Volume 37 Number 1 March 2003 Published by The Raptor Research Foundation, Inc. THE RAPTOR RESEARCH FOUNDATION, INC. (Founded 1966 ) OFFICERS PRESIDENT: Brian A. Millsap SECRETARY: Judith Henckfx VICE-PRESIDENT: David M. Bird TREASURER: Jim Fitzpatrick BOARD NORTH AMERICAN DIRECTOR #1; Jeff Smith NORTH AMERICAN DIRECTOR #2: Laurie J. Goodrich NORTH AMERICAN DIRECTOR #3: Ted Swem INTERNATIONAL DIRECTOR #1: Beatriz Arroyo INTERNATIONAL DIRECTOR #2: Ricardo RodrIguez-Estrella OF DIRECTORS INTERNATIONAL DIRECTOR #3: Steve Redpath DIRECTOR AT LARGE #1: Jemima ParryJones DIRECTOR AT LARGE #2: Petra Bohall Wood DIRECTOR AT LARGE #3: Michael W. Collopy DIRECTOR AT LARGE #4: Carol McIntyre DIRECTOR AT LARGE #5: Robert N. Rosenfield DIRECTOR AT LARGE #6: Daniel E. Varland EDITORIAL STAFF EDITOR: James C. Bednarz, Department of Biological Sciences, P.O. Box 599, Arkansas State University, State University, AR 72467 U.S.A. ASSOCIATE EDITORS James R. Beithoff Marco Restani Clint W. Boat Ian G. Warkentin Joan L. Morrison Troy I. Wellicome Juan Jose Negro BOOK REVIEW EDITOR: Jeffrey S. Marks, Montana Cooperative Research Unit, University of Montana, Missoula, MT 59812 U.S.A. SPANISH EDITOR: Cesar Marquez Reyes, Instituto Humboldt, Colombia, AA. 094766, Bogota 8, Colombia EDITORIAL ASSISTANTS: Rebecca S. Maul, Lynn Hodgens, Joan Clark The Journal of Raptor Research is distributed quarterly to all current members. Original manuscripts dealing with the biology and conservation of diurnal and nocturnal birds of prey are welcomed from throughout the world, but must be written in English. Submissions can be in the form of research articles, short communications, letters to the editor, and book reviews. Contributors should submit a typewritten original and three copies to the Editor. All submissions must be typewritten and double-spaced on one side of 216 X 278 mm (8V2 X 11 in.) or standard international, white, bond paper, with 25 mm (1 in.) mar- gins. The cover page should contain a title, the author’s full name(s) and address (es). Name and address should be centered on the cover page. If the current address is different, indicate this via a footnote. A short version of the title, not exceeding 35 characters, should be provided for a running head. An abstract of about 250 words should accompany all research articles on a separate page. Tables, one to a page, should be double-spaced throughout and be assigned consecutive Arabic numer- als. Collect all figure legends on a separate page. Each illustration should be centered on a single page and be no smaller than final size and no larger than twice final size. The name of the author(s) and figure number, assigned consecutively using Arabic numerals, should be pencilled on the back of each figure. Names for birds should follow the A.O.U. Checklist of North American Birds (7th ed., 1998) or another authoritative source for other regions. Subspecific identification should be cited only when pertinent to the material presented. Metric units should be used for all measurements. Use the 24-hour clock (e.g., 0830 H and 2030 H) and “continental” dating (e.g., 1 January 1999). Refer to a recent issue of the journal for details in format. Explicit instructions and publication policy are outlined in “Information for contributors,”/. Raptor Res,, Vol. 36(4), and are available from the editor. Submit manuscripts to J. Bednarz at the address listed above. COVER: Black-winged Kite {Elanus caeruleus). Painting by Juan Varela (see www.juanvarela.com). Contents Copulation Behavior of a Potentiauly Double-brooded Bird of Prey, the Black- winged Kite {ElANUS CAERULEUS). JuanJ. Ferrero,Juan M. Grande, and Juan J, Negro 1 Mortality, Morbidity, and Lead Poisoning of Eagles in Western Canada, 1986-98. Mark Wayland, Laurie K Wilson, John E. Elliott, Michael J.R. Miller, Trent Bollinger, Malcolm McAdie, Ken Langelier, Jonathan Keating, and Jennifer M.W. Froese 8 Abundance of Soaring Raptors in the Brazilian Atlantic Rainforest. Santi Mahosa, Eduardo Mateos, and Vittorio Pedrocchi 19 Differential Spring Migration of Adult and Juvenile Levant Sparrowhawks {AcCIPITER BREVIPES) Through Eilat, Israel. Reuven Yosef, Lorenzo Fornasari, Piotr Tryjanowski, Marc J. Bechard, Gregory S. Kaltenecker, and Keith Bildstein 31 Vocal Development in American Kestrel {Falco sparverius) Nestlings. John a. Smallwood, Valerie Dudcyek, Siv^ini Gilchrist, Mary Anne Smallwood 37 Morphometrics and Status of Ayres’s Hawk-Eagle in Zimbabwe. Ronald r. Hartley and Peter J. Mundy 44 Breeding Density and Altitudinal Distribution of the Ural, Tawny, and Boreal Owls in North Dinaric Alps (Central Slovenlv) . Ai Vrezec 55 Short Communications Fall Migration of the White-tailed Hawk in Central Bolivia. Cristian oiivo 63 Nocturnal Arrival at a Roost by Migrating Levant Sparrowhawks. Reuven Yosef 64 Nesting Distribution, Food Habits, and Conservation of Osprey on Boavista Island (Archipelago of Cape Verde) . Diego Ontiveros 67 Red-shouldered Hawk and Aplomado Falcon from Quaternary Asphalt Deposits IN Cuba, william Suarez, Storrs L. Olson 71 Subadult and Pale Steppe Eagles Breeding in Mongolia. David h. eius 75 Two Large Bald Eagle Communal Winter Roosts in Utah. Robert Wilson and James a. Gessaman 78 Letters First Breeding Record for Falco peregrinus in Urban Lima, Peru, with Remarks on the Peruvian Breeding Population. Oscar Beingolea and Clayton M. White 84 Lone Harris’s Hawk Kills Great Blue Heron. Hope D. Woodward and R. William Trussell 85 Book Reviews. Edited by Jeff Marks 87 Manuscript Referees 89 The Raptor Research Foundation, Inc. gratefully acknowledges funds and logistical support provided by Arkansas State University to assist in the publication of the journal. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBUCATION OF THE RAPTOR RESEARCH FOUNDATION, INC. VoL. 37 March 2003 No. 1 J Raptor Res. 37(l):]-7 © 2003 The Raptor Research Foundation, Inc. COPULATION BEHAVIOR OF A POTENTIALLY DOUBLE-BROODED BIRD OF PREY, THE BLACK-WINGED KITE (ELANUS CAERULEUS) JuanJ. Ferrero/ Juan M. Grande, and Juan J. Negro^ Department of Applied Biology, Estacion Biologica de Donana (CSIC), Pabellon del Peru, Avda. JVP- Luisa s/n 41013, Sevilla, Spain Abstract. — Raptors show high copulation rates when compared to other birds. This fact has been generally attributed to sperm competition. However, copulation rates in raptors tend to be seasonally bimodal, with an early peak outside the female’s fertile period that cannot be explained by the sperm competition hypothesis. We studied the copulation behavior of the Black-winged Kite {Planus caeru- leus), that, unlike raptors previously studied, may raise two broods consecutively in the same breeding season. The first recorded copulation occurred 44 d prior to the onset of laying and the last one on d 66 after. We observed three seasonal peaks in copulation frequency. The first two peaks (40 and 10 d before egg laying, on average) correspond to the normal pattern in raptors. To explain the first peak, we propose three hypotheses: (1) pair bonding, (2) mate assessment, and (3) territorial sig- naling. Unfortunately, we have been unable to elaborate further predictions that would distinguish among these three hypotheses. The second peak, during the fertile period, may be attributed to sperm competition. The third peak, previously unreported in raptors, occurred 50 d after the onset of laying, the mean time lag between first and second clutches in the area. These late copulations, which were also performed by pairs that laid one clutch, may reflect the propensity of the species to lay second clutches (we recorded six replacement clutches and five second clutches out of 98 breeding attempts) . This third peak of copulations may also have a fertilization function and may be related to sperm competition. Keywords: Black-winged Kite; Elanus caeruleus; copulation behavior, sperm competition; territorial signaling. COMPORTAMIENTO COPULATORIO DE UNA RAPAZ CON DOBLE NIDADA POTENCIAL: EL ELANIO COMUN (ELANUS CAERULEUS). Resumen. — Las aves rapaces suelen presentar altas tasas copulatorias cuando se las compara con otras aves. Este hecho ha sido generalmente atribuido a la competencia espermatica. Sin embargo, la tasa de copulas en aves rapaces tiende a presentar un patron estacional bimodal, con un primer pico antes del periodo en que la hembra es fertil, que no puede ser explicado por esta hipotesis. En este trabajo estudiamos el comportamiento copulatorio del elanio comun {Elanus caeruleus) , que a diferencia de las otras rapaces previamente estudiadas puede sacar adelante dos polladas consecutivas en la misma tem- porada de cria. La primera copula se observe 44 dias antes del inicio de la puesta y la ultima 66 dias despues. Observamos tres picos en la frecuencia copulatoria a lo largo de la temporada de cria. Los primeros dos (alrededor de los dias 40 y 10 antes de la puesta) corresponden al patron observado en otras rapaces. Para explicar el primer pico proponemos que se consideren tres hipotesis: (1) reforza- miento de los lazos de pareja, (2) valoracion de la calidad del companero y (3) senalizacion del terri- torio. Desafortunadamente, no hemos podido elaborar predicciones que permitan distinguir entre estas ^ Present address: Tirso de Molina, 14, Merida 06800, Badajoz, Spain. ^ Corresponding author’s e-mail address: negro@ebd.csic.es 1 2 Ferrero et al. VoL. 37, No. 1 tres hipotesis. El segundo pico, durante el periodo fertil de la hembra podria atribuirse a fenomenos de competencia espermatica. El tercero, observado por primera vez en rapaces, ocurrio 50 dias despues del inicio de la puesta, lo que coincide con el intervalo medio entre primeras y segundas puestas en la zona de estudio. Este tercer pico, que tambien se detecto en aquellas parejas que solo hicieron una puesta, podria reflejar una propension en esta especie a realizar segundas puestas (encontramos seis puestas de reemplazo y cinco segundas puestas en 98 intentos de cria), Este tercer pico de copulas estaria probablemente relacionado con la competencia espermatica. [Traduccion de los autores] The copulation behavior of birds has been a popular research topic among behavioral ecolo- gists in the last two decades (e.g., Birkhead et al. 1987, Birkhead and Mpller 1992). This interest has been triggered by the observation of widely vari- able copulation rates between species (i.e., from one to several hundred copulations in a single breeding season), and the realization, through mo- lecular paternity assessment, that extra-pair fertil- izations are common in certain species (Petrie and Kempenaers 1998). Male birds seem to have evolved two strategies to minimize the risk of paternity loss: (1) close mate guarding (Birkhead 1979) and (2) frequent within-pair copulations to dilute the sperm of pos- sible competitors (Birkhead et al. 1987). Most rap- tor species probably employ the second strategy as males typically provide food to the female through- out the pre-laying period and, therefore, are pre- cluded from guarding their mates effectively (M 0 ller and Birkhead 1992). Raptors show high copulation rates, although the seasonal pattern of copulations in these birds is not consistent with all predictions of the sperm-competition hypothesis. Many raptors show an early peak of copulations 20—65 d before laying, before the females are fer- tile (M 0 ller 1987, Negro et al. 1992, Villarroel et al. 1998, Mougeot 2000, Negro and Grande 2001), and some species copulate after clutch completion, well into the chick-rearing period (Ellis and Powers 1982, Sodhi 1991, Holthuijzen 1992). In addition, extra-pair copulations, and also extra-pair fertiliza- tions, are infrequent in raptors, and this again calls into question the sperm-competition hypothesis (Negro et al. 1992, 1996, Villarroel et al. 1998, Ne- gro and Grande 2001). There are at least three alternative hypotheses for high copulation rates in raptors. The pair-bonding hypothesis (Newton 1979) states that the members of the pair copulate frequently to maintain or to reinforce the pair bond. Copulations are expected to occur all through the breeding period, but are not neces- sarily restricted to this period if pair bonds are maintained through the year. The mate-assessment hypothesis (Tortosa and Redondo 1992, Negro et al. 1996, Villarroel et al. 1998) states that the fe- males may acquire information on male quality through copulations. Assuming that copulations are costly, better males would be able to copulate more often. The territorial signaling hypothesis (Negro and Grande 2001) proposes that raptors signal territory ownership to conspecifics, and pos- sibly to other avian species, by copulating frequent- ly and conspicuously in the defended nesting area early in the breeding season. The Black-winged Kite (Elanus caeruleus; also called the common Black-shouldered Kite in the Old World literature) is a small-sized (ca. 300 g) raptorial bird widely distributed in Africa (Cramp and Simmons 1980). In southern Africa, it breeds continuously, while in other areas of Africa it breeds mostly in the spring and fall (Cramp and Simmons 1980, Brown et al. 1982, Mendelsohn 1983). The species is slowly colonizing southern Europe (Ferrero 1994, Rufino 1994). In Spain, egg-laying peaks in March (Cramp and Simmons 1980, unpubl. data), but in some years some pairs may lay eggs as early as November or as late as July (Ferrero and De Lope 2001). Our aim with this paper is to describe the cop- ulation behavior and pattern of Black-winged Kites. Contrary to all raptorial species previously studied, the Black-winged Kite may raise two broods in a year (Mendelsohn 1983) and could have evolved a distinct seasonal pattern of copulations (i.e., some copulations are expected to occur after the brst clutch) . McrnoDs We studied a breeding population of Black-winged Kites in Extremadura, southwestern Spain. The mam breeding habitat used by Black-winged Kites in the area consists of fragmented semi-open Mediterranean oak for- ests (mainly Quercus rotundifolia) with an understory of cereal crops (cultivated “dehesas”; Ferrero and Onrubia 1998). The stronghold for the European population (es- timated at 1000 breeding pairs [Ferrero 1994]) is located Marcum 2003 Copulation in Black-winged Kite 3 Table 1. Number of breeding pairs for which mating behavior was monitored between 1979 and 2000. Year No. OF Pairs Total Observations (hr) 1979 13 123.4 1980 12 20.7 1987 8 46.9 1988 26 53.3 1989 12 126.9 1993 2 32.7 1995 14 44.2 1996 2 8.4 1997 6 19.2 2000 3 11.5 in the dehesas of Extremadura and the neighboring Alentejo in Portugal. We performed behavioral observations year round on a total of 98 breeding attempts in 79 different nesting territories between 1979 and 2000 (Table 1). We ob- served six renesting attempts after the first clutch failed to hatch (replacement clutches) and five second clutch- es. The observations were performed from vantage points at 200-300 m from the nests, using spotting scopes (20— 60X) and binoculars (lOX). Observers completed 269 observation bouts lasting 10 min to 14 hr (mean ob- servation period: 80 min). Given that it is unlikely to ob- serve copulations during very short observation periods, only those bouts more than 20 min long were used to calculate copulation rate and to plot the seasonal pattern of copulations. We were unable to distinguish between successful and unsuccessful copulations, and therefore we defined a copulation as occurring when a male mounted a female. Copulations as well as other distinct behaviors, including prey transfers among pair members, chasing to intruders, aerial displays, and delivery of nest material, were recorded and timed. Black-winged Kites in this study were not individually marked. This limitation may have resulted in the detec- tion of intruders, and thus, observation of extra-pair cop- ulations. Nonetheless, any intrusions by conspecifics dur- ing observations would be easily detectable, as home ranges of kites are relatively small (2-4 km^, Mendelsohn 1983, Bustamante 1993, J. Ferrero unpubl. data), tree cover in the dehesas is sparse (3-9 trees/ha), and terri- torial birds may be observed continuously. In addition, Black-winged Kites are strongly territorial and intruders are evicted from the nesting areas (Mendelsohn 1983, see Results) . The absence of marks may also have caused pseudoreplication due to the inclusion of data from the same pairs observed for more than one year. However, given the length of the study period (21 yr), and even considering that 13 territories were monitored twice or even three times, the long time lag between two consec- utive observations in the same territory (.5.8 ± 4.8 yr, on average), it is unlikely these territories were occupied by the same individuals. The remaining 65 territories were sampled once. Therefore, we believe the incidence of pseudoreplication is minimal in our analysis. We analyzed the daily variation in copulation rates dur- ing the fertile period. In previous studies on copulation behavior of raptors, the fertile period has been assumed to start around 12 d prior to laying (Bird and Buckland 1976, Negro et al. 1992, Donazar et al. 1994, Mougeot 2000). With no data available for the Black-winged Kite, we have conservatively assumed that the fertile period started 15 d before the onset of laying and ended around 10 d after when the clutch was completed (4-5 eggs layed with an interval of 2 d between consecutive eggs in the clutch [Cramp and Simmons 1980]). For analytical pur- poses and given that Black-winged Kites present activity peaks at dawn and dusk (Cramp and Simmons 1980), we divided daytime into three periods: (1) from sunrise to 3 hr later, (2) midday hours (of variable duration owing to photoperiodic variation throughout the copulatory pe- riod), and (3) from 3 hr before sunset to sunset. As the copulation rate within the different observation bouts fol- lowed a Poisson distribution, we fitted a Generalized Lin- ear Model (GEM) with Poisson errors to test for differ- ences among the three dehned daily periods. To analyze seasonal variation, days of observation were grouped into 5-d intervals relative to the estimated onset of laying (considered as day zero), and thus, indepen- dently of the period of the year in which each pair started to breed. To calculate hatching dates the 8**^ primary feather of the oldest chick of three different broods, whose hatch- ing date was precisely known, was measured twice a week through the chicks’ growing period until they were about to fledge. Using these data a linear regression of the length of the 8'^^ primary feather on age (in days) was calculated (age = 0.1837 primary + 10.277, = 0.9852, P < 0.0001). This regression line was subsequently used to estimate hatching dates for the remaining chicks in the study. Laying date was then estimated from hatching date assuming a 31 d incubation period (del Hoyo et al 1994). Data from nests that failed early, for which we did not know the laying date, were excluded from analysis Copulations during second breeding attempts, for which observations were limited, were not considered to profile the species’ seasonal-copulation pattern. Results Copulation Behavior. Each pair of Black-winged Kites usually copulated at 3-4 exposed perching sites in the nesting territory (not further than 150 m from the nest). These perching sites tended to be high and leafless branches at the top of a tree, which along with the bold coloration of the birds (white breast and belly, bluish grey upperparts) and their noisy vocalizations made copulations very conspicuous. No copulations were recorded at the nest. Copulation duration averaged 11.01 ± 0.45 sec (x ± SE,N= 75). On 161 instances, we recorded the behavior of the kites immediately before copulation took place. For 55 copulations (34.2%) we did not ob- 4 Ferrero et al. VoL. 37, No. 1 3.00 01 Day Figure 1. Mean ± SE frequency of copulation/hr of the Black-winged Kite related to the onset of laying (d 0). The line was adjusted with a normal kernel smoother. serve any prior social behavior among mates. In some cases, two or more characteristic behaviors preceded copulation. The male transferring prey to his mate preceded 32 copulations (19.8%). On those occasions, the male waited until the female finished her meal, then flew directly onto her back and copulated. Thirty-one copulations (19.2%) were preceded by the delivery of sticks to the nest by one member of the pair. On 26 occasions (16.1%) copulations occurred shortly after aggres- sive encounters had taken place with intruding birds, mainly Common Buzzards {Buteo buteo) and Common Ravens {Corvus corax), which were chased away from the breeding area. Intruding Black-winged Kites were observed in 4.8% of the observation periods (13 out of 269 observation bouts). As with the other bird species, intruding kites were invariably expelled by one or both mem- bers of the pair. Only two of these intrusions by conspecifics were immediately followed by copula- tion of the resident pair. Frequency and Timing of Copulations. A total of 216 copulations were observed in 487.3 hr of ob- servation. The first recorded copulation occurred 44 d prior to the onset of laying and the last one on d 66 after the onset of laying. The maximum number of copulations seen during a single obser- vation bout was nine in 4 hr, which occurred 12 d prior to laying a second clutch. Using six pairs for which we have observational data on more than 10 different days, and consid- ering daytime periods of 12 hr and a copulation period of 110 d (Fig. 1) we estimated 600.6 ± 166.2 copulations/female for the whole breeding season and 354.9 ± 98.2 for the first breeding at- tempt (65 d). Combining data from the different pairs under investigation, but excluding those that laid a sec- ond clutch (see Methods), a trimodal pattern emerged in daily copulation rates through the breeding season (Fig. 1). There was a first peak around d —40 and a second peak at d —10. Kites started to copulate again around 25 d after laying, resulting in a third copulation peak around d 50, almost coinciding with the mean dates of replace- ment clutches (53.83 ± 4.04 d after laying date of the first clutch, N = 6) and close to the mean date for second breeding attempts (second laying date, 62 ± 3.76 d after laying date of the first clutch, N = 4). We failed to find differences in copulation rates between the three defined daily periods (i.e., 3 hr post sunrise, midday and 3 hr to sunset), as the change in deviance of the GLM was not signif- icant (P = 0.36). No copulations were detected in twilight, despite the fact that the kites were ob- served hunting frequently at these times. Discussion Patterns of Copulation. Copulations in the Black-winged Kite were conspicuous, and thus, eas- ily detectable. Copulations were distributed throughout the daytime, and were not necessarily associated to other social behaviors such as mate feeding. The two peaks in copulation frequency be- fore and during the presumed fertile period of fe- male Black-winged Kite have previously been ob- served in other raptors (M 0 ller 1987, Negro et al. 1992, Pandolfi et al. 1998, Villarroel et al. 1998, Mougeot 2000). The third peak, at day +50, is re- ported for the first time in raptors, although it was anticipated due to the relative frequency of second clutches among Black-winged Kites. Considering only the copulations performed during a single breeding attempt (presumably the first one in the majority of pairs that we studied) , copulation rates were high and in the range found in other diurnal raptors (see references above, also Balgooyen 1976, Rosenfield et al. 1991, Korpimaki et al. 1996). Our estimate for the Black-winged Kite (around 5.5 copulations/d) is similar to that re- ported by Van Der Merwe (1980) for a pair in South Africa (7 copulations/d in a 15-d period dur- ing nest building) . As we already discussed (see Introduction), the first peak in copulation frequency around 40 d pri- or to laying is unlikely related to fertilization, as it March 2003 CopuiATioN IN Black-winged Kite 5 probably occurs well before the female’s fertile pe- riod. Alternative hypotheses, such as mate-assess- ment or strength of the pair bond hypotheses pre- dict higher copulation rates at the time of pair formation or pair reunion, but are difficult to test. The territory-signaling hypothesis predicts fre- quent and conspicuous copulations when a pair es- tablishes a breeding territory (Negro and Grande 2001), In the Black-winged Kite, copulations are indeed very frequent and almost always occur on conspicuous perches. The frequency of copula- tions immediately preceded by agonistic encoun- ters with other bird species or conspecifics is not negligible (16%). Our data seem to support the territory-signaling hypothesis as a possible expla- nation for the early peak in copulations as sug- gested for most diurnal raptors (Negro and Grande 2001), but we cannot discard either the pair-bond or the mate-assessment hypotheses. The second peak in copulation frequency, at the beginning of the female’s fertile period, may be related to sperm competition pressures. However, some aspects of the kite’s behavior seem to indi- cate low levels of sperm competition during this second peak. First, as in other frequently-copulat- ing birds (Tortosa and Redondo 1992, Birkhead and M 0 ller 1993, Bertran and Margalida 1999), the copulation rate decreased from beginning to end of the fertile period. In the context of sperm com- petition, copulation rates should be highest at the onset of laying (Birkhead 1988, Birkhead and M0ller 1992, but see Birkhead and M0ller 1993). Two intrusions of Black-winged Kites followed by copulation of the resident pair occurred during the fertile period. However, intrusions by conspe- cifics do not seem to be frequent, at least in our study area. Bustamante (1993) found only six in- trusions in 146 hr of observation (0.04 intrusions/ hr), and we found just 13 in 487 hr (0.03 intru- sions/hr), which is similar to that found in the Merlin {Falco columbarius, 0.02—0.05 intrusions/hr, Sodhi 1991). Mougeot (2000) reported higher fre- quencies of intrusions for the Red Kite {Milvus mil- vus, 0.7-4 intrusions/hr). Copulations After the First Nesting Attempt: Re- nesting, Second Clutches or Sequential Polyandry? The most remarkable finding of our study is the existence of a third peak in copulation frequency later in the breeding season, close to the fledging time of the young produced in the first breeding attempt. In some raptors, copulations cease at the end of laying or close to it (Newton 1986, Birkhead and Lessells 1988, Negro et al. 1992, Mougeot 2000). In other species the occurrence of copula- tions during incubation or even during the nest- ling period are not uncommon (Ellis and Powers 1982, Holthuijzen 1992, Donazar et al. 1994, Pan- dolfi et al. 1998). Nonetheless, we know of no oth- er raptor species in which late copulations reached similar rates to those of the pre-laying period, as we found in this study. Copulations after clutch completion are suggested to be a preventive be- havior to speed up a replacement clutch if the first one is lost (Birkhead et al. 1987). But renesting attempts are scarce among raptors, especially in larger ones (Newton 1979, Mundy et al. 1992, Mar- tinez et al. 1998). Second breeding attempts are even rarer, but do occur in some species (Newton 1979, Toland 1985). In the Iberian Peninsula only the Eurasian Kestrel {Falco tinnunculus) has been observed raising a second brood in the same year (Sanchez 1990, Fargallo et al. 1996). When condi- tions permit, the Black-winged Kite may breed two or more times in a year (Mendelsohn 1983). Ap- propriate conditions do not seem to be common in Spain but they do occur, as we observed five second-breeding attempts. In small birds, such as passerines, both replacement clutches and multi- ple breeding attemps are common (e.g., M 0 ller 1985), and therefore, copulations are expected in these species after the first breeding attempt. How- ever, we have found no analysis on the seasonal pattern of copulations in multiple-brooded species. The coincidence of the third peak in copulation rates with the dates of replacement or double clutches may be an adaptation related to the ten- dency of the Black-winged Kite to initiate a second breeding attempt, which they may abort just before laying if food conditions are not adequate. Al- though data on female’s fertility in this period are lacking, it may be argued that these copulations occur in a fertilization context and sperm compe- tition might play a role. The unusual early laying dates that we recorded in some years (i.e., November or December, Fer- rero and De Lope 2001) seem to indicate that in- dependently of seasonal photoperiodic and climat- ic variation of the temperate zones. Black-winged Kites may try to breed if other factors (most prob- ably food abundance) are favorable. Bustamante (1993), in a study carried out in the same area, found that two radiotagged adult fe- males had left the nest well before the nestlings became independant, whereas the male fed the 6 Ferrero et al. VoL. 37, No. 1 young until they dispersed from the territory. In three other pairs (with unmarked birds) this au- thor was unable to conclude if both adult birds remained in the territory or not. We have also found some evidence of only one adult remaining until the young fledge, but this does not exclude the possibility that other pairs remain together and initiate a second breeding attempt. In fact, we did not find any evidence of mate switching (temporal disappearance of one of the pair members) in the six renesting attempts, and at least in one of the second breeding attempts both members of the pair fed the young of the first brood while copu- lating and constructing the second nest in a neigh- boring tree. Unfortunately, as our birds were not marked, we cannot conclude if renesting or second breeding attempts (and therefore the third copu- lation peak) are performed by the same mates, or as occurs in South Africa (and in Australia by the Letter-winged Kite, Elanus scriptus), the female in- volved in the male’s second attempt is a different individual (Mendelsohn 1983, Marchant and Hig- gins 1993). Acknowledgments The late F. Carbajo initiated this study and participated in early observations. We are indebted to P. Gomez, M. Cortes, M. Cabrero and J.A. Roman for their assistance in different periods of fieldwork. V. Pizarro, L. Lozano, and M.G. Calzado shared with us their wide knowledge on kites’ behavior. We thank J. Seoane for the amount of precious time he wasted with us during data analysis. F. de L.ope, J. Aviles, B. Arroyo, D. Parejo, T. Birkhead, and an anonymous reviewer improved an early version of the manuscript with their comments. Fieldwork was carried out with no institutional financial support. Literature Cited Balgooyen, T.G. 1976. Behavior and ecology of the American Kestrel {Falco sparveriush.) in the Sierra Ne- vada of California. Univ. Calif. Publ. Zool. 10.S;l-83. Berkeley, CA U.S.A, Bertran, J. AND A. Margalida. 1999. Copulatory behav- ior of the Bearded Vulture. Condor 101:164—168. Bird, D.M. and R.B. Buckland. 1976. The onset and du- ration of fertility in the American Kestrel. Can. ]. Zool. 54:1595-1597. Birkhead, T.R. 1979. Mate guarding in the Magpie Pica pica. Anim. Behav. 27:866—874. . 1988. Behavioral aspects of sperm competition in birds. Adv. Stud. Behav. 18:35—72. and C.M. Lessells. 1988. Copulation behavior of the Osprey Pandion haliaetus. Anim. Behav. 36:1672- 1682. and A.P. M0I.LER. 1992. Sperm competition in birds. Academic Press, London, U.K. and . 1993. Why do male birds stop copu- lating while their partners are still fertile? Anim. Behav. 45:105-118. , L. Atkin, and A.P. M0LLER. 1987. Copulation be- havior of birds. Behavior 101:101-138. Brown, L.H., E.K. Urban, and K. Newton. 1982. The birds of Africa. Vol. 1. Academic Press, London, U.K. Bu.STAMANTE, J. 1993. The post-fledging dependence pe- riod of the Black-shouldered Kite {Elanus caeruleus). J. EtaptorRes. 24:185-190. Cramr, S. and K.E.L. Simmons. 1980. Handbook of the birds of Europe, the Middle East and North Africa. Vol. 2. Oxford Univ. Press, Oxford, U.K. DEL Hoyo, J., A. Elliott, and J. Sargaeal. 1994. Hand- book of the birds of the world. Vol. IT Lynx Edicions, Barcelona, Spain. Donazar, J.A., O. CEBAI.LOS, AND J.l.. Tella. 1994. Cop- ulation behavior in the Egyptian Vulture. Bird Study 41:37-41. Ellis, D.H. and L. Powers. 1982. Mating behavior in the Golden Eagle in non-fertilization contexts. Raptor Res. 16:134-136. Fargallo, J.A., G. Blanco, and E. Soto-I.argo. 1996. Possible second clutches in a Mediterranean montane population of the Eurasian Kestrel {Ealco tinnunculus) . J. Raptor Res. 30:70-73. Ferrero, JJ- 1994. Situacion del Elanio Azul Elanus caeru- leus en el Mediterraneo, Pages 101-1 15 mj, Muntaner and J. Mayol (Eds.), Biologia y Conservacion de las rapaces mediterraneas. SEO/BirdLife Monografias No. 4. Pozuelo de Alarcon, Madrid, Spain. and F. De Lope. 2001. Breeding phenology of Black-shouldered Kites {Elanys caeruleus) in Fixtremad- ura (SW Spain). Page 65 in M. Ferrer, J.J. Negro, and J. Bustamante (Eds.). 4th Eurasian Congress on Rap- tors. Book of Abstracts. Sevilla, Spain. AND A. Onrubia. 1998. Expansion del area de cria y distribucion actual del Elanio Azul Elanus caeruleus en Espana. Pages 159-171 in B.-U. Meyburg, R.D. Chancellor, and J.J. Ferrero [Eds.], Holarctic birds of prey. WWGBP & ADENEX, Merida, Spain. Hoia iluijZEN, A.M.A. 1992. Frequency and timing of cop- ulations in the Prairie Falcon. Wilson Bull. 104:333- 3.38. Korpimaki, E., K. Lahti, C.A. May, D.T. Parkin, G.B. Powell, P. Toi.onen, and J.H. Wetton. 1996. Copu- latory behavior and paternity determined by DNA fin- gerprinting in kestrels: effects of cyclic food abun- dance. Anim. Behav. 51:945-955. Marchant, S. and P.J. Higgins. 1993. Handbook of Aus- tralian, New Zealand, and Antarctic birds. Vol. 2. Ox- ford Univ. Press. Melbourne, Australia. Martinez, R, G. Bianco, and R. Rodriguez. 1998. Rate, timing, and success of clutch replacement by colonial Griffon Vultures {Cryps pilvus) . Ornis [Ann. 75:145-148. Mendelsohn, J.M. 1983. Social behavior and dispersion of the Black-shouldered Kite. Ostrich 54:1-18. March 2003 Copulation in Black-winged Kite 7 M0LLER, A.P. 1985. Mixed reproductive strategy and mate guarding in a semi-colonial passerine, the swallow Hi- rundo rustica. Behav. Ecol. Sociobiol. 17:401-408. . 1987. Copulation behavior in the Goshawk, Ac- cipiter gentilis. Anim.. Behav. 35:755—763. AND T.R. Birkhead. 1992. A pairwise comparative method as illustrated by copulation frequency in birds. Am. Nat. 139:644-656. Mougeot, F. 2000. Territorial intrusions and copulation patterns in Red Kites, Milvus milvus, in relation to breeding density. Anim. Behav. 59:633-642. Mundy, R, D. Butchart, J. Ledger, and S. Piper. 1992. The vultures of Africa. Acorn & Russel Friedman Books. Johannesburg, South Africa. Negro, JJ- and J.M. Grande. 2001. Territorial signaling: a new hypothesis to explain frequent copulation in raptorial birds. Anim. Behav. 62:803-809. r J-A. Donazar, and F. Hiraldo. 1992. Copulatory behavior in a colony of Lesser Kestrels: sperm com- petition and mixed reproductive strategies. Anim. Be- hav. 43:921-930. , M.R. ViLLARROEL, J.L. Telia, U. Kuhnlein, J.A. Donazar, and D.M. Bird. 1996. DNA fingerprinting reveals a low incidence of extra-pair fertilizations and intraspecific brood parasitism in the Lesser Kestrel. Anim. Behav. 51:935-943. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A. . 1986. The Sparrowhawk. T. &: A.D. Poyser, Cal- ton, UK. Pandolfi, M., R. Pagi-iarani, and G. Olivetti. 1998. In- tra- and extra-pair copulations and female refusal of mating in Montagu’s Harriers. J. Raptor Res. 32:269- 277. Petrie, M. and B. Kempenaers. 1998. Extra-pair paternity in birds: explaining variation between species and populations. Trends Ecol. Evol. 13:52-58. Rosenfield, R.N., j. Bielefeldt, andJ. Cary. 1991. Cop- ulatory and other pre-incubation behaviors of Coo- per’s Hawks. Wilson Bull. 103:656-660. Rufino, R. 1994. Black-shouldered Kite Elanus caeruleus Pages 146-147 in G.M. Tucker and M.F. Heath [Eds ], Birds in Europe, their conservation status. BirdLife International, Cambridge, U.K. Sanchez, A. 1990. Noticiario ornitologico. Arrfco/a 37:335. SODHI, N.S. 1991. Pair copulations, extra-pair copula- tions, and intraspecific nest intrusions in Merlin. Con- dor 93:433-437. Toland, B.R. 1985. Double brooding by American Kes- trel in central Missouri. Condor 87:434— 436. Tortosa, F.S. and D.T. Redondo. 1992. Frequent copu- lations despite low sperm competition in White Storks {Ciconia ciconia) . Behavior 121:^2,87— 3\b. Van DER Merwe, F. 1980. Nest building in the Black- shouldered Kite. Ostrich 6\:\\3—\\4. ViLLARROEL, M., D.M. BIRD, AND U. KUHNLEIN. 1998. Cop- ulatory behavior and paternity in the American Kes- trel: the adaptive significance of frequent copulations. Anim. Behav. 56:289-299. Receivd 25 March 2002; accepted 5 December 2002. / Raptor Res. 37(1):8-18 © 2003 The Raptor Research Foundation, Inc. MORTALITY, MORBIDITY, AND LEAD POISONING OF EAGLES IN WESTERN CANADA, 1986-98 Mark Wayland^ Environment Canada, P&NR, Prairie and Northern Wildlife Research Centre, Saskatoon, SK S7N 0X4 Canada Laurie K. Wilson and John E. Elliott Canadian Wildlife Service, P&YR, Pacific Wildlife Research Centre, Delta, BC V4K 3N2 Canada Michael J.R. Miller Jolaire Ecological Consulting, 210-112^^ St., Saskatoon, SK S7N 1V2 Canada Trent Bollinger Canadian Cooperative Wildlife Health Centre, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, SK S7N 5B4 Canada Malcolm McAdie British Columbia Ministry of the Environment, 2080A Lahieux Rd., Nanaimo, BC V9T 6J9 Canada Ken Langelier Island Veterinary Hospital, 1800 Bowen Rd., Nanaimo, BC V9S 3W4 Canada Jonathan Keating and Jennifer M.W. Froese Environment Canada, P&NR, Prairie and Northern Wildlife Research Centre, Saskatoon, SK S7N 0X4 Canada Abstract. — Between 1986 and 1998, we investigated causes of mortality and morbidity, and assessed lead (Pb) contamination in Bald {Haliaeetus leucocephalus) and Golden eagles [Aquila chrysaetos) in western Canada. The most common diagnoses were trauma (13.0%), electrocution (11.5%), pes- ticide poisoning (7.6%), gunshot wounds (7.3%), and Pb poisoning (6.4%). It was not possible to determine the cause of mortality or morbidity in 29% of the 546 eagles. Excluding undetermined cases and those with an unknown collection date, toxicoses accounted for 40% of the diagnoses in the prairie provinces, but only 19% of those in British Columbia {P < 0.001). Ten percent of eagles {N = 372) had tissue Pb levels consistent with Pb poisoning and 5% had elevated, but non-lethal Pb levels. Age class and species differed significantly among eagles with either background or high (Pb-poisoned and elevated Pb) Pb levels {P ^ 0.01). When data from both species were combined, high Pb levels were detected in 19. .5% of adult and subadult eagles {N = 220) but only 7% of immature eagles {N = 128). Twenty-eight percent of Golden Eagles {N = 39) but only 13% of Bald Eagles {N = 333) had high Pb levels. Proportions of eagles with high and background Pb levels were not different based on the level of waterfowl hunting near the locations of collection {P > 0.48). Golden Eagles with high tissue Pb concentrations tended to be found more often during or soon after the fall hunting season than eagles with background Pb concentrations {P ~ 0.08). The greater incidence of high Pb in Golden Eagles compared to Bald Eagles and the lack of relationship with waterfowl hunting intensity suggest that waterfowl is not the primary source of Pb for eagles in western Canada. Key Words: Bald Eagle, Haliaeetus leucocephalus; Golden Eagle, Aquila chrysaetos; lead] lead poisoning, pesticide poisoning, toxicity. ' E-mail address: mark.wayland@ec.gc.ca 8 March 2003 Eagle Mortauty in Western Canada 9 MORTALIDAD, MORBILIDAD, E INTOXICACION FOR PLOMO DE AgUILAS EN EL OESTE DE CANADA, 1986-98 Resumen. — Entre 1986 y 1998, investigamos las causas de mortalidad y morbilidad, y evaluamos la con- tarainacion por plomo (Pb) en aguilas calvas (Haliaeetus leucocephalus) y aguilas doradas {Aquila chrysaetos) en el oeste de Canada. El diagnostico mas comiin fue trauma (13.0%), electrocucion (11.5%), enve- nenamiento por pesticidas (7.6%), heridas por arma de fuego (7.3%), y envenenamiento con plomo Pb (6.4%). No fue posible determinar la causa de mortalidad o morbilidad en 29% de las 546 aguilas. Excluyendo los casos indeterminados, la toxicosis dio cuenta de 40% de los diagnosticos en las provincias de la pradera, pero unicamente en 19% de los misraos para British Columbia (P < 0.001). Diez por ciento de las aguilas (N = 372) tuvo niveles de Pb en los tejidos consistentes con el envenenamiento por Pb y 5% tuvieron niveles elevados pero no letales de plomo. Las clases de edad y las especies difirieron significativamente tanto entre aguilas con antecedentes como en aquellas con altos niveles de plomo (envenenadas o con Pb elevado) (P^ 0.01). Cuando los datos de ambas especies se combinaron, los altos niveles de plomo se detectaron en 19.5% de las aguilas adultas y subadultas (N — 220) y unicamente en 7% de las aguilas inmaduras (N = 128). Veintiocho por ciento de las aguilas doradas (N = 39) y unicamente 13% de las aguilas calvas {N = 333) tuvieron altos niveles de plomo. Las proporciones de aguilas con antecedentes y altos niveles de plomo no difirieron con base en el nivel de aves acuaticas cazadas cerca de las localidades de coleccion (P s 0.48). Las %uilas doradas con alta concentracion de plomo en los tejidos se hallaron mas frecuentemente durante o un poco despues de la estacion de caza de otoho que las aguilas con antecedentes de concentraciones de plomo (P= 0.08). La mayor incidencia de altos niveles de plomo en aguilas doradas que en aguilas calvas y la carencia de interrelaciones con la intensidad de caza de aves acuaticas, sugiere que las aves acuaticas no son la fuente primaria de plomo para las aguilas del occidente de Canada. [Traduccion de Cesar Marquez] Lead (Pb) poisoning in waterfowl has long been recognized as an environmental problem in North America (Bellrose 1959). More recently, it has been documented in other avian species (Locke and Friend 1992). Since the 1960s, Pb poisoning has accounted for ca. 1-15% of recorded mortality in Bald Eagles {Haliaeetus leucocephalus) and other raptors (Mulhern et al. 1970, Pattee and Hennes 1983, U.S. Fish and Wildlife Service 1986, Pain and Amiard-Triquet 1993, Wayland and Bollinger 1999). The primary source of Pb in poisoned rap- tors is assumed to be shot or bullets used by hunt- ers (Pattee and Hennes 1983, Scheuhammer and Norris 1995, Kendall et al. 1996). Raptors are ex- posed to Pb by consuming prey containing Pb shot or bullet fragments (Pattee and Hennes 1983, Pat- tee et al. 1990, Pain et al. 1993, Harmata and Res- tani 1995, Pain et al. 1997, Mateo et al. 1999). While there is a strong association between the ingestion of Pb ammunition from prey tissues and Pb poisoning in raptors, the importance of Pb shot in waterfowl as the main contributing factor is less certain. Pattee and Hennes (1983) hypothesized that Pb shot in waterfowl was the major source of Pb exposure in Bald Eagles; however, Pb ammu- nition embedded in the tissues of other prey ani- mals may also be important (Kramer and Redig 1997). In the case of Golden Eagles {Aquila chry- saetos), upland game animals are believed to be a more important source of Pb shot than are water- fowl (Pattee et al. 1990). It is important to distin- guish between waterfowl and other prey as the main source of Pb exposure, because Pb ammu- nition has been banned in Canada since 1999 for hunting waterfowl, but not for hunting other types of game. We examined causes of death, illness, and injury in Bald and Golden eagles from western Canada to evaluate Pb poisoning as a cause of mortality and morbidity. In addition, we assessed whether sublethal concentrations of Pb were associated with toxicosis, physical injury, or disease, as sublethal exposure to Pb can predispose birds to other caus- es of death (Burger 1995). We compared the prev- alence of high Pb levels among different age clas- ses of eagles because population levels of some raptors are linked to the survival of adult, repro- ducing birds (Grier 1980, Newton 1988). To assess whether Pb ammunition is the probable, primary source of Pb exposure in eagles in western Canada, we determined whether the prevalence of elevated Pb levels was highest during and soon after the fall hunting season. Moreover, we examined elevated Pb levels in the context of waterfowl hunting in- tensity and interspecific dietary differences to test the hypothesis that consumption of waterfowl is a major source of Pb exposure in eagles. 10 Wayiand et al. VoL. 37, No. 1 Methods Sample Collection. From 1986-98, provincial wildlife agencies and raptor rehabilitation organizations in Man- itoba, Saskatchewan, Alberta, and British Columbia (BC) received dead, sick, and injured Bald and Golden eagles. The sex of birds in rehabilitation centers was determined based on morphometric measurements (Bortolotti 1984a, 1984b), and all dead birds were sexed during nec- ropsy. Plumage characteristics were used to estimate age (Bortolotti 1984b, McCollough 1989), and eagles were designated as immature, subadult, or adult. Birds in re- habilitation centers were diagnosed by the staff of these facilities, not necessarily a veterinarian. Dead eagles were frozen until necropsies could be performed by veterinary pathologists. Blood samples were drawn from live eagles in rehabil- itation centers, placed in vials prewashed with nitric acid, and frozen at — 20°C until subsequent Pb analysis. In ad- dition, plasma samples from 96 birds in rehabilitation centers in BC were analyzed for cholinesterase activity. Kidneys and livers were removed during necropsy and refrozen at — 20°C in glassware prewashed with nitric acid or acetone/hexane. These tissue samples were sent to the National Wildlife Research Centre in Hull, Quebec, for Pb analysis. Brain tissue was collected and frozen for analysis of cholinesterase activity when organophosphate (OP) or carbamate pesticide poisoning was suspected based on evidence found in the field or at the time of necropsy. Such evidence included the presence of insecticide-laced bait or other dead animals at the field site, and copious quantities of meat in the gastrointestinal tract of an eagle carcass in good condition. Diagnoses of pesticide poison- ing were made based on this evidence, residue analysis, and cholinesterase activity <60% of normal levels for that species (Greig-Smith 1991). Stomach and crop contents of 23 eagles that were suspected to have been poisoned were analyzed for residues of OP and carbamate pesti- cides known to be used widely in the study area (Elliott et al. 1996, 1997, T. Bollinger unpubl. data). Brain cho- linesterase activity was determined in 19% of eagles col- lected from BG and in 22% of those from the prairies, using methods similar to those described by Martin et al. (1981) and Hill and Fleming (1982). Lead Analysis. Blood samples (100 pi) were pipetted into 1500 pi Eppendorf micro-centrifuge tubes contain- ing 400 pi of a 0.2% NH 4 H 2 PO 4 plus 0.5% Triton X-100 solution. Each tube was capped and shaken for 10 sec. Pb was determined by graphite furnace atomic absorp- tion spectrometry (GFAAS) using a Perkin-Elmer 3030b spectrophotometer equipped with a HGA-300 graphite furnace and an AS-40 autosampler, according to methods described by Fernandez and Hilligoss (1981). Blood Pb values are expressed as pg Pb/ml blood. Blank solutions were analyzed to verify that Pb was below detection limits. Sample detection limits ranged from 0.005-0.025 pg/ml. Recoveries of spiked samples ranged from 87-110% and averaged 97%. Coefficients of variation of duplicate and triplicate analyses ranged from 7-11% and averaged 9%. Aliquots of liver and kidney samples were weighed to determine wet weights; placed in plastic, nitric acid- washed test tubes; freeze-dried; and their dry weights re- corded. Samples were digested in a solution of deionized water and 70% nitric acid, at a concentration of 0.5 ml solution per 0.1 g dry weight of sample. Pb levels in liver and kidney tissues were expressed on a dry weight basis. Pb was analyzed by flame-atomic absorption spectrometry (AAS) using an atom concentrator tube (ACT-80). The detection limit ranged from 1 . 0 - 2.0 pg/g dry weight. Samples in which Pb was not detected using flame-AAS were analyzed by GFAAS as described above. Detection limits using the GFAAS ranged from 0. 3-1.0 pg/g dry weight. Recovery of Pb from spiked samples and standard reference materials ranged from 68-118% and averaged 93%. Coefficients of variation ranged from 0-19% and averaged 6 %. Pb concentrations greater than 1.0 pg/ml blood, 30 pg/ g dry weight liver, or 20 pg/g dry weight kidney were considered to be diagnostic of Pb poisoning, while con- centrations greater than 0.2 pg/ml blood or 6 pg/ g dry weight liver or kidney indicated elevated Pb exposure (Pattee et al. 1981, Redig et al. 1984, Reichel et al. 1984, Pain et al. 1994, Franson 1996). When liver and kidney Pb levels were reported in the literature on a wet-weight basis, they were converted to dry weight values using mean tissue moisture values determined in this study to be 76.5% for eagle kidney and 71% for eagle liver. For certain analyses, the Pb-poisoned and elevated Pb cate- gories were combined into a single high Pb group. Pb levels lower than those values stated above were consid- ered to be due to background exposure. Data Analysis. Causes of death or morbidity were clas- sified into three categories: ( 1 ) physical injury, including drowning, electrocution, collision, gunshot, and other trauma; ( 2 ) disease and debilitation, including avian cholera, other diseases, emaciation, and degeneration; (3) toxicosis, including pesticide and suspected pesticide poisoning, Pb poisoning, and other poisonings (mercury and strychnine). Two-way contingency table analysis was used to test the relationship between cause of death or morbidity and species. Data from both species were com- bined in subsequent analyses because no difference was detected among proportions of Bald and Golden eagles within the three diagnostic categories (x^ = 0.7, P= 0.70, N = 388) . This was followed by log-linear analysis using the maximum-likelihood technique (PROG CATMOD, SAS Institute 1988) to examine relationships among cause of death or morbidity, region (prairie provinces and BC), and time period. The two time periods exam- ined were May-October, which represented the breeding season, and November-April, which represented autumn migration, wintering, and spring migration. The latter time period included months during and soon after the fall hunting season, which occurs from September-De- cember. The log-linear model included all main effects as well as two-way interactions that included cause of death or morbidity. Separate analysis of the relationship among species and cause of death or morbidity was re- quired because there were too few Golden Eagles {N = 39) to include in the multi-way log linear table. We used analysis of variance of rank-transformed kid- ney Pb level data to evaluate if Pb exposure at back- ground and sublethal levels predisposed eagles to partic- ular causes of death or morbidity. We used logistic regression with maximum likelihood estimation (PROG CATMOD) to determine if Pb levels were related to age Marc:h 2003 Eagle Mortaljty in Wesiern Canada 11 class, date of recovery, waterfowl hunting intensity, or species. Only main effects were included in the model. In addition, we used two-sample Wilcoxon tests to ex- amine relationships between Pb levels and time of year of recovery, relative to the fall hunting season (Septcm- ber-December). Throughout this study, Pvalues less than 0.05 were considered significant. We examined the geographic association among Pb levels in eagles and waterfowl hunting intensity. To do so, we used harvest survey data (Canadian Wildlife Service unpubl. data) from 1990-95 to estimate mean number of waterfowl-hunting-days/year. Waterfowl-hunting-days data were provided as point estimates that represented 30-min by 30-min blocks of land. Geostatistics, a point surface interpolation technique, was used to derive esti- mates of waterfowl hunting activity across the study area (GS+® version 3.11.7, Gamma Design 1999). Geostatis- tics is an acceptable method for estimating data values for locations that were not sampled directly, by examin- ing data from locations that were sampled. A model of spatial correlation was established and used to interpo- late data values at the unknown locations. The first step of geostatistics is to calculate the sample semivariance and use that to estimate the shape of a curve that represents the semivariance as a function of distance. The second step is to use the estimated semi- variance function to determine the weights needed to define the contribution of each sampled point to the in- terpolation. Sample points close to the point for which an estimated value is to be generated contribute the most to the interpolation. The waterfowl-hunting-days data fit an exponential model {F^ = 0.98 and residual sums of squares = 0.052), indicating that the variability in hunt- ing activity among locations was a function of distance. Thus, spatial interpolation of the data was appropriate, and waterfowl-hunting-days were interpolated at a reso- lution of 10 km X 10 km across the study area. Interpo- lation of the waterfowl-hunting-days data for a specific location was done using nearest-neighbor values weight- ed by distance and the degree of autocorrelation present for that distance. A maximum of 16 nearest neighbors lying within a 30-km radius of the specific location was used in the interpolations. IjOW intensity waterfowl hunt- ing areas were designated as those with fewer than 1000 hunting-days/year, and accounted for 87% of the 100 km^ hunting areas. Hunting activity on high intensity wa- terfowl hunting areas ranged from 1000-11 178 hunting- days per year. Eagles with high Pb levels that had been shot {N = 5) were excluded from statistical analyses of Pb contamina- tion because the elevated Pb levels in their tissues may have resulted from embedded Pb shot fragments rather than metabolic uptake of Pb (Wayland et al. 1999). Sam- ples from eagles that had been in rehabilitation centers for >3 wk were not used in analyses because Pb concen- trations normally decline to background levels within that time period (Reiser and Temple 1981). Similarly, ea- gles that received chelating agents to bind Pb were not used in analyses. Results Causes of Death, Injury, or Illness. Cause of death, injury, or illness could not be determined Table 1. Final diagnoses for 546 dead and debilitated Bald and Golden eagles in western Canada, 1986-98 Bruism COI.UMBIA Prairie Provinc:es Diagnosis Bald Eagle Golden Eagle Bai.d Golden Eagle Eagle Total Degeneration or emaciation 21 3 5 3 32 Avian cholera 0 0 6 0 6 Other infectious disease 14 0 8 2 24 Pesticide poisoning 20 0 19 3 42 Lead poisoning 19 1 11 4 35 Other poisoning 2 0 3 0 5 Suspected poisoning'^ 4 0 12 2 18 Collision 34 0 1 2 37 Drowning 15 0 0 0 15 Electrocution 50 0 4 9 63 Gunshot 25 0 14 1 40 Other trauma*^ 33 1 29 8 71 Undetermined 123 4 22 9 158 ^ Poisoning suspected because of additional evidence. Category includes wing injuries, attacks by other eagle.s, leg in- juries, trap-related injuries, unknown origin, soaked feathers, in- ternal injuries, asphyxiation, spinal injuries, fall from nest, and exhaustion. for 158 of 546 eagles that were examined. Thirteen percent of eagles were diagnosed with trauma oth- er than electrocution, drowning, collisions, or gun- shot wounds; 11.5% were electrocuted; 7.6% were pesticide-poisoned; 7.3% were shot; 6.4% were Pb- poisoned, and 25% died or were debilitated by oth- er causes (Table 1 ) . Eagles poisoned by OP or car- bamate insecticides had mean ±SD brain cholinesterase activity of 8.4 ± 6.5 |jLmol/min/g {N = 22). The corresponding values for eagles diag- nosed with other or undetermined causes of death (A^ = 7) were 17.5 ± 2.8 pmol/min/g. Excluding the cases with undetermined cause of death or date of collection, proportional differences in caus- es of death or morbidity of eagles {N = 370) were related to region where they were found {P < 0.001; Fig. 1), but not to time of year {P — 0.09). Poisonings and suspected poisonings accounted for 40% of diagnoses of eagles from the prairie provinces, but only 19% of those from BC. In con- trast, 45% of eagles from the prairies and 66% of those from BC were injured. Lead Exposure and Poisoning. Pb levels were de- 12 Wayland et al. VoL. 37, No. 1 BC Prairies Location Figure 1. Percent frequency of three categories of di- agnosis of death and morbidity, according to the region where eagles were found. Unknown causes of death and morbidity have been excluded. Utilization of a log-linear model showed that cause of death or morbidity was in- fluenced by region. Numbers in parentheses indicate sample sizes. termined in kidney, liver, or blood samples from 372 eagles (Table 2). Most eagles (85%) had back- ground levels of Pb and 15% had high Pb levels in their tissues. The high Pb group included eagles with Pb levels consistent with poisoning (10%) and those with elevated Pb levels (5%). Bald Eagles with toxicoses other than Pb poison- ing had higher Pb levels than those diagnosed with various physical injuries, disease, or debilitation {P — 0.037). For Bald Eagles, median values for kid- ney lead levels in three categories of diagnosis were as follows: toxicoses other than lead poisoning — 1.09 ftg/g (N = 29); physical injuries — 0.27 |xg/g {N = 139); and disease and debilitation — 0.26 pg/ g {N = 37) . There was no difference in kidney Pb levels among groups of Golden Eagles (P = 0.614) . Median values for kidney Pb levels in Golden Ea- gles with toxicoses other than Pb poisoning {N = 4), physical injuries {N = 16), and diseases or de- bilitation {N = 6), were 0.77, 0.74, and 1.27 pg/g, respectively. Age class and species differentiated between ea- gles with background or high Pb levels (P^ 0.01). Adult and subadult eagles had a higher percentage with high Pb levels than did immature eagles. This difference was evident for Bald Eagles (P = 0.005) and when data from both species were combined (P = 0.002) . High Pb levels were detected in 17.6% of adult and subadult Bald Eagles, but in only 5.9% of immature Bald Eagles. When data from both species were combined, 19.5% of adult and sub- adult eagles, and 7.0% of immature eagles had high Pb levels. In Golden Eagles, the percent of adult and subadult birds with high Pb (36.3%) was not different from the percent of immature eagles with high Pb (20.0%, P = 0.58). In comparison with Bald Eagles, a higher percentage of Golden Eagles had high Pb levels. The percent of Bald Ea- gles with high Pb levels in the immature and sub- adult/adult age classes were 5.9% and 17.6%, re- spectively. The percent of Golden Eagles with high Pb levels in these age classes were 20.0% and 36.3%, respectively. Date of recovery and waterfowl hunting intensity failed to provide additional discrimination be- tween the background and high Pb groups (P s 0.19). However, Golden Eagles with high Pb levels tended to be found during or soon after the fall hunting season (Fig. 2), while those with back- ground Pb levels were more often found several months later. This seasonal difference in date of recovery approached significance (Wilcoxon two- sample test, P = 0.08). Date of recovery of Bald Eagles, as related to hunting season, did not differ Table 2. Pb levels in tissues from 372 immature, subadult, adult, and unknown age Bald and Golden eagles collected from western Canada, 1986—98. Bald Eagle Golden Eagle Pb Level Immature Subadult and Adult Unknown Immature Subadult and Adult Unknown Total Background^ 111 163 15 8 14 6 317 Elevated’’ 1 11 1 1 4 0 18 Poisoned*^ 6 24 1 1 4 1 37 "" Pb <6 |JLg/g dry weight kidney or liver; <0.2 (xg/ml blood. Pb 6—20 jJLg/g dry weight kidney; 6-30 |JLg/g dry weight liver; 0.2— 1.0 (ig/ml blood. Pb >20 (ig/g dry weight kidney; >30 ji-g/g dry weight liver; >1.0 \Lg/ml blood. March 2003 Eagle Mortality in Western Canada 13 Nov - Jan Feb - Apr May - Jul Aug - Oct Month Figure 2. Temporal changes in the percent of Bald and Golden eagles with high Pb levels, in relation to the fall hunting season (September-December) . High Pb levels were defined as >0.2 pg/ml blood or >6 p-g/g dry weight liver or kidney tissue. Numbers in parentheses in- dicate sample sizes. between those with background and high Pb levels {P — 0.8). Waterfowl hunting intensity did not af- fect the percent of Bald or Golden eagles with high Pb levels {P ^ 0.48). In areas of low intensity wa- terfowl hunting, 12.8% of Bald Eagles and 32.0% of Golden Eagles had high Pb levels. High Pb levels were evident in 14.1% of Bald Eagles and 21.4% of Golden Eagles recovered from high intensity wa- terfowl hunting areas. Median and quartile Pb levels in kidneys of five eagles with Pb shot in their gastrointestinal tracts were 23.1 p-g/g and 20.8-76.8 p-g/g, respectively, and were higher than in 329 eagles without shot in their gastrointestinal tracts (0.4 pg/g, 0.08-1.4 pg/g, Wilcoxon two-sample test: P < 0.001). Discussion Causes of Death, Injury, or Illness. A sampling bias existed in this study because of the way in which eagles were found and reported. Thus, rel- ative importance of each cause of death or mor- bidity may not be representative of the actual pro- portion at the population level. Birds affected by starvation and disease may be reclusive and suscep- tible to depredation (Wobeser 1994, 1997), thus, decreasing the probability of encounter by hu- mans. Many eagles during this study were found in southwestern BC near major population centers, but few were obtained from northern breeding ar- eas where human populations are sparse. Prevalence of Pb poisoning, electrocution, trau- Table 3. Causes of death or morbidity (percent of all cases) reported in eagles in the United States and in west- ern Canada. Diagnosis U.S.A. 1963- 84^ U.S.A. 1960s- 90s^ Western U.S.A. Canada 1978- 1986- 8U 98^> Trauma 21 23^ 20 22 Pb poisoning 6 27f 8 6 6 OP and carbamate 0.5 ~3 0 8 poisonings Other 5 IJS 1 1 poisonings Electrocution 9 12e 15 11 Gunshot 23 25f 15 19 7.6 Trapping 5 U 6 1 Emaciation 8 u 11 6 Disease 2 u 6 5 5 Undetermined 18 u 6 29 ^ N = 1429 Bald Eagles (National Wildlife Health Laboratory 1985). bjv ~ 4300 Bald and Golden eagles (Franson et al. 1995). ^ N = 29,3 Bald Eagles (Reichel et al. 1984). N = 546 Bald and Golden eagles (this study). Bald Eagles. * Golden Eagles. 8 Unknown. ma, emaciation, and disease as reported in this study in western Canada were similar to those pre- viously reported for the United States (Table 3). The present study found that Pb poisoning ac- counted for 6% of diagnoses, a value that is in agreement with previous reports. Interestingly, we found that carbamate and OP insecticide poison- ings in western Canada accounted for a much higher percent of cases than had been reported in the United States. Gunshot and trapping diagnoses accounted for a smaller percent of cases in western Canada than in the United States. Overall, the pro- portion of human-induced mortality and morbidity in this study (44% of all cases and 62% of cases with known diagnoses) was similar to that in other studies (Reichel et al. 1984, National Wildlife Health Laboratory 1985, Franson et al. 1995, Har- mata et al. 1999). Results suggest humans are di- rectly responsible for a large proportion of eagle deaths, but it is uncertain whether this is impacting eagle populations. There is no evidence that pop- ulations of either Bald or Golden eagles had de- 14 Wayland et al. VoL. 37, No. 1 dined in western Canada through to the early 1980s (Gerrard 1983, Kirk 1996). Although no re- cent data are available, remedial measures to re- duce the numbers of eagles dying from human- related causes may not be necessary for the conservation of eagle populations. In this study, eagles from the prairie provinces were poisoned by insecticides more often than those from BC. This regional variation was likely due to differences in the manner in which carba- mate and OP insecticides were used. Insecticide poisoning in BC likely resulted from the appropri- ate use of granular carbamate and OP insecticides (Elliott et al. 1996, 1997). Furthermore, insecticide poisoning in BC may have been reduced in the latter years of this study by successful efforts that were undertaken to remove the most toxic prod- ucts from the market or to at least reduce their use. In contrast, insecticide poisonings in the prai- rie provinces have been attributed to the illegal use of carbamate and OP insecticides in baits intended to kill coyotes ( Canis latrans; Bollinger and Mineau 1995). Twelve percent (22/177) of the eagles col- lected from the prairie provinces were diagnosed with pesticide poisoning. Assuming that these cases all resulted from efforts to bait and kill coyotes, the effect of this activity is greater in the prairie prov- inces than has been documented for North Amer- ica as a whole (5%, Bortolotti 1984c). The use of poisons to kill predators of livestock continues to be a problem for eagles in the prairie provinces. Lead Exposure and Poisoning. Although 6% of the eagles in this study were initially diagnosed as Pb-poisoned, subsequent tissue analysis showed that 10% of 372 eagles had concentrations indica- tive of Pb poisoning. The principal reason for this difference is that Pb poisoning cannot be properly determined without analyzing tissues for Pb. Thir- teen percent of the eagles from the prairie prov- inces, for which post-mortem examination did not identify a cause of death, had tissue Pb levels in- dicative of Pb poisoning. Cause of death or mor- bidity was undetermined for 158 eagles in this study, and not all birds were analyzed for Pb. Thus, diagnosis without tissue analysis may have under- represented Pb poisoning as a mortality factor. The 10% estimate based on only those eagles for which Pb levels were determined, may be more accurate. At concentrations below those known to cause mortality, Pb is immunotoxic and neurotoxic, can cause behavioral deheits, and impair digestion and feeding (Burger 1995). Thus, elevated concentra- tions of Pb in birds may increase their susceptibility to diseases and accidents, and impair their ability to hunt, obtain food, and digest food. We did not hnd evidence to support an association between elevated Pb levels and the prevalence of disease, emaciation, or physical injury in eagles. Similarly, the known causes of raptor mortality in Great Brit- ain were generally unrelated to elevated Pb expo- sure (Pain et al. 1994). Bald Eagles that were poisoned by other toxic agents, principally OP and carbamate insecticides, had higher kidney Pb concentrations than dis- eased, emaciated, or injured birds. Scheuhammer and Wilson (1990) reported that various cholines- terase-inhibiting insecticides in combination with Pb did not have an additive effect on d-aminole- vulinic acid dehydratase (ALA-d) inhibition when compared with the effects of Pb alone. The ALA-d enzyme is sensitive to Pb and serves as an excellent biomarker of Pb exposure (Scheuhammer and Wil- son 1990). In this study, many insecticide-poisoned Bald Eagles were found in areas where Pb expo- sure was also comparatively high. The association between tissue Pb levels and the prevalence of poi- sonings by other toxic agents probably reflects a greater possibility that birds from these areas will be exposed through their diets to both Pb and in- secticides. We doubt that sublethal Pb exposure in- creases the susceptibility of eagles to insecticides. In this study, high Pb levels were found in a greater proportion of adult and subadult eagles than immature eagles. In the United States, adult females comprised 47% of all Pb-poisoned Bald Ea- gles, but only 25% of the continental population, indicating a particular sensitivity of this age-sex group (U.S. Fish and Wildlife Service 1986). Fre- quent exposure to Pb may result in age-related in- creases in Pb concentrations in soft tissue (Pain et al. 1994), and eagles in western Canada may be exposed to Pb frequently enough to result in age- related Pb accumulation. This may he an impor- tant issue from the population perspective, because eagles are longdived and have a low annual repro- ductive potential, so population levels are linked to the survival of reproducing adults (Grier 1980, Newton 1988). Eagles with Pb shot in their gastrointestinal tracts had much higher Pb levels than those with- out shot in their gastrointestinal tracts. This sug- gests that Pb ammunition is an important source of Pb exposure in eagles in western Canada. Al- though the majority of Pb-poisoned eagles did not March 2003 Eagle Mortality in Western Canada 15 Table 4. Summary of North American studies showing the prevalence of eagles with high Pb and Pb-poisoned tissue concentrations. Species Location Tissue Analyzed (AO High Pb"* (Per- cent) Pb- Poisoned’’ (Per- cent) Reference Bald Eagle Idaho Liver (5) 83 83 Craig et al. 1990 Montana Blood (37) 86 5 Harmata and Restani 1995 Minnesota Blood (25) 96 28 Hennes 1985 Montana & Sas- katchewan Blood (178) 19 1 Miller et al. 1998 Western USA Blood (120) 3" NA'i Wiemeyer et al. 1989 Western Canada Blood, liver, kidney (333) 13 9 This study Golden Eagle Idaho Liver (16) 56 44 Craig et al. 1990 Idaho Blood (178) 42 NA Craig and Craig 1995 Montana Blood (86) 56 2 Harmata and Restani 1995 California Blood (162) 36 2 Pattee et al. 1990 Western Canada Blood, liver, kidney (39) 28 15 This study ^ Pb >6 M-g/g dry weight kidney or liver; >0.2 |JLg/ml blood. Pb >20 Rg/g dry weight kidney; >30 Rg/g dry weight liver; >1.0 Rg/ml blood. Pb >0.4 Rg/ml blood. NA — not available. have Pb shot or fragments in their digestive tracts, there may have been ample time for regurgitation (Pattee et al. 1981) or complete digestion of the Pb fragments. Although the difference was not sta- tistically significant, Golden Eagles with high Pb levels tended to be found during or soon after the fall hunting season, whereas those with back- ground Pb levels were found more often long after hunting seasons had ended. Similar results have been reported for Golden Eagles in California (Pattee et al. 1990), Bald Eagles in the United States (Pattee and Hennes 1983, Wiemeyer et al. 1989) and BC (Elliott et al. 1992), and Western Marsh-Harrier ( Circus aeruginosus) in Europe (Pain et al. 1993, 1997, Mateo et al. 1999). In our study, there was no relationship between Pb levels and the time of year when Bald Eagles were found. However, some Bald Eagles found in late winter or early spring may have died at the onset of winter, but were frozen and buried by snow for several months before being discovered. Eurthermore, ea- gles ingesting Pb ammunition during or shortly af- ter the hunting season may not have died until sev- eral weeks later (Pattee et al. 1981). We tested the hypothesis that consumption of waterfowl is a major source of Pb exposure in ea- gles (Pattee and Hennes 1983) by examining ele- vated Pb levels in the context of waterfowl hunting intensity and interspecific dietary differences. We found no geographic evidence linking high Pb lev- els in eagles to the use of Pb shot for waterfowl hunting, suggesting that waterfowl is not their pri- mary source of Pb. Alternatively, this lack of cor- relation may have resulted from eagles flying long distances in short time periods (McClelland et al. 1994, Brodeur et al. 1996) after ingesting Pb shot. Eagles may have consumed waterfowl and associ- ated Pb shot in areas of high waterfowl hunting intensity, then flown to areas of low waterfowl hunting intensity before dying of Pb poisoning. In addition, the gradual phasing out of lead shot for waterfowl hunting in southwestern BC, beginning in 1998, may have reduced the likelyhood that lead shot-contaminated waterfowl would have been an important source of lead to Bald Eagles, many of which were recovered in that area during this study. Bald and Golden eagles are opportunistic for- agers that eat a wide variety of prey, but in North America, Bald Eagles feed mainly on fish and wa- terfowl (Stalmaster and Plettner 1992, Miller et al. 1998, Restani et al. 2000), while Golden Eagles 16 Wayland et al. VoL. 37, No. 1 feed mainly on mammals and upland game birds (Olendorff 1976). If waterfowl were the primary source of Pb for eagles, one would expect to find a higher percentage of Bald Eagles with high Pb levels than Golden Eagles. In fact, this has been reported in several studies from the western Unit- ed States (Table 4) . In contrast, we found high Pb levels in 13% of Bald Eagles and 28% of Golden Eagles in western Canada. Kramer and Redig (1997) suggested that Pb poisoning in eagles in the north-central United States resulted mainly from their scavenging small mammals and large game. Pb shot and bullet fragments in game birds, small mammals, and large game carcasses constituted the major source of Pb for Golden Eagles (Pattee et al. 1990, Gjershaug 1992, Bezzel and Eiinfstuck 1995, Harmata and Restani 1995, Kendall et al. 1996). Moreover, Pb-poisoned Bald Eagles in the Greater Yellowstone Ecosystem had ingested large caliber bullets while feeding on ungulate carcasses (Harmata et al. 1999). Our results, as well as those from the studies mentioned above, suggest that wa- terfowl is not the primary source of Pb for eagles m large areas of western North America. Banning the use of Pb shot for waterfowl hunting may not significantly reduce Pb poisoning in eagles, and the use of Pb ammunition for hunting mammals and upland game birds may continue to pose a risk to eagles and other raptor species in Canada. In conclusion, we found that various physical in- juries, electrocution, pesticide poisoning, gunshot wounds, and Pb poisoning were the most common diagnoses in Bald and Golden eagles in western Canada. Further monitoring of eagle mortality and populations are required to assess the population level impacts of these mortality factors. Although high Pb levels were associated with other toxicoses, there was no evidence to suggest that sublethal ef- fects of Pb predisposed eagles to injury or disease. In comparison with immature eagles, a higher per- cent of adult and subadult eagles had high Pb lev- els. This difference among age classes may also have repercussions at the population level. We found no significant association between Pb levels and the fall hunting season, but we suspect that eagles in western Canada were exposed incidental- ly to Pb ammunition through the consumption of various prey species. Waterfowl appeared not to be the primary source of Pb for eagles in western Can- ada, and we suspect that Bald and Golden eagles will continue to be Pb-poisoned despite the ban on the use of Pb shot for waterfowl hunting. Acknowledgments We thank the many people who provided carcasses and injured birds. Staff at local wildlife rehabilitation centers and biologists and conservation officers with Alberta Fish and Wildlife, British Columbia Ministry of the Environ- ment, Manitoba Department of Natural Resources, and Saskatchewan Environment and Resource Management were especially helpful. The following individuals provid- ed logistical and laboratory support: G. Babish, S. Lee, F. Leighton, R. McNeil, G. Sans-Cartier, B. Treichel, S. Tru- deau, B. Wakeford, and G. Wobeser. E. Neugebauer con- ducted the Pb analyses. Funding was provided by Envi- ronment Canada (Canadian Wildlife Service) and the Fraser River Action Plan. We thank L. Locke, M. Restani, and A. Harmata for reviewing earlier versions of this manuscript. Literature Cited Bellrose, F. 1959. Lead poisoning as a mortality factor in waterfowl populations. III. Nat. Hist. Surv. Bull. 27: 235-288. Bezzel, E. and H.J. Funestuck. 1995. Lead poisoning as a threat to Golden Eagles Aquila chrysaetos in the northern Alps? J. Ornithol. 136:294-296. Bollinger, T. and P. Mineau. 1995. Pesticide poisonings in eagles — update. Wildlife Health Centre Newsletter, Vol. 3, No. 3. Canadian Cooperative Wildlife Health Centre, University of Saskatchewan, Saskatoon, Can- ada. Bortolotti, G.R. 1984a. Sexual size dimorphism and age-related size variation in Bald Eagles. J. Wildl. Man- age. 48:72-81. . 1984b. Age and sex size variation in Golden Ea- gles. /. Field Ornithol. 55:54—66. . 1984c. Trap and poison mortality of eagles. J. Wildl. Manage. 48: 1 1 73-1 1 79. Brodeur, S., R. Decarie, D.M. Bird, and M. Fuller. 1996. Complete migration cycle of Golden Eagles breeding in northern Quebec. Condor 98:293-299 . Burger, J. 1995. A risk assessment for lead in birds. /. Toxicol. Environ. Health 45:369-396. Craig, E.H. AND T.H. Craig. 1995. Lead levels in Golden Eagles in southeastern Idaho./. Raptor Res. 29:54-55. Craig, T.H., J.W. Connelly, E.H. Craig, and T.L. Parker. 1990. Lead concentrations in Golden and Bald eagles. Wilson Bull. 102:130-133. Elliott, J.E., K.M. Langelier, A.M. Scheuhammer, P.H. SiNCiAiR, AND RE. Whiiehead. 1992. Incidence of lead poisoning in Bald Eagles and lead shot in water- fowl gizzards from British Columbia 1988-91. CWS Prog. Note 200, Canadian Wildlife Service, Ottawa, Canada. , K.M. Langelier, P. Mineau, and L.K. Wilson. 1996. Poisoning of Bald Eagles and Red-tailed Hawks by carbofuran and fensulfothion in the Fraser Delta of British Columbia, Canada./. Wildl. Dis. 32:486-491. , L.K Wilson, K.M. Langelier, P. Mineau, and P.H. Sinclair. 1997. Secondary poisoning of birds of March 2003 Eagle Mortality in Western Canada 17 prey by the organophosphorus insecticide, phorate. Ecotoxicology 6:219—231. Fernandez, FJ AND D. Hilligoss. 1981. An improved graphite furnace method for the determination of lead in blood using matrix modification and the L’vov platform. At. Spectrosc. 2:1-7. Franson, J.C. 1996. Interpretation of tissue lead residues in birds other than waterfowl. Pages 265-279 in W.N. Beyer, G.H. Heinz, and A.W. Redmon-Norwood [Eds.], Environmental contaminants in wildlife: inter- preting tissue concentrations. Lewis Publishers, New York, NYU.S.A. , L. SiLEO, AND N.J. Thomas. 1995. Causes of eagle deaths. Page 68 in E.T. LaRoe, G.S. Farris, C.E. Puck- ett, P.D. Doran, and M.J. Mac [Eds.], Our living re- sources: a report to the nation on the distribution, abundance and health of U.S, plants, animals, and ecosystems. U.S. Department of the Interior, National Biological Service, Washington, DC U.S.A. Gamma Design. 1999. GS+ for Windows 95/98/NT. Ver- sion 3.11.7. Gamma Design Software (http:// www.gammadesign.com) , Plainwell, Ml U.S.A. Gerrard, J.M. 1983. A review of the current status of Bald Eagles in North America. Pages 5-20 in D.M. Bird, N.R. Seymour and J.M. Gerrard [Eds.], Biology and management of Bald Eagles and Ospreys. Harpell Press, Ste. Anne de Bellevue, Canada. Gjershaug, J.O. 1992. Blyhagl og rovfugler. Vart Fuglefau- na 15:10-13. Greig-Smith, P.W. 1991. Use of cholinesterase measure- ments in surveillance of wildlife poisoning in farm- land. Pages 127-150 m P. Mineau [Ed.], Cholinester- ase-inhibiting insecticides: their impact on wildlife and the environment. Elsevier Science Publishers, New York, NYU.S.A. Grier, J.W. 1980. Modeling approaches to Bald Eagle population dynamics. Wildl. Soc. Bull. 8:316-322. Harmata, A.R. and M. Restani. 1995. Environmental contaminants and cholinesterase in blood of vernal migrant Bald and Golden eagles in Montana. Inter- mountain J. Sci. 1:1-15. , G.J. Montopoli, B. Oakleaf, P.J. Harmata, and M. Restani. 1999. Movements and survival of Bald Ea- gles in the greater Yellowstone ecosystem. J. Wildl. Manage. 63:781-793. Hennes, S.K. 1985. Lead shot ingestion and lead residues in migrant Bald Eagles at the Lac Qui Parle Wildlife Management area, Minnesota. M.S. thesis, Univ. Min- nesota, St. Paul, MN U.S.A. Hill, E.E. and W.J. Fleming. 1982. Anticholinesterase poisoning of birds: Field monitoring and diagnosis of acute poisoning. Environ. Toxicol. Chem. 1:27-38. Kendall, R.J., T.E. Lacher, Jr., C. Bunck, B. Daniel, C. Driver, C.E. Grue, E. Leighton, W. Stansley, P.G. Watanabe, and M. Whitworth. 1996. An ecological risk assessment of lead shot exposure in non-water- fowl avian species: upland game birds and raptors. En- viron. Toxicol. Chem. 15:4-20. Kirk, D.A. 1996. Updated status report on the Golden Eagle Aquila chrysaetos in Canada. COSEWIC Status Report. Committee on the status of endangered wild- life in Canada. Ottawa, Canada. Kramer, J.L. and P.T. Redig, 1997. Sixteen years of lead poisoning in eagles, 1980-95: an epizootiologic view. J. Raptor Res. 31:327-332. Locke, L.N. and M. Eriend. 1992. Lead poisoning of avi- an species other than waterfowl. Pages 19-22 in DJ. Pain [Ed.], Lead poisoning in waterfowl. Proc. IWRB Workshop, Brussels, Belgium, 1991. IWRB Spec. Publ. 16. International Waterfowl and Wetlands Research Bureau, Slimbridge, U.K. Martin, A.D., G. Norman, P.I. Stanley, and G.E. West- lake. 1981. Use of reactivation techniques for the dif- ferential diagnosis of organophosphorus and carba- mate pesticide poisoning in birds. Bull. Environ. Toxicol. 26:775-780. Mateo, R., J. Estrada, J-Y Paquet, X. Riera, L. Domin- guez, R. Guitart, and a. Martinez-Vilalta. 1999. Lead shot ingestion by marsh harriers Circus aerugi- nosus from the Ebro delta, Spain. Environ. Pollut. 104. 435-440. McClelland, B.R., L.S. Young, P.T. McClei.iand, J.G. Crenshaw, H.L. At . ten, and D.S, Shea. 1994. Migra- tion ecology of Bald Eagles from autumn concentra- tions in Glacier National Park, Montana. Wildl. Mon- ogr. 125:1-61. McCollough, M. 1989. Molting sequence and aging of Bald Eagles. Wilson Bull. 101:1-10. Miller, M.J.R., M. Restani, A.R. Harmata, G.R. Borto- lotti, and M.E. Wayiand. 1998. A comparison of blood lead levels in Bald Eagles from two regions on the Great Plains of North America. J. Wildl. Dis. 34 704-714. Mulhern, B.M., W.L. Reichel, L.N. Locke, T.G. Lamont, A. Belisle, E. Gromartie, G.E. Bagley, and R.M Prouty. 1970. Organochlorine residues and autopsy data for Bald Eagles. Pestic. Monit. J. 4:141-144. National Wildlife Health Laboratory. 1985. Bald Ea- gle mortality from lead poisoning and other causes, 1963-84. U.S. Fish and Wildlife Service, National Wildlife Health Laboratory, Madison, W1 U.S.A. Newton, I. 1988. Individual performance in sparrow- hawks: the ecology of two sexes. Int. Ornithol. Cong V 125-154. OlENDORFF, R.R. 1976. The food habits of North Amer- ican Golden Eagles. Am. Midi. Nat. 95:231-236. Pan, D.J. and C. Amiard-Triquet. 1993. Lead poisoning of raptors in France and elsewhere. Ecotoxicol. Environ Saf. 25:183-192. , C. Amiard-Triquet, C. Bavoux, G. Burneleau, L. Eon, and P. Nicolau-Guillaumet. 1993. Lead poi- soning in wild populations of marsh harriers Circus 18 Wayland et al. VoL. 37, No. 1 aeruginosus in the Camargne and Charente-Maritime, France. Ibis 135:379-386. , J. Sears, and I. Newton. 1994. Lead concentra- tions in birds of prey in Britain. Environ. Pollut. 50; 173-180. , C. Bavoux, and G. Burneleau. 1997. Seasonal blood lead concentrations in marsh harriers Circus aeruginosus from Charente-Maritime, France: relation- ship to hunting season. Biol. Conserv. 81:1-7. Pattee, O.H. and S.K. Hennes. 1983. Bald Eagles and waterfowl: the lead shot connection. Trans. N. Am. Wildl. Nat. Resour. Conf. 48:230-237. , S.N. Weimeyer, B.M. Mulhern, L. Sileo, and J.W. Carpenter. 1981. Experimental lead shot poison- ing in Bald Eagles./. Wildl. Manage. 45:806-810. , P.H. Bloom, J.M. Scott, and M.R. Smith. 1990. Lead hazards within the range of the California Con- dor. Condor 92:931-937. Redig, P.T., G.E. Duke, S. Schwartz, and E. Lawler. 1984. An investigation into the effects of lead poison- ing on Bald Eagles and other raptors; final report. Minnesota Endangered Species Program Study 200A- 200B, Minneapolis-St. Paul, MN U.S.A, Reichel, W.L., S.K. ScHMELiNG, E. Cromartie, T.E. Kai- ser, A.J. Krynitsky, T.G. Lamont, B.M. Mulhern, R.M. Prouty, C.J. Stafford, and D.M. Swineford. 1984. Pesticide, PCB and lead residues and necropsy data for Bald Eagles from 32 states — 1978-81. Envi- ron. Monit. Assess. 4:395-403. Reiser, M.H. and S.A, Temple. 1981. Effects of chronic lead ingestion on birds of prey. Pages 21-25 mJ.E. Cooper and A.J. Greenwood [Eds.], Recent advances in the study of raptor diseases. Chiron Pub. Ltd., West Yorkshire, U.K. Restani, M., A.R. Harmata, and E.M. Madden. 2000. Nu- merical and functional responses of migrant Bald Ea- gles exploiting a seasonally concentrated food source. Condor 102:561-568. SAS Institute. 1988. SAS stat user’s guide. Release 6.03. SAS Institute, Cary, NC U.S.A. Scheuhammer, A.M. and L.K. Wilson. 1990. Effects of lead and pesticides on d-aminolevulinic acid dehydra- tase of Ring Doves {Streptopelia risoria) . Environ. Toxicol. Chem. 9:1379-1386. AND S.L. Norris. 1995. A review of the environ- mental impacts of lead shotshell ammunition and lead fishing weights in Canada. Occasional Paper No. 88, Canadian Wildlife Service, Ottawa, Canada. Stalmaster, M.V. and R.G. Plettner. 1992. Diets and foraging effectiveness of Bald Eagles during extreme winter weather in Nebraska. J. Wildl. Manage. 56:355- 367. U.S. Eish and Wildijfe Service. 1986. Einal supplemen- tal environmental impact statement on the use of lead shot for hunting migratory birds. Eish and Wildlife Service, Office of Migratory Bird Management, Wash- ington, DC U.S.A. Wayland, M. and T. Bollinger. 1999. Lead exposure and poisoning in Bald Eagles and Golden Eagles in the Canadian prairie provinces. Environ. Pollut. 104: 341-350. , E. Neugebauer, and T. Bollinger. 1999. Con- centrations of lead in liver, kidney, and bone of Bald and Golden eagles. Arch. Environ. Contam. Toxicol. 37: 267-272. Wiemeyer, S.N., R.W. Frenzel, R.G. Anthony, B.R. McClelland, and R.L. Knight. 1989. Environmental contaminants in blood of western Bald Eagles./. Rap- tor Res. 23:140-146. WOBESER, G.A. 1994. Investigation and management of disease in wild animals. Plenum Press, New York, NY U.S.A. . 1997. Diseases of wild waterfowl, 2nd Ed. Plenum Press, New York, NY U.S.A. Received 30 December 2000; accepted 16 November 2002 Associate Editor: Ian Warkentin J. Raptor Res. 37(l):19-30 © 2003 The Raptor Research Foundation, Inc. ABUNDANCE OF SOARING RAPTORS IN THE BRAZILIAN ATIANTIC RAINFOREST Santi Manosa/ Eduardo Mateos, and Vittorio Pedrocchi Departament de Biologia Animal, Universilat de Barcelona, Facultat de Biologia, Avinguda Diagonal, 645, 08028 Barcelona, Catalonia, Spain Abstract. — 18 August-4 September 1998, we conducted 23 3-4 hr point-counts in an Atlantic rainforest area of southeastern Brazil to evaluate the richness and relative abundance of raptors in two adjacent protected areas, Parque Estadual Intervales and the Parque Estadual Turistico do Alto Ribeira. During 88.2 hr, we recorded 334 contacts with raptors, involving 734 individuals of nine .species. Contacts per hour and the number of species tallied showed that the counts were higher between 0900-1200 H (Local Standard Time), and that counts of 3 hr were the most cost effective. Reasonable precision for abundance indices was achieved with samples sizes of 20-30 points, but samples of 12 should give satisfactory results for the more common species, as long as counting points are distributed sufficiently in space. We derived abundance indices for species of raptors most commonly seen in the area. In 14 3-hr counts. Black Vultures (Coragyps atratus) were observed in 100% of them, Mantled Hawks (Leucop- ternis polionota) in 71%, Black Hawk-Eagles (Spizaetus tyrannus) in 50%, Turkey Vultures {Cathartes aura) in 29%, Ornate Hawk-Eagles {Spizaetus ornalus) in 21%, Roadside Hawks {Buteo magnirostris) in 14%, Short-tailed Hawks {Buteo hrachyurus) in 14%, Crested Caracaras {Polyborus plancus) in 14%, and Tiny Hawks {Accipiter superciliosus) in 7%. Bat Falcons {Falco rupgiilaris) and White-tailed Hawks {Buteo albi- caudatus) also were reported in the area, but outside the counting periods. Key Word.S: Brazilian rainforest, point count; richness; Sdo Paulo State. ABUNDANCIA DE RAPACES PLANEADORAS EN LA SELVA ATLANTICA BRASILENA Resumen. — Entre el 18 de agosto-4 de septiembre de 1998, realizamos 23 census pun tales de 3-4 hr de duracion en la selva Atlantica del Brasil, para estimar la riqueza y abundancia de aves de presa en dos zonas protegidas adyacentes, el Parque Estadual Intervales y el Parque Estadual Turistico do Alto Ribeira. Tras 88.2 hr de censo se obtuvieron 334 observaciones de rapaces correspondientes a 734 individuos de 9 especies. El numero de contactos y de especies observadas indican que la mejor hora para realizar los census se situa entre las 0900—1200 H (Hora Local Time) y que su duracion optima es de 3 hr. Una buena precision en las estimas de abundancia se obtiene a partir de tamahos muestrales de 20-30 puntos, pero muestras de 12 puntos pueden ofrecer resultados satisfactorios para las especies mas frecuentes, siempre y cuando los puntos se encuentren bien repartidos por el area de estudio. En catorce census de tres horas llevados a cabo entre las 0900—1200 H, el Zopilote Negro {Coragyps atratus) se observe en el 100% de los census, el Busardo Blanquinegro {Leucopternis polionota) en el 71%, el Aguila-Azor Negra {Spizaetus tyrannus) en el 50%, el Aura Gallipavo {Cathartes aura) en el 29%, el Aguila- Azor Galana {Spizaetus ornatus) en el 21%, el Busardo Caminero {Buteo magnirostris), el Busardo Colicorto {Buteo brachyurus), el Caracara Carancho {Polyborus plancus) en el 14%, y el Gavilancito Americano {Accipiter superciliosus) en el 7% los casos. El Halcon Murcielaguero {Falco rufigularis) y el Busardo Col- iblanco {Buteo albicaudatu.s) fueron observados fuera de los periodos de censo. [Traduccion de los Autores] The Atlantic rainforest of Brazil is one of the most threatened biomes in the world. Only 2-8% of the 10*^ km^ of forest that once covered a narrow stretch of land along the southeastern coast of Bra- zil remains in scattered, small fragments (Fonseca ^ E-mail address: srife@bio.ub.es 1985, Albuquerque 1995, Myers et al. 2000). One of the best-preserved areas of Atlantic rainforest is the Paranapiacaba forest fragment (140 000 ha) (Mateos et al. 2002). This fragment includes some areas of unprotected forest and four protected ar- eas (Parque Estadual Turistico do Alto Ribeira, Parque Estadual Intervales, Parque Estadual Carlos 19 20 Manosa et al. VoL. 37, No. 1 Botelho, and Esta^ao Ecologica Xitue) known as the Paranapiacaba Ecological Continuum (Pisciot- ta 2002), which, all together, form one of the larg- est remaining patches of uninterrupted Atlantic rainforest (Eig. 1). In spite of its fragmentation, the Atlantic rain- forest remains among the richest of all ornitholog- ical areas in the world (Cracraft 1985, Wege and Long 1995, Stattersfield et al. 1998) and is consid- ered a hotspot for biodiversity conservation (Myers 1988, Bibby et al. 1992b, Myers et al. 2000). The area gives refuge to many endemic birds (Myers et al. 2000) , among which are several species and sub- species of raptors such as Grey-headed Kites {Lep- todon cayanensis) , White-necked Hawks {Leucopternis lacernulata) , Mantled Hawks {Leucopternis poliono- ta), and Black Hawk-Eagles {Spizaetus t. tyrannus) (Collar et al. 1992, del Hoyo et al. 1994, Bierre- gaard 1998, Bildstein et al. 1998). This justifies the need for scientific and conservation initiatives to protect this important element of the global bio- diversity (Burnham et al. 1994). Raptors usually live at low densities and are dif- ficult to detect, so that the methods usually em- ployed to evaluate general bird populations are not adequate (Eorsman and Solonen 1984, Thiollay 1989, Bibby et al. 1992a). Accurate species richness and abundance estimates require surveying large areas and conducting a large number of counts. The dense structure of the rainforest adds even more difficulty to the study of these raptors (Thiol- lay 1989), and a great deal of time and effort are often required to prepare and to conduct the counts (Whitacre and Turley 1990, Manosa and Pedrocchi 1997). For these reasons, the conservation status of many woodland raptors in the neotropics is poorly known (Albuquerque 1986, Thiollay 1994), in spite of the fact that more than half of the species of neotropical raptors are endangered (Thiollay 1985, Bildstein et al. 1998). Several research and monitoring programs are currently providing new information about raptors in tropical forests (Thiollay 1989, Vannini 1989, Whitacre and Thor- strom 1992), but only preliminary research has been conducted in the Atlantic rainforest (Willis and Oniki 1981, Guix et al. 1992, Mateos and Man- osa 1996, Manosa and Pedrocchi 1997, Veilliard and Silva 2001, Manosa et al. 2002). Considering the need of improving our knowl- edge on raptor assemblages in Atlantic rainforest areas, as well as the need of achieving maximum accuracy of the results and efficiency of the count- ing effort, the objectives of our research were (1) to evaluate the effect of several variables (time of day, duration of counts, number of counts, selec- tion of counting points) on the efficiency of the point-count method in estimating species richness and abundance indices of raptor assemblages in the Atlantic rainforest, and (2) to derive the best estimate of abundance indices of raptors in a well- preserved area of Brazilian Atlantic rainforest. Material and Methods Study Areas. Counts were conducted in two adjacent areas of the Parque Estadual Turistico do Alto Ribeira (Petar) and the Parque Estadual Intervales (Intervales), within the Paranapiacaba forest fragment in Serra do Mar mountain range, in the state of Sao Paulo, Brazil (Eig. 1). The Serra do Mar extends parallel to the Atlan- tic coast for over 900 km. Its slopes rise to abrupt moun- tain peaks between 800-1100 m in elevation. The Par- anapiacaba forest fragment (Pisciotta 2002) consists of four legally-protected, neighboring reserves, known as the Paranapiacaba Ecological Continuum (1258 km^), plus some adjacent private forest areas. Together, they comprise some 1400 km^ of uninterrupted forest in sev- eral ecological successional stages (Fig. 1). Most of the Paranapiacaba forest fragment is covered by Low Eleva- tion South Hillside Atlantic rainforest (Guix 2002), typi- cal of the altitudinal range between 50-100 m and 1200- 1600 masl, and tree height ranging from 20-30 m. Some surrounding and marginal areas (near 10% of the total area) is planted Araucaria angustifolia, Firms sp. and Eu- calyptus sp. forest, banana plantations, and pastures. Four main vegetation types can be distinguished in the Par- anapiacaba forest fragment: mature forests, which have been subject to no episode of deforestation or intensive selective extraction, or only a single one more than 80- 100 yr ago; late-secondary forests, which suffered the last episode of deforestation or intensive selective extraction 50-80 yr ago; young-secondary forests, which suffered the last deforestation or selective extraction episode 20-40 yr ago; and “capoeiras” or shrublands, which are forest ar- eas that have undergone the last deforestation episodes just 5-15 yr ago. The study area is covered by mature and secondary forest types. In some zones of shallow calcar- eous soil in the Petar site, mature or late-secondary for- ests are lower in stature than those of the same .succes- sional stages in Intervales. The study area in Intervales was the main valley of the Formoso-Piloes River, from just above Base do Carmo to a few kilometers beyond Base do Alecrim (Fig. 1). The Atlantic rainforest in this valley is mature or late-second- ary, with a mean height of 18 m and emergent trees reaching 24—30 m. Pahnito {Euterpe edulis) is abundant but becomes scarcer at the bottom of the valley, where taquara {Merostachys sp.) is widespread. On mountain sides and summits and facing north northwest, the Atlan- tic rainforest is most intact and there is hardly any ta- quaral present. On the highest areas of the valley, near Base do Carmo, human influence is minimal or nonex- istent: mature or old-secondary forests cover the region. March 2003 Soaring Raptors in Atiantic Rainforest 21 4 - 0 2 4 6 8 10 km I I I I I I I I I i I I Figure 1. Location of the study area in Brazil. The limits of the four protected areas conforming the Ecological Continuum of the Paranapiacaba forest fragment within the Sao Paulo State are indicated. (A: Parque Estadual Turistico do Alto Ribeira; B: Parque Estadual, Intervales; C: Parque Estadual, Carlos Botelho; D: Estagao Ecologica Xitue). The study area is outlined and enlarged, showing the location of the 23 counting points indicated by the numbers. 22 Manosa et al. VoL. 37, No. 1 “Capoeiras” occur only along the main road and trails, due to the extraction of palmito palms and wood. The valley bottom near Alecrim has been inhabited and mod- ihed for centuries. Late-secondary forest dominates this region, with a scattering of small banana plantations. Ta- quaral areas around Alecrim are the remains of opened areas used for subsistence farming 15-25 yr ago. The study area in Parque Estadual Turistico do Alto Ribeira was adjacent to the village of Nucleo Caboclos (Fig. 1). The Nucleo Caboclos is 600 masl and is covered by typical Hillside Atlantic rainforest vegetation (Carval- ho et al. 2002) with a high botanical diversity. This high diversity is caused by the existence of calcareous sedi- ments that enhance the growth of forests with a partic- ular floristic composition. Secondary vegetation grows in the margins of dirt tracks and trails; it is characterized by taquaral and shrublands. Field Procedures. The point-count method (Whitacre and Turley 1990, Whitacre et al. 1992a) was used because It was the most effective method in a preliminary survey in 1994 (Manosa and Pedrocchi 1997). Counts were con- ducted during the dry season, between 18 August and 4 September 1998, when some raptors (i.e., Mantled Hawk, Black Hawk-Eagle) are expected to be displaying in the area (Vielliard and Silva 2001). One or two observers stayed on high points of the landscape or perched on the tops of trees, having a view angle of 80-294° and a range of view above the canopy of 1000-4000 m. We selected 23 points along main tracks or footpaths (Eig. 1), which we judged to be representative of the dominant habitat m the area. Seventeen of these points were along 27 km of the Formoso-Piloes Valley and six around the Caboclos area. Each point was selected in order to give an inde- pendent view of a part of the study area. The mean dis- tance between neighboring points was 997 ± 441 m (SD) in Intervales (range = 267-1493) and 1900 ± 1665 m in Petar (range = 600-5161). Even the closest points of- fered non-overlapping views of the study area. The counts were initiated between 0800-1200 H (Local Stan- dard Time, sunrise 0519 H), and la.sted for 2. .5-4 hr. The mean length of the counts was 3.8 ± 0.4 hr (SD). Each point was sampled only once. The 23 counts totalled 88.2 hr of observation, concentrated from 0900-1300 H (72.1 hr) . Counts were conducted in clear weather, except for one that was conducted in light rain but yielded com- parable results. During each count interval, we recorded every raptor in view to an unbounded distance. In this manner, we obtained a list of the minimum number of individuals and groups observed during every count. Computation of Relative Abundance Parameters. We computed the following abundance indices for eacb spe- cies and area: percentage of counting points in which a species was detected (percent presence ±95% Confi- dence Interval [Cl]), mean group size (individuals ± SD), and contacts/hr (±95% Cl). Confidence intervals were computed by bootstrap techniques, using the Resam- pling Slats suite (Bruce et al. 1995): a parameter value was computed for each original counting point. This gave an initial sample of values, which was randomly resam- pled with replacement to obtain a sample size identical to the original one. This process was repeated 10 000 times. Then, the mean and confidence intervals of the 10 000 parameter estimates were obtained. Evaluation of Biases. The effect of the time of day (morning only) on the results of the counts was evalu- ated by analyzing the number of species reported at sev- eral hourly intervals. For species in which the number of observations was large (Mantled Hawk, Black Vulture, Turkey Vulture, and Black Hawk-Eagle), an hourly re- porting pattern was also obtained. An index of activity was computed for each hourly interval and species. Each hourly period of observation was divided into twelve 5- min intervals. During each interval, we recorded the number of individuals in view. The activity index for a given species during a specific hourly interval was the product of the proportion of 5-min intervals in which the species was in view during that period (from 0/12-12/ 12) by the addition of all individuals recorded during the twelve 5-min intervals. To assess the optimal duration of counts, we analyzed how the number of detected species and abundance in- dices estimates changed in relation to count duration. We limited this analysis to the counts initiated at 0900 H and that were extended to 1300 H {N = 9), so that the effect of sampling at different times of day was removed. The relationship between precision of estimates and sample size was modelled using bootstrapping techniques (see above) by taking samples of progressively increasing size (1-50). To evaluate how the precision of the abun- dance indices improved with sample size, the ±95% Cl of these estimates were plotted against the number of counts. To make comparisons between parameters and species possible, precision was expressed as (upper 95% Cl — lower 95% Gl)/mean. To test how the selection of the counting points could affect the results, we divided our initial set of 23 points into two sets of spatially alternating points, a set of 12 (set A) and a set of 1 1 (set B) points. These two data subsets were used to derive two separate estimates of the relative-abundance parameters. Comparison between these two sets should give some indication of the repro- ducibility and accuracy of our results. Computing Standard Abundance Estimates. Standard- ized abundance indices estimates of raptors for the study area as a whole and for each of the two Preserve Areas were derived, considering only the 14 counts in the 0900-1200 H period. The remaining nine counts were omitted because they lasted for less than three hours or because they did not include the above-mentioned peri- od. Resui.ts We recorded 1 1 ray^tor species in die Paranapia- caba forest fragment, two of which, the Bat Falcon {Falco rujigularis) and the Wliite-tailed Hawk {Buteo albicaudatus) , were observed for the hrst time in the area (Table 1). The 23 counts yielded 334 ob- servations of raptors (3.8 contacts/hr) involving 734 individuals, belonging to nine species of rap- tors (Table 2). Only one of the 23 counts (4%) resulted in no raptor observations, four yielded one species, five yielded two species, four yielded March 2003 Soaring Raptors in Atlantic Rainforest 23 Table 1 . Composition of the raptor assemblage in the Paranapiacaba forest fragment, according to the species recorded in several surveys to the area, “x” indicates the species was recorded outside the primary sample periods. Vagrant species, as well as migratory species that are absent from the area in August, are excluded. Willis AND Oniki (1981)^ VlF.I ,I .lARD AND Silva (2001)'^ Manosa et ai.. (1997) This study* Black Vulture ( Coragyps atratus) 49 100% 50% 100% Turkey Vulture ( Cathartes aura) 59% X 29% Grey-headed Kite {Leptodon cayanensis) 9% ?e Rufous-thighed Kite {Harpagus diodon) 18% Grey-bellied Goshawk {Accipiter poliogaster) 14% Tiny Hawk {Accipiter superdliosus) 25% 7% Sharp-shinned Hawk {Accipiter striatus) X Crane Hawk {Geranospiza caerulescens) X Mantled Hawk {Leucopternis polionota) 2 86% 50% 71% Great Black-Hawk {Buteogallus urubitinga) 4% Black-chested Buzzard-Eagle ( Geranoaetus melanoleucus) X Roadside Hawk {Buteo magnirostris) 34 95% X 14% White-rumped Hawk {Buteo leucorrhous) 2 X Short-tailed Hawk {Buteo brachyurus) 7 54% X 14% White-tailed Hawk {Buteo albicaudatus) X Guiana Crested Eagle {Morphnus guianensis) X Black Hawk-Eagle {Spizaetus tyrannus) 68% X 50% Ornate Hawk-Eagle {Spizaetus ornatus) 25% 21% Crested Caracara {Polyborus plancus) 2 18% 14% Chimango Caracara {Milvago chimachima) 17 54% X Laughing Falcon {Herpetotheres cachinnans) 5 X Barred Forest-Falcon {Micrastur ruficollis) 10 77% Collared Forest-Falcon {Micrastur semitorquatus) 4% X Bat Falcon {Falco rufigularis) X Individuals/ 100 hr observation {N = 41.2 hr) reported in the Carlos Botelho area between 24—28 February and 5- -10 July 1979 (Swallow-tailed Kite [Elanoides forficatus excluded] ) . Percent of 22 visits to the Intervales area between August 1988 and December 1992 in which the species were reported, “x” indicates presence during other occasional visits to the area (Osprey [Pandion haliaetus], Swallow-tailed Kite, and Plumbeous Kite [Ictinia plumbea excluded] ) . Percent of point counts {N = 4) in which the species was detected during a visit to Intervales between 1 and 12 August 1994. “x” indicates detection outside the counts. Percent of point counts {N = 14) in which the species was detected during a visit to Intervales and the Petar in August-September 1998. “x” indicates detection outside the counts. Identification was uncertain. three species, six yielded four species, and three yielded five species. Effect of Time of Day. Only Mantled Hawks and Roadside Hawks (Buteo magnirostris) were observed before 0900 H. Although observation effort was minimal, only Black Vultures were consistently ob- served during the afternoon (1300—1600 H). The activity of this species remained more or less con- stant from 1000 H onwards (Fig. 2). Comparatively, the activity of Turkey Vultures was more concen- trated around mid-day. Activity of Mantled Hawks remained constant during the morning and early afternoon, whereas the activity of Black Hawk-Ea- gles tended to be maximum between 0900 and 1200 H. Overall, the period 1000-1200 H em- braced the peak activity periods of the three most common species, as well as all the observations of Short-tailed Hawks {Buteo brachyurus) . A high per- centage of the records of the less detectable spe- cies, such as Tiny Hawks (Accipiter superdliosus, 100%, N = \), Crested Caracaras {Polyborus plancus, 50%, A — 2), and Ornate Hawk-Eagles {Spizaetus ornatus, 50%, A = 4) were concentrated between 0900 and 1000 H. 24 Manosa et al. VoL. 37, No. 1 Table 2. Raptor abundance indices in Parque Estadual Intervales, Parque Estudual Turistico do Alto Ribeira and pooled estimate. Means and ±95% Confidence Intervals are given in parenthesis. N is the number of counts. Intervales (N= 9) Petar {N= 5) Pooled (N= 14) Hours of observation 27 15 42 Number of species 8 8 9 Black Vulture Contacts/hr 2.5 (T5-3.7) 3.3 (2.3-4.2) 2.8 (2.0-3.6) Percent presence 100 (100-100) 100 (100-100) 100 (100-100) Mantled Hawk Contacts/hr 0.7 (0.3-1.3) 0.8 (0.4-1. 2) 0.8 (0.4-1.2) Percent presence 67 (33-100) 80 (40-100) 71 (50-93) Black Hawk-Eagle Contacts/hr 0.6 (0.1-1.2) 0.7 (0.1-1.4) 0.6 (0.2-1. 1) Percent presence 44 (11-78) 60 (20-100) 50 (21-79) Turkey Vulture Contacts/hr^ <0.1 (O.O-O.l) 0.5 (0.1-0.8) 0.2 (0.05-0.4) Percent presence 11 (0.00-33) 60 (20-100) 29 (7-57) Roadside Hawk Contacts/hr 0.2 (0.0-0.6) 0.1 (0.0-0.2) 0.1 (0.0-0.4) Percent presence 11 (0-33) 20 (0-60) 14 (0-36) Ornate Hawk-Eagle Contacts/hr <0.1 (O.O-O.l) 0.1 (0.0-0.3) 0.1 (O.O-O.l) Percent presence 11 (0-33) 40 (0-80) 21 (0-43) Short-tailed Hawk Contacts/hr 0.0 0.2 (0.0-0.5) 0.1 (0.0-0.2) Percent presence 0 40 (0-80) 14 (0-36) Crested Caracara Contacts/hr <0.1 (O.O-O.l) 0.1 (0.0-0.2) 0.05 (O.O-O.l) Percent presence 11 (0-33) 20 (0-60) 14 (0-36) Tiny Hawk Contacts/hr <0.1 (O.O-O.l) 0.0 0.02 (O.O-O.l) Percent presence 11 (0-33) 0 7 (0-21) '' Significant differences between areas at P < 0.05; Kruskal-Wallis test. Effect of Count Duration. The number of spe- cies observed (±95% Cl) tended to increase as the duration of the counts increased from 1 hr (1.4 ± 0.5 species) to 2 hr (2.7 ± 0.9), but leveled off progressively for 3 hr (3.0 ± 0.9) and 4 hr counts (3.2 ± 1.0). Among the four more frequent spe- cies, the cumulative percentage of counts levelled off after 2-3 hr of sampling for the Black Vulture and the Black Hawk-Eagle. For the Turkey Vulture and the Mantled Hawk, this parameter increased slightly, although not signihcantly, during the fourth hour of the count (Fig. 3). Effect of the Number of Counts. The number of species detected increased quickly with the num- ber of counts. Some 20 counts were needed to de- tect 90% of the total species documented, but the rate of addition of new species was low with the addition of more counts (Fig. 4). Precision of abundance indices estimates quickly increased with sample size (Fig. 5) . For the more frequent species (i.e., percent presence 70-100%: Black Vultures, Mantled Hawks), the precision of the abundance estimates falls below one ±95% Cl of the mean after 12 counts. For species with <30% presence March 2003 Soaring Raptors in Atlantic Rainforest 25 a £ 50 counts) , which also would be needed to detect ad- ditional species or to slightly improve the precision of the estimates of more common species. Bias in the selection of counting points may also have an effect on the estimation of species richness and on the accuracy of abundance estimates. This bias can arise from selecting points not represen- tative of the area being surveyed, or from excessive proximity" of sampling points, which may result in double counting of territorial raptors. Comparison of two subsets of our data points suggests that for the most detectable species, species richness was little affected by the reduction of number and placement of counting points, and that relative abundance estimates were also not altered. Counts were conducted at the start of the nest- ing season, when raptors were most active and probably when the point count method was most effective (Whitacre and Turley 1990). However, during the counts we only detected nine of the 24 species potentially present in the area (Table 1). Most of the remaining species seem to be relatively uncommon, but some may have been missed be- cause of the timing of the counts. This may be the case of Grey-headed Kite (Leptodon cayanensis), which generally displays early in the breeding sea- son (October, according to Vielliard and Silva [2001] and Whitacre pers. comm.). Small size may account for the absence from our list of Rufous- thighed Kite {Harpagus diodon), whieh was relative- ly frequent in some previous surveys. The point- count method may also overlook non-soaring speeies, which would explain why Tiny Hawks, Bat Falcons, forest-falcons (Micrastur sp.) , and Crested Eagles {Morphnus guianensis) were not detected. For these species, the use of acoustical luring or pre-dawn listening may be more effective (Whita- cre and Turley 1990). Comparing all raptor surveys available from the area (Willis and Oniki 1981, Mahosa et al. 1997, Vielliard and Silva 2001; Table 1), the Blaek Vul- ture appears as the most common species. The Roadside Hawk emerged as the second most-fre- quently-detected species in Willis and Oniki (1981) and Vielliard and Silva (2001) studies, instead of the Mantled Hawk, which ranked second in our survey. This differenee may indicate temporal changes in species abundance or, more likely, dif- ferences in the methods and habitats sampled. These differences may also explain the surprising absence of the Ornate Hawk-Eagle from the Willis and Oniki (1981) and Vielliard and Silva (2001) surveys. On the other hand, we were unable to de- tect some species that may be relatively common in the area, such as forest-falcons and the Chiman- go Caracara {Milvago chimachima). Other raptor species we may have missed that have been report- ed from other Atlantic rainforest areas are the Bi- colored Hawk {Accipiter bicolor) (Willis and Oniki 1981), the White-necked Hawk and the Harpy Ea- gle (Harpia harpyja) (Albuquerque 1995). The point count method provides only indices of species abundance rather than information on population densities. Theoretically, these measures should be related, but detectability may differ be- tween species, so comparisons of detection rates are most reasonable between species of similar size and behavior (Thiollay 1989). Different soaring propensities, behavior, and body size of the differ- ent species largely affect detectability during point counts. This is especially true because we used an unlimited observation radius, which favors larger species at the expense of smaller ones. Thus, the point-count method may overestimate the abun- dance of larger and wide-ranging species (eagles, vultures) in relation to those of the smaller raptors (Jullien and Thiollay 1996). Both the Intervales and Petar protected areas may support similar raptor assemblages, although some raptors, in particular the Turkey Vulture, tended to be more abundant in Petar than in In- tervales. The relatively high abundanee of open- country species reveals the effect of human settle- ment in the area and the relative proximity of open habitat near the preserves. In spite of this influence, the Mantled Hawk, an endemic species of the Atlantic rainforest, was found to be common in the area. Hawk-eagles also were common, par- ticularly the Black Hawk-Eagle, a specialist of sec- ondary or disturbed forest patches (Jullien and Thiollay 1996, Thiollay 1999). The Ornate Hawk- Eagle, which is thought to be restricted to little- disturbed forest (Jullien and Thiollay 1996), was less frequently observed than the Black Hawk-Ea- gle. This difference in detections of these two spe- cies may reflect the dominance of late-secondary forest in the area, but also the lower soaring pro- pensity (del Hoyo et al. 1994) and detectability of Ornate Hawk-Eagle, even in areas where it is more abundant than the former species (D. Whitacre pers. comm.). Mean percent presence of hawk-ca- 28 Manosa et al. VoL. 37, No. 1 gles was higher and the presence of Buteo hawks lower in the Paranapiacaba fragment than in sev- eral rainforest areas where raptor counts have been conducted in Central America (Jones and Sutter 1992, Whitacre et al. 1992b). Also, the Black Vulture may be relatively more abundant and Tur- key Vulture less abundant than in the Central American study sites. These differences may indi- cate differences in habitats among areas or reflect different degrees of modification among sites. The results of these counts confirm the impor- tance of the ecological continuum of the Parana- piacaba fragment for the conservation of endan- gered raptors in the Brazilian Atlantic rainforest. The area supports relatively abundant populations of the Mantled Hawk, a poorly known species (Thiollay 1985, lUCN 1990), which is considered endangered (Thiollay 1994) or nearly endangered (Collar et al. 1992, del Hoyo et al. 1994). The rel- atively frequent detections of two other poorly- known species, the Ornate and the Black Hawk- Eagles (Bildstein et al. 1998), also is notable. Particularly as the subspecies of Black Hawk-Eagle, which was documented, is an endemic of the Bra- zilian Atlantic rainforest (Bierregaard 1998). Fur- ther fragmentation of the Paranapiacaba forest may severely impact populations of these poorly- studied species. Although changes on the popula- tions of the rare or less detectable species may not be adequately tracked with this general point- count technique, standardization of the counting method would allow the implementation of moni- toring programs with the objective to detect trends of some of the more detectable species, such as Mantled Hawks and hawk-eagles. Importantly, these species are of conservation concern and may be good indicators of habitat changes in the Atlan- tic rainforest relevant to the entire raptor assem- blage. AcKNOWLEDGMENI'S This work is dedicated to our late friend Carles Lopez Adzerias. We also wish to remember Roberto Brirgi, for- mer director of the Parque Estadual Turistico do Alto Ribeira, who died in an accident during this study. We would like to thank Dr. Juan Carlos Guix, Dr. Antoni Ser- ra, and Kattia Pisciotta for making this expedition possi- ble, as well as the staff of Parque Estadual Intervales and Parque Estadual Turistico do Alto Ribeira for all the fa- cilities they provided. Flavia de Campos Martins, Blanche Sousa Pinto, and Montse Ontanon assisted during the field work. The comments of Dr. D.F. Whitacre, Dr. K.L. Bildstein, and an anonymous referee very much im- proved this manuscript. Literature Cited Albuquerque, J.L.B. 1986. Conservation and status of raptors in southern Brazil. Birds Prey Bull. 3:88-94. . 1995. Observations of rare raptors in southern Atlantic rainforest of Brazil. J. Field Ornithol. 66:363- 369. Bibby, C.J., N.D. Burgess, and D.A. Hill. 1992a. Bird census techniques. BTO & RSPB. Univ. Pre.ss, Cambridge, U.K , NJ. Collar, M.J. Crosby, M.F. Heath, C.H. Im- BODEN, T.H. Johnson, AJ. Long, A.J. Sattersfield, and S.J. Thirgood. 1992b. Putting biodiversity on the map: priority areas for global conservation. ICBP, Cambridge, U.K. Bierregaard, Jr., R.O. 1998. Conservation status of birds of prey in the South American tropics. J. Raptor Res. 32:19-27. Bildstein, K.L., W. Schelsky, and J. Zallf.s. 1998. Con- servation status of tropical raptors./. Raptor Res. 32:3- 18. Bruce, R, J.L. Simon, and T. Oswald. 1995. Resampling stats for DOS. Operation guide. Resampling Stats Inc., Arlington, VA U.S.A. Burnham, W.A., D.F. Whitacre, and J.P. Jenny. 1994. The Maya project: use of raptors as tool for conservation and ecological monitoring of biological diversity. Pag- es 257-264 in B.-U. Meyburg and R.D. Chancellor [Eds.], Raptor conservation today. WWGBP and Pica Press, Near Robertsbridge, U.K. Carvalho, M.P., C.H.G. Cezare, C.G. Costa, R.G.T. Da Cunha, S.S. Dornelles, M.S. Filhao, J.C. Guix, A. Hernandez, PC. Lazarin, L, It.orens, E. Mateos, C Miquel, L.M. Petroni, B.S. Pinto, K. Pisciotta, C. Sanchez, I. Oliveras, and A. Serra. 2002. Description of the study areas. Pages 27-48 in E. Mateos, J.C Guix, A. Serra, and K. Pisciotta [Eds.], Censuses of vertebrates in a Brazilian Atlantic rainforest area: the Paranapiacaba forest fragment. Centre de Recursos de Biodiversitat Animal. Universitat de Barcelona, Barcelona, Spain. Coliar, N.J., L.P. Gonzaga, N. Krabble, A. Madrono- Nieto, L.G. Naranjo, T.A. Parker, and D.C. Wege 1992. Threatened birds of the Americas. The ICBP/ lUCN Red Data Book. ICBP, Cambridge, U.K. Cracraft,J. 1985. Historical, biogeography, and patterns of differentiation within the South American avifau- na: areas of endemism. Pages 49-84 in P.A. Buckley, M.S. Foster, E.S, Morton, R.S. Ridgely, and FG. Buck- ley [Eds.], Neotropical ornithology. The American Ornithologists’ Union, Washington, DC U.S.A. del Hoyo, j., A. Elliott, and J. Sargatal (Eds.). 1994. Handbook of the birds of the world. Vol. 2. New world vultures to Guineafowl. I.ynx Ed., Barcelona, Spain. Fonseca, G.A.B. 1985. The vanishing Brazilian Atlantic forest. Biol. Conserv. 34:17-34. Forsman, D. and T. Solonen. 1984. Censusing breeding March 2003 Soaring Raptors in Atiantic Rainforest 29 raptors in southern Finland: methods and results. Ann. Zool. Fenn. 21:317—320. Guix, J.C. 2002. Introduction and objectives. Pages 3-18 in E. Mateos, J.C. Guix, A. Serra, and K. Pisciotta [Eds.], Censuses of vertebrates in a Brazilian Atlantic rainforest area: the Paranapiacaba forest fragment. Centre de Recursos de Biodiversitat Animal. Univer- sitat de Barcelona, Barcelona, Spain. , A.A.J. Tabanez, A.N. Silva, C. Lopez, C. Marti- nez, E. Matheu, F.L. De Souza, K.R. Pisciotta, N. Bradbury, and W.G. Portilho. 1992. Viagem de re- conhecimento cientifico a algumas areas desconheci- das da Fazenda Intervales, Estado de Sao Paolo, dur- ante o periodo de 04 a 16 outubro de 1991. Grupo Estud. Ecol. Sen Doc. 4:38-94. lUCN. 1990. lUCN Red List of threatened animals. lUCN, Cambridge, U.K. Jones, L.E. and J. Sutter. 1992. Resultados y compara- cion de dos ahos de trabajos de censo en tres unida- des del complejo Reserva de la Biosfera Maya/Reser- va de Calakmul. Pages 69-87 in D.F. Whitacre and R.K. Thorstrom [Eds.], Proyecto Maya: uso de aves rapaces y otra fauna como indicadores del medio am- biente, para el diseno y manejo de areas protegidas y para fortalecer la capacidad local para la conservacion en America Latina. The Peregrine Fund, Inc., Boise, ID U.S.A. Jullien, M. and J.-M. Thiollay. 1996. Effects of rain for- est disturbance and fragmentation: comparative changes of the raptor community along natural and human-made gradients in French Guiana. J. Biogeogr. 23:7-25. Manosa, S. and V. Pedrocchi. 1997. A raptor survey in the Brazilian Atlantic rainforest./. Raptor Res. 31:203-207. , E. Mateos, and V. Pedrocchi. 2002. Birds of prey survey (aves: cathartiformes and accipitriformes) in the Paranapiacaba forest fragment. Pages 165-179 in E. Mateos, J.C. Guix, A. Serra, and K. Pisciotta [Eds.], Censuses of vertebrates in a Brazilian Atlantic rainforest area: the Paranapiacaba forest fragment. Centre de Recursos de Biodiversitat Animal. Univer- sitat de Barcelona, Barcelona, Spain. Mateos, E. and S. Manosa. [Eds.] 1996. Memoria de re- sultados del segundo viaje de reconocimiento cienti- fico a regiones poco conocidas de la Fazenda Inter- vales, estado de Sao Paulo, Brasil. Publicaciones de la Universidad de Barcelona, Barcelona, Spain. , J.C. Guix, A. Serra, and K. Pisciotta. [Eds.] 2002. Censuses of vertebrates in a Brazilian Atlantic rainforest area: the Paranapiacaba fragment. Centre de Recursos de Biodiversitat Animal. Universitat de Barcelona, Barcelona, Spain. Myers, N. 1988. Threatened biotas: “hotspots” in tropi- cal forests. Environmentalists , R.A. Mittermeier, C.G. Mittermeier, G.A.B. Da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403:853-858. Pisciotta, K. 2002. The Paranapiacaba forest fragment Pages 19-25 in E. Mateos, J.C. Guix, A. Serra, and K. Pisciotta [Eds.], Censuses of vertebrates in a Brazilian Atlantic rainforest area: the Paranapiacaba forest frag- ment. Centre de Recursos de Biodiversitat Animal Universitat de Barcelona, Barcelona, Spain. Stattersfiei.d, A.J., M.J. Crosby, A.J. Long, and D.C. Wege. 1998. Endemic bird areas of the world: priori- ties for biodiversity conservation. BirdLife Interna- tional, Cambridge, U.K. Thiollay, J.-M. 1985. Falconiforms of tropical rain forest' a review. Pages 155—165 in I. Newton and R.D. Chan- cellor [Eds.], Conservation studies on raptors. Vol. 5. ICBP Tech. Publ., Cambridge, U.K. . 1989. Censusing of diurnal raptors in a primary rain forest: comparative methods and species detect- ability. /. Raptor Res. 23:72-84. . 1994. A world review of tropical forest raptors. Current trends, research objectives, and conservation strategy. Pages 231-239 in B.-U. Meyburg and R.D. Chancellor [Eds.], Raptor Conservation Today. Pica Press, Berlin, Germany. . 1999. Response of an avian community to ram forest degradation. Biodiv. Conserv. 8:513-534. Vannini, J.P. 1989, Neotropical raptors and deforestation notes on diurnal raptors at finca El Faro, Quetzalten- ango, Guatemala./. Raptor Res. 23:27-38. Vielliard, J.M.E. and W.R. Sii.va. 2001. Avifauna. Pages 125-145 in C. Leonel [Ed.], Intervales. Fundagao para a conserva^ao e a produfao florestal do estado de Sao Paulo. Secretaria de esatdo do Meio Ambiente, Governo de Estado de Sao Paulo, Brazil. Wege, D.C. and A.J. Long. 1995. Key areas for threatened birds in the neotropics. BirdLife Conserv. Ser. 5, BirdLife International, Cambridge, U.K, Whitacre, D.F. and C.W. Turley. 1990. Further compar- isons of tropical forest raptor census techniques. Pag- es 71-92 in W.A. Burnham, D.F. Whitacre, and J.P. Jenny [Eds.], Progress Report III. Maya Project: use of raptors as environmental indices for design and management of protected areas and for building local capacity for conservation in Latin America. The Per- egrine Fund, Inc., Boise, ID U.S.A. and R.K. Thorstrom. [Eds.] 1992. Proyecto Maya: uso de aves rapaces y otra fauna como indica- dores del medio ambiente, para el diseno y manejo de areas protegidas y para fortalecer la capacidad lo- cal para la conservacion en America Latina. The Per- egrine Fund, Inc., Boise, ID U.S.A. , L.E. Jones, and J. Sutter. 1992a. Gensos de aves rapaces y de otras aves en el bosque tropical: mejoras hechas a la metodologia. Pages 43-56 in D.F. Whitacre and R.K. Thorstrom [Eds.], Proyecto Maya: uso de aves rapaces y otra fauna como indicadores del medio ambiente, para el diseno y manejo de areas protegidas y para fortalecer la capacidad local para la conserva- 30 Manosa et ae. VoL. 37, No. 1 cion en America Latina. The Peregrine Fund, Inc., Boise, ID U.S.A. , C. Turiey, A.E, Hernandez, and F. Osorio. 1992b. Una cotnparacion de comunidades de aves ra- paces que habitan bosques tropicales primarios y mo- saicos de ranchos agropecuarios que utilizan la tec- nica de Tala y quema del bosque. Analisis de los datos de 1989. Pages 89-101 in D.F. Whitacre and R.K. Thorstrom [Eds.], Proyecto Maya: uso de aves rapaces y otra fauna como indicadores del medio ambiente, para el diseho y manejo de areas protegidas y para fortalecer la capacidad local para la conservacion en America Latina. The Peregrine Fund, Inc., Boise, ID U.S.A. Wll.iJS, E.O. AND Y. Oniki. 1981. Levantamento prelimi- nar de aves en treze areas do estado de Sao Paulo Rev. Bras. Biol. 41:121-135. Received 14 March 2002; accepted 5 October 2002 Associate Editor: Marco Restani J. Raptor Res. 37(l);31-36 © 2003 The Raptor Research Foundation, Inc. DIFFERENTIAL SPRING MIGRATION OF ADULT AND JUVENILE LEVANT SPARROWHAWKS {ACCIPITER BREVIPES) THROUGH EILAT, ISRAEL Reuven Yosee^ International Birding and Research Centre in Eilat, Department of Life Sciences, Ben-Gurion University of the Negev, RO. Box 774, Eilat 88000, Israel Lorenzo Fornasari Department of Environmental Sciences, University of Milano Bicocca, Piazza della Scienza 1, 1-20126, Italy PlOTR TrYJANOWSKI Department of Avian Biology and Ecology, Adam Mickiewicz University, Eredry 10, PL-61-701 Poznan, Poland Marc J. Bechard and Gregory S. Kaltenecker Idaho Bird Observatory, Department of Biology, Boise State University, 1910 University Drive, Boise, ID 83725 US. A Keith Bildstein Hawk Mountain Sanctuary, 1 700 Hawk Mountain Road, Kempton, PA 19529 U.S.A. Abstract.- — ^As many as 50 000 Levant Sparrowhawks (Accipiter brevipes) are counted during migration at the northern end of the Gulf of Aqaba each spring. We present data from 1819 migrants that were captured and ringed at Eilat, Israel: 459 from 1984-88, 21 from 1989-95, and 1345 captured from 1996- 2000. Of these, 396 (22%) were adult females, 631 (35%) were adult males, 359 (20%) were juvenile females, and 433 (24%) were juvenile males. We compare migration timing and body sizes in juvenile (i.e., first-time spring migrants) and adult migrants, and in males and females. Wing chord length and body mass in males and females changed significantly with date of arrival. Further, a significant corre- lation was found for both sexes between wing chord length and body mass in spring. Within age classes, both wing chord and body mass declined significantly with date of ringing. Body mass was also signifi- cantly related to size obtained from PCA analyses (PCI), both in males and females. We computed also standardized residuals of body mass on PCI. Date of passage was also significantly correlated to the standardized residuals, both in males and females. This suggested, testing for allometry vs. isometry, that birds in better than expected ‘condition’ migrated earlier. Moreover, results from analysis of vari- ance revealed that body mass and age were significantly related to the date of passage. The median date of passage for adults preceded that of juveniles by 2.5 days. We believe juveniles on their first spring passage migrate slower than adults and that they are more likely to be later and in poorer body con- dition. Key Words: Levant Sparrowhawk, Accipiter brevipes; age, Eilat, Israel] sex; spring migration. MIGRACION DIFERENCIAL DE PRIMAVERA ENTRE ADULTOS JUVENILES DEL AZOR DEL MED- ITERRANEO oriental {ACCIPITER BREVIPES) A TRAVES DE EILAT, ISRAEL Resumen. — Tantos como 50 000 azores del mediterraneo oriental {Accipiter brevipes) son contados dur- ante su migracion en el limite del golfo de Aqaba cada primavera. Presentamos datos de 1819 emigrantes que fueron capturados y anillados en Eilat, Israel: 459 de 1984—88, 21 de 1989—95, y 1345 capturados entre 1996-2000. De estos, 396 (22%) fueron hembras adultas, 631 (35%) eran machos adultos, 359 (20%) hembras juveniles, y 433 (24%) machos juveniles. Comparamos el tiempo de migracion y los tamahos del cuerpo en emigrantes juveniles (v.gr., azores migrantes que lo hacfan por primera vez en primavera) y adultos, y en machos y hembras. La longitud de la cuerda del ala y la masa del cuerpo en ^ E-mail address: ryosef@eilatcity.co.il 31 32 Yosef et al. VoL. 37, No. 1 machos y hembras cambio significativamente con la fecha de arribo, Por otro lado, se encontro una correlacion significativa para ambos sexos entre la longitud de la cuerda del ala y la masa corporal en primavera. Dentro de las clases de edad, tanto la longitud de la cuerda del ala como la masa del cuerpo declino significativamente con la fecha del anillado. Las masas corporales ademas estuvieron relacion- adas significativamente al tamano obtenido mediante analisis PCA (PCI), tanto en machos como hem- bras. Calculamos ademas residues estandarizados de masa corporal en PCI. La fecha de paso estuvo tambien correlacionada significativamente con los residues estandarizados, tanto en machos como en hembras. Esto sugiere, haciendo pruebas de alometria vs. isometria, que las aves en mejor “condicion” que la esperada, migraban mas temprano. Por otra parte, los resultados del analisis de varianza revelan que la masa del cuerpo y la edad estuvieron relacionadas significativamente con la fecha de paso. La fecha promedio para los adultos precedia a la de los juveniles en 2.5 dias. Creemos que los juveniles en su primera pasada de primavera migran mas lento que los adultos y que son mas propensos a estar retrasados y en condiciones corporales mas pobres. [Traduccion de Cesar Marquez] The Levant Sparrowhawk {Accipiter brevipes) breeds principally within the western Palearctic Re- gion, in southeast Europe, locally through Turkey to northern Iran, and is widespread in southwest Russia and Kazakhastan (Snow and Perrins 1998). It winters in the east Sahel of sub-Saharan Africa, and sporadic reports are received also from the Ethiopian highlands and the southern Arabian Peninsula (Snow and Perrins 1998, Shirihai et al. 2000) . Principal migration routes lie entirely with- in the Middle East (Frumkin et al. 1995, Shirihai et al. 2000) with especially large concentrations found at Eilat during spring (Shirihai and Christie 1992, Yosef 1995). In Israel, the Levant Sparrowhawk is an abun- dant migrant in both spring and autumn and about 90% of the world population of the species passes through Israel within a short period of a fortnight (Shirihai et al. 2000), and this is the only raptor known to migrate at night in the region (Stark and Liechti 1993). Visible migration surveys conducted since 1977 suggest that Eilat is an im- portant stopover site for the species in spring. Eilat is at the northern edge of the Sahara and Sinai deserts, and in spring many northbound migrants stop there to rest and feed (Safriel 1968, Yosef 1996a). Levant Sparrowhawks were trapped and ringed at and around Eilat during spring (mid- April through early May) from 1984—2000 (Clark and Yosef 1997, Yosef and Fornasari 2000). In this paper, we present morphometric data collected af- ter capture to compare age- and sex-related differ- ences in body size and migration timing in this spe- cies. Study Area and Methods Levant Sparrowhawks were captured and ringed at and around Eilat, Israel (29°33'N, 34°57'E), using bow-nets, mist nets, dho-gazas, box traps, and Bal-chatri traps (Clark and Yosef 1997). Al data were pooled because 63% of Levant Sparrowhawks were caught in 7-m-high mist nets in the early mornings when the flocks started the day’s migration. We assumed that these birds repre- sent a sample of the general population because the ma- jority was not trapped with food as bait (Gorney et al 1999). All captured individuals were aged, sexed, mea- sured (including unflattened wing chord) , and weighed Aging was based on plumage, molt, and iris color (Clark and Yosef 1998). In addition, birds divided into two cat- egories: juveniles — in second calendar year (SY) and adults — after second calendar year (ASY) . Data presented in this paper were collected during 1984-88 (Gorney and Yom Tov 1994, Gorney et al. 1999), 1989-95 (Yosef and Fornasari 2000), and 1996-2000 (Clark and Yosef 1997, Shirihai et al. 2000) . We excluded from the analyses five birds (two identified as female and three as male) be- cause according to the biometrics we consider them to have been sexed incorrectly. Owing to differences in weather and other local con- ditions which influence the phenology of migration (me- dian test, = 294.5, df = 13, P < 0.0001), data were standardized between years. By computing the median date of passage for each year, data were collected. Then the dates of all captured hawks were transformed as val- ues before (minus) or after (plus) the median. All the basic statistics were performed according to So- kal and Rohlf (1995). We computed correlation between birds’ measurements and timing of migration (in Julian dates). However, because in accipiters wing chord is an indicator of body size (e.g., Mueller and Meyer 1985, Wyl- lie and Newton 1994), we established a condition index. We used a total body-size measurement, which was ob- tained from the principal component analysis with VAR- IMAX rotation of four (wing chord, culmen, hallux, and tail length) log-transformed, external measurements (Piersma and Davidson 1991) computed separately for the sexes. All morphometric variables had positive and a similar magnitude of loading on the first component (0.318-0.742, eigen value = 1.811, 45.3% of total vari- ance explained for females and 0.234-0.636; eigen value = 1.577, 39,4% of total variance explained for males) We analyzed structural size (PCI) /mass relationships sep- arately within four age/sex classes to test for allometry or isometry. March 2003 Spring Migration of Levant Sparrowhawks 33 240 -| 220 - ® 200 I 180 s ^ 160 - o pa 140 - 120 - 100 -25 (A) After-second-year males y = -0.3255X + 167.85 . P = 0.058 -15 -5 5 15 25 Standarized Year Value (Median = 0) 240 220 - $200 I 180 S 160 O 140 - 120 100 -- -25 (B) Second-year males -15 -5 5 15 25 Standarized Year Value (Median = 0) Figure 1. Body mass of (A) after-seeond-year (ASY) and (B) second-year (SY) male Levant Sparrowhawk in rela- tion to date of capture and passage at Eilat. Regression line is non-significant for SY males. 260 n 240 - 220 r 200 I 180 - 1 “ 1 60 - PP 140 - 120 ^ 100 — -25 (B) Second-year females -15 -5 5 15 25 Standarized Year Value (Median = 0) Figure 2. Body mass of (A) after-second-year (ASY) and (B) second-year (SY) female Levant Sparrowhawk in re- lation to date of capture and passage at Eilat. Regression line is non-significant for SY females. Results A total of 1819 migrant Levant Sparrowhawks was captured; 459 in 1984-88, 21 in 1989-95, and 1345 in 1996-2000. Of these, 396 (22%) were adult females, 631 (35%) were adult males, 359 (20%) were juvenile females, and 433 (24%) were juvenile males. Sex ratio among juvenile birds differed sig- nificantly from 1:1 (y2 = 27.4, P< 0.0001), but was not significantly different in adults (x^ = 3.6, P = 0.06). Hence, sex ratio between the two age cate- gories differed significantly (x^ = 8.4, P = 0.004). Each spring, transients were trapped in Eilat be- tween early March and early May. The two sex and two age groups differed significantly in their me- dian time of migration (SY females median date = 27 April, range = 10 March-15 May; ASY females median date = 24 April, 18 March-16 May; SY males, 27 April, 18 March-21 May; ASY males, 25 April, 28 March-12 April; Rruskal-Wallis test, x^ = 142.36, df = 3, P < 0.0001). A post-hoc Ty\mn' % Q: test (P < 0.05) demonstrated that significantly more ASY individuals of both sex migrate earlier than SY birds. Wing chord length in males (r^oso — —0.207, P < 0.0001) and females (^744 = —0.127, P= 0.001) changed significantly with date of arrival, as did body mass in males {r^Q 22 ~ ^0.141, P < 0.0001; Fig. 1) and females (^740 = —0.137, P < 0.0001; Fig. 2) changed significantly with date of arrival. Further, a significant correlation was found for both sexes between wing chord length and body mass in spring (males y = 0.140 (±0.011) x + 0.367, t — 12.61, P < 0.0001 versus females y = 0.128 (±0.011) X + 0.380, t = 11.17, P < 0.0001). In addition, we found a significant correlation be- tween mass (log transformed before analyses) and wing chord length for juveniles (riooe ~ 0.625, P < 0.0001) and adults (^ 7^3 = 0.665, P< 0.0001). How- ever, there was no statistical difference between the correlation values for the two-age classes (P = 0.152). Body mass was also significantly correlated with size obtained from PGA analyses (PCI), both in males: r 5 g 7 = 0.368, P < 0.0001 and females: ^740 = 34 Yosef et al. VoL. 37, No. 1 Table 1 . Analysis of Variance analyses of biometric val- ues of Levant Sparrowhawks related to migration passage at Eilat, Israel. Source of Variation Sum of Squares df F P Covariates 166.525 2 3.978 0.019 Body mass 134.968 1 6.448 0.011 Wing chord 4.626 1 0.221 0.638 Main effects 139.636 2 51.107 <0.001 Age 1755.369 1 83.856 <0.001 Sex 0.147 1 0.007 0.933 2-way interactions Age*Sex 30.493 1 1.457 0.228 Explained 3301.656 5 31.545 <0.0001 Residual 37 218.904 1778 20.933 Total 40 520.560 1783 22.726 0.330, P< 0.0001. We computed also standardized residuals of body mass on PCI. Date of passage was significantly correlated to the standardized residu- als, both in males: r 59 g = —0.154, P < 0.0001 and females: = —0.110, P = 0.040. This suggested, testing for allometry versus isometry, that birds in better than expected ‘condition’ migrated earlier. Moreover, results from Analysis of Variance re- vealed that only body mass and age were signifi- cantly related to the date of passage (Table 1). Discussion In many raptors, adults migrate earlier in spring than do juveniles (Newton 1979, Gorney and Yom- Tov 1994, Yosef et al. 2002). With an overall 10-yr median trapping date of 25 April for adults versus 27 April Juveniles, our results, which extend an ear- lier 5-yr study of Gorney et al. (1999), confirm that Levant Sparrowhawks in Israel, too, exhibit age-re- lated differences in the timing of migration. Al- though age differences in raptor migration are not completely understood, previous work suggests that such differences occur because breeding pres- sures on adults select for earlier arrival on the breeding grounds (Newton 1979, Gorney and Yom- Tov 1994), juveniles require more time either to initiate or complete their journeys, or both (Gor- ney and Yom-Tov 1994, Gorney et al. 1999) or al- ternatively, juveniles may over-winter farther from their breeding grounds than do adults (Krol 1983). Gorney and Yom-Tov (1994) argued that earlier passage of adult steppe Common Buzzards {Buteo buteo vulpinus) at Eilat suggested that adults were “time selected” migrants, whereas juveniles were “energy-selected” migrants. Because most juvenile birds do not breed, they would not need to reach their “breeding grounds” as early in spring as adults. We offer another explanation: that adults precede juveniles because they are better prepared and more efficient at migration, therefore, are more capable and faster migrants en route. Little is known about the Levant Sparrowhawks on their breeding grounds in Eurasia, and the species has yet to be studied in detail on its wintering grounds in Africa (Shirihai et al. 2000). The fact that adults arrive in Eilat only a few days earlier than juveniles and because adults are heavier than juveniles, in our view, suggests that adults are more capable mi- grants, rather than that the age classes are using different migration strategies. However, it is also possible that age classes winter in separate regions, at different distances from Eilat, resulting in dis- crepancy in phenology between the two classes. However, the latter cannot be verified at present owing to the lack of data and observations for the species from the wintering grounds. Another pos- sible explanation is that the different age classes may have different migration strategies because adults have longer wings and tails in the spring (Yosef and Eornasari 2000). Proportionally longer wing and tail length allow for a greater proportion of time spent in soaring flight, which is in contrast to juveniles who have shorter wing and tail which requires comparatively more flapping flight that requires greater fat reserves and better body con- dition. Soaring migrants such as Levant Sparrowhawks (Spaar 1997, Spaar et al. 1998) typically travel in large flocks, presumably so that individuals can more quickly locate thermals needed to assist their long-distance movements (Kerlinger 1989). Obser- vations of Broad-winged Hawks {Buteo platypterus) in North America suggest that juveniles are more likely to be wind drifted and blown off course than are adults (Hagar 1988, Hoffman and Darrow 1992). It is thought that coastal raptor migrations consist primarily of juveniles of all species because of this fact (Kerlinger 1989). We propose that ju- venile Levant Sparrowhawks pass through Eilat lat- er than adults because they are less efficient mi- grants as shown by their lower body weights and lower ‘condition’ (Clark and Yosef 1997, Gorney et al. 1999). Although Gorney et al. (1999) found no signif- icant associations between condition indices of March 2003 Spring Migration of Levant Sparrowhawks 35 adults and juveniles, we believe that large numbers of Levant Sparrowhawks, particularly immatures, reach Eilat in poor body condition. Many raptors ringed at Eilat have tar and oil residues on their feathers and feet (Clark and Gorney 1987). Per- haps they mistakenly land in oil fields of the Sahara and Sinai deserts in search of fresh water during migration flights (Clark and Gorney 1987). Data collected from 1996-2000 indicate that 81% {N — 43) of oil-contaminated Levant Sparrowhawks were juveniles. We assume that juveniles are more stressed during migration, and therefore, more likely to seek drinking water. Gorney et al. (1999) found no significant asso- ciations within age and sex groups between date of migration and physical condition. In contrast, our study shows that birds with longer wing chords and greater body mass passed through the area earlier than smaller individuals of the same sex and age class. In this study, we find that wing chord was significantly correlated with body mass, suggesting that while on migration, larger birds are heavier. Gorney and Yom-Tov (1994) suggest that the large proportion of immatures ringed at Eilat may have resulted from age differences in migration routes as has been reported for other raptor species (Bild- stein et al. 1983, Yosef 1996b, Yosef and Alon 1997). No evidence suggests that adult Levant Sparrowhawks follow a different migration route than juveniles (Shirihai and Christie 1992, Shirihai 1996, Shirihai et al. 2000). Rather, geography of the region (Shirihai et al. 2000, Zalles and Bild- stein 2000) suggests that northern end of the Gulf of Eilat serves as a concentration point for many western Palearctic migrating raptor species during spring, regardless of age or sex (Spaar et al. 1998). We think that the limited sample size included by Gorney and Yom-Tov (1994) and Gorney et al. (1999) in their study may have led them to conclu- sions not supported by our data set, which includes more than double in the number of birds involved in the previous analyses. In conclusion, the fact that within sex and age classes, heavier and better ‘condition’ individuals are trapped early in the season suggests that in both juveniles and adults, the more efficient mi- grants pass earlier than less efficient migrants, and that adults are more efficient than juveniles. Acknowled gments We thank the many organizations and individuals who have helped the International Birding and Research Cen- tre through the years, W.S. Clark, Dayton Baker of the National Aviary in Pittsburgh, World Wildlife Fund-Inter- national, Sir and Lady Kaye, Mr. and Mrs. Speers, Dr Gerold Dobler and Swarovski Optics, Earthwatch Insti- tute, and Rafi Saar and Montal of Kibbutz Eilot. We thank Mick Marquiss and three anonymous referees for com- ments on an earlier version of the manuscript. Literature Cited Bildstein, K.L., W.S. Clark, D.L. Evans, F. Marshall, L SouCY, and E. Henckel. 1983. Sex and age differenc- es in fall migration of Northern Harriers. /. Field Or- nithol. 55:143-150. Clark, W.S. and E. Gorney. 1987. Oil contamination of raptors migrating along the Red Sea. Environ. Pollul 46:307-313. AND R. Yosef. 1997. Migrant Levant Sparrow- hawks {Acdpiter brevipes) at Eliat, Israel: measurements and timing./. Raptor Res. 31:317—320. AND . 1998. In-hand identification guide to palearctic raptors. International Birdwatching Centre in Eilat. Tech. Publ. 7, Eilat, Israel. Frumkin, R., B. Pinshow, and S. Klelnhaus. 1995. A re- view of bird migration over Israel./. Ornithol. 136:127- 147. Gorney, E. and Y Yom-Tov. 1994. Fat, hydration condi- tion, and moult of steppe buzzards Buteo buteo vulpinus on spring migration. Ibis 136:185-192. , W.S. Clark, and Y. Yom-Tov. 1999. A test of the condition-bias hypothesis yields different results for two species of sparrowhawks {Acdpiter). Wilson Bull 111:181-187. Hagar, J.A. 1988. Broad-winged Hawk: migration. Pages 1-25 m R.S. Palmer [Ed.], Handbook of North Amer- ican birds. Vol. 5. Yale Univ. Press, New Haven, CT U.S.A. Hoffman, W. and H. Darrow. 1992. Migration of diurnal raptors from the Florida Keys into the West Indies. HNIANA Hawk Migration Stud. 17:7—14. Kerlinger, P. 1989. Flight strategies of migrating hawks Univ. of Chicago Press, Chicago, IL U.S.A. Krol, W. 1983. Bird ringing results in Poland. Migration of the buzzards Buteo buteo buteo. Acta Ornithol. 19:137- 151. Mueller, H.C. and K. Meyer. 1985. The evolution of re- versed sexual dimorphism in size — a comparative analysis of the Falconiformes of the western palearc- tic. Curr. Ornithol. 2:65-101. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A. PiERSMA, T. and N.C. Davldson. 1991. Confusion of mass and size. Auk 108:441-444. Safrlel, U. 1968. Bird migration at Eilat, Israel. Ibis 110: 283-320. Shirillai, H. 1996. The birds of Israel. Academic Press, London, U.K. and D. Christie. 1992. Raptor migration at Eilat. Br. Birds 85:141-186. 36 Yosef et al. VoL. 37, No. 1 , R. Yosef, D. Ai.on, G. Kirwan, and R. Spaar. 2000. Raptor migration in Israel and the Middle East — a summary of 30 years of field research. Inter- national Birding & Research Centre in Eilat, Eilat, Is- rael. Snow, D.W. and C.M. Perrins. [Eds.] 1998. The birds of the western palearctic. Vol. 1. Non-passerines. Oxford Univ. Press, Oxford, U.K. SoKAl,, R.R. AND F. J. Rohlf. 1995. Biometry, 3rd Ed, Freeman Press, New York, NY U.S.A. Spaar, R. 1997. Flight strategies of migrating raptors: a comparative study of interspecific variation in flight characteristics. Ibis 139:523-535. , H. Stark, and F. Liechti. 1998. Migratory flight strategies of Levant Sparrowhawks: time or energy minimization? Anim. Behav. 56:1185—1197. Stark, H. and F. Liechti. 1993. Do Levant Sparrowhawks Accipiter brevipes also migrate at night? Ibis 135:233- 236. Wyllie, I. and I. Newton, 1994. Latitudinal variation in the body-size of Sparrowhawks Accipiter nisus within Britain. Ibis 136: 434-440. Yosef, R. 1995. Spring 1994 raptor migration at Eilat, Israel./. Raptor Res. 29:127-134. . 1996a. Raptors feeding on migration at Eilat, Is- rael: opportunistic behavior or migration strategy? / Raptor Res. 30:242-245. . 1996b. Sex and age classes of migrating raptors during the spring of 1994 at Eilat, Israel./. Raptor Res 30:160-164. . and D. Aeon. 1997. Do immature palearctic Egyptian Vultures Neophron percnopterus remain in Af- rica during the northern summer? Vogelwelt 118:285- 289. . AND L. Fornasari. 2000. Biometric differences between age and sex classes of the Levant Sparrow- hawk Accipiter brevipes on migration at Eilat, Israel. Is- rael J. Zool. 46:207-214. , P. Tryjanowski, and K.L. Bildstein. 2002. Spring migration of adult and immature Buzzards {Buteo bu- teo) through Eilat, Israel: timing and body size./. Rap- tor Res. 36:115-120. Zalles, JJ- AND K.L. Bildstein, (Eds.) 2000. Raptor watch: a global directory of raptor migration sites. BirdLife Conservation Series No. 9. BirdLife Inter- national, Cambridge, U.K. and Hawk Mountain Sanc- tuary. Kempton, PA U.S.A. Received 19 December 2001; accepted 9 September 2002 J. Raptor Res. 37(1 ) ;37— 43 © 2003 The Raptor Research Foundation, Inc. VOCAL DEVELOPMENT IN AMERICAN KESTREL {FALCO SPARVERIUS) NESTLINGS John A. Smallwood^ and Valerie Dudajek Department of Biology and Molecular Biology, Montclair State University, Upper Montclair, NJ 07043 US. A. SiVAjiNi Gilchrist Newark Museum, 49 Washington Street, Newark, NJ 07101 US. A. Mary Anne Smallwood Ironia School, Randolph, NJ 07869 US. A. Abstract. — We studied the acoustical characteristics of calls made by nestling American Kestrels {Falco sparverius). A total of 563 vocal samples was obtained from 88 chicks (49 males and 39 females) from 20 broods. Thirteen frequency, three numerical, and two temporal characteristics were measured using audio spectrography. Discriminant function analysis failed to distinguish the calls of male and female chicks, but univariate and principal component analyses suggest that vocal ontogeny proceeds more rapidly in males than in females. The acoustical characteristics of call notes changed in a consistent manner as nestlings matured, and by day 16 chicks produced calls similar to those of adults. Key Words: American Kestrel', Falco sparverius; vocalization', development. DESARROLLO vocal EN POLLUELOS del CERNIGALO {Falco sparverius) Resumen. — Estudiamos las caracteristicas acusticas de las vocalizaciones hechas por polluelos de cerni- calo {Falco sparverius). Un total de 563 muestras de vocalizaciones fueron obtenidas de 88 polios (49 machos y 39 hembras)de 20 nidadas. Trece caracteristicas de frecuencia, tres numericas y dos temporales fueron medidas usando audio espectrografia. El analisis de la funcion discriminante fallo para distinguir los llamados de los polios hembras y machos, pero el analisis univariado y de componentes principales sugiere que la ontogenia vocal precede mas rapidamente en machos que en hembras. Las caracteristicas acusticas de las notas de los llamados, cambiaron de manera consistente con la madurez de los polluelos, y para el dia 16 los polios produjeron llamados similares a los de los adultos. [Traduccion de Cesar Marquez] Although the acquisition of species-specific song has been studied extensively in passerines (e.g., Kroodsma and Miller 1982), much less is known about the development of vocal behavior in non- passerines. The American Kestrel {Falco sparverius) has a simple vocal repertoire, consisting of three main calls in adults: klee, whine, and chitter, and combinations thereof (Willoughby and Cade 1964). Vocalizations of nestlings have been de- scribed (Sherman 1913, Roest 1957, Balgooyen 1976, Smallwood and Bird 2002), but only quali- tatively. Audio spectrography facilitates quantifica- tion of acoustical signals, and the variables derived from spectrographs may be analyzed with both uni- ^ E-mail address: smallwoodj@mail.montclair.edu variate and multivariate statistical treatments. For example, discriminant function analysis was used to distinguish calls among four species of penguins (Thumser et al. 1996) and principal component analysis was used to identify individuals within a flock of Greater Flamingos {Phoenicopterus ruber, Mathevon 1997). Multivariate techniques also have been used to determine gender in birds whose vo- calizations are not readily discernable by human observers (e.g.. Whooping Cranes, Grus americana; Carlson and Trost 1992). No differences in vocali- zations of male and female American Kestrels, ei- ther adults or young, have been reported. The ob- jectives of this study were to analyze acoustical characteristics of calls made by nestling kestrels to (1) determine if gender can be distinguished vo- cally and (2) examine ontogenic changes. 37 38 Smallwood et al. VoL. 37, No. 1 Table I. Principal component analysis for frequency variables used to describe acoustical characteristics of nestling American Kestrel calls. See methods for descrip- tions of variables. Eigenvectors Variable PGl PC2 Low mean frequency 0.357 -0.150 End mean frequency 0.340 -0.161 End maximum frequency 0.322 -0.194 Low maximum frequency 0.316 -0.207 Dominant harmonic 0.306 -0.115 Maximum frequency 0.300 -0.010 High mean frequency 0.283 0.335 High maximum frequency 0.270 0.347 Initial mean frequency 0.256 0:210 Low 25% frequency 0.238 -0.304 High 75% frequency 0.220 0.373 Initial maximum frequency 0.213 0.218 Frequency range -0.044 0.549 Study Area The study area was in rural northwestern New Jersey, bordered to the north and west by the Kittatinny Ridge and Delaware River, and to the east and south by resi- dential and commercial development. This area is char- acterized by mixed agriculture, including corn, hay, and cattle production, and forestland in the ridge and valley physiographic region (Sauer et al. 1997). Eighty-two wooden nest boxes (internal dimensions: 20 X 23 cm floor, ca. 34 cm in height) were erected in open habitats in Sussex County (centered ca. 41°fl'N, 74°38'W) be- tween 1 April 1993 and 6 April 1999, and 103 nest boxes in Warren County (ca. 40°47'N, 75°04'W) between 5 Au- gust 1995 and 19 April 1998; 124 nest boxes were avail- able during the 1999 breeding season. Methods Data Collection. We monitored nest boxes for kestrel breeding activity at 21-28-d intervals between 30 March and 3 August 1999, Once a nest box contained at least one kestrel egg, additional visits were scheduled so that chicks would be observed within 2 d of hatching. Age was determined by body mass (Roest 1957, Balgooyen 1976, Lacombe et al. 1994, Smallwood and Bird 2002); eight chicks were still wet when first observed. To collect vocal samples, we visited nest boxes with chicks at 2— 3-d inter- vals until the oldest chick of a brood was about 22 d old. In this study, area chicks fledge on about day 28 and are prone to premature fledging if disturbed during the pre- ceding week (Smallwood and Natale 1998). We made analog audio recordings with a Marantz PMD 101 portable cassette recorder. To collect vocal samples, we removed all chicks of a brood from their nest box and held each (one at a time) by hand in an upright position ca 12 cm from the recorder’s built-in condenser micro- phone. Most chicks vocalized within a few seconds. If a chick failed to vocalize for 3 min, we ended the recording Table 2. Age-related changes in acoustical characteris- tics of American Kestrel chicks (49 males and 39 females pooled) from northwestern New Jersey, 1999. For each variable, the correlation is between age category {N = 11, 0-1 d through 20-21 d) and the mean value for each age category. See methods for descriptions of variables. Variable n P Notes/call 0.964 0.0001 Internote -0.973 0.0001 Note length 0.916 0,0001 Number of harmonics 0.973 0.0001 Dominant harmonic 0.954 0.0001 Amplitude pulses -0.706 0:0152 Maximum frequency 0.927 0.0001 Initial maximum frequency 0.891 0.0002 High maximum frequency 0.845 0.0010 Low maximum frequency 0.973 0.0001 End maximum frequency 0.936 0;0001 Initial mean frequency 0.900 0.0002 High mean frequency 0.855 0.0008 Low mean frequency 0.964 0.0001 End mean frequency 0.927 0.0001 High 75% frequency 0.827 0.0017 Low 25% frequency 0.882 0.0003 Frequency range -0.809 0.0026 attempt and noted that the chick was silent. We identified individual chicks by coloring the down feathers of the humeral tract with permanent marking pens; gender was determined when primary feathers erupted, about day 8. The analog recordings were digitized at a 44.1-kHz sampling rate using the sound recorder program of Microsoft Windows 98 on a PC platform. We prepared digital spectrographs with Avisoft-SASLab Pro v. 3.4 soft- ware with an effective bandwidth of 647 Hz and a 256- point Fast Fourier Transform (FFT) size. Analytical fea- tures of this program include amplitude spectra of user-defined segments of the acoustical signals. Acoustical Variables. All vocalizations consisted of a se- ries of distinct notes that ranged from about 10-300 msec in duration. We defined a “call” operationally as a se- quence of notes such that the interval between notes of sequential calls was at least 1.5 times greater than the interval between notes within a call; the “intercall” du- rations generally were at least several times greater than the “intracall” durations. We measured the following 18 acoustical variables. (1) NOTES/CALL: the mean num- ber of notes per call. A random numbers table was then used to select one note (excluding the last note of a call; see next variable) from each vocal sample for acoustical analysis. (2) INTERNOTE: the interval (in sec) between the selected note and the following note within the same call, (3) NOTE LENGTH: the duration (in sec) of the selected note. (4) NUMBER OF HARMONICS: We ex- amined a spectrograph of the selected note for distinct frequency bands. We then analyzed the note at the point in time when the maximum number of frequency bands March 2003 Kestrel Chick Vocalizations 39 Age (days) Figure 1. Vocal development of male {N = 38) and female {N — 31) American Kestrel chicks, northwestern New Jersey, 1999. Values are means (±SE) of the first principal component. Principal component analysis was performed on 13 frequency variables; PCI eigenvectors for these variables are presented in Table 1. Time (sec) Figure 2. Age-related changes in the calls of an individua;! male American Kestrel chick from Sussex County, New Jersey, 1-21 June 1999. Digital spectrographs of analog recordings were prepared using 44.1-ikHz sampling rate and Avisoft-SASLab Pro v. 3.4 software (with an effective bandwidth of 647 Hz and a 256-point FFT transform size). Amplitude is indicated by darkness (i.e., black indicates more energy than grey). 40 Smallwood et al. VoL. 37, No. 1 Age (days) Figure 3. Call notes of American Kestrel chicks {N = 69), northwestern New Jersey, 1999. (Top) Call notes are given with increasing rapidity during the nestling stage. Values are means (±SE) of internote durations. (Bottom) The frequency (pitch) of call notes increases during the nestling stage. Values are the frequencies (mean ±SE) that have the highest amplitude within a call note. were separated by the greatest differences in amplitude. (5) DOMINANT HARMONIC: the frequency (in kHz) of the harmonic with the greatest amplitude. (6) AM- PlfTUDE PULSES: the number of distinct amplitude pulses within the selected note. (7) MAXIMUM FRE- QUENCY: the frequency with the highest amplitude, measured from the cumulative amplitude spectrum gen- erated from the entire selected note. We measured the remaining variables from the spectrograph, in which curves were generated denoting specific properties of the signal for each point in time within the selected note. (8) INITIAL MAXIMUM FREQUENCY: the frequency that had the maximum amplitude at the start of the note. (9) HIGH MAXIMUM FREQUENCY: the highest frequency on the maximum amplitude curve within the selected note. (10) LOW MAXIMUM FREQUENCY: the lowest frequency on the maximum amplitude curve within the selected note. (11) END MAXIMUM EREQUENCY: the frequency that had the maximum amplitude at the end of the note. (12) INITIAL MEAN FREQUENCY: the March 2003 Kestrel Chick Vocaijzations 41 0-1 2-3 4-5 6-7 8-9 10-11 12-13 14-15 16-17 18-19 20-21 Age (days) Figure 4. Call notes of American Kestrel chicks {N = 69) from northwestern New Jersey, 1999. (Top) The number of distinct harmonic frequencies increases during the nestling stage. Values are means ±SE. (Bottom) The frequency (pitch) of the dominant harmonic increases during the nestling stage. Values are the frequencies (mean ±SE) of the harmonic that has the highest amplitude within a call note. mean frequency at the start of the note. (13) HIGH MEAN EREQUENCY: the highest frequency on the mean frequency curve within the selected note. (14) LOW MEAN FREQUENCY: the lowest frequency on the mean frequency curve within the selected note. (15) END MEAN FREQUENCY: the mean frequency at the end of the note. (16) HIGH 75% FREQUENCY: the highest fre- quency on the 75th percentile curve within the selected note. The 75th percentile curve denotes, for each point in time, the frequency below which 75% of the acoustical energy is present. This variable, and the next two vari- ables, provide measures of frequency ranges that are not dependent upon recording level. (17) LOW 25% FRE- QUENCY: the lowest frequency on the 25th percentile curve within the selected note. (18) EREQUENCY RANGE: an index of frequency range, 75% HIGH FRE- QUENCY minus 25% LOW FREQUENCY. Statistical Analyses. We divided the data set into 112- d age categories, allowing each age category to be ex- amined separately and ensuring that no individual kestrel was represented by more than one vocal sample per age category. The data were tested for normality. Because we detected significant deviations, we used nonparametric statistical treatments for all univariate comparisons be- tween males and females. The results of comparisons of gender within one age category were not independent of the comparisons within another age category because the same individual birds were represented in each, therefore, we adjusted the P-values using Bonferroni’s probabilities (Snedecor and Cochran 1980). We per- formed a discriminant function analysis, with gender as the single classification variable, separately for each age category. The 13 acoustical variables denoting frequency (measurements in kHz) were subjected to a principal component analysis, and we employed univariate treat- ments to compare males and females with respect to the 42 Smallwood et al. VoL. 37, No. 1 first two principal components separately for each of the 1 1 age categories; again, we adjusted the P-values due to the nonindependence of these tests. Finally, we tested age-related changes in the acoustical properties of vocal- izations with nonparametric correlation analyses; age cat- egory {N = 11.) was correlated with the mean values (of each age category) for each variable. Results Gender Comparisons. A total of 563 vocal sam- ples was obtained from 88 chicks (49 males and 39 females) from 20 broods. We sampled individual chicks from 1-9 times (v = 6.4, mode = 8). The tendency to vocalize was similar for males and fe- males; mean response rates were 89.8% ± 2.17 (SE) and 90.2% ± 2.46, respectively (Z = 0.374, P — 0.71, Wilcoxon rank sums test). Univariate analysis of the 18 acoustical variables (for 11 age categories; 198 comparisons in all) re- vealed three significant differences between males and females, all for age 8-9 d: INITIAL MEAN EREQUENCY {x - 5.26 kHz ± 0.20 and 4.35 kHz ± 0.24, respectively, Z = 2.865, P = 0.046, Wil- coxon rank sums test), LOW 25% EREQUENCY (x = 3.69 kHz ± 0.34 and 2.78 kHz ± 0.24, re- spectively, Z = 2.896, P = 0.042), and DOMI- NANT HARMONIC (x = 5.08 kHz ± 0.28 and 3.59 kHz ± 0.30, respectively, Z = 3.330, P = 0 . 010 ). The discriminant function analyses for each age category had error rates that ranged from 7.9% (age 14—15 d, 1 of 11 females classified as a male and 1 of 15 males classified as a female) to 37.4% (age 20-21 d, 10 of 25 females classified as males and 8 of 15 males classified females). The error rate was not correlated with age category (r^ = — 0.109, P = 0.75). Pooled results of the 11 age- specific analyses had an error rate of 26.0% (63 of 249 samples from females classified as males and 79 of 297 samples from males classified as females) . A principal component analysis of the 13 acous- tical variables denoting frequency generated a first principal component (PCI) that accounted for 50.03% of the sample variability (eigenvalue — 6 504) and a second principal component (PC2) that accounted for an additional 17.92% of the sample variability (eigenvalue = 2.329); thus, over two-thirds of the sample variability was explained by the first two principal components (Table 1). Univariate analysis of PCI and PC2 for 11 age cat- egories (22 comparisons in all) revealed only one signihcant difference between males and females: for age 8-9 d, mean PCI values were 0.58 ± 0.42 and —1.52 ± 0.48, respectively (Z = 3.083, P = 0.022, Wilcoxon rank sums test; Eig. 1) . Age Comparisons. Figure 2 is a composite spec- trograph of the vocalizations of an individual male kestrel chick at ages 0, 6, 13, and 20 d. The day 0 call notes were “thin” (harmonically simple, with a narrow frequency range at any point in time), clear tones that dropped in pitch, sounding like peep or cheep. The day 6 call notes were given more rapidly and had more energy in the upper fre- quencies. The pitch was slurred downward and the notes were more cheep-\\ke. The day 13 call notes were delivered faster still and had more energy concentrated into distinct harmonic frequencies. The pitch, no longer slurred downward, instead was either steady or tremulous, sounding like chee or kee. The day 20 call was similar to that of an adult klee. Distinct harmonic frequencies were well developed and the frequency modulation, al- though distinct in the spectrograph, was rapid and slight, such that the notes did not sound particu- larly tremulous. Each acoustical variable was significantly corre- lated with age category (Table 2) . All correlations were positive, except for INTERNOTE, AMPLI- TUDE PULSES, and FREQUENCY RANGE. The relationship between age category and INTER- NOTE, MAXIMUM FREQUENCY, NUMBER OF HARMONICS, and a DOMINANT HARMONIC are given in Figs. 3 and 4. Discussion Acoustical characteristics of call notes changed in a consistent manner as chicks matured. Notes became longer and increasingly complex with re- spect to harmonic structure, and both the number of notes per call and the rate at which they were delivered increased. In general, vocal characteris- tics changed most rapidly during the first two weeks (Figs. 1, 3, and 4). By about day 16 chicks were able to produce calls that sounded similar to the klee calls of adults. Roest (1957) noted that chicks of this age were able to utter the adult-like ‘‘killy-killy” cry, although some chicks were silent when handled. The discriminant function analyses did not per- form well in distinguishing gender. The pooled er- ror rate of 26.0% was slightly closer to that of ran- dom classification (50% error) than to perfect discrimination. Vocalizations of males and females were essentially indistinguishable from each other March 2003 Kestrel Chick Vocalizations 43 with respect to the acoustical characteristics we measured, except that males appeared to progress toward adult-like vocalizations more rapidly than females, especially during the second week (Fig. 1). The first principal component (derived from frequency variables and most strongly influenced by LOW MEAN FREQUENCY, END MEAN FRE- QUENCY, and END MAXIMUM FREQUENCY; Table 1) differed significantly between 8-9-d-old males and females. Three other frequency vari- ables (INITIAL MEAN FREQUENCY, LOW 25% FREQUENCY, and DOMINANT HARMONIC) also differed significantly at this age. These results sug- gest that males may develop adult-like frequencies (higher pitch; Table 2) sooner than females. Although possible, it is unclear if individual var- iability in vocalization patterns is sufficient for in- dividual recognition by kestrels. Such vocal recog- nition is widespread in both passerines and nonpasserines (e.g., Falls 1982, Stoddard 1996, and citations therein). Additional research is required to document if this behavior occurs in American Kestrels. Acknowledgments We gratefully acknowledge GPU Energy and Sussex Ru- ral Electric Cooperative for allowing us to erect nest box- es on their utility poles, and the Wallkill River National Wildlife Refuge and local landowners for providing us access to field sites. We thank G. Laborda and B. Wyman for assistance in monitoring nest boxes, and C. Schleck for digitizing the analog recordings. We also thank K.L. Bildstein for a thoughtful review of the manuscript. Funding was provided by a Margaret and Herman Sokol Faculty/Student Research Grant and a Separately Bud- geted Research Award, Montclair State University (MSU). Release time for J. Smallwood was provided by the Faculty Scholarship Incentive Program, MSU. Literature Cited Balgooyen, T.G. IQ'Zb. Behavior and ecology of the American Kestrel {Falco sparverius h.) in the Sierra Ne- vada of California. Univ. Calif. Publ Zool. 103:1-83. Carlson, G. and C.H. Trost. 1992. Sex determination of the Whooping Crane by analysis of vocalizations. Condor 94:532-536. Falls, J.B. 1982. Individual recognition by sound in birds. Pages 237-278 in D.E. Kroodsma and E.H. Miller [Eds.], Acoustic communication in birds, Vol. 2. Ac- ademic Press, New York, NYU.S.A. Kroodsma, D.E. and E.H. Miller (Eds.). 1982. Acoustic communication in birds, Vols. 1 and 2. Academic Press, New York, NY U.S.A. Lacombe, D., D.M. Bird, and K.A. Hibbard. 1994. Influ- ence of reduced food availability on growth of captive American Kestrels. Can. J. Zool. 72:2084-2089. Mathevon, N. 1997. Individuality of contact calls in the Greater Flamingo Phoenicopterus ruber and the prob- lem of background noise in a colony. Ibis 139:513- 517. Roest, A.I. 1957. Notes on the American Sparrow Hawk Auk 74:1-19. Sauer, J.R., J.E. Hines, G. Gough, I. Thomas, and B.G. Peterjohn. 1997. The Breeding Bird Survey results and analysis. Version 96.3. Patuxent Wildlife Research Center, Laurel, MD U.S.A. Sherman, A.R. 1913. The nest life of the Sparrow Hawk. Auk 30:406-418. Smallwood, J.A. and C. Natale. 1998. The effect of pa- tagial tags on breeding success in American Kestrels N. Am. Bird Bander 2S-.7A-78. and D.M. Bird. 2002. American Kestrel {Falco sparverius) . In A. Poole and F. Gill [Eds.] , The birds of North America, No. 602. The Birds of North Amer- ica, Inc., Philadelphia, PA U.S.A. Snedecor, G.W. and W.G. Cochran. 1980. Statistical methods, 7th Ed. Iowa State Univ. Press, Ames, lA U.S.A. Stoddard, P.K. 1996. Vocal recognition of neighbors by territorial passerines. Pages 356-374 in D.E. Kroods- ma and E.H. Miller [Eds.], Ecology and evolution of acoustic communication in birds. Cornell Univ. Press, Ithaca, NY U.S.A. Thumser, N.N., J.D. Karron, and M.S. Ficken. 1996. In- terspecific variation in the calls of Spheniscus pen- guins. Wilson Bull 108:72-79. Willoughby, E.J. and T.J. Cade. 1964. Breeding behavior of the American Kestrel (Sparrow Hawk). Living Bird 3:75-96. Received 21 August 2001; accepted 30 June 2002 J Raptor Res. 37(1) :44— 54 © 2003 The Raptor Research Foundation, Inc. MORPHOMETRICS AND STATUS OF AYRES’S HAWK-EAGLE IN ZIMBABWE Ronald R. Hartley^ Zimbabwe Falconers’ Club and The Peregrine Fund Inc., Falcon College, Esigodini, Zimbabwe Peter J. Mundy Department of National Parks and Wild Life Management, PO. Box 2283, Bulawayo, Zimbabwe Abstract. — ^We recorded information on plumage, molt, body mass, and morphology for 28 Ayres’s Hawk-Eagles (Hieraaetus ayresii); 17 were trapped or injured in Zimbabwe between 1981-2002, the other 11 were museum specimens. Juvenile and adult birds had variable plumage coloration. Some adults of either sex had a black eye mask and others a pied appearance (a white eyebrow mark over black eye mask) , while all juveniles had a pale eyebrow mark over a slate eye mask. Adult females were more heavily marked than adult males. Ayres’s Hawk- Eagles in captivity started molt in August-September (N = 2, 13 molts) and changed into adult plumage at the end of their molt at 2 yr of age. The mean body mass for males was 656.2 g (N = 8) and for females 1003.0 g (N = 20). Body mass and wing length did not overlap between sexes. Of nine nests, six were on well-wooded hill slopes and three were in riparian woodland. Nests were 9-15 m above ground, situated below the canopy and nest diameters were 71—129 mm. Although Ayres’s Hawk-Eagle is a rare species, it was not as scarce in Zimbabwe as some observers had claimed. Keywords: Ayres’s Hawk-Eagle, Hieraaetus ayresii; breeding, capture, morphology, plumage, status', Zimbabwe. MORFOMETRIA Y ESTADO DEL AgUILA DE AYRES EN ZIMBABWE Resumen. — Colectamos informacion sobre el plumaje, muda, masa corporal y morfologia de 28 aguilas de Ayres {Hieraaetus ayresii)', 17 fueron atrapadas o heridas en Zimbawe entre 1981—2002, y las otras once fueron especimenes de museo. Las aves juveniles y las adultas tuvieron coloracion de plumaje variable. Algunos adultos de ambos sexos tenian una mascara con un ojo negro y otros una apariencia moteada (una marca de ceja blanca sobre la mascara de ojo negro), mientras que todos los juveniles tuvieron una palida marca de ceja sobre una mascara de ojo pizarra. Las hembras adultas estaban mas fuertemente marcadas que los machos adultos. Las aguilas de Ayres en cautividad inician su muda en Agosto-Septiembre {N = 2, 13 mudas) y viran hacia su plumaje de adultos al final de su muda en el 2 ano de edad. La masa corporal promedio para los machos fue 656.2 g (A = 8) y para las hembras 1003.0 g (A = 20). La masa corporal y la longitud del ala entre sexos no se sobrelapo entre sexos. De nueve nidos, seis estuvieron en pendientes de colinas bien arboladas y tres en bosques riparios. Los nidos estaban entre 9-15 m sobre el piso, situados bajo el dosel y con diametros entre 71-129 mm. Aunque el aguila de Ayres es una especie rara, esta no era tan escasa como algunos observadores han afirmado. [Traduccion de Cesar Marquez] Ayres’s Hawk-Eagle {Hieraaetus ayresii), is a small, dashing raptor that is thought to be inexplicably rare throughout its range (Brown et al. 1982, del Hoyo et al. 1994, Harrison et al. 1997). We follow Amadon and Bull (1988) in using the specific name ayresii for this eagle. Little is known about this species (Steyn 1982, Harrison et al. 1997, Vir- am and Watson 1998), other than some basic de- * E-mail address: dimpshartley@yahoo.com scription of its breeding biology, which has been reported in Kenya (Brown 1966, Dewhurst et al. 1988). Ayres’s Hawk-Eagles may occur in urban ar- eas where they commonly prey on Rock Doves ( Co- lumba livia; Irwin 1981, 1985, del Hoyo et al. 1994, Harrison et al. 1997), other columbids, and small birds (Lendrum 1982, Dewhurst et al. 1988). It is an aerial hunter and may spend much of the day on the wing (Clark 1999). Few nests have been found anywhere, and there are only 12 breeding 44 March 2003 Ayres’s Hawk-Eagle in Zimbabwe 45 Table 1. Body mass (g) and measurements (mm) of Ayres’s Hawk-Eagles. Age Males Females Mean Range SD N Mean Range SD N Mass Juvenile 650.0 620-680 ±22.36 4 988.5 950-1045 ±38.74 10 Adult 662.5 615-714 ±37.94 4 1017.5 879-1150 ±87.02 10 Wing Juvenile 355.0 340-368 ±10.0 8 396.7 374-420 ±13.67 19 Adult 346.0 320-360 ±14.97 5 392.8 375-408 ±11.17 16 Tail Juvenile 189.4 180-198 ±6.0 8 208.4 191-224 ±10.3 18 Adult 185.2 175-195 ±7.29 4 210.6 196-225 ±8.96 17 Tarsus All 63.5 61-67 ±2.02 13 73.8 61-80 ±4.97 19 Toe All 45.2 42-49 ±1.8 12 52.0 46-56 ±2.6 17 Bill All 21.3 20-22 ±1.0 9 25.4 24-31 ±5.78 12 Gape All 25.3 24-28 ±1.3 12 27.9 25-32 ±1.96 12 records for Zimbabwe (Irwin 1981), some of which may be questionable. There are few data on mea- surements and body mass (Maclean 1993), and no details on the progression of Juvenal to adult plum- age (Brown et al. 1982, Steyn 1982). Plumage is variable, and includes primarily light and uniform- ly dark color morphs (Brown 1966, Brown and Davey 1978, del Hoyo et al. 1994, Kemp and Kemp 1998), but there are few data summarizing this var- iation. Furthermore, it is likely that plumage vari- ation and confusion with species such as African Hawk-Eagle (H. spilogaster) , Booted Eagle {H. pen- natus), and Cassin’s Hawk-Eagle {Spizaetus african- us) have led to the Ayres’s Hawk-Eagle being over- looked throughout its range (Brown 1966, Ash 1981, Clark 1999). Here, we present new data on plumage (includ- ing molt), change of eye color, morphology, and body mass of Ayres’s Hawk-Eagles in Zimbabwe. We also summarize sightings and breeding records in Zimbabwe and discuss the status of Ayres’s Hawk- Eagle. Methods Nine Ayres’s Hawk-Eagles were captured and another eight eagles were recovered injured in Zimbabwe. Fifteen of these were weighed and 16 measured (Table 1). Mass measurement was done to the nearest 5 g. Other mea- surements follow Biggs et al. (1978), except for bill, which was taken as the chord from tip of bill to distal edge of the fleshy cere. Vernier calipers accurate to 0.5 mm were used to measure bill, gape, tarsus, and middle toe. Wing and tail length were measured with a wing rule accurate to 1 mm. Wing area, wing span, and mass load- ing follow Mendelsohn et al. (1989). The scientific liter- ature was searched for further data (Brown and Davey 1978, Hartley 1982, Dewhurst et al. 1988, Grimes 1987, Hanmer 1997) and 11 museum specimens in the collec- tion of the Natural History Museum of Zimbabwe (NHMZ) were examined and measured. We also intro- duced data (including mensural) on five specimens from South Africa provided by A.C, Kemp, and one from Zim- babwe (Transvaal Museum, TM), and 10 from the British Museum (BM) . Birds were sexed on the basis of size (Ta- ble 1), males being considered the smaller birds, less than 750 g body mass and 370 mm wing length, respec- tively. Body mass and wing length between sexes were nonoverlapping. Where body mass was not available and wing length was close to 370 mm, measurements of tarsus (<67 mm for male) and mid-toe (<49 mm for male) were also taken into consideration. An eagle in its Juvenal plumage was classed as Juvenile and as an adult after its first prebasic 1 molt (Humphrey and Parkes 1959) was completed. Eye color, plumage, and molt descriptions include the progress of two indi- viduals in captivity from juvenile to adult. R. Hartley has kept two male and two female Ayres’s Hawk- Eagles for extended periods, including a male for five and a female for ten years. Our own records of sightings were included with those of the Zimbabwe Falconers’ Club (ZFC, Har- tley 1993). Breeding records were provided by egg col- lectors and falconers. Laying dates were estimated on the basis of 45 d and 75 d for incubation and nestling peri- ods, respectively (Steyn 1982). Results Captures and Specimens. Of 23 eagles recovered in Zimbabwe, four in South Africa, one in Zambia, and one in Kenya, 18 were Juveniles (shot = 4, collisions = 5, captured = 9) and 11 were adults (shot = 5, collisions = 2, captured = 4). One fe- male was recovered at Kalichero in northern Zam- bia after fighting with a European Honey Buzzard {Pernis «j&morw5) (Benson 1962). In Zimbabwe sev- en eagles were involved in collisions: two with ve- hicles, one with a power line, and four struck fenc- es. Five eagles were captured at pigeon lofts in towns and another five were captured on farm lands Just north of Harare. AJuvenile male (620 g) 46 Hartley AND Mundy VoL. 37, No. 1 Figure 1. Locations of places mentioned in the text in Zimbabwe. was captured after it attacked a tethered adult fe- male Black Sparrowhawk {Accipiter melanoleucus, 920 g), the latter holding onto the eagle. One fe- male was recovered in a chicken pen at Bulawayo (Hartley 1982; Fig. 1). Eight eagles were males (two adults and six juveniles) and 12 were females (two adults and 10 juveniles). Ten of these eagles were released. Six of the specimens in the NHMZ were from Zambia and one was from South 7\frica. While most eagles were seen singly, sometimes they occurred in small groups. Two juveniles (male and female) were trapped in the company of pairs of adults in Harare and Bulawayo, but breeding has not yet been recorded in these towns. A juvenile female was trapped within an hour of trapping an adult male (March) near Harare, and two juvenile females were trapped (April) about an hour apart m the same general area. A juvenile male was seen with an adult (unknown sex) at a loft in Harare (January) and at Triangle (February). As Ayres’s Hawk-Eagle lays mainly in April-May, these juve- niles may have been independent of the adults. Body Mass and Measurements. Based on body mass and wing length, one museum specimen was incorrectly sexed, while 10 records that were un- sexed were now classified accordingly. The wing loading and wingspan of a juvenile male (Table 1 ) was 0.387 g/cm^ (total wing area = 1652.67 cm^) and 97 cm, respectively. These measurements were 90% and 78% of that recorded for females by Men- delsohn et al. (1989). Plumage. There was substantial variation within the juvenile and adult morphs, both between and within sexes (Tables 2 and 3; see photographs in Oberprieler and Gillie 2002). All juveniles and adults had an eye mask (Fig. 2). Juveniles had a pale eyebrow over a slate eye mask. Juveniles had either cream, buff or rufous nape feathers, which were streaked with dark brown or slate as stated by Brown et al. (1982). While the dark brown to slate throat and/or chest streaks extended over the shoulders like two shoulder straps in most (Hartley 1982), some individuals had almost plain colored underparts from throat to abdomen; the degree of streaking and/or blotching on the chest was not related to sex. The underwing coverts were cream, buff or rufous. The upper-wing coverts, back and rump were edged paler, giving a scaled appearance (Brown et al. 1982) that is most pronounced in the March 2003 Ayres’s Hawk-Eagle in Zimbabwe 47 Table 2. Plumage of juvenile Ayres ’s Hawk-Eagles. Eagles Crown Head Forehead/ Eyebrow Eye Mask Throat/Crop/Chest/Belly Males 2 light brown light brown slate light brown with slate streaks on flanks N= 3 slate cream slate cream-buff with slate streaks on flanks Eemales iV= 4 slate buff slate light brown with slate streaks on flanks N= S slate cream slate light brown with fine streaks on flanks N= 2 slate cream slate cream with fine slate streaks fresher plumage of younger birds, as these edges wear off and are less noticeable after a few months. All Ayres’s Hawk-Eagles, including juveniles, showed the characteristic white patches at the front base of the wings the so-called “landing lights.’’ This characteristic is most obvious when the birds are in flight. In captivity, the first basic molt of two individuals was almost complete, with Just a few juvenile body feathers remaining, and the progression was di- rectly into the adult plumage (as suggested by Brown et al. 1982). However, the head pattern of one eagle changed significantly by the beginning of her sixth year, from a black mask (see Eig. 2 in Table 3. Plumage of adult Ayres’s Hawk-Eagles. Hartley 1982 and photograph in Sinclair 1984) to a paler black and white pied or mottled appear- ance (see photograph in Hartley 1989a). Such a change was not shown by three other eagles over similar time periods. Generally, males had less streaked and blotched chests and flanks than did females. However, the head patterns of males and females overlapped from the completely black- head-mask type (illustrated in Hartley 1989b) to the white forehead-eyebrow and black-eye-mask type so that this distinction could not be used as an aid to identification of the sexes contrary to the suggestion of Brown et al. (1982). The upper wing coverts and back and rump feathers were generally Head Eagle Crown Forehead/ Eyebrow Eye/Head Mask Throat/Chest/Belly Males N= 1 black white/black black head white with few black streaks — low density N= 1 black white/white black eye white with black streaks — medium density N= 2 black white/black black head white with fine black srteaks on flanks — medium density Females N= 2^ black white/black black head white with black streaks and blotch- N= 5^ black black/ black black head es white with black streaks and blotch- N= 3^ black black/black black head cs heavily blotched black over white One captive female assumed this plumage by the end of 2 yr, but developed an eye mask by the end of 6 yr. ^ One individual had a white eyebrow. Dark morphs. 48 Hartley and Mundy VoL. 37, No. 1 Figure 2 . Nine-yr old female Ayres’s Hawk-Eagle (left) and 5-yr old male (right). pale edged, and this ‘scaled’ appearance (see Brown and Davey 1978) can help to distinguish Ay- res’s Hawk-Eagle from the superficially similar Af- rican Hawk-Eagle. Six adult females were dark morphs (Table 3), four of these from Zambia. Two of three adult specimens from South Africa were dark morphs (A. Kemp pers. comm.). Timing of Molt. Two captive eagles generally started their molt in August-September and had completed it by the end of March. However, one of these had started body molt (not primary, sec- ondary or tail) by 1 1 May when it was captured as a juvenile (Hartley 1982). To begin body molt well before flight feather molt is usual in African rap- tors (Edelstam 2001). Another 1 1 eagles were recorded molting, each fitting into the period August-March. However, one adult probably started earlier than August, as she had nearly completed her primary molt by the third week of September. By the end of August one eagle had molted three primaries on each wing, while four eagles were proportionately more ad- vanced by mid-September and December respec- tively. Eour other eagles had nearly molted (with flight feathers nearly finished) by mid-Eebruary and another by the end of March. Eye Color. The eye color of recently-fledged ju- veniles was light gray. Six months later this had changed to pale yellow. Two juveniles had pale yel- low eyes, which changed to bright yellow and yel- low-orange, respectively, over the next eight months. In one year, the eye color of one eagle turned deep yellow. We noted no significant dif- ference in eye color between the sexes (but see Brown et al. 1982). Sightings. P. Mundy recorded 23 sightings (11 adults, six juveniles, and six not aged) of these ea- gles within Zimbabwe over a 12-yr period (1985- 96) . All adults were seen singly. Regular sightings were made in all months of the year in the Zambezi Valley and the eastern highlands (Fig. 1; Table 4). In the Chimanimani area, P. Caldwell-Barr (pers. comm.) recorded adults (in pairs and singly) and juveniles, some at- tacking birds in aviaries. This suggests that Ayres’s Hawk-Eagle breeds in these rugged, heavily-wood- ed areas in particular. By contrast, in Harare and Bulawayo, they were largely absent from May to De- March 2003 Ayres’s Hawk-Eagi.e in Zimbabwe 49 Table 4. Summary of sightings and specimens by month of Ayres’s Hawk-Eagles in Zimbabwe. Area J F M A M J J A S O N D TOTAI. Total ruraP Rukomeche, Kariba, Mavuradonha, 6 6 9 8 6 5 5 12 17 12 6 15 107 Mutare and Chimanirnani’^ X X X X X X X X X X X X Bulawayo and Harare'^’'^’*^ .5 9 13 4 2 1 1 4 1 1 41 Bulawayo‘S X X X X X X X X X ■» Vernon (1979), Hartley (1982, 1998a), Irwin (1984, 1985), Howells (1985), Tree (1989, 1990, 1991, 1992, 1994, 1995, 1996, 1997), specimens (this study), Hartley (pers. observ.), Mundy (pers. observ.). •^Hartley (1982), Tree (1989), K. Hustler, R. Naisbitt, and P. Caldwell-Barr (pers. comm.). ^Hartley (1998b). d Irwin (1984), Tree (1989, 1997), Solomon (1995). '^Sightings reported {N ^ 598) in Lendrum (1982). cember. While they may nest in towns in Zimbab- we, a site has not been found, despite intensive searching for raptor nests by the ZFC and others. It is possible that Ayres’s Hawk-Eagles are breeding in heavily-wooded areas relatively close to Bulawayo and Harare, as family groups (i.e., adult male and female, and juvenile) have been seen raiding lofts in these cities. Nest Location. We report six additional nests (including an alternate site for Bangala, see Phil- lips 1978) to those reported in Irwin (1981) for Zimbabwe, and one for southern Zambia (Table 5). Data from four nests indicated that the struc- ture consisted of sticks up to 45-60 cm long and 1. 8-2.0 cm diameter, but generally they are much thinner and smaller. The nests tend to be small (Table 5), flat on top and deep (Brown and Davey 1978). On average they are smaller than nests of the African Hawk-Eagle which tend to be over 1 m in diameter (Steyn 1982). Nests were located in dense, mature woodland in rugged terrain or in thick, riparian habitats. In hilly terrain the nests tend to be located on slopes overlooking ravines, four of which can be described as hidden valleys. Several of these nests have been difficult to spot, except at close range, as little as ca. 4 ra for nest 4 (Table 5). Nests were placed mainly in a vertical fork or sometimes on a horizontal branch, well inside the canopy. Medium sized, smooth barked trees with heavy foliage were preferred, making nests extra difficult to locate. Nests were in trees that were generally in clumps emergent above the overall canopy of woodland. For instance, nest 1 was well concealed on the lat- eral branch of a Syzygium sp. in a clump of riparian forest on the side of a hill. Nest 4 was in a clump of big Spirostachys africanus that covered about 1 ha (Phillips 1978). Nest 3 was in a tall Brachystegia glau- cescens growing in a steep gully inside the hills. However, nest 8 was in a Julbernardia globiflora con- nected with the overall canopy in an area of sub- stantial mature woodland on a slope surrounded by relatively flat terrain. Timing of Laying and Egg Size, In each instance a single egg was laid. A summary of breeding re- cords (ascribed to R. Brooke) was used in Irwin (1981), and this list included four verifiable laying dates in April (reported by C. Vernon) for nest 3 and nest 5 (Phillips 1978), and one laying record for May (Hough 1950). Added to 13 breeding re- cords observed since then, laying occurred in April {N= 10) and May {N = 8). Mean egg sizes for Zimbabwe were 61.22 mm (SD ± 1.41) X 49.85 mm (±1.51, N= 6). The mass of one egg was 74 g. An egg collected in Zambia was small at 51.3 mm X 42.9 mm (48.5 g). Steyn (1982) also reported small egg size from Zambia {N — 3) and we can offer no explanation for this contrast, especially as there is no apparent differ- ence in the size of eagles between Zimbabwe and Zambia. Hunting and Prey. Prey items seen on nests by egg collectors have all been birds, often doves (Columbidae) , with the exception of one Tree Squirrel {Paraxerus cepapi). An adult female was seen eating a Cape Turtle Dove (Streptopelia capi- cola). A released eagle was seen hunting Rock Doves on numerous occasions. She was seen fre- quently soaring high (ca. 100-200 m) overhead and the Rock Doves were attacked from a vertical stoop. She also took a Green Pigeon {Treron calva) chick. She remained in the area for 20 wk and was seen nearly every day. Also, we saw an adult female stoop through the canopy of miombo woodland at Table 5 Nest site characteristics for Ayres’s Hawk-Eagle. 50 Hartley AND Mundy VoL. 37, No. 1 S H Cd »■ 2 Q s id O H n l-i ^ ^ § s s o n ^ o w O g z h w w 2 w w H (Z) Ph S P • pn Oh £ Id T3 1) g O 13 P O O P2 I p "O Ip 03 »o 03 CO p u TP Tj p u p 1 OJ _ u be P bo P ■"' P iS O cS i-i n >-i g 3 g ?g .2 xi .2 -P Sm O P3 s S -■S- "T § i s e s o f K) S o Vi e ai SC3 s “I bo cs t a a t' e e p ea TP p Cti s C) f -C o 03 CM CO o CD 03 t'- O) i> C03 03 rH 03 , — , rH o on 03 CM 00 03 CD CD o V V 1> 03 o> 1> 03 05 rH rH be 03 f-H o rH s_^ . . S PQ CO pj > p CL) 03 .•' Ki IT) Si tT3 W 03 fi 1-H PQ CD I> 00 03 Ct p o TP cc3 ui ^ I D S 00 03 N TP P Sh 3 ct N g o dP u p CJ ceJ P ip in CM o xs ■ r*^ fl 13 CX ITi P 13 Tp Li p o dP 1 • rH u 13 1) p u 13 P CIh • rH u £X • »“H dP rt he Lh Tp U Tp 'S 0 p o 1 Lh Tp 1) Tp 1 range 1 13 be Lh 0 be 4-1 4P ■£P cn ded on Tp 1) Tp P (TO x> u P P 0 1 sx 13 « fH V • rH u O c/3 g 0 1 HH P JTO U U 1) rPn * rH rH 13 D dP (/J dP ‘C dp g dP dP ■E, m i> . — . _.rrs. W 00 1> CO rH CD on (M CM CM X "' — ’ — ’ ^ ^ 03 1-H O 00 o o m CM Oi 1> CM CD CM 00 I o lO 03 ci CD 03 CM CO 1) Lh <5 p p OJ Lh be Lh u > CJ TP P CS g io ■O . g .‘Cl I s v: Q a g g «ka S a ^ I CO 00 CO dp o g 2 cO n TP P , be ^ ‘b, 00 .ij CO --I X3 3 Til J> 03 CJ3 d g ^ £a Ci T—i ^ ^ Oh tiC ^ f3 s ^ E Ah 00 • »-H CO V CM Q C P I o PQ CQ 00 GO CD a dj O S Uh ^ I' u TO in P ce March 2003 Ayres’s Hawk-Eagle in Zimbabwe 51 the top of the Chizarira escarpment and nearly snatch a Black-eyed Bulbul {Pycnonotus harbatus). All of the eagles captured and shot were frequent- ing Rock Dove lofts, with the exception of eagles hunting other doves. In captivity, these eagles show a distinct preference for doves. Rock Doves, and small birds, and do not seem to favor mammals such as squirrels (recorded as prey by Brown and Davey 1978). Discussion Although marked variations in plumage (Brown 1966, Ash 1981) are supported in this study, some trends noted by Brown (1966) have not been ver- ified. Reduced white forehead and eyebrow and greater extent of buff on the forehead is not nec- essarily indicative of adult and juvenile males, re- spectively. Except for a dark-morph, adult female in Mutare (Hartley 1982), none of the adults in this study were as pale or dark as the extremes in Brown and Davey (1978) and the dark morph in Finch-Davies and Kemp (1980). Furthermore, the latter did not display the characteristic landing lights, which is most unusual. Otherwise, the na- ture of streaking, blotching, wing and tail patterns, and color of the soft parts was in accord with Brown et al. (1982); adult males usually much less streaked on the breast than females (Seavy 2000). It is also possible that observers mistake juvenile Ayres’s Hawk-Eagles for Booted Eagles, a rare sum- mer migrant to Zimbabwe. Timing of molt generally accords with Steyn (1982) and Lendrum (1975) who reported Sep- tember to February for birds in Bulawayo. This is also outside of the breeding period (egg-laying) and may further confirm that the peak laying pe- riod for this species is April-May, with chicks fledged by August-September. Hanmer (1997) ex- amined an adult female captured at Nchalo in Feb- ruary that had probably completed the molt. Ayres’s Hawk-Eagle shows marked reversed sex- ual dimorphism (Newton 1979), the male being on average 65% of the mass of the female, and about 89% of female size. Brown (1966) emphasized the size difference, but presented few data, including the only record of 714 g for a male (Brown and Davey 1978), and 879 g and 940 g for females. Gape size is small (27.9 mm for female) for a rap- tor this size, possibly reflecting the preferred diet of small birds (Lendrum 1982, Steyn 1982). Also a specialist aerial hunter of birds, the African Pere- grine Falcon {Falco peregrinus minor) has a gape of 30 mm (female == 700 g, Hartley 2000). Ayres’s Hawk-Eagle lays a single egg from April- May in Zimbabwe and April-July in Zambia (Steyn 1982). Therefore, several breeding records in Ir- Avin (1981) are questionable, including a two-egg clutch. No authenticated record of a two-egg clutch for Ayres’s Hawk-Eagle has been verified by any account and Brown (1966), Brown et al. (1982), Steyn (1982), and Tarboton and Allan (1984) have also questioned such records. Further- more, a single-egg clutch collected on 25 October is outside the range of laying based on our data. However, an experienced egg collector, B. Neuby- Varty (pers. comm.), collected a single egg on 17 September from a nest located in a large tree in thick bush beside an annual stream in mountain- ous habitat. This may have constituted a second clutch after a nest failure (Brown 1966) . An August record in Irwin (1981) from Hough (1950) was ac- tually for a well grown chick, that was also filmed (J. Hough pers. comm.), and therefore, provides an estimated laying date of May. In Zimbabwe nests have been found only in well- wooded areas in very rugged terrain (Phillips 1978), usually in close proximity to hills or moun- tains, but sometimes in riparian forest. In addition, P. Danckwerts (pers. comm.) found two occupied nests along the Eiafue River in Zambia. Middleton (2000) studied a raptor community in a 200 km^ study area of mainly flat terrain, but that included 29 km^ of well wooded, rugged, but smaller kopjes and domed inselbergs south of Phillips’ (1978) site. Despite fieldwork over eight years, no nest of Ayres’s Hawk-Eagle was found, nor were birds seen. To the south-east in a 40 km^ cluster of well wood- ed, rugged sandstone hills Davison (1998) studied a raptor community over two seasons and found 19 pairs of eagles, but none were Ayres’s Hawk- Eagle. Despite the discovery of numerous raptor nests in the Matobo Hills over the past 40 yr by the Black Eagle Survey team, Ayres’s Hawk-Eagle is re- garded as a vagrant (Gargett 1990). The same ap- plies to four other intensively-surveyed areas in well-wooded, domed inselberg terrain (R. Hartley unpubl. data). As Ayres’s Hawk-Eagle soars a great deal (Clark 1999) and displays conspicuously dur- ing courtship (Brown 1966), it is unlikely that they would be overlooked during these surveys. Generally Ayres’s Hawk-Eagles have been absent from Zimbabwe towns during May-^uly (Table 4), while a similar pattern was noted in the Transvaal 52 Hartley and Mundy VoL. 37, No. 1 (Tarboton and Allan 1984) and southern Botswana (Herremans 1994), where the species was absent for the winter months south of Zimbabwe (Harri- son et al. 1997). A pair of Ayres’s Hawk- Eagles bred in a stand of gum trees {Eucalyptus sp.) in Nairobi and another two pairs were also observed (Dewhurst et al. 1988). With the exception of these, plus three observations reported in this pa- per, all previous records in towns have been for solitary birds (Hartley 1982, Lendrum 1982, Tar- boton and Allan 1984), which were probably no- mads (Hartley 1998b), rather than migrants. A great danger to these birds is, in fact, their attraction to towns where they come into frequent contact with domestic and racing Rock Doves and ultimately with the owners (Lockwood 1979, Har- tley 1982, Lendrum 1982, Tarboton and Allan 1984, Herremans and Brewster 1994, Hartley et al. 1996). Some domestic pigeons are worth large sums of money. The ZFC and the Department of National Parks and Wild Life Management (DNPWLM) have established contact with the Zim- babwe Racing Pigeon Association (ZRPA) and it is vital that this be maintained. Although two of these eagles were accused of harassing poultry, these events may have been cases of mistaken intention as Ayres’s Hawk-Eagle does not usually molest poul- try (Hartley 1982, Lendrum 1982, Dewhurst et al. 1988). Hartley et al. (1996) reported habitat de- struction as another negative impact and recom- mended that Ayres’s Hawk-Eagle be closely moni- tored. In Zimbabwe, South Africa, and Kenya, 15 fe- males and three males were recovered in towns. The greater frequency of females is possibly a re- sponse to the size of domestic and racing Rock Doves (300—350 g), which are easily tackled and carried by the heavier females, rather than by smaller males. That the Ayres’s Hawk-Eagle lays only one egg and can produce no more than one youngster re- flects some limitation to available prey. (Newton 1979). This is supported by the apparent patchi- ness of its nesting density, despite Brown’s (1966) suggestion that its home range is ca. 25 km^. The similar Booted Eagle occurs at much higher den- sities in the Cape Province of South Africa (Steyn 1982) and it produces up to two eggs (and both chicks can survive). The Booted Eagle also hunts from a stoop, but it is a generalist with birds con- tributing about 50% of its prey. Small mammals, reptiles (mainly lizards) and insects are also taken which may allow it to capture more prey than does the Ayres’s Hawk-Eagle, and consequently it is more abundant. Another generalist, the African Hawk-Eagle, also occurs much more frequently than Ayres’s Hawk-Eagles and sometimes is rela- tively abundant in hills during the nesting period. Predation (Dewhurst et al. 1988) and poor breeding success (Brown 1974) have been suggest- ed as reasons for the rarity of Ayres’s Hawk-Eagle. Predation can occur frequently on raptors, includ- ing on some large eagles (Davison 1998). We sug- gest that predation may explain the tendency to select relatively concealed positions for nesting. Ay- res’s Hawk-Eagle is the smallest winter-breeding ea- gle, and although it sometimes shares its hill site nesting habitat with African Hawk-Eagle and other eagles. Brown (1966) recorded little antagonism between them at his Eagle Hill study site. Aggres- sive Peregrine (Brown 1966) and Lanner falcons {E biarmicus) (Phillips 1978), on the other hand, have probably interfered with breeding success of the Ayres’s Hawk-Eagle. It is also possible that this eagle owes its rarity to its role as a specialist bird predator, using its equally specialized method of attacking from a stoop. Clark (1999) contended that it is overlooked, because observers are not searching the sky. While it is clear that the Ayres’s Hawk-Eagle is not as scarce as some authorities claim (Brown et al. 1982, Steyn 1982, Maclean 1993, Clark 1999), it is still a rare species. Ackn owledgments This is a contribution from the Raptor Research Pro- gramme of the ZFC, which is a joint project with the DNPWLM and The Peregrine Fund, Inc. (USA) . The fol- lowing contributors are gratefully acknowledged: Falcon College and Peterhouse School Falconry Clubs; A. Barnes, P. Caldwell-Barr, N. Deacon, D. Dudman, A.S. Dunkley, N. Goodwin, R. Geisswein, N. Greaves, A. Gro- enewald, A. Fluelin, O. Mitumbili, R. Naisbitt, R.W. Querl, S. Rankine, K. Wilson (all ZFC); R. Greaves, K. Hustler, J. Hough, R.D. Jeffrey, A.C. Kemp (TM),J. Lo- gan, J. Neuby- Varty, and D. Parkes. We thank the ZRPA who assisted in the trapping of some of the eagles. We thank W.S. Clark, A.C. Kemp, and an anonymous referee for making very helpful comments on the manuscript. Literature Cited Amadon, D. and j. Bull. 1988. Hawks and owls of the world: a distributional and taxonomic list. Proc. West. Found. Vertebr. Zool. 3:294-357. Ash, J.S. 1981. Ayres’ Eagle Hieraaetus dubius in Ethiopia and Somalia. Scopus 5:54-56. Benson, C.W. 1962. Some additions and corrections to a March 2003 Ayres’s Hawk-Eagle in Zimbabwe 53 checklist of the birds of Northern Rhodesia. Occas. Pap. Nat. Mus. South. Rhod. 266:631-652. Biggs, H.C., R. Biggs, and A.C. Kemp. 1978. Measure- ment of raptors. Pages 77-82 in A symposium on Af- rican predatory birds. Transvaal Museum, Pretoria, South Africa. Brown, L.H. 1952. On the biology of large birds of prey in the Embu District, Kenya Colony. Ibis 94:577-620. . 1953. On the biology of large birds of prey in the Embu District, Kenya Colony. Ibis 95:74—114. . 1966. Observations on some Kenyan eagles. Ibis 108:531-572. . 1974. Is poor breeding success a reason for the rarity of Ayres’ Hawk-Eagle? Ostrich 45:145-146. AND P.R.A. Davey. 1978. Natural longevity, as de- termined by plumage variation, in Ayres’ eagle Hi- eraaetus dubius. Bokmakierie 30:27-31. , E.K. Urban, and K. Newman. 1982. The birds of Africa. Vol. 1. Academic Press, London, U.K. Clark, W.S. 1999. Ayres’ Hawk-Eagle Hieraaetus (dubius) ayresii sightings in Cameroon. Bull. ABC 6:115-116. Davison, B. 1998. Raptor communities in hill habitats in south-eastern Zimbabwe. M.S. thesis, Rhodes Univ., Grahamstown, South Africa. DEL Hoyo, J., A. Elliott, and J. Sargatal. 1994. Hand- book of the birds of the world. Vol. 2. New world vul- tures to guineafowl. Lynx Edicions, Barcelona, Spain. Dewhurst, C.F., G.R. Gunningham Van Someren, R.G. Allan, and S. Thomsett. 1988. Observations on the breeding ecology of Ayres’ Eagle Hieraaetus ayresii at Karen, Nairobi, Kenya. Gabar 3:85—93. Edelstam, C. 2001. Raptor moult patterns and age cri- teria. Pages 50-53 in Ferguson-Lees, J. and D.A. Chris- tie [Eds.], Raptors of the world. Houghton Mifflin Co., Boston, MA U.S.A. Finch-Davies, C.G. and A.C. Kemp. 1980. The birds of prey of southern Africa. Winchester Press, Johannes- burg, South Africa. Gargett, V. 1990. The Black Eagle: a study. Acorn Books and Russel Friedman Books, Randburg and Halfway House, South Africa. Grimes, L.G. 1987. The birds of Ghana. An annotated check-list. British Ornithologists’ Union, London, U.K. Hanmer, D.B. 1997. Miscellaneous measurements and molt of non-passerine birds from Mozambique and Malawi: raptors and owls, part 2. Honeyguide 43:214— 218. Harrison, J.A., D.G. Allan, L.G. Underhill, M. Herre- mans, A.J. Tree, V. Parker, and CJ. Brown [Eds.]. 1997. The atlas of southern African birds, Vol. 1. Non- passerines. BirdLife South Africa, Johannesburg, South Africa. Hartley, R. 1982. Notes on Ayres’ Hawk Eagle in the Mutare area. Honeyguide 110:2^27 . . 1989a. Conservation projects of the Zimbabwe Falconers’ Club. Quagga 24:16-18. . 1989b. Ayres’ Eagle. Page 139 in P.J. Ginn, W.G. Mcllleron, and P. le S. Milstein [EdS.], The complete book of southern African birds. Struik-Winchester, Cape Town, South Africa. . 1993. Falconry as an instrument for conservation in Zimbabwe. 7 tc Wilson, R.T. [Ed.], Birds and the Af- rican environment: Proc. VIII Pan-Afr. Ornithol. Congr. Ann. Mus. R. de VAfr. Centrale (Zoologie) 268:105-110. . 1998a. Yellow-billed Kites and other raptors in Tsholotsho and Bulalima-Mangwe Districts. Honey- guide 44:190-198. . 1998b. Raptor migration and conservation in Zimbabwe. Torgos 28:135-150. . 2000. Ecology of Taita, Peregrine and Tanner Fal- cons in Zimbabwe. Pages 87-105 in R.D. Chancellor and B.-U. Meyburg [Eds.], Raptors at risk. WWGBP and Hancock House, Berlin, Germany. , K. Hustler, and PJ- Mundy. 1996. The impact of man on raptors in Zimbabwe. Pages 337-353 m D.M. Bird, D.E. Varland, and J. Jose Negro [Eds], Raptors in human landscapes. Academic Press, Lon- don, U.K. Herremans, M. 1994. More records of the Ayres’ Eagle {Hieraaetus ayresii) in Botswana. Babbler 28:32—33. AND C. Brewster. 1994. The status of the Ayres’ Hawk-Eagle Hieraaetus ayresii in Botswana. Babbler 26- 27: 12-13. Hough, J. 1950. Birds of prey from the eastern districts of Mashonaland. Babbler b:2-6. Howells, W. 1985. The birds of the Dande communal lands, middle Zambezi valley, Zimbabwe. Honeyguide 31:26-48. Humphrey, P.S. and K.C. Parkes.1959. Comments on the study of plumage succession. Auk 80:496-503. Irwin, M.P.S. 1981. The birds of Zimbabwe. Quest Pub- lishing, Salisbury, Zimbabwe. . 1984. Recent reports. Honeyguide . 1985. Recent reports. Honeyguide Kemp, A. and M. Kemp. 1998. Sasol birds of prey of Africa and its islands. New Holland, Cape Town, South Af- rica. Lendrum, A.L. 1975. Observations on Ayres’ Hawk-Eagle. Honeyguide 81:41. . 1982. Ayres’ Hawk-Eagle in Bulawayo, Zimbabwe. Honeyguide 110:15—22. Lockwood, G. 1979. Ayres’ Hawk-Eagle in Parkhurst. Wit- watersrand Bird Club News 105:5. Maclean, G.L. 1993. Roberts’ birds of southern Africa 6* Ed. Trustees John Voelcker Bird Book Fund, Cape Town, South Africa. Mendelsohn, J.M., A.C. Kemp, H.C. Biggs, R. Biggs, and C.J. Brown. 1989. Wing areas, wing loadings and wing spans of 66 species of African raptors. Ostrich 60:35— 42. Middleton, A.G. 2000. The effects of cultivation on the nesting densities of eagle populations in the south- eastern Lowveld of Zimbabwe. Pages 1—40 in A.G. 54 Hartley AND Mundy VoL. 37, No. 1 Middleton, M. Parish, and KJ. Hardy, Lowveld [Eds.], Newcastle University expedition to Triangle, south- east Lowveld, Zimbabwe. Unpubl. report. Newton, I. 1979. Population ecology of raptors. A. 8c T. Poyser, Berkamsted, U.K. Oberprieler, U. and B. Gillie. 2002. Raptor identifica- tion guide for southern Africa. Rollerbird Press, Pi- negowrie and Random House, Parklands, South Af- rica. Phillips, R. 1978. The nesting of Ayres’ Hawk-Eagle Hi- eraaetus dubius in the south-eastern Lowveld. Honey- guide 94:27—30. Seavy, N.E. 2000. Observations at an Ayres’ Hawk-Eagle nest in Kibale National Park, Uganda. J. Raptor Res. 34:59-60. Sinclair, J. 1984. Field guide to the birds of southern Africa. Struik, Cape Town, South Africa. Solomon, D. 1995. Ayres’ Hawk-Eagle in the Mukuvusi woodland. On Safari 9:17. Steyn, P. 1982. Birds of prey of southern Africa. David Philip, Cape Town, South Africa. Tarboton, W.R. and D.G. Allan. 1984. The status and conservation of birds of prey in the Transvaal. Trans- vaal Mus. Monogr. No. 3, Pretoria, South Africa. Tree, A. 1989. Recent reports. Honeyguide'ib\\'2.2>,\^4:. . 1990. Recent reports. Honeyguide 36:46,200. . 1991. Recent reports. Honeyguide . 1992. Recent reports. 38:35,83,197. . 1994. Recent reports. 40:39,108,190. . 1995. Recent reports. 41:34,122,246. . 1996. Recent reports. Honeyguide 42:44,116,175, 231. . 1997. Recent reports. Honeyguide Vernon, C.J. 1979. Notes on birds of prey in Zimbabwe- Rhodesia. Honeyguide 99:28-29. Virani, M. and R.T. Watson. 1998. Raptors in the east African tropics and west Indian Ocean islands; state of ecological knowledge and conservation status. J. Raptor Res. 32:28—39. Received 25 February 2002; accepted 22 September 2002 /. Raptor Res. 37(1) :55-62 © 2003 The Raptor Research Foundation, Inc. BREEDING DENSITY AND ALTITUDINAL DISTRIBUTION OF THE URAL, TAWNY, AND BOREAL OWLS IN NORTH DINARIC ALPS (CENTRAL SLOVENIA) Al Vrezec^ National Institute of Biology, Vecna pot 111, SI-1000 Ljubljana, Slovenia Abstract. — Ural {Strix uralensis). Tawny {Strix aluco), and Boreal owl {Aegolius funereus) density and altitudinal distribution were determined using playback to census owls on Mt. Krim (North Dinaric Alps, central Slovenia). Survey points were selected proportionally by altitude according to the relief of the area (320-1060 masl). Density of Ural Owls was estimated to be 2.2 territories/ 10 km^; high relative to published data from Europe, while densities of Tawny (4.0 territories/ 10 km^) and Boreal owls (2.8 territories/ 10 km^) were in the range or lower than other European data. The Tawny Owls were found at significantly lower altitudes (320—850 masl), while Boreal Owls were at higher altitudes (700—940 masl) than expected. I suggest that Ural Owl territories were located in suboptimal habitat for Tawny Owls. The segregation of these owls by altitude in temperate-zone, continuous-montane forests is either a consequence of competitive exclusion or predation. The similarity in altitudinal distribution between Tawny and Boreal owls was low, suggesting that Tawny Owl territories are not suitable habitat for Boreal Owls. At high altitudes, harsh conditions prevent the Tawny Owl from competing with the Ural Owl; an advantage for the Boreal Owl, which was capable of surviving harsh conditions within Ural Owl territories. Further studies are needed to determine competitive exclusion or predation interactions among these owls. Key Words: Boreal Owl, Aegolius funereus; Tawny Owl, Strix aluco; Ural Owl, Strix uralensis; altitudinal distribution', density, Dinaric Alps', Slovenia. DENSIDAD DE ANIDAMIENTO Y DISTRIBUCION ALTITUDINAL DE LOS BUHOS URAL, LEONA- DO Y BOREAL EN LOS ALPES DINARICOS DEL NORTE (ESLOVENIA CENTRAL) Resumen. — La densidad y distribucion altitudinal de los buhos ural {Strix uralensis), leonado (Strix aluco), y boreal (Aegolius funereus o buho de Tengmalm) se determine usando play back para censar buhos en Mt. Krim (alpes dinaricos del norte, Eslovenia central). Los puntos de conteo fueron seleccionados proporcionalmente de acuerdo con el relieve del area (320-1060 msnm). La densidad de los buhos urales se estimo en 2.2 territorios/10 km^; relativamente alta con respecto a dates publicados en Europa, mientras que la densidad de buhos leonados (4.0 territorios/10 km^) y buhos boreales (2.8 territories/ 10 km^) estuvieron en el range o por debajo de otros datos tornados en Europa. Los buhos leonados se encontraron en alturas significativamente mas bajas (320-850 msnm), mientras que los buhos bore- ales se encontraron altitudinalmente mas arriba (700-940 msnm) de lo esperado. Sugiero que los ter- ritorios del buho ural estaban localizados en habitat suboptimo para buhos leonados. La segregacion de estos buhos por altitud en un bosque montano continue de zona templada es tanto una consecuencia de exclusion competitiva como de depredacion. La similitud en la distribucion altitudinal entre buhos leonados y boreales fue baja, sugiriendo que los territories de buho leonado no proveen de habitat adecuado a los buhos boreales. En elevadas altitudes, las duras condiciones impiden al buho leonado competir con el buho ural; una ventaja para el buho boreal, el cual fue capaz de sobrevivir a dificiles condiciones dentro de los territorios del buho ural. Se requiere de mayores estudios para determinar la exclusion competitiva o las interacciones de depredacion entre estos buhos. [Traduccion de Cesar Marquez] Bird densities and their altitudinal distributions are influenced by habitat quality, competitive behavioral ^ E-mail address: al.vrezec@uni-lj.si mechanisms, food supply, and availability of suitable nest sites (Gill 1995, Zabel et al. 1995, Newton 1998). For owls, defense of an exclusive hunting territory plays an important role; the size of the territory is 55 56 Vrezec VoL. 37, No. 1 often governed by owl mass and prey scarcity (Schoe- ner 1968, Carbone and Gittleman 2002). In the Dinaric Alps (western Balkan Peninsula) the ecology of Ural {Strix uralensis), Tawny {Strix aluco ) , and Boreal owls {Aegolius funereus) is poorly known. Density and some breeding habits of Ural Owls are documented for Slovenia (Mihelic et al. 2000), but only anecdotal data are available from other parts of the region (Kralj 1997). For the Taw- ny and Boreal owls, only distribution is known. Pri- or to this paper, the altitudinal distribution of owls from Dinaric Alps were based on several incidental observations (Tome 1996, Mihelic et al. 2000). In this paper, I present data on density and al- titudinal distribution of the Ural, Tawny, and Bo- real owls from the north part of Dinaric Alps. Of particular value are data on the Ural and Boreal owls, because the study area is at the southern limit of their distribution and is disjunct from the main European population; both species are glacial rel- ics. SriJDY Area The field work was done on Mt. Krim (14°25'55"E, 45°58T.5"N) in a study area covering 140 km^, 10 km south of Ljubljana (central Slovenia), between 1997 and 2000. Mt. Krim is a medium-high mountain (290-1107 masl) with a widely extended plateau. Most of the study area is north facing and is within the Dinaric zoogeo- graphical region (Mrsic 1997), part of the north Dinaric Alps. Clearings or nonforest areas, both natural and man- made, represent 25% of the study area. They are small and dispersed, mostly around the settlements. The mixed forest is widespread (75%), belonging to the association of Dinaric beech forest with fir (Omphalodo-Fagetums. lat.) appearing in different subassoeiations. The most fre- quent subassociation is Omphalodo-Fagetum asaretosum eu- ropaei (syn. Abieti-Fagetum dinaricum clematidetosunv, for de- scriptions see Puncer 1980). Dominant tree species are silver fir {Abies alba), Norway spruce {Picea ahies), and beech {Fagus sylvatica). Most of the trees in the forest have trunk diameter >30 cm. (Furlan 1988, Slovenian Forest Service unpubl.) Meiiiqds To estimate owl density and altitudinal distribution, 41 survey points were selected from the base to the top of Mt. Krim. Survey points were selected proportionally by altitude. Density was estimated in the breeding season 1998 only, but the data for altitudinal distribution were collected between 1997 and 2000. Survey points were lo- cated about 1000 m apart, a distance recommended by Holmberg (1979) and Zuberogoitia and Campos (1998). The detection of owls was enhanced by using call play- back (Forsman 1983, Redpath 1994, Zuberogoitia and Campos 1998, Appleby et al. 1999). Recordings of male territorial calls of Ural, Tawny, and Boreal owls were used (Roche and Mebs 1989). Surveys were conducted from dusk to approximately midnight during the spring and summer, up to three times per month. Playbacks were broadcast for 10 min, followed by a 5-min listening pe- riod, at each survey point; a sampling interval suggested to be adequate for detecting most owls that are occupy- ing a territory (Zuberogoitia and Campos 1998). On a specific survey night, only one species’ call was used dur- ing broadcast sampling. 1 estimated human detection of the playback in forest habitat with the equipment used in the survey at a dis- tance of ca. 500 m. This distance was used to define the effective survey area (0.78 km^) around each survey point. I assumed that each response at a point repre- sented an occupied territory. The presence of two owl territories at one point count was recorded only if two males were detected at the same time. Spontaneous call- ing owls, that were estimated to be outside (>500 m from point) the survey area, were excluded from further anal- ysis. Crude density was estimated as the sum of all survey areas at all altitudes divided by the number of detected owl territories. Because territory-size data were not ob- tained, only approximations of ecological density were possible. Ecological densities (the number per unit of habitat space; Odum 1971) were calculated from survey areas within the lowest and the highest recorded altitude for species; only forest-covered areas were used in the analysis. Similar approaches to approximate owl densities were employed by Penteriani and Pinchera (1990) and Diller and Thome (1999). The owl altitudinal distribution was presented as a rel- ative abundance index. This standardized relative abun- dance was calculated as number of owl territories per survey point in a 100 m altitudinal interval per year. 1 defined altitudinal range with 50% of all detected owl territories as the center of altitudinal distribution for each species. Disproportionate use of a particular altitude by each owl species was tested with Mann-Whitney Utest (Sokal and Rohlf 1995) comparing the altitudinal distri- bution of all survey areas with the altitudinal distribution of survey areas with occupied owl territories. A similarity index of altitudinal distribution between three owl spe- cies was calculated using the MacNaughton-Wolf similar- ity index as suggested by Mikkola (1983) and Korpimaki (1986). Resui.ts In 1998, 343 survey points were completed in 25 nights. Seven Ural Owl, 13 Tawny Owl, and nine Boreal Owl territories were found on 41 survey ar- eas. Crude densities of Ural, Tawny, and Boreal owls were estimated at 2.2, 4.0, and 2.8 territories/ 10 km^ respectively. Estimated ecological density was measurably higher only in the Boreal Owl (Ta- ble 1). In years 1997-2000, 582 survey points were done during 50 nights. Twenty occupied territories of Ural, 23 of Tawny, and 17 of Boreal owls were re- corded. The Ural Owl occurred over the greatest March 2003 Ural, Tawny, and Boreal Owl Density 57 Table 1. Estimated crude and ecological densities of three owl species in 1998 on Mt. Krim. Crude density is the sum of all survey areas divided by the number of detected owl territories (see Methods). Ecological density is the number of owl territories per unit of habitat space (calculated on the basis of sample area) . Ural Owl Tawny Owl Boreal Owl Crude density 2.2 4.0 2.8 (territory/ 10 km^) Ecological density 2.8 5.6 6.9 (territory/ 10 km^) Sample area (km^) 25.1 23.0 13.0 span in altitudinal distribution (410-1060 masl; Fig. 1). The Tawny Owl was found at significant lower elevations, while Boreal Owls occurred at higher altitudes (Table 2, Fig. 1). The greatest sim- ilarity regarding altitudinal distribution was found between Ural and Boreal owls, and the least simi- larity was found between the distribution of the Tawny and Boreal owls (Table 3) . Discussion Ural Owl Density. Density of Ural Owls in North Dinaric Alps (including Mt. Krim) is high relative to other parts of Europe (Table 4) . I suggest three possible explanations for these differences. First, different census methods may be responsible; we counted singing males, while active nests were counted in some other studies. With the playback technique, it is impossible to distinguish between breeding and nonbreeding pairs or even nonmat- ed, but territorial individuals. Pairs can occupy a territory even if they do not breed (Saurola 1989). The proportion of Ural Owl pairs that actually lay eggs varies between 12 and 87% (Pietiainen 1988). « ■o 3 1000-1090 900-990 800-890 700-790 600-690 500-590 400^90 300-390 1000-1090 900-990 800-890 700-790 600-690 500-590 400-490 300-390 1000-1090 900-990 800-890 700-790 600-690 500-590 400-490 300-390 Aegolius funereus Sthx aluco Strix uralensis 0.2 0.4 0.6 0.8 1 1.2 1.4 Relative Abundance [territories per survey point] 1.6 Figure 1. Altitudinal distribution of Ural {N = 20), Tawny {N = 23), and Boreal owl {N = 17) on Mt. Krim. Relative abundance was calculated as the number of owl territories per survey point in each 100 m altitudinal interval per year. 58 Vrezec VoL. 37, No. 1 Table 2. Altitudinal distribution of all survey areas compared to survey areas, where Ural Owls, Tawny Owls, and Boreal Owls were detected in years 1997-2000 on Mt. Krim. Data were compared with Mann-Whitney Utest. Median Altitude (m) Minimum- Maximum Center of Altitudinal Distribution"* U P All survey areas (N = 4) 710 320-1060 520-820 Survey areas with Ural Owls (N = 20) 800 410-1060 650-840 271.0 >0.05 Survey areas with Tawny Owls {N = 23) 490 320-850 410-610 223.5 0.033 Survey areas with Boreal Owls (N= 17) 800 700-940 770-850 163.5 0.037 ® Center of altitudinal distribution contains 50% of all detected owl territories. For that reason density is herein presented as oc- cupied territories and not as breeding pairs. Sec- ondly, density in birds is a function of the size of the study area (Bezzel 1982). Areas with low den- sities of owls were considerably larger than my study area (Table 4). Finally, the amount of avail- able food may be responsible. In Slovenia, a large part of Ural Owl’s diet consists of fat dormouse {Glis glis) (Vrezec 2000b), while in other countries, voles {Microtus spp.) are the predominant prey spe- cies (Sladek 1961/62, Mikkola 1972, 1983, Jader- holm 1987, Korpimaki and Sulkava 1987, Glutz von Blotzheim and Bauer 1994, Czuchnowski 1997, Stiirzer 1998, 1999). Fat dormouse is an abundant small mammal in Slovenian forests (Krystufek 1991) and its mass is approximately four times as much as voles, that is 245 g compared to 64 g (Krystufek 1991, Glutz von Blotzheim and Bauer 1994) . Prey availability and prey body mass are im- portant factors that inversely affect the size of a predator’s territory (Schoener 1968, Zabel et al. 1995) . One consequence of large territories is a relatively lower density of owls (LaHaye et al. 1997). Prey availability on Mt. Krim could result in small territories and may explain the observed Table 3. Similarity in altitudinal distribution between three owl species on Mt. Krim (MacNaughton-Wolf sim- ilarity index following Mikkola [1983] and Korpimaki [1986]). Ural Owl Tawny Owl Boreal Owl 0.65 0.17 Tawny Owl 0.42 high density of the Ural Owl. However, I have no data on fat dormouse density on Mt. Krim to sup- port this suggestion. Tawny Owl Density. According to data from Eu- rope (Table 4) , density of Tawny Owls on Mt. Krim was low; perhaps, because of interspecihc compe- tition with the larger Ural Owls (Mikkola 1983, Konig et al. 1999, Vrezec 2000a). The Tawny Owl is lowland species in Slovenia (Tome 1996), and that was confirmed also on Mt. Krim (Fig. 1). Boreal Owl Density. The Boreal Owl has a rela- tively small territory (Konig et al. 1999). Neigh- boring males can sometimes sing very close to each other without any aggressive interactions (Konig et al. 1999), so ecological densities can exceed 10 pairs/ 10 km^ (Table 4). Three important factors that limit Boreal Owl density in forests were ad- vanced by Locker and Fliigge (1998): (1) presence of suitable nest tree holes of Black Woodpeckers (Dryocopus martins); (2) optimal foraging areas, large clearings or windfall areas; and (3) absence of the Tawny Owl, an important predator of Boreal Owls throughout Europe. In the area of Mt. Krim all factors are optimal, so high ecological densities are no surprise. Low crude density (Table 2) is the consequence of altitudinal limitations of the spe- cies’ distribution, which was probably caused by the presence of Tawny Owls at lower altitudes. Altitudinal Partitioning. The present study has shown that in North Dinaric Alps, competing owl species are segregated by altitude, an important factor in habitat selection for some species such as for Tawny and Boreal owls. The Ural Owl is the dominant species in the owl guild living in my study area, and its distribution is not restricted by March 2003 Urai., Tawny, and Boreal Owl Density (U Qh o Si w G rS G D O 'rt CJ !i 0 P3 ~a 1 Si p o CO w U P 0 CO 4-1 i; >u 'oj 00 O) 03 Si be 03 103 o 13 G :« he 0 '' 05 05 1-H o; .s§ G 1) G be G -S s ^ "C U 1) X ■3 ^ G CO 13 Ph O d 30 TO G csj rS w d, esj N I N H X 13 G oi CO oo" 05 05 1-H 1/5 o d, 3 rt U TO G a .5 S 30 X 05 00 O rH 05 he o O g Si X K eh N W5 0 s 00 C75 05 13 (5 Si OJ 13 X X 13 13 G G Pm £ TO TO G G eh eh 'g 'g 2 2 P 'C 13 13 w u G G 13 13 Pi Pm q 30 3f5 P I CM X 13 S X 05 i> 05 CT3 05 1-H 05 05 1-H G 13 • G 0 0> a 1 r— S 0 CO X G X X p G u 05 eh G N P 05 1-H u a • ^ 13 o" 05 ct -C^ 05 05 X 05 t-M PQ a T3 p2 "0 X G o CM eh g g 1- •- CO U 13 he X TO G ■ N 0 {h 13 X X - 5 CO o T3 ^ G ^ -2i g G Ph 3 eo P O O > £j 3 - 00 30 00 o o o CM Vi > O G eh P TO G eh 00 05 ^ eh eo 00 05 _G X X u X PSPS o X X r2 3 X X I o V • rH 5-h O •M S-l V H-l U o *c3 a & 00 ;z; u Q 0 0 jX CM CM q en CM CM d iM rM 00 >—1 1 1 CM 0 0 d X ii 11 0 q X 1-1 0 J> 1-1 P d 1 1 CM q d P q q ifi Gvi 7 P o q 1-H CO I o P 00 00 CM 1> I 00 o 3fi CM 1T5 00 O ID ■Tt^ CM CM I CM I O 00 o 3f5 30 CM O d CM X I I O 00 P d eiO P o q CM 13 I/} G 1) TO TO G eh CM i> 30 CM I 00 'PP 00 O d o X o o 00 T)^ o d o 1-H I o eh 13 e3 G iw O 13 N 3 180 0 Total 22121 Israel, the only land bridge between three continents, IS at a junction for birds migrating south from Eurasia to Africa in autumn and north to their breeding grounds m spring (Safriel 1968). In spring the Red Sea and the Gulf of Aqaba/Eilat act as a long deflection barrier di- verting many northbound raptors to Eilat (Shirihai and Christie 1992, Yosef 1995, Grieve 1996). Eilat is at the northern edge of almost 2000 km of continuous desert regions of the Sinai and the Sahara. Hence many birds land here to rest after crossing deserts to the south (Saf- riel 1968, Yosef 1998a). Recent studies document that during the northbound migration Levant Sparrowhawks {Accipiter brevipes) concen- trate in the Eilat region in great numbers (e.g., 45 000- 50 000; Safriel 1968, Shirihai 1987, Yosef 1995, Clark and Yosef 1997, Shirihai et al. 2000, Yosef and Fornasari 2000) and migrate north along the Great Rift Valley towards Syria and Lebanon (Frumkin et al. 1995). Levant Sparrowhawks are considered scarce, and Cramp and Simmons (1980) state that information on their distribution, populations, and status is limited. Recoveries of birds banded at Eilat during northbound migrations are from Romania, Ukraine, Russia, and Syria (Yosef 1998b). During the trapping and banding operations of spring 1996, 1997, and 1998, we noted that although we re- mained in the vicinity of the palm {Phoenix dactylifera) plan- tations until dark we seldom saw Levant Sparrowhawks ar- rive, and yet on the next morning we observed large numbers departing from the palmeries on their migration north along the Syrio-African Rift Valley. Stark and Liechti (1993) suggested that Levant Sparrowhawks might resort to flapping flight to reduce time spent on migration (i.e., that they were time minimizers — that minimize overall mi- gration time from Africa to their Asian and European breeding sites). Using radar, they identified the “signa- ture” wing-beat pattern of Levant Sparrowhawks at night in autumn 1991. This led to a prediction that based on Table 2. Number of Levant Sparrowhawks observed ar- riving after dark at palm plantations at Eilat, Israel, and the numbers counted lifting off the next morning. Num- ber in parentheses represents the percentage of birds counted arriving at roost in the dark, or trapped the next morning, in comparison to those counted at lift off from the same area the following morning. Date N Detected AT Roost (Percent) N Counted at Lift Off N Trapped (Percent) 18 April 1136 (31) 19 April 3111 22 (0.7) 19 April 1201 (24) 20 April 4432 36 (0.8) 20 April 3752 (53) 21 April 7018 31 (0.4) 21 April 6726 (70) 22 April 9652 36 (0.4) 22 April 6040 (71) 23 April 8422 44 (0.5) 23 April 90 (16) 24 April 546 19 (3.5) 24 April 3176 (59) 25 April 5344 53 (1.0) TOTAL 22121 (57) 7 38 525 241 (0.6) the number of birds I observed at the palm groves after sunset, I could predict the volume of the flocks that would take off early the next morning, and thereby increase our trapping success the next morning. Methods and Materials I conducted seven night watches during the peak migra- tion period of the Levant Sparrowhawk (18 April-24 April) in the spring 1998 season. Observations were initiated at sunset and continued until no flocks or individual birds were seen for at least 30 min. I used a Swarovski NC2 night scope (X4 magnification) to time the arrival of the flocks and to estimate their numbers. For convenience of calcu- lation I split the observation period into 30 min blocks 1 opted to observe the northern of the two palm plantations owing to logistic constraints along the Israeli-Jordan border Data were recorded on appropriate observation sheets and are presented as mean values ±SD (Table 1). Results and Discussion I observed after-dark arrival by I,evant Sparrowhawk on all seven nights of observation. For the first 60 min after sunset very few arrivals were noted. An increase in the number of arrivals was noted between 90-150 min post- sunset (Table 1). No Levant Sparrowhawks were observed arriving after 180 min post-sunset. Numbers estimated with the night scope were consistently lower than the numbers observed at lift-off at dawn the next morning (Table 2). I attribute this to two major factors — my visi- bility being restricted to the magnification available m the night scope, and that the limited field of vision at night prevented the discovery of flocks that arrived along the shoreline of the Red Sea from the south to the south- ern palmeries and remained undetected until the next morning. However, the early warning of the numbers ar- 66 Short Communications VoL. 37, No. 1 nving at the roost allowed the trapping team to organize the trapping equipment appropriately and during the study period a record total 368 Levant Sparrowhawk was trapped (Table 2) . This is in contrast to the trapping suc- cess during the two years (1996, 1997) prior to the night observations and four subsequent years (1999-2002) when no such observations were conducted and the numbers of Levant Sparrowhawks trapped was less than 200 individuals per season. Data suggest that south of Israel a larger than previously reported proportion of the Levant Sparrowhawk popula- tion resorts to nocturnal flight in order to minimize the time spent over inhospitable areas. Spaar et al. (1998) showed that migratory strategy depend on feeding condi- tions en route and that in good conditions a nonstop flight strategy of soaring-gliding during daylight hours and flap- ping-gliding flight during the night is the time minimizing strategy. However, under poor conditions, soaring-gliding flight when thermal convection is available and roosting during the night is the energy- and time-minimizing strat- egy. The fact that 22 121 Levant Sparrowhawks, i.e., 57% of total observed, were observed arriving at a single roost site up to two hours after sunset suggests that the latter appears to be the case for the Ixvant Sparrowhawk at Eilat. Stark and Leichti (1993) argued that nocturnal migrants minimized time to join larger flocks for the next day’s migration. Spaar et al. (1998) thought that the very short period of their passage in Israel suggests that the migratory timing of the Levant Sparrowhawks was under strong en- dogenous control and that delays in the migratory timing was compensated by nocturnal flights, as has been shown for Tree Pipits {Anthus trivialis, Jenni 1984). In contrast, Kerlinger (1989, 1995) thought that raptors that under- take long crossings of barriers, and are unable to complete them during daylight, would resort to noeturnal migration. However, the above does not entirely explain why a sub- stantial proportion of the population arrives at Eilat several hours after dark and do not stop at other human settle- ments, to the south of Eilat, in the Sinai Peninsula or along the Red Sea coast. Hence, I assume that these species have knowledge, based on previous migrations or innate, of spe- cihc staging areas along the migratory route at which they roost and will resort to nocturnal migration to reach them. The data presented here, which validate previous radar studies based on “flight signature” (e.g., Casement 1996, Stark and Leichti 1993), repre.sent the first time that mi- gratory raptors have been observed and identified at close range with a nightscope arriving at a roost site. During banding operations we noted that although we remained in the vicinity of the palm plantations until dark we seldom saw Levant Sparrowhawks (Accipiter brev- ipes) arrive, and yet on the next morning we observed large numbers departing from the palmeries on their mi- gration north along the Syrio-African Rift Valley. We pre- dieted that based on the number of birds observed at the palm groves after sunset, I could estimate the volume of the flocks that would take off early the next morning, and thereby increase our trapping success the next morn- ing. I conducted seven night watches during the peak migration period of the Levant Sparrowhawk in spring 1998. Observations were initiated at sunset and contin- ued until no flocks or individual birds were seen for at least 30 minutes. I used a Swarovski NC2 night scope (X4 magniheation) to time the arrival of the flocks and to estimate their numbers. For the hrst 60 minutes after sunset very few arrivals were noted. An increase in the number of arrivals was noted between 90-150 min post- sunset. No Levant Sparrowhawks were observed arriving after 180 min post-sunset. The early warning of the num- bers arriving at the roost allowed the trapping team to organize the trapping equipment appropriately and dur- ing the study period a record total 368 Levant Sparrow- hawk was trapped. The data represent the first time that migratory raptors have been observed and identihed at close range with a nightscope arriving at a roost site. Resumen. — Durante las operaciones de marcaje notamos que aunque permanecieramos en la vecindad de las plan- taciones de palma hasta oscurecer, contadas veces vimos arribar azores del mediterraneo oriental {Accipiter brevi- pes), y con todo, a la mahana siguiente vimos grandes numeros partiendo de las palmeras en su migracion ha- cia el norte a lo largo del valle de la falla Sirio-Africana. Predecimos que con base en el numero de aves obser- vadas en el boscaje de palmas despues del ocaso, podria- mos estimar el volumen de las bandadas que despegarfan temprano en la mahana .siguiente, y por tanto incremen- tar nuestro exito de capturas en la mahana siguiente. Lleve a cabo siete rondas nocturnas durante el periodo pico de la migracion del azor del mediterraneo oriental en la primavera de 1998. Las observaciones se iniciaban al caer el sol y continuaban hasta que ninguna bandada o individuo fuera visto por al menos en 30 minutos. Use lentes nocturnos Swarovski NC2 (X4 de aumento) para registrar el tiempo de arribo de las bandadas y para es- timar sus numeros. En los primeros 60 minutos despues del ocaso se notaron muy pocos arribos. Un incremento significativo de ellos sc noth 90-150 min. posteriormente al ocaso. Ningun azor del mediterraneo oriental fue ob- servado arribando despues de 180 min post-ocaso, El av- iso previo de los numeros que arribaban a las perchas, permitio al equipo de captura organizar adecuadamente el dispositivo de captura, y durante el periodo de estudio un total de 368 azores del mediterraneo oriental fueron atrapados. Los datos representan la primera vez en que rapaces migratorias han sido observadas e identificadas de cerca con unos lentes de vision nocturna, arribando un sitio percha. [Traduccion de Cesar Marquez] Acknowi.edgments I thank members of Kibbutz Elot for their cooperation and the Israel Nature Reserves Authority for their help. Dr. G. Dobler, Swarovski Optics, kindly donated the use March 2003 Short Communications 67 of the N2 night scope for this and other nocturnal stud- ies. R. Spaar, R Kerlinger, K. Bildstein, and an anony- mous reviewer improved an earlier version of the man- uscript. Literature Cited Casement, M.B. 1996. Migration across the Mediterra- nean observed by radar. Ibis 108:461-491. CiARK, W.S, AND R. Yosef. 1997. Migrant Levant Sparrow- hawks Accipiter brevipes at Eilat, Israel: measurements and timing./. Raptor Res. 31:317-320. Cramp, S. and K.E.L. Simmons. [Eds.] 1980. Handbook of the birds of Europe, the Middle East and North Africa. Vol. 2. Hawks to bustards. Oxford Univ. Press, Oxford, U.K. Frumkin, R., B. Pinshow, and S. Kleinhaus. 199.5. A re- view of bird migration over Israel./. Ornithol. 136:127- 147. Grieve, A. 1996. Spring raptor movements at Gebel el Zeit, Egypt. Sandgrouse 18:61-63. Jenni, L. 1984. Herbstzugmuster von Vogeln auf dem Col de Bretolet unt besonderer Berucksichtigung nach- brutzeit licher Bewegungen. Ornithol. Beob. 81:183- 213. Kerlinger, P. 1989. Flight strategies of migrating hawks. Univ. of Chicago Press, Chicago, IL U.S.A. . 1995. How birds migrate. Stackpole Books, Me- chanicsburg, PA U.S.A. Pennycuick, C.J. 1972. Soaring behaviour and perfor- mance of some East African birds, observed from a motor-glider. Ibis 114:178—218. Porter, R.F. and 1. Willis. 1968. The autumn migration of soaring birds at the Bosphorous. Ibis 110:520-536. Safriel, U. 1968. Bird migration at Eilat, Israel. Ibis 110: 283-320. Shirihai, H. 1987. The birds of Israel. Oxford Univ. Press, Oxford, U.K. and D.A. Christie. 1992. Raptor migration at Elat. Br. Birds 85:141-186. , R. Yosef, D. Alon, G.M. Kirwan, and R. Spaar 2000. Raptor migration in Israel and the Middle East — a summary of 30 years of field research. Tech Publ. Int. Birding Res. Centre Eilat, Israel. Spaar, R. 1997. Flight strategies of migrating raptors; a comparative study of interspecific variation in flight characteristics. Ibis 139:523-535. , H. Stark, and F. Liechti. 1998. Migratory flight strategies of Levant Sparrowhawks: time or energy minimization? Animal Behav. 56:1185-1197. AND H. Stark. 1996. Migratory flight strategies of Levant Sparrowhawks: time or energy minimization? Anim. Behav. 56:1185-1197. Stark, H. and F. Liechti. 1993. Do Levant Sparrowhawks Accipiter brevipes also migrate at night? Ibis 135:233- 236. Yosef, R. 1995. Spring 1994 raptor migration at Elat, Is- rael. /. Raptor Res. 29:127—134. . 1998a. Two decades of studying bird migration at Eilat: a test case of conflicts and paradoxes. Proc Int. Seminar — Migrating birds know no boundaries. 7orgo.s 28:109-117. . 1998b. Clues to the migratory routes of the east- ern flyway of the western Palearctics — ringing and re- coveries at Eilat, Israel (II- Falconiformes). Vbgelwarte 39:203-208. and L. Fornasari. 2000. Biometric differences be- tween age and sex classes of Levant Sparrowhawk Ac- cipiter breruipes on migration at Eilat, Isr. Israel J. Zool. 46:207-214. Received 14 June 2002; accepted 12 December 2002 /. Raptor Res. 37(l):67-70 © 2003 The Raptor Research Foundation, Inc. Nesting Distribution, Food Habits, and Conservation of Osprey on Boavista Island (Archipelago of Cape Verde) Diego Ontiveros^ Departamento de Biologia Animal y Ecologia, Facultad de Ciencias, Universidad de Granada, E-18071 Granada, Spam Key Words: Osprey, Pandion haliaetus; Boavista-, conser- vation; diet, distribution. The Osprey {Pandion haliaetus) is widely distributed around the world and it has suffered heavily from several human impacts such as persecution, disturbances, and fishery practices (Saurola and Koivu 1987). This raptor eats live fish almost exclusively (Hakkinen 1977, 1978, Saurola and Koivu 1987) and therefore its distribution is restricted to the vicinity of favorable fishing waters; e.g., rivers, lakes, and sea coasts (Poole 1989). In ideal conditions Osprey nests are located close to ’ E-mail address: dontive@ugr.es 68 Short Communications VoL. 37, No. 1 the shoreline. However, timber extraction and shoreline development have reduced the availability of suitable nesting sites, likely causing population declines (Ewins 1997). Because Ospreys are generally wary in areas dis- turbed by human activities, they may locate their nest several kilometers from foraging areas (Saurola 1997). Nevertheless, in many areas. Ospreys have adapted to in- tensive human activity by nesting on powerline poles, windmills, bridges, and other structures (Cramp and Sim- mons 1980, Ewins 1997, Saurola 1997). Naurois (1969) estimated 50 pairs of Osprey inhabited the Cape Verde Islands during the 1960s. Through the years, this population has remained about the same size, with 3-8 pairs on each island and 1-2 pairs on the islets (Hazevoet 1995). However, Eerreira and Palma (2000) described a general decline of this population due to per- secution and human pressure, which has caused aban- donment of several accessible nests (only 31-38 breeding pairs reported in 1998). Apart from this information, there is a study on Ospreys in Cape Verde (Naurois 1987); but the diet, nesting distribution, and specihc con- servation problems of the population on the island of Boavista have never been analyzed. Traditionally, this Os- prey population has remained stable because this island was unknown to travelers (Naurois 1969, Hazevoet 1995). Also, Boavista has been only sparsely populated by hu- mans for centuries. However, the plans for the future tourist expansion will increase human disturbances and may threaten the species. In this paper, I examine the population, nesting distribution, characteristics of the nests, food habits, and the possible influence of tourist activities in the Osprey population on Boavista Island. Study Area and Methods The Cape Verde Islands are situated in the eastern At- lantic between 14°48'-17°22'N and 22°44'-25°22'W, 460- 830 km west of Senegal (Fig. 1). There are 10 islands and several islets. Boavista Island is the third largest island of the archipelago, 620 km^ in area with a perimeter of 1 16 km. The topography is generally flat, the highest eleva- tion (Monte Estancia) being only 387 m, and very dry (mean annual rainfall for Boavista is 91 mm [Kasper 1987]). Because the moderating influence of the sur- rounding ocean temperatures are relatively constant, the amplitude of mean temperatures in different months is seldom more than 6“C. Large areas are covered with sand, forming mobile dunes and barren stony plains, but in the interior there are oases with palms (Sena-Martins ct al. 1986). For this study, I gathered preliminary data on the Boav- ista Osprey population from island people and natural- ists. Because Osprey breeding sites are usually close to water (Ewins 1997, Saurola 1997), I conducted transects by vehicle and on foot close to the shoreline during 1999 (total transect length = 121.2 km). Each time I detected an Osprey, I stopped the vehicle and searched by foot for nests and perches. To analyze the Osprey diet in July 1999, 1 collected prey remains and pellets from nests and perches (Hak- Figure 1. Map of Cape Verde Islands, and their position relative to West Africa. kinen 1978) used by five different pairs. All pellets and 95.4% of the remains were collected near perches, and the rest of the prey remains came from nests. The prey items were identified to the species level (when possible) by comparing them with a local reference collection and fish-identification guides (Muus and Dahlstrpm 1971, Rojo 1988, Muus et al. 1999). Finally, I contacted with the Cape Verde Authorities and the Cape Verde Nature 2000 project coordinator, for information on current and planned tourism development and research projects for Boavista Island. Results and Discussion The Osprey population of Boavista is composed of at least eight pairs that are distributed regularly along the March 2003 Short Communications 69 Table 1. Diet of the Osprey on Boavista Island (Cape Verde). Prey Frequency OF Prey Percent Frequency Parrotfish 127 89.4 Salemas 2 1.4 Bennett’s flying fish 2 1.4 Dentex 5 3.5 Trachichthydae 1 0.1 Unidentified fish 5 3.5 Total 142 shoreline. The mean distance between nests was 12.6 ± 2.1 km (range = 10.5-16.5 km; distance between pairs measured as km of coast from the nests of neighboring pairs) . On Boavista Island, trees are practically non-exis- tent and, thus, Ospreys normally occupy other nest sites. I found nests on small cliffs {N = 7) , on the mast of a beached boat (N = 1), on the ground {N = 2), on the ground on a small islet {N = 1 ) , and one nest on a palm- tree {Phoenix sp.) . The occupied Osprey breeding sites in Boavista Island are close to the shoreline, at a mean distance of 134.8 ± 306.7 ra from the water (range = 0-950 xn; N = 8). This location minimizes energy expenditure (Garher 1972, Henny et al. 1978, Ewins 1997) for Ospreys, which gen- erally forage on the coastline. The availability of suitable nest sites may be limiting local breeding populations (Newton 1979, Ewins 1997) as I found few structures suitable for supporting nests during my survey. The number of Ospreys in Boavista probably has remained stable during the last few years, but increased tourism on this island may eventually cause a population decline (Ferreira and Palma 2000) . I found that the two nests on the ground, which were vulnerable to disturbance, did not have prey remains indicating no recent use. Therefore, the scarcity of alternative nesting cliffs and the human disturbance of some existing nest sites may be a threat to the future of this population. As Ospreys clean and tear the fish before eating them (Moll 1962), only 11 pellets were found in the study area. These pellets corresponded to 11 prey, and the rest of information on food habits came from collected remains. Six different types of prey were identified among the 142 prey items found in the study area (Table 1). The abun- dance of Sparisoma sp. in the diet of this raptor on Boav- ista Island is notable. This genus is the most abundant among the medium-size fish around this island, thriving in water up to 2 m deep (Ontiveros unpubl. data), near the surface, where Osprey usually catch their prey (Poole 1989). This agrees with the fact that Osprey takes the most abundant available prey in each moment (Edwards 1988, Poole 1994). However, my data were in opposition to Bannerman and Bannerman (1968), who described the Osprey as having a wide choice of prey in the sea around the Cape Verde Islands and specified that Serran- us cabrilla was the most important prey in this archipelago. In the past, Ospreys were killed in Europe as a com- petitive consumer of fish (Dennis 1991, Saurola 1997). However, in Boavista the fishing is based on large fish species not consumed by Ospreys, such as tuna ( Thunnus thynnus); thus, this raptor has not been considered as a competitor by the people, the bird being traditionally re- spected. Nevertheless, Osprey eggs and nestlings are con- sumed by people as food on some Cape Verde islands (Ferreira and Palma 2000) . The conservation needs of island fauna, and especially of raptors, are generally more urgent than those of con- tinental species, except where limited distributions on continents mimic an island situation (Virani 1995, Virani and Watson 1998). The Cape Verde Osprey population has recently been estimated at 31-38 pairs (Ferreira and Palma 2000) and, as on Boavista Island, tourism projects on many islands of the archipelago represent a threat to the entire island population. The Cape Verde govern- ment is promoting projects to boost tourism in the coun- try in two ways; (1) building tourist centers in undis- turbed coastal habitats and (2) favoring such activities as watching marine turtles, especially the loggerhead sea turtle (Caretta caretta), the most abundant turtle in the sea surrounding Boavista Island (Lopez-Jurado et al 1999). The Osprey is easily detected on the shoreline of the island and can also be a tourist attraction in the area Nevertheless, the current increasing human population threatens the Osprey population on the island and prob- ably on other islands of Cape Verde. In other places m the world, the stabilization of Osprey breeding habitats has been due largely to intensive management, including nest-site protection (Poole 1989, Spitzer 1989, Houghton and Rymon 1997), which has been proposed for Ospreys on the Cape Verde Islands (Ferreira and Palma 2000). If Ospreys are not persecuted, they tolerate human activi- ties quite well (Houghton and Rymon 1997, Saurola 1997). A growing number of adventure lovers and eco- tourists are expected to visit this desertic island in the next years; maintaining a buffer area surrounding nests free of human disturbances during the breeding period (especially for the most accessible nests), could help to conserve the Osprey population while permitting the tourist activities in the archipelago. Efforts in this direc- tion can serve to make the conservation of Boavista wild- life compatible with economic expansion on the island Resumen. — La distribucion y dieta del Aguila Pescadora {Pandion haliaetus) fue estudiada en la Isla de Boavista (archipielago de Cabo Verde) durante 1999 a traves de transectos lineales en vehiculo y a pie, asi como las amen- azas existentes sobre la poblacion. Se localizo una pob- lacion uniformemente distribuida a traves del perimetro de la isla, con una distancia media entre las 8 parejas 70 Short Communications VoL. 37, No. 1 encontradas de 12.6 ± 2.1 km. Debido a la escasez de vegetacion de la isla y la exigua poblacion humana, el Aguila Pescadora nidified en lugares tan variados como roquedos, islotes, el mastil de un barco encallado, una palmera, y el propio suelo. Los nidos se ubicaron invar- iablemente cerca de la orilla del mar, con una distancia media entre el nido y el agua de 134.8 ± 306.7 m, min- imizando de esta forma el gasto energetico de las aves en sus desplazamientos. El 89.4% de las presas capturadas fue Sparisoma sp., uno de los generos de peces mas abun- dantes en aguas somera. Para evitar la influencia del tur- ismo creciente, es conveniente disenar un plan de pro- teccidn para la isla que lo haga compatible con la conservacion de la especie. [Traduccidn de Cesar Marquez] Acknowledgments 1 wish to thank J.M. Pleguezuelos and J.J. Negro for reviewing the original manuscript and providing valuable suggestions, and L.F. Lopez-Jurado and people of Boav- ista, who kindly helped in finding Osprey locations and shared important data. Two anonymous referees also pro- vided helpful comments on the manuscript. Literature Cited Bannerman, D.A. and W.N. Bannerman. 1968. History of the birds of the Cape Verde Islands. Pages 287-291 in R. Oliver and D. Boyd [Eds.], Birds of the Atlantic Is- lands. Vol. IV. Oliver and Boyd, Ltd. Edinburgh, U.K. Cramp, S. and K.E.L. Simmons. 1980. The birds of the western palearctic. Vol. 11. Oxford Univ. Press, Ox- ford, U.K. Dennis, R, 1991. Ospreys. Colin Baxter Photography Ltd., Lanark, Scotland. Edwards, T.C. 1988. Temporal variation in prey refer- ence patterns of adult ospreys. Auk 105:244—251. Twins, PJ 1997. Osprey (Pandion haliaetus) populations in forested areas of North America: changes, their causes, and management recommendations. J. Raptor Res. 31:138-150. Ferreira, J. and L. Pai.ma. 2000. The Osprey Pandion hal- iaetus in the Cape Verde Islands: distribution, popu- lation trends, and conservation problems. Pages 721- 727 in R.D. Chancellor and B.-U. Meyburg [Eds.], Raptors at risk. Part 9. WWGBP, Hancock House, Mid- rand, South Africa. Garber, D.P. 1972. Osprey nesting ecology in Lassen and Plumas Counties, California. M.S. thesis, California State Univ., Humboldt, CA U.S.A. Hakkinen, I. 1977. Food catch of the Osprey Pandion hal- iaetus during the breeding season. Ornis Fenn. 54: 166-169. . 1978. Diet of the osprey Pandion haliaetus in Fin- land. Ornis Scand. 9:111-116. Hazevoet, C.J. 1995. The birds of the Cape Verde Is- lands. British Ornithologists’ Union, Dorchester, U.K. Henny, C.J., D.J. Dunaway, R.D. Mulette, and J.R. Ko- PLIN. 1978. Osprey distribution, abundance, and sta- tus in western North America: the northern Califor- nia population. Northwest Sci. 52:261-271. Houghton, L.M. and L.M. Rymon. 1997, Nesting distri- bution and population status of U.S. Ospreys 1994. J Raptor Res. 31:44-53. Kasper, J.E. 1987. Ilha da Boa Vista, Cabo Verde: aspectos historicos, socias, ecologicos e economicos; tentativa de analise. Institute Caboverdeano do Livro, Praia. Lopez-Jurado, L.F., J.A. Mateo, and P. Geniez. 1999. Los reptiles de la Isla de Boavista (archipielago de Gabo Verde). Bol. Asoc. Herpetol. Esp. 10:10—13. MoiJ., K.H. 1962. Der Fischadler. Die Neue Brehm Buch- erei 309. A. Ziemsen Verlag, Wittenberg Lutherstadt, Germany. Muus, B.J. and P. Dahlstr0M. 1971. Guia de los peces de mar del Atlantico y del Mediterraneo. Ed. Omega, Barcelona, Spain. , J.G. Nielsen, P. Dahi.str0m, and B.O. N«tr0M. 1999. Sea fish. Scandinavian Fishing Year Book. He- dehusene, Denmark. Naurois, R. 1969. Notes breves sur I’avifaune de I’archipel du Cap-Vert: Faunistique, endemisme, ecol- ogie. Bull. Inst. Fondan. Afr. Noire 31:143-218. . 1987. Le Balbuzzard {Pandion haliaetus L.) aux ilex du Cap Vert. Ann. Mus. Civ. Star. Nat. 86:657-682. Newton, I. 1979. Population ecology of raptors. T. & A.D Poyser, Berkhamsted, U.K. Pooi.E, A.F. 1989. Ospreys: A natural and unnatural his- tory. Cambridge Univ. Press., Cambridge, U.K. . 1994. Family Pan dionidae (Osprey). Pages 42-51 mj. del Hoyo, A. Elliott, andj. Sargatal [Eds.], Hand- book of the birds of the world. Lynx Ed., Barcelona, Spain. Rojo, A.L. 1988. Diccionario enciclopedico de anatomia de peces. Monografias Institute Espanol de Oceono- grafia. Ministerio de Agricultura, Pesca y Alimenta- cion. Madrid, Spain. Saurola, P.L. 1997. The Osprey {Pandion haliaetus) and modern forestry: a review of population trends and their causes in Europe./. Raptor Res. 31:129-137. and J. Koivu. 1987. Saiisksi. Kanta-Hameen Lin- tumiehet, Forssa, Finland. Sena-Martins, D.A., J.M. Moreno, and J.M Gomes. 1986, La desertification aux ilex du Cap-Vert. Ministry of Rural Development and Fisheries, Praia. Spitzer, P.R. 1989. Osprey. Pages 22-29 in B.G. Pendle- ton [Ed.], Proc. Northeast Raptor Management Sym- posium and Workshop. Natl. Wildl. Fed. Sci. Tech Ser. No. 13. Washington, DC U.S.A. Virani, M. 1995. The ecology of the endangered Sokoke Scops Owl Otus ireneae. M.S. thesis, Leicester Univ., Leicester, U.K. and R.T. Watson. 1998. Raptors in the east Afri- can tropics and western Indian Ocean islands: state of ecological knowledge and conservation status. J Raptor Res. 32:28-39. Received 2 May 2002; accepted 17 November 2002 Associate Editor: Juan Jose Negro March 2003 Short Communications 71 J. Raptor Res. 37{l):7l-75 © 2003 The Raptor Research Foundation, Inc. Red-shouldered Hawk and Aplomado Falcon from Quaternary Asphalt Deposits in Cuba William Suarez Museo Nacional de Historia Natural, Obispo 61, Plaza de Armas, La Habana CP 10100 Cuba Storrs L. Olson^ National Museum of Natural History, Smithsonian Institution, Washington, DC 20560 U.S.A. Key Words: Aplomado Falcon-, Falco femoralis; Red-shoul- dered Hawk, Buteo lineatus; Antilles', Cuba; extinctions', fossil birds; Quaternary; West Indies. The fossil avifauna of Cuba is remarkable for its diver- sity of raptors, some of very large size, both diurnal and nocturnal (Arredondo 1976, 1984, Suarez and Arredon- do 1997). This diversity continues to increase (e.g., Sua- rez and Olson 2001a, b, 2003a) and many additional spe- cies are known that await description. Not all of the raptors that have disappeared from Cuba in the Quater- nary are extinct species, however. We report here the first records for Cuba of two widespread living species that are not known in the Antilles today. These fossils were obtained during recent paleontolog- ical exploration of an asphalt deposit. Las Breas de San Felipe, which is so far the only “tar pit” site known in the West Indies. Two fossiliferous localities known as San Felipe I and II occur among extinct and active tar seeps in the floor of the San Felipe Valley, Matanzas Province, 5.5 km west of the town of Marti (ca. 22°57'N, 80°58'W; sheet Marti 4084-IV, 1:50 000 map, X502, Y347; map pub- lished in 1986 by the Instituto Cubano de Geodesia y Cartografia) . The age of the deposits is Quaternary, prob- ably late Pleistocene and early Holocene (Iturralde-Vi- nent et al. 1999, 2000). Although the fossil record of birds in Cuba has hitherto been biased by the fact that almost all specimens have come from cave deposits, the tar seeps of San Felipe provide a much better sample of open-country and aquatic birds that seldom or never are preserved in caves. The list of taxa is extensive and in- cludes among other taxa cranes (Grus), thick-knees (Bu- rhinus), storks (Ciconiidae), waterfowl (Anatidae), crows (Corvus), with a diverse variety of raptors and scavengers being especially abundant (Iturralde-Vinent et al. 2000, Suarez 2000, Suarez and Olson 2003a, b, Suarez unpubl. data) . ^ Corresponding author’s e-mail address: olson.storrs® nmnh.si.edu Material Examined Fossils are from the collections of the Museo Nacional de Historia Natural, La Habana, Cuba (MNHNCu). Mod- ern comparative skeletons included specimens of all of the species of Buteo and Falco in the National Museum of Natural History, Smithsonian Institution, Washington, DC (USNM). The following specimens were used for the tables of measurements: Buteo lineatus 16633— 16634, 17952-17953, 18798, 18846, 18848, 18965, 19108, 19929, 290343, 291174-291175, 291197-291200, 291216, 291860-291861, 291883, 291886, 296343, 321580, 343441, 499423, 499626, 499646, 500999-501000, 610743-610744, 614338; Falco femoralis 30896, 291300, 319446, 622320-622321. Family Accipitridae Genus Buteo Lacepede Red-shouldered Hawk Buteo lineatus (Gmelin) (Fig. 1 A-C) Referred Material Proximal end of right femur (MNHNCu P4614), distal halves of right and left tibiotarsi (MNHNCu P4615, MNHNCu P4616), distal end of left tibiotarsus (MNHNCu P4617) and distal halves of right and left tar- sometatarsi (MNHNCu P4618, MNHNCu P4619). Col- lected in November 1988 by Manuel Iturralde-Vinent, Reinaldo Rojas-Consuegra, and Stephen Diaz-Franco at San Felipe II. Comparisons In size and proportions, these specimens agree with the Red-shouldered Hawk {Buteo lineatus) (Table 1), be- ing larger than the Broad-winged Hawk {B. platypterus) and smaller than the Red-tailed Hawk {B. jamaicensis) , the only two species of Buteo that are year-round residents in Cuba today (Garrido and Garcia Montana 1975). As was the case with fossils from the Bahamas, we took pains to compare the specimens with skeletons of Gray Hawk {B. nitidus), a widespread species of open country that is of approximately similar size and that might be expected to have occurred in the West Indies. But skeletal ele- ments of B. nitidus are consistently more robust than in B. lineatus. 72 Short Communications VoL. 37, No. 1 Figure 1. A-C, modern Red-shouldered Hawk, Buteo lineatus (USNM 17953, on the left in each pair) compared with Cuban fossils of the same species (A, MNHNCu P4614; B, P4615; C, P4618). D-E, modern Aplomado Falcon, Falco femoralis (USNM 291300, on the left in each pair) compared with Cuban fossils of the same species (D, MNHNCu P4606; B, P4609). A, proximal end of right femur in anterior view; B, distal end of left tibiotarsus in anterior view; C, distal end of right tarsometatarsus in anterior view; D, right carpometacarpi in internal view; E, left tarsometatarsi in anterior view. Scale bars = 2 cm. March 2003 Short Communications 73 Table 1. Skeletal measurements (mm) of Cuban fossil and modern Red-shouldered Hawk {Buteo lineatus). Cuban Fossils Modern Measurement Range Mean N Range Mean N Femur Depth of head 4.9 1 4.5-5. 7 5.1 33 Tibiotarsus Least width of shaft at midpoint Distal width 5.6 1 4.8-6. 1 5.5 33 through condyles 9.9-10.9 10.4 2 9.5-11.7 10.9 33 Tarsometatarsus Least width of shaft at midpoint Width of shaft 4.9 1 3. 7-5.0 4.4 31 proximal to meta- tarsal facet 5. 1-5.3 5.2 3 4.1-5.8 5.1 32 Depth of shaft proximal to meta- tarsal facet 4.2-4.4 4.3 2 S.4-4.4 3.9 32 Distal width 12.3 1 11.2-13.3 12.3 32 Depth of middle trochlea 4.9 1 4.4-5.3 4.9 32 Remarks Although the Red-shouldered Hawk now has an en- tirely continental distribution, it has previously been known in the West Indies from a few fossils from cave deposits in the Bahamas (Olson 2000), where it first was described as an endemic genus and species Calohierax quadratus (Wetmore 1937, but see Olson and Hilgartner 1982, Olson 2000). Thus, its occurrence in Cuba might have been predicted. The Bahaman population was prob- ably derived from that of Cuba, as has been the case with many other birds (Brodkorb 1959, Olson and Hilgartner 1982). The Red-shouldered Hawk is ordinarily a species of mesic bottomland forests, so its withdrawal from Cuba and the Bahamas is difficult to understand in light of the fact that ecological conditions in these islands presum- ably have become more mesic since the end of the last glacial period. Potential sources of food were much great- er in Cuba than in the Bahamas, making the disappear- ance of this hawk from Cuba even more enigmatic. The Red-tailed Hawk and the Broad-winged Hawk, each represented by supposedly endemic subspecies {B. jamaicensis solitudinis Barbour and B. platypterus cubanensis Burns), are common on Cuba today (Raffaele et al. 1998, Garrido and Kirkconnell 2000) and both have been re- corded from Quaternary cave deposits on Cuba (Jimenez 1997, Suarez and Arredondo 1997), with the latter being found in the San Felipe II asphalt deposits as well (Suarez unpubl. data.). It hardly seems likely that the disappear- ance of the intermediate-sized Red-shouldered Hawk from Cuba could be related to the disappearance of prey, which would presumably have affected the other species of Buteo as well. Ridgway’s Hawk (Buteo ridgwayi) , endemic to Hispan- iola, is now believed to be a small derivative of B. lineatus (Olson 2000). The prehistoric occurrence of the latter in Cuba suggests that the ancestral stock of Ridgway’s Hawk was probably derived from insular populations of B. lineatus, and most likely from Cuba. Family Falconidae Genus Falco Linnaeus, 1758 Aplomado Falcon Tato/mora/A Temminck, 1822 (Fig. 1 D-E) Referred Materiai. Right carpometacarpus lacking minor metacarpal (MNHNCu P4606) , right carpometacarpus lacking distal end and minor metacarpal (MNHNCu P4607), distal end of left tibiotarsus (MNHNCu P4608), proximal end of left tarsometatarsus (MNHNCu P4609), collected 25 Feb- ruary 2001 by Stephen Diaz-Franco and William Suarez at San Felipe I, area C. Comparisons These specimens agree perfectly in size and characters with the Aplomado Falcon (Falco femoralis) (Table 2). They are much too large for American Kestrel, Merlin, or Bat Falcon (F. sparverius, F. columbarius, F rufigulans) 74 Short Communications VoL. 37, No. 1 Table 2. Skeletal measurements (mm) in fossil and modern Aplomado Falcon (Falco femoralis). Cuban Fossils Modern Measurement Range Mean N Range Mean N Carpometacarpus Total length 41.8 1 37.7-42.5 40.8 4 Proximal width 4.6* 1 4.4-5. 1 4.7 4 Proximal depth Width of major 10.8 1 9.1-11.6 10.7 4 metacarpal at midpoint 3.5-3.6 3.5 2 2.9-3.6 3.3 4 Tibio tarsus Distal width 7.5^ 1 7.0-8.7 7.9 5 Tarsometatarsus Width at level of proximal foramina 6.4 1 5.7-7.0 6.5 5 ^ Estimated. or for the extinct Cuban species F. kurochkini (Suarez and Olson 2001a), and too small for a Peregrine {F. peregri- nus) or Prairie falcon {F. mexicanus). No skeletons were available for Orange-breasted Falcon {F. deiroleucus) , but this species has very different proportions from F. femor- alis, with a proportionately shorter and much more ro- bust tarsometatarsus. Remarks That the Aplomado Falcon once occurred in Cuba is perhaps not unexpected. It is a partially migratory spe- cies with an extremely wide range extending from the southwestern United States to Tierra del Fuego and the Falkland Islands. It inhabits shrub grasslands and .savanna and there is increasing evidence of various species of birds adapted to such conditions in the Quaternary of Cuba. This is the first indication of the species anywhere m the West Indies. Rfsumen. — Procedentes de depositos cuaternarios de as- falto en San Felipe, al norte de la Provincia de Matanzas, se registran por primera vez para Cuba dos especies de rapaces que viven hoy en el continente: Buteo lineatus y Falco femoralis-, este ultimo constituye la primera evidencia de ese taxon en la Subregio n Antillana. [Traduccion de los autorcs] Acknowledgments Travel by W. Suarez to Washington, DC, was made pos- sible by the Alexander Wetmore endowment fund. Divi- sion of Birds, National Museum of Natural History, Smithsonian Institution. W. Suarez is especially grateful to Dr. Helen James (Smithsonian Institution), and Travis and Sydney Olson, for their help and kindness during his visit to Washington, DC. Gilberto Silva Taboada (Mu- seo Nacional de Historia Natural de Cuba) transported fossil specimens to the Smithsonian Institution from Cuba. Photographs are by John Steiner, Smithsonian Photographic Services, and the figure was arranged by Brian Schmidt, Division of Birds. Literature Cited Arredondo, O. 1976. The great predatory birds of the Pleistocene of Cuba. Smithson. Contrib. Paleobiol. 27: I 69-1 88. . 1984. Sinopsis de las aves halladas en depositos fosiliferos pleisto-holocenicos de Cuba. Rep. Invest. Inst. Zool. 17:1—35. Brodkorb, P. 1959. Pleistocene birds from New Provi- dence Island, Bahamas. Bull. Fla. State Mus. Biol. Sci. 4:349-371. Garrido, O.H. and F. Garcia Moniana. 1975. Catalogo de las Aves de Cuba. Academia de Ciencias de Cuba, Habana, Cuba. and a. Kirkconnell. 2000. Field guide to the birds of Cuba. Cornell Univ. Press, Ithaca, NY U.S.A. Iturraj.de-Vinent, M., R.D.E. MacPhee, S. Diaz-Franco, and R. Rojas-Consuegra. 1999. A small “Rancho La Brea” site discovered in Cuba./. Geol. Soc. Jam. 33:20. , , , , W. Suarez, and A. Lom- ba. 2000. Las Breas de San Felipe, a Quaternary as- phalt seep near Marti (Matanzas Province, Cuba). Ca- ribb.J. Sd. 36:300-313. Jimenez Vasquez, O. 1997. Seis nuevos registros de aves fosiles en Cuba. Pitirre 10:49. Olson, S.L. 2000. Fossil Red-shouldered Hawk in the Ba- hamas: Calohierax quadratus Wetmore synonymized with Buteo lineatus (Gmelin). Proc. Biol. Soc. Wash. 113; 298-301. and W.B. Hilgartner. 1982. Fossil and subfossil birds from the Bahamas. Pages 22—56 in S.L. Olson March 2003 Short Communications 75 [Ed.], Fossil vertebrates from the Bahamas. Smithson. Contrib. Paleobiol. 48. Raffaele, H.A., J. Wiley, O. Garrido, A. Keith, and J. Raffaele. 1998. A guide to the birds of the West In- dies. Princeton Univ. Press, Princeton, NJ. SuArez, W. 2000. Contribucion al conocimiento del es- tatus generico del condor extinto (Ciconiiformes: Vulturidae) del Cuaternario cubano. Ornitol. Neotrop. 11:109-122. AND O. Afiredondo. 1997. Nuevas adiciones a la paleornitologia cubana. Pitirre 10:100-102. and S.L. Olson. 2001a, A remarkable new species of small falcon from the Quaternary of Cuba (Aves: Falconidae: Falco). Proc. Biol. Soc. Wash. 114:34-41. AND . 2001b. Further characterization of Caracara creightoni Brodkorb based on fossils from the Quaternary of Cuba (Aves: Falconidae). Proc. Biol. Soc Wash. 114:501-508. and . 2003a. A new species of caracara {Mil- vago) from Quaternary asphalt deposits in Cuba, with notes on new material of Caracara creightoni Brodkorb (Aves: Falconidae). Proc. Biol. Soc. Wash. 16: in press and . 2003b. New records of storks (Cico- niidae) from Quaternary asphalt deposits in Cuba Condor 150-154 Wetmore, a. 1937. Bird remains from cave deposits on Great Exuma Island in the Bahamas. Bull. Mus. Comp. Zool. 80:427-441. Received 22 July 2002; accepted 28 November 2002 J. Raptor Res. lb-11 © 2003 The Raptor Research Foundation, Inc. Subadult and Pale Steppe Eagles Breeding in Mongolia David H. Ellis’^ uses Patuxent Wildlife Research Center, 11410 American Holly Drive, Laurel, AID 20708 U.S.A. BLey Words: Steppe Eagle, Aquila nipalensis. All adult Steppe Eagles {Aquila nipalensis) are report- edly very dark (Ferguson-Lees and Christie 2001). How- ever, the closely related (Wink and Sauer-Gurth 2000) Tawny Eagle (A. rapax) does have a pale adult morph (see Plate 114, Brown and Amadon 1968). Clark (1992) decried the confusion in the scientific literature and in museum collections over the various morphs of the Steppe and Tawny eagles and advanced “criteria for the correct identification of all museum specimens and live birds . . . .” He states conclusively that Steppe Eagles be- come “much darker as adults.” His assertion stems from fieldwork in Israel, India, and Africa, and, more impor- tantly, from handling over 300 museum specimens. His conclusion reaffirms statements by Cramp and Simmons (1980) that subadults are paler than adults and that all very pale birds are young. While it is helpful to examine migrants and wintering birds in evaluating the prevalence of adult morphs, evi- dence to support the claim that no adults are pale must come from the breeding grounds. Even there, if very pale breeders are found, it is necessary to determine if re- ^ Present address: USGS Southwest Biological Science Center, HC 1 Box 4420, Oracle, AZ 85623 U.S.A.; e-mail address: dcellis@theriver.com placement (i.e., newly grown) feathers are light or dark before concluding that the Steppe Eagle has a pale adult morph. Although subadults of some species of Aquila ea- gles are known to at least occasionally breed (e.g., New- ton 1979, Steenhof et al. 1983), I know of no prior record of a subadult Steppe Eagle breeding. During five expeditions to Mongolia from 1994-2000, I found more than 20 Steppe Eagle nests. At one site in arid southeastern Mongolia (115°E, 45°N), we found a very pale bird (Fig. 1). Elsewhere we found two rufous- plumaged birds. All three were attending live young. One of the rufous birds was captured (Ellis et al. 2001) and photographed in hand. The very pale bird was photo- graphed on its nest at a distance of 2 m. In Mongolia, Steppe Eagle adults are generally deep chocolate brown above and below with blackish remiges and rectrices finely barred with black. These dark birds match Clark’s (1992, 1996) descriptions of the dark brown adult plumage. The only consistently present light area in the plumage of dark adults is a broken line of whitish spots on the upper tail coverts. This line is readily visible at great distances when a bird is flying, except when overhead. Some dark birds (probably those molting from subadult plumage) show a dappled line of light brown at the trailing edge of the under wing coverts. Under some light conditions, pale areas at the base of the primaries are apparent on the underside of the wings of some, and probably all, dark adults. The head is gen- 76 Short Communications VoL. 37, No. 1 Figure 1. Pale Steppe Eagle at its nest in southeastern Mongolia. This bird resembles the adult pale morph of the Tawny Eagle as illustrated by J. Harrison (Brown and Amadon 1968:649) and appears identical to a pale subadult or adult Tawny Eagle photographed at the nest by P. Steyn (1973:54-55). Note metal trash in nest, a conspicuous, and sometimes dangerous, feature of Steppe Eagle nests across Mongolia (Ellis and Lish 1999). The lower photo includes a hatchling and 1 egg. This photo also shows two generations of feathers (most conspicuous in the anterior scapulars), feathers of both generations are worn (suggesting that this bird is at least 2 yr old) and both are very light (dem- onstrating that in this individual light plumage was retained). The upper photo shows pale tips on rectrices, “tertials,” and secondaries, a characteristic common to all juvenile Steppe Eagles. This bird fits Clark’s (1992) gray-brown morph, frequently seen in Steppe Eagles for the first 3 or 4 yr. March 2003 Short Communtcations 77 erally uniformly dark, but, as stated by Cramp and Sim- mons (1980), some birds show light (buff) hackles on the nape or back of the head. One extremely pale-head- ed, but otherwise dark, adult from central Mongolia was observed at close range while it perched on the nest rim. It displayed a light area on its head and nape just as pale as in most adult Golden Eagles (A. chrysaeios) . In contrast to the extremely dark adults that are typical for Mongolia, our blond breeder (Fig. 1) was uniformly pale buff on the whole mantle, head, and upper wing coverts. Only the remiges, rectrices, and distal scapulars were heavily pigmented. Even though the rectrices and secondaries were dark, they showed pale tips and pale basal barring and were, therefore, much lighter than those of dark adults. W. Clark (pers. comm.) compared photographs of this bird with his published descriptions of various age classes (Clark 1992, 1996) and concluded that this bird was subadult. Indeed, it closely matches the third/fourth-year Steppe Eagles illustrated by Ferguson- Lees and Christie (2001). The key plumage features leading to the conclusion that the bird was not adult are the pale tips of the rec- trices and secondaries, very evident in the photographs (Fig. 1). In the pale morph of the Tawny Eagle, extensive pale tips reportedly are present only on juveniles. How- ever, Steyn (1973) photographed a breeding, pale morph Tawny Eagle with pale tips of secondaries and rectrices just as seen in our pale bird from Mongolia. Further, Fer- guson-Lees and Christie (2001) illustrate the pale morph adult Tawny Eagle with pale tips. Hence the confusion: was this bird a subadult breeder or do both species have extremely pale adults which show broad, light tips to the rectrices and secondaries? However, the rarity of records of blond Steppe Eagles breeding (i.e., this is the only record now documented) strongly suggests that the pale bird was not an adult. Our two intermediate morph breeders match the ru- fous-tawny, subadult morph described by Clark (1992). As such, these two records substantiate breeding in sub- adult plumage and add the Steppe Eagle to the list of Aquila eagles that breed in subadult plumage. However, there is a remote possibility that the refom-tawny eagles and the blond eagle may have been replacement birds (Phillips et al. 1991), not the biological parents of the young in the nests. Resumen. — Una forma palida y dos oscuras de aguila de las estepas {Aquila rapax) fueron observadas en medio de 20 parejas encontradas anidando en Mongolia. Todas las tres estaban cuidando juveniles vivos. Las caracteristicas del plumaje de las aves de color cafe rojizo — tostado su- gieren que no eran adultos. En consecuencia la anida- cion de subadultos del aguila de las estepas {Aquila ni- palensis) es documentada. La anidacion tambien se documento para un ave de fase palida, pero la edad de esta ave es incierta; tampoco se sabe si esta era la primera forma palida adulta conocida para la especie o, si mas probablemente, esta representa un ave reproductora de 2-, 3-, o 4 ahos de edad. Acknowledgments Our travels were financed largely by NASA-Goddard Space Flight Center and USGS Patuxent Wildlife Re- search Center with supplemental support from the Insti- tute for Raptor Studies, NARC-United Arab Emirates, Howell Wynne, and an anonymous philanthropist. Tsen- geg Purevjav served as my primary translator. Merlin Ellis assisted in the field. William S. Clark kindly improved the manuscript and identified the birds (from photographs) as .subadult. All three JRR reviewers, Rob Simmons, Keith Bildstein, and D. Pepler deserve mention not only for comments on the text but also for improving logic. Literature Cited Brown, L. and D. Amadon. 1968. Eagles, hawks, and fal- cons of the world. McGraw Hill, New York, NYU.S A. Clark, W.S. 1992. The taxonomy of Steppe and Tawny Eagles, with criteria for separation of museum speci- mens and live eagles. Bull. Br. Ornithol. Club 112:150- 157. . 1996. Aging Steppe Eagles. Birding World 9:268- 274. Cramp, S. and K.E.L. Simmons. (Eds.) 1980. Handbook of the birds of Europe, the Middle East, and North Africa. Oxford Univ. Press, Oxford, U.K. Ellis, D.H. and J.W. Lish. 1999. Trash-caused mortality in Mongolian raptors. Ambio 28:536-537. , S.L. Moon, and J.W. Robinson. 2001. Annual movements of a Steppe Eagle {Aquila nipalensis) sum- mering in Mongolia and wintering in Tibet. /. Bombay Nat. Hist. Soc. 98:335—340. Ferguson-Lees, J. and D.A. Christie. 2001. Raptors of the world. Houghton Mifflin, Boston, MA U.S.A. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A. Phillips, R.L., J.L. Cummings, and J.D. Berry. 1991. Re- sponses of breeding Golden Eagles to relocation. Wildl. Soc. Bull. 19:430-434. Steenhof, K., M.N. Kochert, and J.H. Doremus. 1983. Nesting of subadult Golden Eagles in southwestern Idaho. Auk 100:743-747. Steyn, P. 1973. Eagle days; a study of African eagles at the nest. Purnell 8c Sons South Africa, Johannesburg, South Africa. Wink, M. and H. Sauer-Gurth. 2000. Advances in the molecular systematics of African raptors. Pages 135- 147 in R.D. Chancellor and B.-U. Meyburg [Eds.], Raptors at risk. WWGBP/Hancock House, Blaine, WA U.S.A. Received 4 February 2002; accepted 10 November 2002 78 Short Communications VoL. 37, No. 1 J Raptor Res. 37(l):78-83 © 2003 The Raptor Research Foundation, Inc. Two Large Bald Eagle Communal Winter Roosts in Utah Robert Wilson^ and James A. Gessaman Department of Biology, Utah State University, Logan, UT 84322 US. A. Keywords: Bald Eagle; Haliaeetus leucocephalus; migra- tion; roost, Utah; winter. As migratory birds, many Bald Eagles {Haliaeetus leu- cocephalus) breed in the northern portions of the species’ range, and winter in the southern portions (Stalmaster 1987). The location of a particular breeding range is a good predictor of the location of the winter range and the corresponding migratory route. Such associations of winter and summer ranges have been documented in Maine (McCollough 1989), central Canada and U.S. (Gerrard et al. 1978, Griffin et al. 1980, Harmata and Stahlecker 1993), Alaska (Hodges et al. 1987), and the Intermountain West (McClelland et al. 1994). Banding and tracking studies also discovered numerous examples of nesting area fidelity (Gerrard et al. 1978, Jenkins et al. 1999), and winter range fidelity (Gerrard et al. 1978, Har- mata and Stahlecker 1993, McClelland et al. 1994). But such patterns are not rigid; nomadic behavior has also been documented (e.g., Postupalsky 1976, McClelland et al 1994, Jenkins et al. 1999). Numerous studies have described interior populations of wintering Bald Eagles (e.g., Southern 1963, 1964, Knight and Knight 1983, Harmata and Stahlecker 1993, Restani 1997), and many focused on communal roosts (Edwards 1969, Keister and Anthony 1983, Harmata 1984, Isaacs and Anthony 1987, Crenshaw and Mc- Clelland 1989). But only a few studies (e.g., Swisher 1964, Edwards 1969) have examined communal roosts in Utah. The primary aim of this study is to provide occupancy dates, population estimates, locations, and habitat de- scriptions of two large (>200 eagles/night) communal roosts in northern Utah. Study Area The Willard Canyon roost (WCR) is in the Willard Ba- sin which expands to the southeast of the mouth of Wil- lard Canyon (41°25'N, 112°00’W). Vegetation consists mostly of a Douglas-fir (Pseudotsuga menziesii) canopy, with an understory of ninebark {Physocarpus malvaceus), and some white fir {Abies concolor), subalpine fir {A. lasiocar- pa), and limber pine {Pinus flexilis). This vegetation is consistent with the douglas-fir/ ninebark community de- ^ Present address: 2108 Sierra Ridge Ct., Salt Lake City, UT 84109 U.S.A.; e-mail address: bobdogwilson® hotmail.com scribed by Mauk and Henderson (1984). The mouth of the canyon is 4 km from Willard Bay on the west shore of the Great Salt Lake, which has .surface elevations near 1280 m. The Bear River delta is at or near the elevation of the Great Salt Lake and includes the Bear River Mi- gratory Bird Refuge and surrounding area. The Ogden Bay roost (OBR) is located in the Weber River delta (41°12'N, 112°9'W) on the west shore of the Great Salt Lake. The elevation of the delta is only slightly above the elevation of the Great Salt Lake. As a result of high water in the 1980s, numerous deciduous trees along the channels of the lower delta were killed by salt water invasion. Eagles roost in these snags, which are buffered by roadless wetlands nearly 1 km in radius. This site is ca. 23 km southwest of the WCR. Methods The WCR had the highest occupancy of all roosts sur- veyed. Therefore, from the beginning of January to early April, 1998, and from the beginning of December 1998 to early April, 1999, we sampled this site systematically. We surveyed three times a week in the 1997-98 season and twice a week during the 1998-99 season; surveys were conducted from 1230 H until after sunset by one observer located near the mouth of the canyon. Early in the 1997-98 season, we also observed the site between first light and 1000 H. We used 10 X 42 binoculars and a 16-32 X 50 spotting scope to locate and identify eagles at a distance. The exact sampling location varied de- pending upon flight patterns, but the intersection of a private canal road with a gravel pit road was used most frequently. The roost itself could be observed from lo- cations on the ridge north of the canyon from whence we could see the top of the West Fork drainage. When possible, eagles were classified by plumage, according to criteria described by Stalmaster (1987) and Wheeler and Clark (1995). Because eagles that returned to the roost early in the afternoon would sometimes soar above the mouth of the canyon, we only counted eagles seen approaching from the direction of the daily activity centers (DACs). This reduced the chances of redundant counting of individ- uals. The OBR was surveyed every other week during the 1997-98 season, and weekly during the 1998-99 season. However, no data were collected from OBR in March- April 1999. At the WCR, time of return was recorded in 30-min intervals before sunset. Results Occupancy. During the winter of 1997-98, all sites ex- cept WCR had peak numbers in February (Figs. 1 and March 2003 Short Communications 79 250 200 o> Q 150 & 100 50 CM □ Unkno^ ■ Immature □ Ackjit S (D <6 (D 2 lA Figure 1. Number of Bald Eagles at Willard Canyon. 2), while the season high at the WCR of 227 eagles oc- curred 8 March. The mid-winter high of 212 occurred at the WCR on 8 February. This was followed by a decline to 91 preceding the brief influx of eagles in March, and then a sharp decline after the March peak (Fig. 1 ) . Other roosts did not follow a bimodal pattern. The OBR had a season high of 153 eagles on 4 February, after which the number gradually declined to three on 18 April, the last sample (Fig. 2). In the 1998—99 season surveys began 3 December. On this date 58 individuals were already present at WCR, and we saw two eagles enter the roost on 4 November. Num- bers peaked to 211 individuals on 13 December, de- clined, and rose again to a season high on 15 January, when 363 occupied the roost (Fig. 1). Numbers then de- clined, with the exception of 3 March when 101 returned to the roost (Fig. 1). Between 9 January and 18 February the number of eagles at the OBR increased from 48 to the season high of 264 (Fig. 2). Unlike the previous year, eagles appeared to shift from WCR to OBR (Fig. 3). Age Composition. Fromjanuary-March 1998, the max- imum percentage of adults (87.5%) was recorded on 18 January and immatures (88.4%) on 4 March 1998 (Fig. 1). A decline in the percentage of adults began after 8 March. The maximum number of 167 adults was ob- served on 8 February and 119 immatures on 4 March. From December 1998— April 1999, the maximum per- centage of adults (75.5%) was recorded on 15 January and immatures (94.1%) on 30 March (Fig. 1). The max- imum number of 274 adults was observed on 15 January and 72 immatures on 15 January. At the OBR, percentages between December 1997 and April 1998 peaked at 86.9% adults on 4 February and 80 Short Communications VoL. 37, No. 1 Figure 2. M « OI CQ UJ E 3 180 160 140 120 100 80 60 40 20 0 I n n 1 Q ■ oo CM a ■ Immature □Adult t 300 250 v> o> OI 200 •I! UJ 0 150 w 1 100 2 50 I n i 9 DO n 0 D 0 n n n 0 £ Q 6 CM C CD ;z 1 CO 7 CO CM Number of Bald Eagles at Ogden Bay. $ I CO 400 350 300 250 M « O) IQ UJ Salt Creek poisoned Ogden Bay poisoned ■ Willard Canyon □ Ogden Bay o 200 1 1» 3 ^ 100 50 0 t T n Q I CO £ Q cfi CM ! CD 5 a “3 ub SP c CD (D U. I CM lU CD s ch Figure 3. Comparison of eagle numbers using Willard Canyon and Ogden Bay (December 1998-March 1999). March 2003 Short Communications 81 100% immatures on 18 April. The percentage of adult birds at the OBR declined steadily after 10 March. The maximum numbers were 133 adults on 4 February and 39 immatures on 20 February (Fig. 2). During the 1998- 99 season the maximum percentages at the OBR were 91.7% adults on 9 January and 23.6% immatures on 12 February. Numbers peaked at 212 adults and 52 imma- tures on 18 February (Fig. 2). Dlscussion The roost habitat most likely to deteriorate in the near future is that of the OBR. No living trees exist in the roost area, and we detected no evidence of regeneration. Wa- ter management in the upper Weber drainage, in addi- tion to fluctuating water and salt levels of the Great Salt Lake, result in unfavorable growing conditions for trees. Though riparian corridors with living trees exist up- stream of the roost area, they are outside of the Ogden Bay Waterfowl Management Area boundaries, and sub- ject to higher levels of disturbance. The Willard Canyon roost is both the most heavily used and the most sheltered from human disturbance. We de- tected no evidence of habitat degeneration at this site, nor any threat of encroachment. This site is currently on U.S. Forest Service property and, because such large numbers of eagles use this roost consistently, its protec- tion and maintenance should be ensured by Forest Ser- vice policy. Changes in forage abundance also influence roost site selection. Both Edwards (1969) and Keister et al. (1987) observed that eagles tend to use the roost nearest the areas with highest prey densities, and that when prey den- sities changed, eagles selected different roosts according- ly. Shifts from WCR to OBR or elsewhere were not ap- parent during the 1997-98 season, but probably occurred during the 1998-99 season. Between 5 and 15 January 1999 the population of the WCR went from 191 to the season high of 363 eagles, then dropped to 22 by 11 February 1999. Conversely, between 9 January and 18 February the number of eagles at OBR increased from 48 to the season high of 264. Thus, the eagles appeared to shift from WCR to OBR. We suggest that water man- agement at the eagles’ daily-activity centers was the likely explanation for this change. The Bear River Migratory Bird Refuge drained its inner units during the first and second weeks of January trapping and exposing fish in shallow ponds and mudflats. Ponds at Salt Creek Water- fowl Management Area were drained in early January and rotenone was applied on 5 and 6 January (Fig. 3). These events corresponded with the influx of eagles into the WCR. This explanation is supported by the fact that eagles approached the roost mainly from the direction of Bear River Migratory Bird Refuge and Salt Creek Wa- terfowl Management Area. Water levels at Ogden Bay Wildlife Management Area were lowered during the first week of February, and on or near 8 February rotenone was applied to kill carp. Between 6 February and 18 Feb- ruary the number of eagles jumped at OBR from 115- 264. This evidence strongly suggests a migration from one locally-abundant food source to another. Availability of suitable habitat, such as that of the roosts described by Keister and Anthony (1983), Harmata (1984), and Restani (1997) probably explains the loca- tion of observed roosting habitat in northern Utah. The scarcity of roost habitat in close proximity to Bear River Migratory Bird Refuge and other suitable foraging habi- tat is apparent from the geography of the region. The mouth of Willard Canyon, 4 km from Willard Bay and the Great Salt Lake, is the closest portion of the Wasatch Mountains to suitable eagle foraging habitat. Further- more, the top of Willard Basin is the largest basin in that portion of the Wasatch front. Canyons to the south have jagged, rocky sides, and do not open up into sub-alpme basins. Canyons to the north have origins in higher ba- sins, but lack extensive evergreen cover. The canyons on the west slope of the Wellsville mountains (a ridge of the Wasatch beginning 12 km north of Willard Canyon) are steep, narrow, and lacking extensive evergreen coverage. The steep and rocky qualities of the lower portions of Willard Canyon make it difficult for human access when snow and ice are absent, and this canyon is essentially not accessible by people when winter conditions exist. We suggest that eagles would not make a 25 km daily com- mute to Bear River Migratory Bird Refuge (one way) if suitable roost sites existed closer to the foraging areas. Thus, Willard Canyon has the appropriate suite of characteristics including location, altitude, topography, vegetation, and isolation that make it suitable as Bald Ea- gle roosting habitat. Likewise, the OBR is in the grove of trees closest to the foraging areas, and is insulated from human disturbance. Harmata (1984) cited the absence of disturbance as the primary factor in roost site suit- ability. Eagles commuted to and from the roost daily, depart- ing from the roost at first light, and returning from late morning until just after sunset, with rates of return in- creasing later in the afternoon. These patterns are simi- lar to those described by McClelland (1973), Crenshaw and McClelland (1989), and others. Unlike communal roosts in western Montana which are migratory stopovers used during early- and mid-au- tumn (Crenshaw and McClelland 1989, McClelland et al 1994), the Great Salt Lake is a migratory stopover for some individuals, and a southern terminus for others. This is evidenced by population spikes that suggest the passage of birds that winter farther south as well as con- tinuous occupation by some birds between fall and spring migration. Crenshaw and McClelland (1989) observed that peak numbers and percentages of immatures oc- curred earlier than those of adults. We observed opposite patterns: maximum adult numbers and percentages oc- curred in mid-winter, while maximum immature num- bers occurred later (except in March 1998 when maxi- mum numbers of both adults and immatures occurred 82 Short Communications VoL. 37, No. 1 on the same dates at WCR). Maximum immature per- centages always occurred later than those of adults. Often small numbers of immatures were the last eagles to vacate a roost in the spring. Harmata (1984) documented sim- ilar patterns in the San Luis Valley, Colorado. McClelland et al. (1994) speculated that immatures delay spring mi- gration into colder, northern breeding ranges because they cannot breed, and thus, have nothing to gain by migrating early in spring. Occupation dates from early November to early April were very similar to those of winter-range communal roosts in central Utah (Edwards 1969, Platt 1976), in the San Luis Valley (Harmata 1984), and in the Klamath Ba- sin (Keister et al. 1987). The numbers of eagles occupy- ing either WCR or the OBR exceeded all other com- munal roosts documented in Utah. Occupation of roosts m northern Utah begins in early November, at the same time that eagles are vacating roosts in western Montana (Crenshaw and McClelland 1989, McClelland et al. 1994). Furthermore, McClelland et al. (1994) tracked more eagles to Utah from fall concentrations in western Montana, than to any state other than Montana. Many were tracked to the Great Salt Lake area, including Wil- lard Bay. The correspondence of departure and arrival dates between western Montana roosts and northern Utah roosts, and the corroborating radiotelemetry infor- mation, provide evidence that northern Utah is a stop- over or southern terminus for eagles in the “McKen/.ie- Intermountain Flyway” (McClelland et al. 1994). Furthermore, Harmata and Stahlecker (1993) and McClelland et al. (1994) documented multiple examples of winter site fidelity, including some in central Utah. The Willard Canyon and Ogden Bay roosts are among the largest known communal roosts in the lower 48 states, and the largest documented winter roosts within the McKenzie-Intermountain Flyway. They and their associ- ated wetlands provide key habitat for Bald Eagles in in- terior-western North America. Resumen. — Grandes numeros de aguilas calvas {Haliaee- tus leucocephalus) invernantes ocupan dos perchas comun- ales en el norte de Utah: la percha del canon Willard y la percha de la bahia Ogden. Estas estan entre las mas grandes perchas comunales conocidas en los 48 estados mferiores, y es la mas grande percha invernal documen- tada dentro de la via de vuelo inter montahosa Mc- Kenzie. Durante los inviernos de 1997-98 y 1998-99, es- tudiamos estos sitios- percha documentando su localizacion geografica, habitat, y ocupacion. El Canon Willard tuvo un numero maximo anual de 227 aguilas en marzo de 1998 y 363 el 15 de enero de 1999. La percha de la Bahia Ogden tuvo un maximo de 153 aguilas el 2 de febrero de 1998 y 264 hacia el 18 de febrero de 1999. Las aguilas usaron estas perchas regularmente dentro de las estaciones y entre ahos. [Traduccion de Cesar Marquez] Acknowledgments We gratefully acknowledge HawkWatch International for technical support and advice, with special thanks to Steve and Lisa Hoffman. Additional assistance was pro- vided by the staffs of Ogden Bay and Public Shooting Ground Waterfowl Management Areas-Utah Division of Wildlife Resources (UDWR). Funding was provided by the UDWR. Literature Cited Crenshaw, J.G. and B.R. McClelland. 1989. Bald Eagle use of a communal roost. Wilson Bull. 101:626-633. Edwards, C.C. 1969. Winter behavior and population dy- namics of American Bald Eagles in western Utah. Ph.D. dissertation, Brigham Young University, Provo, UT U.S.A. Gerrard, J.M., D.W.A. Whitfield, P. Gerrard, P.N. Ger- rard, and W.J. Maher. 1978. Migratory movements and plumage of subadult Saskatchewan Bald Eagles. Can. Field-Nat. 92:375-382. Griffin, C.R., J.M. Southern, and L.D. Frenzel. 1980 Origins and migratory movements of Bald Eagles win- tering in Missouri./. Field Ornithol. 51:161-167. Harmata, A.R. 1984. Bald Eagles of the San Luis Valley, Colorado: their winter ecology and spring migration. Ph.D. dissertation, Montana State University, Boze- man, MT U.S.A. AND D.W. Stahlecker. 1993. Fidelity of migrant Bald Eagles to wintering grounds in southern Colo- rado and northern New Mexico. J. Field Ornithol. 64 129-134. Hodges, J.I., E.L. Boeker, and A.J. Hansen. 1987. Move- ments of radio-tagged Bald Eagles, Haliaeetus leucoce- phalus, in and from southeastern Alaska. Can. Field- Nat. 101:136-140. Isaacs, F.B. and R.G. Anthony. 1987. Abundance, for- aging, and roosting of Bald Eagles wintering in the Harney Basin, Oregon. Northwest Sci. 61:114—121. Jenkins, J.M., R.E. Jackman, and W.G. Hunt. 1999. Sur- vival and movements of immature Bald Eagles fledged in northern California./. Raptor Res. 33:81-86. Keister, G.P. and R.G. Anthony. 1983. Characteristics of Bald Eagle communal roosts in the Klamath Basin, Oregon, and California, f. Wildl. Manage. 47:1072- 1079. , R.G. Anthony, and EJ. O’Neill. 1987. Use of communal roosts and foraging areas by Bald Eagles wintering in the Klamath Basin. /. Wildl. Manage. 51: 415-420. Knight, S.K. and R.L. Knight. 1983. Aspects of food finding by wintering Bald Eagles. Auk 100:477-484. Malik, R.L. andJ.A. Henderson. 1984. Coniferous forest types of northern Utah. U.S. Forest Service Inter- mountain Forest and Range Experiment Station, Og- den, UT U.S.A. McClelland, B.R. 1973. Autumn concentrations of Bald Eagles in Glacier National Park, Montana. Condor 7b\ 121-123. March 2003 Short Communications 83 , L.S. Young, P.T. McClelland, J.G. Crenshaw, H.L. Allen, and D.S. Shea. 1994. Migration ecology of Bald Eagles from autumn concentrations in Glacier National Park, Montana. Wildl. Monogr. 125. McCollough, M.A. 1989. Molting sequence and aging of Bald Eagles. Wilson Bull. 101:1-10. Platt, J.F. 1976. Bald Eagles wintering in a Utah desert. Am. Birds 30:783-788. Postupalsky, S. 1976. Banded northern Bald Eagles in Florida and other southern states. Auk 93:835-836. Restani, M. 1997. Foraging strategies of migrant Bald Ea- gles exploiting a seasonally concentrated food source. Ph.D. dissertation, Utah State University, Logan, UT U.S.A. Southern, W.E. 1963. Winter population, feeding behav- ior, and seasonal dispersal of Bald Eagles in north- western Illinois. Wilson Bull. 75:42-55. . 1964. Additional observations on wintering Bald Eagle populations; including remarks on bioteleme- tery techniques and immature plumages. Wilson Bull 76:121-137. Stalmaster, M.V. 1987. The Bald Eagle. Universe Books, New York, NY U.S.A. Swisher, Jr., J.F. 1964. A roosting area of the Bald Eagle in northern Utah. Wilson Bull. 76:186—187. Wheeler, B.K. and W.S. Clark. 1995. A photographic guide to North American raptors. Academic Press, San Diego, CA U.S.A. Received 30 December 2000; accepted 29 August 2002 Former Associate Editor: Allen Fish Letters J Raptor Res. 37(1) :84-85 © 2003 The Raptor Research Foundation, Inc. First Breeding Record eor Falco peregrinus in Urban Lima, Peru, with Remarks on the Peruvian Breeding Population The Peregrine Falcon {Falco peregrinus) breeds on every continent except Antarctica, but is absent from large parts of some continents where they occur (most of Amazonia, Saharan regions of North Africa, and central China) and most islands of the Pacific Ocean. In South America the breeding distribution is incompletely known. Through the 1970s, the recorded breeding range for South American peregrines was thought to be from the southern tip of South America north to about 30-35°S (slightly north of Santiago, Chile) with densest known populations in extreme southern Chile, Argentina, and Falkland (Malvinas) Islands (Stresemann and Amadon 1979, Order Falconiformes. Pages 271-425 in E. Mayr and G.W. Cottrell [Eds.], CheckTist of birds of the world, Vol. 1, 2nd Ed. Mus. Comp. ZooL, Cambridge, MA U.S.A.; Cade 1982, Falcons of the world. Comstock, Cornell Univ. Press, Ithaca, NY U.S.A.; and McNutt et al. 1988, Distribution and status of the Peregrine Falcon in South America. Pages 237-253 in T.J. Cade, J H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Falcon populations; their management and recov- ery. The Peregrine Fund, Inc., Boise, ID U.S.A). Based on a pair at a cliff near Yauli, Peru, (3400 m elevation) on 10 September, a date prior to the arrival of Nearctic migrant peregrines, breeding was suggested (Morrison 1939, Ihrs 81:453-486), then circumstantial evidence of breeding was found on a cliff ledge in 1979 near Tacna, Peru (Ellis and Glinski 1980, Condor 82:350-351), and later breeding in Peru was confirmed (Schoonmaker et al. 1985, Condor 87 423—424) as it was also confirmed into Ecuador (Jenny et al. 1981, Condor 83:387) . Therefore, the map of breeding distribution in South America was radically changed by McNutt et al. (1988), so that rather than distribution stopping at about 30-35°S as in most map renditions, the distribution was extended northward along South America west of the Andes Mountains into Ecuador to a latitude near the equator. In fact, in Ecuador, peregrines had been found breeding at Yanayacu as early as 1877, but the record lay hidden in the literature (White 1989, Cowdor 91:995-997) The number of known breeding locations in Peru has increased in the past two decades, perhaps because of a greater search effort, but also there may have been a numerical increase in falcons (Kery 2002,/. Raptor Res. 36:213-217), there are no data to suggest that the South American (Peruvian) populations were reduced or significantly negatively affected by synthetic chemicals (e.g., DDT) (Walker et al. 1973, Antarctic J. 8:29-31) as they were in North America. In addition to local breeders, high number of migrant non-breeding Nearctic peregrines from both the tundra {F p tundrius) and boreal forests (E p. anatum) occur throughout South America, especially along coastal regions and m large populated urban regions during austral spring and summer months (Albuquerque 1978, Rev. Bras. Biol. 38: 727-737; Risebrough et al. 1990, Rev. Bras. Biol. 50:563-574; Silva e Silva 1996, Pap. Avulsos ZooL, Sdo Paulo 39:249- 270; and Silva e Silva 1997, Ararajuha 5:203-208). While there is extensive use of cities by peregrines they do not remain to breed there. Elsewhere, however, peregrines bred occasionally in urban areas of Europe, North America, Australia, and Africa before organochlorine pesticides became extensively used (Cade et al. 1988, Peregrine Falcon populations: their management and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A.; and Cade et al. 1996, Peregrine Falcons in urban North America. Pages 3-13 in D. Bird, D, Varland, andj. Negro [Eds.], Raptors in human landscapes. Academic Press, New York, NY U.S.A). In South America the only previously-known, urban breeder was a pair nesting at a church in the center of Cordoba city, Argentina, in 1981-82 (R. Stranek pers. comm., C. White pers. observ.). With the banning of DDT in the Northern Hemisphere, and as peregrines began to increase there, they invaded urban regions with increasing frequency largely as a result of the release of captive-bred birds which apparently recognized cities as appropriate breeding locations, perhaps related to their early captive experiences (Cade et al. 1996, D. Rockenbauch 1998, Der Wanderfalke in Deutschland. Verlag Christine Holzinger, Ludwigsburg, Germany). But also the invasion into urban environments began to occur, for whatever reason, in regions where peregrines were not reduced by pesticides and where captive-bred birds were not released, such as in most urban centers in Australia (P Olsen pers. comm.), and our Peruvian nesting fits that pattern. Today, resident peregrines are not uncommon in coastal Peru, and they are also observed in interior Peru. While increasing numbers of falcons are sighted during the Austral winter, sightings which exclude wintering non-breeding Nearctic peregrines are also more frequent. For example, in river valleys of central coastal Peru, three pairs were found with nestlings in July 1993. Two pairs produced four young each and a third had three fledglings on 18 July 84 March 2003 Letters 85 (0. Beingolea pers. observ.). In July 1994, two of these pairs raised four young each. A third site inhabited by a pair was not visited. Although large portions of the Peruvian Andes have not been searched, it is likely that the finding of these pairs suggests that the breeding range for peregrines occurs along the entire Andean chain into central Ecuador. It is not known if they extend along the Andes beyond Ecuador or into the Cordillera Central and Oriental of Colombia and Cordillera de Merida of Venezuela; these areas have been extensively surveyed by ornithologists and it appears doubtful. With this increase, peregrines have moved into urban Lima city. In the first week of June 2001, Beingolea was told about a pair of hawks that were excreting on the walls of the skyroom of the abandoned 20-floor, former Hotel Crillon in downtown Lima. Beingolea visited the building and confirmed the presence of a female peregrine beside a cavity, caused by the removal of a large cement block on the upperside of a window’s roof, just below the hotel’s 20th floor skyroom. On 2 July 2001, Beingolea visited the building again and found a male peregrine incubating a single egg; the female was nearby eating a Rock Dove ( Columba livia) . On 8 August, he visited the building and found one eggshell but could not assess the number of eggs or nestlings, but on 14 August there were three nestlings about 5-7 d old On 14 September one young was found dead, possibly due to a Trichomonas infection; the other two seemed healthy Three more pairs were regularly sighted in Lima along with other territorial single individuals during the Austral winter 2001. The sighting of an immature inside Lima during late August 2000 (J. Otero pers. comm.) suggested that in fact they were already breeding inside urban Lima city before our observations. Finally, Beingolea found fledglings between 18 July and late September indicating about a 10 wk span of egg laying for pairs nesting at 12°S. Calculating that fledgling occurs between 11 and 12 wk after onset of eggs (Cade 1988), the earliest laying for the pair having fledglings in 18 July should have taken place during the last week of April (1993) and the latest lying during the first week of July (2001). In 2001, there were a total of six resident pairs within Lima and her outskirts. It is generally thought that subtropical raptors regularly have smaller clutches than populations elsewhere (Newton 1979, Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A.), but the Peruvian nests checked averaged 3.71 young/pair, at the high end of peregrine fledging numbers. A failed attempt, due to predation, also had four eggs, further suggesting that large clutch sizes are common. The breeding season for Peruvian peregrines differs from those in central to southern Chile and Argentina; the latter breed in the Austral spring, but central Peruvian peregrines nest during the Austral fall and winter. Distance and different breeding seasons might restrict gene flow between these populations. Lack of gene flow and different climatic and environmental selection pressures probably explains the morphological differences. Northern birds are paler, less heavily marked on the breasts and smaller, about 800 g for northern and 950 g for southern females and about 550 g for northern males (with one at 480 g) and 650 g for southern males (O. Beingolea unpubl. data, see White 1989). Further studies on the geographic differences within South American populations are needed. D.H. Ellis, W.G. Mattox, and T.L. Maechtle provided useful comments on the manuscript. We thank them. — Oscar Beingolea, Calle La Venturosa 114, Urb. Los Rosales. Lima 33, Peru, and Clayton M. White (corresponding author), Department of Integrative Biology, Brigham Young University, Provo, UT 84602 U.S.A.; e-mail address: clayton_white @byu. edu Received 1 April 2002; accepted 19 October 2002. Associate Editor: Marco Restani J. Raptor Res. 37(l):85-86 © 2003 The Raptor Research Foundation, Inc. Lone Harris’s Hawk Kills Great Blue Heron The Harris’s Hawk {Parabuteo unicinctus) is a neotropical species whose range extends into the southwest U.S., resident from southeast California (Colorado River area; irregularly), south and central Arizona, to southwest and south Texas (Bednarz 1995, In A. Poole and F. Gill [Eds.], The birds of North America, No. 146. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC, U.S.A.). In New Mex- 86 Letters VoL. 37, No. 1 ICO, the species nests across the southern tier of counties and north in the Rio Grande Valley to Sierra County (S. Williams pers. comm.). Published accounts of Harris’s Hawk food habits report the majority of its prey as (in order of dominance) medium- sized to relatively large mammals (particularly rabbits and hares [Leporidae] ) , birds, and reptiles (Mader 1975, Living Bird 14:59-85; Whaley 1986, Raptor Res. 20:1-15; Bednarz 1988a, Condor 90:311-323; Bednarz 1995). Bent (1937, Bull. U S. Natl. Mus. 167:142-147) listed relatively large avian prey items as Common Moorhen {Gallinula chloropus), night- herons {Nycticorax nycticorax, Nyctanassa violacea), Snowy Egret {Egretta thula), and Green-winged Teal {Anas crecca); Whaley (1986) observed remains of a Cooper’s Hawk {Accipiter cooperii) in one nest in Arizona. Here, we report our observations of an attack on a Great Blue Heron {Ardea herodias) by a Harris’s Hawk. At ca. 1630 H on 21 July 2000, a Juvenile (less than 12 mo) Great Blue Heron waded in a puddle near our camp in the Gila National Forest Bird Habitat Area, Grant County, New Mexico. The bird did not fly as we approached to within 5 m, but slowly walked off into the young mesquite {Prosopis glandulosa) growth. We assumed it was in the area to feed on the abundant grasshoppers (Orthoptera) . On at least two prior occasions, we had seen a Great Blue Heron roost in nearby trees. About 2 hr later, we heard a prolonged croaking cry and investigated. An immature Harris’s Hawk flew off the now prostrate heron and perched ca. 50 m away. Its plumage (white in the wings and underparts) indicated that the hawk was an immature bird (Clark and Wheeler 1987, A field guide to hawks of North America, Houghton Mifflin Co., Boston, MA U.S.A.). The heron was still alive, but it died within a minute. We detected no obvious external injuries when we examined it, but no internal exam or necropsy was performed. We are uncertain why it did not fly or defend itself from apparent attack. The hawk remained perched in the tree for about 10 min before flying away. We saw the hawk later that day and on subsequent days, but it never returned to the heron carcass. The age of the bird supports prior observations on hawk hunting behavior. Generally, immature solo hawks (par- ticularly in falconry) exhibit more reckless and daring behavior than birds in adult plumage, often attacking inap- propriately large prey (J. Bednarz and J. Coulson pers. comm.). Perhaps the Harris’s Hawk’s age and inexperience led it to attack the Great Blue Heron. We believe this account to be the first recorded for Harris’s Hawk predation upon a bird as large as the Great Blue Heron. The mean mass of this hawk’s largest known avian prey (Common Moorhen, 334 g; Green-winged Teal, adult males, 364 g; Snowy Egret, 371 g; Cooper’s Hawk, adult females, 529 g; and Black-crowned Night-heron, 883 g) is much smaller than that of the Great Blue Heron (2204-2576 g [Dunning 1993, Body weights of 686 species of North American birds. International Wildlife Rehabilitation Council, Suisun, CA U.S.A.]). A single Harris’s Hawk is thus capable of attacking and killing much larger birds than previously reported. Although the Great Blue Heron is the largest documented avian prey taken by a wild Harris’s Hawk (captive Harris’s Hawks flown in falconry have incidentally captured healthy Great Blue Herons [T. and J. Coulson pers. comm.]), its weight is less than that of adult female black-tailed jackrabbits {Lepus californicus, >3000 g) commonly taken by hunting groups (Bednarz 1988b, Science 239:1525-1527), and occasionally by solitary hunting Harris’s Hawks (Brannon 1980, The reproductive ecology of a Texas Harris’s Hawk [Parabuteo unicinctus harrisi] population. M.S. thesis, University of Texas, Austin, TX U.S.A.) . We gratefully thank J.C. Bednarz for initial information on Harris’s Hawk-heron interactions. The comments of S.H. Stoleson and reviewers J.C. Bednarz, P.H. Bloom, J.O. Coulson, J.D. Ligon, M.A. Patten, and D.A. Zimmerman greatly improved this manuscript. We extend special appreciation to P. Boucher and R. Pope, Gila National Forest, Silver City District, for authorization to conduct research in the Gila National Forest Bird Habitat Area. — Hope D. Woodward, U.S.D.A. Forest Service, Rocky Mountain Research Station, 333 Broadway SE, Suite 115, Abuqueruque, NM 87102-3497 U.S.A., and R. WiUiam TrusseU, P.O. Box 856, Moyie Springs, ID 83845 U.S.A.; e-mail address: woodwardh@yahoo.com Received 13 November 2001; accepted 28 October 2002 BOOK REVIEWS /. Raptor R£s. 37(1) :87 © 2003 The Raptor Research Foundation, Inc. Raptors in the New Millennium. Edited by Reu- ven Yosef, Michael L. Miller, and David Pepler. 2002. International Birding 8c Research Center, Ei- lat, Israel. 276 pp., numerous tables and figures. Softcover, $20.00. — This volume constitutes the proceedings of the joint meeting of the Raptor Re- search Eoundation and the World Working Group on Birds of Prey and Owls that was held in Eilat, Israel, from 2-8 April 2000. Close to 200 biologists from 30 countries attended the meeting, which consisted of 84 oral presentations and 17 posters. In total, the book contains 29 papers (ranging in length from 2 to 22 pages) and 80 abstracts pub- lished in English under eight subject headings; General 8c Techniques (2 papers, 8 abstracts), Diet and Eoraging (1, 3), Reproductive Ecology (4, 12), Migration and Wintering Ecology (10, 8), Popula- tion Status and Ecology (6, 24), Genetics and Tax- onomy (2, 7), Ecotoxicology and Diseases (2, 11), and Conflicts and Solutions (2, 7). As expected, the range of species studied and topics covered is quite large, although taxonomically the papers are dominated by diurnal raptors. Indeed, only four of the 29 papers are devoted to owls (two on Barn Owls [Tyto alba], and one each on the Little Owl [Athene noctua] and the Tawny Owl [Strix aluco]). Among the highlights in the collection is a paper on identifying “real threats” to raptor populations (Kenward) , one on the breeding biology of Great- er Spotted Eagles (Aquila clanga) in Poland (Grasz- ynski et ak), one on modeling habitat suitability for Little Owls in Belgium (Van Nieuwenhuyse and Be- kaert) , one on migration routes and habitat selec- tion by nonbreeding Lesser Kestrels {Falco nauman- ni) in Africa (Pepler), an update on the status of Great Philippine Eagles (Pithecophaga jefferyi) on Mindanao (Miranda et ak), and a global review of lead poisoning in falconiforms (Miller et ak). The proceedings provide a snapshot of raptor research currently underway across a vast portion of the globe, albeit with a bias toward studies of Palearctic falconiforms. The editing is very good considering that English must not have been the first language for many of the authors. Given the range of topics and species covered, and the at- tractive price (which includes shipping) , Raptors in the New Millennium will be a welcome addition to libraries — both personal and public — around the world. To obtain a copy, write the International Birding 8c Research Center, P.O. Box 774, Eilat 88000, Israel (http://www.arava.org/birds-eilat/ index.html). — Jeff Marks, Montana Cooperative Wildlife Research Unit, University of Montana, Missoula, MT 59812 U.S.A. J. Raptor Res. 37(1) :87-88 © 2003 The Raptor Research Foundation, Inc. Birds of the Saskatoon Area. Edited by A.L. Leighton, J. Hay, C.S. Houston, J.F. Roy, and S. Shadick. 2002. Saskatoon Natural History Society Spec. Pubk No. 23. 345 pp., 31 color photographs, 40 line drawings, 14 maps, 8 appendices. ISBN 0- 921104-19-7. Softcover, $39.00 Canadian. — This book presents a compilation of bird records for the Saskatoon Bird Area (SBA) , which encompasses an 11012-km^ block centered near Moon Lake just south-southwest of Saskatoon, the largest city in Saskatchewan. The roughly 60 000 observation cards of birds submitted by hundreds of members of the Saskatoon Natural History Society between 1965 and 1988 form the basis of the book, the pro- ject being a massive undertaking by a dedicated group of individuals committed to understanding the birds of their local area. The result is a detailed picture of the timing of occurrence and status for the nearly 350 species of birds that were recorded. Introductory sections contain information on the history of the project, the physical and biological properties of the region (including some great col- or photos of habitat), a glossary of terms, and a brief guide to interpreting the species accounts. The eight appendices contain, among other 87 88 Book Reviews VoL. 37, No. 1 things, summaries of Breeding Bird Survey data for four time periods (up to 2000), summaries of banding efforts and recoveries/recaptures for the area, and a list of theses from the University of Saskatchewan that involved bird research in the SBA. The species accounts range in length from about one-third of a page to just over two pages and focus on arrival and departure dates, status, and habitat affinities. Twenty-nine species of raptors are treat- ed, 17 falconiforms and 12 owls. Perhaps not sur- prisingly, given that Stu Houston is one of the ed- itors, the raptor accounts are among the most detailed and interesting in the book. For example, it turns out that Saskatoon is one of the best spots on the continent to observe nesting Merlins {Falco columbarius) . In excess of 30 pairs breed within the city limits in some years, and more than 2000 in- dividuals, mostly nestlings, have been banded there. At the nocturnal end of the spectrum. Great Horned Owls {Bubo virginianus) have also received a lot of attention. The species account reports that 1208 nestlings from 401 nests were banded in the SBA between 1961 and 1998. To date, 130 have been recovered or recaptured, 118 in the SBA and the remainder in Alberta, Manitoba, Iowa, Min- nesota, Montana, North Dakota, and South Dako- ta. On balance, each account contains an interest- ing tidbit or two, including some valuable natural history information, on the bird’s presence in the SBA. I must admit that when Birds of the Saskatoon Area arrived, I hesitated to review it for JRR because rap- tors were not its focus. However, as I flipped through the pages, and especially after I read the introductory portions, I realized the importance of drawing attention to a work of this sort. The pub- lication represents the collective efforts of a group of people who are truly engaged in the natural his- tory that surrounds them. So engaged, in fact, that hundreds of them cooperated to gather the infor- mation and make it available to the entire com- munity. Imagine how wonderful it would be if all other cities the size of Saskatoon or larger had among their citizenry a group of people similarly engaged in nature and suitably motivated to com- pile bird records and produce so useful a publi- cation. — Jeff Marks, Montana Cooperative Wildlife Research Unit, University of Montana, Missoula, MT 59812 U.S.A. J. Raptor Res. 37(1) :89 © 2003 The Raptor Research Foundation, Inc. Manuscript Referees The following people reviewed manuscripts for the Journal of Raptor Research in 2002. Peer review plays a vital role in the publishing process and in improving the quality of the Journal. The editorial staff would like to thank the following for reviewing manuscripts this past year. The names of those who reviewed two or more manuscripts are indicated with an asterisk. L. Alterman, D. Andersen, R. Antor, L. Arent, D. Arsenault, J. Balbontin-Arenas, T. Balgooyen, W. Baltosser, M. Bechard, J. Bednarz*, I. Bellocq*, J. Berkelman, K. Bildstein, R. Bierregaard, D. Bird, P. Bloom*, T. Bloxton, G, Boano, P. Bohall Wood, T. Bosakowski, R. Bowman, P Boxall, R. Bradley, U. Brendel, T. Brush, J. Buchanan, D Buehler, T. Cade*, J. Cartron, J. Cerda, J. Chantler, S. Chaplin, H. Chen, A. Clarke, N. Clum, M. Collopy, C. Conway, S. Debus, M. Desmond*, S. DeStefano, S. Deem, D. Dekker*, M. de Lucas, L. Dickson, J. Donazar*, C. Dove, F. Doyle, J. Duncan*, C. Dykstra, J. Elliott, D. Ellis, J. Enderson*, PL. Enriquez, D. Evans, S. Eattorini, M. Eerrer*, L. Eorbes, M. Eorero, E. Forsman, G. Foster*, A. Fowler, J. Fraser*, T. French, R. Frumkin, M. Fuller, V. Galushin, L. Gangoso, H. Garner*, R. Garrott, F. Gehlbach, J. Gehring*, S. Gillihan, R. Glinski, M. Goldstein, T. Grubb*, R. Gutierrez, H. Hakkarainen, P. Hall, A. Harmata*, R. Harper, G. Hayward, C. Henny*, L. Hillstrom, S. Holman, G. Holroyd*, D Holt*, T. Huels, G. Hunt*, E. Inigo-Elias, J. Jackson, F. Jaksic*, S. Janes, A. Jenkins, J. Jimenez, R. Jovani, T. Katzner, R. Eiavanaugh, P. Kerlinger, T. Kimmel, M. Kochert*, E. Korpimaki, P. Koskimies, W. LaHaye, R. Lehman, G. Leonardi, K. Levenstein, J. Lish, M. Louette, A. Lueders, T. Maechtle, L. Marchesi, B. Marcot, S. Markman, J. Marks*, M. Martell*, C. Marti*, J. Marzluff*, W. Mattox, M. McGrady, D. McLeod*, M. McMillian*, E. McNabb, S. Melvin, B. Millsap*, J. Mitani, C. Moorman, M. Morrison*, P. Mundy, R. Murphy, W. Nelson*, R. Nero*, P. Nye, C. Olson, J Orr, J. Pagel, M. Patten, E. Pavez*, V. Penteriani*, M. Perry, J. Plissner, A. Poole, C, Preston*, G. Proudfoot*, J. Rau, P Redig, E. Revilla, R, Reynolds, S. Rinkevich, D, Ripper*, G. Ritchison*, A. Rodriguez, R. Rodriguez-Estrella, R. Rosenheld, L. Salvati*, J. Sanchez-Zapata*, B. Sandercock, J. Schmutz, L. Schueck, M. Seamans, J. Seoane, F. Sergio*, J. Shelnutt, S. Shiraki, J. Sikarskie*, J. Smallwood, R. Smith*, R. Spaar*, B. Spears, J. Squires*, D. Stahlecker*, M. Stalmaster, T. Swem*, S. Talbot, J.M. Thiollay*, P. Thomas, W. Thompson*, D. Tinkler, K. Titus, F. Tseng*, J. Valkama, S. Vanderwall, J. Vargas, D. Varland*, A. Village*, J. Ward, I. Warkentin*, P Weatherhead, D. Whitacre*, C. Williams, W. Whaley, C. White*, K. Wiebe, S. Wiemeyer, C. Wightman, J. Withey, B. Woodbridge, S. Xirouchakis, R. Yates, S Zack, Z. Zhenwang. A Telemetry Receiver Designed with The Researcher in Mind What you've been waiting for! finally, a highly sei.-itivc * chafu.i.; synthi'M/ed telemetry leceiver that weighs less than 13 ounces. Is completply user programmable and offers variable scan rates over all fteguencies. For each animal being tracked, the large ICD display provides not only the frequency (to lOOHz) and channel number, but also a / character alphanumeric comment field and a digital signal strength meter. Stop carrying receivers that are the si/e of a lunch bo* or cost over S1500. The features and performance of the new R-IOOO pocket sired Iclemrtry receiver will impress you. and the price will convince you. Other Features include: • F.ictnry tuned to any 4MH; Aide segment in the K8174MHr Band • Very high sensitivity of KSdBm to ISOdBrn • lUummated display and keypad for use in low light or darkness • User selectable scan rates from MO seconds m 1 u'cond steps • Rechargeable batteries operate the receiver for 12 hours and tan be replaced with standard AA Alkaline batteries in Ifn* field. Both 12vdc and llOvac chargers are included. • 6.1" (16.6cm) high, 2.6" (6.6cm) wide, 1.5" (3.8cm) deep. • 3 year warranty • 1 day delivery $ 695.00 Please specify desired 4MHr wide segment in the 148 174MH/ band Visit out website foi complete specifications, operating manual and information on the R-1000 or oui toll free rnimber to i.ik!*''! ynu' tereivei no'.s. Try the New R-JOOO and YouV Bt Impreised! COMMUNICATIONS SPECIALISTS. INC. ; .t- II S A • 1 . 1 • 1 . ..01 I (V Buteo 3130 Laurel Road Shipman, VA 22971 Books 800-722-2460 allen@buteobooks.com 434-263-8671 www.buteobooks.com BUTEO BOOKS is the largest retailer of Ornithology titles in North America, with over 2000 in-print titles in stock. In-Print Falconry and Birds of Prey Handbook of the Birds of the World. The first seven volumes of this projected sixteen- volume work have been published, including Volume 2, covering the diurnal raptors, and Volume 3, covering owls. They are priced at $185 each. The Raptor Almanac: A Comprehensive Guide to Eagles, Hawks, Falcons, and Vultures. Scott Weidensaul. Lyons Press, 2000. 382 pp. Color photos. Cloth. $40.00 The Harris Hawk: Management, Training and Hunting. Lee W. Harris. Swan Hill Press, 2001. 144 pp. Color and b&w photos and illustrations. Cloth. $32.95 Rare and Out-of-print Falconry Usually available from Buteo Books, the classic reference on diurnal raptors: Brown & Amadon: Eagles, Hawks and Falcons of the World. First English edition, 1968. $300. Birds of North America series individual species accounts for over 500 species, including: Swallow-tailed Kite White-tailed Kite Snail Kite Mississippi Kite Bald Eagle Northern Harrier Sharp-shinned Hawk Cooper’s Hawk Northern Goshawk Common Black-Hawk Harris’ Hawk Red-shouldered Hawk Broad- winged Hawk Swainson’s Hawk White-tailed Hawk Zone-tailed Hawk Hawaiian Hawk Red-tailed Hawk Ferruginous Hawk Crested Caracara American Kestrel Merlin Gyrfalcon Prairie Falcon Bam Owl Flammulated Owl Eastern Screech-Owl Whiskered Screech-Owl Great Homed Owl Snowy Owl Northern Hawk Owl Northern Pygmy-Owl. Fermginous Pygmy-Owl Elf Owl Burrowing Owl Spotted Owl Barred Owl Great Gray Owl Long-eared Owl Short-eared Owl (R) Boreal Owl Northern Saw- whet Owl Available soon: American Kestrel California Condor 2003 ANNUAL MEETING The Raptor Research Eoundation, Inc. 2003 annual meeting will be held on 3-7 September 2003 in Anchorage, Alaska. For information about the meeting see the following website; http://www. alaskabird.org or contact Alaska Bird Observatory (birds@alaskabird.org). Persons interested in predatory birds are invited to join The Raptor Research Foundation, Inc. Send requests for information concerning membership, subscriptions, special publications, or change of address to OSNA, P.O. Box 1897, Lawrence, KS 66044-8897, U.S.A. The Journal of Raptor Research (ISSN 0892-1016) is published quarterly and available to individuals for !$33.00 per year and to libraries and institutions for $50.00 per year from The Raptor Research Foundation, Inc., 14377 117th Street South, Hastings, Minnesota 55033, U.S.A. (Add $3 for destinations outside of the continental United States.) Periodicals postage paid at Hastings, Minnesota, and additional mailing offices. POSTMASTER: Send address changes to The Journal of Raptor Research, OSNA, P.O. Box 1897, Lawrence, KS 66044-8897, U.S.A. Printed by Allen Press, Inc., Lawrence, Kansas, U.S.A. Copyright 2002 by The Raptor Research Foundation, Inc. Printed in U.S.A. 0 This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Raptor Research Foundation, Inc., Awards Lifetime Achievement Awards The Tom Cade Award recognizes an individual who has made significant advances in the area of captive prop- agation and reintroduction of raptors. Nomination packets can be submitted at any time. Contact: Brian Walton, Predatory Bird Research Group, Long Marine Laboratory, University of California, Santa Cruz, CA 95064 U.S.A.; tel. 408-459-2466; e-mail: walton@cats.ucsc.edu. The Fran and Frederick Hamerstrom Award recognizes an individual who has contributed significantly to the understanding of raptor ecology and natural history. Nomination packets can he submitted at any time. Con- tact; Dr. Clint Boal, Texas Cooperative Fish and Wildlife Research Unit, BRD/USGS, Texas Tech University, 15th Street & Boston, Ag Science Bldg., Room 218, Lubbock TX 79409-2120 U.S.A.; tel. (806) 742-2851; e-mail: cboal@ttacs.ttu.edu. Student Recognition and Travel Assistance Awards The James R. Koplin Travel Award is given to a student who is the senior author and presenter of a paper or poster to be presented at the RRF annual meeting for which travel funds are requested. Contact: Dr. Patricia A. Hall, 5937 E. Abbey Rd. Flagstaff, AZ 86004 U.S.A.; tel. 520-526-6222; e-mail: pah@spruce.for.nau.edu. Application Deadline: due date for meeting abstract. The William C. Andersen Memorial Award is given to the students who are senior authors and presenters of the best student oral and poster presentation at the annual RRF meeting. Contact; Laurie Goodrich, Hawk Mountain Sanctuary, 1700 Hawk Mountain Road, Kempton, PA 19529 U.S.A.; tel. 610-756-6961; email: goodrich@hawkmountain.org. Application Deadline: due date for meeting abstract; no special application is needed. Grants For each of the following grants, complete applications must be submitted to the contact person indicated by 15 February. Recipients will be notified by 15 April. The Dean Amadon Grant for $200-400 is designed to assist persons working in the area of distribution and sys- tematics (taxonomy) of raptors. Contact: Dr. Carole Griffiths, 251 Martling Ave., Tarrytown, NY 10591 U.S.A.; tel. 914-631-2911; e-mail: cgriff@liu.edu. The Stephen R. Tully Memorial Grant for $500 is given to support research, management, and conservation of raptors, especially to students and amateurs with limited access to alternative funding. Contact: Dr. Rim Titus, Alaska Department of Fish and Game, Division of Wildlife Conservation, P.O. Box 240020, Douglas, AK 99824 U.S.A.; e-mail: kimt@fishgame. state. ak.us. The Leslie Brown Memorial Grant for up to $1,000 to support research and/or dissemination of information on birds of prey, especially to proposals concerning African raptors. Contact: Dr. Jeffrey L. Lincer, 9251 Golondrina Dr., La Mesa, CA 91941 U.S.A.; e-mail: jefflincer@tns.net.