The Journal of Raptor Research Volume 37 Number 2 June 2003 Published by The Raptor Research Foundation, Inc. THE RAPTOR RESEARCH EOUNDATION, INC. (Founded 1966) OFFICERS PRESIDENT: Brian A. Millsap SECRETARY: Judith Henckel VICE-PRESIDENT: David M. Bird TREASURER: Jim Eitzpatrick BOARD NORTH AMERICAN DIRECTOR #1: Jeff Smith NORTH AMERICAN DIRECTOR #2: Laurie J. Goodrich NORTH AMERICAN DIRECTOR #3: Ted Swem INTERNATIONAL DIRECTOR #1; Beatriz Arroyo INTERNATIONAL DIRECTOR #2: Ricardo Rodri'guez-Estrelia OF DIRECTORS INTERNATIONAL DIRECTOR #3: Steve Redpath DIRECTOR AT LARGE #1: Jemima ParryJones DIRECTOR AT LARGE #2: Petra Bohait. Wood DIRECTOR AT LARGE #3: Michael W. Gou opy DIRECTOR AT LARGE #4: Carol McIntyre DIRECTOR AT LARGE #5: Robert N. Rosenfield DIRECTOR AT LARGE #6; Daniel E. Varland EDITORIAL STAFF EDITOR: James C. Bednarz, Department of Biological Sciences, P.O. Box 599, Arkansas State University, State University, AR 72467 U.S.A. ASSISTANT EDITOR: Rebecca S. Maul ASSOCIATE EDITORS James R. Belthofe Marco Restani Clint W. Boal Ian G. Warkentin Joan L. Morrison Troy I. Wellicome Juan Jose Negro BOOK REVIEW EDITOR: Jeffrey S. Marks, Montana Cooperative Research Unit, University of Montana, Missoula, MT 59812 U.S.A. SPANISH EDITOR: Cesar MArquez Reyes, Institute Humboldt, Colombia, AA. 094766, Bogota 8, Colombia EDITORIAL ASSISTANTS: Jaimie Varner, Joan Ciark 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 (8% 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. Exphcit instructions and publication policy are outlined in “Information for contributors,” J. Raptor Res., Yo\. 36(4), and are available from the editor. Submit manuscripts to J. Bednarz at the address listed above. COVER: Soaring Peregrine Falcon {Falco peregrinus) . Oil painting by Dick Dekker. Contents Peregrine Falcon Predation on Dunlins and Ducks and Kleptoparasitic Interference from Bald Eagles Wintering at Boundary Bay, British Columbia. Dick Dekker 91 The 2000 Canadian Peregrine Falcon Survey. Petra Rowell, Geoffrey l. Hoiroyd, and Ursula Banasch 98 Breeding Season Habitat Use and Ecology of Male Northern Pygmy-Owls. Alan R. Giese and Eric D. Forsman 117 Habitat Use by Swainson’s Hawks on Their Austral Wintering Grounds in Argentina. Sonia B. Canavelli, Marc J. Bechard, Brian Woodbridge, Michael N. Kochert, Juan J. Maceda, and Maria E. Zaccagnini 125 Raptor and Chihuahuan Raven Nesting on Decommissioned Telephone-line Poles IN THE Northern Chihuahuan Desert. Don L. Brubaker, Kathleen L. Brubaker, and Bruce C. Thompson 135 Short Communications Aerial Telemetry Accuracy in a Forested Landscape. Travis l. DeVauit, Warren l. Stephens, Bradley D. Reinhart, Olin E. Rhodes, Jr., and I. Lehr Brisbin, Jr. 147 Perch-site Selection and Spatial Use by Cactus Ferruginous Pygmy-Owls in South- central Arizona. Aaron D. Flesch 151 Interspecific and Intraspecific Kleptoparasitic Interactions of the Bearded Vulture {GyPAETUS BARBATUS) at Nesting Areas. Antoni Margalida and Joan Bertran 157 Gyrfalcon Predation on Mallards and the Interaction of Bald Eagles Wintering in Central Alberta. Dick Dekker and Gordon Court 161 Wintering Snowy Owls Feed on Sea Ducks in the Belcher Islands, Nunavut, Canada. Gregory J. Robertson and H. Grant Gilchrist 164 The Colonization of Sicily by the Black Kite {Milvus migrans) . Maurizio Sara 167 Letters Gyrfalcon Color Variation. Ian Flann 173 Two White-tailed Sea Eagles (Halmeetus ALBiaiiA) Collide with Wind Generators in Northern Germany. Oliver Krone and Christian Scharnweber 174 Talon-locking in the Red-tailed Hawk. Robert W. Dickerman 176 The Raptor Research Foundation, Inc. gratefully acknowledges funds and logistical support provided by Arkansas State University to assist in the publication of the journal. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. PEREGRINE FALCON PREDATION ON DUNLINS AND DUCKS AND KLEPTOPARASITIG INTERFERENCE FROM BALD EAGLES WINTERING AT BOUNDARY BAY, BRITISH COLUMBIA Abstract. — ^At Boundary Bay, British Columbia, wintering Peregrine Falcons {Falco peregrinus) captured 94 Dunlins {Calidris alpina) in 652 hunts. The major hunting techniques were open attacks on flying flocks (62%) and stealth attacks on feeding or roosting flocks (35%). Success rates for these techniques were 9.1% and 23.6%, respectively. Sixty-five Dunlins were taken directly from the edge of flocks; 29 Dunlins were seized after they had split off from flocks or were flying alone. Adult peregrines were significantly more successful than immatures (26.8% vs. 9.0%). Peregrines captured one Green-winged Teal {Anas crecca) and three larger ducks. The teal was carried away, but the larger ducks were pirated by Bald Eagles {Haliaeetus leucocephalus ) . Eagles often joined peregrines that were chasing Dunlins; six eagles succeeded in either capturing the Dunlin or forcing a peregrine to drop its just-caught prey. It is postulated that peregrines wintering at Boundary Bay avoid prey species that are too heavy to be carried out of reach of kleptoparasitic eagles. Female peregrines aggressively chased off other females, but they tolerated males. Females often joined males that were chasing Dunlins; 14 males were forced to surrender their prey to females. Four peregrines pirated Dunlins from Merlins {Falco columbarius) . ETy Words: Bald Eagle, Haliaeetus leucocephalus; Dunlin', Calidris alpina; Peregrine Falcon', Falco pere- grinus; British Columbia coast, ducks; hunting habits; kleptoparasitism. PREDACION DEL HALCON PEREGRINO SOBRE CHORLITOS YPATOS, E INTERFERENCIA KLEP- TOPARASITICA DEL AGUILA CALVA TEMPERANDO EN LA BAHIA DE BOUNDARY, BRITISH COLUMBIA Resumen.— En sitios de destinacion final de la migracion de otoho, el halcon peregrino {Falco peregrinus) capture 94 chorlitos {Calidris alpina) en 652 cacerias/persecuciones, en la Bahia de Boundary, British Columbia. Las tecnicas de caceria mas sobresalientes fueron ataques abiertos sobre parvadas al vuelo (62%) y ataques furtivos sobre parvadas forrajeando o perchando (35%). Las tasas de exito para esta tecnicas fueron 9.1% y 23.6%, respectivamente. Sesenta y cinco chorlos fueron tornados directamente del borde de la bandada; 29 chorlitos fueron atrapados despues de haberse separado de la bandada o estaban volando solos. Los peregrines adultos fueron significativamente mas exitosos que los inmaduros (26.8% vs. 9.0%). Los peregrines capturaron un pato aliverde {Anas crecca) y tres patos mas grandes. El pato aliverde fue llevado/apartado, pero el pato grande fue robado por aguilas calvas {Ftaliaeetus leu- cocephalus) . Las aguilas a menudo se unen a los peregrinos que estan persiguiendo chorlitos; seis aguilas tuvieron exito, ya sea capturando chorlitos o forzando a los peregrinos a soltar su recien atrapada presa. Se ha postulado que los peregrinos que temperan en la Bahia de Boundary evitan especies de presas que son demasiado pesadas para ser llevadas fuera del alcance de aguilas kleptoparasiticas. Las hembras de peregrino persiguen agresivamente otras hembras, pero toleran a los machos. Las hembras a menudo se unieron a los machos que estaban cazando chorlitos; 14 machos fueron forzados a ceder su presa a las hembras. Cuatro peregrinos robaron los chorlitos a los esmerejones {Falco columbarius). VoL. 37 No. 2 J Raptor Res. 37(2):91-97 © 2003 The Raptor Research Foundation, Inc. Dick Dekker^ 3819-112 A Street N.W., Edmonton, AB, T6J 1K4, Canada [Traduccion de Cesar Marquez] ^ E-mail address: tj_dick_dekker@hotmail.com 91 92 Dekker VoL. 37, No. 2 In their nearly worldwide range, Peregrine Fal- cons {Falco peregrinus) prey on a wide variety of birds, although they may locally specialize on very few species (Brown and Amadon 1968, Palmer 1988). The size of the prey taken is partly deter- mined by the sex of the peregrine because females are about one-third heavier than males. This sexual size-dimorphism has spawned several hypotheses, including that it widens the range of prey that can be taken by paired falcons (Selander 1966, Cade 1982). Although both sexes of peregrines may hunt the same prey species, such as the Rock Dove {Columba livia) in human-altered ecosystems (Rat- cliffe 1993, Frank 1994), or murrelets {Brachyram- phus spp.) over the ocean (Beebe 1960, Dekker and Bogaert 1997), in many habitats female pere- grines tend to kill larger prey than do males. For example, in central Alberta, migrant female pere- grines predominantly killed ducks (Anatidae), while the males took sandpipers (Scolopacidae) and small passerines (Dekker 1980, 1988). Pere- grines wintering in coastal agricultural regions of British Columbia and Washington also partitioned the food resource by gender (Anderson and De- Bruyn 1979, Dekker 1987, 1995, 1999). In the estuaries of western North America, a common prey of wintering peregrines is the Dun- lin ( Calidris alpina) . Based on the remains of Dun- lins killed by falcons, researchers in Washington and California reported that the majority are ju- veniles, which might be related to age-related flocking behavior (Kus et al. 1984, Warnock 1994). This paper details the hunting methods of both sexes of peregrines attacking flocks of Dunlins. It also reports on the duck hunting habits of these peregrines with particular reference to kleptopa- rasitic interference from Bald Eagles {Haliaeetus leucocephalus) . Study Area and Methods The study area is at Boundary Bay, which is part of the Fraser River delta (49°05'N, 123°00'W) in southwestern British Columbia. The bay is 16 km across and the inter- tidal zone is roughly 4 km wide at the lowest ebb. The tidal mud flats are bordered by a narrow strip of salt- marsh and a dyke that protects low-lying agricultural fields inland. Boundary Bay is a major stop-over for mi- gratory waterbirds and a wintering refuge for ca. 20 000 ducks and 50 000 Dunlins. The only other shorebird to winter in the bay in some numbers (ca. 1000) is the Black-bellied Plover {Pluvialis squatarola) (Butler 1994, Butler and Kaiser 1995). Bald Eagles are common year- round and increase locally to 50-150 during January- February (Dekker 1999). For a more detailed description of the Fraser delta and its avifauna, see Butler and Camp- bell (1987). Between early November and early February, 1994— 2003, I spent part or all of 151 d in the study area for a total of 940 hr. I walked the dyke or, especially during rain and strong winds, sat in a parked vehicle at a vantage point from where the tide flats were visible. To study the interaction of peregrines and their prey species, 1 used three principal methods: (1) flocks of ducks and shore- birds were monitored for alarm behavior such as sudden flushing; (2) the area was frequently scanned through binoculars to spot flying peregrines; and (3) perched peregrines were observed for varying lengths of time in the hope of seeing them hunt. Peregrines were classified as either adult or immature, based on dorsal color and ventral markings (Palmer 1988). Males and females could only be separated with certainty, respectively, at the lower or higher end of their size range, because recorded weights of the heaviest male peregrines overlap with those of the lightest females, par- ticularly between the subspecies occurring in western North America (Brown and Amadon 1968, White et al. 2002). Some falcons appeared typical of F. p. pealei (Bee- be 1960), others of F p. anatum (Palmer 1988). There is no evidence that peregrines of arctic origin (F. p. tun- drius) winter in the bay. Already by the second week of February sightings of peregrines dropped off sharply, perhaps indicative of a return to breeding sites in the region. Data were recorded in diary form and entered into an annotated table of hunts and kills. The term “hunt” is defined as a completed attack of which the outcome was known (Dekker 1980, 1988). A hunt could include one or more passes or stoops at the same Dunlin or flock. An attack on a flock and subsequent pursuit of a single Dun- lin fleeing that flock were counted as one hunt. However, if the peregrine abandoned the pursuit and presently made another attack on the same or a different flock, this was counted as a second hunt. This definition of a hunt is equivalent to the term “attack” used by Cresswell (1996), but differs from “hunting flight” as formulated by Buchanan et al. (1986) and Buchanan (1996). Wheth- er or not a hunt resulted in a kill was not always imme- diately apparent, especially at distances of >1 km. If the falcon ended a hunt by flying away fast and attacking other birds, without having been chased by a kleptopa- rasite, the preceding hunt was unsuccessful. On the other hand, if the falcon briefly slowed down with lowered feet, or if it retrieved something from the ground or water, then headed directly to shore, it was probable that prey had been caught. Supporting clues were pursuit by con- specifics, eagles or large gulls {Larus spp.). However, all probable captures were deleted from the data base unless they could be confirmed; for instance, by feeding activity. Perched peregrines were viewed through a 20—60 power telescope. Separate success rates were computed for adults and immatures and for the major strategies. Hunt- ing data collected in 1994-98, which were presented in an earlier paper (Dekker 1998), were analyzed further and added to the data obtained in 2000-03. Data sets were compared statistically by G-test of independence (Sokal and Rohlf 1969). June 2003 Peregrine Predation on Dunlins and Ducks 93 Table 1. Success rates of various strategies used by Per- egrine Falcons hunting Dunlins at Boundary Bay, British Columbia. Method Hunts Kills % Success Stealth hunts 229 54 23.6 Open hunts 406 37 9.1 Unknown method 17 3 17.6 Totals 652 94 14.4 Results Dunlin Hunts. In hunting Dunlins, peregrines used two main methods: a low stealth attack on resting or feeding flocks, or an open attack on fly- ing flocks. The stealth approach was used in 35% of hunts and resulted in 57% of total kills. The success rate of stealth hunts was significantly high- er (G = 17.53, P< 0.0001) than open hunts (Table 1 ) . Low stealth attacks were launched from perches or flapping flight, rarely by a stoop from high soar- ing flight. The Dunlins usually flushed at the very last moment and the peregrine might immediately N4/ succeed in seizing its prey. Often one or more Dun- lins dodged the falcon by dropping back onto the mud or water. Most took off at once with the per- egrine occasionally pursuing them again. Others failed to rise and appeared to have been hit by the falcon. These prey, either crippled or dead, were retrieved in an unhurried pass by the peregrine. In 62% of hunts, the falcons attacked Dunlins that had flushed well ahead or that were already airborne over the mudflats or ocean. If the ap- proaching peregrine flew >10 m high, the Dunlins formed dense, globular flocks over the ocean. (See Buchanan et al. 1988.) In the deciding stage of attacks on aerial flocks, individual peregrines dif- fered in method. They commonly maneuvered for position above the spherical flock until they stooped perpendicularly with rigid wings, either held open or tucked in. From a distance, it might seem as if some falcons plunged right through the massed Dunlins. This could never be verified. In- stead, the falcon often skirted the outside edge of the flock. At the terminus of its stoop, the pere- grine typically drew level and attacked the bottom end of the flock (Fig. la). Other peregrines c Figure 1. Strategies used by peregrines in attacks on flocks of Dunlins. The prey is either taken from the bottom, the edge, or the tail end of the flock. A. The peregrine stoops alongside of the flock and aims for the bottom B The peregrine stoops directly at the flock, which “caves in” and flattens out low over the water. C. Pursuit of flushing or low-flying flock. 94 Dekker VoL. 37, No. 2 stooped straight at the flock, which reacted by “caving in” and flattening out over the water while the falcon raked low over them (Fig. lb). One or more Dunlins might splash down and take off again. Others may have been hit by the falcons. If the falcon failed to seize its prey at once, it either: (1) pursued a single Dunlin that split away from the flock; (2) regained altitude and maneuvered for another vertical stoop at the same flock; (3) attacked a different flock some distance away; or (4) ended the hunting sequence by perching or flying away. In stealth hunts, when the peregrine ap- proached in low flight (<1 m), flocks of Dunlins that flushed ahead might stay low as well, particu- larly during stormy weather when they commonly flew very low over the waves. If attacked and over- taken by a falcon. Dunlins in the trailing end of the low-flying flock could be seized (Fig. Ic). Sin- gle birds that dodged by dropping into the water and taking off again might be pursued and swooped at with varying degrees of persistence. Twenty-seven Dunlins were captured after individ- ual pursuit including 3-10 passes. Two lone Dun- lins, flying at 20-25 m, were seized by immature male peregrines that stooped nearly perpendicu- larly from a height of >100 m. In three separate, but unsuccessful instances, immature peregrines persisted in follow-chasing Dunlins >100 m high and far inland. After dodging 20-40 swoops, all three Dunlins descended and escaped into bushes. The main hunting methods of stealth and open attack were used by all peregrines, although open attacks on flying flocks with multiple stoops ap- peared to be more typical of males than females. Immature males could be very persistent, but quite unsuccessful. For instance, on 21 January 2003, an immature male failed to make a capture over a pe- riod of 2.5 hr during which he launched three se- ries of attacks over the low tide line, including a total of about 30 stoops at flocks. On 24 and 25 January 2003, an immature male made 22 and 33 consecutive stoops at flying flocks respectively, all without success. By contrast, on 11 November 2002, an adult male captured two Dunlins 20 min apart, each requiring only one stealth attack on roosting flocks. Of the total 652 hunts directed at Dunlins, 25% were made by adult peregrines, yet they accounted for 47% of the kills (Table 2). The success rate of adults was 26.8%, significantly high- er than the 9.0% rate of immatures (G = 19.92, P - < 0 . 0001 ). Table 2. Success rates of Peregrine Falcons hunting Dunlins in British Columbia, and comparison with other studies of peregrines hunting shorebirds. Hunts Kills % Suc- cess This study (seacoast, winter) Adults 164 44 26.8 Immatures 399 36 9.0 Unidentified 89 14 15.7 Totals 652 94 14.4 Scotland^ (seacoast, winter) 233 25 10.7 Alberta*^ (Prairie Lake, migration times) 569 50 7.2 * Cresswell 1996. Dekker 1988. Unless they were hunting, adult males appeared to spend little time in the study area. On the other hand, adult females often perched for hours in the salt marsh or on inland cultivated land. On most days, two different adult females perched ca. 3 km apart. Evidently territorial, they attacked and chased off other females, either adult or immature. By contrast, the females tolerated males, and often joined them in attacking flocks of Dunlins. Fre- quently, both sexes pursued the same Dunlin, which might be captured by either one. However, if the male made the capture, he was chased and robbed by the female. Fourteen males dropped their prey, others were pursued far away and out of sight. Probably to avoid piracy by conspecifics, some male peregrines, upon capturing a Dunlin, flew to forested headlands across the bay. Four males circled to a high altitude and consumed their prey on the wing. The peregrines also robbed four Merlins {Falco columbarius) of their freshly- caught Dunlins. Peregrines hunted Dunlins throughout the day; 46 kills were made prior to 1200 H and 48 after 1200 H. The respective values for kills per hour of observation (1/8.1 and 1/ 12.4) were not significantly different {G — 2.85, P > 0.05). Peregrines often resumed hunting im- mediately after eating a Dunlin, which took 10-20 min. Duck Hunts. The peregrines rarely hunted ducks in the study area. Only four kills were seen. A Green-winged Teal {Anas crecca ) , that was flushed and briefly pursued by an adult male, was seized in a shallow pool on the edge of the salt marsh. June 2003 Peregrine Predation on Dunlins and Ducks 95 The peregrine carried his prey out of sight low over the marsh. Female peregrines pursued and captured three ducks larger than teal, presumably Northern Pintails {Anas acuta), which were seized on the ground after the ducks dodged the ap- proaching falcon by dropping onto the wet tide flats. Kleptoparasites. All three large ducks captured by peregrines in the study area were forfeited with- out a fight to approaching Bald Eagles. Four other peregrines that chased and harassed ducks for some distance were shadowed by one or more Bald Eagles, which attempted to seize ducks that dodged the peregrine by ditching into the water. Eagles also commonly interfered with peregrines that were hunting Dunlins. At least 50 of 94 pere- grines, which had just caught Dunlins, were im- mediately pursued by one or more eagles. The fal- cons usually managed to stay ahead of eagles. However, three peregrines, closely harassed by 2-4 eagles, released their Dunlins. Two of these falling prey items were retrieved in mid-air by one of the eagles. In at least three cases, eagles pounced on and captured Dunlins that had ditched into the water or dropped into grassy vegetation to dodge the falcon. The following incident is a typical example of intra- and interspecific competition between fal- cons and eagles. In the afternoon of 16 January 2003, a male peregrine stooped at a flock of Dun- lins and pursued a single bird that split off and dodged repeated passes. The male was quickly joined by a female peregrine and four eagles. All of these alternately swooped at, or tried to swoop at the Dunlin which flew erratically I— 10 m high over the water. After a combined total of 10-12 passes, the Dunlin was caught by the female pere- grine, which was immediately set upon by the ea- gles. The peregrine managed to escape and carry her prey far inland. Prey-carrying peregrines were frequently, but vainly, chased by Glaucous-winged Gulls {Larus glances cens) . In addition, gulls often joined peregrines that were pursuing lone Dunlins, and on two occasions gulls pounced on Dunlins that had ditched into the water to escape a swoop- ing peregrine. Discussion The stealth tactics used by the falcons in this study were similar to those reported for migrating peregrines hunting shorebirds in central Alberta, although, due to more favorable weather condi- tions, the migrant falcons launched their stealth attacks more often from high soaring flight (Dek- ker 1980, 1988, 1999). The percentage of stealth hunts in the two studies is, however, quite different (35% at Boundary Bay vs. 77% in Alberta). The explanation can be found in the difference in re- spective habitats. The Alberta falcons hunted over reed-studded shallows of a large inland lake, which afforded ample opportunities for a concealed ap- proach. By contrast, the relatively low use of stealth on the coast of British Columbia reflects the lack of opportunity for effective surprise over the tide flats. The percentage of stealth hunts in this study is nearly equivalent to the 36% reported from the east coast of Scotland (Cresswell 1996). It is possible that stealth hunts, especially suc- cessful ones by adult peregrines, are under-record- ed in this study, resulting in a bias in favor of open hunts. Successful stealth attacks on flocks of Dun- lins roosting near the salt marsh easily escape de- tection by the human observer. Stealth attacks are often initiated from distances of >I km and the terminal portion takes place very low over the ground or water. The light-colored plumage of the adult falcon blends into the background of water and renders it more difficult to detect than a dark immature peregrine. Screened by the massive flushing of hundreds or thousands of Dunlins, the attacking peregrine might carry its prey away be- fore the observer is aware of what happened. Adult falcons are competent in the use of stealth and especially the male tends to keep a low profile, per- haps to avoid attracting the attention of klepto- parasites. In contrast to stealth attacks, open hunts are readily observed because the observer is alerted to the arrival of the predator by the rising of dense flocks of Dunlins. However, during a prolonged se- ries of attacks, it is difficult to keep accurate score of the exact number of falcon hunts. Stoops on flocks can follow each other in rapid succession and the observer often cannot determine if a stoop was aimed at the same or a different flock. To avoid this problem, Buchanan et al. (1986) used the term hunting flight, which is defined as “a perch to perch flight involving one or more cap- ture attempts.” Although the number of kills was very small (A = 7), Buchanan et al. (1986) calcu- lated the success rate of hunting flights as 47% and that of individual capture attempts as 14.6%. Any study of the hunting habits of the Peregrine Ealcon involves a certain amount of subjective in- 96 Dekker VoL. 37, No. 2 terpretation, which might in part explain the vari- ance between studies (White et al. 2002). In my appraisal of open hunts, I kept score of separate attacks in a conservative way, concentrating on the outcome. How the kill was made seemed of more importance than how often the falcon missed. Many hunts that fail may not be all that intense. This possibility was pointed out by Treleaven (1980), who coined the terms high-intensity and low-intensity hunting. Based on the examination of prey remains, a high proportion of shorebirds killed by raptors were reported to be juveniles (Whitfield 1985, War- nock 1994), Kus et al. (1984) suggested that this may be due to age-related differences in Dunlin flocking behavior. 1 postulate that the mechanics of prey selection may be very simple. Sixty-five (69%) of the Dunlins captured during this study were taken from the outside or the tail end of flocks (Fig. 1). Flocking is a well-known predator avoidance behavior of open-country birds. Their drawing together into dense aerial formations is the result of each bird’s instinctive desire to find safety in the center of the flock (Tinbergen 1951). I speculate that adults are more proficient at this than juveniles, which are left on the outside of the flock. Several studies have reported age-related seg- regation of birds in roosting and feeding flocks (Newton 1998). If peregrines selectively remove ju- venile Dunlins from a wintering population be- cause they are easier to capture than adults, then the proportion of vulnerable juveniles in that pop- ulation should decline over the course of the win- ter. Furthermore, juveniles that are attacked on a daily basis and manage to survive for several months, might learn to become more vigilant and avoid capture. These conclusions seem to be sup- ported by the data. The respective hunting success rates of peregrines for November and January de- cline from 18.6% (N = 134) to 12.4% (N = 404), although the difference is not significant (G = 2.30, P> 0.05). The kleptoparasitic habits of Bald Eagles are well known, particularly at the expense of peregrines that hunt ducks or seabirds (Anderson and De- Bruyn 1979, Dekker 1995, Dekker and Bogaert 1997). Bald Eagles also kleptoparasitize Merlins. During this study, Merlins were seen to capture 14 Dunlins, two of which were pirated by pursuing ea- gles. Buchanan (1988) reported similar instances and he suggested that the threat of losing prey to a kleptoparasite resulted in Merlins engaging in hunting flights of shorter duration when potential kleptoparasites were present. Gyrfalcons {Falco rus- ticolus) wintering in Alberta were often robbed of ducks by eagles as well and avoided hunting at lo- calities where Bald Eagles sat on prominent perch- es (Dekker and Court 2003) . In conclusion, I suggest that the ubiquitous pres- ence of eagles at Boundary Bay discourages win- tering peregrines from hunting ducks on the tide flats. The possibility exists that peregrines may hunt ducks more often at night or at inland local- ities where eagle presence is lower than on the coast. Adult female peregrines, which had perched on the coast for much of the day, flew inland at dusk when ducks also leave the coast to feed on inland meadows (Dekker 1999). During Novem- ber, when eagles are far less numerous at Boundary Bay than in January, several immature peregrines hunted ducks persistently in the study area. As the winter progresses however, I surmise that these ju- veniles eventually stop hunting ducks on the tide flats and instead concentrate on prey such as the Dunlin that can easily be transported over long dis- tances, out of reach of kleptoparasitic eagles. Acknowledgments All expenses incurred during this study were privately financed by the author. Accommodation in 2000, 2002, and 2003 was kindly provided by D. Leach. Irma Dekker periodically assisted in the field. Occasional co-observers were D. Hancock, R. Swanston, and R Thomas. G. Court did the statistical tests. The comments of W. Nelson, J. Buchanan, and T. Swem, who refereed the first draft of this paper, were greatly appreciated. Literature Cited Anderson, C.M. and P.D. DeBruyn. 1979. Behavior and ecology of Peregrine Falcons wintering on the Skagit Flats, Washington state. Unpubl. report. Dept, of Game, Olympia, WA U.S.A. Beebe, F.L. 1960. The marine peregrines of the northwest Pacihc coast. Condor 62:14.5-189. Brown, L. and D. Amadon. 1968. Eagles, hawks, and fal- cons of the world. Hamlyn House, Feltham, U.K. Buchanan, J.B. 1988. The effect of kleptoparasitic pres- sure on hunting behavior and performance of host Merlins./. Raptor Res. 22:63-64. . 1996. A comparison of behavior and success rates of Merlins and Peregrine Falcons when hunting Dun- lins in two coastal habitats./. Raptor Res. 30:93-98. , S.G. Herman, and TM. Johnson. 1986. Success rates of the Peregrine Falcon hunting Dunlin during winter. Raptor Res. 20:130-131. , C.T. Schick, L.A. Brennan, and S.G. Herman. 1988. Merlin predation on wintering Dunlins: hunt- June 2003 Peregrine Predation on Dunlins and Ducks 97 ing success and Dunlin escape tactics. Wilson Bull. 100: 108-118. Butler, R.W. 1994. Distribution and abundance of West- ern Sandpipers, Dunlins, and Black-bellied Plovers in the Fraser River estuary. Pages 18-23 in R.W. Butler and K. Vermeer [Eds.], The abundance and distri- bution of estuarine birds in the Strait of Georgia, Brit- ish Columbia. Can. Wildl. Serv. Occ. Paper No. 83. Ottawa, Canada. AND R.W. Campbell. 1987. The birds of the Fraser River delta: populations, ecology, and international significance. Can. Wildl. Serv. Occ. Paper No. 65. Ot- tawa, Canada. ^ AND G.W. Kaiser. 1995. Migration chronology, sex ratio, and body mass of Least Sandpipers in British Columbia. Wilson Bull 107:413-422. Cade, T.J. 1982. The falcons of the world. Cornell Univ. Press, Ithaca, NYU.S.A. Cresswell, W. 1996. Surprise as a winter hunting strategy in sparrowhawks, peregrines, and Merlins. Ibis 138: 684-692. Dekker, D. 1980. Flunting success rates, foraging habits, and prey selection of Peregrine Falcons migrating through central Alberta. Can. Field-Nat. 94:371-382. . 1987. Peregrine Falcon predation on ducks in Alberta and British Columbia. J. Wildl. Manage. 51: 156-159. . 1988. Peregrine Falcon and Merlin predation on small shorebirds and passerines in Alberta. Can. J. ZooL 66:925-928. . 1995. Prey capture by Peregrine Falcons winter- ing on southern Vancouver Island, British Columbia. J. Raptor Res. 29:26-29. . 1998. Over-ocean flocking by Dunlins (Calidris alpind) and the effect of raptor predation at Bound- ary Bay, British Columbia. Can. Field-Nat. 112:694-697. . 1999. Peregrine — bolt from the blue. Hancock House Publishing, Surrey, BC Canada. and L. Bogaert. 1997. Over-ocean hunting by Peregrine Falcons in British Columbia. J. Raptor Res. 31:381-383. AND G. Court. 2003. Gyrfalcon predation on Mal- lards and the interaction of Bald Eagles wintering m central Alberta./. Raptor Res. 37:161-163. Frank, S. 1994. City peregrines. A ten-year saga of New York City falcons. Hancock House Publishers, Surrey, BC Canada. Kus, B.E., P. Ashman, G.W. Page, and L.E, Stenzel. 1984. Age-related mortality in a wintering population of Dunlin. Auk 101:69—73. Newton, 1. 1998. Population limitation in birds. Academ- ic Press, San Diego, CA U.S.A. Palmer, R.S. 1988. Handbook of North American birds Vol. 5. Diurnal raptors. Part 2. Yale Univ. Press, New Haven, CT U.S.A. Ratcliffe, D. 1993. The Peregrine Falcon. Academic Press, San Diego, CA U.S.A. Selander, R.K. 1966. Sexual dimorphism and differential niche utilization in birds. Con^Zor 68:113-151. SOKAL, R.R. and F.J. Rohlf. 1969. Biometry. Freeman and Company, San Francisco, CA U.S.A. Tinbergen, N. 1951. The study of instinct. Oxford Univ. Press, Oxford, U.K. Treleaven, R.B. 1980. High and low intensity hunting m raptors. Z. Tierpsychol. 54:339—345. Warnock, N.D. 1994. Biotic and abiotic factors affecting the distribution and abundance of a wintering popu- lation of Dunlin. Ph.D. dissertation, Univ. California, Berkeley, CA U.S.A. White, C.M., NJ. Clum, T.J. Cade, and W.G. Hunt. 2002. Peregrine Falcon {Falco peregrinus) . In A. Poole and F Gill [Eds.], The birds of North America, No. 660. The Birds of North America Inc., Philadelphia, PA U.S A. Whitfield, D.P. 1985. Raptor predation on wintering waders in southeast Scotland. Ihis 127:544—558. Received 1 July 2002; accepted 15 March 2003 J Raptor Res. 37(2):98-116 © 2003 The Raptor Research Foundation, Inc. THE 2000 CANADIAN PEREGRINE EALCON SURVEY Editors Petra Rowell No. 4 Morgan Crescent, St. Albert, AB T8N 2E2 Canada Geoffrey L. Holroyd^ and Ursula Banasch Canadian Wildlife Service, Room 200, 4999-98 Avenue, Edmonton, AB T6B 2X3 Abstract. — National surveys for Peregrine Falcons {Falco peregrinus) have been conducted in Canada every five years since 1970. This paper discusses the results of the seventh such survey. In 2000, 22 areas in nine provinces and three territories were searched for three peregrine subspecies. Within the con- tinental F. p. anatum range, 374 sites were occupied by territorial peregrines. As in previous surveys, the majority of anatum sites were located in the western boreal ecoregion. However, numbers are increasing in populations south of 58°N. The number of F. p. tundrius in 2000 was similar to previous surveys at Rankin Inlet, Nunavut and Tuktut Nogait National Park, Northwest Territories. Nine sites occupied on the North Slope of the Yukon Territory is a slight increase in this tundrius subgroup. Coastal F. p. pealei, on the Queen Charlotte Islands appear to have been stable for at least the past two decades. Pealei numbers on the north shore of Vancouver Island and the surrounding area were substantially higher in 2000 — largely the result of a more extensive search effort. Reproductive output in 2000 was 1.0 or more young/territorial pair, except in the anatum subpopulations of northern Alberta and the Peel River, Yukon Territory where it was 0.7 and 0.6 young/ territorial pair, respectively. Overall, with stable populations in the north and growing populations in the south, the peregrine recovery is well underway and in some areas, approaching historical size and probably ecological carrying capacity. Keywords; Peregrine Falcon-, Falco peregrinus; status; survey; Canada; reproductive output. PROSPECCION DEL HALCON PEREGRINO CANADIENSE EN EL 2000 Resumen. — Prospecciones nacionales del Halcon Peregrino {Falco peregrinus) han sido llevadas a cabo en Canada cada 5 anos desde 1970. Este articulo discute los resultados de la septima inspeccion. En el 2000, 22 areas en nueve provincias y tres territorios fueron monitoreados para tres subespecies de peregrino. Dentro del rango continental de F. p. anatum, 374 sitios fueron ocupados por peregrinos territoriales. A1 igual que en prospecciones previas, la mayoria de los sitios de anatum estuvieron localizados en la ecoregion boreal del oeste. Sin embargo, los numeros se estan incrementando en las poblaciones al sur de 58°N. El numero de F. p. tundrius en el 2000 fue similar al de surveys previos en Rankin Inlet, en los parques nacionales de Nunavut and Tuktut Nogait de los territorios del Noroeste. Nueve sitios ocupados en la Pendiente norte del territorio de Yukon, es un incremento leve en este subgrupo tundrius. El costero F. p. pealei, en las Islas Queen Charlotte parece que han estado estables por al menos, las pasadas dos decadas. Los numeros de pealei en la costa norte de la Isla de Vancouver y en las areas aledanas, fueron substan- cialmente mas altas en el 2000 — debido, en gran medida a un esfuerzo de busqueda mas extensive. El rendimiento reproductivo en el 2000 fue de 1.0 o mas par joven/territorio, excepto en la subpoblacion de anatum en el norte de Alberta y en el Rio Peel (Peel River), Territorio de Yukon donde fue de 0.7 y 0.6 par joven/territorio respectivamente. Globalmente, con poblaciones estables en el norte y poblaciones crecientes en el sur, la recuperacion del peregrino es buena y en algunas areas, se aproxima al tamano historico y probablemente a la capacidad de carga ecologica. [Traduccion de Cesar Marquez] Peregrine Falcons {Falco peregrinus) bred histor- ically throughout most regions of Canada and are again breeding in many of these same regions to- 1 E-mail address: Geoffrey.Holroyd@ec.gc.ca day. Three subspecies of peregrine occur in Ca- nada (White 1968). Generally, the continental ana- tum subspecies breeds south of the treeline from the Atlantic to the Pacific oceans. The smaller, northern tundrius subspecies nests along arctic riv- 98 June 2003 2000 Peregrine Survey 99 ers, lakes, and coasdine and inland escarpments. The larger, darker western pealei subspecies occu- pies coastal islands and areas of adjacent mainland British Columbia. The biology of Peregrine Fal- cons is described by Hickey (1969), Cade et al. (1988), Erickson et al. (1988), and Brown and Amadon (1989). A decline in peregrine numbers in North Amer- ica started in the late 1940s and continued up to the mid 19V0s (Hickey 1969, Kiff 1988). This de- cline was linked directly to the contamination of peregrines by pesticides such as DDT, dieldrin, and heptachlor epoxide (Ratcliffe 1969, Risebrough and Peakall 1988). These chemicals occurred in eggs, carcasses and some prey species, causing death, eggshell thinning and breakage, and overall population decline. In response, in 1978 the Com- mittee on the Status of Endangered Wildlife in Ca- nada (COSEWIC) classified anatum peregrines as endangered, tundrius as threatened, and pealei as rare (Martin 1979). Captive-breeding of Peregrine Falcons began in Canada in 1972, releases started in 1975 (Fyfe 1988), and the first captive-raised fal- cons bred in the wild in 1977 (White et al. 1990). In 1986, a Recovery Team was formed and in 1987 a Recovery Plan was approved (Erickson et al. 1988). In Canada, researchers began surveying pere- grines prior to 1960 (Beebe 1960, Enderson 1965, Fyfe 1969). From 1970 on, national surveys were carried out every 5 yr to determine nest site oc- cupancy and reproductive rates. The 1965 Madison Peregrine Conference and the 1970 and 1975 North American surveys documented the contin- ued downfall of the peregrine (Hickey 1969, Cade and Fyfe 1970, Fyfe et al. 1976). By 1980, numbers were low, but stabilizing in northern Quebec, the Northwest Territories, (including today’s Nunavut Territory) and the Yukon. Only one site was oc- cupied in the prairies and the boreal forest south of 58°N and east of the Rocky Mountains (White et al. 1990). Surveys on the Labrador coast in the mid-1980s identified a significant breeding popu- lation (Lemon and Brazil 1990). This population expanded to 31 known pairs by 1995. The release of 178 captive-raised young in the Maritimes be- tween 1982 and 1991 (Sam et al. 1994) resulted in the return of this population to six pairs by 1995 (Amirault et al. in press) . By 1985-86, west coast and northern popula- tions also had increased or were stable while in the south a small urban population was established by the reintroduction of captive-raised young (Mur- phy 1990). Starting in the late 1980s, the releases of captive-raised young were made in fewer areas, with more young per release. The 1990 survey again documented stable or increasing northern and coastal populations with smaller gains made in the south (Holroyd and Banasch 1996). In 1992, COSEWIC reclassified tundrius as vulnerable (Bromley 1992). By 1995, stable or increasing sur- vey numbers of anatum met the objectives set pre- viously for six management zones across Canada (Erickson et al. 1988, Banasch and Holroyd in press). In 1999, pealei were retained as a species of special concern (COSEWIC 2000) and anatum were downlisted from endangered to threatened (Johnstone 1999). The purpose of this paper is to present results of the 2000 national survey that documented the population status and reproductive output of Ca- nada’s Peregrine Falcons. By comparing these data with data from previous surveys, we can evaluate the success of past management actions, reevaluate status designations, continue to monitor the com- plete and successful recovery of this species, and determine what management actions are required in the future. General Methods In 2000, the survey included a comparable number of sites with effort similar to previous 5-yr surveys such that population numbers and reproductive data could be compared with the past six national surveys. Although the focus of surveys was to visit historical nest sites, new sites found during and in-between 5-yr surveys were also included. Previous survey results were taken from Cade and Fyfe (1970), Fyfe et al. (1976), Court et al. (1988a), Munro and Van Drimmelen (1988), Murphy (1990), White et al. (1990), Holroyd and Banasch (1996), and Banasch and Holroyd (in press). As with all surveys, how- ever, there are some limitations. Timing of surveys varied from mid-May to the end of July. While every effort is made to conduct 5-yr surveys during the same stage of the breeding cycle, logistics, weather, funding, and time constraints dictate changes from survey to survey. Any effects from a change of mode and timing on results are identified in the regional accounts that follow. Definitions used in the 2000 survey are consistent with those generally accepted and used in previous surveys (Murphy 1990, Holroyd and Banasch 1996). “Nest site” is the actual site of the nest on a cliff, however, the exact location of the nest may vary from year to year; “occu- pied site” or “occupied territory” is a location where one or two territorial adults were present; “territorial pair” is a pair that defended its nesting cliff against other pere- grines or human intruders, or a pair that is persistent m remaining at the site; “breeding pair” is a pair that laid at least one egg; and “successful pair” is a pair that raised 100 Roweli. et al. VoL. 37, No. 2 at least one chick to fledging or that was assumed to have fledged. “Known sites” include all historically-occupied sites and any sites found occupied prior to the survey year. Sites found during the 2000 survey were “new sites.” If single falcons were seen only once and no ter- ritorial behaviour was observed, or birds did not remain at a site, they were considered “sightings.” Subspecies distinctions were assigned to regional pop- ulations on the basis of historical range and were rein- forced by surveyor observations of physical appearance (for an overview of peregrine subspecies distinctions see White and Boyce 1988). These distinctions are consistent with previous surveys. Only young produced in the wild (i.e,, not captive- raised) were included in determining the reproductive output of pairs. Reproductive measurements include the mean number of young fledged or assumed to have fledged/successful pair and the number of young/ terri- torial pair (a pair that actively defends or is persistent at a nest site) . Regional Accounts Labrador, Newfoundland. Contributed by Joe Bra- zil, Inland Fish and Wildlife Division, Department of Forest Resources and Agrifoods, Box 8700, St.John’s, NF AIB 4J6 Canada. Survey area and methods. Along the Labrador coast of Newfoundland, two helicopter crews sur- veyed 63 historical sites. The first survey was con- ducted 8-12 July and covered the southern coastal nests as well as most of the interior sites. The sec- ond survey was conducted 26-31 July covering the northern coastal nests from Hopedale north as well as the rest of the interior sites. Results and discussion. Of 63 known sites sur- veyed, peregrines occupied 22; single birds occu- pied seven; and territorial pairs occupied 15 sites. At the 15 sites, ten had young present, two had nests with eggs and two had females believed to be incubating. Ten pairs successfully produced 24 young for ratios of 2.4 young/ successful pair and 1.6 young/ territorial pair. Survey area and effort was similar to that in 1995 except that more inland sites (63 versus 42) were visited in 2000. The number of occupied sites de- clined from 31 in 1995 to only 22 in 2000. The 1995 figure (31) was the highest number of occu- pied sites ever recorded in Labrador, although this was partly the result of increased search effort in the intervening period between 1990 and 1995 sur- veys (Brazil et al. in press). The 2000 figure (22) was similar to the number of occupied sites in 1990 (21). Declines in 2000 occurred largely in the in- land sites, which were surveyed a month later than in 1995. One particular valley that had 11 of 13 sites occupied in 1999 had only two sites occupied in 2000. A single territorial bird held one of these. A cold, wet, snowy spring together with the pres- ence of fresh white wash (excrement) at 11 sites unoccupied at the time of the survey, suggests weather and early nest failures may have been a factor in this decline. More surveys will be con- ducted in the near future. Bay of Fundy (New Brunswick and Nova Scotia). Contributed by Diane L. Amirault, Canadian Wildlife Service, Atlantic Re^on, P.O. Box 6227, Sackville, New Brunswick E4L 1G6 Canada, and Mark F. Elderkin, Wildlife Division, Nova Scotia Department of Natural Resources, 136 Exhibition Street, Kentville, NS B4N 4E3 Canada. Survey area and methods. In Nova Scotia, surveys included intensive aerial and ground checks of all known nest sites, suitable coastal habitat, and known nesting pairs. On 12-13 June, observers used a Hughes 500 helicopter to survey the Inner Bay of Fundy from Digby through Lower Blomidon and both sides of Cape Chignecto peninsula from Great Village to Joggins; the west coast of Cape Breton Island; and a series of coastal islands in St. Margaret’s Bay on the Atlantic shore including Shut-In-Island, where a pair was believed to have nested successfully in 1999. Known nesting pairs were visited at least twice — early June and mid-July, to establish reproductive success. In New Brunswick, a 1-d helicopter survey of known and potential sites was conducted on 7 July, along the Bay of Fundy coast from Saint John, northeast to the Mary’s Point area, along the Pe- titcodiac River, and Dorchester Cape and Wood Point within Shepody Bay. Several inland sites with potential habitat were also checked with the excep- tion of Swan Lake, which lacks suitable nesting habitat. The canyons surrounding Little Salmon River, although not previously surveyed, were cov- ered because habitat looked suitable. In addition to aerial surveys, young were banded in two occupied nests. A ground search late in the season was conducted in the Bayfield area, New Brunswick, because of repeated reports of Pere- grine Falcon sightings throughout the season, de- spite the lack of suitable nesting sites. Reports from Fundy National Park wardens (Parks Canada) and local naturalists supplemented survey efforts. Results and discussion. In Nova Scotia, five breed- ing pairs were located, all on cliffs within the Inner Bay of Fundy. Four pairs fledged 12 young. The June 2003 2000 Peregrine Survey 101 fifth pair, new in 2000, hatched at least one young that did not survive. In New Brunswick, two known nest sites located within Shepody Bay and one known nest on Grand Manan Island were occupied. Two new territorial pairs were located during aerial surveys — both south of Fundy National Park. A minimum of seven young fledged from these five sites. A sixth pair nested in the Bayfield/Cape Jourimain area, the exact site was not found. A pair of adults was there earlier in the year and one juvenile bird was pres- ent on 8 July. One traditional nest site located in Saint John was not occupied, although two adults were present earlier. In addition to the six con- firmed pairs, another site at Shepody Bay was not surveyed due to the risk of disturbing a Great Blue Heron {Ardea herodias) colony nearby. The number of sites occupied in the Bay of Fun- dy (on both the Nova Scotia and New Brunswick sides) increased from six in 1995 to 11 in 2000 and is probably nearing historical numbers. Stocek and Pearce (1978) documented 13 historical sites in Nova Scotia and New Brunswick combined. Repro- ductive output in 2000 averaged 1.8 young/ terri- torial pair and 2.0 young/ successful pair. Southern Quebec. Contributed by Michel Lepage So- ciete de la Faune et Des Parcs du Quebec, Direction du Developpement de la Faune, 675, Boul. Rene-Lh>esque Est. He Stage PQ GIR 5V7 Canada. Survey area and methods. In 2000, surveys were limited to portions of Quebec located south of 49°N. In all, 63 sites were visited including the group of rock cliffs along the Saguenay River, where three sites were occupied in 1995. No sur- veys were conducted in the Gaspe Peninsula be- cause no previous information was available from there. The survey crew visited all 23 known sites where nesting previously occurred. In addition, the regional staff of the Societe de la Faune et Des Parcs surveyed 28 other high-potential sites. Be- tween two and four visits were made to each oc- cupied site including prior to fledging of young, which enabled the survey crew to evaluate repro- ductive success. Surveys were conducted from ground or from boat. Results and discussion. In 2000, Peregrine Falcons occupied 16 historical and 12 new sites; 25 sites were occupied by a pair and three by a single adult. There were 39 fledged young for 17 pairs. Repro- ductive output was 1.6 young/territorial pair and 2.3 young/ successful pair. The majority of the 12 new sites were discovered by bird watchers and or- nithologists as part of the Inventory Program for Threatened Bird Species in Quebec. Although the number of sites visited in 2000 (68) was less than those visited in 1995 (112), the number of occupied territories rose from 15 m 1995 to 28 in 2000 (87% increase). Pairs increased from 13-25 (92% increase). However, 12 new sites were located through the Inventory Program for Threatened Bird Species in Quebec and may re- flect both different search effort and population growth. The survey confirms that most natural cliffs where Peregrine Falcons are found continue to be occupied annually. With at least six pairs, the Saguenay River area is particularly exceptional. Most of the 28 occupied sites are on rock cliffs located along water. Only eight are on artificial structures such as skyscrapers, bridges or quarries. In at least three cases, one bird wore a red band, indicating it was released during the Canadian Per- egrine Falcon release program. With no Peregrine Falcon releases in Quebec since 1994 and the ma- jority of birds in the 1995 survey without a band, we believe that some of the birds released in On- tario over the last few years have settled in Quebec. Southern Ontario. Contributed by Brian Ratcliff and Ted Armstrong, Ontario Ministry of Natural Re- sources, 435 South James Street, Suite 221, Thunder Bay, ON P7E 6S8 Canada. Survey area and methods. Between 28 February and 1 1 August, surveyors and site monitors visited 40 known sites, 18 new sites, and hundreds of cliff sites in suitable habitat on foot, by boat, and heli- copter (Ratcliff and Armstrong 2000). The total area surveyed extended from Thunder Bay east along Lake Superior to Sault Ste. Marie, east to North Bay, and south to Bruce Peninsula on Lake Huron. Urban centres were searched and local monitoring programs were established for known sites. Results and discussion. There were 53 occupied sites, including 42 with territorial pairs. Attempted nesting was recorded for 31 pairs; 26 produced 68 young for 2.6 young/successful pair and 1.6 young/territorial pair. Most sites (31 or 58%) in 2000 were on cliffs located in the Lake Superior basin. In southern Ontario, urban sites dominated with nine pairs in Toronto (3), Hamilton, Ottawa, London, Mississauga, Nanticoke, and Niagara Falls. A single cliff nest in southern Ontario was located on the Bruce Peninsula, an area that peregrines reoccupied in 2000 after a 70-yr absence. The re- maining 12 sites were located in both urban (6) 102 Rowell et ai.. VoL. 37, No. 2 and rural locations (6) across the Great Lakes ba- sin for a total of 38 cliff and 15 urban sites. Ontario’s Peregrine Falcon {F. p. anatum) pop- ulation declined dramatically in the 1950s and 1960s with the last known nesting attempt in north- ern Ontario in 1963 (Armstrong in press). Captive- raised young were released between 1977 and 1996. Recovery was slow, however, with only a lone adult occupying a single territory in 1985 (Arm- strong in press). In 1986, one pair and a lone bird occupied two territories (Murphy 1990). Two nest- ing pairs and a single territorial bird were observed during 1990 (Holroyd and Banasch 1996). Several additional occupied sites were located between 1990 and 1995 (Armstrong in press). Ontario’s 2000 survey demonstrated a three-fold increase; from 15 occupied sites and 14 territorial pairs in 1995 (Ratcliff and Armstrong 2000). Although the exact proportions are unknown, this increase is partly due to increased search effort in 2000 and partly to natural population growth. An estimate of natural increase can be obtained for the west end of Lake Superior where the same effort (number of sites examined) was expended each year since 1997. There, occupancy has increased from 11 in 1997 to 23 in 2000, a natural growth rate of 109% over three years (28% annually). Similarly, the number of urban sites in southern Ontario in- creased from three in 1995 to 13 in 2000 (growth of 333% or 34% annually). A total of 117 young were banded at nests adja- cent to Lake Superior since 1996 and at least a portion of the current population expansion is from natural reproduction as young birds return as adults to nest. The occurrence of USA-banded birds in Ontario between 1996 and 2000, in partic- ular five in 2000, indicates the Ontario population is at least mixing with or is a part of a larger, ad- jacent population. Of 33 known origin (banded) adults, five (15%) were from the U.S. (Ohio and Pennsylvania), three (9%) were from Canadian re- introduction programs and the remainder (76%) were bred and banded in the wild in Canada. Southern Manitoba. Contributed by Tracy Macona- chie, Manitoba Peregrine Falcon Recovery Project, Box 24, 200 Saulteaux Crescent, Winnipeg, MB R3J 3W3 Canada. Survey area and methods. Because of the lack of records for northern Manitoba (Bechard 1981) and financial constraints, the 2000 survey effort was reduced from 1995, concentrating on known nesting, hacking, or roosting sites in and around urban centers that were involved in the release of 170 young between 1981 and 1996 including Win- nipeg, Brandon, Portage la Prairie, and Gimli (Jones et al. in press) . Results and discussion. In 2000, 21 sites with po- tential habitat in southern Manitoba were sur- veyed. Three nesting sites were occupied and mon- itored throughout the breeding season. These sites had two pairs (one in Winnipeg and one in Bran- don) and a single bird (in Winnipeg). The two pairs hatched five young of which four successfully fledged, all from the nest in Brandon. Therefore, 4.0 young fledged/ successful pair and 2.0 young fledged/ territorial pair. The 2000 data represent a decline in Manitoba peregrines from four pairs in 1995 (three sites in Winnipeg and a fourth in Brandon) (Jones et al. in press) to two in 2000. However, Manitoba per- egrines are part of a larger mid-continental popu- lation (Jones et al. in press). Birds released here have been re-sighted in Saskatchewan, Alberta, and Nebraska. The first female to nest in Manitoba in 1989 was from Minnesota and a female from Iowa replaced her. Despite management, this popula- tion remains small; never exceeding four pairs, and this is unlikely to change. However, some turnover is expected in the near future as the two breeding females are 9 and 11 yr old. Southern Saskatchewan. Contributed by W.J Pat- rick Thompson, Box 234, Clavet, SK SOK OYO Canada. Survey area and methods. Because historical evi- dence for the presence of Peregrine Falcons {F. p. anatum) is scarce in Saskatchewan (Thompson in press) and previous surveys in potential habitat in the north discovered no nest sites, no formal sur- vey was undertaken in Saskatchewan during 2000. However, all previously occupied urban sites were visited and, if occupied, monitored to record nest- ing success. Results and discussion. In 2000, four occupied sites were located in urban areas. In Regina, a 12- yr-old female released in Winnipeg during 1988 paired with a wild 1990 male from Saskatoon. The pair fledged four young, but one young male sub- sequently died from a collision and a young female was turned into the Veterinary College with prob- able secondary carbamate poisoning. In Saskatoon, a pair produced four eggs, but abandoned them after the male disappeared midway through incu- bation. New, since the 1995 survey, are pairs at Moose Jaw and Prince Albert. The Moose Jaw nest site was not located (it moved from the previous June 2003 2000 Peregrine Survey 103 year’s site), but an adult female and fledging fe- male seen in the city in midjuly indicate a pair nested locally. In Prince Albert, a single territorial male was seen several times during the spring and summer. In 2000, three pairs attempted breeding and two pairs fledged young. One pair successfully fledged four young. Though not located, the Moose Jaw site was believed to have fledged at least one young. To date, no occupied rural sites have been confirmed in Saskatchewan. Though anecdotal, no peregrines were seen in the South Saskatchewan River and area in 2000 (B. Hanbridge and R. Ra- fuse pers. comm., and P. Thompson unpubl. data). The Saskatchewan population is part of a larger mid-continental population, but, this population is unlikely to grow in the near future. Alberta, South of 58°N. Contributed by Rob Corri- gan, Alberta Conservation Association (in partnership with Natural Resources Service, Alberta Environment), #111, 4999-98 Avenue, Edmonton, AB T6B 2X3 Ca- nada. Survey area and method. Court (1993) document- ed peregrines as “a relatively common summer res- ident” in central and southern Alberta before their decline in the mid-20th century. In 2000, surveys were conducted from helicopter, boat, and on foot. Coverage was similar to the 1995 survey with 76 of 78 historical and previously known sites (Court 1993, Corrigan 2000) visited as well as suitable sites along several of Alberta’s major river systems. Results and discussion. In 2000, 23 territories were occupied by pairs. Two sites were new including the first rural site on the North Saskatchewan River since the 1960s. Nest sites were located predomi- nately in urban areas or on man-made structures (14). There were nine rural cliff sites, which were located mostly on the Red Deer River (7). All to- gether, 19 successful pairs produced 57 young. Survey results (23 territories) showed an obvious significant increase in the population from 1995 (13 territories). This increase is due in part to the 223 captive-raised young released in central and southern Alberta between 1992 and 1996 (Step- nisky 1996). As well, the 80% (5-9) increase in pairs returning to rural sites indicated a recovery is underway. An average of 3.0 young/ successful pair and 2.5 young/ territorial pair marked an ex- tremely good reproductive year. Of interest was: a Washington coast yearling female recovered near Drumheller, the first known occurrence of a north- ern Alberta banded falcon breeding in the south- ern population (near Edmonton), and the recov- ery of a central Alberta bred peregrine within 100 km of the arctic coast in the Northwest Territories. As Corrigan (2000) points out, such recoveries in- dicate the Alberta population is not closed, and is in fact, mixing with the larger continental popu- lation. Alberta, North of 58°, Wood Buffalo National Park and Adjacent Northwest Territories. Contrib- uted by Rob Corrigan, Alberta Conservation Association (in partnership with Natural Resources Service, Alberta Environment), #111, 4999-98 Avenue, Edmonton, AB T6B 2X3 Canada, Mark Bradley, Wood Buffalo Nation- al Park, Parks Canada, Port Smith, NT, XOE OPO Ca- nada, and Geoffrey Holroyd, Canadian Wildlife Service, Room 200, 4999-98 Ave. Edmonton, AB T6B 2X3 Ca- nada. Survey area and methods. Peregrine Falcons {P. p. anatuni) were first observed in Wood Buffalo Na- tional Park in 1966 (Moore et al. in press). The population low reached three pairs in 1975 and remained below ten occupied sites until 1992- From 1993-95, pairs increased from 12-23 in part due to increased search effort, which located four new sites. The area surveyed in 2000 included 27 known sites in northeastern Alberta and 21 sites within Wood Buffalo National Park (including both the Alberta and Northwest Territories por- tions of the park). In addition, approximately 100 cliffs were surveyed on the granite Canadian Shield east of the Slave River in northeast Alberta and the adjacent Northwest Territories. In 1995, 142 cliffs were surveyed in the same area. Most breeding sites were visited by helicopter, but sites near Fort Chipewyan were visited by boat and a few near Fort Smith were checked on foot. These sur- vey methods were similar to 1995. Results and discussion. In 2000, territorial pairs oc- cupied 29 sites; only 23 attempted breeding. Of those, eight were successful, producing 21 young for a mean of 2.6 young/ successful pair and 0.7 young/ territorial pair. The number of occupied territories in 2000 (29) was higher than 1995 (23). This increase was not due to additional survey effort, but represents ad- ditional pairs occupying sites and suitable cliffs that were surveyed both in 1995 and 2000. Reproduction was extremely poor. While the ma- jority of territorial pairs laid eggs, a number of pairs around the Fort Chipewyan area failed to fledge young. Weather may have been a factor, al- though sustained, cold, wet weather was not re- 104 Rowell et al. VoL. 37, No. 2 corded in June at the Fort Chipewyan weather sta- tion. Reproduction per territorial pair was higher, over 1.5 young, from 1987-99 (G. Holroyd and M. Bradley unpubl. data), so we suggest that 2000 was an anomalous year. Southwestern and Interior British Columbia (BC). Contributed by Michael J. Chutter, British Colum- bia Wildlife Branch, Ministry of Environment, P. O. Box 9374, Stn. Prov. Govt., Victoria, BC V8W 9M4 Canada. Survey area and methods. Peregrine Falcons {F. p. anatum) in BC occupy the interior, lower mainland (Fraser Valley), southeast Vancouver Island, and the Gulf Islands. A helicopter survey of the lower mainland on 15 June visited eight previously known sites in the Fraser River area. Activity at two other sites was confirmed from the ground prior to the helicopter survey. All of the Vancouver Is- land areas, including new areas along the west and east coasts and the Gulf Islands, received more complete and expanded coverage than in previous surveys. The Gulf Islands were surveyed by boat on 15 May and 17 May and by helicopter 29-30 May. Vancouver Island was surveyed by boat 5-8 June and by helicopter on 15-16 June. Results and discussion. During 1996, surveys in the South Okanogan located a territorial pair — the first in the area since 1959 (Chutter in press). From 1998-2000, 30 captive-raised young were re- leased in and near Kelowna (M. Krupa pers. comm.). In 2000, incidental reports confirmed a territorial pair in the Thompson/Nicola area of the southern interior. Two other possible pairs were unconfirmed, one in the Thompson/Nicola area and one in the Williams Lake area. On the lower mainland, six sites were occupied; one site by a single territorial adult and five by territorial pairs. In the southeast Vancouver Island and Gulf Islands, 11 known sites were surveyed along the southeast coast and in the Gulf Islands, plus two new sites and several potentially suitable cliffs. Nine known and two new sites were occupied — nine with territorial pairs and two with single ter- ritorial adults. Reproductive output was not deter- mined. In 1995, Paul DeBruyn visited 38 known anatum sites in BC (Chutter in press). Of these, 21 were on cliffs near lakes, rivers, and coastal areas of the lower mainland and southwest Fraser River area in- cluding the Gulf Islands. Eleven sites were in the southern interior, four in the central interior and one each in the northwest and northeast interior areas. DeBruyn reported 19 sites occupied (17 in the southwest, a new site in the Thompson area of the southern interior and one historical site in the northern part of the central interior) . Thus, in the southern interior, lower mainland and southeast Vancouver Island and Gulf Islands, the 18 occu- pied sites in 2000 was comparable to 1995 results of 19 occupied sites, although areas surveyed may have differed slightly between years. Queen Charlotte, Langara, North Vancouver, and Scott Islands, British Columbia. Contributed by Michael J. Chutter, British Columbia Wildlife Branch, Ministry of Environment, PO. Box 9374, Stn. Prov. Govt., Victoria, BC V8W 9M4 Canada. Survey area and methods. Coastal Peregrine Fal- cons {F. p. pealei) have been surveyed by the British Columbia (BC) Wildlife Branch on the Queen Charlotte Islands since the early 1960s and in the north Vancouver Island area starting in 1980 (Chutter in press). Wayne Nelson surveyed pere- grines of Langara Island annually since 1968 (Nel- son in press). In 1995, 87 sites were occupied in this area including 62 sites in the Queen Charlotte Islands, 7 on Langara Island, 10 on the north shore of Vancouver Island and 8 on Triangle Island. In 2000, the main survey of the Queen Charlotte Is- lands was conducted by boat from 18-28 May with additional surveys by boat of Langara, Reef Island, and Limestone Islands later in May and early June. The west side of Kunghit was checked in June. Results and discussion. All combined, the 2000 survey was probably the most extensive survey of this area to date. In total, 160 of 162 known sites, six new sites, and many other suitable sites were surveyed on the above islands. As well, the north and northwest shore of Vancouver Island, the sur- rounding Scott Islands, other offshore islands, and parts of the adjacent mainland were surveyed in- cluding 31 known, six new, and many potential cliffs. On the Queen Charlotte Islands, including Lan- gara Island, 69 sites were occupied with 51 pairs defending nests. The remaining 18 sites had per- sistent single birds in attendance and may well have been pairs, hence the number of pairs re- ported here is probably an underestimate. As oc- cupancy was the focus of the survey, reproductive data were not collected except on Langara Island, where Wayne Nelson surveyed between 31 May and 7 June. Nelson recorded nine occupied territories with two single adults and seven pairs of which five successful pairs produced nine young. For north Vancouver Island and Triangle Island, 27 of 37 sites June 2003 2000 Peregrine Survey 105 surveyed were occupied: 18 sites had pairs defend- ing or an adult with a young, nine had single birds present. The total number of territories occupied by pea- lei in BC in 2000 (96) increased over the number in 1995 (87) . This increase occurred on the north part of Vancouver Island and other surrounding islands, which was surveyed more intensively in 2000 than in previous surveys. Thus, we believe the increase to be largely due to more complete survey effort. Overall, the BC pealei population is consid- ered to be stable to slightly increasing. Yukon Interior. Contributed by Dave Mossop, Yukon College, Box 2799, Whitehorse, YT YIA 4K4 Canada. Survey area and methods. Peregrine Falcons {F. p. anatum) breed on the cliffs of rivers draining the central Yukon. The 2000 survey was designed to monitor reproduction in the previously best-mon- itored core areas of each sub-population. Two sub- drainages of the Peel River were not surveyed in 2000 resulting in a smaller sample of 31 different sites in 200 compared to 40 in 1995. The sched- uling of a single survey was planned to coincide with observing young approximately 20 d old. Vol- unteers were divided into six teams that accessed sites by boat between 27 June and 20 July. The Southern Lakes area was not surveyed in 2000 ex- cept for a visit to one unoccupied site. Results and discussion. Of 132 sites visited in the central Yukon during 2000 on the Porcupine, Yu- kon, and Peel River drainages, 88 known and 15 new sites were occupied for a total of 103 territo- ries. Of these, 55 pairs showed successful repro- duction. Although a total of 113 territories was recorded in the previous survey, based on the comparison of the same survey areas, a growth in occupancy of approximately 2.5% annually occurred since 1995 (Mossop 2000) . Peregrines on the Porcupine River declined in the late 1960s, but retained a remnant population (Mossop in press) . This population was the first to recover and increased up to 1990. It has remained stable over the last decade. Peregrines on the Peel River declined in the 1960s, but also retained a remnant population. They increased slowly in 1990 with significant growth in 1995 (Mossop in press) . The Yukon River population de- clined throughout the early 1970s and with only one known occupied territory remaining, captive- raised young were fostered there from 1978-92. By 1990, this population was well above known histor- ical levels. Over the past five years, this population appears to have stabilized with 46 pairs identified in both the 1995 and 2000 surveys. The small num- ber of known peregrines in the southern lakes pop- ulation disappeared during the 1970s. In 1990, this population was thought to be extirpated; however, one successfully-occupied territory was located m 1995. The production of young was very low in 2000, but at least 53% of territorial pairs produced fledg- lings. Clearly, monitoring on a 5-yr cycle poses some risk of hitting below average years of repro- duction making the results difficult to interpret. Yukon North Slope. Contributed by Dave Mossop, Yukon College, Box 2799, Whitehorse, YT YIA 4K4 Ca- nada. Survey area and methods. The 2000 Yukon survey was designed to monitor reproduction in sample areas of sub-populations of which the North Slope drainage is the only Yukon area surveyed for F. p. tundrius. Surveys were conducted by helicopter on 11-14 July. Results and discussion. The survey of 16 known and four new territories resulted in the location of nine occupied sites. Seven territorial pairs raised 15 young for a mean of 2.1 young/ successful pair. Peregrines (F. p. tundrius) on the North Slope of the Yukon were thought to be extirpated locally in 1980 (Mossop in press). Reintroductions of cap- tive-raised young were conducted from 1983-85. In the 1990 survey, a single adult occupied a territory that had previously been occupied by a pair with three eggs in 1989 (Holroyd and Banasch 1996). Annual surveys since 1990 showed this population was in the initial stages of recovery. By 1995, there were five known nesting pairs, three produced young (Mossop in press) . The number of occupied territories (9) on the Yukon’s North Slope drain- age increased since the 1995 survey (5 occupied territories) . Within the same area surveyed during both surveys, two new nesting pairs were located and one pair disappeared in 2000 (Mossop 2000). Iwavik National Park staff also cooperated, for the first time, in this year’s effort, locating three new pairs and expanding the area now included in the survey. The North Slope population remains of concern with only seven known pairs found at the 20 territories visited. Mackenzie Valley, Northwest Territories. Contrib- uted by Steve Matthews and Suzanne Carriere, Depart- ment of Resources, Wildlife and Economic Development, Government of the Northwest Territories, Scotia Centre 5th Floor, Yellowknife, NT XI A 3S8 Canada. 106 Rowell et al. VoL. 37, No. 2 Survey area and methods. Surveys for Peregrine Falcons {E p. anatum) along the Mackenzie River have occurred over the last three decades (Shank in press). The 2000 survey covered the Mackenzie Valley from Saline River, about 80 km upstream from Tulita (Fort Norman), to Inuvik, Northwest Territories (NT) . The area surveyed in 2000 is the same as that surveyed in both 1990 and 1995, with some minor exceptions. Ten nest sites recorded in 1995 were not examined in 2000 due to fog, time constraints, or unknown coordinates. We surveyed the entire study area by helicopter from 17-21 July. From 12-22 July, Keith Hodson surveyed by boat the portions of the study area dh rectly on the Mackenzie River. All sites surveyed by boat were also surveyed by helicopter, but not vice versa. Results and discussion. A total of 118 known sites were surveyed. As well, four new sites were found. Of the known sites, 76 (64%) were occupied in 2000. Only 37 pairs of 80 occupied sites were suc- cessful with a total of 80 young fledged from 36 sites. Mean reproduction was 1.0 young/ territorial pair and 2.2 young/successful pair. Mackenzie Valley peregrines increased dramati- cally through the 1980s until about 1990, when they were believed to have reached “ecological car- rying capacity.” Numbers have remained fairly constant during the last decade, despite some an- nual fluctuations in breeding success. In 1995, 83 sites were occupied (Shank in press). In 2000, oc- cupancy (80 sites) was slightly less than in the 1995 (83 sites) and 1990 (88 sites) surveys. However, some apparent decline may be due to the ten known sites not visited in 2000. The surveyed por- tions of the Mackenzie Valley may have reached a maximum occupancy for Peregrine Falcon terri- torial pairs in the 1990s. Similar to the 1995 survey (Shank in press), forest fires might have affected nesting occupancy and success. In 2000, we noted that some sites were lost due to fire: old raven stick nests previously used by peregrines were burned or the cliff face was modified by slumping and mud deposition. In 2000, reproductive output, measured as the mean number of young/ reproducing pair, was not significantly different from the mean of 2.3 record- ed during surveys between 1969 and 1995 {t = —0.21, df — 22, P > 0.5). Despite habitat changes and some annual variability in reproduction dur- ing the last decade, the number of sites occupied by Peregrine Falcons during the last decade in the Mackenzie Valley has been the highest since the survey was initiated in the mid-1960s. Rankin Inlet, Nunavut. Contributed by David Aber- nethy and Josh Hunter, Nunavut Dept, of Sustainable Development, 3rd Floor, Brown Building, P.O. Box 1340, Iqaluit, NU, XOA OHO Canada, and Gordon Court, Natural Resources Service, Alberta Environment, 6909- 116 St. Edmonton, AB T6H 4P2 Canada. Survey area and methods. Peregrine Falcons {E p tundrius) have been under intensive study at Ran- kin Inlet, on the northwest shore of Hudson’s Bay, Nunavut, since 1980 (Court et al. 1988b). Popula- tion size and reproductive performance of all pairs within the circumscribed boundaries of the 450- km^ study area were reported in national peregrine inventories in 1985, 1990, and 1995. In 2000, a sin- gle survey for territorial occupancy and reproduc- tive output at 39 known nesting cliffs was conduct- ed in late July. Results and discussion. In the 2000 survey, 22 pairs of adults and three lone birds were detected at 25 sites. Sixteen successful pairs had a total of 37 young yielding 1.7 young/ territorial pair and 2.3 young/successful pair. Up to 29 pairs (annual mean 24) have occupied territories in the study area in any given year since survey boundaries and protocols were standardized in 1982. Bradley et al. (1997) summarized breed- ing success for the peregrine population over 13 years (1982-94) and found that in most years, 13 territorial pairs fledged young and, on average, 1 .4 young (0.6-2. 5 per year) were fledged/ territorial pair; 2.5 were fledged/successful pair (1. 8-3.1 per year) . Weather conditions and prey abundance are known to have dramatic effects on the reproduc- tion of this population (Court et al. 1988a, Bradley et al. 1997). Comparing these data to historical measures of occupancy and reproductive perfor- mance, we suggest that 2000 was an average to above-average year for the Peregrine Falcons at Rankin Inlet. Tuktut Nogait National Park, NT. Contributed by Joachim Obst, Box 1888, Yellowknife, NT XI A 2P4 Ca- nada. Survey area and methods. In 1988 and 1990, exten- sive ground surveys for raptors were conducted throughout Tuktut Nogait National Park, northeast Northwest Territories. In 1991, 1996, and 1999, supplemental data were collected from visits to ad- ditional nesting sites. In total, 53 peregrine {F. p. tundrius) territories were identified in the park in- cluding 31 sites on the Hornaday River. In 2000, June 2003 2000 Peregrine Survey 107 from 17-26 July, three Parks Canada teams sur- veyed 160 km of the Hornaday River in the western part of Tuktut Nogait National Park by boat and on foot. Surveyors revisited 18 known sites visited previously in 1988 or 1990, as well as four new ter- ritories within the survey area. Sites were observed for 0. 5-3.0 hr using a spotting scope. Results and discussion. The 2000 survey confirmed 13 reproducing pairs, a territorial pair, and a ter- ritorial adult male in 15 of 18 known territories. Occupancy could not be ascertained at three sites due to inadequate survey time, although at two of these, fresh whitewash and/or prey remains were present indicating presence of young. In addition, four new territories with reproducing pairs were found in 2000 for a total of 19 occupied territories. Seven pairs had 18 young for a mean of 2.6 young/ successful pair. Another three pairs were observed with one or more young, however, numbers of young were not confirmed. All of the remaining pairs, except for one territorial pair, were assumed to have young because they displayed vigorous de- fense behavior at known nest sites. Results for 2000 indicate a stable population of Peregrine Falcons on the Hornaday River for the past decade. Although only 15 of 18 territories oc- cupied in 1988 or 1990 were used again in 2000, four new territories in 2000 could represent a shift of pairs to alternate territories. Densities were higher in canyons and average spacing of occupied territories was 1.6 km {N— 10) in one section (Zol- tai et al. 1992). Pairs averaged 3.1 eggs {N — 7) in 1988 and 3.2 eggs {N = 5) in 1990. During all sur- veys, it was not feasible to count the exact number of eggs or young in all nests but most pairs were assumed successful based on their behavior. To clarify the uncertain status of assumed “unoccu- pied” territories in future surveys, more time is needed for ground observations. Tuktut Nogait National Park offers the opportunity to monitor Peregrine Falcons and other raptors at a relatively low cost through ground surveys. Synthesis and Conclusion Since 1970, national surveys have been conduct- ed every five years to locate breeding Peregrine Falcons. Survey results present the known breed- ing population but do not reflect the total pere- grine population size. The total population would include peregrines in areas that were not surveyed, failed breeders that abandoned their sites before surveys, non-breeding subadults and non-breeding adult floaters, which do not have territories. Regions surveyed for anatum peregrines in 2000 (Table 1) were comparable to those surveyed in 1995 but effort within specific areas varied. The 2000 survey effort was less than in 1995 in Mani- toba and Saskatchewan because the northern parts of neither province was searched in 2000. However, effort to identify urban pairs in the southern por- tion of each province was similar to 1995. Similarly, fewer cliffs east of the Slave River in northern Al- berta were surveyed in 2000 than in 1995, but this did not affect numbers in the main study area which had similar survey effort in 1995 and 2000. Anatum surveys in the Yukon were concentrated on core areas and two drainages of the Peel River sur- veyed in 1995 were omitted in 2000. Similarly, 10 sites surveyed in the Mackenzie Valley in 1995 were not surveyed in 2000. Survey effort for anatum was greater in 2000 in Labrador where more inland sites were surveyed. However, surveys were conducted one month later than they were in 1995. Greater effort was also ex- pended in 2000 in the Maritimes, Ontario, and Quebec where park staff and volunteer naturalists contributed sightings that were then verified and included in survey results. As previously noted for Ontario, however, effort for annual surveys of the west end of Lake Superior has been the same since 1997. Effort was also greater in 2000 for the south- ern portion of Vancouver Island, BC. Effort for anatum in southern Alberta, interior BC and the Porcupine and Yukon Rivers, YT, was similar in 2000 to effort in 1995. Area surveyed for tundrius (Table 2) was expand- ed from previous surveys on the Yukon’s North Slope, however, a comparison was provided in the detailed regional account that compared the area covered in both 1995 and 2000. Similarly, data from Tuktut Nogait National Park, added to na- tional results for the first time, were pulled from larger raptor surveys, but represent the findings of a comparable area searched in 1988 and 1990, and 2000. Although the area surveyed was the same at Rankin Inlet, effort was reduced with only one sur- vey (July) conducted compared to two (May and July) in 1995 and many in 1985 and 1990. Ungava Bay tundrius were not surveyed in both 1995 and 2000. Survey efforts for pealei (Table 2) were similar to previous years except on Vancouver Island, where the 2000 survey was the most comprehensive to 108 Rowell et al. VoL. 37, No. 2 Table 1. Sites occupied by anatum Peregrine Falcons in Canada during 2000. ND = no data. Area Known Sites Known Sites Checked Known Sites Occupied New Sites Sites Occupied: Single Birds Pairs Labrador, Newfoundland Bay of Fundy (Nova Scotia, New Bruns- 63 63 22 0 7 15 wick) 15 15 8 3 0 11 Southern Quebec 23 23 16 12 3 25 Southern Ontario 76 76 38 15 11 42 Southern Manitoba 21 21 3 0 1 2 Southern Saskatchewan 4 4 4 0 1 3 Alberta South of 58°N 78 76 21 2 0 23 Interior, British Columbia 17 1 1 0 0 1 Lower Mainland, British Columbia Gulf Island and SE Vancouver Island, 10 9 5 1 1 5 British Columbia 11 11 9 2 2 9 Alberta North of 58°N 48 48 27 2 0 29 Porcupine River, Yukon 45 36 26 9 0 35 Peel River, Yukon 48 28 19 3 0 22 Yukon River, Yukon 57 53 43 3 0 46 Southern Lakes, Yukon 1 ND ND ND ND ND Mackenzie Valley, Northwest Territories 128 118 76 4 0 o 00 Total anatum 645 582 318 56 26 348 ■* Although only a single adult with no young was observed at 26 of these sites, given the timing of surveys (mid-July), these sites are included in the number of territorial pairs. date. Data collected in over half the regions sur- veyed were comparable between the 1995 and 2000 (Table 3). Surveys attempted to cover all “historical or known sites” as well as potential sites of suitable nesting habitat. In the regions surveyed, most (ca. 90%) known sites were visited. Regions with unvis- ited known sites included southern Alberta (2), mainland BC (17), the interior Yukon (34), and Mackenzie Valley (10). Although measures of reproductive success (Ta- bles 4 and 5) and comparisons between 5-yr sur- veys (Tables 3 and 6) are made below, differences in survey effort, timing, and conditions must be taken into account before drawing conclusions. Where single surveys are conducted, only a mini- mum number of occupied territories were docu- mented. Nesting pairs that failed prior to surveys and single adults that have less affinity to their ter- ritory may have been missed or undercounted. Subspecies classification is not always definitive. As well, with the expansion of subspecies’ ranges, introgression may be occurring. Although consid- ered anatum for this paper, the Labrador, New- foundland population is likely on the boundary of tundrius and anatum. In southwestern BC, the Fra- ser River and coastal mainland population is within the historical range of anatum. Peregrines on the Gulf Islands and along the southeast shore of Van- couver Island are also believed to be anatum (W. Nelson and C. White pers. comm.). And finally, known cases of anatum adults in southern Ontario breeding with birds of mixed race from the U.S. seaboard have occurred in recent years. To date, such subspecies classification was based on histor- ical breeding range and physical characteristics. The identification of a genetic marker that could be used to distinguish these subspecies would be extremely worthwhile. With these limitations in mind, it is important to realize that surveys provide a best estimate of the number of breeding peregrines within an area and do not reflect the total population size. Although comparisons of surveys conducted over 30 yr have their limitations, this large body of data provides a comparable picture of peregrine population trends in Canada. Anatum Populations. A total of 374 sites were oc- cupied by anatum in 2000 (with 93% occupied by pairs), the largest number of occupied territories June 2003 2000 Peregrine Survey 109 Table 2. Sites occupied by tundrius and pealei Peregrine Falcons in Canada during 2000. ND = no data. Area Known Sites Known Sites Checked Known Sites Occupied New Sites Sites Occupied: Single Birds Pairs Tundrius Ungava Bay, Quebec 58 0 ND ND ND ND North Slope, Yukon 19 16 5 4 2 7 Rankin Inlet, Nunavut 39 39 25 0 3 22 Tuktut Nogait National Park, Northwest Territories 50 18 15^ 4 1 18 Total Tundrius 166 73 45 8 6 47 Pealei Langara Island, British Columbia 9 9 9 0 2 7 Queen Charlotte Islands, British Columbia 153 151 54 6 16 44 North Vancouver and Scott Islands, British Columbia 24 24 14 6 8 12 Triangle Island, British Columbia 10 7 7 0 1 6 Total Pealei 196 191 84 12 27 69 ® Two additional sites showed evidence of occupancy (fresh whitewash and/or prey remains) but were not confirmed active. recorded to date during the 5-yr surveys. Between 1995 and 2000, the total number of occupied ana- tum territories observed increased by approximate- ly 15% (324 to 374) (Table 3). Although this value is influenced by variation in search effort and tim- ing, it is less than the 36% increase observed be- tween 1990 and 1995 (Banasch and Holroyd in press). This declining rate of increase may be in- dicative of a recovering population that is ap- proaching carrying capacity. Numbers for north- ern areas representing the largest concentration of peregrines (the Mackenzie Valley, Yukon drainages and northern Alberta) were similar to those re- corded in 1995 (219-212) indicating that these re- gions may have stabilized and/ or are approaching saturation. In contrast, the number of occupied territories in southern Canada increased by 54% (105-162). Southern Ontario showed the largest increase, more than tripling its population (15—53) between 1995 and 2000. While some increase was due to increased search effort, at least part of the popu- lation growth is attributed to natural increase. A large number of captive-raised young anatums re- leased in Ontario by hacking through to 1996 are likely responsible for at least a portion of the cur- rent breeding population. Other anatum popula- tions experiencing significant growth in the num- ber of occupied territories included the Bay of Fundy, southern Quebec, and southern Alberta, all nearly doubled in size since 1995. Labrador pere- grines were the only anatum population to register a decline (29%) in the number of occupied terri- tories. The 15% increase in the number of occupied territories (occupied by a single or pair) is greater than the 9% increase in the number of territorial pairs (319 pairs in 1995 to 348 in 2000). While most anatum subpopulations showed increases in the number of territorial pairs, the overall total was reduced by fewer breeding pairs in Labrador and Manitoba and in smaller areas surveyed on the Peel River, Yukon. Newton (1979) described a sta- ble population as one in which “breeding numbers remained absolutely constant or changed by less than 15% of the mean over the period con- cerned.” In 2000, all anatum subpopulations showed stable or increasing growth rates in breed- ing pairs over 1995, except for those in Labrador and southern Manitoba, which both declined by approximately 50% (Table 3). Although there was some regional variation, overall anatum reproductive outputs remained sim- ilar to those of 1995 (Table 6) and were clearly not statistically different between the most recent sur- veys. Mean young/ successful pair rose slightly from Table 3 Number of sites occupied by Peregrine Falcons in selected regions of Canada from 1970 to 2000 Numbers in parentheses indicate the number of territorial pairs present. Numbers in italics indicate 2000 surveys (area and effort) were comparable to 1995 . ND = no data. no Rowell et al. VoL. 37 , No. 2 o o o GU 05 05 o 05 05 CO 00 I 00 05 o 00 05 i> 05 o I> 05 CO CO I CO 05 lO CM CM CO CM CM Q CM CM in CM ^ -N CM ^ ([y 00 CO ^ CT 5 CM m comm^'^^'^'^ — I ^ S' o 0, ^ w Q 0. ^ 1C) ^ ^ ^ rH rM cn CM > 00 ^ rr, Q\ 'o i> CO 05 On j> 'o CM cs| ^ ^ 00 CO 1> CO CM Q I> CM CM 00 J> On CO m On S o CO ^ CM V in ’ — I CM P - ~ - *z. 'z. l>incOOMCMCO CO Q Q Z Z ^ P 1-V 1-V 1> 00 p ^ p In* fh 00 0 CM '■ — ' 'll — * — ^ ^ P CO 00 05 P ^ P rH cO 00 00 CM CO CM On 05 ON o 00 1> CO in CO y"V y , , Q ^ Q ^1. q — V r — V y.— V V y— V y^ y— V y-v y-. ^ 0 CC HH *—w 00 P 00 ^y— 1.^ CM 1— H r—i 0 ^ CM Q '~^'~'Z CM CM Q 1-0 1 —i z 10 CM Z CM P z '' — ' ' — ^ ' ■' ■'• Z ' — '' ^ ^ ' ' — ' ' — ' * — '' cn rH T-^ ^ ^ UD ^ (M (M p CO CO P 05 P 0 th 1-h cn p CM CM P p Q QQ QP 00 00 11 l1 yv y— V. p 10 CM P cn lO Q hH 0 CM P t-H p 0 0 0^0 :q q : z z Q CM P Q 0 Z 2 CO m CO 05 O CM £,Q P 1^ 1^ CM 0 Si P P z z p z CO 00 Q Z P P P P ^ z z z ^ !S s 'O c a 3 I OJ z sO o cS 5 - rt P o cj cn o Z .2 jx, ^ ■•O 1/3 s =* p p ^z PQ u o ■o ^ ao G G u ^ o OJ P O 4-i o o cn c /5 c o 3 - CJ n Si o O O P Gh G O M G P O ^ M „ O P 2 S .21 5 V ‘C 0 *u u pj VD % p PJ 5 p o Z 0 1 .2 'n G o P CJ CS s n 3 flH C 3 Qj >- PQ Z Eh 13 a 'o U P c/5 • ^ pj 'C PQ dS G cS , , y-v y^ CO CM 0 P 0 5 J 00 00 p pH Q q Tfl pH pH pH ii pH 0 00 Q P S Z Z CO p y-S. ^ , p ii CT) iS ND ND p CM p p CM pH P P pH Q Q Q Q p P P Q pH p Z Z Z z P 1> z '' ^ CM CM 05 p 00 V )§ TId § § • fh ^^2 q ^ e2 d OJ pj a cr CU pn X} P T 5 V JT > u p 5/5 cd 0; Sh d d P ^ > u p 5 C O 5/5 u d 5 ^ p p ^ § I j; ^ p s C HH °.a P-) *c« q r-- t-H CM CM i-H i-H t-H o Q Q Oi 1— H d t-H cO y-^ rH Q rH d Q Q Q ' — ' ' — ' ' ' ' — ^ Z Z '' — ^ ' — ^ ' — ^ '' — ^ '' — ^ z ^ '■ ^ z z z I-H CM q q q q q CO q q q q rH CM o O CM CM CM l-H CO CO c4 c4 oi CM CO CM CM CM CM CM CM CM O cn cr> tn 00 I lO 00 cn o 00 O) in CT) O V' q 00 q q q 1> q q 5D q q q rH q q q q GO rH CM rH CM Q Q Q o d ^ Q Q CM rH rH rH rH Q Q p rH ^ ^ ^ ^ o ^ z z ^ ' — ' z '■'— ^ z z z q GO q O q 50 t-H CM CM la rH GO CO CO GO CM CM CM CM nh CM on CM CM rH on CM CM CM CM CM CM rH ^H q q q q lO q q q q rH q q CM rH rH rH rH d rH rH rH CM rH CM rH CM '■ — — ■' — ^ ^ ^ ■' — '' ^ ^ ^ ' 4 q q q q q q q q X q q q q ^H on CM rH CM CM rH rH CM CM on CM CM CM on s 52 00 O in Gsi Ol 00 m q on o o o o o Q 2; Q Q Q Q Q Q ® Z ® Z Z ® o Q Q £ ^ ^ ^ O q o o on CM 00 00 Q Q 2: 2; o q CM o o o Q m q on o o Q Q ^ ^ 2 Z ^ csi • 'PN .£3 a JS fl q o Z s-T O "O oj Sm JD oS Pi • 0 u c/:) 1 " b! PQ 4-1 :> ^ £ CQ o O ^ ’C ■“ Ki U -M S. ci OfO a s !h W -5 -5 IP P o o cn CO ci X! O 'p n p C3 U J= u +J C3 d 05 _P ■q U P3 in ‘P 'G P3 4i ^ 2 "G P-i S «r q U P P !h 4 D 4J •S -S p p O O 05 05 V P •5 -5 p p o o 05 05 rP tn 'C eq nd P 23 [p 'p O P w 05 T3 P P o S sP u PQ 7 ' — ' CO q q ^ X rH CM on CM q q GO* q q Q CM rH rH ^ Q P rH CM Z Q CM ' — ^ z '■ — -' ^ — -' q q q q CM CM rH rH CM CM CM 2 3 S 'o U ‘G PQ S 23 ^ 2 rP in Sm PQ 'p p (t in P d P 53 <4 C3 PC ^ rH ^ |s| Jd“ p 4h P PC *H 4h o Z p g 114 Rowell et al. VoL. 37, No. 2 tion from other raptors, and human disturbance may also play a role in limiting occupancy and re- production in some areas. An effort to understand the impact of such factors will ensure this species’ existence is maintained in the future. Acknowledgments The editors thank the many staff and volunteers who participated in the 2000 survey from coast to coast to coast. In Labrador, survey crew members included G. Goodyear (Universal Helicopter), K. Knox (Jacques Whitford Limited [fWL]), V, Busque (Department of Na- tional Defence (DND), J. Brazil (Inland Fish and Wildlife Division, T. Chubbs (DND) and D. Amirault (Canadian Wildlife Service (CWS)). Additional financial and logis- tical support was provided by P. Trimper ( JWL) , K. Oram (JWL), M^. G. Humphries (DND), the Canadian Coast Guard, C. Bird (Parks Canada) , Inland Fish and Wildlife Division (Labrador), and D. Amirault (CWS). In New Brunswick and Nova Scotia, many naturalists and inter- ested individuals provided information and sightings in- cluding K. McCurdy, D. Baldwin, D. Christie, J. Wilson, J Wolford and A. Robichaud. Survey assistance was pro- vided by the Department of Fisheries and Oceans, Ca- nadian Coast Guard with funding from CWS (Environ- ment Canada), Species at Risk program, Nova Scotia Department of Natural Resources and New Brunswick Department of Natural Resources and Energy. In South- ern Quebec, assistance was provided by P. Laporte (CWS), B. Blais — head of the survey team, P. Fradette — head of the Inventory Program for Threatened Bird Spe- cies in Quebec for the Association Quebecoise des Crou- pes d’Ornithologues, the staff of Societe de la Faune et Des Parcs and many ornithologists who provided survey data. In Southern Ontario, volunteers donated time to check suitable habitat. TransCanada Pipelines and Hydro One Helicopter Services donated helicopter time to sur- vey remote cliffs near Lake Superior. The Canadian Per- egrine Foundation’s website offered information on the urban sites in southern Ontario. Funding for the survey was through Ontario’s Living Legacy Species at Risk Pro- gram. Many organizations and media promoted the sur- vey through publications, newsletters, and websites. In Southern Manitoba, the survey was supported by Mani- toba Conservation, the Manitoba Peregrine Falcon Re- covery Project Committee, and the Manitoba Species-at- Risk Recovery Working Group. Networkx provided an Internet presence. Staff and management of the Radis- son Hotel (Winnipeg) and McKenzie Seeds (Brandon) and many volunteers provided sightings and access. J. Shepherd coordinated Brandon operations. In Southern Saskatchewan, assistance was given by K. Holiday (Regina media contact), J. Triffo (Regina site monitor), J. Kros- hus (Moose Jaw contact person) and many volunteers who built extra nest boxes for urban locations around the province. In Alberta, survey crews included staff from Alberta Environment, Natural Resources Service (G. Erickson, P. Young, K. Froggatt, M. Christensen, G. Court, S. Cotterill and D. Moore, K. Morton, W. Nelson) and Alberta Conservation Association (D. Fairless, B. Pat- terson, D. Park, J. Melzer, and R. Corrigan) the CWS (G. Holroyd); and Parks Canada (M. Bradley, B. Reside and C. Wickenheiser) . For Southwestern and Interior BC, D. Dunbar, Surrey Wildlife Branch, coordinated the helicop- ter survey in the Lower Mainland area. Several individu- als reported additional peregrine sightings, especially J Hobbs who personally investigated a new active site in the interior of the province. The Hahitat Conservation Trust Fund and the Nestucca Trust Fund provided fund- ing assistance. For Coastal BC, Queen Charlotte Island surveys would not have occurred without D. Buries and D. Garner, Parks Canada, who provided the zodiac boat and K. Row- sell and R. Erickson of the Anvil Cove mothership. W Nelson provided nesting and breeding success of pere- grines on Langara Island. G. Schultze, Smither’s Wildlife Branch office, coordinated surveys. The Vancouver Is- land surveys were coordinated by D. Doyle, Nanaimo Ministry of Environment, Lands and Parks (MELP). As well, a number of Wildlife Branch staff, boat pilots (from the Conservation Officer Service) and helicopter pilots (Long Beach Helicopters) assisted surveys. BC Parks al- lowed us access to sensitive seabird colonies. In the Yu- kon, this project was a partnership between the Yukon College and Yukon Department of Renewable Resources. Funding was provided by the Polar Continental Shelf Pro- ject, the Yukon Government endangered species pro- gram and Yukon College Biodiversity Assessment and Monitoring program. N. Speiss, G. Boros, J. Gordon and D. Simpson carried out the survey on the North Slope. As well, 39 volunteer field crew carried out the remaining surveys. In the Mackenzie Valley, NT, funding was provid- ed by the Government of the Northwest Territories, De- partment of Resources, Wildlife and Economic Develop- ment (RWED), and by Environment Canada, Canadian Wildlife Service. K. Hodson, M. Gravel, P. Rivard, and RWED personnel in Wrigley, N. Wells, Fort Good Hope, and Inuvik provided assistance, hospitality and support In Rankin Inlet, Nunavut, assistance was provided by Siu- ling Han of Sustainable Development, Government of Nunavut, The Ahighik Hunters and Trappers Associa- tion, Rankin Inlet, First Air, and A. Karvonen of Reel Girl’s Media Inc. P. Kolit, D. and T. Hamilton, and D. Steele provided additional assistance. In Tuktut Nogait National Park, NT, survey team members included C. Bu- cher, K. Ruben, E. Green, and J. Edwards. I. McDonald, Parks Canada in Inuvik managed logistics. The commu- nity of Paulatuk provided other assistance. The editors would also like to thank Peregrine Falcon Recovery Team members and regional compilers who submitted and reviewed regional summaries. W. Nelson provided thought-provoking commentary on the docu- ment in its entirety. C. White, J. Enderson, and an anon- ymous referee reviewed the manuscript prior to publi- cation. The Canadian Wildlife Service, Environment Canada, provided funding for the production of this re- port. Literature Cited Amirault, D.L., S.P. Flemming, M.F. Elderkin, and G. Sinclair. In press. The 1995 Peregrine Falcon survey in the Bay of Fundy. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occasional Paper. Canadian Wildlife Service, Canada. June 2003 2000 Peregrine Survey 115 Armstrong, T. In press. The 1995 Peregrine Falcon sur- vey in Ontario. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occa- sional Paper. Canadian Wildlife Service, Canada. Banasch, U. and G.L. Holrovd. In press. The 1995 Ca- nadian Peregrine Falcon survey. Occasional Paper. Canadian Wildlife Service, Canada. Bechard, M.J. 1981. Historic nest records of the Pere- grine Falcon in southern Saskatchewan and southern Manitoba. Blue Jay 39:182-183. Beebe, F.L. 1960. The marine peregrines of the northwest Pacific Coast. Conrfor 62:145-189. Bradley, D.M., R.M. Johnstone, G.S. Court, and T. Duncan. 1997. Influence of weather on breeding suc- cess of Peregrine Falcons in the Arctic. Auk 114:786— 791. Brazil, J., D. Lemon, and T. Wellicome. In press. The 1995 Peregrine Falcon {Falco peregrinus) survey in Lab- rador. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Ealcon survey. Occasional Paper. Canadian Wildlife Service, Canada. Bromley, M. 1992. Status report on the tundra Peregrine Falcon {Falco peregrinus tundrius) in Canada. COSE- WIC report. Ottawa, ON Canada. Brown, L. and D. Amadon. 1989. Eagles, hawks, and fal- cons of the world. The Wellfleet Press, Secaucus, NJ U.S.A. Cade, T.J. and R. Fyfe. 1970. The North American pere- grine survey, 1970. Can. Field-Nat. 84:231-245. , J.H. Enderson, C.G. Thelander, and C.M. White. (Eds.) 1988. Peregrine Ealcon populations: their management and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A. Chutter, M.J. In press. The 1995 Peregrine Falcon sur- vey in British Columbia. In U. Banasch and G. Hol- royd [Eds.], The 1995 Canadian Peregrine Falcon sur- vey. Occasional Paper. Canadian Wildlife Service, Canada. Corrigan, R. 2000. Survey of the Peregrine Falcon {Falco peregrinus anatum) in Alberta. Alberta Sustainable Re- source Development, Fish and Wildlife Service, Al- berta Species at Risk Report No. 2, Edmonton, AB Canada. COSEWIC, 2000. Canadian Species at Risk, May 2000. Committee on the Status of Endangered Wildlife in Canada. Environment Canada, Ottawa, ON Canada. Court, G. S. 1993. A review of historical nesting records for the Peregrine Falcon {Falco peregrinus anatum) in Alberta south of 56°N: priorities for surveying a re- covering population. Unpubl. report. Alberta Envi- ronmental Protection, Fish and Wildlife Services, Ed- monton, AB Canada. , D.M. Bradley, C.C. Gates, and D.A. Boag. 1988a. The population biology of Peregrine Falcons in the Keewatin District of the Northwest Territories, Canada. Pages 729-739 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, Inc., Boise, ID U.S.A. , C.C. Gates, and D.A. Boag. 1988b. Natural his- tory of the Peregrine Falcon in the Keewatin District of the Northwest Territories. Arctic 41:17-30. Enderson, J.H. 1965. A breeding and migration survey of the Peregrine Falcon. Wilson Bull. 77:327-339. Erickson, G., R. Fyfe, R. Bromley, G.L. Holroyd, D Mossop, B. Munro, R. Nero, C. Shank, and T. Wiens. 1988. Anatum Peregrine Falcon recovery plan. Can. Wildl. Serv. Publication, Edmonton, AB Canada. Fyfe, R.W. 1969. The Peregrine Falcon in northern Ca- nada. Pages 101-114 m J.J. Hickey [Ed.], Peregrine Falcon populations, their biology and decline. Univ. Wisconsin Press, Madison, WI U.S.A. . 1988. The Canadian Peregrine Falcon recovery program, 1967-1985. Pages 599-610 mT.J. Cade, J.H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Ealcon populations: their management and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A. , S.A. Temple, and T.J. Cade. 1976. The 1975 North American Peregrine Falcon survey. Can. Field- Nat. 90:228-273. Hickey, J.J. (Ed.) 1969. Peregrine Falcon populations, their biology and decline. Univ. Wisconsin Press, Mad- ison, Wl U.S.A. Holroyd, G.L. and U. Banasch. 1990. The reintroduc- tion of the Peregrine Falcon, Falco peregrinus anatum, into southern Canada. Can. Field-Nat. 104:203-208. AND . 1996. The 1990 Canadian Peregrine Falcon {Falco peregrinus) survey. / Raptor Res. 30:145- 156. Johnstone, R.M. 1999. Update COSEWIC status report on aiiatrtm Peregrine Falcon {Falco peregrinus anatum) . Committee on the Status of Endangered Wildlife in Canada. Environment Canada, Ottawa, ON Canada. Jones, R.E., D. Chranowski, J. Duncan, T. Maconachie, AND R. Wheeldon. In press. The 1995 Peregrine Fal- con survey in Manitoba. In U. Banasch and G. Hol- royd [Eds.], The 1995 Canadian Peregrine Falcon sur- vey. Occasional Paper. Canadian Wildlife Service, Canada. Kief, L.F. 1988. Changes in the status of the Peregrine Falcon in North America: an overview. Pages 123-129 mT.J. Cade, J.H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Falcon populations: their management and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A. Lemon, D. and J. Brazil. 1990. Preliminary report of breeding Peregrine Falcons, F.p. tundrius, in Labrador, 1987 and 1988 survey results. Can. Field-Nat. 104:200- 202 . Martin, M. 1979. Status report on Peregrine Falcon {Fal- co peregrinus) in Canada. COSEWIC report. Environ- ment Canada, Ottawa, ON Canada. Moore, D., G. Holroyd, and J. Dixon. In press. The 1995 northern Alberta Peregrine Falcon survey. In U. 116 Rowell et al. VoL. 37, No. 2 Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occasional Paper. Canadian Wildlife Service, AB Canada. Mossop, D.H. 2000. Population status of the Peregrine Falcon in the Yukon Territory, 2000. Northern Re- search Institute, Yukon College, Whitehorse, YT Ca- nada. . In press. The 1995 population status of the Per- egrine Falcon in the Yukon Territory. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occasional Paper. Canadian Wildlife Service, Canada. Munro, W.T. and B. Van Drimmelen. 1988. Status of per- egrines in the Queen Charlotte Islands, British Co- lumbia. Pages 69-72 inT.J. Cade,J.H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Falcon populations; their management and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A. Murphy, J.E. 1990. The 1985-86 Canadian Peregrine Fal- con, Falco peregrinus, survey. Can. Field-Nat. 104:182- 192. Nelson, R.W. In press. The 1995 Langara Island Pere- grine Falcon survey. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occasional Paper. Canadian Wildlife Service, Canada. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A. Ratcliff, B. and T. Armstrong. 2000. Ontario’s Pere- grine Falcon survey, 2000. Unpubl. report. Ontario Ministry of Natural Resources, Thunder Bay, ON Ca- nada. Ratcliffe, D. 1969. Population trends of the Peregrine Falcon in Great Britain. Pages 239-270 in J.J. Hickey [Ed.], Peregrine Falcon populations: their biology and decline. Univ. Wisconsin Press, Madison, WI U.S.A. Risebrough, R. and D. Peakall. 1988. Pages 449-462 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, Inc., Boise, ID U.S.A. Sam, D., S. Boates, P. Austin-Smith, B. Johnson, and G. Dickie. 1994. Status of Peregrine Falcon (anatum) in Nova Scotia: a synopsis of recovery efforts. Nova Sco- tia Department of Natural Resources. Technical Re- port No. 74. Halifax, NS Canada. Shank, C. In press. The 1995 Peregrine Falcon survey in the Northwest Territories. In U. Banasch and G. Hol- royd [Eds.], The 1995 Canadian Peregrine Falcon sur- vey. Occasional Paper. Canadian Wildlife Service, Ca- nada. Stepnisky, D.P. 1996. Southern Alberta Peregrine Falcon reintroduction project — summary and evaluation of the 1992-96 releases. Unpublished report. Alberta Environmental Protection, Natural Resources Service, Wildlife Management Division. Edmonton, AB Cana- da. Stocek, R.F. and P.A. Pearce. 1978. The Peregrine Fal- con in the Maritime provinces. Can. Wildl. Serv., Wildlife Toxicology Division. Manuscript Reports No. 36. Halifax, NS Canada. Thompson, P. In press. The 1995 Saskatchewan Peregrine Falcon survey. In U. Banasch and G. Holroyd [Eds.], The 1995 Canadian Peregrine Falcon survey. Occa- sional Paper. Canadian Wildlife Service, Canada. White, C.M. 1968. Diagnosis and relationships of the North American tundra inhabiting Peregrine Falcons. Auk 85:179-191. White, C.M. and D.A. Boyce, Jr. 1988. An overview of Peregrine Falcon subspecies. Pages 789-810 in T.J. Cade, J.H. Enderson, C.G. Thelander, and C.M. White [Eds.], Peregrine Falcon populations: their manage- ment and recovery. The Peregrine Fund, Inc., Boise, ID U.S.A. , R.W. Fyfe, and D.B. Lemon. 1990. The 1980 North American Peregrine Falcon, Falco peregrinus, survey. Can. Field-Nat. 104:174-181. ZOLTAI, S.C.,J. SiROis, and G.W. Scotter. 1992. a natural resource survey of the Melville Hills region. North- west Territories. Technical Report Series No. 135. Ca- nadian Wildlife Service, Western and Northern Re- gion, Yellowknife, NT Canada. Received 19 April 2002; accepted 6 March 2003 /. Raptor Res. 37 (2) : 1 1 7-1 24 © 2003 The Raptor Research Foundation, Inc. BREEDING SEASON HABITAT USE AND ECOLOGY OE MALE NORTHERN PYGMY-OWLS Alan R. Giese^ Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR 97331 US. A. Eric D. Forsman us. Forest Service, Pacific Northwest Research Station, 3200 SW Jefferson Way, Corvallis, OR 97331 U.S.A. Abstract. — We examined habitat use, nest location, diet, and activity patterns of radio-marked Northern Pygmy-Owls {Glaucidium gnoma) during four breeding seasons (1994-97) in fragmented forests on the Olympic Peninsula, Washington. We observed foraging in all available vegetation categories, but patterns of use were non-random. Structurally diverse and older forests were most heavily used, openings and patches of saplings received the least use, and use of edge was intermediate. We located eight nests, all in woodpecker cavities in patches of structurally-diverse forest. We found no clear evidence that nests were located near edges. Northern Pygmy-Owls were diurnally active, and male owls delivered food to females during nest establishment and incubation, and to both females and nestlings during brooding. Date of fledging varied from midj^une to mid-July. Fledging was synchronous, and minimum estimates of brood size ranged from 1-5. Diet included a mix of small birds, mammals and insects. Our results suggest that the creation of openings by clear-cut logging is unlikely to benefit Northern Pygmy-Owls, and that the replacement of structurally diverse forests with uniform forests may be detrimental. Key Words: Northern Pygmy-Owl, Glaucidium gnoma; fragmentation-, habitat use. ESTACION REPRODUCTIVA, USO DE HABITAT Y ECOLOGIA DEL BUHO PIGMEO MACHO DEL NORTE Resumen. — Examinamos el uso de habitat, la localizacion de nidos, la dieta y los patrones de la actividad de buhos pigmeos {Glaucidium gnoma), durante cuatro estaciones reproductoras (1994—97) en bosques frag- mentados en la peninsula Olympic, en Washington. Observamos el forrajeo en todas las categorias de ve- getacion disponibles, cuyos patrones de uso no estaban determinados por el azar. Los bosques mas viejos y estructuralmente diversos fueron los mas usados, los claros y parches de arboles fueron los mucho menos usados, la utilizacion de hordes fue intermedio. Localizamos ocho nidos, todos en cavidades de carpinteros en parches de bosques estructuralmente diversos. No encontramos ninguna evidencia clara que los nidos estuvieran situados cerca de los hordes. Los buhos pigmeos del norte fueron mas activos durante el dia, los buhos machos entregaron el alimento a las hembras durante el establecimiento de nidos y la incubacion, tanto a hembras y pichones durante el empollamiento. La fecha de crecimiento de plumas en que los polluelos debian abandonar el nido vario a partir de mediados de junio hasta mediados de julio. El creci- miento de plumas fue sincronico y las estimaciones minimas del tamano de la nidada fluctuaron entre 1 y 5. La dieta incluyo una mezcla de pharos, mamiferos e insectos pequehos. Nuestros resultados sugieren que es poco probable que la creacion de claros por la tala beneficie al buho pigmeo del norte y que el reemplazo de bosques estructuralmente diversos por bosques uniformes puede ser peijudicial. [Traduccion de Cesar Marquez] Holt et al. (1990) labeled the Northern Pygmy- Owl {Glaucidium gnoma) North America’s least- studied owl. Thirteen years later, detailed infor- mation for this locally-common inhabitant of ^ Present address: Department of Biology, Arizona State University, Tempe, AZ 85287-4601 U.S.A.; e-mail address: alan.giese@asu.edu western forests remains scant, and published infor- mation is often equivocal. Yet western forests have experienced decades of intense resource extrac- tion pressure (Parry et al. 1983) and it is presently impossible to estimate how such changes might af- fect Northern Pygmy-Owls. Here, we report on Northern Pygmy-Owl habitat use and ecology in forests heavily fragmented by clear-cut logging. 117 118 Giese and Forsman VoL. 37, No. 2 The range of the Northern Pygmy-Owl extends from Alaska to Central America, including forests from the west coast to the Rocky Mountains (Johnsgard 1988). They are diurnally active, use cavity nests, and are generalist predators of small birds, mammals, and insects (Bent 1938, Johnsgard 1988). Northern Pygmy-Owls have been commonly sighted in or near openings, leading to statements that they preferentially foraged in openings (Bent 1938, AOU 1983, Johnsgard 1988), nested near edges (Webb 1982, Reynolds et al. 1989), and might benefit from partial forest clearing (Johns- gard 1988). In contrast, others have suggested that partial forest clearing may be detrimental to North- ern Pygmy-Owls (Marshall 1992). Hayward and Garton (1988) used call-response surveys to examine resource partitioning by small forest owls in Idaho and concluded that the North- ern Pygmy-Owl was a habitat generalist. Although their work is the most extensive study to date, many questions remain about habitat use by this species. We used radiotelemetry to study habitat use for two behaviors, foraging and nesting. We were specifi- cally interested in whether partial forest clearing might benefit this species. We also describe home range size, diet, activity patterns, nest characteris- tics, and nesting behavior. For the purposes of this paper, we hereafter use ‘Northern Pygmy-OwF and ‘owl’ synonymously. Methods The study area was a mixture of federal, state, and pri- vate lands on the northwest corner of the Olympic Pen- insula, Washington. The area was hilly to mountainous with elevations ranging from 50-1350 m. Mean annual precipitation ranged from 150-250 cm. Natural vegeta- tion was dominated by coniferous forests of western hem- lock {Tsuga heterophylla) , Douglas-fir {Pseudotsuga menzie- su) , western red-cedar ( Thuja plicata) , and silver fir {Abies amabilis). Valley bottoms typically included variable amounts of bigleaf maple {Acer macrophyllum) and red al- der {Alnus rubra) (Henderson et al. 1989). Even-aged patches of primarily Douglas-fir resulted from logging and silviculture. We located owls by walking logging roads while vocally imitating their calls. Surveys were conducted during April and May 1994-97, and were not designed to sample the study area equitably. Rather, we concentrated on areas where we had observed owls previously, and surveyed oth- er areas less intensively. When found, owls were captured in mist nets, marked with bands issued by U.S. Geological Survey Bird Banding Laboratory; and fitted with back- pack transmitters (Model BD-2G ca. 2.5 g with harness, Holohil Systems Ltd., Carp, Ontario, Canada). Radio-marked owls were relocated several times each week during the 15-wk transmitter life span. We worked from a sequential list of radio-marked owls, locating as many owls each day as possible (typically two to eight) and beginning where we had finished the day before Thus, each owl was located at all times of the day, because a new owl typically topped the list each day. Location times ranged from about an hour before sunrise to about an hour after sunset. We classified habitat into five vegetation categories based on structural characteristics. The Open-Sapling c-aX- egory included areas dominated by small conifer saplings (1-4 m tall) intermixed with extensive areas (>25% cov- er) of bare ground and shrub cover, mostly on recent dear-cuts. The Early Stem Exclusion category was dominat- ed by young conifers (5-15 m tall) with few openings in the overstory and with dense, overlapping limbs in the understory. The Late Stem Exclusion category was domi- nated by medium-sized conifers (typically 20-40 cm DBH and >15 m tall) with a closed canopy and open under- story The Structurally Diverse category was characterized by a wide variety of tree sizes and a multilayered canopy. It consisted primarily of mature and older forests (typi- cally >80 yr), but also included mid-aged stands charac- terized by high structural diversity. Finally the Edge cat- egory included all areas within 30 m of an intersection between a patch of Open-Sapling, and any of the other categories. Thirty meters was used because pilot-study ob- servations suggested that foraging flights by Northern Pygmy-Owls rarely exceeded this distance. A foraging flight was defined as any flight which included or ended with an attempt to capture prey. Because patchiness in the study area was predominantly the result of clear-cut logging, patches were generally easy to delineate and classify. We delineated patch boundaries using ARC/ INFO (ESRI, Redlands, CA) and digital orthophoto- graphs. We visited each patch and based vegetation clas- sifications for both used and available cover on visual inspection. We used compositional analysis (Aebischer et al. 1993) and program RESELECT (Leban 1994; available at http, //ces.iisc.ernet.in/hpg/envis/resdocl 120.html) to com- pare relative use among vegetation categories. Composi- tional analysis treats the individual as the sampling unit, accounts for the unit-sum constraint of proportions, and allows unique availability for each individual. We used an f^test to determine study-wide deviation from random us- age, and pairwise Hests for differential use between veg- etation categories. Use ratios were calculated by dividing proportionate use by proportionate availability. Proportionate use was defined as the proportion of locations for a given owl in a given vegetation category. All locations were deter- mined by homing to an owl with a hand-held receiver until the owl was located visually, or until triangulation indicated that the owl was directly overhead. Locations were mapped in the field on aerial photographs and sub- sequently digitized using digital orthophotographs. Lo- cations of owls <50 m from an active nest were not in- cluded in the analyses because of the possibility that those locations represented nest activity rather than for- aging. Removal of such locations would introduce bias if they were foraging locations. However, because all nests were located in the most heavily-used vegetation category (see below), this bias would be conservative relative to our conclusions. We assessed the error associated with June 2003 Pygmy-Owl Habitat Use 119 overhead triangulations by mapping the location of 12 transmitters placed in trees by an independent observer. Proportionate availability was defined as the propor- tion of a given vegetation category within the minimum convex polygon (MCP) that encompassed the locations for a given owl. We used the MCP for availability because we believe it best approximated the area in which a breeding owl had the opportunity to forage. A circle cen- tered on the nest was unsatisfactory because our pilot study revealed that nests were not always centrally located within territories, and territorial overlap was minimal. Thus, vegetation within such a circle might receive little use due to the presence of a neighboring owl. Alterna- tively, kernel estimators (Seaman et al. 1998) are more biased toward heavily-used vegetation patches than are MCPs, and may exclude areas that are available but used infrequently. We evaluated the sufficiency of our avail- ability estimates with a post hoc analysis of MCP size in relation to the number of relocations. An inflection point was apparent at ca. 30 locations so we removed owls from further analysis if they either died or left the study area before we had accumulated 30 locations. For comparison, we estimated home range sizes using both 100% MCP and 95% fixed kernel methods. In gen- eral, owls in this study occupied well-defined home rang- es during the breeding season. However, one owl made two excursions of 6 and 1 1 km from its core use area for three days each in April, and then returned and nested near its original trapping location. We removed the six locations collected during these excursions from MCP home range estimation, but included them in the kernel estimates. Reynolds and Linkart (1990) discussed extra- range movements in Flammulated Owls ( Otus flammeolus) and Linkart et al. (1998) removed extra-range move- ments from their home-range determinations. Programs CALHOME (Kie et al. 1996) and KERNELHR 4.28 (Sea- man et al. 1998) were used for MCP and kernel analyses, respectively. Nests were located by observing males delivering food to incubating females. The distance from each nest to the nearest edge (^f„) was measured in the field with a 50-m tape and compared to the mean distance-to-nearest- edge (dp for 100 randomly generated points in the same stand. Random points and associated distances were gen- erated with ARC/INFO. A studentized Z-statistic was cal- culated for each nest (Eq. 1) Z = (d„- dp/s^ (1) where is the standard deviation of the random point distances. We used a one-tailed f-test to test for Z < 0. We climbed to each nest post-fledging and measured cav- ity entrance and tree dimensions. Tree heights >20 m were estimated with a clinometer. We collected pellets and prey remnants from the ground near nests two to three times per wk, and recorded all observations of owls with prey. We pooled pellets and remnants for each col- lection date and nest, and estimated minimum vertebrate prey counts for each pooled sample. Each pellet was treated as independent for counts of insect prey. We re- corded owl behaviors during three dawn-to-dusk nest watches and 34, 2-hr focal animal observations. Results We radio-marked 21 owls during four field sea- sons (1994-97), including 16 males, one female, and four sex-unknowns. We had sufficient data to estimate ranges and conduct habitat use analyses for nine males (Table 1). Of those, six nested and fledged young, one nested and failed to fledge young, one nested and was thought to have failed, and the nesting status of one was undetermined. Radio-marked owls excluded from the analysis of habitat use included four males with fewer than 30 locations (Table 1), four that left the study area, two that died, one whose sex was unknown, and one female. Although we surveyed in consecutive years, we never trapped any previously banded owls. We collected a mean of 49 locations per owl (range = 34^66, = 9) . Estimates of home range size (mean ±SE) were 296 ± 42 ha (N = 9) for the MCP method and 209 ± 28 ha (N = 9) for the fixed kernel method (Table 1). Of all locations, 49% were confirmed visually and 51% were esti- mated by triangulation. Estimated triangulation er- ror (mean ±SE) for transmitters placed in trees was 11.6 ± 2.3 m (N = 12). Use of vegetation categories for foraging was nonrandom (T 45 = 29.41, P< 0.01). The Structur- ally Diverse category was the most used, followed by Late Stem Exclusion, Edge, Early Stem Exclusion, and Open-Sapling (Table 2) . Confidence in the rank as- signments for the Open-Sapling and Structurally Di- verse categories was high, as indicated by low P-val- ues for pairwise comparisons of rank with other vegetation categories. Confidence in the relative ranks of Late Stem Exclusion, Edge, and Early Stem Exclusion was low, as indicated by mostly non-sig- nificant P-values for pairwise comparisons. Our re- sults suggested a dichotomy between ‘forested’ and ‘non-forested’ vegetation categories. We reana- lyzed the data with the Open-Sapling category re- moved and found that the relationships between the remaining categories were qualitatively un- changed. We also analyzed the data with and with- out the one owl whose nesting status was not con- firmed and the results were again qualitatively unchanged. We located eight nests, all of which were in dead trees in cavities excavated by woodpeckers (Table 1). Estimated distance to the nearest edge (x ± SE) was 59 ± 16 m for the nests and 99 ± 3 m for the random points. The studentized difference be- Table 1. Nest characteristics and home range sizes for adult male Northern Pygmy-Owls on the Olympic Peninsula, WA, 1994-1997. 120 Giese and Forsman VoL. 37, No. 2 o o Q 3 w a H C/5 u 2 : Q W U u Q O Q u J P . I o H a a W ' 1 1 Q — tn u I — [ U w 0. c/:i h O K w H H- 1 CT) I> rH tH IT) OO o 05 CM O CD O lO 1—1 no 00 00 lO CD CD CD no 1 1 1 1 CM O no CD 00 CD 00 CM CO CO r-- 00 O o 1 1 1 1 00 1-H 1—1 GM I— H CM oo CM 1 1 1 1 «o o 1—1 05 CD O CD 05 1 1 1 1 ^ no 00 00 CD CD CD 1 1 ^ CO 00 CM nn 00 00 CD 1 1 00 CD o I— 1 o CD tT CO CM CO 1 1 CM 1-H CM 00 CM no CM CM CM 3 GO S OJ C d OJ ^ OJ 3 3 :3 lO (>f 00 O -4 O CM CM CM ^ CM OO 00 CM m I cO '“1 yc I o d o d D CM 00 oo irj d 'Tp u I - -3 d 00 00 u JS t3 'a d V 2 00 ho be V 00 00 o o CM >— i rH 00 o yo CM eo CO CM ID ^ ^ O tD 0> CO T}H W W W W W HH hH HH W HH HH hH HH HH HH iri iTi in m ^ H H H H H ^ CM 00 00 W p ^2^ Id ffi in H m m D D D D u u u 0 0 0 0 0 0 2; z 2 CS o o o 2i ■d D J-* OJ u O 00 m cn 00 no O 2 : no «D CD no CD CO 00 00 00 ^;3^nOno^tOCDcDt;'J>J> oj^oicoiCTJiDico^cpa) pqMpqU^tiSP^U^UeQ l-Jl— ^ hJi— l^C/CndPQ^^PQhJ i> CO 1 o J> 9 Q n I g s s § •d d t I d c D O bJD be V D o a § I ^ § a G 4-J -J ^ P S *5 S .S ^ s a ffi ^ ^ H Q S June 2003 Pygmy-Owl Habitat Use 121 Table 2. Use of vegetation categories by adult male Northern Pygmy-Owls on the Olympic Peninsula, Washington, 1996-97, expressed as logged use-ratio differences (SE). A positive value indicates that relative use for the row category exceeded that of column category. Categories are ranked from most (4) to least (0) used by adding the number of positive use-ratio differences across rows. Vegetation Category Open- Sapling Early Stem Exclusion Edge Late Stem Exclusion Structurally Diverse Rank Open-Sapling -2.02** -2.68** — 2.74** —3.44** 0 (0.51) (0.40) (0.47) (0.33) Early Stem 2.02=^* -0.66 -0.72 -1.42* 1 Exclusion (0.51) (0.48) (0.51) (0.49) Edge 2.68** 0.66 -0.06 -0.76** 2 (0.40) (0.48) (0.35) (0.17) Late Stem 2.74** 0.72 0.06 -0.70 3 Exclusion (0.47) (0.51) (0.35) (0.31) Structurally 3.44** 1.42* 0.76** 0.70 4 Diverse (0.33) (0.49) (0.17) (0.31) '* P < 0.05 from two-tailed t-tests for pairwise differences in log-ratios. ** p < 0.005 from two-tailed t-tests for pairwise differences in log-ratios. tween nests and random points {x ± SE) was Z = —0.51 ± 0.31, and did not provide sufficient evi- dence to show that nests were associated with edges (One-tailed 1>j = 1.64, P = 0.07). All nests were located in Structurally Diverse forest patches, and seven nests were in patches of late successional, (>200 yr old) coniferous forest. The eighth nest was in a relatively young, mixed patch of conifer- ous and deciduous trees that had regenerated nat- urally following logging. Northern Pygmy-Owls consumed a variety of small birds, mammals and insects (Tables 3 and 4) , and males provisioned females and nestlings dur- ing incubation and brooding. In ca. 100 hr of nest observation we saw no indication of females for- aging. We observed females accepting prey items from males and retrieving cached prey items, but not leaving or returning with fresh prey items of their own. During dawn-to-dusk observations at three nests, females were either in the nest cavity or perched within 50 m, and radiotelemetry sug- gested that males visited the nest stand every 1-3 hr. Additionally, during the egg-laying period, fe- males typically perched near the nest while males foraged and delivered food. We documented date of fledging for nine nests (Table 1). At four nests we observed chicks exiting the nest, and in each case, all known chicks from a given nest exited within a 6 hr period. Minimum estimates of brood size varied from one to five based on the maximum number of fledglings observed simultaneously (Ta- ble 1). Discussion We found strong evidence that patterns of use differed from patterns of availability, indicating that owls discriminated between the vegetation cat- egories we defined. Although use was concentrated in structurally-complex forests, we observed forag- ing flights in all vegetation categories. Therefore, locations in seldom-used categories cannot be at- tributed to owls in transit. Use of edges (as defined here) was proportionate to availability, the least used vegetation category consisted primarily of re- cent dear-cuts, and nests did not appear to be as- sociated with edges. Thus, our results suggest that Table 3. Percent composition of the diet of adult male Northern Pygmy-Owls on the Olympic Peninsula, Washing- ton, 1996-97, based on three different methods of data collection. Method N Mammals Birds Insects Total Direct observation 59 45.8 50.8 3.4 100.0 Pellet analysis 83 59.0 18.1 22.9 100.0 Prey remnants 8 12.5 87.5 0.0 100.0 122 Giese and Forsman VoL. 37, No. 2 Table 4. Species of mammals and birds identified as prey of adult male Northern Pygmy-Owls on the Olympic Peninsula, Washington, 1996-97, based on direct observation, pellet analysis and prey remnants. Mammals Birds Shrew (Sorex sp.) Coast mole {Scapanus orarius) Deer mouse {Peromyscus maniculatus) Red-backed vole ( Clethrionomys gapperi) Vole {Microtus sp.) Townsend’s chipmunk {Tamias townsendii) Gray Jay {Perisoreus canadensis) Chestnut-backed Chickadee {Poecile rufescens) Brown Creeper {Certhia americana) Winter Wren ( Troglodytes troglodytes) Golden-crowned Kinglet {Regains satrapa) Thrush {Catharus s\i.) Varied Thrush {Ixoreus naevius) Dark-eyed Junco (Junco hyemalis) White-crowned Sparrow {Zonotrichia leucophrys) Golden-crowned Sparrow {Zonotrichia atricapilla) the creation of openings by clear-cut logging is un- likely to benefit Northern Pygmy-Owls and that the replacement of structurally diverse forests with uni- form forests may be detrimental. This conclusion is subject to the caveat that habitat use may vary temporally, and we cannot exclude the possibility that different types of habitat receive heavier use in different seasons or years. Owls in this study did not use edge habitat or openings heavily, and detection bias is one plausi- ble explanation for this inconsistency with previous anecdotal accounts. Prior to initiating this radio- telemetry study, we observed Northern Pygmy-Owls most frequently in recent dear-cuts and along edg- es. Moreover, our visual confirmation rate was ap- proximately 50% for owls in forests, compared to 100% for owls in openings. Alternatively, edges and openings created by clear-cut logging may differ in important ways from edges and openings in other contexts. For example, transitional vegetation was all but absent in our study area. Also, our defini- tion of a 30-m buffer to define edges was arbitrary, and other definitions of edge might lead to differ- ent conclusions. Lastly, we assumed that the owls we tracked were foraging. While we attempted to strengthen this assumption by focusing on nesting males and eliminating locations near nests, we can- not demonstrate that our data reflect use for for- aging. If owls used different vegetation categories for different activities, important use of some cat- egories might be obscured by our analyses. These alternative explanations should be tested before general conclusions regarding edge associations of Northern Pygmy-Owls are drawn. We also failed to find support for the idea that Northern Pygmy-Owls use nests near edges. How- ever, our sample was small and the results were nearly significant {N = 8, P — 0.07). Furthermore, our approach assumed that potential nests were evenly distributed in forest patches. While it would be useful to know the true distribution of potential nests, estimating such a distribution would be problematic, and might only be relevant if poten- tial nests were in limited supply. Our analyses were also sensitive to scale. The mean distance from a nest to the nearest edge (59 m) may be a biologi- cally meaningful proximity that was not statistically significant in our study due to an abundance of edges. A mean distance of 99 m from random points to the nearest edge gives an indication of the ubiquitousness of edges in our study area. Based on an analysis of forest characteristics at locations where owls responded to a vocal lure, Hayward and Garton (1988) concluded that the Northern Pygmy-Owl was a habitat generalist. At least two plausible and non-mutually exclusive hy- potheses can explain the differences between their conclusions and ours. First, owls might behave dif- ferently on different study areas and second, owls might use different criteria to select habitat for home-range location, calling, and foraging. The owls we studied occupied home ranges that encom- passed a diverse array of vegetation categories, but they predominantly used a subset of those catego- ries. Additionally, calling locations and foraging lo- cations appeared to differ (see below). Two im- portant differences between our study and that of Hayward and Garton are that they apparently sur- veyed for owls at night, whereas our locations were crepuscular and diurnal, and their study included many vegetation types across a broad range of el- evations. June 2003 Pygmy-Owl Habitat Use 123 Our results suggest the need for care when eval- uating habitat associations of Northern Pygmy- Owls from opportunistic sightings or vocal lure sur- veys. First, of 21 owls radio-marked, 11 were initially detected responding to playbacks from for- est edges (unpubl. data) , yet our analyses on a sub- set of nine of these owls did not indicate dispro- portionately heavy use of edge habitat. This can most easily be explained if owls reacted to calling surveys by moving toward the perceived source be- fore vocalizing. Proudfoot et al. (2002) document- ed the movement of Ferruginous Pygmy-Owls (Glaucidium hrasilianum) toward calling stations. In our case, because surveys were conducted from logging roads, this would tend to pull owls toward edges of dear-cuts. However, we cannot exclude the possibility that edges are selected as sentinel perch areas for calling. Second, of the same 21 ra- dio-marked owls, only 12 established territories in- clusive of their original response location (unpubl. data). Lastly, we occasionally detected responses from multiple owls on the same territory at differ- ent times during a season, again suggesting that calling location and territory location may be de- coupled. From our experience, the surest way to document Northern Pygmy-Owls in residence would be to repeatedly detect unsolicited vocali- zations from the same area. None of the owls that we radio-marked were re- located in subsequent years. We did monitor nest- ing owls in consecutive years at the same site in five cases (LB-94/95, 95/96, 96/97, WC-96/97, and BC-96/97; Table 1). Additionally, there were three sites used in one season, but vacant in the follow- ing season (UB-94, LC-95, and SK-96), for a total of eight sites where a resident male was not found at the same site in the following season. Possible explanations include: (1) nest-site fidelity was low, (2) mortality was high, (3) the study area was a population sink, and (4) radio-marking negatively affected the owls we worked with by either increas- ing their mortality or inducing them to bnd new territories. Information on the annual movements of Northern Pygmy-Owls would be useful in eval- uating these hypotheses. Seven of the eight nests we located were in late successional forests. This finding, coupled with high use in the same types of forest suggests that the loss of late-successional forest may negatively affect Northern Pygmy-Owls. However, one nest was in a mid-aged stand that differed from most of the mid-aged stands in our study area by having greater structural diversity. This suggests that log- ging practices that do not result in monocultural plantations may have a lesser impact. Our findings are similar to those of a recent study of habitat use by the Eurasian Pygmy-Owl {Glaucidium passerinum) (Strom and Sonerud 2001). Home range sizes for males in their study (100% MCP, 40—600 ha) overlapped our estimates and habitat use patterns were similar, except that they found that Edge was the highest ranked cate- gory for the Eurasian Pygmy-Owl. However, they defined Edge to be a 10-m-wide strip, where forest vegetation bordered open areas. To facilitate com- parisons, we reanalyzed our data using their defi- nition of Edge. It is important to note that although Edge had the highest relative rank in their study, it did not differ significantly from either of the next two highest ranked cover types (analogous to our Structurally Diverse and Late Stem Exclusion vegeta- tion categories). In our reanalysis, the Structurally Diverse category again had the highest relative rank, followed by Edge, Late Stem Exclusion, Early Stem Exclusion, and Open-Sapling (results not shown) . However, the relative rank of Edge was not significantly different than any category type ex- cept Open-Sapling. Thus, even though the relative rank for Edge increased by one, there was still in- sufficient evidence to conclude that Northern Pyg- my-Owls used edges disproportionately. Our natural history observations support much of the consensus knowledge regarding Northern Pygmy-Owls. We observed synchronous fledging in four instances and the dates of fledging were more similar within years than between (Table 1). The owls that we observed also exhibited a diverse diet, consistent with previous reports (Earhart and John- son 1970, Snyder and Wiley 1976, Holt and Leroux 1996). However, we have no information on rela- tive prey availability, and owls may exhibit prey preferences that would not be apparent in simple tallies. Acknowitdgments These data were collected with the invaluable assis- tance of Elissa Arnheim, Dave Manson, and the Sequim Audubon Club. We are grateful to J. Duncan, G. Hay- ward, T. Katzner, G. Proudfoot, R. Reynolds, and two anonymous reviewers for insightful comments that great- ly improved the final manuscript. This research was sup- ported by a grant from the Washington State Department of Natural Resources (FY97-021) and generous in-kmd support from the U.S. Forest Service Pacific Northwest Research Station. 124 Giese and Forsman VoL. 37, No. 2 Literature Cited Aebischer, N.J., P.A. Roberts, and R.E. Kenward. 1993. Compositional analysis of habitat use from animal ra- dio-tracking data. Ecology 74:1313-1325. American Ornithologists’ Union (AOU). 1983. Check- list of North American birds, 6th Ed. American Or- nithologists’ Union, Washington, DC U.S.A. Bent, A.C. 1938. Life histories of North American birds of prey. Part 2. U.S. Natl. Mus. Spec. Bull. 170:400-439. Earhart, C.M. and N.K. Johnson. 1970. Size dimor- phism and food habits of North American owls. Con- dor 72:251-264. Hayward, G.D. and E.O. Carton. 1988. Resource parti- tioning among forest owls in the River of No Return Wilderness, Idaho. Oecologia 75:253-265. Henderson, J.A., D.H. Peter, R.D. Lesher, and D.C. Shaw. 1989. Forested plant associations of the Olym- pic National Forest. USDA Forest Service Ecol. Tech. Paper R6-ECOL-TP 001-88. Holt, D.W. and L.A. Leroux. 1996. Diets of Northern Pygmy-Owls and Northern Saw-whet Owls in west-cen- tral Montana. Wilson Bull. 108:123-128. , R. Kline, and L. Sui.livan-Holt. 1990. A descrip- tion of “tufts” and concealing posture in Northern Pygmy-Owls. J. Raptor Res. 24:59-63. Johnsgard, P.A. 1988. North American owls, 1st Ed. Smithsonian Institution, Washington, DC U.S.A. Kie, J.G., J.A. Baldwin, and C.J. Evans. 1996. CALHOME: a program for estimating animal home ranges. Wildl. Soc. Bull. 24:342-344. Leban, F. 1994. Program RESELECT. University of Idaho, Moscow, ID U.S.A. Linkhart, B.D., R.T. Reynolds, and R.A. Ryder. 1998. Home range and habitat of breeding Flammulated Owls in Colorado. Wilson Bull. 110:342-351. Marshait, D.B. (Ed.). 1992. Sensitive vertebrates of Oregon. Oregon Department of Fish and Wildlife, Portland, OR U.S.A. Parry, B.T, HJ. Vaux, and N. Dennis. 1983. Changing conceptions of sustained-yield policy on the national forests./. Forestry 81:150—154. Proudfoot, G.A., S.L. Beasom, F. Chavez-Ramirez, and J.L. Mays. 2002. Response distance of Ferruginous Pygmy-Owls to broadcasted conspecific calls. /. Raptcnr Res. 36:170-175. Reynolds, R.T. and B.D. Linkhart. 1990. Extra-pair cop- ulation and extra-range movements in Flammulated Owls. Ornis Scand. 21:74-76. , R.A. Ryder, and B.D. Linkhart. 1989. Small for- est owls. Natl. Wildl. Fed. Sci. Tech. Ser. 12:134-143. Seaman, D.E., B. Griffith, and R.A. Powell. 1998. KER- NELHR: a program for estimating animal home rang- es. Wildl. Soc. Bull. 26:95—100. Snyder, N.F.R. and J.W. Wiley. 1976. Sexual size dimor- phism in hawks and owls of North America. Ornithol. Monogr. 20:1-96. Strom, H. and G.A. Sonerud. 2001. Home range and habitat selection in the Pygmy Owl Glaucidium passe- rinum. Ornis Fenn. 78:145—148. Webb, B. 1982. Distribution and nesting requirements of montane forest owls in Colorado. Colo. Field Ornithol J. 16:76-81. Received 12 April 2002; accepted 26 December 2002 J. Raptor Res. 37(2):125-134 © 2003 The Raptor Research Foundation, Inc. HABITAT USE BY SWAINSON’S HAWKS ON THEIR AUSTRAL WINTERING GROUNDS IN ARGENTINA Sonia B. Canavelli^ INTA, EEA Parana, Ruta 11 km 12, 3100 Parana, Entre Rios, Argentina MarcJ. Bechard Department of Biology, Boise State University, Boise, ID 83725 U.S.A. Brian Woodbridge USD A Forest Service, Klamath National Forest, 1312 Fairlane Rd., Yreka, CA 96097 U.S.A. Michael N. Kochert USGS, Forest and Rangeland Ecosystem Science Center, Snake River Field Station, 970 Fusk St., Boise, ID 83706 USA JuanJ. Maceda Departamento de Ciencias Naturales, Facultad de Ciencias Exactas y Naturales, Universidad de Fa Pampa, Av. Uruguay 151, 6300 Santa Rosa, Fa Pampa, Argentina Maria E. Zaccagnini INTA, EEA Parana, Ruta 11 km 12, 3100 Parana, Entre Rios, Argentina Abstract. — ^We examined the use of agricultural habitats by Swainson’s Hawks’ {Buteo swainsoni) in La Pampa and Santa Fe provinces, Argentina. We found an association of foraging Swainsons’ Hawks with permanent pastures such as fallow, natural, and alfalfa fields. The hawks also used plowed fields for sunning, resting, and preening. Fields planted with annual crops and pastures were used very little, except when they were cut for hay, plowed, and harvested, or when low crop height and cover allowed the hawks to land in fields. The availability of abundant, yet widely-spaced and transient food-sources, such as insect outbreaks, appeared to be the principal factor influencing habitat use by the hawks. Their reliance on agricultural habitats makes Swainson’s Hawks highly vulnerable to pesticide contamination and has contributed to the occurrence of significant mortality events on their wintering grounds. Keywords: Swainson’s Hawk, Buteo swainsoni; agroecosystems', Argentina', habitat use, wintering grounds. uso de hAbitat del aguilucho langostero en sus areas de invernada austral EN ARGENTINA Resumen. — Estudiamos el uso de ambientes agrfcolas por aguiluchos langosteros {Buteo swainsoni) en las provincias de La Pampa y Santa Fe, Argentina. Encontramos una fuerte asociacion de los aguiluchos langosteros con pasturas perennes tal como campos enmalezados, naturales y alfalfa, que utilizaron para alimentarse en el suelo. Los aguiluchos tambien utilizaron campos arados para exponerse al sol, des- cansar y acicalarse. Los lotes de cultivos y pasturas anuales fueron usados muy poco, excepto cuando fueron cortados y enfardados, arados o cosechados, o cuando la baja altura y cobertura de las plantas les permitieron a los aguiluchos posarse en el suelo. La disponibilidad de fuentes de alimento abun- dantes, aunque ampliamente espaciadas y pasajeras, como las mangas de insectos, serfa el factor prin- cipal que explicarfa los patrones observados de uso de habitat en estas escalas. La asociacion de los aguiluchos con ambientes agricolas hace estas aves altamente vulnerables al uso de plaguicidas, y habria contribuido a la ocurrencia de episodios de mortandad masiva en sus areas de invernada. [Traduccion de los autores] ^ E-mail address: scana@parana.inta.gov.ar 125 126 Canavelli et al. VoL. 37, No. 2 Migratory birds use different habitats according to their life history requirements and seasonal en- vironmental changes (Newton 1979, Alerstam 1990, Moore et al. 1995, Morrison et al. 1998). Their mobility and exposure to a variety of envi- ronments on breeding and wintering grounds make it especially challenging to understand their habitat use patterns (Cody 1985, Rappole 1995). In addition, their reliance on multiple habitats at different times of the year makes them particularly vulnerable to impacts from human activities (Rap- pole 1995), particularly agricultural activities, given the reliance that many of these birds have on crop fields and edge areas (Rodenhouse et al. 1995). Among migratory birds, the trophic level and slow reproductive rates of raptors make them es- pecially vulnerable to human-induced, environ- mental changes such as habitat fragmentation and pesticide contamination (Newton 1979, Alerstam 1990, Telia et al. 1998). Although only a small number of raptors are long-distance migrants (Alerstam 1990), most species of raptor performs some kind of migratory movement in at least part of its range. In all cases, these movements are gen- erally associated with variations in prey availability (Newton 1979). The Swainson’s Hawk (Buteo swainsoni), a Neo- tropical migratory raptor that breeds in North America and migrates to South America for the austral summer, is typically associated with open fields such as grasslands. It has adapted to environ- ments with a high proportion of agriculture, both in its breeding and wintering ranges (Mouchard 1996, England et al. 1997). As examples, 17.7% and 75.4% of the area within 1 km of Swainson’s Hawks nests in North Dakota were cultivated crops and pasture/hayland, respectively (Gilmer and Steward 1984), and 42% of the area within forag- ing radius of radio-marked individuals in Califor- nia was active agriculture (Woodbridge 1991). This makes the species vulnerable to impacts from ag- ricultural practices, such as habitat modification and pesticide applications. Understanding the ecology of the Swainson’s Hawk is crucial to reduce these impacts over its entire range. Although the habitat use of the Swainson’s Hawk is well documented on its breeding grounds (Be- chard 1982, Schmutz 1984, 1989, Estep 1989, Woodbridge 1991, Babcock 1995, Smallwood 1995), there is little information on its habitat use on the wintering grounds (England et al. 1997, Herkert and Knopf 1998, Kirk and Hyslop 1998). Figure 1. Approximate wintering range of Swainson’s Hawks in Argentina (from CIPA 1987 and England et al. 1997) and location of the study areas. As part of a more comprehensive project started in the austral summer of 1995-96, this study fo- cused on the analysis of habitat use by wintering Swainson’s Hawks in two areas in La Pampa and Santa Fe provinces of Argentina. Our goal was to obtain information on the ecology of wintering Swainson’s Hawks in Argentina that could help to reduce the potential negative impacts of agricul- ture. Study Areas Fieldwork was conducted in the northeastern portion of La Pampa and central portion of Santa Fe provinces, Argentina (Fig. 1). Between 28 November 1996 and 16 March 1997, a 2250 km^ (50 X 45 km) area was estab- lished around a principal roost at “Chanilao” ranch (35.2°S, 64.0°W) in La Pampa. The area included several June 2003 Habitat Use by Wintering Swainson’s HA^^^cs 127 locations where mass mortalities of Swainson’s Hawks oc- curred during the previous austral summer (Woodbridge et al. 1995, Goldstein et al. 1996, Goldstein 1997). We chose this area because two study teams could cover it during one survey day, which minimized double counting of hawks for estimation of population abundances. In 1998, a 900 km^ sampling area (45 X 20 km) was selected m the central portion of Santa Fe province, on the bor- der with Cordoba province (Fig. 1). Fieldwork was car- ried out in this area between 7 January and 14 March 1998. Both study areas were dominated by agricultural fields. The La Pampa area contained a mosaic of crops includ- ing sunflowers, corn, sorghum, millet, and soybeans, as well as patches of natural habitats such as Poa spp. and Stipa spp. grasslands. The Santa Fe study area was domi- nated (74%) by continuous cropland (mostly soybean), with the remaining of the surface (26%) dedicated to livestock and milk production over non-native pastures. La Pampa area had lower rainfall and slightly lower mean annual temperature than Santa Fe area (600-700 mm, 16°C and 800-900 mm, 17°C, respectively; Canavelli 2000 ). Methods Individutil Patterns of Daily Activity and Habitat Use. Between 16 December 1996 and 13 March 1997, focal observations of individuals were recorded to describe pat- terns of daily activity and habitat use within La Pampa study area. The day was divided into three equal periods; morning (0600-1040 H), midday (1041-1520 H), and af- ternoon (1521-2000 H) based on the length of the day during December (sunshine equal to 14 hr 35 min be- tween 0604 and 2025 H, R. Rodriguez pers. comm.). We randomly assigned these periods to three different days each week. Locations for focal observation points were randomly selected among locations of flocks observed in the area on previous surveys (Canavelli 2000). When a flock of hawks was found, we began a scan with binoculars starting on a randomly selected horizontal an- gle (at 10° intervals) and vertical section (air/ground) in which the group was previously divided, until one indi- vidual was identified. Then, we began an observation bout of 5-min conducted on that individual using 10 X binoculars and a 20-60X spotting scope. A 5-min rest bout followed and then a new individual was randomly selected and observed. This procedure was repeated dur- ing the entire 280-min observation period for a total of 29 observation bouts/observation period. Observations recorded during each observation bout were: general weather conditions (cloud cover, wind speed and direc- tion, and temperature); cover type (wheat, corn, sunflow- er, alfalfa, other improved pasture, native-plant pastures, fallow field, plowed field, woodland, and other) ; behavior of the individual hawk (soaring, active flight, and perch- ing on the ground, fence post, light pole, and in a tree) ; and bird activity (preening, foraging, short runs, short flights, pecking on the ground, extension of feet while soaring, and movement of feet to the bill while soaring) . Individuals were considered to be foraging on the ground if they made short runs, walks, jumps or flights usually followed by pecks to the ground to catch prey (without necessarily using talons). While soaring, hawks were considered to be foraging when they dove, extend- ed one or both feet to catch prey and then transferred it to the bill (Woffinden 1986). The relative frequencies and mean frequencies of be- havioral observations were analyzed using SAS System for Windows (v6.12, 1998). In order to pool all the obser- vations for an estimation of hourly-activity budget per day, the number of observations for each behavior during each hour was scaled based on how much of the hour was completely covered on the observation periods. In this way, a balance was achieved among the unequal num- ber of observations obtained for each hour (as result of the differences in detectability of Swainson’s Hawks at different times of the day) . Then, the percentage of time devoted to each activity was estimated on an hourly basis Habitat Selection at Population Level. Our character- ization of habitat use at the population level (population is defined as the group of Swainson’s Hawks sharing the study area at the same time) was based on systematic sur- veys conducted every 3-4 d along roadways regularly spaced over the two study areas. Twenty-two surveys were conducted in La Pampa between 21 December 1996 and 16 March 1997, and 21 were conducted in Santa Fe be- tween 7 January and 14 March 1998. Eight surveys in La Pampa were conducted using strip transects (bandwidth of 200 and 300 m on each side, five and three surveys, respectively), while 14 surveys in this area and all the surveys in Santa Fe study area were conducted using point counts (variable circular plots; Buckland et al 1993). In the latter case, routes were 45 km long regu- larly spaced every 10 km, with ten 5-min point-count sta- tions spaced at 5 km intervals on each route, totaling 60 point-count stations in La Pampa (six routes) and 30 sta- tions in Santa Fe (three routes; Canavelli 2000). One or two groups of two people conducted the sur- veys in a vehicle driving at 40-60 km/hr during the morning (0600-1200 H) and in the afternoon (1400- 2000 H). In the case of two groups (La Pampa), each group was randomly assigned to cover three routes each day, in order to reduce observer bias on different areas. Additionally, both in La Pampa and in Santa Fe, the start- ing point and route for the survey were randomly estab- lished, but considering each route was equivalently cov- ered on different time periods (morning, midday, and afternoon) in order to standardize the influence of time of the day on the counts (Watson et al. 1996). All hawks detected during each survey were recorded. Data included percent cloud cover and wind speed m qualitative categories; time of day; location using the truck odometer; and distance to the hawks from the ob- server estimated in intervals of 0-10, 11-50, 51-100, and 101-500 m. We also recorded land cover type, including field crops (such a corn, sunflowers, wheat, or sorghum), annual pastures (such as millet), permanent pastures (such as alfalfa), native plant pastures, fallow fields, plowed fields, and woodlands. We recorded the number of hawks (which were counted if there were individual hawks or small groups or estimated a range of numbers for large flocks); behaviors included: soaring, active flight, perched on the ground, on fence posts, on electric light posts, or in trees; bird activities: feeding in the air; feeding on the ground, preening; and resting/ sunning Two behaviors could be recorded for an individual hawk 128 Canavelli et al. VoL. 37, No. 2 □ In the Air i Foraging in Air 6 7 8 9 io ii 12 13 14 15 16 17 is 19 ■ On Ground ■ Foraging on Ground Figure 2. Daily activity pattern of Swainson’s Hawks in La Pampa, Argentina (N = 416 observations). A. Activity budget (as percent of observations for each behavior in a day); B. Distribution of activities at different hours (as percent of observations made each hour) . (e.g., a hawk could be recorded preening and perched on a fence post). Only counts of hawks observed on the ground were included in the habitat-use analysis because it was clear that they were using a particular cover type. We later re- assigned these counts to five cover types: cropland (sun- flower, wheat, and sorghum), annual pastures (millet), permanent pastures (alfalfa and pastures >2 yr-old, nat- ural, fallow and shortgrass fields), plowed fields, and woodlands. Because the hawks usually moved in flocks (not independent individuals), habitat use was analyzed using the number of observations and not the number of hawks observed in each cover type (Thomas and Tay- lor 1990, Alldredge and Ratti 1992). Although we made sporadic observations of hawks using native woodlands and lowlands outside the study area in La Pampa and hawks soaring over urban areas, we made no observations of Swainson’s Hawks on the ground in these areas during surveys. For this reason, these cover types were not in- cluded in the analysis of habitat use. Habitat availability in both areas was obtained from sat- ellite image analysis using remote sensing methodology. The National Institute of Agricultural Technology (INTA) at Castelar (Buenos Aires, Argentina) provided satellite images for both study areas (Mosaic Landsat TM, Path/Rows 228-84 and 228-85 for La Pampa and Path/ Rows 228—82 and 227-82 for Santa Fe) . Images included bands three (red), four (near infrared), and five (mid- dle-infrared) and they were georeferenced to latitude and longitude. Unsupervised and supervised classifica- tions were carried out on each area using IDRISI for Win- dows (v.2.008, Clark University, Eastman 1997). Eighteen land use classes were initially considered in La Pampa and 16 in Santa Fe. Later, we grouped these into six new classes: permanent pastures (alfalfa fields, native-plant pastures, short-grass fields, and fallow fields), cropland (sunflower, sorghum, corn, oat, wheat, stubble wheat, plowed fields, and annual pastures), native woodlands. Eucalyptus spp. woodlands, lowlands, and urban areas. Habitat selection wzis estimated following Neu et al. (1974) and using the program HABUSE (Byers et al. 1984). In addition to Bonferroni’s confidence intervals estimated by this method, Bailey’s intervals were estimat- ed in order to complement those intervals obtained by the program HABUSE (Cherry 1996). Bailey’s intervals are more robust for small sample sizes and provide the best combination of low error rates and interval length on the estimation of confidence intervals (Cherry 1996). Results Individual Patterns of Daily Activi^ and Habitat Use. We recorded 416 activity and habitat use ob- servations over 22 d. In the morning (0600-1040 H), we obtained 204 observations on 10 different days, 139 observations were obtained at midday (1041-1520 H) on 8 d, and 73 were obtained in the afternoon (1520-2000 H) on 7 d. Observations were made in more than one time block in the same day on only 3 d (16 December 1996, 8 and 17 January 1997). Hawks spent most of the day (59%) sunning/ resting, preening, and foraging on the ground (Fig. 2A), mainly in the morning and late after- noon (Fig. 2B). The remainder 41% of the day was spent soaring, gliding, and foraging in the air, par- ticularly at midday. Most hawks (64-100%) stayed on the ground until 0900-1000 H, with tempera- tures 11-29°C (Fig. 2B). At 1100 H, hawks were observed in nearly equal proportions in the air (48%) and on the ground (52%). Between 1200- 1600 H (temperatures 20-38°C), hawks were ob- served mostly soaring (77-82%). When the highest temperatures were registered (midday, 35-38°C), only soaring hawks were observed. At 1700 H, hawks were again observed in nearly equal pro- portions in the air (57%) and on the ground (43%). Finally, between 1800-2000 H (tempera- June 2003 Habitat Use by Wintering Swainson’s Hawks 129 H Perm. Pasture M Cropland ■ Plowed Fields □ Woodland Habitat Type ■ Resting □ Foraging Figure 3. Cover types used by Swainson’s Hawks on the ground during focal observations (N = 279 observations) A, Percentage of observations related to cover type; B. Percentage of observations by behavior in each cover type. Permanent pastures in A include fallow fields, mowed alfalfa, natural fields, and standing alfalfa. tures 20-35°C) , hawks were again mainly observed on the ground (67-78%, Fig. 2B). Swainson’s Hawks used various cover types for different activities at different times during the day. They mainly used plowed fields (N = 100 obser- vations) and permanent pastures {N = 84) to sun, rest, and preen on the ground (Fig. 3A and 3B) during the day. Additionally, they used plowed fields {N — 29) and permanent pastures (fallow, natural, mowed, and standing alfalfa fields; N = 38) to forage on the ground (Fig. 3B). Plowed fields were used for foraging on the ground early in the morning (A^ — 21), while permanent pas- tures were used more at midday and afternoon (N — 20 and 10, respectively). Hawks were observed soaring and foraging in the air over several cover types (N = 92 and 45 observations, respectively), including permanent pastures {N = 36), annual crops (N = 31), plowed fields (N — 28) and annual pastures {N = 3), especially at midday {N = 85) and in the afternoon {N — 26). In woodlands {N = 66 observations), most (76%) of hawks observed were either resting or preening in trees, early in the morning or late in the afternoon (N = 44 and 6, respectively). Based on our observations, a typical day for a Swainson’s Hawk on its wintering grounds was di- vided into three periods: (1) morning, when hawks stayed mostly on the ground, in roosts, or in plowed fields located next to roosts either sunning, resting, preening, and foraging; (2) midday and afternoon, when some hawks still remained on the ground, mostly foraging on permanent pastures, but increasingly left the ground and soared while feeding in the air over crops and pastures; and (3) late afternoon prior to roosting, when hawks are again on the ground in plowed fields and pastures preening and foraging. Hawks were usually ob- served using crop and pasture fields without any disturbance, but on seven sampling occasions, hawks were observed on crop and pasture fields while farm operations such as plowing, mowing or baling, and burning were conducted (four, two, and one occasions, respectively), or cattle were grazing on the fields (one occasion). Habitat Use at the Population Level. Observa- tions from both study areas showed that Swainson’s Hawks primarily used permanent pastures which were not plowed on an annual basis (159 obser- vations, 6265 hawks in La Pampa; 240 observations, 6885 hawks in Santa Fe; Fig. 4). In La Pampa, plowed fields followed permanent pastures in im- portance of use (77 observations, 3793 hawks) while in Santa Fe, annual pastures were the second most used cover type (72 observations, 1511 hawks) . Cropland was more used in La Pampa than in Santa Fe, but in both study areas, observations were less than 15% of the total (36 observations, 3039 bawks and 23 observations, 141 hawks in La Pampa and Santa Fe, respectively). Less than 10% of our observations on both areas (15 observa- tions, 329 hawks in La Pampa and 36 observations. \ 1 130 Canavelli et al. VoL. 37, No. 2 70 60 50 § 30 Q. 20 10 0 Perm. Past. Ann. Past. Crop Plowed Woodland Habitat Type ■ Used □ Available La Pampa + - - + + 70 60 50 ^ 40 c 0) S 30 a. 20 10 0 Perm. Past. Ann. Past. Crop Plowed Woodland Habitat Type ■ Used □ Available Santa Fe Figure 4. Habitat use vs. availability in La Pampa and Santa Fe study areas, Argentina. Habitat use is expressed as percent of observations with hawks on the ground, N = 290 observations in La Pampa and N = 393 observations in Santa Fe), and habitat availability is expressed as percent of each habitat type determined from satellite images. ( + ) = used more than available, ( — ) = used less than available, and (0) = used as available (Table 1). 844 hawks in Santa Fe) were on hawks using wood- lands (Fig. 4). Among permanent pastures, fallow, native-plant, and alfalfa fields were the most commonly used in La Pampa (70, 41, and 40 observations; 1715, 1671, and 2640 hawks, respectively). In alfalfa fields, more observations were made on standing than on mowed alfalfa (23 and 17, respectively), although fewer hawks were observed on standing alfalfa compared to mowed fields (855 and 1785 hawks, respectively). In Santa Fe, 83% (200) of observa- tions made on permanent pastures were in fallow and short-grass fields (129 observations, 2538 hawks and 7l observations, 2456 hawks, respective- ly) . In this case, both number of observations and hawks were greater in mowed than in standing al- falfa (23 observations, 1287 hawks and 16 obser- vations, 603 hawks, respectively) . Sixty-one percent and 77% of observations made on hawks foraging on the ground in La Pampa (14 observations, 2525 hawks) and Santa Fe (23 observations, 2069 hawks), respectively, were on permanent pastures, followed in importance by plowed fields (La Pam- pa: 6 observations, 1524 hawks and Santa Fe: 4 ob- servations, 280 hawks). Fallow (26% of observa- tions with hawks foraging on the ground in La Pampa and 27% in Santa Fe), native-plant (9% in La Pampa), short-grass (43% in Santa Fe), and al- falfa fields (22% in La Pampa and 7% in Santa Fe) were the main permanent pastures used for for- aging on the ground. Groups of hawks foraging on the ground varied between 207 (SE = 65) and 86 (SE = 37) individuals in La Pampa and Santa Ee, respectively. In La Pampa, annual crops and pastures were mostly used when plowed than in other crop stages (65% of observations in these cover types, N= 77, 3793 hawks). Wheat was used when in stubble (8 observations, 1232 hawks), and corn and sunflower were used principally when emerged (cover height <40 cm; 12 observations, 1466 hawks and 15 ob- servations, 340 hawks, respectively). In Santa Ee, 41% of 91 observations made on crops and annual pastures were on mowed millet followed by plowed and soybean fields (24% and 24%, respectively). On both areas, hawks were observed with working farm machinery, both while mowing or baling al- falfa and annual pastures (such as millet, 11 obser- vations, 774 hawks) and plowing fields (7 obser- vations, 333 hawks). Additionally, on six occasions (188 hawks), hawks were observed on fields being grazed by cattle. Cropland and annual pastures comprised most of the agricultural landscape in both study areas, followed by permanent pastures (Eig. 4). Wood- lands contributed little to the general land cover. June 2003 Habitat Use by Wintering Swainson’s Hawks 131 Table 1. Analysis of habitat use versus availability for La Pampa and Santa Fe study areas, Argentina. Probability of disproportional use <0.05. ( + ) = used more than expected, (0) = used as expected, and ( — ) = used less than expected (Byers et al. 1984 and Cherry 1996). Habitat Observed Use (Percent) ^ Use Interval Byers’ Intervals Bailey’s Intervals Expected Use (Percent)'^ La Pampa Permanent pasture 0.548 0.473-0.624 (-I-) 0.439-0.646 ( + ) 0.380 Annual pasture 0.010 0.000-0.026 (-) 0.000-0.047 (-) 0.049 Crops 0.124 0.074-0.174 (-) 0.063-0.201 (-) 0.465 Plowed 0.266 0.199-0.332 ( + ) 0.177-0.361 ( + ) 0.105 Woodland 0.052 0.018-0.085 ( + ) 0.015-0.110 ( + ) 0.002 Santa Fe Permanent pasture 0.611 0.547-0.674 ( + ) 0.530-0.683 ( + ) 0.102 Annual pasture 0.183 0.133-0.233 (-) 0.126-0.248 (-) 0.257 Crops 0.059 0.028-0.089 (-) 0.027-0.102 (-) 0.575 Plowed 0.056 0.026-0.086 (0) 0.025-0.099 (0) 0.065 Woodland 0.092 0.054-0.129 ( + ) 0.051-0.143 (-H) 0.005 * Percent of observations of hawks on the ground/cover type. Percent values based on classified satellite image. Swainson’s Hawks used permanent pastures and plowed fields for foraging and woodlands for roost- ing disproportionately with respect to their avail- ability in each study area = 531.8, F < 0.001 in La Pampa; = 625.6, F < 0.001 in Santa Fe). Byers’ and Bailey’s intervals indicated that, in ad- dition to these cover types, the hawks used plowed fields more than available in La Pampa, but in the same proportion as expected based on availability in Santa Fe (Table 1). Cropland and annual pas- tures were used less than expected based on their availability. Discussion Our data indicated that Swainson’s Hawks relied heavily on plowed fields and permanent pastures (Fig. 3) , with the former cover type being used pri- marily for sunning and resting early in the morn- ing and afternoon, and the latter one being used for foraging during midday (Fig. 2). This pattern of habitat use was probably associated with differ- ences in physical attributes and food availability provided by each habitat type. Plowed and recent- ly-harvested fields usually offer the best conditions for the development of thermals during the morn- ing (R. Rodriguez pers. comm.). Swainson’s Hawks spent their time sunning and resting in these fields until thermals developed in the morning and after they subsided in the afternoon. In addition, open fields may provide a greater availability of insects (such as insect larvae) and worms early in the morning and late in the afternoon, given their movements to the upper layers to feed on vegetal matter. Coleoptera larvae and earthworms move on the soil in response to food availability, temper- ature, moisture and soil type (Murton 1973, Braga da Silva 1995). Finally, farm operations such as plowing and discing expose soil insects to birds, which usually forage around working farm machin- ery, acting as natural insect larvae control (Frana and Imwinkelried 1996). Therefore, insect larvae and worms in plowed fields and cropland would have been more abundant in the upper layers of soil when temperatures were low and moisture was high (early in the morning and late in the after- noon) or when tillage operations (such as plowing and discing) exposed them to the surface. Con- versely, flying insects such as grasshoppers are more active as temperatures rise, becoming most active at midday (Liebermann and Schiuma 1946, Capinera and Sechrist 1982, Salto and Beltrame 1999), the time when we observed the hawks main- ly foraging in permanent pastures. Differences in prey availability in different hab- itats were also indicated by the foraging behaviors we observed while the hawks were on the ground. Swainson’s Hawks are adapted to pursue mobile and exposed species such as rodents and swarms of insects, hoth on the ground and in the air (Eng- 132 Canavelli et al. VoL. 37, No. 2 land et al. 1997) . The hawks we observed in plowed fields usually foraged by standing on the ground until they saw grasshoppers and other in- sects. Then, they used short walks, runs, and pecks to catch prey (without necessarily capturing prey with their talons first), pouncing at and running down the insects (“like domestic turkeys do,” Eng- land et al. 1997). Swainson’s Hawks also foraged this way in permanent pastures, but they typically made short flights and jumps in addition to short runs, apparently to flush grasshoppers. Our observations of activity and behavior are consistent with that of a hawk with an insectivorous diet. Swainson’s Hawks mainly feed on grasshop- pers while wintering, although they also prey on other insects such as insect larvae, beetles, and dragonflies according to their availability (Lieber- mann 1935, Pereyra 1937, Liebermann 1944, White et al. 1989, Jaramillo 1993, Rudolph and Fisher 1993, Woodbridge et al. 1995, Goldstein et al. 1996, Serracin Araujo and Tiranti 1996, Gold- stein 1997, Canavelli et al. 2001). Grasshoppers are usually the dominant aboveground invertebrate in pastures and natural grasslands based on biomass (Capinera et al. 1997). Additionally, agricultural practices, such as plowing, discing, mowing, har- vesting, and burning of grasses and stubble are ex- pected to increase the availability of insect prey in annual crops and pastures, as it happens with mammals, given the modification or removal of cover and prey movement generated by these op- erations (Bechard 1982, Tewes 1984, Caldwell 1986). In fact, both in Santa Fe and La Pampa provinces, Swainson’s Hawks were observed on pas- ture and crop fields when operations such as mow- ing, baling and plowing were conducted. There- fore, the observed pattern of habitat use was as expected for an insectivorous bird preying on tem- porarily abundant, easily captured, and often spa- tially-unpredictable insect prey (Alerstam 1990, Sherry and Holmes 1995). Food availability, as determined by prey abun- dance and vegetative structure, could be the key factor that determines the selection of permanent pastures as foraging sites by Swainson’s Hawks dur- ing winter. On their breeding grounds, hawks usu- ally forage on fields that offer less overall vegetative cover and greater prey availability (such as mowed or irrigated alfalfa and harvested cropland, Be- chard 1982, Estep 1989, Woodbridge 1991). Ob- servations made on different fields in Santa Fe province (Argentina) shown that fallow fields, such as old pastures, alfalfa and short-grass fields had the highest grasshopper abundances, the lowest vegetative cover, and the greatest grasshopper availability of all the cover types used by foraging hawks (Canavelli and Salto unpubl. data). The relationship between food availability and habitat selection by this species helps also to ex- plain the significant mortalities of Swainson’s Hawks that have been reported on the wintering grounds (Woodbridge et al. 1995, Goldstein et al. 1996) . At least 5000 Swainson’s Hawks were killed in pasture and crop fields by the misuse of organ- ophosphorous pesticides (Goldstein et al. 1999), probably after they were drawn to the fields by the abundance and availability of grasshopper prey. Twelve of 19 mortality incidents occurred on alfal- fa fields, with individual reports on corn, wheat, and sorghum fields (Canavelli and Zaccagnim 1996, Goldstein 1997, Goldstein et al. 1999). Al- though in this study, crop fields were sporadically used and not selected as foraging habitats, the combination of insect outbreaks and pesticide ap- plications (increase on food availability of flushed prey caused hy the spraying machine on the plot) could have favored the congregations of hawks on these plots, resulting in the massive mortality inci- dents. Additionally, applications of organophos- phorous pesticides on summer crops and pastures in the region mostly occurred between December and February (Canavelli and Zaccagnini 1996), the time when Swainson’s Hawks were wintering in Ar- gentina (Woodbridge et al. 1995, England et al. 1997) , and during the morning or late in the af- ternoon (Grue et al. 1983, R. Bogino pers. comm.), when hawks were mainly sunbathing and foraging on the ground (Fig. 2). Therefore, behav- ioral traits of Swainson’s Hawks and the current practice of pesticide applications may synergistical- ly increase the potential exposure of hawks to the chemicals, making these raptors particularly vul- nerable to such operations. Acknowledgments Funding was provided by the U.S. Fish and Wildlife Service, the National Fish and Wildlife Foundation, U.S Geological Survey, Forest and Rangeland Ecosystem Sci- ence Center, and the National Institute of Agricultural Technology (INTA, Argentina). Thanks to F. Rivera-Mi- lan for his advice on quantitative aspects of the analysis, to A. Lanusse and S. Salva who offered hospitality, help, and logistic support while in La Pampa, to J. Sarasola, F Marcellino, B. Corro-Molas, V. Maisterena, and S. Baker who helped during the surveys. We also extend thanks to C. Salto (INTA, Rafaela Agricultural Experimental Sta- June 2003 Habitat Use by Wintering Swainson’s Hawks 133 tion) and his students, for support and help with grass- hopper biology and logistics while in Santa Fe, and the many INTA people from several Agricultural Experimen- tal Stations that were involved in different aspects of this study. Finally, we thank J. Schmutz, J. Smallwood, D. Rip- er, and the editor for their valuable review comments, which substantially improved this paper. Literature Cited Altrstam, T. 1990. Bird migration. Cambridge Univ. Press, Cambridge, U.K. Alldredge, J.R. and J.T. Ratti. 1992. Further compari- son of some statistical techniques for analysis of re- source selection./. Wildl. Manage. 56:1-9. Babcock, K.W. 1995. Home range and habitat use of breeding Swainson’s Hawks in the Sacramento Valley of California./. Raptor Res. 29:193-197. Bechard, M.J. 1982. Effect of vegetative cover on forag- ing site selection by Swainson’s Hawks. Conrfor84:153- 159. Braga da Silva, M.T. 1995. Aspectos biologicos, danos e controle de Diloboderus abderus Sturm, 1826 (Coleo- pera, Melolonthidae) em plantio direto. Dissertayao de mestrado, Santa Maria, RS Brasil. Buckland, S.T., D.R. Anderson, K.P. Burnham, and J.L. Laake. 1993. Distance sampling: estimating abun- dance of biological populations. Chapman and Hall, London, U.K. Byers, C.R., R.K. Steinhorst, and RR. Krausman. 1984. Clarification of a technique for analysis of utilization- availability data. /. Wildl. Manage. 48:1050-1053. Caldwell, L.D. 1986. Predatory bird behavior and tillage operations. Condor 88:93-94. Canavelli, S.B. 2000. Abundance, movement and habitat use of Swainson’s Hawks on their wintering grounds, Argentina. M.S. thesis, Univ. Florida, Gainesville, FL U.S.A. http://etd.fcla.edu/etd/uf/2000/ana6132/ etd.PDF and M.E. Zaccagnini. 1996. Mortandad de agui- lucho langostero {Buteo swainsoni) en la region para- peana: primera aproximacion al problema. INTA Tech. Rep. Buenos Aires, Argentina. , J J- Maceda, and A.C. Bosisio. 2001. Dieta del aguilucho langostero {Buteo swainsoni) en su area de invernada (La Pampa, Argentina). Hornero 16:89-92. Capinera, J.L. AND T.S. Sechrist. 1982. Grasshoppers (Acrididae) of Colorado: identification, biology, and management. Colorado State Univ. Agric. Exp. Stn. Bull 584s, Fort Collins, CO U.S.A. , C.W. Scherer, and J.B. Simkins. 1997. Habitat associations of grasshoppers at the MacArthur Agro- Ecology Research Center, Lake Placid, Florida. Fla. Entomol. 80:254—261. Cherry, S. 1996. A comparison of confidence interval methods for habitat use-availability studies. /. Wildl. Manage. 60:653—658. CIPA Seccion Argentina. 1987. La presencia actual del aguilucho langostero en Argentina. Nuestras Aves 13: 13-16. Cody, M. 1985. Habitat selection in birds. Academic Press, New York, NYU.S.A. Eastman, J.R. 1997. IDRISI for Windows. User’s guide. Version 2.0. Clark Univ., Worcester, MA U.S.A. England, A.S., M.J. Bechard, and C.S. Houston. 1997 Swainson’s Hawk {Buteo swainsoni). No. 265. In A Poole and F. Gill [Eds.], The birds of North America The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washing- ton, DC U.S.A. Estep, J.A. 1989. Biology, movements and habitat rela- tionships of the Swainson’s Hawk in the Central Valley of California, 1986-87. Calif. Dept. Fish and Game, Nongame Bird and Mammal Sec. Rep., Sacramento, CA U.S.A. Frana, j. and J.M. Imwinkelried. 1996. El complejo de gusanos blancos en trigo. Publicacion Miscelanea No 74. INTA, EEA Rafaela, Centro Regional Santa Fe, Santa Fe, Argentina. Gilmer, D.S. and R.E. Stewart, 1984. Swainson’s Hawk nesting ecology in North Dakota. Condor 86:12-18 Goldstein, M.I. 1997. Toxicological assessment of a neo- tropical migrant on its nonbreeding grounds: case study of the Swainson’s Hawk in Argentina. M.S. the- sis, Clemson Univ., Clemson, SC U.S.A. , B. Woodbridge, M.E. Zaccagnini, and S.B. Ca- navelli. 1996. An assessment of mortality of Swain- son’s Hawks on wintering grounds in Argentina / Raptor Res. 30:106-107. , T.E. Lacher, Jr., B. Woodbridge, M.J. Bechard, S.B. Canavelli, M.E. Zaccagnini, G.P, Cobb, E J. Sco- LLON, R. Tribolet, AND M.J. HOOPER. 1999. Monocro- tophos-induced mass mortality of Swainson’s Hawks in Argentina, 1995-96. Ecotoxicology 8:201-214. Grue, G.E., W.J. Fleming, D.G. Busby, and E.F. Hill. 1983. Assessing hazards of organophosphate pesti- cides to wildlife. Trans. N. Am, Wildl. Nat. Res. Conf 48:200-220. Herkert, J.R. AND F. Knopf. 1998. Research needs for grassland bird conservation. Pages 273-282 in J.M. Marzluff and R. Sallabanks [Eds.], Avian conserva- tion: research and management. Island Press, Seattle, WA U.S.A. Jaramillo, A.P. 1993. Wintering Swainson’s Hawks in Ar- gentina: food and age segregation. Condor 95:475- 479. Kirk, D.A. and C. Hvslop. 1998. Population status and recent trends in Canadian raptors: a review. Biol. Con- serv. 83:91-118. Liebermann, j. 1935. Aves acridiofagas de la Republica Argentina. Hornero 4:82-90. . 1944. El cincuentenario de la visita de Lawrence Bruner a la Argentina. Hornero 7:556-562. and R. Schiuma. 1946. Las “tucuras” mas perju- diciales de nuestra agricultura y ganaderia. Instituto 134 Canavelli et al. VoL. 37, No. 2 de Sanidad Vegetal, Ministerio de Agricultura de la Nacion. Ano II. Serie B. No. 7. Buenos Aires, Argen- tina. Moore, F.R., S.A. Gauthreaux, Jr., P. Kerlinger, and T. R. Simons. 1995. Habitat requirements during mi- gration: important link in conservation. Pages 121- 144 in T.E. Martin and D.M. Finch [Eds.], Ecology and management of neotropical migratory birds. Ox- ford Univ. Press, London, U.K. Morrison, M.L., B.G. Margot, and R.W. Mannan. 1998. Wildlife-habitat relationships. Concepts and applica- tions, 2nd Ed. Univ. Wisconsin Press, Madison, WI U. S.A. Mouchard, a. 1996. Informacion basica sobre el agui- lucho langostero y su conservacion. Recopilacion e mforme interno. Asociacion Ornitologica del Plata, Buenos Aires, Argentina. Murton, R.K. 1973. Man and birds. Taplinger Publish- ing, New York, NY U.S.A. Neu, C.W., C.R. Byers, and J.M. Peek. 1974. A technique for analysis of utilization-availability data. J. Wildl. Manage. 38:541—545. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD U.S.A. Pereyra, J.A. 1937. Miscelanea ornitologica: aves acridio- fagas. Hornero 6:439. Rappole, J.H. 1995. The ecology of migrant birds. A neo- tropical perspective. Smithsonian Institution Press, Washington, DC U.S.A. Rodenhouse, N.L., L.B. Best, R.J. O’Connor, and E.K. Bollinger. 1995. Effects of agricultural practices and farmland structures. Pages 269-292 in T.E. Martin and D.M. Finch [Eds.], Ecology and management of neotropical migratory birds. Oxford Univ. Press, Lon- don, U.K. Rudolph, D.C. and C.H.C. Fisher. 1993. Swainson’s Hawk predation on dragonflies in Argentina. Wilson Bull 105:365-366. Salto, C. and R. Beltrame. 1999. Manejo y reconoci- miento de tucuras. Centro oeste de Santa Fe y centre este de Cordoba. INTA, Centro Regional Santa Fe, Estacion Experimental Agropecuaria Rafaela. Publi- cacion Tecnica No. 59. Rafaela, Argentina. SCHMUTZ, J.K. 1984. Ferruginous and Swainson’s Hawk abundance and distribution in relation to land use in southeastern Alberta. /. Wildl Manage. 48:1180-1187. . 1989. Hawk occupancy of disturbed grasslands in relation to models of habitat selection. Co«dor91:362- 371. Serracin Araujo, R. and S.I. Tiranti. 1996. Stomach contents of a Swainson’s Hawk from Argentina./. Rap- tor Res. 30:105—106. Sherry, T.W. and R.T. Holmes. 1995. Summer versus win- ter limitation of populations: what are the issues and what is the evidence? Pages 85-120 in T.E. Martin and D.M. Finch [Eds.], Ecology and management of neo- tropical migratory birds. Oxford Univ. Press, London, U.K. Smallwood, K.S. 1995. Scaling Swainson’s Hawk popu- lation density for assessing habitat use across an ag- ricultural landscape. / Raptor Res. 29:172-178. Tella, J.L., M.G. Forero, F. Hiraldo, and J.A. Donazar. 1998. Conflicts between Lesser Kestrel conservation and European agricultural policies as identified by habitat use analysis. Conserv. Biol 12:593-604. Tewes, M.E. 1984. Opportunistic feeding by White-tailed Hawks at prescribed burns. Wilson Bull 96:135—136. Thomas, D.L. and E.J. Taylor. 1990. Study designs and tests for comparing resource use and availability. /. Wildl Manage. 54:322-330. Watson, Ch., E. Atkinson, K. Steenhof, and J. Roten- berry. 1996. Abundance and distribution of raptors and ravens across the ISA benchlands, 1991-1994. Chapter 4F in Final Report. U.S. Dep. Inter. USGS, BRD, Boise, ID U.S.A. White, C.M., D.A. Boyce, and R. Straneck. 1989. Obser- vations on Buteo swainsoni in Argentina, 1984, with comments on food, habitat alteration and agricultural chemicals. Pages 79-87 in B.-U. Meyburg and R.D. Chancellor [Eds.], Raptors in the modern world WWGBP, Berlin, Germany. WOFFINDEN, N.D. 1986. Notes on the Swainson’s Hawk in central Utah: insectivory, premigratory aggregations, and kleptoparasitism. Great Basin Nat 46:302-304. WOODBRIDGE, B. 1991. Habitat selection by nesting Swain- son’s Hawks: a hierarchical approach. M.S. thesis, Oregon State Univ., Corvallis, OR U.S.A. , K. Finley, and S.T. Seager. 1995. An investiga- tion of the Swainson’s Hawk in Argentina. /. Raptor Res. 29:202-204. Received 18 July 2002; accepted 26 February 2003 J. Raptor Res. 3'7(2) :135-146 © 2003 The Raptor Research Foundation, Inc. Raptor and Chihuahuan Raven Nesting on Decommissioned Telephone-line Poles in the Northern Chihuahuan Desert Don L. Brubaker^ and Kathleen L. Brubaker^ New Mexico Cooperative Fish and Wildlife Research Unit and Fishery and Wildlife Sciences Department, New Mexico State University, P.O. Box 30003, MSC 4901, Las Cruces, NM 88003 US. A. Bruce C. Thompson^ us. Geological Survey-Biological Resources, New Mexico Cooperative Fish and Wildlife Research Unit, P.O. Box 30003, MSC 4901, Las Cruces, NM 88003 US. A. Abstract. — ^White Sands Missile Range (WSMR) in south-central New Mexico includes ca. 800 000 ha of northern Chihuahuan Desert, where litde is known regarding raptor and Chihuahuan Raven {Corvus cryptoleucus) nesting. We studied 338 decommissioned telephone-line poles with 22 different cross-arm configurations and six electrical poles in 10 vegetation cover types from 29 March-14 September 1997. We encountered 64 Chihuahuan Raven nests on decommissioned telephone-line poles. We found 27 Swainson’s Hawk {Buteo swainsoni) nests, most often on a configuration with two vertical poles supporting four paired sets of cross-arms. Using a geographic information system analysis, we found Chihuahuan Raven nests similarly distributed in all vegetation cover types, while honey mesquite {Prosopis glandulosa) desertscrub covered a mean of 64.2% of the area within 3 km of each Swainson’s Hawk nest. Decommis- sioned telephone-line poles benefit raptors and Chihuahuan Ravens as nesting platforms, and should be managed and retained where possible in the southwest. We suggest that pole salvage operations in the northern Chihuahuan Desert should retain poles with at least two sets of paired cross-arms, which are suitable raptor and raven nesting sites. For management of Swainson’s Hawks, configurations with two vertical poles supporting four paired sets of cross-arms seemed most suitable in the area we investigated. Key Words: Swainson’s Hawk, Buteo swainsoni; Chihuahuan Raven; Corvus cryptoleucus; Chihuahuan Desert, CIS; nest habitat; nest structure; poles. ANIDACION DE RAPACES Y DEL CUERVO DE CHIHUAHUA EN POSTES DE LINEAS TELEEONI- CAS DESCONTINUDAS EN EL NORTE DEL DESIERTO DE CHIHUAHUA Resumen. — La zona de prueba de misiles de White Sand (WSMR) en el centro-sur de nuevo Mexico incluye ca. 800,000 ha del norte del Desierto de Chihuahua, donde es poco se lo que se sabe sobre la nidacion de rapaces y del cuervo de Chihuahua (Corvus cryptoleucus). Estudiamos 338 postes de lineas telefonicas descontinuadas con 22 configuraciones con estructuras cruzadas (X) diferentes y seis postes electricos en 10 tipos de cobertura vegetal, desde el 29 de Marzo-14 Septiembre 1997. Encontramos 64 nidos de cuervos de Chihuahua en postes de lineas telefonicas descontinuadas. Encontramos 27 nidos de halcones de Swainsoni (Buteo swainsoni) , la mayoria a menudo, en una configuracion con dos postes verticales soportando 4 pares de sets de cruces (X). Usando analisis de sistema de informacion geogra- fica, encontramos nidos de cuervos de Chihuahua similarmente distribuidos en todos los tipos de co- bertura vegetal, mientras que Prosopis glandulosa cubrio una media de 64.2% del area de 3 km de cada nido del gavilan de Swainsoni. Los postes de lineas telefonicas descontinuados benefician a las rapaces y a los cuervos de Chihuahua ya que sirven como plataformas de anidamiento y deberian ser manejadas y retenidas/mantenidas donde sea posible, en el suroeste. Sugerimos que operaciones de postes silves- tres en el norte del Desierto de Chihuahua deberian retener/mantener postes con al menos dos sets de cruces (X) pareados, los cuales son sitios de anidacion apropiados para rapaces y cuervos. Para el manejo del gavilan de Swainsoni, configuraciones con dos postes verticales soportando cuatro sets pa- reados de cruces (X), parecieron ser mas apropiados en el area que investigamos. [Traduccion de Cesar Marquez] ^ Present address: U.S. Fish and Wildlife Service, San Diego National Wildlife Refuge, 13910 Lyons Valley Road, Suite R, Jamul, CA 91935 U.S.A.; e-mail address: DonJBrubaker@rl.fws.gov ^ Present address: U.S. Fish and Wildlife Service, 6010 Hidden Valley Road, Carlsbad, CA 92009 U.S.A. ^ Present address: New Mexico Dept, of Game and Fish, P.O. Box 25112, Santa Fe, NM 87504 U.S.A. 135 136 Brubaker et al. VoL. 37, No. 2 Few investigators have examined raptor and ra- ven nesting on artificial platforms in northern Chi- huahuan Desert vegetation cover types. During a 3-yr study in southern New Mexico, Kimsey and Conley (1988) found desert, irrigated agricultural, and riparian areas with power-transmission-line support poles received more use by raptors than similar areas devoid of poles. Other studies have described raptors nesting on natural substrates in southern New Mexico. Pilz (1983) found Swainson’s Hawks {Buteo swainsoni) in southern New Mexico nesting in grasslands on soaptree yuccas {Yucca data). England et al. (1997) described this hawk as nesting in scattered trees in grasslands, shrublands, and agricultural areas oc- casionally, with some nests used more than 1 yr. Bednarz and Hoffman (1988) found Swainson’s Hawks nesting in large honey mesquite {Prosopis glandulosa) and soapberry trees {Sapindus drum- mondii) dispersed over a landscape covered by hon- ey mesquite, shinnery oak {Quercus havardii), sand sage {Artemisia filifolia) , snakeweed {Gutierrezia sa- rothrae), and creosotebush {Larrea tridentata) in southeastern New Mexico. Eleven nests of Chihua- huan Ravens {Corvus cryptoleucus) and one Swain- son’s Hawk nest in southeastern New Mexico were in mesquites taller than surrounding vegetation (Griffing 1974, Davis and Griffing 1977). Bednarz and Raitt (2002) described this raven as nesting in various shrubs and trees in grasslands, shrublands, and woodlands, and on human-made structures in- cluding poles. Raptor and raven nesting ecology, on artificial and natural substrates, has been studied extensive- ly in other regions of North America. Steenhof et al. (1993) monitored progression of use by Com- mon Ravens {Corvus corax) and raptor species on newly erected power-line-transmission towers in Idaho and Oregon. In the area of these towers, nesting pairs of Common Ravens, Golden Eagles {Aquila chrysaetos). Red-tailed Hawks {Buteo jamai- censis), Ferruginous Hawks {Buteo regalis), and Great Horned Owls {Bubo virginianus) increased during the 9-yr study. In that study, nesting pairs of raptors and Common Ravens gradually aban- doned nearby natural nesting substrates and began nesting on the available towers. Bechard et al. (1990) found Red-tailed Hawk, Ferruginous Hawk, and Swainson’s Hawk nesting substrates in Washington included cliffs, transmis- sion-line towers, windmills, and trees. Red-tailed Hawks nested closest to water, primarily in the tall- est and oldest trees, and in areas with large amounts of shrub and grassland habitat. Swain- son’s Hawks nested entirely in trees close to roads and human structures; nearly half of the nests were in areas where the surrounding habitat was domi- nated by wheat fields. This is consistent with Bed- narz’ (1988) investigation of breeding ecology of Swainson’s Hawks in southeastern New Mexico, where grasses were a key component among the nest habitats of this species. White Sands Missile Range (WSMR) in south- central New Mexico represents an ecologically sig- nificant portion of northern Chihuahuan Desert because of restricted access and relatively low dis- turbance compared to outside the WSMR bound- ary. The northern Chihuahuan Desert within WSMR has been devoid of agricultural activity in- cluding livestock grazing for over 50 years. Addi- tionally, other anthropogenic developments such as buildings and associated infrastructures are lim- ited and confined to isolated areas throughout WSMR. WSMR contains numerous decommissioned tele- phone-line poles and active electrical-transmission- line poles, each consisting of a variety of configu- rations among several vegetation cover types. A pole configuration consists of one to several verti- cal poles supporting one to several paired or un- paired cross-arms. The telephone-line poles pre- sent in 1997 were retained during WSMR-wide pole removal in the early 1990s. Retained poles were those that contained a nest structure during pole removal operations and ca. 10% of other poles that had paired cross-arms, regardless of presence of a nest structure. We investigated decommissioned telephone-line poles to assess use of these struc- tures by nesting raptors and ravens on WSMR. Us- ing a geographic information system, we described the vegetation cover types around each decommis- sioned telephone-line pole. Study Area WSMR is a military and space systems test and evalua- tion facility operated by the U.S. Army. Located in south- central New Mexico, WSMR encloses ca. 8000 km^ of the northern Chihuahuan Desert. Topography of WSMR in- cludes portions of the Tularosa Basin, the Jornada del Muerto, and several mountain ranges. The WSMR por- tion of the Tularosa Basin contains extensive gypsum dunes, lava beds, level to rolling grasslands, small marsh- es, and salt flats. Vegetation includes vast patches of hon- ey mesquite, creosotebush, four-wing saltbush {Atriplex ca- nescens), and grasslands of a variety of species (Dick-Peddie 1993). The Jornada del Muerto of WSMR June 2003 Raptor and Raven Nesting on Poles 137 includes extensive patches of grasslands, honey mesquite, sand sage, and creosote bush. Two prominent mountain ranges exist on WSMR. The San Andres Mountains dominate the west central portion of WSMR, while the Oscura Mountains dominate the northeast. High elevations of both ranges are covered with pinyon pine {Pinus edulis) and one-seed juniper (Ju- niperus monosperma) . The rocky slopes and cliffs of both ranges are covered with mountain mahogany ( Cercocarpus montanus) and piedmont grasslands that give way to level and rolling topography of the Tularosa Basin and Jor- nada del Muerto. Weather conditions over the time of this investigation (29 March-14 September 1997) were compiled from 19 surface atmospheric monitoring stations throughout WSMR. Mean monthly precipitation throughout lowland portions of WSMR, where most decommissioned poles exist, was 3.17 cm. Maximum temperature was 44°C and minimum temperature was — 6°C. Mean temperature over a 24-hr period during this investigation was 19°C. Mean wind speed throughout WSMR was 14 km/hr with peak gusts of l74 km/hr in the Jornada del Muerto and 163 km/hr in the Tularosa Basin. These were typical weather conditions on WSMR. Developed areas of WSMR are largely limited to clus- ters of one to three buildings distributed throughout the range. Historically, the area that is now WSMR contained several cattle, goat, and sheep ranches. However, ranch- ing was terminated when WSMR was officially established in 1945. Since then, grazing has been limited to feral horses and gemsbok ( Oryx gazella) , an African antelope, introduced in the late 1960s, The landscape within WSMR is minimally disturbed, as impacts of weapons and missile testing over the last 50 yr have been limited to specific areas. Methods Pole Examination. All decommissioned telephone-line poles within WSMR were located using maps of the tele- communications network, or were found while conduct- ing other investigations. Decommissioned telephone-line poles in areas of highly-restricted access (bombing, and air-to-ground gunnery ranges) on WSMR were not ex- amined. Fewer than eight poles were known to exist in those areas. Locations for all poles were collected with a global po- sitioning system (GPS) receiver and differentially cor- rected to UTM coordinates (datum = NAD83). Poles were examined by traveling maintenance access roads that paralleled a line of poles. Pole configurations were characterized by the number of poles supporting a num- ber of paired or unpaired cross-arms (Fig. 1). For ex- ample, a pole configuration could include two vertical poles supporting four paired or parallel cross-arms. Cross-arms were numbered from the top down on all poles. Decommissioned telephone-line and electrical poles were examined from 5 April-2 July 1997. If a pole had an occupied raptor or raven nest, the species and date were noted and the pole was revisited during August- September 1997 to measure pole characteristics after the young had fledged. A compass was used to collect a bearing of the long axis of cross-arms on each pole to determine pole ori- entation. Pole orientation was used to determine nest ex- posure (placement of the nest on cross-arms relative to exposure to wind and solar radiation). This measure- ment was taken to investigate whether raptors or ravens nested on a specific side of a vertical pole relative to the pole’s orientation. The bearing was taken in the direction the nest received the most exposure. Heights of pole features were measured with a tele- scoping fiberglass rule to the nearest centimeter. Cross- arm heights, distances between cross-arms, distances be- tween the top of the nest and underside of the above cross-arm, and pole heights were measured only on those decommissioned telephone poles with occupied and un- occupied nests. Herein, decommissioned telephone-line pole configurations are identified using the number of vertical poles and the number of paired cross-arms. For example, a single vertical pole supporting four sets of cross-arms is denoted as a 1X4. Vegetation Characterization, Vegetation cover types within a 3-km radius of each pole were identified using a GIS. These cover types included arid grasslands domi- nated by blue {Bouteloua gracilis), hairy {B. hirsuta), and black {B. eriopoda) grama, mesa dropseed {Sparoholus flex- uosus), giant sacaton {Sporobolus imightii), and tobosa (Pleuraphis mutica); shrublands dominated by four-wing saltbush, creosotebush, honey mesquite, and mountain mahogany; and woodlands dominated by pinyon and ju- niper (Dick-Peddie 1993, Muldavin et al. 1996). From the GPS location of each pole, a polygon with a 3-km radius was placed around eaeh point location. The polygon equated to a vegetation sampling area of 28 1 km^, or 2810 ha in the GIS (Arcinfo, Environmental Sys- tems Research Institute, Redlands, CA). This vegetation assessment area was chosen to permit inclusion of home range sizes for raptors known to nest in the area. Tabu- lation of vegetation cover types with a GIS was conducted using a digital version of the New Mexico Gap Analysis Project (NM-GAP) vegetation data set (Muldavin et al. 1996, Thompson et al. 1996). This digital data set pro- vided vegetation classes mapped at 100-ha resolution with an accuracy of 77% at the most general classification (Thompson et al. 1996). Results Pole Examination. We examined 338 decommis- sioned telephone-line poles of 22 different config- urations on WSMR. Of the 22 configurations, four comprised 86.3% of the available decommissioned poles (Fig. la, b, c, d). Three configurations, con- sisting of 12 poles, were considered incapable of supporting raptor or raven nests because of un- paired cross-arms and overall poor condition (e.g., Fig. le). We located 23 box-style decommissioned telephone-line poles comprising 10 configurations on WSMR (Fig. If). Orientations of decommissioned telephone-line poles were not evenly distributed, with 67.4% of 311 poles oriented east/west or northwest-south- east (Table 1). Box-style configurations were not 138 Brubaker et al. VoL. 37, No. 2 fipj=» g g gL Jp q ,ggggg^D?gs>^.g.sj eeeee iHE g gg.§L fi gg>g&geLjgg. ae&^ £ £ i q 8'^ g^ggggpgggggg, ojx^ool- Iqoq^t^o O c o, c I 5 5 5 a P o D ^eeneeeeeqnee yqngqagqqlJqq yqLJqqqgOCnqq LoUa ^OQa;^aflIiroflj ^ aal^aaasafll las oi c e' ? o C I 38 e C f Figure 1. Examples of decommissioned telephone-line pole configurations found in the northern Chihuahuan Desert on White Sands Missile Range, New Mexico, 1997. used in orientation and exposure analyses because these configurations were oriented in multiple di- rections. Dimensions of decommissioned tele- phone-line pole features such as distance between cross-arms, length of cross-arms, and distance be- tween vertical poles were consistent among pole configurations. Cross-arms were 3 m long and ver- tically spaced 50 cm apart. Vertical poles were spaced 2 m apart in two- and four-pole configura- tions. Decommissioned telephone poles were 7.94— 10.55 m tall with heights of cross-arms varying ac- cordingly. No pole configuration was consistently June 2003 Raptor and Raven Nesting on Poles 139 Table 1 . Orientation of decommissioned telephone-line poles used as nesting substrates by the two most common pole-nesting species in the Chihuahuan Desert on White Sands Missile Range, New Mexico, 1997, Orientation East-West North- EAST- South- WEST North- South North- WEST- SOUTH- east All poles 114" 65 36 96 (N= 311) (36.6)'" (20.9) (11.6) (30.8) Chihuahuan Ravens 21 11 7 22 {N= 61) (34.4) (18.1) (11.5) (36.1) Swainson’s Hawks 12 4 4 6 (N = 26) (46.2) (15.4) (15.4) (23.1) ^ Number of poles having a specific orientation. ^ Percentage of occurrence. taller than others. Within a line of decommis- sioned telephone-line poles, distances between poles ranged from 4.6-1200 m. Raptor and Raven Nests. We detected 64 Chi- huahuan Raven, 27 Swainson’s Hawk, three Red- tailed Hawk, and two Great Horned Owl nests among the decommissioned telephone-line poles (Fig. 2). Four Red-tailed Hawk and two Chihua- huan Raven nests were detected on electrical poles (Fig. 2). Nests on telephone-line poles were con- centrated in the Tularosa Basin, while nests on electrical poles were encountered mostly on the Jornada del Muerto (Fig. 2). Subsequent analyses and discussion focus only on Chihuahuan Raven and Swainson’s Hawk nests on decommissioned telephone-line pole nests. Chihuahuan Raven Nests. Of 19 telephone-line pole configurations available as nesting platforms, Chihuahuan Ravens nested on five (Table 2). Ad- ditionally, for configurations with more than one vertical pole, raven nests were most often between vertical poles (Table 3). Also, 37.1% of raven nests were on cross-arm number two, while cross-arm number four supported 33.9% of the nests (Table 2) . Mean heights of cross-arms two and four were 7.88 m (SD = 0.42) and 7.01 m (SD = 0.67), re- spectively. Analysis of exposure potential for 61 Chihua- huan Raven nests (data missing for three nests) indicated orientation away from direction of pre- vailing wind and afternoon solar radiation. We found that 26.2%, 19.7%, and 16.4% of these nests were exposed to the east, southeast, and northwest, respectively (Table 4). Chihuahuan Raven nests on telephone poles dis- played a similar distribution among vegetation cov- er types and pole configurations as was present on WSMR (Table 5) . Within 3 km of Chihuahuan Ra- ven nests, honey mesquite desert scrub covered a mean of 13.3 km^ or 48.3% of each of the 28.1 km^ polygons around 64 poles containing Chihua- huan Raven nests (Table 5). Black grama/drop- seed grassland and creosotebush desert scrub cov- ered most of the remaining area (Table 5). Analysis of vegetation cover types within 3 km of the four pole configurations most used by nesting Chihuahuan Ravens indicated some differences in composition between the four-cross-arm configu- rations and the five-cross-arm configurations. Area around 1X4 and 2X4 configurations used by Chi- huahuan Ravens had fewer vegetation cover types than the 1X5 and 2X5 configurations used by this species (Fig. 3). Four-wing saltbush and black grama/dropseed grassland covered the greatest area within the 3-km radius polygons around Chi- huahuan Raven nests on lX5s compared to avail- able lX5s. Chihuahuan Raven nests on 2X5s had more honey mesquite desert scrub covering the area around these poles relative to all 2X5s. Swainson’s Hawk Nests. Swainson’s Hawks nest- ed on five configurations of decommissioned tele- phone-line poles (Table 6). Nearly half of the nests (48.2%) were on the 2X4 configuration; however, this configuration comprised 19.2% of all decom- missioned poles (Table 6). Exposure potential for 26 Swainson’s Hawk nests (data missing for one nest) revealed that 34.6% were exposed to the east and 23.1% were exposed to the northwest (Table 4). All Swainson’s Hawk nests on telephone poles were in the Tularosa Basin (Fig. 2). Most (66.6%) of Swainson’s Hawk nests were on cross-arm two (Table 6) . Mean height of cross-arm two on the 1X4 configuration was 7.54 m (SD = 0.55) and 8.27 m (SD - 0.70) for the 2X4 config- uration. Swainson’s Hawk nests were placed most often between vertical poles on configurations with more than one vertical pole (Table 3). Honey mesquite desert scrub covered a mean of 18.1 km^ or 62.8% of the area in each 28.1 km^ polygon around poles with Swainson’s Hawk nests (Table 5). Black grama/dropseed grassland con- 140 Brubaker et al. VoL. 37, No. 2 Figure 2. Distribution of Chihuahuan Raven (C), Swainson’s Hawk (S), Red-tailed Hawk (R), and Great Horned Owl (G) nests on decommissioned telephone-line poles, and nests of Red-tailed Hawks ® and Chihuahuan Ravens © on active electrical transmission line poles in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. tributed 17.6% of the remaining area, or a mean of 4.8 km^ of polygon area (Table 5). Analysis of vegetation cover types within 3 km of the two-pole configurations most used by nesting Swainson’s Hawks indicated that land-cover around the 2X4 configuration was composed mostly of honey mesquite desert scrub (Fig. 4) . Additionally, land-cover within 3 km of the 1X4 configuration was more diverse than that of the 2X4 conhgura- tion (Fig. 4). Discussion Greater use of 2X4 cross-arm configurations by Swainson’s Hawks and 2X5 configurations by Chi- June 2003 Raptor and Raven Nesting on Poles 141 Table 2 . Placement of 62 Chihuahuan Raven nests among cross-arms on five decommissioned telephone-line pole configurations in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997.'* Cross-arm Pole Configurations*' No.** 1X4 2X4 1X5 2X5 4X22 No. (Percent) 1 0 0 0 0 0 0 2 8 4 1 10 0 23 (37.1) 3 4 3 0 3 1 11 (17.7) 4 6 6 6 3 0 21 (33.9) 5 n/ /a y /a 1 6 0 7 (11.3) No. (%) 18 (29.0) 13 (20.9) 8 (12.9) 22 (35.5) 1 (1.6) •* Although 64 Chihuahuan Raven nests were detected on decommissioned telephone poles on WSMR, data were not collected for two nests. ^ Cross-arms are numbered from top down. Percent occurrence of these five configurations among 338 decommissioned poles: 27.8, 26.3, 13.1, 28.6, and 6.8%, respectively huahuan Ravens suggests selection for a particular pole configuration, although we did not specifical- ly test preference. Both configurations were inter- spersed with single pole configurations. Configu- rations with two vertical poles may provide greater protection from wind and sun exposure than would single pole configurations. Nearly all nests, either occupied or unoccupied, were inside or be- tween the two vertical poles. Only nine nests were placed on the outside of two-pole configurations, with one nest on the outside of a box-style config- uration. By contrast, 135 nests were placed inside two-pole and box-style configurations. On portions of WSMR where prevailing winds blow in the direction of the long axis of cross-arms, nests were placed on the leeward side of the pole Table 3. Placement of 36 Chihuahuan Raven and 17 Swainson’s Hawk nests among cross-arms relative to vertical poles on three decommissioned telephone-line pole configurations in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997.** Nest Placement by Pole Configuration Cross-arm 2X4 2X5 4X22 No.** Inside*^ Outside** Inside Outside Inside Outside Totals Chihuahuan Raven 1 0 0 0 0 0 0 0 2 4 0 9 1 0 0 14 3 3 0 3 0 0 0 6 4 5 1 3 0 0 1 10 5 % y 'a 6 0 0 0 6 Totals 12 1 21 1 0 1 36 Swainson’s Hawk 1 0 0 0 0 0 0 0 2 7 1 1 1 0 0 10 3 1 0 0 0 1 0 2 4 3 1 1 0 0 0 5 5 y '2i % 0 0 0 0 0 Totals 11 2 2 1 1 0 17 ** Only configurations with two or four vertical poles are included in this tabulation. ** Cross-arms are numbered from top down. Represents nests placed inside or between vertical poles. Represents nests placed outside vertical poles. 142 Brubaker et al. VoL. 37, No. 2 Table 4. Percent of primary exposure directions among 61 Chihuahuan Raven and 26 Swainson’s Hawk nests on decommissioned telephonedine poles in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. Primary Exposure Percent of Nests Chihuahuan Raven {N= 61) Swainson’s Hawk {N = 26) North 3.3 7.7 Northeast 8.2 7.7 East 26.2 34.6 Southeast 19.7 0.0 South 8.2 7.7 Southwest 9.8 7.7 West 8.2 11.5 Northwest 16.4 23.1 relative to the origin of the winds. Nests, regardless of cross-arm, were placed against the poles in all cases, further suggesting some form of protection from wind or sun. Smith and Murphy (1982) sug- gested that raptor species tend to select nest sites with exposures that offer an optimum microcli- mate within reasonable variations. Steenhof et al. (1993) found Common Ravens nesting in the dens- est sections of steel electrical transmission line tow- ers, and they speculated that nests in these sections provided better protection from high winds that caused nest destruction in other sections of the towers. We found no nests on the topmost cross-arm on telephone poles. Mosher and White (1976) indi- cated that exposure of young raptors to extreme temperatures and direct sun may be a major source of thermoregulatory stress during early stages of nestling development. Schmutz et al. (1984) found Swainson’s and Ferruginous hawks preferred shel- tered nest platforms relative to non-sheltered plat- forms in Alberta, Canada. Over half of the Swain- son’s Hawk nests and more than a third of Chihuahuan Raven nests detected on decommis- sioned telephone-line poles on WSMR were on cross-arm two. This may indicate selection of a high nest platform for prey detection and predator avoidance, while maintaining some protection from exposure. Swainson’s Hawks in southern New Mexico nest in a variety of vegetation cover types. Bednarz and Hoffman (1988) reported Swainson’s Hawks nest- ing in large honey mesquite and soapberry trees among scrub habitat in southeastern New Mexico. Pilz (1983) described the habitat of Swainson’s Hawks nesting in soaptree yuccas in southern New Mexico as sparsely covered by honey mesquite; however, black grama, hairy grama, and red three- awn (Aristida oligantha) grasslands occurred throughout the area. Table 5. Percentage of vegetation cover types present in the area encompassed by all 3-km radius polygons sur- rounding 338 decommissioned telephone-line poles and poles with Chihuahuan Raven and Swainson’s Hawk nests in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. Percent Vegetation Cover Type Among Species and Pole Status Vegetation Cover Types Chihuahuan Raven Nests {N = 64) Swainson’s Hawk Nests {N = 27) Poles With Unoccupied Nests {N= 101) Poles With No Nests {N= 140) All Poles {N = 338) Pinyon/juniper — Closed^ 0.1 0.1 0.00 6.6 2.8 Pinyon/juniper — Open*^ 1.1 0.9 0.8 6.1 3.1 Mountain mahogany chaparral 0.7 0.1 0.4 1.9 1.1 Four-wing saltbush desert scrub 8.1 9.5 6.4 4.4 6.0 Creosotebush desert scrub 12.4 5.0 8.1 18.1 13.2 Honey mesquite desert scrub 48.3 62.8 61.6 36.4 47.9 Blue/hairy grama grassland 0.5 0.2 0.7 1.3 0.8 Black grama/ dropseed grassland 23.3 17.6 17.1 19.9 19.8 Tobosa/ giant sacaton grassland 1.9 3.7 1.6 0.5 1.4 Barren/rock outcrop/playa 4.6 0.2 3.2 4.7 3.8 Canopy cover >60%. ^ Canopy cover = 25% to 50%. Percentage of Total Area Covered June 2003 Raptor and Raven Nesting on Poles 143 Nests on 1X4’s (A/ = 18) All 1X4‘s (N = 95) ■ Nests on 1X5’s (A/ = 8) ■ All 1X5's (A/ = 43) 90 n 80 . ■ Nests on 2X5's (A/ = 22) 70- ■ AII2X5's(A/ = 89) Piny on/ Piny on/ Mountain Fourwing Creosote Honey Blue/haby Black Tobosa/ Barren kin^r- kiniper- mahogany saltbush desert mesquite grama grama/ giant Ciosed Open desert scrub desert grassland drops eed sacaton scrub scrub grassland grassland Vegetation Cover Types Figure 3. Percentage of total area for 10 vegetation cover types within 3-km of Chihuahuan Raven (CHRA) nests on two 5-cross-armed configurations of decommissioned telephone-line poles in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. 144 Brubaker et al. VoL. 37, No. 2 Table 6. Placement of 27 Swainson’s Hawk nests among cross-arms on five decommissioned telephone-line pole configurations in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. Pole Conugurations Total Cross-arm No.^ lX4b 2X4 1X5 2X5 4X13 (Percent) 1 0 0 0 0 0 0 2 6 8 2 2 0 18 (66.6) 3 1 1 0 0 1 3 (11.1) 4 1 4 0 1 0 6 (22.2) 5 % % 0 0 0 0 Total (Percent) Percent occurrence of 8 (29.6) 13 (48.2) 2 (7.4) 3 (11.1) 1 (3.7) configuration among all poles (N = 338) b 28.1 19.2 12.7 26.3 6.8 ^ Cross-arms are numbered from top down. ^ Percent values do not account for 23 poles with configurations other than those included in table. On WSMR, land cover within 3 km of all Swain- son’s Hawk nests on decommissioned telephone- line poles was dominated by honey mesquite desert scrub. Furthermore, land cover within 3 km of the 2X4 telephone-pole configuration was 87.4% hon- ey mesquite desert scrub compared to 63.5% for all lX4s along the same route. The 1X4 pole con- figuration was interspersed among 2X4s along routes, however, fewer of these poles were used by Swainson’s Hawks as nest poles. This suggests that for Swainson’s Hawks, pole configuration coupled with vegetation cover type are important features for nest site selection. Chihuahuan Ravens nested more often on 2X4 and 2X5 configurations than on 1X4 and 1X5 con- figurations even though the single pole configu- rations were interspersed among the two-pole con- figurations along routes, but there were no clear differences between used and unused poles. This suggests that Chihuahuan Ravens may select poles 90 T 80 ' 70 ' 60 ' § o 0 1 ■s 50 H o 40 - *5 ^ 30 ' 20 - c § 0> Q. 10 Nests on 1X4's (N = 7) AII1X4's (N = 95) Nests on 2X4's (N = 13) All 2X4's (N = 65) I ■■ Hnyon/ juniper - Closed Rnyon/ Mountain Four-wing Creosotebush htoney Blue/hairy Black Tobosa/ juniper - mahogany saltbush desert scrub mesquite grama grama/ giant Open desert scrub desert scrub grassland dropseed sacaton grassland grassland Vegetation Cover Types Barren Figure 4. Percentage of total area for 10 vegetation cover types within 3-km of Swainson’s Hawk (SWHA) nests on two 4-cross-armed configurations of decommissioned telephone poles in the northern Chihuahuan Desert, White Sands Missile Range, New Mexico, 1997. June 2003 Raptor and Raven Nesting on Poles 145 based on configuration rather than the land-cover around nest poles. The two-pole configurations probably afford better protection from exposure to wind and sun. Box-style configurations on WSMR were suitable nesting platforms for raptors and Chihuahuan Ra- vens, with 8-22 paired cross-arms providing a dense assortment of potential nesting platforms. The 23 box-style configurations comprised 6.8% of the total available decommissioned poles on WSMR, but they supported 4.3% of nests. Box-style configurations may not be used readily for nest placement by raptors and Chihuahuan Ravens on WSMR for three reasons: (1) density of the config- uration reduces visibility for detection of predators or intruding conspecifics; (2) predators capable of climbing, such as ring-tailed cats {Bassariscus astu- tus) or bobcats {Felis rufus), may be more apt to climb on these configurations successfully because of the dense supportive network of cross-arms; and (3) box-style configurations had telecommunica- tion wiring criss-crossing inside the configuration. All one and two vertical pole configurations en- countered on WSMR had the wiring removed. Box- style configurations still had wiring inside, and in some instances, outside the box, thus possibly cre- ating obstacles to flight. We examined essentially all (>97%) of the de- commissioned telephone-line poles on WSMR dur- ing 1997, thus we considered these data to repre- sent the effective population of poles and related nests rather than a sample. Further, nest structures on poles represent nest placement that has pre- vailed for several years. Therefore, we interpreted directly observed characteristics of poles and sur- rounding vegetation as indicative of nesting asso- ciated with the population of poles that year. We have not applied statistical tests because samples were not compared. We recognize that longer-term investigation can sample several breeding seasons and draw inferences about comparisons among years. Our inferences are biologically based on ex- tensive observation of poles and nests overall, they are not grounded in specific statistical applications. Our subsequent statements of implications should be viewed with that perspective in mind. Conservation Implications Decommissioned telephone-line poles and elec- trical poles are acceptable nesting sites for raptors and Chihuahuan Ravens in the northern Chihua- huan Desert. Before WSMR was established, Chi- huahuan Ravens, Swainson’s Hawks, Red-tailed Hawks, and Great Horned Owls probably nested on natural substrates and possibly on an occasional windmill. Since introduction of telephone and electrical poles, raptors and Chihuahuan Ravens have adapted to these substrates for nesting. This investigation focused on raptors and Chi- huahuan Ravens nesting on decommissioned poles. However, other avian species including Log- gerhead Shrikes {Lanius ludovicianus) , Western Kingbirds (Tyrannus verticalis), and Scott’s Orioles {Icterus parisorum) also nested on these poles. West- ern Kingbirds and Scott’s Orioles sometimes shared the same nest structure with Swainson’s Hawks. Cavities in the vertical poles, possibly ex- cavated by breeding Ladder-backed Woodpeckers {Picoides scalaris), have provided nest cavities for Ash-throated Flycatchers {Myiarchus cinerascens) and, possibly, Elf Owls (Micrathene whitneyi). Like snags in forests or saguaros (Carnegiea gigantea) in arid landscapes, these decommissioned poles are used by a variety of migratory species in the Chi- huahuan Desert on WSMR. Removal of decommis- sioned poles could be detrimental to the present population of raptors and Chihuahuan Ravens nesting on these poles. If poles are removed, rap- tor and ravens may turn to nesting on structures used for testing operations on WSMR and, thus, become vulnerable to nest disturbance. Also, other birds associated with poles and raptor nests would be displaced. We suggest that pole-management strategies be developed and implemented by land stewards in southwestern deserts with property where pole lines remain and may be altered or dismantled in the future. Such strategies should provide for pro- tection, maintenance, and prospective replace- ment of important pole structures. We found that Swainson’s Hawks used 2X4 pole configurations disproportional to other configurations available. Management options for Swainson’s Hawks nesting on poles likely are benefited by maintenance and retention of the 2X4 configuration as a nest plat- form for this species where suitable natural sub- strates are limited or absent. To increase potential use of poles by birds, pole salvage operations in the southwest should strive to retain poles with two or more sets of paired cross- arms. These sets of cross-arms should be placed one above the other to provide a suitable nesting platform and shade structure for raptors and Chi- huahuan Ravens. 146 Brubaker et al. VoL. 37, No. 2 Acknowledgments This research was primarily funded by WSMR. Partial financial support was provided by the New Mexico Agri- cultural Experiment Station. Daisan Taylor facilitated ac- cess to WSMR and provided historical information re- garding pole salvage operations. Deedee Jefferson provided information regarding pole-line routes and construction details. Scott Lerich provided helpful infor- mation and historical perspectives regarding nesting rap- tors and ravens on WSMR. Dennis Eschrich provided data from surface atmospheric monitoring stations throughout WSMR. Barbara Nolen and James Wakeman provided GPS hardware, software, and helpful knowledge regarding global positioning satellites. David Garber and Ralph Campbell facilitated GIS operations to tabulated vegetation cover type data. Stuart Munson-McGee aided translation of GIS data to database files and spread- sheets. Jim Bednarz, Scott Lerich, and an anonymous ref- eree provided extensive assistance with manuscript re- view. Literature Cited Bechard, M.J., R.L. Knight, D.G. Smith, and R.E. Fitz- NER. 1990. Nest site and habitats of sympatric hawks {Buteo spp.) in Washington. J. Field Ornithol. 61:159- 170. Bednarz, J.C. 1988. A comparative study of the breeding ecology of Harris’ and Swainson’s Hawks in south- eastern New Mexico. Conrfor 90:311-323. AND S.W. Hoffman. 1988. The status of breeding Swainson’s Hawks in southeastern New Mexico. Pages 253-259 in R.L. Glinski, B.G. Pendleton, M.B. Moss, M.N. LeFranc, Jr., B.A. Millsap, and S.W. Hoffman [Eds.], Proceedings of the Southwest Raptor Manage- ment Symposium and Workshop. Natl. Wildl. Fed. Sci. Tech. Ser. No. 11. Washington, DC U.S.A. AND RJ. Raitt. 2002. Chihuahuan Raven {Corvus cryptoleucus) . In A.. Poole and F. Gill [Eds.], The birds of North America, No. 606. The birds of North Amer- ica, Inc., Philadelphia, PA U.S.A. Davis, C.A. and J.P. Griffing. 1977. Nesting of the White- necked Raven [Chihuahuan Raven] in southeastern New Mexico. Research Report, New Mexico State Uni- versity Agricultural Experiment Station, No. 231. Las Cruces, NM U.S.A. Dick-Peddie, W.A. 1993. New Mexico vegetation: past, present, and future. Univ. New Mexico Press, Albu- querque, NM U.S.A. Engiand, A.S., M.C. Bechard, and C.S. Houston. 1997. Swainson’s Hawk {Buteo swainsoni). In A. Poole and F. Gill [Eds.], The birds of North America, No. 265. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Griffing, J.P. 1974. Scissor-tailed Flycatchers and Swain- son’s Hawks nesting in the same tree. Southwest. Nat 19:111-112. Kimsey, B. and M.R. Conley. 1988. Habitat use by raptors in south-central New Mexico. Pages 197-203 in R.L. Glinski, B.G. Pendleton, M.B. Moss, M.N. LeFranc, Jr., B.A. Millsap, and S.W. Hoffman [Eds.], Proceedings of the Southwest Raptor Management Symposium and Workshop. Natl. Wildl. Fed. Sci. Tech. Ser. No. 11. Washington, DC U.S.A. Mosher, J.A. and C.M. White. 1976. Directional expo- sure of Golden Eagle nests. Can. Field-Nat. 90:356-459 Muldavin, E.M., PJ- Crist, B.C. Thompson, AJ. Peters, M. Eve, B. Middleton, J. Eggerton, and D.L. Gar- ber. 1996. A vegetation classihcation and land cover legend for application to New Mexico gap analysis. New Mexico Cooperative Fish and Wildlife Research Unit, Las Cruces, NM U.S.A. PiLZ, W.R. 1983. Nesting ecology and diet of Swainson’s Hawk in the Chihuahuan Desert, south-central New Mexico. M.S. thesis. New Mexico State Univ., Las Cru- ces, NM U.S.A. ScHMUTz, J.K., R.W. Fyfe, D.A. Moore, and A.R. Smith 1984. Artificial nests for Ferruginous and Swainson’s Hawks./. Wildl. Manage. 48:1009-1013. Smith, D.G. and J.R. Murphy. 1982. Nest site selection in raptor communities of the eastern Great Basin Desert. Great Basin Nat. 43:395-404. Steenhoe, K., M.N. Kochert, and J.A. Roppe. 1993. Nest- ing by raptors and Common Ravens on electrical transmission line towers./. Wildl. Manage. 57:271-281. Thompson, B.C., PJ. Crist, J.S. Prior-Magee, R.A. Deit- ner, D.L. Garber, and M.A. Hugfies. 1996. Gap anal- ysis of biological diversity conservation in New Mexico using geographic information systems. Research Gom- pletion Report. New Mexico Cooperative Fish and Wildlife Research Unit, New Mexico State Univ., Las Cruces, NM U.S.A. Received 30 December 2002; accepted 1 February 2003 Short Communications J. Raptor Res. 37(2):147-151 © 2003 The Raptor Research Foundation, Inc. Aerial Telemetry Accuracy in a Forested Landscape Travis L. DeVault^ Department of Forestry and Natural Resources, Purdue University, West Lafayette, IN 47907 U.S.A. Warren L. Stephens and Bradley D. Reinhart Savannah River Ecology Laboratory, Drawer E, Aiken, SC 29802 U.S.A. Olin E. Rhodes, Jr. Department of Forestry and Natural Resources, Purdue University, West Lafayette, IN 47907 U.S.A. I. Lehr Brisbin, Jr. Savannah River Ecology Laboratory, Drawer E, Aiken, SC 29802 U.S.A. Key Words: aerial telemetry, geographic information system] radiotelemetry, telemetry error. Radiotelemetry is one of the more widely used tools for investigations of animal movements, home range size, habitat use, and survival. Radiotelemetric estimations of animal locations, however, are not without error. Bearing errors associated with ground-based triangulated esti- mates of radio-tagged animals can be in excess of 20 de- grees (Hupp and Ratti 1983, Garrott et al. 1986), and depending on the aim of the study, excessive error may preclude meaningful analyses of data (White and Garrott 1990). Some researchers have chosen to ignore the issue of error in telemetry data (see Hupp and Ratti 1983, Saltz 1994, Withey et al. 2001), but ultimately the value of telemetry data is severely diminished when the mag- nitude of error is not investigated. Aerial tracking may be more desirable than ground- based methods in studies involving telemetry of wide- ranging birds and other highly-mobile animals, or in dense forests where signal “bounce” may limit the effec- tiveness of ground-based telemetry (Gilmer et al. 1981, Marzluff et al. 1994). Aerial tracking allows researchers to avoid much of the potential error associated with ground-based tracking because fewer obstructions lie be- tween the receiver and transmitter. However, aerial telem- etry does not always provide “line of sight” radio fixes, especially in heavily forested landscapes. Although many studies have elucidated environmental factors contribut- ing to bearing error from ground-based telemetry, no studies have documented how environmental factors in- fluence conventional aerial telemetry accuracy. ^ E-mail address: tdevault@fnr.purdue.edu In this study, we evaluated the accuracy, precision, and sources of error for an aerial telemetry protocol designed for investigations of home range size and movement pat- terns of raptors and other wide-ranging birds at a heavily forested site in South Carolina, U.S.A. At this site, ground-based telemetry for wide-ranging animals is im- practical due to substantial forest cover and extensive roadless areas. Methods Study Area. This study was conducted at the 78 000 ha Savannah River Site (SRS), near Aiken, South Carolina. The SRS is a nuclear facility owned and operated by the United States Department of Energy, and is designated as a National Environmental Research Park. Approxi- mately 64% of the SRS has been planted in loblolly pine {Pinus taeda), longleaf pine {P. palustris), and slash pine {P. elliottii. Workman and McLeod 1990), which are man- aged for timber production by the United States Forest Service. An additional 15% of the land cover is classified as bottomland hardwood (Workman and McLeod 1990) . Although most of the SRS is forested, there are several industrial areas located throughout the site. Overhead powerlines of various sizes are present near industrial ar- eas and along major roads. Elevation at the SRS ranges from 30 masl or less on the southwestern portion of the site near the Savannah River to 115 m on the northern portion of the site (White and Gaines 2000). Field Procedures. Transmitters (Holohil Systems Ltd , Model AJ-2B, 164—165 MHz) were attached to small trees or wooden stakes 1 m from ground level at 25 locations throughout the SRS. We placed transmitters (hereafter, beacons) arbitrarily in locations that varied with respect to habitat type and distance from overhead powerlines We considered four habitat types, which were deter- mined by visual inspection and by referencing a digital habitat map of the SRS (Wiggins-Brown et al. 2000): tree- 147 148 Short Communications VoL. 37, No. 2 less habitat {N = 3 beacons), open-canopy pines (12), closed-canopy pines (4), and deciduous hardwoods (6). The number of beacons placed in each habitat type roughly corresponded to the relative proportion of each habitat type on the SRS. Distances between beacons and the nearest overhead powerline ranged from 26-5477 m (x = 1438; SD = 1620). Beacon locations were recorded in Universal Transverse Mercator (UTM) coordinates with a Trimble Pro XR Global Positioning System (GPS) unit (sub-meter accuracy). We used a Cessna 172 airplane equipped with two-el- ement Yagi antennas attached to each wing strut to esti- mate beacon locations (Gilmer et al. 1981). Beacon lo- cations were estimated during three, 2-hr flights during the spring of 2001 after peak leaf emergence (10, 20, and 24 April). The same pilot, observer, and aircraft per- formed all flights. We generally flew 60-180 m from ground level at an air speed of 140—175 km/hr. We esti- mated beacon locations with a Garmin 12 CX handheld GPS unit (10-30 m accuracy; see Marzluff et al. 1994 for further description of aerial telemetry methods) . We es- timated beacon locations in random order. The pilot and observer were blind to beacon locations; they were given only a list of radio frequencies and had no prior knowl- edge of their placement. We generally followed steps out- lined in Samuel and Fuller (1996) for aerial telemetry: once a signal was detected, the observer indicated the general direction of that signal to the pilot. An attempt then was made to keep the signal on one side of the airplane while circling the signal source. A visual estimate of the beacon location was made after circling the signal source several times and monitoring signal strength. We then flew directly over the estimated location and re- corded the UTM coordinates with the GPS unit. Data Analyses. UTM coordinates of actual and esti- mated beacon locations were imported into a Geograph- ic Information System (GIS; ArcView 3.2; Environmental Systems Research Institute, Inc., Redlands, CA) as point themes (Fig. 1). The GIS was used to determine the dis- tances from actual beacon locations to estimated loca- tions, and the distances from actual beacon locations to the nearest overhead powerline. It was not necessary to correct linear distances with respect to topography due to the relative uniformity of the terrain at the scale of the measured distances. We examined the effects of habitat at the beacon lo- cation and proximity of the nearest overhead powerline on linear accuracy of location estimates. Linear accuracy was defined as the distance between the actual beacon locations and the locations estimated with the handheld GPS from the airplane. We used one-way analysis of var- iance to test whether the habitat within which the beacon was located influenced the linear error of estimates. The LSD procedure (SPSS 1999) was used for pair-wise post hoc comparisons between variables. We used simple linear regression to examine the influence of the proximity of the nearest overhead powerline to beacon locations on linear error. Confidence areas (95%) were computed by centering a circle with a radius of (1.96) X (SD of the linear estimate of error) at beacon location estimates (White and Garrott 1990). All analyses were performed using SPSS 10.0 (SPSS 1999). Results We located 24 of the 25 beacons (96%) from the air. Location of one beacon was confounded by interference on its frequency from other sources on the SRS; this bea- con was excluded from the analyses. Mean linear distance error for the 24 location estimates was 191 m (range = 22-1011 m, SD = 197 m; Table 1). The 95% confidence circles were 47 ha in size (Fig. 1). Twenty-two of 24 (92%) of the actual beacon locations fell within the confidence circles. One-way analysis of variance showed that linear error differed among habitat types {P — 0.02). Post hoc comparisons indicated linear error differences between beacons placed in deciduous hardwoods (x error = 401 m) and each of the other three habitat types (open pine, 117 m; dense pine, 130 m; treeless habitat, 124 m). The proximity of overhead powerlines to beacons had no dis- cernable influence on linear error {W- = 0.05, P= 0.32). Discussion According to Samuel and Fuller (1996), “. . . few data are available from tests of the precision and accuracy of aerial radio-tracking, but ±100-200 m is probably the best commonly achieved accuracy.” Marzluff et al. (1994) reported a mean linear error of 409 m, with an associated 95% confidence circle of 112 ha around each point es- timate. Carrel et al. (1997) used several location-record- ing methods to achieve linear errors of 73-386 m at un- known beacon locations. Hoskinson (1976) reported a linear error of 7-70 m with two different pilots, but he flew exceedingly slowly (95-115 km/hr), dangerously low to the ground (15-30 m from ground level), and circled each beacon for at least 5 min. Hoskinson’s (1976) re- sults probably reflect the minimum error possible for ae- rial telemetry with conventional equipment, but his methods were unrealistic for most research situations. In all radio-tracking studies, the required level of ac- curacy depends on the study objectives. Our mean linear error of 191 m and the associated 95% confidence circles of 47 ha are probably sufficient for investigations of home range and local movement patterns of wide-rang- ing species, although this level of accuracy would be in- adequate for many fine-scale analyses. For example, Cole- man and Fraser (1989) reported annual home range sizes of 14 881 ha for Black Vultures {Coragyps atratus) and 37 072 ha for Turkey Vultures {Cathartes aura) in Pennsylvania and Maryland. Home range estimates for Ferruginous Hawks {Buteo regalis) have exceeded 9000 ha in south-central Washington (Leary et al. 1998). Our er- ror rates would be relatively inconsequential when ex- amining home range size and broad-scale movements of such birds. However, depending on habitat heterogene- ity, an accurate description of microhabitat use at specific time intervals probably would not be possible. Whereas factors such as pilot skill and attitude (Hos- kinson 1976), wind speed (Hoskinson 1976, T. DeVault pers. observ.), and airplane altitude and speed (Caughley 1974) are known to influence aerial telemetry accuracy, June 2003 Short Communications 149 20 Kilometers SRS Habitat Types Treeless Habitat Open-canopy Pines Dense-canopy Pines Deciduous Hardwoods Figure 1. Map of the Savannah River Site, South Carolina, depicting major habitat types, location of overhead powerlines (thin lines) and beacon locations (solid circles). The circles representing beacon locations also represent the relative size of the confidence circles (47 ha). This figure was modified from a 1999 digital habitat map (30 m resolution) with 33 original habitat types (Wiggins-Brown et al. 2000) . 150 Short Communications VoL. 37, No. 2 Table 1. Mean linear distance error and coverage by major habitat type of beacon locations estimated with aerial telemetry. Habitat Type (N) Mean Linear Distance Error SE Coverage^ Treeless habitat (3) 124 54.3 100 Open-canopy pines (11) 117 15.1 100 Closed-canopy pines (4) 130 26.2 100 Deciduous hardwoods (6) 401 127.9 67 Total (24) 191 40.3 92 ^ Percentage of actual beacon locations that fell within confidence circles. our data indicate that the forest type within which a trans- mitter is located also can influence accuracy, and should be considered when assessing aerial telemetry accuracy. Rempel et al. (1995) and Rumble and Lindzey (1997) demonstrated that forest type and tree density affected GPS collar observation rate and accuracy, so it follows that signal bounce from vegetation could influence ac- curacy of conventional transmitters detected by aerial te- lemetry as well. Conversely, our results do not support evidence that the presence of overhead powerlines de- creases telemetry accuracy (Withey et al. 2001). However, due to low sample sizes, we were unable to examine the influence of powerlines on mean linear error within each habitat type, even though we found differences in mean linear error across habitat types. Thus, this analysis may be somewhat unreliable. Marzluff et al. (1994) expressed concern about the ability of aerial telemetry to accurately represent move- ments and home ranges of Prairie Falcons {Falco mexican- us) because the researchers often could not obtain radio fixes until the birds were perched, and often already back at their nests. Pilot studies at the SRS have shown that this problem does not exist to a large extent when track- ing soaring birds, like buteo hawks and vultures (T. DeVault unpubl. data). Such birds move throughout their home ranges much more slowly than falcons, facil- itating rapid and unbiased location estimates. However, we tested only stationary beacons, thus our results may not be applicable to moving instrumented birds. Fur- thermore, it should be noted that the inferential value of our data may be limited because we did not place beacons randomly throughout the study site. Our study suggests that aerial telemetry is an effective method for some radio-tracking applications in areas where ground-based methods are not feasible due to ex- tensive forest cover and the need for large distances be- tween receiver and transmitter (i.e., when tracking wide- ranging birds). Although aerial telemetry does not provide completely unobstructed paths from the trans- mitter to the receiver, it appears that the obstructions do not influence the quality of the radio signal to a large extent, even in heavily forested landscapes. Resumen. — Evaluamos la precision y las fuentes de error de la telemetria aerea en la zona boscosa del Rio Savan- nah en Carolina del Sur, U.S.A. Los radiotransmisores fueron ubicados en 25 localidades en un sitio de 78.000 ha. Usamos un aeroplano/avion Cessna 172, equipado con antenas duales para localizar los transmisores. El er- ror medio lineal para telemetria aerea fue de 191 m, y el circulo de confianza del 95%, fue de 47 ha. Veintidos de 24 (92%) de las localidades actuales de los transmi- sores estuvieron dentro del circulo de confianza. El tipo de habitat influencio la precision aerea: el error lineal entre las localidades de transmisores actual y estimada, fue mayor para los transmisores ubicados en bosques de- ciduos y en otros tipos de habitats. La proximidad de lineas de energia no tuvo un efecto significativo en la precision de la telemetria aerea. Especialraente en habi- tats de bosques densos, la telemetria aerea provee una alternativa precisa y practica a la telemetria en tierra. [Traduccion de Cesar Marquez] Acknowledgments We thank Michael Ross from Augusta Aviation for his expert piloting. Robert A. Garrott, William L. Thompson, John M. Marzluff, John C. Withey, James M. Novak, Rob- ert T. Stoudt, and Julie L. Weston offered advice on var- ious drafts of this manuscript, and Joey E. Counts provid- ed logistical support. This project was supported in part by contract DABT63-96-D-0006 between Purdue Univer- sity and the United States Air Force Bird/Wildlife Air- craft Strike Hazard Team. We also acknowledge the sup- port of the Purdue University Department of Forestry and Natural Resources, and the United States Depart- ment of Energy through contract DE-FC09-96SR18546 with the University of Georgia’s Savannah River Ecology Laboratory. Literature Cited Carrel, W.K., R.A. Ockenfels, J.A. Wennerlund, and J.C. DeVos, Jr. 1997. Topographic mapping, LORAN- C, and GPS accuracy for aerial telemetry locations. J. Wildl. Manage. 61:1406—1412. Caughley, G. 1974. Bias in aerial survey./ Wildl. Manage. 38:921-933. June 2003 Short Communications 151 Coleman, J.S. and J.D. Fraser. 1989. Habitat use and home ranges of Black and Turkey Vultures. J. Wildl. Manage. 53:782-792. Garrott, R.A., G.C. White, R.M. Bartmann, and D.L. Weybright. 1986. Reflected signal bias in bioteleme- try triangulation systems./. Wildl. Manage. 50:747—752. Gilmer, D.S., L.M. Cowardin, R.L. Duval, L.M. Mech- lin, C.W. Schaiffer, and V.B. Kuechle. 1981. Proce- dures for the use of aircraft in wildlife biotelemetry studies. U.S. Dept. Inter. Resource Publ. 140. Wash- ington, DC U.S. A. Hoskinson, R.L. 1976. The effect of different pilots on aerial telemetry error./. Wildl. Manage. 40:137-139. Hupp, J.W. andJ.T. Ratti. 1983. A test of radio telemetry triangulation accuracy in heterogeneous environ- ments. Proc. Int. Wildl. Biotelemetry Conf. 4:31-46. Leary, A.W., R. Mazaika, and M.J. Bechard. 1998. Factors affecting the size of Ferruginous Hawk home ranges. Wilson Bull. 110:198-205. Marzluff, J.M., M.S. Vekasy, and C. Coody. 1994. Com- parative accuracy of aerial and ground telemetry lo- cations of foraging raptors. Cowiior 96:447-454. Rempel, R.S., A.R. Rodgers, and K.F. Abraham. 1995. Performance of a GPS animal locations system under boreal forest canopy./. Wildl. Manage. 59:543-551. Rumble, M.A. and F. Lindzey. 1997. Effects of forest veg- etation and topography on Global Positioning System collars for elk. Resour. Tech. Inst. Symp. 4:492-501. Saitz, D. 1994. Reporting error measures in radio loca- tion by triangulation: a review. /. Wildl. Manage. 58* 181-184. Samuel, M.D. and M.R. Fui.i.er. 1996. Wildlife radiote- lemetry. Pages 370—418 in T.A. Bookhout [Ed.], Re- search and management techniques for wildlife and habitats, 5th Ed. The Wildlife Society, Bethesda, MD U.S.A. SPSS. 1999. SPSS base 10.0 applications guide. SPSS, Inc , Chicago, IL U.S.A. White, D.L. and K.F. Gaines. 2000. The Savannah River Site: site description, land use, and management his- tory. Stud. Avian Biol. 21:8-17. White, G.C. and R.A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press, San Diego, CA U.S.A. Wiggins-Brown, H., J.E. Binder, III, K.K. Guy, and T.E. Rea. 2000. 1999 Habitat map of the Savannah River Site. Savannah River Ecology Laboratory, Aiken, SC U.S.A. Withey, J.C., T.D. Bloxton, and J.M. Marzluff. 2001. Effects of tagging and location error in wildlife radio- telemetry studies. Pages 43-75 m J.J. Millspaugh and J.M. Marzluff [Eds.], Radio tracking and animal pop- ulations. Academic Press, San Diego, CA U.S.A. Workman, S.K. and K.W. McLeod. 1990. Vegetation of the Savannah River Site: major community types SRO-NERP-19. Savannah River Ecology Laboratory, Aiken, SC U.S.A. Received: 22 July 2002; accepted 19 February 2003 /. Raptor Res. 37(2):151-157 © 2003 The Raptor Research Foundation, Inc. Perch-site Selection and Spatial Use by Cactus Ferruginous Pygmy-Owls in South-central Arizona Aaron D. Flesch^ Wildlife and Fisheries Science Program, School of Renewable Natural Resources, 325 Biological Sciences East, University of Arizona, Tucson, AZ 85721 U.S.A. Key Words: Ferruginous Pygmy-Owl] Glaucidium brasi- lianum; cactorum; Altar Valley, Arizona] endangered species] habitat selection] perch site. Cactus Ferruginous Pygmy-Owls {Glaucidium brasilian- um cactorum) are federally endangered in Arizona and therefore of signihcant conservation and management interest (U.S. Fish and Wildlife Service 1997). Concern for pygmy-owls has resulted in major efforts in conser- ^ E-mail address: flesch@ag.arizona.edu vation planning including a focal role in the Sonoran Desert Conservation Plan, proposed designation of criti- cal habitat, and recent release of a Draft Recovery Plan by the U.S. Fish and Wildlife Service (USFWS 2002, 2003). Descriptions of areas occupied by pygmy-owls in Arizona are limited to anecdotal accounts from the late 1800s and early to mid 1900s (e.g., Fisher 1893, Brenin- ger 1898, Gilman 1909, Phillips et al. 1964), a recent study by Richardson (2000), and unpublished reports. No published information exists on characteristics and size of areas used by pygmy-owls in semidesert grasslands in Arizona. 152 Short Communications VoL. 37, No. 2 Information on habitat selection and composition of areas used by pygmy-owls is important for recovery ef- forts, guiding development and land-use activities, and prioritizing prospective conservation reserves. My objec- tives were to compare conditions at and around calling perches used by pygmy-owls with the surrounding envi- ronment, and describe the size, shape, and composition of occupied areas. Methods Study Area. The Altar Valley (centered at 31°40'N, 111°20'W) is southwest of Tucson, Arizona and bordered by Mexico to the south, State Highway 86 to the north, and eight mountain ranges to the east and west. Vegeta- tion consists primarily of semidesert grassland ranging from open savannah to shrub-invaded thornscrub and woodland (Brown 1982). Woodlands of mesquite {Proso- pis velutina) and catclaw acacia {Acacia greggii) are com- mon along drainages. Upland vegetation consists of semi- desert grassland, desertscrub, thornscrub, oak {Quercus sp.) woodland, and oak savannah. Broadleaf trees other than netleaf hackberry ( Cdtis reticulata) and saguaro cacti (Carnegiea gigantea) are rare in the southern and central portions of the valley. The study area included the Buen- os Aires National Wildlife Refuge (NWR), Arizona State Trust land, and private land. Site Location and Home Range Delineation. Between March and May 1999, 1 surveyed ca. 8300 ha in the south- ern and central Altar Valley using recorded, conspecific territorial calls to elicit responses from pygmy-owls. I doc- umented seven areas and four nests occupied by pygmy- owls. 1 revisited four of these areas between April and August 1999 to locate perch sites, assess size and shape of home ranges, and measure features at and around substrates used for calling and randomly-selected sub- strates. I visited occupied areas during early mornings and eve- nings when pygmy-owls are most active and vocal (Proud- foot and Johnson 2000). During each visit, I marked the perch substrate where I initially detected a pygmy-owl and recorded time, date, sex, and activity. I determined sexes by the type of vocalization, duration of calling, and behavior of owls (Proudfoot and Johnson 2000). During visits when owls were exceptionally active (moved every 1-10 min and used >5 perches), 1 also marked an ad- ditional substrate used 30 min after owls left sites where 1 detected them initially. 1 did not solicit responses or flush owls during visits. I mapped perch sites as points on topographic maps and determined their coordinates with a global positioning system (GPS) receiver. I calcu- lated minimum convex polygons (MCP) that included all perch sites to estimate the size and shape of home ranges (Mohr and Stumpf 1966). 1 generated and visited ran- dom coordinates within each MCP and marked the near- est potential perch substrate (woody plant or saguaro cav- ity >2 m tall). Although three of four males were banded, my observations suggested that only one male occupied each home range. Vegetation Measurements. I measured vegetation fea- tures around perch substrates used for calling and ran- dom substrates at three spatial scales: perch substrate, microhabitat, and mesohabitat. I recorded species of each substrate, measured height with a clinometer or measuring pole, and basal diameter with a tape. For co- lonial shrubs with multiple stems, I measured basal di- ameter of each cluster of stems including open space. I measured canopy diameter of each substrate by averag- ing the widest horizontal canopy dimension with a per- pendicular measurement across the canopy. I measured distance from the base of each substrate to closest drain- age and classified the vegetation community around sub- strates as woodland (xeroriparian vegetation along drain- age), savannah (scattered trees), grassland (open with occasional tree) , or desertscrub (in rocky uplands) . To quantify vegetation at the microhabitat scale, I mea- sured vegetation within 0.003-ha circular plots (3-m ra- dius) centered on perch and random substrates. I re- corded distance to, and height of, nearest woody or succulent plants >3 m tall in four, 90° quarters denoted by the cardinal directions (Cottam and Curtis 1956). Within each plot, I recorded species, height, and basal diameter of all woody trees, shrubs, and succulents >2 m tall rooted within plots and listed other woody species present. I used a vertical line-intercept method (Mills et al. 1991) to measure vegetation cover and volume. Six, 3-m transects, based on a random bearing and radiating 60° apart were used per plot. I placed a 17-mm diameter pole at 1-m intervals along transects and recorded deci- meter intervals with vegetation (any rooted plant mate- rial) within 5 cm of the pole. I also recorded ground cover (litter, bare ground, grass, forb, or rock) and all woody or succulent species present within four vegetation strata (0-0.5 m groundcover, 0.51-2 m midstory, 2.01-5 m low canopy, 5+ m high canopy). I used these data to calculate total vegetation volume, vegetation volume with- in 1-m strata, percent vegetation cover in four strata, and ground cover. To quantify vegetation at the mesohabitat .scale, I mea- sured vegetation within 0.07-ha circular plots (15-m ra- dius) around four to five randomly-selected perch sub- strates per home range. I only described vegetation in desertscrub and woodland communities because pygmy- owls rarely used grassland or savannah. Methods were similar to those described for 0.003-ha plots, but arrange- ment and quantity of vertical line-intercept samples dif- fered. 1 placed eight 15-m transects 45° apart and placed the pole 2.4, 8.9, 12.1, and 14.6 m from center points. Line-intercept points were spaced unevenly to reduce over-sampling plot centers. All other measurements fol- lowed 0.003-ha plot protocols. Analyses. I used all perch sites and a GPS receiver to measure area, perimeter, and maximum length of home ranges. I compared used and available resources to mea- sure habitat selection by pygmy-owls (Manly et al. 1995). I used nonparametric procedures (Kruskal-Wallis) to test for differences in vegetation conditions (Daniel 1978). To test for selection of common perch species and veg- etation communities, I used Pearson’s Chi-square good- ness-of-fit tests (Sokal and Rohlf 1995). I used stepwise (P < 0.25 to enter, P < 0.10 to remain) discriminant function analysis (DFA) to determine which combination of 29 variables best discriminated between used and ran- dom sites. I then used canonical DFA on selected vari- ables and canonical scores (CS) to determine which var- iables accounted for most discriminatory power. I tested June 2003 Short Communications 153 Table 1. Spatial metrics of four cactus Ferruginous Pygmy-Owl home ranges, Altar Valley, Arizona, 1999. Variable Male 1 Male 2 Male 3 Pair Mean SE No. of locations 24 27 15 25 22.8 2.7 Area (ha) 9.9 11.7 47.3 18.5 21.9 8.7 Perimeter (m) Maximum distance between 1743 1830 3193 2832 2400 391 points (m) 783 803 1359 1347 1073 162 the model’s explanatory ability with multivariate analysis of variance (MANOVA) . When necessary, variables were transformed using log(x) or log(x + 1) to better meet assumptions of parametric procedures (Sokal and Rohlf 1995). I performed univariate comparisons separately for each individual and multivariate analysis for all individ- uals combined. Results Between 10 April and 15 August 1999, 1 made 83 visits to four home ranges, three occupied by unpaired males and one by a nesting pair. Unpaired males vocalized on 96, 71, and 67% of visits and the paired male on 56% of visits. Home Ranges. A total of 24, 27, 15, and 25 perch sites were located within each home range. Home ranges av- eraged 21.9 ha in area (SE = 8.7) and ranged from 9.9- 47.3 ha (Table 1). Three home ranges were distributed along large drainage segments where maximum distance between sites ranged from 783-1347 m (x = 978, SE = 185). A fourth home range comprised four smaller drain- ages 160-375 m apart. The female was always observed within 950 m of the nest, but remained within ca. 150 m throughout incubation. All home ranges were occupied throughout the study period except for the largest, where the male was last detected on 11 June. Three home ranges included one or more linear stands of xeroriparian woodland dominated by mesquite and catclaw acacia, mesquite-savannah on flats above woodlands, and desertscrub restricted to rocky uplands on one side. A fourth home range comprised semi-mesic riparian woodland dominated by ash {Fraxinus velutina), mesquite, catclaw acacia, nedeaf hackberry, and red bar- berry {Berberis haematocarpa) . Two home ranges included a dry water catchment or corral. The one nest was locat- ed in an ash tree along the main drainage. Microhabitat Selection. I located 27, 18, and 13 perch substrates used for calling and 31, 29, and 30 random substrates within home ranges occupied by single males. I did not measure sites for the paired male due to a lim- ited number of calling perches (N = 9). Pygmy-owls did not use vegetation communities in proportion to avail- ability within all home ranges (x^ — 7.22, 0.027) (Eig. 1). Woodlands were used in greater proportion (x^ — 4.01, P ^ 0.045), whereas savannah was used less than expected (x^ ^ 6.91, P 0.009). Desertscrub was used in proportion to availability (x® ^ 2.36, P S: 0.124). Perch substrates were taller and had larger basal and canopy diameters than random substrates within all home ranges (x^ — 8.55, P < 0.003; Table 2). Use of perch species was disproportionate to availability in one home range (x^ = 11.36, P = 0.078), but not in the other two (x^ — 8.16, P s 0.226). Pygmy-owls used mesquite, saguaro, blue paloverde (Cercidium florida), ocotillo {Fou- quieria splendens), catclaw, and whitethorn acacia {Acacia constricta) for calling perches. Mesquite was used for 70— 92% of perches overall and 93-100% of perches in wood- lands, but mesquite use reflected availability (x^ — 1.97, P > 0.160). Male pygmy-owls called from inside saguaro cavities during six visits (7.8%) and were observed roost- ing in these same cavities during day and night. Total vegetation volume and vegetation volume >2 m above ground were greater around perch substrates with- in all home ranges (x^ — 4.44, P < 0.035). Vegetation volume 0—1 m above ground averaged 18.7% greater around random sites in two home ranges (x^ — 3.60, P < 0.058), but only 9.5% greater in the other (x^ = 1.42, P = 0.23). Vegetation volume 1-2 m above ground did not differ from random sites (x^ — 1.39, P & 0.24). Mi- crohabitat around perch substrates had greater plant spe- cies richness, and higher density and basal area of plants ^2 m tall (x^ ^ 3.38, P ^ 0.066); these patterns were mostly significant. Ground cover around perch substrates tended to have more litter and less bare ground than random sites (Table 2). Perch substrates were 2.9 times closer to drainages {x = 29.2 m, SE = 10.5) in one home range, where cover of upland desertscrub was limited. Within other home ranges, perch substrates outside of desertscrub averaged only 6.3 (SE = 1.1) and 6.4 m (SE = 1.5) from drainages. All perch substrates were within 160 m of a drainage. Height, basal and canopy diameters, mean distance to nearest neighbor plants, distance to drainage, vegetation volume 1-2 m and >6 m above ground, density of trees and shrubs, and ground cover of grass and bare ground discriminated calling perches from random sites (Wilk’s Lambda — 0.346; df = 10, 137; P < 0.0001). Factors that accounted for most variation in the model included perch substrate height (CS = 2.09), perch basal diameter (CS = 1.12), and density of trees and shrubs (CS = 0.64) Mesohabitat. Occupied woodlands {N =11 plots) had an open to semi-open canopy of mesquite, catclaw acacia, and, occasionally, blue paloverde; moderate cover 2-5 m 154 Short Communications VoL. 37, No. 2 80 1 60 J 40 20 0 - ■ Use □ Available 80 n Woodland Desertscrub Savannah Figure 1, Use and availability of vegetation communities around perch substrates used for calling within three cac- tus Ferruginous Pygmy-Owl home ranges, Altar Valley, Ar- izona, 1999. “-I-” indicates use > expected {P < 0.05), “ — ” indicates use < expected, and no sign indicates no difference. above ground (x = 49.9%, SE = 3.7, range = 30-64), and low cover above 5 m (x = 7.9%, SE = 3.2, range = 0-27). Woodlands had moderate cover 0.5-2 m above ground (x = 59.3%, SE = 2.2, range = 48-76) composed mainly of desert hackberry {Celtis pallida) , wolfberry {Ly- cium sp.), catclaw acacia, mesquite, graythorn {Ziziphus obtusifolia) , and cholla {Opuntia sp.). Desert hackberry and wolfberry often formed scattered patches of dense midstory vegetation under larger trees (Table 3). Cover 0-0.5 m above ground was moderately high (x = 71.4%, SE = 2-2, range = 61-81) and often composed of grass. Ground cover was composed mainly of bare ground (x = 54.0%, SE = 3.6), litter (x = 31.3%, SE = 3.1), and grass (x = 9.8%, SE = 3.1). Upland desertscrub {N = 3 plots) was dominated by ocotillo and blue paloverde (Table 3). Adult saguaros were present in or immediately around all plots and har- bored cavities. Desertscrub plots had high cover below 0.5 m above ground (x = 88.0%, SE = 1.7, range = 85— 91), moderate cover 0.5-2 m (x = 49.7%, SE = 6.9, range = 36-58), and low cover above 2 m (x = 10.0%, SE = 2.0, range = 0-12). Ground cover was composed mainly of rock (x = 66.0%, SE = 12.7), bare ground (x = 15.7%, SE = 4.7), and litter (x = 13.3%, SE = 5.9). Discussion Spatial use by these four pygmy-owls conformed to the arrangement of landscape and vegetation features within home ranges. Woodland size and shape appeared to cor- respond with home range boundaries and was less de- veloped outside of home ranges on both upstream and downstream sides. The largest home range may have been underestimated because the male was not detected after 11 June. In Texas, area used (based on MCP) by nine paired males from one week before to one week after incubation ranged from 1.3-23.1 ha, whereas an un- mated male used 110 ha during the same period (Proud- foot and Johnson 2000). Additionally, five families (adults and three fledglings/family) used from 9. ,3-59. 5 ha be- tween fledging and dispersal. In Arizona, preliminary es- timates of three home ranges used during the 1998 breeding season (based on MCP) were 8.1, 14.2, and 89.0 ha (Arizona Game and Fish Department unpubl. data). Perch substrates used for calling were generally the largest trees available. Although heights of calling loca- tions within substrates were not measured, observations indicated that owls often called from the upper third of substrates. Calling from near the tops of large trees is likely more audible than when closer to the ground. Se- lection of elevated calling perches has been documented for other bird species (Knopf et al. 1990) and likely pro- motes advertisement to females and aids in territorial maintenance and defense. Calling from inside saguaro cavities may aid in advertisement of potential nest cavities to females (Proudfoot and Johnson 2000). Saguaros and upland desertscrub are rare in the south- ern and central Altar Valley, but presence of these types within three home ranges augmented diversity of vege- tation and habitat features. Cavities of sufficient size for nesting were rare in woodland trees except where large broadleaf species were present. Cavities created by Gila Woodpeckers {Melanerpes uropygialis) and Gilded Flickers {Colaptes chrysoides) were rare except in saguaros. Pres- ence of large columnar cacti also appears to be a key June 2003 Short Communications 155 Table 2. Means and standard errors for habitat variables at and around calling perches (0.003 ha plots) of cactus Ferruginous Pygmy-Owls {N = 3) and random sites within home ranges, Altar Valley, Arizona, 1999. Means and standard errors based on V = 27, 18, and 13 used, and V = 31, 30, and 29 available plots. Used Available ScAi.E Variable Mean SE Mean SE P < 0.05^ Perch substrate Height (m) 5.6 0.6 3.2 0.1 3 Basal diameter (cm) 17.8 0.7 8.8 0.6 3 Canopy diameter (m) 6.8 0.3 4.0 0.2 3 Distance to drainage (m) 31.8 2.2 47.1 19.3 0 Microhabitat Ground cover bare (%) 22.6 0.9 39.7 5.6 3 Litter 57.4 7.9 36.3 1.1 2 Grass 5.6 2.7 11.7 2.8 0 Total veg. volume (hits) 245.6 17.1 160.9 10.3 3 Veg. volume 0-1 m 61.1 4.1 71.9 1.2 1 1-2 m 50.6 1.2 45.7 2.8 0 2-3 m 57.9 5.6 26.3 1.5 3 3-4 m 37.8 3.5 9.4 2.2 3 4—5 m 20.9 4.5 5.6 2.6 3 5-6 m 10.9 5.8 2.4 1.3 3 >6 m 6.6 6.3 0.7 0.6 1 Density’’ (no.) 3.4 0.4 1.3 0.2 3 Height’’ (m) 2.5 0.1 1.6 0.1 2 Basal diameter’’ (cm) 19.4 3.4 7.9 1.4 2 Plant sp. richness (no.) 6.6 0.2 4.8 0.7 3 Nearest neighbor height (m) 4.1 0.1 3.8 0.1 1 Distance (m) 7.0 1.3 11.1 1.2 2 Number of three possible comparisons within home ranges where P < 0.05. Plants >2 m tall. Table 3. Density and height of vegetation in 0.07-ha plots centered on perch substrates in three cactus Ferruginous Pygmy-Owl home ranges, Altar Valley, Arizona, 1999. Species Woodland {N = 11) Desertscrub {N = 3) Densitv® Height (m) Density^ Height (m) Mean SE Mean SE Mean SE Mean SE Mesquite 12.5 1.9 3.9 0.2 1.0 0.6 2.6 0.4 Catclaw acacia 16.4 2.5 2.8 0.1 Blue paloverde 2.3 0.5 3.8 0.4 4.3 2.0 2.6 0.1 Wolfberry 3.1 0.8 2.4 0.1 3.0 1.2 2.2 0.1 Desert hackberry 10.2 3.0 3.2 0.2 Graythorn 1.2 0.4 2.6 0.2 Whitethorn acacia 0.8 0.4 2.8 0.4 1.7 1.7 2.3 0.1 Saguaro 2.3 1.5 6.5 0.5 Cholla 1.9 0.5 2.7 0.2 0.3 0.3 2.0 2.0 Ocotillo 27.0 2.5 3.0 0.1 All species 47.9 6.0 3.2 0.1 40.0 2.5 3.1 0.1 Mean number of individuals within 0.07-ha plots. 156 Short Communications VoL. 37, No. 2 factor that influences pygmy-owl distribution in neigh- boring Sonora, Mexico (Flesch 2003). Vegetation cover in occupied woodlands was moderate and patchy, and vegetation volume near the ground was often moderate around perch substrates. This structure provided good horizontal and vertical visibility that may be important for the perch-and-pounce or sit-and-wait hunting strategies of pygmy-owls. Woodlands along drain- ages provided cover for hunting, roosting, and escape, whereas saguaros and desertscrub vegetation in uplands contributed potential nest and roost cavities and cover. Although this study focused on selection of features at and around calling perches, I observed pygmy-owls using many of these same patches of vegetation for hunting and roosting. Therefore, I suspect these same vegetation patches are important for other aspects of pygmy-owl life history. Management of pygmy-owls in the areas studied should stress retention of large trees, structurally diverse patches of woodland and desertscrub vegetation, and po- tential cavity-harboring substrates. Resumen. — Se describen el tamaho y composicion de cuatro areas ocupadas por machos (N = 3) y una pareja de la especie en peligro, buho pigmeo de los cactaceas ( Glaucidium brasilianum cactorum) en el valle de Altar en el sureste de Arizona durante 1999. En estas areas, se compararon los arboles, sitios, y areas de perchado 11a- mado {N = 27, 18, y 13) contra sitios al azar. Las areas usadas (poligono minimo convexo) midieron de 9.9-47.3 ha y incluyeron las comunidades de bosque, matorral de- sierto, sabana, y pastizales. El uso de la comunidad bos- cosa, excedieron disponibilidad, y en las comunidades de sabana y pastizal el uso fue menor a lo disponible. Los arboles de percha y llamado presentaron mayor altura {P < 0.001), diametro basal {P< 0.001), y diametro de copa {P = 0.003), que los substratos disponibles. Los sitios de perchado (0.003 ha) fueron mas altos en la riqueza de plantas, volumen total de vegetacion y el volumen de la vegetacion >2 m sobre el suelo que los sitios al azar (P ^ 0.035). La altura, el diametro basal, la distancia de las plantas vecinas mas cercanas S3 m de altura, y la distan- cia del perchado al cauce del drenaje mas cercanas, dis- tmguen mejor los sitios de perchado llamado de los sitios aleatorios. La retencion de arboles grandes, parches bos- cosos y matorral desierto estructuralmente diversa, y los substratos con potencial para contener cavidad pueden ayudar en los esfuerzos de recuperacion y manejo para esta especie en Arizona. [Traduccion de Gabriel Valencia Ortega] ACKNO WITD GMENTS This study was supported by USFWS Cooperative Agreement No. 1448-00002-99-G943 and Arizona State Land Department right of entry permit No. 29-104780- 000. I thank F. Baucom, M. Wrigley, and T. Gatz (USFWS, Arizona Ecological Services Field Office) for supporting this work. W. Shifflett (Buenos Aires NWR) provided es- sential logistical support. D. Abbate and R. Wilcox (Ari- zona Game and Fish) and G. Proudfoot (Texas A&M) banded owls and assisted in locating perch sites. T. Wood and A. Smith assisted with vegetation measurements. R. Steidl assisted with multivariate analyses. I also thank R Steidl, T. Brush, B. Marcot, and B. Spears for suggestions on earlier versions of the manuscript. Literature Cited Breninger, G.F. 1898. The Ferruginous Pygmy-Owl. Os- prey 2:128. Brown, D.E. 1982. Biotic communities of the American Southwest: United States and Mexico. Desert Plants 4: 1-342. Cottam, G. and J.T. Curtis. 1956. The use of distance measures in phytosociological sampling. Ecology 37. 451-460. Daniel, W.W. 1978. Applied nonparametric statistics. Houghton Mifflin Co., Boston, MA U.S.A. Fisher, A.K. 1893. The hawks and owls of the U.S. in relation to agriculture. U.S. Department of Agricul- ture. Ornithol. Mamm. Bull. 3:1-210. Flesch, A.D. 2003. Distribution, abundance, and habitat of cactus Ferruginous Pygmy-Owls in Sonora, Mexico. M.S. thesis, Univ. Arizona, Tucson, AZ U.S.A. Gilman, M.F. 1909. Some owls along the Gila River of Arizona. Condor 11:145-150. Knopf, F.L., J.A. Sedgwick, and D.B. Inkley. 1990. Re- gional correspondence among shrubsteppe bird hab- itats. Condor 92:45-53. Manly, B.F., J.L.L. MacDonald, and D.L. Thomas. 1995. Resource selection by animals: statistical design and analysis for field studies. Chapman and Hall, London, U.K. Mills, G.S., J.B. Dunning, Jr., andJ.M. Bates. 1991. The relationship between breeding bird density and veg- etation volume. Wilson Bull. 103:468-479. Mohr, C.O. and W.A. Stumpf. 1966. Comparison of methods for calculating areas of animal activity. J Wildl. Manage. 30:293—304. Phillips, A.R., J.T. Marshall, Jr., and G. Monson. 1964. The birds of Arizona. Univ. Arizona Press, Tucson, AZ U.S.A. Proudfoot, G.A. and R.R. Johnson. 2000. Ferruginous Pygmy-Owl {Glaucidium brasilianum cactorum). In A Poole and F. Gill [Eds.], The birds of North America, No. 498. The Birds of North America, Inc., Philadel- phia, PA U.S. A. Richardson, M.E. 2000. Characteristics of occupied cac- tus Ferruginous Pygmy-Owl {Glaucidium brasilianum cactorum) habitat at Organ Pipe Cactus National Mon- ument, Arizona. M.S. thesis, Arizona State Univ., Tem- pe, AZ U.S.A SOKAL, R.R. and F.J. Rohlf. 1995. Biometry, 3rd Ed. W.H Freeman, New York, NY U.S.A. United States Fish and Wildlife Service. 1997. Final rule. Endangered and threatened wildlife and plants; June 2003 Short Communications 157 determination of endangered species status for the cactus Ferruginous Pygmy-Owl {Glauddium brasilia- num cactorum) in Arizona. Federal Register 62(46): 10730-10747. United States Fish and Wii.dlife Service. 2002. Pro- posed rule. Endangered and threatened wildlife and plants; designation of critical habitat for the Arizona distinct population segment of the cactus Ferruginous Pygmy-Owl {Glauddium brasilianum cactorum) Novem- ber 27, 2002. Federal Register 67(229) :71032-71064 United States Fish and Wildufe Service. 2003. Cactus Ferruginous Pygmy-Owl {Glauddium brasilianum cac- torum) draft recovery plan. Albuquerque, NM U.S A Received 2 May 2002; accepted 15 March 2003 Associate Editor: Clint Boal J. Raptor Res. 37(2): 157-1 60 © 2003 The Raptor Research Foundation, Inc. Interspecific and Intraspecific Ki.eptoparasitic Interactions oe the Bearded Vulture ( Gypaetus barbatus) at Nesting Areas Antoni Margalida’ and Joan Bertran Group of Study and Protection of the Bearded Vulture, Apdo. 43, E-25520, El Pont de Suert, Lleida, Spain Key Words: Bearded Vulture, Gypaetus barbatus; inter- specific interactions', intraspecific interactions', kleptoparasitism', nesting areas. Kleptoparasitism is the stealing of previously procured food from heterospecifics or conspecifics (Brockmann and Barnard 1979). This behavior is quite widespread among birds, especially among seabirds (Furness 1987). Although kleptoparasitic interactions in vultures have been reported in the literature (Brockmann and Barnard 1979, Pascual and Santiago 1991, Bertran and Margalida 1996, Margalida and Heredia 2002) information on this behavior at nesting areas is scarce. This may be due to the fact that vulture species generally interact around the carcass (vMvarez et al. 1976, Konig 1983, Blanco et al. 1997, Mundy et al. 1992) and that they deliver food, which they carry in their crop to the nest, making theft of this food by other birds difficult. The Bearded Vulture {Gypaetus barbatus) is a solitary and territorial osteophagous vulture that inhabits moun- tain areas of the southern Palearctic and the Afrotropical region (del Hoyo et al. 1994). In contrast to the ecology of other vultures, this species presents some features that may favor kleptoparasitism: (1) the carrying of large bones or bone fragments that are visible to other species, (2) the repeated breaking actions that take place in the ossuaries, and (3) the predictability of food sources where prey items accumulate (ossuaries, perches, and nests). Nevertheless, kleptoparasitic events involving the Bearded Vulture have only been reported occasionally (Elosegi 1989). ^ E-mail address: margalida@gauss.entorno.es In the eastern Pyrenees, the Bearded Vulture popula- tion occurs in a high population density (Margalida et al. 2003) and with an abundant food supply (Margalida et al. 1997). These factors suggest a low frequency of klep- toparasitism events because this feeding strategy is fa- vored when food is less abundant (Stillman et al. 1997). On the other hand, a low frequency of interspecific in- teractions would be expected as consequence of low ben- efits that could be obtained by heterospecifics from steal- ing a specialized food such as bone remains. In this note, we document some interspecific and m- traspecific kleptoparasitic interactions of the Bearded Vulture at nesting areas and we analyze the factors af- fecting this behavior. Materiai. and Methods Fieldwork was undertaken between 1991-97 in the cen- tral Pyrenees (Catalonia, northeast Spain) during a larger study of the breeding biology of this species (Margalida and Bertran 2000). Eight focal pairs were studied and we recorded incidental observations of another seven pairs during the pre-laying, incubation, and chick-rearing pe- riods (September-July) . Bearded Vulture nests are situated on rocky cliffs at altitudes between 650 m and 2100 m. Among the species which coexist with the Bearded Vulture and which often interact with it are the Golden Eagle {Aquila chrysaetos) (10 territories), the Egyptian Vulture {Neophron per cnop- terus) (four territories), the Common Raven {Gorvus cor- ax) (14 territories), and the Eurasian Griffon {Gyps ful- vus) (nine territories). At the same time, there are also intraspecific interactions with individuals of various age classes (immatures, <3 yr; subadults 4—5 yr; adults >6 yr) that often visit the nesting areas. Observations were made using 20-60 X telescopes from vantage points that allowed a good view of nesting sites 158 Short Communications VoL. 37, No. 2 Table 1. Inter- and intraspecific kleptoparasitic actions of Bearded Vultures observed in nesting areas in Spain. Host Kleptoparasite Bearded Vulture (Adult) Golden Eagle Bearded Vulture (immature) 19 0 Bearded Vulture (subadult) 2 0 Bearded Vulture (adult) 1 1 Golden Eagle 2 0 Griffon Vulture 13 0 Common Raven 16 0 (300-600 m). Notes were made of all interactions in which food was stolen, the species involved, and the sit- uation in which they occurred (in flight, on the nest, at an ossuary, or on a perch). In all intraspecific interac- tions observed, we recorded the individual’s age, which was determined following Bertran and Margalida (1996). Results We observed 54 interactions consisting of food theft, 26 of which took place at ossuaries, 24 at the nest, 3 in flight, and 1 on a perch. Twenty-two of all interactions were intraspecific, while the remaining 32 were interspe- cific. No significant difference was observed among the situations where the two types of interactions took place (X ^2 “ 1.5, P = 0.47). In 96.3% of all interactions Beard- ed Vulture adults acted as hosts and only in two did they act as kleptoparasite (Table 1). Intraspecific Interactions. Thirteen of all intraspecific thefts took place at the ossuaries, eight at the nest, and only one in flight. The frequency of thefts observed was higher than expected in places where prey remains ac- cumulated, such as nests or ossuaries. Direct kleptopar- asitic actions in flight were avoided (x^i = 9.28, P < 0 001). Twenty-one actions (95.5%) were carried out by non-breeding individuals. Birds that were less than 3 yr old took part in 19 of all actions and this age group used this feeding strategy more often than expected (x^g = 13.01, P < 0.001). The only kleptoparasitic action in flight took place between two adult birds in a territory held by a polyandrous quartet. The parasitic individual followed insistently a bird that was carrying the bone re- mains, flapping its wings vigorously. After an aerial chase of 15 min, the adult bird that was carrying the remains dropped its prey, which the other adult bird collected on the ground. Interspecific Interactions. Of 32 interspecific interac- tions observed in which food was stolen, in only one case was the Bearded Vulture the kleptoparasitic species: the vulture robbed a Golden Eagle of the prey it was holding m Its talons while perched in the vicinity of its nest. Once m flight, the eagle chased the vulture and tried to recov- er the prey, which fell into a wooded area from which it could not be recovered. In the remaining 31 interactions, the Bearded Vulture acted as host. In flight it was klep- toparasitised only twice, both by Golden Eagles which chased and took from the vultures’ talons the prey they were carrying. On one of these occasions, the prey was dropped and the eagle retrieved it before it touched the ground. Of the other 29 occasions, the Bearded Vulture suffered 16 of the thefts at the nest by ravens and 13 of the thefts at ossuaries by Eurasian Griffons. As occurred with intraspecific interactions, interspecific events of kleptoparasitism mostly occurred in those places where food was accumulated/gathered, with kleptoparasitic chases in flight being significantly infrequent (x^i = 12.44, P< 0.001). Discussion Our results suggest that the non-breeding population of Bearded Vultures, perhaps due to limited foraging ef- ficiency (Brown 1988, Bertran and Margalida 1996), as has been suggested with other species (Fisher 1985, Cal- dow et al. 1999), are making use of the spatial and tem- poral predictability of food resources by becoming klep- toparasites. All thefts suffered at the nest by breeding pairs of Bearded Vulture took place during chick-rearing, a period when prey items often accumulate at the nest sites. Thefts at ossuaries or in flight occurred during win- ter (pre-laying and incubation periods), a time when food availability is reduced and weather may greatly limit the activities of foraging and locating food. For those age groups (principally <3 yr) that are more dependent on predictable food sources such as feeding stations (Here- dia 1991), this might be a foraging strategy used much more regularly. These results are in agreement with the idea that immature or inexperienced birds may compen- sate for their less-effective foraging abilities by kleptopar- asitism (Wunderle 1991). To the contrary, kleptoparasi- tism by adults could be an opportunistic foraging behavior. Nevertheless, our observations were mainly done close to the nest and do not include observations during foraging. This accounts for the fact that breeding adults were the host bird in 96% of all observed events. Nevertheless, the abundant food avalaible and the rela- tively infrequent number of stolen prey, suggest that klep- toparasitism amounts to a small cost for the breeding pairs, without any measurable impact on breeding suc- cess (see Margalida et al. 2003). As a result of the cost/benefit rate, two factors would determine that the species that attempted stealing would resort to this indirect strategy: the territorial behavior of the Bearded Vulture (Margalida and Bertran 2000, Ber- tran and Margalida 2002) and the accumulation of prey remains in nesting areas. Dominance of adults over im- matures is a well documented phenomenon in raptors (Newton 1979), but a reverse-dominance pattern also has been observed (Rodriguez-Estrella and Rivera-Rodriguez 1992). In the case of conspecifics, plumage coloration of June 2003 Short Communications 159 Bearded Vulture adults could act as status signal (Negro et al. 1999). This signal could be used by territorial adults to displace other immature Bearded Vultures not by at- tacking them, but simply by signalling their status while approaching them (Bautista et al. 1998). On the other hand, the Bearded Vulture’s low wing loading and its large wingspan give this species great dominance in flight (Donazar 1993) and make it difficult for an opponent to steal food successfully. In the case of conspecifics, the fact that younger birds are less skillful in flight would mean that they would be less successful in actions of direct pi- racy, so that the energetic cost of those attempts might be greater than the likely benefits obtained from those actions (Fisher 1985). Concerning the interspecific interactions, the prey re- mains consumed by Bearded Vultures mostly consisting of bone remains, would be energetically inadequate for many raptor species (e.g., Golden Eagles). In the case of Eurasian Griffons, they would be occasionally more in- terested in obtaining bone remains at ossuaries (Bertran and Margalida 1997) in order to compensate for a pos- sible lack of calcium in their diet (Mundy and Ledger 1976). The dominance of the Eurasian Griffon on the ground, given its larger size and the accumulation of bone fragments and splinters at ossuaries, would favor the strategy of obtaining calcium at these sites (Bertran and Margalida 1997). Eurthermore, the Bearded Vul- ture’s attacks of intruders in the vicinity of the nest throughout the breeding season (Brown 1988, Margalida and Bertran 2000) would act as deterrent and would make food storages near the nest the least convenient for stealing. The success in aggressive encounters appears determined by the body size and condition, and the pre- vious possession of the disputed resource (Bautista et al. 1998). In contrast, those species with higher aerial ma- neuverability but with smaller size, such as ravens, would have to focus their actions at the nest, where prey re- mains also accumulate. Obtaining prey remains there may be less costly for those birds because; (1) adults are gradually less often present at the nest as the breeding season progresses (Brown 1990, Margalida and Bertran 2000) and (2) prey items present in the nest have a high- er meat content as consequence of differential require- ments in nutrients for the chick (Margalida and Bertran 1997, Margalida and Bertran 2001). Finally, kleptoparasitism could be a strategy of dem- onstrating dominance or competence (Bautista et al. 1998). The fact that in our study area the Bearded Vul- ture density is high (Margalida et al. 2003), heterospe- cifics compete by the same nest sites (Margalida and Garcia 1999), and food supply is abundant, suggest that kleptoparasitic actions could be determined by other re- sources than food. For example, demonstrating domi- nance or competence by nest sites or in breeding terri- tories may result in social benefits (Caldow et al. 1999, Yates et al. 2000). Resumen. — Documentamos interacciones interespecifi- cas e intraspecificas de cleptoparasitismo en el quebran- tahuesos en los sectores de nidificacion. De un total de 54 interacciones, 26 tuvieron lugar en rompederos, 24 en el nido, tres en vuelo y uno en posaderos. Veintidos de las interacciones fueron intraspecificas y las 32 restan- tes interespecificas con cuervos {Corvus corax), buitres leonados {Gyps fulvus) y aguilas reales {Aquila chrysaetos). La mayoria de las interacciones tuvieron lugar en zonas donde el aliraento se acumulaba (nidos y rompederos) evitando las acciones directas en vuelo. En el caso de los adultos, este comportamiento podria ser una accion oportunista pero para las aves de <3 anos si podria tra- tarse de una estrategia alimenticia. Puesto que la dispo- nibilidad trofica es suficiente, la poblacion reproductora esta incrementandose y las especies heterospecificas com- piten por los mismos sectores de nidificacion, el clepto- parasitismo podria estar relacionado con otros factores no relacionados con la obtencion de alimento. [Traduccion de los autores] Acknowi.edgments We would like to thank J.A. Donazar, P.J. Mundy, JJ. Negro, J.R. Garrido, J.M. Arcos, S. Xirouchakis, and an anonymous reviewer for their comments on the manu- script and C. Carboneras and S. Hardie for translating the text into English. Funding for the research was pro- vided by Departament de Medi Ambient of Generalitat de Catalunya and Fundacio Territori i Paisatge. Literature Cited Alvarez, E, L. Arias de Reyna, and F. Hiraldo. 1976 Interactions among avian scavengers in southern Spain. Ornis Scand. 7:215-226. Bautista, L.M., J.C. Alonso, and J.A. Ai.onso. 1998. For- aging site displacement in common crane flocks Anim. Behav. 56:1237-1243. Bertran, J. and A. Margalida. 1996. Patron anual de observaciones de quebrantahuesos ( Gypaetus barbatus) de diferentes grupos de edad en los sectores de ni- dificacion. Alauda 64:171-178. and a. Margalida. 1997. Griffon Vultures {Gyps fulvus) ingesting bones at the os.suaries of Bearded Vultures {Gypaetus barbatus) , J. Raptor Res. 31:287-288 AND A. Margaijda. 2002. Territorial behavior of the Bearded Vulture in response to the Griffon Vul- ture. /. Field Ornithol. 73:86-90. Blanco, G., J.M. Traverso, J. Marchamalo, and F. Mart- Inez. 1997. Interspecific and intraspecific agression among Griffon and Cinereous Vultures at nesting and foragings sites./. Raptor Res. 31:77-79. Brockmann, H.J. and C.J. Barnard. 1979. Kleptoparasi- tism in birds. Anim. Behav. 27:487-514. Brown, C.J. 1988. A study of the Bearded Vulture Gypae- tus barbatus in southern Africa. Ph.D. dissertation, Univ. Natal, Pietermaritzburg, South Africa. . 1990. Breeding biology of the Bearded Vulture 160 Short Communications VoL. 37, No. 2 in southern Africa, Part II: the nestling period. Ostrich 61:33-42. Galdow, R.W.G., J.D. Goss-Gustard, R.A. Stillman, S.E.A. LE V. BIT DURELL, R. SwlNFEN, AND T. BrEGN- BALLE. 1999. Individual variation in the competitive ability of interference-prone foragers: the relative im- portance of foraging efficiency and susceptibility to interference./. Anim. Ecol. 68:869-878. DEL Hoyo, J., A. Ellioit, andJ. Sargatal. (Eds.). 1994. Handbook of the birds of the world. Vol. 2, New world vultures to guineafowl. Lynx Ed., Barcelona, Spain. Donazar, J.A. 1993. Los buitres ibericos: biologia y con- servacion. J.M. Reyero [Ed.], Madrid, Spain. Elosegi, I. 1989. Vautour fauve {Gyps fulvus), Gypaete barbu {Gypaetus barbatus) , Percnoptere d’Egypte (Neophron percnopterus) : synthese bibliographique et recherches. Acta Biologica Montana. Serie documents de travail, 3. GBEA, Pau, France. Fisher, D.L. 1985. Piracy behavior of wintering Bald Ea- gles. Condor 87:246-251. Furness, R.W. 1987. Kleptoparasitism in seabirds. Pages 77-101 mJ.P. Groxall [Ed.], Seabirds: feeding ecology and role in marine ecosystems. Gambridge Univ. Press, Gambridge, MA U.S.A. Heredia, R. 1991. Alimentacion y recursos alimenticios. Pages 79-89 in R. Heredia and B. Heredia [Eds.], El Quebrantahuesos (Gypaetus barbatus) en los Pirineos. IGONA, Goleccion Tecnica, Madrid, Spain. Konig, G. 1983. Interspecific and intraspecific competi- tion for food among old world vultures. Pages 153- 171 in S.R. Wilbur and J. A. Jackson [Eds.], Vulture biology and management. Univ. Galifornia Press, Berkeley and Los Angeles, GA U.S.A. Margalida, a. and j. Bertran. 1997. Dieta y seleccion de alimento de una pareja de Quebrantahuesos (Gypaetus barbatus) en los Pirineos durante la crianza. Ardeola 44:191-197. and j. Bertran. 2000. Breeding biology of the Bearded Vulture Gypaetus barbatus:. minimal sexual dif- ferences in parental activities. Ibis 142:225-234. AND J. Bertran. 2001. Function and temporal var- iation in the use of ossuaries by Bearded Vultures (Gypaetus barbatus) during the nestling period. Auk 118:785-789. AND D. GarciA. 1999. Nest use, interspecific rela- tionship and competition for nests in the Bearded Vulture Gypaetus barbatus in the Pyrenees: influence on breeding success. Bird Study 46:224-229. , D. Garcia, J. Bertran, and R. Heredia. 2003. Breeding biology and success of the Bearded Vulture (Gypaetus barbatus) in northeastern Spain. Ibis 145: 244-252. , D. GarcIa, and R. Heredia. 1997. Estimacion de la disponibilidad trofica para el Quebrantahuesos ( Gypaetus barbatus) en Gataluna (NE Espana) e impli- cacions sobre su conservacion. Dohana Acta Vert. 24: 235-243. and R. Heredia. 2002. Interspecific interaction between Lammergeier Gypaetus barbatus and Black Vulture Aegypius monachux predation or kleptoparasi- tism? Sandgrouse 24:138-139 . Mundy, R, D. Butchart, J. Ledger, and S. Piper. 1992. The vultures of Africa. Academic Press, London, U.K. AND J.A. Ledger. 1976. Griffon vultures, carni- vores, and bones. S'. Afr. J. Sci. 72:106-110. Negro, JJ-. A. Margalida, F. Hiraldo, and R. Heredia. 1999. The function of cosmetic colouration of Beard- ed Vultures: when art imitates life. Anim. Behav. 58: F14-F17. Newton, I. 1979. Population ecology of raptors. T. & A.D. Poyser, Berkhamsted, U.K. Pascual, J. and J.M. Santiago. 1991. Egyptian Vultures steal food from nestling Griffon Vultures./. Raptor Res. 25:96-97. RodrIguez-Estrella, R. and L. Rivera-Rodriguez. 1992. Kleptoparasitism and other interactions of Grested Garacara in the Gape Region, Baja Galifornia, Mexico. /. Field Ornithol. 63:177-180. Stillman, R.A., J.D. Goss-Gustard, and R.W.G. Galdow. 1997. Modelling interference from basic foraging be- haviour. / Anim. Ecol 66:692-703. Wunderle, J.M., Jr. 1991. Age-specific foraging proficien- cy in birds. Curr. Ornithol. 8:273—324. Yates, M.G., R.A. Stili.man, and J.D. Goss-Gustard. 2000. Gontrasting interference functions and foraging dispersion in two species of shorebird (Gharadrii)./ Anim. Ecol. 69:314—322. Received 18 April 2002; accepted 30 December 2002 Associate Editor: Juan Jose Negro June 2003 Short Communications 161 J. Raptor Res. 37 (2): 161-1 63 © 2003 The Raptor Research Foundation, Inc. Gyrfalcon Predation on Mallards and the Interaction of Bald Eagles Wintering in Central Alberta Dick Dekker^ 3819-112A Street, Edmonton, AB, T6J 1K4 Canada Gordon Court Fish and Wildlife Division, SRD, Government of Alberta, O.S. Longman Building, 6909-116 Street, Edmonton, AB, T6H 4P2 Canada Key Words: Gyrfalcon; Falco rusticolus; Bald Eagle, Hal- iaeetus leucocephalus; Mallard; Anas platyrhynchos; for- aging success; kleptoparasitism; predation. The food habits of Gyrfalcons {Falco rusticolus) have been studied mainly by the collection and identification of prey remains (Palmer 1988, Poole and Boag 1988, Cade et al. 1998). Field observations of Gyrfalcons capturing prey are few and anecdotal (White and Weeden 1966, Bengtson 1971, Dobler 1989, Garber et al. 1993). The larg- est data sets of hunts and kills by wintering Gyrfalcons orig- inate from urban study areas. Jenning (1972) and Dekker and Lange (2001), respectively, detailed the capture of 17 Mallards {Anas platyrhynchos) and 15 Rock Doves {Columba livid) in Stockholm and Edmonton. We present a compa- rable sample of Mallard kills by Gyrfalcons wintering in a rural region of Alberta. Bald Eagles {Haliaeetus leucocephal- us) usually forage over water, and their kleptoparasitic hab- its are well known, particularly at the expense of the Per- egrine Falcon {Falco peregrinus) (Anderson and DeBruyn 1979, Dekker 1987, 1995). In this paper, additionally, we describe observations of eagles hunting ducks over land and pirating prey from Gyrfalcons. Study Area and Methods The latitude of the study area is 53°N and the climate IS cold continental. The North Saskatchewan River, that flows through Edmonton, Alberta, is frozen from Novem- ber to May except for a stretch of roughly 10 km down- stream from the city. The open water attracts 1 000 — 2000 Mallards that stay all winter. In fall, they make daily feed- ing flights of 2-10 km over gently undulating agricultural plains. After the ground is covered with >10 cm of snow, the ducks abandon the stubble fields and congregate at farms where cattle are being fed with grains or silage. Gyrfalcons are migrants and winter residents in central Alberta. Earliest and latest records are 25 September and 14 April, although the majority of sightings date from November to mid March (Dekker 1983, Court 1999). Bald Eagles winter in south-central Alberta where lakes or rivers remain ice free (Godfrey 1986). ^ E-mail address: tj_dick_dekker@hotmail.com To study the interaction between ducks and Gyrfalcons we used three principal methods. (1) Sitting in a parked vehicle, we monitored feeding ducks for sudden alarm behavior caused by the arrival of a predator. (2) Erom a vantage point overlooking the river valley, we frequently scanned the skies throueh binoculars, waitine for the Mallards to leave the river. (3) Gyrfalcons, either flying or perched, were observed in anticipation that they may initiate the pursuit of prey. Over three winters, 1999- 2002, we visited the study area on 95 days (3-5 hr/d) and sighted one or more Gyrfalcons on 56 days. A “hunt” was defined as an attempt by a Gyrfalcon at capturing a duck of which the outcome was known (Dek- ker 1980). The term “kill” indicates that we saw the Gyr- falcon pursue and seize a duck, or that we located the falcon on its prey shortly after the hunt. A “probable kill” indicates that a Gyrfalcon, in close pursuit of a duck, chased it down to the ground and disappeared from view behind trees or sloping terrain in the distance. Addition- al clues were provided by Bald Eagles that flew to and descended in the same locality of the suspected Gyrfal- con kill. Results and Discussion Gyrfalcons. During each winter, 1999-2002, we sighted four to seven different Gyrfalcons, ranging from very dark immatures to partly white adults (Court 1999, Dek- ker and Lange 2001). Some recognizable individuals stayed in the study area for a period of several days or weeks; others were sighted only once. Gyrfalcons are known to hunt low over the ground and seize avian prey just after it flushes. The target prey may be first spotted from an elevated perch (White and Wee- den 1966, Bengtson 1971, Palmer 1988, Glum and Cade 1994) . A similar strategy of surprise was employed by the Gyrfalcons in this study. Their sudden arrival, flying low over the ground, caused the ducks to fly up. After se- lecting an individual duck, the Gyrfalcon chased it on an erratic, twisting course. If the hunt failed, the falcon perched on a post or tree and then launched another hunt. We saw a total of 42 low surprise hunts of which five resulted in a kill (Table 1). In four additional instanc- es, we just missed seeing the attack and discovered the falcon on a fresh kill. Six successful low-surprise hunts 162 Short Communications VoL. 37, No. 2 Table 1. Hunting methods of Gyrfalcons preying on Mallards in Alberta, Canada. Hunting Method Hunts >— 1 in Prob- able Kills Stealth attack on feeding ducks 42 5 — High, open attack on flying ducks 25 10 3 Unknown approach 3 1 1 Totals 70 16 4 were reported to us by the associate observers listed in the Acknowledgments. A very different strategy was employed by Gyrfalcons that flew out to meet airborne ducks at altitudes of 50- 300 m and still >2 km away. Leaving a perch, the Gyr- falcon began a long climb, gradually increasing speed. At the falcon’s approach, the ducks turned away and split up. In the terminal stage of the hunt, the Gyrfalcon se- lected a single duck that appeared to realize its vulnera- bility and descended in an attempt to reach cover. Twen- ty-five of these open hunts resulted in 10 confirmed and three probable kills (Table 1). Most pursuits ended in a close tail-chase. In the two exceptions, the Gyrfalcon flew 20-100 m higher than the duck and stooped at it nearly vertically. In both cases, the duck dodged the stoop and landed, followed by the fal- con. One was a confirmed kill, the other a probable kill. We saw no evidence that the falcons struck and knocked down their prey. In all observed captures, the falcon seized its prey in mid-air or on the ground. The majority (77%) of ducks that were closely pursued escaped into low vegetation protruding from the snow, or into bushes or roadside trees. In two instances, the falcon landed on the road shoulder and, in a futile at- tempt to flush the duck, walked into the snow-drifted ditch under the trees. Eight ducks landed on the weedy shoulders or medians (10-15 m wide) of divided high- ways, and two of these ducks were seized by the falcon while heavy traffic passed by. In other cases, the falcon repeatedly swooped at the duck but failed to grab it. One drake, observed at close range, defended itself by lunging with gaping bill at the swooping raptor. After several failed passes, the Gyrfalcon perched in the vicinity. When the drake took off, it was immediately pursued until it again took cover. E. Pletz saw a Gyrfalcon pursue a duck until it went down on an open, snow-covered field. The falcon swooped at the duck, but did not seize it and land- ed 1-2 m away. Such stand-offs between Gyrfalcons and Mallards have been described by others (Jenning 1972). Gyrfalcons can be similarly reluctant to grab tethered de- coy pigeons that refuse to flush (Dekker and Lange 2001, E. Pletz pers. comm.). A true assessment of the hunting success rate of the Gyrfalcons in this study is difficult. Based on confirmed kills, the success rate was 22.9%. The addition of proba- ble kills brings the rate up to 28.6%. This is more than twice as high as the 10.6% success rate in 141 hunts by Gyrfalcons preying on Rock Doves (Dekker and Lange 2001). Mallards are less maneuverable than pigeons and usually escape from falcons by plunging into water. In this study, the Gyrfalcons outflew and forced down any Mallard they pursued with persistence. Although the long-range, high altitude interception of flocks of ducks by Gyrfalcons has not been described in the literature, similar direct-climbing attacks on other avian prey have been reported and the Gyrfalcon’s capacity in this regard is well-known to falconers (Cade 1982). An identical mode of hunting, launched from a perch and aimed at the interception of high-flying ducks, was employed by male peregrines (F. p. pealei) wintering in British Colum- bia (Dekker 1995, 1999). Although some Gyrfalcons observed in the study area were seen hunting Rock Doves or found feeding on Gray Partridges {Perdix perdix ) , their primary prey appeared to be the (locally-wintering) Mallards. Probably in response to predation risk, the ducks delayed their feeding flights During fall, the flocks routinely traveled to the stubble fields near noon. However, after they had been attacked often, they might not leave the river until close to sun- down. For instance, on 18 February 2002 the first flock (20-25 ducks) left the river at 1500 H. As soon as they detected an approaching Gyrfalcon, the flock turned back and escaped into water. The falcon remained in the area, often switching perches. Two hours later, the ducks suddenly left the river again in multiple flocks totalling hundreds of birds. At the approach of the Gyrfalcon, only the flock under direct attack returned to the river, pursued by the falcon, while other flocks continued on to a feeding area about 8 km away. Of nine Mallard kills of which the sex was known, sev- en were drakes and two hens. In two instances, a Gyrfal- con ate only part of a drake and returned to the remains the following morning. Two other falcons consumed all flesh from the carcass in 35-45 min, leaving only the head, the pelvis, and the lower portion of the legs. As reported by a farmer who shot a number of raiding Mal- lards, by late winter these ducks were emaciated. Bald Eagles. During the freeze-up period, the last of the Bald Eagles migrating through central Alberta active- ly hunt ducks in water holes of lakes (Dekker 1984) However, the food habits of the Bald Eagles wintering in the study area have not been studied in detail. Each year, we counted three to six eagles perched on trees along the open stretch of the river. One eagle was seen to catch a Mallard in the river. In late fall, the eagles also perched on trees overlooking the stubble fields where Mallards congregated. Some eagles actively hunted the ducks in fast, low-surprise attacks. However, all of 12 chases of ducks that flushed just ahead of eagles were unsuccessful One immature eagle pounced on a drake that had turned back and landed again 20-25 m away. Associate June 2003 Short Communications 163 observers reported three additional captures of Mallards of which one was taken in flight (F. Whiley) . After the fields were covered in snow and the ducks commuted daily to cattle feedlots, the eagles left the river and perched on trees in view of the feeding ducks. We suspect that they were also watching for opportunities to steal prey from other raptors. We never saw Gyrfalcons attack ducks on farms, where eagles sat waiting on prom- inent perches. However, if falcons hunted in the distance, it was common to see eagles fly toward that direction, apparently searching the fields. We saw four eagles rob Gyrfalcons, which released their ducks at once. Two ad- ditional cases of kleptoparasitism were reported by asso- ciates (E. Pletz, F. Whiley). The Gyrfalcons did not de- fend their prey against the eagles. Other raptors seen to hunt the ducks were Northern Goshawks {Accipiter gentil- is) (six low attacks), Prairie Falcon {Falco mexicanus) (one flush and long chase), and Snowy Owls {Nyctea scandiaca) (two low attacks) . After watching a Gyrfalcon feed on its kill, one Snowy Owl scavenged the remains. Note added in proof. During February and March of 2003, we saw 35 hunts (24 stealth and 11 open attacks), resulting in three probable and four confirmed kills. Ad- ditionally, the falcons seized two ducks, on or near the ground, which were released again moments later. Three confirmed kills were pirated by Bald Eagles. All of five known falcon kills were hen Mallards. Resumen. — Patos de la especie Anas platyrhynchos que mi- graron a un estrecho no congelado rio abajo de la ciudad de Edmonton, Alberta, intercambiaron diariamente los campos de siembra y las areas de forraje de ganado, don- de fueron atacados por halcones gerifaltes {Falco rustico- lus). Setenta intentos de caza dieron lugar a 16 muertes confirmadas y cuatro probables. La mayoria {11%) de los patos perseguidos por los halcones escaparon a la cap- tura, cubriendose en el suelo, en arbustos, o a lo largo de los hordes de la carretera. Las presas fueron atrapadas en tierra o en aire. Los gerifaltes utilizaron dos metodos principales de caza; (1) ataques bajos por sorpresa contra patos que se alimentaban; (2) intercepciones de largo alcance de patos en vuelo alto. Las tasas de exito para estos metodos de caza fueron 12% y 40%, respectivamen- te. Las aguilas calvas {Haliaeetus leucocephalus) persiguie- ron a los patos sobre campos de cultivo, teniendo un exito relativo. Adicionalmente, se apoderaron de patos que habfan sido capturados por los halcones. [Traduccion de Cesar Marquez] Acknowledgments All expenses incurred during this study were privately financed by the authors. We thank the following associate observers for sharing their observations; Erhard Pletz, Gerald Romanchuk, and Fred Whiley. T.J. Cade, R. Rit- chie, and R.B. Weeden made much appreciated com- ments on the manuscript. Literature Cited Anderson, C.M. and P.D. DeBruyn. 1979. Behavior and ecology of Peregrine Falcons wintering on the Skagit Flats, Washington State. Unpubl. Report Dept, of Game, Olympia, WA U.S.A. Bengtson, S.A. 1971. Hunting methods and choice of prey of Gyrfalcons, Falco rusticolus, at Myvatn in Ice- land. Ibis 113;468-476. Cade, T.J. 1982. The falcons of the world. Cornell Univ. Press, Ithaca, NY U.S.A. , P. Koskimies, and O.K. Nielsen. 1998. Gyrfalcon {Falco rusticolus). Birds of the western palaearctic 2’1 Oxford Univ. Press, Oxford, U.K. Cl.UM, N.J. AND T.J. Cade. 1994. Gyrfalcon {Falco rustico- lus) in A. Poole and F. Gill [Eds.], The birds of North America, No. 114. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A. Court, G. 1999. Edmonton’s year of the Gyr. Edmonton Nat. 27:29-30. Dekker, D. 1980. Hunting success rates, foraging habits, and prey selection of Peregrine Falcons migrating through central Alberta. Can. Field-Nat. 94:371-382 . 1983. Gyrfalcon sightings at Beaverhills Lake and Edmonton, 1964-1983. Alberta Nat. 13:103. . 1984. Migrations and foraging habits of Bald Ea- gles in east-central Alberta, 1964—1983. Blue Jay 42’ 199-205. . 1987. Peregrine Falcon predation on ducks in Alberta and British Columbia. J. Wildl. Manage. 51: 156-159. . 1995. Prey capture by Peregrine Falcons winter- ing on southern Vancouver Island, British Columbia. J. Raptor Res. 29:26-29. . 1999. Bolt from the blue — ^wild peregrines on the hunt. Hancock House Publishers, Surrey, BC, Canada andJ. Lange. 2001. Hunting methods and success rates of Gyrfalcons and Prairie Falcons preying on fe- ral pigeons (Rock Doves) in Edmonton, Alberta. Can. Field-Nat. 115:395-401. Dobler, F.C. 1989. Wintering Gyrfalcon {Falco rustico- lus) habitat utilization in Washington. Pages 61-70 in B.-U. Meyburg and R.D. Chancellor [Eds.], Rap- tors in the modern world. WWGBP, Berlin, Germany. Garber, C.S., B.D. Mutch, and S. Platt. 1993. Obser- vations on wintering Gyrfalcons {Falco rusticolus) hunt- ing Sage Grouse {Centrocercus urophasianus) in Wyo- ming and Montana, U.S.A. / Raptor Res. 27:171-172 Godfrey, W.E. 1986. The birds of Canada. Revised Ed. Na- tional Museum of Natural Sciences, Ottawa, Canada Jenning, W. 1972. laktagelser rorande en overwintrande jaktfalk {Falco rusticolus). Vdr Fdgelvdrld 32:1-8. Pai.mer, R.S. 1988. Handbook of North American birds. Vol. 5. Diurnal raptors, part 2. Yale Univ. Press, New Haven, CT U.S.A. Poole, K.G. and D.A. Boag. 1988. Ecology of Gyrfalcons in the central Canadian arctic: diet and feeding be- havior. Can. J. Zool. 66:334-344. White, C.M. and R.B. Weeden. 1966. Hunting methods of Gyrfalcons and behavior of their prey. Condor 68; 517-519. Received 3 July 2002; accepted 24 January 2003 164 Short Communications VoL. 37, No. 2 J. Raptor Res. 37 (2): 164-1 66 © 2003 The Raptor Research Foundation, Inc. Wintering Snowy Owls Feed on Sea Ducks in the Belcher Islands, Nunavut, Canada Gregory J. Robertson^ Canadian Wildlife Service, 6 Bruce Street, Mount Pearl, NL, AIN 4T3 Canada H. Grant Gilchrist Canadian Wildlife Service, 1125 Colonel By Drive, Raven Road, Carleton University, Ottawa, ON Canada Key Words: Snowy Owl, Nyctea scandiaca; diet, polynyas; sea ducks; wintering. Snowy Owls {Nyctea scandiaca) have a broad wintering distribution that covers much of Canada and northern United States (Kerlinger et al. 1985). The owls consis- tently winter in the northern Great Plains while their movements to the east and west are more irruptive con- sisting mostly of immature birds (Kerlinger and Lein 1986). Snowy owls have a varied diet of mammals and birds, but usually focus on lemmings {Lemmus spp. and Dtcrostonyx spp.) during the breeding season and other small rodents and birds that are abundant during the non-breeding season (Watson 1957, Boxall and Lein 1982a, Parmelee 1992). Most research on wintering Snowy Owls has been con- ducted in the Canadian Prairies (Boxall and Lein 1982a, 1982b, Kerlinger and Lein 1988a, 1988b), but there have been relatively few studies of Snowy Owls wintering fur- ther north. The Belcher Islands, Nunavut, Canada (56°00'-57°30'N, 79°30-80°00'W) are known breeding and wintering areas for Snowy Owls (Todd 1963, Free- man 1970, Manning 1976). While studying the ecology of sea ducks wintering around the Belcher Islands, we also recorded Snowy Owls (Gilchrist and Robertson 2000) . Herein, we expand on our observations of Snowy Owls, describing their distribution and their relationship with wintering Common Eiders {Somateria mollissima) and Long-tailed Ducks {Clangula hyemalis). Methods Field work was conducted during three separate 3-4 wk periods during the winters of 1998 and 1999 (March 1998, January 1999, March 1999) on the Belcher Islands, southeastern Hudson Bay. Throughout the winter, areas of near-shore, open water are present in the Belcher Is- lands archipelago. These open-water areas are of two types: (1) small recurring polynyas in areas of strong tidal currents, and (2) water that is adjacent to ice-floe edges and remains free of ice only during offshore winds (Na- kashima and Murray 1988, Gilchrist and Robertson 2000 ) . During each winter, we traveled by snowmobile, with ^ E-mail address: greg.robertson@ec.gc.ca local Inuit guides, to various open-water areas around the Belcher Islands (Gilchrist and Robertson 2000). Desti- nations depended mainly on past and present weather conditions, and on information from other Sanikiluaq hunters who reported locations of recent sea duck sight- ings. After arriving at each area of open water, we estimated how many individuals of each species were present. If ducks were sighted and were reasonably close to the ice edge (<500 m) , we remained at the same site to collect behavioral data and information on the population of birds using the site; otherwise, we moved on to another open-water area. If conditions were suitable at the end of observations, our guides attempted to collect sea ducks. We did not always see Snowy Owls immediately upon ar- rival as we inspected surrounding ice ridges with spotting scopes or binoculars after observations of sea ducks were completed. Sometimes we saw owls after they flew; there- fore, we believe our sightings underestimate the actual number of Snowy Owls present. Snowy Owl numbers at water bodies were best de- scribed by a Poisson distribution (coefficient of disper- sion = 1.63), so we used Poisson regressions to examine relationships between duck and Snowy Owl numbers. The data were observational, so we conducted the anal- ysis using a model-selection framework (Burnham and Anderson 1998) rather than a traditional hypothesis-test- ing framework. In the model-selection framework, can- didate models are constructed, and their ability to fit the data, without including too many parameters, is evaluat- ed using Akaike’s Information Criterion (AIC). To ex- plain the number of Snowy Owls sighted, we constructed four candidate models. The first model included terms for numbers of Common Eiders and Long-tailed Ducks present, the second only eiders, the third only Long- tailed Ducks, and the last included only an intercept term. As some observations were from the same water body, we included the location of sightings as a repeated measure in the Poisson regression, and used a corrected AIC (termed AIC^) that used number of locations rather than number of sightings. Medians, means, and ranges are also presented to describe the data (see Gilchrist and Robertson 2000). Results and Discussion Snowy Owls were regularly seen at several open-water areas. On 56 trips to 14 different water bodies. Snowy Owls were seen 15 times. The observed number of Snowy June 2003 Short Communications 165 Table 1. Model fitting results of Poisson regressions, with numbers of ducks as the predictor (s) and number of Snowy Owls as the response. Models with the lowest AJCc values indicate the most parsimonious (i.e., best fit- ting) model. K is the number of parameters, while De- viance is a measure of model fit. Model K Deviance AICc Common Eiders + Long-tailed Ducks 3 62.78 71.18 Common Eider 2 68.06 73.15 Long-tailed Duck 2 69.91 75.00 Intercept only 1 84.31 86.65 Owls (x = 0.4, range 0-1) was positively related to the observed number of Common Eiders (x = 1169, range = 0-12 500) and to the observed number of Long-tailed Ducks (x = 55, range = 0-700) (Table 1). On two different occasions we saw a Snowy Owl take a carcass of a Long-tailed Duck directly off the water im- mediately after the duck was shot by our Inuit guides. On another occasion, when our guides shot a male Common Eider, a Snowy Owl made a few passes over the carcass but did not come close enough to seize it, presumably because we were standing directly at the floe edge trying to retrieve the eider. Our Inuit guides described addi- tional instances where they observed owls taking crippled eiders off the water (Nakashima and Murray 1988). In all cases, the owls appeared from the ice-edge and flew out over the water. Owls often perched on ice ridges piled up at the landfast ice-edge, and appeared to watch sea ducks from there. Owl pellets and what appeared to be raptor-killed carcasses (only breast bone and wings re- maining) of both eiders and Long-tailed Ducks were also found at ice perches along the floe edge. To our knowl- edge, there are no other raptors wintering in the Belcher Islands (Gilchrist and Robertson 2000), although some of these remains may have been left by scavenging Arctic foxes {Alopex lagopus) and Common Ravens (Corvus cor- ax) , We were able to travel continuously along two of the floe edges. The first was ca. 10 km long, and we saw four owls; the second was ca. 15 km long, and we saw three owls. The owls appeared to be spaced out evenly along these flow edges, at approximately 2-5 km intervals. Snowy Owls wintering in the Prairie Region of Canada form winter territories (Boxall and Lein 1982b), as do Snowy Owls wintering in Massachusetts (D. Holt pers. comm.). It is therefore possible that Snowy Owls form temporary territories along the floe edges that have open water. Snowy Owls apparently associate with concentrations of Common Eiders and Long-tailed Ducks. Eiders spend considerable time loafing and resting on the ice, espe- cially at night or when currents are strong. Out of water, eiders may be particularly vulnerable to Snowy Owls and other predators, such as Arctic fox, as they cannot dive to escape (Nakashima and Murray 1988). While Long- tailed Ducks (700-800 g) may be easier prey for Snowy Owls, Common Eiders may be a more important source of food because they are larger (>2000 g) and more abundant in the Belcher Islands. Also, large numbers of Common Eiders (lOOs-lOOOs) occasionally starve when open-water areas freeze over (Nakashima and Murray 1988, Gilchrist and Robertson 2000), making the eiders easy prey. Although Snowy Owls are known to prey primarily on small mammals, they will consume waterbirds when that prey is plentiful. On the coast of southwestern British Columbia, wintering Snowy Owls consumed a wide vari- ety of waterbirds, with Horned Grebes {Podiceps auritus) being the most common prey item, followed by Buffle- head (Bucephala albeola) (Campbell and MacColl 1978). Although Campbell and MacColl (1978) demonstrated that birds in 400-800 g weight class were preferred, birds as large as White-winged Scoters {Melanitta fused) at 1300 g, Mallards {Anas platyrhynchos) at 1100 g, and Glaucous- winged Gulls (Larus glaucescens) at 1200 g, were also pre- sent in the diet. In Alaska, non-breeding Snowy Owls con- sumed many Ancient Murrelets {Synthliboramphus antiquus) , as well as other alcids and ducks (Williams and Frank 1979). Snowy Owls migrating along the St. Lawr- ence River fed heavily on murres ( Uria spp.) (Bent 1938) . Portenko (1989) describes that Snowy Owls wintering on St. Lawrence Island, Bering Sea, fed exclusively on win- tering ducks, specifically Long-tailed Ducks and King Ei- ders {Somateria spectabilis) . Snowy Owls wintering in the Belcher Islands also appear to focus their winter foraging in marine areas where sea ducks are plentiful. We suggest that Snowy Owls are important members of the avian community wintering in the Belcher Islands. Our observations show that Snowy Owls prey upon abun- dant wintering sea duck populations and that their dis- tribution and abundance within the region are positively associated with concentrations of sea ducks. Sea ducks have to move around the Belcher Islands archipelago as open-water areas change locations with varying currents and winds. Snowy Owls likely move with the sea ducks, following them around the archipelago and possibly es- tablishing temporary territories. As with the Great Plains of North America (Kerlinger and Lein 1986), the Belcher Islands, and perhaps other areas in the Arctic with re- curring open-water bodies and wintering sea ducks, are likely used consistently by Snowy Owls (Portenko 1989) Resumen. — Documentamos la presencia del buho nival {Nyctea scandiaca) durante la migracion de otono en las Islas Belcher, Nunavut en 1998-99. En 56 vi^es a los si- tios de estudio, algunos buhos fueron relacionados con Somateria mollissima y con Clangula hyemalis presentes en los cuerpos de agua abiertos. Lo buhos fueron observa- dos, capturando en el agua a los patos de cola larga he- 166 Short Communications VoL. 37, No. 2 ridos. Tambien fueron detectados en perchas de filos de hielo con restos de patos o con egragopilas cercanas. Las areas de aguas abiertas de las Islas Belcher y las de las “polynias” recurrentes, asi como los hordes de los tem- panos de hielo, pueden ser utilizados consistentemente como areas de destino de la migracion de otoho, por el buho nival. [Traduccion de Cesar Marquez] Acknowledgments We thank our guides, Johnassie Inuktaluk, Noah Mee- ko, Elijah Oqaituk and Johnny Oqaituk, for great com- pany while taking us around the Belcher Islands. Discus- sions of the animals, ice and survival were much appreciated and certainly educating. Lucassie Arragutai- naq Jr., the Sanilikuaq Hunters and Trappers Association and the Municipality of Sanikiluaq provided invaluable logistical support. Dwayne Searle and employees of the Mitiq Cooperative made our stay in Sanikiluaq enjoyable. Maureen Kay provided logistical support from Yellow- knife. This study received generous funding from the Nu- navut Community Development Fund, the World Wildlife Fund, Wildlife Habitat Canada and the Canadian Wildlife Service. G. Robertson was supported by Atlantic Coop- erative Wildlife Ecology Research Network while under- taking this research. R Boxall, L. Dickson, D. Holt, and T Wellicome all provided helpful comments that certain- ly improved this manuscript. Literature Cited Bent, A.C. 1938. Life histories of North American birds of prey. Part 2. Smithsonian Institution United States National Museum Bulletin, Washington, DC U.S.A. Boxall, P.C. and M.R. Lein. 1982a. Feeding ecology of Snowy Owls (Nyctea scandiaca) wintering in s. Alberta. Arctic 35:282-290. AND . 1982b. Territoriality and habitat se- lection of female Snowy Owls {Nyctea scandiaca) in winter. Can. J. Zool. 60:2344—2350. Burnham, K.P. and D.R. Anderson. 1998. Model selec- tion and inference — a practical information-theoretic approach. Springer-Verlag, New York, NY U.S.A. Campbell, R.W. and M.D. MacColl. 1978. Winter foods of Snowy Owls in southwestern British Columbia. J. Wildl Manage. 42:190-192. Freeman, M.M.R. 1970. The birds of the Belcher Islands, N.W.T, Canada. Can. Field-Nat. 84:277-290. Gilchrist, H.G. and G.J. Robertson. 2000. Observations of marine birds and mammals wintering at polynyas and ice edges in the Belcher Islands, Nunavut, Ca- nada. Arctic 53:61-68. Kerlinger, R, M.R. Lein, and B. Sevtck. 1985. Distribu- tion and population fluctuations of wintering Snowy Owls {Nyctea scandiaca) in North America. Can.J. Zool. 63:1829-1834. and . 1986. Differences in winter range among age-sex classes of Snowy Owls Nyctea scandiaca in North America. Ornis Scand. 17:1-7. and . 1988a. Causes of mortality, fat con- dition, and weights of wintering Snowy Owls. J. Field Ornithol. 59:7-12. and . 1988b. Population ecology of Snowy Owls during winter on the Great Plains of North America. Condor 90:866-874. Manning, T.H. 1976. Birds and mammals of the Belcher, Sleeper, Ottawa, and King George Islands, Northwest Territories. Canadian Wildlife Service Occasional Pa- per Series No. 28, Ottawa, ON Canada. Nakashima, DJ. and D.J. Murray. 1988. The Common Eider {Somateria mollissima sedentaria) of eastern Hud- son Bay: a survey of nest colonies and Inuit ecological knowledge. Environmental Studies Revolving Funds Report, No. 102. Ottawa, ON Canada. Parmelee, 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. PORTENKO, L.A. 1989. Birds of the Chukchi Peninsula and Wrangel Island. Vol. 2. Smithsonian Institution Libraries and the National Science Foundation, Wash- ington, DC U.S.A. Todd, W.E.C. 1963. The birds of the Labrador Peninsula Univ. Toronto Press, Toronto, ON Canada. Watson, A. 1957. The behaviour, breeding, and food- ecology of the Snowy Owl Nyctea scandiaca. Ibis99:419- 462. Williams, RL. and L.G. Frank. 1979. Diet of the Snowy Owl in the absence of small mammals. Condor81:218— 214. Received: 2 November 2001; accepted 17 February 2003 Associate Editor: Troy Wellicome June 2003 Short Communications 167 J. Raptor Res. 37(2):167-172 © 2003 The Raptor Research Foundation, Inc. The Colonization of Sicily by the Black Kite {Milvus migrans) Maurizio Sara^ Department of Animal Biology, University of Palermo, 18 via Archirafi, 90123, Palermo, Italy Key Words: Black Kite, Milvus migrans; Red Kite, Milvus milvus; colonization', Sicily. The Black Kite {Milvus migrans) is a locally common and widespread migratory breeder throughout the Pale- arctic (Bijlsma 1997). This species was much more abun- dant in the recent past and it has been classified as vul- nerable within Europe (Vinuela and Sunyer 1994). In Italy, the species’ stronghold is located in the lake district of the Italian Alps, where eight of nine populations sur- veyed in the 1990s were found to be declining (Sergio and Boto 1999, Sergio et al. 2002). Materials and Methods The Sicani Mountain, in central Sicily, is an area of ca. 1500 km^, where the raptor population has been contin- uously monitored for the last 30 years. The raptor surveys have included both the wintering population, by road census, and the breeding pairs by direct observation and counts. To monitor the wintering birds of prey some 6000 km of road were surveyed by vehicle for over two decades (Massa 1980, Sara et al. 1993, Sara 1996a and references therein). The breeding population was monitored dur- ing the first (Massa 1985) and second Regional Atlas of breeding birds (Lo Valvo et al. 1993) and then continu- ously from 1999-2001. From January 1990-August 2001, 130 excursions have been carried out to count the local population of Black and Red kites {M. milvus) totalling 620 hr of observation. Local ornithologists, studying other raptors in the area m 1980-2001 also provided 340 records of kite observa- tions. All the records on the Black Kite {N = 86 out of 470) were extracted from these databases and chronologically ordered to investigate the trend of population settle- ment. Black Kites were aged in the field following char- acteristics in Forsman (1999). Several factors facilitated this analysis, namely the con- tinuous monitoring of the area, the low densities of win- tering and breeding kites, and their preferential use of some easily-monitored sites (e.g., dumps, roosts, cliffs). In addition, the review of the ornithological literature of past centuries allowed the reconstruction of the historical status of the species. Results and Discussion Historical Status (1840-1900). The Black Kite was known in Sicily since the 19th century. The vernacular ^ E-mail address: mausar@unipa.it name of this species was ‘Nigghiu di passci (migrating kite) or 'Nigghiu niuru’ (black kite). Sicilian ornitholo- gists (Benoit 1840, Mina Palumbo 1853, 1857, Doderlem 1869-74, 1893) mostly considered the species as rare and occasional. However, Maltese Schembri (1843), reported Black Kites as common in inland Sicily. Doderlein (1869- 74) considered the latter report a mistake and wrote in 1869: “. . . indigenous in Northern Africa, very rare in Sicily; Schembri considers it as common but I think he is ivrong. I have had notice of only one specimen around Palermo, which was stuffed, and I am not sure it exists in Sicily." The living and working locations of the past ornithologists, who report- ed the species’ rarity, were distributed along the current migration route (Fig. 1) and cast very few doubts on this past status. Recent Status (1950-2000). Today, the Black Kite is a common passage migrant along the Northern coasts of Sicily and the Messina Strait, from early March to late May. During spring, it was ranked as the second species crossing the Messina Strait in order of abundance (Di- marca and lapichino 1984, Galea and Massa 1985, Agos- tini 1992). Thiollay (1977) estimated that 15 000 was the minimum number of Black Kites crossing the Mediter- ranean Sea, from Tunisia (Cape Bon) to Sicily. The spe- cies is more numerous on the autumnal passage, from early August to late October (lapichino and Massa 1989). Huge concentrations of up to 1000 birds were recorded recently during late August to mid September over the Egadi Islands (Agostini et al. 2000) . The Colonization Process. Faunistic turnover is ex- pected to occur on islands (MacArthur and Wilson 1967); but within birds, Passerines are most commonly involved (Diamond 1969, 1972, Blondel 1986). Because the past ornithological literature reported a striking dif- ference in the Black Kite status from what occurs today, I analyzed the information related to colonization in de- tail. The first confirmed breeding attempt in Sicily oc- curred in 1979 (Massa 1980), though earlier records of possible breeding were reported in the 1950s (Mebs 1957). Black Kites in Sicily breed on bare cliffs or, more often, on trees and big shrubs (mainly Quercus ilex) grow- ing on the cliff face. Until now, only one nest is known on a tree in a woodlot, as commonly reported elsewhere in continental Europe (Sergio and Boto 1999). During the road census completed during the winters 1977-80 (Massa 1980) the species was found to be ab- 168 Short Communications VoL. 37, No. 2 A,B Lakes and streams • “ Winter roost ' • Winter record \ ^ Summer roost I.. / • t: T •; t r ^ ^ ) r:/ + •A ^ ■ B f • 1 J' '' 1 1 f 1 \ ' 1 . i # t i 1 7 ‘ ' ' M- 0 20 km Figure 1. Map of Sicily with superimposed UTM grid of 100 km^ cells, showing the main breeding and wintering area (A) of the Black Kite {Milvus migrans) and the other aquatic habitats potentially suitable for breeding (B). The main winter and summer roosts were in the core of the breeding range. The 2nd Regional Adas in 1992 and 2000 recorded 10 breeding squares (large black dot = confirmed breeding; medium black dot = probable; small dot = possible) . The census of 2000 showed differences in breeding status but not in the recorded number of squares. The large arrows show the main migratory routes of Black Kites over Sicily during late summer and autumn. Palermo (PA), Castelbuono (CB), and Messina (ME) are the cities where the 19th century ornithologists, who described the past species rarity, lived. sent, but 10 yr later (1987-92), the censuses in the same area recorded a few birds overwintering regularly, with a mean of 0.3 individuals/ 100 km (Sara et al. 1993). It is thus likely that, since the late 1980s, some birds formed a small group of year-long residents that, year after year, began to attract other individuals from migrating flocks in the spring. The area of breeding and overwintering, in central Si- cily, is today about 1000 km^ with a core area of ca. 700 km^. However, other records of this kite were reported from elsewhere on the island. The kite population is probably concentrated in the vicinity of the lacustrine and riverine system (Fig. 1; area A). The breeding area and the number of Black Kite pairs currently seem to be (1999—2001) rather stable, ca. 25—35 residents corre- sponding to 6-10 territorial pairs, plus a flock of 20-25 summer birds. The mean number of young fledged per successful pairs was 2.25 ± 0.96 (SE, N = 4) , a relatively high productivity level (Sergio and Boto 1999). System- atic exploration of area B (Fig. 1) is needed to complete a survey for the species on the island. During the winter 1990-91, one Black Kite was ob- served for the first time in a roost of Red Kites {Milvus milvus), which was located in the middle of the area A (Fig. 1). In the following years, a small group was regu- larly recorded in the Red Kite roost from early Septem- ber to late March (Fig. 2). The observation of juveniles of the year and second-year individuals in the vrinter roost was very recent (1999). Central Sicily is a suitable habitat for Black Kites, which usually select aquatic and open habitats for foraging and for breeding (Bijlsma 1997). This area has a large num- ber of artificial dams and large streams (Sosio, Platani, and Belice), open croplands, grazing, and some large Mediterranean mixed-oak woods. This suitable habitat lies along the migration route of the passage Black Kite flocks moving inland, and which cross the island more or less through its center en route to the southern coasts and later to pass over the Sicilian channel. I propose that after kites established this new migratory route in the early 1900s, some pairs found and began to breed in suit- able habitats (Mebs 1957, Massa 1980). Initially these kites were probably present only during spring and sum- mer; later some individuals presumably became resident and were recorded also during winter. The last step of the colonization process was the estab- lishment of a summer roost in a patch of pines {Pinus pinea) near a garbage dump (Fig. 1). Since 1998, a group June 2003 Short Communications 169 25 □ Summer roost Figure 2. Mean number of Black Kites recorded in two winter; area A, Fig. 1) in western Sicily (1992-2001). of non-breeders arrived in late spring and settled in this roost near the dump. The number of Black Kites record- ed in summer at the dump increased from a mean of 6.6 individuals to 19 non-breeding birds in 2001 (Fig. 3). In early spring, only adults were recorded. Numbers de- creased to 50% in the other seasons when the second- year individuals formed half of the summer population. roosts (garbage dump in summer and Red Kite roost in Young of the year hatched in the vicinity of the roosts were observed from late August to December (Table 1 ) . The maximum number of individuals was observed in late August, when also the residents and their young con- centrated in the summer roost. From very late August to early September most of the resident population leaves the area and departures continue until early October. In Figure 3. Mean number of Black Kites at the roosts in western Sicily (area A, Fig. 1) from November 1989-Septem- ber 2001. 170 Short Communications VoL. 37, No. 2 Table 1. Age class composition of Black Kites in western Sicily (area A, Fig. 1). Sample size is in parentheses including 74% of the total observed kites (N = 255) recorded during 24 visits between June 1998 and June 2001. Nov-Feb Mar-May JUN-JUL Aug-Oct Adults 0.50 (11) 1.00 (17) 0.50 (51) 0.51 (25) Second-year birds 0.36 (8) 0.00 (0) 0.50 (50) 0.33 (16) Young of the year 0.14 (3) 0.00 (0) 0.00 (0) 0.16 (8) this regard, I made two direct observations (October 1994 and August 2000) of the arrival and mixing between one flock of migrating Black Kites and the group of res- idents in the area. In August 2000, a flock of some 120 migrating kites was recorded to arrive in late afternoon and to settle in at the roosting site that already hosted 40 residents. A visit to the roost a week later revealed the presence of only 15 birds, and that decreased to four wintering birds in November. I suggest that the dump area is a stop-over site in central Sicily for migrating kites and that its presence may have facilitated the coloniza- tion process. It seems the first step of the colonization by Black Kites involved breeding, later followed by their presence in winter (breeders becoming residents) and hnally pres- ence in summer. This last step was probably favored by the return of birds originally bred in Sicily. Forero et al. (1999) have shown that Black Kites in Southern Spain are highly philopatric, with most of individuals coming back to breed near their natal site, but they suggested that this behaviour can vary depending on ecological conditions. A first phase of colonization involving the presence of singles or a few birds before the onset of breeding prob- ably occurred in the period when sporadic observations of this species were reported (Mebs 1957, Massa 1980). Thus the numerical increase recorded from summer 1999 would probably represent the second stage of the colonization process, involving the establishment of a sur- plus population. 25 - ■ Red kite □ Black kite Figure 4. Number of territorial Black and Red Kite pairs recorded in 1980 and 2000 in central Sicily, Italy. Apart from the process of recolonization following lo- cal extinction caused by human persecution or pesti- cides, relatively few status changes have been recorded in raptors (Newton 1979). For example, Newton described the sedentarization of Swainson’s Hawks {Buteo swain- soni) . Moreover, the regular year-long presence of the for- merly summer breeding Red Kite populations in France (Valet 1975), Switzerland (Juillard 1977), and Germany (George 1996 and references therein) is well known. These authors suggest the Red Kite sedentarization was related to the increasing occurrence of mild winters with- out snow that allowed accessibility to the local food sup- plies (e.g., more active prey, development of garbage dumps) . The hrst overwintering of Black Kite in central Sicily occurred during four years (1987-90) of severe drought and warm winters. In the same period, the wintering of Lesser Kestrels {Falco naumanni) in the area was also re- corded (Sara 1996a). The sedentarization of Black Kite may have been initiated by such climate change, and one can speculate that global warming, increasing the mean local winter temperatures in recent decades, may have favored the extension of the species’ wintering range northward. The colonization process since 1980 may have also been facilitated by the recent decline of the local Red Kite population (Fig. 4) . The ecological niche left empty by declining resident Red Kite pairs has likely been oc- cupied by the Black Kite, a new colonizer expanding its range northward. Human persecution and poor land management (e.g., sumner hres, poisoning, and shooting) resulting in di- rect kite mortality and lower prey populations are prob- ably the main factors currently limiting the density of Black Kites in Sicily (Sara 1996b) and, thus, slowing the colonization process in the last 50 years. Resumen. — Describe la colonizacion de los milanos ne- gros {Milvus migrans) en Sicilia. Los antiguos ornitologos del siglo IX consideraban esta especie como ocasional y rara en Sicilia. Gomunmente, el milano negro ha sido clasificado como la segunda especie mas comun obser- vada en el estrecho de Messina en primavera y es inclu- sive mas abundante durante la migracion de otoho. Hab- itats adecuados de ambientes acuaticos y de espacios abiertos estan disponibles a lo largo de la ruta migratoria del milano, lo cual pudo haber favorecido la coloniza- June 2003 Short Communications 171 cion. Esta especie se encuentra reproduciendose y per- maneciendo durante el invierno en un area aproximada de 100 km^ localizada en el centre de Sicilia. El primer registro reproductive confirmado fue reportado en 1979, luego prosiguio una colonizacion mas amplia. La estadia tardia fue registrada durante los censos vehiculares en 1987-92, estos milanos no fueron observados durante las investigaciones previas (1977-80). En la porcion central del area del area de estudio, un individuo fue observado entrando a un dormidero del milano rojo (Milvus milvus) durante el invierno de 1990-91 y un pequeno grupo de cuatro a siete milanos negros es regularmente observado ahora. La ultima fase del proceso de colonizacion fue el establecimiento de un dormidero de verano sobre un vertedero de basura, el cual aumento de una media de 6.6 al9 individuos no reproductivos en 1998-2001. La poblacion total hacia finales de Agosto es de 50-60 aves. [Traduccion de Cesar Marquez] Acknowledgments I thank all the local ornithologists, mainly Bruno Mas- sa, Massimiliano Di Vittorio, Salvatore Grenci, and Ame- deo Falci, who gave me access to their data and Laura Zanca for accompanying me in the field and for drawing Fig. 1. Two referees, Fabrizio Sergio and Manuela Forero, improved the paper with useful comments and critiques. Literature Cited Agostini, N. 1992. Spring migration of Honey Buzzard (Pernis apivorus) at the Strait of Messina in relation to atmospheric conditions./. Raptor Res. 26:93-96. , D. Lagozzo, and M. Panuccio. 2000. The island of Marettimo (Italy) , important bird area for the au- tumn migration of raptors. Avocetta 24:95-99. Benoit, L. 1840. Ornitologia Siciliana. Stamperia G. Fiu- mara Messina, Italia. Bijlsma, R.G. 1997. Black Kite. Pages 132-133 in W.J.M. Hagemeijer and M.J. Blair [Eds.], The EBCC atlas of European breeding birds, their distribution and abun- dance. T. & A.D. Poyser, London, U.K. Blond el, J. 1986. Biogeographie evolutive. Masson et Gie, Paris, France. Diamond, J.M. 1969. Avifaunal equilibria and species turnover rates on the Channel Islands off California. Proc. Natl. Acad. Sci. U.S.A. 64:57—63. . 1972. Biogeographie kinetics: estimation of relax- ation times for avifaunas of southwest Pacific Islands. Proc. Natl. Acad. Sci. U.S.A. 69:3199-3203. Dimarca, a. and C. Iapichino. 1984. La migrazione dei Falconiformi sullo Stretto di Messina. Primi dati e problemi di conservazione. Lega Italiana Protezione Uccelli, Parma, Italy. Doderlein, P. 1869-74. Avifauna del Modenese e della Sicilia. Giorn. Sci. Nat. Economiche, Palermo, Italy. . 1893. Avifauna Sicula. Rapaces. Atti R. Accad. Sci. Lett. Arti Palermo 2:1-33. Forero, M.G., J.A. Donazar, J. Bias, and F. Hiraldo. 1999. Causes and consequences of territory change and breeding dispersal distance in the Black Kite. Ecol- ogy 80:1298-1310. ForsmAN, D. 1999. The raptors of Europe and the Middle East. T. & A.D. Poyser, l^ondon, U.K. Galea, C. and B. Massa. 1985. Notes on the raptor mi- gration across the central Mediterranean. ICPB Tech Publ. 5:257-261. George, K. 1996. Trends of numbers wintering Red Kites Milvus milvus L. between Harz mountains and Hakel till 1994 Pop. Greifvogel Eulenarten SA99-205. Iapichino, C. and B. Massa. 1989. The birds of Sicily British Ornithologists’ Union. Check-list No. 11, Lon- don, U.K. JuiLLARD, M. 1977. Observation sur I’hivernage et le dor- toirs du Milan royal Milvus milvus (L.) dans le nord- ouest de la Suisse. Nos Oiseaux 34:41-57. Lo Valvo, M., B. Massa, and M. SarA. 1993. Uccelli e paesaggio in Sicilia alle soglie del terzo millenmo. Nat. Sicil. I7(suppl.):3-37l. MacArthur, R.H. and E.O. Wilson. 1967. The theory of island biogeography. Princeton Univ. Press, Prince- ton, NJ U.S.A. Massa, B. 1980. Ricerche sui rapaci in un’area-campione della Sicilia. Nat. Sicil. 4:59-72. . (Ed.). 1985. Atlas faunae Siciliae. Aves. Nat. Sicil. 9(n. speciale): 1-274. Mebs, T. 1957. Ornitologische Beobachtungen in Sizi- lien. Voge/weft 78:169-176. MinA PA1.UMBO, F. 1853. Catalogo degli uccelli delle Ma- donie. 1. Atti Accad. Sci. Let. Arti Palermo 2:1-32. . 1857. Catalogo degli uccelli delle Madonie. II Atti Accad. Sci. Lett. Arti Palermo 3:1-45. Newton, I. 1979. Population ecology of raptors. T. & A D. Poyser, Berkhamsted, U.K. Sara, M. 1996a. Wintering raptors in the central Medi- terranean basin. Pages 345-359 mj. Muntaner andj. Mayol J. [Eds.], Biologia y conservation de las rapaces Mediterraneas. Monografias No. 4. SEO, Madrid, Spain. . 1996b. Sicile: Statut des rapaces diurnes. Fond d ’Interventions pour les Rapaces 28:27-28. SarA, M., L. Zanca, G. Sorci, and B. Massa. 1993. Cen- simenti di rapaci in un’area campione della Sicilia. Pages 275-282 in M. Lo Valvo, B. Massa, and M. Sara [Eds.], Uccelli e paesaggio in Sicilia alle soglie del terzo millennio. Nat. Sicil. 17(suppl.):3— 371. Schembri, a. 1843. Quadro comparativo. Le Ornitologie di Malta, Sicilia, Roma, Toscana, Liguria, Nizza e la provincia di Gard. Tipografia Anglo-Maltese, Malta. Sergio, F. and A. Boto. 1999. Nest dispersion, diet, and breeding success of Black Kites {Milvus migrans) m the Italian pre-Alps. / Raptor Res. 33:207-217. Sergio, F., P. Pedrini, and L. Marchesi. 2002. Reconcil- ing the dichotomy between single species and ecosys- tem conservation: Black Kites {Milvus migrans) and eutrophication in pre-alpine lakes. Biol. Conserv. 110: 101 - 111 . 172 Short Communications VoL. 37, No. 2 Thiollay, J.M. 1977. Importance des populations de ra- paces migrateurs au Mediterranee occidentale. Alau- da 45:115-121. Valet, G. 1975. La sedentarization du Milan Royal Milvus milvus en Auxois. Alauda 43:263-269. ViNUELA, J. AND C. SuNYER. 1994. Black Kite Milvus mi- grans. Pages 148-149 in G.M. Tucker and M.F. Heath [Eds.], Birds in Europe: their conservation status. BirdLife Conservation Series No. 3. Cambridge, U.K. Received 29 November 2001; accepted 22 Eebruary 2003 Letters J. Raptor Res. 37(2) :173-174 © 2003 The Raptor Research Foundation, Inc. Gyrfalcon Color Variation A literature survey dealing with Gyrfalcons (Falco rusticolus) reveals that references to the various Gyrfalcon color variants as morphs is becoming more common (e.g., Sibley 2000, The Sibley guide to birds, Alfred A. Knopf, Inc., New York, NYU.S.A.). Such references often make the assumption that the three common Gyrfalcon color groups are comparable to the true red and gray morphs of the Eastern Screech-Owl {Otus asio) and to the blue and Snow Goose {Chen caerulescens) morphs. In my opinion, this is not correct. Ford (1945, Biol. Rev. 20:73-88), emphasized that morphs must first be phenotypically distinguishable with discon- tinuities in the phenotype. He also stated that polymorphism must not be used to include continuous or quasi- continuous variation. Huxley (1955, Acta Int. Ornithol. Congr. 11:309-328) proposed use of the term morphism to denote genetic poly- morphism opposed to other kinds of polymorphism (e.g., seasonal, cyclic, geographic). He referred to Ford’s defi- nition of genetic polymorphism as the coexistence in one interbreeding population of two or more sharply distinct and genetically-determined forms. Further, Huxley pointed out that the least abundant form should be present m numbers too great to be due solely to recurrent mutation, and proposed that such forms be called morphs. Huxley also suggested removal of the adjective ‘"sharply” from the definition to accommodate rare cases of continuous polymorphism. He argued that in continuous morphisms, such as in the sea-plantain {Plantago maritima), there are no sharp discontinuities, and the excessive genetic variability constituting the continuous polymorphism is maintained by selection. He stressed the fact that every morphism must involve some balance of selective advantage and disad- vantage, some underlying genetic basis, and that its discontinuities must be determined either genetically or devel- opmentally. The Gyrfalcon would be an example of continuous polymorphism if that were acceptable terminology. I suggest that the term continuous polymorphism is a contradiction that should be avoided. It has been established for many years that Gyrfalcon color variation extends in a continuum from white to almost black, although most birds are classed as either white, gray, or dark (Th.N. Krabbe 1934, Medd. Dan. Naturh. Foren 98:4—107; see Plates III to VI). In a study of 205 specimens collected from European Russia across to eastern Siberia, Ellis et al. (1996,/. Raptor Res. 26:81-88) found the same continuum. They classified the birds as color variants, and found that climate was a better predictor of color than latitude; the area with the highest percentage of white birds was eastern Siberia. Glum and Cade (1994, In A. Poole and E.B. Gill [Eds.], The birds of North America, No. 114. The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U.S.A.) state the stripe on white birds, when expressed, consists of a few dark streaks next to the shafts of the feathers, giving a kind of speckled appearance. Glum and Cade also asked “what level of plumage variation among morphs is justified?’’ This use of the term morph was an error (T. Cade pers. comm.). In North America, there is a coarse progression in the ratio of light birds to dark birds from northeast to west; birds of northern Greenland and the eastern high arctic are usually white, those of the central arctic are usually 50 50 white:gray, and those in the Yukon and Alaska are mainly gray, a few dark. At Ungava Bay, however, “the whitest and blackest varieties breed together in the same area along with every kind of intermediate’’ and there are white birds in western Alaska (Cade 1960, Univ. Calif. Publ. Zool. 63:151-290). In Eurasia there are no white birds in Eennoscandia and western Russia, but the occurrence of white birds increases eastward until they make up ca. 50% of the population in northeastern Siberia and Kamchatka (Cade et al. 1998,/. Birds Western Palearctic . In the Lower Kolyma, northeastern Siberia, pairs of white Gyrfalcons were limited to the maritime coastline with cliffs, whereas pairs of gray birds were limited to the timberline. The white Gyrfalcons were preying on ducks and waders, while the gray Gyrfalcons were eating grouse {Lagopus sp.). The hunting strategy of the two color variants was also different (E. Potapov pers. comm.). Because Gyrfalcons do not exhibit two or more distinct forms, the various Gyrfalcon color variants are not morphs according to Huxley’s definition. Why then do we see more and more references to Gyrfalcons as morphs in bird 173 174 Letters VoL. 37, No. 2 guides and scientific papers? The error stems from at least two sources. Huxley, apparently, was under the impression that all Gyrfalcons fell into one of three color categories; therefore, he referred to them as morphs. Second, some authors have expanded the original definition of morph. “Morph — term used for recognizably different forms of a species, usually color related. Color morphs are dark, rufous, and light.” (Wheeler and Clark 1995, A photographic guide to North American raptors, Academic Press, San Diego, CA U.S.A.). Frank Beebe (pers. comm.) agrees that the concept of three Cyrfalcon morphs is misleading to some ornithologists and birdwatchers. For example, at some locations white Cyrfalcons have all-white tails, while at other locations they have conspicuously barred tails. A few field guides refer to the white, gray, and dark phases of the Cyrfalcon. For some the term phase implies a change with time, and sometimes it refers to synchronization; therefore, its use with reference to Cyrfalcons is not appropriate. It is akin to referring to blue phase Rock Doves ( Columba livia) . What is the appropriate terminology for Cyrfalcons? Cade (pers. comm.) now prefers color variants. Thomson (1964, [Ed.], a new dictionary of birds, McCraw-Hill, New York, NY U.S.A.) suggested the term form, and provided the following definition: “Form, in taxonomy, a loose or deliberately neutral term for a species or sub-division thereof, non-committal as regards rank or status to be assigned to it.” This problem in semantics is best resolved by understanding the causes for the unusual color variation. Cade (pers comm.) hypothesized that the three broad color groups are a result of geographic isolation in Pleistocene refugia during the past 100 000 years or so. He believes that the white birds evolved in isolation in the Ellesmere Island- north Greenland refugium. A melanistic form may have evolved in another refugium around Labrador; while the original or ancestral gray gyrs were restricted to habitats south of the continental ice sheets in North America and Eurasia. When the ice retreated, the breeding ranges of the various forms merged and, because no reproductive isolating mechanisms had evolved, the three color variants (originally geographic in origin), have freely interbred to produce the present distribution and range of variants we see in the Gyrfalcon’s plumage. In any case, it is clear that there is a color continuum in Gyrfalcons, and not in screech owls, and that the term morph is being used for both patterns of morphological variation. I suggest the term morph is most appropriate for discontinuous variants and not the Cyrfalcon. I would like to thank T.J. Cade for his assistance over the years, and E. Potapov and referees D. Bird, A. Jenkins, and P. Koskimies for their helpful comments. — Ian Flann, 1067 Wiseman Crescent, Ottawa, Ontario, KIV 8J3, Canada. Received 12 December 2001; accepted 11 January 2003 / Raptor Res. 37(2): 174-1 76 © 2003 The Raptor Research Foundation, Inc. Two White-tailed Sea Eagles {Haliaeetus albicilla) Collide with Wind Generators in Northern Germany Recently, the issue has arisen whether wind generator structures may have a significant impact on bird populations The sustainable use of wind energy has led to a substantial increase in the number of wind-power plants in Germany. In the last 5 yr, the number of wind turbines has doubled. In the year 2000, wind power plants generated more than 6113 MW power, or 2.4% of the total energy power consumption in Germany. Especially windy areas in the northern parts of Germany, where large numbers of waders and water birds live, are favored by this development. Studies on these bird groups revealed that wind power plants have a substantial effect on the behavior of these birds through disturbance, harassment, and loss of habitat, rather than the direct mortality due to collisions (Exo 2001, Natur und Landschaftsplanung 33:323) . The White-tailed Sea Eagle {Haliaeetus albicilla) is still listed as a threatened species in the red data book of Germany (Witt et al. 1998, In: Binot et al. Rote Liste gefdhrdeter Tiere Deutschlands, Bundesamt fur Naturschutz, 40-47), although the population has doubled in the last 10 yr, reaching 381 territorial pairs in 2001. The core population that has June 2003 Letters 175 Figure 1. Map of the Baltic Sea with its surrounding states. Inset showing strike locations; i.e., wind power plants (asterisks) and breeding sites (closed circles) in northeastern Germany. supported the increase and expansion of this eagle population lies within the federal state of Mecklenburg-Western Pomerania in northwestern Germany with 174 territorial pairs in 2001 (Kollmann et al. 2002, Corax 19:1—14). A White-tailed Sea Eagle was found dead in a wind power plant in northeastern Germany in the federal state of Mecklenburg-Western Pomerania on 26 January 2002. The park consisted of two rows with 20 single turbine towers, each 65 m tall with a rotor of 41 m in diameter (maximum height = 85 m) and is located north of the town of Wolgast in the coastal area of the Baltic Sea (Fig. 1). Prior to the collision, it was very windy, with mean wind speeds between 30 and 40 km/hr and maximum speeds of 90 km/hr. Necropsy at the Institute for Zoo and Wildlife Research (IZW), Berlin, revealed a displacement of the thoracic vertebral column with fractures of two vertebrae and several broken ribs on both sides. The adult female eagle was in very good body condition and no signs of a disease were identified. The accident may have been caused by a squall, which took the eagle into the propeller while it was hunting. Another White-tailed Sea Eagle was direcdy observed when hit by a propeller of a wind generator in a plant consisting of seven generators of the same type mentioned above and located in one row in Mecklenburg-Western Pomerania near the town of Anklam, also in the coastal area of the Baltic Sea on 1 April 2002 (Fig. 1). Residents walked near this wind power plant in sunny weather (17-18°C) when they heard a dull noise at 0920 H GET; they turned and saw the eagle falling to the ground. No behavior to avoid the strike by the rotor was observed, because the event was first recognized when the sound of the collision was heard. The eagle was a subadult 4-yr-old male and it was brought injured to a veterinarian who diagnosed a multiple fracture of the right radius and ulna. The veteri- narian attempted to align the bones, but was unsuccessful. Subsequently, the bird was euthanatized due to the development of the severe osteomyelitis in the right ulna and radius. The eagle was ringed in June 1998 in an eyrie on the island of Usedom ca. 25 km northeast of the location where the bird collided with the wind turbine. This is the first report of White-tailed Sea Eagles being killed by wind turbines. Previous studies on causes of death of White-tailed Sea Eagles in Germany did not mention casualties by wind generators (Oehme 1966, Falke, 13:40-47; Struwejuhl and Latendorf 1997, Vdgelwelt, 118:95-100; Krone et al. in press, Proc. Conf. Sea Eagle 2000, Bjorko, Sweden). The northeastern part of the federal state of Mecklenburg-Western Pomerania borders on the Baltic Sea, with waters draining from the island of Usedom and the Peene stream from the mainland in close proximity. The waters are rich in fish and waterfowl, providing a food source for several breeding pairs of eagles, and supporting high numbers of migrating and wintering individuals. The landscape largely consists of cultivated areas interspersed with patches of deciduous and mixed forests. The climate of the region is strongly influenced by the Baltic Sea resulting in a delayed spring, relatively cold and wet summers and mild winters with little snow. The mean temperature of the year is 8.1°C with a mean rain fall of 584 mm. The preceding wind direction is southwest. Incidents such as these are likely to occur in areas with large raptor populations and a high prey base for raptors. Estep (1989, Unpubl. Rep. California Energy Commission, Sacramento, California) documented collisions with wind turbines by 72 raptors of seven species at two Wind Resource Areas (WRA) in California (Altamont Pass and Tehachapi Pass) . The Golden Eagle (Aquila chrysaetos) and the Red-tailed Hawk {Buteo jamaicensis) were the most frequently reported species, comprising 86% of the collision incidents. Fatalities were reported year round, affecting local as well as migrating and wintering birds. During a 4-yr investigation of 179 radio tagged Golden Eagles, 61 deaths were re- corded, of which 23 were caused by wind turbine blade strikes. Non-territorial Golden Eagles often visited the WRA, 176 Letters VoL. 37, No. 2 whereas breeding eagles rarely entered the WRA (Hunt et al. 1999, Report to the National Renewable Energy Lab- oratory, XAT-5-13174-01, XAT-6-1 6459-01, Predatory Bird Research Group, University of California, Santa Cruz, CA U S.A.). However, a review of different sources of avian collision mortality in the United States indicated that death associated with wind plants is much lower than other sources of collision mortality (Erickson et al. 2001, National Wind Coordinating Committee, Washington, DC U.S.A.). The high levels of raptor mortality at Altamont were ex- plained by large raptor populations, a high prey base for raptors, and the large size of the wind plant. A study performed in the Campo de Gibraltar region in Spain, a major passway of bird migration to Africa, also demonstrated the severe impact of a wind farm on large birds (Montes and Jaque 1995, Summary of final report, Soc. Espan Ornitol.). Of 82 birds found after collisions with wind generators, five raptors were affected, of which the Griffon Vulture ( Gyps fulvus) was most common (43 collisions) . In contrast to the wind power facility in California, the wind power plants in Mecklenburg-Western Pomerania are much smaller, but this is an important area for migrating and wintering White-tailed Sea Eagles in the Baltic Sea region. Systematic studies on wind turbines to examine their full range effects on behavior, reproductive success, and mortality of raptors are strongly needed for Germany, as there are plans to enlarge the total energy generated by wind in Germany in the near future. We are grateful to E. Seemann from the Mueritz Museum Waren and to the editor and anonymous referees for their helpful suggestions. — Oliver Krone, Institute for Zoo and Wildlife Research, P.O. Box 601103, D-10252 Berlin, Germany; e-mail address: krone@izw-berlin.de and Christian Scharnweber, State Office for Environment and Nature, Dorfstr. 86, D-17392 Putzar, Germany. Received 3 July 2002; accepted 16 March 2003 J. Raptor Res. 37(2) :176 © 2003 The Raptor Research Eoundation, Inc. Talon-locking in the Red-tailed Hawk Talon-locking by two Red-tailed Hawks {Buteo jamaicensis) in flight is widely accepted (e.g., Ferguson-Lees and Christie 2001, Raptors of the world, Houghton Mifflin, Boston, MA and New York, NYU.S.A.), although it appears to have been well described in the literature only once (Warren 1890, Report on the birds of Pennsylvania, 2nd Ed., Harrisburg, PA U.S.A.), who observed it during fall migration. He was cited by Bent (1937, Life histories of North American birds of prey, Part 1, U.S. Natl. Mus. Bull. 167, Washington, DC U.S.A.) and Palmer (1988, Handbook of North American birds, Vol. 5, Diurnal raptors, Yale Univ. Press, New Haven, CT U.S.A.), who also briefly described the behavior seen in the spring, which he interpreted as one bird being a territory holder and the other an interloper Palmer was cited by Preston and Beane (1993, in A. Poole and F. Gill [Eds.], The birds of North America, No. 52 The Academy of Natural Sciences, Philadelphia, PA and The American Ornithologists’ Union, Washington, DC U S.A.). Voelker (1969, Loon 41:90-91) witnessed courting birds locking bills in flight and falling to the ground, and he quotes another observer who reported courting birds locking bills or feet and falling to the ground. Both of those observations were made in March. On 24 January 2000, two Red-tailed Hawks {Buteo jamaicensis calurus) were recovered under an electrical transformer near the entrance to the National Guard Armory in Santa Fe, New Mexico by the Public Service Company of New Mexico. They had been electrocuted and partially burned; all four of their talons were locked together. One was m normal definitive plumage, the other was a dark-phase bird with a sub-definitive banded tail. Both birds were females with ovaries measuring 8 X 20 and 8X15 mm, respectively. They were too burned to be preserved. This incident of talon-locking between female Red-tailed Hawks occurred in January, suggesting aggressive inter- action, in which one hawk was attempting to displace another hawk that was perched on the pole with a transformer. With talons locked the two birds made contact with two wires resulting in electrocution. I wish to thank Anne Sanchez of the Public Service Company of New Mexico for presenting these birds to the Museum of Southwestern Biology, and the reviewers of this note for their helpful suggestions. — Robert W. Dickerman, Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM 87131 U.S.A.; e-mail address: bobdickm@unm.edu Received 15 July 2002; accepted 23 March 2003 A Telemetry Receiver Designed with The Researcher in Mind What you've been waiting for! fiiuUy, a hi^tUy 9^9 i4.icmetiy (eceivci ihit ivtlQhs tcu thjn 13 ouncn. is completdy user programmable and offers variable scan rates over all frequencies. For each animal being tracked, the large ICO display provides not only the frequency (to 100Hz) 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 Si 500. The features and performance of the new R-1000 pocket sized telemetry receiver will impress you. and the price will convince you. Othtr Features Include: ^ • Factory tuned to any 4MHz wide segment in the 148-174MHz Band • Very high sensitivity of 'laBdBm to -ISOdBm • Illuminated display and keypad for use in low light or darkness • User selectable scan rates from 1-SO seconds in 1 second steps • Rechargeable batteries operate the receiver for 12 hours and can be replaced with standard AA Alkaline batteries in the held. Both I2vdc and llOvac chargers arc included. • 6.r (15.5cm) high. 2.6" (6.6cm) wide. 1.5" (3.8cm) deep. • 3 year warranty • 1 day deUvery $ 695.00 Please specify desired 4MHz wide segment in the 1AB-174MHZ band Visit our website for complete specifications, operating manual and information on the R-IOOO or call our toll-free number to ofdiM your receiver now Try the New R-1000 and Youll 9 # Impressed! COMMUNICATIONS SPECIALISTS. INC iV.-'.t I.'f; ' otn.' IJ S.A. fHOf)) Oijr ,)•, (, A d.-’roS •, • i ■’ 1 •• • 9^8 • oV i • •.)* 1 7 ; '<7. i.V'...7 • H'’/.' nS'.7 • t'ittt' .so C'ni ,( o'Ti -.c'O Buteo Books 3130 Laurel Road 800-722-2460 allen@buteobooks.com Shipman, VA 22971 434-263-8671 www.buteobooks.com BUTEO BOOKS is the largest retailer of Ornithology titles in North America, with over 2000 in-print titles in stock. In-Print Falconry and Birds of Prey Handbook of the Birds of the World. The first seven volumes of this projected sixteen- volume work have been published, including Volume 2, covering the diurnal raptors, and Volume 3, covering owls. They are priced at $185 each. The Raptor Almanac: A Comprehensive Guide to Eagles, Hawks, Falcons, and Vultures. Scott Weidensaul. Lyons Press, 2000. 382 pp. Color photos. Cloth. $40.00 The Harris Hawk: Management, Training and Hunting. Lee W. Harris. Swan Hill Press, 2001. 144 pp. Color and b&w photos and illustrations. Cloth. $32.95 Rare and Out-of-print Falconry Usually available from Buteo Books, the classic reference on diurnal raptors: Brown & Amadon: Eagles, Hawks and Falcons of the World. First English edition, 1968. $300. Birds of North America series individual species accounts for over 500 species, including Swallow-tailed Kite White-tailed Kite Snail Kite Mississippi Kite Bald Eagle Northern Harrier Sharp- shinned Hawk Cooper’s Hawk Northern Goshawk Common Black-Hawk Harris’ Hawk Red- shouldered Hawk Broad-winged Hawk Swainson’s Hawk White-tailed Hawk Zone-tailed Hawk Hawaiian Hawk Red-tailed Hawk Ferruginous Hawk Crested Caracara American Kestrel Merlin Gyrfalcon Prairie Falcon Bam Owl Flammulated Owl Eastern Screech-Owl Whiskered Screech-Owl Great Homed Owl Snowy Owl Northern Hawk Owl Northern Pygmy-Owl. Fermginous Pygmy- Owl Elf Owl Burrowing Owl Spotted Owl Barred Owl Great Gray Owl Long-eared Owl Short-eared Owl (R) Boreal Owl Northern Saw- whet Owl Available soon: American Kestrel California Condor 2003 ANNUAL MEETING The Raptor Research Foundation, Inc. 2003 annual meeting will be held on 3—7 September 2003 in Anchorage, Alaska. For information about the meeting see the following website: http://www. alaskabird.org or contact Alaska Bird Observatory (birds@alaskabird.org). Persons interested in predatory birds are invited to join The Raptor Research Foundation, Inc. Send requests for information concerning membership, subscriptions, special publications, or change of address to OSNA, P.O. Box 1897, Lawrence, KS 66044-8897, U.S.A. The Journal of Raptor Research (ISSN 0892-1016) is published quarterly and available to individuals for $33.00 per year and to libraries and institutions for $50.00 per year from The Elaptor Research Foundation, Inc., 14377 1 17th 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 6604T8897, U.S.A. Printed by Allen Press, Inc., Lawrence, Kansas, U.S.A. Copyright 2002 by The Raptor Research Foundation, Inc. Printed in U.S.A. 0 This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Raptor Research Foundation, Inc., Awards Lifetime Achievement Awards The Tom Cade Award recognizes an individual who has made significant advances in the area of captive prop- agation and reintroduction of raptors. Nomination packets can be submitted at any time. Contact: Brian Walton, Predatory Bird Research Group, Long Marine Laboratory, University of California, Santa Cruz, CA 95064 U.S.A.; tel. 408-459-2466; e-mail; walton@cats.ucsc.edu. The Fran and Frederick Hamerstrom Award recognizes an individual who has contributed significantly to the understanding of raptor ecology and natural history. Nomination packets can be submitted at any time. Con- tact: Dr. Clint Boal, Texas Cooperative Fish and Wildlife Research Unit, BRD/USGS, Texas Tech University, 15th Street & Boston, Ag Science Bldg., Room 218, Lubbock TX 79409-2120 U.S.A.; tel. (806) 742-2851; e-mail; cboal@ttacs.ttu.edu. Student Recognition and Travel Assistance Awards The James R. Koplin Travel Award is given to a student who is the senior author and presenter of a paper or poster to be presented at the RRF annual meeting for which travel funds are requested. Contact: Dr. Patricia A. Hall, 5937 E. Abbey Rd. Flagstaff, AZ 86004 U.S.A.; tel. 520-526-6222; e-mail: pah@spruce.for.nau.edu. Application Deadline: due date for meeting abstract. The William C. Andersen Memorial Award is given to the students who are senior authors and presenters of the best student oral and poster presentation at the annual RRF meeting. Contact: Laurie Goodrich, Hawk Mountain Sanctuary, 1700 Hawk Mountain Road, Kempton, PA 19529 U.S.A.; tel. 610-756-6961; email: goodrich@hawkmountain.org. Application Deadline: due date for meeting abstract; no special application is needed. Grants For each of the following grants, complete applications must be submitted to the contact person indicated by 15 February. Recipients will be notified by 15 April. The Dean Amadon Grant for $200-400 is designed to assist persons working in the area of distribution and sys- tematics (taxonomy) of raptors. Contact: Dr. Carole Griffiths, 251 Martling Ave., Tarrytown, NY 10591 U.S.A.; tel. 914-631-2911; e-mail: cgriff@liu.edu. The Stephen R. Tully Memorial Grant for $500 is given to support research, management, and conservation of raptors, especially to students and amateurs with limited access to alternative funding. Contact: Dr. 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; jefflincer@tns.net.