The Journal of Raptor Research Volume 36 Number I March 2002 Published by The Raptor Research Foundation, Inc, THE RAPTOR RESEARCH EOUNDATION, INC. (Founded 1966 ) OFFICERS PRESIDENT; Brian A. Millsap SECRETARY; Judith Henckel VICE-PRESIDENT; Keith L. Bildstein TREASURER; Jim Fitzpatrick BOARD NORTH AMERICAN DIRECTOR #1: Philip Detrich NORTH AMERICAN DIRECTOR #2; Laurie J. Goodrich NORTH AMERICAN DIRECTOR #3: Jeff R Smith INTERNATIONA!. DIRECTOR#!: Eduardo Inigo-Elias INTERNAT!ONAL DIRECTOR #2: Ricardo Rodriquez-Estrelia OF DIRECTORS INTERNATIONAL DIRECTOR #3: Beatriz Arroyo DIRECTOR AT LARGE #1: Jemima PajrryJones DIRECTOR AT LARGE #2: Petra Bohall Wood DIRECTOR AT LARGE #3: Michael W. Coi.lopy DIRECTOR AT LARGE #4: Carol McIntyre DIRECTOR AT LARGE #5; Robert N. Rosenfield DIRECTOR AT LARGE #6: Ed Henckel 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. Belthoff Juan Jose Negro Clint W. Boat Marco Restani Cole Crocker-Bedford Ian G. Warkentin Joan L. Morrison 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, Aitison Fowler, 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. 35(4), and are available from the editor. Submit manuscripts to J. Bednarz at the address listed above. COVER: Eurasian Eagle-Owl {Bubo bubo). Painting by Josef Niederlechner, Anton-Mangold-Weg 8, 82362 Weilheim, Germany. Contents Editor’s Page; Working Toward Excellence. James c. Bednarz 1 Survival of Florida Burrowing Owls Along an Urban-Development Gradient. Brian A. Millsap 3 Biases Associated with Diet Study Methods in the Eurasian Eagle-Owl. Luigi Marchesi, Paolo Pedrini, and Fabrizio Sergio 11 Description of a New Subspecies of the Egyptian Vulture (Accipitridae: Neo- phron percnopterus) from the Canary Islands. Jose Antonio Donazarjuanjose Negro, Cesar Javier Palacios, Laura Gangoso, Jose Antonio Godoy, Olga Ceballos, Fernando Hiraldo, and Nieves Capote 17 Density, Nest Sites, Diet, and PRODUCTivnY of Common Buzzards (Buteo buteo) IN THE Italian Pre-Alps. Fabrizio Sergio, Alberto Boto, Chiara Scandolara, and Giuseppe Bogliani 24 Vitamins E and A, Carotenoids, and Fatty Acids of the Raptor Egg Yolk. Nigel W.H. Barton, Nicholas C. Fox, Peter F. Surai, and Brian K. Speake 33 Preliminary Ground and Aerial Surveys for Orange-breasted Falcons in Central America. Russell Thorstrom, Richard Watson, Aaron Baker, Serena Ayers, and David L. Anderson 39 Dispersing Vulture Roosts on Communication Towers. Michael l. Avery, John s. Humphrey, Eric A. Tillman, Kimberly O. Phares, and Jane E. Hatcher 45 Raptor Abundance and Habitat Use in a Highly-Disturbed-Forest Landscape in Western Uganda. Nathaniel E. Seavy and Christine K. Apodaca 51 Trophic Niche of North American Great Horned Owls. Lee A. Cromrich, Denver W. Holt, and Shawne M. Leasure 58 Short Communications Social Organization of a Trio Of Bearded Vultures (Gypaetus barsatus): Sexual aisid Parental Roles. Joan Bertran and Antoni Margalida 66 Genetic Evidence of Alloparental Care of a Female Lesser Kestrel in an At .te n Nest. Pedro J. Cordero, Jose M. Aparicio, and David T. Parkin 70 Nesting of Long-eared Owls Along the Lower Big Lost River, Idaho; A Comparison of 1975-76 and 1996-97. Natalie A Fahler and Lester D. Flake 73 Behavioral and Physical Development of a Nestling Crested Eagle (Morphnus guianensis) . David F. Whitacre, Juventino Lopez Avila, and Gregorio Lopez Avila 77 Spring Weather and Breeding Success of the Eurasian Kestrel (Falco tinnunculus) in Urban Rome, Italy. LucaSalvati 81 Fatal Caryosk>ra Infection in a Free-living Juvenile Eurasian Kestrel {Falco tinnunculus) .OMycy Krone 84 Book Review. Edited by Jeffrey S. Marks 87 Manuscript Referees 89 The Raptor Research Foundation, Inc. gratefully acknowledges funds and logistical support provided by Arkansas State Univeixity to assist in the publication of the journal. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. VoL. 36 March 2002 No. 1 J Raptor Res. 36(l);l-2 © 2002 The Raptor Research Foundation, Inc. EDITOR’S PAGE WORKING TOWARD EXCELLENCE James C. Bednarz, Editor Perhaps you noticed that the Journal of Raptor Research has gone through a minor adjustment dur- ing the last year. Editor Marc Bechard, after a half decade of dedicated efforts to improve the Journal, stepped down and passed the awesome responsi- bility of managing our flagship research publica- tion to me and a small, part-time editorial staff at Arkansas State University. During Marc’s tenure as Editor, the Journal of Raptor Research (JRR) took on a new look and prospered. JRR’s impact factor, a measure of the scientific “impact” of peer-re- viewed periodicals based on the number of cita- tions compiled by the Institute of Scientific Infor- mation in Journal Citation Reports, has grown exponentially over the last few years. Although the value of “impact factors” in ranking the quality of scientific journals is debatable, it represents one measure of the exposure that articles published in a particular journal will receive and their potential influence on science. Currently, JRR?, impact fac- tor is on par with several respected international journals in the disciplines of ornithology (e.g., Wil- son Bulletin) and organismal biology (e.g., Wildlife Society Bulletin) . As editor, I pledge to you that the current editorial staff will continue to strive toward publishing the best journal of raptor science in the world. As I anticipated, the transition of editorial re- sponsibilities from Marc’s shop to my lab set us back a little bit. Our new Editorial Assistant, Re- becca Maul, and I received the official authoriza- tion from the Board of the Raptor Research Foun- dation, Inc. to begin processing manuscripts at the end of February 2001. Although we got right to work, we were faced with an instant backlog of manuscripts. In addition, Rebecca and I had lots to learn about the ins and outs of publishing a scientific journal. Thankfully, we received much help during the transition from Marc Bechard, Joan Clark (the long-time proofreader of JRR), the folks at Allen Press, and the Department of Biolog- ical Sciences at Arkansas State University. We great- ly appreciated this assistance and support during the transition process, that enabled us to produce our first issue (Volume 35, Number 3) only slightly behind schedule. The second issue of JRR, edited and produced by the new editorial office (Volume 35, Number 4), showcased the proceedings of the Second In- ternational Burrowing Owl Symposium. It present- ed a huge challenge for our meager resources. Even though we received outstanding cooperation from Special Editors of this issue, Troy Wellicome and Geoff Holroyd, and the many authors who contributed to this effort, this unique volume rep- resents the largest single issue ever of the JRR. Throughout the production of these proceedings, all contributors, cooperators, and the entire edi- torial team endeavored to produce the best schol- arly product possible. All contributions went through thorough reviews and multiple edits, and several submitted manuscripts that were not deemed suitable for publication in JRR were re- jected to maintain a high standard of peer review. If you have not looked at this special issue of JRR, I encourage you treat yourself and review some ex- cellent science on Burrowing Owls {Athene cunicu- laria) . As new editor-in-chief, I am currently very satis- fied with the quality and format of our journal and have elected to maintain most of the specifications put in place by my recent successors. However, we will continue to tweak things in an effort to im- prove the presentation of raptor science and better 1 2 Bednarz VoL. 36, No. 1 accommodate both authors and readers. For ex- ample, we now publish e-mail addresses of the cor- responding author for each article published. This contemporary contact information will, hopefully, further encourage constructive interactions be- tween scientists and conservationists interested in raptor biology. We also have instituted an optional electronic submission procedure (/. Raptor Res. 35: 409) that should speed up the processing of man- uscripts. Our current priority objective is to accel- erate the turn-around time from manuscript re- ceipt to publication. Let me emphasize, however, our current turn-around time in publishing papers is faster than most other ornithological and organ- ismal biology journals: less than 8 months to a final decision on acceptability and most manuscripts are in print within 14 months after the date of sub- mission. In 2002, we intend to reduce these turn- around intervals as much as feasible. You may also see other minor changes in future issues and we welcome suggestions and comments from both au- thors and readers concerning further improve- ments to the JRR (jrr@astate.edu). As I have discovered over the past 10 months, publishing a quarterly, peer-reviewed journal in raptor biology is a monumental undertaking. I want to express special thanks for the contribu- tions of our current staff of Associate Editors: Jim Belthoff, Clint Boal, Cole Crocker-Bedford, Joan Morrison, Juan Jose Negro, Marco Restani, Ian Warkentin, and Troy Wellicome. Our Book Review Editor, Jeff Marks, and Spanish Editor, Cesar Mar- quez Reyes, continue to do outstanding work. We currently have two part-time Editorial Assistants, Rebecca Maul and Allie Fowler, who accomplish the tough, day-to-day legwork that gets the JRR out on time. Finally, I want to express my appreciation to all the referees, whose thoughtful evaluations make the quality of the Journal. Publication of the JRR is a cooperative effort involving the entire membership and raptor scientific community, who provide the original research and peer review that fuels the forward progress of science. When we re- quest that you contribute to the quality of the JRR by refereeing a manuscript, I strongly encourage you to agree and to take part in this vital process. I look forward to working with all of you in the near future! J. Raptor Res. 36(1): 3— 10 © 2002 The Raptor Research Foundation, Inc. SURVIVAL OF FLORIDA BURROWING OWLS ALONG AN URBAN-DEVELOPMENT GRADIENT Brian A. Millsap^ Bureau of Wildlife Diversity Conservation, Florida Fish and Wildlife Conservation Commission, 620 S. Meridian St., Tallahassee FL 32399-1600 U.S.A. Abstract. — I estimated survival rates of a Florida Burrowing Owl {Athene cunicularia floridana) popula- tion on a 35.9-km^ study area in Lee County, Florida, 1987-91 to determine if there was a relationship between annual survival and development density. The study area spanned a residential development density gradient ranging from <2% to >74% of lots with houses. Survival rates were estimated from a sample of 119 banded adult male, 152 adult female, and 310 juvenile Burrowing Owls using Cormack- Jolly-Seber capture-recapture models and Fisher’s maximum likelihood method of parameter estimation. Survival rates were estimated separately for sections of the study area where <20% of 0.2-ha lots were developed (i.e., homes built), 20-39% were developed, 40-60% were developed, and >60% were de- veloped. The most parsimonious models indicated that adult male survival was 81% (SE = 0.04) where <20% or >60% of lots were developed, and 62% (SE = 0.05) elsewhere. Adult female survival was 52% (SE = 0.06) where 40-60% of lots were developed and 69% (SE = 0.04) elsewhere. Juvenile survival was 18% (SE = 0.08) where <20% of lots were developed, 43% (SE = 0.08) where 20—39% of lots were developed, 28% (SE = 0.04) where 40-60% of lots were developed, and 11% (SE = 0.03) where >60% of lots were developed. 1 speculate that the inverse relationship between adult and juvenile survival across development zones reflected the greater recruitment opportunities that existed for juveniles in development zones where adult mortality was high. Patterns of adult mortality more closely tracked trends in the rate of home construction than the density of homes, suggesting factors associated with home building and the attendant landscape changes might result in high Burrowing Owl mortality. Key Words: Burrowing Owl; demography; development; Florida; mark-recapture; Athene cunicularia flori- dana; survival; urban wildlife management. Supervivencia de Athene cunicularia floridana en un gradiente de desarrolo urbano Resumen. — Estime las tazas de supervivencia de una poblacion de buhos cavadores de la florida {Athene cunicularia floridana) en un area de estudio de 35.9-km^ en el condado Lee, Florida, 1987-91 para determinar si habia una relacion entre la supervivencia anual y la densidad del desarrollo. El area de estudio comprende un gradiente de densidad de desarrollo residencial que va desde <2% a >74% de lotes con casas. Las tasas de supervivencia fueron estimadas partir de una muestra de 119 machos adultos anillados. 152 hembras adultas, y 310 buhos cavadores juveniles usando los modelos de captura recap- tura de Cormack-Jolly-Seber y el metodo de maxima probabilidad de la estimacion de parametros de Eisher. Las tasas de supervivencia fueron estimadas separadamente por secciones del area de estudio que estaban bajo desarrollo asi: <20% de los lotes de 0.2 ha (V. Gr. construcciones familiares), 20- 39%, 40—60%. Y >60 % en desarrollo. La mayoria de modelos de parsimonia indican que la supervi- vencia de los machos adultos fue 81% (SE = 0.04) en donde <20% o >60% de los lotes estaban desarrollados, y 62% (SE = 0.05) en cualquiera de los otros. La supervivencia de los juveniles fue 18% (SE = 0.08) en donde <20% de los lotes estaban desarrollados, 43% (SE = 0.08) donde 20-39% de los lotes se habian desarrollado, 28% (SE = 0.04) donde 40-60% de los lotes estaban desarrollados, y 11% (SE = 0.03) donde >60% de los lotes se habian desarrollado. Especulo que la relacion inversa entre la supervivencia de adultos y juveniles a lo largo de las zonas desarrolladas reflejan las mayores oportunidades que tienen los juveniles para restablecerse en zonas desarrolladas en donde lamortalidad de los adultos fue elevada. Los patrones de mortalidad de adultos mostraron tendencias mas cercana- mente asociadas a la tasa de construccion de hogares que a la densidad de los mismos, sugiriendo que los factores asociados a la construccion y a los consecuentes cambios del paisaje podrian dar como resultado una alta mortalidad de buhos cavadores. [Traduccion de Cesar Marquez] 1 E-mail address: millsab@fwc. state. fl.us 3 4 Miixsap VoL. 36, No. 1 The disjunct population of Burrowing Owls {Athene cunicularia floridana) in Florida was histor- ically closely associated with native prairies in the central peninsula (Rhodes 1892, Nicholson 1954). The species began a rapid range expansion in the state in the 1950s as human settlement converted vast areas of former woodland to pasture and home development (Ligon 1963, Courser 1979, Millsap 1996). Today, many of Florida’s Burrowing Owl populations occur in suburban neighbor- hoods, airports, and industrial parks (MacKenzie 1944, Neill 1954, Ligon 1963, Courser 1979). Sev- eral urban Florida Burrowing Owl populations have shown declines, and some urban populations have been extirpated (Courser 1976, Consiglio and Reynolds 1987) . The collapse of a few highly visible urban populations prompted concern for the spe- cies’ status in Florida, and it was listed as a Species of Special Concern by the Florida Fish and Wildlife Conservation Commission in 1979 (Millsap 1996). Subsequent studies have shown that a complex relationship exists between Burrowing Owl popu- lation status and development in some urban are- as. Wesemann and Rowe (1987) showed that Bur- rowing Owl nest density, as well as arthropod and anole {Anolis spp.) prey populations, were highest where houses occupied from 54-60% of the land- scape in Cape Coral, Lee County, Florida. Millsap and Bear (2000) reported that productivity (num- ber of young fledged per occupied breeding site) in this same population increased with increasing housing development until 45-60% of the land- scape was developed. Productivity seemed to de- cline where development exceeded 70%. The objective of this paper is to determine how Burrowing Owl survival rates varied along the de- velopment density gradient on this same Cape Cor- al study area. I also assess implications of variation in survival rates relative to trends in density and productivity described in the previous studies (We- semann and Rowe 1987, Millsap and Bear 2000). Study Area and Methods Study Area. I conducted this work from 1 January 1987-10 July 1991 on a 35.9-km2 study area (of which 32.7 km^ was suitable Burrowing Owl habitat) in Cape Coral, Lee County, Florida, latitude 81°99'N, longitude 26°57'W (Fig. 1). Climate in Cape Coral is subtropical, with an annual mean temperature of 23.LC. Precipita- tion averages 125.7 cm annually, and 75% of rainfall oc- curs between May-September (climate data from NOAA climatological data summaries for Fort Myers, Florida, 20 km southeast of the study area) . The area was historically unsuitable Burrowing Owl habitat, but it was rendered favorable when wetland filling and land clearing for de- velopment occurred in the early 1950s (Zeiss 1983). The study area consisted of filled upland subdivided into 0.2-ha lots suitable for homes, and dissected by ac- cess roads and saltwater canals. With the exception of a golf course (that was excluded from the study area due to access restrictions), three school campuses, and four recreation fields, the entire upland portion of the study area was subdivided for development. Groups of unde- veloped lots of various configurations are interspersed with single-family homes throughout the study area. Va- cant lots were maintained as grasslands by regular mow- ing by city maintenance crews. Developed lots usually contained manicured lawns of fibrous mats of sod with landscaped beds of trees and shrubs. The ratio of homes to vacant lots varied across the study area, with highest development in the eastern sections (where up to 74% of lots had homes built on them) and lowest in western sections (where as few as 2% of lots had homes on them). Definitions. Burrows attended by one or more adult owls or decorated with shredded paper and grass were consid- ered occupied nest sites. A nest site was the area within 88 m (% the mean inter-nest distance [Millsap and Bear 2000] ) of a burrow where a nest attempt occurred, or where a single adult Burrowing Owl not known to be breeding else- where was seen on three or more occasions between 1 Jan- uary-10 July. The term survival ($) does not distinguish between individuals that survived from one year to the next and those that permanently emigrated {e) from the study area, unless it is specifically noted that adjustments to ac- count for e were made. Recapture probability {p) is the probability of encountering a previously banded individual known to be alive in year i. Methods. I obtained measures of the percent of lots that were developed for each of 14 legal sections (2.59 km^) on the study area in each year of the study from the city of Cape Coral. Nearly all upland throughout the study area was divided into 0.2-ha lots, so the percent of developed lots provided a reliable relative measure of the extent of home development in each section. Develop- ment was not necessarily uniform, and school campuses and clusters of undeveloped lots provided pockets of open space even in the most densely developed areas. Consequently the percent of developed lots is best con- sidered an index to the relative level of development around nest sites at a landscape scale. I had insufficient data to assess survival of banded Bur- rowing Owls in each section, so I grouped owls into four development zones according to the percent of lots de- veloped in the section where they were banded or last encountered. Development zones and area were: (1) zone 1, >60% development, 8.0 km^; (2) zone 2, 40-60% development, 5.1 km^; (3) zone 3, 20-39% development, 7.5 km^; and (4) zone 4, <20% development, 12.1 km^. I chose the break point at 60% based on changes in owl population density and productivity that occurred at this level of development (Wesemann 1986, Wesemann and Rowe 1987, Millsap and Bear 2000). Surveys to locate occupied nest sites were conducted from 1987-91 as described in Millsap and Bear (2000) using a team of trained volunteers. From 1987-90 this team banded 20-25% of breeding adult and juvenile Bur- rowing Owls in each section on the study area. Nest sites March 2002 Survival of Urban Burrowing Owls 5 a b t Development 2.3 2.6 2.6 26 2.5 2.6 1 2.1 . 217 ® 27 . 6 % 397 . 58 % : 66 % 63 % + 37 . ♦ 37 o ♦ 77 . ♦ 21 % ♦ 157 . ; * 11 % 1 1 i 1 i 2.6 1 ♦ 13 % 1.7 2.6 2.6 2.6 1 : 2.2 1.1 ^ ’J 27 . 37 . 11 % 337 . 1 72 % ; 577 . 747 . ♦ 37 . ♦ 37 . ♦ 8 % ♦ 237 . 1 ^ 7 % \ ! ^ 10 % 1 ♦ 10 % , j 1 I 1 i 1 . . . t Zone 1 (>60%) Zone 2 (40-60%) " Zone 3 (20-39%) Zone 4 (<20%) km Figure 1. Map of the Burrowing Owl study area in Cape Coral, Lee County, Florida, showing (a) development zones and (b) distribution of nest sites for the period 1987-90. In a, numbers from top to bottom are: (top) number of km^ of suitable Burrowing Owl habitat, (middle) mean percent of lots with homes for period 1987-90, and (bottom) percent increase in home density from 1987-90. were randomly selected for banding from the pool of occupied sites each year, but some owls not selected a priori were also banded. Owls were captured with noose carpets placed at the burrow entrance and by hand at night with flashlights. Approximately 98% of adults at- tending known nests on the study area were checked for bands annually from 1988-91, allowing me to generate annual survival estimates for the periods 1987-88, 1988- 89, 1989-90, and 1990-91. When banded owls were ob- served, we confirmed identification by reading band numbers with spotting scopes or binoculars or by retrap- ping. Members of the public who reported encounters with banded owls were queried to determine the details of the encounter. The sex of breeding adults could usu- ally be determined at a distance by plumage (males were paler than females due to increased sun-bleaching) or behavior (Millsap and Bear 1997). Breeding females with eggs or small young could be distinguished in the hand by the presence of a large, vascularized incubation patch. We were unable to determine the sex of nesdings when they were initially banded, but sex was determined for those that were subsequendy encountered as breeders. Capture-recapture data were analyzed using the Cor- mack-Jolly-Seber (CJS) family of models, which produce estimates and estimated standard errors (SE) of and p. Goodness-of-fit tests in Program RELEASE (Burnham et al. 1987) were used to assess the adequacy and utility of the basic CJS model for my data. Parameter estimates were calculated using Program SURGE (Pradel et al 1990). I calculated estimates of and p separately for owls banded as adults (which were further separated by sex) and nestlings (which were not identified to sex upon initial banding). For both adults and young, I examined parameters over categorical time intervals (subscript t) and development zones (subscript z). For birds banded as juveniles, I also examined variation with age (subscript a). My objective was to find the model with the simplest structure and fewest parameters that still accounted for significant variability in the data. Model notation follows Lebreton et al. (1992). I initiated model testing for adults with a global model of where t denotes a time effect over the four recapture years (1988-91), z denotes an effect over the four development zones, and * denotes interaction between time and development ef- fects. Thus, my adult global model tested separate { pf,^ for each sex over all four recapture occasions and all four development strata. The global model for owls band- ed as nestlings was {^a 2 *t 4 *z 4 \ Pa 2 *t 4 *z 4 \^ where a denotes an age effect over two age classes (juveniles and >l-yr-old) For the subscripts t, z, and a, I distinguished between nonspecific categorical groups with the additional sub- script n, and specific categories with the subscript n' . For example, the notation z4 denotes a model where devel- opment zone effect is partitioned across all four zones, whereas the model subscripted zT, z2' = z3' = z4' de- 6 Millsap VoL. 36, No. 1 notes a model where development zone effect was par- titioned between zone 1 and zones 2-4 pooled. I tested reduced nested variations of these models against the global models using Akaike’s Information Cri- terion (AIC) to distinguish the most parsimonious model from among those tested (Lebreton et al. 1992). The probability that each of the six highest-ranked nested models (based on AIC scores) was the best model was estimated by the AIC weight (w) for each model (Burn- ham and Anderson 1998). In 1988-89, my team and I searched for banded Bur- rowing Owls that had dispersed and settled at nest sites in a 3.2-km-wide band immediately north of study area, as well as south of study area to the southern terminus of the Cape Coral peninsula. I used these data to adjust my estimates of $ to account for permanent emigration using the formula in Burnham et al. (1996). I contacted persons who reported dead banded owls to determine the cause of death whenever possible. Many recoveries were reported by a local wildlife rehabilitation center (Care and Rehabilitation of Wildlife, Inc.), and center veterinarians routinely conducted necropsies on banded Burrowing Owls. Results The percent of lots with homes ranged from <2% to 74% across the study area, as measured at the section level (Fig. 1). The extent of develop- ment changed over the course of the study in all sections, hut the greatest increase in the percent of developed lots was in moderately-developed parts of the study area. In the 20-39% and 40-60% development zones, from 10-23% of lots that were undeveloped at the start of the study had homes on them when the study ended. My team and I banded 581 Burrowing Owls in- volved in 785 breeding attempts on 264 discreet nest sites on the study area from 1987-90 (Table 1, Fig. 1). For the purposes of survival analyses, I assigned each banded owl to the development zone where the owl was located the preceding time it was encountered. Recapture Probabilities and Survival. No model incorporating variation in p with year or develop- ment zone was a satisfactory fit (Table 2), so I pooled data. Overall estimates of p were relatively high; adult males = 91% (SE = 3%), adult females = 87% (SE = 4%), and juveniles = 86% (SE = 5%). No models that incorporated variation among years in were a good ht, so I pooled data over years for survival analyses. The best overall estimate of fb was 71% (SE = 3%) for adult males, and 64% (SE = 3%) for adult females. The unadjusted overall estimate of O for juveniles from the 2 age-class model was 21% (SE = 3%), but this did not account for known emi- Table 1. Capture-recapture data set used to estimate survival of Florida Burrowing Owls from Cape Coral, Lee County, Florida, 1987-91. See Fig. 1 for zone descrip- tions. Number „ „ Number Recaptured Banded y^i yr^+i yru2 y^i+a yri+4 S Adult male 119 83 44 22 2 Zone 1 36 26 19 12 1 Zone 2 35 27 11 4 0 Zone 3 35 22 8 4 1 Zone 4 13 8 6 2 0 X Adult female 152 80 49 22 4 Zone 1 46 25 18 9 3 Zone 2 48 21 12 4 1 Zone 3 38 22 12 5 0 Zone 4 20 12 7 4 0 S Juvenile 310 55 26 13 4 Zone 1 83 7 1 1 0 Zone 2 113 20 11 6 3 Zone 3 91 26 12 6 1 Zone 4 23 2 2 0 0 gration. Five of 35 Burrowing Owls banded as nes- tlings on the study area in 1987-88 were known to have survived to breed and settled at nest sites away from the study area, yielding an estimated e of 0.14 (SE = 0.06) for juveniles. Adjusting for e, $ for age 0—1 yr = 24%. Survival increased among owls banded as nestlings at >1 yr of age to 62% (SE = 6%). No Burrowing Owls banded as adults on the study area were found nesting off the study area, so there was no basis for adjusting to account for breeding dispersal. Patterns of Survival. The adult male survival model with the lowest AIC and fewest parameters pooled together the >60% and <20% develop- ment zones, and pooled together the 20-39% and 40-60% development zones (Table 3). The AIC weight for this model was low overall, but was over twice that of the next best model. The best adult female survival model pooled together the >60%, 20-39%, and <20% development zones, and had an AIC weight 2.5 times that of the next best mod- el. There was little difference in AIC weights among four of the six best survival models for Bur- rowing Owls banded as nestlings, so there was no clear basis for pooling survival estimates for any development zones. The small sample size in the <20% development zone prevented calculation of March 2002 Survival of Urban Burrowing Owls 7 Table 2. Comparison of six best capture-recapture mod- els based on Akaike’s Information Criterion (AIC) for adult male, adult female, and juvenile Florida Burrowing Owls, Cape Coral, Lee County, Florida, 1987-91. Models are listed in order of decreasing fit, based on AIC weights (oj). Model^ Deviance Kb AIC Adult males ' = z4',z2'— zS'5 327.170 3 333.2 0.57 i^zl',z2'=z3',z4'’ 326.820 4 334.8 0.26 P] 326.617 5 336.6 0.11 I^d' = z2',z3',z4'; P) 331.092 4 339.1 0.03 {^z^' = z2' = z3^,zP> P] 334.143 3 340.1 0.02 {4>; p} 336.978 2 341.0 0.01 Adult females {4’z2',z1'=z3'=z4'; P) 410.502 3 416.5 0.54 {4^z1'=z4',z2'=z3'! p} 412.349 3 418.3 0.22 p} 409.447 5 419.5 0.12 {^z4',zl'=z2'^z8'> P) 415.049 3 421.0 0.06 p} 417.193 2 421.2 0.05 i^t4'*z4r P\ 404.259 17 438.3 <0.01 Juvenile {4‘a2*zr=z4',z2'=z34 P) 430.780 5 440.8 0.29 {4^ffi2*zl',z2>z3'=z44 P) 431.317 5 441.3 0.23 {4^a2*z3',zl = z2'=z4'> P\ 431.715 5 441.8 0.18 \^a2*zV,z2' = z3',z4'l P) 428.033 7 442.0 0.16 14>«2*z4; p\ 424.541 9 442.5 0.13 l^«2; p] 441.740 3 447.7 0.001 Model notation is as follows; = survival; p = recapture prob- ability; = categorical age, where n denotes the number of age classes (when n = 2 the model uses 2 age classes, one for owls <1 yr old and one for owls >l-yr-old); or t„. = time, where n denotes the number of nonspecific time categories modeled, and n' denotes a specific time category (i.e., 1' = 1988, 2' = 1989, 3' = 1990, and 4' = 1991); or z„, = development zone, where n denotes the number of nonspecific development zones mod- eled and n' denotes a specific development zone (i.e., 1 ' = >60% development, 2' = 40-60% development, 3' = 20-39% devel- opment, and 4' <20% development). K = number of parameters estimated in the model. (o = AIC weight, which is the estimated probability the partic- ular model is the best of the suite of models evaluated (Burnham and Anderson 1998). meaningful survival estimates for the >1 yr age class for this zone. Overall, estimates of survival for adult males (from the adult model) and juveniles (from the 2 age-class model) were strongly inversely correlated across development zones (Spearman’s rank r — — 0.89). All five juvenile emigrants captured off the study area came from nest sites in the >60% (4 individuals) and 40-60% (1 individual) develop- ment zones. Causes and Timing of Mortality. Cause of death was estimated for 27 of 41 (65.9%) banded owls. Nineteen (70.3%) were hit by cars, six (22.2%) were the victims of predation (three by domestic dogs or cats, three by other raptors), one (3.7%) was killed during home construction on the nest lot, and one (3.7%) died of an unknown illness. The seasonal distribution of recoveries of owls >1 yr of age {N = 30) was not uniform (x^s = 17.2, P -- 0.001) (Fig. 2). Significantly {P < 0.10) more recoveries of adults were reported during the breeding season than expected, and significantly {P < 0.10) fewer were recovered in the fall. No marked differences were evident between males and females. The seasonal distribution of recover- ies of juveniles {N — 11) was also not uniform (x ^3 — 10.5, P = 0.02), with more recovered at and shortly after fledging in spring than expected {P < 0.10) and fewer recovered than expected in fall {P < 0 . 10 ). Discussion Survival rates of Burrowing Owls have not been widely investigated. The only other similarly de- rived survival estimates in the literature are from a migratory population in central Colorado (Lutz and Plumpton 1997), where adult survival aver- aged 39% per yr over a 4-yr period, but varied among years (range = 18-71%), and nestling sur- vival to 1 yr averaged 12%. Clayton and Schmutz (1997) estimated over-summer survival rates at 83% for adult female, 46% for adult male, and 48% for juvenile migratory Burrowing Owls in Al- berta and Saskatchewan using radiotelemetry, but additional mortality would be expected in this pop- ulation on migration and during winter. Estimates of annual survival from band resightings unadjust- ed for emigration range from 37-57% for adults in a migratory population in Saskatchewan (James et al. 1997), to 30% for juveniles and 81% for adults in a sedentary population in Oakland, Cal- ifornia (Thomsen 1971). Estimates for Cape Coral fall within the upper limits of survival from these previous studies, and are most comparable to esti- mates for the Oakland, California population. The Cape Coral Burrowing Owl population shares sev- eral other traits with the Oakland, California pop- ulation, among them relatively low productivity (Millsap and Bear 2000) and high nest site and mate fidelity (Millsap and Bear 1997). These simi- 8 Millsap VoL. 36, No. 1 Table 3. Estimates of annual survival of Florida Burrowing Owls in Cape Coral, Lee County, Florida, 1987-91 from best-fit models using Program SURGE (see Table 2 for models and model selection criteria) . Estimated Annuai. Survival (SE) Model Class^ >60% Development 40-60% Development 20-39% Development <20% Devei.opment Adult models Adult male 0.81 (0.04) 0.62 (0.05) 0.62 (0.05) 0.81 (0.04) Adult female 0.69 (0.04) 0.52 (0.06) 0.69 (0.04) 0.69 (0.04) Age-class models Age 0-1 yr'’ 0.11 (0.03) 0.28 (0.04) 0.43 (0.08) 0.18 (0.08) Age >1 yr 0.46 (0.02) 0.60 (0.09) 0.63 (0.09) — ^ Adult models are based on birds first captured as breeding adults. Age-class models use birds first banded as nestlings at their natal burrow. ’’ Age 0-1 yr survival was adjusted to account for measured emigration following the approach in Burnham et al. (1996). larities might reflect the absence of seasonal mi- gration and mild climate common to both study sites. Vehicle collisions were an important source of mortality for both juvenile and adult Burrowing Owls in Cape Coral. Most road mortality I observed was on residential streets with reduced speed limits (i.e., ^56 km/hr). Highway collision was also found to be a substantial mortality factor in North Dakota (Konrad and Gilmer 1984), Saskatchewan (Haug and Oliphant 1987, Clayton and Schmutz 1997), and Alberta (Clayton and Schmutz 1997), and it is identified as a principal mortality factor in the Canadian Burrowing Owl Recovery Plan (Hjertaas 1997). In radiotelemetry studies of sur- vival, predation has also been shown to be an im- portant source of mortality (Clayton and Schmutz Juvenile Adult Sep Dec Mar Jun to to to to Nov Feb May Aug Figure 2. Histogram of recoveries of deceased banded Burrowing Owls reported by the public by season from Cape Coral, Lee County, Florida, 1987-91. 1997) , and 1 suspect it was more important in Cape Coral than indicated by band recoveries. This was particularly true in heavily-developed areas where cover for predators was high. During the course of the study, 1 observed, in order of decreasing fre- quency, Cooper’s Hawks {Accipiter coopeni), Fish Crows ( Corvus ossifragus ) , house cats, Merlins {Falco columbarius) , and Peregrine Falcons {F. peregrinus) capture Burrowing Owls. Although band recoveries by the public may pro- vide a misleading picture of the timing and season- ality of Burrowing Owl mortality, my results suggest that adult Burrowing Owls are at higher risk of mortality in spring while breeding. This implies that there is a cost of reproduction to Burrowing Owls on the study area, although 1 have insufficient data to compare annual survival for breeders with nonbreeders. The high mortality of juveniles at about the time of fledging is not unexpected. The drop in mortality in both adults and young in fall is surprising, because it is at this time that mortality from predation by migrant raptors should proba- bly increase. As noted earlier, however, this kind of mortality would not be detectible through band re- coveries reported by the public. The inverse relationship between survival rates for adult males, the group for which conclusions regarding survival were least apt to be confounded by emigration because of high nest site fidelity (Millsap and Bear 1997), and juveniles is curious. I suspect this occurred because high adult mortal- ity in moderate-development zones created more opportunities for surviving banded juveniles to set- tle near their natal nest sites where they had a high probability of being encountered. Tbis hypothesis March 2002 Survival of Urban Burrowing Owls 9 is consistent with the high natal philopatry ob- served in this study population (Millsap and Bear 1997) , and is further supported by the limited data on juvenile emigration, which showed higher rates of movement off the study area by juveniles from heavily-developed areas than from less-developed areas. Because emigration appeared to affect ap- parent juvenile survival greatly, actual survival might have been much different. The only conclu- sion, I believe, that can be drawn safely about ju- venile survival is that it was as high as 43% in some parts of the study area. It is not immediately apparent from the available data why adult male, and, to a lesser extent, adult female survival was lowest in moderately-developed parts of the study area. Both Burrowing Owl nest site density and productivity were positively associ- ated with home development in the <20%, 20- 39%, and 40-60% development zones (Millsap and Bear 2000); hence, adult survival was lowest in ar- eas where both density and productivity were rel- atively high. Although it is conceivable that low adult survival was directly related to pressures as- sociated with high adult population density in moderate development zones, this is not consistent with the high rate of juvenile recruitment in these areas. If there was strong competition among adults for nest sites, fewer rather than more 1-yr- olds would be expected to find breeding vacancies to fill (Newton 1991). A possible explanation that better fits the available data is that the compara- tively rapid rate of home construction in the 20- 39% and, in particular, the 40-60% development zones (Fig. la) caused, either directly or indirectly, higher adult mortality. More work is needed to as- certain whether or not this is the case, and to de- termine the mechanism of the effect. Such work could have significant implications for future con- servation of urban Florida Burrowing Owl popu- lations. Acknowtldgments These observations were a product of a cooperative Burrowing Owl monitoring project between the Florida Game and Fresh Water Fish Commission and Audubon Society of Southwest Florida. I am indebted to C. Bear for her assistance in all facets of this project. The work would not have been possible without the dedicated as- sistance and skill of 23 Audubon Society and Lee County School System volunteers. I also gratefully acknowledge T. Wesemann and M. Rowe for freely sharing the results of their work, and for encouraging us to undertake this project. Earlier drafts of the manuscript were greatly im- proved by reviews of D. Cobb, J. Gore, M. Rowe, S. Shar- pley-Evans, and two anonymous reviewers. The Fish and Wildlife Conservation Commission funded this project through the Florida Nongame Wildlife Trust Fund. Literature Cited Burnham, K.R, D.R. Anderson, G.C. White, C. Brownie, AND K. H. Pollock. 1987. Design and analysis meth- ods for fish survival experiments based on release-re- capture. Am. Fish. Soc. Monogr. 5. , D.R. Anderson, and G.C. White. 1996. Meta analysis of vital rates of the Northern Spotted Owl Pages 92-101 in E.D. Forsman, S. DeStefano, M.G Raphael, and R.J. Guiterrez [Eds,], Studies in Avian Biology Number 17. and D.R. Anderson. 1998. Model selection and inference. A practical information-theoretic ap- proach. Springer-Verlag, New York, New York, U.S.A Clayton, K.M. and J.K. Schmutz. 1997. Burrowing Owl {Speotyto cunicularia) survival in prairie Canada. Pages 107-110 in J.R. Duncan, D.H. Johnson, and T.H Nicholls [Eds.], Biology and conservation of owls of the northern hemisphere. U.S. Dep. of Agri. For Serv. Gen. Tech. Rep. NC-190. CoNSiGLio, B. AND G. REYNOLDS. 1987. Broward’s Burrow- ing Owl watchers. Fla. Nat. 60:3-5. Courser, W.D. 1976. A population study of the Burrow- ing Owl near Tampa, Florida. M.S. thesis, Univ. of South Florida, Tampa, Florida, U.S.A. . 1979. Continued breeding range expansion of the Burrowing Owl in Florida. Am. Birds 33:143-144 Franklin, A.B., RJ. Gutierrez, B.R. Noon, and J.P. Ward, Jr. 1996. Demographic characteristics and trends of Northern Spotted Owl populations in north- western California. Pages 83-91 in E.D. Forsman, S DeStefano, M.G. Raphael, and R.J. Guiterrez [Eds], Studies in Avian Biology Number 17. Haug, E.A. AND L. Oliphant. 1987. Breeding biology of Burrowing Owls in Saskatchewan. Pages 269-271 m G.L. Holroyd, W.B. McGillivray, P.H.R. Stepney, D.M Ealey, G.C. Trottier, and K.E. Eberhart [Eds.], Endan- gered species in the prairie provinces, Provincial Mu- seum of Alberta Occasional Paper Number 9. Hjertaas, D.G. 1997. Recovery plan for the Burrowing Owl in Canada. Pages 107-111 m J.L. Lincer and K. Steenhof [Eds.], The Burrowing Owl, its biology and management including the proceedings of the first international Burrowing Owl symposium. J. Raptor Res. Report 9. James, P.C., T.J. Ethier, and M.K. Toutloef. 1997. Pa- rameters of a declining Burrowing Owl population m Saskatchewan. Pages 34-37 in J.L. Lincer and K Steenhof [Eds.], The Burrowing Owl, its biology and management including the proceedings of the first international Burrowing Owl symposium. J. Raptor Res. Report 9. Konrad, P.M. and D.S. Gilmer. 1984. Observation on the 10 Millsap VoL. 36, No. 1 nesting ecology of Burrowing Owls in central North Dakota. Prairie Nat. 16:129-130. Lebreton, J.D., K.P. Burnham, J. Clobert, and D.R. An- derson. 1992. Modeling survival and testing biologi- cal hypothesis using marked animals: a unified ap- proach with case studies. Ecol. Monogr. 62:67-118. Ligon, J. D. 1963. Breeding range expansion of the Bur- rowing Owl in Florida. Auk 80:367-368. Lutz, R.S. and D.L. Plumpton. 1997. Metapopulation dy- namics of a Burrowing Owl {Speotyto cunicularia) pop- ulation in Colorado. Pages 255-259 in J.R. Duncan, D.H. Johnson, and T.H. Nicholls [Eds.], Biology and conservation of owls of the northern hemisphere. United States Department of Agriculture Forest Ser- vice General Technical Report NC-190. Mackenzie, E.S. 1944. Burrowing Owl in Hernando County. Fla. Nat. 17(4):72. Millsap, B. 1996. Florida Burrowing Owl. Pages 579—587 mJ.A. Rodger, H.W. Kale II, and H.T. Smith [Eds.] Rare and endangered biota of Florida. Vol. V. Birds. Univ. Presses of Florida, Gainesville, Florida, U.S.A. AND C. Bear. 1997. Territory fidelity, mate fidelity, and dispersal in an urban-nesting population of Flor- ida Burrowing Owls. Pages 91-98 mJ.L. Lincer and K. Steenhof [Eds.], The Burrowing Owl, its biology and management including the proceedings of the first international Burrowing Owl symposium. J. Rap- tor Res. Report 9. and C. Bear. 2000. Density and reproduction of Burrowing Owls along an urban development gradi- ent. /. Wildl. Manage. 64:33-41. Neill, W.T. 1954. Notes on the Florida Burrowing Owl, and some new records for the species. Fla. Nat. 27- 67—70. Newton, I. 1991. The role of recruitment in population regulation. Proc. Internatl. Ornithol. Congress 20:1689- 1699. Nicholson, D.J. 1954. The Florida Burrowing Owl; a van- ishing species. Fla. Nat, 27:3-4. Pradel, RJ , J Clobert, and J.D. Lebreton. 1990. Re- cent developments for the analysis of multiple cap- ture-recapture data sets: an example concerning two blue tit populations. Ring 13:193-204. Rhodes, S.N. 1892. The breeding habits of the Florida Burrowing Owl {Speotyto cunicularia floridana). Auk 9. 1 - 8 . Thomsen, L. 1971. Behavior and ecology of Burrowing Owls on the Oakland Municipal Airport. Condor 73: 177-192. Wesemann, T. 1986. Factors influencing the distribution and abundance of Burrowing Owls {Athene cunicular- ia) in Cape Coral, Florida. M.S. thesis, Appalachian State Univ., Boone, North Carolina, U.S.A. and M. Rowe. 1987. Factors influencing the dis- tribution and abundance of Burrowing Owls in Cape Coral, Florida. Pages 129-137 in L.W. Adams and D.L Leedy [Eds.], Integrating man and nature in the met- ropolitan environment. National Institute for Urban Wildlife, Columbia, Maryland, U.S.A. Zeiss, B. 1983. The other side of the river: historical Cape Coral. B. Zeiss, publisher. Cape Coral, Florida, U.S.A, Received 27 October 2000; accepted 26 July 2001 Associate Editor: Ian Warkentin J. Raptor Res. 36(1) :1 1-16 © 2002 The Raptor Research Foundation, Inc. BIASES ASSOCIATED WITH DIET STUDY METHODS IN THE EURASIAN EAGLE-OWL Luigi Marchesi and Paolo Pedrini Trento Natural History Museum, SecMon of Vertebrate Zoology, Raptor Conservation Research Unit, via Calepina 14, 38100 Trento, Italy Fabrizio Sergio^ Edward Grey Institute of Field Ornithology, Department of Zoology, South Parks Road, Oxford 0X1 3PS, U.K. Abstract. — The diet of the Eurasian Eagle-Owl {Bubo bubo) was assessed by collecting pellets and prey remains near the nesting cliffs of 21 pairs in the central-eastern Italian Alps between 1993-97. Taxo- nomic and prey mass composition of the diet was compared between two methods of analysis, pellets and prey remains, to assess biases associated with these techniques. When compared with pellets, remains overestimated avian occixrrence, underestimated mammals, and completely failed to detect fish occur- rence {P < 0.0001). Large prey were also over-represented in remains {P < 0.002). Overall, pellets gave a more realistic and diverse picture of Eurasian Eagle-Owl diet, but failed to detect 26 avian species and 12 avian families identified in remains. Biases associated with the two methods may be lowered by pooling items collected by both methods, assuming the minimum possible number of individuals per species per collection event. However, care must be taken to show the relative contribution of each method in the pooled sample. Further research is needed to quantify biases in diet study methods, by using controlled feeding of captive owls. Similar biases may apply to the study methods commonly employed to assess the diet composition of other owls and predatory birds. Key Words: bias-. Bubo bubo; diet, diet assessment methods-, Eurasian Eagle-Owl, Italy. Sesgos asociados con los metodos de estudio de dieta de Bubo bubo Resumen. — La dieta de Bubo bubo fue evaluada mediante colecta de egagropilas y restos de presas cerca de las cornisas de anidacion de 21 parejas en el centra oriente de los alpes italianos entre 1993-97. La composicion taxonomica y la masa de las presas de la dieta fueron comparadas entre dos metodos de analisis en las egagropilas y los restos de presas, para evaluar los sesgos asociados con estas tecnicas. Cuando fueron comparados con las egagropilas, los restos sobre estimaron la ocurrencia de las aves, subestimaron los mamiferos, y fallaron completamente en detectar la ocurrencia de peces (P< 0.0001). Las presas grandes ademas fueron sobre representadas en los restos(P < 0.002). En conjunto, las ega- gropilas dieron una imagen mas real y diversa de la dieta del buho, pero fallaron en detectar 26 especies de aves y 12 familias de aves identificadas en los restos. Los sesgos asociados con los dos metodos pueden ser disminuidos utilizando mancomunadamente los items colectados por ambos metodos, asumiendo el minimo numero posible de individuos por especie por evento de coleccion. Sin embargo, debe tenerse cuidado para la contribucion relativa de cada metodo en la muestra mancomunada. Es necesaria mayor investigacion para cuantificar los sesgos en los metodos de estudio de dieta, usando alimentacion controlada en buhos cautivos. Sesgos similares pueden aplicar al estudio de metodos comunmente empleados para evaluar la composicion de dieta de otros buhos y otras aves de presa. [Traduccion de Cesar Marquez] The diet of raptors has been assessed by analysis of stomach contents, pellets, or prey remains and by observation/photography of prey captured or 1 Corresponding author’s present address; Trento Natu- ral History Museum, Section of Vertebrate Zoology, Rap- tor Conservation Research Unit, via Calepina 14, 38100 Trento, Italy. E-mail address; fabrizio.sergio@zoo.ox.ac.uk delivered to the nest (Marti 1987, Rosenberg and Cooper 1990). Biases associated with such methods have been examined in at least 11 species of di- urnal raptors (Newton and Marquiss 1982, Collopy 1983, Goszczynski and Pilatowski 1986, Simmons et al. 1991, Bielefeldt et al. 1992, Mersmann et al. 1992, Mahosa 1994, Oro and Telia 1995, Real 1996, Sanchez-Zapata and Calvo 1998, Redpath et al. 11 12 Marches: et al. Voi. 36, No. 1 2001). Some common conclusions emerged from these studies: (1) remains usually overestimate the amount of large and conspicuous prey in the diet, such as large birds, large fish, or medium to large mammals; (2) pellets tend to overestimate the oc- currence of medium to small prey, such as small mammals and passerine birds; and (3) when com- pared to direct-observation methods, remains seem to yield a more biased description of diet than pellets, but allow the detection of many un- usual prey types not recorded in pellets, and the recognition of more items to the species level than pellets. Assessment of biases in diet study methods in nocturnal raptors presents additional problems, because of the difficulty of direct observation of prey capture or prey deliveries to the nest that usu- ally occur at night. The Eurasian Eagle-Owl {Bubo bubo) is a noctur- nal top predator, with a generalist diet, locally spe- cialized in medium-sized birds and mammals (Hir- aldo et al. 1976, Donazar et al. 1989). Due to its frequent predation on other diurnal and noctur- nal raptors, Eurasian Eagle-Owl populations can be a limiting factor for those of other birds of prey (Mikkola 1983, Sergio et al. 1999a, 1999b). The Eurasian Eagle-Owl diet has been extensively stud- ied and recently reviewed (Mikkola 1983, Cramp 1985, Donazar et al. 1989, Penteriani 1996). Diet assessment has been carried out through analysis of pellets and/ or remains, but no studies on biases associated with such methods have ever been pub- lished for this species. A correct evaluation of Eur- asian Eagle-Owl diet is particularly important for two reasons: (1) to obtain a better understanding of its habitat use, diet composition, and conserva- tion requirements; and (2) to obtain a more pre- cise assessment of the impact of Eurasian Eagle- Owl predation on other raptors, or other conservation priority species. The aim of this work was to compare pellet contents with uneaten prey remains and to determine the most accurate meth- od to assess the diet composition of this species. In particular we expected: (1) remains to overesti- mate large and conspicuous prey; (2) pellets to overestimate small items; and (3) the two methods to differ in their degree of taxonomic accuracy of prey identification. Study Area Eurasian Eagle-Owls were surveyed in a 1330-km^ study plot, located in the central-eastern Italian Alps (46°04'N, lEOS'E) (Marches! et al. 1999). The area supported a population of 23-25 pairs. Elevation ranged from 70- 2400 m. The landscape was characterized by mountain slopes covered by broad-leaved woodland interspersed with extensive cliffs. The valley floors were intensively cul- tivated or urbanized. Methods Pellets and prey remains were collected near the nest- ing cliffs of 21 pairs between 1993-97, by carefully search- ing the area near and under the nest, and at traditional roosting places. Collections were carried out at regular intervals throughout the year, so as to avoid biases caused by seasonal variations in the diet (Oro and Telia 1995). Prey items were identified by comparison to a reference collection at the Trento Natural History Museum. For each method and when pooling prey items based on both methods, items were identified assuming the minimum possible number of individuals per collection event. For example, if two individuals of a prey type were identified in pellets but only one individual was detected in remains from the same collection event, we registered two items to the pooled sample. Prey mass was calculated based on information provid- ed by Perrins (1987) and Macdonald and Barrett (1993). Eurasian Eagle-Owls usually capture juvenile individuals of prey species larger than a lagomorph (Donazar et al 1989). Thus, half the mean adult mass of these species was employed, following Donazar et al. (1989). In addi- tion, to avoid affecting the calculations of mean prey mass by few unusually heavy prey, no items were assigned a mass of more than 2500 g (ca. half the weight of an adult red fox, Vulpes vulpes, the heaviest local prey) . Statistical Analysis. Comparison of prey taxa between the two methods of diet assessment was performed by means of analysis on contingency tables (Sokal and Rohlf 1981). To avoid cells with inadequate expected fre- quencies, prey items were grouped in the following cat- egories: family Muridae, Gliridae, Erinaceidae, other mammals, Accipitridae, Strigidae, Phasianidae, Colum- bidae, Rallidae, Turdidae, Corvidae, other birds, uniden- tified birds, and fish. Hereafter, we refer to such groups as “main prey categories.” Mean prey mass was compared between methods by means of t-tests (Sokal and Rohlf 1981). Mean number of species identified per detected family was compared between methods with Matched Pairs Mests (Sokal and Rohlf 1981). Dietary breadth within each diet analysis- method was estimated through the Simpson’s index, cal- culated as S p/, where is the relative proportion of each prey category within the sample (Simpson 1949) The index ranges between 0-1, with higher values indi- cating lower diet diversity. Dietary overlap between dif- ferent methods was estimated through the Pianka’s in- dex, ranging from zero (no overlap) to one (complete overlap; Pianka 1973). When comparing pellets and re- mains, different analyses were carried out for birds, mam- mals, and overall vertebrates composition to gain further insights into differences between the methods. When multiple comparisons were carried out on the same data set, the sequential Bonferroni correction was used to ad- just the significance level (Rice 1989). Means are given with 1 SE, all tests are two-tailed, and statistical signifi- cance was set at P < 0.05. March 2002 Biases in Eurasian Eagi.e-Owl Diet Studies 13 Table 1. Percentage frequency of main prey categories of Eurasian Eagle-Owls in the Italian Alps (1993-97), as determined by two methods of diet analysis and by the combination of both methods (pooled). Prey Category Pellets Remains Pooled'^ N{%) % Mass N{%) % Mass N(%) % Mass Mammals: 461 (76) 83.1 53 (28.2) 45.4 493 (65.6) 69.9 Muridae 203 (33.5) 29.3 0 (0.0) 0.0 203 (27.0) 21.2 Gliridae 173 (28.5) 11.9 4 (2.1) 0.5 175 (23.3) 8.7 Erinaceidae 63 (10.4) 25.8 45 (23.9) 35.2 92 (12.3) 27.3 Other mammals'^ 22 (3.6) 16.2 4 (2.1) 9.7 23 (3.1) 12.6 Birds: 130 (21.5) 13.4 135 (71.8) 54.6 243 (32.3) 28.4 Accipitridae 9 (1.5) 1.9 9 (4.8) 4.5 15 (2.0) 2.6 Strigidae 5 (0.8) 0.9 16 (8.5) 4.9 18 (2.4) 2.1 Phasianidae 5 (0.8) 4.2 9 (4.8) 11.3 12 (1.6) 6.2 Columbidae 0 (0.0) 0.0 24 (12.8) 8.7 24 (3.2) 3.4 Rallidae 0 (0.0) 0.0 7 (3.7) 2.0 7 (0.9) 0.8 Turdidae 85 (14.0) 4.4 23 (12.2) 2.2 96 (12.8) 3.6 Corvidae 5 (0.8) 2.0 26 (13.8) 13.6 29 (3.9) 6.1 Other birds'^ 3 (0.5) 1.1 18 (9.6) 7.4 21 (2.8) 3.7 Unidentified birds 18 (3.0) 3 (1.6) 21 (2.8) Fish 15 (2.5) 2.3 0 (0.0) 0.0 15 (2.0) 1.7 Total 606 188 751 Simpson’s index 0.23 0.22 0.13 0.18 0.16 0.16 ^ Calculated by pooling pellets and prey remains and assuming the smallest possible number of individuals per prey species per collection event. Includes families Talpidae, Canidae, Felidae, and Leporidae. ^ Includes families Podicipedidae, Ardeidae, Anatidae, Falconidae, Tetraonidae, Laridae, Cuculidae, Apodidae, Picidae, Hirundinidae, Sturnidae, and Laniidae. Results Comparison of Pellets and Prey Remains. All vertebrates. Pellets and prey remains differed significantly in frequency of mammals, birds, and fish recorded (x^g “ 194:, P < 0.0001; Table 1). Frequency of main prey categories differed be- tween the two methods (x^is = 421, P < 0.0001; Table 1). Twelve species and six families recorded in pellets went undetected in remains. Twenty-six species and 12 families found in remains were not recorded in pellets. Of the main prey categories, two were undetected in pellets and two in remains (Table 1). Fewer items were identified at the spe- cies level in pellets than in remains (76% and 98%; X^i = 76, P < 0.0001). Mean number of identified species per detected family did not vary signifi- cantly between methods (0.97 ± 0.2 and 1.37 ± 0.3, — —1.44, P — 0.16). Mean prey mass was significantly lower in pellets than in remains (332 ±16 and 552 ± 37, respectively; % 88 ,i 85 ~5.3, P < 0.002; Fig. 1). Diet diversity was higher in re- mains than in pellets (Table 1). Pianka’s index of overlap was 0.30 by number and 0.59 by mass. Mammals. Frequency occurrence of taxonomic groups significantly differed between pellets and remains (x^s — 134, P < 0.0001; Table 1). Nine species, four families and one main prey category recorded in pellets went completely undetected in remains, while all species, families, and main prey categories identified in remains were represented in pellets (Table 1 ) . A lower frequency of items was identified at the species level in pellets than in re- mains (78% and 91%; x^i ~ 20, P — 0.0001). Mean number of identified species per detected family was higher in pellets than in remains (1.56 ± 0.3 and 0.56 ± 0.2, respectively; ^ = 3.00, P = 0.034). Mean prey mass was significantly lower in pellets than in remains (333 ±17 and 534 ± 34, respec- tively; ^573 200 “ —5.3, P < 0.003). Dietary breadth was higher in pellets than in remains; the Simpson index was 0.36 in pellets and 0.73 in remains. Pian- ka’s index of overlap between the two methods was 0.29 by number and 0.68 by mass. 14 Marches: et al. VoL. 36, No. 1 33-64 129-256 513-1024 Prey Mass (g) Figure 1, Mass distribution of prey captured by Eurasian Eagle-Owls in the central-eastern Italian Alps (1993-97), as estimated by pellet analysis, remains analysis, or by pooling pellets and remains. Prey items were grouped by mass categories using an exponential distribution in base two. Birds. Taxonomic composition significantly dif- fered between pellets and remains (x^8 ~ 127, P< 0.0001; Table 1). Twenty-six species and 12 families identified in remains went completely undetected in pellets, while only three species and one family identified through pellet analysis were not detect- ed in remains. Two main prey categories went un- recorded in pellets, but were common in remains (Table 1) . A higher percentage of items were iden- tified at the species level in remains than in pellets (82% and 2%, respectively; = 206, P < 0.0001; Table 1). The mean number of identified species per detected family was significantly higher in re- mains than in pellets (1.89 ± 0.3 and 0.68 ± 0.3, respectively; tig — —4.46, P < 0.003). Mean avian prey mass was significantly higher in remains than in pellets (423 ± 38 and 254 ± 37, respectively; ^ 32,112 ~ “3.1, P = 0.002). The Simpson’s index was 0.57 in pellets and 0.13 in remains, indicating a more diverse diet in the latter method. Pianka’s index of overlap between the two methods was 0.56 by number and 0.69 by mass. Comparison of the Pooled Sample with Pellets and Remains. All species, families, and main prey categories recorded in pellets or remains were ob- viously detected in the pooled sample. A higher percentage of items were identified at the species level in the pooled sample than in pellets (81% and 76%, respectively; x^i = 5, P = 0.019) and in remains than in the pooled sample (98% and 81%, respectively; x^i = 49, P< 0.0001). Mean number of identified species per detected family was higher in the pooled sample than in pellets (1.77 ± 0.2 and 0.97 ± 0.2, respectively; = —4.00, P < 0.001) and remains (1.37 ± 0.2, ^30 = 3.29, P = 0.003). Mean prey mass did not differ between the pooled sample and pellets (368 ± 15 and 332 ± 16, ^ 730, 588 “ “1.62, P = 0.12; Fig. 1) and was sig- nificantly lower in the pooled sample than in re- mains (552 ± 37; ^ 730,185 = 5.21, P< 0.001). Diet diversity in the pooled sample was intermediate be- tween that in pellets and remains (Table 1). Over- lap between the pooled sample and pellets was 0.99 by number and 0.94 by mass. Overlap between the pooled sample and remains was 0.46 by num- ber and 0.79 by mass. Discussion The direct observation of prey capture or deliv- ery to the nest is considered the least biased meth- od of diet analysis (Simmons et al. 1991, Bielefeldt et al. 1992) . However, this method is very time-con- suming, often unfeasible for many species, and par- ticularly poorly suited to the study of Eurasian Ea- gle-Owl diet, because of this species’ nocturnal habits and generally inaccessible cliff nest sites. In- direct methods, such as analysis of pellets and re- mains, are thus required. Due to the above diffi- culties, we were unable to compare pellet and remains analyses with direct observation of prey delivered to the nest. However, comparison of the two indirect methods suggested that both of them incorporated inherent biases. Overlap in frequency of main prey categories be- tween the two methods was extremely low. When compared to pellets, remains overestimated birds, underestimated mammals, and failed to detect the presence of fish in the diet. Large prey were also overrepresented in remains when compared to pel- lets. Biases in remains toward underestimation of fish occurrence and overestimation of bird occur- rence and of large prey have been detected in oth- er studies which compared remains with direct ob- servations (Simmons et al. 1991, Bielefeldt et al. 1992, Mersmann et al. 1992, Real 1996). Such bi- ases are probably caused by the different conspic- uousness and rates of deterioration of the different body parts of different taxa (Goszczynski and Pi- latowski 1986, Mersmann et al. 1992). For exam- ple, avian pluckings are generally more colorful and conspicuous than other vertebrates’ remains March 2002 Biases in Eurasian Eagle-Owl Diet Studies 15 (Bielefeldt et al. 1992); in addition, pluckings of large birds are generally more conspicuous and characterized by lower decay rates than those of smaller birds (Goszczynski and Pilatowski 1986, Newton and Marquiss 1982). Overall, remains gen- erally consisted of large, easily-identifiable body parts. As a result, they were more frequently iden- tifiable to the species level than prey items in pel- lets. Comparison of pellets and remains within dif- ferent vertebrates’ prey groups showed biases to be affected by an interaction between different meth- ods of analysis and different prey taxa. Within mammals, remains failed to detect the presence of the whole family Muridae, which accounted for 33.5% of the items in pellets. Remains were strong- ly dominated by the conspicuous, large skins of hedgehogs {Erinaceus europaeus ) . Within birds, pel- lets failed to record the presence of 26 species and 12 families identified in remains. Thus, remains gave a more complete and diverse picture of avian diet, but still overestimated prey size. Einally, fish were recorded only in pellets. Thus, pellets seemed to yield a more balanced and realistic picture of Eurasian Eagle-Owl diet, but remains seemed to be a useful complementary tool to assess avian occur- rence. As an additional advantage, pellets yielded a much higher number of prey items than remains, despite equal search effort by the researchers. Pel- lets were thus a more cost-effective method to col- lect large numbers of prey. Research Implications. Because of the difficulty of direct-observation methods and the biases in- herent in pellets and remains analyses, Eurasian Eagle-Owl diet should be assessed through multi- ple techniques, including the pooling of samples derived from different methods of analysis. In our study, all species and families identified in pellets and remains were represented in the pooled sam- ple. When compared to pellets and remains, the pooled sample showed an intermediate mean prey mass, diet diversity, and frequency of items identi- fied at the species level. Because of the different sample sizes of pellets and remains, the overlap was high between the pooled sample and pellets, and less so between the pooled sample and remains. By reflecting more the taxonomic and prey-mass com- position of pellets and by adding to it the addition- al avian prey detected through remains, the pooled sample lowered biases associated with each method and provided additional insights in Eurasian Eagle- Owl diet composition. The use of pooled samples consisting of pellets and remains has been pro- posed for other raptors (Goszczynski and Pilatows- ki 1986, Simmons et al. 1991, Manosa 1994, Oro and Telia 1995) and has been demonstrated to yield a relatively close fit to diet composition as- sessed by direct observation (Simmons et al. 1991, Manosa 1994, but see Redpath et al. 2001). Eurther research is needed to assess precise bi- ases associated with different methods of diet anal- ysis by means of controlled feeding of captive Eur- asian Eagle-Owls, as carried out for other species (Mersmann et al. 1992, Real 1996). In the absence of such data, we suggest that future studies of Eur- asian Eagle-Owl diet be carried out by using mul- tiple techniques in a complementary way. Pooling samples obtained through pellet and remains can reduce biases, but care should be taken to show the relative contribution of each method to the pooled sample in terms of taxonomic and prey mass composition (Table 1). In the past, pooled samples have been presented without specifying the relative contribution of pellet items and re- mains to the overall sample (e.g., Olsson 1979, Hir- aldo et al. 1975) . Such inaccuracy can produce bi- ases in different studies, making them difficult to compare and adding statistical noise to review anal- yses of diet composition (Donazar et al. 1989). Re- view studies on predation rate of Eurasian Eagle- Owls on raptors or other conservation sensitive species (e.g., Serrano 2000) should take into ac- count biases inherent to different diet-analysis techniques; studies based exclusively on remains or pellets are likely to overestimate or underestimate, respectively, Eurasian Eagle-Owl impact on other large avian species. Such review analyses should be carried out ideally on pooled samples and rerun for each diet analysis technique separately, to com- pare their results. Finally, as similar biases are likely to apply to many other species, we suggest that sim- ilar approaches to diet assessment be used on oth- er owls or other avian predators. Acknowledgments We thank B. Arroyo, G. Blanco, W. Cresswell, J.J. Ne- gro, V. Penteriani, and C.M. Perrins for comments on a first draft of the paper. Literature Cited Bielefeldt, J., R.N. Rosenfield, andJ.M. Papp. 1992. Un- founded assumptions about the diet of the Cooper’s Hawk. Condor 94:427-436. CoLLOPY, M.W. 1983. A comparison of direct observa- tions and food collection of prey remains in deter- 16 Marchesi et al. VoL. 36, No. 1 mining the diet of Golden Eagles. J. Wildl. Manage. 47:360-368. Cramp, S. 1985. Handbook of the birds of Europe, the Middle East and North Africa. VoL 4. Oxford Univ. Press, Oxford, U.K. Donazar, J.A., E. HIRAI.DO, M. Delibes, and R.R. Estrel- la. 1989. Comparative food habits of the Eagle Owl Bubo bubo and the Great Horned Owl Bubo virginianus in six palearctic and nearctic biomes. Ornis Scand. 20: 298-306. COSZCZYNSKI, J. AND T. PiLATOWSKl. 1986. Diet of Com- mon Buzzard {Buteo buteo L.) and Goshawk {Accipiter gentilis L.) in the nesting period. Ekol. Pol. 34:655-667. H 1 RAI.D 0 , E, J. Andrada, and F.F. Parreno. 1975. Diet of the Eagle Owl (Bubo bubo) in mediterranean Spain. Donana Acta Vertebrata 2:161-177. , F., F. Parreno, V. Andrada, and F. Amores. 1976. Variations in the food habits of the European Eagle Owl {Bubo bubo). Donana Acta Vertebrata 3:137-156. MacDonald, D. and P. Barrett. 1993. Mammals of Brit- ain and Europe. HarperCollins, London, U.K. Manosa, S. 1994. Goshawk diet in a mediterranean area of northeastern Spain./. Raptor Res. 28:84-92. Marchesi, L., P. Pedrini, and P. Galeotti. 1999. Densita e dispersione territoriale del Gufo reale Bubo bubo in Provincia di Trento (Alpi centro-orientali) . Avocetta 23:19-23. Marti, C.D. 1987. Raptor food habit studies. Pages 67- 80 in B.A. Giron Pendleton, B.A. Millsap, K.W. Cline and D.M. Bird [Eds.], Raptor management tech- niques manual. Natl. Wildl. Fed. Washington, DC U.S.A. Mersmann, T.J., D.A. Buehler, J.D. Fraser, and J.K.D. Seegar. 1992. Assessing bias in studies of Bald Eagle food habits./. Wildl. Manage. 56:73-78. Mikkoia, H. 1983. Owls of Europe. T. & A.D. Poyser, Cal- ton, U.K. Newton, I. and M. Marquiss. 1982. Food, predation, and breeding season in Sparrowhawks {Accipiter nisus). J. Zool. Bond. 197:221-240. Olsson, V. 1979. Studies on a population of Eagle Owls, Bubo bubo (L.), in southeast Sweden. Viltrevy 11:3-99. Oro, D. and J.L. Telia. 1995. A comparison of two meth- ods for studying the diet of the Peregrine Falcon. /. Raptor Res. 29:207-210. Penteriani, V. 1996. The Eagle Owl (in Italian). Edagri- cole, Bologne, Italy. Perrins, C.M. 1987. Birds of Britain and Europe. Collins, London, U.K. Pianka, E.R. 1973. The structure of lizard communities. Annu. Rev. Ecol. Sysi. 4:53-74. Real, J. 1996. Biases in diet study methods in the Bo- nelli’s Eagle./ Wildl. Manage. 60:632-638. Redpath, S.M., R. Clarke, M. Madders, and S.J. Thir- GOOD. 2001. Assessing raptor diet: comparing pellets, prey remains, and observational data at Hen Harrier nests. Conrfor 103:184— 188. Rice, W.R. 1989. Analyzing tables of statistical tests. Evo- lution 43:223-225. Rosenberg, K.V. and Cooper, R.J. 1990. Approaches to avian diet analysis. Stud. Avian Biol. 13:80-90. Sanchez-Zapata, J.A. and J.F. Calvo. 1998. Importance of birds and potential bias in food habit studies of Montagu’s Harriers {Circus pygargus) in southeastern Spain./. Raptor Res. 32:254-256. Sergio, F., L. Marchesi, and P. Pedrini. 1999a. Black Kite {Milvus migrans) density and productivity in relation to predation pressure by the Eagle Owl {Bubo bubo). Buteo (Suppl.):30. , , , and F. Rizzolli. 1999b. II Gufo reale Bubo bubo come potenziale fattore limitante per tre specie di rapaci diurni. Avocetta 2S:11 4:. Serrano, D. 2000. Relationship between raptors and rab- bits in the diet of Eagle Owls in southwestern Europe: competition removal or food stress? / Raptor Res. 34: 305-310. Simmons, R.E., D.M. Avery, and G. Avery. 1991. Biases in diets determined from pellets and remains: correc- tion factors for a mammal and bird-eating raptor. /. Raptor Res. 25:63-67. Simpson, E.H. 1949. Measurement of diversity. Nature 163:688. SOKAL, R.R. and F.J. Rohlf. 1981. Biometry. W.H. Free- man, New York, NY U.S.A. Received 2 December 2000; accepted 30 October 2001 J. Raptor Res. 36(l):l7-23 © 2002 The Raptor Research Foundation, Inc. DESCRIPTION OF A NEW SUBSPECIES OF THE EGYPTIAN VULTURE (ACCIPITRIDAE: NEOPHRON PERCNOPTERUS) FROM THE CANARY ISLANDS Jose Antonio DonazarJ Juan Jose Negro, Cesar Javier Palacios AND Laura Gangoso Department of Applied Biology, Estacion Biologica de Donana, Consejo Superior de Investigadones CAentificas, Avda AD Luisa s/n 41013 Sevilla, Spain Jose Antonio Godoy Laboratory of Molecular Ecology, Estacion Biologica de Donana, Consejo Superior de Investigadones Cientificas, Avda M" Luisa s/n 41013 Sevilla, Spain Olga Ceballos and Fernando Hiraldo Department of Applied Biology, Estacion Biologica de Donana, Consejo Superior de Investigadones Cientificas, Avda M“ Luisa s/n 41013 Sevilla, Spain Nieves Capote Laboratory of Molecular Ecology, Estacion Biologica de Donana, Consejo Superior de Investigadones Cientificas, Avda M* Luisa s/n 41013 Sevilla, Spain Abstract. — On the basis of four study skins from museum collections and 37 live birds from the island of Fuerteventura, we describe a new subspecies of the Egyptian Vulture (Neophron percnopterus majorensis) from the Canary archipelago. Canarian Egyptian Vultures are significantly larger than Western European and North African individuals. In addition, some genetic differentiation may exist; analyses of the mi- tochondrial DNA control region revealed that there are haplotypes exclusive to the Canary Islands. The current population of Egyptian Vultures in the Canary Islands is 25-30 breeding pairs restricted to the islands of Fuerteventura and Lanzarote, and this subspecies is therefore endangered. Keywords: Egyptian Vulture', Canary Islands', control region', mitochondrial DNA', Neophron percnopterus majorensis; subspecies. Descripcion de una nueva subespecie de alimoche (Neophron percnopterus majorensis) nativa del archipie- lago de Canarias Resumen. — Describimos una nueva subespecie de alimoche {Neophron percnopterus majorensis) nativa del archipielago de Canarias, sobre la base de 4 pieles de museo y 37 individuos capturados vivos. Los alimoches canarios son significativamente mayores que los del oeste de Europa y norte de Africa. Ade- mas, presentan diferenciacion genetica: un analisis de la Region Control del ADN mitocondrial mostro que existen haplotipos exclusivos de las Islas Canarias. La poblacion actual de alimoches canarios es de 25-30 parejas restringidas a las islas de Fuerteventura y Lanzarote, y esta, por tanto, gravemente amenazada. [Traduccion de los autores] The genus Neophron has only a single species, the Egyptian Vulture {Neophron percnopterus), which lives in dry ecosystems of the Palearctic, Ethiopic, and Indo-Malayan biogeographic regions. Two sub- species have been recognized: N. p. percnopterus oc- ^ E-mail address: donazar@ebd.csic.es curring in most of the range of the species: Eu- rope, Africa, Middle East, central Asia, and northwest India, and N. p. ginginianus in most of the Indian subcontinent (Brown and Amadon 1968, del Hoyo et al. 1994). Eollowing Brown and Amadon (1968) the latter is slightly smaller, having yellow as opposed to dark-brown bill and compar- atively-weaker feet and claws. 17 18 Donazar et al. VoL. 36, No. 1 The Canary Islands are an archipelago of vol- canic origin formed in the past 20 million years close to the western African coast (100 km for Fuerteventura, the nearest island) . The islands have a subtropical climate; fauna and flora are re- lated to other Macaronesian archipelagos (Madei- ra, Azores) and to the Mediterranean region. Twenty-seven percent of the vascular plants and 50% of the invertebrates are endemic (Juan et al. 2000). In addition, seven avian species are consid- ered to be endemic of the archipelago (Blanco and Gonzalez 1992). Endemic subspecies have already been described for three birds of prey living in the islands: Common Buzzard (Buteo buteo insularum, Foericke 1903), Eurasian Sparrowhawk {Accipiter msus granti, Sharpe 1890), and Eurasian Kestrel {Falco tinnunculus canariensis, in the West islands, Koenig 1890; and F. t. dacotiae in the East islands, Hartert 1913). Three other raptor species exist in the archipelago without known distinct subspecific status: Egyptian Vulture, Osprey {Pandion haliae- tus), and Barbary Falcon (Falco pelegrinoides) (del Hoyo et al. 1994). In Macaronesia (Palearctic subregion including the Atlantic Islands) , the Egyptian Vulture inhabits the Canary and Cape Verde archipelagos (Banner- man 1963, Bannerman and Bannerman 1965, 1968) . In the Canary Islands this bird was common historically; the species formerly occurred in the islands of La Gomera, Tenerife, Gran Canaria, Fuerteventura and Lanzarote (Martin 1987). At present, the Egyptian Vulture persists only in the two easternmost islands (Fuerteventura and Lan- zarote) with a total population estimated at 25-30 pairs in the year 2000 (pers. observ.). We have observed strong differences in the mor- phology (see below) of this population in relation to that of Western Europe and North Africa. In addition, analyses of the mitochondrial DNA con- trol region revealed that haplotypes exist exclusive to the Canary Islands. This result suggests that the population has been isolated from others for a very long time. Consequently, we propose the recogni- tion of this population at the subspecific level. In this paper, we describe the Canarian Egyptian Vul- tures as Neophron percnopterus majorensis subsp. nov. Description Holotype. This specimen is an immature (sec- ond plumage) male collected on 22 October 1913 in Toston (northwestern coast of Fuerteventura) by D. Bannerman. It is located in the British Museum of Natural History (BMNH) collection (Tring Cat- alogue No. 199). Geographical Distribution. The subspecies is currently endemic to the islands of Fuerteventura and Lanzarote (eastern Canary Islands) . It almost certainly occurred in the remaining islands of the archipelago until extirpated in the 20th century. There are three specimens in the British Museum of Natural History (Tring) which were collected at Gran Canaria (BMNH No. 121) and Tenerife (BMNH No. 231 and BMNH No. 232, respectively) at the beginning of the 20th century. Description. Plumage patterns of color of Neo- phron percnopterus majorensis resembles the nomi- nate subspecies N. p. percnopterus (Brown and Ama- don 1968, Cramp and Simmons 1980). Adult individuals of the Canarian subspecies typically show white plumage impregnated by rufous col- oration, especially in the crown, nape, median wing coverts, breast, and tail. This coloration is var- iable among individuals and seems to be acquired from iron oxides derived from the local soils, which are rich in iron compounds; evidently, this is an ecological attribute associated with this pop- ulation and not a taxonomic characteristic. Can- arian Egyptian Vultures are sedentary and thus differ behaviorally from Western European popu- lations, which are long-distance migrants (Cramp and Simmons 1980). Measurements. We captured and measured 37 wild individuals in 1999—2000 (all birds released after capture), and compared these data from in- dividuals from continental Spain (Table 1). Mea- surements (mm) of the holotype (BMNH-Tring 199) are: flattened wing chord 516.3 mm; tarsus 88.2 mm; tail; 240.6 mm; bill length with cere 63.5 mm; bill length to the distal edge of cere: 31.8 mm. Morphological Comparisons. N. p. majorensis dif- fers from the nominate subspecies on the basis of larger body measurements. Comparisons with live Iberian individuals (Table 1) revealed significant differences for every trait when the effects of age and sex were controlled. Differences were partic- ularly large for body mass, with the island birds being 18% heavier than Iberian vultures. Tail feathers and wing chord were about 4-8% longer in Canarian vultures. Smaller (ca. 2%), but still sig- nificant differences were detected in the length of the tarsus, primary-feather, bill, and bill with cere. Gigantism is well known to occur on islands (Pe- tren and Case 1997, Grant 1998), which has been Table 1. Biometry of live-trapped Canarian vs. European (Iberian) Egyptian Vultures. For each measurement (see methods) mean, range, standard deviation (SD), and sample size (N) are given; % = percent of increase in measurements of Canarian vs. Iberian individuals. Statistical comparisons were performed through three-way Analysis of Variance including effects of age (preadult, adult) , sex, and locality. Significance of each factor is shown (P) . March 2002 Subspecies of Egyptian Vulture 19 Pi w X h o in H w eg b w u o in g S W Pm S M u p W p CZ5 T-H p o OJ o o d iCi m CO lO o o o o o o o u JJ X X X OD QD (U (U -H o p p p p p p p d d d d d d d V V V ^ CO p J> ^ p p 00 csi CO P 00 P CM 1^ ^ 00 CO Ci O 00 r^ CO ^ CO oq p ^ p p p p CM cd CM c6 CO ^ CM lO 1-H ^ ^ o ^ lo oi ^ GO 00 CO ^ ^ CM oc lO ^ CM CO I o lO lO I p d 4 o I o I I O CM CM ^ ^ d p p P p p r-; CO cd 05 00 CM ^ ^ CD 00 CT) ^ CO CO CD CO CM J> CD CM CM lO lO GO GO CO 00 CO CO CO lO CD CO CO O MO O CM GO CO CM O liO CM r-t CO CM ^ CO MO lO ^ ^ GO CD QO CM 0> MO d CM GO CD O 00 GO MO ^ I I I O MO ID ci rr\ ' m r>N. cr^ K ^ S I I o I CM MO p p P p p p p cd MO cd MO P CM CM ^ 00 ^ 00 CO GO CO MO GO CM CM bJ3 ^ t/3 in * H t3 u o d; u b£) fl u c3 a a be D c ^ I 8 sh rt -rf ^ P pq PQ attributed to the absence of dominant species (Thaler 1973). We lack measurements of birds from neighbor- ing “Sahara,” where the small resident population seems to have been extirpated (see below). We have examined one live bird from the former Spanish Sahara (now Morocco), captured in 1975, which is currently at the Zoo of Jerez (Spain). This bird’s measurements (mm) were: wing chord = 505, tail = 256, tarsus = 75.9, primary = 376, bill length — 31.1, bill cere — 59.4, and mass = 1820 g. These measurements are similar to those of con- tinental birds from the Iberian Peninsula (Table 1 ). Measurements from existing museum specimens are not reliable for the description of the Canarian subspecies as there is only a single skin from Fuer- teventura and the taxonomic status of the extir- pated Egyptian Vultures from other Canary Islands is unknown. In addition, problems relative to the condition of skins, small sample size, and the ex- istence of variability linked to age and sex preclud- ed statistical analyses. Nonetheless, we took mea- sures of flattened wing chord (the measure that can be taken with least error in our experience) from specimens deposited at the BMNH (Tring) (Table 2). The four Canarian Egyptian Vulture skins showed the highest values for wing-chord for the entire range. Their values were well above those of individuals from neighboring populations in the western Mediterranean, continental Africa, and Cape Verde Islands. This suggests that birds from the Canary Islands constituted a single sub- species, which is today restricted to Fuerteventura and Lanzarote. There are no additional skins de- posited in other museums, which precludes further comparisons. It seems interesting that Cape Verde Egyptian Vultures are smaller than Canarian birds. Hazevoet (1995) reported that these birds are sim- ilar in plumage to individuals from African popu- lations; he considered them as belonging to the nominal subspecies. Etimology Scientific Name. The name of the proposed new subspecies is derived from “Majorata,” the ancient name of Fuerteventura Island. It was so called by the Spanish conquerors, since the main native guanche tribe of the island was known as the “Ma- jos.” At present, the inhabitants of the island are still called “majoreros.” English Name. According to the species geo- 20 Donazar et al. VoL. 36, No. 1 Table 2. Flattened wing chord measurements (mm) of Egyptian Vulture {Neophron percnopterus) skins kept in the British Museum of Natural History (Tring). Region Subspecies Median Range N Canary Islands N. p. majorensis 505.7 490.0-516.3 4 Cape Verde Islands N. p. percnopterus 489.6 456.0-515.6 6 Western Mediterranean^ N. p. percnopterus 490.0 472.0-510.0 14 East Africa*^ N. p. percnopterus 480.4 450.0-518.0 13 Middle EasK N. p. percnopterus 501.3 485.5-505.0 4 South Arabia^^ N. p. percnopterus 475.7 351.1-491.3 4 Indian subcontinent N. p. percnopterus 493.5 471.0-510.0 14 Indian subcontinent N. p. ginginianus 460.6 441.5-495.3 23 ■* Spain, France, Algeria, Morocco. Abyssinia, Somalia, Kenya. ^ Egypt, Palestine, Turkey. Oman, Socotra, Arabia. graphical distribution of this subspecies we pro- pose the name Canarian Egyptian Vulture. Genetic Differentiation Methods. We used sequence data of the control region of mitochondrial DNA from 11 Egyptian Vultures of different provinces (two from Fuerte- ventura, two from the Balearics, four from conti- nental Spain, one from western Sahara [Morocco] and two from India [N. p. ginginianus^ (Table 3) . Analyses of mtDNA have been particularly com- mon for studies of subspecies and closely related species because of the relatively rapid rate of evo- lution and ease of analysis relative to nuclear (chromosomal) DNA (Wilson et al. 1985, Avise et al. 1987). Total DNA was extracted from blood samples fol- lowing Gemmell and Akiyama (1996) with some modifications. Preliminary work in our laboratory indicated that the order of genes in the mitochon- drial molecule of Egyptian vultures is consistent with the new avian gene order recently described by Mindell et al. (1998). We thus targeted for am- plification by PCR a potentially hypervariable frag- ment located between the conserved sequence block called Fbox and the flanking Thr-tRNA (for primer sequences and PCR conditions please con- tact authors) . Sequencing reactions were run in an ABI-377 automatic sequencing system (Applied Biosystems) in an external laboratory. Sequences were manually edited and aligned. Genetic dis- tance calculations and phylogenetic analyses were performed with the program PAUP-Version 4.0b4a (Swofford 2000). Previous analysis of fam- ilies of captive birds showed a strict maternal in- heritance of the analyzed sequences, indicating that these were of mitochondrial origin and not from nuclear insertions. Results. A fragment of 459 base pairs of control region was used in the sequence analyses. Sequenc- es for the 11 individuals were all different, i.e., each individual presented a unique haplotype (Fig. 1). A total of 53 polymorphic sites were observed (11.5% of the 459 bp sequenced). Absolute num- bers of pair-wise differences ranged from 2 (0.44%) between samples from Canary Island and 35 (7.63%) between Iberian and Indian samples. Net mean Kimura’s 2 parameter distances between In- dian and other groups ranged from 0.066 to 0.077, while the range of distances between non-Indian groups was 0.005 (Iberian and Balearic) to 0.020 (Canarian and Saharian). Consequently, phyloge- netic analyses of these sequences with using differ- ent algorithms approaches (maximum parsimony, minimum evolution, and maximum likelihood) consistently grouped the two Indian sequences in a clade widely separated from the rest of Iberian, Balearic, Canarian or Saharian sequences, with bootstrap values close to 100% (Fig. 1). The two Canarian sequences also consistently grouped to- gether as a subclade within non-Indian sequences, with bootstrap values up to 92%. A clade of Bale- aric sequences was obtained with some algorithms, with values of up to not higher than 62%. Finally, the only Saharian sequence included was more closely related to Iberian or Balearic than to Can- arian sequences. The sequences are deposited at the EMBL databank under accession numbers AJ305147 to AJ305150 {N. p. majorensis) , March 2002 Subspecies oe Egyptian Vui.ture 21 — Canaryl 92 Canary 2 Iberia 1 Iberia 3 Iberia 2 Balearic 1 62 Balearic 2 Sahara 1 Iberia 4 D.tXB changes 1CD India 1 India 2 Figure 1. Phylogenetic relationships among the Egyptian Vulture Control Region sequences identified in this study Minimum evolution tree, constructed with the program PAUP using Kimura’s 2 parameter distances and the neigh- bor-joining algorithm followed by a branch-swapping procedure. Values above branches represent support from 100 bootstrap replications (only values above 50% are shown). to AJ305162 (A^. p. percnopterus) , and j^ 305163 to AJ305166 {N. p. ginginianus) . Diagnosis The new subspecies can be distinguished from individuals belonging to the nominal subspecies percnopterus on the basis of its larger size (Table 1, 2); the difference is strong with respect to Iberian Egyptian Vultures, which constitute the main west- ern Palearctic population. These latter birds mi- grate along the African coast (Cramp and Sim- mons 1980) and, thus, have some chance of 22 Donazar et al. VoL. 36, No. 1 straggling to the Canary Islands. Morphological differences between majorensis and percnopterus of western European and African populations (in- cluding Gape Verde Islands) are as marked as those existing between the subspecies ginginianus and percnopterus in central Asia (Table 2). A wider population genetic screening of the spe- cies will be needed to assess levels of genetic vari- ability and estimations of gene flow among sub- populations. Meanwhile, the analysis of Control Region sequences in a limited number of Egyptian Vultures shows limited low genetic diversity within the Canary Islands populations versus the larger Iberian popnlations. On the other hand, the Can- arian population shows unique mitochondrial hap- lotypes that group monophyletically within N. p. percnopterus non-Indian sequences. Both results are consistent with colonization and further expansion in the islands by individuals from the mainland and suggest limited gene flow between Canarian and other populations. A differentiated evolution of the Canarian population of Egyptian Vultures would have been favored by the isolation resulting from the 100 km-wide sea corridor existing be- tween the eastern coast of Fuerteventura and that of the African continent. Possibility of individual interchange with other breeding populations ap- pears to be low. The current population in the western Sahara seems to be virtually extinct ( J. Donazar unpubl. data), and therefore the main possibility of genetic exchange would seem to come from Iberian birds migrating along the Af- rican coast and straggling to the islands. We cannot mle out the possibility that some Iberian individ- uals reach the island and eventually interbreed with local individuals. However, there is strong ev- idence that Fuerteventura’s Egyptian Vultures con- stitute an ecologically-isolated population with well- differentiated morphological and genetic characteristics. Summarizing, we propose that the naming of this new subspecies is justified. Description of new subspecies, and even of avian species, has been re- cently done on the basis of only biometric ap- proaches and taking in account only a few speci- mens (e.g.. Smith et al. 1991, Forero and Telia 1997, Safford et al. 1995, Preleuthner and Gamauf 1998, Yosef et al. 2000). In some cases the distance between these new taxa and those closely related were lower than that found in our study. We cannot discard that further genetic analyses might permit the discovery of additional haplotypes which may show them closer to continental birds. This would not invalidate, however, the existence of a differ- entiated Canarian lineage as it is shown in this pa- per. Conservation As was stated above, the entire Canarian Egyp- tian Vulture population is restricted to Fuerteven- tura and Lanzarote, where no more than 30 terri- tories remain occupied. Total population has been estimated at around 130 birds (pers. observ.). Breeding success has been extremely low in recent years (ca. 0.5 fledglings/pair in 1998-2000). Cur- rent threats to this population include mortality from power lines due to collision and electrocution (12 cases during the study period; see also Lorenzo 1995), poisoning (4 cases), and lead intoxication by ingestion of lead bullets (pers. observ.). Theft of eggs and young at the nest, and other human activities may account for some cases of nest de- sertion during the breeding period (Palacios 2000). Conservation measures should be directed preferentially to prevent casualties related to power lines, lead intoxication, and illegal poisoning. Re- inforcement with individuals from continental ar- eas is not recommended, on the basis of the ge- netic differences showed by the Canarian birds with respect to those of neighboring populations. Acknowledgments The Consejeria de Medio Ambiente del Cabildo Insu- lar de Fuerteventura and the Project REN 2000—1556 GLO funded this research. Our special thanks to Manuel Miranda and Carlos Alba who stimulated the project and provided logistic support. Gorgonio Diaz and Maria Asuncion Delgado (Consejeria de Politica Ambiental, Gobierno de Canarias) gave scientific banding permits We thank very much Manuel de la Riva, Jose A. Sanchez Zapata, Gema Garcia, Reyes Lopez-Alonso, Manuel Vaz- quez, Yoharaa Enriquez, Ana Calero, Gema Mosquera, Nemesio Hernandez, Damaso Santana, Francisco Garcia, Jose Hernandez, and Eduardo Castilla for assistance throughout the research. DNA sequencing as well as the sexing of five-trapped birds was done in the Laboratory of Molecular Ecology (Estacion Biologica de Donana). Drs. Mark Adams and Robert Prysjones from the British Museum of Natural History (Tring) and Manuel Barcell (Zoo of Jerez de la Frontera) permitted access to speci- mens in their care. Dean Amadon, Jose L. Telia, and Clay- ton M. White provided valuable criticism of the manu- script. fjTERATURE CiTED Avise, J.C., J. Arnold, R.M. Ball, E. Bermingham, T. Lamb, J.E. Neigel, C.A. Reeb, and N.C. Saunders. 1987. In- traspecific phylogeography: the mitochondrial DNA March 2002 Subspecies of Egyptian Vulture 23 bridge between population genetics and systematics. Ann. Rev, Ecol. Syst. 18:489-522. Bannerman, D.A. 1963. Birds of the Atlantic Islands. Vol. I. A history of the birds of the Canary Islands and of the Salvages. Oliver & Boyd, Edinburgh & London, U.K. AND W.M. Bannerman. 1965. Birds of the Atlantic Islands. Vol. II. A history of the birds of Madeira, the Desertas, and the Porto Santo Islands. Oliver & Boyd, Edinburgh & London, U.K. and . 1968. Birds of the Atlantic Islands. Vol. IV. A history of the birds of the Cape Verde Is- lands. Oliver & Boyd, Edinburgh & London, U.K. Blanco, J.C. and J.L. Gonzalez. 1992. Libro rojo de los vertebrados de Espana. Ministerio de Agricultura, Pes- ca y Alimentacion, Madrid, Spain. Brown, L.H. and D. Amadon. 1968. Eagles, hawks and falcons of the world. Country Life Books, London, U.K. Cramp, S. and K.E.L. Simmons. 1980. Handbook of the birds of the western Palearctic. Vol. II. Oxford Univ. Press, Oxford, U.K. DEL Hoyo, J., a. Elliott, and J. Sargatal. 1994. Hand- book of the birds of the world. Vol. 2. Lynx Editions, Barcelona, Spain. Forero, M.G. AND J.L. Telia. 1997. Sexual dimorphism, plumage variability and species determination in nightjars: the need for further examination of the Nechisar Nigh^ar Caprimulgus solala. Ibis 139:407—409. Gemmell, N.J. and S. Akiyama. 1996. An efficient method for the extraction of DNA from vertebrate tissue. Trends Genet. 12:338-339. Grant, P.R. (Ed.). 1998. Evolution on islands. Oxford Univ. Press, New York, NY U.S.A. Hazevoet, C.J. 1995. The birds of the Gape Verde Island. British Ornithologists’ Union, Dorset, U.K. Juan, C., B.C. Emerson, P. Oromi, andG.M. Hewitt. 2000. Colonization and diversification: towards a phylogeo- graphic synthesis for the Canary Islands. Trends Ecol. Evolution 15:104-109. Lorenzo, J.A. 1995. Estudio preliminar sobre la mortali- dad de aves por tendidos electricos en la isla de Fuer- teventura (Islas Canarias). £co/og7a 9:403-407. Martin, A. 1987. Atlas de las aves nidificantes en la isla de Tenerife. Instituto de Estudios Canarios. Monogra- fia XXXII. Tenerife. Mindell, D.P., M.D. Sorenson, and D.E. Dimcheff. 1998 Multiple independent origins of mitochondrial gene order in birds. Proc. Natl. Acad. Sci. U.S.A. 95:10693— 10697. Paiacios, C.J. 2000. Decline of the Egyptian Vulture {Neo- phron percnopterus) in the Canary Islands. J. Raptor. Res. 34:61. Petren, K. and T.J. Case. 1997. A phylogenetic analysis of body size evolution and biogeography in chuckwallas (Sauromalus) and other iguanines. Evolution 51:206— 219. Preleuthner, M. and a. Gamauf. 1998. A possible new subspecies of the Philippine Hawk-Eagle {Spizaetus philippensis) and its future prospects./. Raptor Res. 32: 126-135. Safford, .S. Ash, J.W. Duckwoth, M.G. Telfer, and C. Zewdie. 1995. A new species of nightjar from Ethi- opia. Ibis 137:301-307. Smith, E.F.G., P. Arctander, J. Fjeldsa, and O.G. Amir. 1991. A new species of shrike (Laniidae: Lanianus) from Somalia, verified by DNA sequence data from the only known individual. lUs 133:227-235. SwOFFORD, D.L. 2000. Phylogenetic analysis using parsi- mony (PAUP) . Version 4. Sinauer Associates, Sunder- land, MA U.S.A. Thaler, L. 1973. Nanisme et gigantisme insulaires. La Recherche 4:741-750. Wilson, A.C,, R.L. Cann, S.M. Carr, M. George, U.B Gyllensleu, K.M. Helm-Bychowski, R.G. Higuchi, S R. Palumbi, E.M. Prager, R.D. Sage, and M. Stoneking. 1985. Mitochondrial DNA and two perspectives on evolutionary genetics. Biol. J. Linn. Soc. 26:375-400 Yosef, R., G. Verdoorn, A. Helbig, and I. Seibold. 2000. A new subspecies of the Booted Eagle from southern Africa, inferred from biometrics and mitochondrial DNA. Pages 43-46 in R.D. Chancellor and B.-U. Mey- burg [Eds.], Raptors at risk. WWGBP, Hancock House, U.K. Received 24 January 2001; accepted 23 September 2001 J Raptor Res. 36(l):24-32 © 2002 The Raptor Research Foundation, Inc. DENSITY, NEST SITES, DIET, AND PRODUCTIVITY OE COMMON BUZZARDS (BUTEO BUTEO) IN THE ITALIAN PRE-ALPS Fabrizio Sergio 1 Edward Ckey Institute of Field Ornithology, Department of Zoology, South Parks Road, Oxford 0X1 3PS, U.K. Alberto Boto, Chiara Scandolara, and Giuseppe Bogliani Dipartimento di Biologia Animate, Piazza Botta 9, 27100 Pavia, Italy Abstract. — ^We studied a Common Buzzard (Buteo buteo) population of 32-35 territorial pairs between 1993 and 1999 in a 113-km^ plot located in the central Italian pre-Alps. Density progressively increased from 28 to 31 pairs/ 100 km^. Territorial pairs were regularly dispersed with a mean distance from the nearest occupied nest of 1108 m {N = 108). Eighty-one percent of 108 used nest sites were on cliffs, while the remaining 19% were placed on mature trees. Each year, 16-21% of the nests built by Common Buzzards were taken over by migratory Black Kites (Milvus migrans). Mean laying date was 9 April (earliest = 7 March, latest 30 April, N = 45). Mean clutch size was 2.32 {N = 19). Hatching success was 91% {N ~ 33 eggs from 14 clutches). Mean brood size at hatching was 2.14 {N = 14). Eighty-nine percent of the territorial pairs laid eggs {N = 37) and 72% raised at least one chick to fledging {N = 100). Mean number of fledged young was 1.07 per territorial pair {N = 100), 1.11 per reproductive pair {N = 33), and 1.49 per successful pair {N = 72), with no significant differences among years. Diet was dominated by medium to small passerines, small mammals, and snakes. Recorded density and productivity were comparable and often higher than those reported for other European populations. Human persecution was high until the 1970s, but is currently unimportant. Future conversion of young coppice stands to mature forest could further favor pre-Alpine populations of Common Buzzards. Keywords: breeding success-, Buteo buteo; Common Buzzard', density, diet] forestry, Italy, pre-Alps. Densidad, sitios nido, dieta y productividad de los gavilanes comunes {Buteo buteo) en los Pre Alpes Italianos Resumen. — Estudiamos una poblacion de gavilanes comunes {Buteo buteo) de 32-35 parejas territoriales entre 1993 y 1999 en una parcela de 113 km^ localizada en los pre Alpes del centro de Italia. La densidad incremento progresivamente de 28 a 31 parejas/ 100 km^. Las parejas territoriales estuvieron dispersas regularmente con una distancia media al nido mas cercanamente ocupado de 1108 m {N = 108). Ochenta y uno por ciento de los 108 sitios nidos usados estaban en cornisas, mientras que el restante 19% estaban ubicados en arboles maduros. Cada aho, 16-21% de los nidos construidos por gavilanes comunes tornados en posesion por milanos negros migratorios {Milvus migrans). La fecha media de postura fue 9 de abril (los primeros = 7 marzo, los mas tardios 30 de abril, N = 45) . El tamano medio de la postura fue 2.32 {N = 19). El exito en la postura fue 91% {N = 33 huevos de 14 nidadas). El tamano medio de la nidada en la postura fue 2.14 {N = 14). Ochenta y nueve por ciento de las parejas territoriales pusieron huevos {N ~ 37) y 72% sacaron adelante al menos un polluelo hasta volanton {N = 100). El numero promedio de jovenes volan tones fue 1.07 por pareja territorial {N = 100), 1.11 por pareja reproductiva {N = 33), y 1.49 por pareja exitosa {N = 72), sin diferencias significativas entre ahos. La dieta fue dominada por passeriformes medianos a pequenos, pequehos mamiferos, y culebras. La densidad y productividad registradas fueron comparables y a menudo mas altas que aquellas repor- tadas para otras poblaciones europeas. La persecucion humana fue alta hasta los 70’s, actuahnente no es importante. La futura conversion de los bosqnecillos jovenes a bosques maduros podria favorecer mayormente a las poblaciones pre-alpinas de gavilanes comunes. [Traduccion de Cesar Marquez] ' Present address: Raptor Conservation Research Unit, Museo Tridentino di Scienze Naturali, Via Calepina 14, 38100 Trento, Italy. E-mail address: fabrizio.sergio@zoo.ox.ac.uk 24 March 2002 Common Buzzard Breeding Ecology 25 The Common Buzzard {Buteo buteo) is one of the most abundant European raptors (Bijlsma 1997), Except for a possible decrease in Sweden (Ryttman 1994) , its populations are generally stable or in- creasing, and in some areas still recovering from declines caused by pesticide poisoning and human persecution in the 1950s and 1960s (Taylor et al. 1988, Bijlsma 1997) . Eactors currently limiting den- sity, productivity, and range expansion, include low availability of food and nest sites, direct persecu- tion, and poisoning (Newton et al. 1982, Elliott and Avery 1991, Gibbons et al. 1994, Graham et al. 1995) . In Italy, the Common Buzzard occurs from sea level to an elevation of 1800 m in the Alps (Canova 1992). However, despite its abundance and wide distribution, the species’ breeding ecology and population trends are virtually unknown. The few existing estimates refer mainly to the density of ter- ritorial pairs and are usually based on low sample sizes (Canova 1992). In this paper, we present data on density, nest spacing, diet, and productivity of a sedentary population of Common Buzzards stud- ied for seven years in the Italian pre-Alps. Study Area The study area is a 113-km^ plot located along the Ital- ian margin of Lake Lugano, within the central Italian pre-Alps (45°55'N, 8°50'E). Altitude ranges from 275- 1125 mask The landscape is characterized by forested mountain slopes interspersed with medium-sized cliffs and rare patches of herbaceous and scrub vegetation, caused by frequent burning. Overall, open areas were scarce, mainly due to human modifications, and concen- trated on the valley floors. The area included 16 small villages, all located on the valley floors. Seventy-one per- cent of the area was covered by woodland, 13% by urban areas, 9% by water bodies, 6% by natural grassland, and 1 % by farmland. Dominant tree species in the forest included sweet chestnut {Castanea saliva), downy oak {Quercus pubescens) , sessile oak {Quercus petraea), European hop-hornbeam {Ostrya carpinifolia) , and locust tree {Robinia pseudoaca- cia) . Forests were managed for timber production pri- marily by means of stool shoots regeneration (coppice system; Matthews 1989), with a rotation of 20-30 yr. Ma- ture trees were often maintained as single individuals or in small clumps as seed bearers (coppice with standards; Matthews 1989). However, most of the woodland had been recently felled and consisted of a homogeneous cover of young second growth forest. Some young wood- land patches were being converted to mature woodland, but at the time of study mature forest was still concen- trated on a few steep slopes. Except for forestry operations, human activities were mainly confined to lowlands and mostly absent from the mountain slopes. Climate is temperate continental with wet springs and dry summers (Pinna 1978). Further de- tails on the area can be found in Sergio and Boto (1999) Methods Common Buzzards were surveyed between 1993 and 1999. We censused territorial pairs during the pre-incu- bation period, between 1 February and 15 April, by ob- serving territorial displays and transfers of nest material. Common Buzzards typically refurbish many alternate nests each year, before selecting the one which they even- tually use (Cramp and Simmons 1980). We put effort into finding all the active alternate nests of each pair every year. An alternate nest was defined as active when it contained greenery or freshly broken branches during the preincubation period, and was defined as used when eggs were laid in it. Whenever possible, nests were visited at least three times: (1) about a week after the mean local laying date to assess clutch size; (2) just after hatching to estimate hatching success, brood size, and hatching date; and (3) when the nestlings were older than 45 d to record the number of fledged young (nestlings usually fledge at 50- 55 d; Cramp and Simmons 1980). Nests were checked by climbing the nest tree, descending cliffs with a rope, or watching the nest from a vantage point up the slope with a 20-60X telescope. To minimize the risk of disturbance, nest desertion, or egg/chick predation by Black Kites {Milvus migrans) or Ravens ( Corvus corax) , only nests that could be checked very rapidly were visited during incu- bation/early hatching. Thus, estimates of clutch size, hatching success, number of laying pairs and brood size represented a subsample of nests. Hatching date was es- timated by backdating from the feather development of nestlings first observed when <15 d old, by observations at eight focal nests and reference to information con- tained in Tubbs (1974), Melde (1976), and Cramp and Simmons (1980). Laying date was estimated by subtract- ing 34 d, the median incubation period (Cramp and Sim- mons 1980), from hatching date. Prey remains found m the nest cup during each nest visit were identified assum- ing the minimum possible number of individuals per col- lection event, and by reference to a reference collection and information contained in Debrot (1982). Terminology follows Steenhof (1987): a reproductive pair is one which laid ^1 egg, a successful pair is one which raised &1 nestling until >45 d old, and breeding success is the percentage of successful territorial pairs. A nest area is an area where >1 alternate nest is found within any one year, but where only one pair nests each year (Sergio and Boto 1999, Sergio and Bogliani 1999). Statistical Methods. The degree of regularity of nest dispersion was estimated by means of the G-statistic (Brown 1975), calculated as the ratio between the geo- metric and arithmetic mean of the squared nearest neighbor distances (NND) between used nests and vary- ing between 0 and 1. Values close to 1 (>0.65) indicate a regular dispersion of nest sites (Brown 1975). Statistical significance of the deviation from randomness toward regularity of nest spacing was assessed by means of the test proposed by Clark and Evans (1954). To minimize the bias caused by the NNDs of pairs located along the border of the study area, we applied the correction sug- gested hy Donnelly (1978). Details of mathematical pro- 26 Sergio et al. VoL. 36, No. 1 Table 1. Density, nest spacing, and regularity of nest dispersion of a Common Buzzard population in the Italian pre-Alps (1993-99). Means are given ±SE. Year Territorial Pairs/ 100 km2 (Ah Mean Nearest Neighbor Distance (m) (AO G-Stati.stic z ph 1993 28 (32) 1041 ± 98 (16) 0.776 8.9 <0.001 1994 28 (32) 1057 ± 132 (15) 0.660 9.1 <0.001 1995 29 (33) 1074 ± 108 (17) 0.721 9.5 <0.001 1996 29 (33) 1028 ± 131 (16) 0.614 9.0 <0.001 1997 29 (33) 1381 ± 140 (13) 0.785 13.2 <0.001 1998 29 (33) 1082 ± 134 (13) 0.696 9.6 <0.001 1999 31 (35) 1134 ± 88 (18) 0.818 10.7 <0.001 Total 29 (7)" 1108 ± 44 (108) 0.703 8.06 <0.001 Number of territorial pairs censused in the study area each year. Statistical significance of the deviation of nest spacing pattern from randomness toward regularity (Krebs 1998). ^ Grand mean for the 7 years of study. cedures can be found in Krebs (1998). To meet the as- sumptions of normality, NNDs were loge transformed, and laying dates were square root transformed prior to parametric tests. All means are given with SE, all tests are two-tailed, and statistical signihcance was set at P < 0.05. Results Density and Nest Dispersion. The number of ter- ritorial pairs increased from 32 to 35 through the study period. Density correspondingly increased from 28 to 31 pairs/100 km^ (Table 1). Mean NND did not vary significantly among years (ANOVA, = 1.06, P — 0.39), and was on average 1108 ± 44 m (range — 400-2500 m, N = 108; Table 1). The G-statistic indicated a regular dispersion of nest sites in all years except 1996 (Table 1). The spacing pattern significantly deviated from ran- domness toward regularity in all the study period (Krebs 1998, Table 1). Nest Sites. Mean altitude of used nests was 585 ± 16 m (range = 270-870 m, N = 108) and did not vary significantly among years (// ~ 0.43, P — 0.86). Mean altitude of cliff nests was higher than that of tree nests (608 ± 15 ra and 483 ± 46 m, respectivly; T) me “ 10.53, P = 0.002). Fourteen to 15 pairs were closely monitored every year until we were reasonably sure to have detected all their active alternate nests. On average, these pairs had three active alternate nests (range = 1-7; Table 2), with no year-to-year variation in their mean num- ber (77,96 = 0.02, P = 1.0). Overall, we censused 377 active alternate nest-years; 76% of them were positioned on cliffs and 24% on trees, with no sig- nificant among-year variation in the two propor- tions (x^ = 1.37, df = 6, P = 0.97; Table 2). Cliff nests accounted for 81% of 108 used nest-years, with no year-to-year variation in their frequency of occurrence (x^ = 5.53, df = 6, P = 0.48; Table 2). Of 52 nests which were used at least once during the seven years of study, 15 were placed on trees, 13 on bare rock ledges, and 24 at the base of trees growing from the cliff faces. Of 15 tree nests, seven were placed on sweet chestnut, two on Scotch pine {Pinus silvestris), two on oak {Quercus spp.), and one each on spruce fir (Picea excelsa), Weymouth pine {Pinus strobus), common lime {Tilia europaea), and European ash (Fraxinus excelsior). The mean height of these 15 nests on trees was 15 ± 1 m. Five pairs had alternate nests on both cliffs and trees, and laid eggs in both types of nests in differ- ent years. The mean number of years that a nest was consecutively occupied was 1.2 ± 0.1 for tree nests (range = 1-3, N = 15) and 2.1 ± 0.3 for cliff nests (range = 1-7, N = 37); the difference be- tween the two was significant (Mann-Whitney U test, z = -2.07, P= 0.038). Each year, 16-21% of the active alternate nests were taken over by migratory Black Kites on their arrival (18 March to beginning of April; Sergio and Boto 1999); this percentage did not vary signifi- cantly among years (x^ = 2.88, df — 6, P = 0.82; Table 2). To assess whether Common Buzzards may have selected cliff or tree nests, we compared the frequency of cliff nesting between used nests and active alternate nests. We removed all nests taken over by Black Kites from the sample of active alternate nests, as these were actually not available to buzzards. There was no significant selection for Table 2. Mean number of active alternative nests/pair, percentage of cliff nests, and percentage of nests taken over by Black Kites in a Common Buzzard March 2002 Common Buzzard Breeding Ecology 27 w cn + 1 C OJ 'Sd aj C QJ 05 05 I oo 05 05 c« &H a § a b c V u CJ iH _0 ‘■C 'B CLh o Oh o H 05 05 05 00 05 05 e- 05 05 I to 05 05 lO 05 05 05 05 on 05 05 > CM o --- an + 1 O i-H o an an o +1 an an an O + 1 an an an d + 1 1—i CO d ^ +1 b: o CO d ^ +1 s C35 d ^ + 1 g CO d ^ +1 b: o 00 'c3 Oh OJ fl cu ■B' ai S !h O V > •a o rt O 0 01 o J> C30 o on ^ aO 00 CO C35 lO ' — 1 OO t- ilO O an 00 05 oo oo 1T5 05 GM J> 05 -tft o 50 ^ to an (M 105 i05 on on m- tn- Ip T — [ I> (05 to liO J> 00 'O ccj rt !h 1 ) 05 00 tD J> 05 (M O (M 05 C 1-1 OJ ^ fi as c/5 •5 u OJ c/5 O C/5 V as d R be be o; .S 0 0 ’-R B as te R t:^ R u 2 C/3 QJ c/5 c/3 c/3 V i; R QJ R OJ ■ Tj g • ^ 1 < CJ 'R u R L> w w R c/3 R R s o; C/5 tu QJ u V u U Uh U V U Ph Rh PL, 05 05 00 o q 05 d d 1-H O (M + 1 1 O U- +1 oO 1— H 05 05 05 o d 1— H 1— H 00 1-H 05 d d oT CO 1 05 05 nH 19 68 +1 1T5 + 1 00 CO 05 CO a cn d 1-H c/3 00 R, 1-H 00 d d QJ S-i 05 05 1-^ 86 +1 +1 Oh cO R CO R I— H ^H iR R •M HH OO 00 CM I-H Lh W d d R 05 CM m +1 +1 QJ 05 I-H U 1 — 1 00 QJ o -R t-H ^H •hj .s CM 00 R CM CJ to d d 3 05 05 18 83 + 1 + 1 cx GO 0 t-; Oh I-H 1—H t3 Sh R N N o CM ^H R CQ 105 d d 05 1-H an + 1 + 1 R 05 1-H an 0 ! 1 CO j> R to t-H i-H g d 1-H i-H 0 U CM 00 R CM 1-H 0 d d cn c/3 QJ 05 05 CO 1-H 62 + 1 lO +1 00 CJ CJ R 00 an d 1-H c/> OJ .5 t£i lO •HH CM ^H CJ R OO d d T3 05 CO +1 +1 0 05 r-H R Lh Oh QJ 00 CM j> S-H t-H I-H C+H o c/3 u 5-h !h w U U R rj R. O, _R c/3 s_, d 'R u 05 c/3 u 'r be bo a C/3 X5 d 'R OJ u 0 ^Lh OJ 15 M J 3 Oh d bJD 0 d bo fl R R 1 ’r 0 a cH !/3 c/3 u CJ S in 9 ’c3 t+H ^ 3 s- L-, 0 "r 0 3 0 R • • 1-H ■ d c/5 o Lh O CO JJ Lh V R rH HH R QJ a 0 ■ rH S-H Uh id "5 fi y R CJ d u d :3 R Lh V -*-i (U !« QJ PR s 28 Sergio et al. VoL. 36, No. 1 cliff or tree nests within any of the seven study years (x^ < 1.71, df = 6, P > 0.19). Breeding Season. Birds were observed on their territories all year. Mean laying date did not vary significantly among years (Kruskal-Wallis x^ ~ 11.15, df = 6, P = 0.08). First egg laying dates ranged from 7 March to 30 April, averaging 9 April (SE = 1.60 d, N — 45). No cases of replacement clutches were observed in any year, even after breeding failures occurred early in the breeding season. The mean date of the first flight of a nes- tling in a brood was 19 June (SE = 2.76 d, earliest = 4 June, latest = 5 July, A = 14 broods). Productivity. Mean clutch size was 2.32 ± 0.13 {N = 19). Hatching success was 91% (N — 33 eggs from 14 clutches). Brood size at hatching was 2.14 ± 0.18 {N — 14). Thirty-three of 37 pairs that were monitored laid eggs, and raised a mean of 1.11 ± 0.15 young per pair. There was no year-to-year var- iation in the percentage of successful territorial pairs (x^ = 5.16, df = 6, P = 0.52; Table 3). Overall breeding success was 72% (Table 3). The mean number of fledged young per territorial pair was 1.07 (Table 3), with no significant among-year dif- ferences (p 0,93 ~ 1.52, P — 0.18). The mean num- ber of fledged young per successful pair was 1.49 (Table 3), and did not vary significantly among years (Pe.es ~ T60, P == 0.16). Causes of failure were usually unknown, apart from two cases of par- tial brood predation by Black Kites, and one case in which a young was electrocuted just after fledg- ing. The number of fledged young declined with lay- ing date, but not significantly (r — —0.17, A — 44, P = 0.27). There was no significant correlation be- tween the number of fledged young and nest site elevation (r = 0.04, A = 99, P = 0.67) or NND (r = 0.11, N — 88, P = 0.30). The mean number of fledged young did not differ between cliff nests and tree nests (Pi gs = 0.002, P = 0.97). Diet. Diet was dominated by birds, mammals, and reptiles which accounted for 46, 29, and 21% of 142 prey remains collected, respectively, in the nests of 20 pairs (Table 4) . We were able to assess the age of 36 avian prey individuals: 19% were nes- tlings, 72% were recently fledged juveniles, and 8% were adult individuals. Discussion Eighty-one percent of the nests used for breed- ing in our area were placed on cliffs. In most of Europe, Common Buzzards are tree nesters, and Table 4. Diet of breeding Common Buzzards in the Ital- ian pre-Alps (1993-99), as estimated by food remains (A = 142) collected from nests. Remains collected during 67 visits to 25 nests. Prey Category Number of Items (%) Birds 66 (46) Blackbird {Turdus merula) 26 (18) Eurasian Jay {Garrulus glandarius) 21 (15) Others” 8 (6) Unidentified Passeriformes 11 (8) Mammal 41 (29) Common Mole {Talpa europaea) 8 (6) Muridae spp.'’ 12 (8) Others'^ 21 (15) Reptiles'* 30 (21) Amphibians" 4 (3) Fish 1 (1) “Includes: European Robin {Erithacus rubecula) {N = 2), Green Woodpecker {Ficus viridis) {N = 2), Eurasian Sparrowhawk {Ac- cipiter nisus) {N = 1), Great Spotted Woodpecker {Dendrocopos major) (A = I), Chaffinch {Fringilla coelebs) (N = 1). '’Includes: Pitymys spp. {N ~ 2), bank vole (Clethrionomys glareolus) (N = 1), wood mouse {Apodemus sylvaiicus) {N = 1), yellow necked mouse {Apodemus flavicollis) {N = 1), house mouse {Mus musculus {N = 1), unidentified Muridae (A = 5). “^Includes: red squirrel {Sdurus vulgaris) {N — 4), Crocidura (A= 1), weasel {Mustela nivalis) (A= 1), dormouse {Myoxusglis) (A = 1), brown hare {Lepus europaeus) (A = 1), unidentified mammal (A = 13). Includes: western whip snake {Coluber viridiflavus) {N= 6), Aes- culapian snake {Elaphe longissima) (A = 7), unidentified Colu- bridae ( A = 11), common wall lizard {Podarcis muralis) (A= 4), slow worm {Anguis fragilis) ( A = 2). '^Includes: common toad {Bufo bufo) {N^ = 4). the availability of woodland can be a key factor lim- iting population density (Dare and Barry 1990, Bijlsma 1993, Halley 1993). Cliff nesting does oc- cur throughout their range, but generally at low frequency, and in areas with limited tree availability (Dare 1995). In our study area, buzzards tended to select nest trees within woodland patches that were more mature than those around random trees (E. Sergio and C. Scandolara unpubl. data). Among alternate nest sites, no preference was evident be- tween cliff and tree nests, even though cliff nests were occupied for higher numbers of consecutive years than tree nests. High frequency of cliff nest- ing may have been caused by low availability of suf- ficiently mature woodland patches, even though single tall trees were relatively abundant and wide- spread in the study area. This is consistent with March 2002 Common Buzzard Breeding Ecoi.ogy 29 o cn I o 00 Cu o w _0 '■J3 3 !=L, o Tj d N N d cq o o U "O OJ u o ;g ■p u 3 "0 O a, rt Q u (j G d !h O "C G c/5 OJ !h D 'c/5 (U Q Id jj 3 w u f4 O c/^ P!i UJ Cl, o z ;2 S o p u j tM P 0 d :z w CO U ^ R p — iP D c/!) P O S g 03 o z p CO c/c W u o , , p 03 Q w CO 05 05 05 05 i-H 0 o c/5 P o ffi 0 3 -p 3 S CO S 00 05 CO lO IP id CT5 05 05 05 1-H t-H IP 0) 15 05 oc oc 05 Tj T3 CJ 05 05 IP 05 05 z 0 o • r-< c 05 05 c; g w m IP IP 00 05 05 05 05 OD Oi cTi d G t-h 4:5 d G 0 0 !U t:^ p d 5 h U u > > P P d 0 'Xi >Z*- ex P P p , ., GO , — ^ , V CO 05 — ^ V5 O o 03 O ccj s c/5 IP 05 05 T— ( G G d a z: o CO 05 ■U 05 ffi ^ z G O • U (D ffi Ccj -i-J o cci p O > Q 15 P cd Ph » i-H d ■ p 15 j-t 0 15 PP 0 CO Th IP C'l(M(Xl>O0O0(P(^CO 'c^^ GO CO C5| ooo5ipo'ti'foo'ti IPIPCOCOCOCMCOJ>0 1 — I Pd 1 — I 1 — I 1 — I CM CM 1 — I CM i> ^ 00 CO ip 1-H r-( IP -'t' CO CO O ^ CM O ^ Pd ^ CO o CO 00 00 J> CO CO CM ^ ^ ^ O O CD p* CO cp* CO O^ ^ ^ M GO^ ^ OO IP J> J> Ip CO 00 05 00 CD ^ o o o o ^ ^ ^ ^ ^ O O CO ^ ip* CO rH CD c^ Oi IP CD i> Pi CO CM c-H i> 00 O CO i:-' GO ip CO ip CD CO CO Pd 05 GO CO o o J> o o o PJ i> CM , , CM pt^ ffi 00 00 pt< CO 05 1-H 05 0 05 0 H H N g r-- I-H X CO 1-H CM I— H Z on Ph 0 0 05 0 to GO X CM d 0 CM C5 q q q i> q q pi CM pi pi pi 00 pi pi pi 0 0 y--V. ^ ^ — , ,, — V 00 n y — N g CD I> Pi X CM j> (O p 0 HH g r— 1 i—H rH 05 X I-H Pi 1— H 1-H z Uh ■■._^ * — ' 05 ^H i> q q I-H q q 1-H i-H I-H 5-H CM 1-H I-H 0 pi 1-H u fj a CM oT (d CM S' ip pi' o' S' GO c/5 CM I-H I-H I-H 4 ; 0 PT I-^ i-r ; z w '—' ' *o X '■ — ' CO 00 0 CM 5—1 J> X GO CO 00 OD p CM i“H iO r-H Pt^ rH CM c/5 C/5 c/3 C/3 C/5 C/3 c/3 G G G G c3 cd cd P P P P 4_t ^ -4_> -M -M cJ d s Ch z z d d 3 d z d q d o o o o o o o a a a a a a a o Vi T3 d -p 3 be - H a ffi g o u p JG d o" D d OJ O o T5 Td ■o T3 d d d d d p p p p p 3 3 3 3 3 0 0 0 0 0 H H X H X c/3 c/3 !/3 V5 .B .S .B .S T 3 T 3 T 3 'O d d d d d p P P P P X 0 0 0 0 0 1 ) u 1 ) D u ex ex ex ex ex d • ^ d o dS & o d iz> D X +J d be p o d ^ ■5 a ® ^ V Vi a Q -g S X 3 ^ c ^ c/3 ^ ^ be W 3 d w ^ [jJ p (h d 15 d d o d t/5 ^ .a a 3 ^ c/3 ^ o p d d d d o o 1 ffi X 2 t/) u X X a .* d d 3 p p 0 1.1 ^ OJ 15 > (1.0 0* d p d. ^ &. U —1 3 .a ^ 05 d d a. o .a < d p p ^ 05 P 05 '~l cd 05 Sh 1 X a z V be d -4-H != a; u V Oh qj 0) u y d V a N :=*^ Q fd fcsj V (J u d o ce C/D ces 30 Sergio et ai.. VoL. 36, No. 1 buzzards responding more to the structural char- acteristics of breeding woodland patches than to the micro-characteristics of individual nesting trees. Alternatively, cliff nesting may be a response to the recent history of persecution and nest rob- bing suffered by the species in this area (Bianchi et al. 1969). From discussion with local villagers, we know of at least three territories where nestlings were regularly collected up until the 1970s, and we have indirect evidence of nest robbing at one easily accessible site during our study. In a high-perse- cution area of Sicily, only two tall and inaccessible tree nests out of seven were not robbed of chicks (Cairone 1982) . Cliff nests are generally less acces- sible than tree nests to humans and cliffs allow buz- zards to place their nests higher from the gronnd than trees. In our study area, cliff nests were also on average at a higher elevation than tree nests, affording additional advantages in terms of dis- tance from sources of human disturbance, which are mostly located at low altitude in the valley floors. Thus, the interaction between the selective pressure associated with potential nest robbing and the low availability of matnre woodland patches may cause the local high frequency of cliff nesting, a pattern also observed in the local Black Kite pop- ulation (Sergio and Boto 1999). Each buzzard pair had on average three alternate nests, and up to seven, within its nest area. This is in agreement with data from other parts of Europe; Tubbs (1974) reported an average of 3.2 alternate nests per nest area (range 1-14) for the New Forest of England. In our study area, some nests, especially on cliffs, were used for a number of consecutive years. However, most nests were used for only one or two years. Such frequent nest switching was pos- sibly enhanced by competition with Black Kites, but has also been reported in other Common Buz- zard populations free of such competition (Tubbs 1974, Cramp and Simmons 1980). Despite the sporadic persecution, the observed density and productivity were in the range of that reported for other European populations (Table 5). In Europe, Common Buzzard breeding densi- ties peak in areas of lowland traditional farmland interspersed with abundant mature woodlots (Bijls- ma 1997). Density in the Italian pre-Alps was only slightly lower to that found in such optimal agri- cultural habitats (Bijlsma 1993, Kostrzewa 1996, Dare 1998, Goszczynski 1997), and higher than any published estimate for mountainous areas (Dare and Barry 1990, Halley 1993, Graham et al. 1995, Penteriani and Faivre 1997; Table 5). Productivity was also comparable or higher than those reported for other mountainous environments (Dare 1995, Swann and Etheridge 1995) and for some lowland areas (Kostrzewa 1996, Dare 1998; Table 5). The diet of the study population was diverse, as typical for this species (Cramp and Simmons 1980), and dominated by birds, small mammals, and snakes. We caution that diet analyses based on prey remains tend to overestimate large or con- spicuous prey species compared to analysis of pel- lets or direct observations of prey delivered to the nest (e.g., Goszczynski and Pilatowski 1986, Red- path et al. 2001, Marches! et al. 2002). However, preliminary results of the analysis of 366 pellets gave a picture of diet composition similar to that obtained by the analysis of remains in the nest (F. Sergio and C. Scandolara unpubl. data). Overall, the high frequency of reptiles conhrms the impor- tance of such prey for Common Buzzards in Med- iterranean countries and at southern latitudes (Cramp and Simmons 1980, Haberl 1995). Finally, the frequent occurrence of typical woodland spe- cies in the diet agreed with our many qualitative observations of individuals hunting by sit-and-wait tactics within woodland habitats. On such occa- sions, buzzards usually perched on intermediate- height branches scanning the forest floor and can- opy for periods of 2-5 min, before moving to another perch on a nearby tree (pause-travel tactic; Widen 1994). The exploitation of a wide range of habitats, the selection of suitable nest sites inaccessible to hu- mans, and the adoption of a diverse opportunistic diet allowed Common Buzzards to settle at a rela- tively good density and reproduce successfully with- in the heavily wooded landscape of the central Ital- ian pre-Alps. The local breeding population was stable or slightly increasing in number. No strong threats were apparent; persecution was sporadic and the continued succession of coppice woodland to mature forest could further increase available nesting and foraging habitat. The role of habitat availability, weather, diet, and competition with Black Kites as potential factors limiting density and breeding performance of the local Common Buz- zard population is currently under investigation. Acknowi.edgmeni s We thank F. Saporetti, P. Pavan, and W. Guenzani for informing us of the location of some nests. We thank L Marchesi for the identification of avian prey remains, and P. Pavan and A. Scandolara for help with the fieldwork March 2002 Common Buzzard Breeding Ecology 31 We thank J.C. Bednarz, M. Donaghy Cannon, W. Cres- swell, L. Gomulski, J.L. Quinn, C.M. Perrins, and an anonymous referee for commenting on a first draft of the paper. Part of the research was funded by the Trento Museum of Natural History. Literature Cited Austin, G.E. and D.C. Houston. 1997. The breeding performance of the Buzzard Buteo buteo in Argyll, Scot- land and a comparison with other areas in Britain. Bird Study 44:146—154. Bianchi, E., L. Martire, and A. Bianchi. 1969. Gli uccelli della provincia di Varese (Lombardia). Riv. ItaL Or- nitol. 39:384—401. Bijlsma, R.G. 1993. Ecologische atlas van de Nederlandse roofvogels. Schuyt, Haarlem, Netherlands. . 1997. Buzzard. Pages 160-161 in W.J.M. Hage- meijer and M.J. Blair [Eds.], The EBCC Atlas of Eu- ropean breeding birds, their distribution and abun- dance. T. & A.D. Poyser, London, U.K. Brown, D. 1975. A test of randomness of nest spacing. Wildfowl 26:102-103. Cairone, a. 1982. Successo riproduttivo di Gheppio, Gril- laio e Poiana nel territorio di Roccapalumba (Sicilia). Avocetta 6:35—40. Canova, L. 1992. Poiana. Pages 569-576 in P. Brichetti, P. De Eranceschi and N. Baccetti [Eds.], Eauna dTtalia. Edizioni Calderini, Bologna, Italy. Clark, PJ- AND E.C. Evans. 1954. Distance to nearest neighbor as a measure of spatial relationships in pop- ulations. Ecology 35:445—453. Cramp, S. and K.E.L. Simmons. 1980. Handbook of the birds of Europe, the Middle East and North Africa. Vol. 2. Hawks and Bustards. Oxford Univ. Press, Ox- ford, U.K. Dare, P.J. 1995. Breeding success and territory features of Buzzards Buteo buteo in Snowdonia and adjacent up- lands of north Wales. Welsh Birds 1:69-78. . 1998. A Buzzard population on Dartmoor, 1955- 1993. Devon Birds 51:4—31. andJ.T. Barry. 1990. Population size, density and regularity in nest spacing of Buzzards Buteo buteo in two upland regions of North Wales. Bird Sudy Al\'2A— 29. Debrot, S. 1982. Atlas des polls de mammiferes d’Europe. Institut de Zoologie de I’Universite de Neu- chatel, Neuchatel, Switzerland. Donnelly, K. 1978. Simulations to determine the vari- ance and edge-effect of total nearest neighbor dis- tance. Pages 91—95 in I. Hodder [Ed.], Simulation methods in archeology. Cambridge Univ. Press, Lon- don, U.K. Dvorak, M., A. Ranner, and H. Berg. 1993. Atlas der Brutvogel Osterreichs. Bundesministerium fiir Um- welt, Jugend und Eamilie. Wien, Austria. Elliott, G.D. and M.I. Avery. 1991. A review of reports of Buzzard persecution 1975-1989. Bird Study 38:52- 56. Gibbons, D., S. Gates, R.E. Green, R.J. Euller, and R.M Euller. 1 994. Buzzards Buteo buteo and Ravens Corvus corax in the uplands of Britain: limits to distribution and abundance. Ibis 137:S75-S84. Goszczynski, J. 1997. Density and productivity of Com- mon Buzzard Buteo buteo and Goshawk Accipiter gentilis populations in Rogow, Central Poland. Acta Ornithol- ogica 32:149-155. AND T. Pilatowski. 1986. Diet of Common Buz- zard {Buteo buteo h.) and Goshawk {Accipiter gentilis h ) in the nesting period. Ekologia Polska 34:655-667. Graham, I.M., S.M. Redpath, and S.J. Thirgood. 1995. The diet and breeding density of Common Buzzards Buteo buteo in relation to indices of prey abundance Bird Study 42:165—173. Haberi., W. 1995. Zum Beutespektrum des Mausebus- sards, Buteo buteo (Accipitridae) , im Waldviertel (Nie- dorosterreich) . Egretta 38:124—129. Halley, D.J. 1993. Population changes and territorial dis- tribution of Common Buzzards Buteo buteo in the Cen- tral Highlands, Scotland. Bird Study 40:24-30. Henrioux, P. and J. Henrioux. 1995. Seize ans d’etude sur les rapaces diurnes et nocturnes dans E Quest le- manique (1975-1990). Nos OAmwx 43:1— 26. Hohmann, U. 1995. Untersuchungen zur Raumnutzung und zur Brutbiologie des Mausenbussards {Buteo bu- teo) im Westen Schleswig-Holsteins. Corax 16:94-104. Kostrzewa, a. 1991. Interspecific interference competi- tion in three European raptor species. Ethology Ecology & Evolution 3:127-143. . 1996. A comparative study of nest-site occupancy and breeding performance as indicators for nesting- habitat quality in three European raptor species. Ethol- ogy Ecology & Evolution 8:1—18. Krebs, C.J. 1998. Ecological methodology. HarperCollins, New York, NY, U.S.A. Marchesi, L., P. Pedrini, and F. Sergio. 2002. Biases as- sociated with diet study methods in the Eagle Owl. J Raptor Res. 36:11-16. Matthews, J.D. 1989. Silvicultural systems. Oxford Univ. Press, Oxford, U.K. Melde, M. 1976. Der Mausebussard. Die Neue Brehm- Biicherei, Wittenberg Lutherstadt, Germany. Newton, L, RE. Davis, and J.E. Davis. 1982. Ravens and Buzzards in relation to sheep-farming and forestry m Wales./. Appl. Ecol. 19:681-706. Penteriani, V. and B. Faivre. 1997. Breeding density and landscape-level habitat selection of Common Buz- zards {Buteo buteo) in a mountain area (Abruzzo Ap- pennines, Italy). / Raptor Res. 31:208-212. Pinna, M. 1978. L’atmosfera e il clima. U.T.E.T., Torino, Italy. Redpath, S.M., R. Clarke, M. Madders, and S.J. Thir- good. 2001. Assessing raptor diet: comparing pellets. 32 Sergio et al. VoL. 36, No. 1 prey remains, and observational data at hen harrier nests. Condor 103; 184— 188. Ryttman, H. 1994. Overlevnadsberackningar och fbrsdk att skatta populationsutvecklingen hos fiskgjiuse Pan- dion haliaetus, ormvrak Buteo buteo och sparvhok Accip- iter nisus i Sverige. Ornis Svecica 4:159-172. Sergio, F. and G. Bogliani. 1999. Eurasian Hobby den- sity, nest area occupancy, diet, and productivity in re- lation to intensive agriculture. Condor 101:806-817. and a. Boto. 1999. Nest dispersion, diet, and breeding success of Black Kites {Milvus migrans) in the Italian pre-Alps./. Raptor Res. 33:207-217. SrEENHOF, K. 1987. Assessing raptor reproductive success and productivity. Pages 157-170 in B.A. Giron Pen- dleton, B.A. Millsap, K.W. Cline, and D.M. Bird [Eds.] , Raptor management techniques manual. Natl. Wildl. Fed., Washington, D.C., U.S.A. Swann, R.L. and B. Etheridge. 1995. A comparison of breeding success and prey of the Common Buzzard Buteo buteo in two areas of northern Scotland. Bird Study 42:37—43. Taylor, K., R. Hudson, and G. Horne. 1988. Buzzard breeding distribution and abundance in Britain and Northern Ireland in 1983. Bird Study 35:109-118. Tubbs, C.R. 1974. The Buzzard. David & Charles, Newton Abbot, U.K. Widen, P. 1994. Habitat quality for raptors: a field exper- iment. /. Avian Biol. 25:219-223. Received 3 September 2000; accepted 29 April 2001 J Raptor Res, 36(l):33-38 © 2002 The Raptor Research Foundation, Inc. VITAMINS E AND A, CAROTENOIDS, AND EATTY ACIDS OE THE RAPTOR EGG YOLK Nigel W.H. Barton^ and Nicholas C. Fox The Falcon Facility, National Avian Research Center, P.O. Box 19, Carmarthen, Wales, SA39 5YL, U.K. Peter F. Surai and Brian K. Speake Avian Science Research Centre, Scottish Agricultural College, Auchincruive, Ayr, Scotland, KA6 5FTW, U.K. Abstract. — ^A captive population of falcons was fed a diet containing a known quantity of vitamin A (retinol) and vitamin E (a-tocopherol) for 6 wk prior to and during egg laying. Infertile eggs were analyzed for vitamin A, vitamin E, carotenoid, and fatty acid composition. Mean daily vitamin intake was 29 mg VitE (35IU) and 1157 |xg VitA (3363IU). Adjusted mean egg yolk content for infertile, unincubated eggs was 314 |xg/g ct-tocopherol and 3.06 pg/g VitA. A distinctive feature of the raptor egg yolk is a very high proportion of arachidonic acid that is probably a reflection of their carnivorous diet. A small number of plasma samples were also available from egg-laying falcons. Mean plasma vitamin E was 32.2 pg/ml and plasma vitamin A 1.02 pg/ml. Keywords: raptor nutrition', egg yolk, vitamins; fatty acids; plasma. Vitaminas E y A, carotenoides, y acidos grasos de la yema de huevos de rapaces Resumen. — Una poblacion cautiva de halcones fue alimentada con una dieta que contenia una cantidad conocida de vitamina A (retinol) y vitamina E (a-tocoferol) por seis semanas antes y durante la postura de huevos. En los huevos infertiles fue analizada la composicion de vitamina A, E carotenoides y acidos grasos. La entrada media diaria de vitamina fue 29 mg VitE (35IU) y 1157 pg VitA (33631U). La media ajustada para el contenido de yemas de huevos infertiles y no incubados fue 314 pg/g a-tocopherol y 3.06 pg/g VitA. Una caracteristica distintiva de la yema de huevo de rapaces es una proporcion muy alta de acido araquidonico lo que probablemente es un reflejo de su dieta carnivora. Estuvieron di.s- ponibles tambien un pequeho numero de muestras de plasma de halcones durante la postura. El pro- medio de vitamina E en el plasma fue 32.2 pg/ml y 1.02 pg/ml de vitamina A. [Traduccion de Cesar Marquez] In recent years attention has focused on the breeding of certain raptor species in captivity as a means of conservation (Cade 1988, Fox and Fox 1993); one example is the Fiji Peregrine Falcon, Falco peregrinus nesiotes (D. Brimm pers. comm,). Productivity is dependent on good-quality eggs and without baseline data for raptor species it is diffi- cult to assess egg quality in a breeding project. There are few data on the egg composition of wild or captive Falconiformes and virtually nothing is known about the fatty acid and the antioxidant profiles of the yolks of these species. Wild raptors have a predetermined clutch size and nutrients from the female, deposited in the egg prior to lay- ing, provide all the necessary nutrition for the em- bryo to develop and for the chick to survive for a ^ E-mail address: nigel-barton@easynet.co.uk few days after hatching. Depending on the species, clutch size in Falconiformes is usually no more than five eggs. However in captivity, this number can be increased to as many as 14 eggs in one sea- son by techniques of egg pulling and clutch pulling (Weaver and Cade 1991) and the requirement for nutrients is therefore much higher. Ideally, captive populations should be fed the same prey items that they would eat in the wild (Glum et al. 1997). Be- cause this is often impractical, it is important to provide a varied, balanced diet with, if necessary, additional supplements of vitamins and minerals to ensure that the egg has sufficient nutrients needed to support successful development (Clum et al. 1997, Fox and Barton 2000). In the yolk, vitamin E and carotenoids are lipid- soluble antioxidants which protect the developing embryo and chick against peroxidative damage, 33 34 Barton et ai.. VoL. 36, No. 1 Table 1. Vitamin and water content of food items fed to captive raptors (Forbes and Flint 2000). Food Type Sample Size VitA IU/100 g DM VitE IU/100 g DM Percent Water Content Whole day-old chick 200 497 40.7 76.1 De-yolked, day-old chick 200 363 21.4 78.5 VitE-enhanced quail 100 3633 10.1 66.6 regulate aspects of cell differentiation, and pro- mote the function of the immune system (Surai and Speake 1998, Surai 1999). These are stored in the maternal liver and mobilized during the laying cycle. Studies on the Lesser Black-backed Gull {La- rus fuscus) have shown that the concentrations of vitamin E and carotenoids in the yolk decrease with each egg laid as the maternal reserves become depleted (Royle et al. 1999). It is therefore possible that captive breeding programs which involve the extension of clutch sizes to increase productivity may result in eggs which are deficient in these an- tioxidants with potentially-harmful consequences for embryonic survival. There is also evidence that some captive-breeding projects for raptors and oth- er birds may not be producing at maximum capa- bility due to an inadequate dietary provision of vi- tamin E (Nichols and Montalli 1987, Dierenfeld et al. 1989). For example, captive Peregrine Falcons fed on whole quail {Coturnixsp^.) achieved plasma vitamin E concentrations of only about 3 |xg/ml compared with about 26 p.g/ml in wild counter- parts; injection or dietary supplementation of the quail with vitamin E was necessary for the captive peregrines to attain plasma vitamin E levels similar to those typical of the wild birds (Dierenfeld et al. 1989). Thus, supplementation of prey items with vitamin E may be needed to achieve optimal re- productive performance in captive raptors. The polyunsaturated fatty acids of yolk lipids, es- pecially the long-chain polyunsaturates arachidon- ic (20:4n-6) and docosahexaenoic (22:6n-3) acids, have vital roles in the functional development of certain embryonic tissues, particularly the brain and retina (Speake et al. 1998). For many avian and reptilian species, eggs produced in captivity of- ten display markedly reduced levels of n-3 poly- unsaturates, vitamin E, and carotenoids in compar- ison with eggs laid in the wild. Noble et al. (1996) and Speake et al. (1999a) suggested that these dif- ferences may be related to low hatchabilities in the captive situation. The main aim of the present study was to eval- uate the yolk concentrations of vitamins E and A and carotenoids in seven raptor species and one hybrid as part of a viable captive breeding program in which the female parents were fed on vitamin E-enriched quail and day-old chickens ( Gallus gal- lus) . Since there is currently very little information on the effects of a carnivorous diet on yolk lipids, the fatty acid composition of the yolk is also re- ported. Methods Eggs were collected from captive raptors held at the Falcon Facility, U.K. Infertile, unincubated eggs were tak- en from imprint falcons that laid eggs prior to insemi- nation. Other eggs analyzed were infertile, but had been incubated for 14 days, at which time infertility was con- firmed. All the females were fed a diet of vitamin-E-sup- plemented quail, day-old cockerels, vitamin supplements (Nekton E and Nekton S — Gunter Enderle, Pforzheim, Germany) and cod-liver oil. Nekton E (100 g) and Nek- ton S (100 g) mixed and dissolved in 1 litre water pro- duces an injectable solution containing 5.66 mg/ml VitE (6.8 lU) and 229 [xg/ml VitA (666 lU). Vitamin content of food items (Table 1) was used to determine total daily vitamin E and vitamin A intake. Wet weights for quail, whole chick, and de-yolked chick were 200 g, 40 g, and 30 g, respectively. Over the winter from 1 August-1 February, female peregrines, Sakers {F. cher- rug) , Gyrfalcons {K rusticolus) , Gyrfalcon X Saker hybrids {E rusticolus X E cherrug), a Common Buzzard {Buteo bu- teo) and a Harris’s Hawk {Parabuteo unicinctus) were fed 6 d of the week with up to eight de-yolked, day-old cock- erels and 1 d with rabbit {Oryctolagus cuniculus) . The es- timated content of the daily food intake was 7 mg or 8 4 lU vitamin E and 45 gg or 131 lU vitamin A. New Zea- land Falcons {F. novaezeelandiae) and a Barbary Falcon {E pehgrinoides) were fed a smaller amount of the same diet, but the same concentration of supplements on a body mass basis. The diet was changed on 1 February to three whole day-old chicks and half a quail enhanced with vitamin E, slightly less for the smaller Barbary and New Zealand fal- cons. Each chick was supplemented daily with an inject- able solution of Nekton E and Nekton S containing 5.66 mg VitE/ml. Each chick was injected with 1 ml of the solution. Daily vitamin E intake from 1 February over 6 wk to the start of egg-laying and during egg-laying was calculated as 29 mg or 35 lU/day. Daily vitamin A intake was 1157 fxg or 3363 lU, although from 1 March this was increased to ca. 1650 (xg or 4800 lU vitamin A three March 2002 Composition of Raptor Egg Yolk 35 Table 2. Mean vitamin levels (|xg/g) in raptor egg yolk from infertile, unincubated eggs and selected captive raptor species (SD in parentheses). N is the number of eggs. Species N a-ToCOPHEROL y-TOCOPHEROL Retinol Carotenoids Saker 4 310 (53.8) 10.63 (4.7) 3.1 (1.0) 46.1 (25.8) Peregrine 7 326 (72.0) 9.1 (1.3) 3.8 (0.5) 32.1 (14.3) New Zealand Falcon 5 212 (67.3) 5.94 (0.8) 2.4 (0.4) 48.3 Barbary Falcon 1 247 10.78 3.3 — Common Buzzard 4 443 (75.0) 16.84 (1.6) 2.49 (0.5) 53.8 (4.0) times/wk when a cod-liver oil supplement was added. Weekly vitamin A intake was therefore about 27 000 lU. Eggs were delivered to the Scottish Agricultural Col- lege, Department of Biochemistry and Nutrition where the yolk was separated from albumin. Vitamin A, vitamin E, carotenoid, and fatty acid levels were determined. Eggs were analyzed from eight peregrines, four Sakers, two Gyrfalcons, four New Zealand Falcons, one Harris’s Hawk, one Common Buzzard, one Barbary Falcon and five Gyrfalcon X Saker hybrids. Vitamins A and E were determined by the method of McMurray et al. (1980). To determine carotenoid levels, 2 ml of tissue or yolk homogenate (20% in 0.01 M phosphate buffer, pH 7.4) were mixed with 2 ml ethanol. Hexane (5 ml) was then added and the mixture was shaken vigorously for 5 min. The hexane phase containing the carotenoids was sepa- rated by centrifugation and collected. The extraction was repeated twice more with 5 ml hexane. Hexane extracts were combined and carotenoids were determined from absorption at 446 nm. For lipid extraction, yolk samples were homogenized in an excess of chloroform:methanol (2T, v/v) and extracts of total lipid were prepared. The extracts were subjected to thin layer chromatography on silica gel G using a solvent system of hexanerdiethyl ether; formic acid (80:20:1, v/v) and the band corresponding to phospholipid was eluted from the silica with methanol. The total lipid extract as well as the isolated phospholipid fraction was transmethylated and the fatty acid composi- tion was determined by gas-liquid chromatography (Speake et al. 1999a). The phospholipids are the major lipids found in cell membranes that are transferred from the yolk to the chick during embryogenesis. During the course of routine veterinary investigations, plasma vitamin E and vitamin A levels were also mea- sured in two Saker Falcons and one peregrine during the laying cycle. Results For unincubated eggs from the falcons (Saker, peregrine, Gyrfalcon), mean a-tocopherol level was 320 |Tg/g (A^ — 11, SD = 60.6); mean a-to- copherol 11.24 pg/g (A^ = 11, SD = 3.2); mean retinol 3.55 |Jtg/g (N — 11, SD ^ 0.77); mean ca- rotenoids 36.7 pg/g {N = 9; SD — 17.5; Table 2). Using the peregrine data of seven incubated (Ta- ble 3) and seven non-incubated eggs (Table 2) , in- cubating eggs significantly reduces the levels of a- tocopherol (^2 ~ 3.25, P< 0.01) and retinol (^^2 = 5.71, P< 0.01). Fatty acid composition of the raptor egg yolk in- cluded saturates (16:0 and 18:0), monounsaturates (16:ln-7, 18;ln-9, and 18:ln-7) and polyunsatu- rates (18:2n-6, 20:4n-6, and 22:6n-3). The fatty acid profiles of eggs from different raptors were similar (Table 4, 5) and only buzzard egg yolk composi- tion had distinctive features, including the highest proportion of linoleic acid (18:2n-6) and lowest proportions of 16:ln-7 and 18:ln-7 acids compared to other raptor eggs. As with the total lipid, the phospholipid fraction of buzzard eggs was charac- terized by the highest proportion of 18:2n-6 and the lowest proportions of monounsaturated fatty acids compared to the other raptor species studied. A notable feature of the fatty acid profiles of total lipid and phospholipid was the very high propor- tion of arachidonic acid. Table 3. Mean vitamin levels (|xg/g) in raptor egg yolk from infertile eggs artificially incubated for 14 days (SD in parentheses). A is the number of eggs. Species N a-TOCOPHEROL 7-Tocopherol Retinol Carotenoids Saker 17 290 (61.3) 7.9 (1.9) 2.24 (0.7) 40.0 (8.18) Peregrine 7 261 (56.4) 8.84 (2.2) 2.15 (0.5) 37.0 (9.7) Gyrfalcon 2 291 6.4 3.8 38.8 New Zealand Falcon 2 174 5.07 1.85 38.1 Gyr/ Saker 5 234 (81) 6.2 (3.0) 2.7 (1.1) 37.1 (16.3) Harris’s Hawk 3 193 (25.1) 4.9 (0.2) 3.14 (0.9) 23.8 (7.1) 36 Barton et al,. VoL. 36, No. 1 Table 4. Mean fatty acid composition of the total lipids extracted from egg yolk as a percentage of total extracted fatty acids (mean ± SD; N is number of eggs). Fatty Acids Saker Peregrine Buzzard New Zeaiand FAI.CON Hybrids Harris’s Hawk 16:0 26.54 ± 0.25 27.14 H- 0.16 26.53 0.16 26.58 0.14 28,25 ± 0.49 26.2 16:ln-7 3.60 ± 0.14 3.36 + 0.13 1.86 + 0.12 3.03 -h 0.21 3.73 ± 0.45 2.71 18:0 6.61 ± 0,07 7.07 -h 0.13 7.37 A- 0.17 7.16 A- 0.11 6.45 ± 0.61 14.8 18:ln-9 40.28 ± 0.16 38.95 0.27 37.24 -h 0.62 38.15 -h 0.73 39.42 ± 0.81 9.21 18:ln-7 3.27 ± 0.06 3.07 0.10 2.32 -h 0.03 3.23 ± 0.05 3.16 ± 0.24 2.6 18:2n-6 9.22 ± 0.16 9.40 A- 0.34 13.95 A- 0.62 9.46 + 0.83 9.32 ± 0.74 10.41 20:4n-6 5.72 ± 0.08 4.79 + 0.65 6.23 + 0.08 7.31 A- 0.13 5.39 ± 0.30 5.79 22:6n-3 1.65 ± 0.05 1.40 + 0.17 1.82 0.10 2.22 + 0.07 — 2.73 N 20 14 4 4 5 2 From the three falcons where plasma vitamin levels were measured, mean a-tocopherol was 32.2 p.g/ml and mean vitamin A was 1.02 pg/ml. Discussion Sufficient eggs were available to provide sum- mary statistics, but the number of individuals of each species limited interspecific comparisons. In- cubation of infertile eggs decreases fat-soluble vi- tamin concentrations with vitamin A being the most sensitive to this process. Thus, for future anal- yses it is recommended to use fresh, unincubated eggs. In birds, the level of vitamin E in the egg yolk and embryonic tissues reflects its level in the food (Surai 1999). There are some species-specific dif- ferences in vitamin E accumulation and transfer to the egg yolk with chicken ( Gallus gallus domesticus) being more effective compared to turkey {Meleagris gallopavo) , duck {Anas platyrhynchos) , or goose (An- ser anser) (Surai et al. 1998). Raptor eggs contain a very high a-tocopherol concentration compared to other avian species (Dierenfeld et al. 1989), probably reflecting dietary vitamin E supplemen- tation. It has been suggested that increased vitamin E supplementation may have a positive effect on falcon reproductive performances (Dierenfeld et al. 1989). In chickens, recent studies also show a positive effect of vitamin E on immune system de- velopment and a protective effect in stress condi- tions (Surai 1999). For certain species such as the Gyrfalcon which often show poor immune respons- es particularly in captivity or under stressful situa- tions, adequate dietary vitamin E levels would be essential for good health. Information is available on the concentration of vitamin E in the yolks of a range of avian species in the wild. These currently include the Lesser Black-backed Gull (L. fuscus) (Royle et al. 1999), the Canada Goose {Branta canadensis) (Speake et al. 1999a) and the Emperor Penguin {Aptenodytes forsteri) (Speake et al. 1999b) which all have vita- min E concentrations of about 80 |xg/ g fresh yolk. Table 5. Mean fatty acid composition of the phospholipid fraction extracted from egg yolk as a percentage of total fatty acids (mean ± SD; N is number of eggs) . Fatty Acids Saker Peregrine Buzzard New Zealand Falcon Hybrids Harris’s Hawk 16:0 24.47 0.21 23.39 ± 0.24 26.28 A- 0.11 24.99 0.25 24.24 + 0.73 25.91 16;ln-7 1.08 + 0.05 0.91 ± 0.05 0.47 A- 0.03 0.74 ± 0.06 1.14 A- 0.14 0.78 18:0 18.88 A- 0.23 20.28 ± 0.18 19.39 A- 0.10 19.93 A- 0.16 18.51 A- 1.05 19.89 18:ln-9 19.20 A- 0.21 17.62 ± 0.48 10.99 A- 0.50 14.11 A- 0.59 18.97 A- 1.00 12.58 18:ln-7 2.50 A- 0.04 2.24 ± 0.05 1.86 A- 0.01 2.34 A- 0.04 2.39 A- 0.17 2.10 18:2n-6 6.98 + 0.14 6.96 ± 0.45 12.51 A- 0.29 6.76 + 0.34 7.54 A- 0.54 10.59 20:4n-6 19.10 0.24 20.49 ± 0.26 21.30 A- 0.13 22.78 0.15 19.16 A- 0.91 19.02 22:6n-3 4.31 -h 0.14 4.51 ± 0.18 4.13 + 0.29 4.77 + 0.10 4.63 A- 0.27 6.28 N 20 14 4 4 5 2 March 2002 Composition of Raptor Egg Yolk 37 The raptor eggs of the present study contained vi- tamin E at concentrations averaging 330 fxg/ g and would therefore seem to be very well provisioned with this vitamin. Thus, feeding the female parent with quail and chickens enriched with vitamin E is a successful strategy for achieving high levels of this vitamin in the egg. Despite the small sample size, the mean plasma vitamin E levels of 32.2 |xg/ml are similar to the levels which the Peregrine Fund, Boise, ID U.S.A. measured in their captive peregrine population. They achieved these levels either by feeding quail that had been injected with vitamin E (220 lU/kg quail) or quail which had been raised on a diet containing 220 lU/kg feed. Wild peregrines on mi- gration had plasma vitamin E levels of 26.3 [xg/ml compared to captive peregrines at the Peregrine Fund with 3.4 [xg/ml (Dierenfeld et al. 1989). Be- cause migratory individuals probably have levels lower than breeding individuals, the levels mea- sured in migratory falcons were taken as a mini- mum requirement for a healthy, captive-breeding population (Dierenfeld et al. 1989). From the three falcons in this study that plasma was taken, egg composition was also analyzed and the fertile eggs from these individuals were all viable and pro- duced healthy offspring. There is no reason to as- sume that the infertile egg composition was any different. The concentration of the carotenoids in the rap- tor yolks was similar to that in the first-laid eggs of L. fuscus (Royle et al. 1999) and higher than the values reported for B. canadensis (Speake et al. 1999a) and A. forsteri (Speake et al. 1999b). The vitamin A content of the raptor yolks was slightly less than the value reported for L. fuscus (Royle et al. 1999). The salient feature of the fatty acid profiles of the raptor yolks is the very high proportion of ar- achidonic acid in the phospholipid fraction which IS about five times higher than the values reported for the domestic chicken and for various wild gra- nivorous and herbivorous birds (Speake and Thompson 1999). This may be a consequence of carnivory because the edible parts of many animals are a rich source of arachidonic acid (Phetteplace and Watkins 1989, Li et al. 1998). The proportion of docosahexaenoic acid in the phospholipid of the raptor yolks was similar to the values for eggs of the chicken and many wild birds, but less than the level reported for the piscivorous A. forsteri (Speake and Thompson 1999). In conclusion, the prey items supplemented with vitamin E were a very effective means of fortifying the yolks of raptors with this antioxidant. Vitamin E deficiency reduces hatchability in the quail (Kling and Soares 1980) and has been identified as a cause of late embryo mortality in an estab- lished raptor breeding program (Dierenfeld et al. 1989). Achieving adequate levels of vitamin E, ca- rotenoids, and polyunsaturated fatty acids in the yolk may be essential for the efficient reproduction of birds in captivity. Further studies should focus on analyzing egg composition of wild falcons be- cause such a comparison would give important in- formation for improvement of the falcon diets in captivity. Acknowitdgments We thank the Environmental Research and Wildlife Development Agency, Abu Dhabi and the Scottish Exec- utive Rural Affairs Department for funding this work Thanks also to Nancy Clum, David Bird and an anony- mous reviewer for commenting on the manuscript. Literature Cited Cade, T.J. 1988. The breeding of peregrines and other falcons in captivity: an historical summary. Pages 539- 547 in T.J. Cade, J.H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Falcon populations, their management and recovery. The Peregrine Fund, Boise, ID U.S.A. Calle, P.P., E.S. Dierenfeld, and M.E. Robert. 1989. Se- rum alpha-tocopherol in raptors fed vitamin E sup- plemented diets, f Zoo Wild. Med. 20:62-67. Ci.UM, N.J., M.P. Fitzpatrick, and E.S. Dierenfeld. 1997 Nutrient content of five species of domestic animals commonly fed to captive raptors./. Raptor Res. 31:267- 272. Dierenfeld, E.S., C.E. Sandfort, and W.C. Satterfield 1989. Influence of diet on plasma vitamin E in captive Peregrine Falcons./. Wildl. Manage. 53:160—164. Forbes, N.A. and C.J. Flint. 2000. Raptor nutrition. Ho- neybrook Farm and Lansdown Veterinary Clinic, U K Fox, N.C. and B.I.M. Fox. 1993. The problems in estab- lishing a domestic, self-sustaining international pop- ulation of New Zealand Falcons. Pages 194—200 in M.K. Nicholls and R. Clarke [Eds.], Biology and con- servation of small falcons. The Hawk and Owl Trust, London, U.K. AND N.W.H. Barton. 2000. Nutrition Handbook Bird of Prey Management Series, Carmarthen, U.K. Kling, L.J. and J.H. Soares, Jr. 1980. Vitamin E deficiency in the Japanese quail. Poultry Sci. 59:2352-2354. Li, D., a. Ng, and AJ. Sinclair. 1998. Contribution of meat fat to dietary arachidonic acid. Lipids 33:437- 440. McMurray, C.H., W.J. Blanchflower, and D.A. Rice. 38 Barton et al. VoL. 36, No. 1 1980. Influence of extraction techniques on the de- termination of a-tocopherol in animal feedstuffs. J. Assoc. Off. Anal. Chem. 63:1258-1261. Nichols, D.K. and R.J. Montali. 1987. Vitamin E defi- ciency in captive and wild piscivorous birds. Pages 419—421 in Proc. Annu. Meet. Am. Assoc. Zoo. Vet., Turtle Bay, Hawaii U.S.A. Noble, R.C., B.K. Speake, R.J. McCartney, C.M. Foggin, AND D.C. Deeming. 1996. Yolk lipids and their fatty ac- ids in the wild and captive ostrich. Comp. Biochem. Phy- siol. 113B:753-756. Phetteplace, H.W. AND B.A. Watkins. 1989. Effects of var- ious n-3 lipid sources on fatty acid compositions in chicken tissues./. Food Comp. Anal. 2:104-117. Royle, N.J., RE. SuRAi, R.J. McCartney, and B.K. Speake. 1999. Parental investment and egg yolk lipid compo- sition in gulls. Fund. Ecol. 13:298—306. Speake, B.K., A.M.B. Murray, and R.C. Noble. 1998. Transport and transformations of yolk lipids during development of the avian embryo. Prog. Lipid Res. 37: 1032. and M.B. Thompson. 1999. Comparative aspects of yolk lipid utilization in birds and reptiles. Poult. Avian Biol. Rev. 10:181-211. , P.F. SuRAi, R.C. Noble., J.V. Beer, and N. Wood 1999a. Differences in egg lipid and antioxidant com- position between wild and captive pheasants and geese. Comp. Biochem. Physiol. 124B, 1:101—107. , F. Decrock, P.F. Surai, and R. Groscoias. 1999b. Fatty acid composition of the adipose and yolk lipids of a bird with a marine based diet, the Emperor Pen- guin {Aptenodytes forsteri) . Lipids 34:28S-290. Surai, P.F. 1999. Vitamin E in avian reproduction. Poult Avian Biol. Rev. 10:1—60. , I.A. Ionov, E. Kuchmistova, R.C. Noble, and B.K. Speake. 1998. The relationship between the levels of a-tocopherol and carotenoids in the maternal feed, yolk and neonatal tissues: comparison between the chicken, turkey, duck and goose./. Sci. Food and Agric 76:593-598. and B.K. Speake. 1998. Distribution of carotenoids from the yolk to the tissues of the chick embryo. / Nutr. Biochem. 9:645-651. Weaver, J.D. and TJ. Cade. 1991 Falcon propagation. A manual on captive breeding. The Peregrine Fund, Inc., Boise, ID U.S.A. Received 30 December 2000; accepted 15 September 2001 J. Raptor Res. 36(l):39-44 © 2002 The Raptor Research Foundation, Inc. PRELIMINARY GROUND AND AERIAL SURVEYS EOR ORANGE-BREASTED EALCONS IN CENTRAL AMERICA Russell Thorstrom,^ Richard Watson, Aaron Baker, Serena Ayers AND David L. Anderson The Peregrine Fund, 5668 West Flying Hawk Lane, Boise, ID 83709 U.S.A. Abstract. — Ground and aerial surveys for Orange-breasted Falcons {Falco deiroleucus) were conducted during March-June 1999 in Honduras, and March-May 2000 in El Salvador, Honduras, Nicaragua, Costa Rica, and Panama, all within the likely range of this species. Sixty-six cliffs were checked by ground surveys. No Orange-breasted Falcon was found in Central America. We examined 55 cliffs by fixed-wing aircraft during 24 hr of flying. No Orange-breasted Falcon was seen in El Salvador, Honduras, Nicaragua, Costa Rica, and Panama. In Honduras and Panama, we checked 262 limestone cliffs by helicopter during 47 hr of flying and one Orange-breasted Falcon was seen in a limestone canyon in the Darien Province of Panama, and one sighting of two birds, possibly Orange-breasted Falcons, 11 km south in the same region. The apparent absence of cliff-nesting Orange-breasted Falcons in seemingly suitable habitat in much of Central America is inexplicable and warrants further surveys but indicates that the population in Guatemala and Belize is geographically and genetically isolated from the South American birds. Keywords: Falco deiroleucus; Orange-breasted Falcon', Central America', survey', cliffs. Estudios preliminares terrestres y aereos para los Halicones de Pecho Naranja en Centro America Resumen. — Estudios terrestres y aereos para el halcon de pecho naranja {Falco deiroleucus) fueroll con- ducidos durante marzo-junio de 1999 en Honduras, y marzo-mayo 2000 en El Salvador, Honduras, Nicaragua, Costa Rica, y Panama, todos dentro del rango probable de la especie. Sesenta y seis cornisas fueron revisadas en los estudios terrestres. No se encontro ningun halcon de pecho naranja en Centro America. Examinamos 55 precipios con aeronaves de ala fija durante 24 horas de vuelo. Ningun halcon de pecho naranja fue visto en El Salvador, Honduras, Nicaragua, Costa Rica, y Panama. En Honduras y Panama, revisamos 262 cornisas de piedra caliza en helicoptero durante 47 horas de vuelo y un halcon de pecho naranja fue visto en una canon de piedra caliza en la provincia del Darien en Panama, y un avistamiento de dos aves, posiblemente halcones de pecho naranja, 11 km al sur en la misma region. La aparente ausencia de halcones de pecho naranja anidando en cornisas en un habitat hipoteticarnente adecuado en la mayorfa de Centro America es inexplicable y requiere mayores estudios para indicar que la poblacion en Guatemala y Belice esta aislada geografica y geneticamente de las aves de Sur America. [Traduccion de Cei^r Marquez] The Orange-breasted Falcon {Falco deiroleucus) is a little known and, perhaps, rare falcon (Baker 1998) . Although distributed in Central and South America, the species is local (del Hoyo et al. 1994), sparsely distributed and difficult to detect (Cade 1982), and probably threatened by habitat alter- ation (Jenny 1989, Baker 1998, Baker et al. 2000). The historical distribution of Orange-breasted Fal- cons in Central America is summarized in Baker et al. (2000). Apart from 19 nest sites documented recently in Guatemala and Belize (Jenny 1989, ^E-mail address: rthorstrom@peregrinefund.org Baker 1998, Baker at al. 2000) most reliable re- cords for Central America are more than 20 years old, and none of them involve breeding. Possibly, the species has been extirpated from parts of its former breeding range (Cade 1982) . This species may warrant special status due to the isolation of the known population in Central America from larger populations in South America (Collar and Andrew 1988, Baker et al. 1992, Collar et al. 1994, Baker 1998, Baker et al. 2000). Substantial areas of seemingly suitable habitat for Orange-breasted Falcons exist in Belize, Gua- temala, Honduras, Nicaragua, Costa Rica, and Pan- ama among limestone mountains, yet recent ob- 39 40 Thorstrom et al. VoL. 36, No. 1 servations of this species are almost nonexistent throughout this region. We conducted ground and aerial surveys for Orange-breasted Falcons from El Salvador to Panama to improve knowledge of its present breeding range in Central America. Study Area and Methods The breeding season for Orange-breasted Falcons in Guatemala and Belize begins with courtship in January and February. Eggs are laid in late March and early April, and young fledge mainly in June and July at the begin- ning of the wet season (Baker 1998). Nesting pairs are particularly aggressive toward congeners and other rap- tors during the courtship period and the post-fledging dependence period in June and July. We chose March- June for the aerial surveys to take advantage of the con- spicuousness of breeding pairs tending their nest and young. Ground surveys were conducted from March— May m 1999-2000. Because most of the Pacific slope from El Salvador through Costa Rica is deforested, significantly drier, and has yielded few, if any. Orange-breasted Falcon records historically, we concentrated our survey effort on the Atlantic slope (Fig. 1). Honduras. Ground surveys were conducted in April— May 1999 within a 15-km radius of the Tawahka Indian village of Krausirpe (15°02'N, 84°52W) in the Sierra del Warunta (Anderson 1999) . The small area of ground cov- erage was limited by walking distance and the lack of a trail system in this region. Ground surveys were conducted by D. Anderson ac- companied by 2-3 Tawahka guides to locally-known, lime- stone cliffs (Fig. 1). We searched for falcons using 7 X 42 binoculars and a 16-24 zoom spotting scope from 0700—1230 H and 1530-1730 H. Eighteen cliffs and one large limestone mountain were searched during 1 7 days Cliff surveys consisted of looking for whitewash on cliff faces, falcons perched on cliffs and on trees atop cliffs, flying birds, prey remains at the bases of cliffs, and lis- tening for vocalizing falcons. Aerial surveys were conducted by R. Watson and R Thorstrom in a larger area of the Sierra del Warunta and adjacent Montanas de Golon (ca. 14°50' — 15°05'N, 84°45-85°05'W) (Fig. 1). Three survey flights totaling 8.83 hr were made on 25-26 June 1999 in a Robinson 44 helicopter equipped with a Global Positioning System (GPS) navigational system. Survey routes were chosen at first by visiting the same cliffs surveyed from the ground by D. Anderson, as located by GPS and marked on 1. 50 000 maps of the region. These cliffs were visited dur- March 2002 Orange-breasted Falcon Survey 41 Table 1. Summary by country of suitable cliffs surveyed for Orange-breasted Falcons {Falco deiroleucus) from the ground, by fixed-wing aircraft and helicopters in 1999-2000. Country Number of Suitable Cliffs Surveyed FROM THE Ground Number of Suitable Cliffs Observed During Fixed-wing Aircraft Surveys (hr) Number oe Suitable Cliffs Ob.served During Helicopter Surve\x (hr) El Salvador 12 15 (3) 0 Honduras 41 25 (5.75) 62 (8.5) Nicaragua 0 0 (4) 0 Costa Rica 9 15 (6) 0 Panama 4 0 (5) 200 (38.5) Total 66 55 (23.75) 262 (47) ing the first part of the first flight, after which survey routes were chosen by sight, keying on potentially suit- able nesting habitat (e.g., exposed cliffs with forest be- low) and major topographic features such as valleys and rivers. Our survey route was tracked by GPS and mapped between flights to ensure the entire mountain range was surveyed. Surveys were conducted by flying the helicopter within 20-50 m of the cliff face, at about its highest point so two observers could look down on the cliff face and look into ledges and potholes. We were also able to see the top of the cliff and potential perches, such as snags and overhanging trees. Some cliffs were too tall for an ade- quate view of the entire cliff, so a second, third or even more passes were made at lower elevations in such cases. The helicopter flew slowly past the cliff face while ob- servers scanned the face and all potential perch sites. The helicopter made a second pass or hovered in front of cliffs that needed more time to survey thoroughly. The helicopter provided ample time in front of cliffs, and ma- neuverability to complete visual surveys in narrow valleys and canyons. El Salvador to Costa Rica. Fixed-wing aerial surveys to locate cliffs were conducted by A. Baker and S. Ayers from March-May 2000 (Fig. 1). Areas of potential habitat for aerial surveys were identified from topographic maps where relief might indicate cliffs. We communicated with individuals (both locally and abroad) known to have ex- perience and information in and about the region. In addition, surveys were limited to below 2000 masl (me- ters above sea level) because this species is not known to occur above ca. 1000 masl. Surveys were conducted from a variety of small single-engine planes and, in one case, a twin-engine, fixed-wing plane chartered locally in each country. All flights were conducted at between 200-500 m above the ground at speeds of between 160-250 kmph, suitable altitudes and speeds for surveying the area for cliffs. When cliffs were spotted we circled for a closer look, recorded the locality with a GPS receiver, and noted cliff size, surrounding habitat, and accessibility from the ground. Ground surveys were conducted by A. Baker and S. Ayers after cliffs were located from the air and consisted of spending several hours on the ground in view of and within hearing distance of any falcons on the cliff. Cliffs were selected for ground survey based on accessibility and quality in terms of cliff size, height of cliff above surrounding forest, and extent of surrounding forest. Panama. Between 12-26 April 2000, R. Thorstrom searched from a Robinson 44 helicopter for Orange- breasted Falcons on cliffs, rocks, and escarpments, using 10 X 42 binoculars and a Garmin 12 XT. GPS receiver. We visited five m^yor areas in Panama: Pacific side of the Darien region, the Maje range in central-eastern Panama, the Chagres area of the Atlantic Colon region, Bocas del Toro, Veraguas, Code in the Atlantic western region, and the central range at Volcano Baru on the Chiriqui (Pa- cific side) (Fig. 1). A. Baker and S. Ayers spent 9.25 hr in a helicopter on 22-23 May revisiting the three sites where R. Thorstrom had recorded an Orange-breasted Falcon on 26 April and other unidentified falcon activity. Fixed-wing aircraft flights were made by A. Baker and S. Ayers to survey Chagres National Park and surrounding areas and all of San Bias from Nujagandi to the Colombia border on both sides of the coastal range totaling 5 hr (3 flights including 1 uncompleted flight due to weather) . Ground surveys were made to Madden Dam by R. Thorstrom, and three cliffs were checked near Fortuna Reservoir north of David by A. Baker and S. Ayers (Fig. 1). Results Honduras. Eighteen cliffs and one large lime- stone mountain (Cerro Wampu) were searched from the ground during 17 days between 12 April- 8 May 1999 (Table 1), dates corresponding to the incubation/nestling period in Guatemala and Be- lize (Baker 1998). One additional cliff was searched earlier on 15 March. Cliff size {N — 18) ranged from 30-250 m tall {x = 102 m) and 35- 500 m wide (x = 153 m). Cliff surfaces were bro- ken by crevices, ledges, caves, and epiphytic bro- meliads, providing many potential nest surfaces as well as abundant shade. Cliffs were oriented pre- dominately northward, with eight cliffs facing north, five northwest, one northeast, and four east. 42 Thorstrom et al. VoL. 36, No. 1 We checked an additional 22 cliffs by ground sur- vey in El Boqueron National Park, El Chile Nation- al Park, Sierra de Agalta National Park, Rio Can- grejal, and Celaque National Park. Sixty-two distinct cliff faces were surveyed by he- licopter in the Sierra del Warunta and Montanas de Colon ranges, as well as one isolated mountain in the Rio Platano Biosphere Reserve (en route to the main survey site) called Cerro Chachahuate (Table 1). We observed four Bat Falcons {Falco ru- figularis), but no Orange-breasted Falcons. El Salvador to Costa Rica. We detected 55 ap- parently-suitable cliff faces during 23.75 hr of fixed-wing aerial surveys (Table 1). Of these, we surveyed 41 cliffs from the ground (Table 1). Cliffs that were not surveyed from the ground were un- reachable on foot due to their remote nature, ex- treme topography, and flooded rivers. All cliffs were of igneous origin and ranged from 50-700 m high and 50-1000 m wide (mode = ca. 100 m high X 200 m wide). Although all of these cliffs were located in forested areas, the surrounding forest was patchy, and the patches were mostly small. No suitable cliffs were found in Nicaragua where we conducted fixed-wing surveys (Table 1). No Orange-breasted Falcons were found in El Salva- dor, Honduras, Nicaragua, or Costa Rica. We are confident both that no Orange-breasted Falcons occurred on cliffs in the areas that we surveyed and that we surveyed thoroughly almost all suitable cliffs in these countries. Panama. During 38.5 hr of helicopter survey, we visited over 200 cliffs, rock slopes, and escarpments (Table 1). No suitable cliffs were found in the Cha- gres National Park or San Bias province of Panama where we conducted fixed-wing surveys (Table 1). In the Darien province, we observed one Orange- breasted Falcon on 26 April at 1304 H (07°44'N, 078°05'W) when a large dark-backed, heavy-winged falcon flew five times around the helicopter inside a box-like canyon at 610 masl (Fig. 1). This falcon had slightly larger and thicker wings with slower wing beat than a typical female Bat Falcon. On 13 April at 1030 H, two large dark-backed falcons were flushed off a cliff face by the helicopter (07°39'N, 078°05'W) and descended and disappeared into a ravine, allowing a 2 s observation of the darting falcons. We believe these two disappearing falcons were Orange-breasted Falcons. This site was at 770 m elevation, and the cliff contained one slightly overhung large pothole with a potential scrape and whitewash. We returned to the same site on 26 April at 1100 H but failed to detect the falcons at the site or on the cliff extension to the north (07°40’N, 078°05’W). In the Code province, on 22 April 2000 at 1700 H an unidentified falcon nest was located (08°4TN, 080°40'W). This nest site was a pothole with two cavities at the base of an overhung rock. It contained two nestlings with dark rufous on the sides of neck, blue ceres, and no down on the head. On the first helicopter pass by this site a small dark-backed falcon was flushed from the face and disappeared. On the second pass a larger dark- silhouetted falcon flew above the helicopter and disappeared. On 23 April, we visited this site at 1300 H and found no young present in the pot- hole. One silhouetted heavy-billed falcon was ob- served flying from the helicopter. At 1500 H, we returned to this site after investigating other cliffs in the area; here we located a Bat Falcon on a nest with two eggs on the opposite side of the cliff face formerly containing the unidentified nestlings and at a distance of ca. 200-300 m. During a follow-up helicopter survey in Panama we confirmed the presence of an adult female Orange-breasted Falcon at one site in the Darien. At another Darien cliff, we flushed either a female Bat Falcon or male Orange-breasted Falcon from a ridge-top perch. At the Code site, we observed either a female Bat or male Orange-breasted Fal- con flying away from the helicopter. During 38.5 hr of helicopter time, we recorded eight observa- tions of Bat Falcons perched in trees or evading the helicopter near cliffs and two nests; one in the Maje mountains and the other in the Code region. Discussion Orange-breasted Falcons formerly occurred in a broad range from southern Mexico south through Central America and into South America, where the species was found east of the Andes as far south as northern Argentina (del Hoyo et al. 1994). Re- cords of the species, however, are sparse through- out its range. The most concentrated records, in- cluding the only breeding records from Central America, come from Guatemala and Belize (Baker 1998). Baker (1998) studied 13 breeding pairs in Belize and six pairs in Guatemala from 1991-97, of which the closest neighboring pairs were 1.7 km apart. If our survey in Central America was in sim- ilarly suitable habitat, we estimate there would be sufficient cliffs, spaced far enough apart, to sup- port numerous breeding pairs. In Nicaragua, no March 2002 Orange-breasted Fai.con Survey 43 cliffs were found in forested habitat and in El Sal- vador, no cliffs were associated with primary for- ests. The fact that we found no Orange-breasted Falcons in El Salvador, Honduras, Nicaragua, or Costa Rica suggests that either the habitat is not suitable, they are not using cliffs as nest sites, or we overlooked this species’ presence. The only Orange-breasted Falcons detected during these surveys were in the Darien province of Panama. This area contains the northernmost foothills of the Andes of South America. The presence of these falcons in the Darien suggests that these birds are at the northern limit of the South Amer- ican population. Panama appears to have many cliffs in the Code and Darien regions surrounded by intact primary forest, perhaps as many as the Belize and Peten, Guatemala region. It seems very likely that Panama contains a breeding population of Orange-breast- ed Falcons. The presence of this species in Belize, Guatemala, and the Darien province of Panama, where an abundance of cliffs in conjunction with large tracts of forest remains, lends support to this hypothesis. One pair of palm-nesting Orange-breasted Fal- cons was recorded in Peten, Guatemala (Baker 1998) and equally suitable palm trees and other canopy emergent trees potentially suitable as nest- ing sites occur throughout Central America. Due to the rarity of this behavior we suspect that it may result from the size of the population in the Gua- temala and Belize region relative to a limited num- ber of suitable nesting cliffs, leading to the occa- sional use of trees for nesting sites. The seeming absence of cliff-nesting Orange- breasted Falcons from El Salvador, Nicaragua, and Costa Rica is not surprising because of the low mu- seum records of the species from the region. A very low abundance of suitable cliffs in Costa Rica and El Salvador, and no cliffs in Nicaragua com- pounded by extensive deforestation and dry areas, especially in El Salvador and most of Honduras, is a likely explanation. The lack of nesting Orange- breasted Falcons in parts of Honduras is surprising, because there are many limestone cliffs with an ex- tensive tract of primary forest. Detection of four Bat Falcons during 8.5 hr of helicopter surveys pro- vides evidence that our observation methods should have revealed the Orange-breasted Falcons if they were present. We offer the following speculative explanations for the apparent absence of Orange-breasted Fal- cons from most of Central America, with the un- derstanding that they might influence further study. Populations of this falcon may have been negatively affected in the period before DDT was banned in the United States (prior to 1972) as a result of feeding on contaminated migratory and resident birds, but this is not consistent with the presence of the Belize/Guatemala population. DDT and other organochlorines were used for years in developing countries after being banned in the U.S., particularly for pest control on cotton crops. Cotton was grown extensively along the Pa- cific seaboard of Central America from southern Mexico through Guatemala, El Salvador, Hondu- ras, Nicaragua, and into Costa Rica, expanding rap- idly in the 1950s and 1960s, and reaching peak pro- duction in 1978 (Murray 1994). Because pesticide contamination has never been associated with a de- cline of Orange-breasted Falcons and DDT has been used primarily on the Pacific coastal side, however, we suggest it is unlikely that this was the main cause for extirpating this species from most of Central America. Alternatively, the lack of detection of nesting fal- cons may be a result of our focusing only on cliffs in areas where Orange-breasted Falcons may be nesting in trees. However, while Orange-breasted Falcons are sometimes difficult to locate when nesting in trees, when nesting they are usually near their nests and quite vocal making detection a bit easier. In 7 yr, staff of The Peregrine Fund’s Maya Project located 1 tree nest in Guatemala and Belize (Baker 1998). In Ecuador, searches by boat, air, and truck during five months resnlted in observa- tions of 15 Orange-breasted Falcons and four nests in emergent trees (Cade 1982). Furthermore, Orange-hreasted Falcons may have been more dif- ficult to detect from the ground in Central Amer- ica if April and May correspond to the incubation period, as in Guatemala (Baker 1998), which could have contributed to our lack of detections. Orange-breasted Falcons are regarded as threat- ened in Central America because of their small population size and the probability that breeding populations in Guatemala and Belize are geneti- cally isolated from the South American population (Collar and Andrew 1988, Collar et al. 1994, Baker 1998). The lack of detection of Orange-breasted Falcons in apparently all suitable habitat in Hon- duras and Costa Rica is an important observation that should stimulate further study. If the species is sensitive to habitat alteration, as suggested by 44 Thorstrom et al. VoL. 36, No. 1 some authors (Jenny and Cade 1986, Jenny 1989, Baker 1998), it could be a useful indicator of na- tive-plant community alteration in the remaining lowland tropical rain forests of Central America. We recommend additional studies to understand nest site selection in areas where this bird may nest in emergent trees as in Ecuador and possibly Brazil (Whittaker 1996) or in palms as in Guatemala. Re- sults would be useful in implementing surveys for tree-nesting Orange-breasted Falcons in the Sierra del Warunta range of Honduras, La Mosquitia re- gion of Honduras and Nicaragua, and the Code, Bocas del Toro, Veraguas, and Chagres regions of Panama. Acknowledgments We thank the Wolf Creek Foundation for support for this study and Robert Berry to whom we express our sin- cere appreciation. We thank Lloyd Kiff, Tom Cade, Steve Sherrod, Jim Bednarz, and Jim Enderson for their com- ments on drafts of the manuscript. Thanks to Edwin Ur- riola for enduring and participating in the long helicop- ter flights. Special thanks to Thomas Jakits, pilot, and Captain Thomas Exenberger, owner of Helipan Corpo- ration. Literature Cited Anderson, D.L. 1999. Tawahka Project, Honduras. 1999 Field Season Report. The Peregrine Fund, Boise, ID U.S.A. Baker, A.J. 1998. Status and breeding biology, ecology, and behavior of the Orange-breasted Falcon (Falco dei- roleucus) in Guatemala and Belize. M.S. thesis, Brig- ham Young Univ., Provo, UT U.S.A. ,J.P. Jenny, and D.F. Whitacre. 1992. Pages 217- 224 in D.F. Whitacre and R.K. Thorstrom [Eds.], Orange-breasted Falcon reproduction, density, and behavior in Guatemala and Belize. Progress Report V, 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 Peregrine Fund, Inc., Boise, ID U.S.A. , Whitacre, D., O.A. Aguirre-Barrera, and G White. 2000. The Orange-breasted Falcon Falco dei- roleucus in Mesoamerica: a vulnerable, disjunct popu- lation? Bird Cons. Int. 10:29-40. Cade, T.J. 1982. The falcons of the world. Comstock/ Cornell Univ. Press, Ithaca, NYU.S.A. Collar, NJ. and P. Andrew. 1988. Birds to watch. The ICBP World Check-list of Threatened Birds. Interna- tional Council for Bird Preservation, Cambridge, U.K. , M.J. Crosby, and A.J. Statterseield. 1994. Birds to watch 2: world list of threatened birds. Birdlife Conservation Series No. 4. Page Bros. (Norwich) Ltd U.K. del Hoyo, J., a. Elliott, and J. Sargatal. 1994. Hand- book of birds of the world. Vol. 2. New World vultures to guineafowl. Lynx Edicions, Barcelona, Spain. Jenny, J.P. 1989. Observations of the Orange-breasted Fal- con [Falco deiroleucus) in the northern Peten from 1979 through 1989. Pages 93-97 in W.A. Burnham, J.P. Jenny, and C.W. Turley [Eds.], Progress Report II, 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 Peregrine Fund, Inc., Boise, ID U.S.A. and TJ. Cade. 1986. Observations on the biology of the Orange-breasted Falcon Falco deiroleucus. Pages 119—124 in R.D. Chancellor and B.-U. Meyburg [Eds.], Birds of Prey Bulletin No. 3. World Working Group on Birds of Prey and Owls, Berlin, Germany. Murray, D.L. 1994. Cultivating crisis: the human cost of pesticides in Latin America. Univ. of Texas Press, Aus- tin, TX U.S.A. Whittaker, A. 1996. First records of the Orange-breasted Falcon Falco deiroleucus in central Amazonian Brazil, with short behavioural notes. Cotinga 6:65-68. Received 1 March 2001; accepted 19 November 2001. J. Raptor Res. 36(1):45— 50 © 2002 The Raptor Research Foundation, Inc. DISPERSING VULTURE ROOSTS ON COMMUNICATION TOWERS Michael L. Avery\ John S. Humphrey, Eric A. Tillman and Kimberly O. Phares USDA/APHIS, Wildlife Services, National Wildlife Research Center, Florida Field Station, 2820 East University Avenue, Gainesville, FL 32641 U.S.A. Jane E. Hatcher USDA/APHIS, Wildlife Services, Florida State Director’s Office, 2820 East University Avenue, Gainesville, FL 32641 U.S.A. Abstract. — Communication towers provide attractive roost sites for Black ( Coragyps atratus) and Turkey vultures {Cathartes aura). The birds’ roosting activity creates problems, however, for tower operators, nearby businesses, and adjacent homeowners. To alleviate these problems, at six sites in northern Florida we evaluated the effectiveness of suspending vulture carcasses or taxidermic effigies from towers to disperse vulture roosts. In each case, vulture numbers decreased immediately after installation of the stimulus, and roosts declined 93-100% within nine days. The effect was independent of the composition of the roost and occurred regardless of which vulture species was used as the carcass or effigy. At one site, the roost was substantially reduced using a commercial plastic goose decoy painted to resemble a Turkey Vulture. At three sites, the deterrent effect persisted up to 5 mo even after the carcass or effigy was removed from the tower. Hanging a vulture carcass, taxidermic effigy, or even an artificial decoy, from a tower creates an unfavorable roosting environment for vultures and offers a simple, effective means to manage problem-roost situations. Key Words: Cathartes aura; communication tower, Coragyps atratus; effigy, roost dispersal, vultures. Dispersion de perchas para gallinazos en torres de comunicacion Resumen. — Las torres de comunicacion proveen unos sitios de percha atractivos para los gallinazos comunes {Coragyps atratus) y los de cabeza roja {Cathartes aura). El uso de perchas de las aves crea problemas para los operadores de las torres, negocios cercanos y casas familiares adyacentes. Para aliviar estos problemas, en seis sitios del norte de la Florida evaluamos la efectividad de suspender esqueletos de gallinazos o figuras disecadas e las torres para dispersar las perchas de los gallinazos. En cada caso, el numero de gallinazos disminuyo inmediatamente despues de la instalacion del estimulo, y las perchas declinaron 93-100% en nueve dfas. El efecto fue independiente de la composicion de la percha y ocurrio sin importar cual especie de gallinazo fuera usada como el esqueleto de la figura. En un sitio, la percha fue sustancialmente reducida usando un sehuelo comercial plastico de ganso pintado para simular un gallinazo negro. En tres sitios, el efecto disuasivo persistio por mas de cinco meses aun despues de que los esqueletos o las figuras fueran rempvidas de la torre. Colocar un esqueleto de gallinazo, una figura disecada, o aun un sehuelo artificial en una torre, crea un efecto desfavorable para que los gallinazos puedan perchar y ofrece un medio simple y efectivo para manejar situaciones prob- lematicas con las perchas. [Traduccion de Cesar Marquez] Recent estimates suggest that in the United States there are nearly 45 000 communication and broadcast towers taller than 61 m, and industry projections suggest that 10000 more are likely to be built in the next decade (Evans and Mannville 2000, Tollefson 2001). Vulture populations also are increasing. Analyses of Breeding Bird Survey data ^ E-mail address: Michael.L.Avery@aphis.usda.gov (1980-99) indicate that Black Vultures {Coragyps atratus) are increasing at an annual rate of 2.9% in Florida and 2.4% nationwide, and Turkey Vultures {Cathartes aura) are increasing annually by 1.2% in Florida and 1.8% throughout the country (Sauer et al. 2000). Vultures sometimes roost on communication and broadcast towers and similar structures. Stolen (1996) recorded as many as 130 vultures roosting on a microwave tower in east-central Florida. In 45 46 Avery ex al. Vox. 36, No. 1 Table 1. Various types of stimuli were evaluated as means to disperse vulture roosts on towers at six sites in northern Florida. Tower Site Height (m) Vultures at Site Initial No. % BLVU Stimulus'^ Date Instali.ed Macclenny A (F)*^ 105 100-200 80 BLVU carcass 25 Sep 2000 Macclenny (G) 85 25-40 65 TUVU carcass 28 Nov 2000 (F) 31 15-25 65 BLVU effigy 28 Nov 2000 Waldo (F) 83 140-170 90 BLVU effigy 30 Oct 2000 Durbin (F) 78 40-60 85 BLVU effigy 27 Nov 2000 Jacksonville (F) 45 100-150 40 BLVU effigy 29 Jan 2001 Niceville (G) 92 50-150 25 Goose decoy 27 Feb 2001 TUVU effigy 15 Mar 2001 ^ BLVU — Black Vulture; TUVU — Turkey Vulture. (F) — free-standing; (G) — guyed. Two towers, ca. 45 m apart. Texas, Buckley (1998) observed 4—136 Black and Turkey vultures roosting on power transmission line support structures. Kirk and Mossman (1998) state that Turkey Vultures may roost on commu- nication towers “especially on warm, still nights” but provide no documentation of this activity. Defecations by roosting vultures interfere with the operation of expensive equipment and create unsafe and unpleasant conditions for workers who climb towers to service and install equipment. In addition, businesses and homeowners adjacent to a vulture roost site are adversely affected by vulture droppings and the unpleasant odor that results. Given current trends in vulture populations and tower construction, it is probable that roosting on towers by vultures will become more widespread, and the need for effective, nonlethal solutions to this problem will increase as well. Pyrotechnics and other noisemakers are disruptive to neighboring businesses and homeowners and provide short- term relief at best. Physical and chemical deter- rents applied to perching substrates would be im- practical because of the expansive perching area available on a tower. Furthermore, they would in- terfere with operation and maintenance activities on the tower. Visual deterrents such as reflecting tape and scare-eye balloons seemed impractical and probably ineffective based on previous evalu- ations with other species (Tobin et al. 1988, Tipton et al. 1989). One method that appears to have some promise is hanging a vulture carcass or effigy in the roost. This technique was suggested on a fact sheet on vulture management in Virginia, but no support- ing data were presented (M. Lowney pers. comm.). Trials in Ohio demonstrated that Turkey Vultures in a tree roost and on an abandoned tower, dis- persed when freeze-dried Turkey Vulture effigies were suspended at each site (T. Seamans pers. comm.). These promising results with Turkey Vul- tures have not been duplicated for Black Vultures, however. The only Black Vulture effigy trial that we are aware of comes from a newspaper article (Tam- pa Tribune-Times, 20 February 1994). On a Virgin- ia farm where Black Vultures reportedly attacked and killed several ducks, a Black Vulture carcass suspended near a farm pond deterred the vultures for “about two hours.” Our principal objective in this study was to de- termine whether whole carcasses or taxidermic ef- figies would disperse Black Vultures from roosts on towers. Secondarily, we examined responses of Black Vultures to Turkey Vulture effigies, and vice versa. In addition we conducted a limited trial to evaluate a plastic goose decoy as a vulture dispersal agent. Methods In northern Florida, we conducted trials at six sites (Table 1). The towers were not selected at random but were determined by requests for assistance from the tow- er owners. There was considerable variability among the structures (Fig. 1). Black Vultures were predominant roosting species at most sites (Table 1). We monitored vulture numbers at each site 3 d before and 9 d after installation of the vulture carcass, taxider- mic effigy, or goose decoy. At a given site, we counted roosting birds at the same time each day, either early in the morning (0630-0830 H) or late in the afternoon (1630-1830 H). At four sites, we counted all of the birds March 2002 Vulture Roost Dispersal 47 Figure 1. Towers used as roost sites by vultures in northern Florida: A — Macclenny A; B — Macclenny B; C — ^Waldo; D — -Jacksonville; E — Durbin; F — Niceville. on the tower at the start of the daily observation period and then recorded all vultures that arrived or departed during the next 2 hr. We then derived a maximum daily vulture count for each of the four sites. At the Macclenny B site, we counted vultures once in the morning (0800- 0830 H), and at Durbin, cooperators counted all the vul- tures they could see on the tower each day at 1700 H. Cooperators were asked to be consistent and to count all vultures roosting on the tower at the same time each day. Total numbers of vultures are reported without regard to species. Professional climbers installed the carcass, effigy, or de- coy so that it hung freely and was able to swing and twist m the wind without becoming entangled in the structure. Installation always occurred at midday to avoid any con- tact with vultures using the site. We secured the ends of a short leather strap to the legs of the carcass, effigy, or decoy and clipped a fishing tackle swivel to the strap. The other end of the swivel was tied to a length of coated twine 1.5 m— 3.5 m long, and then secured to the tower at the specified location by whatever means the climber felt appropriate. At two sites, the climbers installed pulley systems so the stimulus could be recovered and replaced or redeployed if necessary. The taxidermist prepared the vulture effigies so that one wing extended beyond the head and the other wing was folded. The plastic goose decoy was painted to resemble a Turkey Vulture and had the wings outstretched perpendicular to the body. For analysis, we grouped data into one 3-d pretreat- ment period and three 3-d posttreatment periods. For each study site, we calculated a mean vulture count for each of the four periods. We analyzed these data using Friedman’s test (Steel and Torrie 1980) to compare the number of vultures recorded during pretreatment with 48 Avery et al. VoL. 36, No. 1 those present after the stimulus was installed. The four time periods were treatments and the six study sites were blocks. At some sites, we deviated from the general procedures to collect additional information not included in the data analysis. At Macclenny B, 3 km north of Macclenny A, there is a guyed 85-m communications tower (Fig. IB) as well as a free-standing 31-m Doppler radar tower 45 m away. On 28 November, a Turkey Vulture carcass was in- stalled on the gfuyed tower, ca. 75 m above the ground, and a taxidermic Black Vulture was installed ca. 25 m up on the Doppler tower. We used different stimuli on each tower because we did not know which would be more effective or if just one would suffice for both structures. Through March 2001, both Macclenny sites were checked for vultures at 0800-0830 H, an average of 3 d weekly. On 13 November 2000, to see if vultures would reoc- cupy the structure, we removed the effigy on the Waldo tower and counted birds there on 14-17 and 20-23 No- vember. Thereafter, irregular visits to the tower were made for 2 mo to document any additional vulture activ- ity. The guyed structure at Niceville consists of two vertical masts, 92 m and 73 m, connected by three horizontal crosspieces. On the morning of 27 February 2001, a climber installed a plastic Canada Goose {Branta canaden- sis) decoy that we painted to resemble a Turkey Vulture. The decoy was suspended from the uppermost horizontal crosspiece, ca. 70 m above the ground. Because there were still vultures on the tower, on 15 March 2001 we replaced the decoy with a taxidermic Turkey Vulture ef- figy- Results Vulture Dispersal. During pretreatment, the mean daily number of vultures at the six sites var- ied from 29-157 {x = 89, SE = 21). After instal- lation of the vulture carcass, taxidermic effigy, or goose decoy, vulture numbers declined markedly (P = 0.002, Friedman’s test, S = 15.10, 3 df). Ex- cluding the Niceville site, numbers of roosting vul- tures were reduced 93-100% by day 12 (Fig. 2). At Niceville, the presence of a goose decoy caused vul- ture numbers to decline, although not as dramat- ically as with a vulture carcass or effigy. Replace- ment of the goose decoy with a Turkey Vulture effigy (15 March 2001) then dispersed the residual roosting population. Nine days into the treatment period at Macclen- ny A, there were no vultures on the tower. The carcass deteriorated over time, and by mid-Novem- ber 2000 all that remained attached to the tower were the legs and back. Nevertheless, through March 2001, vultures did not reoccupy the tower. We obtained the same long-term response at Mac- clenny B, despite the fact that the Turkey Vulture carcass installed on the guyed tower fell off on day 175- Figure 2. Roosting vultures were counted at six tower sites during 12-day study periods. Following the count on day 3, a vulture carcass was installed on the Macclenny A and Macclenny B towers, a plastic goose decoy painted to resemble a vulture was installed at Niceville, and a taxi- dermic vulture effigy was installed at the other sites. 7. The Black Vulture taxidermic effigy on the near- by Doppler unit remained in place throughout the study. At Waldo, the average maximum daily count on the tower prior to installation of the vulture effigy was 157 (SE = 9, V = 3), compared to 12 vultures (SE = 7, N= 9) with the effigy in place, and 9 (SE = 5, N = S) after it was taken dovm. Even with the effigy no longer in place, regular checks of the tow- er in the morning and afternoon revealed no vul- tures through March 2001. Behavioral Observations. Vultures that encoun- tered an effigy or carcass hanging from a tower typically circled the structure and flew close to the effigy or carcass. Most birds did not land on the tower. Among vultures that did, there was no ob- vious trend or preference to be above, below, or at the same level as the stimulus. Many that landed on a tower peered at the carcass or effigy for up to several minutes and then departed. Those that stayed on the tower preened and interacted with other vultures in what appeared to us to be a nor- mal manner. Vultures that perched on the tower did not demonstrate overt avoidance of the effigy or carcass. Many perched very close with no obvi- ous concern. Several times we noted that all the vultures on the tower flew up and departed the area en masse. On at least one occasion this was due to the arrival of a Red-tailed Hawk {Buteo ja- March 2002 Vulture Roost Dispersal 49 maicensis ) , but usually the reason for a mass depar- ture was not apparent. Discussion Not every available tower is occupied by vultures. It is not known what features of a tower attract roosting vultures. We noted many unoccupied tow- ers of seemingly identical design as those in this study. Site microclimate is likely an important fac- tor in vultures’ choice of a roost site (Thompson et al. 1990). Birds roosting on a tower are not pro- tected by branches and surrounding vegetation like they would be in a tree roost. Exposure to the ambient conditions is possibly offset by birds’ abil- ity to roost closer together on a tower than they could in a tree roost with consequent thermoreg- ulatory benefits (Buckley 1998). Because towers are higher than surrounding trees, vultures prob- ably can enter and depart the roost more easily. Furthermore, wind striking the structure might create updrafts, called obstruction currents, that facilitate the birds’ flight near the tower (Thomp- son et al. 1990). The towers we studied are near heavily-traveled roads or highways. Roosting close to roads could be advantageous for vultures be- cause of thermals generated from the pavement and the availability of road kills (Thompson et al. 1990). Vulture roosts can form in response to tempo- rary availability of local food resources (Sweeney and Fraser 1986, Coleman and Fraser 1989). Among these study sites, the Waldo tower is within 2 km of a small pig farm frequented by Black Vul- tures that sometimes preyed upon newborn piglets. The owner of the farm informed us that the num- ber of vultures at his farm declined substantially after we installed the effigy and dispersed birds at the Waldo tower. This observation supports the no- tion that local food availability can be a determin- ing factor in the formation of vulture roosts on towers. From the consistent responses that we recorded, it is obvious that the presence of a dead vulture hanging by its feet makes a tower less suitable as a vulture roost site. In every trial, there was imme- diate reduction in numbers of roosting birds, fol- lowed soon by abandonment of the roost site, re- gardless of the species composition of the roost and regardless of the species of vulture carcass or effigy. Even the installation of a Canada Goose de- coy caused substantial reduction, although not abandonment, at one site. It is not clear what features of the effigies are offensive to the vultures. Taste, tactile, and aural cues can be ruled out because vultures never con- tacted the effigies and the effigies produced no sounds. Conceivably, the odor of a decaying vul- ture carcass could be perceived by other vultures as a signal to stay away from the area. However, we observed similar responses with intact carcasses, taxidermic effigies, and a plastic decoy. The odors produced by these stimuli are, no doubt, sufficient- ly distinct for vultures to discriminate them. Thus, at this time, we think it unlikely that odor cues are important. Rather, we feel that visual cues are pre- dominant. This is supported by observations of many perched vultures peering at the effigy hang- ing from the tower and by vultures circling the tow- er, flying close to the effigy, and then departing. The more challenging task is determining what vi- sual attributes are most salient to the vultures. Pos- sibilities include size, shape, color, orientation, movement, and height on tower. In this study, we did not experiment or manipulate these variables because our goal was to solve the problems of our cooperators, not to isolate the factors that might be essential to the effectiveness of this roost dis- persal technique. Particularly noteworthy was the degree to which the repellent effect of the effigy or carcass persist- ed after the stimulus was removed. Months after the carcass at the Macclenny A tower had rotted away, no vultures occupied the tower. Similarly, the carcass installed on the Macclenny B tower fell off after 4 days yet vultures continued to avoid the structure. At this site, the presence of a Black Vul- ture effigy on the 31-m Doppler tower might have contributed to the absence of vultures on the taller tower 45 m away. Finally, at Waldo, we intentionally removed the Black Vulture effigy, and regular monitoring disclosed no reoccupation of the tower through March 2001, over 4 mo later. We did not intentionally remove effigies at other sites because of commitments to our cooperators, but it is cer- tainly of interest to determine the relationship be- tween length of vultures’ exposure to the stimulus and the duration of their avoidance responses. Our findings would have been strengthened by the inclusion of unmanipulated vulture roosts as controls. However, we feel that pretreatment ob- servations at each site provide sufficient evidence that the roosts would have persisted had we not intervened. Vulture roosts can be ephemeral (Sweeney and Fraser 1986, Coleman and Fraser 50 Avery et al. VoL. 36, No. 1 1989), but it is unlikely that each of the roosts we studied happened to disperse coincidentally with the installation of the carcass, effigy, or decoy. Management Implications. Suspending a vulture effigy or carcass in a tower appears to be a quick, effective means to rid the structure of roosting vul- tures. Once the stimulus is properly installed, the only problem likely to be encountered is possible entanglement of the support line with the struc- ture. This can be avoided by keeping the support line to an appropriately short length. The extent to which the effigy/carcass approach to manage- ment of nuisance vulture roosts can be extended to other types of roosts remains to be determined. Initial trials that we have conducted in vulture tree roosts affecting residential neighborhoods have been promising. In each case the roost has dis- persed, although the response by the vultures was not as rapid as we observed in the tower roosts (M. Avery unpubl. data). There are constraints to the general use of a vul- ture carcass or taxidermic effigy. Both species of vultures are protected by Federal laws and it is un- lawful to possess them without a permit from the U.S. Fish and Wildlife Service. Therefore, this tech- nique can only be used under supervision of the appropriate authorities. Also, the hanging of a vul- ture carcass or taxidermic effigy could be distaste- ful to the public. If this technique is used in areas of high visibility, then it might be prudent to con- tact local conservation or birding groups so that the carcass or effigy is not mistaken for a bird that accidentally became entangled in the tower. Final- ly, prolonged exposure to the weather deteriorates the carcass or effigy. We feel the development of an effective, dura- ble, readily available alternative is essential to the widespread use of this vulture management meth- od. The trial we conducted at Niceville with the Canada Goose decoy was an encouraging step in this direction. The decoy cost about $25.00 (U.S.), and we made only minor changes in its appear- ance, yet vulture use of the tower was reduced 60% after the decoy was installed. This suggests that suc- cessful roost dispersal can be accomplished without the use of actual carcasses or taxidermic effigies. The focus of future field trials will be the evalua- tion of various commercial decoy alternatives. Acknowledgments We are grateful for the cooperation and support from SpectraSite, Inc. (S. Sirignano), Pinnacle Towers, Inc. (G. McMillan), and AT&T Wireless, Inc. (L. Gorodetzer) . J Weaver and B. Owens assisted with data collection. B Millsap reviewed the manuscript. Taxidermic effigies were prepared by T. Gilliard. Literature Cited Buckley, N.J. 1998. Interspecific competition between vultures for preferred roost positions. Wilson Bull. 110. 122-125. Coleman, J.S. and J.D. Fraser. 1989. Habitat use and home ranges of Black and Turkey vultures. J. Wildl Manage. 53:782-792- Evans, W.R., and A.M. Manville, II (Eds.). 2000. Avian mortality at communication towers. Transcripts of Proceedings of the Workshop on Avian Mortality at Communication Towers, 11 August 1999, Cornell Univ., Ithaca, NY U.S.A. http://www.towerkill.com and http:/ / migratorybirds.fws.gov/issues/ towers/ agenda. Kirk, D.A. and M.J. Mossman. 1998. Turkey Vulture {Ca- thartes aura). In A.. Poole and F. Gill [Eds.], The Birds of North America, No. 339. The Birds of North Amer- ica, Inc. Philadelphia, PA U.S.A. Sauer, J.R., J.E. Hines, I. Thomas, J. Fallon, and G Gough. 2000. The North American Breeding Bird Survey, Results and Analysis 1966-1999. Version 98.1, US Geologic Survey Patuxent Wildlife Research Cen- ter, Laurel, MD U.S.A. http://www.mbr-pwrc.usgs/ gov/bbs. Steel, R.G.D. and J.H. Torrie. 1980. Principles and Pro- cedures of Statistics, 2nd Ed. McGraw-Hill Book Co New York, NY U.S.A. Stolen, E.D. 1996. Roosting behavior and foraging ecol- ogy of Black Vultures in central Florida. M.S. thesis, Univ. of Central Florida, Orlando, FL U.S.A. Sweeney, T.M. and J.D. Fraser. 1986. Vulture roost dy- namics and monitoring techniques in southwest Vir- ginia. Wildl. Soc. Bull. 14:49-54. Thompson, W.L., R.H. Yahner, and G.L. Storm. 1990 Winter use and habitat characteristics of vulture com- munal roosts. J. Wildl. Manage. 54:77-83. Tipton, A.R., J.H. Rappole, D.B. Johnson, J. Hobbs, P. Schulz, S.L. Beasom, and J. Palacios. 1989. Use of monofilament line, reflective tape, beach-balls, and pyrotechnics for controlling grackle damage to citrus. Pages 126-128 in Ninth Great Plains Wildlife Damage Control Workshop Proceedings. USDA Forest Service General Technical Report RM-171, Fort Collins, CO U.S.A. Tobin, M.E., P.P. Woronecki, R.A. Dolbeer, and R.L. Bruggers. 1988. Reflecting tape fails to protect rip- ening blueberries from bird damage, Wildl. Soc. Bull 16:300-303. Tollefson, C. 2001. Reducing fatal bird collisions with nation’s communications towers. People, Land & Water 8(2):17. Received 29 May 2001; accepted 1 October 2001 Associate Editor: Clint Boal /. Raptor Res. 36(1) :5 1-57 © 2002 The Raptor Research Foundation, Inc. RAPTOR ABUNDANCE AND HABITAT USE IN A HIGHLY-DISTURBED-FOREST LANDSCAPE IN WESTERN UGANDA Nathaniel E. Seavy^ and Christine K. Apodaca Department of Zoology, 223 Bartram Hall, RO. Box 118523, University of Florida, Gainesville, FL 32611-8525 U.S.A. Abstract. — ^We conducted roadside raptor surveys in an area of western Uganda that included undis- turbed forest, agricultural mosaics, and tea plantations. Between September 1997-April 1998, we sur- veyed three transects (51 km of roadway) twice each month, once during the morning and again in the afternoon. During these surveys, we detected 14 falconiform species and 77 individuals. We detected significantly more raptors during morning than afternoon surveys. Wahlberg’s Eagles (Aquila wahlbergi) and Long-crested Eagles (Lophaetus occipitalis) were observed most freqnently and were sighted consis- tently throughout the study. Most raptors observed were resident species, whereas Palearctic migrants comprised less than 25% of all raptors, noteworthy considering their abundance in other African re- gions. We did not detect an equal number of raptors in all habitat types. Agriculture mosaics accounted for 61% of the habitat we surveyed and 75% of all raptor detections were in these habitats. In contrast, tea plantations were 14% of the area surveyed, but only 3% of all raptors were detected there. Based on these results, we suggest that tea plantations may be suboptimal habitat for larger, open-habitat raptors. We did not detect large, forest-dwelling eagles outside of large areas of undisturbed forest, which are probably critical to their persistence in these landscapes. Keywords: Uganda', roadside survey; conservation; habitat use, deforestation; Aquila wahlbergi; Wahlberg’s Eagle, Lophaetus occipitalis; Long-crested Eagle. Abundancia de repaces y uso de habitat en un paisaje de bosque altaraente disturbado en el oeste de Uganda Resumen. — Hicimos estudios de rapaces alrededor de las carreteras en un area del oeste de Uganda que incluia bosque sin intervencion, mosaicos agricolas, plantaciones de te. Entre septiembre de 1997— abril de 1998, estudiamos tres transeptos (51 km de carreteras) dos veces cada mes, una vez en la mahana y otra en la tarde. Durante estos estudios detectamos 14 especies falconiformes y 77 individuos. Detectamos signihcativamente mas especies durante los estudios de la manana que en los de la tarde. Las aguilas de Wahlberg {Aquila wahlbergi) y las aguilas de cresta alargada {Lophaetus occipitalis) fueron observadas mas frecuentemente y fueron avistadas consistentemente a lo largo del estudio. La mayoria de rapaces observadas fueron especies residentes, mientras que las migratorias palearticas comprendie- ron menos del 25% de todas las rapaces, esto es digno de mencionarlo considerando su abundancia en otras regiones africanas. No detectamos un numero igual de rapaces en todos los tipos de habitats. Los mosaicos agricolas dieron cuenta de 61% del habitat que estudiamos y 75% de todas las detecciones de rapaces estuvieron en esos habitats. En contraste, las plantaciones de te fueron 14% del area estu- diada, pero unicamente 3% de todas las rapaces fueron detectadas alii. Basados en estos resultados, sugerimos que las plantaciones de te pueden ser habitats sub optimos para las rapaces mas grandes y de habitats abiertos. No detectamos grandes aguilas residentes de bosque fuera de grandes areas de bosque no pertnrbado, los cuales probablemente son criticos para su supervivencia en estos paisajes. [Traduccion de Cesar Marquez] When compared to Nearctic and Palearctic spe- cies, Afrotropical raptors are relatively unstudied. This deficiency was recently illustrated by Virani and Watson (1998); describing the state of knowl- 1 E-mail address: nseavy@zoo.ufl.edu edge of 79 raptors breeding in East Africa, they considered only 6.3% of these species “well- known,” while rating 60.8% “unknown.” For many of these unknown species, little or no information exists on distribution, abundance, breeding biolo- gy, or feeding ecology — basic information for iden- 51 52 Seavy and Apodaca VoL. 36, No. 1 tifying populations that are threatened by habitat alteration. Determination of populations that are threatened is an important aspect of conservation, especially in areas where forests are being cleared to support growing human populations. With ca. 104 people per km^, Uganda has one of the densest populations in sub-Saharan Africa and it continues to grow at a yearly rate of 2.6% (FAO 1999). This population has driven wide- spread deforestation. A hundred years ago, ca. 13% of Uganda was covered by moist, broadleaf forest, but only 3% of this forest remains (Sayer et al. 1992). The remaining forested areas, many in reserves and national parks, are under increasing pressure from growing rural populations. Thus, in- formation on raptor abundance and habitat use both inside and outside of these protected areas is important for management and conservation de- cisions. Roadside raptor surveys have been widely used to study relative abundance, community composi- tion, and habitat use of raptors (e.g., Woffinden and Murphy 1977, Ellis et al. 1990). Although lim- itations of roadside surveys are well-recognized (Millsap and LeFranc 1988, Bunn et al. 1995), the method offers an efficient means of describing rap- tor communities. Roadside raptor surveys have been conducted in East Africa (Brown 1971, Thio- llay 1978, Sorley and Andersen 1994), but these efforts have occurred mostly in savanna habitats that are characterized by high densities of large mammals, vultures, and eagles. Here we present results from eight months of roadside raptor sur- veys in the Kabarole District of western Uganda. We describe the abundance and diversity of raptors in mature forest, agriculture/forest mosaic, agri- culture/grassland mosaic, tea plantation, second- ary vegetation, villages, papyrus swamp, and lake- shore habitats. Study Area and Methods The study area (0°13'-0°41'N, 30°19'-30°32'E) lies just north of the equator at the foot of the Ruwenzori Moun- tains in western Uganda (Fig. 1), This area is volcanic in origin and characterized by rolling hills. Elevation ranges from 1100 m in the south to 1590 masl in the north. Although historically forested, much of the land has been cleared for subsistence farming and banana {Musa spp.) and tea (Thea sivensis) plantations to support the growing human population. Within this agricultural landscape, forest fragments remain, especially on steep slopes around crater lakes (Onderdonk and Chapman 2000), but the only extensive forested area is within the 766 km^ Kibale National Park (KNP) . Yearly rainfall, measured at the Makerere University Biological Field Station in KNP, averaged 1778 mm between 1990-98 (L. Chapman and C. Chapman unpubl. data) and peak periods of rainfall occur in April and October (Struhsaker 1997). We conducted roadside raptor surveys along three transects (19, 15, and 17 km in length; Fig. 1) between September 1997-April 1998. We surveyed transects twice each month, once in the morning (between 0800-1000 H) and again in the afternoon (between 1500-1700 H), usually on the same day. Driving at speeds of 20-40 km/ hr with five observers in the vehicle, we stopped briefly to identify raptors on both sides of the transect and re- corded the time, distance from the transect, habitat type, and activity (perched or flying) for each bird. During heavy rain we did not conduct surveys. To quantify habitat availability, we visually estimated habitat composition (10% increments) within 100 m of either side of the road of each km traveled. For this anal- ysis, we recognized eight habitat types: mature forest, ag- riculture/forest mosaic, agriculture/grassland mosaic, tea plantation, secondary vegetation, villages, papyrus swamp, and lakeshore. Mature forest -wsis primarily encoun- tered when transects passed through portions of Kibale National Park. This forest is considered moist, evergreen forest, transitional between lowland rainforest and mon- tane forest (Struhsaker 1997). Canopy height ranges from 25-30 m, with a few trees as tall as 55 m, and com- mon tree species include Diospyros abyssinica, Markhamia platycalyx, Celtis durandii, Uvariopsis congensis, and Bosqueia phoberos (Chapman et al. 1997). Agriculture/forest mosaic was characterized by a patchwork of remnant forest frag- ments surrounded by subsistence agriculture (primarily bananas, millet, cassava, and corn), fallow fields domi- nated by elephant grass (Pennisetum purpureum, 3-5 m tall), cattle pasture, pine and eucalyptus plantations, and scattered houses. Similar to this habitat, agriculture/ grass- land mosaic was also a patchwork of farmland, cattle pas- ture, and forest plantations, but uncultivated land was covered by large areas of elephant grass, and there were no remaining natural forest fragments. Agriculture/grass- land mosaic occurred in the southern portion of the study area. Tea plantations are large fields of tea shrubs grown in rows and trimmed regularly to maintain a uniform height of about 1 m. This crop is one of the few intensive, large-scale agricultural industries in western Uganda. Four other habitat types — secondary vegetation, villages, pa- pyrus swamp, and shores of crater lakes — together comprised <10% of the area surveyed. Because the number of raptors detected per transect was low, we pooled survey results from the three transects and considered them a single transect, though surveyed on different days. We considered monthly surveys inde- pendent samples and used a Wilcoxon signed-ranks test to compare distributions of raptor count data from morning and afternoon surveys. We investigated the re- lationship between monthly rainfall and raptor detec- tions (morning and afternoon surveys pooled) using Spearman rank correlation. We used Chi-square good- ness-of-fit tests to compare observed and expected num- bers of raptors among months and habitat types (all months pooled). We recognize that pooling survey re- sults violates conditions of statistical independence, but March 2002 Surveys in a Disturbed-Forest Landscape 53 Figure 1. our limited sample size prevented us from using other techniques to compare data. To meet the assumptions of the Chi-square goodness- of-fit test for the habitat analysis, we combined habitat types that made up less than 5% of the transect length (secondary vegetation, villages, papyrus swamp, and lake- shore) into one category, so all categories had >5 ex- pected detections. Differences between observed use and availability may occur if raptor detectability varies system- atically among habitat types. To evaluate whether detect- ability varied among habitats, we compared detection dis- tances (Kruskal-Wallis test) for mature forest, agriculture/ forest mosaic, and agriculture/ grassland mo- saic. Significance level for all tests was P = 0.05. Finally, based on eight months of pooled data, we pre- sent three measures of relative abundance to facilitate comparison to other roadside raptor literature: (1) indi- viduals detected per km traveled, (2) km traveled per individual detected, and (3) Woffinden and Murphy’s (1977) Index of Relative Abundance (IRA), calculated cis: No. ind. of each species observed IRA = X 1000 No. km traveled Results and Discussion We observed 14 falconiform species and 77 in- dividuals during our study (Table 1 ) . During morn- ing surveys we detected more raptors (median = 5.5, X = 6.5 raptors/survey) than during afternoon surveys (median = 2.5, x — 3.1, Wilcoxon signed- ranks test, z — —2.54, N — S, P < 0.05). This dif- ference is consistent with temperate zone studies demonstrating that roadside raptor surveys are af- fected by raptor activity patterns (Bunn et al. 1995). The mean number of detections per month was 8.0 raptors, ranging from 5 (December) to 20 (October). The number of detections in October was inflated when we observed 13 Common Buz- zards (Buteo buteo vulpinus) migrating southward. When these sightings were omitted there was no significant difference between the number of rap- tors detected each month (xS “ 4.25, P > 0.05) . There was no relationship between monthly rain- 54 Seavy and Apodaca VoL. 36, No. 1 Table 1. Numbers of raptors detected, detection frequencies, and relative abundance indices of raptors during roadside surveys in Kabarole district, Uganda. Species No. Observed Indiv. Per km Km Trav. Per Indiv. Index of Rel. Abund.^ Black-shouldered 'Kit^{Elanus caeruleus) 3 0.004 272.0 3.7 Black Kite {Milvus migrans) 5 0.006 163.2 6.1 African Fish-Eagle {Haliaeetus vocifer) 1 0.001 816.0 1.2 Palm-nut Vulture ( Gypohierax angolensis) 3 0.004 272.0 3.7 African Harrier-Hawk {Polyboroides typus) 4 0.005 204.0 4.9 Unidentified harrier {Circus spp.) 1 0.001 816.0 1.2 African Goshawk {Accipiter tachiro) 2 0.002 408.0 2.5 Lizard Buzzard {Kaupifalco monogrammicus) 2 0.002 408.0 2.5 Common (Steppe) Buzzard {Buteo buteo vulpinus) 14 0.017 58.3 17.2 Wahlberg’s Eagle {Aquila wahlbergi) 18 0.022 45.3 22.1 Long-crested Eagle {Lophaetus occipitalis) 16 0.020 51.0 19.6 Crowned Hawk-eagle {Stephanoaetus coronatus) 2 0.002 408.0 2.5 Gray Kestrel {Falco ardosiaceus) 2 0.002 408.0 2.5 Northern Hobby {Falco subbuteo) 1 0.001 816.0 1.2 Unidentified hobby {F. subbuteo or cuvieri) 3 0.004 272.0 3.7 ^Analysis of relative abundance after Woffinden and Murphy (1977). fall and the number of raptors detected (r^ = 0.13, N — S, P — 0.76). Omitting the October flight of Common Buzzards did not change the significance of this relationship (r^ — —0.12, N = 8, P = 0.77). Raptors were not equally abundant in all habitats (X\ - 19.0 , P < 0.05). However, the detection dis- tances for mature forest (median = 0, = 0.9 m, N = 9), agriculture/forest mosaic (median = 40, X — 44.7 m, N = 31) and agri culture/ grassland mosaics (median = 40, x — 64.2 m, N = 27) were significantly different (Kruskal-Wallis, “ 20.26, P < 0.05). As a result, raptor detectability was un- doubtedly reduced in forest habitats, but when for- est habitat was excluded from the analysis, raptor detections were still not equally abundant in all habitats (x^g ^ 15.57, P< 0.05). Therefore, we do not believe that this result simply reflects differ- ential detectability. Agricultural/forest mosaic was the most com- mon habitat along the route and accordingly had the greatest species richness (N = 8) and total de- tections {N — 31; Table 2). Agriculture/grassland mosaic had the second highest species richness {N = 6) and total number of detections {N — 27; Ta- ble 2). However, detections per km were higher in agriculture/grassland mosaic (0.17 raptors/km) than in agriculture/forest mosaic (0.09 raptors/ km) . Tea plantations accounted for 14% of the area surveyed, but only 3% of all raptor detections. In this habitat, we detected 0.02 raptors/km and only a single species, the small, insectivorous Gray Kes- trel {Falco ardosiaceus; Table 2). Larger, vertebrate- eating raptors were conspicuously absent from tea plantation habitats, despite the fact that these fields were often bordered by forest fragments or crossed by rows of utility poles, providing suitable perches. In mature forest we detected only Crowned Hawk-Eagles {Stephanoaetus coronatus) and Com- mon Buzzards. All observations {N — 7) of Com- mon Buzzards in this habitat occurred during a single survey in October while this species was pass- ing through on migration. Common Buzzards nor- mally winter in open savanna habitats throughout Africa, and their occurrence in this forested area was probably an unusual event associated with mi- gration. In contrast, Crowned Hawk-Eagles are breeding residents in KNP, where they prey pri- marily upon arboreal monkeys (Skorupa 1989, Mi- tani et al. 2001). The absence of Crowned Hawk- Eagles outside of mature forest suggests that they may be sensitive to deforestation in this region of Africa, perhaps because many primates are absent from forest fragments outside of the park (Onder- donk and Chapman 2000) . The least common habitat, lakeshore, occurred only where one transect passed a small crater lake March 2002 Surveys in a Disturbed-Forest Landscape 55 Table 2. Number of raptors observed in eight habitat types during roadside surveys in Eiabarole district, Uganda. Species Habitat Type' a. AFM AGM TP MF VI SV PS LS Black-shouldered Kite 3 Black Kite 3 2 African Fish-Eagle 1 Palm-nut Vulture 3 African Harrier-Hawk 3 1 Unidentified harrier 1 African Goshawk 2 Lizard Buzzard 2 Common Buzzard 2 3 7 1 1 Wahlberg’s Eagle 2 16 Long-crested Eagle 13 3 Crowned Hawk-Eagle 2 Gray Kestrel 2 Northern Hobby 1 Unidentified hobby 3 Total species 8 6 1 2 2 1 2 Total individuals 31 27 2 9 3 1 4 Raptors per km 0.09 0.17 0.02 0.06 0.10 0.05 0.0 1.25 Km habitat surveyed 337.6 160.0 110.4 148.8 28.8 22.4 4.8 3.2 Percent of total survey length 41.4 19.6 13.5 18.2 3.5 2.8 0.6 0.4 ^Habitat types: AFM = Agriculture/forest mosaic, AGM = Agriculture/grassland mosaic, TP = Tea plantation, ME = Mature forest, VI = Village, SV = Secondary vegetation, PS = Papyrus swamp, and LS = Lakeshore. and accounted for <1% of the habitat surveyed. However, this habitat was the only area where we observed an African Fish-Eagle (Haliaeetus vocifer) and Palm-nut Vultures ( Gypohierax angolensis; Table 2 ). Long-crested Eagles {Lophaetus occipitalis) and Wahlberg’s Eagles {Aquila wahlbergi) were the most abundant raptors, accounting for 34 of 77 (44%) raptors detected (Table 2). We observed no true vultures during our surveys, despite the fact that Brown (1971) recorded four vulture species during roadside surveys in Queen Elizabeth National Park, less than 50 km south of our study area. Most (75%) of the raptors we observed were either res- idents or intra-African migrants. The other 25% of detections were Palearctic migrants, but this per- centage is inflated by migrating Common Buzzards that did not have a sustained presence in the area. Thus, the proportion of migrant species wintering in the area was much less than 25%, even though our surveys took place during the Palearctic winter when northern migrants would be present. Deter- mining the exact proportion of wintering migrants is difficult, because some or all of the unidentified raptors (e.g., Circus spp. and Falco spp.) may have been migratory and some species, such as Black Kites (Milvus migrans) , may have been represented by both migrant (M. m. migrans) and resident (M. m. parasitus) subspecies. However, even if we as- sume that all unidentified species were wintering Palearctic migrants, these individuals account for only 8% of all raptor detections when Common Buzzards are omitted. Based on raptor surveys in Mexico, Guiana, and Ivory Coast, Thiollay (1985) demonstrated that when tropical forests are fragmented and convert- ed to open agricultural habitats, resident species tolerant of open habitats and northern migrants replace resident forest species. Similarly, in a re- view of roadside raptor surveys in South Africa, Cade (1969) recognized that Palearctic migrants were abundant in the disturbed agricultural habi- tats of Transvaal and Orange Eree State, where they accounted for 33-95% of all raptor detections, but were relatively rare in Kruger National Park and Kalahari Gemsbok Park, where they accounted for 56 Seavy and Apodaca VoL. 36, No. 1 only 2-4% of all detections. In savanna habitats of Uganda, Palearctic migrants comprised only 11% of the raptors detected by Brown (1971) and 31- 43% of raptors detected by Thiollay (1978). Most of our study area was deforested and highly-dis- turbed, and although the commonly-detected res- idents were associated with open, agricultural mo- saic habitats, we did not observe a large proportion of wintering migrants. Roadside raptor surveys generally work well in open habitats; however, they are less effective in forested habitats (Millsap and LeFranc 1988). This may be especially problematic in tropical forests where raptor communities are notoriously difficult to sample and complete surveys often involve spe- cial techniques (Thiollay 1989, Whitacre et al. 1992). Thus, the low number of raptors we detect- ed in mature forest is probably not indicative of overall raptor abundance and species richness in this habitat. For instance, we did not detect Cas- sin’s Hawk-Eagles {Spizaetus africanus) nor Ayres’ Hawk-Eagles (Hieraaetus ayresii) . Both are large, un- common, forest-dwelling eagles that breed in Ki- bale National Park (Skorupa et al. 1985, Seavy 2000). Similarly, smaller raptors that use easily overlooked perch sites are often difficult to detect (Millsap and LeFranc 1988) and these biases should be considered when interpreting our survey results. Species that were observed in the study area, but not detected on our transects, were Os- prey {Pandion haliaetus), Banded Snake-Eagle (GV- caetus cinerascens) , Black-chested Snake-Eagle {Cir- caetus pectoralis), African Marsh-Harrier (Circus ranivorus) , Black Goshawk (Accipiter melanoleucus) , Little Sparrowhawk (Accipiter minullus) , Martial Ea- gle (Polemaetus bellicosus) , and African Hobby (Falco cuvieii) . The results of our surveys provide two conclu- sions with implications for raptor conservation in East African forest landscapes. Eirst, in the domi- nant agricultural habitat we detected species asso- ciated with agricultural habitats, but very few for- est-associated species. Thus, although some raptors may be numerous in these cleared habitats, forest- associated species appear to be uncommon, re- stricted to large tracts of forest, and sensitive to deforestation. If the large, forest-dwelling raptors are entirely absent from open habitats surrounding KNP, then these species may be confined to the forest inside the park. Second, our results suggest that tea plantations may provide unique foraging opportunities for small insectivorous raptors, but unsuitable habitat for many larger, open-habitat raptors. In short, not all open habitats are of equal value to all open-habitat raptors. These results should be considered by managers and researchers concerned with raptor conservation and protected area design in forested landscapes of East Africa. Acknowledgments Invaluable assistance in the field was provided by E. Aliganyiri, S. Balcomb, A. Ermosi, T. Lawrence, and J. Mugenyi. We thank L.J. Chapman and C.A. Chapman for support of this project. We express our gratitude to the Makerere University Biological Field Station for provid- ing logistical support in Kibale National Park. J. Piascik prepared the figure of the study area. C.A. Chapman, R.S. Duncan, P. Mundy, J. Paul, D.F. Whitacre, and two anonymous reviewers provided comments on drafts of the manuscript. Literature Cited Brown, L. 1971. African birds of prey. Houghton Mifflin Co., Boston, MA U.S.A. Bunn, A.G., W. Klein, and K.L. Bildstein. 1995. Time- of-day effects on the numbers and behavior of non- breeding raptors seen on roadside surveys in eastern Pennsylvania./. Field Ornithol. 66:544-552. Cade, T.J. 1969. The status of the peregrine and other Falconiformes in Africa. Pages 289-321 in].]. Hickey [Ed,], Peregrine Falcon populations, their biology and decline. Univ. of Wisconsin Press, Madison, WI U.S.A. Chapman, C.A., L.J. Chapman, R.W. Wrangham, G. Isa- birvt-Basuta, and K. Ben-David. 1997. Spatial and temporal variability in the structure of a tropical for- est. Afr. J. Ecol. 35:287-302. Ellis, D.H., R.L. Glinski, and D.G. Smith. 1990. Raptor road surveys in South America. /. Raptor Res. 24:98- 106. FAO. 1999. State of the world’s forests. 1999. United Na- tions, Rome, Italy. Millsap, B.A. and M.N. LeFranc, Jr. 1988. Road transect counts for raptors: how reliable are they? / Raptor Res. 22:8-16. Mitani, J.C., WJ. Sanders, J.S. Lwanga, and T.L. Wind- FELDER. 2001. Predatory behavior of Crowned Hawk- Eagles (Stephanoaetus coronatus) in Kibale National Park, Uganda. Behav. Ecol. Sociohiol. 49:187-195. Onderdonk, D.A. AND C.A. Chapman. 2000. Coping with forest fragmentation: the primates of Kibale National Park, Uganda. Int. /. Primatol. 21:587-611. Saver, J.A., C.S. Harcourt, and N.M. Collins. 1992. The conservation atlas of tropical forests: Africa. Macmil- lan, U.K. Seavy, N.E. 2000. Observations at an Ayres’ Hawk-Eagle nest in Kibale National Park, Uganda. /. Raptor Res. 34:59-60. Skorupa, J.P. 1989. Crowned Eagles Stephanoaetus corona- March 2002 SuRVE\^ IN A Disturbed-Forest Landscape 57 tus in rainforest: observations on breeding chronology and diet at a nest in Uganda. Ibis 131:294-298. , J. Kalina, T.M. Butwski, G. Tabor, and E. Kel- i.OG, 1985. Notes on the breeding biology of Cassin’s Hawk-Eagle Hieraaetus africanus. Ibis 127:120-122. SORLEY, C.S. and D.E. Andersen. 1994. Raptor abun- dance in south-central Kenya in relation to land-use patterns. Afr. J. Ecol. 32:30-38. Struhsaker, T.T. 1997. Ecology of an African rainforest: logging in Kibale and the conflict between conserva- tion and exploitation. Univ. Presses of Elorida, Gaines- ville, FL U.S.A. Thiollay, J.-M. 1978. Population structure and seasonal fluctuations of the Falconiformes in Uganda National Parks. E. Afr. Wildl J. 16:145-151. . 1985. Composition of Falconiform communities along successional gradients from primary rainforest to secondary habitats. Pages 181-190 ml. Newton and R.D. Chancellor [Eds.], World conference on birds of prey: conservation studies on raptors. ICBP Tech. Publ. No. 5, Cambridge, U.K. . 1989. Censusing of diurnal raptors in a primary rainforest: Comparative methods and species detect- ability./. Raptor Res. 23:72-84. Virani, M. and R.T. Watson. 1998. Raptors in the east African tropics and western Indian Ocean islands: state of ecological knowledge and conservation status. J. Raptor Res. 32:28-39. Whitacre, D.F., L.E. Jones, andJ. Sutter. 1992. Census- ing raptors and other birds in tropical forest: further refinements of methodology. Pages 39-52 in D.E. Whi- tacre and R.K. Thorstrom [Eds.], Maya Project Report No. 5. The Peregrine Fund, Boise, ID U.S.A. Woffinden, N.D. and J.R. Murphy. 1977. A roadside rap- tor census in the eastern Great Basin — 1973-74. Rap- tor Res. 11:62-66. Received 25 January 2001; accepted 21 November 2001 J Raptor Res. 36(1):58~65 © 2002 The Raptor Research Foundation, Inc. TROPHIC NICHE OF NORTH AMERICAN GREAT HORNED OWLS Lee a. Cromrich,^ Denver W. Holt, and Shawne M. Leasure Owl Research Institute, PO. Box 39, Charlo, MT 59824 Abstract. — The trophic niche of Great Horned Owls {Bubo virginianus) was summarized using 22 North American studies reporting >100 prey items each. Twenty-one of these studies were reviewed from the published literature, and one, our Montana data, is presented here for the first time. More than 92 species from four taxonomic classes have been recorded from 19 2*78 prey items. Mammals constituted >93.3% of prey from all studies, with six studies reporting 100% mammalian prey. Food-niche breadth ranged from 2.09-19.15 {x = 5.1*7) for combined studies, 2.12-19.15 (x = 6.29) for breeding seasons, and 2.09-4.72 (x = 3.50) for non-breeding seasons. Evenness values ranged from 0.408-0.840 (a = 0.620) for combined studies, 0.420-0.703 {x = 0.596) for breeding seasons, and 0.408-0.724 {x = 0.609) for non-breeding seasons. Estimated masses of individual prey species ranged between 2 and 6300 g. Birds were only a minor part of the owl diet, although a variety of species were eaten. Key Words: diet, food-niche breadth. Great Horned Owl; Bubo virginianus; North America. Nicho trofico del Gran Buho Cornado Americano Resumen. — El nicho trofico de {Bubo virginianus) fue corapendiado usando 22 estudios norte americanos reportando >100 Items presa cada uno. Veintiuno de esos estudios fueron revisados en la literatura publicada, y uno, nuestros datos de Montana, se presentan aqui por primera vez. Mas de 92 especies de cuatro clases taxonomicas han sido registradas a partir de 19 278 Items presa. Los mamiferos cons- tituyeron >93.3% de presas en todos los estudios, con seis estudios reportando 100% de presas ma- mlferas. La amplitud del nicho alimenticio estuvieron en el rango de 2.09-19.15 {x = 5.17) para estudios combinados, 2.12-19.15 {x = 6.29) para estaciones reproductivas, y 2.09-4.72 (x = 3.50) para tempo- radas no reproductivas. La masa estimada de especies presa individualmente estuvo entre 2 y 6300 gr. Las aves fueron tan solo una parte menor de la dieta del buho, aunque una variedad de especies fueron consumidas. [Traduccion de Cesar Marquez] The Great Horned Owl {Bubo virginianus) is per- haps the most widely-distributed owl in North America (Houston et al. 1998, Holt et al. 1999). Numerous studies of the food habits of Great Horned Owls have been conducted in North America and it has been considered to be an op- portunistic feeder. Indeed, the Great Horned Owl has been reported to have the broadest diet of any North American owl species (Marti and Kochert 1996, Houston et al. 1998). However, the owl’s tro- phic niche has not been reviewed continent-wide. Earhart and Johnson (1970) summarized principal food habits of Great Horned Owls from published literature, but did not identify prey to the species level, provide prey numbers, or discuss their con- clusions. Jaksic and Marti (1984) made compari- ^ E-mail address: owlmontana@charlo.net sons between a few Neotropical and Nearctic lo- calities, but compared owl diets from only two regions in North America. Our paper summarizes the trophic niche of Great Horned Owls from 22 North American studies; 21 from published litera- ture, and one, our original Montana data. Our objectives were to: (1) determine Great Horned Owl trophic niche from west-central Mon- tana and (2) compare trophic niche among North American studies. Methods In Montana, we collected pellets and prey remains an- nually from 10 territories in the Missoula and Mission valleys during the breeding and non-breeding seasons from 1987-95. Prey was identified using local dichoto- mous keys for mammals (Hoffmann and Pattie 1968) and by comparing feather and body parts of prey with mu- seum specimens at the Philip L. Wright Zoological Mu- seum (University of Montana). Numbers and proportions 58 March 2002 Great Horned Owl Trophic Niche 59 of prey types were then compared between breeding and non-breeding seasons for Montana. Comparisons of tro- phic niche were then made among other available North American data sets. We defined trophic niche as the relationship between owls and their prey. We followed Marti’s (IQS'/) defini- tions for trophic diversity in which a broad food-niche breadth (FNB) has a high number of prey species, which are nearly equally distributed, and a narrow food-niche breadth has a low number of prey species unequally dis- tributed. However, we found no method to determine the statistical significance between narrow and wide food- niche breadth. We compared owl trophic niche from 22 North American studies with >100 prey items each (Ta- bles 1-3). We then divided these studies into breeding season and non-breeding season diets. To compare tro- phic niche among studies, we primarily used prey iden- tified to the species. Prey identified to genus were in- cluded if they occurred frequently or exhibited an unusual body mass. Insects, arachnids, and unidentified reptiles, birds, and mammals were eliminated from tro- phic niche comparisons because they were either not identified to genus or occurred only rarely (<1%) in the diet. Food-niche breadth {H') was calculated for each study using the antilog of the Shannon-Weiner diversity index (Marti 1987). We used this equation because it is linearly related to the number of prey categories in the sample. Evenness was calculated using Alatalo’s (1981) modifi- cation of Hill’s (1973) equation; Evenness = {N^ ~ 1)/ {Ni — 1), where — exp H' and Evenness values range from zero to one. An evenness value of one indicates prey proportions in the diet are equal. We com- pared food-niche breadth and evenness values from all studies as well as those from breeding and non-breeding seasons using the Mann-Whitney Gtest (Eowler and Co- hen 1990). Spearman rank correlation (Fowler and Cohen 1990) was used to examine the relationship between the num- ber of mammalian species and number of prey items with food-niche breadth values among studies. We did so to determine if wider food-niche breadth values were asso- ciated with increased numbers of prey or species in the diet. The Spearman rank correlation was also used to examine the relationship between number of prey items and food-niche breadth because food-niche breadth val- ues can fluctuate with sample size, thus influencing the results. A relative-size category of the main prey classes eaten by the owls was derived using body mass estimates of mammals (Whitaker 1992) and birds (Dunning 1984). Standard mean prey biomass estimates were not calculat- ed based on species because of unfounding factors. For example, standard prey biomass estimates are usually de- rived from the adult age class and do not consider other age classes in the population. Further, mean prey biomass estimates generally give whole carcass masses and do not consider that only specific portions of some medium to large prey are eaten (Holt 1993, 1994). Results The Montana study yielded 4350 prey items: 2696 from the breeding season and 1654 from the Table 1. The number of individual prey consumed by Great Horned Owls during breeding and non-breeding seasons in Montana from 1987-95. Species No. Breeding Season OF Prey Non-Breeding Season Mammals Microtus pennsylvanicus 1264 902 Microtus montanus 1158 470 Microtus spp. 72 159 Peromyscus maniculatus 37 52 Thomomys talpoides 54 39 Ondatra zibethicus 13 14 Mustela frenata 1 1 Sylvilagus nuttallii 2 — Tamias amoenus — 1 Tamiasciurus hudsonicus 2 — Glaucomys sabrinus — 1 Birds Sturnus vulgaris 30 4 Phasianus colchicus 12 4 Fulica ajnericana 6 1 Pica pica 7 1 Turdus migratorius 3 — Colaptes aumtus 2 1 Xanthocephalus xanthocephalus 12 — Anas platyrhynchos 2 2 Anas spp. 1 — Bomby cilia spp. 1 — Sturnella neglecta 1 — Asia otus 1 — Porzana Carolina 1 — Rallus limicola 1 — Bonasa umbellus 1 — Agelaius phoeniceus — 1 Other Catostomus spp. 4 — crayfish 6 1 squawfish 2 — Total 2696 1654 = 4350 non-breeding season. Although collectively the owls ate a wide variety of prey {N = 28), they ate predominately small mammals, particularly voles (Table 1). During the breeding season, the owls consumed 28 species of prey. Of these however, they ate pre- dominately small mammals, especially Microtus voles (92.5%, N = 2494). During the non-breeding season, the owls ate only 16 species of prey, again consuming predominately Microtus \o\e^ (92.5%, N 60 Cromrich et al. VoL. 36, No. 1 Table 2. Landscape diets of Great Horned Owls in North America. Percent of prey in taxonomic classes calculated from 22 studies, representing 19 278 prey items. No. OF Prey Osteichthyes Percent Crustacea Aves Mammalia Location Source Breeding season 2696 0.2 0.2 3.09 6.6 MT This study 1896 — — — 100.0 OR Maser 8c Brodie 1966 1300 — — 1.4 98.6 ID Marti & Kochert 1996 398 — 1.5 2.8 95.7 UT Smith & Murphy 1973 356 0.8 — 2.2 96.9 WY Craighead & Craighead 1969 276 3.3 — 31.5 65.2 NY, NJ, CT Bosakowski & Smith 1992 209 — — — 100.0 MB Bird 1929 142 — 9.2 20.4 70.3 MI Craighead Sc Craighead 1969 119 — — 30.3 69.7 OH Springer & Kirkley 1978 Non-breeding season 1845 — — 1.4 98.6 MI Craighead 8c Craighead 1969 1654 — 0.1 1.0 98.9 MT This study 756 2.5 — 1.9 95.6 MT Seidensticker 1968 584 — — 0.3 99.7 GA Rudolph 1978 210 — — 2.4 97.6 IN Kirkpatrick 8c Conway 1947 161 — — — 100.0 NE Rickart 1972 122 — — — 100.0 YT Weir & Hanson 1989 Breeding and non-breeding seasons 2571 - 0.1 4.4 95.2 WI Errington 1932 2152 — — 1.7 98.1 CO Marti 1974 809 0.1 — 1.7 98.1 WA Knight & Jackman 1984 568 — — — 100.0 CA Barrows 1989 273 — 0.7 21.2 78.0 WI Orians 8c Kuhlman 1956 178 — — — 100.0 OK Tyler & Jensen 1981 ^ 1531). The decreased prey species diversity dur- ing the non-breeding season reflected the fewer species of prey available during the fall and winter months in Montana. The 22 studies combined yielded 19 278 prey items (Table 2) from eight western, six central, and three eastern states, and two Canadian provinces. Studies from New York and Pennsylvania (Latham 1950), and Alberta (Rusch et al. 1972, Mclnvaille and Keith 1974, Adamcik et al. 1978) were also re- viewed but omitted from trophic calculations be- cause dominant prey species were not always iden- tified to genus or species. The owls consumed >92 prey species from four taxonomic classes: Osteichthyes, Crustacea, Aves, and Mammalia (Table 2). Mammals composed ^65.2% of the prey from each study, constituting 93.3% of the total prey from all studies. Six studies reported 100% mammalian prey (Table 2). Although the owls preyed on a broad number of species overall, Microtus {N — 10 studies), Peromys- cus (N — 6), Perognathus {N = 2), Sigmodon (N — 1), and Lepus (N = 1) species represented the highest percentage of prey in all studies. Overall, food-niche breadth values ranged from 2.09-19.15 (x = 5.17, SD ± 3.61) (Table 3). Food-niche breadth values for the breeding season (range = 2.12-19.15, X - 6.29, SD ± 5.39, N = 9) and non- breeding season (range = 2.09-4.72, x 3.50, SD ± 0.82, N = 7) were similar. Food-niche breadths were not significantly different (Mann-Whitney U = 24.5, P > 0.05) between seasons. The broadest food-niche breadth (FNB — 19.15) was from New York, New Jersey, and Connecticut, where 15 mammal species constituted 65.2% of the diet, with Peromyscus representing 14.3% (Bosa- kowski and Smith 1992) (Table 3). Fourteen other mammal, 20 bird, and two fish species comprised the remainder. The broad FNB in this study, com- pared to other studies, may be explained by the March 2002 Great Horned Owl Trophic Niche 61 Table 3. Trophic parameters calculated from twenty-two studies representing 19 278 prey items. No. OF Prey Food-niche Breadth Evenness Location Source Breeding season 2696 3.27 0.644 MT This study 1896 2.12 0.465 OR Maser &c Brodie 1966 1300 8.12 0.686 ID Marti & Kochert 1996 398 4.47 0.420 UT Smith & Murphy 1973 356 2.85 0.566 WY Craighead & Craighead 1969 276 19.15 0.670 NY, NJ, CT Bosakowski & Smith 1992 209 2.94 0.687 MB Bird 1929 142 4.55 0.527 MI Craighead & Craighead 1969 119 9.10 0.703 OH Springer & Kirkley 1978 Non-breeding season 1845 2.94 0.669 MI Craighead &: Craighead 1969 1654 3.43 0.649 MT This study 756 3.91 0.622 MT Seidensticker 1968 584 2.09 0.724 CA Rudolph 1978 210 4.72 0.631 IN Kirkpatrick & Conway 1947 161 3.84 0.558 NE Rickart 1972 122 3.56 0.408 YT Weir & Hanson 1989 Breeding and non-breeding season 2571 4.89 0.629 WI Errington 1932 2152 6.36 0.605 CO Marti 1974 809 5.27 0.602 WA Knight &: Jackman 1984 568 3.89 0.840 CA Barrows 1989 273 6.98 0.604 WI Orians & Kuhlman 1956 178 5.87 0.720 OK Tyler & Jensen 1981 large number of bird, as well as mammal, species included in the diet, or the relatively small sample size (N= 276). Food-niche breadth calculated from Ohio (Springer and Kirkley 1978) was also broad (9.10) compared to other studies (Table 3). In this study, six mammal species constituted 69.7% of the owl’s diet with Microtus representing 26.1%. Six mammal and 12 bird species represented the remainder of the diet. The narrowest FNBs came from Califor- nia (2.09), Oregon (2.12), Wyoming (2.85) and Manitoba (2.94), respectively (Table 3). In all these cases, small mammals dominated the diet (Bird 1929, Maser and Brodie 1966, Craighead and Craighead 1969, Rudolph 1978). Evenness values overall ranged from 0.408—0.840 (x = 0.620, SD ± 0.101). Evenness values for the breeding (range — 0.420-0.703, x = 0.596, SD ± 0.105, N ~ 9) and non-breeding season (range — 0.408-0.724, A = 0.609, SD ± 0.102, N = 7) were also similar (Table 3). Evenness was not signifi- cantly different (Mann-Whitney U = 30, P > 0.05) between seasons. A weak positive correlation existed between the number of mammalian species in the diet and food-niche breadth values (r^ = 0.299, P < 0.01). A weak negative relationship occurred between the number of prey items and food-niche breadth val- ues (r, = —0.207, P > 0.01), suggesting that sam- ple sizes were not influencing the results. Mammal prey biomass ranged from 2 g, (masked shrew {Sorex cinereus^) to 6300 g, (striped skunk [Me- phitis mephitis} ) (Whitaker 1992) . The m^ority of prey ranged from 2-1800 g and included shrews, voles, mice, rats, pocket gophers, squirrels, and rabbits. The dominant prey from each study, Microtus, Peromyscus, Perognathus, Sigmodon, and Lepus ranged in body mass from 16-85 g, 10-43 g, 16-47 g, 80-120 g, and 1800- 3600 g, respectively (Whitaker 1992). Other medium- sized mammals, including yellow-bellied marmot {Marmota flaviventris) and white-tailed jackrabbit {Le- 62 Cromrich et al. VoL. 36, No. 1 pus toimsendii) rarely occurred in the diet and ranged from 2200-4500 g. Birds were not a major part of the owl’s diet, but a wide variety of species were eaten. Waterfowl, shorebirds, pheasants and allies, and passerines represented the majority of bird prey. Several owl species were also reported as prey in nine studies: Northern Saw-whet Owl (Aegolius acadicus) (Bosa- kowski and Smith 1992), Long-eared Owl {Asio otus) (Marti 1976, Holt this study), Barn Owl {Tyto alba) (Knight and Jackman 1984), and Eastern Screech-Owl {Otus asio) (Errington 1932, Orians and Kuhlman 1956, Craighead and Craighead 1969, Bosakowski and Smith 1992). Body masses of avian prey ranged from: 318-1100 g, waterfowl; 74- 415 g, shorebirds; 178-1317 g, pheasants and allies; 88-580 g, owls; and 29-458 g, passerines (Dunning 1984). Passerines constituted most of the avian prey. Discussion Great Horned Owls are generally considered to be opportunistic feeders, preying on a broader range of species than any other North American owl (Craighead and Craighead 1969, Voous 1988, Marti and Kochert 1996, Houston et al. 1998). Bo- sakowski and Smith (1992) reported such unusual species as a raccoon {Procyon lotor), opossum {Di- delphis virginiana) , and a Red-shouldered Hawk {Buteo lineatus); Marti (1974) reported a yellow-bel- lied marmot and black-tailed prairie dog {Cynomys ludovicianus); Errington (1932) reported a striped skunk; and Rudolph (1978) reported a Brazilian free-tailed bat {Tadarida brasiliensis) . Llinas-Gutier- rez et al. (1991) reported a wide variety of arach- nids, insects, and reptiles in the owl diet, and Roh- ner and Doyle (1992) reported a Great Horned Owl feeding on an adult Northern Goshawk {Ac- cipiter gentilis ) . The moderate trophic niche (high number of prey species unequally distributed [see Methods section] ) of the Great Horned Owl reported here- in somewhat contrasts with previous studies (see text). Excluding predominately insectivorous owl species, the Great Horned Owl’s moderate food- niche breadth (opportunistic feeding) aligns it with species such as the Burrowing Owl {Speotyto cunicularia) (Haug et al. 1993), Spotted Owl (Strix occidentalis) (Gutierrez et al. 1995), and Eastern Screech-Owl (Gehlbach 1995), for example. Spe- cies apparently more opportunistic than Great Horned Owls include the Ferruginous Pygmy-Owl {Glaucidium brasilianum) (Proudfoot 1997) and Northern Pygmy-Owl {Glaucidium gnoma) (Holt and Leroux 1996, Holt and Petersen 2000). More specialized species include the Snowy Owl {Nyctea scandiaca) (Watson 1957, Parmelee 1992), Short- eared Owl {Asio flammeaus) (Holt 1993, Holt and Leasure 1993), Northern Saw-whet Owl (Holt et al. 1991, Cannings 1993), and Barn Owl (Marti 1989, 1992). Other Bubo species have a trophic niche similar to the Great Horned Owl. Herrera and Hiraldo (1976) reported FNB values ranging from 2.40- 6.68 (x = 4.13, SD ± 0.01) for the Eurasian Eagle- Owl {Bubo bubo) in Europe. Jaksic and Marti (1984) found that Great Horned Owls and Eurasian Eagle- Owls followed a similar trophic pattern in North American and European shrubland. Donazar et al. (1989), however, reported limited dietary conver- gence between these two species. They attributed discrepancies in trophic diversity to the differences between similar North American and European bi- omes, variations in the composition and abun- dance of prey types, and differences in tbe body masses of Great Horned Owls and Eurasian Eagle- Owls. The moderate trophic niche of Great Horned Owls could be the result of several factors, includ- ing prey species size, diversity, density, availability, and distribution. Marti (1974) suggested that al- though owls can capture a broad range of prey siz- es, an optimum size exists in terms of how efficient- ly a particular individual prey item can be found and caught. He argued that very small prey is only efficient for Great Horned Owls if it can be caught quickly and easily. Marti felt that prey density, ease of killing, overlap of time of activity between pred- ator and prey, and learning by individual owls all determine what proportions of the diet a particular prey species will comprise. The Great Horned Owl’s moderate food-niche breadth may also reflect the habitat or time of day in which they forage. Open areas the Great Horned Owl inhabits are frequented by mice, voles, lagomorphs, and gophers, which may emerge during the owl’s optimal feeding periods of evening, night, and early morning (Maser et al. 1970) . The community structure of predators with- in a particular habitat may also affect food-niche breadth. Marti et al. (1993) found that although predators in an area may utilize prey resources in different fashions, patterns of resource use do emerge, particularly in terms of predator size. March 2002 Great Horned Owl Trophic Niche 63 The diet of Great Horned Owls may vary de- pending upon the particular region the owls in- habit (Marti et al. 1993). Hayward et al. (1993) found that coastal Great Horned Owls in Washing- ton fed exclusively on birds during the summer months. Bosakowski et al. (1989) reported owls liv- ing in the deciduous forests of New Jersey, New York, and Connecticut preyed more heavily upon birds than those living in open coniferous forests of the western United States. Desert owls fed on a variety of arachnids, insects, and reptiles because of their availability and abundance in that biome (Jaksic and Marti 1984, Barrows 1989, Llinas-Gu- tierrez et al. 1991). Jaksic and Marti (1984) found that the diversity of Great Horned Owl prey at the class level was very low in the temperate regions and very high in the desert regions of North America. They be- lieved this difference reflected the greater repre- sentation of mammals in the diet of temperate owls, thus resulting in a moderate trophic niche. Great Horned Owls may respond opportunisti- cally to the local profile of prey sizes and densities (Jaksic and Delibes 1987, Jaksic 1988). Llina- Gutierrez et al. (1991) suggested that lagomorphs and rodents were the dominant prey species in their study compared with other desert studies in the region because of their high abundance. Rusch et al. (1972) reported that the diet of Great Horned Owls was strongly affected by changes in the numbers of snowshoe hare {Lepus americanus) . They found that in years with high snowshoe hare populations, owls exhibited higher predation rates on snowshoe hare and lower predation rates on mice and voles. The data herein support the general conclusion that Great Horned Owls prey on a wide range of species. However, the data also show convincingly that Great Horned Owls feed primarily on only three to four species of voles and mice under most conditions, indicating a moderate food-niche breadth. Acknowledgments We thank the Flathead Indian Reservation, Montana Department of Fish, Wildlife and Parks, United States Fish and Wildlife Service, and many private landowners for access to lands. Marc Bechard, Carl Marti, and Fabian Jaksic made thoughtful comments on the manuscript. Literature Cited Adamcik, R.S., A.W. Todd, and L.B. Keith. 1978. De- mographicand dietary responses of Great Horned Owls during a snowshoe hare cycle. Can. Field-Nat. 92 156-166. Aiatalo, R.V. 1981. Problems in the measurement of evenness in ecology. Oikos 37:199-204. Barrows, C.W. 1989. Diets of five species of desert owls West. Birds 20:1-10. Bird, R.D. 1929. The Great Horned Owl in Manitoba. Can. Field'Nat. 43:79-83. Bosakowski, T, R. Speiser, and D.G. Smith. 1989. Nest- ing ecology of forest-dwelling Great Horned Owls, Bubo virginianus, in the eastern deciduous forest bi- ome. Can. Field-Nat. 103:65-69. AND D.G. Smith. 1992. Comparative diets of sym- patric nesting raptors in the eastern deciduous forest biome. Can.]. Zool. 70:984—992. Cannings, RJ. 1993. Northern Saw-whet Owl (Aegohus acadicus). In A. Poole and F. Gill [Eds.], The birds of North America, No. 42. The Academy of Natural Sci- ences, Philadelphia, PA and The American Ornithol- ogists’ Union, Washington, DC U.S.A. Craighead, JJ- AND F.C. Craighead. 1969. Hawks, owls and wildlife. Dover Publications, New York, NYU.S A. Donazar, J.A., F. Hiraldo, M. Delibes, and R.R. Estrel- la. 1989. Comparative food habits of the Eagle Owl Bubo bubo and the Great Horned Owl Bubo virginianus in six Palearctic and Nearctic biomes. Ornis Scand. 20: 298-306. Dunning, J.B. 1984. Body weights of 686 species of North American birds. West. Bird Band. Assoc. Mono. No. 1 Earhart, C.M. and N.K. Johnson, 1970. Size dimor- phism and food habits of North American owls. Con- dor 72:251-264. Errington, P.L. 1932. Food habits of southern Wisconsin raptors. Condor 34:176-186. Fowler, J. and L. Cohen. 1990. Practical statistics for held biology. Great Britain Redwood Press Limited, Manchester, U.K. Gehlbach, F.R. 1995. Eastern Screech-Owl {Otus asio) In A. Poole and F. Gill [Eds.], The birds of North Amer- ica, No. 165. The Academy of Natural Sciences, Phil- adelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Gutierrez, R.J., A.B. Franklin, and W.S. Lahave. 1995 Spotted Owl {Strix occidentalis) . In A. Poole and F. Gill [Eds.], The birds of North America, No. 179. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Haug, E.A., B.A. Millsap, and M.S. Martell. 1993. Bur- rowing Owl (Speotyto cunicularia) . In A. Poole and E Gill [Eds.], The birds of North America, No. 61. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Hayward, J.L.,J.G. Galusha, and G. Frias. 1993. Analysis of Great Horned Owl pellets with Rhinoceros Auklet remains. Auk 110:133-135. 64 Cromrich et al. VoL. 36, No. 1 Herrera, C.M. and F. Hiraldo. 1976. Food-niche and trophic relationships among European owls. Ornis Scand. 7:29-41. Hill, M.O. 1973. Diversity and evenness: a unifying no- tation and its consequences. Ecology 54:427—432. Hoffmann, R.S. and D.L. Pattie. 1968. A guide to Mon- tana mammals. University Press, Missoula, MT U.S.A. Holt, D.W., E. Andrews, and N. Claflin. 1991. Non- breeding season diet of Northern Saw-whet Owls, Ae- golius acadicus, on Nantucket Island, Massachusetts. Can. Field-Nat. 105:382-385. . 1993. Trophic niche of nearctic Short-eared Owls. Wilson Bull. 105:497-503. and S.M. Leasure. 1993. Short-eared Owl {Asia flammeus). In A. Poole and F. Gill [Eds.], The birds of North America, No. 62. The Academy of Natural Sci- ences, Philadelphia, PA and The American Ornithol- ogists’ Union, Washington, DC U.S.A. . 1994. Effects of Short-eared Owls on Common Tern colony desertion, reproduction and mortality. Colon. Waterbirds 17:1-6. AND L.A. Leroux. 1996. Diets of Northern Pygmy- Owls and Northern Saw-whet Owls in west-central Montana. Wilson Bull. 108:123-128. , R. Berkley, C. Deppe, P.L. Enriquez-Rocha, P.D. Olsen, J.P. Petersen, J.L. Rangel-Salazar, and K. Wood. 1999. Strigida species accounts. Pages 153-242 in], del Hoyo, A. Elliott, andj. Sargatal, [Eds.], Hand- book of birds of the world, Vol. 5. Lynx Edicions, Bar- celona, Spain. and J.L. Petersen. 2000. Northern Pygmy-Owl {Glaucidium gnoma) . In A. Poole and F. Gill [Eds.], The birds of North America, No. 494. The birds of North America, Inc., Philadelphia, PA U.S.A. Houston, C.S., D.G. Smith, and C, Rohner. 1998. Great Horned Owl {Bubo virginianus) . In A. Poole and E. Gill [Eds.], The birds of North America, No. 372. The Birds of North America, Inc., Philadelphia, PA U.S.A. Jaksic, F.M. and C.D. Marti. 1984. Comparative food habits of Bubo owls in Mediterranean-type ecosys- tems. Condor 86:288-296. and M. Delibes, 1987. A comparative analysis of food niche relationships and trophic guild structure in two assemblages of vertebrate predators differing in species richness: causes, correlations and conse- quences. Oecologia 71:461-472. . 1988. Trophic structure of some nearctic, neo- tropical and palearctic owl assemblages: potential roles of diet opportunism, interspecific interference and resource depression./. Raptor Res. 22:44—52. Kirkpatrick, C.M. and C.H. Conway. 1947. The winter foods of some Indiana owls. Am. Midi. Nat. 38:755- 766. Knight, R.L. and R.E. Jackman. 1984. Food niche rela- tionships between Great Horned Owls and Common Barn Owls in eastern Washington. Auk 101:175-179. Latham, R.M. 1950. The food of predaceous animals in northeastern United States. Pennsylvania Game Gom- mission, Harrisburg, PA U.S.A. Llinas-Gutierrez, J., G. Arnaud, and M. Acevedo. 1991 Food habits of Great Horned Owl {Bubo virginianus) in the Cape Region of lower California, Mexico. J Raptor Res. 25:140-141. Marti, C.D. 1974. Feeding ecology of four sympatric owls. Condor 76:45-61. . 1976. Review of prey selection by Long-eared Owl. Condor 78:331-336. . 1987. Raptor food habits studies. In B.A. Giron Pendleton, B.A. Millsap, K.W. Cline, and D.M. Bird [Eds.], Raptor management techniques manual. Nat Wildl. Fed. Sci. & Tech. Ser. No. 10. Washington, DC U.S.A. . 1989. A long-term study of food-niche dynamics in the Common Barn Owl: comparisons within and between populations. Can.]. Zool. 66:1803-1812. . 1992. Barn Owl {Tyto alba). In A. Poole, P. Stet- tenheim, and F. Gill [Eds.], The birds of North Amer- ica, No. 1, The Academy of Natural Sciences, Phila- delphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. . K. Steenhof, M.N. Kochert, and J.S. Marks 1993. Community trophic structure: the roles of diet, body size, and activity time in vertebrate predators Oikos 67:6-18. , E. Korpimaki, and E.M. Jaksic. 1993. Trophic structure of raptor communities: a three-continent comparison and synthesis. Ch. 2. /iiD.M. Power [Ed.], Cxxrrent Ornithology, Vol. 10. Plenum, NYU.S.A. and M.N. Kochert. 1996. Diet and trophic char- acteristics of Great Horned Owls in southwestern Ida- ho./. Field Ornithol. 67:499-506. Maser, C. and E.D. Brodie, Jr. 1966. A study of owl pellet contents from Linn, Benton, and Polk counties, Oregon. Murrelet 47:9-14. , E.W. Hammer, and S.H. Anderson. 1970. Com- parative food habits of three owl species in central Oregon. Murrelet 51:29-33. McInvaille, W.B. and L.B. Keith. 1974. Predator-prey re- lations and breeding biology of the Great Horned Owl and Red-tailed Hawk in central Alberta. Can Field-Nat. 88:1-20. Orians, G. and F. Kuhlman. 1956. Red-tailed Hawk and Great Horned Owl populations in Wisconsin. Condor 58:371-385. Parmalee, D.F. 1992. Snowy Owl {Nyctea scandiaca). In A Poole and F. Gill [Eds.], The birds of North America, No. 10. The Academy of Natural Sciences, Philadel- phia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Proudfoot, G.A. 1997. Food habits of nesting Ferrugi- nous Pygmy Owls in southern Texas. Wilson Bull. 109. 741-748. Rickart, E.A. 1972. An analysis of Barn Owl and Great March 2002 Great Horned Owl Trophic Niche 65 Horned Owl pellets from western Nebraska. Prairie Nat. 4:35-38. Rohner, C. and F.I. Doyle. 1992. Food stressed Great Horned Owl kills adult Goshawk: exceptional obser- vation or community process? J. Raptor Res. 26:261- 263. Rudolph, S. 1978. Predation ecology of coexisting Great Horned and Barn owls. Wilson Bull. 90:134-137. Rusch, D.H., E.C. Meslow, P.D. Doerr, and L.B. Keith. 1972. Response of Great Horned Owl populations to changing prey densities./. Wildl. Manage. 36:282-296. Seidensticker, J.C. 1968. Notes on the food habitats of the Great Horned Owl in Montana. Murrelet Smith, D.G. andJ.R. Murphy. 1973. Breeding ecology of raptors in the Eastern Great Basin of Utah. Brigham Young University. Science Bull. Biol. Series 18:1-76. Springer, M.A. and J.S. Kirkley. 1978. Inter- and intra- specific interactions between Red-tailed Hawks and Great Horned Owls in central Ohio. Ohio J. Sci. 78 323-328. Tyler, J.D. and J.F. Jensen. 1981. Notes on foods of Great Horned Owls {Bubo virginianus) in Jackson County, Oklahoma. Proc. Okla. Acad. Sci. 61:28-30. Voous, K.H. 1988. Owls of the northern hemisphere MIT Press, Cambridge, MA U.S.A. Watson, A. 1957. The behavior, breeding and food ecol- ogy of the Snowy Owl {Nyctea scandiaca) . Ibis 99:419- 462. Weir, D. and A. Hanson. 1989. Food habits of Great Horned Owls, Bubo virginianus, in the northern taiga of the Yukon Territory and Alaska. Can. Field-Nat. 103: 12-17. Whitaker, J.O, 1992. The Audubon Society field guide to North American mammals. Alfred A. Knopf, New York, NY U.S.A. Received 27 October 2000; accepted 15 September 2001 Short Communications J Raptor Res. 36(1):66— 70 © 2002 The Raptor Research Foundation, Inc. Social Organization of a Trio of Bearded Vultures ( Gypaetus barbatus) : Sexual and Parental Roles Joan Bertran and Antoni Margalida^ Group of Study and Protection of the Bearded Vulture (GEPT), Apdo. 43, E-25520, El Pont de Suert, (Lleida), Spain Key Words: Bearded vulture, Gypaetus barbatus; coopera- tive polyandry, parental roles', homosexual matings', copulation behavior, Spain. In birds, cooperative polyandry (Oring 1986) is an un- usual mating strategy. In raptors it only happens exten- sively in the Galapagos Hawk {Buteo galapagoensis) (Faa- borg and Bednarz 1990), but polyandry has been described as occasionally occurring in some other spe- cies, that are usually considered monogamous (e.g., Telia 1993, Arroyo 1996). The Bearded Vulture {Gypaetus barbatus) is a large, ter- ritorial vulture which nests on rocky cliffs in certain mountain areas of the southern Palearctic and of the Af- rotropical regions (Hiraldo et al. 1979), and that feeds mainly on bones. The species is generally considered mo- nogamous, but in the Pyrenees, polyandrous trios are rel- atively common (Heredia and Donazar 1990). The Pyr- enees (Spanish and French sides) hold a small isolated population of Bearded Vultures which constitutes the bulk of the breeding population in the western Palearc- tic. In this population, the percentage of territories oc- cupied by trios has increased gradually since 1979, when the first case was reported; 11.5% of 56 territories were recorded in 1988 (Heredia and Donazar 1990); and 15.2% of 92 were observed in 2000 (Heredia and Mar- galida 2001, M. Razin pers. comm.). These trios generally remain stable from one year to the next, but occasionally they may break up (pers. observ.). Trios tend to be ob- served in traditional breeding areas with the highest food availability, but their productivity is similar to that of mo- nogamous pairs (Heredia and Donazar 1990). The factor explaining the existence of trios remains obscure, al- though it seems probable that the formation of such groups might be favored by a reduced availability of suit- able sites for reproduction (Heredia and Donazar 1990). Observing the Bearded Vulture is a difficult task, and this, coupled with its low reproductive success (Heredia and Margalida 2001), accounts for the little information ^ Corresponding author’s e-mail address: margalida® gauss.entorno.es available on the structure in such formations. In addi- tion, trios were not known until recently. The first obser- vations carried out by Heredia and Donazar (1990) in- dicated that the two male members of a trio copulate with the female and that both males provide food for chicks. However, no detailed data are available to quan- tify the behavior and parental contribution of each in- dividual in these groups. In this paper, the copulatory behavior and sexual re- lationships in a Pyrenean Bearded Vulture trio are de- scribed in detail for the first time, and the individual in- vestment of the three birds in parental activities are examined. Study Area and Methods We conducted the present study in the Catalonian Pre- Pyrenees (northeastern Spain), between 1991 and 1992. This is an area with isolated calcareous massifs (maxi- mum altitude of 2070 masl) and Mediterranean vegeta- tion. The trio studied was formed in 1986 after an adult bird joined the territory already occupied by a reproductive pair. No successful breeding occurred between 1986-89, after which this group first bred as a trio in 1990. The study site was visited in bouts of two consecutive days at 5-d intervals resulting in a total of 699 hr of observation distributed as follows: pre-laying period (October-De- cember 1991) 212 hr; incubation (December 1991-Feb- ruary 1992) 101 hr; and nestling period (April-July 1991 and February-May 1992) with 253 and 133 hr, respec- tively. We made observations with spotting scopes from vantage points at a distance of about 300 m from the nests. Birds were sexed using size criteria and based on dif- ferences noted in their parental activities. Females are slightly larger than males (Hiraldo et al. 1979) and tend to be more intensely colored (Negro et al. 1999). In this respect, the male that appeared to be subordinate (male B) was smaller and showed a paler ventral coloring com- pared to the other individuals (the female and male A). Males are more active than females in supplying material to the nest and in territorial defense behavior (Margalida and Bertran 2000a, Margalida and Bertran 2000b). The identification of the birds was based on comparison of the molt patterns and the individual markings and pat- terns on pectoral bands and crowns. The long molting 66 March 2002 Short Communications 67 Pre-laying Post-laying Day Figure 1. Temporal distribution of individual copulation behavior and success in a trio of Bearded Vultures in relation to the day on which incubation began (day 0). Shading indicates the time period during which behaviors were observed with probable contact of the cloaca; the proportions and percentage (in parentheses) of successful copulations are indicated. process facilitated identification of the birds during much of the study period. Bearded Vultures usually copulate on an exposed area of the cliff where they are nesting (Bertran and Margal- ida 1999). During the sexual activity period, all copula- tion attempts and related behavior were recorded. We noted the individuals involved in each copulation at- tempt and whether this was successful or not. A copula- tion was recorded as being successful behaviorally if clo- acal contact was observed. We noted whether the female accepted or refused a copulation attempt. The relative frequency of copulations was estimated daily as the num- ber of attempts hr f The temporal distribution of cop- ulations was examined in relation to the day on which incubation began (day 0). Laying was estimated with a maximum error of 5 d. Clutch size in this species is ha- bitually two eggs, which are laid at intervals of 3-7 d, and incubation starts as soon as the first egg is laid (Hiraldo et al. 1979, Brown 1990, pers. observ.). We quantified the individual contribution to the following parental tasks: nest building, territorial defense, incubation, brooding, and feeding of the chick. We recorded prey deliveries, chases against conspecifics or birds from other species (principally Griffon Vultures [Gyps falvus] and Common Ravens [ Corvus corax] , see Margalida and Bertran 2000a) , and deliveries of nest material. Data were analyzed with nonparametric tests because data were not normally dis- tributed (Sokal and Rohlf 1981). Values are presented as means ±SD. Results The trio showed an extended copulation period (114 d) . Similar to monogamous pairs (Bertran and Margalida 1999), the hrst copulation attempts were noted in No- vember and continued until the early stages of chick rear- ing in March. We observed a total of 123 copulation at- tempts (including both successful and aborted ones). Of these, the female copulated with one of the males in 74% of cases, whereas in the remaining 26%, the two males copulated with one another (homosexual copulations) . During the pre-laying period, male A’s copulation at- tempts with the female were more frequent than those of male B (male A: 0.33 ± 0.26, V = 40 copulation at- tempts h"^ vs. male B: 0.18 ± 0.21 copulation attempts h"^ N = 21; Mann-Whitney C-test, P = 0.047). Most cop- ulation attempts by male A (90%, N = 40) took place m the presence of male B, while male A witnessed male B’s copulation attempts in 48% of the cases (V = 21; Fisher exact test, P = 0.003). During 25% of 16 cases in which male A tried to copulate with the female while sharing the same perch with male B, the latter interfered aggres- sively in the copulation attempt; although on only one occasion did he succeed in interfering with the attempt. In contrast, male A did not demonstrate any aggressive- ness when he witnessed copulation attempts by male B The response of the female to the sexual advances of the two males differed substantially. Male A obtained suc- cessful copulations from the early attempts (54 d before laying) and these continued until 12 d after egg-laying took place. Successful copulations by this male made up 52.6% and 30% during pre-laying and after egg-laying periods, respectively (Fig. 1). In contrast, the female did not accept most copulations initiated by male B during the pre-laying stage: the male was rejected in 15 of 21 copulation attempts. However, successful copulations by male B were observed around the onset of incubation, and these continued for 24 d longer than those of male A. Copulations between male B and the female were per- formed when male A was at the nest. Both males displayed bisexual behavior, although the proportion of homosexual mounts attempted by male B was significantly greater (male B: 45.4% of 66 copulation attempts vs. male A: 3.5% of 57 copulation attempts; y} = 25.82, df = 1, P < 0.001). None of the observed at- tempts of homosexual mounts were forced or refused and 97% of these occurred in the presence of the female. On four occasions, male A was mounted immediately af- 68 Short Communications VoL. 36, No. 1 Table 1. Frequency and percentage of territorial de- fense attacks by individual group members observed dur- ing pre-laying and post-laying periods. Pre-iamng Post-laying Individuals (V = 49) {N= 131) Male A 32 (63.3%) 43 (32.8%)* Male B 3 (6.1%) 61 (46.6%)* Female 14 (28.6%) 27 (20%) * Significantly different {P < 0.05) in relation to preceding pe- riod based on Chi-square test. ter male B tried to interrupt the former’s mating attempt with the female. We observed only apparent cloacal con- tact (60% of cases, N = 10) when male B took the initia- tive for mating. Male B’s behaviorally successful copula- tions obtained from male A and from the female occurred over the same chronological period (Fig. 1). Significant differences were found in the material sup- plied to the nest in relation to the number of individual visits: male A and the female were mostly responsible for this task, delivering material during 67.6% and 58.5% of their 71 and 51 visits, respectively, as Male B only deliv- ered nesting material during 21.4% of 17 visits (x^ = 14.09, df = 2, P< 0.001). We registered a total of 180 instances of territorial de- fense that took place between October 1991 and April 1992. The majority of the attacks (66.1%) were directed at Griffon Vultures ( Gyps fulvus) and at Common Ravens {Corvus corax) (24.4%). The remainder (9.5%) were di- rected to other raptors and corvids. By periods, we de- tected significant differences related to contrasting ter- ritorial behavior shown by both males. In male A a higher than expected frequency of attacks was observed during the pre-laying period, contrary to that seen during post- laying. In contrast, male B increased his territorial be- havior after laying took place (P < 0.05, for both cases; Table 1). The daily investment of the three adults in relation to incubation of the clutch did not differ significantly (male A. 38.5 ± 22.7%; male B: 34.4 ± 22.3%; female: 27.1 ± 20.6%, Kruskal-Wallis H = 1.85, df = 2, P = 0.397). How- ever, the amount of time invested by male B at the nest during brooding was significantly less, both in 1991 (male A: 37.08 ± 18.14%; male B: 13.23 ± 14.36%; female 49.69 ± 20.68%; Kruskal-Wallis H = 35.919, df = 2, P < 0.001) and in 1992 (male A: 46.55 ± 14.25%; male B 19.73 ± 12.38%; female: 33.72 ± 14.96%; Kruskal-Wallis H = 25.238, df = 2, P< 0.001). The distribution of prey deliveries to the nest by the three individuals did not differ significantly from a theoretical distribution in which all individuals delivered the same number of items (X^ = 1.30, df = 2, P > 0.05; Table 2). However, signifi- cant differences were obtained when comparing the fre- quency of feedings with that expected from the individ- ual contributions (x^ = 14.97, df — 2, P < 0.001; Table 2) with male B feeding the chick less than the other two individuals (x^ = 14.12, df = 2, P < 0.001). Discussion This first detailed record of behavior of a reproductive trio of Bearded Vultures suggests a social structure con- sisting of an alpha pair and a subordinate male. Levels of aggressiveness were not high between the two males in the trio. However, some conflicts associated with sexual activity were observed, and these were always initiated by the presnmed subordinate male. Conflicts between the male birds were only recorded during heterosexual cop- ulation attempts, and these were always initiated by the subordinate male. A remarkable aspect of our results is the tolerance shown by the alpha male, and the fact that he accepted a number of mountings by the subordinate male. A similar behavior has been noted in two other Pyrenean groups (Margalida et al. 1997, unpubl. data) In other cooperative breeders, homosexual behavior is seen as a means of appeasement that would increase co- hesion within the group (Heg and van Treuren 1998). Cooperative polyandry in the Bearded Vulture may not provide reproductive compensations in the short term, but may in the long term when overall survival is consid- ered (Heredia and Donazar 1990). In a long-living spe- cies such as the Bearded Vulture, being in good condi- tion would permit a greater number of breeding attempts. This species has a long breeding cycle, during which the adult birds are committed to providing food and to chick-rearing, while they also have to obtain food for themselves (Margalida and Bertran 2000a). There- fore, some direct benefit for breeding pairs should be Table 2. Frequency and percentage of contributions of prey items delivered to the nest and feedings to the chick by individual group members. Individuals Prey Deliveries {N = 60) Expected Feedings (V= 108) Expected Male A 24 (40.0%) 20 49 (45.4%) 36 Male B 17 (28.3%) 20 13 (12.0%) 36 Female 19 (31.7%) 20 46 (42.6%) 36 March 2002 Short Communications 69 expected from the collaboration of helpers, because searching for food, which is spatially and temporarily un- predictable, remains a difficult task. By remaining in a territory, subordinate males might have a possibility of acquiring an alpha breeding position within the group. By using a helper strategy, B males would be in a position to inherit the territory and the female. This would be particularly important for birds in pre-adult plumage at breeding age, if suitable breeding territories are not avail- able (Heredia and Donazar 1990). Young birds might be capable of evaluating the individual quality of the differ- ent territories in the course of their dispersive move- ments (Bertran and Margalida 1996). Short-term direct benefits of the sexual relationship with the male birds seem more obvious for the females. This hypothesis is based on the fact that during the pre-laying stage, the female only permitted successful copulations by male A, whose total contribution to the breeding effort was larg- er. Yet, at the time of egg-laying, the female apparently accepted the secondary male furtively, and this coincided chronologically with the latter’s increased help contri- bution (principally territorial defense and provisioning of prey to the nest). In the Bearded Vulture, both sexes share parental duties, but females tend to participate more actively in the direct care of the offspring (Margal- ida and Bertran 2000a) . Thus, polyandrous females may benefit from the males’ investment: both males provide parental care and, consequently, the female has more time to fly around (see Birkhead 1998). Female Bearded Vultures may, because of their larger size, actively refuse males in unwanted copulation at- tempts (pers. observ.). There may be a relationship be- tween size, coloration, and dominance in this species. For instance, it has been suggested that coloring of the ven- tral area might act as a status signal (Negro et al. 1999). The coloring is acquired deliberately (Negro and Mar- galida 2000) and its intensity is directly related to the birds’ age and probably to their status as well. Females, which are slightly larger than males and may be behav- lorally dominant, show more intense coloring (Negro et al. 1999). In this trio, coloring intensity increased pro- gressively from the secondary male (markedly paler) to the female. A probable relationship between social rank and coloring intensity was also observed among other vul- ture groups as well (Negro et al. 1999). Our results suggest a defined hierarchical social rank within groups. In that context, females, due to their larg- er size and behavioral characteristics, would be able to exert social control within such groups. This would open new and interesting perspectives in the study of these social groups, which should be examined in larger sam- ples. Resumen. — El Quebrantahuesos {Gypaetus barbatus) oca- sionalmente forma grupos cooperativos (trios). Durante 1991 y 1992 examinamos la conducta reproductora de un trio Pirenaico (NE Espana). Nuestras observaciones indican que el grupo estuvo formado por una pareja alfa y un macho secundario. El trio exhibio un periodo inu- sualmente extenso de copulas donde la hembra copulo preferentemente con el macho alfa, pero tambien accep- to aparentemente de forma furtiva al macho secundario, en concreto tras la puesta. Eueron observadas algunas copulas macho-macho comportamentalmente exitosas, si bien el significado de estas es desconocido. Nuestros re- sultados sugieren que la hembra pudo beneficiarse a tra- ves de las copulas de la inversion parental suministrada por ambos machos. En este contexto, les hembras, de- bido a sus caracteristicas fisicas y comportamentales po- drian ejercer un control social dentro de estos grupos. [Traduccion de los autores] Acknowledgments We wish to thank J. Real for valuable advice during the analysis of the data. Helpful comments of different as- pects of this work have been provided by R. del Amo, J. Bertran, S. Manosa, and J. Moya. Thanks to J.C. Bednarz, J.A. Donazar, J.J. Negro, J. Vinuela, and an anonymous referee for their reviews of earlier versions of this man- uscript. In addition, we thank E. Aubarell, J. Canut, C Cuberes, and their families for the logistic support and their hospitality. Thanks are due to C. Carboneras and S Cahill for translating the manuscript into English. This study was supported by the Departament de Medi Am- bient of Generalitat de Catalunya. Literature Cited Arroyo, B. 1996. A possible case of polyandry in Mon- tagu’s Harrier, y. Raptor Res. 30:100-102. Bertran, J. and A. Margalida. 1996. Patron anual de observaciones de Quebrantahuesos {Gypaetus barba- tus) de diferentes grupos de edad en los sectores de nidificacion. Alauda 64:171-178. AND A. Margalida. 1999. Copulatory behavior of the Bearded Vulture. Cora2 hr; is wet from rain, as the female spends a good portion of the day away; afterwards, chick spreads its wings in the sun. Chick first seen to tear off and eat a few bites on its own, though with difficulty. 65 Chick jumps and flaps frequently; pounces on and seizes a stick (nest material) with its tal- ons. 72 Chick weighs 1630 g (ca. adult female mass); wing chord is 245 mm, total length 450 mm, and wingspan 870 mm; eyes are dark gray, feet buffy yellowish, and cere, facial skin, and beak black. 80 Chick often feeds on its own. March 2002 Short Communications 79 Table 1. Continued. Nestling Age (days) Development ai. Feature Firsi Noted on This Day 81 Adult female begins remaining away from nest most of time and periodically brings prey to the nest. Chick adept at plucking prey and feeding itself. Chick now weathers frequent heavy rains on its own. 92 On 28 August, chick equipped with a backpack radiotransmitter. Chick weighs 1697 g, wing chord is 400 mm, total length .590 mm, and wingspan 1140 mm. 93 Chick pouncing, seizing and pecking at nest material and prey remains and exercising wings more frequently and vigorously; commonly jumps, flapping, from rim to rim of the nest, and periodically tugs at the antenna of its radiotransmitter. When wet from rain, the nestling often spreads its wings in the sun. 100 Chick frequently moves nest material aronnd with its bill; practice hunting and flapping have reached a fever pitch. 114 On 19 September, nestling flies around within nest tree and to other trees; fledged at 109— 114 d of age (in nest at age 108 d). (1995) was near the center of the park. Tikal National Park is a tropical lowland site at 17°N latitude in Guate- mala’s Peten Department. Tikal’s environment is de- scribed by Schulze and Whitacre (1999). Because nest No. 1 failed prior to hatching, we report here on obser- vations at nest No. 2, which produced one fledgling. We discovered nest No. 2 on 8 May 1995, while the adults were incubating. The nest was 16.4 m high in a live Job- illo tree {Astronium graveolens). At a similar height in a tree 74 m from the nest, we built a wooden observation platform, using poles cut from the forest, a 1 .2 X 0.6 m piece of plywood, and baling wire. Construction of the platform took several hours during three days, from 8- 10 May. The Crested Eagles gave no indication of being disturbed by the construction process. On 1 1 May we began dawn-to-dusk observations, usu- ally 13-14 hr in duration, using a spotting scope. We Figure 1. Mass gain of hatchling male Crested Eagle (band number = 220107) in the Oklahoma City Zoolog- ical Park. Data from zoo records, courtesy of Barbara Howard. logged 979.3 hr of observation on 83 d; observation pe- riods averaged 11.8 hr in duration. We observed this nest during the last three weeks of the incubation period and until the chick fledged at 109-114 days of age. Using a backpack arrangement and 6 mm teflon ribbon, we placed an 18 g radiotransmitter (216 KHz; Holohil Sys- tems Ltd., Carp, Ontario) on the juvenile once it neared fledging age. We monitored the radio-tagged fledgling until 16 mo of age, when we ended fieldwork. Results The nestling, later judged by size to be a female, was not hatched as of 26 May and was very tiny and weak the morning of 29 May; we estimated she hatched 28 May, which was designated as day 0. The chick’s physical and behavioral development during the 114 d brood-rearing period are described in Table 1, Post-Ffedging Period. At first, the fledgling returned to the nest to sleep and receive prey, but within a few days it spent little time in the nest, flying frequently be- tween various trees up to 100 m from the nest, and re- turned to the nest mainly to receive prey from the adults At this time, the female often fed the chick, bill to bill. With radiotelemctry, we followed the fledgling to the age of nearly 16 mo, when she still remained dependent on the adults. Unlike some other raptor species at Tikal, the fledgling did not indulge in protracted food-begging while adults were away from the nest area, but rather, it called mainly when they approached. At least as late as day 141 (16 October), the female still occasionally fed the chick bit-by-bit when she delivered prey. Data on a Captive-reared Nestling. Because no growth curves are available for wild birds, here we report data for a male (surgically sexed) hatched at the Oklahoma City Zoo (Fig. 1). This male weighed 51 g on hatching day, showed essentially linear mass gain after two weeks, and stabilized at ca, 1117 g at age 42-53 d (Fig. 1). Al- lowing 5-10% additional mass gain thereafter (Newton 80 Short Communications VoL. 36, No. 1 1979:120, T. Cade pers. comm.), its predicted adult weight is ca. 1170-1230 g. At 21 days, this chick’s talons were beginning to turn from white to gray, and covert feathers were emerging from quills. At four weeks, pri- mary feathers were emerging from sheaths and at six weeks the chick was very alert and interested in his sur- roundings. On day 51 he stood in the nest, on day 72 he sat on the nest rim, and on day 76 he was found on the floor and returned to the nest. On day 94 the primaries were hard-penned, though the tail was still growing. On day 103, the chick flew across the room to another perch, and two days later was flying about the room (B. Howard pers. comm.). Dlscussion Developmental mileposts accord closely between the captive-reared and wild chicks. Both were able to stand for some time during their seventh week, and both made their first significant flights at 103-114 days. The Crested Eagle nestling at Tikal developed at a slower rate than did Ornate Hawk-Eagle {Spizaetus ornatus) nestlings (Whi- tacre et al. in press b), which in turn developed more slowly than did Black Hawk-Eagle nestlings {S. tyrannus; Whitacre et al. in press c). We observed chicks first stand- ing up for prolonged periods at 5 wk in the Black Hawk- Eagle, during the seventh week in the Ornate Hawk-Ea- gle, and at 8 wk in the Crested Eagle. Chicks first flapped m place vigorously at 4 wk in the Black Hawk-Eagle, 5 wk m the Ornate Hawk-Eagle, and during the seventh week m the Crested Eagle. The age at which we noted chicks first able to feed themselves was less variable; this oc- curred at 8 wk in the Black Hawk-Eagle and during the ninth week for both the Ornate Hawk-Eagle and Crested Eagle. A Black Hawk-Eagle chick brst walked out onto limbs near the nest early in the fourth week and did so commonly by the end of the fourth week. We observed this behavior at 9 wk in the Ornate Hawk-Eagle and 16 wk in the Crested Eagle. First flights within the nest tree were observed during the eighth week for the Black Hawk-Eagle, the tenth week in the Ornate Hawk-Eagle, and at 16 wk in the Crested Eagle. Fledging from the nest tree took place during the tenth week in the two hawk-eagles and at 16 wk in the Crested Eagle. The relative speed of development of the above three species accords with their relative body sizes. Black Hawk- Eagles (most rapid to develop) are also smallest, with mean adult female mass of 1115 g (Whitacre et al. in press c). Ornate Hawk-Eagles, with intermediate rate of development, are intermediate in size, with females av- eraging 1450 g (Whitacre et al. in press b). Crested Ea- gles, slowest to develop, have a mean adult female mass of about 1850 g (Whitacre et al. in press a). The nestling periods of these two Crested Eagles (103- 105 d for the captive bird and 109-114 d for the wild bird) are notably long for a raptor of this size. The Crest- ed Eagle falls well above the curve relating nestling pe- riod to female mass in a wide range of falconiforms (Newton 1979:119). Indeed, the nestling period we doc- umented is equivalent to that of the Philippine Eagle (Pi- thecophaga jefferyi) , Crowned Hawk-Eagle (Stephanoaetus co- ronatus), and Martial Eagle {Polemaetus bellicosus), all substantially larger birds than the Crested Eagle (Newton 1979:344). It is unclear why the Crested Eagle should have a nestling development period as long as these larg- er tropical eagles. Further data are needed in order to more confidently estimate the duration of the nestling period in Morphnus. Resumen. — Se observe dos nidos de Aguila Monera Morphnus guianensis en Peten, Guatemala. Reportamos sobre el desarrollo comportamiento de un polluelo. Tara- bien presentamos dates sobre el crecimiento y compor- tamiento de un juvenil en cautiverio. Lajuvenil silvestre, una hembra, eclosiono aproximadamente 28 Mayo, de- signado dia 0. La primera vez que observamos el polluelo arreglarse las plumas fue en dia 16; en dia 17 defeco hacia la orilla del nido, y siempre lo hizo asl despues. Dia 18 intento pararse por primera vez, y en dia 23 pudo pararse brevamente. En dia 24 estiro las alas de manera estereotipica. A partir del dia 25, respondio agresivamen- te cuando el macho trajo presa, y a especies tal como buitres. A partir del dia 37 el polluelo logro pararse por un buen rate, y batio las alas en ejercicio. En dia 59, logro alimentarse ella mismo por primera vez, y quedo parada por dos horas. Cuando el polluelo tenia 81 dias, la hem- bra adulta empezo a ausentarse la mayor parte del dia, supuestamente cazando, y trajo presas al nido periodi- camente. A partir del dia 93, el polluelo frecuentemente salto, aleteando, de una orilla del nido a la otra, mane- Jaba palitos del nido en su pico, y las atacaba con las garras como que fueran presa. Entre dias 109 y 114, void fuera del arbol del nido por primera vez, pero siguio regresando al arbol del nido para recibir presas. Usando radio-telemetria, seguimos lajuvenil hasta 16 meses de edad, cuando terminamos el estudio. En aquella fecha, la juvenil quedaba todavia dependiente en los adultos. Hacemos comparaciones entre la rapidez de desarrollo del Aguila Monera, el Aguilucho de Penacho {Spizaetus ornatus), y el Aguilucho Negro (S. tyrannus). El desarrollo del juvenil de Morphnus fue lento en comparacion con otros rapaces de igual tamano. [Traduccion de los autores] Ac;KNOWI T.DGMENTS This is an offering of The Peregrine Fund’s Maya Pro- ject; we are grateful to the many individuals and foun- dations that provided funding for that project. We are grateful to Barbara Howard for providing records for Crested Eagles at the Oklahoma City Zoological Park and to R. Bierregaard, J. Bednarz, and an anonymous review- er for helpful comments on the manuscript. Literature Cu ed Bierregaard, R.O., Jr. 1984. Observations of the nesting biology of the Guiana Crested Eagle {Morphnus guia- nensis). Wilson Bull. 96:1-5. March 2002 Short Communications 81 Newton, I. 1979. Population ecology of raptors. T. & A.D. Poyser, Berkhamsted, U.K. Schulze, M.D. and D.F. Whitacre. 1999. A classification and ordination of the tree community of Tikal Na- tional Park, Peten, Guatemala. Bull. Fla. Mus. Nat. Hist. 41:169-297. Starck, J.M. AND R.E. Ricklefs. 1998. Avian growth and development: evolution within the altricial-precocial spectrum. Oxford University Press, Oxford, U.K. Whitacre, D. R, J. Lopez Avila, and G. Lopez Avila. In press a. Crested Eagle {Morphnus guianensis). In D.F. Whitacre (Ed.), Raptors of the maya forest: ecology of a neotropical raptor community. Cornell University Press, Ithaca, NY U.S.A. , J.A. Madrid M., H.D. Madrid M., R. Cruz E., CJ Flatten, and S. Funes A. In press b. Ornate Hawk- Eagle {Spizaetus ornatus). In D. Whitacre (Ed.), Rap- tors of the maya forest: ecology of a neotropical raptor community. Cornell University Press, Ithaca, NY U.S.A. , J. Lopez A., G. Lopez Avila, S. Funes Aviia, CJ Flatten, and J.A. Madrid M. In press c. Black Hawk- Eagle {Spizaetus tyrannus). In D. Whitacre (Ed.), Rap- tors of the maya forest: ecology of a neotropical raptor community. Cornell University Press, Ithaca, NY U.S.A. Received 2 December 2000; accepted .SO October 2001 J. Raptor Res. 36(l):81-84 © 2002 The Raptor Research Foundation, Inc. Spring Weather and Breeding Success of the Eurasian Kestrel {Falco tinnunculus) IN Urban Rome, Italy Luca SalvatE Piazza F. Morosini 12, 1-00136 Rome, Italy Key Words: Eurasian Kestrel', Falco tinnunculus; breeding success] weather conditions] urban habitats] Mediterranean ar- eas. The breeding biology of the Eurasian Kestrel {Falco tin- nunculus) has been well-studied in northern and central Europe mainly focusing on the influence of prey fluctu- ations on clutch size and productivity (e.g.. Village 1990, Plesnik and Dusik 1994, Valkama et al. 1995) as well as on the influence of weather conditions on timing of breeding (Kostrzewa and Kostrzewa 1990, 1991). In Med- iterranean Europe, few studies addressed these aspects (Gil-Delgado et al. 1995) and relevant accounts on kes- trel breeding success are by Rizzo et al. (1993), Gil-Del- gado et al. (1995), Fargallo et al. (1996), and Aviles et al. (2000). Here, I provide data on the breeding success of kestrels in two different habitats of Rome, central Italy, through 5 yr. I studied between-year differences in breed- ing success in relation to spring weather and I compare my results with data collected 15 yr earlier from the same population (Sommani 1986). Study Area and Methods I conducted fieldwork from March 1996-July 2000 in Rome, Latium, central Italy (41°53'N, 12°28'E). The area IS characterized by developed areas, urban parks, open- ^ E-mail address: picoidesmajor@yahoo.com lands (mainly dry pastures and cereal crops) , and small oak woods (mainly Quercus ilex). The two census plots included one strictly urban area (inner city) and one sub- urban, built-up area (Appia Antica park) . Breeding den- sity was 1.9 pairs/km^ {N = 86 pairs) in the urban area and 0.6 pairs/km^ {N = 34 pairs) in the suburban area (Salvati et al. 1999). For census procedure to locate breeding pairs see Salvati et al. (1999, 2000). Nests were monitored weekly from the pre-incubation period. Visits were increased to 2-3 d intervals during the nesting period. Laying date for each nest was deter- mined by subtracting the mean incubation period of the species (28 d; Aviles et al. 2000) from the hatching date Hatching date was determined taking into account that all eggs hatch in 4 d (Aviles et al. 2000). Fledging date was defined as the first day when all fledglings leave the nest. Young generally stay for 5-10 d in the vicinity of the nest using perches previously frequented by the parents, but rarely come back to the nest during daylight (Komen and Myer 1989, Bustamante 1994). As the interval of nest visits was 2-3 d, an error of ±1 d should be assigned to fledging date. Clutch size and laying date were recorded for a subsample of breeding pairs, because many nests were inaccessible for an exact count of eggs or chicks during the early stages of breeding (Salvati et al. 1999). I measured percent egg productivity as the number of fledglings in a nest divided by the total number of eggs laid in that nest. Breeding parameters for the years 1979— 85 were obtained through the same technique from Som- mani (1986). As weather variables, 1 used mean monthly rainfall and 82 Short Communications VoL. 36, No. 1 Table 1. Breeding success and fledgling dales of Eurasian Kestrels in Rome (1979-85, 1996-2000). Year Fledglincls per Breeding Pair N Mean Fi.edging Date (1 = 1 June) N Seasonal Decline in No. of Fildgings/Pair WTTH DATE"^’^ 1979-85'^ 2.98 ± 1.10 40 16.53 ± 6.41 38 — 1996 3.08 ± 0.57 25 13.28 ± 4.54 11 r = -0.61, df = 9* 1997 3.07 ± 0.83 27 15.43 ± 6.93 23 r = -0.44, df = 21* 1998 3.06 ± 1.01 36 23.97 ± 8.33 33 r = -0.38, df = 31* 1999 3.06 ± 1.00 46 27.11 ± 11.92 45 r = -0.69, df = 43** 2000 2.98 ± 0.84 54 28.28 ± 10.43 54 r = -0.63, df = 52** 1996-2000 3.04 ± 0.88 188 24.43 ± 10.84 166 r = -0.56, df = 154** Pearson correlation between fledging date and number of fledglings per pair (probability levels; * P < 0.05, ** P < 0.001). Fledging dates were log-transformed to obtain a normal distribution. Data recalculated from Sommani (1986). mean minimum temperatures in March-June from 1979—85 and from 1996-2000. There were no correla- tions between monthly rainfall and minimum tempera- tures (Spearman’s rank correlation test, P> 0.1). All me- teorological data were obtained from the station of Collegio Romano, which is located within the study area. All breeding parameters were compared by means of t- tests and analysis of variance (ANOVA) in order to inves- tigate differences among years. Logarithmic transforma- tion was performed on fledging date to correct for deviations from normality. Relationships between each weather variable and breeding parameters were examined using Spearman’s rank correlation analysis. In all tests (two-tailed), a minimum probability level of P< 0.05 was accepted. Statistical analyses were performed using STA- TISTICA software. Results are presented as mean ±SD. Results Mean clutch size was 4.43 ± 0.94 eggs (range = 3—6 eggs, N = 14), and mean laying date was 22.15 April ± 13.13 d (N = 13). Mean number of fledglings per suc- cessful pair was 3.08 ± 0.82 (range = 2-6 young, N = 186), and mean fledging date was 24.43 June ± 10.84 d (N = 166). Percent egg productivity was 80.6% (N = 14). As two pairs failed breeding (percentage of successful pairs = 98.9%, N= 188), mean number of fledglings per breeding pair was 3.04 ± 0.88 (N = 188). Seasonal de- cline in number of fledglings per pair was observed in all years of study (Table 1). I did not detect between-year differences in mean number of fledglings per breeding pair (^4 IH 3 = 0.09, P = 0.985), whereas mean fledging date showed significant differences among years (^ij ,q] = 6 78, P 0*0001). Pooling all data collected m urban Rome from 1979-85 and from 1996-2000, the difference m mean number of fledglings per breeding pair was not signihcant (%26 “ —0.37, P = 0.709). Pairwise correla- tions between mean fledging date and all weather vari- ables were not significant {P> 0.1), with only April rain- fall bordering the significance level (r, = 0.56, P = 0.058, N = 12). Discussion Breeding parameters of kestrels in Rome are similar to those observed in other Mediterranean areas (Rizzo et al. 1993, Gil-Delgado et al. 1995, Aviles et al. 2000). Since my study was restricted to pairs breeding in cavities other than nest-boxes, the overall reproductive success may be lower compared with that of populations breeding in nest-boxes, likely due to the potentially high predation rates (Aviles et al. 2000) . However, the number of kestrel predators is generally low in cities, thus reducing the probability of nest predation. Many authors have reported that large annual varia- tions in kestrel productivity are linked to fluctuations in rodent density and to unfavorable weather conditions during breeding in northern and central Europe (Kos- trzewa and Kostrzewa 1990, 1991, Village 1990, Valkama et al. 1995). In southern Europe, some studies have shown slight between-year differences in laying date (Avi- les et al. 2000) and number of fledglings per pair (Gil- Delgado et al. 1995). Interestingly, voles generally show small fluctuations in density in the Mediterranean basin (Paradis and Cmedon 1993, Rizzo et al. 1993). In these areas kestrels usually feed on alternative prey (Rizzo et al. 1993, Gil-Delgado et al. 1995, Piattella et al. 1999), thus reducing the influence of rodent fluctuations on productivity. Therefore, in southern Europe weather con- ditions may assume a role in determining annual varia- tions in kestrel breeding success. Mild climate (i.e., high temperatures and low rainfall) during spring triggers laying and favors chick rearing (Gil-Delgado et al. 1995). In Rome, the between-year sta- bility of kestrel nesting success may confirm the impor- tance of mild and stable weather conditions during breeding, as already observed for the Tawny Owl {Strix aluco) in the same area (Ranazzi et al. 2000), although it seems plausible that in the Mediterranean basin, high rainfall during spring may force kestrels to delay laying, as suggested by the weak relationship between mean fledging date and April rainfall. I suggest that the annual March 2002 Short Communications 83 variation in mean fledging date is probably not affected by weather conditions. Factors linked to high population levels in Rome (Salvati et al. 1999) more likely could af- fect variations in laying date. Following Aviles (2000) hypothesis, in dry and semi- arid landscapes of southern Europe, high rainfall in the spring may result in higher prey availability (e.g., insects; Rizzo et al. 1993, Aviles et al. 2000), contributing to an- nual fluctuations in breeding success. Although data from this study area do not support Aviles and co-work- ers’ hypothesis, 1 suggest that the improved reproductive output related to rainfall also depends on the feeding habits of kestrels in each area. In those Mediterranean areas, like Rome, where kestrels prey mainly on birds (Piattella et al. 1999), the effect of spring rainfall could be negligible. Whereas in most arid European areas, where kestrels feed mainly on insects (Gil-Delgado et al. 1995), the Aviles and co-workers’ hypothesis seems to be plausible to explain the between-year differences in breeding performance of this raptor. Resumen. — Estudie las diferencias entre ahos en el exito reproductivo del cernicalo euroasiatico (Falco tinnuncu- lus) en relacion con el clima primaveral en un periodo de cinco anos, y compare mis resultados con datos colec- tados 15 anos atras en la mi misma poblacion. El tamano medio de la postura fue 4.43 ± 0.94 huevos, y la fecha media de la postura fue 22.15 Abril ± 13.13 d. El pro- medio de volantones por pareja exitosa fue 3.08 ± 0.82 y la fecha promedio del primer vuelo fue 24.43 Junio ± 10.84 d. La productividad de huevos fue 80.6%. La di- ferencia entre anos en el promedio de volantones por pareja reproductora no fue significante, en tanto que la fecha promedio del primer vuelo mostro variaciones sig- nificativas entre ahos. No hubo correlacion entre la fecha promedio del primer vuelo y las variables climaticas. El exito reproductivo de los cernicalos en Roma es com- parable a los observados en otras areas del mediterraneo. El clima moderado durante la primavera puede ser el detonador de las posturas y favorecer la cria de los pi- chones. Por otro lado, en paisajes semi aridos, donde los cernicalos se alimentan principalmente de insectos, la alta precipitacion en primavera puede dar como resul- tado la alta disponibilidad de presas, contribuyendo a las fluctuaciones anuales en el exito reproductivo. Los datos de esta area de estudio, donde los cernicalos comen pri- mordialmente aves, no dan soporte a esta hipotesis. Sin embargo esta hipotesis puede explicar las variaciones anuales en el exito reproductivo de las rapaces insecti- voras en algunas areas del mediterraneo. [Traduccion de Cesar Marquez] Acknowledgments Help in field work from A. Manganaro and L. Ranazzi was essential for the author. Thanks are also due to C. Beltrano for providing meteorological data. Literature Cited Aviles, J.M.,J.M. Sanchez, and A. Sanchez. 2000. Breed- ing biology of the Eurasian Kestrel in the steppes of southwestern Spain./. Raptor Res. 34:45-48. Bustamante, J. 1994. Behavior of colonial Common Kes- trels {Falco tinnunculus) during the post-fledging de- pendence period in southwestern Spain./. Raptor Res 28:79-83. Fargallo, J.A., C. Blanco, and E. Soto-Largo. 1996 Possible second clutches in a Mediterranean montane population of the Eurasian Kestrel {Falco tinnunculus) f. Raptor Res. 30:70—73. Cil-Dei gado, J.A., J.A. Verdfjo, and E. Barba. 1995. Nes- tling diet and fledgling production of Eurasian Kes- trels {Falco tinnunculus) in eastern Spain./. Raptor Res. 29:240-244. Komen, J. and E. Myer. 1989. Observations on post-fledg- ing dependence of Kestrels {Falco tinnunculus rupico- lus) in an urban environment./. Raptor Res. 23:94—98. Kostrzewa, a. and R. Kostrzewa. 1990. The relationship of spring and summer weather with density and breeding performance of the Buzzard Buteo buteo, Coshawk Accipiter gentilis and Kestrel Falco tinnunculus Ibis 132:550-559. Kostrzewa, R. and A. Kostrzewa. 1991. Winter weather, spring and summer density, and subsequent breeding success of Eurasian Kestrels, Common Buzzards and Northern Coshawks. Auk 108:342-347. Paradis, E. and C. Cuedon. 1993. Demography of a Med- iterranean microtine: the Mediterranean pine vole, Microtus duodecimcostatus. Oecologia 95:47-53. Piattella, E., L. Salvati, A. Manganaro, and S. Fatio- rini. 1999. Spatial and temporal variations in the diet of the Kestrel {Falco tinnunculus) in urban Rome / Raptor Res. 33:172—175. Plesnik, j. and M. DusIk. 1994. Reproductive output of the Kestrel Falcon Falco tinnunculus in relation to small mammal dynamics in intensively cultivated farmland. Pages 61-65 in Meyburg B.-U. and R D. Chancellor [Eds.], Raptor conservation today. WWCBP/The Pica Press, London, U.K. Ranazzi, L., A. Manganaro, and L. Salvati. 2000. The breeding success of Tawny Owls {Strix aluco) in a Med- iterranean area: a long-term study in urban Rome / Raptor Res. 34:322-326. Rizzo, M.C., L. Migliore, and B. Massa. 1993. Insects, small mammals and breeding performance of farm- land populations of the Common Kestrel {Falco tin- nunculus). Pages 11—18 mM.K. Nicholls and R. Clarke [Eds.] , Biology and conservation of small falcons. Pro- ceedings of the 1991 Hawk and Owl Trust Confer- ence, The Hawk and Owl Trust, London, U.K. Saiaati, L., a. Manganaro, S. Fattorini, and E. Piat- tella. 1999. Population features of Kestrels Falco tin- nunculus in urban, suburban and rural areas in cen- tral Italy. Acta Ornithol. 34:53—58. 84 Short Communications VoL. 36, No. 1 , , AND . 2000. Responsiveness of nesting Eurasian Kestrels Falco tinnunculus to call play- backs. /. Raptor Res. 34:319-321. SOMMANi, E. 1986. Note sulla biologia di alcune coppie di Gheppio, Falco tinnunculus, present! in Roma. Riv. Ital. Ornitol. 56:40-52. Valkama, J., E. Korpimaki, and R Tolonen. 1995. Habi- tat utilization, diet and reproductive success in the kestrel in a temporally and spatially heterogeneous environment. Ornis Fenn. 72:49—61. Village, A. 1990. The kestrel. T. & A.D. Poyser, London, U.K. Received 27 October 2000; accepted 3 September 2001 J Raptor Res. 36(l):84-86 © 2002 The Raptor Research Foundation, Inc. Fatal Caryospora Infection in a Free-living Juvenile Eurasian Kestrel {Falco tinnunculus) Oliver Krone^ Institute for Zoo and Wildlife Research, P.O. Box 601103, D-10252 Berlin, Germany Key Words: coccidiosis; Caryospora kutzeri; Protozoa; Eur- asian Kestrel; Falco tinnunculus. Infections in birds of prey by Caryospora spp. are a com- mon and often serious problem in captive breeding sta- tions (Heidenreich 1996). In British breeding centers, nestlings of Merlins {Falco columbarius) fall ill due to in- fections with C. neofalconis (Forbes and Simpson 1997). Typically, symptoms are displayed at an age of 28 to 55 d and may include regurgitation, hemorrhagic feces, de- pression, and reduced appetite. Peracute or acute death with or without clinical signs may occur also. The possible explanation for this disease in young birds at this partic- ular age is waned maternal immunity and incomplete de- velopment of their own active immunity. In experimen- tally-infected adult Eurasian Kestrels (F tinnunculus) the prepatency of C. neofalconis was 8-10 d and the patency 10-93 d, and for C. kutzeri 8-13 d and 4-34 d, respectively (Boer 1982). The developmental cycle can be either di- rect or indirect, In captivity caryosporans apparently uti- lize the direct life cycle, possibly also using paratenic hosts, such as earthworms (Heidenreich 1996); free-liv- ing birds of prey acquire infection by feeding on infected prey (Cawthorn and Stockdale 1982). Previously, 16 species of Caryospora have been de- scribed in raptors, 10 from birds in Europe and six from North America, including one which also occurs in Ven- ezuela (Upton et al. 1990, Kltth 1994). Because most pub- lications on Caryospora (Yamikoff and Matscboulsky 1936, Wetzel and Enigk 1937, Schellner and Rodler 1971, Boer 1982) consider only captive birds of prey, tbe distribution and significance of Caryospora in free-living birds of prey ^ E-mail address: krone@izw-berlin.de remains unclear. No Caryospora oocyst could be found in 72 free-living Merlins examined in Great Britain (Forbes and Fox 2000), nor in 247 birds of prey (including 35 Eurasian Kestrels, four Hobbys [it subbuteo] and 22 Per- egrine Falcons [F. peregrinus] from Germany [Krone 1998]). However, C. hoeri was found in seven of 15 free- living Eurasian Kestrels from Germany in another study (Kliih 1994). Furthermore, in free-living Eurasian Kes- trels from Austria, oocysts of C. falconis and oocysts of Caryospora spp. were diagnosed (Kutzer et al. 1980). Case Report A juvenile Eurasian Kestrel observed in Berlin on 29 August 2000 showed distinct signs of a general weakness. The bird was conspicuous, it demonstrated a reduced- flight distance, and when chased away, the bird flew only short distances. On 30 August 2000 the kestrel was cap- tured and a hemorrhagic diarrhea was reported. On the morning of 31 August 2000 the kestrel died. Post-mortem findings indicated a heavy protozoan infection which lead to death from associated severe dehydration and ca- chexia. As the bird was banded its history was known. The bird and its clutch mates had been banded on 11 July 2000 at an age of 18-19 d in a nesting box. The bird was found less than 500 m away from the nesting box. The necropsy of the 69-d-old male Eurasian Kestrel revealed a poor condition and a mass of 101 g. A heavy Caryospora spp infection (Fig. 1) resulting in a severe hemorrhagic en- teritis was documented during the examination of the digestive tract. The highest level of oocysts (ca. 100/visual field at magnification of 200X) were detected in the first third of the jejuno-ileum. Oocysts were mixed with potassium dichromate solu- March 2002 Short Communications 85 tion and spomlated in Petri dishes at room temperature (22-24°C) within three days. Discussion The clinical signs including the cachexia are similar to those described in captive-bred birds, in which the dis- ease is well-known. This is the first record of a fatal car- yosporan infection in a free-living bird of prey species in Europe. In his survey, Cawthorn (1993) did not identify any cases of a fatal coccidiosis in a raptor, but stated that clinical coccidiosis is uncommon in free-living birds of prey. Measurements of oocysts and sporocysts (Table 1) were within the range for C. kutzeri given by Boer (1982). Due to their round oocysts, C. falconis and C. boeii were excluded as potential candidates. C. megafalconis is too large and C. neofalconis too small to be considered as can- Table 1. Measurements of Caryospora kutzeri. Length (pm) Width (pm) Oocysts {N = 15 ) X 37.58 32.54 SD 3.07 1.77 Range 32 . 5 - 43.13 30 . 0 - 35.0 Length-width-ratio 1.15 Sporocysts {N = 15 ) X 24.17 21.96 SD 2.45 2.66 Range 18 . 75 - 28.75 18 . 75 - 28.75 Length-width-ratio 1.1 Residual body {N = 3 ) X 10.8 10.8 didates in this case. C. henryaevjere described with a triple oocyst wall. C. kutzeri is a specific parasite of the genus Falco as determined by cross infection experiments with the genera Buteo, Accipiter, Milvus, Bubo and Asio where transmission was unsuccessful (Boer 1982). Species iden- tification solely based on measurements of oocysts and sporocysts is difficult and uncertain. Five of 10 Caryospora species from Europe are still described inadequately or have only been found on one occasion. More reliable identification would be derived from cross-transmission experiments and a comparison using molecular genetics The extremely low prevalence of species of Caryospora in free-living birds of prey from central Europe contrasts with the high prevalence found in captive birds of prey. Boer (1982) diagnosed a prevalence of 9% of caryospor- an oocysts in fecal samples of captive Falconiformes {N = 628) from Germany. In contrast to Caryospora, oocysts of Sarcocystis/Frenkelia spp, occur at much higher preva- lence (31.4%, N = 194) in free-living birds of prey from Germany (Krone 1998). One reason could be that Car- yospora spp. are found predominantly in young birds at an age of 28-55 d and these birds are examined less fre- quently than older ones in the wild. Another explanation could be the geographic distribution of the protozoa The parasite may not occur naturally in central Europe, but may have been introduced with birds used for fal- conry (mainly Saker Falcons [T cherru^). This would im- ply that the parasite has its natural nidus in the distri- bution range of the Saker Falcon and its appropriate intermediate host. If an infected bird is imported into a breeding station, the parasite can infect many other fal- cons either due to its direct developmental cycle or pos- 86 Short Communications VoL. 36, No. 1 sibly via an intermediate host. Cawthorn and Stockdale (1982) have demonstrated that mice {Mus musculus) can act as experimental intermediate hosts for C. bubonis from the Great Horned Owl {Bubo virginianus) in North America. Nothing is known about possible intermediate hosts of Caryospora spp. in Europe. Currently, we do not know whether endemic areas ex- ist around captive breeding facilities where falcons are frequently exchanged, with a high possibility for caryo- sporan infections to be introduced to wild birds. In cap- tive-breeding programs in which birds of prey, especially of the genus Falco, are planned to be reintroduced into the wild, birds should be checked for Caryospora spp. be- fore release. Rfsumen. — Una infeccion cariosporica fatal fue diagnos- ticada en un macho inmaduro de cernicalo euroasiatico {Falco tinnunculus) encontrado moribundo en Berlin, Alemania. Las sehales clinicas de esta enfermedad coc- cidial incluyen diarrea hemorragica, depresion, y una dis- tancia de vuelo reducida. El ave murio debido a la alta infeccion de Caryospora kutzeri dando como resultado des- hidratacion y caquexia. En aves de presa silvestres en Eu- ropa Caryospora spp. no ha sido registrada, mientras que las infecciones en aves cautivas son comunes. [Traduccion de Cesar Marquez] Acknowledgments I am grateful to A. Hallau for submitting the carcass and to S. Kupko and K. Koch who provided the life his- tory of the falcon. I would like to thank N. Forbes and H. Hofer for reading the manuscript and for their help- ful comments, and the two reviewers P.T. Redig and R.J. Cawthorn for their suggestions improving the quality of the manuscript. Excellent technical assistance was provid- ed by K. Ernst. Literai ure Cited Boer, B. 1982. Untersuchungen iiber das Vorkommen von Kokzidien bei Greifvogeln und die Entwicklung von zwei Caryospora-hrX.e:n der Falken {Caryospora neo- falconis n. sp. und Caryospora kutzeri n. sp.). Vet. Med. Diss. Hannover, Germany. Cawthorn, R.J. 1993. Cyst-forming coccidia of raptors: Significant pathogens or not? Pages 14-20 m P.T. Re- dig, J.E. Cooper, J.D. Remple, and B. Hunter [Ens.], Raptor Biomedicine. University of Minnesota Press, Minneapolis, U.S.A. AND P.H.G. Stockdale. 1982. The developmental cycle of Caryospora bubonis in the Great Horned Owl, Bubo virginianus. Can. J. Zool. 60:152—157. Forbes, N.A. and G.N. Simpson. 1997. Caryospora neofal- conis: an emerging threat to captive-bred raptors in the United Kingdom./. Avian Med. Surg. 11:110—114. AND M.T. Fox. 2000. Control of endemic Caryos- pora species infestation of captive raptors. Pages 173- 179 in Proceedings of the Annual Conference of the Association of Avian Veterinarians, Portland, OR U.S.A. Heidenreich, M. 1996. GreifVogel. Krankheiten, Hal- tung, Zucht. Blackwell Wissenschafts-Verlag, Berlin, Germany. Ki.uh, P.N. 1994. Untersuchungen zur Therapie und Prophylaxe der CaryoVora-Infektion der Falken (Fal- coniformes: Falconidae) mit Toltrazuril sowie die Beschreibung von zwei neuen Caryospora-A.rto^n der Falken (C. megafalconis n. sp. und C. boeri n. sp.) Vet. Med. Diss. Hannover, Germany. Krone, O. 1998. Endoparasiten (Faunistik, Epizootiolo- gie, Pathogenitat) bei wildlebenden Greifvogeln aus drei verschiedenen Gebieten Deutschlands. Vet. Med. Diss. Berlin, Germany. Kutzer, E., H. Frey, and J. Kotremba, 1980. Zur Parasi- tenfauna osterreichischer Greifvogel (Falconiformes). Angew. Parasitol. 21:183-205. ScHELLNER, H.P. AND S. Rodler. 1971. Ein Trematoden- und Kokzidienbefall bei einem Luggerfalken {Falco jugger). Jb. Dt. Falkenorden 1970/71:90-93. Upton, S.J., T.W. Campbell, M. Weigel, and R.D. Mc- Kown. 1990. The Eimeriidae (Apicomplexa) of rap- tors: review of the literature and description of new species of the genera Caryospora and Eimeria. Can. J. Zool. 68:1256-1265. Wetzel, R. and K. Enigk. 1937. Caryospora falconis n. sp. (Eimeridea) aus dem Wanderfalken. SB Ges. naturf. Freunde Berlin v. 19. Januar:6-9. Yamikoff, W.L. and S.N. Matschoulsky. 1936. The coc- cidian of birds of prey./. R. Alicrosc. Soc. 56:372-375. Received 15 February 2001; accepted 6 June 2001 Associate Editor: Ian Warkentin BOOK REVIEW /. Raptor Res. 36(l):87-88 © 2002 The Raptor Research Foundation, Inc. The Raptor Almanac. By Scott Weidensaul. 2000. The Lyons Press, New York, NY. ix + .882 pp., nu- merous color photographs, figures, and tables. ISBN 1-58574-170-1. Cloth, $40.00.— Scott Weiden- saul has assembled an impressive compendium of facts about raptors in this attractive and compre- hensive volume. Designed “for the birder or nat- uralist who wants to go beyond the fundamentals,” the almanac includes information about evolution, behavior, migration, conservation, and just about every other topic related to raptors. The book is packed with amusing anecdotes and interesting trivia that one would be hard-pressed to find any- where else in a single volume. The well-written text is accompanied by numerous charts, graphs, and an excellent collection of stunning photographs taken by several well-known wildlife photogra- phers. I found that the book had answers to almost all the questions that nonbiologists typically ask about diurnal raptors: How big is a Bald Eagle? How long does an Osprey live? How fast does a Peregrine Falcon fly? How can I build a kestrel nest box? The title, “Raptor Almanac,” is somewhat mis- leading, however, because Weidensaul does not dis- cuss owls (raptors by anyone’s definition), but he does discuss New World vultures (now classed as ciconiiforms) . The subtitle on the cover more ac- curately describes the book as “A Comprehensive Guide to Eagles, Hawks, Falcons, and Vultures.” I am puzzled as to why the author chose to include New World vultures as “raptors” now that they have been reclassified. Weidensaul actually pre- sents the latest DNA evidence that they are not rap- tors, but he then proceeds to treat them as if they were. The book seems to devote a disproportionate amount of space to the New World vultures; a high percentage of text and figures is about them, often because they are exceptions to so many “rules” about “true raptors.” Although worldwide in scope, the book had a decided emphasis on taxa from North America and Europe. Raptors from Af- rica, Asia, and Australia were mentioned through- out the text but were noticeably underrepresented in the charts and tables. The North American bias was reflected by the fact that some (not all) mea- surements were given in English rather than metric units. The book is divided into four main sections. The introductory section basically defines raptors and describes anatomical and physiological features that make them unique. The section on ecology and natural history comprises the heart of the vol- ume (>120 pages) and covers behavior, social structure, courtship, nesting, diet, migration, lon- gevity, and mortality. A chapter on conservation in- cludes case histories of several endangered raptors and a review of effective management techniques. A final chapter on raptor-human relationships pre- sents a unique collection of information about how human religious and economic perceptions about raptors have changed over time; this last section has an interesting list of biblical references to rap- tors as well as a number of suggestions on how people can help raptors. Appendix I has nine ta- bles with interesting etymological information about raptors, and Appendix II provides a compre- hensive reference for English and scientific names as well as general breeding distribution of all 310 species of diurnal “raptors.” The layout of the book is pleasing, with attractive headings. Many handy (but unnumbered) charts, maps, and tables break up the text. It would have been helpful to have a list of tables and figures for easy reference. Several of the illustrations were pre- pared by the multitalented author, himself; others were adapted from other sources. Some of these “adaptations” did not go smoothly, as seen by the unfortunate whiteout smudge in the drawing on page 66. An error like this stands out against the highly professional quality of the other figures and the outstanding quality of the color photographs. I spotted only one typographical error (in the chart on p. 109) and one grammatical error (bot- tom of p. 153) . It is usually easy to find fault with an account designed for a popular audience, but I found al- most no exaggerations or erroneous statements in The Raptor Almanac. Weidensaul avoids restating popular myths. For example, his section on mate 87 88 Book Review VoL. 36, No. 1 fidelity is accurate and balanced and makes it clear that raptors do not always “mate for life.” In fact, most of what Weidensaul says is on target. As a scientist, my biggest disappointment with The Raptor Almanac was that Weidensaul did not provide references for the statements he makes. I realize that this practice makes it easier for a non- scientist to read the text, but it is very frustrating for the serious biologist who wants to verify state- ments or to get more information about a specific topic. Weidensaul provides only a very incomplete “Selected Bibliography” at the end of the book. A more effective compromise would have been to in- clude a list of references after each chapter. Some of the material in the charts and tables is refer- enced; these references suggest that Weidensaul re- lied heavily on the secondary literature (e.g., Brown and Amadon 1968, Palmer 1988, Johnsgard 1990) for facts. This practice works well 95% of the time, but doing so exclusively means missing out on some of the more recent and important studies. Unfortunately, it also can perpetuate unsubstanti- ated information. For example, Weidensaul men- tions the notion that the sex of Golden Eagles can be determined by tail bands, even though more recent work has shown this technique to be invalid. Readers may be disappointed to learn that the almanac, with a publication date of 2000, does not include the most up-to-date information on many topics. Data on California Condor chick produc- tion, number of occupied Bald Eagle territories, and number of Peregrine Falcon nesting pairs have not been updated since the early to mid- 1990s. Information on the legal status of some spe- cies, like the Bald Eagle and the Peregrine Falcon, is no longer accurate. When Weidensaul stated (p. 242) that experimental reintroductions of Golden Eagles had been unsuccessful, he apparently was unaware of recent successful nesting by released birds in Georgia and Tennessee. Weidensaul relies on very old sources for several important topics. For example, he features Craighead and Craighead 1956 in his discussions of diet and home range, and uses Bent 1937 for his data on chronology. Despite these shortcomings, WeidensauFs book will serve as a useful reference for anyone who wants or needs to know about raptors. Weidensaul deserves credit for taking on and successfully com- pleting an enormous job. I would recommend The Raptor Almanac as a textbook for introductory courses about raptor biology. The beautiful pho- tographs, by themselves, would make this book wel- come on any coffee table. — Karen Steenhof, USGS Forest and Rangeland Ecosystem Science Center, Snake River Field Station, 970 Lusk Street, Boise, ID 83706 U.S.A. J Raptor Res. 36(1) :89 © 2002 The Raptor Research Foundation, Inc. Manuscript Referees The following people reviewed manuscripts for the Journal of Raptor Research in 2001. 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 arc indicated with an asterisk. L. Alterman, D. Amadon, T. Augspurger*, L. Ayers*, A. Baker, J. Barclay, G. Barrowclough, J. Bates, J. Bednarz*, J. BelthofP, J, Berkelman, R. Bierregaard*, K. Bildstein*, D. Bird, W. Boarman, J. Bollmer, T. Bosakowski, E. Botelho, B. Bower, N. Brown*, J. Bustamante, T. Cade*, M. Donaghey Cannon*, R. Cawthorn, J. Cely, D. Centili, W. Clark*, D. Collister, C. Conway*, P. Cordero, J. Coulson, S. Crocoll, P. Dare, D. D’Auria, R. Davies, J. Dawson, K. DeSmet*, M. Desmond*, T. De Vries*, J. Donazar*, J. Duncan, D. Ellis*, J. Enderson*, R. Rodriguez Estrella, D. Evans, C. Finley, L. Flake, D. Forrester, G. Foster, A. Fowler, J. Fraser, M. Fuller*, FI. Garner, F. Gehlbach, J. Gehring, R. Gerhardt, J Gervais, S. Gillihan, R. Glinski, L. Goodrich, K. Grandison, T. Grubb, B. Hanbidge, A. Harmata*, D. FFarvey, K Hasselblad, G. Hayward, G. Holroyd*, D. Holt*, J. Hoth, D. Houston, S. Houston, B. Howe, M. Huff, S. Hutchings*, E. Inigo-Elias, E. Jaksic*, P. James, A. Jenkins, P. Jenny, B. Johnson*, R. Johnson, S. Jones, A. Kemp, P. Kennedy, P. Kerlinger, T. ICimmel, D. Kirk, N. Kjellen, N. Korfanta, A. Kostrzewa, O. Krueger, D. Krueper, B. Lehman*, S. L.erich, D. Lieske, J. Lincer*, L. Locke, M. Louette, D. Low, S. Lutz, T. Mabee, W. Mannan, J. Marks, M. Martell*, C. Marti*, J. Marzluff, M. McGrady, D. McICinnon, R. Mesta*, B. Millsap*, P. Mineau, M. Mitchell, J. Morrison, M. Mossman, H Mueller*, P. Mundy*, R. Murphy*, R. Nelson, B. Olenick, J. Pagel, J. Parker, S. Patla, E. Pavez, V. Penteriani*, D Plumpton*, S. Porter, D. Prescott, P. Radley, C. Rains, J.-L. Rangel-Salazar, E. Rave*, P. Redig, L. Rejt, M. Restani*, B Ritchie, G. Ritchison, P. Enriquez Rocha, D. Rosenberg*, R. Rosenfeld, M. Rowe, M. Ruane, L. Salvati, G. Santolo*, J Schmutz*, S. Schmutz, N. Seavy, C. Shackelford, J. Sharp, S. Sherrod*, D. Shyry*, J. Sidle*, J. Simonetti, B. Smith*, J Smith, R. Spaulding*, M. Stalmaster, K. SteenhoP, R. Steidl, A. Stewart, L. Takats, S. Talbot, J. Telia*, J.-M. Thiollay*, R. Thorstrom, L. Todd*, E. Toyne, C. Trost, L. Trulio, E. Tseng, V. Vanegas, D. Varland, A. Village, J. Walters, B Walton, T. Wellicome*, K. Wiebe, C. White, S. Wiemeyer*, R. Yosef, B. Zink, P. Zwank*. A Telemetry Receiver Designed with The Researcher in Mind What you've been waiting for! finally, a highly iensitivc 999 channel synthesized teicmetry receiver that weighs less than 13 ounces. Is completely user programmable and offers variable scan rates over all frequencies. For each animal being tracked, the large LCD display provides not only the frequerKy (to lOOHz) and channel number, but also a 7 character alphanumeric comment field and a digital signal strength meter. Stop carrying receivers that are the size of a lunch box or cost over SISOO. The features and performance of the new R-1000 pocket sized telemetry receiver will Impress you, and the prke will convince you. Other features Include: • Factory tuned to any 4MHz wide segment in the l4«-174MHz Band • Very high sensitivity of -liBdBm to •ISOdBm • Illuminated display and keypad for use in low light or darkrress • User selectable scan rates from 1-30 seconds in I second steps • Rechargeable batteries operate the receiver for 12 hours and can be replaced with standard AA Alkaline batteries in the field. Both 12vdc and llOvac chargers are included. • 6.1 2.6 1.5 (15.5cm) high, (6.6cm) wride. (3.8cm) deep. • 3 year warranty • 1 day delivery S695.00 Please specify desired 4MHz wide segment in the 148-174MHZ band Visit our website for complete specifications, operating manual artd information on the R-1000 or call our toll-free rMjmber to order your receiver now. Try the New R-1000 and You'll Be Impressed! COMMUNICATIONS SPECIALISTS. INC. 426 West Taft Avenue • Orange, CA 92865-4296 • 1-714-998-3021 • Fax 1-714-974-3420 Entire U.S.A. (800) 854-0547 • Fax (800) 850-0547 • http://www.com-spec.com BUTEO BOOKS The following Birds of North America Species Accounts are available through Buteo Books, 3130 Laurel Road, Shipman, VA 22971. TOLL-FREE ORDERING: 1-800-722-2460; FAX: (804) 263-4842. E-mail; alien® buteobooks.com Barn Owl (1). Carl D. Marti. 1992. 16 pp. Boreal Owl (63). G.D. Hayward and RH. Hayward. 1993. 20 pp. Broad-winged Hawk. (218). L.J. Goodrich, S.C. Crocoll and S.E. Senner. 1996. 28 pp. Burrowing Owl (61). E.A. Haug, B.A. Millsap and M.S. Martell. 1993. 20 pp. Common Black-Hawk (122). Jay H. Schnell. 1994. 20 pp. Cooper’s Hawk (75). R.N. Rosenfield and J. Bielefeldt. 1993. 24 pp. Crested Caracara (249). Joan L. Morrison. 1996. 28 pp. Eastern Screech-owl (165). Frederick R, Gehlbach. 1995. 24 pp. Ferruginous Hawk (172). MarcJ. Bechard and Josef K. Schmutz. 1995. 20 pp. Flammulated Owl (93). D. Archibald McCallum. 1994. 24 pp. Great Gray Owl (41). Evelyn L. Bull and James R. Duncan. 1993. 16 pp. Great Horned Owl (372). C. Stuart Houston, Dwight G. Smith, and Christoph Rohner. 1998. 28 pp. Gyrfalcon (114). Nancy J. Glum and Tom J. Cade. 1994. 28 pp. Harris’ Hawk (146). James C. Bednarz. 1995. 24 pp. Long-eared Owl (133). J.S. Marks, D.L. Evans and D.W. Holt. 1994. 24 pp. Merlin (44). N.S. Sodhi, L. Oliphant, P. James and I. Warkentin. 1993. 20 pp. Mississippi Kite (402). James W. Parker. 1999. 28 pp. Northern Saw-whet Owl (42). Richard J. Cannings. 1993. 20 pp. Northern Goshawk (298). John R. Squires and Richard T. Reynolds. 1997. 32 pp. Northern Harrier (210). R. Bruce MacWhirter and Keith L. Bildstein. 1996. 32 pp. Northern Hawk Owl (356). James R. Duncan and Patricia A. Duncan. 1998. 28 pp. Red-shouldered Hawk (107). Scott T. Crocoll. 1994. 20 pp. Red-tailed Hawk (52). C.R. Preston and R.D. Beane. 1993. 24 pp. Short-eared Owl (62). D.W. Holt and S.M. Leasure. 1993. 24 pp. Snail Kite (171). P.W. Sykes, Jr., J. A. Rodgers, Jr. and R.E. Bennetts. 1995. 32 pp. Snowy Owl (10). David F. Parmelee. 1992. 20 pp. Spotted Owl (179). R.J. Gutierrez, A.B. Franklin and W.S. Lahaye. 1995. 28 pp. Swainson’s Hawk (265). A. Sidney England, MarcJ. Bechard and C. Stuart Houston. 1997. 28 pp. Swallow-tailed Kite (138). Kenneth D. Meyer. 1995. 24 pp. Turkey Vulture (339). David A. Kirk and Michael J. Mossman. 1998. 32 pp. White-tailed Hawk (30). C. Craig Farquhar. 1992. 20 pp. White-tailed Kite (178). Jeffrey R. Dunk. 1995. 16 pp. Buteo Books stocks all published species accounts, not only those covering raptors. The current list in taxo- nomic order may be viewed at: http:/ /www. buteobooks.com Buteo Books stocks the Handbook of the Birds of the World. The first five volumes of this projected 12-volume work have been published including: Volume 2 : New World Vultures to Guineafowl (1994) covering the diurnal raptors and Volume 5: Barn Owls to Hummingbirds (1999) covering owls. These volumes are priced at $185 each plus shipping and handling. Usually available from Buteo Books, the clas.sic reference on diurnal birds of prey; Brown, Leslie and Dean Amadon. Eagles, Hawks and Falcons of the World. Country Life Books, 1968. Two volumes. First English edition in brown cloth. Fine in slipcase. $300.00 and other editions at lesser prices. 2002 ANNUAL MEETING The Raptor Research Foundation, Inc. 2002 annual meeting will be held in conjunction with the Third North American Ornithological Conference on 24-28 September in New Orleans, Louisiana. For information about the meeting see the following website: http://www.tulane.edu/~naoc-02/ or contact Dr. Tom Sherry (tsherry@tulane.edu). 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 11 7th 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. © This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Raptor Research Foundation (RRF), 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 be submitted at any time. Contact: Dr. Clint Boal, Texas Cooperative Fish and Wildlife Research Unit, BRD/USGS, Texas Tech Univer- sity, 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; e-mail: 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. TuUy 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. Kim 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; )efflincer@tns.net.