harvard university

Ernst Mayr Library of the Museum of Comparative Zoology

MCZ

LIBRARY

FEB 15 2011

HARVARD

UNIVERSITY

U U-

^he Wilson Journal

of Ornithology

Volume 118, Number 1, March 2006

Published by the Wilson Ornithological Society

THE WILSON ORNITHOLOGICAL SOCIETY FOUNDED DECEMBER 3, 1888

Named after ALEXANDER WILSON, the first American Ornithologist.

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

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

Second Vice-President — E. Dale Kennedy, Biology Dept., Albion College, Albion, MI 49224, USA; e-mail: dkennedy@albion.edu

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

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

Treasurer — Melinda M. Clark, 52684 Highland Dr., South Bend, IN 46635, USA; e-mail: MClark@tcservices.biz

Elected Council Members — Robert C. Beason, Mary Gustafson, and Timothy O’Connell (terms expire 2006); Mary Bomberger Brown, Robert L. Curry, and James R. Hill, III (terms expire 2007); Kathy G. Beal, Daniel Klem, Jr., and Douglas W. White (terms expire 2008).

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 Journal of Ornithology is sent to all members not in arrears for dues.

THE WILSON JOURNAL OF ORNITHOLOGY (formerly The Wilson Bulletin )

THE WILSON JOURNAL OF ORNITHOLOGY (ISSN 1559-4491) 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 OSNA, 5400 Bosque Blvd., Ste. 680, Waco, TX 76710.

All articles and communications for publications 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 OSNA, 5400 Bosque Blvd., Ste. 680, Waco, TX 76710. Phone: (254) 399-9636; e-mail: business@osnabirds.org. Back issues or single copies are available for $12.00 each. Most back issues of the journal are available and may be ordered from OSNA. Special prices will be quoted for quantity orders. All issues of the journal published before 2000 are accessible on a free Web site at the Univ. of New Mexico library (http://elibrary.unm. edu/sora/). The site is fully searchable, and full-text reproductions of all papers (including illustrations) are available as either PDF or DjVu files.

COVER: Wilson’s Snipe ( Gallinago delicata). Illustration by Scott Rashid.

FRONTISPIECE. An adult White-masked Antbird ( Pithys castaneus ) above and a juvenile below. Previously known from only the type specimen, the species was rediscovered in 2001 in northwestern Department Loreto, Peru (see p. 13). Original painting (watercolor and gouache) by Daniel F. Lane.

VOL. 118, NO. 1

ne Wilson Journal of Ornithology

Published by the Wilson Ornithological Society March 2006 PAGES 1—130

The Wilson Journal of Ornithology 118(1): 1-2, 2006

MESSAGE FROM THE EDITOR: THE NEW WILSON JOURNAL OF ORNITHOLOGY

This issue of your journal — 118(1), March 2006 — is the debut issue of The Wilson Journal of Ornithology. As indicated in the insert letter that came with your December 2003 issue, the Wilson Council, Wilson Society officers, and I spent considerable time over the last year de- bating— and eventually agreeing on — the need to update the journal’s name and appearance. We believe that the new name maintains the tradition of honoring Alexander Wilson, more clearly reflects the journal’s theme and content, and is more contemporary. In addition to the new journal name, the front and back covers have been redesigned, the title page is new, and we have added a new feature to The Wil- son Journal of Ornithology.

The front cover of each issue will portray a different illustration of one of the species named after Alexander Wilson. Pen and ink or halftone artwork was solicited from over two dozen artists, and we selected those illustra- tions that we believe demonstrate both orni- thological and artistic merit. The Wilson’s Snipe on the March cover is a halftone by artist Scott Rashid. Pen and ink illustrations of the Wilson’s Phalarope, Wilson’s Plover, and Wilson’s Storm-Petrel will appear on the covers of the June, September, and December issues, re- spectively. The fifth species named after Al-

exander Wilson, Wilson’s Warbler, will appear on the cover of each issue in a logo designed by George Miksch Sutton, and the Wilson’s Warblers that appeared on the cover from 1962 to 2005 — also by G. M. Sutton — will now ap- pear on the title page of the first article in each issue.

The back cover (Contents) has also been re- designed, to make it more aesthetically pleas- ing and easier to read. A new feature, “Once Upon a Time in American Ornithology,” de- buts, as well. This feature will put forward the observations and reflections of naturalists from times past — to afford retrospection and to re- mind us all of the exhilaration that comes from being afield and how it once was in American ornithology. I encourage Wilson Ornithologi- cal Society members and other readers of the journal to submit favorite historical field ac- counts (including a brief introductory state- ment) for consideration of publication in a fu- ture issue.

I realize that such cosmetic modifications will have little long-term effect on subscrip- tions, membership, or the ornithological sci- ence offered in The Wilson Journal of Orni- thology. Combined with a renewed commit- ment and more substantive changes behind the scenes, however, I believe that the publication

1

2

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

of this issue does mark a new beginning for the Wilson Ornithological Society and its jour- nal: (1) the journal has been published on time beginning with the June 2005 issue, (2) most authors are receiving an initial decision on their work within 3-4 months, (3) the time from manuscript submission to publication now averages only about 12 months, and (4) manuscript submissions are up >20% from 2004. I sincerely hope that you, the readers and authors, welcome the new look and the

improvements we continue to make to The Wilson Journal of Ornithology. I thank Wilson Council and officers; Keith Parsons, Karen Ridgway, and the graphics department at Allen Press; Teri Kman; and The Wilson Journal of Ornithology Editorial Office staff — Beth Dillon, Alison Goffredi, and Cynthia Melcher. All were instrumental in the execution and realization of the new design changes and in helping to bring The Wilson Journal of Ornithology back on schedule. — James A. Sedgwick, Editor.

The Wilson Journal of Ornithology 1 18(1):3-12, 2006

VARIATION IN MASS OF FEMALE PROTHONOTARY WARBLERS

DURING NESTING

CHARLES R. B LEM 123 AND LEANN B. BLEM1 2

ABSTRACT. — Over an 18-year period (1987-2004), we examined variation in body mass of female Protho- notary Warblers ( Protonotaria citrea) captured throughout their nesting cycle. As is typical for many small passerine birds, body mass was greatest during egg laying and decreased throughout incubation and feeding of young. Mass decreased significantly between the onset of incubation and fledging of both first and second broods. Mass loss was gradual during incubation, noteworthy during the first 2 days of feeding nestlings, but did not continue to decrease throughout the feeding period. Mass lost while raising the first brood was regained before initiating the second brood. Mass of female warblers, adjusted for effects of nest attempt, year, clutch size, and day and stage of nesting, increased slightly with age. Body mass of nesting female warblers varied significantly with day of the nest cycle during incubation but not during egg laying or feeding of young. Mass was associated with clutch size during incubation in both first and second broods, but was not associated significantly with brood size when females were feeding nestlings. Frequency of food delivery to nestlings was associated nega- tively with female body mass. Females typically made more feeding trips per day than males. Feeding rates were correlated among pairs; that is, females with higher rates of delivery were mated to males that made a higher number of trips. Received 18 February 2005, accepted 21 October 2005.

Mass loss is often used as an index of re- productive costs in birds (see review in Mer- kle and Barclay 1996), largely because it is a consistent factor in patterns of avian life his- tory. During the breeding season, female pas- serine birds typically gain mass in the period before egg laying, maintain or gradually lose a small amount during incubation, and then lose a significant amount of mass during brooding (e.g., Ricklefs 1974; Freed 1981; Moreno 1989a, 1989b). A similar pattern of change during breeding has been documented in several passerine birds (e.g.. Freed 1981, Ricklefs and Hussell 1984, Hillstrom 1995, Merila and Wiggins 1997). Researchers have hypothesized that mass loss may be a proxi- mate response to energetic demands (e.g., Nice 1937, Hussell 1972, Askenmo 1977). Specifically, mass loss should be greatest dur- ing periods when energy demands are great- est, particularly near fledging when nestlings have acquired the ability to thermoregulate, and are relatively large. According to this hy- pothesis, mass loss should be a function of brood size. A second hypothesis suggests that decreased mass reduces the energy required

1 Dept, of Biology, Virginia Commonwealth Univ., 1000 W. Cary St., Richmond, VA 23284-2012, USA.

2 Current address: Flathead Lake Biological Station, 311 Bio Station Lane, Poison, MT 59860, USA.

3 Corresponding author; e-mail: cblem@saturn.vcu.edu

for flight when food demands of nestlings are greatest, thus reducing energy requirements of females and increasing the efficiency of feed- ing the young (e.g.. Freed 1981, Norberg 1981, Hinsley 2000). In this instance, body mass should decrease shortly after eggs hatch and should be independent of brood size. A final hypothesis is that mass loss results from degeneration of female reproductive tissues during the nesting cycle (Ricklefs 1974, Rick- lefs and Hussell 1984), and should not pro- gressively occur during incubation or feeding of young. Some studies have eliminated the tissue degeneration hypothesis because gonad- al atrophy is over before the period when mass loss is greatest (Moreno 1989a, 1989b; Merkle and Barclay 1996). It is difficult to isolate these three hypotheses, however, and some researchers have not found them to be mutually exclusive (e.g., Hillstrom 1995, Mer- ila and Wiggins 1997).

The question that usually has been ad- dressed is: “Is mass loss evidence of energy demand and/or does it reduce costs of flight and enhance parental fitness?” It has been shown that energy expenditure is related sig- nificantly to rates of nest visitation, but not always in a linear manner (Bryant 1988). Fur- thermore, decreased body mass of adults rear- ing young may enhance their fitness through reduction of energy demand during the period of feeding nestlings. Our study examined

3

4

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

measurements of body mass of female Pro- thonotary Warblers ( Protonotaria citrea) ob- tained over an 1 8-year period. With these data, we attempted to answer three questions: (1) How does female body mass in this species vary over the breeding season? (2) Does body mass vary significantly among stages of nest- ing and among years? (3) What are the roles of brood size, stage of reproduction, and nest attempt in determining body mass in this spe- cies?

METHODS

Study area and measurement of mass. — Be- ginning in March 1987, we placed wooden nest boxes along tidal creeks in swamp forest on and near Presquile National Wildlife Ref- uge (37° 20' N, 77° 15' W) near Hopewell, Virginia (Blem and Blem 1991, 1992, 1994). The dominant tree species were black gum (Nyssa sylvatica), red maple ( Acer rubrum), and ash ( Fraxinus sp.). Tidal amplitude in the swamp during spring tides was >1 m. Nest boxes were placed on metal poles at approx- imately 100-m intervals along creek banks. Box dimensions were 28LX9WX6D cm and the entrance hole was 3.8 cm in diameter (see Blem and Blem 1991). We determined optimal nest-box sites during the first 2 years of the study (Blem and Blem 1991) and boxes were adjusted accordingly to maximize their usage by warblers. The number of nest boxes used in the study was gradually increased from 141 in 1987 to 320 in 2004.

The contents of boxes were documented 6— 20 times during the breeding season each year, depending upon the demands of other inves- tigations of reproductive output. Females were captured as they exited nest boxes, weighed to the nearest 0.1 g on a portable electronic balance, and banded with federal bands. No warbler in these analyses was weighed twice per stage, and usually not more than once dur- ing the same nest attempt. Midday (10:00- 14:00 EST) masses (g) did not vary signifi- cantly with time of day (mass = -0.04 hr + 16.3, P = 0.49, R 2 = 0.008, n = 2,124). Only midday masses were used in the following analyses. We recorded dates of first eggs and clutch sizes for those nests visited often enough that we could be certain of the timing. Clutch size throughout the study was consid- ered to be the number of eggs present at the

onset of incubation. We converted first egg (nest start) dates into Julian days for analysis. Prothonotary Warblers generally produce two clutches each season (Petit 1989), and second clutches typically include fewer eggs (Blem et al. 1999). We therefore divided nests with eggs in two groups — “first nests,” in which first eggs were laid from 25 April through 20 May, and “second nests,” in which first eggs were laid after 20 May (see Petit 1989). Some of the second nests may have been replace- ment clutches for first nests that had been dep- redated, but we are certain that many of them were produced by females that had success- fully fledged young (Podlesak and Blem 2001, 2002). We used 20 May as the separation date because it represents a major hiatus in laying and is the date after which few first clutches have been laid at our study site. It also was used because of the length of time necessary for Prothonotary Warblers to complete one nesting cycle (approximately 27 days) after a mean potential starting date of 24 April (Blem and Blem 1992). We divided nesting into three phases: laying (and egg formation), incuba- tion, and feeding young. The first phase ended with the first day of incubation and included birds that were building nests as well as laying eggs. The second phase began with the first egg and ended with hatching (Fig. 1).

Feeding visits. — In 2002, we recorded feed- ing visits by warblers at individual boxes dur- ing first broods by means of battery-powered remote video cameras with programmable, portable videocassette recorders. We obtained >500 hr of nest-activity records at eight nests (four broods of three young and four broods of five young) on days 7 through 10. Video cameras were small and camouflaged and did not noticeably alter behavior of the warblers. Individual visits (see Figs. 2-3) were tran- scribed from replays of the recordings in the lab. We totaled all feeding visits made by both parents from dawn-to-dark for all 4 days. We could not accurately assess prey size from the recordings, but we did count the number of items — mostly caterpillars — that were dis- tinctly larger than 2 cm (“large prey”), as judged by the entry hole in the nest box. Fe- male warblers were weighed 2 days before nestlings fledged.

Analyses. — Over the 18-year period, we ob- tained 2,124 measurements of body mass from

20

18

16

14

12

20

18

16

14

12

}dy r

and Blent • MASS OF FEMALE PROTHONOTARY WARBLERS

5

First broods

n= 145

i • *•.}( .i. .....

t •• • *! • ir

n = 1 ,343

n = 237

Second broods

0 -5

Egg laying

5 10

Incubation

Days

15 20 25

Feeding nestlings

(g) of female Prothonotary Warblers during nesting in eastern Virginia, 1987-2004 (day tion). Numerous circles are hidden under duplicate values ( n = 2,124).

6

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

FIG. 2. Feeding visits/nest/day made by female Prothonotary Warblers during days 7-10 of feeding nestlings versus female body mass at the end of incubation, eastern Virginia, 2002. Open circles represent broods of three nestlings; solid circles represent broods of five nestlings. Nest visitation was a function of female body mass, regardless of brood size.

977 different adult female warblers. For anal- ysis, we partitioned these measurements among nesting attempts (first and second nests, n — 1,344 and 780, respectively) and stages of nesting (egg formation and laying, incubating, and feeding). The number of mea- surements in each stage-year combination var- ied from 24 during laying in second nests to 1,344 during incubation in first nests. Clutch size varied from two to six eggs and ages of females ranged from 1 to 8 years.

To examine differences in mass between nests and among stages of nesting and brood sizes (adjusted for day of nesting), we used univariate ANCOVA with multiple indepen- dent variables in PROC GLM (SAS Institute, Inc. 2000). Brood size, nest attempt, age, stage of nesting, and their interactions were considered fixed (categorical) effects in vari- ous models. Day of nesting (range = —9 to 24; 0 = day of onset of incubation) was a continuous variable. Analysis of covariance

was done using the PROC GLM procedure because the data set was unbalanced among effects (Zar 1999). Type III sums of squares were used, adjusting significance of each fac- tor for the effects of all other variables. Single comparisons of means were done by means of appropriate Mests based on tests of equality of variances (SAS Institute, Inc. 2000). Few females were measured more than once during the same stage of nesting in a given nest in the same year; therefore, we did not use re- peated measures analyses. Because some of the associated variables were not measured with each measurement of body mass, sample sizes vary among analyses. All r-tests were two-tailed. Means are presented ± SD. Statis- tical significance was set at P < 0.05.

RESULTS

Body mass. — In the following analyses and comparisons, we assumed that patterns found between specific points along a regression

Blem and Blem • MASS OF FEMALE PROTHONOTARY WARBLERS

7

Female visits/nest/day

FIG. 3. Feeding visits/nest/day by mated pairs of Prothonotary Warblers during days 7-10 of feeding nest- lings, eastern Virginia, 2002. Open circles represent broods of three nestlings; solid circles represent broods of five nestlings. Males brought food less often than females, but the frequency of male visits/nest/day was a function of that of females.

were representative of patterns deduced from single measurements of numerous females. This was confirmed in our observations of multiple measurements of a few single fe- males (CRB unpubl. data).

Body mass of female Prothonotary War- blers varied over the breeding season in the typical passerine pattern. That is, variation was greatest during egg laying, mass de- creased gradually during incubation, and then there was a noteworthy decrease in mass im- mediately after the eggs hatched (Fig. 1). Af- ter the decline immediately after hatching, adult female mass did not change over time throughout the period of feeding nestlings. Mean body masses did not differ between nest attempts during egg formation and laying (first nests: 16.9 ± 1.2, n = 143; second nests: 16.8 ± 1.9, n = 93, Fh235 = 0.20, P = 0.65), but did differ between nests during incubation (first nests: 16.2 ± 0.9, n = 1,225; second nests: 15.6 ± 0.9, n = 304, FU526 = 6.7, P = 0.011) and during the feeding phase (first

nests: 15.2 ± 1.0, n = 238; second nests: 14.9 ± 0.8, n = 121; F1>358 = 6.7, P = 0.012). Mass did not vary with day of nesting in the laying or feeding stages of either nesting attempt, but it did decline significantly with day of incu- bation (first nests: Ful3 42 = 18.0, P < 0.001; second nests: F1303 = 33.5, P < 0.001).

As judged by the collective scatter of in- dividual masses over time, females collective- ly lost 10.1% of their body mass between the onset of incubation and fledging of first broods and 1 1 .3% in second broods. Much of this loss appeared to occur during the first 2 days of feeding nestlings (5.4 and 7.7%, re- spectively). Mass lost during first broods was regained before the initiation of second broods. Body mass extremes were 11.9 g for an incubating bird and 21.0 g for a female during the early days of egg laying.

When the data set including all variables was considered (n = 1,814; Fig. 1), mass var- ied significantly with nest attempt, stage of nesting, clutch size (2-6), female age (1-8

8

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

TABLE 1. Analysis of covariance of body mass of female Prothonotary Warblers in eastern Virginia, 1987- 2004 (n = 1,814). All two-way and three-way interactions were statistically insignificant except for nesting attempt X stage of nesting. Clutch sizes were 2-6 and ages were 1-8 years. Days of nesting ranged from -9 through 24.

Source	df	F	P > F
Nesting attempt	1	7.6	0.006
Stage of nesting	2	27.0	<0.001
Clutch size	4	10.4	<0.001
Age	5	6.8	<0.001
Day of nest cycle	1	35.7	<0.001
Year	17	2.6	0.015
Nesting attempt X stage of nesting	1	2.8	0.050

years), day of the nest cycle, and year (Table 1). There was a significant interaction between nesting attempt (first/second nest) and stage of nesting, but no other two-way and three-way interactions were statistically significant. When stages of nesting were analyzed indi- vidually, body mass during the laying and feeding stages did not differ among clutches/ broods of different sizes and mass did not vary significantly with day of nesting in these stages.

Body mass adjusted for effects of nest at- tempt, year, clutch size, and day and stage of nesting varied significantly with female age (Fu 213 = 15.0, P < 0.001; Table 2). Unad- justed masses indicated that much of this change occurred between birds in their first year (SY birds) and all older age classes (ASY). Measurements of mass were obtained from a large range of ages, including 64 mea- surements that exceeded the published maxi- mum age (5 years 1 1 months) for the species (Kennard 1975).

During incubation, mass was significantly

TABLE 2. Least-squares means of body mass among incubating female Prothonotary Warblers dur- ing mid-incubation (days 3-8) as a function of age (years) in eastern Virginia, 1987-2004 (/? = 1,540). All means were adjusted for the effects of nest attempt, clutch size, and day and stage of nesting.

Age	Mean mass (g)	n
1	16.0	275
2	16.3	565
3	16.4	420
4	16.4	147
5	16.1	80
>6	16.1	48

associated with day of nesting and clutch size (Table 3). Mass tended to decrease gradually throughout incubation. Birds with larger clutches during first nesting attempts tended to have greater body mass; birds with small clutches in second nests had the lowest body mass.

Feeding visits. — Total nest visits per day made by females during days 7-10 of feeding nestlings was a function of female body mass, regardless of brood size (three young; Fl3 = 13.8, P = 0.023, R2 = 0.80; five young: F13 = 15.5, P = 0.034, R2 = 0.85; Fig. 2). Males brought food less often than females (three young: x2 = 38.2, df = 1, P < 0.052; five

TABLE 3. Analysis of covariance of body mass among female Prothonotary Warblers in eastern Vir- ginia, 1987-2004 by stage of nesting ( n = 2,124 in all analyses). Clutch and brood sizes were 2-6 and ages were 1-6 years; days of nesting ranged from —9 through 24 (day 0 = first day of incubation).
Source	df	F	P > F
Egg formation and laying (/?	= 169)
Nesting attempt	1	0.9	0.34
Clutch size	4	2.2	0.092
Day of nesting	1	0.2	0.70
Age	5	1.7	0.13
Incubation {n = 1,647) Nesting attempt	1	52.3	<0.001
Clutch size	4	9.3	<0.001
Day of nesting	1	40.4	<0.001
Age	5	6.3	<0.001
Feeding nestlings ( n = Nesting attempt	308) 1	4.3	0.039
Brood size	4	1.0	0.45
Day of nesting	1	0.3	0.58
Age	5	1.3	0.26

Blem and Blem • MASS OF FEMALE PROTHONOTARY WARBLERS

9

TABLE 4. Mean visitation rates (no./day total) for days 7-10 of nestling development in	± SD) of male eastern Virginia	and female Prothonotary Warblers (percent of l, 2002.
	Female visits		Male visits
Brood size	Per nest	Per nestling	Per nest	Per nestling
3 {n = 4) 5 (n = 4)	306 ± 95 (63.8) 396 ± 148 (56.5)	102.0 79.2	171 ± 40 (36.2) 295 ± 108 (43.5)	57.0 59.0

young: x2 = 12.1, df = 1, P < 0.054; Table 4), but frequency of male visits per day was a function of that of females (female visits = 1.0 ± 1.06 X male visits; R2 = 0.75, Fl3 = 17.7, p = 0.006; Fig. 3). Female feeding trips per nestling decreased with brood size (x2 = 9.3, df = 1, P < 0.05; Table 4), but male trips per nestling did not decrease (x2 = 0.034, df = 1, P > 0.05). The percentage of total pa- rental visits made by males declined from a high of 44.0% on day 7 to a low of 34.8% on day 10. Males brought significantly more “large prey items” to the nest than did fe- males (males: 330, females: 210; x2 = 26.7, df = 1, p < 0.05). These prey items were mostly Hexagenia sp. mayflies and lepidop- teran caterpillars. There was no significant dif- ference in the number of larger prey delivered by males to different brood sizes (175 in broods of three, 155 in broods of five; x2 = 1.2, df = 1, P > 0.05).

DISCUSSION

Body mass clearly is associated with stage of breeding activity in small passerines (Freed 1981, Ricklefs and Hussell 1984, Cichon 2001), and each stage — egg formation and laying, incubation, and feeding of nestlings is characterized by a different pattern of mass change (e.g., Fig. 1). Mass change of female Prothonotary Warblers in our study was sim- ilar to that reported in several other studies of passerine species (e.g.. Freed 1981, Ricklefs and Hussell 1984, Johnson et al. 1990, Hills- trom 1995). During egg laying, body mass varied greatly with follicle formation and re- lease of eggs, then declined progressively throughout incubation (Fig. 1), and dropped sharply at hatching. Female mass then re- mained relatively constant throughout the pe- riod of feeding nestlings. Mass changes in Prothonotary Warblers during egg laying and incubation were similar to those of all small passerines and require little explanation. Mass

loss at hatching is more complex and differs among species. Because the significance of this loss is uncertain, the behavior and com- positional dynamics of females requires closer scrutiny.

Two potential hypotheses have been pro- posed to explain mass loss of female birds during feeding of nestlings: (1) energy de- mand (cost of reproduction hypothesis = re- serve mobilization hypothesis; Cavitt and Thompson 1997), and (2) long-term benefits from reduction of power demands for flight during feeding (mass adjustment hypothesis = flight efficiency hypothesis). Forming and lay- ing eggs, incubating, and feeding nestlings re- quires additional collection and expenditure of energy, whereas adjusting mass to save energy expended in flight during the numerous trips made while feeding young is an adaptive loss.

It has become obvious that body mass can vary as a result of energy demand during ex- treme years (Merila and Wiggins 1997) or with larger broods (Nur 1984). It appears to be axiomatic that reserves should be depleted during times of high-energy demand and it is well known that body mass and energy re- serves are closely related (Blem 1990). Part of the variation in mass within stages of the nest cycle may result from differences in an- nual factors, such as temperature extremes, in- clement weather (Merila and Wiggins 1997), or brood number (De Laet and Dhondt 1989). Because of our large sample size, we were able to detect annual variation within the in- cubation period of first nests, largely by elim- inating much of the variation associated with several other variables. Others (e.g., Johnson et al. 1990) have likewise found significant annual variations in mass of breeding birds, and extreme environmental conditions in ex- ceptional years have important influences on body mass (Merila and Wiggins 1997).

Not all studies, however, have shown that energy demand is an important factor in body

10

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

mass. For example, larger broods are not al- ways associated with greater mass loss of fe- males (Pinkowski 1978, this study), even though energy expenditure by females in- creases with brood size (Sanz et al. 1998). Furthermore, food-supplementation studies have provided mixed results. Food supple- ments did not affect female mass, brood mass, or length of the nestling period among House Wrens ( Troglodytes aedon\ Cavitt and Thompson 1997) or Northern Wheatears ( Oenanthe oenanthe; Moreno 1989a). How- ever, food-supplemented female Mountain Bluebirds ( Sialia currucoides; Garcia et al. 1993) maintained greater body mass and fledged larger young than females receiving no food supplementation. Some studies have found that female mass is a negative function of brood size (Nur 1984, Merila and Wiggins 1997), and that energy demand during first broods may influence the probability of hav- ing a second brood in some species (De Laet and Dhondt 1989). In Prothonotary Warblers, it appears that many females totally recover lost mass fairly quickly between nest attempts. It has been suggested that species breeding in different environments may respond different- ly to stress associated with increased energy requirements and there may not be selection for adaptive mass loss (Cavitt and Thompson 1997).

The pattern of mass change in female Pro- thonotary Warblers in our study does not sup- port the cost of reproduction hypothesis, but it does support the mass adjustment hypothe- sis. Important supporting observations includ- ed ( 1 ) the regular loss of mass after hatching in both nesting attempts, (2) the lack of influ- ence of brood size on female mass, (3) no increasing loss in female mass as young grew and when feeding activity levels were great- est, (4) more feeding trips made by females that weighed less, and (5) little evidence that males adjusted their feeding efforts to offset demands on females. Trivers (1972) predicted that, within breeding pairs, females would provide the largest proportion of nestling care because they had a larger share of investment of energy than males. In our study, female Prothonotary Warblers made more feeding trips than males (both broods). Male Protho- notary Warblers, however, brought a greater proportion of large prey, which may have sig-

nificantly offset female effort during later stages in the nesting cycle even though males made fewer trips as nestlings neared fledging.

The mass adjustment hypothesis suggests that birds benefit from mass loss due to de- creased wing loading (e.g.. Freed 1981, Nor- berg 1981, Ricklefs and Hussell 1984, Cavitt and Thompson 1997). Energy saved by mass reduction may enable parent birds to raise more young faster or produce fledglings with greater mass. Observations supporting the mass adjustment hypothesis include (1) great- er loss of mass before the period of maximum energy requirement (e.g., Freed 1981, Ricklefs and Hussell 1984, Merkle and Barclay 1996, this study), (2) loss of mass independent of brood size (e.g., Freed 1981, this study) or length of incubation (Sanz and Moreno 1995, this study), and (3) no increase in body mass among food-supplemented females feeding nestlings (Cavitt and Thompson 1997). In our study, mass loss of females during incubation was correlated with clutch size, but mass of females feeding nestlings was not affected by brood size, nor did female mass decrease throughout nestling development. If increased energy demand is important, then female mass should decline significantly as nestlings grow, although it is possible that males may “pick up the slack.” That is, male warblers might feed young more frequently or with higher- quality food in large broods than small, thus reducing energy demands on females and al- lowing them to maintain their mass and fit- ness. Our observations weakly support these ideas. Males did bring more large prey items than females, but this did not vary with brood size or with nestling age. Furthermore, males made fewer visits late in the nesting cycle than females. This pattern is nearly identical with that documented for Willow Tits ( Poecile montanus ; Rytkonen et al. 1996). Similar studies have shown that nest visitation rates may be greater in males of some species (Grundel 1987), greater in females of others (Pinkowski 1978, Conrad and Robertson 1993), or may not differ between the sexes (Best 1977, Knapton 1984, Omland and Sher- ry 1994). The significance of the age:body mass relationship during the reproductive pe- riod is not clear. We are aware of few studies that have demonstrated an age effect on mass (see De Laet and Dhondt 1989, Merila and

Blem and Blem • MASS OF FEMALE PROTHONOTARY WARBLERS

1 1

Wiggins 1997). In our study, however, female age had a significant effect on body mass, even after mass was adjusted for the effects of many other variables.

Mass variation of female birds during nest- ing obviously is a complex phenomenon. Deeper insight into mass variations will be ob- tained only with studies that combine mea- sures of body composition, condition of re- production tracts, and measures of hormone levels with stage of nesting. While time-con- suming, collecting large data sets over nu- merous years is well worth the trouble, but would be even more valuable if simultaneous studies could be carried out at several sites over the range of the species.

ACKNOWLEDGMENTS

We thank the officials of the Eastern Virginia Rivers National Wildlife Refuge (NWR) Complex for per- mission to conduct this study at Presquile NWR. The North American Bluebird Society financially support- ed box construction. More than 100 students, friends, and faculty colleagues assisted in this project and we thank them all, especially A. S. and K. C. Seidenberg and J. R. and R. J. Reilly for their continuous help over many years. We thank L. B. Williams, K. R. Guis- inger, and D. S. Stevens for transcribing visits of war- blers from long, boring tape recordings. The Virginia Society of Ornithology and several of its chapters helped fund our efforts. This is Rice Center for Envi- ronmental Life Sciences Research Contribution No. 001.

LITERATURE CITED

Askenmo, C. 1977. Effects of addition and removal of nestlings on nestling weight, nestling survival, and female weight loss in the Pied Flycatcher Fi- cedula hypoleuca (Pallas). Ornis Scandinavica 8: 1-8.

Best, L. B. 1977. Patterns of feeding Field Sparrow young. Wilson Bulletin 89:625-627.

Blem, C. R. 1990. Avian energy storage. Current Or- nithology 7:59-1 14.

Blem, C. R. and L. B. Blem. 1991. Nest box selection by Prothonotary Warblers. Journal of Field Orni- thology 62:299-307.

Blem, C. R. and L. B. Blem. 1992. Prothonotary War- blers nesting in nest boxes: clutch size and timing in Virginia. Raven 63:15-20.

Blem, C. R. and L. B. Blem. 1994. Composition and microclimate of Prothonotary Warbler nests. Auk 111:197-200.

Blem, C. R., L. B. Blem, and C. I. Barrientos. 1999. Relationships of clutch size and hatching success to age of female Prothonotary Warblers. Wilson Bulletin 111:577-581.

Bryant, D. M. 1988. Energy expenditure and body

mass changes as measures of reproductive costs in birds. Functional Ecology 2:23-34.

Cavitt, J. F. and C. F. Thompson. 1997. Mass loss in breeding House Wrens: effects of food supple- ments. Ecology 78:2512-2523.

Cichon, M. 2001. Body-mass changes in female Col- lared Flycatchers: state-dependent strategy. Auk 1 18:550-552.

Conrad, K. F. and R. J. Robertson. 1993. Patterns of parental provisioning by Eastern Phoebes. Condor 95:57-62.

De Laet, J. F. and A. A. Dhondt. 1989. Weight loss of the female during the first brood as a factor influencing second brood initiation in Great Tits Parus major and Blue Tits P. caeruleus. Ibis 131: 281-289.

Freed, L. A. 1981. Loss of mass in breeding wrens: stress or adaptation? Ecology 62:1 179-1 186.

Garcia, P. F. J., M. S. Merkle, and R. M. R. Barclay. 1993. Energy allocation to reproduction and main- tenance in Mountain Bluebirds ( Sialia currocoi- des ): a food supplementation experiment. Cana- dian Journal of Zoology 71:2352-2357.

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

Hillstrom, L. 1995. Body mass reduction during re- production in the Pied Flycatcher Ficedula hypo- leuca: physiological stress or adaptation for low- ered costs of locomotion? Functional Ecology 9: 807-817.

Hinsley, S. A. 2000. The costs of multiple patch use by birds. Landscape Ecology 15:765-775.

Hussell, D. J. T. 1972. Factors affecting clutch size in Arctic passerines. Ecological Monographs 42: 317-364.

Johnson, R. K., R. R. Roth, and J. T. Paul, Jr. 1990. Mass variation in breeding Wood Thrushes. Con- dor 92:89-96.

Kennard, J. H. 1975. Longevity records of North American birds. Bird-Banding 46:55-73.

Knapton, R. W. 1984. Parental feeding of nestling Nashville Warblers: the effects of food type, brood-size, nestling age, and time of day. Wilson Bulletin 96:594-602.

Merila, J. and D. A. Wiggins. 1997. Mass loss in breeding Blue Tits: the role of energetic stress. Journal of Animal Ecology 66:452-460.

Merkle, M. S. and R. M. R. Barclay. 1996. Body mass variation in breeding Mountain Bluebirds Sialia currucoides : evidence of stress or adapta- tion for flight? Journal of Animal Ecology 65: 401-413.

Moreno, J. 1989a. Body-mass variation in breeding Northern Wheatears: a field experiment with sup- plementary food. Condor 91:178-186.

Moreno, J. 1989b. Strategies of mass change in breed- ing birds. Biological Journal of the Linnean So- ciety 37:297-310.

Nice, M. M. 1937. Studies in the life history of the

12

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

Song Sparrow. I. Transactions of the Linnean So- ciety of New York 4:1-247.

Norberg, R. A. 1981. Temporary weight decrease in breeding birds may result in more fledged young. American Naturalist 118:838-850.

Nur, N. 1984. The consequences of brood size for breeding Blue Tits. I. Adult survival, weight change and the cost of reproduction. Journal of Animal Ecology 53:479-496.

Omland, K. E. and T. W. Sherry. 1994. Parental care at nests of two age classes of male American Red- start: implications for female mate choice. Condor 96:606-613.

Petit, L. 1989. Breeding biology of Prothonotary War- blers in riverine habitat in Tennessee. Wilson Bul- letin 101:51-61.

Pinkowski, B. C. 1978. Feeding of nestling and fledg- ling Eastern Bluebirds. Wilson Bulletin 90:84-98.

Podlesak, D. W. and C. R. Blem. 2001. Factors af- fecting growth of Prothonotary Warblers. Wilson Bulletin 113:263-272.

Podlesak, D. W. and C. R. Blem. 2002. Determina- tion of age of nestling Prothonotary Warblers. Journal of Field Ornithology 73:33-37.

Ricklefs, R. E. 1974. Energetics of reproduction in birds. Pages 152-291 in Avian energetics (R. A.

Paynter, Jr., Ed.). Publications of the Nuttall Or- nithological Club, no. 15. Cambridge, Massachu- setts.

Ricklefs, R. E. and D. J. T. Hussell. 1984. Changes in adult mass associated with the nesting cycle in the European Starling. Ornis Scandinavica 15: 155-161.

Rytkonen, S., K. Koivula, and M. Orell. 1996. Pat- terns of per-brood and per-offspring provisioning efforts in the Willow Tit Parus montanus. Journal of Avian Biology 27:21-30.

Sanz, J. J. and J. Moreno. 1995. Mass loss in brood- ing female Pied Flycatchers Ficedula hypoleuca : no evidence for reproductive stress. Journal of Avian Biology 26:313-320.

Sanz, J. J., J. M. Tinbergen, M. Orell, and S. Ryt- konen. 1998. Daily energy expenditure during brood rearing of Great Tits Parus major in north- ern Finland. Ardea 86:101-107.

SAS Institute, Inc. 2000. SAS/STAT user’s guide, ver. 8.2. SAS Institute, Inc., Cary, North Carolina.

Trivers, R. L. 1972. Parental investment and sexual selection. Pages 136-179 in Sexual selection and the descent of man (B. Campbell, Ed.). Aldine Press, Chicago, Illinois.

Zar, J. J. 1999. Biostatistical analysis. Prentice-Hall, Upper Saddle River, New Jersey.

The Wilson Journal of Ornithology 1 18(1): 13— 22, 2006

THE REDISCOVERY AND NATURAL HISTORY OF THE WHITE-MASKED ANTBIRD ( PITHYS CASTANEUS)

DANIEL F. LANE,1 6 THOMAS VALQUI H.,1 2 JOSE ALVAREZ A.,1 2 3 JESSICA ARMENTA,25 AND KAREN ECKHARDT4 5 6

ABSTRACT. — In July 2001, a Louisiana State University Museum of Natural Science expedition rediscovered the White-masked Antbird ( Pithys castaneus ) at a site along the Rfo Morona in northwestern Departmento Loreto, Peru. Prior to this rediscovery, the species was known only from the type specimen, taken in 1937, and nothing was recorded concerning its natural history. The lack of additional specimens led to speculation that P. castaneus was a hybrid. Here, we present data demonstrating that the White-masked Antbird is a valid species, and we report the first observations of its behavior, habitat, morphology, and voice. Received 14 January 2005, accepted 1 1 October 2005.

In 1938, Berlioz (1938) described a distinc- tive new species of antbird in the genus Pith- ys— until then considered monotypic — from a single specimen collected by Ramon Olalla on 16 September 1937 at “Andoas, lower [Rio] Pastaza, eastern Ecuador.” This new species, the White-masked Antbird ( Pithys castaneus ), has remained one of the most intriguing mys- teries of Neotropical ornithology for over 60 years (see David and Gosselin 2002 for gen- der of scientific name). Besides the collector, no biologist had ever seen the bird alive, and there is no information on the species’ natural history or preferred habitat. The type locality, “Andoas,” is particularly intriguing in that at least three sites in the Pastaza area bear this name (Stevens and Traylor 1983, Paynter 1993), and according to T. Mark {in lift.), we may never really know the true location of the type locality.

The type specimen, a male (contra Ridgely and Tudor 1994), is housed at the Paris Mu- seum in France. According to Berlioz (1938, 1948), it was part of a collection that included three specimens of White-plumed Antbird (P. albifrons peruvianas ) and therefore appeared

1 Louisiana State Univ. Museum of Natural Science, 1 19 Foster Hall, Baton Rouge, LA 70803, USA.

2 Dept, of Biological Sciences, Louisiana State Univ., Baton Rouge, LA 70803, USA.

3 Inst, de Investigaciones de la Amazonia Peruana (IIAP), Av. Quinones Km. 2.5, Iquitos, Peru.

4 Museo de Historia Natural de la Univ. Nacional Mayor de San Marcos, Apartado 14-0434, Lima, Peru.

5 Current address: Dept, of Biological Sciences, P.O. Box 413, Lapham Hall, Univ. of Wisconsin, Milwau- kee, WI 53201, USA.

6 Corresponding author; e-mail: dlane@lsu.edu

to be a sympatric congener. It differed from P. albifrons in its larger size, its lack of any gray on the body, and its lack of elongated plumes on the face or throat.

Decades passed without any additional re- cords of P. castaneus. Subsequent authors doubted the validity of the species, and many suggested that it represented nothing more than a hybrid of P. albifrons and another ant- bird species (Sibley and Monroe 1990, Schu- lenberg and Stotz 1991, Collar et al. 1992, Stattersfield and Capper 2000, Ridgely and Greenfield 2001b). Willis (1984) and person- nel at the Philadelphia Academy of Natural Sciences (ANSP; Collar et al. 1992, Ridgely and Tudor 1994) searched without success for P. castaneus along the upper Rio Pastaza in Peru and Ecuador, respectively.

Thus, when our Louisiana State University Museum of Natural Science (LSUMZ) orni- thological field team visited several sites in northwestern Departamento Loreto, Peru, from May through July 2001 , it was with great surprise that we found P. castaneus to be fair- ly common at one of our field sites. The main goal of our fieldwork was to inventory the avifauna of two isolated patches of varillal (white sand) forest (see Whitney and Alvarez 1998; Alvarez and Whitney 2001, 2003). One of these forest patches was in the interfluvium between the Morona and Santiago rivers in northern Peru, north of the Rio Maranon, only about 60 km west of the Rfo Pastaza, and it was there that we found P. castaneus.

Remarkably, while reviewing specimen ma- terial at the Museo de Historia Natural de la Universidad Mayor San Marcos (MUSM),

13

14

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

FIG. 1 . Known localities for Pithy s castaneus in northwestern Departmento Loreto, Peru. The star represents suspected location of “Andoas,” the type locality, on the Rio Pastaza (Berlioz 1938). The square represents the location of the species’ rediscovery in July 2001 on the west bank of the Rio Morona (04° 17' S, 77° 14' W). The Cordillera Campanquis lies between the ribs Morona and Santiago, immediately to the west of our field site.

Lima, in November 2002, we discovered two additional specimens of P. castaneus (one adult and one juvenile). These specimens were reportedly taken somewhere in the Cordillera Campanquis region on the border of Depart- mentos Amazonas and Loreto between the Morona and Santiago rivers (see Fig. 1), in the mid- to late 1990s by a Peruvian anthro- pologist, Andres Treneman (I. Franke J. pers. comm.). Unfortunately, no additional speci- men data are available, and the collector could not be contacted for additional information.

METHODS

Locality. — We established a campsite on the west bank of the Rio Morona about 54 km north-northwest of its mouth (04° 17' S, 77° 14' W; Fig. 1), Departmento Loreto. The

study site was on the south side of the mouth of Quebrada Cashacano, a right-bank tributary of the Rfo Morona, about 2.3 km north of the village of Tierra Blanca. We observed and made a general collection of birds at this site between 2 and 21 July 2001. Our camp was set up in a clearing of a homestead abandoned about 30 years earlier and which, reportedly, has been reinhabited since our visit (B. Walker pers. comm.). A preexisting trail, used for the harvest of palm fronds for thatched-roof con- struction, led directly into white-sand forests for about 2 km. Another trail, cut along the bluff above the Morona, connected the camp with the village of Tierra Blanca. From this trail, at least another three trails also entered the varillal forest. Additional trails were cut near camp for census routes and net lanes;

Lane et al. • REDISCOVERY OF WHITE-MASKED ANTBIRD

15

most trails were in varillal, but three also en- tered the adjacent varzea (seasonally inundat- ed) forest. We also found two patches of richer clay-soil terra firme forest north and south of the surveyed varillal forest patch, into which we cut two trails.

Habitat. — Most of the forest where P. cas- taneus was observed — particularly away from major water bodies — grew on very moist, white-sand soils. Numerous areas of wet, swampy conditions indicated a high water ta- ble. The terrain was without significant relief, but throughout the varillal forest were many small depressions where water accumulated (particularly after rains), presumably pits re- sulting from tree-falls. The soil consisted of rather coarse sand with stones of up to 5 cm in diameter (up to 15 cm in the small creeks that transected the forest interior). Using a nat- ural cut at the Rio Morona riverbank for ref- erence, the sandy soil is approximately 4 m deep at the river’s edge. Typical of many var- illal forests, a thick layer of dead leaves and humus covered the forest floor (Ruokolainen and Tuomisto 1993, 1998; Richards 1996). The forest canopy of the varillal was relative- ly even, with a height of about 20 to 30 m. The relative absence of buttressed trees is typ- ical of varillal forests (Richards 1996); how- ever, many such trees were present in more humid forest areas at the Morona site. As has been noted in other varillal forests (Anderson 1981, Richards 1996), there were few lianas, and epiphytic growth was negligible.

Data collection. — We collected specimens using mist nets and shotguns. Permits for specimen collection were issued by Peru’s In- stitute Nacional de Recursos Naturales (IN- RENA). Specimens were deposited into the collections of LSUMZ and MUSM. Skull os- sification, gonad information, and presence of fat in prepared specimens were determined following standard LSUMZ specimen prepa- ration protocol. Natural history information was acquired through opportunistic (not sys- tematic) encounters with P. castaneus. Spec- trograms of voice recordings were prepared using Canary sound analysis software (Charif et al. 1995).

Specimens examined. — Pithy s castaneus : Peru: Loreto; west bank of Rio Morona, —54 km NNW of the mouth, 140 m elevation (04° 17' S, 77° 14' W) (LSUMZ 172973, 172974,

172975, 172976 [skeleton and partial skin], 172977, 172978, 172979 [skeleton and partial skin], MUSM 23504, 23505, 23506, 23507; DFL 1646 [skeleton, uncataloged], TVH 399 [alcohol, uncataloged]).

Pithys albifrons : Ecuador: Pastaza; Coco- naco, 300 m elevation (LSUMZ 83237); Peru: Amazonas; Huampami, —215 m elevation (LSUMZ 84917), Chiriaco, -320 m elevation (LSUMZ 78514, 88018, 88019, 88022); Lor- eto; Libertad, S bank of Rio Napo, 80 km N of Iquitos, 120 m elevation (LSUMZ 1 10094, 110096, 110097, 110098, 110099, 110100, 1 10102, 1 10103, 1 10104, 1 10105); 157 km by river NNE of Iquitos, N of Rio Napo, 110 m elevation (LSUMZ 110106, 110109, 110112, 110113).

Gymnopithys leucaspis: Peru: Loreto; west bank of Rio Morona, —54 km NNW of the mouth, 140 m elevation (04° 17' S, IT 14' W) (LSUMZ 172985); Quebrada Oran, -5 km N of Rio Amazonas, 85 km NE of Iquitos, 1 10 m elevation (LSUMZ 119884, 1 19885, 119886, 119887, 119890, 119891, 119892, 119893).

Phlegopsis erythroptera: Ecuador: Sucum- bios; Limoncocha, 300 m elevation (00° 24' S, 76° 37' W) (LSUMZ 70916, 70917, 70919, 83314). Peru: Loreto; W bank of Rio Morona, —54 km NNW of the mouth, 140 m elevation (04° 17' S, IT 14' W) (LSUMZ 173001); 1.5 km S of Libertad, S bank of Rio Napo, 80 km N of Iquitos, 120 m elevation (LSUMZ 110213, 110215, 110217); 1 km N of Rio Napo, 157 km by river NNE of Iquitos, 110 m elevation (LSUMZ 110219); lower Rio Napo region, E bank of Rio Yanayacu, —90 km N of Iquitos, 120 m elevation (LSUMZ 115573).

Rhegmatorhina melanosticta: Peru: Ama- zonas; headwaters of Rio Kagka (of Rio Ce- nepa), —790 m elevation (04° 16' S, 78° 09' W) (LSUMZ 88028, 88029); San Martin; -15 km by trail NE of Jirillo on trail to Balsa- puerto, 1,350 m elevation (LSUMZ 116947); Huanuco; —35 km NE Tingo of Marfa, Ha- cienda Santa Elena, —1,000 m elevation (LSUMZ); Pasco; Abra Aguachini, —30 km SW of Puerto Bermudez, 1,020 m elevation (LSUMZ 130274); Pasco; Puellas, km 41 on Villa Rica-Puerto Bermudez highway, 950 m elevation (LSUMZ 106073, 106074, 106078).

16

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

RESULTS

Specimen data. — We collected 13 speci- mens of Pithys castaneus during our visit to the Rio Morona site. We prepared nine as study skins (from which several trunk skele- tons were saved), three as complete skeletons (from which two partial skin specimens were saved), and one was preserved whole in al- cohol. Mass and lengths of flat-wing, tail, tar- sus, and culmen (from distal edge of the nares to bill tip) of all specimens are presented and compared with measurements of the P. cas- taneus holotype and other Peruvian ant swarm-following antbirds (Table 1).

Three of the 12 specimens in “adult” plum- age (LSUMZ 172973, MUSM 23504, MUSM 23507) still possessed a bursa of Fabricius and one had an incompletely ossified skull (75% ossification), suggesting that first basic plum- age is acquired quickly and is nearly indistin- guishable from definitive plumage (but see be- low). One specimen (LSUMZ 172978) was a male still largely in juvenal plumage (skull os- sification 50%, bursa 8X6 mm). Of the 12 specimens dissected, only 2 — both with im- mature characters — were reported to have subcutaneous fat deposits: “trace fat” in one and “light fat” in the other. Six of 12 speci- mens dissected exhibited trace or light body molt. Seven individuals had asymmetrical wing molt, and seven had asymmetrical tail molt. Stomach contents were reported as “in- sect parts” for all specimens in which the stomachs were not empty. The guts of two specimens were infested with nematodes.

Variation in the series. — Twelve speci- mens— 5 males and 7 females — exhibited similar plumage, with no sexual dichroma- tism. All these adults appeared to match the description of P. castaneus and the photos of the holotype very closely. Of the specimens in “adult” plumage, two that appeared to be in their first year (see above) have very sparse, light-grayish scaling on the center of the throat (unmarked white in all other individu- als), suggesting that it may be an age-related character. Otherwise, plumage characters were uniform among all the “adult” specimens. The juvenal-plumaged bird differs in being washed with colder brown overall, particular- ly on the breast, vent, and center of the back. Furthermore, the white of the juvenile’s face

dddddbod

r- (N o

+1 +1 +1 +| +i +| +i +i +i +i "qqa^qooqaoi;^ ri n' d O ri ri ri ^

in oo io o o o o o

ro +| +i +i +i +i <N

o i; q in o mi oi —i no

04 04 04 Ol 04

On On ~ On 04 — : — ' o

+1 +1 +1 +1 +1

oj cq ^ no on

oi oi b *n cd

in m m

O O- O »/"> CO

04 +| +| +| +| +| 00

oo no >n >n >n

h n 1; q h

oi — i oi — oi

+1 +1 +1 +1 +1

q o in o;

On d oo oo rd

04 CO < —

04

— no m O O i/n

a C p &

a. a

£1

5 c M S C « 2 ^ ^ c .5

s c

.a £

"q "q 5.

£ £ £ £ £ 6'

Gymnopithys leucaspis (females) 5 23.8 ± 2.3 71.9 ± 1.3 42.9 ± 2.1 25.8 ± 1.2

Phlegopsis erythroptera (males) 5 58.4 ± 5.2 91.6 ± 1.7 63.3 ± 2.4 33.5 ±1.7

Phlegopsis erythroptera (females) 5 58.2 ± 7.0 88.0 ± 1.3 59.0 ± 1.0 32.0 ± 1.4

Rhegmatorhina melanosticta (males) 4 30.0 ±1.9 81.3 ± 5.2 53.0 ± 2.0 27.6 ±1.2

Rhegmatorhina melanosticta (females) 5 33.0 ± 4.7 78.0 ± 2.8 49.8 ± 2.5 27.4 ± 0.8

Lane et al. • REDISCOVERY OF WHITE-MASKED ANTBIRD

17

TABLE 2. Number of individuals per species attending army ant swarms ( Eciton burchelli and Labidus praedator ) with Pithys castaneus , Departmento Loreto, Peru, July 2001. Columns represent individual swarms. Only swarms observed for >15 min were included.

Date (ant swarma)

4 July	6 July	6 July	8 July	10 July	11 July	12 July	14 July	17 July
(E)	(E)	(E)	(E)	(L.)	(E)	(E)	(L)	(E)

Pithys castaneus	2	4	3	3	1	1	4	4	3
Pithys albifrons	3	5	—	—	—	—	—	—	—
Phlegopsis erythroptera	—	2	—	—	—	—	—	—	—
Gymnopithys leucaspis	5	4	2	2	—	3	2	4	4
Hylophylax poecilinota	—	2	2	1	1	—	—	—	—
Percnostola arenarum	1	—	1	—	—	—	1	1	2
Dendrocolaptes certhia	1	3	—	—	—	—	—	—	—
Dendrocincla merula	—	—	—	—	—	—	—	1	1
Xiphorhynchus ocellatus	2	2	—	—	—	1	—	1	1
Deconychura longicauda	1	—	—	—	—	—	—	—

a E = Eciton burchelli , L = Labidus praedator.

was restricted to the area between the eye and gape and a longitudinal line along the center of the throat. This specimen’s dark head mark- ings were more extensive than those on defin- itive-plumaged birds, and they were a duller, sooty, dark brown (see frontispiece).

Soft-part colors were relatively uniform across most specimens. The irides were brown or dark brown (all soft-part colors taken from tag data recorded at time of preparation) in nine specimens with adult characters, dark gray-brown in the three specimens with first- basic characters, and dark gray in the juvenile. Thus, iris color evidently changes from gray to dark brown as an individual ages. In all specimens, the maxilla was blackish-slate with a silvery-white tomium, the latter con- stricted at mid-bill in some individuals. Man- dible coloration varied more. Most adults had a mostly silvery-white tomium with blackish- slate color on the gonys and base of the man- dible (except the tomium). Approximately the distal half of the juvenile’s bill was silvery- white, and the mouth interior was dark gray. The tarsus color of adult individuals was brownish-orange or ochre-orange; the juve- nile’s tarsi were dirty yellow with a gray tinge. The toes were dirty yellow, pale orange, or dull saffron yellow; the claws of the juvenile bird were gray.

Behavioral observations. — Our initial ob- servations of P. castaneus were made by TVH and DFL at a swarm of Eciton burchelli army ants on 4 July 2001, when the first specimens were collected. Based on our observations, we

were confident in labeling P. castaneus a pro- fessional army ant-follower ( sensu Willis 1967). We never saw it foraging away from army ant swarms and observed it attending swarms of two army ant species: Eciton bur- chelli and Labidus praedator. For at least 1 2— 15 min on 8 July, JAA observed a single in- dividual of P. castaneus with a female Scale- backed Antbird ( Hylophylax poecilinota ) fol- lowing a swarm of L. praedator ants that oc- cupied less than 10 m2 of the forest floor. The bird’s behavior was similar to that of others observed following swarms of E. burchelli. Both the P. castaneus and the H. poecilinota individuals left the swarm for 3-4 min, only to return later. Also observed attending swarms of L. praedator (although independent of the above observation) were Allpahuayo Antbirds ( Percnostola arenarum), a species previously unknown as an ant-follower (Isler et al. 2001, Zimmer and Isler 2003), and Bi- colored Antbirds ( Gymnopithys leucaspis). On four occasions on different days, we observed a single individual of P. castaneus quietly passing through the forest without foraging, suggesting movement between ant swarms or between an ant swarm and a nest (Willis 1981). In Table 2, we present the attendance of regular ant-following species observed at swarms at the Morona site.

Most often, P. castaneus was observed at or near the broad front of a moving ant col- umn. Individuals tended to perch <0.5 m above ground and frequently dropped to the forest floor to investigate leaf litter or capture

18

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

arthropods. Birds often were observed attend- ing a swarm for 5 to 15 min at a time and then leaving the swarm (at least once while carrying a food item) for roughly equal peri- ods of time. On at least one such occasion, a pair of P. castaneus was observed joining a family group of G. leucaspis moving between what appeared to be two column heads (about 30 m apart) of a single E. burchelli ant swarm. Willis (1981) reported similar behavior for P. albifrons. On another occasion, a single indi- vidual was seen moving around a standing hollow tree in which a swarm of E. burchelli had bivouacked the previous evening, but had not yet started its morning activity.

Most of the professional ant-following thamnophilids at the Morona site regularly made exaggerated tail “pounding” or “wag- ging” movements (terms following Zimmer and Isler 2003) while foraging at ant swarms, especially upon returning to a perch after pouncing on a prey item, or when agitated by the presence of an observer. P. castaneus was not observed regularly using any such tail movement. Only once or twice did we notice an individual pound its tail, usually after a pouncing attack on prey; the tail movement was made once and not repeated. By contrast, DFL noted that the G. leucaspis almost con- stantly wagged their tails laterally, although this contrasts with the published observations of others (e.g., Zimmer and Isler 2003). In ad- dition, DFL observed both P. albifrons and the Reddish-winged Bare-eye ( Phlegopsis er- ythroptera ) regularly pounding their tails downward (also see Willis 1981, 1984; Zim- mer and Isler 2003). We were unable to de- termine whether such tail movements are in- tended as a form of inter- or intraspecific “body language” among swarm attendants, as a sign of agitation, or as a form of flushing insect prey. Nevertheless, the relative lack of such tail-moving behavior in P. castaneus seems noteworthy. Willis (1968) reports that the monotypic genus Skutchia also lacks ste- reotypic tail-moving behavior, but other ob- servers contest this (B. M. Whitney pers. comm.).

In our observations of ant-following birds at the Morona site (Table 2), we noted several occurrences of one ant-following species sup- planting another near the leading edges of ant swarms and took this to represent a domi-

nance hierarchy among the attendant species (see Willis 1967, 1981). From our observa- tions, we concluded that the dominance hier- archy (from most to least dominant) was Phle- gopsis erythroptera , Pithys castaneus , and G. leucaspis. Other swarm-attending antbirds, in- cluding Pithys albifrons, noticeably avoided the leading edge of the swarm when any of the other professional ant-followers were pre- sent. Our observations of the last species agree with those of Willis (1981), who also termed P. albifrons a subordinate ant swarm attendant. Since the dominance hierarchy sug- gested above has a positive correlation to overall body size, we suggest that size may be the ultimate cause (or, alternatively, a proxi- mate cause — i.e., a source of maintenance) of the hierarchy (see Table 1).

Voice. — We recorded at least seven distinct vocalizations from P. castaneus (Fig. 2), in- cluding a mewed whistle that rises in frequen- cy (Fig. 2A). This is a single note often given quietly, although occasionally it can be quite loud, and we suspect represents the species’ “loudsong” (such as it is). To our knowledge, P. albifrons does not give a true loudsong (sensu Willis 1967, Isler et al. 1998, Isler and Whitney 2002, Zimmer and Isler 2003) as do most other thamnophilids. However, the spe- cies is known to produce a vocalization sim- ilar to that described above for P. castaneus : a rarely heard, weak, mewing whistled vocal- ization that falls in frequency and is suspected to serve as a song (Willis 1981, Isler and Whitney 2002; Fig. 2B). The whistled notes of the loudsong of P. castaneus appear to be punctuated by occasional quiet, chiming notes (Fig. 2C), perhaps an integral part of the loud- song. Song intervals can be as short as 2 sec but often are longer.

P. castaneus also produced two vocaliza- tions when alarmed or when agitated by play- back of what we believed was the species’ song (see below). These notes of agitation were interspersed with sharp chattered “chit!” calls (Fig. 2D), similar to the “chip” calls de- scribed for P. albifrons by Willis (1981). An- other vocalization given by agitated birds was a louder, higher-pitched “chirr,” with the in- dividual notes more distinct (Fig. 2E) than in the undisturbed chirr call (see below). Occa- sionally, the agitated chirr commenced with a chit note (Fig. 2F). While giving these vocal-

Lane et al. • REDISCOVERY OF WHITE-MASKED ANTBIRD

19

FIG. 2. Sound spectrograms of antbird vocalizations. Unless otherwise noted, all recordings were made by D. F. Lane at our Rfo Morona locality, Departmento Loreto, Peru, July 2001. (A) “Song” of Pithys castaneus. (B) “Song” of Pithys albifrons (T. A. Parker, III, and G. F. Budney, from Isler and Whitney 2002). (C) “Chime” of Pithys castaneus. (D) “Chit” of Pithys castaneus. (E) Agitated “Chirr” of Pithys castaneus. (F) “Chit-chirr” of Pithys castaneus. (G) “Mew” of Pithys castaneus (J. Alvarez A.). (H) “Chirr” of Pithys castaneus. (I) “Chirr” of Pithys albifrons. (J) “Chirr” of Gymnopithys leucaspis. (K) “Chirr” of Phlegopsis erythroptera.

izations of agitation, one male (sex confirmed by collection), was observed perched on a horizontal branch at the edge of a treefall gap about 2 m above the ground. This was the highest we ever observed the species to perch, and was likely an agitation response to play- back of the song. On one occasion, a distinct, quiet, mewing “eaaah” call was given by two individuals while close to one another; we in- terpret this as some sort of contact call or “softsong” within the pair (Fig. 2G).

The most common vocalization was a call given by individuals while foraging at ant swarms. This was a deep chirr call (terms fol- lowing Willis 1967, Zimmer and Isler 2003; Fig. 2H), similar to vocalizations given by most professional ant-following thamnophil-

ids when attending ant swarms, and suspected to be a means of maintaining individual for- aging space at the swarm (Willis 1967; M. L. Isler in litt.). When the chirr of P. castaneus was heard simultaneously with those of most of the other species of professional ant-fol- lowers at a swarm, it sounded generally loud- er, of lower overall frequency, and descended less obviously (see Fig. 2H-2K). Only the chirr call of Phlegopsis erythroptera (Fig. 2K) reaches a frequency as low as that of Pithys castaneus , but the former can be distinguished easily by a higher, more metallic introductory sound and a more sharply descending com- ponent. The chirr call of Phlegopsis erythrop- tera was louder than that of Pithys castaneus on occasion, but this appeared to be influ-

20

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

enced by emotional state and was not always the case.

Playback experiments using recordings of the suspected song elicited varying reactions from individuals: some responded immediate- ly, giving agitated calls and posing on exposed perches that were higher than typical perches (see above), while others approached silently to investigate. On two occasions, individuals approached only after 2-3 min of playback. Playback of chirr calls resulted in a quiet, cu- rious approach at best.

DISCUSSION

Taxonomic status of the species. — Whereas the generic allocation of Pithys castaneus has been considered dubious, we believe that phe- notypic characters such as the species’ song- like vocalization, its bold chestnut plumage, black hood and white face, and its saffron- yellow legs all suggest a close relationship with P. albifrons. Furthermore, R. T. Brum- field and J. G. Tello (unpubl. data) have been building a molecular phylogeny of the Tham- nophilidae, and have found P. castaneus and P. albifrons to be sister taxa.

Potential habitat specialization. — Based on our observations, we suspect that P. castaneus is restricted to varillal forests. We should note, however, that we observed and mist-net- ted P. castaneus individuals that had followed ant swarms from varillal into varzea forest immediately adjacent to our campsite, and twice we recorded individuals on richer, hilly terra firme forest within 300 m of typical var- illal habitat. We never encountered Hairy- crested Antbird ( Rhegmatorhina melanosticta ) at the Morona site and wonder whether it may be replaced by the similarly sized P. casta- neus (see Table 1) in the region or (more like- ly) habitat. We can find no evidence that R. melanosticta inhabits the region between the rios Santiago and Pastaza, but it is quite pos- sible that this is due to poor sampling as it is to true absence. If R. melanosticta competi- tively excludes P. castaneus outside the Mo- rona-Pastaza varillal forest, this may explain the restricted distribution of the latter species. Furthermore, if varillal forest habitat was not included in the searches conducted by Willis and the ANSP expedition along the Pastaza, their failure to encounter the species may be

explained by the possible habitat specializa- tion of P. castaneus.

Potential distribution of Pithys castaneus. — Landsat imagery, complemented with infor- mation from Instituto de Investigaciones de la Amazonia Peruana personnel and local peo- ple, shows what we interpret to be fairly large blocks of varillal forest embedded within a quadrangle formed by the Rio Maranon to the south, the Rio Morona to the east, the Rio Mayuriaga to the north, and the Cordillera Campanquls to the west. Besides this area, P. castaneus populations are likely to occur in similar forest along the Rio Pastaza in Loreto and probably into Ecuador. At present, we have no information about the existence of varillal forest at the latter sites. However, some indicator species of varillal forest have been found along the upper Rio Pastaza in Ec- uador (e.g.. Pompadour Cotinga, Xipholena punicea, and Red-fan Parrot, Deroptyus accip- itrinus\ Ridgely and Greenfield 2001a), sug- gesting that the area probably supports varillal forest habitat. We suspect that once such for- ests along the upper Rio Pastaza are located and surveyed, the mystery of the true position of the “Andoas” collecting locality finally will be unraveled.

Conservation. — The west bank of the Rio Morona, including the areas of varillal forest where our work was conducted, are part of the recently created Zona Reservada Santiago Co- maina, created in 1999. According to Peruvian legislation, its new status is temporary, but supposedly, it will be ranked as a definitive conservation unit in the future (National Park, National Reserve, National Sanctuary, or Communal Reserve). However, local leaders of the Federacion de Comunidades Indlgenas del Rio Morona informed us that they strongly oppose the creation of a reserve and will fight to prevent this action.

A branch of the North-Peruvian oil pipeline that transports oil from the upper Rio Pastaza passes through a large portion of varillal for- est as it crosses the Rio Mayuriaga on its way to the Rio Maranon. At present, this has meant the destruction of only a 50-m-wide swath of forest along the pipeline. However, an oil spill could have drastic consequences for this rath- er delicate habitat, particularly with its flat ter- rain and poor drainage. Furthermore, the pipe- line itself could represent a potential dispersal

Lane et al. • REDISCOVERY OF WHITE-MASKED AN'I BIRD

21

barrier for P. castaneus. It is also worth men- tioning that there are several plans to connect Ecuador’s Amazonian road network to the Rio Maranon. Anecdotal evidence suggests that many bird species of interior forest understory are averse to crossing large openings or other similar breaks, such as rivers or roads (Zim- mer and Isler 2003). Thus, gaps such as those associated with roads and pipelines may pose barriers to gene flow in populations of these understory species.

Population estimate. — During our stay we surveyed about 8 km2 of white-sand forests and encountered between six and eight differ- ent army ant swarms of E. burchelli and two of L. praedator. Based on our extrapolations, we estimate the number of P. castaneus to be between 18 and 26 individuals in the area we surveyed. If we consider the immediate area (the Morona-Santiago interfluvium) covered with varillal, then the population estimate of P. castaneus would be —1,300-2,500 individ- uals. Prior to our rediscovery of P. castaneus, the species was considered to be rare, with a very restricted global distribution, and prob- ably threatened (Bibby 1992, Stattersfield and Capper 2000). Considering the population es- timates and the potential threats presented here, we recommend changing the species’ status from Data Deficient to Vulnerable, ac- cording to the ranking criteria presented in Stattersfield and Capper (2000). If a road or any other invasive construction project threat- ens the white-sand forests between the rlos Morona and Santiago, then the species’ status should be upgraded to a category of higher risk.

Since our rediscovery of P. castaneus in July 2001, and our discovery of the two Tre- neman specimens in MUSM, we have been informed of two subsequent observations of P. castaneus by colleagues who visited our Morona site. Observers visited the site 22-24 June 2002 and 24 May 2003 (M. Levy, J. Nilsson, M. Sokol, and B. Walker pers. comm.). Both parties saw the species, but the 2002 observation was of multiple individuals and the observers regarded the species as “one of the most common birds” at the site. During the 2003 visit, however, only one in- dividual was observed, possibly because swarms of army ants were not easily encoun- tered then (an artifact of the season?).

ACKNOWLEDGMENTS

We thank J. P. O’Neill as the initiator and organizer of the 2001 expedition. Funding for the 2001 expedi- tion was received from the Coypu Foundation and a donation from the late R. B. Wallace. Additional fund- ing was provided to JAA by a Fulbright Scholarship. Permits for fieldwork were granted by INRENA, and we particularly appreciate the efforts of M. Prieto C. and R. Acero V. of that institution. D. Huachaca, M. Sanchez S., A. Urbay T., M. Pizango, M. Tenazoa, H. Pizango, F. Salazar, and R. Sandoval all provided lo- gistical support in the field. M. L. and P. R. Isler gra- ciously allowed us access to their data and recording collections, and freely provided the fine maps and sonograms for our figures. M. Levy, J. Nilsson, M. Sokol, and B. Walker all related information from their visits to the Morona site. T. Mark provided us with his Andoas manuscript, and he and J. W. Fitzpatrick pro- vided information and photos of the holotype at the Paris Museum. L. B. McQueen produced the illustra- tion that allowed us to recognize Pithys castaneus while in the field. T. S. Schulenberg provided us with rare reprints and data collected on Peruvian antbirds. I. Franke J. kindly allowed us access to the ornitho- logical collection in her care at MUSM, and assisted us in trying to find more details of the two “mystery” Pithys castaneus specimens there. This manuscript benefited from comments by M. L. and P. R. Isler, J. V. Remsen, Jr., T. S. Schulenberg, B. Walker, B. M. Whitney, K. J. Zimmer, and two anonymous reviewers.

LITERATURE CITED

Alvarez A., J. and B. M. Whitney. 2001. A new Zimmerius tyrannulet (Aves: Tyrannidae) from white sand forests of northern Amazonian Peru. Wilson Bulletin 113:1-9.

Alvarez A., J. and B. M. Whitney. 2003. New dis- tributional records of birds from white-sand for- ests of the northern Peruvian Amazon, with im- plications for biogeography of northern South America. Condor 105:552-566.

Anderson, A. B. 1981. White-sand vegetation of Bra- zilian Amazonia. Biotropica 13:199-210.

Berlioz, J. 1938. Pithys castanea, sp. nov. Bulletin of the British Ornithologists’ Club 58:90-91. Berlioz, J. 1948. Note critique sur le genre Pithys Vieillot (Formicariides). L’Oiseau et la Revue Frangaise d’Ornithologie 18:1-4. [in French] Bibby, C. J. 1992. Putting biodiversity on the map: priority areas for conservation. International Council for Bird Preservation, Girton, Cambridge, United Kingdom.

Charif, R. A., S. Mitchell, and C. W. Clark. 1995. Canary 1.2. Cornell Laboratory of Ornithology, Ithaca, New York.

Collar, N. J., L. P. Gonzaga, N. Krabbe, A. Mad- rono Nieto, L. J. Naranjo, T. A. Parker, III, and D. C. Wege. 1992. Threatened birds of the Amer- icas: the ICBP/IUCN Red Data Book, 3rd ed., part

22

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 1 18, No. 1, March 2006

2. Smithsonian Institution Press, Washington, D.C.

David, N. and M. Gosselin. 2002. The grammatical gender of avian genera. Bulletin of the British Or- nithologists’ Club 122:257-282.

Isler, M. L., J. Alvarez A., P. R. Isler, and B. M. Whitney. 2001. A new species of Percnostola antbird (Passeriformes: Thamnophilidae) from Amazonian Peru, and an analysis of species limits within Percnostola rufifrons. Wilson Bulletin 1 13: 164-176.

Isler, M. L., P. R. Isler, and B. M. Whitney. 1998. Use of vocalizations to establish species limits in antbirds (Passeriformes: Thamnophilidae). Auk 115:577-590.

Isler, P. R. and B. M. Whitney. 2002. Songs of the antbirds: Thamnophilidae, Formicariidae, and Conopophagidae (CD). Library of Natural Sounds, Cornell University, Ithaca, New York.

Paynter, R. A., Jr. 1993. Ornithological gazetteer of Ecuador. Harvard University Press, Cambridge, Massachusetts.

Richards, P. W. 1996. The tropical rainforests: an eco- logical study. Cambridge University Press, Cam- bridge, United Kingdom.

Ridgely, R. S. and P. J. Greenfield. 2001a. The birds of Ecuador: field guide. Cornell University Press, Ithaca, New York.

Ridgely, R. S. and P. J. Greenfield. 2001b. The birds of Ecuador: status, distribution, and taxonomy. Cornell University Press, Ithaca, New York.

Ridgely, R. S. and G. Tudor. 1994. The birds of South America. University of Texas Press, Austin.

Ruokolainen, K. and H. Tuomisto. 1993. La vege- tacion de terrenos no indundables (tierra firme) en la selva baja de la Amazonia peruana. Pages 1 39 — 1 53 in Amazonia Peruana: vegetacion humeda tropical en el llano subandino (R. Kalliola, M. Pu- hakka, and W. Danjoy, Eds.). Amazon Project of the University of Turku, Jyvaskyla, Finland.

Ruokolainen, K. and H. Tuomisto. 1998. La vege-

tacion natural de la zona de Iquitos. Pages 253- 365 in Geoecologia y desarrollo amazonico. Es- tudio integrado de la zona de Iquitos, Peru (R. Kalliola and S. Flores P, Eds.). Annales Univer- sitatis Turkuensis Series A II, vol. 1 14.

Schulenberg, T. S. and D. F. Stotz. 1991. The taxo- nomic status of Myrmeciza stictothorax (Todd). Auk 108:731-733.

Sibley, C. G. and B. L. Monroe, Jr. 1990. Distribu- tion and taxonomy of birds of the world. Yale University Press, New Haven, Connecticut.

Stattersfield, A. J. and D. R. Capper (Eds.). 2000. Threatened birds of the world: the official source for birds on the IUCN red list. BirdLife Interna- tional, Cambridge, United Kingdom, and Lynx Edicions, Barcelona, Spain.

Stevens, L. and M. A. Traylor, Jr. 1983. Ornitho- logical gazetteer of Peru. Harvard University Press, Cambridge, Massachusetts.

Whitney, B. M. and J. Alvarez A. 1998. A new Herpsilochmus antwren (Aves: Thamnophilidae) from northern Peru and adjacent Ecuador: the role of edaphic heterogeneity of terra firme forest. Auk 115:559-576.

Willis, E. O. 1967. The behavior of Bicolored Ant- birds. University of California Publications in Zo- ology 79:1-132.

Willis, E. O. 1968. Taxonomy and behavior of Pale- faced Antbirds. Auk 85:253-264.

Willis, E. O. 1981. Diversity in adversity: the behav- iors of two subordinate antbirds. Arquivos de Zoologia 30:159-234.

Willis, E. O. 1984. Phlegopsis erythroptera (Gould 1855) and relatives (Aves, Formicariidae) as army ant followers. Re vista Brasiliera Zoologia 2:165- 170.

Zimmer, K. J. and M. L. Isler. 2003. Family Tham- nophilidae (typical antbirds). Pages 448-681 in Handbook of birds of the world, vol. 8: broadbills to tapaculos (J. del Hoyo, A. Elliott, and D. Chris- tie, Eds.). Lynx Edicions, Barcelona, Spain.

The Wilson Journal of Ornithology 1 1 8( 1 ):23 — 35, 2006

NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS IN SAND SAGEBRUSH PRAIRIE OF SOUTHWESTERN KANSAS

JAMES C. PITMAN,'-1 2 3 4-8 CHRISTIAN A. HAGEN,1-5 BRENT E. JAMISON,1-6 7 ROBERT J. ROBEL,1 THOMAS M. LOUGHIN,2 AND ROGER D. APPLEGATE57

ABSTRACT. — Despite the fact that the Lesser Prairie-Chicken ( Tympanuchus pallidicinctus) is a species of conservation concern, little is known about its nesting ecology, particularly in sand sagebrush (. Artemisia filifolia) habitats. To find and monitor nests, we captured and equipped 227 female Lesser Prairie-Chickens with trans- mitters (87 yearlings, 1 17 adults, and 23 of unknown age) from 1997 to 2002 in southwestern Kansas. Apparent nest success was similar for yearlings (31%, n = 74) and adults (27%, n = 97) but differed marginally ( P = 0.090) between first nests (29%) and renests (14%). An estimated 31% of females that were unsuccessful in their first nesting attempt initiated a second nest. The probability that a female would initiate a second nest after failure of the initial attempt was negatively influenced by the day of incubation on which the initial attempt failed. Over 95% of all nests were initiated and completed between 5 May and 2 July. The primary cause of nest failure was predation by coyotes ( Canis latrans ) and gopher snakes ( Pituophis melanoleucus ). Mean clutch size, egg fertility, hatching success, nesting and renesting frequency, and incidence of interspecific parasitism were all similar across years and between yearlings and adults. Distances between nest sites were used as an index to nest-site fidelity between first nests and renests and for across-year nesting attempts. Mean distances between first nests and renests were similar for yearlings (1,071 m) and adults (1,182 m). Mean distance between nests constructed by the same female in subsequent years (918 m) did not differ between age classes or success of the first year’s nest. Most females (80%) nested closer to a lek other than the lek where they were captured. Received 24 January 2005, accepted 21 September 2005.

Range-wide, Lesser Prairie-Chickens ( Tym- panuchus pallidicinctus ) have declined by an estimated 97% since the 1800s (Crawford 1980, Taylor and Guthery 1980). In Kansas, Lesser Prairie-Chickens are most abundant in the western part of the state — south of the Ar- kansas River in mixed and shortgrass prairie dominated by sand sagebrush {Artemisia fili- folia). They also occur in mixed grass prairie north of the Arkansas River, but this habitat is generally devoid of sand sagebrush. Lesser

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

2 Dept, of Statistics, Kansas State Univ., Manhattan, KS 66506, USA.

3 Survey and Research Office, Kansas Dept, of Wildlife and Parks, P.O. Box 1525, Emporia, KS 66801, USA.

4 Current address: Survey and Research Office, Kan- sas Dept, of Wildlife and Parks, P.O. Box 1525, Em- poria, KS 66801, USA.

5 Current address: Oregon Dept, of Fish and Wild- life, 61374 Parrell Rd., Bend, OR 97702, USA.

6 Current address: Missouri Dept, of Conservation, P.O. Box 368, Clinton, MO 64735, USA.

7 Current address: Tennessee Wildlife Resources Agency, Ellington Agricultural Center, P.O. Box 40747, Nashville, TN 37204, USA.

8 Corresponding author; e-mail: jimp@wp.state.ks.us

Prairie-Chickens currently occupy 31 of 39 counties believed to compose their historical distribution in Kansas, but counts of leks and individual birds suggest that Lesser Prairie- Chickens have experienced significant de- clines since 1964 (Jensen et al. 2000).

The mechanisms responsible for Lesser Prairie-Chicken population declines have not been identified; however, aspects of nesting ecology could be influential (Peterson and Sil- vy 1996, Wisdom and Mills 1997). Thus, identifying age-specific variation in nesting variables is important to understanding a spe- cies’ demography or life-history strategy (Pat- ten et al. 2005). Most research on Lesser Prai- rie-Chicken nesting ecology has been con- ducted in sand shinnery oak {Quercus havar- dii ) habitats in New Mexico and Texas (Davis et al. 1979, Haukos and Broda 1989, Riley et al. 1992). The objectives of our study were to provide baseline information on age-specific variation in nesting ecology, record fidelity to previous nest sites (within-year renests and across-year attempts), and document nest-site locations relative to leks of Lesser Prairie- Chickens in sand sagebrush prairie of south- western Kansas. We examined annual varia- tion and the effects of age on reproductive pa- rameters and nest-site placement.

23

24

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

METHODS

Study area. — From 1997 to 2002, we stud- ied Lesser Prairie-Chickens inhabiting sand sagebrush habitat south of the Arkansas River in Finney County, Kansas (37° 52' N, 100° 59' W). We initiated field work on a 7,700-ha area in 1997 and on a nearby 5,600-ha area in 2000; we continued work on both areas through summer 2002. Vegetation was similar in both areas; sand sagebrush was the most conspicuous vegetation present and was inter- spersed with grasses, including little bluestem ( Schizachyrium scoparium), needle-and- thread ( Stipa comata), sand lovegrass (. Era - grostis trichodes ), sixweeks fescue ( Vulpia octoflora ), blue grama ( Bouteloua gracilis ), sand dropseed ( Sporobolus cryptandrus ), sideoats grama ( B . curtipendula), and western wheatgrass ( Agropyron smithii). The most common forb species were Russian thistle ( Salsola kali), western ragweed ( Ambrosia psilostachya ), sand lily ( Leucocrinum montan- um), and common sunflower ( Helianthus an- nuus). Each study area was bounded almost entirely by center-pivot irrigated cropland and grazed seasonally by livestock. Annual pre- cipitation averaged 50 cm (U.S. Department of Commerce 2003) and ranged from 42 cm (2000) to 59 cm (1997) during our study.

Locating and monitoring nests. — Using walk-in funnel traps, we captured female Lesser Prairie-Chickens on leks from mid- March through mid-April (Haukos et al. 1990). Except in 1997 (when age was not de- termined), we classified captured birds as yearlings (—10 months of age) or adults (>21 months of age) by examining the primaries (Copelin 1963). We equipped birds with 11-g necklace-style transmitters (life expectancy = 6-12 months; models from AVM Instrument Company, Colfax, California; Advanced Te- lemetry Systems, Isanti, Minnesota; and Ho- lohil Systems, Carp, Ontario) and released them on-site immediately after capture. Each day, we determined locations of transmitter- equipped birds by triangulating bearings col- lected from a truck-mounted, null-peak telem- etry system. Bird locations were determined until transmitter failure, emigration from the primary study areas, or bird death. When birds emigrated from our study area, we re-located them by extensive ground searches or from

fixed-wing aircraft. We monitored females that moved off our study area two to three times per week throughout the nesting season.

Using a hand-held antenna, we found nests by approaching transmitter-equipped females when their locations had remained unchanged >3 consecutive days. If the female was in- cubating, she was flushed so the eggs could be counted and the clutch examined for inter- specific parasitism (Hagen et al. 2002). We marked nest locations with flags (1997) or transmitters (1998-1999) at a distance of 5 m from the nest bowl (Jamison 2000), or we re- corded locations with a global positioning sys- tem (2000-2002). Nest sites were not visited again until the female departed the site with a brood or until the nest was depredated or abandoned. This technique allowed us to es- timate apparent nest success only. Because we did not determine nest status throughout in- cubation, we did not estimate daily survival of eggs or nests according to the Mayfield method (Mayfield 1975).

After the departure of each nesting female, we classified nest fate as successful (produced at least one chick), unsuccessful, or aban- doned. Beginning in 2000, we opened un- hatched eggs to determine whether embryos had developed. If the nest was depredated, we systematically searched the area within a 10- m radius for tracks, scat, or eggshell frag- ments to help determine the predator’s identity (Sargeant et al. 1998).

Statistical analyses. — We recorded clutch size and estimated the start of incubation for yearling and adult nests. We defined the start of incubation as the first day on which we detected no changes in the female’s telemetry locations — typically, 3-5 days before a nest was located. We estimated the initiation date of each nest by backdating from the start of incubation by 1 day for each egg in the clutch (Coats 1955). We also calculated apparent nest success (the proportion of all known nests producing at least one chick X 100), hatching success, egg fertility, percentage of females attempting a nest, percentage of females re- nesting, and the incidence of interspecific par- asitism— separately for yearlings and adults. We defined hatching success as the number of eggs hatched divided by initial clutch size (Westemeier et al. 1998b). We defined percent fertility as the number of eggs hatching or

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

25

containing a developed embryo divided by the total number of eggs in the nest bowl at the time of hatching. We estimated incubation length as the time (days) between the start of incubation and the date when a female left the nest with a brood (as determined from telem- etry locations). We estimated nesting frequen- cy as the percentage of females that attempted a nest. Females that did not attempt a nest and died before 31 May were excluded from our estimate of nesting frequency. Because we documented some first nesting attempts after 31 May, it was uncertain whether birds dying prior to this date would have subsequently at- tempted a nest. Interspecific parasitism was reported as the percentage of nests containing eggs of both Lesser Prairie-Chickens and oth- er bird species. Interspecific nest parasitism was previously described for the 1 997 to 1 999 field seasons (Hagen et al. 2002); here, we summarize all records of parasitism from 1997 to 2002. The percentage of females at- tempting to renest was estimated as the per- centage of females known to have incubated and lost a first clutch and that subsequently incubated a second. Because of some small expected cell counts, we used a Fisher’s exact test for all comparisons (Agresti 1996). In ad- dition, we used two-tailed f-tests for unequal variances (Zar 1999) to compare clutch size, incubation date, hatch date, and incubation length between yearlings and adults.

We used logistic regression to assess the re- lationship between the likelihood of renesting and (1) age class, (2) clutch size of the initial nest attempt, and (3) day into incubation when the initial attempt failed. We excluded data from 1997 because we did not identify age class of birds that year. Initially, we fit seven a priori models to data associated with 59 failed first nest attempts recorded from 1998 to 2002. We considered all four additive mod- els and main effect models for each of the three independent terms. We used the mini- mization of Akaike’s Information Criterion for small sample sizes (AICc.) to rank the models (Burnham and Anderson 1998). All models where AAICc < 2 were considered to be com- peting models (Burnham and Anderson 1998). Because age class was not included in any of the competing models (all AAICc > 2), we excluded this variable and developed models using an expanded data set ( n = 69) that in-

cluded failed first nest attempts recorded from 1997 to 2002. We used the same model pro- cedures previously described to fit three of our a priori models that included the main effects (1) clutch size and (2) day of incubation on which the initial attempt failed.

We calculated distances between first nests and renests, nesting attempts in multiple years, and distances from nest sites to the lek of capture and the nearest lek. We used anal- ysis of variance (ANOVA) to determine whether year or age class influenced the dis- tance between an initial nest site and the re- nest location and the affinity of nesting fe- males to lek sites (capture lek and nearest lek). We also used ANOVA to determine whether age class or success of the first-year nest af- fected distance between nest sites in subse- quent years. For these analyses, we excluded all data from 1997 because we did not identify age class that year; however, we included pooled age-class data from 1997 in the data tables to provide an overview of nesting pa- rameters for the duration of our study. We in- terpreted simple effects with two-sample t- tests when significant interactions were found (Zar 1999). We considered all differences sig- nificant when P < 0.05 and marginally sig- nificant when 0.05 < P < 0.10. We report parameter estimates and means as ± SE (or SD as noted).

RESULTS

Nesting ecology. — We captured 227 female Lesser Prairie-Chickens and fitted them with transmitters (87 yearlings, 117 adults, and 23 of unknown age). We found 209 nests (77 yearling, 103 adult, and 29 unknown-age). The percentage of females initiating a nest was similar ( P = 0.50) for yearlings (94%) and adults (92%; Table 1). We determined fate for 196 of 209 (94%) nests; apparent nest suc- cess was 26 ± 3% (51 of 196). The remaining nests were either abandoned (2%, n = 5) or success could not be determined from evi- dence remaining at the nest site (4%, n = 8). Nest success did not differ across years (x2 — 6.95, df = 5, P = 0.22) or between age classes for first nests (P = 0.60) or renests (P = 0.82; Table 1). An estimated 31% of all females that were unsuccessful in their first nesting attempt initiated a second nest, and this percentage did not differ (P = 0.85) between yearlings and

TABLE 1 . Lesser Prairie-Chicken nesting statistics (mean ± SE), by nesting attempt and age, compiled over a 6-year period in the sand sagebrush prairie of

26

THE WILSON JOURNAL OF ORNITHOLOGY

Vol. 118, No. 1, March 2006

—

cn

o

—

NO

o

—

cn

o

CN

—

CN

+1

+ 1

+1

+ 1

+1

ON

o

in

00

■St

q

cn

o

NO

CN

__

cn

CN

r-i

00

cn

—

cn

>n

CN

NO

ON

cn

—

CN

in

CN

r- — — LT;

in no Tf

m o> >o

cn cn cn

no cn o — no CN On Tt h -H h -H O — LT, m — — <N

ON

t-*

3r

CN

o

cn

>n

o

>n

CN

O

ON

o

00

cn

r-

—

>n

CN

in

ON

in

00

r-

o

d

o

d

CN

ON

q

NO

as

o

NO

in

o

—

00

o

—

cn

in

o

cn

CN

cn

NO

CN

+1

+ 1

_

cn

ON

o

r-~

cn

CN

>n

CN

cn

CN

U

—

CN

U

CN

NO

r-

On

ON

cn

—

CN

>n

in

NO

CN

00 — NO ■'t vO IT) M

r- oo r- (N — — —

o cn cn

ON CN CN CN

cn cn

M n

oo in o

CN

On

o

0

NO

CN

>n

q

»n

o

O

CN

—

o

CN

—

in

d

cn

00

CN

+1

+ 1

+1

+ 1

+1

+ 1

in

00

■sj-

r-

in

CN

as

n

cn

<n

''t

cn

i>

NO

—

00

cn

>n

cn

f"

ON

cn

—

CN

m

in

r»

CN

o — r" o o o in (N

\r " h o o h h n r, x h r, h CN

<U ^

^ <u _ 2

o o o ^

5 §

U c/$

<u

•? I §

! P

52 3 ^

8 g a

| 'a £ -O M -f ° _

« = | 3

z u w ac

Let 60 Of) c

g 2 s 5

^ cfl c/3

g* s* ^ ^ e*

a Includes females of unknown age.

b Females that attempted a nest; females that did not attempt a nest and died before 31 May were excluded. c n = number of failed first nests.

d Nests were parasitized by either Ring-necked Pheasants or Northern Bobwhites.

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

27

adults (Table 1). However, success of renests (14%) was marginally less than success of ini- tial nests (29%; X2 = 3.31, df = 1, P = 0.090). No females were known to have initiated a third nest in the same year. Mean hatch date (all years combined) was 1 June for first nest- ing attempts and 22 June for renests (Fig. 1), with a mean incubation length of 26.7 days (Table 1). More than 95% of all nests were initiated and completed between 5 May and 2 July (Fig. 1).

Mean clutch size did not differ between yearlings and adults for either first nesting or renesting attempts (Table 1). Mean clutch size was 7.6 ± 0.4 eggs for renests, significantly less (f188 = 1 1.77, P < 0.001) than the mean clutch size (12.0 ± 0.1 eggs) of first nests. Overall hatching success was 74 ± 2% and did not differ between yearlings and adults. Likewise, egg fertility was similar between the two age classes, with 94 ± 1% of all eggs containing a developed embryo (Table 1).

Six of 209 (3%) Lesser Prairie-Chicken nests were parasitized by other bird species. Four of the six nests contained Lesser Prairie- Chicken and Ring-necked Pheasant ( Phasi - anus colchicus ) eggs, and eggs of both species hatched in two of these nests. One nest was parasitized by a Northern Bobwhite ( Colinus virginianus\ 10 prairie-chicken eggs and 1 quail egg), and the remaining nest was para- sitized by both Ring-necked Pheasant and Northern Bobwhite (3 prairie-chicken eggs, 1 pheasant egg, and 1 quail egg). Both of these latter nests were depredated before hatching.

Nest predators. — Most (>80%) known pre- dation events occurred >3 days after our ini- tial nest visit (mean = 10.2 days ± 6.9 SD). We assigned predator species to 112 of 161 (70%) unsuccessful Lesser Prairie-Chicken nests. Coyotes ( Canis latrans ) depredated the majority (64%) of the nests and were the pri- mary cause of nest predation during most years (Table 2). Snakes were responsible for the loss of 31% and 42% of the unsuccessful Lesser Prairie-Chicken nests in 2001 and 2002, respectively. Most of the snake preda- tion was probably by Gopher snakes (. Pituo - phis melanoleucus ) because they were the most observed snake species on our study ar- eas. Other causes of nest loss included pre- dation by ground squirrels ( Spermophilus spp.) and trampling by cattle (Table 2).

Renesting probability. — The probability of a Lesser Prairie-Chicken renesting was influ- enced by both clutch size and the day of in- cubation on which the initial attempt failed. An additive model including both terms was the highest-ranking (AAICc = 0.00; AICc = 80.90), but the model including only date of failure also had considerable support (AAIC( = 1.48). The model including only clutch size was not supported (AAICc = 15.24). Females incubating initial nests later into incubation tended to have a lower probability of renesting (Gdate = -0.18, 95% Cl = -0.28 to -0.08; Fig. 2). Females laying a larger clutch in the initial nest attempt tended to be more likely to renest (Bclutch = 0.31); however, the magni- tude of this effect was not clear because the confidence interval overlapped zero (95% Cl = —0.01 to 0.63). The odds of a female at- tempting to renest decreased by 16.2% with each day into incubation of the initial attempt and increased 20.2% with each one-egg in- crease in clutch size (Fig. 2).

Nest-site location. — Between 1997 and 2002, we found 28 renests (Table 3). Distance between first nests and renests (1,271 m) was not influenced by age class (Flf23 = 1.69, P = 0.21) or year (F4>23 = 1.65, P = 0.21); there was no interaction effect (F2i23 — 1.82, P = 0.19; 1998-2002 data). Similarly, the distance between nests initiated by the same female in subsequent years (mean = 918 m, n = 15; Table 3) was not influenced by age class (FU4 = 0.16, P = 0.70) or success of the first-year nest (FU4 = 0.05, P = 0.82); there was no interaction effect (FU4 = 0.00, P = 0.98).

The distance from a nest to the nearest lek (mean = 691 m, n = 194; Table 4) was not influenced by year (F4164 = 1.11, P = 0.36) or age class (F U64 = 0.00, P = 0.99), nor was there an interaction effect (F4164 = 1.41, P = 0.23; 1998-2002 data). Of 184 nests, 147 (80%) were located closer to a lek other than the lek where the female was last captured. Ten nests (5%) were located >10 km from the lek at which the incubating female was cap- tured (median = 20.6 km, range = 10.6-56.5 km). The female nesting 56.5 km from her lek of capture was successful in her nesting at- tempt. The distance from nest site to the lek where the female was captured (mean = 3,082 m, n = 184; Table 4) was not influenced by age class (FU58 = 0.12, P = 0.73) or year

Percentage of nests Percentage of nests

28

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

Weekly interval

FIG. 1. Percentage of Lesser Prairie-Chicken first nests (A) and renests (B) in southwestern Kansas that were initiated, incubated, depredated, and hatched, by weekly intervals, 1997-2002. Mean dates for each variable are listed at the top of each figure.

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

29

TABLE 2. Probable causes of predation of Lesser Prairie-Chicken nests in the sand sagebrush prairie of southwestern Kansas, 1997-2002.

Depredation (%)

Predator	1997 (n = 24)	1998 (n = 12)	1999 in = 20)	2000 in = 44)	2001 in = 36)	2002 in = 26)	Total3 in = 161)
Coyote	71	100	70	34	22	27	45
Ground squirreP	4	0	0	11	0	0	4
Snakec	13	0	5	1 1	31	42	19
Cattle	0	0	5	2	3	0	2
Unknown	13	0	20	41	45	31	30
a Percentage of all nests destroyed by each predator.

b We did not differentiate between thirteen-lined ground squirrels and spotted ground squirrels. c Gopher snakes appeared to be the most abundant snake species.

(F4158 = 1.25 P = 0.29), and there was no interaction effect (F4158 = 1.33, P = 0.26;

1998-2002 data).

DISCUSSION

Although rainfall during the primary 4- month nesting period (April through July) var-

ied substantially during the 6 years of our study (range — 22.3-38.3 cm), we document- ed little annual variation in Lesser Prairie- Chicken nesting activity. Our ability to detect annual variation, however, may have been hin- dered by relatively small sample sizes within years, especially in the early years of the

FIG. 2. Probability of Lesser Prairie-Chickens initiating renests after failure of the initial nest attempt in southwestern Kansas, 1997-2002. Probabilities are plotted for various clutch sizes (8, 10, 12, 14) and the day of incubation when the initial nest attempt failed.

30

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

TABLE 3. Evidence of nest-site fidelity as shown by mean distances (m) between nests for Lesser Prairie- Chickens in southwestern Kansas, 1997-2002. Within- and across-year distances are presented by age class and nest fate.

Within-year* Across yearsb’c

Category n Distance SE n Distance SE

Age class Yearling	11	1,071	327	6	1,170	599
Adult	13	1,182	263	9	750	365
Nest fated
Successful	—	—	—	6	712	438
Unsuccessful	—			9	1,055	453
Totale	28	1.271	218	15	918	316

a Distance between the first nest and the renest.

bFor two females that initiated >1 nest within a year, the mean coordinates of those nests were used to calculate the distance to the nest site in subsequent years.

c Nests for one female were located in non-consecutive years; all other nests were located in consecutive years. d Nest fate refers to fate of first nests. e Age of four females was undetermined.

study. Additionally, we observed little age- specific variation — except that yearlings had slightly smaller clutches and marginally later hatch dates for first nest attempts than did adults.

For all known nests, initiation began in ear- ly May; peak hatching was 1 June for first nests and 22 June for renests (Fig. 1). Similar dates of nest initiation (mid-April through late May) and hatching (late May through mid- June) have been reported from studies throughout the species’ range (Giesen 1998, Patten et al. 2005). Mean incubation length was 26.7 days (this study). Because nest at-

tentiveness of grouse increases throughout the laying period (Giesen and Braun 1979), we may have overestimated incubation length by misidentifying the start of incubation. How- ever, the time required to hatch Lesser Prairie- Chicken eggs in an incubator (24—26 days; Coats 1955, Sutton 1968) was only slightly less than our estimate for eggs incubated by wild birds.

The success of all nests averaged 26% in our study, substantially less than estimates from New Mexico (42%) and Oklahoma (40%; Patten et al. 2005), but similar to the 28% reported by Giesen (1998) for 10 studies

TABLE 4. Distances (m) between Lesser Prairie-Chicken nest sites and leks in southwestern Kansas, 1997- 2002.

Nest site to lek of capture Nest site to nearest lek

Category n Median Mean ± SE n Median Mean ± SE

Year

1997	25	1,528	1,647	±	226	26	556	557 ± 52
1998	14	1,134	1,727	±	529	14	577	546 ± 71
1999	24	2,357	2,317	+	332	25	726	701 ± 55
2000	56	1,282	2,874	-t-	1,006	56	675	742 ± 53
2001	37	1,396	3,241	±	983	41	727	740 ± 54
2002	28	2,333	5,901	±	1,366	32	631	703 ± 65
Age
Yearling	68	1,893	3,580	±	853	68	633	702 ± 48
Adult	91	1,258	3,104	±	591	97	675	718 ± 32
Total	184a	1,427	3,082	4-	432	194b	632	691 ± 25

a Includes 25 nests of females of unknown age. b Includes 29 nests of females of unknown age.

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

31

conducted throughout the range of the Lesser Prairie-Chicken. Giesen (1998) suggested that nest success from those 10 studies was nega- tively biased due to observer disturbance at nest sites. Negative bias in our study was like- ly only slight because females were flushed from their nests only once. Westemeier et al. (1998a) reported that flushing incubating Greater Prairie-Chickens (T. cupido ) once did not result in reduced nest success. Also, the number of days between our initial nest visits and predation events averaged >10 days. In addition, only 2% of the nests in our study were abandoned — a much smaller percentage than the 25% reported by Riley et al. (1992) for Lesser Prairie-Chickens in New Mexico. Further, one of five nests abandoned during our study was abandoned 9 days after the re- searcher’s visit, indicating that it probably was not due to human disturbance.

The percentage of females initiating a sec- ond nest during our study (31%) was between previous estimates for Lesser Prairie-Chickens in New Mexico (15%) and Oklahoma (79%; Patten et al. 2005), and it was less than the 83% reported for Greater Prairie-Chickens (Svedarsky 1988) and the 67% estimated for Sharp-tailed Grouse (T. phasianellus\ Roers- ma 2001). The percentage of Greater Sage- Grouse ( Centrocercus urophasianus ) initiat- ing a renest was highly variable (5 to 87%) throughout their range (Schroeder et al. 1999), and most estimates were less than what we observed for Lesser Prairie-Chickens. Our models indicated that the low probability of Lesser Prairie-Chickens renesting in south- western Kansas was influenced by the length of incubation before their clutches were dep- redated (>50% of unsuccessful initial clutches were incubated >12 days prior to predation). Similarly, Schroeder (1997) reported that Greater Sage-Grouse in Washington whose initial nests failed late in incubation were less likely to renest than those whose nests failed earlier in incubation. Clutch size of the initial nesting attempt was also somewhat associated with renesting probability in our study; how- ever, the magnitude of this effect was unclear. The positive relationship that we observed may have been due to increased fitness asso- ciated with females laying larger clutches or the possibility that we misclassified some re- nests as initial nest attempts. We speculate that

the latter was not a common occurrence dur- ing our study, but our methods did not allow us to locate nests that were depredated prior to the onset of incubation.

Few prairie grouse researchers have report- ed nest success separately for first nest at- tempts and subsequent renestings. Bergerud and Gratson (1988) hypothesized that preda- tion of grouse nests was density-dependent and that renests would be more successful than first nest attempts due to lower nest den- sities. They also believed that nest success should improve as new vegetative cover ap- pears throughout the nesting season. Success of first and second nesting attempts of Lesser Prairie-Chickens in Kansas, however, does not support Bergerud and Gratson’s (1988) hy- potheses, as first nest attempts were margin- ally more successful than renestings. Like- wise, Greater Prairie-Chicken nests initiated in Kansas prior to 30 April (presumably first at- tempts) were more successful than nests ini- tiated after 1 May (presumably renests; Robel 1970). Initial nesting attempts for Attwater’s Greater Prairie-Chicken ( T c. attwateri) also were more successful than renests in 4 of 5 years (Lutz et al. 1994). Similar nest success for first attempts and subsequent renestings has been reported for Greater Prairie-Chickens in Colorado (Schroeder and Braun 1992) and Greater Sage-Grouse in Washington (Schroe- der 1997) and Alberta, Canada (Aldridge and Brigham 2001). The only support for Berge- rud and Gratson’s (1988) hypothesis comes from studies on Sharp-tailed Grouse in Min- nesota and North Dakota, where success was higher for second attempts than first attempts (Christenson 1970, Schiller 1973). In our study, Lesser Prairie-Chicken nests initiated after 15 May were less successful (11.9%, n = 42) than earlier nests (31.5%, n = 143), regardless of nesting attempt. We speculate that nests initiated after 1 5 May were less suc- cessful due to an increase in predator efficien- cy later in the nesting season, corresponding to changes in the structure and composition of vegetation. Cattle grazing began on our study area around 15 May, and, after that date, grass cover and visual obstruction decreased sub- stantially (JCP unpubl. data). Grazing coupled with normal drought conditions during the summer months in southwestern Kansas may result in declining habitat quality, and, there-

32

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

fore, the poor success of renesting Lesser Prai- rie-Chickens. Land management practices that maintain taller and denser vegetation structure later into the nesting season may promote the overall nesting success of Lesser Prairie- Chickens.

Clutch size in Kansas averaged 11.3 eggs in 191 completed clutches — greater than that reported in New Mexico (8.7) and Oklahoma (10.8; Patten et al. 2005) or in 60 completed clutches located in other states occupied by Lesser Prairie-Chickens (10.4; Giesen 1998). Our study is the first to document substantially different mean clutch sizes for first nests (12.0 eggs) and renests (7.6 eggs). Merchant (1982) reported mean clutch size for initial and sec- ond nesting attempts, but his estimates were similar for both (9.8 and 10.7 eggs, respec- tively). In our study, the percentage of eggs containing a developed embryo was 94% and hatching success was 74%. Egg fertility has not been reported previously for the Lesser Prairie-Chicken, but hatching success of eggs was estimated at >90% across three studies (see Giesen 1998). The lower hatching suc- cess observed in our study reflects partial nest losses that occurred in 32 of 48 (67%) suc- cessful nests.

Identifying nest predators from nest re- mains is difficult because patterns of egg breakage overlap among, and even within, predator species (Lariviere 1999). Uncertain- ties were reduced on our study area, however, because coyotes and gopher snakes were the only common species capable of preying on Lesser Prairie-Chicken nests. Studies in New Mexico and Texas revealed that Chihuahuan Ravens ( Corvus cryptoleucus ), badgers ( Tax- idea taxus), striped skunks ( Mephitis mephi- tis), and ground squirrels were the primary predators of Lesser Prairie-Chicken nests (Da- vis et al. 1979, Haukos and Broda 1989, Riley et al. 1992). However, few corvids, badgers, or striped skunks were observed on our study area, and, although ground squirrels were abundant (estimated from casual roadside ob- servations), they were identified as important nest predators during only 1 year (2000).

Davis et al. (1979) documented snakes preying on Lesser Prairie-Chicken nests in New Mexico. We found little evidence for snake predation of nests during the early years of our study (Jamison 2000), but snake abun-

dance appeared to increase (estimated from casual roadside observations), as did nest pre- dation by snakes, in the later years (Pitman 2003). Snakes may have been responsible for most partial-nest depredations because of the lack of eggshell fragments at partly depredat- ed nests. Also, three incubating Lesser Prairie- Chickens were likely killed by snakes because their intact carcasses were found with a thin film of mucus covering the heads. In each case, it appeared as if a snake had tried to swallow the bird.

Interspecific nest parasitism has been re- ported for Greater Prairie-Chickens and Sharp-tailed Grouse (Leach 1994, Westemeier et al. 1998b), but had not been reported for Lesser Prairie-Chickens before our work in Kansas (Hagen et al. 2002). Only 6 of 209 (3%) nests were parasitized by Ring-necked Pheasants and/or Northern Bobwhites, and 2 of the 6 (33%) nests produced Lesser Prairie- Chicken chicks. Hatching success of eggs in these two nests was 72%, similar to the 74% estimated for 46 unparasitized nests (Hagen et al. 2002). Our study provided no evidence that nest parasitism negatively affected nest suc- cess or hatchability of Lesser Prairie-Chick- ens.

Bergerud and Gratson (1988) hypothesized that successful female grouse would nest in the same area in the subsequent breeding sea- son. In southwestern Kansas, female Lesser Prairie-Chickens nested within 712 m of the site of their previous year’s nest site (if suc- cessful). This degree of philopatry is similar to that reported for Greater Sage-Grouse in Wyoming (Berry and Eng 1985) and Idaho (Fischer et al. 1993). Greater Sage-Grouse in Washington showed less philopatry to a pre- vious year’s successful nest location, moving an average of 1 ,600 m in the subsequent nest- ing season (Schroeder and Robb 2003).

The association between lek location and nest placement has important management im- plications for identifying critical nesting hab- itat. Bradbury (1981) hypothesized that fe- male home ranges included only one lek and that >50% of all females should locate their nests nearer to that lek than other nearby leks. Studies of Greater Sage-Grouse and Sharp- tailed Grouse have provided support for this hypothesis (Bradbury et al. 1989, Giesen 1997). In Colorado and Minnesota, however.

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

33

only 23 of 89 (26%; Schroeder 1991) and 7 of 18 (39%; Svedarsky 1988) Greater Prairie- Chickens nested closer to their lek of capture than to other leks, respectively. Similarly, in Idaho Wakkinen et al. (1992) found 92% of Greater Sage-Grouse nests within 3 km of a lek, but only 55% were within 3 km of the lek of capture. Our Lesser Prairie-Chicken nesting data also do not support Bradbury’s (1981) hypothesis: 80% of our females (147 of 184) nested closer to a lek other than that on which they were captured. More impor- tantly, we located >80% of all nests within 1 km of a known lek site; thus, we believe that providing secure nesting habitat within 1 km of a lek site is an important management strat- egy*

Our study provides the first comprehensive description of Lesser Prairie-Chicken nesting ecology in terms of age-specific reproductive effort. Our estimates of Lesser Prairie-Chick- en nesting parameters should be viewed as ap- proximations, however, because our method- ology did not allow us to locate nests that were destroyed during the laying process. Nevertheless, our estimates provide a much better understanding of Lesser Prairie-Chick- en demography in sand sagebrush habitats. The low nest success we observed (26%) is troubling, especially if >50% nest success is required for population stability (Westemeier 1979). Sensitivity analyses have revealed that nest success is one of the most influential de- mographic parameters affecting population growth of prairie grouse (Peterson and Silvy 1996, Wisdom and Mills 1997, Hagen 2003). Thus, habitat management designed to en- hance nest success of Lesser Prairie-Chickens in southwestern Kansas should be a priority. Similar information on nesting ecology from Lesser Prairie-Chicken populations in other states and habitat types is needed to identify regional and site-specific conservation needs and to aid in the development of range-wide population models.

ACKNOWLEDGMENTS

We thank the private landowners of southwestern Kansas and the Sunflower Electric Power Corporation for property access. C. C. Griffin, G. C. Salter, T. G. Shane, T. L. Walker, Jr., and T. J. Whyte assisted with fieldwork. This study was supported by Kansas State University, Division of Biology; Kansas Agricultural Experiment Station (Contribution No. 04-41 1-J); Kan-

sas Department of Wildlife and Parks; Federal Aid in Wildlife Restoration Projects W-47-R and W-53-R; and Westar Energy, Inc. Finally, we thank J. W. Con- nelly, M. A. Schroeder, and three anonymous review- ers for comments on earlier drafts of this manuscript.

LITERATURE CITED

Agresti, A. 1996. An introduction to categorical data analysis. John Wiley and Sons, New York. Aldridge, C. L. and R. M. Brigham. 2001. Nesting and reproductive activities of Greater Sage- Grouse in a declining northern fringe population. Condor 103:537-543.

Bergerud, A. T. and M. W. Gratson. 1988. Adaptive strategies and population ecology of northern grouse, vol. II: theory and synthesis. University of Minnesota Press, Minneapolis.

Berry, J. D. and R. L. Eng. 1985. Interseasonal move- ments and fidelity to seasonal use areas by female Sage Grouse. Journal of Wildlife Management 49: 237-240.

Bradbury, J. W. 1981. The evolution of leks. Pages 138-169 in Natural selection and social behavior (R. D. Alexander and D. W. Tinkle, Eds.). Chiron Press, New York.

Bradbury, J. W., R. M. Gibson, C. E. McCarthy, and S. L. Vehrencamp. 1989. Dispersion of displaying male Sage Grouse: II. The role of female disper- sion. Behavioral Ecology and Sociobiology 24: 15-24.

Burnham, K. P. and D. R. Anderson. 1998. Model selection and inference: a practical information- theoretic approach. Springer, New York. Christenson, C. D. 1970. Nesting and brooding char- acteristics of Sharp-tailed Grouse ( Pedioecetes phasianellus jamesi ) in southwestern North Da- kota. M.Sc. thesis. University of North Dakota, Grand Forks.

Coats, J. 1955. Raising Lesser Prairie Chickens in captivity. Kansas Fish and Game 13:16-20. Copelin, F. F. 1963. The Lesser Prairie Chicken in Oklahoma. Technical Bulletin no. 6. Oklahoma Wildlife Conservation Department, Oklahoma City.

Crawford, J. A. 1980. Status, problems, and research needs of the Lesser Prairie-Chicken. Pages 1-7 in Proceedings of the Prairie Grouse Symposium (P. A. Vohs and F. L. Knopf, Eds.). Oklahoma State University, Stillwater.

Davis, C. A., T. Z. Riley, R. A. Smith, H. R. Suminski, and M. J. Wisdom. 1979. Habitat evaluation of Lesser Prairie-Chickens in eastern Chaves County, New Mexico. New Mexico Agriculture Experi- ment Station, Las Cruces.

Fischer, R. A., A. D. Apa, W. L. Wakkinen, K. P. Reese, and J. W. Connelly. 1993. Nesting area fidelity of Sage Grouse in southeastern Idaho. Condor 95:1038-1041.

Giesen, K. M. 1997. Seasonal movements, home rang- es, and habitat use by Columbian Sharp-tailed

34

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

Grouse in Colorado. Special Report, no. 72. Col- orado Division of Wildlife, Denver.

Giesen, K. M. 1998. Lesser Prairie-Chicken ( Tympan - uchus pallidicinctus). The Birds of North Ameri- ca, no. 364.

Giesen, K. M. and C. E. Braun. 1979. Nesting be- havior of female White-tailed Ptarmigan in Col- orado. Condor 81:215-217.

Hagen, C. A. 2003. A demographic analysis of Lesser Prairie-Chicken populations in southwestern Kan- sas: survival, population viability, and habitat use. Ph.D. dissertation, Kansas State University, Man- hattan.

Hagen, C. A., B. E. Jamison, R. J. Robel, and R. D. Applegate. 2002. Ring-necked Pheasant parasit- ism of Lesser Prairie-Chicken nests in Kansas. Wilson Bulletin 114:522-524.

Haukos, D. A. and G. S. Broda. 1989. Northern Har- rier ( Circus cyaneus ) predation of Lesser Prairie- Chicken ( Tympanuchus pallidicinctus). Journal of Raptor Research 23:182-183.

Haukos, D. A., L. M. Smith, and G. S. Broda. 1990. Spring trapping of Lesser Prairie-Chickens. Jour- nal of Field Ornithology 61:20-25.

Jamison, B. E. 2000. Lesser Prairie-Chicken chick sur- vival, adult survival, and habitat selection and movements of males in fragmented rangelands of southwestern Kansas. M.Sc. thesis, Kansas State University, Manhattan.

Jensen, W. E., D. A. Robinson, Jr., and R. D. Apple- gate. 2000. Distribution and population trend of Lesser Prairie-Chicken in Kansas. Prairie Natural- ist 32:169-175.

Lariviere, S. 1999. Reasons why nest predators cannot be inferred from nest remains. Condor 101:71 8— 721.

Leach, S. W. 1994. Mallard parasitizes Sharp-tailed Grouse nest. Blue Jay 52:144-146.

Lutz, R. S., J. S. Lawrence, and N. J. Silvy. 1994. Nesting ecology of Attwater’s Prairie-Chicken. Journal of Wildlife Management 58:230-233.

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

Merchant, S. S. 1982. Habitat use, reproductive suc- cess, and survival of female Lesser Prairie- Chickens in two years of contrasting weather. M.Sc. thesis, New Mexico State University, Las Cruces.

Patten, M. A., D. H. Wolfe, E. Shochat, and S. K. Sherrod. 2005. Habitat fragmentation, rapid evo- lution, and population persistence. Evolutionary Ecology Research 7:235-249.

Peterson, M. J. and N. J. Silvy. 1996. Reproductive stages limiting productivity of the endangered At- twater’s Prairie-Chicken. Conservation Biology 4: 1264-1276.

Pitman, J. C. 2003. Lesser Prairie-Chicken nest site selection and nest success, juvenile gender deter- mination and growth, and juvenile survival and dispersal in southwestern Kansas. M.Sc. thesis, Kansas State University, Manhattan.

Riley, T. Z., C. A. Davis, M. Ortiz, and M. J. Wis- dom. 1992. Vegetative characteristics of success- ful and unsuccessful nests of Lesser Prairie-Chick- ens. Journal of Wildlife Management 56:383-387.

Robel, R. J. 1970. Possible role of behavior in regu- lating Greater Prairie-Chicken populations. Jour- nal of Wildlife Management 34:306-312.

Roersma, S. J. 2001. Nesting and brood rearing ecol- ogy of Plains Sharp-tailed Grouse ( Tympanuchus phasianellus jamesi ) in a mixed-grass/fescue ecoregion of southern Alberta. M.Sc. thesis, Uni- versity of Manitoba, Winnipeg.

Sargeant, A. B., M. A. Sovada, and R. J. Green- wood. 1998. Interpreting evidence of depredation of duck nests in the prairie pothole region. U.S. Geological Survey, Northern Prairie Wildlife Re- search Center, Jamestown, North Dakota, and Ducks Unlimited, Memphis, Tennessee.

Schiller, R. J. 1973. Reproductive ecology of female Sharp-tailed Grouse ( Pedioecetes phasianellus ) and its relationship to early plant succession in northwestern Minnesota. Ph.D. dissertation. Uni- versity of Minnesota, St. Paul.

Schroeder, M. A. 1991. Movement and lek visita- tion by female Greater Prairie-Chickens in re- lation to predictions of Bradbury’s female pref- erence hypothesis of lek evolution. Auk 108: 896-903.

Schroeder, M. A. 1997. Unusually high reproduc- tive effort by Sage Grouse in a fragmented hab- itat in north-central Washington. Condor 99: 933-941.

Schroeder, M. A. and C. E. Braun. 1992. Seasonal movement and habitat use by Greater Prairie- Chickens in northeastern Colorado. Special Re- port, no. 68. Colorado Division of Wildlife, Den- ver.

Schroeder, M. A. and L. A. Robb. 2003. Fidelity of Greater Sage-Grouse Centrocercus urophasianus to breeding areas in a fragmented landscape. Wildlife Biology 9:291-298.

Schroeder, M. A., J. R. Young, and C. E. Braun. 1999. Sage Grouse ( Centrocercus urophasianus). The Birds of North America, no. 425.

Sutton, G. M. 1968. The natal plumage of the Lesser Prairie Chicken. Auk 85:69.

Svedarsky, W. D. 1988. Reproductive ecology of fe- male Greater Prairie-Chickens in Minnesota. Pag- es 193-239 in Adaptive strategies and population ecology of northern grouse, vol. 2 (A. T. Bergerud and M. W. Gratson, Eds.). University of Minne- sota Press, Minneapolis.

Taylor, M. A. and F. S. Guthery. 1980. Status, ecol- ogy, and management of the Lesser Prairie-Chick- en. General Technical Report RM-77, USDA For- est Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado.

U.S. Department of Commerce. 2003. National Oce- anic and Atmospheric Administration. National Climatic Data Center, www.ncdc.noaa.gov/ (ac- cessed 7 January 2003).

Pitman et al. • NESTING ECOLOGY OF LESSER PRAIRIE-CHICKENS

35

Wakkinen, W. L., K. P. Reese, and J. W. Connelly. 1992. Sage Grouse nest locations in relation to leks. Journal of Wildlife Management 56:381- 383.

Westemeier, R. L. 1979. Factors affecting nest success of prairie chickens in Illinois. Proceedings of the Prairie Grouse Technical Council 13:9-15. Westemeier, R. L., J. E. Buhnerkempe, and J. D. Brawn. 1998a. Effects of flushing nesting Greater Prairie-Chickens in Illinois. Wilson Bulletin 110: 190-197.

Westemeier, R. L., J. E. Buhnerkempe, W. R. Ed- wards, J. D. Brawn, and S. A. Simpson. 1998b. Parasitism of Greater Prairie-Chicken nests by Ring-necked Pheasants. Journal of Wildlife Man- agement 62:854-863.

Wisdom, M. J. and L. S. Mills. 1997. Sensitivity anal- ysis to guide population recovery: prairie-chick- ens as an example. Journal of Wildlife Manage- ment 61 :302-312.

Zar, J. H. 1999. Biostatistical analysis, 4th ed. Prentice Hall, Englewood Cliffs, New Jersey.

The Wilson Journal of Ornithology 118(1 ):36 — 4 1 , 2006

A COMPARATIVE BEHAVIORAL STUDY OF THREE GREATER SAGE-GROUSE POPULATIONS

SONJA E. TAYLOR1 3 AND JESSICA R. YOUNG1 2 3

ABSTRACT. — We compared male strut behavior of the genetically distinct Lyon, Nevada/Mono, California Greater Sage-Grouse ( Centrocercus urophasianus ) population with that of two proximal populations: Nye, Ne- vada, and Lassen, California. We measured strut rates and nine acoustic components of the strut display in all three populations. Male strut rates did not differ among populations. Acoustic components of the Lyon/Mono and Lassen populations were similar, whereas the Nye population was distinct. The genetically distinct Lyon / Mono population was more similar behaviorally to the Nye population than the genetically similar Nye and Lassen populations were to each other. Overall, the Lyon/Mono population did not exhibit detectable differences in male strut behavior. Reproductive isolation through sexual selection does not appear to have occurred in the Lyon/Mono population. Received 27 September 2004, accepted 19 October 2005.

Two recent studies based on mitochondrial gene sequence (Benedict et al. 2003, Oyler- McCance et al. 2005) and nuclear microsat- ellite markers (Oyler-McCance et al. 2005) re- vealed a genetically distinct population of Greater Sage-Grouse ( Centrocercus urophas- ianus) on the Nevada/California border (Lyon, Nevada/Mono, California). Those studies in- dicated that the Lyon/Mono Greater Sage- Grouse population is more genetically distinct from other Greater Sage-Grouse populations than is the newly described (Young et al. 2000) Gunnison Sage-Grouse (C. minimus) species. Several factors, including the appar- ent genetic and geographic isolation of Lyon/ Mono sage-grouse from other populations, the degradation and loss of sagebrush (Artemisia spp.) habitat, and an overall population de- cline, have made this a population of interest from both evolutionary and conservation per- spectives.

Morphological (Hupp and Braun 1991) and behavioral studies (Young et al. 1994) of Gun- nison Sage-Grouse provided evidence that sexual selection had driven speciation in the isolated populations of sage-grouse in south- western Colorado and southeastern Utah. The use of both mitochondrial (Kahn et al. 1999) and nuclear markers (Oyler-McCance et al.

1 Rocky Mountain Center for Conservation Genetics and Systematics, Dept, of Biological Sciences, Univ. of Denver, Denver, CO 80208, USA.

2 Western State College of Colorado, Dept, of Nat- ural and Environmental Sciences, Gunnison, CO 81231, USA.

3 Corresponding author; e-mail: sonja_taylor@comcast.net

1999) supported the morphological and be- havioral data and led to species designation for the Gunnison Sage-Grouse (American Or- nithologists’ Union 2000, Young et al. 2000). A similar approach would determine whether the genetic distinctiveness of the Lyon/Mono population has been manifested morphologi- cally and/or behaviorally as it has in Gunnison Sage-Grouse. If so, it could potentially lead to a taxonomic reclassification.

Male mating success and mate-choice cues (Gibson and Bradbury 1985), territoriality (Gibson and Bradbury 1987), components of female choice (Gibson et al. 1991), and male strutting behavior (Young et al. 1994) have been studied previously in the Mono sage- grouse population. However, with the excep- tion of Young et al. (1994), there have been no comparative studies among populations. Young et al. (1994) compared secondary sex- ual characteristics from male strut displays among three populations — one Gunnison Sage-Grouse population (Gunnison Basin, Colorado) and two Greater Sage-Grouse pop- ulations (Mono, California, and Jackson, Col- orado). The structure of the Gunnison male strut display was strikingly different from that of the other two populations. However, the comparison of the similarly structured strut display between males from Mono and Jack- son indicated statistically significant differ- ences in most of the acoustic measures.

In light of the genetic distinctiveness of Lyon/Mono sage-grouse and the behavioral results of Young et al. (1994), we undertook a further examination of male strut display be- havior. We compared the Lyon/Mono popu-

36

Taylor and Young • GREATER SAGE-GROUSE BEHAVIOR

37

FIG. 1. Current Greater Sage-Grouse distribution in California and Nevada, and locations of three sample populations (modified from Schroeder et al. 2004).

lation with two proximal populations of Greater Sage-Grouse (Fig. 1). We tested the hypothesis that the Lyon/Mono population’s behavior is measurably different from that of other Greater Sage-Grouse populations and may, in fact, be considered a separate taxon given the genetic differences. Alternatively, although the Lyon/Mono population appears genetically isolated, behaviorally it may not be significantly different from other Greater Sage-Grouse populations, indicating that sex- ual selection resulting in pre-mating isolating mechanisms has not occurred.

METHODS

The three populations we studied are from the southwestern edge of the Greater Sage- Grouse range in Nevada and California (Fig. 1). Behavioral measurements of male strut

displays were taken at five leks. Greater Sage- Grouse in Lyon County, Nevada, and Mono County, California, form a connected, inter- breeding population (Lyon/Mono). Record- ings were completed between 9 and 17 April 2001 at three leks from the Lyon/Mono pop- ulation: Lyon County, Nevada (Desert Creek 2 lek; 38° 42' N, 1 19° 18' W; 1,603 m), south- ern Mono County, California (Long Valley 1 lek; 37° 42' N, 118° 48' W; 2,124 m), and northern Mono County, California (Biedeman lek; 38° 12' N, 119°6'W; 2,447 m). Of the three recorded Lyon/Mono leks, the Desert Creek and Biedeman leks are farthest apart (123 km). Lassen County, California (Eastside lek; 40° 18' N, 120° 0'W; 1,490 m), is ap- proximately 250 km north and Nye County, Nevada (Roadside lek; 38° 42' N, 1 16° 47' W; 2,121 m), is approximately 215 km east of the

38

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

N

X

>s

u

c

a>

D

cr

a>

■
Air sac pops		jL-Whistle minimum
■ / . Whistle peak
* Whistlp start ^ *	k
• « ■ — * ** ^1

Time (sec)

FIG. 2. Typical sonagram of a Greater Sage-Grouse male strut display. The two air sac pops, whistle start frequency, whistle peak, and whistle minimum are labeled. See Table 1 for all acoustic components (modified from Young et al. 1994).

Lyon/Mono population; recordings at these sites were completed between 3 and 1 1 April 2002. The number of males sampled from each of the five leks was as follows: Desert Creek 2 (n = 6), Long Valley 1 ( n = 9), Biedeman ( n — 9), Eastside ( n = 11), and Roadside ( n = 14); therefore, the sample size for the Lyon/Mono population was n = 24.

Males perform a ritualized strut display in which they take a few steps forward and brush their wings twice against their esophageal pouch producing loud swishing noises (Fig. 2). Following these wing movements, males compress air sacs and produce syringeal sounds to complete a single strut display (Hjorth 1970). Male strut displays were re- corded and compared using the methods of Young et al. (1994) with the following mod- ifications. Only adult males were monitored, and these were distinguished from juveniles in the field by the presence of a clear white upper breast on adults. Individual males were iden- tified by their tail patterns (Wiley 1973). At least 15 struts per male were recorded using a Sony DCR TRV720 digital camcorder and a Sennheiser MKH70-P48 microphone. Sounds of individual struts were digitized at 22 kHz using Canary 1.2.4 sound analysis software (Cornell Laboratory of Ornithology, Ithaca, New York).

We measured nine acoustic components (Table 1, Fig. 2) and calculated population means derived from individual male averages

for each component. An estimate of repeat- ability ([r = s2a/(s2 + s2a)]; Lessells and Boag 1987) was used to measure the proportion of within-individual variation within populations for each component. Repeatabilities range from 0 (low) to 1.0 (high). High repeatabilities indicate that the measured trait varies little within individuals relative to the population variation, suggesting that the trait could re- spond to sexual selection.

To calculate strut display rate, we timed be- tween-strut intervals using Etholog 2.2, an ethological transcription tool (Ottoni 2000). The display rate for each male was based on 7—40 consecutive struts in which no more than 30 sec had lapsed between struts. Females were present on all leks during strut-rate mea- surements, but any male included in the strut- rate analyses had to have females within 20 m of them during recording. This criterion lowered the sample sizes (number of males) for population strut-rate estimation (Fig. 3). At the Lassen and Lyon leks, measurements were taken as one female moved throughout the leks. The southern Mono, northern Mono, and Nye leks all had multiple females visiting leks over the various days that measurements were taken.

We used analysis of variance (ANOVA) to assess differences among populations for each acoustic component and strut rate. We then used the GT2-method (Hochberg 1974) to make unplanned comparisons among popula-

Taylor and Young • GREATER SAGE-GROUSE BEHAVIOR

39

TABLE 1. Nine measured acoustic components of male Greater Sage-Grouse strut display in three popu- lations from Nevada and California. Males were recorded while strutting during spring 2001 and 2002.

Lyon, Nevada/Mono, Lassen, California Nye, Nevada

California ( n = 24) (n = 11) (n = 14)

Acoustic

Measured variable	component	Mean	SE	Mean	SE	Mean	SE	pa
First pop to whistle peak (msec) Whistle peak to whistle	1	73.41	0.37	73.85	0.65	70.30	0.52	<0.001
minimum (msec)	2	40.21	0.28	39.81	0.32	41.69	0.61	0.012
Pop to pop (msec) Whistle start frequency	3	199.89	0.73	199.64	0.97	192.24	0.88	<0.001
(Hz)	4	861.17	7.61	861.65	10.97	930.19	20.19	<0.001
Whistle peak (Hz)	5	2,619.83	21.06	2,657.32	23.09	2,873.84	42.85	<0.001
Whistle minimum (Hz) Whistle start to peak dif-	6	533.58	5.89	514.48	7.56	637.26	9.63	<0.001
ference (Hz) Whistle peak to mini-	7	1,771.61	20.69	1,795.22	23.94	1,944.72	35.09	<0.001
mum difference (Hz) Whistle start to mini-	8	2,096.48	21.90	2,151.64	17.61	2,241.51	39.14	0.002
mum difference (Hz)	9	333.90	11.33	353.70	13.99	290.38	16.80	0.020

a ANOVA.

tion means with unequal sample sizes for acoustic components. This method uses the studentized maximum modulus distribution m to compute a minimum significant difference (MSD). The significance level for the ANOVA was set at P — 0.05 and for the GT2- method it was lowered from P = 0.05 to P = 0.017 using a Bonferroni correction (a" = a/ k; Sokal and Rohlf 1995) for multiple tests. We used a" = 0.01 when referring to the stu- dentized maximum modulus m critical values table (GT2-method).

RESULTS

All nine acoustic components of the strut display differed among populations (ANOVA,

8.5

^ 8.0 j/5 D

2, 7.5

Q)

03

| 7.0 w

6.5 6.0

Lyon/Mono Lassen Nye

(ii)

I

F 2,31 = 3.97, P = 0.029

(16)

f

t

(7)

*

all P < 0.05; Table 1). The acoustic compo- nents of the males’ displays were similar be- tween Lyon/Mono and Lassen, whereas those of Nye males’ displays were consistently dis- tinct from those of the other two populations. Nye differed from both Lyon/Mono and Las- sen for acoustic components 1 and 3-7 (GT2- test, all P < 0.01). For component 8, Nye dif- fered only from Lyon/Mono (GT2-test, P < 0.010). All other pairwise population compar- isons for minimum significant differences were not significant (GT2-test, all P > 0.01).

Repeatability estimates of the acoustic com- ponents ranged from 0.41 to 0.84 in Lassen, 0.57 to 0.96 in Nye, and 0.35 to 0.91 in Lyon/ Mono (Table 2). The highest repeatability es- timate for all three populations was for whistle peak (component 5).

Strut rates (struts/min) differed (F2<31 = 3.97, P = 0.029) among populations (Fig. 3). However, pairwise comparisons between pop- ulations indicated that none were significant (GT2-test, all P > 0.01). Lassen males had the highest strutting rate (7.84 struts/min), where- as males from Nye had the lowest strutting rate (6.92 struts/min). Lyon/Mono males had an intermediate strutting rate (7.21 struts/min).

FIG. 3. Means (with standard error bars) and ANOVA result for strut rates of male Greater Sage- Grouse from three populations: Lyon, Nevada/Mono, California; Lassen, California; and Nye, Nevada. Sam- ple sizes (number of males) are in parentheses.

DISCUSSION

We measured behavioral traits and second- ary sexual characteristics that are related to sexual selection in sage-grouse, which could

40

THE WILSON JOURNAL OL ORNITHOLOGY • Vol. 118, No. 1, March 2006

TABLE 2. Repeatability estimates of strut display acoustic components within individual males from three Greater Sage-Grouse populations in California and Nevada. Males were recorded while strutting dur- ing spring 2001 and 2002.

Acoustic component	Lyon, Nevada/ Mono, California n = 24	Lassen, California n = 11	Nye, Nevada n = 14
1	0.51	0.78	0.65
2	0.35	0.44	0.62
3	0.64	0.74	0.65
4	0.57	0.67	0.79
5	0.91	0.84	0.96
6	0.57	0.68	0.79
7	0.53	0.80	0.88
8	0.74	0.49	0.87
9	0.41	0.41	0.57

therefore lead to divergence. Based on behav- ioral differences in male strut displays, our study did not support the idea that the genet- ically distinct Lyon/Mono population should be considered for separate taxonomic status. The Lyon/Mono and Lassen populations were similar to each other, while the Nye popula- tion was the most unique across nine acoustic components of male mating displays. How- ever, across six components (1-4, 6, 9), the Nye versus Lassen populations were either more different or as different as Nye versus Lyon/Mono populations (Table 1). Even though the Lyon/Mono population is geneti- cally distinct, male mating behaviors are more similar to those of the Nye population than those of the genetically similar Nye and Las- sen populations are to each other (Table 1).

The repeatability estimates generally varied widely across populations. However, three acoustic components (3, 5, and 9) were rela- tively comparable among the three popula- tions. The high repeatability estimates for components 3 (pop to pop) and 5 (whistle peak) indicate that these traits vary little with- in individual males relative to the variation within populations and could potentially re- spond to selection. Young et al. (1994) also found high repeatability estimates for whistle peak, which has been shown to be related to female mate choice (Gibson and Bradbury 1985, but see Gibson et al. 1991). A low re- peatability for component 9 (whistle start to minimum difference) is most likely the result of high levels of variability within individuals

rather than a lack of genetic variation or in- accuracies in measurement (Boake 1989). Nye had the highest repeatability estimates for sev- en of the nine acoustic components, suggest- ing low variation in the acoustic measure- ments, despite samples being taken across several days with multiple females being pres- ent.

Although strut rates did differ among pop- ulations, pairwise comparisons of strut rate did not differ statistically between popula- tions. This result agrees with the observations of Young et al. (1994), who found that strut rates did not differ between two Greater Sage- Grouse populations — Mono, California, and Jackson, Colorado. Strut rates may vary with time of day, time of season, and proximity of females (R. M. Gibson pers. comm.); there- fore, variation in strut rate within and between males may outweigh differences in strut rates among populations except in strong cases of population divergence.

Our results suggest that the Lyon/Mono population does not exhibit any appreciable behavioral differences in male mating displays from other Greater Sage-Grouse populations. The Lyon/Mono population is significantly different genetically from the Lassen popula- tion (Benedict et al. 2003, Oyler-McCance et al. 2005), yet behaviorally, the Lyon/Mono and Lassen populations have similar acoustic strut components and strut rates. The impli- cations of the slight behavioral differences ob- served in the Nye population on female mate choice may be determined upon further be- havioral observations that include additional leks, years, and populations. It is possible that there are measurable differences in acoustic components of the strut display between most populations, but these differences are gener- ally minimized by gene flow.

The Lyon/Mono population is genetically more diverse and distinct than the Gunnison Sage-Grouse species (Kahn et al. 1999, Oyler- McCance et al. 1999, Benedict et al. 2003, Oyler-McCance et al. 2005). Using mitochon- drial DNA sequence, Benedict et al. (2003) estimated that the Lyon/Mono population has been isolated from other Greater Sage-Grouse populations for tens of thousands of years. Yet, neither local adaptation to ecological or environmental factors, nor genetic drift, nor sexual selection has led to detectable pheno-

Taylor and Young • GREATER SAGE-GROUSE BEHAVIOR

41

typic (behavioral) differences in this popula- tion. Reproductive isolation does not appear to have occurred through sexual selection in the Lyon/Mono population as it has in the Gunnison Sage-Grouse species.

ACKNOWLEDGMENTS

Funding and support for this project was provided by the California Department of Fish and Game, Quail Unlimited, the National Fish and Wildlife Foundation, Western State College of Colorado, and the Bureau of Land Management. We are grateful to D. S. Blanken- ship, F. A. Hall, T. L. Russi, J. Fatooh, S. L. Nelson,

R. M. Gibson, W. F. Mandeville, and T. Slatauski for logistical and field support. We appreciate M. K. Bollig and M. D. Kascak for their assistance with graphics. We thank C. E. Braun, J. W. Connelly, R. M. Gibson,

S. J. Oyler-McCance, K. P. Reese, and J. St. John for helpful comments on the manuscript. Finally, we thank S. L. Thode and R. D. Taylor for patience and support.

LITERATURE CITED

American Ornithologists’ Union. 2000. Forty-sec- ond supplement to the American Ornithologists’ Union check-list of North American birds. Auk 117:847-858.

Benedict, N. G., S. J. Oyler-McCance, S. E. Taylor, C. E. Braun, and T. W. Quinn. 2003. Evaluation of the eastern ( Centrocercus urophasianus uro- phasianus ) and western ( Centrocercus urophasi- anus phaios) subspecies of sage-grouse using mi- tochondrial control-region sequence data. Conser- vation Genetics 4:301-310.

Boake, C. R. B. 1989. Repeatability: its role in evo- lutionary studies of mating behaviour. Evolution- ary Ecology 3:173-182.

Gibson, R. M. and J. W. Bradbury. 1985. Sexual se- lection in lekking Sage Grouse: phenotypic cor- relates of male mating success. Behavioral Ecol- ogy and Sociobiology 18:117-123.

Gibson, R. M. and J. W. Bradbury. 1987. Lek orga- nization in Sage Grouse: variations on a territorial theme. Auk 104:77-84.

Gibson, R. M., J. W. Bradbury, and S. L. Vehren- camp. 1991. Mate choice in lekking Sage Grouse revisited: the roles of vocal display, female site fidelity, and copying. Behavioral Ecology 2:165- 180.

Hjorth, I. 1970. Reproductive behavior in Tetraoni-

dae, with special reference to males. Viltrevy 7: 183-596.

Hochberg, Y. 1974. Some generalizations of the T- method in simultaneous inference. Journal of Mul- tivariate Analysis 4:224-234.

Hupp, J. W. and C. E. Braun. 1991. Geographic var- iation among Sage Grouse in Colorado. Wilson Bulletin 103:255-261.

Kahn, N. W., C. E. Braun, J. R. Young, S. Wood, D. R. Mata, and T. W. Quinn. 1999. Molecular anal- ysis of genetic variation among large- and small- bodied Sage Grouse using mitochondrial control region sequences. Auk 116:819-824.

Lessells, C. M. and P. T. Boag. 1987. Unrepeatable repeatabilities: a common mistake. Auk 104:1 16- 121.

Ottoni, E. B. 2000. EthoLog 2.2: a tool for the tran- scription and timing of behavior observation ses- sions. Behavior Research Methods, Instruments, & Computers 32:446-449.

Oyler-McCance, S. J., N. W. Kahn, K. P. Burnham, C. E. Braun, and T. W. Quinn. 1999. A popula- tion genetic comparison of large- and small-bod- ied Sage Grouse in Colorado using microsatellite and mitochondrial markers. Molecular Ecology 8: 1457-1465.

Oyler-McCance, S. J., S. E. Taylor, and T. W. Quinn. 2005. A multilocus population genetic sur- vey of Greater Sage-Grouse across their range. Molecular Ecology 14:1 293- 1310.

Schroeder, M. J., C. A. Aldridge, A. D. Apa, J. R. Bohne, C. E. Braun, S. D. Bunnell, J. W. Con- nelly, et al. 2004. Distribution of Sage Grouse in North America. Condor 106:363-376.

Sokal, R. R. and F. J. Rohlf. 1995. Biometry: the principles and practice of statistics in biological research. W. H. Freeman and Company, New York.

Wiley, R. H. 1973. Territoriality and non-random mat- ing in Sage Grouse, Centrocercus urophasianus. Animal Behavior Monographs 6:85-169.

Young, J. R., C. E. Braun, S. J. Oyler-McCance, J. W. Hupp, and T. W. Quinn. 2000. A new species of Sage-Grouse (Phasianidae: Centrocercus ) from southwestern Colorado. Wilson Bulletin 112:445- 453.

Young, J. R., J. W. Hupp, J. W. Bradbury, and C. E. Braun. 1994. Phenotypic divergence of second- ary sexual traits among Sage Grouse, Centrocer- cus urophasianus , populations. Animal Behaviour 47:1353-1362.

The Wilson Journal of Ornithology 1 1 8( 1 ):42— 52, 2006

FIRST KNOWN SPECIMEN OF A HYBRID BUTEO: SWAINSON’S HAWK ( BUTEO SWAINSONI) X ROUGH-LEGGED HAWK (B. LAGOPUS) FROM LOUISIANA

WILLIAM S. CLARK1 2 3 ’ AND CHRISTOPHER C. WITT-

ABSTRACT. — We report a specimen that appears to be a hybrid between Swainson’s Hawk ( Buteo swainsoni) and Rough-legged Hawk ( B . lagopus ), which, to our knowledge, is the first hybrid specimen for the genus. There are few reports of hybridization between Buteo species, most of which have been observations of inter- specific nesting pairs. The specimen described herein was collected in Louisiana and initially identified as a Rough-legged Hawk because of its feathered tarsi and the dark bellyband and carpals. A DNA sequence from the maternally inherited mitochondrial ND6 gene was identical to a published sequence for Swainson’s Hawk. Nuclear DNA sequences from two introns contained only five variable sites among a panel of five potential parental taxa, but the hybrid sequence was most consistent with parentage by Rough-legged and Swainson’s hawks. The feathered tarsi of the hybrid strongly suggested that the father was either a Rough-legged or Fer- ruginous hawk ( B . regalis ), the only North American raptors other than Golden Eagle ( Aquila chrysaetos ) that have feathered tarsi. Plumage and size characters were inconsistent with those of Ferruginous Hawk, and, other than the darkly pigmented leg feathers, were intermediate between the light morphs of Swainson’s and Rough- legged hawks. The breeding range of Swainson’s Hawk in Alaska and northern Canada is poorly known, but it overlaps that of the Rough-legged Hawk in at least a few locations, albeit at low densities, which may be a factor in hybridization. The occurrence of this hybrid is evidence of the potential for interbreeding between North American members of the genus Buteo , most of which are genetically closely related. Such hybridization could have implications for genetic diversity, adaptation, or the evolution of reproductive barriers. In any case, such hybrids present field and museum identification problems. Received 6 December 2004, accepted 3 October 2005.

Few documented cases of hybridization ex- ist between any 2 of the 27 or so species in the genus Buteo. Hybrid combinations have been reported for Long-legged Buzzard ( B . rufinus ) and Upland Buzzard (B. hemilasius ) in Asia (Pfander and Schmigalew 2001), Common Buzzard (B. buteo ) and Long-legged Buzzard in Europe (Dudas et al. 1999), and Red-shouldered Hawk (B. lineatus) and Gray Hawk ( Asturina nitidus ) in North America (Lasley 1989). Additionally, an adult Swain- son’s Hawk (B. swainsoni ) bred for more than 8 years with a presumably escaped South American Red-backed Hawk (Red-backed Buzzard, B. polyosoma ) in Colorado, USA, and produced offspring in some years (Allen 1988, Wheeler 1988); a Red-tailed Hawk ( B . jamaicensis ) that escaped from a falconer bred with a Common Buzzard in Scotland (Murray 1970). However, to our knowledge, there are

1 2301 S. Whitehouse Cir., Harlingen, TX 78550, USA.

2 Dept, of Biological Sciences and Museum of Nat- ural Science, Louisiana State Univ., 119 Foster Hall, Baton Rouge, LA 70803, USA.

3 Corresponding author; e-mail; raptours@earthlink.net

no museum specimens of the offspring of such unions. Thus, it was with great interest that we found a specimen of an apparent hybrid in the Louisiana State University Museum of Natural Science (LSUMNS), Baton Rouge. It is a juvenile male, has feathered tarsi and mostly dark carpal patches, was collected near Baton Rouge, Louisiana, and was identified as a Rough-legged Hawk (B. lagopus ). Its plum- age appears almost the same as that of a prob- able hybrid between the same two species, first seen and photographed in November 2002 by Martin Reid near Ft. Worth, Texas; WSC observed and took photos of that bird in January 2003.

Herein we present a description of the pu- tative hybrid Buteo based on its morphology, plumage, and mitochondrial and nuclear DNA sequences. A comparison of the hybrid to a set of potential parental Buteo taxa led to the conclusion that it descended from the mating of a female Swainson’s Hawk with a male Rough-legged Hawk. Although not shown on some published range maps, Swainson’s Hawks breed sparsely throughout at least a part of the Rough-legged Hawk’s breeding range in far-northern North America.

42

Clark and Witt • HYBRID BUTEO SPECIMEN

43

METHODS

WSC noted that the specimen, LSUMZ 159785, which was stored with a handful of juvenile light-morph Rough-legged Hawks, differed from them and was much like a pre- sumed hybrid he had seen and photographed near Ft. Worth, Texas in January 2003. After a comparison of this specimen with those of juvenile Rough-legged and Swainson’s hawks, he determined that it might be a hybrid. The specimen had been collected on 4 November 1994 in East Baton Rouge Parish, Highway 30 at Burtville, Louisiana, by S. W. Cardiff and D. L. Dittmann. A tissue sample was de- posited in the LSUMNS Collection of Genetic Resources (catalog No. B23743). The speci- men was sexed internally as a male (left testis 7X11 mm) and was in juvenal plumage; the skull was 75% ossified.

We used a DNEasy tissue kit (Qiagen, Va- lencia, California) to extract DNA from frozen muscle tissue of the putative hybrid specimen, and one specimen of each of the following taxa: Rough-legged Hawk, Swainson’s Hawk, Red-tailed Hawk, Harlan’s Red-tailed Hawk ( B . jamaicensis harlani), and Ferruginous Hawk. We amplified the mitochondrial ND6 gene for the hybrid specimen in 25 jjlI PCR reactions using Amplitaq Gold (Applied Bio- systems [ABI], Foster City, California) with the primers tPROfwd and tGLUrev (Haring et al. 1999). For all six specimens, we amplified two nuclear loci, as follows: (1) intron 5 and flanking exon regions of the cytosolic ade- nylate kinase gene (AK1) using the primers AK5b + and AK6c- (Shapiro and Dumbacher 2001), and (2) intron 3 and flanking exon re- gions of the Z-chromosome-linked muscle- specific receptor tyrosine kinase gene (MUSK) using primers designed by F. K. Barker: MUSK-E3F (CTTCCATGCACTAC AATGGGAAA) and MUSK-E4R (CTCTGA ACATTGTGGATCCTCAA). Standard PCR reactions were run on an MJ Research PTC- 200 thermal-cycler under the following tem- perature regime: initial denaturation at 95° C for 8 min; 35 cycles of 92° C for 20 sec, 55° C for 60 sec, 72° C for 60 sec; and a final extension at 72° C for 10 min. For MUSK, the annealing temperature was adjusted to 50° C. Negative control reactions were used for all extractions and PCR to insure against contam-

ination. PCR products were purified using a Qiagen Gel Extraction Kit (Qiagen, Valencia, California). Cycle-sequencing reactions were carried out in both directions using the prim- ers described above in quarter- or sixteenth- volume reactions with a Big Dye Terminator Cycle Sequencing Kit (ver. 2 or 3.1, ABI). Cy- cle-sequencing products were purified using Sephadex columns. Purified samples were electrophoresed on an ABI 377 or 3100 au- tomated sequencer. Sequences were assem- bled and edited using Sequencher 4.2.2 (Gene Codes Corporation, Ann Arbor, Mich- igan). The ND6 sequence was compared with published sequences for various Buteo species (Riesing et al. 2003).

We compared morphology and plumage of the hybrid to a panel of five potential parental taxa. We followed the “contradictory charac- ters” approach of Rohwer (1994) to eliminate potential pairs of parental taxa for which char- acters of the presumed hybrid fall outside of the range of variation. We assembled standard measurements of body mass, wing chord (un- flattened), exposed culmen, and hallux (Bald- win et al. 1931) for juvenile males of potential paternal taxa from banding data for Swain- son’s, Rough-legged, and eastern Red-tailed hawks ( B . j. borealis), and from museum specimen data for western Red-tailed ( B . j. calurus ), Harlan’s Red-tailed, and Ferruginous hawks. We performed two stepwise discrimi- nant function analyses with these four mor- phological variables using SPSS ver. 11.5 (SPSS, Inc. 2002). In both stepwise analyses, we used 0.05 probability of F for entry and 0.10 probability of F for removal of each var- iable, set equal prior probabilities of group membership, and used within-group covari- ance matrices. The three Ferruginous Hawk specimens were not included in the analysis due to small sample size, and the single Har- lan’s Red-tailed Hawk individual was includ- ed in the western Red-tailed Hawk group. The first discriminant function analysis included Rough-legged, Swainson’s, eastern Red-tailed, and western Red-tailed hawks as groups. All four morphological variables were significant and included in the final model, and three sig- nificant discriminant functions were generat- ed. The putative hybrid individual and the three Ferruginous Hawks were then classified using these discriminant functions. In the sec-

44

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

ond discriminant function analysis, we only included Rough-legged and Swainson’s hawks as groups. Only mass, wing chord, and cul- men were significant and included in the final model, and only one discriminant function ex- plained 100% of the variation between the two groups. The putative hybrid was then again classified according to this discriminant func- tion. To account for possible shrinkage of mu- seum specimens relative to live birds (Winker 1993), we repeated all analyses under the as- sumption of a 3% reduction in size due to dry- ing. The adjustment for shrinkage had no sub- stantive effect on the results. Finally, with re- spect to plumage characters, we compared the specimen with juvenile male Swainson’s and Rough-legged hawks, including pigmentation of the head, upperparts, breast, belly, tail, and legs, and emargination of the seventh primary (P7).

RESULTS

The mitochondrial DNA sequence of the putative hybrid, totaling 558 bp, was an iden- tical match to a published sequence from a Swainson’s Hawk collected in New Mexico (Table 1; GenBank accession No. AY2 13028). The sequence was 0.76% divergent from its nearest relative, the Galapagos Hawk (B. ga- lapagoensis ), and 3.23-3.58% divergent from the only sympatric congeners: Red-tailed, Fer- ruginous, and Rough-legged hawks (Clark and Wheeler 2001, Riesing et al. 2003; Table 1). Mitochondrial haplotypes are shared between mothers and their offspring because the mi- tochondrial genome is non-recombining and maternally inherited (Lansman et al. 1983). The identical mtDNA sequences of the spec- imen and a known Swainson’s Hawk strongly suggests that the maternal parent was a Swain- son’s Hawk.

The nuclear AK1 sequence of the putative hybrid, totaling 542 bp, was identical to se- quences from the Swainson’s, Rough-legged, eastern Red-tailed, Harlan’s, and Ferruginous hawks. The complete lack of variation at this locus prevents the elimination of any of these taxa as potential parents. The nuclear MUSK sequence, totaling 599 bp, contained five var- iable sites for the six taxa included in this study (Table 2). Among the five variable sites was a substitution unique to the Ferruginous Hawk sample (T; site no. 480), and another

Louisiana State University Museum of Natural Science, Baton Rouge.

Clark and Witt • HYBRID BUTEO SPECIMEN

45

TABLE 2. Variable sites on the 599 bp MUSK gene sequence for the presumed Buteo hybrid and five other buteos. The sites span part of exon 3, the entire intron 3, and part of exon 4, corresponding to positions 131 1922- 1312509 of the Gallus gallus chromosome Z genomic contig (GenBank NW 060751). Both states (i.e., A/T and A/G) are reported for heterozygous sites, as inferred by unambiguous double peaks on chromatograms.

			Variable position
65	113	157	452	480
Hybrid	A/T	A/G	c	A/G	c
Swainson’s Hawk	T	A	c	G	c
Rough-legged Hawk	A/T	A/G	c	G	c
Ferruginous Hawk	T	A	c	G	T
Eastern Red-tailed Hawk	T	A	T	G	c
Harlan’s Red-tailed Hawk	T	A	T	G	c

that was shared only by the eastern Red-tailed and Harlan’s Red-tailed hawks (T; site no. 157). At two other sites (nos. 65 and 1 13), the Rough-legged Hawk and the hybrid were both heterozygous (A/T and A/G), with one exclu- sively shared state and one state in common with all other taxa (Table 2). The fifth variable site (no. 452) was heterozygous in the hybrid specimen only. Heterozygotes were inferred when chromatograms showed strong signal and unambiguous double peaks of nearly equal height.

We identified the paternal parent using phe- notypic characters. Red-tailed Hawk, includ- ing Harlan’s Hawk, can be eliminated as the putative father because it always has unfeath- ered tarsi. It seems unlikely that two species with bare tarsi would produce a hybrid with feathered tarsi. Further, the Red-tailed Hawk’s culmen is considerably larger than that of the hybrid (Table 3). Finally, juvenile Red-tailed

Hawks share few plumage characters with the hybrid (Wheeler and Clark 1995, Clark and Wheeler 2001); we would not expect, for ex- ample, a hybrid Red-tailed Hawk X Swain- son’s Hawk juvenile to have the heavy, dark bellyband (Fig. 1) or the dark carpal patches of the hybrid.

Both Ferruginous and Rough-legged hawks have feathered tarsi and are the most likely paternal candidates of the hybrid specimen. However, Ferruginous Hawks have noticeably wider gapes (Bechard and Schmutz 1995) and longer bills, wings, and halluces than the hy- brid (Table 3). The measurements of the hy- brid are far closer to those of Swainson’s Hawk than to Ferruginous Hawk, suggesting that the bird is not intermediate in size as would be expected in an FI hybrid between these two species. In contrast, the measure- ments for body mass and wing chord are in- termediate between juvenile male Swainson’s

TABLE 3. Comparison of measurements (mean ± SE) of the hybrid Buteo specimen with juvenile male Rough-legged, Swainson’s, Ferruginous, eastern Red-tailed, western Red-tailed, and Harlan’s Red-tailed hawks. Body mass and wing chord of the hybrid are intermediate between Rough-legged and Swainson’s hawks. Culmen and hallux are closest to Swainson’s Hawk.

	n	Body mass (g)	Wing chord (mm)	Culmen (mm)	Hallux (mm)
Hybrid	i	702.0	381.0	19.3	21.4
Swainson’s Hawk3	20	638.3 ± 16.8	378.5 ± 2.4	21.4 ± 0.3	21.7 ± 0.4
Rough-legged Hawkb	39	860.8 ± 12.6	398.2 ± 1.6	21.5 ± 0.1	23.9 ± 0.2
Ferruginous-Hawkc	3	1,091.4 ± 14.3	413.7 ± 1.8	25.0 ± 0.3	25.6 ± 0.3
Eastern Red-tailed Hawkd	24	825.4 ± 15.8	351.8 ± 1.9	27.2 ± 0.2	24.1 ± 0.2
Western Red-tailed Hawke	12	905.5 ± 30.3	374.4 ± 2.9	24.2 ± 0.3	27.7 ± 0.4
Harlan’s Red- tailed Hawkf	1	932.0	365.0	23.5	26.0

3 Unpublished banding data from Texas and New Jersey, sex determined by size. b Unpublished banding data from New York, sex determined by size.

e MVZ (Museum of Vertebrate Zoology, University of California, Berkeley) specimen data from California. d Unpublished banding data from New Jersey, sex determined by size.

e MVZ specimen data from British Columbia, California, Arizona, New Mexico, and Nevada. f MVZ specimen data from British Columbia.

46

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

FIG. 1 . Specimens showing ventral view of the hybrid Buteo (center), compared with juvenile male Rough- legged Hawk (left) and juvenile male Swainson’s Hawk (right). Characters of the hybrid are intermediate.

and Rough-legged hawks (Table 3). Finally, the plumage characters of both light- and dark-morph juvenile Ferruginous Hawks do not match those of the specimen (Wheeler and Clark 1995, Clark and Wheeler 2001); a hy- brid Ferruginous Hawk X Swainson’s Hawk juvenile, for example, would not be expected to have the dark bellyband (Fig. 1) nor the dark carpal patches of the hybrid.

Most plumage characters of the hybrid specimen are similar to those of juvenile male Swainson’s or Rough-legged hawks, or inter- mediate between them (Figs. 1-2, Table 4).

The notching of P7 is also intermediate (Fig. 3). This feather has a noticeable abrupt wid- ening or “notch” on the trailing edge for Rough-legged Hawk (same for Ferruginous and Red-tailed hawks) but not for Swainson’s Hawk. The widening begins 93 mm from the tip on a juvenile male specimen Rough-legged Hawk (Fig. 3 A), widening about 15 mm at an angle of 70° to the feather shaft. P7 on a ju- venile male Swainson’s Hawk specimen be- gan widening gradually 47 mm from the tip and lacked a distinctive notch (Fig. 3B). The hybrid’s P7 began widening 59 mm from the

Clark and Witt • HYBRID BUTEO SPECIMEN

47

FIG. 2. Specimens showing dorsal view of the hybrid Buteo (center), compared with juvenile male Rough- legged Hawk (left), and juvenile male Swainson’s Hawk (right). Characters of the hybrid are intermediate.

tip with a notch and widened about 9 mm at a 60° angle (Fig. 3C).

In the first discriminant function analysis, which included Rough-legged, Swainson’s, eastern Red-tailed, and western Red-tailed hawks as groups, the first two discriminant functions explained 96.2% of the variation be- tween the groups (Fig. 4A). The first function correlated strongly with culmen (r = 0.651) and wing chord (r = —0.513) and explained 80.1% of the variance. The second function correlated strongly with hallux (r = 0.814) and body mass (r = 0.646) and explained

16.1% of the variance. Using both functions, the hybrid was classified as a Rough-legged Hawk with 3 1 .2% probability, as a Swainson’s Hawk with 68.8% probability, and as an east- ern or western Red-tailed Hawk with 0% probability. In the second discriminant func- tion analysis, which included only Rough-leg- ged and Swainson’s hawks as groups, one dis- criminant function explained 100% of the var- iation between the groups (Fig. 4B). This function correlated strongly with mass (r = 0.875) and wing chord (r = -0.580). Using this function, the hybrid was classified as a

48

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

TABLE 4. Comparison of plumage characters of the hybrid Buteo specimen with juvenile male Rough- legged and Swainson’s hawks. Characters of the hybrid are intermediate or like one or the other of the parent species.
Character		Rough-legged Hawk	Swainson’s Hawk	Hybrid
Crown		Pale	Dark	Dark, pale streaks
Superciliary		None	Rufous	Buffy
Malar		Narrow	Wide	Wide
Back feathers		Brown, pale sides	Dark brown, pale tips	Dark brown, pale tips and sides
Breast		Lightly streaked	Heavily streaked	Heavily streaked
Belly		Solidly dark	Buffy	Dark with pale edges
Legs		Feathered, lightly marked	Bare	Feathered, darkly marked
Uppertail		White base, dusky tip, no bands	Gray-brown, dark bands	Narrow white base, gray-brown, dark bands
Primary, outer web	Grayish cast	Dark	Grayish cast
Primary, inner	web	Pale, no barring	Darker, barring	Pale, barring

Rough-legged Hawk with 45.4% probability and as a Swainson’s Hawk with 54.5% prob- ability.

DISCUSSION

Based on mtDNA, we conclude that the mother of this putative hybrid is a Swainson’s Hawk. The most likely paternal candidates are raptors with feathered tarsi. Rough-legged and

Ferruginous hawks. The latter was eliminated because of its plumage characters, much larg- er size, and unique MUSK intron haplotype.

Independent lines of evidence converged on the identification of the specimen as a hybrid between Swainson’s and Rough-legged hawk. The combination of morphological and mo- lecular characters, as in the diagnosis of a hy- brid manakin ( Ilicura X Chiroxiphia) by Ma-

FIG. 3. Notching of primary 7. (A) Rough-legged Hawk, (B) hybrid, and (C) Swainson’s Hawk. The pos- terior margin of each P7 is highlighted in white. (Scale is not the same on each figure.) The shape of P7 of the hybrid is intermediate and unlike those of any Buteo species.

Clark and Witt • HYBRID BUTEO SPECIMEN

49

• Eastern Red-tailed Hawk

a Western Red-tailed Hawk o Rough-legged Hawk (RLHA)

O Harlan's Hawk & Swainson's Hawk (SWHA)

Discriminant function 1

FIG. 4. Discriminant function analyses comparing juvenile males of Buteo species. In panel (A), plots of points along the first two significant discriminant func- tions are from an analysis that included Rough-legged, Swainson’s, eastern Red-tailed, and western Red-tailed hawks as groups. These two discriminant functions ex- plained 96.2% of the variation between the groups. The Harlan’s Hawk was included in the western Red- tailed Hawk group, but was plotted with a unique sym- bol. The hybrid individual and three Ferruginous Hawks were then classified and plotted using these dis- criminant functions. In panel (B), points are plotted according to a discriminant function from an analysis that only included Rough-legged and Swainson’s hawks as groups. One discriminant function explained 100% of the variation between the two groups. The hybrid was classified and plotted according to this dis- criminant function.

rini and Hackett (2002), is a powerful method for the identification of avian hybrids. In par- ticular, the comparison of a single mtDNA se- quence to the growing database of published sequences is an outstanding tool for identifi- cation of the maternal parent. In this case, the mtDNA sequence of the hybrid strongly sug-

gests that its maternal parent was a Swain- son’s Hawk. The mother could have been a species other than Swainson’s Hawk only if the mitochondrial identity were a mere artifact of incomplete lineage sorting. We consider this possibility unlikely because the mitochon- drial study of Riesing et al. (2003) demon- strated that geographically heterogeneous samples of five Rough-legged, two Ferrugi- nous, nine Red-tailed, and three Swainson’s hawks are each reciprocally monophyletic, and the divergence levels between Swainson’s Hawk and each of its sympatric congeners are greater than 3%.

The paucity of variation in the two nuclear introns illustrates the difficulty of using nu- clear DNA to diagnose hybrids among closely related species. Intraspecific variation and lack of lineage sorting pose significant challenges to the conclusive identification of hybrid in- dividuals, and these problems are compound- ed when potential parental taxa cannot be thoroughly sampled at the population level. Despite these difficulties, our sample of a sin- gle individual for each potential parental tax- on yielded some variation that was consistent with the identification of Rough-legged Hawk as the paternal species. The eastern Red- tailed, Harlan’s Red-tailed, and Ferruginous hawk samples each contained single substi- tutions on the MUSK intron that were not found in the hybrid. In contrast, only the Swainson’s and Rough-legged hawk samples were completely compatible with parentage of the hybrid. Importantly, two heterozygous po- sitions in the hybrid each contained a state that was shared exclusively with the Rough- legged Hawk sample.

Plumage and morphological characters of the hybrid specimen were generally interme- diate between those of juvenile males of the parent species. This pattern is born out by the discriminant function analyses and is consis- tent with the characters of hybrids between other species of birds (e.g.. Graves 1990, Roh- wer 1994, Marini and Hackett 2002). How- ever, the coloration of the tarsi feathers was not intermediate. Juvenile male Rough-legged Hawks have buffy tarsal feathers with sparse, dark markings, whereas Swainson’s Hawks have bare tarsi. The hybrid specimen has tar- sal feathers with heavy, dark barring, clearly not intermediate. The expectation that hybrid

50

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

traits fall within the range of traits expressed by the parental taxa is based on the assump- tion that most traits are additive and polygenic (Falconer 1989) and is implicit in most hybrid diagnoses. Nonetheless, hybrids can also ex- press traits that are extreme relative to those of the parental taxa (Rieseberg et al. 1999). It is possible that the darkly pigmented tarsal feathers could be one such transgressive trait, caused by complementary gene action, over- dominance, or epistasis. Swainson’s and Rough-legged hawk populations are known to possess genetic variation that results in differ- ences in the quantity and distribution of mel- anin-based plumage pigments (Clark and Wheeler 2001). Rohwer (1994) reported other examples of characters that were not inter- mediate between those of the parental species. The culmen, and, to a lesser degree, the hallux of the hybrid were slightly smaller than our Swainson’s and Rough-legged hawk measure- ments for those characters, providing another potential example of a non-intermediate char- acter. However, specimen shrinkage could at least partly account for this difference.

The Swainson’s Hawk breeds in an un- known amount of the breeding range of the Rough-legged Hawk in far northwestern North America. This is the extreme northern periphery of their distribution, and they occur at very low densities in taiga habitat where they are sympatric with the Rough-legged Hawk (England et al. 1997, Bechard and Swem 2002, Sinclair et al. 2003). This could increase the possibility that a female Swain- son’s Hawk could fail to find a conspecific mate. Given the broad overlap in distribution between Swainson’s, Red-tailed, and Ferrugi- nous hawks, the lack of documented instances of hybridization or interspecific pairings be- tween any two of these three species suggests behavioral barriers to reproduction. Such bar- riers may not exist between Swainson’s and Rough-legged hawks, which overlap only marginally and may have come into sympatry only recently. This hybrid pairing is consistent with the model of Short (1969), who proposed that hybridization is most likely to occur at the edges of a species’ range.

Swainson’s Hawks are rare during Novem- ber in the area where the hybrid individual was found; there is only one November record for East Baton Rouge Parish, despite intensive

coverage by birdwatchers and collectors (LSUMNS data). Although Lowery (1974) in- dicated that Rough-legged Hawk is a regular winter visitor to Louisiana, and several sub- sequent sight-based reports lacking photos have been accepted by the Louisiana Bird Re- cords Committee, the only physical evidence substantiating the occurrence of a Rough-leg- ged Hawk in Louisiana is a specimen collect- ed on 12 March 1933 at Grand Isle (LSUMZ 4803). The present hybrid occurred at a place (and time) unexpected for either species — Rough-legged Hawks should occur farther north and Swainson’s Hawks farther south. This intermediate migratory behavior, as well as a myriad of other ecological differences be- tween Swainson’s and Rough-legged hawks, suggests potential sources of reduced fitness in hybrids. Hybridization can provide a mech- anism for gene flow between species, partic- ularly if hybrids are interfertile with parental species and do not suffer reduced fitness (Ar- nold 1992). Alternatively, hybrid unfitness can reinforce behavioral pre-mating barriers through natural selection (Saetre et al. 1997), particularly in taxa such as Swainson’s and Rough-legged hawks that may have recently come into secondary contact.

Hybrids between raptor species are reported infrequently, most likely because they are rare, but also because they are difficult to di- agnose in the field and are underrepresented in collections. That this specimen went unrec- ognized for 9 years after being collected un- derscores the field and museum identification problems posed by hybrids. Hybrids have been reported between Red Kite ( Milvus mil- vus) and Black Kite (A/, migrans) in Sweden (Sylven 1977), a possible hybrid Rueppell’s Vulture ( Gyps rueppellii ) and Cape Vulture ( G . coprotheres ) in Botswana (Borello 2001), Brown Goshawk {Accipiter fasciatus ) and Grey Goshawk (A. novaehollandiae ) in Aus- tralia (Olsen 1995), Shikra (A. badius ) and Le- vant Sparrowhawk (A. brevipes) in Israel (Yosef et al. 2001), Pallid Harrier ( Circus ma- crourus ) and Montagu’s Harrier (C. pygargus ) in Finland (Forsman 1995), Western Marsh Harrier (C. aeruginosus ) and Eastern Marsh Harrier (C. spilonotus ) in Siberia (Fefelov 2001), and Greater Spotted Eagle ( Aquila clanga) and Lesser Spotted Eagle (A. poma- rina) in Latvia (Bergmanis et al. 1996). We

Clark and Witt • HYBRID BUTEO SPECIMEN

51

were unable to locate a copy of Suchelet (1897), who apparently reported a hybrid be- tween Common Buzzard and Rough-legged Hawk. Most unusual were intergeneric hy- brids reported between Black Kite and Com- mon Buzzard near Rome, Italy, that produced rather strange-looking offspring (Corso and Glidi 1998). Equally unusual was a pairing between Gyrfalcon ( Falco rusticolus ) and Per- egrine Falcon (F. peregrinus), in which both members of the pair were females (Gjershaug et al. 1998). The hybrid Turkey Vulture X Black Vulture reported by Mcllhenny (1937) was later determined to be a practical joke (Jackson 1988). Most instances of hybridiza- tion listed above were determined at the nests by observing that the adults were different species, although one was a hybrid captured for banding (Yosef et al. 2001) and another was identified using field observations and photographs (Corso and Glidi 1998).

To our knowledge, our report is the first of a hybrid specimen arising from two Buteo species, and, perhaps, the first hybrid speci- men for any raptor. It provides the first con- clusive documentation of hybridization be- tween two native North American members of the genus Buteo. A pairing of a Red-shoul- dered Hawk with a Gray Hawk (Lasley 1989) produced a downy chick, but it did not fledge, and there were neither photographs nor spec- imens from this union.

ACKNOWLEDGMENTS

This study was facilitated by the Collections of Birds and the Collection of Genetic Resources at the LSU Museum of Natural Science, and the Collection of Birds at the Museum of Vertebrate Zoology, Uni- versity of California, Berkeley. We thank S. W. Cardiff and D. L. Dittmann for finding and collecting this un- usual specimen and P. Bloom, A. Hinde, Braddock Bay Raptor Research, and Cape May Raptor Banding Pro- ject for sharing measurements of juvenile male Swain- son’s, Rough-legged, and Red-tailed hawks with us. F. K. Barker provided previously unpublished primers for the MUSK gene. J. V. Remsen, Jr., J. Schmutz, T. Swem, S. M. Witt, and an anonymous reviewer made helpful comments on previous drafts.

LITERATURE CITED

Allen, S. 1988. Some thoughts on the identification of Gunnison’s Red-backed Hawk ( Buteo polyo- soma) and why it’s not a natural vagrant. Colorado Field Ornithologist Journal 22:9-14.

Arnold, M. L. 1992. Natural hybridization as an evo-

lutionary process. Annual Review of Ecology and Systematics 23:237-261.

Baldwin, S. P, H. C. Oberholser, and L. G. Worley. 1931. Measurements of birds. Scientific Publica- tions of the Cleveland Museum of Natural Histo- ry, no. 2. Cleveland, Ohio.

Bechard, M. J. and J. K. Schmutz. 1995. Ferruginous Hawk ( Buteo regalis). The Birds of North Amer- ica, no. 172.

Bechard, M. J. and T. R. Swem. 2002. Rough-legged Hawk ( Buteo lagopus). The Birds of North Amer- ica, no. 641.

Bergmanis, U., A. Pertins, M. Strads, and I. Krams.

1996. Possible case of hybridization of the Lesser Spotted Eagle and the Greater Spotted Eagle in eastern Latvia. Putni daba 6.3: 1-6. [In Latvian, with an English summary.]

Borello, W. 2001. Possible hybrid vulture at Man- nyelanong Cape Vulture Gyps coprotheres colony, southeastern Botswana. Babbler 38:19-21.

Clark, W. S. and B. K. Wheeler. 2001. Hawks of North America, 2nd ed. Peterson Field Guide Se- ries, no. 35. Houghton Mifflin, Boston, Massachu- setts.

Corso, A. and R. Glidi. 1998. Hybrids between Black Kite and Common Buzzard in Italy in 1996. Dutch Birding 20:226-233.

Dudas, M., J. Tar, and I. Toth. 1999. Natural hy- bridization of Long-legged Buzzard ( Buteo rufi- nus) and Common Buzzard ( B . buteo) in the Hor- tobagy National Park. Temeszet 5-6:8-10. [In Hungarian]

England, A. S., M. J. Bechard, and C. S. Houston.

1997. Swainson’s Hawk ( Buteo swainsoni). The Birds of North America, no. 265.

Falconer, D. S. 1989. Introduction to quantitative ge- netics, 3rd ed. Longman Scientific Technical, Es- sex, United Kingdom.

Fefelov, I. 2001. Comparative breeding ecology and hybridization of Eastern and Western Marsh Har- riers Circus spilonotus and C. aeruginosus in the Baikal region of eastern Siberia. Ibis 143:587- 592.

Forsman, D. 1995. Male Pallid and female Montagu’s Harrier raising hybrid young in Finland in 1993. Dutch Birding 17:102-106.

Gjershaug, J. O., A. O. Folkestad, and L. O. Gok0yr. 1998. Female-female pairing between a Peregrine Falcon Falco peregrinus and a Gyrfal- con F. rusticolus in two successive years. Fauna Norvegica Series C, Cinclus 21:87-91.

Graves, G. R. 1990. Systematics of the “Green-throat- ed Sunangels” (Aves: Trochilidae): valid taxa or hybrids? Proceedings of the Biological Society of Washington 103:6-25.

Haring, E., M. J. Riesing, W. Pinsker, and A. Ga- mauf. 1999. Evolution of a pseudo-control region in the mitochondrial genome of Palearctic buz- zards (genus Buteo). Journal of Zoological Sys- tematic and Evolutionary Research 37:185-194.

Jackson, J. A. 1988. Turkey Vulture. Page 27 in Hand-

52

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

book of North American birds, vol. 4 (R. S. Palm- er, Ed.). Yale University Press, New Haven, Con- necticut.

Lansman, R. A., J. C. Avise, and M. D. Huetel. 1983. Critical experimental test of the possibility of “pa- ternal leakage” of mitochondrial DNA. Proceed- ings of the National Academy of Sciences USA 80:1969-1971.

Lasley, G. 1989. Texas. American Birds 43:505.

Lowery, G. H., Jr. 1974. Louisiana birds. Louisiana State University Press, Baton Rouge.

Marini, M. A. and S. J. Hackett. 2002. A multifac- eted approach to the characterization of an inter- generic hybrid manakin (Pipridae) from Brazil. Auk 119:1114-1120.

McIlhenny, E. A. 1937. Hybrid between Turkey Vul- ture and Black Vulture. Auk 54:384.

Murray, J. B. 1970. Escaped American Red-tailed Hawk nesting with a Buzzard in Midlothian. Scot- tish Birds 6:34-37.

Olsen, P. 1995. Australian birds of prey. The Johns Hopkins University Press, Baltimore, Maryland.

Pfander, P. and S. Schmigalew. 2001. Extensive hy- bridization of Long-legged Buzzard Buteo rufinus and Upland Buzzard B. hemilasius. Omithologis- che Mitteilung 53:344-349. [In German]

Rieseberg, L. H.. M. A. Archer, and R. K. Wayne. 1999. Transgressive segregation, adaptation and speciation. Heredity 83:363-372.

Riesing, M. J., L. Kruckenhauser, A. Gamauf, and E. Haring. 2003. Molecular phylogeny of the ge- nus Buteo (Aves: Accipitridae) based on mito- chondrial marker sequences. Molecular Phyloge- netics and Evolution 27:328-342.

Rohwer, S. 1994. Two new hybrid Dendroica war-

blers and new methodology for inferring parental species. Auk 111:441-449.

Saetre, G. P, T. Moum, S. Bures, M. Kral, M. Ada- mjan, and J. Moreno. 1997. A sexually selected character displacement in flycatchers reinforces premating isolation. Nature 387:589-592.

Shapiro, L. H. and J. P. Dumbacher. 2001. Adenylate kinase intron 5: a new nuclear locus for avian sys- tematics. Auk 118:248-255.

Short, L. L. 1969. Taxonomic aspects of avian hy- bridization. Auk 86:84-105.

Sinclair, P. H., W. A. Nexon, C. D. Eckert, and N. L. Hughes (Eds.). 2003. Birds of the Yukon Ter- ritory. University of British Columbia Press, Van- couver, British Columbia, Canada.

SPSS, Inc. 2002. SPSS for Window, ver. 1 1.5.0. SPSS Inc., Chicago, Illinois.

Suchelet, A. 1897. Des hybrids a L’etat sauvage, part 4: birds of prey. Librairie. J. B. Bailliere & fils, Paris, France. [In French]

Sylven, M. 1977. Hybridization between Red Kite Milvus milvus and Black Kite M. migrans in Swe- den in 1976. Var Fagelvarld 36:38-44. [In Swed- ish, with an English summary.]

Wheeler, B. K. 1988. A Red-backed Hawk in Colo- rado. Colorado Field Ornithologist Journal 22:5-8.

Wheeler, B. K. and W. S. Clark. 1995. A photo- graphic guide to North American raptors. Aca- demic Press, London, United Kingdom.

Winker, K. 1993. Specimen shrinkage in Tennessee Warblers and “Traill’s” Flycatchers. Journal of Field Ornithology 64:331-336.

Yosef, R., A. J. Helbig, and W. S. Clark. 2001. An intrageneric Accipiter hybrid from Eilat, Israel. Sandgrouse 23:141-143.

The Wilson Journal of Ornithology 1 1 8( 1 ):53 — 58, 2006

NOCTURNAL HUNTING BY PEREGRINE FALCONS AT THE EMPIRE STATE BUILDING, NEW YORK CITY

ROBERT DeCANDIDO1 34 AND DEBORAH ALLEN1 2 3 4

ABSTRACT. — We report on nocturnal hunting by Peregrine Falcons ( Falco peregrinus) at the Empire State Building in Manhattan, New York City. From 4 August through 13 November 2004, we saw Peregrine Falcons on 41 of 77 nights of observation. During this period, they hunted migrating birds on 25 evenings, with the first hunting attempt occurring an average of 119 min after sunset. Peregrine Falcons made 111 hunting attempts and captured 37 birds (33% success). Hunting success was highest in September, but was most often observed in October. Peregrines hunted migratory birds at night more frequently in autumn than in spring. Peregrines were significantly more likely to be present on autumn nights when >50 migrants were passing by the Empire State Building. Although the lights associated with skyscrapers are believed to disorient migrating birds and result in many bird-to-skyscraper collisions each year. Peregrine Falcons are able to take advantage of the situation. Skyscrapers provide hunting perches at altitudes often flown by nocturnal migrants, and disorientation caused by the lights sometimes results in birds circling skyscrapers and possibly becoming more vulnerable to predation by falcons. Received 26 January 2005, accepted 11 October 2005.

Several diurnal raptor species, including Black-shouldered Kite ( Elanus axillaris ), Bald Eagle ( Haliaeetus leucocephalus), and Lesser Kestrel ( Falco naumanni ), forage at night (see Kaiser 1989, McLaughlin 1989, Negro et al. 2000). Others, such as Turkey Vulture ( Ca - thartes aura). Osprey ( Pandion haliaetus ), Northern Harrier ( Circus cyaneus ), and Le- vant Sparrowhawk ( Accipiter brevipes), have been observed flying or migrating at night (Tabor and McAllister 1988, Russell 1991, Yosef 2003, DeCandido et al. 2006).

Peregrine Falcons ( Falco peregrinus) are considered nocturnal migrants in some parts of the world (Cochran 1985, Ellis et al. 1990), and they are known to hunt at night (Clunie 1976, Russell 1998). With increased numbers of peregrines nesting and wintering in cities, biologists are beginning to document noctur- nal activity by these falcons in all seasons. Recently, there have been reports of urban peregrines feeding young and/or hunting at night in North America (Cade and Bird 1990, Wendt et al. 1991, Cade et al. 1996), England (Crick et al. 2003), France (Marconot 2003), Germany (Schneider and Wilden 1994, Klad-

1 Hawk Mountain Sanctuary, Acopian Center for Conservation Learning, 410 Summer Valley Rd., Or- wigsburg, PA 17961, USA.

2 P.O. Box 1452, Peter Stuyvesant Station, New York, NY 10009, USA.

3 Current address: 1831 Fowler Ave., The Bronx, NY 10462, USA.

4 Corresponding author; e-mail: rdcny@earthlink.net

ny 2001), Netherlands (van Dijk 2000, van Geneijgen 2000), Poland (Rejt 2000, 2001, 2004a), Hong Kong (Feare et al. 1995), and Taiwan (K. Y. Huang and L. L. Severinghaus unpubl. data). However, direct observation and analysis of nocturnal hunting by Peregrine Falcons, particularly during migration, is rare in the literature.

In New York City, New York, the number and distribution of Peregrine Falcons has changed considerably since such observations were first recorded in the late 1920s. Before the era of DDT (until 1946), from autumn through early spring, lone female peregrines were much more common at skyscrapers than males (Herbert and Herbert 1965). Peregrine Falcons rarely nested in the city, and nocturnal activity by these falcons was not reported in any season (Herbert and Herbert 1965). Be- ginning in the mid-1990s, however, more pairs of Peregrine Falcons have begun residing year-round in Manhattan (and the metropoli- tan area) than previously noted (B. A. Loucks pers. comm., C. Nadareski unpubl. data.). To- day, most, if not all, of the seven pairs of per- egrines that nest in Manhattan remain on ter- ritory year-round. Here, we report our obser- vations of Peregrine Falcon activity at night during the 2004 southbound bird migration at one location in New York City.

METHODS

Most of our observations of Peregrine Fal- cons and nocturnal migrants occurred during

53

54

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

the southbound migration, from 4 August to 13 November 2004; we made observations on 77 of 102 evenings during that period. In spring 2004, we observed northbound mi- grants on 33 evenings from 19 April through 25 May. In spring 2002, we made observa- tions on only 2 evenings (8 May and 15 May).

We made our observations from the outside observation deck (elevation —325 m above ground level) of the Empire State Building (ESB), located in midtown Manhattan in New York City. We arrived each evening approxi- mately 15-30 min prior to sunset. Bird mi- gration, on average, began 30-90 min after sunset. Any Peregrine Falcon activities de- fined as nocturnal occurred after nautical twi- light (1 hr after sunset). We were able to con- duct our study until 22:45 EST each evening (August through October) and until 23:45 in November; the observation deck of the build- ing was closed to all visitors after these times. In spring 2004, we observed from just before sunset until 22:45 each evening, and in spring 2002, we observed from 19:00 until 21:00. During fall migration, the northwest corner of the building provided the best vantage point to count the greatest number of migrating birds, and in spring, we observed migrants from the southwest corner of the observation deck. These locations afforded unobstructed views to the horizon and the sky above. We used 10X binoculars to follow peregrines when they made long flights in pursuit of prey. It was possible to observe migrating birds and the activities of peregrines because the upper floors of the building were illumi- nated with (external) upward-directed halogen lights, and the spire above us was illuminated with (internal) florescent lights. We could not identify the majority of migrants to species because the external halogen lights washed out most plumage details. However, this light- ing array permitted us to count migrants up to —30-60 m above the highest point (445 m agl) of the ESB, and up to 30 m (perpendic- ular) from the observation deck. We estimated that the building’s lights allowed us to see per- egrines chasing small birds in flight up to 60- 80 m distant.

Count protocols to assess nocturnal bird mi- gration in 2004 followed those described in Bildstein and Zalles (1995) for migrating rap- tors. An individual was considered a migrant

if it passed south-to-north (or north-to- south) across an imaginary east-west line at the site, and continued north (or south) out of sight. On 2 evenings during southbound migration, when >100 birds simultaneously circled the ESB, we estimated the maximum number of birds circling per hour and recorded it as the number of migrants seen for that hour. We de- fined the peak of migration as the several-day period in which we counted the highest num- ber of migrants. For both northbound and southbound migration, total counts presented here do not include migrating waterfowl, her- ons, or gulls.

We defined a hunting attempt as one in which a Peregrine Falcon approached to with- in 1 m of its intended prey. On a few occa- sions, peregrines made repeated stoops at the same prey, but did not capture or gain control of it. Each of these stoops was considered a separate hunting attempt. Several times, we observed a peregrine strike a bird but fail to seize it. We classified these as unsuccessful hunting attempts.

We defined the peak period of Peregrine Falcon activity as that during which we ob- served falcons at the ESB during the greatest number of consecutive nights. We used cor- relation statistics (Microsoft Excel 2003) to analyze data collected during this peak period. We compared (a) the time of arrival of the first migrant after sunset with the arrival of the first Peregrine Falcon, and (b) the time of arrival of the first migrant with the time of the first peregrine hunting attempt. Means are present- ed as ± SD.

RESULTS

During southbound migration in 2004, we saw the first Peregrine Falcon at night on 4 August and the last one on the evening of 9 November. During this time, at least two adult peregrines (male and female), as well as im- mature^), used the ESB as a hunting perch. Peregrines were seen hunting or flying at night on 53% (41 of 77) of the evenings we spent at the ESB (Table 1). Falcons were signifi- cantly more likely to be present on evenings when >50 migrants were counted in migra- tion (x2 = 14.7, df = 1, P = 0.001; Table 1). Of the 67 nights we observed migrating birds, peregrines hunted migrants on 25 nights (37%), made 111 hunting attempts, and cap-

DeCandido and Allen • NOCTURNAL HUNTING BY PEREGRINE FALCONS

55

TABLE 1. Summary of nocturnal hunting behavior by Peregrine migrants present after sunset in autumn 2004 at the Empire State Build	Falcons in relation ing. New York.	to the number of
		Number classes of migrant passerines		Total
0	1-10	11-50	51-100	101-250	251 +
No. nights migrants counted	10	9	23	10	13	12	77
No. nights peregrines present	1	1	12	8	9	10	41
No. nights peregrines hunted	—	0	8	3	7	7	25
No. hunting attempts	—	0	29	17	15	50	111
No. successful hunts	—	0	8	7	8	14	37
Hunting success	—	—	28%	41%	53%	28%	33%
No. nights male observed hunting	—	0	5	2	5	6	18
No. nights female observed hunting	—	0	2	1	1	1	5
No. nights unknown sex observed hunting	—	0	1	0	1	1	3

tured prey 37 times (33% success). All of the migrants we observed being captured or chased were in the warbler-to-oriole size class.

The peak of Peregrine Falcon activity oc- curred from 26 September through 14 October 2004. During that time, we conducted obser- vations on 17 nights; on 16 of those nights we observed Peregrine Falcons, and on 1 1 nights we observed them hunting (70 total hunts, 21 prey captures, 30% success). During this pe- riod, the first migrant birds were observed 65 ± 20 min after sunset (range = 42-1 14 min); Peregrine Falcons arrived 91 ±41 min after sunset (range = 47-190 min), and made their first hunting attempt 45 ± 59 min later (range = 61-284 min), or approximately 136 min af- ter sunset. There was no correlation between passage of the evening’s first migrant and the arrival of a Peregrine Falcon at the ESB (r2 = 0.10, P = 0.73) or between passage of the first migrant and the time of a peregrine’s first hunting attempt (r2 = 0.15, P — 0.24).

Nocturnal hunting success was greatest in September (12 of 27, 44%) and lowest in No-

vember (1 of 8, 13%; Table 2). On 10 October from 20:12 to 20:42, a male Peregrine Falcon made 25 hunting attempts and captured 9 birds (36%), caching the birds on the ESB tower after each kill. Throughout the autumn, we observed Peregrine Falcons capture only migratory birds, although a few Rock Pigeons ( Columba livia), and at least two bat species. Little Brown ( Myotis lucifugus ) and Red ( Las - iurus borealis ) bats, were present on some evenings. We could identify only two prey species: a Baltimore Oriole (. Icterus galbula) captured on 23 August, and a Yellow-billed Cuckoo ( Coccyzus americanus) taken on 9 October. On 3 and 9 November, despite high numbers of American Woodcocks ( Scolopax minor ) migrating past the ESB tower (36 counted each night), no peregrines were ob- served.

In autumn 2004, most bird migration oc- curred at eye-level and above the observation deck. We counted 10,826 migrating birds, and the peak of the migration occurred from 5 to 11 October when 3,871 migrants (36% of the

TABLE 2. Summary of nocturnal hunting behavior and success by Peregrine Falcons during four autumn months in 2004 at the Empire State Building, New York.
	Aug	Sep	Oct	Nov	Total
No. hunting attempts	16	27	60	8	111
No. successful hunts	6	12	18	1	37
Hunting success	38%	44%	30%	13%	33%
No. nights one peregrine present	10	11	10	3	34
No. nights ^2 peregrines present	0	3	4	0	7
No. nights hunting observed	5	9	10	1	25
No. nights male made a hunting attempt	5	7	5	1	18
No. nights female made a hunting attempt	0	2	3	0	5
No. nights unknown sex made a hunting attempt	—	1	2	—	3

56

THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 118, No. 1, March 2006

fall flight) were counted, averaging 1 14 birds/ hr on these 7 evenings. In spring 2004, we counted 3,359 migrants during 33 nights of observation. The peak of the migration oc- curred from 6 to 15 May when 1,752 migrants (52% of the spring flight) were counted, av- eraging 51 birds/hr on these 10 evenings. Lone Peregrine Falcons were observed on 2 evenings: 24 April (0 migrants counted) and 22 May (79 counted), but no hunting attempts were observed on either night. On 15 May 2002, we observed an adult female peregrine make 10 unsuccessful hunting attempts on mi- grants from 20:15 until 21:00.

In the breeding season of 2004, a pair of Peregrine Falcons may have attempted to nest on the ESB (B. A. Loucks pers. comm.). It is possible that this pair executed many of the hunting attempts we observed in autumn 2004. During 5 evenings between 26 Septem- ber and 7 October, we saw an adult male and an adult female peregrine perched near one another, each vocalizing with the “eechup” or “creaking” call, and the “wailing” calls (see Ratcliffe 1980). On 3 October, we observed three adults (a male, his mate, and a second female) perched for <5 min within