Wilson Bul-
letin received 245 manuscripts including 6 that were
provided by Jed Burtt for the Symposium on Neotrop-
ical Ornithology. This is a continued increase in the
number of manuscripts submitted over the past several
years. The acceptance rate was 40% for these manu-
scripts, similar to that of the past. The review process
took an average of 13 weeks and most of manuscripts
were returned to the authors with reviewers’ comments
2-4 months after receipt. We have been using E-mail
for much of the correspondence with authors and re-
viewers. This has resulted not only in monetary sav-
ings from postage, but also has resulted in faster turn
around for manuscripts; in some cases the author re-
ceived a decision within a month of submission.

As a result of the new format and page size of The
Wilson Bulletin, the journal has gone from approxi-
mately 800 pages per volume to 600 pages. The cur-
rent volume contains the same number of manuscript
pages, figures, and tables as previous issues. Most of
the difference in length is because many of the figures
and tables require less space with the 2-column format
than they did with the single column format. The use
of electronic manuscript (disk) submission has resulted
in fewer corrections in the galley stage. Three issues
in 1998 contained color frontispieces and 3 of the 4
issues that are out or in the works for 1999 contain
color frontispieces. We have had some problems with
the printer to get them to make the frontispieces to
bleed to the edge of the paper, but I think those have
been resolved.

I greatly appreciate the assistance of the editorial
board, Clait Braun, Richard Conner, and Kathy G. Beal
for their timely advise on many manuscripts, especially
Kathy for her advice on manuscripts with difficult sta-
tistical problems and for doing the index. Editorial as-
sistants Tara Baideme, Melanine Daniels, John Lamar,
Dante Thomas, and Doris Watt assisted in tracking and
checking the many manuscripts. The State University

Doris J. Watt, Treasurer

of New York at Geneseo and the Biology Department
continue to support the editor and the running of the
editorial office in many ways.

Robert C. Beason, Editor
The reports of the standing committees are as follows;

REPORT OF THE MEMBERSHIP
COMMITTEE

In October 1998, Laurie Goodrich took over as
membership chair from John Smallwood, WOS Sec-
retary. Current members of the WOS membership
committee are Laurie Goodrich, Chair, Hawk Moun-
tain (PA), Jim Ingold of Louisiana State University
(LA), John Smallwood of Montclair State University
(NJ), Amanda Rodewald of Pennsylvania State Uni-
versity (PA), Christine Howell of University of Mis-
souri-Columbia (MO), and Daniel Ingold of Musking-
um College (OH).

The membership poster was displayed during the
last year at the joint OSNA societies meeting in St.
Louis (MO), April 1998, the Raptor Research Foun-
dation meeting in Salt Lake City (UT), October 1998,
and the Pennsylvania Wildlife Society meeting in Wil-
liamsport (PA), March 1999. Brochures were also dis-
played at a few other meetings, including the Penn-
sylvania Society of Ornithology in Wellsboro, May
1998.

Since October 1998, the chair has received at least
ten inquiries from people who have not received their
journal. Most inquiries occurred between March and
May. All inquiries have been forwarded to Allen Press
and/or the OSNA Director, Anthony Bledsoe. Most
people seem to be locating a contact person for mem-
bership via the web site. The chair has not received
any other inquiries for brochures or information during
her tenure.

The membership brochure needs to be reprinted. Be-
fore we undertake a large run, the committee would
like to have Council review the text and layout, etc.,
to make any necessary changes. Ideally, we need to

604

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

make these changes before the August AOU meeting.
Please return all suggestions to Laurie Goodrich or
John Smallwood by June 30. If anyone needs a copy
of a brochure, please contact Laurie by reply to this e-
mail, and she will send one directly. The committee
would like to request that any Council Member at-
tending a meeting other than the AOU consider help-
ing out with the display. The chair will ship the poster
to the site, and the member would only have to put it
up and then repack and ship it back at end of the con-
ference. To volunteer for a meeting, contact Laurie
Goodrich, goodrich@hawkmountain.org. If unable to
put up a display, please consider taking a stack of bro-
chures for distribution tables.

Laurie J. Goodrich, Chair

REPORT OF THE UNDERGRADUATE
OUTREACH COMMITTEE

Two people have joined the committee since our last
meeting in April, 1998. Thomas Knight of Denver,
CO, and Yves de Repentigny, Departement de Scienc-
es Biologiques, Universite de Montreal.

The committee is pleased to announce a Wilson Bul-
letin paper stemming from the workshop on teaching
held at the Guelph meeting in 1992. The citation is:
Burtt, E. H., Jr., and W. H. Wilson, Jr. 1999. A survey
of ornithology courses in North America. Wilson Bull.

1 1 1:287-293.

The Guide to Graduate Programs in Ornithology
that the Committee has compiled continues to be a
valuable resource for undergraduates seeking to pursue
graduate research in ornithology. The guide can be
found on the Wilson Ornithological Society home page
(http://www.ummz.lsa.umich.edu/birds/studies.html).
Some users have found that the information provided
for some programs is outdated. The committee will be
working this year on updating the material in the on-
line guide.

Jed Burtt continues to compile laboratory exercises
in ornithology. These exercises will be made available
at a reasonable cost for any instructor requesting them.

W. Herbert Wilson, Jr., Chair

REPORT OF THE JOSSELYN VAN
TYNE MEMORIAL LIBRARY
COMMITTEE

I am very pleased to be the new Chair of the Jos-
selyn Van Tyne Memorial Library (JVTML) Commit-
tee. As such I would like to take this opportunity to
thank William A. Lunk for his tireless leadership of
the Library Committee. He has been chair of this com-
mittee for 40 years. Since 1958 many changes have
occurred (e.g., computers have become a standard re-
search tool) and under Dr. Lunk’s chairmanship the
Library has grown and modernized along with the
times. Dr. Lunk deserves our heart-felt thanks for help-
ing to ensure that the Library has been able to provide
WOS members and others with ornithological infor-
mation that they need. During my tenure as chair, I

hope to do as good a job, but I do not think I will be
able to do it for 40 years!

My philosophy for managing the JVTML has two
parts. First, the Library needs to continue acquiring
volumes, thereby ensuring that it stays current and thus
as useful as possible. Second, it needs to be as acces-
sible as possible to WOS members and others, while
maintaining the integrity of the Library.

With respect to acquiring volumes, the following
has happened over the past calendar year:

• Loans of library materials to members included
102 transactions to 46 people or institutions. These
loans included 417 books, journals and photocop-
ied articles, with many of the articles going to au-
thors of BNA accounts.

• A total of 201 publications were received from 167
organizations or individuals. These included 120
exchanges, 23 subscriptions, and 24 gifts.

• A few journals have been added, primarily from
Africa, and others dropped because they have ei-
ther ceased or merged.

• 758 items were donated by members and friends.
These donations included 17 books, 522 journal
issues, and 219 reprints, reports and misc. items.

• 17 members and friends donating materials includ-
ed: C. Braun, D. Klem, Jr., J. Hinshaw, O. Komar,
J. Marks, R. Meek, R. Payne, D. Pioli, W. Post, W.
J. Richardson, T. Shane, J. Spendelow, R. Tashian,
E. Walters, R. Whiting, J. Youngman, and H. Zer-
nickow.

• Five institutions also donated materials: L. Birch
for the Edward Grey Institute of Field Ornithology,
J. Buki for the Hungarian Institute of Ornithology,
L. Kiff for the Peregrine Fund, E Lohrer for Arch-
bold Biological Station, and G. Penn for Point
Reyes Bird Observatory.

• New items purchased for $781.52 from the New
Book Fund and for the $742.88 credit with Buteo
Books included 19 books, journal issues, CD’s and
tapes.

• The sale of 209 books, 45 journal issues and 5
color plates, all of which were duplicates, resulted
in $5,082.30 plus a credit of $2900 with Buteo
Books.

• Gifts to other institutions included 216 journal is-
sues to Hungarian Institute of Ornithology, 38 jour-
nal issues to Walter Thiede, and 3 journal issues to
The Peregrine Fund.

As to the accessibility of the materials to members
and others, the following has been accomplished:

• The web site (http://www.ummz.lsa.umich.edu/
birds/wos.html) continues to be enhanced. Journals
currently received are listed on the site as well as
how to access the University of Michigan's on-line
catalogue, which can be used to check holdings.

• The large number of duplicate books for sale from
Helen Lapham’s bequest will soon be listed on the
web. When it is available, a notice will be posted
in the 0,SNA Newsletter.

Annual rkport

605

Due to lack, of space, we need to reduce our stock
of back issues of The Wilson Bulletin. We plan to keep
around 50 to 75 copies of each issue where possible.
We would like to sell (give away, actually) surplus
copies for the cost of postage and handling. Thus far
we have “sold” over 1300 issues. Certainly we do not
want surplus issues to end up in the recycle bin and
ideally we would like these issues to go to institutions
in lesser developed countries. I do not know the best
way to determine which institutions could use them.
Any ideas to help us solve this problem would be most
appreciated.

Needs for the 1999/2000 academic year:

• We need help in identifying institutions in lesser
developed countries that could use copies of back
issues. We currently have suiplus copies of most
issues, which will allow us to provide nearly com-
plete runs.

• An additional part-time student is needed to help
the Library work to reduce significantly its surplus
of back issues of the Bulletin, to get more infor-
mation up on the web, and to send out the large
numbers of orders we will hopefully generate (via
the web) from members and friends wanting to buy
duplicate copies from Helen Lapham’s bequest. I
do not know the particulars of how to obtain funds
to hire a student. I assume that funds would need
to be run through the University of Michigan,
which would mean overhead would be applied.
The amount of overhead would depend on what
the Society’s bylaws say or what is customary for
the Society to pay. (The full indirect-cost rate at
the University is 52.5%, but can range from that
to 0%, depending on what the funder requires.) We
would like to have a student who works at $7.00
an hour for an average of 15 hours a week over 1 1
months, and with 8.55% in benefits (i.e., FICA).
Thus a total of $5,000 (direct cost) would be need-
ed. Certainly if we can find a work/.study student,
then we would need only around 2/3 of that or
$3,400. We anticipate that sales from Helen La-
pham’s bequest will provide that amount of money
and more, but it will not be available ahead of
time.

My job has been quite ea.sy due to the help of Janet
Hinshaw, who manages the day-to-day tasks in the Li-
brary. Indeed, she is the one who provided me with
the figures I am reporting below.

In closing, I want to offer my thanks to those people
who make the JVTML work .so efficiently: Joann Con-
stantinides, the secretary for the Bird Division in the
Museum of Zoology, who handles many of the library
requests; Kari Chciuk, the library work/.study student,
who for 2 years has handled most of the day-to-day
filing and routine jobs, as well as sorting thousands of
reprints; and of course to THE most important person
connected with the Library, Janet Hinshaw, who pro-
vided me with the figures I reported above and who
literally “runs the show” by making sure the library
is kept up-to-date, and as useful to members and

friends as possible. All of these people make my job
quite easy.

Terry L. Root, Chair

The Committee on the Scientific Program, consist-
ing of William E. Davis, Jr., chair, and John C. Kricher,
presented the following program, assisted by session
moderators and workshop organizers Edward H. Burtt,
Jr., Richard N. Conner, Janet Hinshaw, Jerome A.
Jackson, Sara R. Monis, John A. Smallwood, and Su-
san M. Smith.

PAPER SESSIONS

D. J. Albano, Univ. of Massachusetts, Amherst, MA,
“Partial migration in the Belted Kingfisher (Cer-
yle o Icy on).”

J. C. Barlow and S. N Leckie, Royal Ontario Museum,
Toronto, ON, Canada, “Winter frugivory in Gray
Vireos — do eastern populations eat berries too?”
R. E. Brown, USFWS Southern Research Station, Nac-
ogdoches, TX, “Habitat, nest site, and nest box
selection by the Prothonotory Warbler in eastern
Texas.”

R. N. Conner, D. Saenz, and D. C. Rudolph, Southern
Research Station, Nacogdoches, TX, “The value
of Red-cockaded Woodpecker cavity trees for
studying initial attack of pines by southern pine
beetles.”

R. L. Curry and J. Wotanis, Villanova Univ., Villan-
ova, PA, “Northward movement of the Carolina/
Black-capped Chickadee contact zone in south-
eastern Pennsylvania: inferences from song pat-
terns.”

T. ’V. Dietsch and A. H. Mas, Univ. of Michigan, Ann
Arbor, Ml, “Relating Neotropical birds and veg-
etative structure to certification criteria for coffee
agroecosystems in Chiapas, Mexico.”

P. C. Frederick, Univ. of Florida, Gainesville, FL,
“Community structure and population dynamics
of breeding wading birds in the Everglades.”

M. J. Forman, E. H. Burtt, Jr., and J. M. Ichida, Ohio
Wesleyan Univ., Delaware, OH, “Microorganisms
in the plumage of Galliformes.”

M. J. Hartley, Univ. of Maine, Orono, ME, “Effects
of partial cutting on avian community composi-
tion.”

F. Huettmann, Atlantic Cooperative Wildlife Ecology
Research Network, Fredericton, NB, Canada,
“Wintering Razorbills, AIca torda. and auk as-
semblages in the lower Bay of Fundy, Canada.
Results from two winter surveys 1997/98 and
1998/99.”

J. M. Ichida, J. M. Mann, and E. H. Burtt, Jr., Ohio
Wesleyan Univ., Delaware, OH, “The nest, the
nestling, and the microbe.”

J. A. Jackson, Whitaker Center, Florida Gulf Coast
Univ., Ft. Myers, FL, and W. E. Davis, Jr., Boston
Univ., Boston, MA, “Great Egrets, catfish, and the
economics of birds and humans.”

R. Z. Jennings, Univ. of Texas., Austin, TX, “Spatial

606

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

and temporal variation in the distributions of Ca-
lypte hummingbirds along an elevational transect.
Riverside County, CA.”

J. M. Jonas and J. A. Smallwood, Montclair State

Univ., Upper Montclair, NJ, “The influences of
weather and migration pressure on the growth of
nestling American Kestrels in northwestern New
Jersey.”

E. L. Kershner, J. W. Walk, and R. E. Warner, Univ.
of Illinois, Urbana, IL, “The effect of predator
removal on grassland bird nest success.”

D. L. King, R. M. DeGraaf, USDA Forest Service
Northeastern Forest Experiment Station, and C. R.
Gritfin, Univ. of Massachusetts, Amherst, MA,
“Do predation rates on artificial nests accurately
reflect predation rates on natural bird nests?”

P. A. McDowell, J. M. Ichida, and E. H. Burn, Jr., Ohio
Wesleyan Univ., Delaware, OH, "Streptomyces
and the evolution and microecology of avian
plumage.”

K. E. Miller, Univ. of Florida, Gainesville, FL, “Nest-

ing success of the Great Crested Flycatcher in nat-
ural nests and in nest boxes; predation rates in-
crease with nest box age.”

J. J. Nocera and P. D. Taylor, Acadia Univ., Wolfville,

NS, Canada, “Behavior of failed and non-breed-
ing Common Loons during the summer breeding
season.”

K. A. Peters, North Carolina State Univ., Raleigh, NC

“Swainson’s Warbler habitat selection in a man-
aged botttomland hardwood forest in South Car-
olina.”

Z. Richards and D. Capen, Univ. of Vermont, Burling-
ton, VT, “Landscape effects on the Black-throated
Blue Warbler {Denciroica caende.scensy. a com-
parison of demographics in large forest isolates
and contiguous forest.”

D. C. Rudolph, R. N. Conner, R. H. Johnson, W. G.
Ross, S. J. Burgdorf, R. R. Schaefer, and D.
Saenz, Southern Research Station, Nacogdoches,
TX, “Red-cockaded Woodpecker cavity trees,
southern pine beetle mortality, and pine resin
characteristics.”

D. Saenz, R. N. Conner, C. S. Collins, and D. C. Ru-
dolph, Southern Research Station, Nacogdoches,
TX, “Long-term use of artificial cavities by Red-
cockaded Woodpeckers in eastern Texas.”

J. A. Sedgwick, Biological Resources Div., Fort Col-
lins, CO, “Geographic variation in the song of the
Willow Flycatcher."

C. M. Somers and R. D. Morris, Brock Univ., St. Cath-
arines, ON, Canada, “Bird depredation of grapes
in a Niagara vineyard: do predictable trends ex-
ist?”

R. Stiehl and A. Farmer, Midcontinent Ecological Sci-
ence Center, Fort Collins, CO, “Richness of Pip-
ing Plover foraging habitat on Fire Island, NY.”

J. W. Walk, E. L. Kershner, and R. E. Warner, Univ.
of Illinois, Urbana, IL, “Area and age of habitat
patches and nesting succe.ss in grassland birds.”

J. R. Walters, Virginia Polytechnic Institute and State

Univ., Blacksburg, VA, “Experimental studies of
effects of cavities on territory quality in Red-
cockaded Woodpeckers.”

W. H. Wilson, Jr., Colby College, Waterville, ME,
“Arrival dates of Maine migratory breeding birds:
a trans-century comparison.”

POSTERS

R. C. Banks, National Museum of Natural History,
Washington, DC, “Variation in Greater White-
fronted Geese in the Central Fly way.”

M. T. Bradley and S. R. Morris, Canisius College, Buf-
falo, NY, “Is tail feather shape a reliable indicator
of age in warblers and thrushes?”

E. H. Burtt, Jr., and P. Y. Burtt, Ohio Wesleyan Univ.,
Delaware, OH, “Ice damage to feathers.”

M. Carey and J. Mills, Univ. of Scranton, Scranton,
PA, “Nest-site selection and breeding biology of
Field Sparrows in a rapidly changing old field
habitat.”

M. C. Herceg, J. A. Weiner, and S. R. Morris, Canisius
College, Buffalo, NY, “A seasonal comparison of
the stopover patterns and behavior of migratory
passerines on Appledore Island, Maine.”

J. H. Hull and A. R. Buckelew, Jr., Bethany College,
Bethany, WV, “Group anting in the Common
Grackle.”

J. L. Ingold, Louisiana State Univ., Shreveport, LA,
“Winter birds of small remnant prairies in the Pin-
ey Woods of northern Louisiana.”

E. D. Kennedy and A. McCauley, Albion College, Al-
bion, MI, “Acquisition by Albion College of
prints from 10 original copper plate engravings
used in Alexander Wilson’s “American Ornithol-
ogy”-”

G. J. Robertson, G. Chapdelaine, and R. D. Elliot, Ca-
nadian Wildlife Service, A. W. Diamond and F.
Huettmann, Univ. of New Brunswick, Fredericton,
NB, Canada, “Population status and trends of Ra-
zorbills in Canada.”

C. R. Smith, J. T. Weber, and M. E. Richmond, Cornell

Univ., Ithaca, NY, “Using gap analysis informa-
tion to guide planning for con.servation of birds in
New York state; a comparison of science-based
and expert-opinion approaches.”

D. Westmoreland and M. Moseley, US Air Force
Academy, Colorado Springs, CO, “The cost of
bright egg coloration in American Robin nests.”

D. W. White and E. D. Kennedy, Albion College, Al-
bion, MI, “Do House Wrens identify target eggs
by sight or by feel?”

J. B. Whittier and D. M. Leslie, Jr., US' Geological
Survey, Biological Resources Division, Oklahoma
Cooperative Fish and Wildlife Research Unit,
Oklahoma State Univ., Stillwater, OK, “The effect
of drought on nest success in Least Terns.”

ATTENDANCE

COLORADO: For! CoUin.s, Jim Sedgwick, Dick
Stiehl; Colorado Sprinf>.s, David Westmoreland.

Annual report

607

CONNECTICUT; New Britain, Sylvia Halkin; Sharon,
Elyse Glover.

DISTRICT OF COLUMBIA: Washington, Dick
Banks.

FLORIDA: Fort Myers, Jerry Jackson; Gainesville,
Mary Clench, Peter Frederick, Karl Miller; Tal-
lahassee, Fran James.

ILLINOIS: Bushell, Larry Hood; Centralia, Pricilla
McDowell; Collinsville, Kimberly Peters; Grays-
lake, Scott Hickman; Urbana, Eric Kershner, Jef-
frey Walk.

INDIANA; Notre Dame, Doris Watt.

KANSAS; Manhattan, Dave Rintoul.

LOUISIANA: Shreveport, James Ingold.

MAINE; Bar Harbor, Goodale Wing; Belgrade, Don
Mairs; Fairfield, Miriam Bennett; Farmington,
Sarah Sloane; Orono, Mitschka Hartley; Rich-
mond, Peter Vickery; Unity, Ed Beals; Waterville,
Larkspur Morton, Neal Taylor, Herb Wilson; Wil-
ton, Wendy Howes.

MARYLAND: Chevy Chase, Ellen Paul; St. Mary’s
City, Ernie Willoughby.

MASSACHUSETTS; Boston, Ted Davis; South Had-
ley, Susan Smith.

MICHIGAN: Albion, Dale Kennedy, Doug White; Ann
Arbor, Tom Dietsch; Chelsea, Janet Hinshaw;
Kalamazoo, Richard Brewer.

NEW HAMPSHIRE; Keene, Jon Atwood; West Swa-
zey, Lewis Kibler.

NEW JERSEY: Belleville, Jeffrey Jonas; Cape May,

Tom Parsons; Edison, Robert Colburn; Upper
Montclair, John Smallwood.

NEW YORK: Buffalo, Sara Morris; Cheetkowaga, Je-
anette Weiner; Geneseo, Bob Beason; Grand Is-
land, Maria Bradley; Ithaca, Charles Smith; yo/?n-
son City, Michael Herceg; Utica, Judy McIntyre.

NORTH CAROLINA: Asheville, Lou Weber; Chape!
Hill, Helmut Mueller, Nancy Mueller.

OHIO: Columbus, Sandy Gaunt, Toby Gaunt; Dela-
ware, Jed Burtt, Martin Forman, Jann Ichida; Sun-
bury, Kathy Wildman.

OKLAHOMA; Stillwater, Joanna Whittier.

PENNSYLVANIA: Berwyn, Phil Street; Doylestown,
Jennifer Niese; Kempton, Laurie Goodrich; Scran-
ton, Michael Carey; Villanova, Bob Curry;
Swarthmore, Janet Williams.

RHODE ISLAND; Wakefield, Cecil Kersting.

TEXAS: Austin, Rachel Jennings; Belton, John Cor-
nelius; Galveston, Dwight Peake; Nacogdoches,
Raymond Brown, Dick Conner, Craig Rudolph,
Dan Saenz.

VERMONT; Northfield, Bill Barnard.

VIRGINIA: Arlington, Marcus Koenen; Blacksburg,
Jeff Walters.

WEST VIRGINIA: Bethany, Jay Buckelew; Morgan-
town, George Hall.

CANADA; Fredericton New Brunswick, Falk Huett-
man; St. Catherine’s Ontario, Christopher Somer;
Toronto Ontario, Jon Barlow, Sheridan Leckie;
Wolfville Nova Scotia, Joseph Nocera.

Wilson Bull, 111(4), 1999, pp. 608-609

Acknowledgments

The following individuals graciously served as referees for the volume of The Wilson Bulletin. 1 am deeply
grateful for their assitance and advice — Robert C. Reason, Editor.

C. S. Adkisson, D. G. Ainley, J. L. B. Albu-
querque, T. A1 worth, E. Ammon, D. Ander-
son, J. G. T. Anderson, T. R. Anderson, B. A.
Andres, C. D. Ankney, R. D. Applegate, D.
Arcese, W. J. Arendt, F. Arengo, K. A. Arnold,
R. A. Askins, J. E. Austin, M. L. Avery, M.
C. Baker, N. E. Baldaccini, R. C. Banks, D.
R. Barber, E K. Barker, J. R. Bart, J. M. Bates,
E. Bayne, K. G. Beal, D. L. Beaver, J. C. Bed-
narz, B. M. Beehler, S. R. Beissinger, L. D.
Beletsky, J. R. Belthoff, W. Belton, T. M. Ber-
gin, R W. Bergstrom, S. L. Berman, L. B.
Best, J. Bielefeldt, K. L. Bildstein, M. A.
Bishop, J. G. Blake, C. R. Blem, W. M. Block,
J. Blondel, C. W. Boal, C. E. Bock, E. K. Bol-
linger, E. S Botelho, R. Bowman, G. T. Bra-
den, C. E. Braun, J. D. Brawn, R. M. Brigh-
am, D. J. Bright-Smith, M. C. Brittingham, E.
W. Brooks, B. T. Brown, C. R. Brown, K. M.
Brown, A. H. Brush, T. Brush, A. Burger, D.
E. Burhans, B. E. Byers, V. J. Byre, D.F. Cac-
camise, L. D. Caldwell, S. B. Canale, R. A.
Canterbury, T. Caraco, M. D. Carey, T. W.
Carpenter, N. R. Carrie, E. L. Caton, E. Cham-
berlain, B. R. Chapman, M. K. Chase, A. R.
Clark, R. G. Clark, L. M. Coburn, J. A. Col-
lazo, D. M. Collister, R. N. Conner, P. G. Con-
nors, C. J. Conway, E Cooke, C. J. Counard,
R. J. Craig, D. A. Cristol, J. E Cully, R. L.
Curry, D. R. Curson, T. W. Custer, E J. Cuth-
bert, T. V. Dailey, G. C. Daily, S. J. Daniels,
M. P. Darveau, W. B. Davison, D. K. Dawson,
R. D. Dawson, B. Day, S. Debus, R. M. De-
graaf, J. R. Des Lauriers, T. DeSanto, T. L.
Devault, A. W. Diamond, V. M. Dickison, D.

R. Diefenbach, R. H. Diehl, T. V. Dietsch, J.
J. Dinsmore, R. C. Dobbs, S. L. Dodd, T. Don-
ovan, A. W. Doolittle, S. Dove, R. D. Drob-
ney, D. C. Duffy, J. R. Duncan, J. B. Dunning,
C. R. Dykstra, S. W. Eaton, J. R. Eberhard, L.

S. Eberhardt, M. E. Eddins, W. R. Eddleman,
R. T. Engstrom, A. J. Erskine, R. M. Erwin,

C. Estades, M. R. Evans, T. J. Evans, J. R.
Faaborg, S. Faccio, G. H. Farley, G. Farns-
worth, M. B. Fenton, E. J. Finck, L. D. Flake,

D. J. Flaspohler, P L. Flint, T. B. Ford, M. S.

Foster, A. D. Fox, A. B. Franklin, J. D. Fraser,
H. Freifeld, L. E. Friesen, R. W. Furness, S.
W. Gabrey, J. L Ganey, K. L. Garrett, R. J.
Gates, A. S. Gaunt, D. G. Gawlik, E R. Gehl-
bach, R. P Gerhardt, J. C. Gering, J. A. Ges-
saman, H. G. Gilchrist, C. E. Gill, S. Gill, J.
Glahn, R. S. Gnam, N. M. Gobris, M. Goch-
feld, C. B. Goguen, T. Goldsmith, G. Good-
ing, L. J. Goodrich, C. E. Gordon, B. M. Gott-
fried, C. Graham, W. Gram, J. G. Granlund,

C. L. Gratto-Trevor, G. A. Green, C. H.
Greenberg, J. S. Greenlaw, J. W. Grier, C. R.
Griffin, C.S. Griffiths, K. Groschupf, T. C.
Grubb, R. Grundel, J. A. Grzybowski, K. J.
Gutzwiller, S. J. Hackett, J. H. Haffer, T. M.
Haggerty, S. M. Haig, J. P Hailman, E R.
Hainsworth, D. A. Hall, M. J. Hamas, P B.
Hamel, I. K. Hanski, R. G. Harper, M. J. Har-
tley, D. Haskell, E. W. Hein, H. T. Hendrick-
son, C. J. Henny, G. R. Hepp, J. R. Herkert,
F Hertel, S. K. Herzog, D. O. Hill, S. L. Hilty,
A. Holmes, D. W. Holt, P D. Hooper, J. P.
Hoover, S. L. Hopp, D. J. Horn, C. S. Hous-
ton, J. D. Hubbard, J. Hughes, W. G. Hunt, D.
J. T. Hussell, R. L. Hutto, L. D. Igl, D. J.
Ingold, M. L. Isler, E M. Jaksic, D. A. James,
E C. James, R. A. James, A. P. Jaramillo, R.

L. Jarvis, D. A. Jenni, W. E. Jensen, D. H.
Johnson, L. S. Johnson, M. D. Johnson, J. Jo-
livette, I. L. Jones, J. Jones, K. L. Jones, S.
Jones, D. G. Jorde, S. M. Joy, R. Kane, T. L.
Kast, W. Kendall, R. A. Kennamer, E. D. Ken-
nedy, R L. Kennedy, D. M. Keppie, L. H. Ker-
mott, H. Khanna, A. A. Kinsey, M. Klaassen,
P K. Kleintjes, J. T. Klicka, D. S. Klute, R.
W. Knapton, M. T. Koenen, W. D. Koenig, R.
R. Koford, O. Komar, N. Krabbe, A. W. Krat-
ter, D. G. Krementz, J. C. Kricher, K. S. Kritz,

D. E. Krodsma, J. D. Lambert, R. B. Lanctot,
R. S. Lange, D. V. Fanning, W. E. Lanyon, S.
Lariviere, B. Lauro, J. J. Lawler, R E. Lederle,

M. R. Lein, R. E. Lemon, R. A. Lent, M. L.
Leonard, D. J. Levey, E. T. Linder, B. D. Link-
hart, C. D. Littlefield, J. R Loegering, B. A.
Loiselle, E. R. Loos, P. E. Lowther, R. C.
Lundquist, J. E. Lyons, T. R. Mace, R. H. E

608

ACKNOWLEDGMENTS

609

Macedo, R. E. Maier, T. J. Maier, B. P Mar-
ansky., M. A. Marin, M. A. Marini, J. S.
Marks, C. D. Marti, P. M. Mayer, H. L. Mays,
J. C. Mazourek, K. M. Mazur, T. McCoy, K.
P McFarland, W. B. McGillivray, J. W. Mc-
Intyre, K. McKay, M. R. McLandress, D. B.
McNair, R. McNeil, S. M. Melvin, B. D. Mey-
burg, J. M. Meyers, E. H. Miller, K. E. Miller,

B. A. Millsap, M. Milonoff, D. Mock, A. P
M0ller, R. Montgomery, D. Moore, E R.
Moore, D. C. Morimoto, S. R. Morris, M. L.
Morrison, S. G. Mosa, R. S. Mulivihill, M. T.
Murphy, J. J. Negro, D. A. Nelson, S. A. Nes-
bitt, G. J. Niemi, M. Nogales, E. Nol, C. J.
Norment, R. L. Norton, J. P. O’Neill, P Olsen,
L. W. Oring, C. P. Ortega, H. Ouellet, R. B.
Owen, K. C. Parsons, D. N. Pashley, C. A.
Paszkowski, P. W. Paton, R. B. Payne, G. Pen-
delton, C. M. Perrins, B. G. Peterjohn, T. A.
Peterson, J. Pieman, P. J. Pietz, W. H. Piper,

B. J. Ploger, D. W. Pogue, W. Post, B. Poulin,
A. N. Powell, H. W. Power, S. Pribil, R. O.
Prum, K. L. Purcell, W. H. Pyle, P Radewald,
S. A. Raouf, J. H. Rappole, S. E. Reinert, D.

L. Reinking, L. Reitsma, J. V. Remsen, M. K.
Reynolds, R. E. Ricklefs, R. S. Ridgely, C. M.
Riley, J. D. Rising, G. Ritchison, J. H. Rivera,

C. S. Robbins, M. B. Robbins, G. J. Robert-
son, W. D. Robinson, N. L. Rodenhouse, H.

M. Rogers, R. N. Rosenfield, R. R. Roth, S.
I. Rothstein, S. M. Russell, M. R. Ryan, D.
Saenz, L. Santisteban, R. A. Sargent, G. Sat-
tler, J. A. Savidge, T. L. Scarlett, P E
Schempf, A. Scheuerlein, S. R. Schmidt, J. K.

Schmutz, G. D. Schnell, E. A. Schreiber, T.
Schulenberg, D. Schulter, J. A. Sedgwick, G.
Seutin, C. E. Shackelford, D. H. Shedd, M. A.
Shields, W. G. Shriver, M. Siders, T. R. Si-
mons, C. G. Sims, J. A. Smallwood, C. R.
Smith, D. G. Smith, J. N. M. Smith, M. K.
Sogge, S. S. Soukup, D. A. Spector, J. A.
Spendelow, S. E H. Spofford, D. W. Stead-
man, K. Steenhof, M. A. Stem, R. B. Stiehl,
C. M. Stinson, S. H. Stoleson, D. E Stotz, R

C. Stouffer, W. D. Svedarsky, D. L. Swanson,
P W. Sykes, D. A. Tallman, W. K. Taylor, D.
L. M. Teixeira, R. C. Telfair, E. J. Temeles, G.

D. Therres, G. R. Thomas, C. E Thompson,
C. W. Thompson, J. E. Thompson, S. C.
Thompson, T. L. Tibbits, M. E. Tobin, T Tra-
cy, C. L. Trine, J. D. Tyler, T J. Underwood,
C. Van Riper, W. M. Vander Hagen, D. W.
Verser, P. D. Vickery, M. A. Villard, G. H.
Visser, C. M. Vleck, K. H. Voous, E Vuilleu-
mier, S. J. Wagner, J. W. Walk, G. E. Wals-
burg, C. Ward, D. Ward, I. G. Warkentin, K.
S. Warner, B. D. Watts, L. M. Weber, L. C.
Wemmer, C. G. West, G. B. West, D. West-
moreland, C. J. Whelan, D. Whitacre, C. M.
White, M. J. Whitfield, R. C. Whitmore, D.
A. Wiedenfeld, E A. Wilkinson, D. E. Willard,
J Williams, L. E. Williams, C. N. Willis, W.
Wiltschko, J. C. Wingfield, D. W. Winkler, M.
Winter, L. L. Wolf, D. H. Wolfe, R B. Wood,
R W Woodward, B. L. Woodworth, A. L.
Wright, R. H. Yahner, P. H. Yaukey, Y Yom-
Tov, R. Yosef, S. Zack, G. D. Zenitsky, M. C.
Zicus

Wilson Bull., 111(4), 1999, pp. 610-630

Index to Volume 111, 1999

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. Nomenclature follows the AOU Check-list of North American Birds (1998).
Reference is made to books reviewed, and announcements as they appear in the volume.

Ahurria ahurri, 564-569
abundance

effect of wind turbines on upland nesting birds,
100-104

of breeding birds in old-growth forests, 89-99
of Sitta pusilla after pine plantation thinning, 56-60
Accipiter hicolor, 225
hrevipes, 181-187
cooperii, 7—14, 558
gentilis, 432-436, 475
ni.sus, 181-187
stricitus, 7—14, 92, 558, 575
superciliosus,

Acrochordopus burmeisteri, 412
Actophilornis africanus, 264
Aegolius funereus, 273
age

effect on breeding performance in Dendroica pete-
chia, 381—388

Agelaius phoeniceus, 84, 86, 100-104, 105—114, 131,
254, 348, 554, 555, 558
tricolor, 425
Agriornis Uvula, 530
Aimophila aestivalis, 56
Aix sponsa, 1-6, 108, 109, 469
Alden, Peter, reviews by, 595, 595-596, 596
Alectrurus tricol. I'll
Alouatta palliata, 125

Alvarez A., Jose, see Krabbe, Niels, Morton L. Isler,

Phyllis R. Isler, Bret M. Whitney, , and

Paul J. Greenfield

Alworth, Tom, and Isabella B. R. Scheiber, An incident
of female-female aggression in the House
Wren, 130-132

Amazon, Blue-fronted, see Amazona aestiva
Amazona aestiva, 225, 408
finschi, 488-493
vinacea, 408

Amhassis gymnocephalus, 19
Ammodramus hairdii, 389—396
henslowii, 515-527
humeralis, 228
leconteii, 1 00— 1 04
savannarum. 1 00— 1 04
Anahacerthia amaurotis, 410
Anairetes parulus, 530
Anas acuta, 468, 470

hahamensis haliamensis, 56 1
discor.s, 108, 109, 111, 468, 469

platyrhynchos, 108, 109, 111,469
ruhripes, 108
spp., 17

Anderson, James T, and Thomas C. Tacha, Habitat use
by Masked Ducks along the Gulf Coast of Tex-
as, 119-121

Angehr, George R., Rapid long-distance colonization
of Lake Gatun, Panama, by Snail Kites, 265-
268

Ani, Greater, see Crotophaga major
Smooth-billed, see Crotophaga ani
Anjos, Luiz Dos, and Roberto Bo^on, Bird commu-
nities in natural forest patches in southern Bra-
zil, 397-414
announcement

Lincoln Park Zoo Scott Neotropic and Africa/Asia
Funds, 456

Anodorhynchus hyacinthinus, 225
Anser alhifrons, 167

albifrons frontalis, 166—180
caerulescens, 166, 167, 177
canagica, 1 66
rossii, 166

Ant-Tanager, Red-crowned, see Habia rubica
Antbird, Black-chinned, see Hypocnemoides melano-
pogon

Black-throated, see Myrmeciza atrothorax
Caura, see Percnostola caruensis
Dusky, see Cercomacra tyrannina
Jet, see Cercomacra nigricans
Rio Branco, see Cercomacra carbonaria
Silvered, see Sclateria naevia
Yapacana, see Myrmeciza disjuncta
Anthus spragueii, 389—396
Antilophia galeata, 222
Antwren, Ornate, see Myrmotherula ornatua
Rusty-backed, see Formicivora rufa
White-eyed, see Myrmotherula leucophthalma
Anumhius annumbi, 410
Aphanotriccus capitalils, 124-128
Aphelocoma californica, 252
coerulescens, 285, 348
Aquila adalberti, 475

chrysaetos, 437, 472-477
Ara chloropterus, 225
maracana, 225

Aragones, Juan, Luis Arias de Reyna, and Pilar Re-
cuerda. Visual communication and sexual se-
lection in a nocturnal bird species, Caprimul-

610

INDEX TO VOLUME I 1 I

611

guM ruficollis, a balance between crypsis and
conspicLiousness, 340-345
Aramides cajanea, 225
saracurti, 407
Aratinga aurea, 225
leucophthcdmus, 225
Archilochus colubris. 92, 94
Arenaria interpres, 263

Arendt, Wayne J., see Woodworth, Bethany L., John

Faaborg, and

Arius mudagascariensis, 19

armadillo, nine-banded, see Dasypus novemcinclus
Arremon aurantiirostris, 234
flavirostris, 228, 233
Arremonops rufivirgatus, 234
arrival patterns

of Vireo griseus, 46—55
Asia capensis,

flanitneus [5rr/x domingensis, Asia portoricensis,
Asia domingensis}, 274, 303—313 (Frontis-
piece)
otus, 273-276
spp., 218

Asthenes modesta, 530
Ateles geojfroyi, 125

Athene cunicularia, 273, 310, 434, 569—571
Attagis goyi, 264

Avery, Michael L., Carrie L. Schreiber, and David G.
Decker, Fruit sugar preferences of House
Finches, 84—88
avifauna

at the Paraguayan Cerrado site, Parque Nacional
Seram'a San Luis, depto. Concepcion, 216—
228

of Tabasco, 229—235

Avocet, American, see Recurvirostra americana
European, see Recurvirostra avosetta
awards and grants

Lincoln Park Zoo Scott Neotropic and Africa/Asia
Funds, 456
Aythya affinis, 465
americana. 108, 109
marila, 465—471

Backus, Leslie K., see Bowman, Reed, David L. Leon-
ard, Jr., , and Allison R. Mains

badger, see Taxidea tcvca

Baeolophus hicolor, 92, 94, 139-143, 373, 554, 555,
558

inornatus. 25 1 , 374
spp., 368

wollweheri, 373, 374

Baicich, Paul J., and Colin J. O. Harrison, A guide to
the nests, eggs, and nestlings of North Ameri-
can birds, reviewed, 295—296
Bananaquit, see Coereha fiaveola
banding

returns of Vireo griseus. 46-55
Banks, Joshua B., see Breitwisch, Randall, Amy J.
Schilling, and

Banks, Richard C., see Dove, Carla J., and

Barau, Armand, see Barre, Nicolas, , and Chris-

tian Jouanin

Barre, Nicolas, Armand Barau, and Christian Jouanin,
Oiseaux de la Reunion, reviewed, 300-301
Barrientos, Claudia I., see Blem, Charles R., Leann B.
Blem, and

Barrow, Mark V., Jr., A passion for birds, American
ornithology after Audubon, reviewed, 152-153
Barrow, Mark V., Jr., review by, 453-454
Bartramia longicauda, 263
Baryphthengus martii, 439
ruficapillus, 221, 226

Basileuterus culicivorus. 216, 220, 401, 413
flaveolus, 220, 228
hypoleuciis. 216, 220, 228
leucoblepharus, 401, 413
Batara cinerea, 410

Bay, Michael D., The type B song of the Northern
Parula: structure and geographic variation
along proposed sub-species boundaries, 505-
514

Beaman, Mark, and Steve Madge, The handbook of
bird identification for Europe and the western
Palearactic, reviewed, 597-599
Becard, Gray-collared, see Pachyramphus major
Beehler, Bruce M., review by, 596—597
behavior
aggressive

female-female in Troglodytes aedon, 130—132
response of Poecile atricapillus and Poecile car-
olinensis to calls, 363—367
anti-predator

“snorkeling” by Jacana Jacana chicks, 262-265
courtship

of Theristicus caudatus, 1 18—119
interspecific

with foraging Picoides borealis. 346-353
learning

of artificial nest location by predators, 536-540
nesting

of Amazona finschi, 488-493
responses of Vireo belli to brood parasitism, 499-
504

parental

of bigamous Cardinalis cardinalis. 283-286
postfiedging

in Aquila chrysaetos. 472—477
roosting

in pre-migratory Progne subis. 354-362
siblicide

role of food at Accipiter gentilis nests, 432-436
Beletsky, Les, The ecotravellers' wildlife guide to Be-
lize and northern Guatemala, reviewed, 594—
595

Beletsky, Les, The ecotravellers’ wildlife guide to Cos-
ta Rica, reviewed, 594-595
Beletsky, Les, The ecotravellers' wildlife guide to
tropical Mexico, reviewed, 594-595
Bellbird, Bare-throated, see Procnias nudicollis

Benz, Brett W., see Watson, David M., and

Berkelman, James, James D. Fraser, and Richard T.

612

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Watson, Madagascar Fish-Eagle prey prefer-
ence and foraging success, 15-21
Biatas nigropectiis, 410

Bielefeldt, John, see Trexel, Dale R., Robert N. Ro-
senfield, , and Eugene A. Jacobs

biology

of nesting Spiza americana and Ammodramus hen-
slowii, 515-527

Bishop, K. David, see Hill, Norman R, and

Bittern, American, see Botaiirus lentiginosus
Least, see Ixobrychus exilis
Blackbird, Brewer’s, see Euphagu.s cyanocephalus
Chopi, see Gnorimopsar chopi
Red-winged, see Agelaiu.s phoeniceus
Rusty, see Euphagiis carolinus
Tricolored, see Agelaius tricolor
Yellow-headed, see Xanthocephalus .xanthocephalus

Blair, M. J., see Hagemeijer, W. J. M., and

Blarina brevicauda, 239

Blem, Charles R., Leann B. Blem, and Claudia I. Bar-
rientos, Relationships of clutch size and hatch-
ing success to age of female Prothonotary War-
blers, 577-580

Blem, Charles, R., review by, 294

Blem, Leann B., see Blem, Charles R., , and

Claudia I. Barrientos

Bloom, Peter H., see Goldstein, Michael I., ,

Jose H. Sarasola, and Thomas E. Lacher
Bluebird, Eastern, see Sicilia sialis
Mountain, see Sialia ciirrucoides
Western, see Sialia mexicana
Bobolink, see Doiichonyx oryzivoru.s
Bobwhite, Northern, see Colinus virginianus
Bock, Walter J., review by, 152—153

Bo^on, Roberto, see Anjos, Luiz Dos, and

Bollinger, Eric K., see Kershner, Eric, L., and

Bombycilla cedroriim. 84, 92, 94, 558
Bonasa iimbeiiu.s, 536

Boobook, Southern, see Ninox novaeseelandiae

Boone, C. A., see, Hopp, S. L., A. Kirby, and

Bosque, Carlos, and Emilio A. Herrera, “Snorkeling”
by the chicks of the Wattled Jacana, 262-265
Botauru.x lentiginosus, 107, 108, 111
pinnatus.

Bowman, Reed, David L. Leonard, Jr., Leslie K. Back-
us, and Allison R. Mains, Interspecific inter-
actions with foraging Red-cockaded Wood-
peckers in south-central Florida, 346-353
Brachyramphus marmoratus, 257-261
Brant, Black, see Branta bernicia nigricans
Branta bernicia bernicia. 166
bernicia nigricans, 167, 468
canadensis, 1 08, 1 66— 1 80, 181, 468, 469
canadensis occidentalis, 167
Breitwisch, Randall, Amy J. Schilling, and Joshua B.
Banks, Parental behavior of a bigamous male
Northern Cardinal, 283-286
Brewer, Gwenda L., reviews by, 447—448, 450—452
Bristlefront, Slaty, see Meridaxis ater, 199
brood parasitism

responses of Vireo belli, 499—504

Bromley, Robert G., Carriere, Suzanne, , and

Gilles Guathier
Brotogeris chiriri, 225
tirica, 408

Brown, Charles R., Swallow summer, reviewed, 294
Bubo virginianus, 218, 226, 274, 529, 530
Bubuicus ibis, 225, 433

Bugden, Shawn C., and Roger M. Evans, The devel-
opment of a vocal thermoregulatory response
to temperature in embryos of the domestic
chicken, 188-194

Bunting, Indigo, see Passerina cyanea
Lark, see Calaniospiza melanocorys
Burhans, Dirk E., and Frank R. Thompson III, Habitat
patch size and nesting success of Yellow-
breasted Chats, 210-215
Burhinus spp., 344

Burr, Timothy, see Reed, J. Michael, Elizabeth M.
Gray, Dianne Lewis, Lewis W. Oring, Richard

Coleman, , and Peter Luscomb

Burtt, Edward H., Jr., and W. Herbert Wilson, Jr., A
survey of undergraduate ornithology courses in
North America, 287—293
Buteo aibicaudatus, 407
buteo vuipinus, 182
jamaicensis, 558
lagopus, 558
lineatus, 89-99, 348
magnirostris, 225, 407
piatypterus, 230-231, 254
polyosoma, 530
spp., 218

swainsoni, 428—432, 434
Buteogallus meridionaUs, 218, 225
urubitinga, 218

Butler, Robert W., The Great Blue Heron: a natural
history and ecology of a seashore sentinel, re-
viewed, 445
Butorides striatus, 225
Buttonquail, see Tumix sp.

Buzzard, Steppe, see Buteo buteo vuipinus
Buzzard-Eagle, Gery, see Geranoaetus meianoieucus

Cacicus chrysopteriis, 218, 228, 413
haemorrhous, 222, 413
meianicterus, 84

Cacique, Golden-winged, see Cacicus chrysopterus
Red-rumped, see Cacicus haemorrhous
Yellow-winged, see Cacicus meianicterus
Cadman, Michael D., see Friesen, Lyle E., Valerie E.

Wyatt, and

Cairina moschata, 225
Calaniospiza melanocorys, 416
Calcarius mccownii, 416
ornatus, 389-396
Calidris fuscicollis, 225, 263
himantopus, 263
Callipepla californica, 530
Campephilus melanoleucos, 226
Camptostoma ohsoletum, 226, 412
Campylopterus curvipennis, 232

INDEX TO VOLUME I I I

613

excellens. 229, 232, 234
heinileucurits, 232

Campylorhynchiis brunneicapillus, 1 29
Ccinis familiaris, 56 1
Caprimulgus anthonyi, 344
candiciins. 222, 343, 344
carolinensis, 344
longirostris, 530
nigrescens, 340, 344
pan’ulus, 218, 226, 344
rufus. 218, 219, 226
vociferus, 229, 231

Caracara, Chimango, see Mitvago chimango

Crested, see Caracara plancus [Caracara lutosus,
Caracara cheriway, Polyborus tharus]
Mountain, see Phalcoboenus megalopterus
Caracara plancus, 225, [Caracara lutosus, Caracara
cheriway, Polyborus tharus] 330-339, 437
Carhonelle, Montserrat, see Johnsgard, Paul A., and

Cardinal, Northern, see Cardinalis cardinalis
Red-crested, see Paroaria coronata
Cardinalis cardinalis, 76-83, 137, 141, 142, 283-286,
558

Cardisoma guanhani, 561—564
Carduelis magellanica, 228, 402, 414
pinus, 558

tristis, 93, 105, 108, 111, 558
Cariatna cristata, 225
Carpodacus mexicanus, 84—88
purpureas, 93, 94, 558

Carriere, Suzanne, Robert G. Bromley, and Gilles
Guathier, Comparative spring habitat and food
use by two Arctic nesting geese, 166-180

Carroll, Adele E, see Gilbert, William M., and

Casiornis rufa, 227
Casmerodius albus, 530

Cassidy, Kelly M., see Smith, Michael R., Philip W.
Mattocks, Jr., and

Castrale, John S., Edward M. Hopkins, and Charles E.
Keller, Atlas of breeding birds of Indiana, re-
viewed, 294-295
cat, feral, see Felis catus
Catbird, Gray, see Dumetella carolinensis
Cathartes aura, 225, 407
Catharus fuscescens, 216, 220, 227, 575-576
guttatus, 92, 94, 558
ustulatus, 92, 94, 95
Catherpes mexicanus, 129, 199
Catoptrophorus semipalnuilus, 263, 561
Cebus capucinus, 125
Celeiis lugubris, 226
censusing methods

temporal differences in point counts, 139-143
Cepphus columba, 257—261
Cercomacra carbonaria, 207
nigricans, 207
serx’a, 207
tyrannina, 207

Certhia aniericana, 89—99, 245, 246, 558
Chaetura meridioncdis, 226

pelagica, 23 I
.spp., 35
vauxi, 229, 23 1
Chamaeza campanisoma, 4 1 0
ruficauda, 410

Charadrius alexandrinus, 56 1
melodus, 121-123, 321-329
spp., 344

vociferus, 424—426, 576
wilsonia, 561

Chat, Yellow-breasted, see Icteria virens
Chelydra serpentina, 115
Chencaerulescens caerulescens, 465, 468
Chickadee, Black-capped, see Poecile atricapillus
Carolina, see Poecile carolinensis
Mountain, see Poecile [Paru,s] gambeli
chicken, see Callus gallus
Chilia melanura, 530

Chinn, Robert W., see Smith, Kimberly G., W. Marvin
Davis, Thomas E. Kienzle, William Post, and

chipmunk, eastern, see Tamias striatus
Merriam’s, see Eutamias merriami
Chiroxiphia caudata, 411
Chlidonias niger, 108, 109
spp., 560

Chloroceryle aenea, 232
americana, 226
inda, 226

Chlorostilbon aureoventris, 409
Chordeiles gundlachii, 311, 561
minor, 226, 340
rapes tris, 343, 344

Chuck-will’s-widow, see Caprimulgus carolinensis
Circus cyaneus, 108, 109, 274, 558
Cistothorus palustris, 105-1 14, 254
platensis, 100-104, 105-114,558
Claravis pretiosa, 225

Clark, William S., see Gorney, Edna, , and Yor-

an Yom-Tov

Clethrionomys gapperi, 236, 239
Clibanornis dendrocolaptoides, 410
Cnemotriccus fusctus, 227, 41 1
Coccyzus americanus, 141, 408
melacoryphus, 226
minor, 231, 310

Cock-of-Rock, Guianan, see Rupicola rupicola
Coereba flaveola, 3 1 1

Colaptes auratus, 92, 108, 110, 141, 142, 348, 554,
558

campestris, 226, 409
melanochloros, 409
pitius, 530

Coleman, Richard, see Reed, J. Michael, Elizabeth M.

Gray, Dianne Lewis, Lewis W. Oring, ,

Timothy Burr, and Peter Luscomb
Colibri serrirostris, 409

Colinus virginianus, 236, 326, 541, 542, 545, 554,
555, 558

Collazo, Jaime A., see Saracco, James E, and

Colonia colonus, 4 1 1

614

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

colonization

of Lake Gatun, Panama, by Rostrhamus sociabilis,
265-268

Colorhamphus parx’irostris, 530
Coluber constrictor flaviventris, 524
Cohirnba araiicana, 530
cayennensis, 225, 408
leucocephala, 310
livia, 86, 289, 558
maculosa, 408
picazuro, 408
plumbea, 408
speciosa, 408
Columbina
passe rina, 561
picui, 225
squarnmata, 225
talpacoti, 225, 408
communication

visually in Caprimulgus riificollis, 340—345
community

in old-growth conifer-hardwood forest, 89-99
response to 1997-1998 El Nino in Chile, 527-535
response to weather, photoperiod, and year, 550-558
condition

among sex/age classes in migrant Regulus calen-
dula. 61-69

of bait trapped Accipiter nisus and Accipiter brevi-
pes, 181-187

Conirostrum speciosum, 228

Conner, Richard N., D. Craig Rudolph, Richard R.
Schaefer, Daniel Saenz, and Clifford E. Shack-
elford, Relationships among Red-cockaded
Woodpecker group density, nestling provision-
ing rates, and habitat, 494-498
Conopophaga lineata, 410
conservation

impact of wind turbines on upland nesting birds,
100-104

Contopus cinereus, 221, 411
virens, 92, 108, 110
Coot, American, see Fulica americana
Coragyps atratus, 225, 407

Corx’us brachyrhynchos, 89—99, 276—278, 538, 543,
544, 546, 554, 558
cora.x, 86, 92, 94, 1 15, 245, 276-278
ossifragus, 543, 544
Coryphospingus cucullatus, 228
Corylhopis delandi, 226

Coturni.x japonica, 236—242, 289, 326, 415—420, 433,
539, 547

sp., 236-242, 332, 323
Coulter, Malcolm C., review by, 593-594
Cowbird, Bay-winged, see Molothrus badius
Bronzed, see Molothrus aeneus
Brown-headed, see Molothrus ater
Giant, see Scaphidura oryzivora
Screaming, see Molothrus rufoaxillaris
Shiny, see Molothrus bonariensis
Cox, Robert R., Jr., see Whitt, Michael B., Harold H.
Prince, and

crab, giant white land, see Cardisoma guanhani

Crake, Columbian, see Neocrex colombianus
Paint-billed, see Neocrex erythrops
Ruddy, see Laterallus ruber
Cranioleuca obsoleta, 397-414
pallida, 397-414
Crax fasciolata, 218, 225
Creeper, Brown, see Certhia americana
Crescentchest, Collared, see Melanopareia torquata
Crossbill, Red, see Loxia curxnrostra
Crotophaga ani, 226, 311, 562
major, 226, 264

Crow, American, see Corvus brachyrhynchos
Pish, see Corvus ossifragus
Red-ruffed Pruit, see Pyroderus scutatus
Crypturellus obsoletus, 407
parvirostris, 225
undulatus, 225

Cuckoo, Mangrove, see Coccyzus minor
Squirrel, see Piaya cayana
Yellow-billed, see Coccyzus americanus
Cuculus canorus, 213
Culicivora caudacuta, 222
Curaeus curaeus, 530
Curassow, Bare-faced, see Crax fasciolata
Northern Helmeted, see Pauxi pauxi
Curlew, Long-billed, Numenius americanus
Currasow, Bare-faced, see Crax fasciolata
Custer, Thomas W., review by,

Cuthbert, Prancesca J., Brian Scholtens, Lauren C.
Wemmer, and Robyn McLain, Gizzard contents
of Piping Plover chicks in northern Michigan,
121-123

Cyanocitta cristata, 92, 94, 96, 139—143, 276—278,
348, 349, 538, 543, 544, 546, 558
stelleri, 245

Cyanocorax caeruleus, 402, 412
chrysops, 412
cristatellus, 217, 227
cyanomelas, 221
yncas, 233

Cyanoliseus patagonus, 530
Cyclarhis gujanensis, 228, 401, 402, 413
Cyprinus carpio, 17, 19, 20
Cypseloides niger, 30-47
Cypsnagra hirundinacea, 228

Dasypus novemcinctus, 443-444
Davis, S. K., D. C. Duncan, and M. Skeel, Distribution
and habitat associations of three endemic grass-
land songbirds in southern Saskatchewan, 389—
396

Davis, William E., Jr., reviews by, 294-295, 296-297,
299-300, 445, 448-449

Davis, W. Marvin, see Smith, Kimberly G., ,

Thomas E. Kienzle, William Post, and Robert
W. Chinn

de Reyna, Luis Arias, see Aragones, Juan, , and

Pilar Recuerda

Dean, Kurtis L., see Swanson, David L., Eric T. Lik-
nes, and

INDEX TO VOLUME I I 1

615

Decker, David G., see Avery, Michael L., Canie L.
Schreiber, and

DeGraat, Richard M., Thomas J. Maier, and Todd K.
Fuller, Predation of small eggs in artificial
nests: effects of nest position, edge, and poten-
tial predator abundance in extensive forest,
236-242

Delichon urhica, 579
Dendragapus obsciinis, 245
Dendrohates auratus, 439-440
tricolor, 439

Dendrocincla anabatina, 443
tiirdimi, 409

Dendrocolaptes platyrostris, 226, 409
Dendrocygna arboreo, 561
bicolor, 3 1 1

Dendroica caerulescens, 74, 89-99, 213
castanea, 93

coronata, 93, 141, 233, 558
discolor, 136
fusca, 93, 94, 96, 505
graciae, 5 1 1
magnolia, 93, 94
occidentalis, 95
pensylvanica, 93

petechia, 105-114, 136, 210, 290, 311, 381-388
pinus, 93, 94, 346, 348, 349
townsendi, 95
virens, 93, 94, 505
density

of Picoides borealis groups, 494-498
DeSucre-Medrano, Atahualpa, see Perez-Villafana,
Monica, Hector Gomez de Silva G., and

Dewey, Sarah R., see Estes, Wendy A., and

Patricia L. Kennedy
Dickcissel, see Spiza americana
Didelphis marsupialis, 546
diet

fruit sugar preference in Carpodacus mexicanus,
84-88

gizzard contents of Charadrius melodus chicks,
121-123

in spring in Anser albifrons frontalis and Branta
canadensis, 1 66— 1 80

in Strix occidentalis, 22-29

of Cypseloides niger in California, 30—47
Dillon, M. Beth, see Skagen, Susan K., Thomas R.

Stanley, and

dispersal

invasion of southern Florida by Asio flammeus from
the Antilles, 303-313

of Recun’irostra americana among western Great
Basin wetlands, 314—320
distribution

invasion of southern Florida by Asio flammeus from
the Antilles, 303-313

of endemic grassland songbirds in southern Sas-
katchewan, 389-396

of Pauxi pauxi and Aburria aburri in Venezuela,
564-569

Diuca diuca, 531, 533
dog, domestic, see Canis familiaris
Doliclionyx oryzivorus, 100-104, 108
Dove, Carla J., and Richard C. Banks, A taxonomic
study of Crested Caracaras (Falconidae), 330-
339

Dove, Mourning, see Zenaida macroura
Rock, see Columba livia
White-tipped, see Leptotila verreauxi
Zenaida, see Zenaida aurita
Drymophila malura, 410
Dryocopus lineatus, 226, 403, 409
pileatus, 92, 94, 245

Duck, American Black, see Anas rubripes
Masked, see Nomonyx dominicus
Wood, see Aix sponsa

Dugger, Bruce D., see Dugger, Katie M., , and

Leigh H. Frederickson

Dugger, Katie M., Bruce D. Dugger, and Leigh H.
Frederickson, Annual survival rates of female
Hooded Mergansers and Wood Ducks in south-
eastern Missouri, 1-6
Dumetella carolinensis, 108, 254

Duncan, D. C., see Davis, S. K., , and M. Skeel

Dysithamnus mentalis, 222, 410

Eagle, Bald, see Haliaeetus leucocephalus

Crowned Solitary, see Haryphaliaetus coronatus
Golden, see Aquila chrysaetos
Spanish Imperial, see Aquila adalberti
White-tailed, see Haliaeetus albicilla
Edwards, Ernest Preston, A field guide to the birds of
Mexico and adjacent areas (Belize, Guatemala,
and El Salvador), reviewed, 588

egg

dimensions and shell characteristics in Pufftnus na-
tivitatis, 421-422

ejection in Sialia currucoides, 440-442
laying time of Molothrus aeneus, 137—139
predation in artificial nests, 236—242
relation of age in female Protonotaria citrea to
hatching success, 577—580
variation in size and composition in Aytliya marila.
465-47 1

Egret, Cattle, see Bubulcus ibis
Egretta thula, 530

Eider, Common, see Somateria mollissima
Eira barbarra, 125
Elaenia albiceps, 121, 529, 530
chiriquensis, 4 1 2
flavoguster, 226
mesoleuca, 401—402, 412
obscura, 412
pan’irostris, 227, 412
Elaenia, Gray, see Myiopagis caniceps
Elanus caeruleus, 27
leucurus, 225, 529, 530
Electron platyrhynchum, 440
Emberizoides herbicola, 228
Embernagra platensis, 228
Empidonax albigularis, 229, 232

616

THE WILSON BULLETIN • VoL III, No. 4. December 1999

difficilis. 95
euleri, 4 1 1
flaviventris, 92, 95
hammondii, 95
minimus. 92, 94, 96
oberholseri, 74

traillii. 108, 585-588, 589-592
traillii e.xtimus, 573-575
varius, 227

vire.scens, 92, 94, 95, 542
Empidonomus varius, 41 1
Engstrom, R. Todd, review by, 297-298
Eremophila cdpestris, 416, 558
Erwin, R. Michael, review by, 454—456
Escalante R, Patricia, see Winker, Kevin, Stefan Ania-

ga Weiss, Juana Lourdes Trejo R, and

Esker, Terry L., see Walk, Jeffrey W., , and

Scott A. Simpson
Esnx masquinongy, 1 1 5
niger, 1 15

Estes, Wendy A., Sarah R. Dewey, and Patricia L.
Kennedy, Siblicide at Northern Goshawk nests:
does food play a role?, 432-436
Eucometis penicillata, 228, 443
Eumomota superciliosa, 439, 440
Euphagus carolinus, 558
cyanocephalus, 108

Euphonia, Antillean, see Euphonia musica
Chestnut-bellied, see Euphonia pectoralis
Euphonia chalybea, 413
chlorotica, 228
musica, 413
pectoralis, 413

Euscarthmus rufomarginatiis , 22 1
Eutamias merriami, 252

Evans, Michael I., see Tucker, Graham M., and

Evans, Roger M., see Bugden, Shawn C., and

Exum, Jay H., see Harper, Craig A., and

Eaaborg, John, review by, 147-148

Eaaborg, John, see Woodworth, Bethany L., ,

and Wayne J. Arendt
Ealco elenorae, 185
femoralis, 407, 529, 530
mexicanus, 558
peregrinus, 530

sparx’erius, 225, 269-271, 311, 407, 434, 530, 558
tinnunculus, 21 , 475
Falcon, Aplomado, see Ealco femoralis
Eleonora's, see Ealco elenorae
Peregrine, see Ealco peregrinus
Prairie, see Ealco mexicanus

Faucett, Rob. C., see Robbins, Mark B., , and

Nathan H. Rice
Eelis cams. 376
Ficedula hypoleuca. 579

Finch, Deborah, see Stole.son, Scott H., and

Finch, Deborah, see Yong, Wang, and

Finch, Diuca, see Diuca diuca

House, see Carpodacus mexicanus
Glive, see Eysurus castaneiceps
Purple, see Carpodacus purpureus

Sooty-faced, see Eysurus crassirostris
Zebra, see Poephila gullata
Finck, Elmer J., see Stapanian, Martin A., Christopher
C. Smith, and

Fire-eye, White-shouldered, see Pyriglena leucoptera
Firewood-gatherer, see Anumbius annumbi
Fish-Eagle, African, see Haliaetus vicifer
Madagascar, see Haliaeetus vociferoides
fisher, see Maries pennanti
Fleming, Robert L., review by 149-150
Flicker, Northern, see Colaptes auratus
Flint, Paul L., and J. Barry Grand. Patterns of variation
in size and composition of Greater Scaup eggs:
are they related?, 465-47 1
Fluvicolo leucocephala, 227
Flycatcher, Acadian, see Empidonax virescens

Crowned Slaty, see Griseotyrannus aurantioatro-
cristatus

Dusky, see Empidonax oberholseri
Great Crested, see Myiarchus crinitus
Hammond’s, see Empidonax hammondii
La Sagra’s, see Myiarchus sagrae
Least, see Empidonax minimus
Matinan, see Muscicapa sanfordi
Ochre-bellied, see Mionectes oleagineus
Pacific Slope, see Empidonax difficilis
Pied, see Ficedula hypoleuca
Sepia-capped, see Eeptopogon amaurocephalus
Southwestern Willow, see Empidonax traillii e.xti-
mus

Sulphur-rumped, see Myiobius sulphureipygius
Tawny-chested, see Aphanotriccus capitalis
White-throated, see Empidonax albigidaris
Willow, see Empidonax traillii
Yellow-bellied, see Empidonax flaviventris
Yellow-olive, see Tolmomyias sulphurescens
Foliage-gleaner, Chestnut-capped, see Hylocryptus
rectirostris

Ochre-breasted, see Philydor lichtensteini
Planalto [Russet-mantled], see Philydor dimidiatus
food

fRiit sugar preference in Carpodacus mexicanus,
84-88

role in siblicide at Accipiter gentilis nests, 432—436
spring use by Anser albifrons frontalis and Branta
canadensis, 1 66— 1 80

foraging

behavior

cooperative effort in Phalcoboenus megalopterus,
437-439

of Cypseloides niger in California, 30-47
of Percnostola caurensis and Myrmeciza di.yunc-
ta. 195-209

possible suet cutting by Corvus cora.x, 276—278
Seiurus aurocapillus follows Dasypus novemcinc-
tus, 434-444

ecology

of Haliaeetus vociferoides, 15—21
Forest-Falcon, Barred, see Micrastur ruflcollis
Eormicivora rufa, 221
Corpus .xanthopterygius, 408

INDEX TO VOLUME 1 1 1

617

Foster, Janet, Working for wildlife: the beginning of
preservation in Canada, reviewed, 453-454
fox, red, see Vidpes vulpes

Franzreb, Kathleen E., Factors that influence translo-
cation success in the Red-cockaded Woodpeck-
er, 38-45

Fraser, James D., see Berkelman, James, , and

Richard T. Watson

Frederickson, Leigh H., see Dugger, Katie M., Bruce
D. Dugger, and

Friesen, Lyle E., Valerie E. Wyatt, and Michael D.
Cadman, Nest reuse by Wood Thrushes and
Rose-breasted Grosbeaks, 132-133
Friesen, Lyle E., Valerie E. Wyatt, and Michael D.
Cadman, Pairing success of Wood Thrushes in
a fragmented agricultural landscape, 279-281
frog, black-and-green poison dart, see Dendrobates au-
ra tus

phantasmal poison dart, see Dendrobates tricolor
Fidica americana, 108, 469

Fuller, Todd K., see DeGraaf, Richard M., Thomas J.

Maier, and

Furnarius rufus, 226, 402, 410

Gallinago gallinago, 231, 263, 344
media, 344
paragLiaiae, 225
spp., 344

Gallinida chloropus, 108
Gallinule, Purple, see Porphyrida martinica
Gallus gallus, 188-194, 289
sp., 236-242

Gaston, Anthony J., and Ian Jones, The auks, re-
viewed, 593—594

Gauthier, Gilles, see Carriere, Suzanne, Robert G.
Bromley, and

Gauthreaux, Sydney, Jr., see Russell, Kevin R., and

Gavia immer, 115—116, 116-117
geographic variation

of type B song in Panda americana, 505-514
Geositta rufipennis, 530
Geothlypis aequinoctiaUs, 228, 413
trichas, 93, 94, 100-104, 105-114
Geotrygon chrysia, 310, 311
montana, 408

Geranoaetus melanoleucus, 530

Gervais, Jennifer A., and Daniel K. Ro.senberg, West-
ern Burrowing Owls in California produce sec-
ond broods of chicks, 569—57 1
Gilbert, William M., and Adele E Carroll, Singing in
a mated female Wilson’s Warbler, 134—137
Glaucidium brasilianum, 218, 226, 408
nanum, 530

Glaucomys sabrinus, 21
sp., 239
volans, 38, 269
Glossogobius giurus, 19
Glyphorynchus spirurus, 232
Gnorimopsar cliopi, 220, 221, 228, 413
Goldfinch, American, see Carduelis tristis

Goldstein, Michael L, Peter H. Bloom, Jose H. Sara-
sola, and Thomas E. Lacher, Post-migration
weight gain of Swainson’s Hawks in Argentina,
428-432

Gomez de Silva G., Hector, see Perez- Vallifana, Mon-
ica, , and Atahualpa DeSucre-Medrano

Goose, Canada, see Branta canadensis

Greater White-fronted, see Anser albifrons frontalis
Lesser Snow, see Chen caerulescens caerulescens
gopher, northern pocket, see Thymomas bottae
Gorney, Edna, William S. Clark, and Yoran Yom-Tov,
A test of the condition-bias hypothesis yields
different results for two species of sparrow-
hawks (Accipiter), 181-187
Goshawk, Northern, see Accipiter gentilis
Grackle, Common, see Quiscalus quiscula

Grand, J. Barry, see Flint, Paul L., and

grants

see awards and grants
Grassquit, Black-faced, see Tiaris bicolor

Gray, Elizabeth M., see Reed, J. Michael. ,

Dianne Lewis, Lewis W. Oring, Richard Cole-
man, Timothy Burr, and Peter Luscomb
Grebe, Pied-billed, see Podilymbus podiceps

Greenberg, Russell, see Schelhas, John, and

Greenfield, Paul J., see Krabbe, Niels, Morton L. Isler,
Phyllis R. Isler, Bret M. Whitney, Jose Alvarez
A., and

Griseotyrannus aurantioatrocristatus, 218, 227
Grosbeak, Rose-breasted, see Pheucticus ludovicianus
Ground-Dove, Common, see Columbina passerina
Ruddy, see Columbina talpacoti
Grouse, Blue, see Dendragapus obscurus
Ruffed, see Bonasa umbellus
growth

patterns in Himantopus mexicanus knudseni, 478-
487

Guan, Wattled, see Aburria aburri
Gubernetes yetapa, 227
Guillemot, Pigeon, see Cepphus columba
Guira guira, 226, 408
Gull, Hening, see Larus argentatus
Laughing, see Larus atricilla
Ring-billed, see Larus delawarensis
Gutierrez, R. J., see Smith, Richard B., M. Zachariah
Peery, , and William S. Lahaye

Habia rubica, 413
habitat

association

of endemic grassland songbirds in southern Sas-
katchewan, 389-396

breeding

association of group density and nestling provi-
sioning rates in Picoides borealis, 494—
498

use of Ly thrum salicaria, 105-114
patch size

in Icteria virens, 210—215

618

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

selection

by Setophaga niticilla in aspen-dominated forest
fragments, 70-75
use

by Nomonyx domicus, 1 19-121
in spring by Anser albifrons frontalis and Branta
canadensis, 166—180

winter

of Sterna aleutica, 559-560
use of burned and unburned coniferous forests,
243-250

Hagemeijer, W. J. M., and M. J. Blair, eds. The EBCC
atlas of European breeding birds; their distri-
bution and abundance, reviewed, 154-156

Haig, Susan M., Plissner, Jonathan H., , and

Lewis W. Oring
Haliaeetus albicilla, 15

leucocephalus, 15-21, 115-116,260
vocifer, 19
vociferoides, 1 5—2 1
Hall, George A., review by, 452-453
Haney, J. Christopher, Hierarchical comparisons of
breeding birds in old-growth conifer-hardwood
forest on the Appalachian plateau, 89-99
Haplospiza unicolor, 414
Harpagus bidentatus, 443

Harper, Craig A., and Jay H. Exum, Wild Turkeys (Me-
leagris gallopavo) renest after successful hatch,
426-427

Harpyhaliaetus coronatus, 218, 225
Harrier, Northern, see Circus cyaneus

Harrison, Colin J. O., see Baicich, Paul J., and

Hawk, Broad-winged, see Buteo platypterus
Cooper’s, see Accipiter cooperii
Great Black, see Buteogallus urubitinga
Harris’s, see Parabuteo unicinctus
Red-shouldered, see Buteo lineatus
Red-tailed, see Buteo jamaicensis
Roadside, see Buteo magnirostris
Rough-legged, see Buteo lagopus
Savannah, see Buteogallus meridionalis
Sharp-shinned, see Accipiter striatus
Swainson’s, see Buteo swainsoni
Hawk-Eagle, Ornate, see Spizaetus ornatus
Hawk-Owl, Bismarck, see Ninox variegata
Brown, see Ninox scutulata
Cinnabar, see Ninox ios
Manus, Ninox meeki
Moluccan, see Ninox squamipila
New Britain, see Ninox odiosa
Ochre-bellied, see Ninox ochracea [perx’ersa]
Philippine, see Ninox philippensis
Solomons, see Ninox jaccpdnoti
Speckled, see Ninox punctulata
Sumba, see Ninox rufolfi

Heinrich, Bernd, Planning to facilitate caching; possi-
ble suet cutting by a Common Raven, 276-278
Heliobletus contaniinatus, 4 1 0
Hemithraupis guira, 228
ruficapilla. 4 1 3

Hemitriccus inargaritaceiventer, 227

obsoletus, 412

Hendrickson, Herbert T, reviews by, 148, 151
Herbert, Percy N., Evidence of egg ejection in Moun-
tain Bluebirds, 440-442
Heron, Brown, see Nycticorax nycticorax
Herpetotheres cachinnans, 225, 407
Herpsilochmus atricapillus, 226

Herrera, Emilio A., see Bosque, Carlos, and

Heterospizias meridionalis, 407
Heterotis niloticus, 17, 19

Higgins, Kenneth E, see Leddy, Krecia L., ,

and David E. Naugle

Hill, Norman R, and K. David Bishop, Possible winter
quarters of the Aleutian Tern?, 559-560
Himantopus himantopus leucocephalus, 485
mexicanus, 478, 479
me.xicanus knudseni, 478-487
novaezealandiae, 485
Hirundinea ferruginea, 411
Hirundo fulva, 3 1 1

Hoag, David J., Hybridization between Clay-colored
Sparrow and Field Sparrow in northern Ver-
mont, 581-584

Hoffman, Wayne, Glen E. Woolfenden, and P. William
Smith, Antillean Short-eared Owls invade
southern Florida, 303-313

Holmes, Derek, and Karen Phillipps, The birds of Su-
lawesi, reviewed, 151

Hopkins, Edward M., see Castrale, John S., ,

and Charles E. Keller

Hopp, S. L., A. Kirby, and C. A. Boone, Banding re-
turns, arrival pattern, and site-fidelity of White-
eyed Vireos, 46—55

House-Martin, Common, see Delichon urbica
Houston, C. Stuart, Barred Owl nest in attic of shed,
272-273

Howell, Steve N. G., A bird-finding guide to Mexico,
reviewed, 595

Hubbard, John P. A critique of Wang Wong and
Finch’s field identification of Willow Flycatch-
er subspecies in New Mexico, 585-588
human impact

influence of human scent on mammalian predators,
415-420

Hummingbird, Giant, see Patagona gigas
Ruby-throated, see Archilochus colubris
hybridization

between Spizella pallida and Spizella pusilla in
northern Vermont, 581-584
Hydrophasianus chirurgus, 264
Hylocharis chrysura, 226
sapphirina, 222

Hylocichla mustelina, 92, 94, 132-133, 141, 279-281,
379, 538, 539

Hylocryptus recti rostris, 222
Hylopezus nattereri, 410
Hylophilus poicilotis, 413
Hylorchilus sumichrasti, 1 28- 1 30
Hymenops perspicillatus, 227

INDEX TO VOLUME 1 I 1

619

Hypocneinoides melanopo^on, 205
spp., 207

Ibis, Buff-necked, see Theristicus caudalus
Icteria virens, 84, 210—215
Icterus hidlockii, 251—256
cayanensis, 228
gal hula, 1 08
Ictinia pluiuhea, 225
Idioptilon nidipendulum, 412
Inezia inornata, I'll

Isler, Morton L., see Krabbe, Niels, , Phyllis R.

Isler, Bret M. Whitney, Jose Alvarez A., and
Paul J. Greenfield

Isler, Phyllis R., see Krabbe, Niels, Morton L. Isler,

, Bret M. Whitney, Jose Alvarez A., and

Paul J. Greenfield
Ixobrychus exilis, 105—1 14
Ixoreus naevius, 245

Jacana, African, see Actophilornis africanus
Lesser, see Microparra capensis
Pheasant-tailed, see Hydrophasianus chirurgus
Wattled, see Jacana jacana
Jacana jacana, 262—265

Jacobs, Eugene A., see Trexel, Dale R., Robert N. Ro-

senfield, John Bielefeldt, and

Jaeger, Parasitic, see Stercorarius parasiticus
Jahn, Olaf, Maria Eugenia Jara Viteri, and Karl-L.
Schuchmann, Connecticut Warbler, a North
American migrant new to Ecuador, 281-282
Jaksic, Eabian M., and Ivan Lazo, Response of a bird
assemblage in semiarid Chile to the 1997—1998
El Nino, 527-535
Jay, Blue, see Cyanocitta cristata

Bushy-crested, see Cyanocorax melanocyaneus
Curl-crested, see Cyanocorax cristatellus
Gray, see Perisoreus canadensis
Green, see Cyanocorax yncas
Stellar’s, see Cyanocitta stelleri
Johnsgard, Paul A., and Montserrat Carbonell, Ruddy
ducks and other stiff-tails: their behavior and
biology, reviewed, 450—452

Jones, Ian, see Gaston, Anthony J., and

Jones, Jason, Cooperative foraging in the Mountain
Caracara in Peru, 437—439

Jouanin, Christian, see Barre, Nicolas, Armand Barau,
and

Junco, Dark-eyed, see Junco hyemalis
Junco hyemalis, 93, 94, 558

Keehn, Shannon, see Sodhi, Navjot S., Cynthia A.
Paszkowski, and

Keith, Allan R., The birds of St.Lucia, West Indies,
reviewed, 147—148
Keith, Allan R., review by, 597-599
Keller, Charles E., see Castrale, John S., Edward M.
Hopkins, and

Kennedy, Patricia L., see Estes, Wendy A., Sarah R.
Dewey, and

Kennedy, Patricia L., see O’Toole, Laura T, ,

Richard L. Knight, and Lowell C. McEwen
Kershner, Eric, L., and Eric K. Bollinger, Aggressive
response of chickadees towards Black-capped
and Carolina calls in central Illinois, 363-367
Kestrel, American, see Falco spar\>erius
Common, see Falco tinnunculus
Kienzle, Thomas E., see Smith, Kimberly G., W. Mar-
vin Davis, , William Post, and Robert

W. Chinn

Killdeer, see Charadrius vociferus
King, Ben E, Checklist of the birds of Eurasia, re-
viewed, 596-597

King, David L, Mortality of an adult Veery incurred
during the defense of nestlings, 575-576
Kingbird, Eastern, see Tyrannus tyrannus
Gray, see Tyrannus dominicensis
Tropical, see Tyrannus melancholicus
Western, see Tyrannus verticalis
Kingfisher, Pygmy, see Chloroceryle aenea
Kinglet, Golden-crowned, see Regulus satrapa
Ruby-crowned, see Regulus calendula
kingsnake, prairie, see Lampropeltis calligasster cal-
ligaster

Kirby, A., see Hopp, S. L., , and C. A. Boone

Kiskadee, Great, see Pitangus sulphuratus
Kite, Black, see Milvus migrans

Black-shouldered, see Elanus caeruleus
Double-toothed, see Harpagus hidentatus
Snail, see Rostrhamus sociabilis
White-tailed, see Elanus leucurus
Knight, Richard L., see O’Toole, Laura T, Patricia L.

Kennedy, , and Lowell C. McEwen

Knipolegus cyanirostris, 41 1
lophotes, 41 1
nigerrimus, 41 1

Krabbe, Niels, Morton L. Isler, Phyllis R. Isler, Bret
M. Whitney, Jose Alvarez A., and Paul J.
Greenfield, A new species in the Myrmotherula
haematonota superspecies (Aves; Thamnophil-
idae) from the western Amazonian lowlands of
Ecuador and Peru, 157—165
Kreisel, Karen J., and Steven J. Stein, Bird use of
burned and unburned coniferous forests during
winter, 243-250

Kricher, John C., reviews by, 148-149, 151, 594-595
Kricher, John, A Neotropical companion: an introduc-
tion to the animals, plants, and ecosystems of
the New World tropics, reviewed, 146-147
Kuletz, Katherine J., and John E Piatt, Juvenile Mar-
bled Murrelet nurseries and the productivity in-
dex, 257-261

Lacher, Thomas E., see Goldstein, Michael L, Peter H.

Bloom, Jose H. Sarasola, and

Lagopus lagopus, 274

Lahaye, William S., see Smith, Richard B., M. Za-

chariah Peery, R. J. Gutierrez, and

Lampropeltis calligaster calligaster, 524
Lanins excubitor, 285

ludovicianus, 285, 348, 558

620

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Lanner, Ronald M., Made for each other: a symbiosis
of birds and pines, reviewed 297—298
Lapwing, Northern, see Vanellus vanellu.s
Lark, Homed, see Eremophila alpestri.s
Larsson, Hans, see Olsen, Klaus Mailing Olsen and

Loru.s argentatus, 115, 192
atricilla, 561
dekiwarensis, 192
Liitercillus ruber. 231
spp., 423

luithrotricciis euleri. Til

Laso, Ivan, see Jaksic, Fabian M., and

leaftosser, see Sclerurus spp.

Leddy, Krecia L., Kenneth E Higgins, and David E.
Naugle, Effect of wind turbines on upland nest-
ing birds in conservation reserve program
grasslands, 100-104
Legatus leucophaius, 227

Lein, M. Ross, see Wiebe, Myra O., and

Leistes siiperciliaris, 228

Lemon, R. E., see Lozano, G. A., and

Leonard, David L., Jr., see Bowman, Reed, ,

Leslie K. Backus, and Allison R. Mains
Lepiclocolaptes cmgustirostris, 221, 226, 409
fu.'icatus, 409
squamatus, 397—414
Leptastheniira aegithaloides, 530
.setaria, 410
struikcita, 410
Leptodon cayanensis, 225
Leptopogon cimciurocephalu.<;. 111, 232, 412
Leptotila rufaxilla, 225, 408
verreauxi, 225, 402, 408
Leucochloris albicollis, 409
Leuconerpes Candidas, 409

Levey, Douglas J., Foraging Ovenbird follows arma-
dillo, 443-444

Lewis, Dianne, see Reed, J. Michael, Elizabeth M.
Gray, , Lewis W. Oring, Richard Cole-

man, Timothy Burr, and Peter Luscomb

Liknes, Eric T, see Swanson, David L., , and

Kurtis L. Dean

Limnothlypis swainsonii, 229, 233, 234
lizard, tegu, see Tupinambis sp.

Lochmias nematiira, 410

Longspur, Chestnut-collared, see Calcarius ornatiis
McCown’s, see Calcarius mccownii
Loon, Common, see Gavia immer
loosetrife, purple, see Lythrum salicaria
Lophodytes cucullalus, 1—6
Lowther. Peter E., review by, 146
Loxia curv’iroslra, 245

Lozano, G. A., and R. E. Lemon, Effects of prior res-
idence and age on breeding performance in
Yellow Warblers, 381-388
l.urocalis semitorquatus, 218, 226, 408
Luscomb, Peter, see Reed, J. Michael, Elizabeth M.
Gray, Dianne Lewis, Lewis W. Oring, Richard

Coleman, Timothy Burr, and

Lysurus caslaneiceps, 1 26

crassirostris, 1 24— 1 28
Lythrus salicaria, 105—114

Machetornis rixosus, 227, 4 1 1

Mack, Tara, see Paruk, James D., Dean Seanfield, and

Mackenziaena leachii, 410

Madge, Steve, see Beaman, Mark, and

Magpie, Black-billed, see Pica pica

Mahan, Carolyn G., see Yahner, Richard H., and

Maier, Thomas J., see DeGraaf, Richard M., ,

and Todd K. Fuller

Mains, Allison R., see Bowman, Reed, David L. Leon-
ard, Jr., Leslie K. Backus, and

Mallard, see Anas platyrhynchos
management

response of Sitta pusilla to pine plantation thinning,
56-60

use of burned and unburned coniferous forests dur-
ing winter, 243-250

Manakin, Helmeted, see Antilophia galeata
Margarops fuscatus, 3 1 1

Marin, Manuel, Food, foraging, and timing of breeding
of the Black Swift in California, 30-47
Marks, Jeffrey S., and Alison E. H. Perkins, Double
brooding in the Long-eared Owl, 273-276
Martes pennanti, 240

Martin, Brown-chested, see Phaeoprogne tapera
Crag, see Ptyonoprogne rupestris
Purple, see Progne subis

Marzluff, John M. and Rex Sallabanks, eds.. Avian
conservation, reviewed, 449-450
Master, Terry L., Predation by Rufous Motmot on
black-and-green poison dart frog, 439—440
mate choice

color preference in female Cardinalis cardinalis,
76-83

Matthynsen, Erik, The nuthatches, reviewed, 452—453

Mattocks, Philip W., Jr., see Smith, Michael R., ,

and Kelly M. Cassidy

Mayer, Paul M., see Staus, Nancy L., and

McEwen, Lowell C., see O’Toole, Laura T, Patricia L.

Kennedy, Richard L. Knight, and

McLain, Robyn, see Cuthbert, Francesca J., Brian

Scholtens, Lauren C. Wemmer, and

Meadowlark, Eastern, see Sturnella magna
Western, see Sturnella neglecta
Mearns, Barbara and Richard Mearns, The bird collec-
tors, reviewed, 144—146

Mearns, Richard, see Mearns, Barbara and

Megalops cyprinoides, 19
Megarynchus pitangua. 111, 4 1 1
Melanerpes Candidas, 226

carol inns, 139-143, 346-353, 558
erythrocephalus, 141, 142, 348, 551, 558
flavifrons, 409
fonnicivorus, 252, 254
Melanopareia torquata, 222
Meleagris gallopavo, 426-427, 555, 558
spp., 17

INDEX TO VOLUME 1 1 I

621

Melospiza i>eori>iana, 105-1 14
lincolnii. 234, 554, 555, 558
melodici. 108, 285, 558
Mephifis mephitis, 416

Merganser, Common, see Mergus merganser
Hooded, see Lophodytes ciicidlatus
Mergus merganser, 92
Meridaxis ater, 199
methods

critique of field-identification of Empidonax trailii
subspecies, 585-588
Response, 589-592

temporal differences in point counts, 139-143
Metriopelia melanoptera, 530
Micrastur ruficollis, 225, 407
semitorquatus, 125
Microparra capensis, 264
Microtiis pennsylvanicus, 239
spp., 274
migration

arrival patterns of Vireo griseiis, 46-55
differences in timing among sex/age classes of Reg-
ains calendula, 61—69
Milvago chimachima, 225, 407
chimango, 530

Milvus migrans, 16, 17, 18, 475
Mimus gundlachii, 311
polyglottos, 285, 348
saturninus, 227
thenca, 530
triurus, 228

Mionectes oleagineus, 232, 234
Mniotilta varia, 93

Mockingbird, Bahama, see Mimus gundlachii
Northern, see Mimus polyglottos
Molothrus aeneus, 137-139

ater, 72, 93, 94, 96, 100-104, 108, 110, 133, 137-
139, 210, 252, 289, 440, 441, 499, 502, 515-
527, 558, 577
badius, 137-139, 220

honariensis, 137-139, 220, 221, 228, 311, 376, 379,
413, 441

rufoaxillaris, 138, 220, 221, 228
Momotus momota, 219, 226, 439, 440
monkey, see Cebus capucinus, Ateles geoffroyi, Al-
ouatta palliate
squirrel, see Saimiri oerstedi

Monson, Gale, see Russell, Stephen M., and

Moorhen, Common, see Gallinula chloropus
Morris, Sara R., review by, 295-296
mortality

of an adult Catharus fuscescens during nestling de-
fense, 575-576

of Charadrius melodus from Ocypode quadrata,
321-329

Motmot, Blue-crowned, see Momotus momota
Broad-billed, see Electron platyrhynchum
Rufous, see Baryphthengus martii
Rufous-capped, see Baryphthengus ruficapillus
Turquoise-browed, see Eumomota superciliaria
mouse, see Peromyscus spp.

deer, see Peromyscus maniculatus, Peromyscus sp.

house, see Mus musculus

northern grasshopper, sec Onychomys leucogaster
white-footed, see Peromyscus leucopus
woodland jumping, see Napaeoz.apus insignis
Murrelet, Marbled, see Brachyramphus marmoratus
Mus musculus, 23
Muscicapa sanfordi, 463
Muscipipra vetula, 4 1 1
Musci.saxicola macloviana, 530
muskellunge, see Esox masquinongy
Mustela frenata, 239, 416
Myadestes unicolor, 229, 233, 234
Mycteria americana, 225

Myiarchus crinitus, 91, 92, 108, 110, 141, 232, 269,
346, 348
sagrae, 3 1 1
swainsoni, 227
tyrannulus, 227
Myiobius sulphureipygius, 232
Myiodynastes maculatlus, 227, 411
Myiopagis caniceps, 226, 412
viridicata, 226

Myiophobus fasciatus, 227, 411
Myiopsitta monachus, 491
Myiornis auricularis, 221, 226, 412
Myrmeciza atrothorax, 205, 207
disjuncta, 195-209

Myrmotherula fjeldsaai, sp. nov., 157—165 (Frontis-
piece)

haematonota, 157—165 (Frontispiece)
leucophthalma, 157-165 (Frontispiece)
ornata, 157
sororia, 157-165

spodionota, 157-165 (Frontispiece)

Napaeozapus insignis, 239

Naughton, M. B., see Whittow, G. C., and

Naugle, David E., see Leddy, Krecia L., Kenneth F.

Higgins, and

Nemosia pileata, 228
Neocrex colombianus, 423
erythrops, 422-424
Neophron percnopterus, 475
Neotoma fuscipes, 22-29, 252
nest

arthropod fauna of Falco span’erius boxes, 269—27 1
influence of human scent on mammalian predators,
415-420

of Strix varia in attic of shed, 272-273
predation along three edge types, 541-549
predators of open and cavity in oak woodlands,
251-256

renest after successful hatch in Meleagris gallopavo,
426-427

reuse by Hylocichla mustelina and Pheucticus lu-
dovicianus, 132—133

nest-site

habitat of Accipiter striatus and Accipiter cooperii
in Wisconsin, 7-14

unusual for Empidonax trailii extimus, 573—575

622

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

nesting

behavior of Amazona finschi, 488-493
biology of Spiza americana and Ammodramus hen-
slowii, 515-527

continuous in Tyto alba, 571-573
fall and winter records for Charadrius vociferus in
the southern United States, 424-426
of four species of the Caribbean slope of Costa Rica,
124-128

success of Icteria virens, 210-215
nestling

mortality of an adult Catharus fuscescens during de-
fense of, 575-576

provisioning rates in Picoides borealis, 494-498
Neubig, Jeffrey R, and John A. Smallwood, The “sig-
nificant others” of American Kestrels; cohabi-
tation with arthropods, 269-27 1
Night-Heron, Black-crowned, see Nycticorax nycticor-
a.x

Yellow-crowned, see Nyctanassa violacea
Nighthawk, Antillean, see Chordeiles gundlachii
Nacunda, see Podager nacimda
Sand-colored, see Chordeiles rupestris
Nightjar, Blackish, see Cciprimulgus nigrescens
Little, see Caprimidgus parvulus
Red-necked, see Caprimulgus rujicollis
Rufous, see Caprimulgus rufus
Scrub, see Caprimulgus anthonyi
White-winged, see Caprimulgus candicans
Nino.x connivens, 462

ios, sp. nov., 457-464 (Frontispiece)

jacquinoti, 462

meeki, 462

natal is, 458

novaeseelandiae, 462

ochracea [jjerver.sa], 457-464 (Frontispiece)

odiosa, 462

philipensis, 457-464

punctulata, 457, 462

rudolfi, 458, 461

rufa, 462

scutulata, 462

squamipila, 457-464 (Frontispiece)
strenua, 462
superciliaris, 462
theomacha, 458, 462
variegata, 462

Nomonyx dominicus, 1 19—121, 31 I
Nothoprocta perdicaria, 530
Nothura maculosa. 225
Nucifraga columbiana. 245
Numenius phaeopus, 263

Nutcracker, Clark's, see Nucifraga columbiana
nuthatch, see Sitta spp.

Nuthatch, Brown-headed, see Sitta pusilla
Red-breasted, see Sitta canadensis
White-breasted, see Sitta carolinensis
Nyctibius grandis, 1 24- 1 28
griseus, 2 1 8, 226

Nycticorax nycticorax, 105, 225, 530, 531
Nyctidromus albicollis, 218, 226, 231, 344
Nyctanassa violacea, 562

Nystalus chacuru, 226
maculatus, 226

O’Connor, Raymond J., review by, 154—156
Odontoplwrus capueira, 407

Olsen, Klaus Mailing Olsen and Hans Larsson, Skuas
and Jaegers: a guide to the skuas and jaegers
of the world, reviewed, 298-299
Onychomys leucogaster, 416
Ophicephalus striatus, 17, 19
Oporornis agilis, 281—282
Philadelphia, 93, 281

opossum, Virginia, see Didelphis marsupialis
Oreochromis spp., 17

Oring, Lewis W., see Plissner, Jonathan H., Susan M.
Haig, and

Oring, Lewis W., see Reed, J. Michael, Elizabeth M.
Gray, Dianne Lewis, , Richard Cole-

man, Timothy Burr, and Peter Luscomb
Oriole, Baltimore, see Icterus galbula
Bullock’s, see Icterus bullockii
O’Toole, Laura T, Patricia L. Kennedy, Richard L.
Knight, and Lowell C. McEwen, Postfledging
behavior of Golden Eagles, 472-477
Otus asio, 254

atricapillus, 221, 226
choliba, 226, 408

Ovenbird, see Seiurus aurocapillus
Owl, Barking, see Ninox connivens
Barn, see Tyto alba
Barred, see Strix varia
Boreal, see Aegolius funereus
Burrowing, see Athene cunicularia
Great Gray, see Strix nebulosa
Great Horned, see Bubo virginianus
Long-eared, see Asio otus

Northern Spotted, see Strix occidentalis caurina
Papuan Boobook, see Ninox theomacha
Powerful, see Ninox strenua
Rufous, see Ninox rufa

Short-eared, see Asio fiammeus [Strix domingensis,
Asio portoricensis, Asio domingensis^
Spectacled, see Pulsatrix perspicillata
Spotted, see Strix occidentalis
Tawny, see Strix aluco
Ural, see Strix uralensis
White-browed, see Nino.x superciliaris
0.xyruncus cristatus, 227

Pachyramphus castaneus, 227, 41 1
major, 232

polychopterus, 227, 401, 411
validus, 227
y/m/fv. 227, 41 1

Painted-Snipe, see Rost rat ula semicollaris
Pandion haliaetus, 19
Parabuteo unicinctus, 218, 437, 530
Parakeet, Maroon-bellied, see Pyrrhura frontalis
Monk, see Myiopsitta monachus
Plain, see Brotogeris tirica
Reddish-bellied, see Pyrrhura frontalis

INDEX TO VOLUME 1 I 1

623

Parker. Timothy H., Responses of Bell’s Vireos to
brood parasitism by the Brown-headed Cow-
bird in Kansas, 499-504
Parkes, Kenneth C., review by, 144-146
Paroaria coronata, 218, 228
Parrot, Lilac-crowned, see Amazona jinschi
Reddish-bellied, see Pyrhura frontalis
Paruk, James D., Dean Seanfield, and Tara Mack, Bald
Eagle predation on Common Loon chick, 1 15-
116

Paruk, James D., Territorial takeover in Common
Loons (Gavia immer), 116—117
Panda americana, 93, 505—514
pitiaywni, 228, 397-414
Parula, Northern, see Panda americana
Tropical, see Parula pitiayumi
Pants caeruleus, 368
major, 368, 579
spp., 368

Passer domesticus, 228, 236—242, 251, 269, 289, 440—
442, 558

Passerculus sandwichensis, 100—104, 108, 234

Passerella iliaca, 558

Passerina cyanea, 93, 141, 211, 212, 214

Paszkowski, Cynthia A., see Sodhi, Navjot S., ,

and Shannon Keehn
Patagona gigas, 529, 530
Pauraque, see Nyctidromus alhicollis
Pauxi pauxi, 564—569

Peer, Brian D., and Spencer G. Sealy, Laying time of
a Bronzed Cowbird, 137-139

Peery, M. Zachariah, see Smith, Richard B., ,

R. J. Gutierrez, and William S. Lahaye
Pelecanus erythrorhynchos, 437
occidentalis, 437

Pelican, American White, see Pelecanus erythrorhyn-
chos

Brown, see Pelecanus occidentalis
Penelope ohsciira, 401, 407
Percnostola [Schistocichla] caurensis, 195-209
[Schistocichla] leucostigma, 197, 206
lophotes, 206
rufifrons, 206

[Schistocichla] schistacea, 206
Perez-Vallafina, Monica, Hector Gomez de Silva G.,
and Atahualpa DeSucre-Medrano, Sexual di-
morphism in the song of Sumichrast’s Wren,
128-130

Perisoreus canadensis, 245

Perkins, Alison E. H., see Marks, Jeffrey S., and

Peromyscus leucopus, 25, 28, 236—242, 269, 538
rnanicidatus, 236, 416
spp., 252, 543

Phacellodomus rufifrons, 226
Phaeomyias rnurina, 226
Phaeoprogne tapera, 227, 359
Phaethornis pretrei, 409
Phalacrocorax brasilianus, 225
Phalcohoenus megalopterus, 437-439
Phasianus colchicus, 108, 554, 555, 558

Pheasant, Ring-necked, see Phasianus colchicus
Pheucticus ludovicianus, 93, 94, 96, 108, 132-133

Phillipps, Karen, see Holmes, Derek, and

Philydor dimidiatus, 2 1 6-228
lichtensteini, 221, 226
rufus, 226, 410
Phimosus infuscatus, 225
Phloeoceastes rohustus, 403, 409
photoperiod

response of a Kansas winter bird community to,
550-558

Phrygilus alaudinus, 531
fruticeti, 53 1
gayi, 530

Phyllomyias fasciatus, 402, 412
reiseri, 222

Phylloscartes oustaleti, 412
ventralis, 412
Phytotoma rara, 530, 531

Piatt, John E, see Kuletz, Katherine J., and

Piaya cayana, 226, 408
Pica pica, 245, 274
Picoides arcticus, 243—250

borealis, 38-45, 56, 346-353, 494-498
lignarius, 530
mixtus, 226

pubescens, 91, 92, 141, 142, 245, 246, 277, 346—
353, 558

tridactylus, 243-250

villosus, 89-99, 244, 245, 246, 277, 346, 348, 558
Piculus aurulentus, 409
chrysochloros, 218, 226
Picumnus cirratus, 226
nebulosus, 409
spp., 403
temminckii, 409

Pigeon, White-crowned, see Columba leucocephala
pike, northern, see Esox niger
Pintail, Northern, see Anas acuta

White-cheeked, see Anas bahamensis bahamensis
Pionopsitta pileata, 408
Pionus maximiliani, 225, 408
Pipilo crissalis, 251, 252
erythrophthalmus. 348, 538
maculatus. 558

Pipit, Sprague’s, see Anthus spragueii
Pipra fasciicauda, 227
Pipraeidea melanonola, 413
Piranga fiava, 228
leucoptera, 233, 234
olivacea, 93, 94, 95
rubra. 141, 348
Pipromorpha rufiventris, 4 1 2
Pitangus sulphuratus, 227, 401, 411
Pitylus fulginosus, 414

Plains-wanderer, see Pedionomus torquatus

Plantcutter, Rufous-tailed, see Phytotoma rara

Platycichla flavipes, 412

Platypsaris rufus, 4 1 1

Platyrinchus mystaceus, 221 , 41 1

Plissner, Jonathan H., Susan M. Haig, and Lewis W.

624

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Oring, Within- and between-year dispersal of
American Avocets among multiple western
Great Basin wetlands, 314-320
Plover, Piping, see Charadrius melodus
Snowy, see Charadrius alexandrinus
Wilson’s, see Charadrius wilsonia
plumage

color preference in female Cardirtalis cardinalis,
76-83

Podager nacunda, 226, 344
Podilymbus podiceps, 105-1 14

Poecile atricapillus. 92, 94, 245, 277, 363-367, 368,
374, 554, 558

carolinensis, 141, 363-367, 374
gambeli, 245, 368-375
spp., 368

Poephila guttata, 251
Pogonotriccus eximius, 412
Polioptila dumicola, 221
Polyborus plancus, 407
Polystictus pectoralis, 222
Pooecetes gramineus, 213
Poospiza lateralis, 414
Porphyrula flavirostris, 264
rnartinica, 264
Porzana albicollis, 225
Carolina, 107, 108
spp., 423

Post, William, see Smith, Kimberly G., W. Marvin Da-
vis, Thomas E. Kienzle, , and Robert W.

Chinn

Potoo, Common, see Nyctibius griseus
Great, see Nyctibius grandis
Pratt, Jerome J., The Whooping Crane; North Ameri-
ca’s symbol of conservation, reviewed, 151
predation

by Baryphthengus martii on Dendrobates auratus,
439-440

influence of human scent in shortgrass prairie, 415-
420

of Haliaeetus leucocephalus of Gavia immer chick,

1 15-1 16

of open and cavity nesting birds in oak woodlands,
251-256

of small eggs in artificial nests, 236-242
on artificial nests in three edge types, 541—549
potential learning of artificial nest locations, 536-
540

relationship of Ocypode quadrata abundance and
Charadrius melodus mortality, 321—329
“snorkeling” as escape behavior in Jacana jacana,
262-265

prey

preference in Haliaeetus vociferoides, 15-21

Prince, Harold H., see Whitt. Michael B., , and

Robert R. Cox, Jr.
proceedings

of eightieth annual meeting, 596-603
Procnias nudicollis, 219, 221, 227, 41 I
Procyon lotor, 2, 416, 538, 542, 546

productivity

juvenile Brachyrarnphus marmoratus nurseries,
257-261

Progne chalybea, 221
spp., 359
subis, 354-362

Prolonotaria citrea, 141, 577-580
Psarocolius decumanus, 228
Pseudoleistes guirahura, 228
Pteroglossus castanotis, 226
Pteroptochos megapodius, 530
Ptyonoprogne rupestris, 358
Puffinus nativitatus, 421—422
Pulsatrix perspicillata, 218, 226

Purcell, Kathryn L., and Jared Verner, Nest predators
of open and cavity nesting birds in oak wood-
lands, 251-256

Pygmy-Owl, Ferruginous, see Glaucidium brasilianum
Pygmy-Tyrant, see Euscarthmus rufomarginatus
Eared, see Myiornis auricularis
Pygochelidon cyanoleuca, 530
Pyrhura frontalis, 219
Pyriglena leucoptera, 410
Pyrocephalus rubinus, 221
Pyroderus scutatus, 220, 227
Pyrope pyrope, 530
Pyrrhocoma ruficeps, 413
Pyrrhura frontalis, 221, 225

Quail, California, see Callipepla californica
Japanese, see Coturnix japonica
quail, see Coturnix sp.

Quail-Dove, Key West, see Geotrygon chrysia
Ruddy, see Geotrygon montana
Quiscalus quLscula, 84, 543, 544, 546, 554, 555, 558

racer, yellow-bellied, see Coluber constictor flaventris
racoon, see Procyon lotor
Rail, King, see Rallus elegans
Virginia, see Rallus limicola
Rallus elegans, 107
limicola, 1 05- 1 1 4
sanguinolentus, 530
Ramphastos discolorus, 409
toco, 226

Ramphocelus carbo, 402

Rasmussen, Pamela C., A new species of hawk-owl
Ninox from North Sulawesi, Indonesia, 457-
464

rat, see Rattus spp.

Rattus spp., 376, 561

Raven, Common, see Con’us corax

record

first Oporornis agilis for Ecuador, 281-282
of breeding Neocrex erythrops in Costa Rica, 422—
424

renest after successful hatch in Meleagris gallopavo,
426-427

second brood after successful fledging in Athene
cunicularia, 569—571

Recuerda, Pilar, see Aragones, Juan, Luis Arias de
Reyna, and

INDEX TO VOLUME I I 1

625

Recurvirostra avosetta, 485
americana. 314—320
Redhead, see Aythya americana
Redstart, American, see Setophaga niticilla
Reed, J. Michael, Elizabeth M. Gray, Dianne Lewis,
Lewis W. Oring, Richard Coleman, Timothy
Bun; and Peter Luscomb, Growth patterns of
Hawaiian Stilt chicks, 478-487
Regiihis calendula. 61-69

satrapa. 89-99, 245, 246, 554, 555, 558
release

factors affecting success in Picoides borealis, 38—
45

Renton, Katherine, and Alejandro Salinas-Melgoza,
Nesting behavior of the Lilac-crowned Panot,
488-493

report

survey of undergraduate ornithology courses in
North America, 287-293
reproduction

double brooding in Asia otus, 273—276
double brooding in Athene cimicidaria, 569-571
effect of prior residence and age in Dendroica pe-
techia, 381-388

nesting biology of Spiza americana and Ammodra-
niLis henslowii, 515—527

nurseries and the productivity index in Brachyram-
phus marmoratus, 257—261
pairing success of Hylocichla mustelina in a frag-
mented agricultural landscape, 279-281
relationship of clutch size and hatching success to
age in female Protonotaria citrea, 577—580
responses of Vireo belli to brood parasitism, 499—
504

timing in Cypseloides niger, 30-47
success in Strix occidentalis, 22—29
Restall, Robin, Munias and mannikins, reviewed, 148
Rhea americana, 222, 225
Rhea, Greater, see Rhea americana
Rhynchotus maculosa, 225

Rice, Nathan H., see Robbins, Mark B., Rob. C. Fau-
cett, and

Rice, Nathan H., Courtship behavior of the Buff-
necked Ibis (Theristicus caudatus), 1 18-1 19
Rising, James D., A guide to the identification and
natural history of the sparrows of the United
States, reviewed, 445—446
Robbins, Chandler S., review by, 146-147
Robbins, Mark B., Rob. C. Faucett, and Nathan H.
Rice, Avifauna of a Paraguayan Cerrado local-
ity: Parque Nacional Serram'a San Luis, depto.
Concepcion, 216-228
Robin, American, see Turdus migratorius
Black, see Turdus infuscatus
White-throated, see Turdus assimilis
Rogers-Price, Vivian, John Abbot’s birds of Georgia:
selected drawings from the Houghton Library,
Harvard University, reviewed, 299-300

roost

pre-migratory in Progne subis, 354-362

Rosenberg, Daniel K., .see Gervais, Jennifer A., and

Rosenfield, Robert N., see Trexel, Dale R., ,

John Bielefeldt, and Eugene A. Jacobs
Rostratula semicollaris, 218, 225
Rostrhamus sociabilis, 218, 225, 265—268
Rowley, Ian, and Eleanor Russell, Fairy-wrens and
grasswrens Maluridae, reviewed, 448-449

Rudolph, D. Craig, see Conner, Richard N., ,

Richard R. Schaefer, Daniel Saenz, and Clif-
ford E. Shackelford
Rupicola rupicola, 200

Russell, Eleanor, see Rowley, Ian, and

Russell, Kevin R., and Sidney A. Gauthreaux, Jr., Spa-
tial and temporal dynamics of a Purple Martin
pre-migratory roost, 354—362
Russell, Stephen M., and Gale Monson, The birds of
Sonora, reviewed, 595—596

Sabrewing, Long-tailed, see Campylopterus excellens
Violet, see Campylopterus hemileucurus
Saenz, Daniel, see Conner, Richard N., D. Craig Ru-
dolph, Richard R. Schaefer, , and Clif-

ford E. Shackelford
Saimiri oerstedi, 443

Salinas-Melgoza, Alejandro, see Renton, Katherine,
and

Sallabanks, Rex, see Marzluff, John M. and

Salpinctes obsoletus, 199
Saltator atricollis, 217, 228
similis, 228, 414

Saltator, Black-throated, see Saltator atricollis
Buff-throated, see Saltator maximus
Sandpiper, Stilt, see Calidris himantopus
Upland, see Bartramia longicauda
White-rumped, see Calidris fuscicollis
Sapphire, Rufous-throated, see Hylocharis sapphirina
SapsLicker, Yellow-bellied, see Sphyrapicus varius
Saracco, James E, and Jaime A. Collazo, Predation on
artificial nests along three edge types in a North
Carolina bottomland hardwood forest, 541-549
Sarasola, Jose H., see Goldstein, Michael I., Peter H.

Bloom, , and Thomas E. Lacher

Sarcoramphus papa. 225
Satrapa icterophrys, 4 1 1
Scaphidura oiyzivora. 138, 228
Scatophagus tetracanthus, 1 9
Scaup, Greater, see Aythya marila
Scelorchilus albicollis, 530

Schaefer, Richard R., see Conner, Richard N., D. Craig

Rudolph, , Daniel Saenz, and Clifford

E. Shackelford

Scheiber, Isabella B. R., see Alworth, Tom, and

Schelhas, John, and Russell Greenberg,

eds.. Forest patches in tropical landscapes, re-
viewed, 148-149
Schijfornis virescens, 41 1

Schilling, Amy J., see Breitwisch, Randall, ,

and Joshua B. Banks
Schoeniophylax phryganophila, 226

626

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Scholtens, Brian, see Cuthbert, Erancesca J., ,

Lauren C. Wemmer, and Robyn McLain

Schreiber, Carrie L., see Avery, Michael L., ,

and David G. Decker

Schuchmann, Karl-L., see Jahn, Olaf, Maria Eugenia

Jara Viteri, and

Sciuru.s carolinensis, 269, 542, 546
griseus, 25
Sclateria naevia, 204
Sclerurus scansor, 410
spp., 199, 205

Screech-Owl, Eastern, see Otiis asio
Variable, see Otus atricapilhis
Scrub-Jay, Elorida, see Aphelocoma coerulescens
Western, see Aphelocoma californica
Scytalopu.s ma^ellanicu.s, 530

Sealy, Spencer G., see Peer, Brian D., and

Seanheld, Dean, see Paruk, Janies D., , and

Tara Mack

Seedeater, Dark-throated, see Sporophila ruficoUis
Seiurus aurocapilliis, 93, 94, 96, 240, 279, 443-444
Sephanoide.s galeritus, 530

Seriema, Red-legged, see Lurocalis semitorquatus
Serpophaga cinerea, 124—128
suh.scistata, 4 1 2

Setophaga ruticilla, 70-75, 93, 94, 96
sexual

dimorphism

in song of Hylorchilus surnichrasti, 128—130
selection

in Caprimulgus ruficoUis, 340—345
Shackelford, Clifford E., see Conner, Richard N., D.
Craig Rudolph, Richard R. Schaefer, Daniel
Saenz, and

Shearwater, Christmas, Puffinus nativitatus
shrew, masked, see Sorex cinereiis

northern short-tailed, see Blarina hrevicauda
smokey, see Sorex fumeus
Shrike, Loggerhead, see Lxmius hidovicianus
Northern, see Lanins excuhitor
Sialia currucoides, 440—442
mexicana, 243, 253, 441
.sialis, 269, 346, 348, 441, 558
spp., 441

Sicalis fiaveola, 228, 414
sp., 530

Silva, Jose, Notes about the distribution of Pauxi pauxi
and Ahurria ahurri in Venezuela, 564-569
Simpson, Scott A., see Walk, Jeffrey W., Terry L Es-

ker, and

Sirysles sihilator, 227

Siskin, Hooded, see Carduelis magellanica
Pine, see Carduelis pinus
site-fidelity

in Vireo g rise us. 46—55
Sitla canadensis, 89-99, 245, 246, 277, 558
carolinen.si.s, 92, 94, 245, 246, 363-367, 558
pu.silla, 56-60, 346-353
spp., 204

Siltasomus griseicapillus, 22 1 , 226, 409

.Skagcn, Susan K., Thomas R. Stanley, and M. Beth

Dillon, Do mammalian nest predators follow
human scent trails in the shortgrass prairie?,
415-420

Skeel, M., see Davis, S. K., D. C. Duncan, and

skunk, striped, see Mephitis mephitis
Skutch, Alexander E, Life of the flycatcher, reviewed,
147

Smallwood, John A., see Neubig, Jeffrey P, and

Smith, Charles R., review by, 445-446

Smith, Christopher C., see Stapanian, Martin A.,

, and Elmer J. Einck

Smith, Kimberly G., W. Marvin Davis, Thomas E.
Kienzle, William Post, and Robert W. Chinn,
Additional records of fall and winter nesting by
Killdeer in southern United States, 424-426
Smith, Michael R., Philip W. Mattocks, Jr., and Kelly
M. Cassidy, Breeding birds of Washington
state: location data and predicted distributions,
reviewed, 296-297

Smith, P. William, see Hoffman, Wayne, Glen E.
Woolfenden, and

Smith, Richard B., M. Zachariah Peery, R. J. Gutier-
rez, and William S. Lahaye, The relationship
between Spotted Owl diet and reproductive
success in the San Bernadino Mountains, Cal-
ifornia, 22-29

Smith, Winston Paul, and Daniel J. Twedt, Temporal
differences in point counts of bottomland forest
landbirds, 139-143
snake, garter, see Thamnophis sirtalis
Snipe, Common, see Gallinago gallinago
Sodhi, Navjot S., Cynthia A. Paszkowski, and Shan-
non Keehn, Scale-dependent habitat selection
by American Redstarts in aspen-dominated for-
est fragments, 70-75

Solitaire, Slate-colored, see Myadestes unicolor
Somateria mollissima, 1, 3, 4
Sora, see Porzana Carolina
Sorex cinereus, 239
fumeus, 239

Sparrow, American Tree, see Spizella arhorea
Bachman’s, see Aimophila aestivalis
Baird’s, see Ammodramus hairdii
Brewer’s, see Spizella breweri
Chipping, see Spizella passerina
Clay-colored, see Spizella pallida
Field, see Spizella pusilla
Fox, see Passerella iliaca
Grasshopper, see Ammodramus savannarum
Harris’, see Zonotrichia querula
Henslow’s, see Ammodramus henslowii ■

House, see Passer domesticus
Le Conte’s, see Ammodramus leconteii
Lincoln’s, see Melospiza lincolnii
Olive, see Arremonops rufivirgatus
Orange-billed, see Arremon aurantiirostris
Rufous-collared, see Zonotrichia capensis
Savannah, see Passerculus sandwichensis
Song, see Melospiza melodia
Swamp, see Melospiza georgiana

INDEX TO VOLUME 1 I 1

627

Vesper, see Pooecetes gramineiis
White-crowned, see Zonotrichia leucophrys
White-throated, see Zonotrichia alhicollis
Sparrowhawk, Eurasian, see Accipiter nisus
Levant, see Accipiter brevipes
species nova

Myrmotherula haematonota, 157-165
Ninox ios, 457—464
Speotyto cunicularia, 226, 530
Spermophiliis heecheyi, 252, 253
tridecemlineatus. 416
Sphyrapiciis variiis, 92, 94, 348, 558
Spindalis zena, 3 1 1

Spinetail, White-lored, see Synallaxis alhilora
Spiza americana, 100-104, 515-527
Spizaetus ornatiis. 218, 225
Spizella arborea, 554, 555, 558
breweri. 581

pallida. 100-104, 581-584
passerina, 93, 94, 581, 584
pusiUa, 214, 554, 558, 581-584
Sporophiia bouvreuil, 228
caendescens. 228, 414
collaris, 228
hypoxantha, 228
plumbea, 228
ruficoUis, 218, 222, 228

squirrel, California ground, see SpermophUus beecheyi
flying, see Glaucomys sp.
gray, see Sciurus caroiinen.'iis
northern flying, see Glaucomys sabrinus
red, see Tamiasciurus grahamensis and Tamiasciu-
rus hudsonicus

southern flying, see Glaucomys volans
thirteen-lined ground, see SpermophUus tridecemli-
neatus

western grey, see Sciurus griseiis

Stanley, Thomas R., see Skagen, Susan K., ,

and M. Beth Dillon

Stapanian, Martin A., Christopher C. Smith, and Elmer
J. Finck, The response of a Kansas winter bird
community to weather, photoperiod, and year,
550-558

Starling, European, see Sturnus vulgaris
Staus, Nancy L., and Paul M. Mayer, Arthropods and
predation of artificial nests in the Bahamas: im-
plications for subtropical avifauna, 561-564

Stein, Steven J., see Kreisel, Karen J., and

Stelgidopteryx ruficoUis, 221
Stephanophorus diadematus. 401, 402, 413
Stephanoxis lalandi, 409
Stercorarius parasiticus, 437
Sterna aleutica. 559—560
anaethetus. 560
antillarum, 231
bergii, 560
dougallii, 563
forsteri, 108, 109
fuscata, 560
hirundo, 560
sandvicensis, 23 1

superciliaris, 344

Stilt, Black, see Himantopus novaezealandiae
Black-necked, see Himantopus mexicanus
Hawaiian, see Himantopus mexicanus knudseni
Pied, see Himantopus himantopus leucocephalus
Stoleson, Scott H., and Deborah M. Finch, Unusual
nest sites for southwestern Willow Flycatchers,
573-575

Streptopelia risoria, 25 1
Strix aluco, 273
hylophila, 408
nebulosa, 21
occidentalis, 22—29, 569
occidentaiis caurina, 22
uraiensis, 273
varia, 89-99, 272-273
Sturnella loyca, 530
magna, 254, 555, 558
neglecta, 100-104, 416, 554, 555, 558
Sturnus vulgaris, 84, 86, 87, 253, 269, 289, 359, 558,
579

Sungrebe, see Heliornis fulica

Suriri suiriri, 226

survival

of Lophodytes cucullatus and Aix sponsa in south-
eastern Missouri, 1-6
of Vireo latimeri, 376-380
Swallow, Bahama, see Tachycineta cyaneoviridis
Bam, see Hirundo rustica
Cave, see Hirundo fulva
Tree, see Tachycineta bicolor
Swanson, David L., Eric T. Liknes, and Kurtis L.
Dean, Differences in migratory timing and en-
ergetic condition among sex/age classes in mi-
grant Ruby-crowned Kinglets, 61-69
Swift, Black, see Cypseioides niger
Vaux’s, see Chaetura vauxi
SyivUagus spp., 25
Synallaxis albescens, 226
albilora, 222
cinerascens. 410
ruficapilla. 410
.•ipixi. 410

Syndactyla rufosuperciliata, 410
Svrigma sibilatrix, 225, 407

Tacha, Thomas C., see Anderson, James T, and

Tachuri, Bearded, see Polystictus pectoralis
Tachycineta bicolor. 108, 233, 269
cyaneoviridis. 3 1 1
leucopyga. 530
leucorrhoa, 221
Tachvphonus coronatus, 413
rufus, 228

Taeniopygia guttata. 237
Tamias striatus, 221 , 239, 538
Tamiasciurus hudsonicus. 239, 269
Tanager, Gray-headed, see Eucometis pencillata
Rufous-headed, see Hemithraupis ruficapilla
Sayaca, see Thraupis sayaca

628

THE WILSON BULLETIN • Vol. Ill, No. 4, December 1999

Scarlet, see Piranga olivacea
Stripe-headed, see Spiudcilis zena
Summer, see Piranga rubra
White-winged, see Piranga leucoptera
Tangara preciosa, 413
Tapera naevia, 226, 408
Ta.xidea ta.xa, 416
taxonomy

new species in the Myrmotherida haematonota su-
perspecies, 157-165
of Caracara spp., 330-339

of Percno.stola caurensis and Myrmeciza disjuncta,
195-209

tayra, see Eira barbara
Teal, Blue-winged, see Ana.s discor.s
Tern, Aleutian, see Sterna aleutica
Black, see Chlidonia.s niger
Bridled, see Sterna anaethetus
Crested, see Sterna bergii
Common, see Sterna hirundo
Forster's, see Sterna for.steri
Least, see Sterna antillarum
Roseate, see Sterna dougalli
Sandwich, see Sterna sandvicensis
Sooty, see Sterna fiiscata
Yellow-billed, see Sterna superciliari.s
territory

takeover in Gavia immer, 1 16—1 17
Tersina viridi.x, 413
Thalurania furcata, 226
glaucopis, 409

Thamnophihis caerule.xcens, 226, 404, 410
Thamnophi.x .xirtalis, 440
Theri.sticu.s caudatus. 118-119, 225, 407
thermoregulation

vocalization of Gallu.s gallu.s embryos in response
to temperature, 188-194
Thinocorus spp., 264

Thompson, Frank R., Ill, see Burhans, Dirk E., and

Thrasher, Brown, see Toxostoma rufum
Pearly-eyed, see Margarop.s fu.scatu.s
Thraiipi.s bonarien.si.s, 413
.sayaca, 228, 40 1 , 402, 4 1 3
Thrush, Hermit, see Catharus guttatii.x
Swainson’s, see Catharus ustulatus
Varied, see Ixoreus naevius
Wood, see Hylocichia mustelina
Thryomanes bewickii, 558
Thryothorus leucotis, 205

ludovicianus. 141, 142, 555, 558
modest us. 229, 233
Thymomas bottae, 25, 28
Tiaris bicolor. 31 1
fuliginosa. 414
Tilapia spp., 17
Tit, Blue, see Parus caerideus
Great, see Parus major
Titmouse, Bridled, see Baeolophus woUweberi
Plain, see Baeolophus inornatus
Tufted, see Baeolophus bicolor

Tityra cayana. 221 , 41 1
inquisitor, 227

Todd, Frank S., Natural history of the waterfowl, re-
viewed, 446-448
Todirostrum plumbeiceps, 412
Tolmomyias sulphurescens, 227, 412
Towhee, California, see Pipilo crissalis
Eastern, see Pipilo erythrophthalmus
Spotted, see Pipilo maculatus
Toxostoma rufum. 558
translocation

factors affecting success in Picoides borealis, 38-
45

Trejo R, Juana Lourdes, see Winker, Kevin, Stefan Ar-
riaga Weiss, , and Patricia Escalante P.

Trexel, Dale R., Robert N. Rosenfield, John Bielefeldt,
and Eugene A. Jacobs, Comparative nest site
habitats in Sharp-shinned and Cooper’s hawks
in Wisconsin, 7—14
Trichothraupis melanops, 228, 413
Tringa soiitaria, 225

Troglodytes aedon, 130-132, 227, 253, 254, 403, 412,
530

sissonii, 128

troglodytes, 89-99, 128, 245, 558
Trogon curucui, 226
rufus, 409
surrucura, 409

Tucker, Graham M., and Michael I. Evans, Habitats
for birds in Europe: a conservation strategy for
the wider environment, reviewed, 454-456
Tupinambis sp., 218
Turdus albicollis, 412

amaurochalinus, 227, 401, 402, 412

assimilis, 233

faicklandii, 530

flavipes, 412

infuscatus, 229, 233, 234

migratorius, 84, 86, 87, 92, 94, 141, 142, 441, 558
nigriceps, 412
rufiventris, 227, 401, 402
turkey, see Meleagris spp.

Turkey, Wild, see Meleagris gallopavo
Turnix sp., 251

Turnstone, Ruddy, see Arenaria interpres
turtle, snapping, see Chelydra serpentina
Turtle-Dove, Ringed, see Streptopelia risoria

Twedt, Daniel J., see Smith, Winston Paul, and

Tyrannulet, Reiser’s, see Phyllomyias reiseri
Torrent, see Serpophaga cinerea
Tyranniis dominicensis, 310
melancholicus, 227, 402, 411
savana, 221 , 4 1 1
tyrannus, 108, 346, 348, 349
vertical is, 252

Tyrant, Cock-tailed, see Alecturus tricolor
Sharp-tailed, see CuUcivora caudacuta
White-tufted, see Elaenia albiceps
Tyto alba. 27, 273, 408, 530, 571-573
Upucerthia dumetaria, 530
ruficauda, 530

INDEX TO VOLUME 1 1 I

629

use

of Lythrion salicaria dominated habitat, 105-114

Vanelhis chilensis, 225, 530
vanellus. 131, 484
Veery, see Cathuriis fuscescens
Veit, Richard R., review by, 298—299
Veniliornis passerinus, 226
spilogaster. 397-414
Vermivora luciae, 577
ruficcipilla. 229, 233

Verner, Jared, see Purcell, Kathryn L., and

Vireo altiloqiius, 310

atricapilliis, 46—55, 378, 379
helm, 51, 52, 378, 379, 499-504
bellii pusilliis, 46, 499
crassirostris, 311
gilvus, 378

griseus, 46—55, 141, 378
latimeri, 376—380

olivaceus, 52, 91, 93, 94, 141, 228, 279, 378, 413
pallens, 232

solitarius, 92, 94, 95, 233
vicinior, 378

Vireo, Bell’s, see Vireo hellii

Black-capped, see Vireo atricapillus
Black-whiskered, see Vireo altiloquus
Blue-headed, see Vireo solitarius
Gray, see Vireo vicinior
Least Bell’s, see Vireo hellii pusillus
Mangrove, see Vireo pallens
Puerto Rican, see Vireo latimeri
Red-eyed, see Vireo olivaceus
Thick-billed, see Vireo crassirostris
Warbling, see Vireo gilvus
White-eyed, see Vireo griseus

Viteri, Maria Eugenia Jara, see Jahn, Olaf, , and

Karl-L. Schuchmann
vocalization

aggressive response of Poecile atricapillus and Poe-
cile carolinensis to calls, 363-367
of Callus gallus embryos in response to tempera-
ture, 188—194

of Percnostola caurensis and Myrmeciza disjuncta.
1 95-209

sexual dimorphism in song of Hylorchilus sumi-
chrasti, 128-130

singing in mated female Wilsonia pusilla, 134-137
type B song of Parula americana, 505-514
use of song types by Poecile gamheli, 368-375
Volatinia jacarina, 228, 414
vole, see Microtus spp.

pine, see Microtus pennsylvanicus
red-backed, see Clethrionornys gapperi
Vuilleumier, Eran9ois, review by, 300-301
Vulpes vulpes, 417, 576
Vultur gryphus, 530
Vulture, Black, see Coragyps atratus
Egyptian, see Neophron percnopterus
Turkey, see Cathartes aura

Walk, Jeffrey W., Ten y L Eskcr, and Scott A. Simpson,
Continuous nesting of Barn Owls in Illinois,
571-573

Warbler, Bay-breasted, see Dendroica castanea
Black-and-white, see Mniotilta varia
Black-throated Blue, see Dendroica caerulescens
Black-throated Green, see Dendroica virens
Blackburnian, see Dendroica fusca
Chestnut-sided, see Dendroica pensylvanica
Connecticut, see Oporornis agilis
Golden-crowned, see Basileuterus culicivorus
Grace’s, see Dendroica graciae
Hermit, see Dendroica occidentalis
Hooded, see Wilsonia citrina
Magnolia, see Dendroica magnolia
Mourning, see Oporornis Philadelphia
Nashville, see Vermivora ruficapilla
Pine, see Dendroica pinus
Prairie, Dendroica discolor
Prothonotary, see Protonotaria citrea
Swainson’s, see Limnothlypis swainsonii
Townsend’s, see Dendroica townsendi
White-bellied, see Basileuterus hypoleucus
Wilson’s, see Wilsonia pusilla
Yellow, see Dendroica petechia
Yellow-rumped, see Dendroica coronata
Watson, David M., and Brett W, Benz, The Paint-billed
Crake breeding in Costa Rica, 422-424
Watson, Richard T, see Berkelman, James, James D.
Eraser, and

Watts, Bryan D., see Wilson, Michael D., and

Waxwing, Cedar, see Bomhycilla cedrorum
weasel, long-tailed, see Mustela frenata
weather

response of a bird assemblage to the 1997-1998 El
Nino in Chile, 527—535

response of a Kansas bird community, 550-558
weight

post-migration gain of Buteo swainsoni in Argenti-
na, 428-432

Weiss, Stefan Arriaga, see Winker, Kevin, , Ju-

ana Lourdes Trejo P, and Patricia Escalante P.
Wemmer, Lauren C., see Cuthbert, Francesca J., Brian

Scholtens, , and Robyn McLain

Wheatley, Nigel, Where to watch birds in Asia, re-
viewed, 149-150

Whimbrel, see Numenius phaeopus
Whip-poor-will, see Caprimulgus vociferus
Whistling-Duck, Fulvous, see Dendrocygna hicolor
West Indian, see Dendrocygna arhorea
Whitney, Bret M., see Krabbe, Niels, Morton L. Isler,

Phyllis R. Isler, , Jose Alvarez A., and

Paul J. Greenfield

Whitt, Michael B., Harold H. Prince, and Robert R.
Cox, Jr., Avian use of purple loosestrife dom-
inated habitat relative to other vegetation types
in a Lake Huron wetland complex, 105-114
Whittow, G. C., and M. B. Naughton, Christmas
Shearwater egg dimensions and shell charac-
teristics on Laysan Island, northwestern Ha-
waiian Islands, 421-422

630

THE WILSON BULLETIN • Vol. II], No. 4, December 1999

Wiebe, Myra O., and M. Ross Lein, Use of song types
by Mountain Chickadees {Poecile gambeli),
368-375

Willet, see Catoptrophonis semipabnatus

Wilson, Michael D., and Bryan D. Watts, Response of
Brown-headed Nuthatches to thinning of pine
plantations, 56-60

Wilson, W. Herbert, Jr., see Burtt, Edward H., Jr., and

Wilsonia citrina, 93, 94, 95
pusilla, 134-137

Winker, Kevin, Stefan Arriaga Weiss, Juana Lourdes
Trejo R, and Patricia Escalante R, Notes on the
avifauna of Tabasco, 229-235
Winter, Maiken, Nesting biology of Dickcissels and
Henslow’s Sparrows in southwestern Missouri
prairie fragments, 515-526

Wolcott, Donna L., and Thomas G. Wolcott, High mor-
tality of Piping Plovers on beaches with abun-
dant ghost crabs: correlations, not causation,
321-329

Wolcott, Thomas G., see Wolcott, Donna L., and

Wolfenbarger, L. Lareesa, Female mate choice in
Northern Cardinals: is there a preference for
redder males?, 76-83

Wood-Pewee, Eastern, see Contopus virens
Woodcreeper, Great Rufous, see Xiphocolaptes major
Narrow-billed, see Lepidocolapte.s angu.stirostris
Olivaceous, see Sitta.somu.s griseicapillii.s
Tawny-winged, see Dendrocincia anabatina
Wedge-billed, see Glyphorhynchu.s spirurus
Woodpecker, Acorn, see Melanerpes formicivorus
Black-backed, see Picoide.s arcticus
Downy, see Picoide.s pube.scen.s
Golden-green, see Piculu.s chry.sochloro.s
Hairy, see Picoide.s villo.su.s
Lineated, see Dryocopii.s lineatii.s
Pileated, see Dryocopii.s pileatii.s
Red-bellied, see Melanerpe.s carolinu.s
Red-cockaded, see Picoide.s boreali.s
Red-headed, see Melanerpe.s erythrocephahi.s
Three-toed, see Picoide.s tridactylu.s
woodrat, dusky-footed, see Neotoma fu.scipe.s
Woodworth, Bethany L., John Faaborg, and Wayne J.
Arendt, Survival and longevity of the Puerto
Rican Vireo, 376-380

Woolfenden, Glen E., see Hoffman, Wayne, ,

and P. William Smith

Wren, Bewick’s, see Thryomanes bewickii
Buff-breasted, see Thryothorus leiicotis
Cactus, see Campylorhynchu.s bruimeicapillus
Canyon, see Catherpes mexicamis
Carolina, see Thryothorus hidovicianus
House, see Troglodytes aedon
Marsh, see Cistothorus palustris
Plain, see Thryothorus modestus
Rock, see Salpinctes obsoletus
Sedge, see Cistothorus platensis
Sorocco, see Troglodytes sissonii
Sumichrast’s, see Hylorchilus sumichrasti
Winter, see Troglodytes troglodytes

Wyatt, Valerie E., see Friesen, Lyle E., , and

Michael D. Cadman

Xanthocephalus xanthocephalus, 108, 254, 558
Xenops minutus, 410
rut Hans, 410

Xenopsaris albinucha, 218, 227
Xenopsaris, White-naped, see Xenopsaris albinucha
Xiphocolaptes albicolis, 409
major, 221, 226
Xolmis cinerea, 227, 411
velata, 227

Yahner, Richard H., and Carolyn G. Mahan, Potential
for predator learning of artificial arboreal nest
locations, 536-540

Yellowthroat, Common, see Geothlypis trichas
Yom-Tov, Yoram, see Gorney, Edna, William S. Clark,
and

Young, Wang and Deborah M. Fitch, Response, 589-
592

Young, Bruce E., and James R. Zook, Nesting of four
poorly-known bird species on the Caribbean
slope of Costa Rica, 124-128

Zenaida auriculata, 408, 530
aurita, 310, 311

macroura, 92, 94, 96, 139-143, 310, 558
Zimmer, Kevin J., Behavior and vocalizations of the
Caura and Yapacana antbirds, 195-209
Zonotrichia albicollis, 141, 554, 555, 558
capensis, 228, 402, 414, 530, 533
leucophrys, 554, 555, 558
querula, 213, 558

Zook, James R., see Young, Bruce E., and

This issue of The Wilson Bulletin was published on 14 December 1999.

PUBLISHED BY THE WILSON OKMTHOI-OGIUAI. SOCIETY

VOLUME 111 1999 QUAKTEKLY

EDITOR: ROBERT C. SEASON
EDITORIAL BOARD: KATHY G. BEAL
CLAIT E. BRAUN
RICHARD N. CONNER
INDEX EDITOR: KATHY G. BEAL
ASSISTANT EDITORS: TARA BAIDEME
JOHN LAMAR
DANTE THOMAS
DORIS WATT

The Wilson Ornithological Society
Founded December 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist

President John C. Kricher, Biology Department, Wheaton College, Norton, Massachusetts
02766; E-mail: JKricher@wheatonma.edu.

First Vice-President— William E. Davis, Jr., College of General Studies, 871 Commonwealth
Ave., Boston University, Boston, Massachusetts 02215; E-mail: WEDavis@bu.edu.

Second Vice-President — Charles R. Blem, Dept, of Biology, 816 Park Ave., Virginia
Commonwealth Univ., Richmond, Virginia 23284; E-mail: cblem@saturn.vcu.edu.

Editor — Robert C. Beason, Department of Biology, State University of New York, 1 College
Circle, Geneseo, New York 14454; E-mail: WilsonBull@geneseo.edu.

Secretary — John A. Smallwood, Department of Biology, Montclair State University, Upper
Montclair, New Jersey 07043; E-mail: Smallwood@mail.montclair.edu.

Treasurer — Doris J. Watt, Department of Biology, Saint Mary’s College, Notre Dame,
Indiana 46556; E-mail: DWatt@saintmarys.edu.

Elected Council Members — Charles F. Thompson and Sara R. Morris (terms expire 2000),
Jonathan L. Atwood and James L. Ingold (terms expire 2001), Robert A. Askins and
Jeffrey R. Walters (terms expire 2002).

DATES OF ISSUE OF VOLUME 1 1 1
OF THE WILSON BULLETIN

NO. 1 — I March 1999
NO. 2 — 10 May 1999
NO. 3 — 10 August 1999
NO. 4 — 14 December 1999

CONTENTS OF VOLUME 1 1 1

NUMBER I

MAJOR PAPERS

ANNUAL SURVIVAL RATES OF FEMALE HOODED MERGANSERS AND WOOD DUCKS IN
SOUTHEASTERN MISSOURI Katie M. Dugger, Bruce D. Dugger, and Leigh H. Kredrickson

COMPARATIVE NEST SITE HABITATS IN SHARP-SHINNED AND COOPER’S HAWKS IN WIS-
CONSIN Dale R. Trexel, Robert N. RosenfieUl, John Bielefeldt, and Eugene A. Jacob.s

MADAGASCAR FISH-EAGLE PREY PREFERENCE AND FORAGING SUCCESS

James Berkelman, James D. Fraser, and Richard T. Watson

THE RELATIONSHIP BETWEEN SPOTTED OWL DIET AND REPRODUCTIVE SUCCESS IN THE

SAN BERNARDINO MOUNTAINS, CALIFORNIA

Richard B. Smith. M. Zachariah Peery, R. J. Gutierrez, and William S. Lcduiye

FOOD, FORAGING, AND TIMING OF BREEDING OF THE BLACK SWIFT IN CALIFORNIA

Manuel Marin

FACTORS THAT INFLUENCE TRANSLOCATION SUCCESS IN THE RED-COCKADED WOOD-
PECKER Kathleen E. Franzreb

BANDING RETURNS, ARRIVAL PATTERN, AND SITE-FIDELITY OF WHITE-EYED VIREOS --

S. L. Hopp, A. Kirby, and C. A. Boone

RESPONSE OF BROWN-HEADED NUTHATCHES TO THINNING OF PINE PLANTATIONS

Michael D. Wilson and Bryan D. Watts

DIFFERENCES IN MIGRATORY TIMING AND ENERGETIC CONDITION AMONG SEX/AGE

CLASSES IN MIGRANT RUBY-CROWNED KINGLETS

David L. Swanson, Eric T. Liknes, and Kurtis L. Dean

SCALE-DEPENDENT HABITAT SELECTION BY AMERICAN REDSTARTS IN ASPEN-DOMINAT-
ED FOREST FRAGMENTS Navjot S. Sodhi, Cynthia A. Paszkowski. and Shannon Keehn

FEMALE MATE CHOICE IN NORTHERN CARDINALS: IS THERE A PREFERENCE FOR REDDER
MALES? _ Wolfenbarger

FRUIT SUGAR PREFERENCES OF HOUSE FINCHES

Michael L. Avery, Carrie L. Schreiber, and David G. Decker

HIERARCHICAL COMPARISONS OF BREEDING BIRDS IN OLD-GROWTH CONIFER-HARD-
WOOD FOREST ON THE APPALACHIAN PLATEAU J- Christopher Haney

EFFECTS OF WIND TURBINES ON UPLAND NESTING BIRDS IN CONSERVATION RESERVE
PROGRAM GRASSLANDS Krecia L. Leddy, Kenneth F. Higgins, and David E. Naugle

AVIAN USE OF PURPLE LOOSESTRIFE DOMINATED HABITAT RELATIVE TO OTHER VEGE-
TATION TYPES IN A LAKE HURON WETLAND COMPLEX

Michael B. Whitt. Harold H. Prince, and Robert R. Cox. Jr.

SHORT COMMUNICATIONS

BALD EAGLE PREDATION ON COMMON LOON CHICK

James D. Paruk. Dean Seanfield. and Tara Mack

TERRITORIAL TAKEOVER IN COMMON LOONS (GAVIA IMMER) James D. Paruk

COURTSHIP BEHAVIOR OF THE BUFF-NECKED IBIS (THERISTICUS CAUDATUS)

Nathan H. Rice

HABITAT USE BY MASKED DUCKS ALONG THE GULF COAST OF TEXAS

James T. Ander.son and Thomas C. Tacha

GIZZARD CONTENTS OF PIPING PLOVER CHICKS IN NORTHERN MICHIGAN

Francesca J. Cuthbert. Brian Scholtens. Uiuren C. Wemmer. and Robyn McLain

NESTING OF FOUR POORLY-KNOWN BIRD SPECIES ON THE CARIBBEAN SLOPE OF
COSTA RICA Bruce E. Young and James R. Zook

SEXUAL DIMORPHISM IN THE SONG OF SUMICHRAST’S WREN —

Monica Perez-ViUafana. Hector Gomez de Silva G.. and Atahualpa De Sue re -Medrano

AN INCIDENT OF FEMALE-FEMALE AGGRESSION IN THE HOUSE WREN —

Tom Alworth and Isabella B.R. Scheiber

NEST REUSE BY WOOD THRUSHES AND ROSE-BREASTED GROSBEAKS

Lyle E. Eriesen. Valerie E. Wyatt, and Michael D. Cadman

SINGING IN A MATED FEMALE WILSON’S WARBLER

William M. Gilbert and Adele F. Carroll

LAYING TIME OF THE BRONZED COWBIRD Brian D. Peer and Spencer G. Sealy

TEMPORAL DIFFERENCES IN POINT COUNTS OF BOTTOMLAND FOREST LANDBIRDS

Winston Paul Smith and Daniel J. Twedt

ORNITHOLOGICAL LITERATURE

100

105

115

1 16

118

1 19

121

124

128

130

132

134

137

139

144

NUMBER 2

MAJOR PAPERS

A NEW SPECIES IN THE MYRMOTHERULA HAEMATONOTA SUPERSPECIES (AVES; THAM-
NOPHILIDAE) FROM THE WESTERN AMAZONIAN LOWLANDS OF ECUADOR AND PERU

Niels Krabbe, Morton L. Isler, Phyllis R. Isler, Bret M. Whitney,

Jose Alvarez A., and Paul J. Greenfield

COMPARATIVE SPRING HABITAT AND FOOD USE BY TWO ARCTIC NESTING GEESE

Suzanne Carriere, Robert G. Bromley, and Gilles Gauthier

A TEST OF THE CONDITION-BIAS HYPOTHESIS YIELDS DIFFERENT RESULTS FOR TWO
SPECIES OF SPARROWHAWKS {ACCIPITER)

Edna Gorney, William S. Clark, and Yoram Yom-Tov

THE DEVELOPMENT OF A VOCAL THERMOREGULATORY RESPONSE TO TEMPERATURE IN

EMBRYOS OF THE DOMESTIC CHICKEN Shawn C. Bugden and Roger M. Evans

BEHAVIOR AND VOCALIZATIONS OF THE CAURA AND THE YAPACANA ANTBIRDS

Eevin J. Zimmer

HABITAT PATCH SIZE AND NESTING SUCCESS OF YELLOW-BREASTED CHATS

Dirk E. Burhans and Erank R. Thompson, III

AVIFAUNA OF A PARAGUAYAN CERRADO LOCALITY: PARQUE NACIONAL SERRANIA SAN

LUIS, DEPTO. CONCEPCION Mark B. Robbins, Rob. C. Eaucett, and Nathan H. Rice

NOTES ON THE AVIFAUNA OF TABASCO

Kevin Winker, Stefan Arriaga Weiss, Juana Lourdes Trejo P., and Patricia Escalante P.

PREDATION OF SMALL EGGS IN ARTIFICIAL NESTS: EFFECTS OF NEST POSITION, EDGE,

AND POTENTIAL PREDATOR ABUNDANCE IN EXTENSIVE FOREST !

Richard M. DeGraaf Thomas J. Maier, and Todd K. Fuller

BIRD USE OF BURNED AND UNBURNED CONIFEROUS FORESTS DURING WINTER

Karen J. Kreisel and Steven J. Stein

NEST PREDATORS OF OPEN AND CAVITY NESTING BIRDS IN OAK WOODLANDS

Kathryn L. Purcell and Jared Verner

SHORT COMMUNICATIONS

JUVENILE MARBLED MURRELET NURSERIES AND THE PRODUCTIVITY INDEX

Katherine J. Kuletz and John E. Piatt

“SNORKELING” BY THE CHICKS OF THE WATTLED JACANA

Carlos Bosque and Emilio A. Herrera

RAPID LONG-DISTANCE COLONIZATION OF LAKE GATUN, PANAMA, BY SNAIL KITES

George R. Angehr

THE “SIGNIFICANT OTHERS” OF AMERICAN KESTRELS: COHABITATION WITH AR-
THROPODS Jeffrey P. Neubig and John A. Smallwood

BARRED OWL NEST IN ATTIC OF SHED C. Stuart Houston

DOUBLE BROODING IN THE LONG-EARED OWL

Jeffrey S. Marks and Alison E. H. Perkins

PLANNING TO FACILITATE CACHING: POSSIBLE SUET CUTTING BY A COMMON RA-
VEN Bernd Heinrich

PAIRING SUCCESS OF WOOD THRUSHES IN A FRAGMENTED AGRICULTURAL LAND-
SCAPE Lyle E. Eriesen, Valerie E. Wyatt, and Michael D. Cadman

CONNECTICUT WARBLER, A NORTH AMERICAN MIGRANT NEW TO ECUADOR

Olaf Jahn, Maria Eugenia Jara Viteri, and Karl-L. Schuchmann

PARENTAL BEHAVIOR OF A BIGAMOUS MALE NORTHERN CARDINAL

Randall Breitwisch, Amy J. Schilling, and Joshua B. Banks

SPECIAL REPORT

A SURVEY OF UNDERGRADUATE ORNITHOLOGY COURSES IN NORTH AMERICA ..L . _

Edward H. Burtt, Jr. and W. Herbert Wilson, Jr.

ORNITHOLOGICAL LITERATURE

157

166

181

188

195

210

216

229

236

243

251

257

262

265

269

272

273
276
279
281
283

2&7

294

NUMBER 3

ANTILLEAN SHORT-EARED OWLS INVADE SOUTHERN ELORIDA -

Wayne Hoffman, Glen E. WoolfenJen, and P. William Smith 303

WITHIN- AND BETWEEN-YEAR DISPERSAL OE AMERICAN AVOCETS AMONG MULTIPLE

WESTERN GREAT BASIN WETLANDS

Jonathan H. Plissner, Susan M. Haig, and Lewis W. Oring 3 1 4

HIGH MORTALITY OE PIPING PLOVERS ON BEACHES WITH ABUNDANT GHOST CRABS;

CORRELATION, NOT CAUSATION Donna L Wolcott and Thomas G. Wolcott 321

A TAXONOMIC STUDY OE CRESTED CARACARAS (EALCONIDAE)

Carla J. Dove and Richard C. Banks 330

VISUAL COMMUNICATION AND SEXUAL SELECTION IN A NOCTURNAL BIRD SPECIES, CA-
PRIMULGUS RUFICOLLIS, A BALANCE BETWEEN CRYPSIS AND CONSPICUOUSNESS -
Juan Aragones, Luis Arias De Reyna, and Pilar Recuerda 340

INTERSPECIEIC INTERACTIONS WITH FORAGING RED-COCKADED WOODPECKERS IN

SOUTH-CENTRAL FLORIDA

Reed Bowman, David L. Leonard, Jr., Leslie K. Backus, and Allison R. Mains 346

SPATIAL AND TEMPORAL DYNAMICS OF A PURPLE MARTIN PRE-MIGRATORY ROOST

Kevin R. Russell and Sidney A. Gauthreaux, Jr. 354

AGGRESSIVE RESPONSE OF CHICKADEES TOWARDS BLACK-CAPPED AND CAROLINA

CHICKADEE CALLS IN CENTRAL ILLINOIS Eric L. Kershner and Eric K. Bollinger 363

USE OF SONG TYPES BY MOUNTAIN CHICKADEES {POECILE GAMBELL)

Myra O. Wiebe and M. Ross Lein 368

SURVIVAL AND LONGEVITY OF THE PUERTO RICAN VIREO

Bethany L. Woodworth, John Faaborg, and Wayne J. Arendt 2sl()

EFFECTS OF PRIOR RESIDENCE AND AGE ON BREEDING PERFORMANCE IN YELLOW WAR-
BLERS Lozano and R. E. Lemon 381

DISTRIBUTION AND HABITAT ASSOCIATIONS OF THREE ENDEMIC GRASSLAND SONG-
BIRDS IN SOUTHERN SASKATCHEWAN S. K. Davis, D. C. Duncan, and M. Skeel 389

BIRD COMMUNITIES IN NATURAL FOREST PATCHES IN SOUTHERN BRAZIL

Luiz Dos Anjos and Roberto Bogon 397

DO MAMMALIAN NEST PREDATORS FOLLOW HUMAN SCENT TRAILS IN THE SHORTGRASS

PRAIRIE? Susan K. Skagen, Thomas R. Stanley, and M. Beth Dillon 415

SHORT COMMUNICATIONS

CHRISTMAS SHEARWATER EGG DIMENSIONS AND SHELL CHARACTERISTICS ON LAY-

SAN ISLAND, NORTHWESTERN HAWAIIAN ISLANDS

G. C. Whittow and M. B. Naughton 421

THE PAINT-BILLED CRAKE BREEDING IN COSTA RICA

_ David M. Watson and Brett W. Benz 422

ADDITIONAL RECORDS OF FALL AND WINTER NESTING BY KILLDEER IN SOUTHERN

UNITED STATES Kimberly G. Smith, W. Marvin Davis, Thomas E. Kienzle,

William Post, and Robert W. Chinn 424

WILD TURKEYS (MELEAGRIS GALLOPAVO) RENEST AFTER SUCCESSFUL HATCH

Craig A. Harper and Jay H. E.xum 426

POST-MIGRATION WEIGHT GAIN OF SWAINSON’S HAWKS IN ARGENTINA

Michael I. GoULstein, Peter H. Bloom, Jose H. Sarasola, and Thomas E. Lacher 428

SIBLICIDE AT NORTHERN GOSHAWK NESTS: DOES FOOD PLAY A ROLE?

Wendy A. Estes, Sarah R. Dewey, and Patricia L. Kennedy 432

COOPERATIVE FORAGING IN THE MOUNTAIN CARACARA IN PERU Ja.son Jones 437

PREDATION BY RUFOUS MOTMOT ON BLACK-AND-GREEN POISON DART FROG

Terry L. Master 439

EVIDENCE OF EGG EJECTION IN MOUNTAIN BLUEBIRDS Percy N. Hebert 440

FORAGING OVENBIRD FOLLOWS ARMADILLO Douglas J. Levey 443

ORNITHOLOGICAL LITERATURE 445

NUMBER 4

A NEW SPECIES OF HAWK-OWL NINOX FROM NORTH SULAWESI, INDONESIA

(j Rasmussen

PATTERNS OF VARIATION IN SIZE AND COMPOSITION OF GREATER SCAUP EGGS: ARE

THEY RELATED? — Paul L. Flint and J. Barry Grand

POSTFLEDGING BEHAVIOR OF GOLDEN EAGLES

Laura T. O'Toole, Patricia L. Kennedy, Richard L. Knight, and Lowell C. McEwen

GROWTH PATTERNS OF HAWAIIAN STILT CHICKS

Michael Reed, Elizabeth M. Gray, Dianne Lewis, Lewis W. Oring, Richard Coleman,

Timothy Burr, and Peter Luscomb

NESTING BEHAVIOR OF THE LILAC-CROWNED PARROT

- Katherine Renton and Alejandro Salinas-Melgoza

RELATIONSHIPS AMONG RED-COCKADED WOODPECKER GROUP DENSITY, NESTLING

PROVISIONING RATES, AND HABITAT

Richard N. Conner, D. Craig Rudolph, Richard R. Schaefer,

Daniel Saenz, and Clifford E. Shackeljbrd
RESPONSES OF BELL’S VIREOS TO BROOD PARASITISM BY THE BROWN-HEADED COWBIRD

IN KANSAS Timothy El. Parker

THE TYPE B SONG OF THE NORTHERN PARULA: STRUCTURE AND GEOGRAPHIC VARIATION
ALONG PROPOSED SUB-SPECIES BOUNDARIES Michael D. Bay

NESTING BIOLOGY OF DICKCISSELS AND HENSLOW’S SPARROWS IN SOUTHWESTERN

MISSOURI PRAIRIE FRAGMENTS Maiken Winter

RESPONSE OF A BIRD ASSEMBLAGE IN SEMIARID CHILE TO THE 1997-1998 EL NINO

- - - Fabian M. Jaksic and Ivan Lazo

POTENTIAL FOR PREDATOR LEARNING OF ARTIFICIAL ARBOREAL NEST LOCATIONS

---- - - Richard H. Yahner and Carolyn G. Mahan

PREDATION ON ARTIFICIAL NESTS ALONG THREE EDGE TYPES IN A NORTH CAROLINA

BOTTOMLAND HARDWOOD FOREST James F. Saracco and Jaime A. Collazo

THE RESPONSE OF A KANSAS WINTER BIRD COMMUNITY TO WEATHER, PHOTOPERIOD,
AND YEAR Martin A. Stapanian, Christopher C. Smith, and Elmer J. Finck

SHORT COMMUNICATIONS

POSSIBLE WINTER QUARTERS OF THE ALEUTIAN TERN?

- Norman P. Hill and K. David Bishop

ARTHROPODS AND PREDATION OF ARTIFICIAL NESTS IN THE BAHAMAS: IMPLICATIONS

FOR SUBTROPICAL AVIFAUNA Nancy L. Staus and Paul M. Maver

NOTES ABOUT THE DISTRIBUTION OF PAUXI PAUXI AND ABURRIA ABURRI IN VENE-
ZUELA Jose L. Silva

WESTERN BURROWING OWLS IN CALIFORNIA PRODUCE SECOND BROODS OF CHICKS

Jennifer A. Gervais and Daniel K. Rosenberg

CONTINUOUS NESTING OF BARN OWLS IN ILLINOIS

leffery W. Walk, Terry L. Esker, and Scott A. Simpson

UNUSUAL NEST SITES FOR SOUTHWESTERN WILLOW FLYCATCHERS

Scott H. Stoleson and Deborah M. Finch

MORTALITY OF AN ADULT VEERY INCURRED DURING THE DEFENSE OF NESTLINGS

David /. King

RELATIONSHIPS OF CLUTCH SIZE AND HATCHING SUCCESS TO AGE OF FEMALE PRO-

THONOTARY WARBLERS Charles R. Blem, Leann B. Blem, and Claudia /. Barrientos

HYBRIDIZATION BETWEEN CLAY-COLORED SPARROW AND FIELD SPARROW IN
NORTHERN VERMONT David J. Hoag

COMMENTARY

A CRITIQUE OF WANG YONG AND FINCH’S FIELD-IDENTIFICATIONS OF WILLOW FLY-
CATCHER SUBSPECIES IN NEW MEXICO John P Hubbard

RESPONSE Wang Yong and Deborah M. Finch

ORNITHOLOGICAL LITERATURE

PROCEEDINGS OF THE EIGHTIETH ANNUAL MEETING

ACKNOWLEDGMENTS

INDEX TO VOLUME I I I

CONTENTS TO VOLUME I I I

457

465

472

478

488

494

499

505

515

527

536

541

550

559

561

564

569

572

574

576

577
581

585

589

593

600

608

610

THE WILSON BULLETIN

Editor ROBERT C. BEASON

Editorial Board KATHY G. BEAL

Department of Biology

State University of New York

1 College Circle

Geneseo, NY 14454

E-mail: WilsonBull@geneseo.edu

CLAIT E. BRAUN
RICHARD N. CONNER

Review Editor WILLIAM E. DAVIS, JR.

127 East Street

Foxboro, Massachusetts 02035

Editorial Assistants TARA BAIDEME

Index Editor KATHY G. BEAL

JOHN LAMAR
DANTE THOMAS
DORIS WATT

616 Xenia Avenue
Yellow Springs, Ohio 45387

SUGGESTIONS TO AUTHORS

See Wilson Bulletin, 1 10:152-154, 1998 for more detailed “Instructions to Authors.”
http://www.ummz.lsa.umich.edu/birds/wilsonbull.html
Submit four copies of manuscripts intended for publication in The Wilson Bulletin, 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 (Seventh Edition,
1998) insofar as scientific names of U.S., Canadian, Mexican, Central American, and West Indian birds
are concerned. Abstracts should be brief but quotable. 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 Scientific Style and Format Manual” (AIBS 1994). Photographs for illustra-
tions should have good contrast and be on glossy paper. Submit prints unmounted and provide a brief but
adequate legend for each figure with all captions on a single page. 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 Ornitho-
logical Societies of North America, P.O. Box 1897, Lawrence, KS 66044-8897.

MEMBERSHIP INQUIRIES

Membership inquiries should be sent to Laurie J. Goodrich, Route 2 Box 301 A, New Ringgold, PA
17960-9445; E-mail: goodrich@hawkmountain.org.

CONTENTS

A NEW SPECIES OF HAWK-OWL A/AQYFROM NORTH SULAWESI, INDONESIA

— - — - - Pamela C. Rasmussen

PATTERNS OF VARIATION IN SIZE AND COMPOSITION OF GREATER SCAUP EGGS: ARE

THEY RELATED? — Paul L. Flint and J. Barry Grand

POSTFLEDGING BEHAVIOR OF GOLDEN EAGLES

— - Laura T. O Toole, Patricia L. Kennedy, Richard L. Knight, and Lowell C. McEwen

GROWTH PATTERNS OF HAWAIIAN STILT CHICKS

J- Michael Reed, Elizabeth M. Gray, Dianne Lewis, Levins W. Oring, Richard Coleman,

Timothy Burr, and Peter Luscomb

NESTING BEHAVIOR OF THE LILAC-CROWNED PARROT

- - Katherine Renton and Alejandro Salinas-Melgoza

RELATIONSHIPS AMONG RED-COCKADED WOODPECKER GROUP DENSITY, NESTLING

PROVISIONING RATES, AND HABITAT

Richard N. Conner, D. Craig Rudolph, Richard R. Schaefer,

Daniel Saenz, and Clifford E. Shackelford
RESPONSES OF BELL’S VIREOS TO BROOD PARASITISM BY THE BROWN-HEADED COWBIRD

IN KANSAS Timothy H. Parker

THE TYPE B SONG OF THE NORTHERN PARULA: STRUCTURE AND GEOGRAPHIC VARIATION

ALONG PROPOSED SUB-SPECIES BOUNDARIES Michael D. Bay

NESTING BIOLOGY OF DICKCISSELS AND HENSLOW’S SPARROWS IN SOUTHWESTERN

MISSOURI PRAIRIE FRAGMENTS Maiken Winter

RESPONSE OF A BIRD ASSEMBLAGE IN SEMIARID CHILE TO THE 1997-1998 EL NINO

- - - Fabian M. Jaksic and Ivan Lazo

POTENTIAL FOR PREDATOR LEARNING OF ARTIFICIAL ARBOREAL NEST LOCATIONS

Richard H. Yahner and Carolyn G. Mahan

PREDATION ON ARTIFICIAL NESTS ALONG THREE EDGE TYPES IN A NORTH CAROLINA

BOTTOMLAND HARDWOOD FOREST James F. Saracco and Jaime A. Collazo

THE RESPONSE OF A KANSAS WINTER BIRD COMMUNITY TO WEATHER. PHOTOPERIOD,
AND YEAR Martin A. Stapanian, Christopher C. Smith, and Elmer J. Finck

SHORT COMMUNICATIONS

POSSIBLE WINTER QUARTERS OF THE ALEUTIAN TERN?

- Norman P Hill and K. David Bishop

ARTHROPODS AND PREDATION OF ARTIFICIAL NESTS IN THE BAHAMAS: IMPLICATIONS

FOR SUBTROPICAL AVIFAUNA Nancy L. Staus and Paul M. Mayer

NOTES ABOUT THE DISTRIBUTION OF PAUXI PAUXI AND ABURRIA ABURRI IN VENE-
ZUELA Jose L. Silva

WESTERN BURROWING OWLS IN CALIFORNIA PRODUCE SECOND BROODS OF CHICKS

Jennifer A. Gervais and Daniel K. Rosenberg

CONTINUOUS NESTING OF BARN OWLS IN ILLINOIS

Jeffery W. Walk. Terry L. Esker, attd Scott A. Simpson

UNUSUAL NEST SITES FOR SOUTHWESTERN WILLOW FLYCATCHERS

Scott H. Stoleson and Deborah M. Finch

MORTALITY OF AN ADULT VEERY INCURRED DURING THE DEFENSE OF NESTLINGS

- David 1. King

RELATIONSHIPS OF CLUTCH SIZE AND HATCHING SUCCESS TO AGE OF FEMALE PRO-

THONOTARY WARBLERS Charles R. Bletn, Leann B. Blern, and Claudia /. Barrientos

HYBRIDIZATION BETWEEN CLAY-COLORED SPARROW AND FIELD SPARROW IN
NORTHERN VERMONT David J Hoag

COMMENTARY

A CRITIQUE OF WANG YONG AND FINCH’S FIELD-IDENTIFICATIONS OF WILLOW FLY-
CATCHER SUBSPECIES IN NEW MEXICO John P. Hubbard

RESPONSE Wang Yong and Deborah M. Finch

ORNITHOLOGICAL LITERATURE

PROCEEDINGS OF THE EIGHTIETH ANNUAL MEETING

ACKNOWLEDGMENTS —

INDEX TO VOLUME I I I

CONTENTS TO VOLUME I I I

457

465

472

478

488

494

499

505

515

527

536

541

550

559

561

564

569

572

574

576

577
581

585

589

593

600

608

610

3 2044

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The Wilson bulletin"

See other formats