'g^lL

.6L

HARVARD UNIVERSITY

Ernst Mayr Library of the Museum of Comparative Zoology

The WIsoti Bulletin

PUBLISHED BY THE WILSON ORNITHOLOGICAL SOCIETY

VOL. 116, NO. 1

(ISSN 0043-5643)

MARCH 2004

PAGES 1-118

THE WILSON ORNITHOLOGICAL SOCIETY FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

President — Charles R. Blem, Dept, of Biology, Virginia Commonwealth Univ., Richmond, VA 23284, USA; e-mail: cblem@saturn.vcu.edu

First Vice-President — Doris J. Watt, Dept, of Biology, Saint Mary’s College, Notre Dame, IN 46556, USA; e-mail: dwatt@saintmarys.edu

Second Vice-President — James D. Rising, Dept, of Zoology, Univ. of Toronto, Toronto, ON M5S 3G5, Canada; e-mail: rising@zoo.utoronto.ca

Editor — James A. Sedgwick, U.S. Geological Survey, Fort Collins Science Center, 2150 Centre Ave., Bldg. C, Fort Collins, CO 80526, USA; e-mail:wilsonbulletin@usgs.gov

Secretary — Sara R. Morris, Dept, of Biology, Canisius College, Buffalo, NY 14208, USA; e-mail: morriss@canisius.edu

Treasurer — Martha Vaughan, 9 Bennets Neck Dr., Pocasset, MA 02559, USA; e-mail: jkricher@wheatonma.edu Elected Council Members — D. Craig Rudolph and John A. Shetterly (terms expire 2004), R. Todd Engstrom and E. Dale Kennedy (terms expire 2005), and Robert C. Beason and Mary Gustafson (terms expire 2006).

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 Univ. of Michigan Museum of Zoology, was established in concurrence with the Univ. of Michigan in 1930. Until 1947 the Library was maintained entirely by gifts and bequests of books, reprints, and ornithological mag- azines 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. Terry L. Root, Univ. 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 Univ. 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 Univ. of Michigan, 1109 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 10th St., 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.

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 48109.

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. Back issues or single copies are available for $12.00 each. Most back issues of the Bulletin are available and may be ordered from OSNA. Special prices will be quot- ed for quantity orders. All issues of The Wilson Bulletin published before 2000 are accessible on a free web site at the Univ. of New Mexico library (http://elibrary.unm.edu/wilson/). The site is fully searchable, and full-text reproductions of all papers (including illustrations) are available as either PDF or DjVu files.

© Copyright 2003 by the Wilson Ornithological Society Printed by Allen Press, Inc., Lawrence, Kansas 66044, U.S. A.

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

FRONTISPIECE. Five species (Least Flycatcher, Empidonax minimus', Veery, Catharus fuscescens', Wood Thrush Hylocichla mustelina; Common Yellowthroat, Geothlypis trichas', and Scarlet Tanager, Piranga olivacea) showing signifi- cant declines, 1970-2001, in spring and/or fall capture rates at Manomet Center for Conservation Sciences, Massachusetts, and which also showed declining population trends in at least one of three Breeding Bird Survey physiographic strata (south- ern New England, northern New England, and eastern Spruce-Hardwoods) that were considered likely sources of the migrants passing through Manomet. Painting by Dale Crawford.

THE WILSON BULLETIN

A QUARTERLY JOURNAL OF ORNITHOLOGY Published by the Wilson Ornithological Society

VOL. 116, NO. 1 March 2004 PAGES 1-118

Wilson Bulletin, 116(1), 2004, pp. 1-16

32 YEARS OF CHANGES IN PASSERINE NUMBERS DURING SPRING AND FALL MIGRATIONS IN COASTAL MASSACHUSETTS

TREVOR L. LLOYD-EVANSi 3 AND JONATHAN L. ATWOOD' ^

ABSTRACT. — Using standardized mist-net captures collected over a 32-year period (1970-2001), we ex- amined changes in the capture rates of passerines recorded in coastal Massachusetts during fall (78 species) and spring (72 species) migration. Capture rates of 45 species of fall migrants (58%) declined significantly between early (1970-1985) and late (1986-2001) years of the study; 36 species of spring migrants (50%) showed sig- nificant declines. Only Carolina Wren {Thryothorus ludovicianus). Tufted Titmouse (Baeolophiis hicolor). North- ern Cardinal {Cardinalis cardinalis), and Orchard Oriole (Icterus spurius) showed significant increases during spring migration; fall sampling indicated that Carolina Wren, Tufted Titmouse, Black-throated Blue Warbler (Dendroica caerulescens), and Northern Cardinal had significantly higher capture rates. Of 37 species included in the migration monitoring data but not reliably represented by Breeding Bird Survey (BBS) data in any of the northeastern physiographic strata, 23 (62%) showed significant declines at Manomet during at least one of the two migration periods. There were significant correlations in percent changes in migrant capture rates between fall and spring. BBS trends reported from the southern New England and northern New England physiographic strata were correlated with changes in migrant capture rates. However, there were also inconsistencies between results obtained by the two monitoring approaches, suggesting that factors in addition to actual changes in breeding populations may be reflected in the migration capture data. Received 8 July 2003, accepted 26 March 2004.

Monitoring passerine population changes through counts collected along migratory routes has been attempted often (Hussell 1981, Gauthreaux 1992, Hagan et al. 1992, Hus.sell et al. 1992, Peach et al. 1998, Ballard et al. 2003) despite a variety of issues that sometimes make the results of such studies difficult to interpret. In particular, detecting true changes in breeding populations may be confounded by weather effects that produce

' Manomet Center for Conservation Sciences, P.O. Box 1770. Manomet. MA 02.345, USA.

* Current address: Conservation Biology Program, Dept, of Environmental Studies, Antioch New Paiglatid Graduate School, 40 Avon St., Keene, Nil 03431, USA.

’Corresponding author; e-mail; tlloyd-e vans (ofinanomet.org

dramatic differences among years in the num- bers of a particular species that appears during migration at a specific site (Gauthreaux 1971. Moore and Simons 1992, Dunn and Hussell 1995); while “fallouts” may provide exciting birding conditions, they also underscore the substantial stochastic element associated with any migration monitoring scheme. Habitat changes at a migration site also may cause apparent shifts in species' abundances that arc unrelated to true population levels (Remsen and Good 1996). Furthermore, the specific breeding populations actually represented by samples of migrants are almost always un- known (Dunn and Hus.sell 1995). and con- ceivably may vary from year-to-year at a par- ticular migration site under the influence of differing weather conditions, riius, there is lit-

I

2

THE WILSON BULLETIN • Vol. 116, No. 1, March 2004

tie doubt, as some have pointed out (Butcher et al. 1993, Sauer 1993, Remsen and Good 1996), that monitoring changes in breeding populations through counts of migrants ob- tained by mist-net captures is risky business.

Still, most long-term field observers will quickly counter that something is happening to the numbers of migrating land birds in east- ern North America (Robbins et al. 1989, Ter- borgh 1989, Askins et al. 1990), and that these perceived changes are not easily discounted simply by the effects of weather variations or local habitat change. In fact, although short- term fluctuations in numbers of migrants re- corded at a site may be completely meaning- less, we contend that studies of longer dura- tion— despite their inherent complications — may yet help to elucidate true population changes simply by virtue of their long-term perspective.

In this paper we present results, to date, of one of North America’s longest migration monitoring efforts, conducted at Manomet Center for Conservation Sciences (formerly Manomet Bird Observatory, MBO) from the late 1960s to the present. A preliminary anal- ysis of some of these data was presented by Hagan et al. (1992); herein, we extend the scope of this earlier work in terms of years, seasons, and species included. For 78 species in fall and 72 species in spring we examine, for the 32-year period 1970-2001, (a) changes in the numbers of individuals captured at Manomet’s banding station in coastal Massa- chusetts, and (b) similarities in patterns of an- nual fluctuations of capture rates among spe- cies. We also compare changes in capture rates with estimates of population trends ob- tained through a very different type of moni- toring study, the North American Breeding Bird Survey (BBS), which also has operated over this extensive time period (Robbins et al. 1986, Sauer 1993, Sauer et al. 2001).

METHODS

Manomet Center for Conservation Scienc- es, located on the western side of Cape Cod Bay, Plymouth County, Massachusetts (41° 50' N, 70° 30' W), is characterized by brushy, second-growth deciduous woodland, bordered on the east and south by a steep, eroding coastal bluff and on the west and north by brushy wetlands. Dominant tree species on the

7-ha plot include black cherry (Prunus sero- tina), shadbush (Amelanchier sp.), red maple {Acer ruhrum), white oak (Quercus alba), and pitch pine (Pinus rigida). Common catbrier (Smilax rotiindifolia), bay berry (Myrica pen- sylvanica), staghorn sumac {Rhus typhina), honeysuckle {Lonicera morrowi), arrowwood {Viburnum recognitum), and poison ivy {Tox- icodendron radicans) are principal understory species.

Habitat succession was, for the most part, unchecked during the study period, but the site’s location on an exposed coastal bluff re- sulted in annual natural “pruning” by harsh winter storms that probably reduced the de- gree of change in habitat structure over time. Small fields and grassland borders within the study site are mowed routinely. Historic pho- tos of the area indicate that during the early 1920s most of the study area consisted of open sheep pastures, but by the time banding operations were begun in 1966 the site had already acquired the brushy, second-growth condition that characterizes it today. An in- dividual black cherry tree was photographed in 1966, with a bander for height comparison, in a net lane about 1 0 m inland from the ocean bluff. By 2003, the tree had grown an esti- mated 25% in height, probably typical for the exposed coastal net lanes.

From 45 to 50 nylon mist nets (12 m long, 2.6 m high, 4 panels, 36 mm extended mesh) were operated annually from 1970 to 2001, inclusive; because of less complete coverage and imprecise records regarding capture effort expended during the first 4 years of the ob- servatory’s existence (1966-1969), we ex- cluded these years from analysis. Nets were kept at fixed locations throughout the study. Opening and closing times of nets were re- corded and used for calculating daily capture effort (Robbins 1968); except for closures during adverse weather conditions, generally nets were operated from 0.5 hr prior to sunrise to 0.5 hr after sunset. Thus, 50 nets open for 12 hr equals 600 net hr. Sampling was con- ducted 5—7 days per week during spring (15 April- 15 June) and fall (15 August- 15 No- vember) migration.

During the study period, 205,454 individ- uals of 159 species were banded. Records used in this analysis were selected from the overall database using criteria described be-

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

3

low. Only passerines are considered here; sci- entific names and abbreviation codes for spe- cies referenced in the text are provided in the Appendix. Willow and Alder flycatchers were combined, as were Bicknell’s and Gray- cheeked thrushes. Palm Warbler races were treated separately as “Yellow” and “West- ern” Palm warblers. Captures of hybrid “Brewster’s” {n = 3) and “Lawrence’s” {n = 2) warblers were counted as Blue-winged Warblers. Repeat captures were eliminated. Locally breeding birds, identified on the basis of well-developed brood patches or cloacal protuberances, were eliminated, as were spring captures of hatching-year (HY) individ- uals. Species that were captured, by season, in fewer than 15 of the 32 years, were eliminat- ed.

For each species, by season, migration win- dows were calculated as falling between the L‘ and percentiles of all capture dates across all years; any records outside these windows were excluded. These cutoff values are provided in the Appendix. For example, during fall migration, 98% of all captures of American Redstarts occurred from 17 August to 12 October. Any banding activity that took place within this window was considered to represent a legitimate sampling day for this species; days that yielded no redstart captures, but on which nets were open, contributed a value of zero to the overall calculation of cap- ture rate. Any redstart captures that occurred before 17 August or after 12 October were excluded.

For each species (by year and season), we calculated a mean capture rate weighted by the number of hours of mist netting that oc- curred on each contributing date. That is, in calculating mean seasonal and annual capture rates for a species, the rate obtained on a day when nets were open for 400 net hr was given more emphasis than a rate obtained on a day when only 10 net hr of sampling took place. We used Wilcoxon 2-sample tests to examine long-term trends by testing (for each species, by season) the hypothesis that mean capture rates were equal between Early (1970-1985) and Late (1986-2001) years of the study.

Spearman rank correlations were used to assess concordance between each species' fall and spring capture rates, and between the per- cent change in mean capture rates (liarly ver-

sus Late) for each species and the population trends provided by BBS data (Sauer et al. 2001). These authors commendably cautioned that “Small sample sizes, low relative abun- dance on survey routes, imprecise trends, and missing data all can compromise BBS results. Often, users do not take these problems into account when viewing BBS results, and use the results inappropriately.” When we refer to BBS trends in this paper, we conservatively include only instances where the BBS “Re- gional Credibility Measure” was in the best- sampled, “blue” category. That is, BBS trends considered by Sauer et al. (2001) to include “important deficiencies” (red) and “deficiencies” (yellow) were not used in the correlation analyses.

Presentation of graphs showing changes in capture rates for each species and season com- bination in this study would require 150 in- dividual figures. Although obviously beyond the space limitations of this publication, these results are provided online at www. manomet.org. Here, in order to visually sum- marize major patterns of variation within this large set of data, we calculated 3-year moving averages based on annual mean capture rates, then standardized each of these values as a percent of the maximum rate encountered for each species among all years (by season). Next, we used Ward’s minimum variance clus- tering approach, as implemented by JMP sta- tistical software (SAS Institute, Inc. 2001 ). to identify, for each season, an arbitrary six groups of species that exhibited similar year- to-year fluctuations in capture rates. Finally, we plotted means and standard errors, calcu- lated from the moving averages for species belonging to each of these clusters.

RESULTS

Of 72 species captured during spring mi- gration, 60 (83%) had lower mean capture rates during 1986-2001 than during 1970- 1985 (Table 1). These declines were signifi- cant (P < 0.05) in 36 species. I weLe species showed higher capture rates during 1986- 2001 than during 1970-1985; in four of these (Carolina Wren, ruf'ted 'fitmouse. Northern Cardinal, and Orchard Orit)le), the increases from F^arly to Late Sami')! ing periotls were sig- nificant ’ < 0.01 ).

I)uri..g fall migration. 69 ot 78 s|')ecies

4

THE WILSON BULLETIN • Vol. 116, No. 1, March 2004

TABLE 1. Mean capture rates and percent change between Early (1970-1985) and Late (1986-2001) sam- pling periods during spring and fall migrations. Population trend data from Breeding Bird Survey (BBS) pre- sented for comparison.

		Spring capture rate^		Fall capture rate^		BBS‘’
Species	Early	Late	(% change)	Early	Late	(% change)	SH	nNE	sNE
Eastern Wood-	0.766	0.464	(-39)	0.183	0.141	(-23)	D	D	d
Pewee Yellow-bellied	1.539	1.336	(-13)	0.455	0.297	(-35)*	[i]	I
Elycatcher Acadian Ply-	0.234	0.206	(-12)						[d]
catcher Willow/Alder Ply-	3.269	3.730	(14)	0.754	0.557	(-26)	[i]	I	m
catcher
Least Elycatcher	0.844	0.674	(-20)	0.866	0.328	(-62)**	D	D	D
Eastern Phoebe	0.210	0.200	(-5)	0.531	0.574	(8)	[I]	[i]	[i]
Great Crested Ply-	0.535	0.813	(52)				D	I	D
catcher Eastern Kingbird	0.342	0.280	(-18)	0.477	0.108	(-77)*	D	d	D
White-eyed Vireo	0.360	0.155	(-57)**	0.143	0.092	(-36)			[I]
Blue-headed Vireo	0.313	0.265	(-15)	0.461	0.610	(32)	I	I	[i]
Warbling Vireo				0.131	0.074	(-44)	[i]	I	i
Philadelphia Vireo				0.379	0.208	(-45)**	[i]	[i]
Red-eyed Vireo	1.316	0.783	(-40)*	4.317	2.834	(-34)*	I	[d]	[d]
Blue Jay	7.071	2.767	(-61)**	2.326	1.289	(-45)*	[i]	i	D
Black-capped	3.176	0.773	(-76)	37.479	18.411	(-51)	I	I	i
Chickadee Tufted Titmouse	0.162	0.593	(266)**	3.672	6.520	(78)*	[i]	[I]	[I]
Red-breasted Nut-				0.291	0.092	(-68)	I	I	[i]
hatch White-breasted				0.156	0.204	(31)	[i]	I	i
Nuthatch
Brown Creeper	0.471	0.148	(-69)**	2.750	1.320	(-52)***	[i]	[i]	[d]
Carolina Wren	0.043	0.146	(240)***	0.072	0.546	(658)***			[I]
House Wren	0.368	0.166	(-55)*	0.269	0.182	(-32)	[d]	[D]	[D]
Winter Wren				0.325	0.224	(-31)	[I]	[i]	[i]
Golden-crowned	0.454	0.943	(108)	5.176	3.981	(-23)	I	[i]
Kinglet Ruby-crowned	4.793	3.014	(-37)	2.964	1.917	(-35)	D	[i]
Kinglet Blue-gray Gnat-	0.724	0.385	(-47)**	0.344	0.255	(-26)		[I]	[I]
catcher
Veery	1.617	0.722	(-55)**	0.909	0.534	(-41)*	[D]	D	d
Gray-cheeked/	0.415	0.140	(-66)**	0.342	0.190	(-44)**
Bicknell’s Thrush
Swainson’s Thrush	4.708	2.069	(-56)**	2.181	0.996	(-54)**	[D]	[d]
Hermit Thrush	3.545	3.706	(5)	3.022	2.548	(-16)	[I]	[i]	[d]
Wood Thrush	1.211	0.398	(-67)***	0.306	0.113	(-63)***	D	[D]	D
American Robin	0.767	0.420	(-45)**	7.925	3.382	(-57)**	i	[d]	d
Gray Catbird	32.243	23.340	(-28)**	24.028	17.410	(-28)**	[D]	D	I
Northern Mock-	0.176	0.203	(15)	0.671	0.327	(-51)*	[I]	[I]	[I]
ingbird
Brown Thrasher	0.893	0.364	(-59)***	0.400	0.111		[D]	[D]	D
Cedar Waxwing	0.499	0.882	(77)	0.474	0.314	(-34)	i	I	I
Blue-winged War-				0.228	0.234	(3)		[d]	D
bler Tennessee Warbler	0.938	0.048		0.381	0.069	( — 82)***	[i]	[d]
Orange-crowned				0.244	0.157	(-36)	[d]

Warbler

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

5

TABLE 1. Continued.

		Spring capture rate^		Fall capture rate^		BBS^
Species	Early	Late	(% change)	Early	Late	(% change)	SH	nNE	sNE
Nashville Warbler	0.304	0.122	(-60)*	0.666	0.428	(-36)*	i	[d]	[d]
Northern Parula	0.555	0.287	(-48)**	0.116	0.050	(-57)*	[I]	[i]	[i]
Yellow Warbler	1.574	1.162	(-26)*	0.528	0.168	(-68)**	[i]	d	[I]
Chestnut-sided	0.292	0.171	(-41)	0.162	0.126	(-22)	d	[D]	d
Warbler
Magnolia Warbler	5.105	5.572	(9)	0.998	0.881	(-12)	I	d	[i]
Cape May Warbler				1.087	0.077		[i]	[i]
Black- throated	0.910	0.861	(-5)	0.549	0.781	(42)*	[i]	i	[i]
Blue Warbler
Yellow-rumped	1.285	0.965	(-25)	45.991	17.639	(-62)***	I	I	[I]
(Myrtle) War- bler
Black-throated	0.208	0.098	(-53)*	0.325	0.250	(-23)	[nc]	i	[I]
Green Warbler Blackburnian War-	0.155	0.090	(-42)*	0.093	0.028	(-70)**	i	[d]	[d]
bier
Prairie Warbler	0.318	0.235	(-26)	0.249	0.187	(-25)		[i]	[D]
Palm Warbler (western) Palm Warbler (yel-	0.706	0.900	(28)	0.543	0.132	(-76)***	[I]
low) Bay-breasted War-	0.338	0.121	(-64)	1.822	0.254	(-86)***	[D]	[i]
bler
Blackpoll Warbler	2.881	1.384	(-52)**	14.753	4.268	^ — '71	[d]	[i]
Black-and- White	5.244	3.310	(-37)**	1.643	0.802	(-51)**	i	d	d
Warbler American Redstart	7.394	4.777	(-35)**	6.351	2.889	( — 55)***	d	d	[I]
Ovenbird	2.991	2.057	(-31)*	0.726	0.586	(-19)	[nc]	I	nc
Northern Water-	3.424	2.091	(-39)	1.341	0.654	(-51)***	[d]	[nc]	[nc]
thrush Connecticut War-				0.232	0.151	(-35)	[d]
bler
Mourning Warbler	1.688	1.531	(-9)	0.447	0.244	(-45)**	[d]	[d]
Common Yellow-	9.441	6.769	T to 00 *	2.294	1.287	_44)***	d	D	D
throat Wilson’s Warbler	2.733	1.310	(-52)**	1.150	0.735	(-36)**	[i]
Canada Warbler	4.548	2.378	(-48)**	0.925	0.596	(-36)*	d	d	[d]
Yellow-breasted				1 .334	0.645	(_52)***			(17|
Chat
Scarlet Tanager				0.418	0.108	( — 74)* + *	Id]	D	Id]
Eastern Towhee	3.453	1.148	(-67)***	1 . 1 35	0.264	(-77)***	ini	1)	1)
American Tree Sparrow Chipping Sparrow	0. 1 65	0.076	(-54)	0.448	0.140	(-69)**	[d]	in	HI
Field Sparrow	0.144	().()30	(-79)**	0.478	0.104	(-78)***	Id]	D	D
Savannah Sparrow	0.314	0.096	(-70)**				ID]	|i|	nn
Fox Sparrow				0.181	0.073	(-60)*	Idl
Song Sparrow	1.174	0.589	(-50)*	2.829	1.952	( 31)*	ID]	im	D
Idncoln's Sparrow	0.744	0.418	(-44)	0.314	0.208	( -34)	li|
Swamp Sparrow	2.624	1 .349	(-49)	1 .476	1.447	(-2)	i	i	in
White-throated	17.076	14.091	(-17)	1 3.389	7.580	( 43)**	I)	1)	im
Sparrow White-crowned	0. 1 94	0.098	(-50)	0.337	0.145	( 57)*
Sparrt)w Dark-eyed (Slate-	0.9 1 5	0.379	( 59)**	4.126	1.474	( 64)**^	I)	d	|d|

colored) Junco

6

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

TABLE 1. Continued.
		Spring capture rate“		Fall capture rate^		BBS»’
Species	Early	Late	(% change)	Early	Late	(% change)	SH	nNE	sNE
Northern Cardinal	0.285	0.764	(168)***	0.615	1.444	(135)***	[I]	I	[I]
Ro.se-breasted	0.199	0.046	(-77)*	0.101	0.034	(-66)**	[D]	[i]	[D]
Grosbeak Indigo Bunting	0. 1 25	0.048	(-61)*	0.076	0.056	(-26)	i	d	D
Red-winged	1.219	0.641	(-47)**				[D]	[d]	D
Blackbird Common Crackle	1.412	1.044	(-26)				d	D	D
Brown-headed	0.634	0.259	(-59)**				[D]	D	d
Cowbird
Orchard Oriole	0.170	0.502	(194)**						Id]
Baltimore Oriole	2.671	1.247	(-53)**	1.100	0.676	(-39)*	[D]	[i]	[D]
Purple Finch				1.213	0.168	(-86)***	D	D	[D]
House Finch	0.136	0.1 16	(-15)	0.375	0.249	(-34)	[I]	[I]	[I]
American Gold-	1.175	0.953	(-19)	0.233	0.390	(67)	[d]	[i]	[i]

finch

^ Based on weighted means of capture rates, by year and season (n = 16 in both Early and Late periods). % Change = (Late - Early )/Early X 100. Significant differences between mean Early and Late capture rates (Wilcoxon 2-sample test) indicated by asterisks: * P < 0.05, ** p < 0.01, *** P < 0.001.

Based on Sauer et al. (2001 ) analysis of 1966-2000 BBS data from physiographic strata 28 (SH, eastern Spruce-Hardwoods), 27 (nNE, northern New England), and 12 (sNE, southern New England). D = significant (P < 0.05) decline; d = non-significant (P > 0.05) decline; 1 = significant increase; i = non-significant increase; nc = no change. Symbols in brackets [ J indicate that Sauer et al. (2001) considered these trend estimates unreliable due to “deficiencies” or “important deficiencies” in sampling. Blanks indicate physiographic regions where a given species was not represented in BBS trend data.

(88%) had lower capture rates during Late years of the study than during Early years (Ta- ble 1); these differences were significant (P < 0.05) in 45 species. Nine species had higher capture rates during 1986-2001 than during

Percent change in captures between Early and Late periods (fall)

FIG. 1. Correlations between spring and fall mi- gration periods for percent change in capture rates be- tween Early and Late periods of the study (P < 0.001, ri = 63 species). Three apparent outliers (CARW, Car- olina Wren; ETTI, Tufted Titmouse; and NOCA, Northern Cardinal) shown as solid circles.

1970-1985; in four of these (Carolina Wren, Tufted Titmouse, Black-throated Blue War- bler, and Northern Cardinal), the differences were signihcant (P < 0.05).

Percent changes in mean capture rates from Early to Late years of the study were posi- tively correlated between spring and fall mi- grations (Rho = 0.55, P < 0.001, n = 63 spe- cies; Fig. 1). Exclusion of three outliers (Car- olina Wren, Tufted Titmouse, and Northern Cardinal) that showed dramatic increases in capture rates during both migration periods did not substantially alter the strength of the observed correlation (Rho = 0.48, P < 0.001, n = 60 species). There were no species that showed significant increases in capture rate during one season and significant decreases in the other.

Because of uncertainty regarding the loca- tion of breeding populations represented by migrants in coastal Massachusetts, we com- pared our results with BBS trends from three physiographic regions (southern New Eng- land, northern New England, and eastern Spruce-Hardwoods) that we considered the most likely sources of the majority of mi- grants observed at Manomet (Fig. 2). Captures of spring migrants were significantly (P <

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

7

FIG. 2. Location of Manomet Center for Conser- vation Sciences (MBO) study site relative to three northeastern physiographic strata used in analysis of Breeding Bird Survey data.

0.05) and positively correlated with BBS trends from northern New England; during fall migration, we found significant positive correlations between capture rates and BBS trends from both southern and northern New England physiographic strata (Table 2).

Four species that breed at high latitudes or high elevations [Gray-cheeked/BicknelTs Thrush, Palm Warbler (western), American Tree Sparrow, and White-crowned Sparrow) were represented in the migration monitoring data but not by BBS analyses; all of these spe- cies showed significantly declining capture rates {P < 0.05) between Early and Late pe- riods of the study. Thirty-three species repre- sented in the migration monitoring data were considered by Sauer et al. (2001) to be rep- re.sented unreliably by BBS data in any of the northeastern physiographic strata (Table 1); 19 of these species (Philadelphia Vireo, Brown Creeper, House Wren, Blue-gray Gnatcatcher, vSwainson’s Thrush, Northern Mockingbird, Tennessee Warbler, Northern Parula, Cape May Warbler, Bay-breasted Warbler, Black- poll Warbler, Northern Waterthrush, Mourning Warbler, Wilson's Warbler, Yellow-breasted Chat, Savannah Sparrow, Fox Sparrow, Rose- breasted Grosbeak, and Baltimore Oriole) showed significant declines at Manomet dur- ing at least one of the two migration periods.

TABLE 2. Correlations between percent change in mean capture rates (Early versus Late sampling peri- ods) and Breeding Bird Survey trends (Sauer et al. 2001) from three physiographic regions. BBS results with “deficiencies” or “important deficiencies” have been excluded from analysis (see text).

	Physiographic region'
sNE	nNE	SH
Spring	0.36 (0.087)*’	0.45 (0.011)	0.17 (0.402)
	n = 23	n = 3\	n = 26
Fall	0.50 (0.018)	0.47 (0.006)	0.34 (0.087)
	n = 22	n = 33	n = 26

® sNE = southern New England, nNE = northern New England, SH = eastern Spruce-Hardwoods.

^ Spearman rank correlation (P-value).

while capture rates of 3 (Tufted Titmouse, Carolina Wren, and Orchard Oriole) signifi- eantly increased during fall and spring migra- tions (Table 1).

Apparent inconsistencies between trends based on migration captures at Manomet and BBS data were greatest for the eastern Spruce-Hardwoods stratum and least for the southern New England stratum. This pattern was true during both spring (Fig. 3) and fall (Fig. 4) migration periods. Spring migration captures indicated significant {P < 0.05) de- clines in three species for which BBS analyses found significant increases: Red-eyed Vireo (eastern Spruce-Hardwoods), Ovenbird (northern New England), and Gray Catbird (southern New England). Fall migration cap- tures significantly declined in four species whereas BBS analyses showed significant in- creases: Red-eyed Vireo and Yellow-rumped (Myrtle) Warbler (eastern Spruce-Hard- woods), Yellow-bellied Flycatcher and Yel- low-rumped (Myrtle) Warbler (northern New England), and Gray Catbird (southern Nev\ England).

For each migration period, cluster analysis was used to identify an arbitrary six groups of species that shared general patterns of change in capture rates across years (Figs. 5 aiul 6). This approach allowed us to summari/e trend data visually for a large number of species. However, we note that similarities in capture rates among members of a group do not nec- essarily mean that sharetl trends were caused by similar proximate factors. In some cases cluster membership may, in fact, rcllcct the inlUicnce of sharetl ecology, f or example.

8

THE Wll.SON BUIXETIN • Vol. 1 16, No. I, March 2004

c

Q>

O

Q.

(/)

■o

C

<u

DECLINE- Decline INCREASE- Increase

significant (n=5) significant (n=7)

(n=8) (n=6)

DECLINE- Decline INCREASE- Increase

significant (n=8) significant (n=4)

(n=10) (n=9)

(n=13) (n=2)

BBS trends (percent of species)

■ DECLINE-significant S INCREASE-significant □ Decline □ Increase

c

<u

o

k.

0)

(f)

•o

c

<D

significant (n=4) significant (n=7) (n=8) (n=7)

DECLINE- Decline INCREASE-

significant (n=8) significant

(n=10) (n=11)

■LB

Increase

(n=4)

DECLINE- Decline INCREASE- Increase

significant (n=5) significant (n=3)

(n=11) (n=2)

BBS trends (percent of species)

■ DECLINE-significant 0 INCREASE-significant □ Decline □ Increase

FIG. 3. Comparison of trends in capture rate based on spring migration monitoring at Manomet relative to trends derived from BBS data (Sauer et al. 2001) in (A) spruce-hardwoods, (B) northern New England, and (C) southern New England physiographic strata. “DE- CLINE-significant,” P < 0.05; “Decline,” P > 0.05; “INCREASE-significant,” P < 0.05; “Increase,” P > 0.05. For example, of 13 species showing significant declines according to BBS data from southern New England, 70% showed significant declines in Manomet capture rates, and 20% showed declines in Manomet capture rates that were not statistically significant.

capture rates of Blackpoll Warbler, Northern Parula, Tennessee Warbler, Cape May War- bler, Blackburnian Warbler, and Bay-breasted Warbler peaked during the mid to late 1970s

FIG. 4. Comparison of trends in capture rate based on fall migration monitoring at Manomet relative to trends derived from BBS data (Sauer et al. 2001) in (A) spruce-hardwoods, (B) northern New England, and (C) southern New England physiographic strata. “DECLINE- significant,” P < 0.05; “Decline,” P > 0.05; “IN- CREASE-significant,” P < 0.05; “Increase,” P > 0.05.

(Fig. 6F); many, if not all, of these species likely responded to a widespread outbreak of spruce bud worm (Choristoneura fume rif ana Clem.) in eastern North America during this time period (Hagan et al. 1992). Carolina Wren and Northern Cardinal, two species known to have shown dramatic regional pop- ulation increases during the last decades (Ha-

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

9

cc

CD

:>^

CO

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

AMRO, BLPW, BTNW, CAWA, GOTH, INBU, LISP. T NAWA. NOPA

I

-II -r^ I

NOWA, SWTH, VEER, WCSP, WEVI, WIWA, WOTH

n 1 r

1970 1975 1980 1985 1990 1995 2000

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

- D ACFL, AMGO, AMRE, BAWW,

^ BGGN, BLJA, BTBW, COGR,

^ COYE, EAKI

GRCA, MAWA, MOWA, NOMO,

- OVEN, PRAW, TRFL, WTSP,

YBFL, YWAR

1.0-	D T		BLBW, BRCR, EAPH, HETH, MYWA, RCKI,
0.8-		]	r scju, sosp, sovi, Lf SWSP
0.6-		■l ,
0.4- 0.2-			j Tirii^n
0.0-

1970 1975 1980 1985 1990 1995 2000

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

II I

BAOR, BBWA, BHCO, BRTH, CHSP, OSWA, EAWP, FISP, HOWR, LEFL, REVI, RSTO, RWBL, SAVS, TEWA

'-mnnj

In

Midpoint (3-year moving average)

FIG. 5. Major patterns of change in spring capture rates of 72 species in coa.stal Massachusetts, 197()-2()01 . Error bars represent ± 1 SE. Species contributing to each plot are indicated with four-letter banding codes; .see Appendix.

gan el al. 1992), were grouped together during both spring (Fig. 5E) and fall (Fig. 6C) mi- grations.

We speculate that at least some of the clus- tering results (and, therefore, underlying trend patterns) may rellect local weather conditions that would have influenced capture rates of species with similar migration periods. There were significant differences among mean mi- gration dates for each of the six clusters (Fig. 7; Wilcoxon rank sum test; spring: y- = 19.34, df = 5, r = ().()()2; fall: y- = 16-12, df = 5, P = ().()()7). During spring, most species as- signed to clusters A and I) (Fig. 5 A, I)) were relatively early migrants, with mean migration dates of 7 May (SFI = 4.5 days) and 3 May (SF = 3.4 days), respectively; both of these groups showed somewhat elevated capture rates during the mid to late l9S0s, possibly

suggesting that during several years in this time period weather conditions caused larger- than-normal numbers of these species to be present in coastal Massachusetts. Similarly, most species assigned to fall cluster A (F^4g. 6A) were relatively late migrants, with a mean migration date of 9 October (SF = 3.3 days); the relatively high capture rates that charac- terized this group during the early 1970s may have reflected local weather conditions that alTected any sjiecies with a peak migration pe- riod in early October.

Nonetheless, we hesitate to try and pro\ ide further explanations for the species “member- ships" in each of these groupings, histeail. we ITiefer to emphasize a more general perspec- tive. noting that only one of the six trend plots from each migration perioil (spring: big. 5F; fall: l ig. b(') showetl obvious increases in

10

THE WILSON BULLETIN • Vol. ! 16, No. 1, March 2004

1.0

0.8

0.5

INBU, GCKI, SCJU, WPWA, RSTO, ATSP, FISP, BRTH, BRCR, WIWR, NAWA, WTSP, LISP

1.0-

0.8

0.6-

0.4-

p BTBW, SOVI, EAPH, SWSP, HETH, ^ RCKl, SOSP, PRAW, BWWA,

^ OVEN, EAWP, BGGN

III,

g, 0.2 2

g 0.0

1970 1975 1980 1985 1990 1995 2000

CEDW, OCWA, COYE, MOWA,

0.2 - HOWR, TRFL, REVI, PHVI, MAWA,

GRCA, BTNW, CAW A, CSWA

0.0-

1970 1975 1980 1985 1990 1995 2000

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

- p WOTH, YBCH, MYWA, YBFL, BLJA, _ ^ GCTH, WCSP, FOSP, SWTH, LEFL,

AMRO, SCTA, RBNU, BCCH, PUFI

VEER, WEVI, CONW,

- YWAR, WIWA, AMRE,

NOWA, BAWW

1.0 0.8 0.6 0.4 0.2 0.0

1970 1975 1980 1985 1990 1995 2000

n	WBNU, ETTl, BAOR,
LJ	IT NOMO, WAVI, RBGR,
	-I I TT	EAKI, HOFI
ji’	' A

1.0

0.8-1

0.6

0.4-1

0.2

BLPW, NOPA, TEWA, CMWA, BLBW, BBWA

li

0.0

1970 1975 1980 1985 1990 1995 2000

Midpoint (3-year moving average)

LIG. 6. Major patterns of change in fall capture rates of 78 species in coastal Massachusetts, 1970-2001. Error bars represent ± 1 SE. Species contributing to each plot are indicated with four-letter banding codes; see Appendix.

capture rates. Four of the plots from each mi- gration period (spring: Fig. 5B-D, F; fall: Fig. 6A-B, E— F) showed decreasing trends in cap- ture rates. One plot from each migration pe- riod was characterized by peak capture rates during the early to mid 1980s, with compa- rably low rates before and after this time pe- riod (spring: Fig. 5A; fall: Fig. 6D).

DISCUSSION

The Breeding Bird Survey is widely rec- ognized as a primary source of information regarding conservation priorities for North American birds (Geissler and Noon 1981, Butcher et al. 1993, Smith et al. 1993, James et al. 1996, Carter et al. 2000), yet relatively few studies have attempted to validate its con- clusions via independent, alternative monitor- ing schemes. Hussell et al. (1992) compared

a migration index from 1961 to 1988 at Long Point, Ontario with BBS trends in that prov- ince and obtained positive correlations, as did Francis and Hussell (1998) in Ontario. Other multiple-year comparisons with BBS data have included intensive counts in Quebec (Jobin et al. 1996) and migration monitoring at Southeast Farallon Island, California (Pyle et al. 1994) and at Point Reyes, California (Ballard et al. 2003). In this paper we present results from a long-term study based on stan- dardized mist-net capture efforts during fall and spring migrations in coastal Massachu- setts, and compare these data with estimates of population trends obtained by Sauer et al. (2001) in their analysis of BBS data.

At first glance it would appear that there is good agreement between our results and BBS analyses. There were strong correlations be-

Lloyd-Evans and Atw’ood • 32 YEARS OF PASSERINE MIGRATION COUNTS

11

FIG. 7. Mean migration dates during spring and fall for clusters derived from capture trends. Cluster letters correspond with those shown in Fig. 5 (spring) and Fig. 6 (fall). Error bars represent ± 1 SE.

tween population trends observed in each of the three BBS strata considered here and changes in Manomet capture rates between 1970-1985 and 1986-2001, suggesting that both methods do, in fact, reflect changes in regional breeding populations. For example. Least Flycatcher was the only species to de- cline significantly in all three northeastern BBS strata, and it showed a significant decline in capture rate during fall at Manomet. Of 10 species for which significant declines were noted in two of three northeastern BBS strata, we found significant declines in capture rates during at least one of the two migration sea- sons for 7 (Eastern Kingbird, Wood Thrush, Common Yellowthroat, Eastern Towhee, Field Sparrow, White-throated Sparrow, and Purple Finch); 2 of the other species (Eastern Wood- Pewee and Common Crackle) declined non- significantly at Manomet, while Great Crested Flycatcher showed a non-significant increase based on migration data. Of 2.3 species for which the BBS showed significant population declines in at least one of the three physio- graphic strata considered here, 18 (789f ) also showed significant declines in capture rates during spring and/or fall migration.

Yet the situation is more complex than these comparisons might suggest. In many cases our study failed to detect increasing population trends indicated by the BBS. Of 16 species shown by Sauer et al. (2001) to have had sig- nificant increases in at least one of the phys- iographic strata considered here, we found sig- nificantly increased capture rates in only 1 (Northern Cardinal). Furthermore, we observed significant declines in capture rates during spring and/or fall migration for five species found by the BBS to be exhibiting significant population increases in at least one of the three physiographic strata [Yellow-bellied Flycatch- er, Red-eyed Vireo, Gray Catbird, Yellow-rum- ped (Myrtle) Warbler, and Ovenbird].

In our study we found significantly declin- ing capture rates during one or both migration periods in 54 of 87 species (62%), but only 5 species (6%) showed significant increases. Among the 37 of these species for which re- liable BBS results were available from at least one of the northeast’s physiographic strata, Sauer et al. (2001) found significant declines in 22 cases (59%) and significant increases in 15 (41%); Great Crested Flycatcher and Gray Catbird showed opposite significant trends in different physiographic strata. These contrasts suggest that factors in addition to changes in breeding populations may be confounding the relationship with capture rates observed dur- ing migration.

We especially note that the patterns we de- scribe here could have emerged if captures of most species we sampled during migration were somehow being reduced, over time, by factors unrelated to actual changes in breeding populations. For example, long-term changes in climate conceivably could cause shifts in regional weather patterns that, in turn, might systematically affect the number of migrants appearing in coastal Massachusetts (Moore et al. 1993). However, we are not aware of any evidence of long-term increases in migration captures at established banding operations east of the Mississippi that might be expected if actual migration patterns were changing. Or. as the vegetation at Manomet has matured since 1970, some species of migrants may now move through the study area at heights where they simply avoid making contact with the nets (2.6 m in height) (Remsen and Good 1996); species that woukl continue to be ac-

12

THB WILSON BULLETIN • VoL 1 16, No. /, March 2004

tive primarily within 3 m of the ground, even in the presenee of higher eanopy cover, might be avoiding the site because of its generally more forested aspect (Moore et al. 1993).

Conversely, the BBS results may them- selves be subject to error due to the effects of roadside bias (Temple and Wiens 1989, Keller and Fuller 1995) or short count period (Welsh 1995, Jobin et al. 1996); thus; the trend esti- mates by Sauer et al. (2001) may not neces- sarily provide a “gold standard” by which to validate Manomet’s migration count results. It is also quite possible that a species could be increasing in one BBS stratum and decreasing in another, or showing conflicting trends with- in different regions of a single stratum — any of which could confuse the relationship be- tween trends shown by the BBS and migration monitoring data sets. One of the three BBS strata considered here, the eastern Spruce- Hardwood forest, is so large (353,538 km^; Rosenberg and Hodgman 2()()0) that presen- tation of a single trend to represent this entire area seems fraught with uncertainty at least equal to our lack of knowledge about the de- tailed breeding locations of migrants passing through Manomet.

At this point we have no way of further assessing these possible explanations. Certain- ly capture rates of migrants at Manomet dur- ing spring and fall have, in many cases, changed substantially from 1970 to 2001, and the vast majority of these changes have been declines. Migration count data from other studies also indicate long-term declines in New England birds; for example. Hill and Ha- gan (1991) found that spring surveys of 26 Neotropical migrants in Middlesex and Essex counties of Massachusetts declined, on aver- age, nearly 1% per year from 1954 to 1987. Personal comments from several banders fa- miliar with the location for 30+ years all in- dicate that there are fewer birds in recent years at Manomet and in New England generally.

Many of the declines documented at Man- omet coincide with declines in breeding pop- ulations reported by the most reliable BBS data. Nonetheless, there are some apparent in- consistencies between results of the two anal- yses that we cannot explain. It appears likely that a combination of factors have influenced the number of migrants captured at Manomet since 1970. We believe, however, that the pre-

ponderance of data suggests long-term popu- lation declines in a wide variety of both Neo- tropical and shorter-distance migrants that greatly exceed the few increases that have been observed.

ACKNOWLEDGMENTS

It is impossible for us to name all of the contributors to this project, many of whom have given their time faithfully since the late 1960s. Hosts of students and volunteers have foregone sleep and decent salaries in order to spend their days walking net lanes. The trust- ees and friends of Manomet Center for Conservation Sciences made this work possible through unfailing personal and financial assistance. We deeply appreciate the support that all of you have given; from Cranberry Hill to Stage Point, your enthusiasm and dedication will always endure. Thank you. C. J. Ralph, C. S. Rob- bins, and an anonymous reviewer provided helpful comments on a preliminary draft of the manuscript. We dedicate this paper to K. Anderson and those initial banders whose vision and passion gave birth to Man- omet Bird Observatory.

LITERATURE CITED

Askins, R. a., J. F. Lynch, and R. Greenberg. 1990. Population declines in migratory birds in eastern North America. Current Ornithology 7:1-57. Ballard, G., G. R. Geupel, N. Nur, and T. Gardali. 2003. Long-term declines and decadel patterns in population trends of songbirds in western North America, 1979-1999. Condor 105:737-755. Butcher, G. S., B. Peterjohn, and C. J. Ralph. 1993. Overview of national bird population monitoring programs and databases. Pages 192-203 in Status and management of Neotropical migratory birds (D. M. Finch and P. W. Stangel, Eds.). General Technical Report RM-229. U.S. Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado.

Carter, M. F, W. C. Hunter, D. N. Pashley, and K. V. Rosenberg. 2000. Setting conservation priori- ties for landbirds in the United States: the Partners in Flight approach. Auk 177:541-548.

Dunn, E. H. and D. J. T. Hussell. 1995. Using mi- gration counts to monitor landbird populations: re- view and evaluation of current status. Current Or- nithology 12:43-48.

Francis, C. M. and D. J. T. Hussell. 1998. Changes in numbers of landbirds counted in migration at Long Point Bird Ob.servatory, 1961-1997. Bird Populations 4:37-66.

Gauthreaux, S. a., Jr. 1971. A radar and direct visual study of passerine spring migration in southern Louisiana. Auk 88:343-365.

Gauthreaux, S. A., Jr. 1992. The use of weather radar to monitor long-term patterns of trans-Gulf migra- tion in spring. Pages 96-100 in Ecology and con- servation of Neotropical migrant landbirds (J. M.

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

13

Hagan, III, and D. W. Johnson, Eds.). Smithsonian Institution Press, Washington, D.C.

Geissler, P. H. and B. R. Noon. 1981. Estimates of avian population trends from the North American Breeding Bird Survey. Studies in Avian Biology 6:42-5 1 .

Hagan, J. M., T. L. Lloyd-Evans, J. L. Atwood, and D. S. Wood. 1992. Long-term changes in migra- tory landbirds in the northeastern United States: evidence from migration capture data. Pages 1 15- 130 in Ecology and conservation of Neotropical migrant landbirds (J. M. Hagan, III, and D. W. Johnson, Eds.). Smithsonian Institution Press, Washington, D.C.

Hill, N. P. and J. M. Hagan, III. 1991. Population trends of some northeastern North American land- birds: a half-century of data. Wilson Bulletin 103: 165-182.

i Hussell, D. j. T. 1981. The use of migration counts for monitoring bird population levels. Studies in Avian Biology 6:92-102.

Hussell, D. J. T, M. H. Mather, and P. H. Sinclair. 1992. Trends in numbers of tropical- and temper- ate-wintering migrant landbirds in migration at Long Point, Ontario, 1961-1988. Pages 101-114 in Ecology and conservation of Neotropical mi- grant landbirds (J. M. Hagan, III, and D. W. John- son, Eds.). Smithsonian Institution Press, Wash- ington, D.C.

\ James, E C., C. E. McCulloch, and D. A. Wiedenfeld. 1996. New approaches to the analysis of population trends in land birds. Ecology 77:13-27.

! JoBiN, B., J.-L. DesGranges, and C. Boutin. 1996.

! Comparison of BBS and intensive surveys at se- , lected BBS stops. Bird Populations 3:14-25.

! Kf-:ller, C. M. E. and M. R. Fuller. 1995. Compar- j ison of birds detected from roadside and off-road I point counts in the Shenandoah National Park. Pages 111-116/7? Monitoring bird populations by point counts (C. J. Ralph, J. R. Sauer, and S. Droe- ge, Eds.). General Technical Report PSW-149. U.S. Forest Service, Pacific Southwest Research Station, Albany, California.

Moori:, E R., S. a. Gauthreaux, Jr., P. Kerlinger, AND T. R. Simons. 1993. Stopover habitat: man- agement implications and guidelines. Pages 58- 69 in Status and management of Neotropical mi- gratory birds (D. M. Pinch and P W. Stangel, Eds.). General Technical Report RM-229. U.S. Forest Service, Rocky Mountain Forest and Range Experiment Station. Fort Collins. Colorado.

Moore, F: R. and T. R. Simons. 1992. Habitat suit- ability and stopover ecology of Neotropical land- bird migrants. Pages 345-355 in F'cology and con- servation of Neotropical migrant landbirds (J. M. Hagan, III, and D. W. Johnson. Fkis.). Smithsonian Institution Press, Washington. D.C’.

Pi ACH, W. J.. S. R. Baii 1 II., and D. F:. Bai mi k. 1998. Long-term changes in the abundance of passerines

in Britain and Ireland as measured by constant effort mist-netting. Bird Study 45:257-275.

Pyle, R, N. Nur, and D. F. DeSante. 1994. Trends in nocturnal migrant landbird populations at South- east Farallon Island, California, 1968-1992. Stud- ies in Avian Biology 15:58-74.

Remsen, j. V., Jr., and D. A. Good. 1996. Misuse of data from mist-net captures to assess relative abundance in bird populations. Auk 1 13:381-398.

Robbins, C. S. 1968. Net hours: a common denomi- nator for the study of bird populations. Eastern Bird-Banding Association News 31:31-35.

Robbins, C. S., D. Bystrak, and P. H. Geissler. 1986. The Breeding Bird Survey: its first fifteen years, 1965-1979. U.S. Fish and Wildlife Service Re- search Publication 157. Washington, D.C.

Robbins, C. S., J. R. Sauer, R. S. Greenberg, and S. Droege. 1989. Population declines in North American birds that migrate to the Neotropics. Proceedings of the National Academy Sciences 86:7658-7662.

Rosenberg, K. V. and T. P. Hodgman. 2()()0. Partners in Flight bird conservation plan for eastern spruce-hardwood forest (physiographic area 28), draft 1.0. Online at <http://www.blm.gov/wildlife/ plan/pL28_10.pdf> (accessed 5 May 2003).

SAS Institute, Inc. 2001. JMP Start Statistics, 2nd ed. Duxbury-Thomson Learning, Pacific Grove, Cal- ifornia.

Sauer, J. R. 1993. Monitoring goals and programs of the U.S. Fish and Wildlife Service. Pages 245- 251 in Status and management of Neotropical mi- gratory birds (D. M. Finch and P. W. Stangel, Eds.). General Technical Report RM-229. U.S. Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado.

Sauer, J. R., J. E. Hines, and J. Fallon. 2001. The North American Breeding Bird Survey, results and analysis I966-2()0(), ver. 2001.2. USGS Pa- tuxent Wildlife Research Center, Laurel, Mary- land. Online at <http://www.mbr-pwrc.usgs.gov/ bbs/bbs.html> (accessed 24 June 2001).

Smith, C. R., D. M. Pi:nce, and R. J. O'Connor. 1993. Status of Neotropical migratory birds in the Northeast: a preliminary assessment. Pages 172- 188 in Status and management of Neotropical mi- gratory birds (D. M. Finch and P. W. Stangel. Fkls.). General Technical Report RM-229. U.S. Forest .Service, Rocky Mountain Forest and Range Experiment .Station, Fort C’tfilins, C'olora(.lo.

Ti:mim.e, S. a. and.). A. Wii:ns. 1989. Bird populations and environmental changes: can birds be bio-in- dicators? American Birds 43:260 270.

Ti.rborgfi, j. 1989. Where have all the birds gone? Princeton University Press. Princeton. New Jersey.

Wi;i..SH. D. A. 1995. An overview of the Ontario Forest Bird Monitoring Program in ( anada. Pages 93 97 in Monitoring bird populations by point eounts (C. .1. Ralph. .1. K. .Sauer, aiul .S. Droege. Eds.). General rechnieal Report PSW-149. U.S. Forest Serviee. Pacilie Southwest Research Station. Al bany. California.

14

THE WILSON BULLETIN • Vol. 1 16, No. J, March 2004

APPENDIX. Banding codes, scientific names, and migration periods of species referred to in text. Lor each season, the limits of sampling window (P' and 99'^^ percentiles) are given in parentheses following the mean date of migration (all years combined). Dashes ( — ) indicate species-season combinations (such as fall Acadian Llycatcher) that failed to meet analysis criteria described in Methods.

Code	Common name	Scientific name		Spring
EAWP	Eastern Wood-	C out opus virens	31	May (13 May-14 Jun)	10
	Pewee
YBLL	Yellow-bellied Ply-	Empidonax flaviven-	02	Jun (22 May-15 Jun)	06
	catcher	tris
ACLL	Acadian Flycatcher	Empidonax virescens	31	May (13 May-15 Jun)	—
TRLL	Willow/Alder Fly-	Empidonax traillii &	02	Jun (19 May-15 Jun)	02
	catcher	E. alnorurn
LELL	Least Flycatcher	Empidonax minimus	21	May (05 May-1 1 Jun)	04
EAPH	Eastern Phoebe	Sayornis phoebe	25	Apr (15 Apr-05 Jun)	21
GCEL	Great Crested Fly-	Myiarchus crinitus	06	Jun (12 May-15 Jun)	—
	catcher
EAKI	Eastern Kingbird	Tyrannus tyrannus	25	May (10 May- 15 Jun)	25
WEVI	White-eyed Vireo	Vireo griseus	21	May (29 Apr-15 Jun)	15
SOVI	Blue-headed Vireo	Vireo solitarius	10	May (26 Apr-31 May)	05
WAVI	Warbling Vireo	Vireo gilvus	—		1 1
PHVI	Philadelphia Vireo	Vireo philadelphicus	—		16
REVI	Red-eyed Vireo	Vireo olivaceus	30	May (14 May-13 Jun)	20
BLJA	Blue Jay	Cyanocitta cristata	15	May (20 Apr-1 1 Jun)	30
BCCH	Black-capped	Poecile atricapillus	08	May (16 Apr-08 Jun)	14
	Chickadee
ETTI	Tufted Titmouse	Baeolophus bicolour	28	Apr (15 Apr-09 Jun)	12
RBNU	Red-breasted Nut-	Sitta canadensis	—		23
	hatch
WBNU	White-breasted Nut-	Sitta carolinensis	—		07
	hatch
BRCR	Brown Creeper	Certhia americana	25	Apr (15 Apr-07 Jun)	09
CARW	Carolina Wren	Thryothorus liidovi-	16	May (15 Apr-14 Jun)	06
		cianus
HOWR	House Wren	Troglodytes aedon	15	May (26 Apr- 13 Jun)	12
WIWR	Winter Wren	Troglodytes troglo-	—		11
		dytes
GCKI	Golden-crowned	Regulus satrapa	22	Apr (15 Apr-06 May)	15
	Kinglet
RCKI	Ruby-crowned	Regains calendula	29	Apr (17 Apr- 17 May)	13
	Kinglet
BGGN	Blue-gray Gnat-	Polioptila caerulea	01	May (17 Apr- 19 May)	09
	catcher
VEER	Veery	Catliarus fuscescens	20	May (05 May-08 Jun)	1 1
GCTH	Gray-cheeked/B ick-	Catharus minimus <&	27	May (14 May- 12 Jun)	01
	nell’s Thrush	C. bicknelli
SWTH	Swainson’s Thrush	Catharus ustulatus	26	May (12 May-10 Jun)	24
HETH	Hermit Thrush	Catharus guttatus	29	Apr (16 Apr- 19 May)	20
WOTH	Wood Thrush	Hylocichla mustelina	16	May (04 May-06 Jun)	18
AMRO	American Robin	Turdus migratorius	02	May (15 Apr- 13 Jun)	26
GRCA	Gray Catbird	Dumetella carolinen-	19	May (03 May-12 Jun)	09
NOMO	Northern Mocking-	sis Mimus polyglottos	08	May (17 Apr-07 Jun)	13
	bird
BRTH	Brown Thrasher	Toxostoma rufuni	10	May (20 Apr-05 Jun)	25
CEDW	Cedar Waxwing	Bombycilla cedrorum	26	May (21 Apr- 15 Jun)	02
BWWA	Blue-winged War-	Vermivora pinus	—		03
	bler
TEWA	Tennessee Warbler	Vermivora peregrina	23	May (13 May-03 Jun)	20
OCWA	Orange-crowned	Vermivora celata	—		15

Warbler

Fall

Sep (16 Aug- 10 Oct) Sep (17 Aug-27 Sep)

Sep (16 Aug-30 Sep)

Sep ( 1 7 Aug-05 Oct) Sep (16 Aug-25 Oct)

Aug ( 1 5 Aug-20 Sep) Sep (15 Aug-25 Oct) Oct (10 Sep-29 Oct) Sep (17 Aug-07 Oct) Sep (23 Aug-21 Oct) Sep (22 Aug-25 Oct) Sep (16 Aug-09 Nov) Oct (23 Aug-1 1 Nov)

Oct (31 Aug- 10 Nov) Sep (18 Aug-02 Nov)

Oct (17 Aug- 14 Nov)

Oct ( 1 1 Sep-04 Nov) Sep (15 Aug-05 Nov)

Sep (17 Aug-22 Oct) Oct (18 Sep- 10 Nov)

Oct (23 Sep- 12 Nov)

Oct (18 Sep- 11 Nov)

Sep (16 Aug-03 Nov)

Sep (20 Aug-10 Oct) Oct (13 Sep-03 Nov)

Sep (30 Aug-22 Oct) Oct (26 Sep- 14 Nov) Sep (18 Aug-26 Oct) Sep (16 Aug- 12 Nov) Sep (15 Aug-18 Oct)

Sep (16 Aug-12 Nov)

Sep (15 Aug-31 Oct) Oct (17 Aug- 10 Nov) Sep (16 Aug-24 Oct)

Sep (19 Aug-28 Oct) Oct (25 Sep- 14 Nov)

Lloyd-Evans and Atwood • 32 YEARS OF PASSERINE MIGRATION COUNTS

15

APPENDIX. Continued.
Code	Common name	Scientific name		Spring		Fall
NAWA	Nashville Warbler	Vermivora rufica- pilla	16	May (30 Apr-10 Jun)	23	Sep (17 Aug-31 Oct)
NOPA	Northern Parula	Parula arnericana	19	May (02 May-09 Jun)	29	Sep (25 Aug-30 Oct)
YWAR	Yellow Warbler	Dendroica petechia	21	May (05 May- 10 Jun)	29	Aug (15 Aug-02 Oct)
CSWA	Chestnut-sided War- bler	Dendroica pensyl- vanica	22	May (03 May-12 Jun)	06	Sep (17 Aug-22 Oct)
MAWA	Magnolia Warbler	Dendroica magnolia	24	May (10 May- 10 Jun)	18	Sep (25 Aug-22 Oct)
CMWA	Cape May Warbler	Dendroica tigrina	—		05	Sep (16 Aug- 13 Oct)
BTBW	Black-throated Blue Warbler	Dendroica caerules- cens	18	May (05 May-04 Jun)	25	Sep (23 Aug-25 Oct)
MYWA	Yellow-rumped (Myrtle) Warbler	Dendroica c. coron- ata	06	May (16 Apr-26 May)	18	Oct (24 Sep- 15 Nov)
BTNW	Black-throated Green Warbler	Dendroica virens	22	May (03 May-13 Jun)	22	Sep (21 Aug-31 Oct)
BLBW	Blackburnian War- bler	Dendroica fusca	26	May (13 May- 10 Jun)	09	Sep (21 Aug-19 Oct)
PRAW	Prairie Warbler	Dendroica discolor	13	May (26 Apr-04 Jun)	06	Sep (16 Aug-21 Oct)
WPWA	Palm Warbler (west- ern)	Dendroica p. palma- rum	—		06	Oct (08 Sep- 12 Nov)
YPWA	Palm Warbler (yel- low)	Dendroica p. hy- pochrysea	28	Apr (16 Apr- 14 May)	—
BBWA	Bay-breasted War- bler	Dendroica castanea	23	May (13 May-07 Jun)	04	Sep (17 Aug-10 Oct)
BLPW	Blackpoll Warbler	Dendroica striata	28	May (12 May- 15 Jun)	26	Sep (03 Sep-29 Oct)
BAWW	Black-and-White Warbler	Mniotilta varia	15	May (30 Apr-05 Jun)	07	Sep (15 Aug- 18 Oct)
AMRE	American Redstart	Setophaga rut ic ilia	28	May (12 May- 13 Jun)	09	Sep (16 Aug- 13 Oct)
OVEN	Ovenbird	Seiurus aurocapilla	19	May (03 May-05 Jun)	08	Sep ( 16 Aug-24 Oct)
NOWA	Northern Water- thrush	Seiurus novebora- censis	19	May (03 May-05 Jun)	07	Sep (16 Aug-17 Oct)
CONW	Connecticut Warbler	Oporornis agilis	—		19	Sep (31 Aug- 16 Oct)
MOWA	Mourning Warbler	Oporornis Philadel- phia	03	Jun (21 May-15 Jun)	09	Sep ( 15 Aug- 17 Oct)
COYE	Common Yellow- throat	Geothlypis trichas	22	May (06 May-10 Jun)	1 1	Sep (16 Aug-27 Oct)
WIWA	Wilson’s Warbler	Wilsonia pusilla	23	May (1 1 May-08 Jun)	I 1	Sep (21 Aug-20 Oct)
CAW A	Canada Warbler	Wi Ison ia canadensis	28	May (13 May-1 1 Jun)	01	Sep ( 16 Aug-28 Sep)
YBCH	Yellow-breasted Chat	Icteria virens	—		19	Sep (21 Aug-06 Nov)
SCTA	Scarlet Tanager	Pi ranga oli vacea	—		13	Sep ( 1 6 Aug-2 1 Oct)
RSTO	Eastern Towhee	Pipilo erythrophthal-	08	May (20 Apr-05 Jun)	27	Sep ( 16 Aug-05 Nov)
ATSP	American Tree Sparrow	nius Spizella arhorea	—		05	Nov ( 16 Oct- 16 Nov)
CHSP	Chipping Sparrow	Spizella passe rina	09	May (21 Apr-03 Jun)	—
FISP	Field Sparrow	Spizella pusilla	07	May ( 19 Apr-1 2 Jun)	21	Oct (02 Sep- 14 Nov)
SAVS	Savannah Sparrow	l\isse rcul us sand- wichensis	07	May (16 Apr-31 May)	—
FOSP	Fox Sparrow	Ihisserella iliaca	—		29	Oct (08 Oct 14 Nov)
SOSP	Song Sparrow	Melospiza melodia	25	Apr ( 1 5 Apr-09 Jun)	29	Sep (16 Aug-0^) Nov)
LISP	Lincoln's Sparrow	Melospiza lincolnii	22	May (05 May -09 Jun)	01	Oct (03 Sep 29 Oct)
SWSP	Swamp Sparrow	Melospiza georgiana	1 1	May ( 1 7 Apr-04 Jun)	12	Oct ( 16 Sep 06 Nov)
WTSP	White-throated Sparrow	Zonal richia (dhicol- lis	04	May (18 Apr-22 May)	10	Oct (13 Sep 12 Nov)
WCSP	White-crowned Sparrow	Zonotrichia leuco- f}hrys	14	May (30 Apr 26 May)	12	Oct (20 Sep 31 Oct)
SC I LI	Dark-eyed (Slate- colored) .lunco	.lunco h. hyenudis	21	Apr ( 1 5 Apr 1 7 May)	18	( )ct ( 1 4 Sep 1 4 No\ )

16

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

APPENDIX. Continued.

Code	Common name	Scientific name			Spring	Fall
NOCA	Northern Cardinal	Cardinal is cardinalis	04	May	(15	Apr-12 Jun)	03 Oct (16 Aug- 12 Nov)
RBGR	Rose-breasted Gros- beak	Pheucticiis ludovici- anus	18	May	(26	Apr-05 Jun)	12 Sep (18 Aug-24 Oct)
INBU	Indigo Bunting	Passerina cyanea	25	May	(25	Apr-14 Jun)	30 Sep ( 1 9 Aug-3 1 Oct)
RWBL	Red-winged Black- bird	Agelaius phoeniceus	1 1	May	(18	Apr-12 Jun)	—
COGR	Common Crackle	Quiscalus cjuiscula	09	May	(18	Apr- 13 Jun)	—
BHCO	Brown-headed Cow- bird	Molothrus ater	03	May	(15	Apr-13 Jun)	—
OROR	Orchard Oriole	Icterus spurius	18	May	(10	May-03 Jun)	—
BAOR PULI	Baltimore Oriole Purple Pinch	Icterus gaihula Carpodacus purpu-	20	May	(09	May-14 Jun)	28 Aug (15 Aug-09 Oct) 03 Oct (21 Aug-05 Nov)
HOLI	House Pinch	Carpodacus mexi- canus	08	May	(15	Apr- 14 Jun)	12 Sep (16 Aug-16 Nov)
AMGO	American Goldfinch	Carduelis tristis	19	May	(18	Apr- 15 Jun)	25 Oct (20 Aug- 15 Nov)

Wilson Bulletin, 116(1), 2004, pp. 17-22

SUNRISE NEST ATTENDANCE AND AGGRESSION BY LEAST BELL’S VIREOS FAIL TO DETER COWBIRD PARASITISM

BRYAN L. SHARP' ’ AND BARBARA E. KUS^-*

ABSTRACT. — We video-recorded three, natural, brood-parasitism events by Brown-headed Cowbirds (Mol- othrus ater) at nests of Least Bell’s Vireos (Vireo bellii pnsillns). All instances occurred near dawn, during both egg-laying and incubation stages of the nesting cycle. In each case, an adult vireo was on the nest when the female cowbird arrived. Both members of each parasitized pair vigorously attacked the intruding cowbird, but in no encounter did a pair of vireos successfully defend its nest from parasitism. Thus, Least Bell’s Vireos in our study were unable to prevent a female cowbird from parasitizing their nests once the cowbird had reached the nest. Received 16 October 2003, accepted 1 April 2004.

There are several ways in which potential hosts may protect a nest from the detrimental impacts of brood parasitism. Two strategies may prevent a cowbird from laying: nest sit- ting (rushing to the nest and sitting in it; Hob- son and Sealy 1989, Gill and Sealy 1996) and aggression directed at the intruding cowbird (Robertson and Norman 1977, Briskie et al. 1990). The effectiveness of such behaviors in deterring parasitism is unclear. The hndings of Sealy et al. (1998) suggest that these behav- iors do not thwart parasitism. However, J. M. Budnik (pers. comm.) video-recorded a Brown-headed Cowbird {Molothnis ater) lay- ing an egg on the back of a midwestern Bell’s Vireo {Vireo bellii bellii) that refused to leave the nest, and several times observed vireo pairs preventing cowbirds from laying by physically attacking and driving them away (Budnik 1999, Budnik et al. 2001).

After a nest has been parasitized, the most effective anti-parasite responses by the host are to ( I ) eject the egg from the nest ( Roth- stein 1976, Sealy and Bazin 1995), or (2) bury the egg (Clark and Robertson 1981, wSealy 1995). A third strategy is that of abandoning the nest and re-nesting, but the effectiveness

' Dept, of Biology, San Diego Stale Univ., 5500 C’ampanilc Dr., San Diego, ('A 021X2, USA.

- U.S. (ieological .Survey, Western Lcological Re- .search ('enter, 5745 Kearny Villa Rcl., San Diego, ('A 02123, USA.

' ('urrent atkiress: Dept, of Itiological and I’hysical Sciences, ('olumbus Stale (’ommunity ('ollege, 550 I'.. Spring St., ('olumbus, OH 43215. IkSA.

' ('orrespoiuling author: e-mail: barbaiii_kus(«4isgs.gov

of this strategy varies (Graham 1988, Hosoi and Rothstein 2000, Kus 2002).

Nest-attendance behavior by potential hosts during times when cowbirds may lay eggs is an important consideration when examining a host’s defense capability (Neudorf and Sealy 1994, Clotfelter and Yasukawa 1999). Scott (1991) and Neudorf and Sealy (1994) found that Brown-headed Cowbirds lay near (usually before) sunrise during the laying stage of the nesting cycle. A host that is vigilant near the nest at these times may be able to defend its nest better than a host that is not attending its nest (Neudorf and Sealy 1994, Clotfelter and Yasukawa 1999). However, Sealy et al. (2000) found that sunrise nest attendance was a func- tion of the onset of incubation rather than a nest-defense strategy.

Least Bell’s Vireo {V. b. piisillit.s) is a fed- erally endangered songbird that is heavily par- asitized throughout most of its current range in California (Franzreb 1989; Kus 1999, 2002). Consequently, extensive cou bird-trap- ping programs have been instituted within the geographic range of this \ ireo. As Least BelLs Vireo is a recent host, co-occurring with cow- birds only during the last century (Maylield 1965), it may lack natural defenses against parasitism. Least Bell’s Vireos only rarely bury cowbird eggs (BliK unpubl. data), and, like many other small hosts, inclutling the western subspecies ol' Warbling Vireo (\'. g/7- vii.\ swain.sonii: ,S. G. .Scaly pets, comm.). Least Bell's Vireos are probably unable to grasp- or puncture-eject cowbird eggs (Roth- stein 1975a, .Spaw and Rohwer 1987, Rohwer and .Spaw 1988). I .east l^ell's Vireos abtuulon parasitized nests, but they do so at a mueh

17

18

THE WILSON BULLETIN • Vol. 1 16, No. !, March 2004

lower rate (mean = 29%, n = 207; Kus 1999) than the nominate subspecies, V. h. hellii [74%, n = 43 (Parker 1999); 51%, n = 63, (Budnik et al. 2001)], and nest desertion and re-nesting by Least Bell’s Vireos is not cur- rently an effective defense against the impacts of parasitism (Kus 2002). Nothing is known about the ability of Least Bell’s Vireos to deter parasitism through nest defense.

We video-recorded natural parasitism events of Least Bell’s Vireos by Brown-head- ed Cowbirds to examine this host’s ability to defend its nest and to analyze host responses after being parasitized. The specific questions we addressed were (1) when do Brown-head- ed Cowbirds lay their eggs in Least Bell’s Vir- eo nests with regard to time of day and stage in the nesting cycle, (2) are Least Bell’s Vir- eos attending the nest when cowbirds arrive to lay an egg, (3) how do Least Bell’s Vireos interact with cowbirds at the nest, and (4) do Least Bell’s Vireos accept or reject a parasitic egg? We addressed these questions to evaluate whether Least Bell’s Vireo may be able to adapt to the pressure of brood parasitism.

METHODS

We monitored 129 Least Bell’s Vireo nests along the San Luis Rey River in northern San Diego County, California, during the 2000 breeding season as part of a larger, long-term study of vireo demographics. Cowbirds were not trapped within our study site during this season. We placed cameras at 19 nests of 13 different vireo pairs between 28 April and 7 July. These nests were in early stages of the nesting cycle when Brown-headed Cowbirds typically parasitize nests (Lowther 1993, Sca- ly 1995). Videotapings were made from 1 day before clutch initiation through the midpoint (day 6) of incubation. We extended videotap- ing through day 6 of incubation to find if par- asitism might occur later in the cycle. Cam- eras were of two types: Fuhrman Microcams and Christensen Sentinel Systems. Lenses were placed within 1 m of nests, camouflaged by surrounding vegetation. The camera was connected by cable to the recorder, which was hidden 10-20 m away from the nest. The field of view for cameras included the nest and 5— 50 cm radii around the nest, depending on where we were able to locate the camera lens. We videotaped nest activity continuously in

time-lapse mode (20 frames/sec) until either the young fledged, or the nest was depredated or abandoned for other reasons, at which time we moved the camera to another nest.

We reviewed video recordings to determine whether an adult vireo was on or near the nest prior to parasitism, whether vireos defended the nest during parasitism attempts, and whether vireos accepted or rejected cowbird eggs after being parasitized. We noted the du- ration of each encounter between a female cowbird and the pair of vireos (sensu the “lay- ing bout’’ described in Sealy et al. 1995) as well as the actual time each cowbird spent poised in a laying posture on the nest. We re- viewed 2-hr segments (from 1 hr before sun- rise to 1 hr after sunrise) for each day a cam- era was located at a nest, up to the midpoint of incubation, targeting the time of day (Scott 1991, Neudorf and Sealy 1994) and stage of the nesting cycle (Lowther 1993, Sealy 1995) when cowbirds typically parasitize nests. We also noted whether an adult vireo roosted on its nest overnight. If an adult was not on the nest within 1 hr before sunrise, we assumed no adult roosted on the nest, and we recorded the time an adult first visited the nest in the morning relative to time of sunrise. We ob- tained sunrise times from the U.S. Naval Ob- servatory website.

We considered incubation to begin the day after the last egg was laid, although incubation begins with the penultimate egg (BEK pers. obs.). We chose this chronology to allow us to designate days as either laying days or in- cubation days. Eor nests parasitized during laying, we left cowbird eggs in place until clutches were complete (6-10 days after par- asitism), and then removed them as authorized by federal and state permits. Means are re- ported ± SE.

RESULTS

We videotaped three parasitism events by Brown-headed Cowbirds (Table 1). Two events occurred before sunrise, and one took place shortly after sunrise, yielding a mean parasitism-event time of sunrise — 1 1.0 ± 6.4 min (Table 1). One nest was parasitized on day 4 of incubation, whereas the others were parasitized during laying (Table 1). No nest was abandoned.

In each parasitism event, an adult vireo was

Sharp and Kiis • VIREO-COWBIRD INTERACTIONS AT NESTS

19

TABLE 1. Characteristics of the nesting stage of videotaped parasitism events at Least Bell’s Vireo nests, San Luis Rey River, San Diego County, California, 28 April-7 July 2000.^*

Date of parasitism	Time of parasitism	Duration of encounter	Time cowbird in nest, laying	Nesting stage when parasitized	Clutch size when parasitized	Completed clutch size (vireo eggs only)
6 May	SR'’ - 21 min	42 sec	23 sec	Day 2 of laying	1	4
15 May	SR - 13 min	35 sec	33 sec	Day 3 of laying	2	4
12 June	SR + 1 min	43 sec	23 sec	Day 4 of incubation	2	2

^ Parasitism event at a fourth nest was not recorded on tape due to technical difficulties. ^ SR = sunrise.

Still on the nest from the nocturnal roosting/ incubation period. Because birds were not marked, we were unable to identify the gender of individuals. Attendant vireos remained on the nest, but were forced off the nest after be- ing pecked repeatedly by the female cowbird. Once ousted from the nest, the vireo attacked the cowbird and was joined by its mate within a few seconds, suggesting the mate was near- by. Although we could not record sound in time-lapse mode, we could see that the vireos were scolding, and we presume that this vo- calization is what attracted the second adult vireo. Both vireos scolded, jumped on the cowbird, struck the cowbird while flying by it, tried to pull the cowbird away from the nest, and repeatedly pecked the intruder. One cowbird left the nest on its first attempt, but returned within 6 sec and successfully laid an egg. Encounters between female cowbirds and hosts averaged 40.0 ± 2.5 sec (Table 1). The mean time each cowbird spent on the nest poised in a laying posture was 26.3 ± 3.3 sec (Table 1 ). No host eggs were removed by cowbirds.

After parasitism, one member of each vireo pair returned to its nest 7-18 sec (/? = 3 pairs)

after the cowbird left, and resumed normal nest attendance 21-146 sec after parasitism. Upon returning to the nest to brood, the adult inspected the contents prior to sitting in the nest. Pairs parasitized during laying continued laying until completing the typical clutch of four vireo eggs (Brown 1993, Kus 1999). Egg-laying continued at a normal rate of one egg per day, with the exception of one female that did not lay her fourth egg until 5-7 days after her third egg (6-8 days after parasitism). None of the vireo pairs removed the parasitic egg. In the 1 14 “nest-days” of tapes we re- viewed (n = 19 nests), there were no encoun- ters with cowbirds other than the three that resulted in parasitism. Thus, we found no in- stances in which Least Bell’s Vireos prevented a cowbird from parasitizing a nest once the cowbird had reached the nest.

Of the 1 14 nest-days of tapes, there were only 1 1 instances in which an adult \ ireo was not on the nest overnight. Of these 1 1 cases, 4 occurred before laying began, 6 were during laying, and 1 occurred on day 2 of incubation (Table 2). There were no instances in which an adult did not roost on a nest overnight from day 3 of incubation forward. Linear regression

TABLE 2. Frequency of overnight roosting on the nest and arrival times of non-roosting adult Least Bell's

Vireos over the nesting cycle, San Luis Rey River, San Diego County, California. 28 April-7 July 2()()().

Day 4 Da\ I

Day before lay*' Day I lay Day 2 lay Day .t lay lay me ' Da> 2 me

Adult on nest over- 0/4 3/5 11/13 0/10 5/6 1.5/15 LV14

night/total nests observed

Mean arrival time of +26.0 13.1 +30.0 5.0 +16.5 * 0.5 +26 *4 0

non-roosting adult relative to SR'’ 1.

SI: (min)

" lay laying, ine ineubalion '’SR sunrise.

20

THE WILSON BULLETIN • VoL 1 16, No. /. March 2004

of arrival time (relative to sunrise) versus day of nesting cycle demonstrated that adults that did not roost on the nest overnight arrived at the nest earlier in the day as the nesting cycle progressed {t = -3.859, P = 0.018; Table 2).

DISCUSSION

Our recordings represent the first video documentation of parasitism by Brown-head- ed Cowbirds of Least Bell’s Vireos. Two of the recorded instances of parasitism occurred during the laying stage of the nesting cycle, as is most often reported in the literature (Lowther 1993, Sealy 1995). The third nest was not parasitized until midway through the incubation stage. Although most parasitism occurs during the laying stage, instances of parasitism during the incubation stage are not uncommon (Lowther 1993, Sealy 1995).

As reported elsewhere (Scott 1991, Neudorf and Sealy 1994), Brown-headed Cowbirds laid their eggs around sunrise. This timing of laying is believed to be an adaptation for lay- ing when host adults do not typically attend nests (Scott 1991); however, Neudorf and Sea- ly (1994) and Sealy et al. (2()()0) found that some hosts do attend their nests at sunrise. Clotfelter and Yasukawa (1999) suggested that onset of nocturnal incubation early in the nesting cycle may be a strategy by which Red- " winged Blackbirds (Agelaius phoeniceus) pre- vent parasitism. Birds roosting or incubating overnight will be on the nest at dawn, when cowbirds lay and, thus, will be able to defend the nest better than if they were away from the nest during a cowbird visit (Neudorf and Sealy 1994, Clotfelter and Yasukawa 1999).

An adult vireo was roosting at the nest or incubating before each parasitism event we videotaped and an adult roosted overnight on nearly every nest-day we observed. All 1 1 in- stances in which an adult did not roost on the nest overnight occurred before or during lay- ing, with the exception of one instance that occurred on day 2 of incubation (Table 2). This pattern, the earlier arrival times of non- roosting adults as the nesting cycle progressed (Table 2), and the vireos’ inability to defend their nest when present at the time of a cow- bird visit, support the conclusion of Sealy et al. (2000): nest attentiveness is more a func- tion of the onset of incubation than an anti- parasite strategy.

Vireos responded aggressively to the fe- male cowbird in each instance of parasitism. However, sitting in the nest and attacking the cowbird did not prevent parasitism, contrary to observations that midwestern Bell’s Vireos escaped parasitism with similar behavior (Budnik 1999, Budnik et al. 2001). We are uncertain as to why defense tactics that were effective in the midwestern subspecies failed to deter parasitism in Least Bell’s Vireos.

Least Bell’s Vireos quickly accepted cow- bird eggs. Possible reasons for acceptance may be ( 1 ) lack of recognition of the cowbird egg (Rothstein 1975b, 1990; Sealy 1996; Tak- asu 1998), or (2) vireos may recognize cow- bird eggs as foreign but are unable to eject them. If the latter is the case, nest abandon- ment is the only response to parasitism avail- able to Least Bell’s Vireos. Although nest abandonment was not observed in our sample of videotaped nests. Least Bell’s Vireos are known to abandon parasitized nests (Kus 1999), albeit it at a low rate (Kus 2002).

When Least Bell’s Vireos accept cowbird eggs, the end result, without human interven- tion (i.e., removal of cowbird eggs from nests), is the failure to produce vireo young (Kus 1999, 2002). Least Bell’s Vireos have never been observed to fledge vireo and cow- bird young from the same nest (BEK unpubl. data). We recognize that our sample size is small (such data are extremely difficult to ob- tain); however, all of the vireos in our study failed to deter parasitism. Given the apparent inability to avert parasitism even when attend- ing the nest at sunrise, and the ineffectiveness of nest abandonment and re-nesting in re- sponse to parasitism (Kus 2002), Least Bell’s Vireos may lack behavioral defenses sufficient to prevent parasitism or its negative impacts once a female cowbird has reached a vireo nest.

ACKNOWLEDGMENTS

We thank P. Beck, B. Peterson, and D. Kisner for monitoring nests of Least Bell’s Vireos. B. Peterson and M. Wellik assisted with camera set up and the changing of batteries to run the cameras. K. Burns, D. Deutschman, P Pryde, B. Peterson, P Beck, J. Wells, K. Lerree, D. Kisner, and J. Rourke assisted in devel- opment of the project. J. Guarderas, R. Ligueroa, V. Mestas-Romero, S. Sullivan, R. del Rio, M. Giometti, and J. Mascarin provided additional field assistance. Much gratitude is extended to R. Veillette for review-

Sharp and Kus • VIREO-COWBIRD INTERACTIONS AT NESTS

21

ing hours of videotapes. Earlier versions of this man- uscript were greatly improved by comments from J. V. Briskie, S. I. Rothstein, S. G. Sealy, and two anony- mous referees. We thank the Bureau of Reclamation and Arizona Game and Fish Department for the loan of five camera systems used in this study. This work was funded by the California Department of Transpor- tation, District 1 1 , and by grants to the senior author from the Garden Club of America (c/o Cornell Labo- ratory of Ornithology), the Frank M. Chapman Fund of the American Museum of Natural History, Sigma Xi, Los Angeles Audubon Society, and the San Diego State University Evolutionary Biology Program Area.

LITERATURE CITED

Briskie, J. V., S. G. Sealy, and K. A. Hobson. 1990. Differential parasitism of Least Flycatchers and Yellow Warblers by the Brown-headed Cowbird. Behavioral Ecology and Sociobiology 27:403- 410.

Brown, B. T. 1993. Bell’s Vireo (Vireo bellii). The Birds of North America, no. 35.

Budnik, J. M. 1999. Demography and factors influ- encing nesting success of Bell’s Vireo in grass- land-shrub habitat. M.Sc. thesis. University of Missouri, Columbia.

Budnik, J. M., D. E. Burhans, M. R. Ryan, and F. R. Thompson, 111. 2001. Nest desertion and apparent nest protection by Bell’s Vireos in response to cowbird parasitism. Condor 103:639-643.

Clark, K. L. and R. J. Robertson. 1981. Cowbird parasitism and evolution of anti-parasite strategies in the Yellow Warbler. Wilson Bulletin 93:249- 258.

Clotfelter, E. D. and K. Yasukawa. 1999. The func- tion of early onset of nocturnal incubation in Red- winged Blackbirds. Auk 116:417-426.

Franzreb, K. E. 1989. Ecology and conservation of the endangered Least Bell’s Vireo. Biological Re- port 89(1), U.S. Fish and Wildlife vService, Wash- ington, D.C.

Gill, S. A. and S. G. Sealy. 1996. Nest defense by Yellow Warblers: recognition of a brood parasite and an avian nest predator. Behaviour 133:263- 282.

Graham, D. S. 1988. Responses of five host species to cowbird parasitism. Condor 90:588-591. Hob.son, K. a. and S. G. Sealy. 1989. Responses of Yellow Warblers to the threat of cowbird parasit- ism. Animal Behaviour 38:510-519.

Hosoi, S. A. AND S. I. R()THsri:iN. 2000. Nest desertion and cowbird parasitism: evidence for evolved re- sponses and evolutionary lag. Animal Behaviour 59:823-840.

Kus, B. H,. 1999. Impacts ol Brown-headed Cowbird parasitism on the productivity ot the endangered Least Bell’s Vireo. Studies in Avian Biology 18: 160-166.

Kus, B. F2 2002. Fitness consec|uences of nest ilcser- tit)ii in an endangercti host, the Least Bell's Vireo. C’ondor 104:795 802.

Lowther, P. E. 1993. Brown-headed Cowbird {Mol- othnis ater). The Birds of North America, no. 47.

Mayfield, H. E 1965. The Brown-headed Cowbird, with old and new hosts. Living Bird 4:13-29.

Neudorf, D. L. and S. G. Sealy. 1994. Sunrise nest attentiveness in cowbird hosts. Condor 96:162- 169.

Parker, T. H. 1999. Response of Bell’s Vireos to brood parasitism by the Brown-headed Cowbird in Kansas. Wilson Bulletin 1 1 1:499-504.

Robertson, R. J. and R. F. Norman. 1977. The func- tion and evolution of aggressive behavior towards the Brown-headed Cowbird (Molotlirus ater). Ca- nadian Journal of Zoology 55:508-518.

Rohwer, S. and C. D. Spaw. 1988. Evolutionary lag versus bill-size constraints: a comparative study of the acceptance of cowbird eggs by old hosts. Evo- lutionary Ecology 2:27-36.

Rothstein, S. I. 1975a. An experimental and teleo- nomic investigation of avian brood parasitism. Condor 77:250-271.

Rothstein, S. I. 1975b. Evolutionary rates and host defenses against avian brood parasitism. American Naturalist 109:161-176.

Rothstein, S. I. 1976. Experiments on defenses Cedar Waxwings use against cowbird parasitism. Auk 93:675-691.

Rothstein, S. I. 1990. A model system for coevolu- tion: avian brood parasitism. Annual Review of Ecology and Systematics 21:481-508.

Scott, D. M. 1991. The time of day of egg-laying by the Brown-headed Cowbird and other icterines. Canadian Journal of Zoology 69:2093-2099.

Sealy, S. G. 1995. Burial of cowbird eggs by parasit- ized Yellow Warblers: an empirical and experi- mental study. Animal Behaviour 49:877-889.

Sealy, S. G. 1996. Evolution of host defenses against brood parasitism: implications of puncture-ejec- tion by a small passerine. Auk I 13:346-355.

Sealy, S. G. and R. C. Bazin. 1995. Low freciuency of observed cowbird parasitism on Eastern King- birds: host rejection, effective nest defense, or par- asite avoidance? Behavioral Ecology 6:140-145.

Se:aly, S. G., D. G. McMa.sti;r. S. A. Ciii.i., and D.

L. Neudorf. 2000. Yellow Warbler nest attentive- ness: antiparasite strategy or onset of incubation? Pages 169-177 in Ecology and management ol cowbirds and their hosts: studies in the conser- vation of North American passerine birds (J. N.

M. Smith, r. L. C'ook, ,S. I. Rothstein, S. K. Rob- inson, and S. G. Sealy, FaIs.). Uni\ersity ol fexas Press, Austin.

Si;ai.y, S. G., I). L. Ni iidori, and D. P. Hiil. 1005. Rapitl laying by Brown-hcatIctI ('owbinls Molot/i- rn.s ater aiul other parasitic birtls. Ibis 137:76 84.

Si Ai.Y, S. (i.. I). L. Ni i DORi , K. A. Hobson, and S. A. Gii I . 1008. Nest tlelense by potential hosts of the Brow ii-heaileil Cow binl: methoilologieal ap proaches, benelits of ilefense, aiul coe\olutit>n. Pages 104 21 I in Parasitic birds and their hosts: stiulies in coevolution tS. I. Rothstein aiul ,S. K.

22

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

Robinson, Eds.). Oxford University Press, New York.

Spaw, C. D. and S. Rohwer. 1987. A comparative study of eggshell thickness in cowbirds and other passerines. Condor 89:307-318.

Takasu, E 1998. Why do all hosts not show defense against avian brood parasitism: evolutionary lag or equilibrium? American Naturalist 151:193- 205.

Wilson Bulletin, 1 16(1), 2004, pp. 23—30

EFFECTS OF FOOD SUPPLEMENTATION ON FEMALE NEST ATTENTIVENESS AND INCUBATION MATE FEEDING IN TWO SYMPATRIC WREN SPECIES

AARON T. PEARSE,'-’’ JOHN F. CAVITT,' '» AND JACK F. CULLY, JR.^

ABSTRACT. — We examined effects of incubation mate feeding on female incubation behavior and correlates of fitness by providing female Bewick’s Wrens {Thryomanes bewickii) and House Wrens {Troglodytes aeclon) with food supplements. Males of these species vary in their rates of feeding; Bewick’s Wrens feed their incu- bating mates frequently, whereas House Wrens seldom engage in this behavior. Average length of incubation bout and nest attentiveness (proportion of time spent on the nest) were higher for supplemented female Bewick’s Wrens and House Wrens compared to controls. Furthermore, mates of supplemented Bewick’s Wrens provisioned females at lower rates than controls, and their rate of feeding was inversely correlated with ambient temperature. Incubation length and hatching success were not significantly different between treatments for either species. These results suggest that incubation mate feeding can increase female nest attentiveness and perhaps enhance fitness of both males and females. In House Wrens, potential tradeoffs between the benefits of parental care and opportunities to obtain additional mates may explain why males rarely feed incubating females. Received I July 2003, accepted 15 March 2004.

In species that exhibit parental care, there is often a division of labor between sexes, with one sex primarily attending the nest. Consequently, trade-offs between offspring development and survival versus parental con- dition can exist if nest attentiveness is con- strained by parental food limitation (Royama 1966). Food brought to the attending adult by the nonattending mate may ameliorate food limitation, and thus, offset these trade-offs (Smith et al. 1989). Feeding of incubating fe- males by mates occurs in more than 40% of North American passerines (Kendeigh 1952) and is most pronounced in cavity nesters. Nonetheless, considerable variation in the rate of incubation mate feeding exists (Martin and Ghalambor 1999).

Traditionally, incubation feeding was thought to maintain the pair bond between mates (Lack 1940, Kluyver 1950, Andrew 1961) or represent a premature attempt by males to feed nestlings (Skutch 1953, Nolan

' Div. of Biology, Kansas State Univ., Manhattan, KS 66506, USA.

^ USGS-BRI) Kansas Coop. Wildlife Research Unit, Div. of Biology, 204 Leasure Hall, Kansas State Univ., Manhattan, KS 66506, USA.

%’urrent address: Dept, of Wildlife and I'isheries, Box 9600 Thompson Hall. Mississippi .State Univ.. Mississippi State, MS 39762. USA.

■* Current address: Dept, of Zoology, Weber .State Univ., 2505 University C’ir., Ogden, U T X440S, USA. ^Corresponding author; e-mail: atp33(f'hnsstate.etlu

1958, Ricklefs 1974, Johnson and Kermott 1992). Both of these hypotheses have been challenged, and it has been suggested that food delivered to females constitutes an es- sential nutritional contribution (i.e., the food limitation hypothesis; von Haartman 1958, Royama 1966, Krebs 1970, Smith 1980, Nils- son and Smith 1988).

Experimental tests of the potential adaptive benefit of incubation feeding (in terms of fe- male attentiveness and hatching success) are relatively rare (e.g., Nilsson and Smith 1988, Moreno 1989, Smith et al. 1989). In this study, we examined effects of food supple- ments on female incubation behavior and cor- relates of fitness in two sympatric, secondary cavity-nesting species, Bewick’s Wren {Thryo- maues bewickii) and House Wren {Troglo- clyte.s aedon). Incubation mate feeding is com- mon in Bewick's Wrens (Miller 1941 ), where- as male House Wrens rarely feed their mates during incubation (Johnson and Kermott 1992). We increased food available to incu- bating females by pro\ iding food supplements inside nest boxes (Nilsson and Smith 1988. Smith et al. 1989). Jhis allowed females to have sole access to food without lea\ing nest cavities, simulating incubation feeding. II mate I'eeding constitutes an important contri- bution to females, we predicted that lootl sup- plements would enhance nest attentiveness. If additional foot! enhances female altenti\ eness. hatching success shouki increase and tluration

23

24

THE WILSON BULLETIN • VoL ! 16, No. /, March 2004

of incubation should be reduced at supple- mented nests relative to controls. If males do monitor female attentiveness, we also predict- ed they would adjust their rates of mate feed- ing accordingly and reduce their rate of in- cubation feeding to highly attentive females provided with food supplements (Smith et al. 1989).

METHODS

Study area and species. — We conducted this study from early April through late Au- gust 1997 at the Konza Prairie Biological Sta- tion, 10 km south of Manhattan, Kansas (see Zimmerman 1993 for site description). We monitored 152 nest boxes along gallery for- ests, attenuated gallery forests, and rock-out- crop shrub communities of Konza (Kennedy and White 1996).

Bewick’s Wrens are common summer res- idents and occasional winter residents of the attenuated gallery forest (Zimmerman 1993). In Kansas, Bewick’s Wrens are double-brood- ed; first nests are initiated in early April and second nests are initiated in late May (Farley 1987). Bewick’s Wrens are socially monoga- mous with only a few suspected cases of po- lygyny; mean clutch size for this population is 6.1 eggs (Kennedy and White 1997). Only female Bewick’s Wrens incubate eggs, but males feed their incubating mates and assist in feeding nestlings (Miller 1941).

House Wrens are common summer resi- dents of Konza, using both gallery and atten- uated gallery forests (Zimmerman 1993). They are double-brooded, initiating first nests in early May and second nests in late June. House Wrens are considered socially monog- amous (Johnson 1998), but the percentage of males that attract secondary females can be as high as 14% in some populations (Soukup and Thompson 1997a). Mean clutch size of first broods is 6.2 eggs, slightly larger than second broods (5.9 eggs; E. D. Kennedy pers. comm.). Only females incubate, but males generally assist in feeding nestlings. The rate of incubation feeding in a Wyoming popula- tion of House Wrens was found to be extreme- ly low (0.2 feedings/hr; Johnson and Kermott 1992), but there are few data on this behavior for other populations.

General procedures. — We checked all nest boxes once weekly from early April until late

July to determine clutch initiation dates. Ac- tive nests were then visited every 1—2 days to determine presence and number of eggs or nestlings. Initiation of incubation was deter- mined by egg temperature (warm versus cold) and female behavior. Near the expected hatch dates, nests were visited daily to determine hatch date and hatching success.

Food supplementation experiment. — All nests discovered during egg laying were ran- domly allocated to either food supplemented or control treatments. Nests allocated to the food supplementation treatment were supplied with 15 g of live mealworm larvae {Tenehrio molitor) every day (06:00-10:00 CST) during incubation. This amount of food was chosen based on the estimate that a 10.6 g wren ex- pends ~61 kJ/day (Dykstra and Karasov 1993). Assuming that the energy content of mealworms is 1 1 .59 kJ/g (calculated from Bell 1990) and a wren’s assimilation efficien- cy of mealworms is 0.65 (Kacelnik 1984), a female would need to consume 8.2 g of meal- worms to satisfy daily energy requirements. Therefore, a 15-g supplement represents a substantial energy contribution to incubating females. Food supplements were placed in plastic feeding dishes (35 mm film canisters; diameter 3.33 cm, height 4.75 cm) hung inside nest boxes above the nest rim (cf. Nilsson and Smith 1988). This allowed us to simulate male provisioning at the nest entranee and enabled the female to obtain food without leaving the nest cavity.

In most cases, mealworms delivered to nests were consumed before our next visit. If food remained in the canister, dead larvae were removed and replaced with fresh larvae. Videotaped observations revealed that three female House Wrens occasionally removed mealworms from their nest box (see also Johnson and Kermott 1992). This behavior was never observed at Bewick’s Wren nests. It is unclear whether these female House Wrens consumed larvae outside their nest boxes or removed them without consuming them. Thus, we performed two exploratory analyses when comparing the effects of treat- ment on House Wrens, one using all nests and the second excluding data from nests where females removed mealworms. Because the re- sults were similar, we present combined data.

To identify the importance of food avail-

Fearse et al. • FOOD EFFECTS ON NEST ATTENTIVENESS

25

ability to female nest attentiveness, incubation behavior was monitored for 2-4 hr per nest from 07:00 to 12:00 by battery-operated video cameras. Each nest was recorded twice: once during early incubation (incubation day 1-6) and once during late incubation (incubation day 7-12). We observed nests twice to in- crease observation time and reduce effects of potential anomalous observations. Sampling early and late also allowed us to test whether nest attentiveness changed during the incu- bation period. Tripods were placed 5-10 m from a nest box one day before taping to ac- climate adults to the disturbance. From re- cordings we determined average length of in- cubation bout (time inside the nest box), av- erage length of recess bout (time outside the nest box), female nest attentiveness (propor- tion of time inside the nest box), and frequen- cy of mate feedings at the nest. Because vid- eotaped observations were used, we could not determine the number of mate feedings that may have occurred away from the nest site, out of camera range. It should be noted, how- ever, that in some populations the frequency of House Wren mate feedings away from the nest is extremely low (Johnson and Kermott 1992). The extent to which Bewick’s Wren males may feed females away from the nest is not known. Temperature at time of taping was obtained from hourly data recorded at a weather station located at the Konza head- quarters.

Data analysis. — All statistical analyses were performed using the ^Statistical Analysis .System (SAS Institute, Inc. 1999). Compari- sons of clutch size and clutch initiation be- tween food-supplemented and control nests for each species were performed using r-tests (PROC FfEST). Correlations between tem- perature at the time of observation and female nest attentiveness (both species) and male feeding rate (Bewick’s Wrens only) were cal- culated using PR()(’ CT)RR.

Four dependant variables describing incu- bation behavior were analyzed in the food supplementation experiment: lengths of incu- bation and recess bouts, nest attentiveness, and mate-feeding rate. Mean incubation- and recess-bout lengths were calculated lor each videotape sessiofi by di\ itling incubation- and recess-bout lengths by the numbei ol incuba- tion and recess bouts taken, respectively. Fe-

male nest attentiveness was defined as the pro- portion of time the female spent in the nest box. Mate-feeding rate (feedings/hr; Bewiek’s Wrens only) was caleulated by dividing the frequency of mate feedings for a videotape session by the total time. We did not calculate mate-feeding rate for House Wren males be- cause we only observed three instances of this behavior during our videotape sessions. All behavioral response variables were analyzed using repeated measures ANOVA, with food supplementation as the independent variable of interest and species as a blocking variable. Repeated measures ANOVA was used to ae- count for correlation between multiple nest observations of a single nest (PROC MIXED). Least-squared (LS) means and associated P- values were obtained using the LSMEANS statement and PDIFF option. Temperature at the time of taping was used as a covariate in the analysis of mate-feeding rate because a significant correlation was found both in our and other studies (Nilsson and Smith 1988, Smith et al. 1989, Halupka 1994). Differences between early and late incubation behavior were analyzed by species using paired /-tests (PROC UNIVARIATE).

Two dependant variables correlated with ht- ness — incubation length and hatching suc- cess— were analyzed to determine whether food supplementation potentially increased fit- ness. Incubation length (INCL) was calculated using hatch date (HD), clutch size (CS), and clutch initiation date (CID):

INCL = HD - |CID + (CS - 1)|.

Analysis of variance was used lo examine dif- ferences in incubation length with the same independent variables as described above (PROC MIXED). Hatching success was ana- lyzed using a generalized linear mixed model approach, which is et|uivalent to a mixed model logistic regression (CLIMMIX Macro: Wolfinger and O'C’onnell 1993). I he success or laihire of indi\idual eggs from successful nests (one or more eggs hatched) were re- sponse variables, and nests were consideied a cluster sample because responses of iiulividual eggs w ithin a nest may be conelated. Standard errors t)l hatching success were calculated us- ing the Delta method from staiulartl errors computetl on the U)gil scale (Littell et al. 199b).

26

THE WILSON BULLETIN • Vol. 116, No. I, March 2004

TABLE 1. Least-square means of ineubation behavior of Bewick’s Wrens (BW) and House Wrens (HW) that were, or were not, supplied with additional food, northeast Kansas, summer 1997.

Food supplemented

Control

Species	Variable			SE		X	SE
BW	Incubation bout'’		51.56	6.89		35.33	6.16
	Recess bouL	12	9.77	1.70	15	12.77	1.52
	Nest attentiveness^		0.82	0.04		0.70	0.04
HW	Incubation bout*’		30.21	5.97		13.15	6.16
	Recess bouL	16	4.78	1.47	15	5.43	1.52
	Nest attentiveness^'		0.81	0.04		0.71	0.04

^ Number of videotape sessions.

^Average amount of time females spent in the nest box without leaving (min).

Average of time females spent out of the nest box before returning (min).

^ Proportion of time females spent in the nest box.

Male House Wrens without food were ob- served visiting nest boxes of their mates. We calculated total visits by Bewick’s Wren and House Wren males (total visits = feeding trips -f nonfeeding trips) and compared total visits of the two species at control nests using re- peated measures ANOVA (PROC MIXED). An alpha value of 0.05 was selected to deter- mine significant differences for all tests.

RESULTS

We observed 15 Bewick’s Wren nests (7 food supplemented, 8 control) for 53.4 hr, and 17 House Wren nests (9 food supplemented, 8 control) for 64 hr. Neither clutch size nor date of clutch initiation (mean difference, 95% Cl) differed between supplemented and con- trol Bewick’s Wren nests [clutch size: 0.13 eggs (-1.37, 1.63); clutch initiation date: -2.5 days (-23, 18)]. Similar results were ob- served for supplemented versus control House Wren nests [clutch size: 0.28 eggs (-0.63, 1.19); clutch initiation date: 2 days (—14, 18)]. No differences were detected between early and late incubation behavior for either species {P > 0.25).

Temperature was not correlated with female nest attentiveness for either species (Bewick’s Wren: r = 0. 16, P = 0.42; House Wren: r - 0.03, P = 0.86), but was negatively correlated with feeding rates of male Bewick’s Wrens (r = —0.66 P < 0.001). Consequently, temper- ature was used as a covariate in the analysis of feeding rate. No significant correlation was found between Bewick’s Wren mate-feeding rate and female nest attentiveness (r = —0.23, P = 0.24).

Eood-supplemented females had signifi-

cantly longer average incubation bouts (P, 2s = 6.97, P = 0.013) compared to females in control nests (Bewick’s Wren, P = 0.090; House Wren, P = 0.057; Table 1). Average length of recess bout was not significantly dif- ferent (Pi. 28 = 1.38, P = 0.25) between sup- plemented and control nests (Bewick’s Wren, P = 0.20; House Wren, P = 0.76; Table 1). Eemale Bewick’s and House wrens were 18 and 14% (respectively) more attentive to their nest when food-supplemented compared to control females (Pi.28 ^ 8.55, P = 0.007; Be- wick’s Wren, P = 0.034: House Wren, P = 0.068; Table 1). Male Bewick’s Wrens made an average of 1.1 fewer mate feedings/hr to food-supplemented females compared to con- trol females (P| 13 = 5.06, P = 0.042; Pig. lA). Furthermore, male Bewick’s Wrens made 0.162 more mate feedings/hr for every 1°C drop in ambient temperature (Pi n = 21.92, P < 0.001). Experimental food supplementation did not explain variation in incubation length (^1.22 0.52, P = 0.48; Bewick’s Wren, P =

0.28; House Wren, P = 0.93; Table 2) or hatching success (Pi. 22 ^ 0.70, P = 0.41; Be- wick’s Wren, P = 0.59; House Wren, P 0.52; Table 2).

Male House Wrens from control nests made an average of 1.88 (SE = 0.48) total visits/hr to their nest box during our videotape ses- sions. This did not differ from the average rate of total visits for Bewick’s Wren [1.56 (SE = 0.50) total visits/hr; Pi ,4 = 0.17, P = 0.68; Fig. IB].

DISCUSSION

The results of our study support the food limitation hypothesis: food provided to incu-

Pearse et al. • FOOD EFFECTS ON NEST ATTENTIVENESS

27

EIG. 1. (A) Eeeding rate (feedings/hr ± 1 SE) of

male Bewick’s Wrens at food-supplemented and con- trol (no food supplementation) nests during incubation, adjusted using ANCOVA for a mean ambient temper- ature of 16.78° C. (B) Total visit rate (feedings + non- feedings/hr ± 1 SE) of male Bewick’s Wrens and House Wrens at control nests.

bating females affects their parental effort. Additional food provided to females increased average length of incubation bout and nest at- tentiveness, suggesting that nest attentiveness is partially determined by the amount of en-

ergy available to the female. Our study sup- ports the results of Smith et al. (1989), who found that nest attentiveness in Pied Flycatch- ers (Ficediila hypoleuca, a species that exhib- its mate-feeding behavior) was greater when females were provisioned with additional food.

Bewick’s Wren males adjusted rates of in- cubation feeding to supplemented females: fe- males provided with additional food were fed less often than females not receiving food sup- plements. Smith et al. (1989) also reported lower male feeding rates to food-provisioned female Pied Flycatchers. Additionally, higher rates of mate feeding in Bewick’s Wren males were observed as ambient temperature de- creased. This response also has been observed in other species exhibiting incubation feeding (Nilsson and Smith 1988, Smith et al. 1989, Halupka 1994). Our results suggest that pro- visioning incubating females is costly to male Bewick’s Wrens and that they regulate their rate of feeding depending on female nest at- tentiveness and nutritional state.

Providing adult females of either species with additional food did not result in signifi- cant reductions in length of incubation period or in increased hatching success relative to controls, although in Bewick’s Wrens there was a trend toward a shorter incubation period for supplemented females (Table 2). Other re- searchers have documented that mate feeding during incubation can influence these vari- ables (Lyon and Montgomerie 1985, Nilsson and Smith 1988). Averaging 2 years of data (17 nests), Nilsson and Smith (1988) reported significantly earlier hatching ( 18.9 hr) in food- provisioned Blue Tit {Purus caendeus) nests than in controls. We were unable to measure time of hatching with such precision, but after

TABLE 2. Least-square means of' incubation length and hatching success of Bewick's Wrens (BV\ ) and House Wrens (HW) that were supplied with additional food, or not, in northeast Kansas in summer IUd7.

t-ood supplemented

Cotilrol

Species	Variable	/r*	V	sr		'	SI
BW	Incubation length*’	6	12.83	0.77	1	14.00	0.71
	Hatching success^		0.78	0. 1 .s		0.74	0.13
HW	Incubation length*’	7	1 1 .67	0.77	6	1 1 .57	0.71
	Hatching success'		0.96	0.16		0.82	0.10

■' Sample si/e of nests used in each analysis. Days of incubation needed to hatch a clutch Propiirtion of successfully hatchetl eggs.

28

THK WII.SON BUI.LETIN • Vol. 1 16, No. I, March 2004

converting our data to hours, we found that hatching was 28 hr earlier in supplemented Bewick’s Wren nests than in control nests. Al- though hatching success was not significantly different between treatment and control nests, it was greater for both species when given ad- ditional food: 0.20 and 0.95 additional eggs hatched in food-provisioned Bewick’s and House wren nests, respectively. In Blue Tits, Nilsson and Smith (1988) found a significant increase (6.5%) in hatching success among food-provisioned nests. Even though we did not detect a statistical difference in hatching success, at a population level this observed difference might be of ecological importance. Furthermore, if our study had been conducted in years with poor food availability or cooler temperatures, htness benefits of additional food might have been more apparent (the long term mean temperature for Manhattan, Kansas for May through July 1897-1994, was 18.3°, 23.7°, and 26.6° C; mean temperatures for May through July 1997 were 16.8°, 24.0°, and 27.3° C).

Enhancing nest attentiveness through incu- bation mate feeding could have other benefits (other than reduced incubation length or in- creased hatching success), such as serving to reduce intra- and interspecihc nest destruc- tion. Nest guarding has been shown to reduce nest predation in other species (Simons 1988, Cavitt 1998), and time available for guarding can be limited by food availability (Cavitt 1998). Thus, if nest destruction by House Wrens is an important source of nest loss for Bewick’s Wrens, increased nest attentiveness may further enhance fitness by reducing the probability of nest destruction by House Wrens. Kennedy and White (1996) reported that the percent of failed Bewick’s Wren nests caused by House Wrens on our site in other years ranged from 33 to 100%. During our study, however. House Wrens destroyed only one Bewick’s Wren nest; thus, we could not test this hypothesis with our data. The nest- destruction hypothesis does not explain the lack of incubation feeding observed in House Wrens, because they are also vulnerable to nest destruction by conspecifics (Johnson 1998). Yet, House Wrens may use other strat- egies, such as the coordination of nest-guard- ing activities (Ziolkowski et al. 1997), to re- duce nest destruction by conspecifics.

If providing additional food to female House Wrens can enhance nest attentiveness and, potentially, male fitness, why don’t males feed their incubating mates more frequently? Several hypotheses have been proposed to ex- plain the lack of incubation mate feeding (Mo- reno 1989, Johnson and Kermott 1992). The predation hypothesis (Lyon and Montgomerie 1987) proposes that species with a greater risk of nest predation should have lower rates of incubation feeding than species with lower predation risks, because increased trips to the nest may attract attention of predators and in- crease predation risk (Skutch 1949; Martin 1992, 1996). In fact, incubation feeding rates in a suite of coexisting species was inversely correlated with predation rate (Martin and Ghalambor 1999). Predation is not a likely ex- planation in the Konza population because we commonly observed male House Wrens vis- iting their nests during the incubation period without delivering food. Total number of vis- its made by House Wren males was not sig- nificantly different from the total number of trips made by male Bewick’s Wrens (Fig. IB). Thus, the occurrence of non-feeding visits by male House Wrens is not consistent with the nest predation hypothesis.

An alternative explanation for the differ- ence between male House Wrens and Be- wick’s Wrens is that although food provided to the female is beneficial, other activities may provide greater gains in male htness (Lifjeld and Slagsvold 1986, Lifjeld et al. 1987). Male House Wrens might, for example, increase ht- ness by seeking extra-pair copulations and at- tracting additional mates. Johnson and Ker- mott (1992) discounted this hypothesis be- cause mate-feeding rates did not differ signif- icantly between males that attempted to attract additional mates and those that did not. How- ever, because House Wren incubation feeding rates are extremely low and variable, detecting any signihcant difference between males that vary in this behavior would be difhcult. Male House Wrens frequently invest time and en- ergy intruding onto adjacent territories (2.02 ± 0.41 intrusions/hr) to obtain extra-pair cop- ulations (Johnson and Kermott 1989). In an Illinois population of House Wrens, Soukup and Thompson (1997b) found a high rate of extra-pair paternity (—27% of all nests sam- pled) and documented that approximately

Pearse et al. • FOOD EFFECTS ON NEST ATTENTIVENESS

29

14% of males were polygynous. Thus, other activities may enhance fitness of male House Wrens more than improving female attentive- ness via mate feeding.

Our results demonstrate that food provided by males to incubating females can be an im- portant factor influencing nest attentiveness and may enhance fitness. The disparity in male mate-feeding rates between these species most likely reflects differences in benefits to male fitness. The ability to maintain high lev- els of nest attentiveness may have a great ef- fect on the fitness of male Bewick’s Wrens because House Wrens are important nest pred- ators. In contrast, benefits of increased nest attentiveness to the fitness of male House Wrens may be outweighed by the benefits of participation in other activities, such as extra- pair copulations and polygamous mating.

ACKNOWLEDGMENTS

We greatly appreciate the generosity of E. D. Ken- nedy and D. W. White for their willingness to share data, ideas, and nest boxes. We thank N. A. Bargmann for assistance in the field and T. Alworth, E. D. Ken- nedy, and two anonymous reviewers for their com- ments on drafts of this manuscript. Portions of this research were funded by an NSF Doctoral Dissertation Improvement Grant (DEB-9520335) to JEC, an NSE- REU Grant (DBI-953 1 3 10), an NSF LTER Grant (DEB-901 1662) to the Division of Biology, and sup- port from the Kansas Cooperative Fish and Wildlife Research Unit.

LITERATURE CITED

Andre-:w, R. J. 1961. The displays given by passerines in courtship and reproductive fighting; a review. Ibis 103:315-348.

Bhi.l, G. R 1990. Birds and mammals on an insect diet: a primer on diet composition analysis in re- lation to ecological energetics. Studies in Avian Biology 13:416-422.

Cavi'H, J. H!, 11. 1998. fhe role of food supply and nest predation in limiting reproductive success of Brown Thrashers Croxostonia rufutn): effects of predator removal, food supplements and predation risk. Ph.D. dissertation, Kansas State University, Manhattan.

Dykstka, C. R. and W. II. Kakasov. 1993. Nesting energetics of House Wrens Cl roylodytes action) in relation to maximal rates of energy flow. Auk I 10: 481-491.

l AKi.t.Y, G. II. 1987. (’omparative breeding strategies of two coexisting passerines: Bell's Vireo (Virco hcllii) and Bewick's Wren ( 1 hrxoniancs Iwwickii). M.Sc. thesis, Kansas State University, Manhattan. llAt.iit’KA, K. 1994. Incubation leeditig in Meadovs Pip

it Anthus pratensis affects female time budget. Journal of Avian Biology 25:251-253.

Johnson, L. S. 1998. House Wren {Troglodytes ae- don). The Birds of North America, no. 380.

Johnson, L. S. and L. H. Kermott. 1989. Territorial intrusions in the House Wren Troglodytes aedon: evidence for the sperm competition hypothesis. Ornis Scandinavica 20:89-92.

Johnson, L. S. and L. H. Kermott. 1992. Why do male House Wrens feed their incubating mates so rarely? American Midland Naturalist 127:200- 203.

Kacelnek, a. 1984. Central place foraging in Starlings (Stunius vulgaris). I. Patch residence time. Journal of Animal Ecology 53:283-299.

Kendeegh, S. C. 1952. Parental care and its evolution in birds. Illinois Biological Monographs, no. 22.

Kennedy, E. D. and D. W. Whete. 1996. Interference competition from House Wrens as a factor in the decline of Bewick’s Wrens. Conservation Biology 10:281-284.

Kennedy, E. D. and D. W. Whete. 1997. Bewick’s Wren (Thryomanes Bewickii). The Birds of North America, no. 315.

Ke.uyver, H. N. 1950. Daily routines of the Great Tit (Paras m. major). Ardea 38:99-135.

Krebs, J. R. 1970. The efficiency of courtship feeding in the Blue Tit Paras caeraleas. Ibis 112:108- 1 10.

Lack, D. 1940. Courtship feeding in birds. Auk 57: 169-178.

Lefjee.d, j. T. and T Se.agsvoet). 1986. The function of courtship feeding during incubation in the Pied Elycatcher Ficedala hypoleaca. Animal Behaviour 34; 1441-1453.

Lefjele), j. T, T. Slagsvoe.E), ane^ G. Stenmark. 1987. Allocation of incubation feeding in a polygynous mating system: a study on pied flycatcheE's Pice- dala hypoleaca. AniEiial Behaviour 35:1663- 1669.

LeTTEE.L, R. C., G. a. MEE.t.EKEiN, W. W. Serocp, ANE) R. D. WoE.EENGER. 1996. SAS system for Eiiixed Eiiodels. SAS Institute, Inc., Cary, NoEth Carolina.

Lyon, B. E. ane) R. D. M()NTC.()Me:ree. 1985. Incuba- tion feeding in Snow Buntings: feEiiale manipu- lation or indiiect Eiiale paiental caie.’ BehavioEal Fxology and Sociobiology 17:279-284.

Lyon, B. E. ane) R. D. Mone(;ome;ree;. 1987. Iwolog- ical coEielates of incubation feeding: a compaia- tive study of high aECtic linches. Ecology 68:713 722.

Mareen, T I:. 1992. hitCEaction of nesE picdati(>n aEul food liEiiitation in icpEoduclivc stiategies. ('lUTcnt Ortiithology 9: 1 63 197.

Mareen, f. I-. 1996. Life history evolution in tidpteal atul south (etnpeiate birds: uhat ilo \se teall\ know ’ .lournal ol A\ ian Biology 27;2(i3 272.

Mareen, f. Ii. \ne) (’. K. Gtt \i aneeeor. I999. Males feeding females tluiing incubation. I. Retiuiieil b\ microclimate or eonstraineil by nest pieilation. American Natuialist I53;I3I 1 39.

30

THE WILSON BULLETIN • Vol. 116, No. 1, March 2004

Miller, E. V. 1941. Behavior of the Bewiek Wren. Condor 43:81-99.

Moreno, J. 1989. Energetic constraints on uniparental incubation in the Wheatear Oenanthe oenanthe (L.). Ardea 77:107-1 15.

Nilsson, J.-A. and H. G. Smith. 1988. Incubation feeding as a tactic for early hatching. Animal Be- haviour 36:641-647.

Nolan, V., Jr. 1958. Anticipatory food-bringing in the Prairie Warbler. Auk 75:263-278.

Ricklefs, R. E. 1974. Energetics of reproduction in birds. Pages 152-297 in Avian Energetics (R. A. Paynter, Ed.). Nutall Ornithological Club, Cam- bridge, Massachusetts.

Royama, T. 1966. A re-interpretation of courtship feeding. Bird Study 13:116-129.

SAS Institute, Inc. 1999. SAS/STAT user’s guide, ver. 8. SAS Institute, Inc., Cary, North Carolina.

Simons, L. S. 1988. Experimental studies in the repro- ductive ecology of Cactus Wrens. Ph.D. disserta- tion, Arizona State University, Tempe.

Skutch, a. E 1949. Do tropical birds rear as many young as they can nourish? Ibis 91:430-458.

Skutch, A. E 1953. How the male bird discovers the nestlings. Ibis 95:1-37.

Smith, H. G., H. Kallander, J. Hultman, and B. Sanzen. 1989. Eemale nutritional state affects the

rate of male incubation feeding in the Pied Ply- catcher Ficedula hypoleuca. Behavioral Ecology and Sociobiology 24:417-420.

Smith, S. M. 1980. Demand behavior: a new interpre- tation of courtship feeding. Condor 82:291-295.

SouKUP, S. S. AND C. E Thompson. 1997a. Social mat- ing system and reproductive success in House Wrens. Behavioral Ecology 9:43-48.

SouKUP, S. S. AND C. P. Thompson. 1997b. Social mat- ing system affects the frequency of extra-pair pa- ternity in house wrens. Animal Behaviour 54: 1089-1 105.

VON Haartman, L. 1958. The incubation rhythm of the female Pied Plycatcher {Ficedula hypoleuca) in the presence and absence of the male. Ornis Pennica 35:71-76.

WOLFINGER, R. AND M. O’CoNNELL. 1993. Generalized linear mixed models: a pseudo-likelihood ap- proach. Journal of Statistical Computation and Simulation 48:233-243.

Zimmerman, J. L. 1993. The birds of Konza: the avian ecology of the tallgrass prairie. University of Kan- sas Press, Lawrence.

ZiOLKOwsKi, D. J., Jr., L. S. Johnson, K. M. Hannam, AND W. A. Searcy. 1997. Coordination of female nest attentiveness with male song output in the cavity-nesting House Wren Troglodytes aedon. Journal of Avian Biology 28:9-14.

Wilson Bulletin, 116(1), 2004, pp. 31^0

RED-COCKADED WOODPECKER NESTLING PROVISIONING AND REPRODUCTION IN TWO DIEFERENT PINE HABITATS

RICHARD R. SCHAEFER,' 2 RICHARD N. CONNER,' D. CRAIG RUDOLPH,' AND

DANIEL SAENZ'

ABSTRACT. — We obtained nestling provisioning and reproductive data from 24 Red-cockaded Woodpecker (Picoides borealis) groups occupying two different pine habitats — longleaf pine (Pinus palustris) and a mixture of loblolly (P. taeda) and shortleaf pine {P. echinata) — in eastern Texas during 1990 and 1991. Habitat data were collected within 800 m of each group’s cavity-tree cluster. Feeding trips per nest and prey biomass per feeding trip were significantly greater in loblolly-shortleaf pine habitat. There were few significant correlations between reproductive/provisioning and habitat variables in either pine habitat. Pines dying from infestation by southern pine beetles (Dendroctonus frontalis) were more common in loblolly-shortleaf than in longleaf pine habitat. In addition, adult male Red-cockaded Woodpeckers weighed more in loblolly-shortleaf pine habitat. Indices of southern pine beetle abundance in loblolly-shortleaf pine habitat were negatively correlated with number of feeding trips per nestling, but positively correlated with prey biomass delivered to nestlings. We hypothesize that the greater abundance of southern pine beetles and associated arthropods in loblolly-shortleaf pine habitat, and the resulting higher frequency of dying pines containing an abundant food source, were as- sociated with an elevated prey biomass available to both nestling and adult Red-cockaded Woodpeckers. Received 29 June 2003, accepted 20 April 2004.

The Red-cockaded Woodpecker {Picoides borealis) is a cooperatively breeding species that lives in family groups of two or more individuals (Ligon 1970, Walters et al. 1988). Groups include a breeding pair, young of the year, and often one to three other adults, which serve as “helpers.” Helpers are usually male offspring from previous nestings and as- sist the breeding pair with caring for nestlings (Ligon 1970, Lennartz and Harlow 1979).

I Red-cockaded Woodpeckers are endangered ' (U.S. Department of Interior 1970) and inhab-

it open, mature pine (Pinus spp.) habitats of the southeastern United States. Populations have become fragmented and isolated due to severe habitat alterations (Costa and Escano 1989, Rudolph and Conner 1994). Cutting of \ old-growth pine forests and elimination of re- curring lire across most of the woodpecker’s I range are major causes of the species’ decline j (Jackson 1971, Lennartz ct al. 1983). Histor- I ically, fire maintained suitable foraging and , nesting habitat. Several studies have reported j positive indirect effects of fire on Red-cock- I

' ' Wildlife Habitat and Silviculture I.ab. (in cooper-

' ation with the Arthur I’emple C’ollege of lorestry, Ste- phen F Austin State Univ.), Southern Research Station, i USDA, b'orest Service, .S06 Hayter St., Nacogdoches. TX 7596,5, USA.

‘Corresponding author; e-mail: rschaeferO I (?Ms.fed.us

aded Woodpecker fitness through increased arthropod abundance (Provencher et al. 1998,

2001) , increased grass and/or forb ground cover (James et al. 1997), and reduced hard- wood midstory vegetation (Walters et al.

2002) .

Red-cockaded Woodpeckers are known to select larger and older pines as foraging sub- strates (Engstrom and Sanders 1997, Zwicker and Walters 1999, Walters et al. 2002); such pines are believed to support more arthropods (Hanula et al. 2000), particularly during the breeding season (Conner et al. 2004). Young pine forests may offer suboptimal foraging habitat by providing a reduced prey base, es- pecially in areas surrounding cavity tree clus- ters (stands of cavity trees occupied by Red- cockaded Woodpecker groups) that have been clearcut or contain dense plantations of young (<30 years) pines. Foraging and provisioning of nestlings may be more difficult in young pine forests, which could have a negative ef- fect on the survivorship of adults and nest- lings (Ligon 1970, 1971).

Logically, jney availability during llie nest- ing season has an impact on Red-cockadcd Woodpecker reproductive success and adult nutrition, rhere is little information regarding comparisons of arthropod tiensities and bio- mass between longleaf pine UJ'nus palustris) and loblolly-shortleaf pine (!\ taedti-P. erlu'n-

31

32

THE WILSON BULLETIN • VoL II 6, No. I, March 2004

ata) habitats. During nesting season, differ- ences in prey availability among habitats dominated by different pine species can im- pact both reproduction and adult nutrition of Red-cockaded Woodpeckers.

The southern pine beetle {Dendroctonus frontalis) is responsible for considerable pine mortality, especially during cyclic epidemics (Conner et al. 2001). Infestations can poten- tially destroy Red-cockaded Woodpecker for- aging habitat and cavity trees. However, dur- ing non-epidemic beetle years, woodpeckers can beneht by concentrating foraging activity on dying pines that provide an arthropod-rich food source (Hooper and Lennartz 1981, Schaefer 1996, Bowman et al. 1997). Such ephemeral food sources, while unpredictable, can provide nutritional benefits to both nest- lings and adults.

Our objectives were to ( 1 ) compare repro- ductive and provisioning effort in longleaf pine and loblolly-shortleaf pine habitats, (2) determine whether habitat variation affected reproduction and nestling provisioning, and (3) use body mass of adults to assess nutri- tional status of birds in longleaf pine and lob- lolly-shortleaf pine habitats.

METHODS

Study areas. — We collected reproductive, nestling provisioning, and vegetation data dur- ing the 1990 and 1991 nesting seasons. Study sites were on the Davy Crockett National For- est (DCNF) and the Angelina National Forest (ANF) in eastern Texas (see Conner and Ru- dolph 1989 for area descriptions). We chose 24 study sites (i.e., 24 woodpecker groups), 8 at DCNF and 16 at ANF. Sites were selected based on the dominant pine species; 1 1 sites were located in longleaf pine and 13 were lo- cated in loblolly-shortleaf pine habitat.

Reproduction and nestling provisioning. — All Red-cockaded Woodpeckers captured at each of the 24 study sites were banded (U.S. Fish and Wildlife Service band and 2-3 color bands) for individual recognition. Birds were visually identified in the field with the aid of binoculars and a 20X spotting scope mounted on a tripod.

Nest monitoring began during the first week of April, about 2 weeks before nesting was expected to commence. If an adult occupied the nest cavity when checked, the tree was

climbed using sectional aluminum ladders; eggs were then counted. If the clutch did not appear complete (normally two to four eggs comprise a complete clutch), it was checked again in a few days. When nestlings were de- tected, the nest tree was again climbed and young were counted and aged (Ligon 1971).

Provisioning data were collected when nest- lings were 8, 20, and 23 days of age. The nest cavity of each woodpecker group was ob- served for a 3-hr period in the morning, be- ginning when the breeding male exited the nest. We recorded identity of the adult bring- ing food, size of each prey item, and time of each feeding. Prey size was visually estimated and categorized as small, medium, or large. An item was considered small if barely visible in the adult’s beak. A medium-sized item was estimated at less than one-half of the beak’s length. A large item was estimated at more than one-half of the beak’s length. We assume that any bias toward larger prey inherent in this procedure was equal among the two pine habitats.

We attempted to obtain a biomass value for each size category. Since it was not possible to collect samples of prey items delivered to nestlings, we collected arthropods similar to those observed being provided in both pine habitats. Samples were obtained from the boles of dead loblolly and shortleaf pines killed by southern pine beetles. These arthro- pods were separated into small, medium, and large size categories using the same criteria used during provisioning observations. We collected 30 individuals of each size category, determined wet weight (mg), and calculated average weight for each size category. Rela- tive values for prey biomass were calculated using the mean weight of each size category (small = 1 1.3 mg, medium = 45.6 mg, large = 197.4 mg).

Vegetation and stand area measure- ments.— Habitat data were collected at each study site within an 800-m radius centered on each woodpecker group’s cluster of cavity trees. Forest compartment stand maps were obtained from the ANF and DCNF district of- fices for those compartments falling within the 800-m radius. Each compartment is comprised of forest stands of varying size. Five dominant or codominant pines were selected within each forest stand within the 800-m radius by

Schaefer et al. • RED-COCKADED WOODPECKER REPRODUCTION AND HABITAT

33

choosing the nearest tree in a random direc- tion from hve arbitrary points well-dispersed within the stand. Habitat measurements were taken within an 11.2-m radius (0. 04-ha cir- cular plot) centered on each of these five trees (Conner 1980), and means were used to char- acterize habitat within the forest stand.

Stand age was determined by coring each central tree at breast height (1.3 m) with an increment borer and counting growth rings of the cores. We added 3 years for loblolly pine and shortleaf pine, and 5 years for longleaf pine to account for growth to breast height (Conner and O’Halloran 1987). Stands were categorized as 0-29, 30-49, 50-69, 70-89, or >90 years old. Tree diameter (cm) was mea- sured at breast height (dbh) with calipers and categorized as 0-30, 30.1-40, 40.1-50, or

50.1- 70 cm. Surrounding canopy height and midstory height (m) were measured with a range finder. Canopy height was placed into categories of 0-12, 12.1-21, 21.1-27, or

27.1- 33 m.

Midstory density was visually estimated and placed into one of five categories: none, sparse, moderate, dense, or very dense. Mid- story conditions were considered suitable if height was <3 m regardless of density, or if density was none to sparse regardless of height. A one-factor metric basal area prism was used to measure basal area (mVha) of pine overstory, hardwood overstory, pine midstory, and hardwood midstory. Pine and hardwood overstory basal areas were placed into categories of 0-9, 9.1-15, 15.1-20, 20.1- 25, or 25.1-30 nT/ha. Pine and hardwood midstory basal areas were categorized as 0-3,

3.1- 6, 6.1-9, or 9.1-12 mVha. The area (ha) of each forest stand within 800 m of each nest tree was measured from compartment stand maps with a digitizer, and the percentage of area occupied by each habitat category cal- culated.

Measurements of southern pine beetle abundance. — Data on southern pine beetle abundance during 1990 and 1991 were ob- tained from the U.S. l orest Service for each forest compartment where study sites were lo- cated. All other causes of mature pine mor- tality were assumed to be ecpial betweem long- leaf and loblolly-shortleaf pine habitats. I hree variables were used as indices of southern pine beetle abundance in comparing beetle ac-

tivity in longleaf pine versus loblolly-shortleaf pine: (1) the number of active beetle spots (one or more contiguous beetle-infested trees), (2) the number of trees infested (dying pines with fading or red needles, and all or most bark remaining), and (3) the number of hect- ares affected by infestation. A total for each variable was calculated for the entire forest compartment, even if only a portion of the compartment fell within the 8()()-m radius cir- cle.

Adult Red-cockaded Woodpecker body mass. — Each adult woodpecker was weighed to the nearest 0.5 g with a lOO-g spring scale. Body mass was obtained throughout the year, except during nesting; each bird was weighed once. Birds were captured either in the morn- ing just before exiting the roost cavity, or in the evening just after entering. We realize there is both seasonal and temporal (24-hr) variability in the body mass of a given indi- vidual. For each of the two pine habitats, body masses were pooled by sex.

Data analysis. — Data were analyzed using SAS (SAS Institute, Inc. 1988). A significance level of P = 0.05 was used in all hypothesis testing. In tests involving habitat variables, stands 0-29 years old, most of which were clearcuts and young pine plantations, were not included in evaluations of available foraging habitat because these stands are considered unsuitable for Red-cockaded Woodpecker for- aging (U.S. Fish and Wildlife Service 2003). However, the 0-29 year stand age category is included for comparative purposes.

The 24 Red-cockaded Woodpecker groups observed produced a total of 37 successful (i.e., one or more fiedglings) nests during the two nesting seasons. For statistical analyses, a 2-year average of each reproductive variable was used for each group to avoid a repealed measures violation. Comparisons of reproduc- tive variables between pine habitats using re- pealed measures analyses were not possible because of instances of small sample sizes withiti years due to some groujis not nesting for various reasons, especially in longleaf pine habitat.

Pearson correlation coefficients were used to explore relationships ol reiirotlucliNc and pro\isioning variables with habitat variables. Iwo-tailed /-tests were used to com|-)are re- produclive pci lormance and provisioning ef-

34

THE WILSON BULLETIN • Vol. 116, No. 1, March 2004

fort between longleaf and loblolly-shortleaf pine. A medium effect size of 0.5 (Cohen 1988) was used in power analyses for statis- tically non-significant variables.

Two-way ANOVAs (pine type X habitat variable) on ranked data were used to compare category distribution of each habitat variable (tree age, diameter at breast height, canopy height, suitable/unsuitable midstory, pine overstory basal area, hardwood overstory bas- al area, pine midstory basal area, and hard- wood midstory basal area) between longleaf (/? = 11) and loblolly-shortleaf {n ~ 13) pine habitats. If the interaction indicated different distributions between the pine habitats, Wil- coxon rank-sum tests were used for each hab- itat variable category to test for differences between longleaf and loblolly-shortleaf pine.

Two-tailed Mests were used to compare southern pine beetle abundance and body mass of adult Red-cockaded Woodpeckers be- tween the pine habitats. Pearson correlation coefficients were used to examine relation- ships between southern pine beetle abun- dance, and provisioning effort and reproduc- tive performance, within each pine habitat. Adult male body mass, including that of both helpers and breeders, was treated separately from adult female body mass due to differing foraging strategies (Ligon 1968, Hooper and Lennartz 1981).

RESULTS

Nesting ejfort in relation to pine habitat.— During the two nesting seasons, 24 Red-cock- aded Woodpecker groups had a total of 37 successful (i.e., one or more fledglings) nests. For various reasons, not all groups success- fully nested. One longleaf clutch was depre- dated and the group did not renest. One lob- lolly-shortleaf group disappeared altogether between years. Eggs at three nests failed to hatch (one in longleaf, two in loblolly-short- leaf). Breeding pairs at five longleaf sites ap- peared to forgo nesting during 1 of the 2 years. Although unlikely, some clutches may have been initiated and then depredated im- mediately before we detected them. If so, the birds did not appear to renest.

Twelve (80%) of 15 successful nests in longleaf pine habitat lacked helpers, and only one helper was present at the remaining 3 (20%) nests. Eleven (50%) of 22 successful

nests in loblolly-shortleaf pine habitat lacked helpers. Of the remaining 11 nests, 10 (45%) had one helper and 1 (5%) had two helpers (one male and one female). For all 24 groups (2 years combined) the average number of helpers per group was 0.4.

During the two nesting seasons, 17 clutches were produced in longleaf pine and 24 in lob- lolly-shortleaf pine. Clutch size was not re- corded in two instances, once in each habitat. The remaining 16 longleaf nests produced a total of 51 eggs {x = 3.19 eggs/clutch), and 23 loblolly-shortleaf nests produced a total of 78 eggs {x = 3.39 eggs/clutch). Hatching suc- cess based on clutch size was 75% in longleaf pine and 87.3% in loblolly-shortleaf. Hatching success, as measured by the number of nest- lings hatched from eggs surviving through the incubation period, was 85.7% (36 nestlings from 42 eggs; /? = 13 nests) in longleaf habitat and 89.9% (62 nestlings from 69 eggs; n = 20 nests) in loblolly-shortleaf habitat. Two clutches in each pine habitat failed to hatch, leaving a total of 15 and 22 broods produced in longleaf and loblolly-shortleaf, respective- ly. The 15 broods in longleaf produced 24 fledglings (T = 1.60 fledglings/nest), and the 22 in loblolly-shortleaf produced 42 fledglings (T = 1.91 fledglings/nest). The initial number of nestlings could not be counted for two broods in each pine habitat. Fledging success subsequent to hatching was 55.6% for the re- maining 13 broods in longleaf, and 62.9% for the remaining 20 broods in loblolly-shortleaf habitat.

Considering only woodpecker groups that produced one or more fledglings, all nest pro- ductivity measures (with the exception of par- tial brood loss) and number of adults were slightly higher in loblolly-shortleaf than in longleaf pine habitat; only feeding trips per nest and prey biomass per feeding trip were statistically greater (Table 1). Power analyses revealed that sample sizes in each pine habitat were too small to detect biological signifi- cance (medium size effect of 0.5, power = 0.2) for statistically non-significant variables.

A two-way ANOVA was calculated to eval- uate the contribution of group size to the num- ber of feeding trips to nests in loblolly-short- leaf and longleaf pine habitats. There was no significant interaction (F2 i6 = 0.43, P — 0.66) between group size and pine habitat in relation

Schaefer et al. • RED-COCKADED WOODPECKER REPRODUCTION AND HABITAT

35

TABLE 1. Red-cockaded Woodpecker nesting effort (mean : shortleaf {n = 13) pine habitats in eastern Texas during 1990 and 1	t SD) in longleaf (n = 1991."’	10)*" and loblolly-
Nesting variable	Longleaf	Loblolly-shortleaf	t	p
Group size	2.3 ± 0.4	2.5	± 0.5	1.43	0.17
Clutch size	3.3 ± 0.6	3.5	± 0.5	0.91	0.37
Initial brood size^	2.9 ± 0.7	3.0	± 0.6	0.53	0.61
Brood size'-’	1.7 ± 0.5	2.0	± 0.4	1.68	0.1 1
Feedings per nesE	31.4 ± 9.7	43.3	± 11.3	2.66	0.015
Feedings per nestling"’	19.7 ± 7.1	22.7	± 5.4	1.15	0.26
Prey biomass per trip (mg)""	73.2 ± 31.0	1 16.7	± 36.3	3.03	0.006
Number of fledglings	1.7 ± 0.5	1.9	± 0.4	1.41	0.17
Partial brood loss‘s*	0.3 ± 0.2	0.2	± 0.2	0.59	0.56

“ Only groups that produced at least one fledgling are included in these analyses.

Sample size = 10 because one longleaf pine group did not produce at least one fledgling during either year.

The total number of nests producing fledglings was 15 at longleaf pine and 22 at loblolly-shortleaf pine sites over the two nesting seasons. The 2 years were averaged for each nesting variable, for each group.

Initial brood size and partial brood loss are each unknown in one instance in longleaf pine habitat. Therefore /z = 9 in longleaf for these two nesting variables.

® Means based upon observations at 8, 20, and 23 days of nestling age at each nest.

Nestling loss between hatching and day 8.

to the number of feeding trips per nest, and group size alone did not influence {F^ k, = 0.41, P = 0.75) number of feeding trips per nest. These results indicate that differences between the two pine habitats, and not group size, were responsible for the greater number of feeding trips made to nests in loblolly- shortleaf pine habitat.

Mean number of feeding trips per nest was significantly greater in loblolly-shortleaf pine habitat, but mean number of feeding trips per nestling was similar, indicating that individual nestlings were fed at about the same frequen- cy in both pine habitats (Table 1 ). However, average prey biomass per feeding trip was sig- nificantly greater in loblolly-shortleaf than in longleaf pine (Table 1), indicating that nest- lings in the former received more food. There were few significant correlations among re- productive/provisioning variables and habitat variables in either pine habitat. Of note was the lack of significant relationships between any of the habitat variables and prey biomass within either pine habitat. Thus, the habitat variables we measured had little or no rela- tionship with size of prey items delivered to nestlings.

Comparison of longleaf and lohloUy-short- leaf pine habitats. — I'he percentage of area occupied by forest stands <30 years old was greater in loblolly-shortleaf than in longleaf pine habitat (Z = -3.22, = O.OOl; E'ig. lA).

This was the result of extensive clear-cutting that occurred during the 1970s and 1980s, as

well as southern pine beetle control cuts in loblolly-shortleaf pine study sites. Cutting sel- dom occurred in longleaf pine study sites. Forest stands in the 30-49 year (Z = 2.71, P = 0.007) and 50-69 year (Z = 2.12, P 0.034 ) age categories occupied more area in longleaf pine habitat, whereas stands in the 70-89 year (Z = —3.62, P < 0.001) age cat- egory occupied more area in loblolly-shortleaf pine habitat (Fig. lA). There was no differ- ence between pine habitats in the percentage of area occupied by the 90-120 year age cat- egory (Z - -1.50, P = 0.13; Fig. lA). This oldest stand-age category constituted only a small percentage of area within the 8()()-m ra- dius in both pine habitats.

Loblolly-shortleaf pine contained a higher frequency of stands in the largest dbh cate- gory of 50.1-70 cm (Z = -2.78, P = 0.006) and highest canopy height category of 27.1- 33 m (Z = -3.72, P < 0.001) than did long- leaf pine (Figs. IB and 1C). Conversely, the smaller dbh category of 30.1-40 cm (Z = 3.63, P < 0.001) and shorter canopy height categories of 12.1-21 m (Z = 2.11. =

0.035) and 21.1-27 m (Z = 2.32, P = 0.021) were more common in longleaf pine (lugs. IB and 1C).

Comparison of midstory botvseen pine hab- itats revealed that the percentage of area \\ ith suitable midstory conditions was greater in longleaf pine (Z = 3.74, P < 0.001) aiul the percentage of area occupietl by unsuitable midstory coiulitions was greater in loblolly-

36

I'HE WILSON BULLETIN • VoL 1 16, No. 1, March 2004

A

70 1 60 - 50 -

Tree age (years)

■ 0-29

□ 30-49

□ 50-69 H 70-89

□ 90-120

40

Longleaf Loblolly-shortleaf

Diameter at breast height (cm)

Canopy height (m)

D

70 1 60 - 50 - 40 - 30 - 20 - 10 - 0 —

Longleaf Loblolly-stx)rtleaf

Midstory

■ Suitable

□ Unsuitable

□ Stands <30 years

EIG. 1. Mean percentage of area within 800 m of Red-cockaded Woodpecker ca\ ity-tree clusters occupied by each category of (A) tree age, (B) diameter at breast height, (C) canopy height, and (D) midstory in longleaf pine (n = 11) and loblolly-shortleaf pine (n = 13) sites in eastern Texas, 1990- 1991.

shortleaf pine (Z = -2.17, P = ().()3(); Fig. ID). When habitat of all ages (i.e., including stands <30 years old) within 800 m of wood- pecker nest trees was considered, the average percentage of area with unsuitable midstory was 51% for longleaf and 93% for loblolly- shortleaf pine.

Pine overstory basal area was similar be- tween pine habitats with the exception of the 20.1—25 m^/ha category, which occupied a greater percentage of area in longleaf pine (Z = 2.62, P = 0.009; Fig. 2A). Trees in the 0- 9 m-/ha hardwood overstory basal area cate- gory occupied a greater percentage of area in longleaf than in loblolly-shortleaf pine (Z — 3.05, P = 0.002). Few forest stands containing overstory hardwoods were within any basal area category greater than 0-9 m^/ha in either pine habitat (Fig. 2B).

No significant differences were found in any pine midstory basal area category be- tween the two pine habitats (F3 gg == 1.96, P

= 0.13; Fig. 2C). The percentage of area oc- cupied by the relatively low hardwood mid- story basal area category of 0-3 mVha was greater in longleaf pine (Z = 3.97, P < 0.001 ). The percentage of area occupied by the great- er hardwood midstory basal area categories of 3.1-6 nF/ha (Z = -2.89, P = 0.004), 6.1-9 m-/ha (Z = -2.13, P = 0.033) and 9.1-12 nW ha (Z = —1.96, P = 0.050; Fig. 2D) were all greater in loblolly-shortleaf pine.

Southern pine beetle influence. — The num- ber of active beetle spots, beetle-infested trees, and total hectares infested with beetles were all significantly greater in loblolly-shortleaf pine habitat (Table 2). At loblolly-shortleaf nests in which at least one fledgling was pro- duced {n — 22), number of active beetle spots (r = 0.48, P = 0.022), beetle trees (r = 0.45, P = 0.036), and infested hectares {r = 0.67, P < 0.001) were positively conelated with prey biomass delivered to nestlings. Number of beetle spots (a* = —0.57, P — 0.006), beetle

Schaefer et al. • RED-COCKADED WOODPECKER REPRODUCTION AND HABITAT

37

A

Pine overstory (m^/ha)

B

Hardwood overstory (m^/ha)

Long leaf Loblolly-shortleaf

Hardwood midstory (m7ha)

PIG. 2. Mean percentage of area within 800 in of Red-cockaded Woodpecker cavity-tree clusters occupied by each category of (A) pine overstory basal area, (B) hardwood overstory basal area, (C) pine midstory basal area, and (D) hardwood midstory basal area in longleaf pine {n = 11) and loblolly-shortleaf pine {n = 13) sites in eastern Texas, 1990-1991.

trees (r = -0.60, P = 0.003), and infested hectares (r = —0.51, P — 0.016) were nega- tively correlated with number of feeding trips per nestling. No significant correlations were found between indices of beetle abundance and the remaining reproductive and provision- ing variables. At longleaf nests in which at least one fledgling was produced (/? = 15), no significant correlations were found between indices of beetle abundance and any of the reproductive and provisioning variables.

Adult nutritional status. — Body mass was used to compare separately the nutritional sta- tus of adult male and female Red-cockaded Woodpeckers in longleaf and loblolly-short- Icaf pine habitats. Body mass of adult males was significantly greater (/ = —2.25, =

().()30) in loblolly-shortleaf (.V = 48.5 g ± 2.3 vSD, n = 27) than in longleaf pine (.v = 46.9 g ± 2.7 SI), n = 18). Adult females averaged only slightly heav ier in loblolly-shortleaf (.v = 46.6 g ± 2.2 SI), n = 17) than in longleaf

pine (a' = 45.3 g ± 2.0 SD, n — 13), and the difference was not statistically significant (/ == -1.59, P = 0.12).

DISCUSSION

Canopy trees in loblolly-shortleaf pine hab- itat were generally older, taller, and larger in diameter than in longleaf pine. Suitable mid- story conditions for Red-cockaded Woodpeck- ers were more widespread in longleaf than in loblolly-shortleaf pine. .Soil-type differences and more effective prescribed burning in long- leaf pine areas had a strong influence on dif- ferences in midstory condition between the two pine habitats (Conner and Rudolph 1989). Red-cockaded Woodpeckers are known to have an aversion to a well-developed stratum of midstory vegetation associated with both nesting (C'onner and Rudolph 1989, Loch et al. 1992) and foraging habitat (Rudolph et al. 2002; Walters et "al. 2()()(), 2002). rhus. it might be expected that nest productivity of

38

THE WILSON BULLETIN • Vol. 116, No. I, March 2004

TABLE 2. Southern pine beetle abundance (mean ± 25) pine sites in eastern Texas during 1990 and 1991.“	SD) at longleaf {n	= 22) and loblolly-shortleaf {n =
Variable	Longleaf	Loblolly-shortleaf	t	p
Number of beetle spots	0.32 ± 0.57	2.80 ± 2.80	4.33	<0.001
Number of infested trees	2.91 ± 5.13	65.32 ± 65.58	4.74	<0.001
Number of infested hectares	0.01 ± 0.04	0.63 ± 0.80	3.84	<0.001

The 2 years were not combined for southern pine beetle analyses due to the potential for substantial year-to-year changes in beetle abundance indices.

woodpeckers in habitat with an abundance of midstory vegetation (i.e., loblolly-shortleaf pine) would be lower than in longleaf pine.

Despite less suitable midstory conditions in loblolly-shortleaf pine habitat, woodpecker groups there performed at least as well repro- ductively as groups in longleaf pine, but only feeding trips per nest and relative prey bio- mass delivered to nestlings were significantly greater in the former. Our sample sizes were too small to detect biologically significant dif- ferences between pine habitats for the remain- ing reproductive and provisioning variables.

Helpers were more common in loblolly- shortleaf groups, but only once was there >1 per group. Other studies indicate that groups with helpers fledge significantly more young than groups without helpers (Lennartz et al. 1987, Walters 1990). In this study, increased group size did not significantly influence the number of feeding trips per nest even though helpers assisted with nestling provisioning. However, helpers may enhance reproductive success by assisting with incubation, brooding and feeding nestlings, territory defense, and defense against predators.

The relative biomass of arthropod prey de- livered to nestlings was significantly greater in loblolly-shortleaf than longleaf pine habitat. At those loblolly-shortleaf sites where south- ern pine beetles were more abundant, adult Red-cockaded Woodpeckers made fewer feed- ing trips per nestling but delivered larger prey items. Access to larger prey items may benefit adults by reducing nestling provisioning ef- fort.

The smallest mean for provisioned biomass per feeding trip (76.9 mg) for any nest in lob- lolly-shortleaf habitat was greater than that for 9 of the 15 nests in longleaf habitat. We know from field observations that adults from at least three of the six nests in longleaf habitat with large values for mean prey biomass per

feeding trip had access to one or more (exact number unknown) nearby dying pines. These trees were often loblolly pines located on wet- ter sites (i.e., streams or baygalls) within long- leaf pine habitat, and were dying from either lightning strikes or southern pine beetle infes- tations. During provisioning observations, we noticed adults spending considerable time traveling between the direction of the dying pines and the nest. Thus, the high values of biomass provisioned to nestlings appear to be at least partially dependent on the local avail- ability of dying pines that have an abundant supply of arthropod prey. A great number of arthropod species are attracted to such dying pines, which provide an abundance of food for Red-cockaded Woodpeckers (Ligon 1968, Hooper and Lennartz 1981, Conner et al. 2001).

Adult and larval southern pine beetles are fairly small prey items for Red-cockaded Woodpeckers. However, the adults and larvae of larger wood boring beetles (e.g., Ceram- bycidae and Buprestidae), which are attracted to pines infested by southern pine beetles, pro- vide much larger prey items for foraging woodpeckers. Red-cockaded Woodpeckers have been observed to forage for as long as 55 min on small groups of dying pines in- fested with arthropods before moving on to a healthy tree (Schaefer 1996). Dying pines pro- vide an important food source for Red-cock- aded Woodpeckers throughout the year, par- ticularly during the nesting season when young woodpeckers are being fed.

We suggest that the greater abundance of southern pine beetles in loblolly-shortleaf pine habitat and the resulting higher frequency of dying pines containing a diverse and abundant arthropod community are associated with el- evated prey biomass. Dying pines were com- paratively rare in longleaf pine habitat be- cause this species is more resistant to southern

Schaefer et al. • RED-COCKADED WOODPECKER REPRODUCTION AND HABITAT

39

pine beetle infestation; this is due to its ability to produce copious amounts of resin and to the different physical properties of its resin (Hodges et al. 1979). Increased prey avail- ability, in terms of biomass, is one indication of increased territory quality. Thus, the quality of foraging habitat at our loblolly-shortleaf pine study sites was greater than that at long- leaf pine sites. That adult male Red-cockaded Woodpeckers weighed more in loblolly-short- leaf pine habitat suggests, at least in eastern Texas, that they are nutritionally more fit than I those in longleaf pine habitat.

The abundant food source available to Red- I cockaded Woodpeckers in dying pines is tran- I sient. During epidemic years southern pine beetles can devastate large areas of pine for- est, including Red-cockaded Woodpecker for- aging habitat and entire cavity-tree clusters.

! However, during non-epidemic years, when southern pine beetle attacks are confined to single trees or small groups of pines, prey j availability may increase for Red-cockaded ' Woodpeckers. Territory quality influenced by ! the presence of ephemeral southern pine bee- j tie infestations will fluctuate and can be un- , predictable.

j ACKNOWLEDGMENTS

We thank C. K. Evans, R Fenci, S. W. Lower, C. E. Shackelford, and M. Watson for assistance with field work, and C. E. Braun, R. T Engstrom, E C. James, B. R. Parresol, E J. Sanders, J. R. Walters and an anon- ymous reviewer for constructive comments on an early draft of the paper. We also thank N. E. Koerth for i statistical and editorial assistance.

LITERATURE CITED

Bowman, R., D. E. Leonard, L. M. Richman, and L.

I K. Backus. 1997. Demography of the Red-cock-

I aded Woodpecker at the Avon Park Air Force

Range. Report Number F()86()2-96-D()01 5. Arch- i bold Biological Station, Lake Placid, Florida.

Cohen, J. 1988. Statistical power analysis for the be- havioral sciences, 2nd ed. Lawrence Erlbaum As- sociates, Hillsdale. New Jersey.

Conner, R. N. 1980. Foraging habitats of woodpeck- ers in southwestern Virginia. Journal of Field Or- nithology .^1:1 19-127.

j CoNNtiR, R. N.. C. S. C7)I.i.ins, D. Saen/, T. Triu s, R. R. ScHAEt-ER, AND D. C. RuDoi.i'H. 2004. Arthro- pod density and biomass in longleaf pines: effects I of pine age and hardwood midstory, in Retl-cock-

) aded Woodpecker symposium IV (R. (’osta anti .S.

J. Daniels, lids.). Hancock Mouse Publishers,

I Blaine, Washington. In press.

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 Bulletin 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. Research Paper SO-250. U.S. Forest Ser- vice, Southern Forest Experiment Station, New Orleans, Louisiana.

Conner, R. N., D. C. Rudolph, and J. R. Walters. 2001. The Red-cockaded Woodpecker: surviving in a fire-maintained ecosystem. University of Tex- as Press, Austin.

Costa, R. and R. E. E Escano. 1989. Red-cockaded Woodpecker: status and management in the south- ern region in 1986. Technical Publication R8-TP- 12. U.S. Forest Service, Threatened and Endan- gered Species Program, Southern Region, Atlanta, Georgia.

Engstrom, R. T. and E J. Sanders. 1997. Red-cock- aded Woodpecker foraging ecology in an old- growth longleaf pine forest. Wilson Bulletin 109: 203-217.

Hanula, j. L., K. E. Franzreb, and W. D. Pepper. 2000. Longleaf pine characteristics associated with arthropods available for Red-cockaded Woodpeckers. Journal of Wildlife Management 64:60-70.

Hodges, J. D., W. W. Elam, W. E Watson, and T. E. Nebeker. 1979. Oleoresin characteristics and sus- ceptibility of four southern pines to southern pine beetle (Coleoptera: Scolytidae) attacks. Canadian Entomologist 111:889-896.

Hooper, R. G. and M. R. Lennartz. 1981. Foraging behavior of the Red-cockaded 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. Pages 4-29 in The ecology and management of the Red-cockaded Woodpecker: proceedings of a symposium at Oke- fenokee National Wildlife Refuge (R. L. Thomp- son, Ed.). Bureau of Sport Fisheries and Wildlife and Tall Timbers Research Station. Tallahassee. Florida.

James, E C., C. A. Hess, and D. Kuerin. 1997. Spe- cies-centered environtnental analysis: indirect ef- fects of (ire history on Red-cockatletl Woodpeck- ers. Ecological Applicatitms 7:1 18 129.

Lennar rz. M. R. and R. F! IlARtow. 1979. Fho role of parent atid helper Retl-cockadeil Woodpeckers at the nest. Wilson Bulletin 91:331 33.S.

LflNNARtZ, M. R.. R. G. llOOITR, \ND R. F; II\RL()V\. 1987. Sociality anti C(n>perative breeding tfi Red- cockadetl Woodpeckers U’iioides horealis). Be- ha\ioral Ideology and SociobioUigy 20:77 88.

I.I NNARIZ. M. R.. II. A. KmGHI. .1. P. Mt C’ll RI . \ND V. A. Rt Dis. 1983. .Status ol Red-cockadetl Wotni- peeker nesting habitat iti the south. Pages 13 19 in Red-coekadetl WotKipecker symposium II: pro-

40

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

ceedings (D. A. Wood, Ed.). Elorida Game and Fresh Water Fish Commission, Tallahassee.

Ligon, J. D. 1968. Sexual differences in foraging be- havior in two species of Demlrocopus woodpeck- ers. Auk 85:203-215.

Ligon, J. D. 1970. Behavior and breeding biology of the Red-cockaded Woodpecker. Auk 87:255-278.

Ligon, J. D. 1971. Some factors influencing numbers of the Red-cockaded Woodpecker. Pages 30-43 in The ecology and management of the Red-cock- aded Woodpecker: proceedings of a symposium at Okefenokee National Wildlife Refuge (R. L. Thompson, Ed.). Bureau of Sport Fisheries and Wildlife and Tall Timbers Research Station, Tal- lahassee, Florida.

Loeb, S. C., W. D. Pepper, and A. T. Doyle. 1992. Habitat characteristics of active and abandoned Red-cockaded Woodpecker colonies. Southern Journal of Applied Forestry 16:120-125.

Provencher, L., K. E. M. Galley, B. J. Herring, J. Sheehan, N. M. Gobris, D. R. Gordon, G. W. Tanner, J. L. Hardesty, H. L. Rodgers, J. P. Mc- Adoo, M. N. Northrop, S. J. McAdoo, and L. A. Brennan. 1998. Post-treatment analysis of res- toration effects on soils, plants, arthropods, and birds in sandhill systems at Eglin Air Force Base, Florida. Annual report. Natural Resources Divi- sion, Eglin Air Force Base, Niceville, Florida.

Provencher, L., A. R. Litt, K. E. M. Galley, D. R. Gordon, G. W. Tanner, L. A. Brennan, N. M. Gobris, S. J. McAdoo, J. P. McAdoo, and B. J. Herring. 2001. Restoration of fire-suppressed longleaf pine sandhills at Eglin Air Force Base, Florida. Final report. Natural Resources Division, Eglin Air Force Base, Niceville, Florida.

Rudolph, D. C. and R. N. Conner. 1994. Forest frag- mentation and Red-cockaded Woodpecker popu- lation: an analysis at intermediate scale. Journal of Field Ornithology 65:365-375.

Rudolph, D. C., R. N. Conner, and R. R. Schaefer. 2002. Red-cockaded Woodpecker foraging behav-

ior in relation to midstory vegetation. Wilson Bul- letin 114:235-242.

SAS Institute, Inc. 1988. SAS/STAT user’s guide: re- lease 6.03. SAS Institute Inc., Cary, North Caro- lina.

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 University, Nacogdoches, Texas.

U.S. Department of Interior. 1970. Listing of Red- cockaded Woodpecker as endangered. Federal Register 35: 1 6047.

U.S. Fish and Wildlife Service. 2003. Recovery plan for the Red-cockaded Woodpecker (Picoides bo- realis): second revision. U.S. Fish and Wildlife Service, Atlanta, Georgia.

Walters, J. R. 1990. Red-cockaded Woodpeckers: a ‘primitive’ cooperative breeder. Pages 67-101 in Cooperative breeding in birds: long-term studies of ecology and behavior (P. B. Stacey and W. D. Koenig, Eds.). Cambridge University Press, Unit- ed Kingdom.

Walters, J. R., S. J. Daniels, J. H. Carter, III, and P. D. Doerr. 2002. Defining quality of Red-cock- aded Woodpecker foraging habitat based on hab- itat use and fitness. Journal of Wildlife Manage- ment 66:1064-1082.

Walters, J. R., S. J. Daniels, J. H. Carter, III, P. D. Doerr, K. Brust, and J. M. Mitchell. 2000. For- aging habitat resources, preferences and fitness of Red-cockaded Woodpeckers in the North Carolina sandhills. Final report. Fort Bragg project. Virgin- ia Polytechnic Institute and State University, Blacksburg, and North Carolina State University, Raleigh.

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. AND J. R. WALTERS. 1999. Selection of pines for foraging by Red-cockaded Wood- peckers. Journal of Wildlife Management 63:843- 852.

Wilson Bulletin, 116(1), 2004, pp. 41-47

EFFECTS OF NEST PREDATION AND BROOD PARASITISM ON POPULATION VIABILITY OF WILSON’S WARBLERS IN COASTAL CALIFORNIA

JENNIFER C. MICHAUD,'-’ THOMAS GARDALE^-* * NADAV NUR,^ AND

DEREK J. GIRMAN'

ABSTRACT. — We studied the consequences of nest predation and brood parasitism on a population of Wil- son’s Warblers {Wilsonia pusilla) breeding in coastal riparian woodlands in northern California. We monitored 90 warbler nests from 1997 to 2000; only 16 of these produced Wilson’s Warbler young. Of 74 failed nests, 73% (54/74) failed due to nest predation. Overall, 33% (30/90) of the nests were parasitized by Brown-headed Cowbirds {Molothrus ater). Nest success, as calculated by the Mayheld method, was 0.085 and notably lower than values reported for other warbler species. We used a simple demographic population model — under sce- narios of high, average, and low productivity and survival — to evaluate the viability of this population and found it to be at risk of local extirpation without immigration. This was due to the combined effects of high levels of nest predation and the impacts of brood parasitism. Received 16 June 2003, accepted 6 April 2004.

Across their range, breeding populations of Wilson’s Warbler {Wilsonia pusilla) have been declining at both regional and local scales over the past few decades (Ammon and Gil- bert 1999, Sauer et al. 2001). According to Breeding Bird Survey data, Wilson’s Warblers across the North American continent have been declining on average 2.0% per year dur- ing the period 1980 to 2000, and populations along the Pacihc coast have been declining on average 1.8% per year over the same time pe- riod (Sauer et al. 2001 ). In contrast, data from a single site in coastal California indicate that the breeding population there is stable (Chase et al. 1997).

Population declines in breeding songbirds have been attributed to a variety of factors, including, but not limited to, loss, degrada- tion, and fragmentation of habitat and asso- ciated factors that affect reproductive success and survival. There is evidence that the Wil- son’s Warbler population in coastal California is regulated primarily by breeding productiv- ity (Chase et al. 1997); however, the factors that limit productivity are unknown. The lead- ing causes of low reproductive success in songbirds are nest predation by vertebrate

I ' Dept, of tbology. Sonoma State Univ., 1801 12 C'o- I tati Ave., Rohnert I’ark, C’A 94928. USA.

I ' Point Reyes tbrcl Observatory. 4990 Slioreline Hwy., Stinson Beach. (^A 94970. USA.

* Current aticiress: Prunuske ('hatham. Inc.. P.O. Box 828, Occiilental. ('A 95465. USA.

' C'orrespoiuling author; e-mail; tgartlaliC” prbo.org

predators and brood parasitism by Brown- headed Cowbirds {Molothrus ater, Britting- ham and Temple 1983; Martin 1992a, 1992b). High levels of nest predation and brood par- asitism have been implicated in the decline of many songbird populations by directly affect- ing productivity and, ultimately, population dynamics (e.g.. Pease and Grzybowski 1995).

While there have been few studies pub- lished documenting the breeding ecology and life history characteristics of western popula- tions of Wilson’s Warbler (Stewart 1973, Stewart et al. 1977, Ammon and Gilbert 1999), little work has been done to explore causes of recent declines and, more specifi- cally, factors limiting reproductive success. Population declines in the past have been at- tributed to loss and degradation of riparian breeding habitat (Ammon and Gilbert 1999). However, few estimates of reproductive suc- cess exist and, to our knowledge, no Mayfield ( 1975) estimates of nest success have been re- ported. There are even fewer accounts of cow- bird parasitism and its effects on reproductive success of Wilson's Warblers.

In this study, we report on the breeding bi- ology and population viability of a coastal population of Wilson's Warblers breeding in Marin County, C'alifornia. Our objectiNcs were to ( I ) examine the effects of cowbird parasitism and nest predation on warbler re- productive success, and (2) (.Icxelop a simple demographic population model to assess the viability of this local population.

41

42

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

METHODS

Study areas. — Our study was conducted in the Golden Gate National Recreation Area (GGNRA) in coastal Marin County, Califor- nia, just north of the San Francisco Bay area. Fieldwork occurred from mid-April to early August 1997 to 2000 along two riparian woodlands, Fagunitas Creek (38° 02' N, 122° 45' W) and Redwood Creek (37° 51' N, 122° 34' W). The Fagunitas Creek site contained two plots and the Redwood Creek site three, including Muir Beach. Nest monitoring at Muir Beach was conducted only from 1997 to 1999. Each study plot was approximately 3.6 ha in size.

All study sites were similar in vegetation type and typical of riparian communities in the sun'ounding area. Red alder {Alnus rubra) and willow (Sali.x spp.) dominated the sites with lesser amounts of box elder {Acer ne- guudo), California bay (Umbellularia califor- nica), California buckeye (Aesculus califor- nica), and coast live oak (Quercus agrifolia). Understory species consisted primarily of Cal- ifornia blackberry {Rubus ursinus), Himilayan blackberry (R. discolor), poison oak {Toxico- dendron diversilobum), and fern species.

The areas surrounding the study sites were largely oak-bay woodlands and coastal scrub. At Fagunitas Creek, there was light livestock grazing in fields adjacent to the creek and our plots. We chose to treat the Muir Beach plot separately (even though it is part of Redwood Creek) because (1) it is divided by a road; (2) there is a residential community (—150 homes), small horse stable, and a tavern im- mediately adjacent; (3) there is a public picnic area and 175-car parking lot within and ad- jacent; (4) it is a heavily used recreational area with over 400,000 visitors per year (National Park Service unpubl. data); and (5) it is the only plot where unsupervised domestic dogs {Can is doniesticus) and cats {Felis domes ti- cus) were seen.

Nest searching and monitoring. — Wilson’s Warbler nests were located and monitored us- ing guidelines described by Martin and Geu- pel (1993). We located nests by observing pa- rental behavior and systematically searching the vegetation. Nests were monitored every 1— 4 days until nest failure or fledging. A nest was considered successful if it fledged at least

one warbler young. Fledging was assigned based on the condition of the nest (e.g., matted rim and/or fecal matter and no signs of dep- redation, stage of the nesting cycle), and/or evidence of fledglings within close proximity of the nest near the expected fledging date. We considered a nest to have failed if it was aban- doned or depredated (disappearance of nest contents) prior to the expected fledging date.

Nests were considered parasitized if at any stage in the nesting cycle they contained a cowbird egg or nestling. We considered nests to have failed from cowbird parasitism if only cowbird eggs or nestlings were observed in the nest during all observations, nests were abandoned during egg laying and cowbird eggs were present, or if only cowbird eggs or nestlings were present in the nest after warbler eggs or young were observed. Parasitized nests were considered successful if warbler young fledged from the nest.

Reproductive success. — Nest survival prob- abilities were calculated using the Mayfield (1975) method with the standard error esti- mator developed by Johnson (1979). The Mayfield method is based on nest losses di- vided by the total number of days nests were observed and, thus, at risk of failure. Survival probabilities were calculated for each stage of nesting (egg laying, incubation, and nestling) and for the entire nesting period. Estimates were based on a 26-day nesting period (4 egg- laying days, 12 incubation days, and 10 nest- ling days) as determined by our nest monitor- ing data. When calculating “exposure” days (the total number of observation days) for nests with known fates, we used the midpoint between the last observed active date and the first observed inactive date; for nests with un- known fates, the last active date was used for counting exposure days (Last-Active B meth- od in Manolis et al. 2000). We compared dif- ferences in nest success probabilities among nesting stages, study sites, and between par- asitized and unparasitized nests with Program CONTRAST (Sauer and Williams 1989).

We also calculated the following compo- nents of reproductive success: clutch size, clutch-initiation date, hatching success, nest- ling number, fledging success, and fledgling number. These were calculated for both cow- birds and warblers separately, with the excep- tion of clutch initiation. Clutch size was based

Michaud et al. • REPRODUCTIVE SUCCESS OE WILSON’S WARBLERS

43

on the maximum number of eggs present throughout egg laying. Clutch-initiation dates were estimated based on the first egg laid for a nesting attempt or backdated to calculate when the first egg was laid. Hatching and fledging success were defined as the total number of nestlings and fledglings, respec- tively, divided by clutch size. The maximum number of young observed between hatching and fledging was the nestling number. Fledg- ling number was based on the number of nest- lings seen during the last nest check prior to the estimated fledging date for successful nests. Comparisons between parasitized and unparasitized nests were made for all com- ponents of reproductive success.

Population trajectory. — To evaluate popu- lation viability, we developed a simple de- mographic population model following Pul- liam (1988) and Donovan et al. (1995). We calculated lambda (A.) values using the follow- ing equation:

X = Pa + PjB

where X is the finite rate of increase. Pa rep- resents adult survival, Pj represents juvenile survival from fledging to the first breeding season, and B is a measure of productivity representing the number of female offspring produced per year. This last component is composed of three sub-components: the num- ber of nesting attempts X probability a nesting attempt is successful X number of female young produced per successful nest; the sec- ond sub-component incorporates our Mayfield estimates of nest success. PjB is a measure of recruitment rate based on the number of new female recruits produced per year.

Under this model, which calculates a finite (annual) rate of increase, the population is considered a sink (X < 1) if Juvenile recruit- ment is less than adult mortality. Conversely, if juvenile recruitment (i.e., PjB) is greater than adult mortality it is considered a source (X > 1 ), and if the two are equivalent the pop- ulation is stable (X = 1; Pulliam 1988). For our calculations, we used = 0.503 (± 0.035 SK, 95% CT = 0.435-0.571), the adult survival estimate from coastal Marin County (Chase et al. 1997). Because direct estimates of juvenile survival arc largely unknown (Gar- dali et al. 2003), we used a conservative es- timate of juvenile survival and assumed it was

70% that of adults (0.352; Powell et al. 1999, Perkins and Vickery 2001). Based on our field data, the number of female offspring per suc- cessful nest was 1.7. The formula was evalu- ated using two, three, and four nesting at- tempts due to the variability of the number of nesting attempts per season across the war- blers’ range (Ammon and Gilbert 1999). Ad- ditionally, we used our overall nest survival estimate and its 95% confidence interval, and juvenile and adult annual survival to evaluate the population under the observed average, best-, and worst-case scenarios.

Statistical analysis. — Nominal logistic re- gression was used to test for differences in predation and parasitism frequency between sites (Nur et al. 1999). Comparisons of pre- dation rates and nest abandonment between parasitized and unparasitized nests were made using the Pearson chi-square test. We used or- dinal logistic regression to determine whether clutch size, hatching success, nestling number, fledging success, and fledgling number were significantly different between parasitized and unparasitized nests. Due to small sample size, data were combined across years and we ex- cluded those nests that were abandoned or depredated prior to the end of egg laying (// = 14). All statistical analyses were performed using JMP software (Sail and Lehman 1996) and means are presented as ± SE.

RESULTS

We located and monitored 90 Wilson's Warbler nests over the course of our study. The earliest nest initiation date (first egg laid) was 17 April, and the last was 10 July. Most nests (68%) were built in blackberry shrubs. 18% in ferns, and 12% in eight other plant species. Mean nest height was 50.4 ± 2.5 cm. Only 16 nests were successful (18%) and 74 (82%) failed to produce young (Table 1).

Causes of nest failure. — Nest predation was responsible for more nest failures than failure caused by cowbird parasitism, 73.0^/f (54/74) versus 13.5% (10/74), respectively. Of 54 depredated nests. 48 were completely depre- dated, resulting in the loss of the entire eluteh or brood, and 6 were partially depretlaletl and subsecjuently abandoned, fhe remaining nests were abaiuloned due to unknown eauses (//

6), failetl due to weather (// 3). or were

aecidentally destroyed (// 1 ). Of the 10 nests

44

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

TABLE 1. Nest outcome and causes of failure for Wilson’s Warblers breeding in coastal Marin County, California, 1997-2000. BHCO refers to Brown-headed Cowbird parasitism.

	All nests	Parasitized nests	Unparasitized nests
Nest outcome
Total number of nests	90	30	60
Successful	16	1	15
Unsuccessful	74	29	45
Percent successful	18%	3%	25%
Causes of nest failure
Depredated'’	54	18	36
BHCO^	10	10	—
Abandoned (unknown)	6	—	6
Abandoned (weather)	3	1	2
Abandoned (other)	1	—	1

Nests from which at least one warbler fledged.

^ Includes four nest failures due to parasitism and subsequent depreda- tion.

Includes four nests fledging BHCO, four lost to predation, and two abandoned.

that failed due to cowbird parasitism, four fledged cowbird young, four were subsequent- ly depredated, and two were abandoned (Table 1).

Nest predation differed among sites for all years combined (x^ 6.58, df = 2, P -

0.043). Predation rates were highest at Muir Beach (91.7%), and lower at Lagunitas Creek (54.5%) and Redwood Creek (61.8%; Table 2). The frequency of predation did not differ between parasitized (73%) and unparasitized (60%) nests (x^ = 1.55, df = 1, P = 0.21). Likewise, there was no difference in the fre- quency of nest abandonment for parasitized (25%) and unparasitized (75%) nests (x^ = 0.27, df = 1, P = 0.60).

Brood parasitism. — Cowbirds parasitized

TABLE 2. Erequency of nest predation and cow- bird parasitism observed in Wilson’s Warblers at three sites in coastal Marin County, California, 1997-2000.

		Ne.st predation	Parasitism
Site		%	n	%	n
Lagunitas Creek {n =	44)	54.5	24	25.0	1 1
Muir Beach {n = 12)		91.7	1 1	83.3	10
Redwood Creek {n =	34)	61.8	21	26.5	9
All sites in — 90)		62.2	56	33.3	30

33.3% (30/90) of all nests, 3% (1/30) of which were successful, as compared to 25% (15/60) of unparasitized nests. The frequency of par- asitism (all years combined) was greater at Muir Beach (83.3%) than at Lagunitas Creek (25.0%) and Redwood Creek (26.5%; x^ ^ 14.97, df = 2, P < 0.001; Table 2). Of the 30 parasitized nests, 26 contained one cowbird egg and 4 had two cowbird eggs. The mean number of cowbird eggs and nestlings per par- asitized nest was 1.16 ± 0.07 and 0.77 ± 0.12, respectively (Table 3). At least one cowbird fledged from each of four parasitized nests, and two cowbirds fledged from one nest, for an overall mean of 0.22 ± 0.10 cowbirds fledged per parasitized nest (Table 3). Warbler young fledged from only one parasitized nest. An inactive nest was parasitized after being depredated.

Clutches in parasitized nests contained few- er warbler eggs (1.52 fewer) than unparasit- ized nests (x^ = 43.13, df = 1, P < 0.001; Table 3). Moreover, the percent of warbler eggs that hatched was lower in parasitized nests (35.7%) than in unparasitized nests (63.8%; x' = 5.55, df = 1, P = 0.019; Table 3) and we found fewer warbler nestlings in

TABLE 3. Eive estimates of reproductive success for parasitized and unparasitized Wilson’s Warbler nests (mean ± SE), coastal Marin County, California, 1997-2000.

Wilson’s Warbler

	Parasitized nests	Unparasitized nests	Combined	Cowbird
Clutch size	1.96 ± 0.16	3.48 ±0.12	2.92 ±0.13	1.16 ± 0.07
Hatching success'’	0.36 ± 0.09	0.64 ± 0.07	0.53 ± 0.06	0.64 ± 0.09
Number of nestlings	0.71 ± 0.19	2.31 ± 0.26	1.72 ± 0.20	0.77 ± 0.12
Fledging success'’	0.04 ± 0.04	0.28 ± 0.06	0.19 ± 0.04	0.18 ± 0.08
Number of fledglings	0.1 1 ± 0.1 1	1.06 ± 0.25	0.71 ± 0.17	0.22 ± 0.10

^ Hatching success: total number of nestlings/clutch size. Fledging success: total number of fledglings/clutch size.

Michaud et al. • REPRODUCTIVE SUCCESS OF WILSON’S WARBLERS

45

TABLE 4. Daily survival and total nest success (Mayfield 1975) for Wilson’s Warblers breeding in riparian woodlands, coastal Marin County, California, 1997-2000.

Number of Exposure Daily survival Nest success

nests days Losses (SE, 95% Cl) (95% Cl)

Lagunitas Creek 44 436 30 0.931(0.012,0.907-0.955) 0.157 (0.080-0.302)

Muir Beach 12 95 12 0.874 (0.034,0.807-0.940) 0.030 (0.004-0.203)

Redwood Creek 34 219 24 0.890(0.021,0.859-0.932) 0.049 (0.014-0.159)

All sites 90 731 66 0.910 (0.011,0.889-0.930) 0.085 (0.047-0.154)

parasitized nests (x^ = 14.76, df = 1, P = 0.001). Similarly, fledging success differed between parasitized (3.6%) and unparasitized (27.6%) nests (x^ = 8.79, df = P = 0.003; Table 3) and parasitized nests fledged fewer warbler young than unparasitized nests (x^ = 9.27, df = 1, = 0.002). Parasitized and un-

parasitized nests averaged 0.71 ± 0.19 and 2.31 ± 0.26 nestlings, respectively.

Daily sunnval and nest success. — The May- field (1975) estimate of nest success was 8.5% (95% Cl = 0.047-0.154; Table 4). Differences among sites in daily survival rates were mar- ginally significant (x^ = 4.72, df = 2, P = 0.094; Table 4). Daily survival rates of para- sitized and unparasitized nests did not differ (X^ = 1.54, df = \, P = 0.21). Daily survival was lowest during the nestling stage (0.87 ± 0.022), and slightly greater during egg laying (0.93 ± 0.023) and incubation (0.93 ± 0.013; X^ = 5.88, df - 2, P = 0.053).

Population trajectory. — The demographic population model suggests that this population of Wilson’s Warbler is not self-sustaining in the absence of immigration from other popu- lations. We used a value of 1 .7 female off- spring per successful nest (based on our held

data) and evaluated the model under a variety of scenarios. Using conservative estimates of survival and productivity and two nesting at- tempts, X = 0.46 (Table 5). When intermediate estimates of survival and productivity for birds attempting three nests per season were used, X = 0.62 (Table 5). Under the best-case scenario (high survival and productivity, four nesting attempts per season), X = 0.98 and approaches the value ( 1 ) required for a stable population.

DISCUSSION

Nest success in this study was extremely low. The proportion of successful nests (0.18) was slightly greater than that reported from inner-coastal California (0.16) and far lower than eight other estimates (0.33-0.93; sum- marized by Ammon and Gilbert 1999).

Since no Mayfield (1975) estimates of nest survival exist for the Wilson’s Warbler, we were unable to compare our survival estimates to those of previous studies. While estimates of nest success may be variable across habitat types, years, and between species, our esti- mates were notably lower than those reported for other warbler species. For example, May-

TABLE 5. Lambda values (finite rate of increase) for population models using mean and 959f confidence intervals for survival and productivity (i.e., low, mean, and high productivity and survival) with two. three, and four nesting attempts for Wilson’s Warbler, coastal Marin County, California, 1 997-2()0(). Number of female offspring/successful nest set at 1.7 (see text).

Number of nesting attempts	Lev	kv productivity* (lower Cl)	Mean productivity*’	High productivity (upper Cl 1
2	5	4	2	?.	4	2	5	4
Low survival (lower Cl)	0.46	0.47	0.49	0.50	0.53	0.57	0.59	0.67	0.75
Mean survival'	0.53	0.55	0.56	0.58	0.62	0.66	0.68	0.77	0.87
High survival' (upper Cl)	0.60	0.62	0.64	0.66	0.70	0.75	0.78	0.88	0.98

“ I.ow proiluclivity: Mayfield success (1.047.

^ Mean productivity; Mayfield success 0 ()S5

' High proiluctivity: Mayfield success; 0 154

'•* Adult survival - ().4,L5. )uvende survival 0..105.

Adult survival - 0.50.L juvenile surv ival 0 .tSJ?

• Adult survival ^ 0.571. juvenile survival 0 4(K)

46

THE WILSON BULLETIN • Vol. 1 16, No. I, March 2004

field (1975) estimates for the Worm-eating Warbler {Helmitheros vennivorus) range from 0.37 to 0.50 in Virginia (Dececco et al. 2000), 0.44 for the Hooded Warbler (Wilsonia citri- mi) in South Carolina (Moorman et al. 2002), and 0.50 for Orange-crowned Warbler {Ver- mivora celata) and 0.58 for Virginia’s Warbler (V. virginiae) in Arizona (Martin 1992a). The lowest estimate for any warbler (summarized by Martin 1992a) is 0.20 for the Kirtland’s Warbler (Dendroica kirtlandii).

Predation appeared to be the primary cause of nest failure for Wilson’s Warblers breeding in coastal riparian woodlands in Marin Coun- ty. However, we documented a relatively high rate of brood parasitism and believe that the combined effects of parasitism and predation explain the poor reproductive success. For ex- ample, the difference in nestling number be- tween parasitized and unparasitized nests was large and contributed to reduced reproductive success. Our results are similar to those of other studies in that they point to nest preda- tion and brood parasitism as the leading caus- es of nesting failures in songbird populations (reviewed by Martin 1992a).

While predation and parasitism were high at all study sites, they were significantly great- er at Muir Beach than at the other sites. Al- though our sample size of nests at Muir Beach was low, we suspect that the higher levels of predation and parasitism at that site may have resulted from the nests’ close proximity to a public picnic area, a horse stable, and a small residential community. Predator densities and predation pressure are higher in areas near suburban landscapes (Wilcove 1985, Andren 1992). At Muir Beach, it is likely that several native and non-native predators are in greater abundance than at our other sites. For exam- ple, house cats were only observed at Muir Beach, and several corvid species frequent the picnic area and, perhaps, the nearby bird feed- ers. Additionally, raccoons (Procyon lotor) may have been more abundant at Muir beach, as they are commonly known to forage from trash cans. In addition, cowbirds may have benefited from the horse pasture, feeders, and mowed picnic area at Muir Beach.

We observed significantly lower reproduc- tive success in warbler nests parasitized by cowbirds. Clutch sizes in parasitized nests were smaller, probably the result of egg-re-

moval behavior by female cowbirds, as were hatching and fledging success. In general, par- asitized nests failed entirely. We observed only one instance of warbler young fledging from a parasitized nest.

Brown-headed Cowbirds experienced poor reproductive success in Wilson’s Warbler nests due to high rates of nest predation; cow- bird fledging success was notably lower than that reported for several other species of cow- bird hosts (reviewed by Ortega 1998), sug- gesting that warblers in this region may not be optimal cowbird hosts.

The Wilson’s Warbler is considered an un- common cowbird host (Ammon and Gilbert 1999), yet it was one of the most common host species at our study sites (Point Reyes Bird Observatory unpubl. data). The propor- tion of nests parasitized in our study (33%), together with estimates for Santa Barbara and San Luis Obispo counties, California (55%, n = 11; Friedmann et al. 1977), suggest that Wilson’s Warbler is a common cowbird host in coastal California.

High levels of nest predation combined with brood parasitism are adversely affecting this population of Wilson’s Warbler. Based on our demographic population model, this local population constitutes a sink; efforts to iden- tify source populations within the region are needed to determine whether the regional pop- ulation can be sustained. High levels of nest predation combined with the effects of brood parasitism point to the underlying causes of recent population declines. Low reproductive success appears to be associated with the proximity of nesting sites to human habitation (i.e., our Muir Beach site), although success was poor at all of our sites. Nest-monitoring data from other sites in coastal California are needed to understand the metapopulation dy- namics of this species. Wilson’s Warblers are not restricted to riparian habitats in coastal Marin County and studies that compare repro- ductive success among different habitats (e.g., riparian versus coniferous forest) would be useful.

ACKNOWLEDGMENTS

We thank P. Northen, D. Stokes, D. Crocker, M. Ste- phens, D. Outlaw, and J. Hull for guidance and sug- gestions, and S. Abbott, D. Humple, S. Laird, C. Rin- toul, S. Scoggin, J. White, and Point Reyes Bird Ob-

Michaud et al. • REPRODUCTIVE SUCCESS OF WILSON’S WARBLERS

47

servatory (PRBO) staff for assistance in the field. We are grateful to D. Hatch and other members of the GGRNA for supporting the project and allowing ac- cess to the study sites. W. M. Gilbert and two anony- mous reviewers provided suggestions that greatly im- proved this manuscript. Financial assistance was pro- vided to J. C. Michaud by the Sonoma State University (SSU) Biology Alumni Student Research Fund, Sigma Xi Grants-In-Aid of Research Fund, and the SSU School of Natural Sciences Student Opportunity Fund, and to PRBO by GGNRA, Golden Gate National Parks Association, Osher Foundation, and the Marin County Audubon Society. This paper is PRBO contribution # 1179.

LITERATURE CITED

Ammon, E. M. and W. M. Gilbert. 1999. Wilson’s Warbler {Wilsonia pusilla). The Birds of North America, no. 478.

Andren, H. 1992. Corvid density and nest predation in relation to forest fragmentation: a landscape perspective. Ecology 73:794-804.

Brittingham, M. C. and S. A. Temple. 1983. Have cowbirds caused forest songbirds to decline? BioScience 33:31-35.

Chase, M. K., N. Nur, and G. R. Geupel. 1997. Sur- vival, productivity, and abundance in a Wilson’s Warbler population. Auk 1 14:354-366.

Dececco, J. a., M. R. Marshall, A. B. Williams, G. A. Gale, and R. J. Cooper. 2000. Comparative seasonal fecundity of four Neotropical migrants in middle Appalachia. Condor 102:653-663. Donovan, T. M., F R. Thompson, III, J. Faaborg, and J. R. Probst. 1995. Reproductive success of mi- gratory birds in sources and sinks. Conservation Biology 9:1380-1395.

Friedmann, H., L. F. Kief, and S. I. Rothstein. 1977. A further contribution to knowledge on the host relations of the parasitic cowbirds. Smithsonian Contributions to Zoology, no. 235. Smithsonian Institution Press, Washington, D.C.

Gardali, T, D. C. Barton, J. D. White, and G. R. Geupel. 2003. Juvenile and adult survival of Swainson’s Thrush (Cathariis iistulatus) in coastal California: annual estimates using capture-recap- ture analyses. Auk 120:1 188-1 194.

Johnson, D. H. 1979. Estimating nest success: the Mayfield method and an alternative. Auk 96:651- 661.

Manolis, j. C., D. F. Andf:rson, and b! J. Ci thbert. 2000. Uncertain nest fates in songbird studies and variation in Mayfield estimation. Auk 117:615- 626.

Martin. T. F. 1992a. Breeding productivity consider- ations: what are the appropriate habitat features for management? Pages 455-473 in Ixology aiul

conservation of Neotropical migrant landbirds (J. M. Hagan, III, and D. W. Johnston, Eds.). Smith- sonian Institution Press, Washington, D.C.

Martin, T. E. 1992b. Nest predation and nest sites. BioScience 43:523-532.

Martin, T. E. and G. R. Geupel. 1993. Nest-monitor- ing plots: methods for locating nests and moni- toring success. Journal of Field Ornithology 64: 507-519.

Mayfield, H. F. 1975. Suggestions for calculating nest success. Wilson Bulletin 87:456-466.

Moorman, C. E., D. C. Guynn, Jr., and J. C. Kilgo. 2002. Hooded Warbler nesting success adjacent to group-selection and clearcut edges in a southeast- ern bottomland forest. Condor 104:366-377.

Nur, N., S. L. Jones, and G. R. Geupel. 1999. Statis- tical guide to data analysis of avian monitoring programs. Biological Technical Publication BTP- R6001-1999. U.S. Fish and Wildlife Service, Washington, D.C.

Ortega, C. P. 1998. Cowbirds and other brood para- sites. The University of Arizona Press, Tucson.

Pease, C. M. and J. A. Grzybowski. 1995. Assessing consequences of brood parasitism and nest pre- dation on seasonal fecundity in passerine birds. Auk 112:343-363.

Perkins, D. W. and P. D. Vickery. 2001. Annual sur- vival of an endangered passerine, the Florida Grasshopper Sparrow. Wilson Bulletin 113:211- 216.

Powell, L. A., M. J. Conroy, D. G. Krementz, and J. D. Lang. 1999. A model to predict breeding season productivity for mulitbrooded songbirds. Auk 116:1001-1008.

Pulliam, H. R. 1988. Sources, sinks, and population regulation. American Naturalist 132:652-661.

Sall, j. and a. Lehman. 1996. JMP Statistics, ver. 3. 1 . SAS Institute, Inc. Cary, North Carolina.

Sauer, J. R., J. E. Hines, and J. Fallon. 2001. The North American Breeding Bird Survey, results and analysis 1966-2000, ver. 2001.2. USGS Pa- tuxent Wildlife Research Center, Laurel, Mary- land. Online at <http://www.mbr.nbs.gov/bbs/>.

Sauer, J. R. and B. K. Williams. 1989. Generalized procedures for testing hypotheses about sur\i\al and recovery rates. Journal of \Vildlife Manage- ment 53:137-142.

Sti:wari, R. M. 1973. Breeding behavior and life his- tory of the Wilson's Warbler. Wilson Bulletin 85: 21-30.

Sn;wART, R. M., R. P. Hi ndi rson. and K. Dari ing. 1977. Breeding ecology of the Wilson’s Warbler in the high Sierra Ne\atla. C'alifornia. Li\ing BirtI 16:83-102.

WiLCovi;, I). S. 1985. Nest pretlation in forest tracts and the tlecline of migratory songbirds. Ideology 66:1211 1214.

Wilson Bulletin, 1 16(1), 2004, pp. 48-55

SPATIAL FORAGING DIFFERENCES IN AMERICAN REDSTARTS ALONG THE SHORELINE OE NORTHERN LAKE HURON DURING SPRING MIGRATION

ROBERT J. SMITH,’ 4 “^ MICHAEL J. HAMAS,’ DAVID N. EWERT,^ AND MATTHEW E. DALLMAN’-^

ABSTRACT. — Lowland coniferous forests adjacent to northern Lake Huron provide important stopover habitat for landbirds during spring migration. Large numbers of aquatic insects emerging from nearshore waters of northern Lake Huron appear to be an important food source. In this study we compared the foraging behavior of a long-distance landbird migrant, the American Redstart {Setophaga niticilla), in areas with high densities of emergent aquatic insects to areas with few or no emergent aquatic insects to assess the significance of these arthropods as an early spring food source. Redstarts foraged differently in shoreline habitats relative to inland habitats of similar vegetation composition. Both males and females gleaned significantly more in shoreline habitats as compared to inland areas of similar vegetation composition, and inland birds performed more sally strikes than birds at the shoreline. Both sexes also varied the use of tree species in which they foraged. Redstarts used northern white cedar {Thuja occidentalis) more at shoreline than inland, while inland redstarts foraged in deciduous trees more than at the shoreline. We suggest that differences in foraging between shoreline and inland locations were responses to differences in prey types and abundance, most notably the presence of emergent aquatic insects (Diptera: Chironomidae) in shoreline habitat. Our results complement those of previous work, suggesting that midges provide a critical early season resource for landbirds migrating through Michigan’s eastern Upper Peninsula during spring. Received 30 September 2003, accepted 26 March 2004.

During spring and early summer in temper- ate North America, large numbers of aquatic insects that have metamorphosed into sexually mature adults often amass in terrestrial habi- tats adjacent to riparian and lacustrine systems (Armitage 1995, McCafferty 1998). These in- vertebrates are relatively weak fliers (e.g., Ko- vats et al. 1996) and tend to be restricted to nearshore terrestrial habitats. In Michigan’s eastern Upper Peninsula, midges (Diptera: Chironomidae) are the predominant aquatic arthropods when migratory landbirds stop during spring migration. These invertebrates swarm profusely in shoreline areas while be- ing virtually nonexistent inland (DNE unpubl. data).

Recent evidence suggests that lowland co- niferous forests adjacent to northern Lake Hu- ron provide important stopover habitat for spring migrants. More landbirds are found in

' Dept, of Biology, Central Michigan Univ., Mt. Pleasant, MI 48858, USA.

-The Nature Conservancy, 101 E. Grand River, Lansing, MI 48906, USA.

^ The Nature Conservancy, 707 Main St. W., Ash- land, WI 54806, USA.

Current address: Dept, of Biology, The Univ. of Scranton, Scranton, PA 18510, USA.

^ Corresponding author; e-mail: smithr9 @ scranton.edu

nearshore, midge-abundant habitats than in- land habitats with comparable vegetation (DNE unpubl. data). Further, Black-throated Green Warblers {Dendroica virens) forage in and use habitat elements differently in midge- abundant habitats compared to inland, midge- depauperate habitats (Smith et al. 1998), and American Redstarts {Setophaga ruticilla) and Black-throated Green Warblers forage at high- er rates in midge-abundant areas relative to habitats with no midges (Seefeldt 1997).

Here we describe shoreline/inland differ- ences in the foraging behavior of a long-dis- tance landbird migrant, the American Red- start, during spring migration in Michigan’s eastern Upper Peninsula. By focusing on spring migration, we were able to examine bird behavior when shoreline/inland differenc- es in arthropod abundances were most dra- matic (Smith et al. 1998; DNE unpubl. data). We compared redstart foraging behavior in ar- eas with high densities of midges to areas with few or no emergent midges to assess the sig- nificance of these arthropods as a food source during early spring.

METHODS

Study area. — The study area included ap- proximately 80 km of shoreline, extending

48