(ISSN OH 92-1016) The OF Raptor Research Volume 26 March 1992 Number 1 Contents Letter 1 Reproductive Parameters for Free Ranging American Kestrels {Falco spar verius) Using Nest Boxes in Montana and Wyoming. Anne Hoag Wheeler 6 Observations on the Behavior of Surplus Adults in a Red-shouldered HAWK Population. Michael D. McCrary, Peter H. Bloom and Marjorie J. Gibson 10 Home Range, Habitat Use and Behavior of Prairie Falcons WINTERING IN EAST-CENTRAL COLORADO. Gary Beauvais, James H. Enderson and Anthony J. Magro 13 Northward Post-fledging Migration of California Bald Eagles, w. Grainger Hunt, Ronald E. Jackman, J. Mark Jenkins, Carl G. Thelander and Robert N. Lehman 19 Determining Sex of Eastern Screech-Owls Using Discriminant FUNCTION Analysis. Dwight G. Smith and Stanley N. Wiemeyer 24 Diet Shifts of Black-chested Eagles (Geranoaetus melanoleucus) from Native Prey to European Rabbits in Chile. Eduardo F. Pavez, Christian A. Gonzalez and Jaimie E. Jimenez 27 Short Communications Methods of Locating Great Horned Owl Nests in the Boreal Forest. Christoph Rohner and Frank I. Doyle 33 Food Habits of the Short-eared Owl (Asio flammeus ) in Southern South America. Jaime R. Rau, Marcelo C. Villagra, Marta L. Mora, David R. Martinez and Maria S. Tilleria 35 Golden Eagles Feeding on Fish. Bryan T. Brown 36 Greater Yellow-headed Vulture (Cathartes melambrotus) Locates Food by Olfaction. Gary R. Graves 38 Letters 40 Thesis Abstracts 44 News and Reviews 47 The Raptor Research Foundation, Inc. gratefully acknowledges a grant and logistical support provided by the University of Saskatchewan to assist in the publication of the journal. ******************** 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 Jim Fitzpatrick, Treasurer, 12805 St. Croix Trail, Hastings, Minnesota 55033, U.S.A. The Journal of Raptor Research (ISSN 0892-1016) is published quarterly and available to individuals for $24.00 per year and to libraries and institutions for $30.00 per year from The Raptor Research Foundation, Inc., 12805 St. Croix Trail, Hastings, Minnesota 55033, U.S.A. Add $3 for destinations outside of the continental United States. Second class postage paid at Hastings, Minnesota, and additional mailing offices. Printed by Allen Press, Inc., Lawrence, Kansas, U.S.A. Copyright 1992 by The Raptor Research Foundation, Inc. Printed in U.S.A. THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. Vol. 26 March 1992 No. 1 J. Raptor Res. 26(1): 1-5 © 1992 The Raptor Research Foundation, Inc. Letters RRF — The First Quarter Century “The idea of some sort of organization of persons interested in raptors in a broad sense had been talked about for some time. Existing organizations which include concerned people were considered too general (ornithological societies) or too specialized (falconry associations). It was felt that an organization devoted exclusively to the birds of prey would be desirable. It was also felt that the number of concerned people was getting larger and that the changes going on in the status of many species of birds of prey and their environment justified an organizational approach.” 1 The “Raptor Research Foundation” first met on 2 September 1965 in Madison, Wisconsin, with 12 members from three countries attending. 2 On 9 February 1966 a “Certificate of Incorporation” 3 was issued in Turner County, South Dakota, and thus an organization called The Raptor Research Foundation, Inc. (RRF) was born. In 1991 the organization had been in existence for a quarter century, and so I am including that time period in this annual President’s Report. According to Article 8 of that 1 Vz page document, the Certificate of Incorporation, “the number of directors constituting the initial board of directors shall be three; and the names . . . are as follows: Donald V. Hunter, Jr., Paul F. Springer, Byron E. Harrell.” We hereby acknowledge the foresight of these three gentlemen, the official founding fathers. We might say that the organization has digressed, for it had an early “corporate headquarters” (Fig. 1) but no longer has one. According to “Article 3” the stated purpose was listed as, “The purposes shall be to stimulate, coordinate, direct and conduct research in the biology and management of the birds of prey, and to promote a better public understanding and appreciation of the value of these birds.” While the stated purposes have changed somewhat from the stimulating, coordinating, directing, and conducting of research to the disseminating of information on birds of prey, the traditional role of research within the organization has persisted as evidenced by the healthy number of scientific papers given at each of the annual conferences and the research supported by grant monies given out by RRF. Documents produced by the RRF started with the Raptor Research News (Volumes 1-5). In 1967 there was a “minimum contribution” 4 of SI. 00 for receiving Raptor Research News which was raised to $2.00 to meet the cost of production of this mimeographed publication. As the News was issued quarterly, this meant that the cost to a member was 25^ per issue. The production cost now has risen to $5.50 per issue and the Journal is, of course, printed by Allen Press, Inc., one of the leading publishers of scientific documents in the world. The Journal has changed from News to a refereed scientific publication. It is still, however, edited by an unsalaried editor. Raptor Research News became Raptor Research (Volumes 6-20), and other publications have included or do include The Journal of Raptor Research (Volumes 21 to present), Raptor Research Reports (Numbers 1-8), Raptor Research Abstracts (Volumes 1 [ 1 ]— 1 [2]), Breeding Project Information Exchanges (Numbers 1-128), The Kettle (even-numbered years since 1984), other information exchanges (17 in all), reports and supplements that have disseminated information in the printed word All of these documents have been indexed by Olendorff. 5 With regard to the major publication of RRF, The Journal 1 D.V. Hunter, Jr. and B.E. Harrell. 1967. A short history of the Raptor Research Foundation. Raptor Research News 1(1) 1-3. 2 Raptor Research News. 1967. The First General Meeting, September 2, 1965, Madison, Wisconsin. Raptor Research News 1(1):3 - 3 Certificate of Incorporation. Filed 9 February 1966 in Turner County, SD. 4 D.V. Hunter, Jr. and B.E. Harrell. 1967. Editor’s notes. Raptor Research News 1 (4): 1 . 5 R.R. Olendorff. 1989. The Raptor Research Foundation, Inc. bibliographic index (1967-1986). Raptor Research Report No. 7. 1 2 Letters Vol. 26, No. 1 Figure 1. The early “corporate headquarters” of The Raptor Research Foundation, Inc. This Vermillion (South Dakota) office was an old military facility donated to the University of South Dakota and used by Byron Harrell to conduct early Foundation business. oj Raptor Research , Mike Collopy, RRF’s Vice President, was appointed Chairman of an ad hoc Committee on Publication Policy and under his excellent leadership a very hard-working committee sent out a lengthy survey form and, based on a very good response from you, the members, drew up some recommendations which were adopted by the Board of Directors and the Regular Members at the general business meeting in Tulsa, Oklahoma. You have already seen one change in the Journal, the publishing of the Annual Conference Abstracts so that those who cannot attend the Annual Conference can still keep in touch with what research is being reported at the conference. There are more changes planned. The membership has grown from 12 members from three countries 6 to about one thousand, including subscription memberships (Fig. 2 and Table l) 7 from about 40 countries, all sharing an interest in birds of prey. The number of directors on the board has grown from the original three to twelve who are presently from Canada, Chile, England, Mexico and the United States. When it was suggested at the 1982 Annual Conference in Salt Lake City that perhaps The Raptor Research Foundation should add a directorship to represent members living outside of Canada and the United States, there was some pessimism that it probably would not work because of the difficulty for Directors from other lands to travel to the United States or Canada; it was agreed to try it. When the expansion of the Board of Directors from six to twelve was discussed at the 1989 Conference in Veracruz, Mexico, there was also some pessimism about the proposed board of twelve being too large to be workable. At the very successful meeting in Tulsa, Oklahoma 8 (1991), there were board members from the five countries mentioned above, and eleven of the twelve board members making up the current Board of Directors were actively involved in conducting the business of the RRF. Until the 6 Raptor Research News. 1967. The First General Meeting, September 2, 1965, Madison, Wisconsin. Raptor Research News 1 ( 1 ): 3 . 7 Data from RRF Treasurer, J. Fitzpatrick, as of 14 February 1992. 8 J-K. Schmutz. 1991. 1991 Annual Meeting. / Raptor Research 25(4):148-149. March 1992 Letters 3 Table 1. RRF membership for 1991 by country. Country No. of RRF Members Country No. of RRF Members United States of America 748 Philippines 2 Canada 61 Republic of China 2 England 26 Sweden 2 Germany 19 Zimbabwe 2 Spain 15 Columbia 1 Mexico 13 Cuba 1 Australia 12 Czechoslovakia 1 Japan 8 Denmark 1 Italy 7 Ecuador 1 Norway 7 Hong Kong 1 New Zealand 6 Iceland 1 Scotland 6 Israel 1 South Africa 5 Namibia 1 Argentina 5 Netherlands 1 Brazil 5 Poland 1 France 5 Portugal 1 Chile 4 Tanzania 1 USSR (former) 4 United Arab Emirates 1 Switzerland 3 Yugoslavia 1 Finland 2 Total 986 Greece 2 1988 conference in Minneapolis, Minnesota, only one Raptor Research conference had been held outside of the USA, and that was the 1981 conference in Montreal, Canada. But in 1989 that changed and the RRF met in Veracruz, Mexico. What has been the result of that meeting in Mexico? Look at your last issue of the Journal and you will note that of the 12 articles reported, there was 1 from Paraguay, 1 from Chile, and 3 from Mexico. Increased membership within the RRF from Latin American countries has grown considerably. I just received a letter that had been sent to Gary Duke urging RRF to consider Costa Rica as a site for a conference, and Gary had penciled on a note saying he thought we should plan a meeting there! The Annual Conference Guidelines Committee under the chairmanship of Michael Kochert has been busy and the 1992 conference will be in Bellevue, Washington, on 10-15 November. There will be an all-day symposium on the Spotted Owl and another symposium on the Burrowing Owl. The 1993 meeting will be in Charlotte, North Carolina, so how about 1994? At the urging of, and the efforts of, At-Large Board Member Robert Kenward, RRF will sponsor a meeting in the U.K., along with The Hawk and Owl Trust, in the second week of September 1993. This is a very brief look at a few of the activities of the RRF over the first quarter of a century. How about the future? At this point I shall diverge from the traditional President’s Report and ask the members to contemplate what they think the next 25 years will bring to the world of raptors! What will the raptor world be like in the year 2017? In order to stimulate your thinking I shall offer some thoughts of my own. I shall use some information about the past to “project” to the future. Two major factors that I see directly affecting what that world will be like are a) the ever increasing human population and b) the increasing consumerism within that human population. Since 1966 the human population has grown from 3 336 000 000 to over 5 333 000 000. 9 Examining briefly the interrelationship between humans and raptors in a general way we can see (Fig. 3) 10 that raptors have suffered disproportionately with regard to endangerments. 9 United Nations. 1990. Demographic yearbook. U.N. Department of International Economic and Social Affairs, Statistical Office, New York. 10 Compiled from several sources but mainly G. Nilsson. 1986. The endangered species handbook. The Animal Welfare Institute, Washington, DC. 4 Letters Vol. 26, No. 1 Figure 2. Membership history for The Raptor Research Foundation, Inc., for its first quarter century. That humans have been either the direct or indirect causes of these conditions can be seen by looking at the documentation for the demise of the species that have become extinct or that are teetering on the edge. This depauperization of raptor populations by humans is not confined to raptors but goes to flora and fauna in general. It has been projected that “by the end of this century, species will be vanishing at the rate of 100 per day, due largely to the destruction of tropical rain forests.” 11 Put another way, “researchers estimate that as many as 25 percent of all species inhabiting earth in the mid-1980s will have disappeared by 2015 if current deforestation trends hold.” 12 As just mentioned, it is proceeding at a most rapid rate in the tropics. Consider the fact that “although closed tropical forests cover only 7% of the earth’s land surface, they contain at least half — and possibly up to 90% — of the world’s species. It is impossible to estimate the exact proportion of species that reside there, since so little is known about tropical organisms and ecosystems Perhaps only 10% of all tropical species have been described to date. Current estimates place about half of all vascular plant and vertebrate species in tropical forests, and among invertebrates, the percentage may be considerably higher. Species diversity in temperate forests differs strikingly from that in tropical forests. Typically, 40-100 species of trees occur on a single hectare of tropical rainforest, compared to only 10-30 on a hectare of forest in the eastern United States. Some 700 species of trees can be found in 10 ha in Borneo, the same as in all of North America. One tree in Peru may contain as many species of ants as occur in all of the British Isles.” 13 Extinctions have gone on previous to man, of course, but “humanities’ impact on species extinction rates goes back thousands of years, but over the last century — especially over the last several decades — the human factor has increased dramatically. For instance, among all birds and mammals, we would expect an extinction only once every 100 to 1000 years in the absence of humans. However, the actual extinction rate for birds and mammals between 1850 and 1950 was one per year — as much as 1000 times greater than the background rate. Predicting future extinction rates in response to the massive habitat disruption that accompanies deforestation is difficult at best. However, a useful rule of thumb is that if a habitat is reduced by 90% in area, roughly half of its species will be lost.” 14 " L. Kaufman. 1987. Why the ark is sinking? Pages 1-41 in L. Kaufman and K. Mallory [Eds.], The last extinction MIT Press, Cambridge, MA. 12 K.R. Miller, W.V. Reid and C.V. Barber. 1991. Deforestation and species loss: responding to the crisis. Pages 78- 111 in J.T. Mathews [Ed.], Preserving the global environment: the challenge of shared leadership. The American Assembly, Columbia University. World Resources Institute, Washington, DC. 13 Ibid. 14 Ibid. March 1992 Letters 5 Raptor Endangerments Observed 'Expected" Non-raptorial Birds Diurnal Raptors Nocturnal Raptors Figure 3. The numbers of raptor species that are currently endangered as compared to birds in general. This shows that raptors are especially vulnerable to endangerment resulting from their need for large home ranges and especially susceptible to the results of biomagnification. “Overconsumption by the world’s fortunate is an environmental problem unmatched in severity by anything but perhaps population growth . . . measured in constant dollars, the world’s people have consumed as many goods and services since 1950 as all previous generations put together.” 15 That the “haves” contribute disproportionately to the environmental degradation caused by overconsumption, as compared to the “have nots” is attested to by the statement that “about 1 billion people do most of their traveling, aside from the occasional donkey or bus ride, on foot, many of them never going more than 100 kilometers from their birthplaces.” 16 Hopefully none of the “haves” would object to the sharing with the “have nots” those basic needs for survival beyond that of subsistence level existence but on the upper end of the scale for the “haves” it should be noted that “beyond the environmental costs of acquisitiveness, some perplexing findings of social scientists throw doubt on the wisdom of high consumption as a personal and national goal: rich societies have had little success in turning consumption into fulfillment. Regular surveys by the National Opinion Research Center of the University of Chicago reveal, for example, that no more Americans report they are ‘very happy’ now than in 1957. The share has fluctuated around one-third since then, despite a doubling of personal consumption expenditures per capita.” 17 What lies ahead? I recall vividly an evening that I spent talking with William C. Dilger. Bill was into wine making and so we were sampling some of the wines that he had made. Our conversation turned to where the world was heading, with regard to the human population, and he suggested an analogy. In making wines, he related to me, you determine whether you get a dry wine or sweet wine by the amount of sugar that is in the mixture. With an excess of sugar the yeast organisms multiply rapidly and when the population reaches a point where the metabolic wastes of the very large numbers start killing off the organisms, they will all eventually be killed from the metabolic waste and the remaining excess remains and the resulting wine is a sweet one. If there is 'insufficient sugar in the mixture to allow the population to reach a point where metabolic wastes kill off the population those organisms thrive to the point of all of the sugar being utilized, the organisms then die of “starvation” and the resulting wine is a dry one. Which do you prefer? A sweet wine or a dry one? — Richard J. Clark, President. 15 A. During. 1991. Asking how much is enough. Pages 153-169 in State of the world 1991 — a Worldwatch Institute Report on progress toward a sustainable society. W.W. Norton and Go., New York. 16 Ibid. 17 Ibid. J Raptor Res. 26(1 ):6-9 © 1992 The Raptor Research Foundation, Inc. REPRODUCTIVE PARAMETERS FOR FREE RANGING AMERICAN KESTRELS {FALCO SPAR VERIUS) USING NEST BOXES IN MONTANA AND WYOMING Anne Hoag Wheeler 1 Department of Zoology and Physiology, University of Wyoming, Laramie, WY 82071 Abstract. — An average of 37 nest boxes were monitored annually for use by American Kestrels ( Falco sparverius ) in an area north of Sheridan, Wyoming, 1977-85. Of 248 occupied territories, 213 were productive with a total of 899 young fledging. During three consecutive years of the study, mean clutch size was 4.7 (N = 105, range 1-7, SE = 0.1), and the average number of chicks to hatch per clutch was 3.7 (N = 100, range 0-6, SE — 0.2), Second clutches were laid in 64.3% of the instances when first clutches failed (N = 14, SE = 0.1). Although second clutch sizes were not smaller, they were less likely to hatch than first clutches. The ratio of males to females fledged was virtually 1:1 (N = 281, SE = 0.03) over the course of this eight-year study. Parametros de reproduction para Halcon Gernicalo silvestre ( Falco sparverius ) usando nidos en cajas, en Montana y Wyoming Extracto. — Un promedio de 37 nidos artificiales usados por Halcones Cernicalo {Falco sparverius), fueron controlados anualmente en una area al norte de Sheridan, Wyoming, 1977-85. De 248 territorios ocupados, 213 fueron productivos con un total de 899 crias. Durante 3 anos consecutivos de estudio, el tamano medio de la nidada fue de 4.7 ( N = 105, entre 1-7 huevos, SE [error estandard] = 0.1), y el numero promedio de crias a criar por cada nido fue de 3.7 {N = 100, entre 0-6 crias, SE = 0.2). Segundos intentos de anidar fueron hechos en 64.3% del total de nidadas sin exito (N =14, SE = 0.1). Si bien las segundas nidadas no fueron mas pequenas, ellas tenian menos posibilidades de exito que las primeras nidadas. En el curso de 8 anos de estudio, la proportion en las crias machos vs. hembras fue virtualmente 1:1 (N = 281, SE = 0.03). [Traduction de Eudoxio Paredes-Ruiz] Few studies in the Northern Great Plains have consistently collected reproductive data on wild rap- tors (Anonymous 1979, Phillips et al. 1990). With the advent of energy development throughout much of the region, monitoring of reproductive parameters and a host of more specialized studies on raptors have been initiated (Phillips and Beske 1990). The progressive compilation and evaluation of data has led to innovative management strategies that accom- modate both raptors and industry (Postovit et al. 1982, Fala et al. 1985). Although American Kestrels ( Falco sparverius ) oc- casionally were found nesting in snags and in sand- stone outcrops, natural nesting sites on the study area were believed to be limiting the distribution of nesting pairs. Forty-two nest boxes were constructed and situated on tracts where natural nesting sites 1 Present address: Nyack Flats, West Glacier, MT 59936- 0144. were lacking (Dahmer et al. 1984). These boxes were monitored from 1977-85, except 1979. Dahmer et al. (1984) showed that revegetated mined lands in Montana and Wyoming could attract and sustain breeding populations of American Kes- trels provided nesting sites were available. Their research further showed that during a five-year pe- riod, reproductive rates of kestrels nesting on revege- tated lands were comparable to those of kestrels nest- ing on native habitats. This paper presents three additional years of data, gathered between 1983-85, on reproductive rates from the same study area in- cluding clutch size, hatching success, and sex ratios. Summary reproductive data are presented here for the entire eight years. Further, I tested the hypoth- eses that, in a wild population where egg-removal techniques were not employed, mean clutch size of first and second clutches would be equal, and that first clutches would have hatching success equal to second clutches. 6 March 1992 American Kestrel Reproduction 7 Table 1. American Kestrel reproduction data 1977-85, from Big Horn County, Montana, and Sheridan County, Wyoming. Year Total Territories Young Fledged/ Occ. Terr. Total Young Fledged Nest Boxes Occupied Active Pro- ductive 1977-78 and 1980-82 3 187 155 (149) b 147 129 3.7 550 1983 35 31 (30) 30 27 3.8 115 1984 36 36 (36) 36 31 3.3 117 1985 40 36 (33) 33 26 3.5 117 1977-78 and 1980-85 298 258 (248) 246 213 3.6 899 a Data from Dahmer et al. (1984). Nests were not monitored in 1979. b The number of nests monitored for productivity is enclosed in parentheses. Study Area and Methods Areas surveyed included approximately 91 km 2 in both Big Horn County, Montana, and Sheridan County, Wy- oming. The topography was rolling plains approximately 1200 m in elevation, transected by perennial and ephem- eral drainages. Sagebrush ( Artemisia tndentata ) grassland, native grassland, and strip-mined lands reclaimed to grass- land ( Agropyron spp.) were the dominant habitat types. Riparian areas along major drainages included Cotton- wood (Populus deltoides), Box Elder (Acer negundo ), Green Ash ( Fraxinus pennsylvanica), and willow ( Salix spp.). Rocky, north facing sidehills supported Ponderosa Pine (Pinus ponderosa) and Rocky Mountain Juniper (Juniperus scopulorum). Confusion exists with terminology used in describing parameters of raptor reproduction. I use the reproductive terminology as described by Postupalsky (1974). Nest boxes were built and erected in locations lacking nesting sites for kestrels, as described by Dahmer et al. (1984). Boxes were monitored for reproductive status and reproductive success, clutch size, hatching success, and sex ratios of fledglings. On occasion, wind damaged a nest box or its support pole. Boxes were not included in the total number of boxes available during a given breeding season if damage oc- curred to the box prior to the breeding season (prior to 1 March). If damage occurred to a nest box after 1 March, it was considered a natural stochastic event, and data were included in productivity calculations. Data from nest boxes were not included in productivity calculations if the num- ber of young fledged was unknown. In one instance in 1983 and in three instances in 1985, boxes that were occupied were not adequately checked for complete pro- ductivity data. Between 1 March and late June the vicinity around each box was checked for occupancy at least four times. Once occupancy was determined, boxes were checked at least every ten days to determine the status of the pair. If the pair had laid eggs, frequent box checks were made to determine clutch size, number of nestlings, and reproduc- tive success. Data are summarized on a yearly basis with the eight- year averages calculated using data from individual re- productive attempts. A comparison of the difference be- tween means was made using Student’s * -tests (Dixon and Massey 1983). Standard errors were calculated using the sample sizes ( N) given in each case. Although some kestrels that fledged from the area returned to breed, banding data from this study documented low mate and nest site fidelity (A.H. Wheeler, unpubl.). Thus, each breeding attempt was considered an independent occurrence. Results and Discussion Reproductive data from the first five years of the study are summarized in Table 1. During the next three breeding seasons, a maximum of 36 nest boxes were monitored annually for a total of 1 1 1 potential breeding territories (Table 1). The percentage of occupied territories was high, averaging 92.9% (range 88.6-100%, SE = 0.6) between 1983 and 1985. Of the 8 territories unoccupied during this period, 2 had lone birds in attendance. Combining data from 1977- 85 with the exception of 1979 shows occupancy av- eraging 87.4% (N = 298, range 52.3-100%, SE = 0.9). Low occupancy occurred during the first breed- ing season after nest boxes were situated. Only 52.3% of nest boxes were used during the first breeding season after initial installation in 1977. Thereafter, the occupancy rate was always >85.0%. Once American Kestrels found nest sites and so- licited a mate, most pairs laid eggs. The percentage of occupied territories that were active between 1983 and 1985 averaged 96.2% (N = 103, range 91.7- 100%, SE = 0.4), while combined data from 1977- 85 yielded a slightly lower average of 95.2% (N = 258, range 84.9-100 %, SE = 0.3). The percentage of occupied territories resulting 8 Anne Hoag Wheeler Vol. 26, No. 1 Table 2. Clutch size and hatchability of American Kestrels 1983-85, from Big Horn County, Montana, and Sheridan County, Wyoming. First Clutches Second Clutches Year Eggs x ± SE Range (N) Hatchlings x ± SE Range (N) Eggs x ± SE Range (N) Hatchlings x ± SE Range (TV) 1983 5.0 ± 0.1 4.2 ± 0.3 5.0 ± 1.0 1.5 ± 1.5 3-7 (28) 0-6 (26) 4-6 (2) 0-3 (2) 1984 4.5 ± 0.2 3.2 ± 0.4 4.4 ± 0.4 3.4 ± 0.7 1-6 (35) 0-5 (35) 3-5 (5) 2-5 (5) 1985 4.8 ± 0.2 4.2 ± 0.3 3.0 ± 2.0 0.0 1-6 (33) 0-6 (30) 1-5 (2) 0(2) 1983-85 4.8 ± 0.1 3.8 ± 0.2 4.2 ± 0.5 2.2 ± 0.7 1-7 (96) 0-6 (91) 1-6 (9) 0-5 (9) in fledglings varied from year to year. The 1983- 85 average was 85.0% ( N = 99, range 78.8-90.0%, SE = 0.6), while the eight-year average was slightly higher (87.6%, N = 248, range 78.8-93.3%, SE = 0.5). Kestrels averaged 3.5 fledglings per occupied territory ( N = 99, range 3. 3-3. 8, SE = 0.03) for 1983-85 with a mean of 4.2 young per active ter- ritory (N = 97, range 1-6, SE = 0.04). Long-term productivity averaged 3.6 fledglings per occupied territory ( N = 248, range 3.2-4. 1, SE = 0.02) with a mean brood size of 4.2 young per active territory (N = 213, range 1-6, SE = 0.02). Clutch size ranged from 1-7 eggs from which 0- 6 chicks hatched (Table 2). Only one 7-egg clutch was found during all years of this study. This female laid a second clutch of 6 eggs after the first clutch failed. Willoughby and Cade (1964) and Heintzel- man and Nagy (1968) reported clutch sizes within the range of 1-6 eggs as have Roest (1957) and Smith et al. (1972). Evidence from egg-removal experi- ments suggests the American Kestrel is an indeter- minate layer (Porter and Wiemeyer 1972, Porter 1974). Eggs were not removed from nest boxes by the researcher during this study, but data from this particular pair of birds supports this concept. During all three years, some American Kestrel pairs laid a second clutch after their first clutch failed. Severe weather including low temperatures, high precipitation, and high winds apparently caused some nest failures. Data on clutch sizes and hatching success are segregated into first or second clutches (Table 2). In 9 of 14 instances where the first clutch failed, kestrels laid a second clutch. This illustrates the need for rechecking nest sites after the initial clutch fails in order to generate accurate productivity estimates. Willoughby and Cade (1964) reported an average clutch of 3.7 eggs for 12 captive pairs. Heintzelman and Nagy (1968), studying 13 wild pairs using nest boxes, reported a slightly larger mean clutch size ( x = 4.2). Neither team mentions second clutches in the papers cited here. Combining data from both first and second clutches from nest boxes on this study area, the mean clutch size was 4.7 eggs (N = 105, SE = 0.1). First clutches averaged 4.8 eggs (N = 97, range 1-7, SE = 0.1), while second clutches did not differ significantly in size ( x = 4.2, N = 9, range 1-6, SE = 0.5, t = 2.26, P > 0.05). The average number of chicks hatched per clutch from all clutches monitored was 3.7 ( N = 100, range 0-6, SE = 0.2). Hatching success of first clutches (x = 3.8, N = 91, SE = 0.2), however, was significantly greater than that of second clutches (x = 2.2, N = 9, SE = 0.7, t = 2.26, P < 0.05). Of 453 eggs from 97 clutches, 81.2% (368 young) hatched. Of 421 eggs from first clutches, 82.7% hatched while only 62.5% of 32 second clutch eggs hatched. Standard errors for both clutch size and hatching success were sig- nificantly greater for second clutches. Hatchability of eggs on this study area was higher than the rate reported by Heintzelman and Nagy (1968) for Pennsylvania where 78% of 55 eggs hatched from 13 clutches. Despite the observation that eggs from second clutches were less likely to hatch, the percentage of chicks hatching from second clutches that survived March 1992 American Kestrel Reproduction 9 to fledging (90.0%, N = 20) was similar to the per- centage of chicks from first clutches (91.3%, N — 346). Sex ratios of fledgling American Kestrels varied from the extreme of 5 females from a clutch of 5 eggs to an even count of males and females. The ratio of female to male fledglings over the course of the eight-year study was not significantly different from 1 to 1 (N = 281, males =141, females = 140, SE = 0.03). Acknowledgments This study was funded by Kiewit Mining and Engi- neering Company, Decker and Big Horn Coal Companies, Sheridan, Wyoming. Their land management practices allowed for placement of the nest boxes and ensured the security of the boxes into the future. John Berry, Tom Dahmer, Brent Stettler, and the late Mark Raphelson aided with both data collection and maintenance of the nest boxes. Nick Forrester gathered much of the repro- duction data up until 1983 and initiated banding. Grati- tude is extended to Sam Scott for always going-to-bat for corporate approval and funding of environmental research. Lyman McDonald and John Berry reviewed the manu- script and made valuable suggestions. Literature Cited Anonymous. 1979. Snake River birds of prey area. Spe- cial report to the Secretary of the Interior, Bureau of Land Management, U.S. Department of the Interior, Boise, ID. Dahmer, T.D., N.C. Forrester, J.M. Lockhart and T.P. McEneaney. 1984. Nest box use by American Kestrels on and around western surfaced-mined lands. Pages 210-213 in Proceedings of a Symposium on Is- sues and Technology in the Management of Impacted Western Wildlife. Technical Publication No. 14, Thorne Ecological Institute, Boulder, CO. Dixon, W.J. and F.J. Massey, Jr. 1983. Introduction to statistical analysis. McGraw-Hill Publishing Co., New York. Fala, R.A., A. Anderson and J.P. Ward. 1985. High- wall-to-pole Golden Eagle nest site relocations. Raptor Research 19:1-7. Heintzelman, D.S. and A.C. Nagy. 1968. Clutch siz- es, hatchability rates, and sex ratios of Sparrow Hawks in eastern Pennsylvania. Wilson Bull. 80:306-311. Phillips, R.L. and A.E. Beske. 1990. Distribution and abundance of Golden Eagles and other raptors in Campbell and Converse Counties, Wyoming. Tech- nical Report 27, Fish and Wildlife Service, U.S. De- partment of the Interior, Washington, D.C. , A.H. Wheeler, J.M. Lockhart, T.P. Mc- Eneaney and N.C. Forrester. 1990. Ecology of Golden Eagles and other raptors in southeastern Mon- tana and northern Wyoming. Technical Report 26, Fish and Wildlife Service, U.S. Department of the Interior, Washington, D.C. Porter, R.D. 1974. Experimental alterations of clutch- size of captive American Kestrels Falco sparverius. Ibis 117:510-515. AND S.N. Wiemeyer. 1972. Reproductive pat- terns in captive American kestrels. Condor 74:46-53. Postovit, HR., J.W. Grier, J.M. Lockhart and J Tate. 1982. Directed relocation of a Golden Eagle nest site. J. Wild l. Manage. 46:1045-1048. Postupalsky, S. 1974. Raptor reproductive success: some problems with methods, criteria, and terminology. Pages 21-31 in F.N. Hamerstrom, Jr., B.E. Harrell and R.R Olendorff [Eds.], Management of raptors. Raptor Re- search Foundation, Inc., Vermillion, SD. Roest, A. I. 1957. Notes on the American Sparrow Hawk Auk 74:1-19. Smith, D.G., C.R. Wilson and H.H. Frost. 1972. The biology of the American Kestrel in central Utah. Southwest, Nat. 17:73-83. Willoughby, E.J. and T.J. Cade. 1964. Breeding be- havior of the American Kestrel. Living Bird 3:75-96. Received 3 May 1991; accepted 17 October 1991 J. Raptor Res. 26(1):10-12 © 1992 The Raptor Research Foundation, Inc. OBSERVATIONS ON THE BEHAVIOR OF SURPLUS ADULTS IN A RED-SHOULDERED HAWK POPULATION Michael D. McCrary Department of Biology, California State University, Long Beach, CA 90840 Peter H. Bloom 1 National Audubon Society , 13611 Hewes Avenue, Santa Ana, CA 92705 Marjorie J. Gibson 1 National Audubon Society, 13611 Hewes Avenue, Santa Ana, CA 92705 Abstract. — We describe the behavior of two adult male Red-shouldered Hawk ( Buteo hneatus) “floaters” in response to the temporary removal of two resident males from their breeding territories in coastal southern California. We also provide observations of a radio-tagged floater moving through eight adjacent Red-shouldered Hawk home ranges. Observaciones de conducta de adultos sobrantes en una poblacion de aguilillas de la especie Buteo Hneatus Extracto. — Describimos la conducta de dos aguilillas de la especie Buteo Hneatus, machos adultos y sin territorio de reproduccion, que responden a la ausencia temporal de dos aguilillas machos residentes, los que han sido alejados de sus territorios de reproduccion en la costa sur de California. Tambien proveemos observaciones sobre una aguililla sin pareja y sin nido, que fue radiocontrolada, mientras se desplazaba a traves del area de reproduccion adyacente de ocho pares de rapaces de esta especie. [Traduction de Eudoxio Paredes-Ruiz] “Floaters” are non-breeding individuals that are capable of establishing a territory and breeding if resources become available (Brown 1969). With the exception of the study by Bowman and Bird (1986) and anecdotal data by Newton (1979), there is little published information on the behavior of floaters in raptor populations. The replacement of individuals or entire pairs that were shot during the breeding season has provided most of the anecdotal evidence for the existence of surplus nonbreeders (Ratcliffe 1980). Time to re- placement has varied from within the same day to 10 wk (Newton 1979). Bent (1937) noted that a Red-shouldered Hawk ( Buteo Hneatus ) of unknown sex was “promptly” replaced after being shot. We describe the attempted replacement of two breeding adult male Red- shouldered Hawks temporarily re- moved from their territories, and the behavior of an apparent adult male floater equipped with a radio- transmitter. 1 Present address: Western Foundation of Vertebrate Zo- ology, 1100 Glendon Avenue, Los Angeles, CA 90024. Study Area and Methods Our study area was located on Camp Pendleton Marine Corps Base, San Diego County, and adjacent Rancho Mis- sion Viejo, Orange County, California. Dominant vege- tation is oak woodland, riparian woodland, coastal sage scrub, and grassland (Bloom 1989). Observations were made at Camp Pendleton of two floaters that attempted to replace two resident adult male hawks, both captured on 17 February 1979 before the egg laying stage (McCrary 1981). Both resident hawks were captured in their own adjacent territories and held 2 and 3 hr, respectively, while we attached back-mounted trans- mitters. Observations of a floating adult male (Fig. 1) captured on 1 July 1987, during the post-fledging/dis- persal period, and affixed with a radiotransmitter were made on adjacent Rancho Mission Viejo (Bloom 1989). Results We observed a surplus Red-shouldered Hawk adult in each of the captured males’ territories at Camp Pendleton. Both intruders were probably males since they were smaller than the females. Reactions of the resident males upon release back into their territory varied. One male attacked and struck the intruder immediately, dislodging it from its perch 10 March 1992 Surplus Red-shouldered Hawks 11 ♦ ♦ ♦<*** ★ 1 2 ‘n KILOMETERS M5 HOME RANGE ♦ M5 LOCATIONS ★ NEIGHBORING NESTS * M5 ORIGINAL TRAPPING LOCATION Figure 1. Home range and locations of adult male Red- shouldered Hawk M5 on Rancho Mission Viejo in re- lation to other Red-shouldered Hawk nesting territories. on a utility pole. This attack was followed by the rapid departure, in opposite directions, of both hawks from the vicinity of the nest. The resident male was later seen about 100 m south of his nest, but the intruder was not found. Prior to the resident male’s release, the resident female and intruder vocalized almost continuously. Within 5 min of the release of the resident male all calling ceased. In the second territory both the resident female and the intruder also vocalized almost continuously. Immediately after release, the resident male perched in a tree near his nest and began calling. The female and the intruder soared over the nest and, after 1- 2 min, the resident male also began soaring. At one point the resident male perched briefly in a tree where he was attacked several times by the intruder. A few minutes later the resident female and intruder perched next to each other on the same branch of a tree near the nest. When the intruder assumed a near horizontal posture similar to that of females during copulation, the female immediately knocked it from the perch. The intruder immediately re- turned to the branch and assumed the same position next to the female, who knocked it off the branch again. Both birds circled over the nest site for ap- proximately 2-3 min, then the female joined the resident male on the nest where they copulated for 8 sec. After perching nearby, the intruder circled over the resident pair for 3-4 min and soared off 2 km north before disappearing. No surplus male hawks were observed at 5 other territories where we captured and temporarily held 5 resident adult males, nor did any surplus females attempt to displace 10 resident adult females at 10 territories where they were captured and held 1.5- 3 hr for transmitter placement (McCrary 1981, Bloom 1989). The hawk (M5) captured on Rancho Mission Viejo and equipped with a radiotransmitter con- formed to the definition of a floater (Brown 1969, Newton 1979). In contrast to seven other adult male Red-shouldered Hawks that were radiotagged and tracked several months/bird, and that maintained distinct territories averaging slightly more than 1 km 2 , this bird ranged over an area of 35.7 km 2 in just 2 mo (Bloom 1989). Except for very slight over- lap between some adjacent pairs, none of the paired, breeding birds wandered into more distant home ranges (McCrary 1981). The floater’s home range between 1 July and 31 August 1987, however, in- cluded portions of the home ranges of eight breeding pairs that it visited from 1-3 times (Fig. 1). Of the 20 d it was monitored during the 2 mo period, 169 daylight hours of 14 d were spent in the area in which it was trapped, and 71 daylight hours of 11 d were spent in the territories of other pairs. During the 2 mo period, it was observed vocalizing only once (100 m from where trapped), and was quiet when in the home ranges of other pairs. On four nights it roosted in two of these home ranges. Discussion Although limited to a few birds, our observations show the existence of floaters in a southern Califor- nia Red-shouldered Hawk population. The Red- shouldered Hawk is a highly territorial species (Bent 1937, Henny et al. 1973), and the existence of float- ers within a population is thought to be an important factor in explaining the function of territoriality (Brown 1969). In most instances described in the literature only 12 McCrary et al. Vol. 26, No. 1 female raptors were removed and replaced (Newton 1979); little documentation is available for males (Village 1983). In a mate replacement experiment conducted on a wild population of American Kestrels (Falco sparverius) during the incubation period, 8 of 16 males were replaced within 18-144 hr (Bowman and Bird 1987). The rapidity of male replacement may be related to the availability of floaters in the population, the proximity of floaters to a vacated territory, stage of the breeding cycle, seasonality, or sexual differences in female behavior or vocaliza- tions. Bowman and Bird (1986) found that only some of the mates lost in their study were replaced. Sur- plus birds may cue in on factors influencing their future reproductive potential before “deciding” to replace (Smith 1978, Saether and Fonstad 1981), or replacement may merely depend on the proximity of suitable recruits when a mate is lost (Dare 1961). Bowman and Bird (1986) concluded that nonbreed- ers did not actively monitor territories for replace- ment potential but were non-randomly distributed in areas of suitable habitat. The behavior of the radio-tagged floater suggests that vacant territories are detected by periodically searching territories and waiting for an opening. The speed at which the other two floaters became aware of vacancies and moved in suggests they were alerted to a possible opening by some aspect of the female’s behavior. Floaters may increase their chance of de- tecting such a signal by active searching. Future research should focus on obtaining the critical ob- servation of a patrolling surplus nonbreeder actually replacing a lost member of the breeding population. Acknowledgments Our work was supported by Rancho Mission Viejo, Spence Porter of Communication Specialists, the Western Foundation of Vertebrate Zoology, and the El Dorado Audubon Society. We sincerely appreciate the cooperation of the Natural Resources Office, Camp Pendleton Marine Corps Base. The field assistance of E.H. Henckel, J.L. Henckel, and J.T. Jennings is gratefully acknowledged. We thank D.M. Bird, R. Bowman, L.F. Kiff, M.L. Mor- rison, J.R. Murphy, R.W. Nelson and D.G. Smith for constructive comments. Literature Cited Bent, A.C. 1937. Life histories of North American birds of prey. U.S. Nat. Mus. Bull. 67:1-409. Bloom, P.H. 1989. Red-shouldered Hawk home range and habitat use in southern California. M.Sc. thesis, California State University, Long Beach, CA. Bowman, R. and D.M. Bird. 1986. Ecological corre- lates of mate replacement in the American Kestrel. Condor 88:440-445. AND . 1987. Behavioral strategies of American Kestrels during mate replacement. Behav Ecol. Sociobiol. 20:129-135. Brown, J.L. 1969. Territorial behavior and population regulation in birds: a review and re-evaluation. Wilson Bull. 81:293-329. Dare, P. 1961. Ecological observations on a breeding population of the Common Buzzard ( Buteo buteo). Ph.D. thesis, Exeter University, Exeter, U.K. Henny, C.J., F.C. Schmid, E.M. Martin and L.L Hood. 1973. Territorial behavior, pesticides and the population ecology of Red-shouldered Hawks in cen- tral Maryland, 1943-1971. Ecology 54:545-554. McCrary, M.D. 1981. Space and habitat utilization by Red-shouldered Hawks ( Buteo lineatus elegans) in southern California. M.Sc. thesis, California State University, Long Beach, CA. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. Ratcliffe, D. 1980. The Peregrine Falcon. Buteo Books, Vermillion, SD. Saether, B.E. and T. Fonstad. 1981. A removal ex- periment showing unmated females in a breeding pop- ulation of Chaffinches. Anim. Behav. 29:637-639. Smith, S.M. 1978. The “underworld” in a territorial sparrow: adaptive strategy for floaters. Am. Nat. 112: 571-582. Village, A. 1983. The role of nest-site availability and territorial behavior in limiting the breeding density of kestrels. J. Anim. Ecol. 51:415-426. Received 23 July 1991; accepted 14 November 1991 J Raptor Res. 26(1):13-18 © 1992 The Raptor Research Foundation, Inc. HOME RANGE, HABITAT USE AND BEHAVIOR OF PRAIRIE FALCONS WINTERING IN EAST-CENTRAL COLORADO Gary Beauvais and James H. Enderson Department of Biology, Colorado College, Colorado Springs, CO 80903 Anthony J. Magro 174 East Lake Shore Drive, Barrington, IL 60010 Abstract. — In the two winters 1988-90 we radiotagged 18 Prairie Falcons ( Falco mexicanus ) in east- central Colorado. The home ranges of 10 birds (mostly males), tracked for 17 to 70 d, averaged 30.2 km 2 (minimum convex polygon) and 10.4 km in maximum linear dimension. This home range is far less than reported for the nesting season. Home ranges included significantly more cultivated or fallow land than predicted according to availability. Horned Lark ( Eremophila alpestris) densities were significantly higher in these habitats than in areas less frequently used by falcons. Larks, almost the exclusive prey of falcons, were attacked in low, high-speed flights, and 13.5% of attacks were successful. Old injuries seen in captured falcons and frequent loss of transmitters from entanglement on barbed wire fences suggest that such attacks are hazardous to falcons. Prairie Falcons were nearly as abundant as Ferruginous Hawks ( Buteo regalis), and about half as numerous as Rough-legged Hawks ( B . lagopus). Intraspecific or inter- specific aggression was rare compared with reports for Prairie Falcons during the breeding season. Espacio habitado, uso del habitat y conducta de halcones de la especie Falco mexicanus, invernando en la zona central este de Colorado Extracto.- — Durante 2 inviernos 1988-90 hemos radiocontrolado 18 Falco mexicanus en la zona central este de Colorado. El espacio habitado por 10 aves (la mayoria machos), seguidas entre 17 y 70 dias, promedio 30.2 km 2 (poligono convexo minimo) y 10.4 km en dimension lineal maxima. Esta area habitada es mucho menor que la registrada para la estacion reproductora. El area de habitation incluyo signifi- cativamente mas tierra cultivada o cultivable que lo pronosticado de acuerdo con la disponibilidad. Las densidades de poblacion de las aves de la especie Eremophila alpestris fueron significativamente mas altas en estos habitats que en areas usadas con menos frecuencia por halcones. Las alondras, que son presas casi exclusivas de los halcones, fueron atacadas en vuelos bajos de alta velocidad, y 13.5% de los ataques tuvieron exito. Halcones capturados con heridas viejas, y la frecuencia en la perdida de los transmisores, al entramparse en los cercos de alambres de puas, sugieren que tales ataques son azarosos para los halcones. Los Falco mexicanus fueron casi tan abundantes como los halcones de la especie Buteo regalis , y casi en la mitad del numero de los halcones de la especie Buteo lagopus. La agresion intraespecie e interespecie fue rara, comparada con los informes que se refieren a los halcones de la especie Falco mexicanus durante la estacion reproductora. [Traduction de Eudoxio Paredes-Ruiz] Falcons in temperate regions seldom have been studied in winter. An exception is the abundant American Kestrel ( Falco sparverius) which was stud- ied for habitat selection (Koplin 1973, Bohall-Wood and Collopy 1986, Smallwood 1987). This paucity may result from wide dispersal and difficult study logistics when raptors use unpredictable food re- sources (Newton 1979). Prairie Falcons (F. mexi- canus), individually marked with holes in remiges, were sighted repeatedly in winter in north-central Colorado. Although the extent of the falcons’ move- ments were estimated, no correlations with habitat types or prey availability were made (Enderson 1964). More recently, telemetry has made a more complete description of movements of wide-ranging falcons feasible. Band recoveries of Prairie Falcons in winter show dispersal of adult and hatch-year individuals east- ward from breeding habitat in the Rocky Mountains and Great Basin to the Great Plains. We determined the size of winter home ranges, territoriality, and habitat use on the Plains where previous roadside counts had shown Prairie Falcons to be the second or third most abundant winter raptor (Enderson 13 14 Beauvais et al. Vol. 26, No. 1 1965, Johnson and Enderson 1972, Bauer 1982). We predicted that habitat use by Prairie Falcons would be correlated with prey abundance and dis- tribution. Study Area This study was conducted about 48 km east of the Rocky Mountains near Ellicott, Colorado. The area has gentle drainages with no permanent streams. Elevation varies between 1700-2070 m. Short-grass prairie was most prev- alent and was dominated by Blue Grama ( Bouteloua grac- ilis). Sand Sage ( Artemisia jilifoha) prevailed on sand hills. Overgrazing has resulted in abundant Yucca ( Yucca glau- ca ), Bushy Buckwheat ( Eriogonum effusum), and Snake- weed ( Guiierrezia sarothrae ) dispersed among Blue Grama, the latter often grazed shorter than 3 cm. Cultivated areas included Winter- wheat ( Triticum anetiuum), Milo ( Sor- ghum sp.) and sod farms. Because rainfall averaged only 32 cm annually, much cultivated land was fallow and was dominated by open stands of Common Sunflower (He- hanthus annuus) , Russian Thistle ( Salsola australis), Sweet- clover ( Melilotus sp.), and Tansy-aster ( Machaeranthera sp.). Horned Larks ( Eremophila alpestris) were ubiquitous and the only conspicuous small passerine. Methods Prairie Falcons were captured with noose-harnesses on Rock Doves ( Columba hvia) or bal-chatri traps (Berger and Mueller 1959) in November-February 1988-90. Each falcon was fitted with a 216-mHz transmitter weighing 15 g, sewn with nylon dental floss to the two central rectrices. Battery life was about 3 mo. Tracking was by hand-held yagi antennas, using aircraft when signals could not be obtained from the ground. When possible, falcons were approached until they could be seen with a spotting scope. Triangulation was used when the bird could not be seen. Generally, we sought to relocate each bird every 3 or 4 d unless a signal could not be found after repeated searches. This interval was chosen to provide individuals time to move between observations, thus reducing the pos- sibility that the position on one observation would influ- ence strongly the position on a second observation. We tracked mainly in mornings, expending about 430 hr. Three falcons were followed from sunrise to sunset on four sep- arate days to check daily movement. Winter range was determined by minimum convex polygon technique (Mohr 1947) because observations were scattered and not ar- ranged linearly. Ranges were plotted on vegetation maps, and the areas of four cultivated vegetational types, and of rangeland were totalled for each bird. The number and location of Horned Larks were re- corded in winter 1989-90 along a 124 km route including all vegetation types. Counts were made on 12 d in a 6-wk period and larks within 300 m of the road were counted after flocks flushed. To speed the counting before the flock disappeared, each flock was assigned one of the following size categories: 1-12, 13-24, 25-36, 37-50, 51-65, and >65. However, in our analysis of abundance we assigned each flock an index value of 6, 18, 30, 44, 58, or 100 individuals, respectively. We used an abundance index instead of analysis on a per flock basis because observations suggested that flock sizes are not fixed and may be variable by habitat and through time. Aerial photos, augmented by observations in the field, were used to create a vegetation map specific to the 124 km route. The position and index value of each flock of larks was entered on the map. Prairie Falcon locations and abundance index values for Horned Larks were categorized by vegetation type and compared using chi-square analysis of use compared to habitat availability. Simultaneous analysis of habitat use was used to determine habitat preference or avoidance (Neu et al. 1974). Results We caught and radiotagged 17 males and 1 fe- male Prairie Falcon; 11 were caught in 1988-89 and seven in 1989-90 (Table 1). One bird was given a second transmitter after the first was lost. All sight- ings, including first encounters with falcons that were given radios, included 57 males, 18 females, and 33 not identified as to sex. Locations of 1 0 falcons tracked at least seven times during periods from 17-70 d were included in the analysis. Fewer locations or days apparently biased home-range estimation downwards (Table 1). The remaining eight tagged birds, including five hatch- year males, could be found for 10 d or fewer after tagging. We suspect they either moved far from the area or lost their transmitters. With one exception, these eight falcons showed small range areas in the brief periods they were tracked. In all, we found six functioning transmitters on the ground near barbed-wire fences, five of which were still attached to tail feathers. The tips of the antennas were tightly coiled suggesting they were struck as the bird barely cleared the top wire in their characteristic high-speed, low-level attacks on prey. One transmitter was attached to the top wire of a fence. None were found below fence posts, suggesting that radios were lost due to collision. We saw two Prairie Falcons with missing central feathers near the capture sites, but could not be sure they were previously tagged. Foraging Ranges and Habitat Use. Minimum home ranges, the area enclosed when outermost points are connected to form a polygon, averaged 30 km 2 (range 12.3-68.0) after one range of 583 km 2 for an adult male was excluded (Table 1). The average maximum linear dimension for the nine ranges was 10.4 km (range 5.8-17.8) excluding one 42 km mea- surement. March 1992 Winter Ecology of Prairie Falcons 15 Table 1. Prairie Falcons radiotagged in El Paso County, Colorado, during winters 1988-90. Measurements are rounded to nearest unit. Age , 3 Sex Capture Date No. Days Observed Locations Recorded Range Area (km 2 ) Longest Axis (km) AM 9 Dec 1989 1 1 — — IM 16 Dec 1989 1 1 — — IM 13 Dec 1989 2 2 — 1 IM 3 Dec 1988 5 3 3 4 IM 1 Jan 1989 4 3 5 4 AM 10 Dec 1988 7 5 14 6 IM 12 Jan 1989 8 6 1 2 AM 10 Jan 1989 10 7 6 5 AM 4 Feb 1989 27 7 22 10 AM 3 Jan 1989 17 8 68 18 AF 25 Nov 1989 70 8 12 6 IM 21 Jan 1989 32 12 22 11 IM 16 Dec 1989 36 14 33 9 AM 1 Jan 1989 55 14 34 12 AM 15 Jan 1989 54 17 27 9 IM 18 Jan 1990 61 19 16 8 IM 30 Dec 1988 41 20 38 11 AM Means (± 21 Jan 1990 Standard Error) b 40 22 583 30.2 ± 5.5 42 10.4 ± 3.8 a A = Adult; I = Hatch year. b Excluding adult male captured 21 January 1990 and individuals observed for fewer than 17 d. Each of 141 locations for the 10 Prairie Falcons was categorized by vegetation type. This observed habitat use was compared to expected habitat use based on habitat availability. We rejected the null hypothesis that falcons frequented habitat in pro- portion to its availability (x 2 = 139.8, df = 4, P < 0.001). Simultaneous analysis of habitat use revealed that falcons frequented milo fields and fallow land significantly more than expected (P < 0.05; Fig. 1). Rangeland was used significantly less than expected (P < 0.05). Two falcons were followed from sunrise to sunset one day each, and another for two non-consecutive days. On all four occasions the birds were most sed- entary in midday. Two falcons flew over a path about 10 km long, and one, whose range was 42 km in maximum dimension (Table 1), flew about 55 km on each of the two days he was followed. This bird was one of two adult males that visited the only rock outcrops in the study area used by nesting Prairie Falcons in recent decades. Interactions with Prey. Prairie Falcons attacked prey 40 times and all but 3 attacks involved Horned Larks. Other prey included two mice (Cricetidae), and a Western Meadowlark ( Sturnella neglecta). Of the 40 attacks, 6 were successful, resulting in the capture of 5 larks and 1 vole ( Microtus sp.). Larks seemed attracted to roadsides, perhaps be- cause of taller and more diverse vegetation, enhanc- ing our counts. Larger flocks were encountered more often in the few days when snow cover was complete. Most Horned Larks occurred in larger flocks rather than small ones. In the 12 lark censuses, 236, 93, 61, 29, 80 and 48 flocks were observed with 1-12, 13-24, 25-36, 37-50, 51-65 and >65 larks, re- spectively. Although fewer large flocks were seen, these flocks were subject to the majority of falcon attacks. Of attacks on larks by Prairie Falcons, 24 (65%) were at flocks exceeding 50 individuals (23% of all flocks). Frequencies of vegetation types along the route used to count larks were similar to those in falcon winter ranges, with rangeland predominating (Fig. 1). We rejected the null hypothesis that larks in- habited vegetation types in relation to vegetation frequency (x 2 = 485.5, df = 4, P < 0.001). Simul- taneous analysis of habitat use revealed that Horned Larks used all cultivated habitats significantly more 16 Beauvais et al. Vol. 26, No. 1 cd o 100 -i 80 - 60 - 40 - 20 - Prairie Falcons ■ ReUrtrv* Abundance i2 Relatrve Use W-Wheat Milo Sod Farm Fallow Rangeland Table 2. Raptors seen in El Paso County, Colorado, during 28 d, November-February 1989-90. All roadside counts were made before noon on calm days. Total distance traveled = 2051 km. SPECIES a km/lNDIVIDUAL Rough-legged Hawk 22 Ferruginous Hawk 51 Prairie Falcon 54 Golden Eagle 73 Northern Harrier 171 American Kestrel 228 Merlin 256 a Bald Eagles ( Haliaeetus leucocephalus) were seen three times and Red-tailed Hawks were seen two times; 59 raptors were too distant to identify. Habitat Type Habitat Type Figure 1. Habitat use by Prairie Falcons and Horned Larks. Symbols above the bars indicate selection: + = greater use than expected, — = less use {P < 0.05) and 0 = no difference. as Prairie Falcons (Table 2). Ferruginous Hawks (Buteo regalis), Golden Eagles (Aquila chrysaetos ), and Northern Harriers ( Circus cyaneus ) were also common. Fence posts, electric distribution and trans- mission line poles were numerous and the latter have been shown to increase the local density of raptors on this study area (Stahlecker 1978). Prairie Falcons interacted with conspecifics only once, when an adult male flew about 1 km to make a shallow dive at another, probably a male. The former landed on a fence post when the other flew slowly away. Several times we saw falcons perched within a few hundred meters of each other without aggression. Overlap between winter ranges deter- mined by telemetry was insignificant, but we ob- served the ranges of adjacent birds for only a few weeks. Interspecific aggression was seen rarely, although other raptors were common. Rough-legged Hawks were attacked three times by falcons after we had released Rock Doves as bait in the immediate vicin- ity. Common Ravens ( Corvus corax ) were chased twice under similar circumstances. Perched falcons were the subjects of shallow stoops to within 1 m by a Merlin ( Falco columbarius ), a Northern Harrier, and a Rough-legged Hawk. The falcons did not fly. than expected ( P < 0.05) and frequented rangeland significantly less than expected ( P < 0.05). Fur- thermore, 39% of all lark flocks over 50 individuals were found in cultivated fields comprising only about 10% of the total habitat. Aggression and the Abundance of Other Rap- tors. Rough-legged Hawks ( B . lagopus ) were the most common raptors and were twice as abundant Discussion Most field studies of raptors in winter were in modified and structurally variable habitats which created irregular distributions of both raptors and prey. In southeastern Idaho, where cultivated farm- land and rangeland were interspersed, Rough- legged Hawks preferred the former where voles were abundant. Golden Eagles preferred rangeland where March 1992 Winter Ecology of Prairie Falcons 17 Black-tailed Jackrabbits ( Lepus cahfornicus) were available (Craig et al. 1986). Preston (1990) found that Red-tailed Hawks ( Buteo jamaicensis) and Northern Harriers were distributed nonrandomly in central Arkansas in winter; harriers responded to prey abundance and cover density, and the hawks to prey abundance and perch availability. Patches of vegetation highest in prey, but with deep cover, were used by both species less frequently than pre- dicted. Male and female American Kestrels winter- ing in Florida defended territories in different hab- itats (Bohall-Wood and Collopy 1986); males used closed habitats that seemingly restricted access to insect prey compared to females in more open areas (Smallwood 1987). In the winters 1960-62 Prairie Falcons in Colorado fed primarily on Horned Larks, but habitat was uniformly winter-wheat and no test of habitat was attempted (Enderson 1964). Of the 18 Prairie Falcons we radiotagged in this study, 8 could not be tracked for long, mainly because transmitters failed or fell off after the birds collided with fences. Transmitter failure probably contrib- uted to an underestimate of the time wintering fal- cons actually remained on the study area. Some of the falcons we saw or captured were probably tran- sient and failed to remain and establish a winter range. In 1960-62, nine Prairie Falcons, wing- marked mostly in November in a winter- wheat area of north-central Colorado, remained an average of 67 d (range 20-117; Enderson 1964) compared to 43 d (17-70) in the present study. Of the eight falcons we were able to track only briefly, five were radiotagged on 16 December or earlier. Cultivated areas, including fallow land, attracted Horned Larks, which in turn probably attracted Prairie Falcons that fed on them. Larks may be able to locate more food, including weed seeds, in such areas. Increased cover associated with fallow fields may also lower wind and enhance shelter compared to the open prairie community. On the other hand, cover provided by Milo fields and fallow land may have concealed falcons attacking the high concen- trations of larks found there. The relationship between lark and Prairie Falcon habitat use was not perfect. Horned Larks used wheat and sod areas more than predicted, but falcons seemed to show no preference. Larks are easily seen on sod and wheat; we may have overestimated their abun- dances there compared to other types. We believe Prairie Falcons oriented to the larger flocks of larks, exposed while moving in short flights as the larks fed. Although Kenward (1978) showed a Goshawk ( Accipiter gentilis ) was less successful attacking larger flocks of Wood Pigeons ( Columba palumbus), Prairie Falcons did not seem distracted by many, panicked Horned Larks. Prairie Falcons that found concentrations of larks did not move far and sometimes sat for several hours unless flushed by automobiles. Once falcons discovered dense con- centrations of Horned Larks, they remained in the area. We discounted perch availability as a factor in winter range size. Poles of various types are abun- dant at borders of rangeland pastures and cultivated areas alike, and such areas are small compared to the size of a falcon’s range. Although many range- land areas have few fences, power poles were always present. The average home range in this study was 30.2 km 2 for nine Prairie Falcons (Table 1). In the nesting period in southern Idaho, 18 adult Prairie Falcons used an average range of 107 km 2 (Dunstan et al. 1978). In the Mohave Desert, Harmata et al. (1978) found mean home range size during nesting was 72 km 2 for males (N — 3) and 47 for females (TV = 3). The average maximum dimension of range in this study was 10.4 km excluding that of one wandering adult male, and was 8.8 km for 11 marked Prairie Falcons in north-central Colorado in the winter of 1960-61 (Enderson 1964). These values were un- expectedly small considering these falcons can fly 1 0 km in less than 7 min. The two largest ranges we measured were of adult males that were not associated with cultivated hab- itats. One of these birds was in a region that was entirely rangeland. The other traveled so widely, he probably encountered cultivated areas and Horned Lark concentrations but did not exploit them for long. Both males visited potential nest cliffs in their long flights. Where there was no cultivated land, Prairie Fal- cons have been more sparsely distributed. In south- eastern Colorado, one Prairie Falcon was seen per 200 km driven in prairie habitat in a 3400 km road- side survey during the winters of 1983-88 (Andersen and Rongstad 1989), compared to one individual per 54 km in this study. We cannot fully explain the unequal sex ratio of falcons (1 female : 3.2 males) seen in this study. Sex is generally easy to determine in this species and individuals not identifiable were usually far away. Prey larger than larks were rare, a factor that may 18 Beauvais et al. Vol. 26, No. 1 have tended to exclude female falcons. Larks may- be too small for females to exploit regularly. Prairie Falcons were found in a ratio of 1:0.6 in winter in northern Colorado wheat country where larger pas- serines and Ringed-necked Pheasants ( Phasianus col- chicus) were common (Enderson 1964). Only 5 in 37 (13.5%) attacks on Horned Larks were successful. Larks weigh about 9% the body weight of a male Prairie Falcon and appeared vastly more maneuverable in flight. High-speed attacks low to the ground result in surprise and none of the misses we saw were followed by second attempts. We caught three males with substantial old wounds on a patagium, foot, or cere, suggesting attacks in fenced country are hazardous. Frequent loss of transmitters at fences underscores the risk, Roalkvam (1985) summarized 13 reports on the hunting effectiveness of Peregrine Falcons ( Falco peregrinus) outside the breeding season and found a mean success rate of 12.7%. Buchanan et al. (1986) found a 14.6% success rate for peregrine first- attempts on Dunlins ( Calidris alpina). Prairie Falcons ate mainly Horned Larks in this study, a sharp contrast to their varied diet in the breeding season. In an extreme case 39 species of reptiles, birds, and mammals were used by 19 pairs in the Mohave Desert (Boyce 1985). Aggression among Prairie Falcons was not seen and winter territories, as described in American Kes- trels (Smallwood 1987), seem unlikely. Prairie Fal- cons were conspicuously passive given the almost continual presence of hawks or other falcons. In the nesting season, Prairie Falcons are perhaps more aggressive. Dunstan et al. (1978) recorded 10 intra- specific and 5 interspecific aggressive encounters while following radio-tagged nesting adults. Acknowledgments C. Hecock, S. Kempers, and M. Robert provided as- sistance with the field work. A. Harmata, I. Newton and C.M. White made valuable comments on the manuscript, and J. Newberry helped with its revision. Literature Cited Andersen, D.E. and O.J. Rongstad. 1989. Surveys for wintering birds of prey in southeastern Colorado: 1983-1988. Raptor Res. 23:152-156. Bauer, E.N. 1982. Winter roadside raptor survey in El Paso County, Colorado, 1962-1979. Raptor Res. 16: 10-13. Berger, D.D. and H. Mueller. 1959. The bal-chatri: a trap for birds of prey. Bird-Banding 30:18-26. Bohall-Wood, P. and M.W. Collopy. 1986. Abun- dance and habitat selection of two American Kestrel subspecies in northcentral Florida. Auk 103:557-563. Boyce, D.A. 1985. Prairie Falcon prey in the Mohave Desert, California. Raptor Res. 19:128-134. Buchanan, J.B., S.G. Herman and T.M. Johnson. 1986. Success rates of Peregrine Falcon ( Falco pere- grinus) hunting Dunlin ( Calidris alpina) during winter Raptor Res. 20:130-131. Craig, E.H., T.H. Craig and L.R. Powers. 1986. Habitat use by wintering Golden Eagles and Rough- legged Hawks in southeastern Idaho. Raptor Res. 20 69-71. Dunstan, T.C., J.H. Harper and K.B. Phipps. 1978 Habitat use and hunting strategies of Prairie Falcons, Red-tailed Hawks and Golden Eagles. Final Report, Bureau of Land Management, Denver, CO. Enderson, J.H. 1964. A study of the Prairie Falcon in the central Rocky Mountain region. Auk 81:332-352 . 1965. Roadside raptor count in Colorado. Wil- son Bull. 77:82-83. Harmata, A., J.E. Durr and H. Geduldig. 1978 Home range, activity patterns, and habitat use of Prai- rie Falcons nesting in the Mojave Desert. Report No. YA-512-CT8D-43, Bureau of Land Management, U.S. Department of the Interior, Denver, CO. Johnson, D. and J.H. Enderson. 1972. Roadside rap- tor census in Colorado — winter 1971-72. Wilson Bull 84:489-490. Kenward, R.E. 1978. Hawks and doves: factors af- fecting success and selection in goshawk attacks on Woodpigeons. J. Anim. Ecol. 47:449-460. Koplin, J.R. 1973. Differential habitat use by sexes of American Kestrels wintering in northern California. Raptor Res. 7:39-42. Mohr, C.O. 1947. Table of equivalent populations of North American small mammals. Am. Midi. Nat. 37. 223-249. Neu, C.W., C.R. Byers and J.M. Peek. 1974. A tech- nique for analysis of utilization-availability data. / Wildl. Manage. 38:541-545. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. Preston, C.R. 1990. Distribution of raptor foraging in relation to prey biomass and habitat structure. Condor 92:107-112. Roalkvam, R. 1985. How effective are hunting pere- grines? Raptor Res. 19:27-29. Smallwood, J.A. 1987. Sexual segregation by habitat in American Kestrels wintering in south-central Flor- ida: vegetative structure and responses to differential prey availability. Condor 89:842-849. Stahlecker, D.W. 1978. Effect of a new transmission line on winter prairie raptors. Condor 80:444-446. Received 27 August 1991; accepted 15 November 1991 J. Raptor Res. 26(l):19-23 © 1992 The Raptor Research Foundation, Inc. NORTHWARD POST-FLEDGING MIGRATION OF CALIFORNIA BALD EAGLES W. Grainger Hunt and Ronald E. Jackman BioSystems Analysis, Inc., 303 Potrero, 29-203, Santa Cruz, CA 95060 J. Mark Jenkins Ecological Services Group, Department of Engineering Research, Pacific Gas and Electric Company, 3400 Crow Canyon Road, San Ramon, CA 94583 Carl G. Thelander and Robert N. Lehman BioSystems Analysis, Inc., 303 Potrero, 29-203, Santa Cruz, CA 95060 Abstract. — We radiotracked five Bald Eagles ( Haliaeetus leucocephalus) on migration from natal areas in northern California. All flights were northerly in direction and were probably aimed at salmon carrion associated with spawning runs in British Columbia and Alaska. Two migrating juveniles traveled north- west to the Pacific Ocean, then north along the coast, and three eagles migrated north along the crest of the Cascade Mountains. These observations indicate a functional migration with food as its target. Migracion hacia el norte de Aguilas Cabeciblancas ( Haliaeetus leucocephalus) jovenes que dejan el nido, en California EXTRACTO. — Hemos radiocontrolado cinco Aguilas Cabeciblancas {Haliaeetus leucocephalus ) jovenes que han dejado el nido, en migracion desde su area natal en California del norte. Todos los vuelos fueron en direccion norte, probablemente en busca de carroiia de peces salmon que desovan en Colombia Britanica y Alaska. Dos de estas aguilas migrantes viajaron en direccion noroeste hacia el Oceano Pacifico, y luego hacia el norte a lo largo de la costa; y las otras tres de ellas, migraron hacia el norte a lo largo de la cumbre de las montanas Cascade. Estas observaciones indican migraciones funcionales que tienen el alimento como objetivo. [Traduction de Eudoxio Paredes-Ruiz] A northward migration of post-fledging Bald Ea- gles {Haliaeetus leucocephalus) from temperate lati- tudes has been suspected since Broley (1947) re- ported recoveries of birds banded as nestlings in Florida appearing in the northeastern United States and Canada. Beebe (1974:44) suggested that juvenile Bald Eagles from southern British Columbia leave natal areas soon after fledging and travel to salmon spawning runs along the coastal rivers of northern Canada and Alaska. Similarly, Servheen and En- glish (1979) speculated that young eagles moved north with adults to early salmon runs after leaving nests on Puget Sound, Washington, and that the timing of salmon spawning influenced eagle movement in the region. Hodges et al. (1987) demonstrated that post-fledging eagles remained on their natal Chilkat River in southeast Alaska until salmon availability declined in December; the eagles then migrated southward. We describe post-fledging local move- ments, departure, and migration of five juvenile Bald Eagles radiotagged as nestlings in California. Re- sults are consistent with the hypothesis of a north- ward, post-fledging migration, presumably directed at post-spawn salmon carrion in British Columbia and Alaska. Methods During 15-20 June 1984-86, we attached radiotrans- mitters to five nestling Bald Eagles (age 8-10 wk) in their nests in California; three along the Pit River near Fall River Mills (40° 59'N 121° 28' W) and two in the Eagle Lake vicinity (40° 33'N 120° 46' W) near Susanville. All nests were in ponderosa pine {Pinus ponderosa) forests near water. The birds wore back-pack style, 2-stage transmit- ters weighing 18-28 g, attached by 13 mm-wide teflon tubular ribbon. Ribbons were connected over the carina with cotton thread to permit eventual transmitter loss. We monitored pre-migratory behavior from the ground, and used fixed-wing aircraft to track eagles from their natal areas. Upon locating an eagle en route, we landed at air- ports ahead of the eagles’ course to reduce air time. We followed them until adverse weather or lack of airports prevented further tracking. 19 20 Hunt et al. Vol. 26, No. 1 For supplemental information on post-fledging behavior and departure, we radiotagged 1 1 additional nestlings and one fledgling in northern California (11 June to 25 July 1983-86) with back-pack or single-stage tail-mount trans- mitters (5 g, tied/glued to a single tail feather at least two- thirds emerged from sheath). We did not track the mi- grations of these 11 birds. Results and Discussion In the weeks following fledging in July, we mon- itored activities of the radio-tagged juveniles. They tended to remain near natal nests during the first week and utilized mostly snags and nest trees for perching. The young eagles soon moved to nearby lakes or rivers within natal territories during the day, often returning to roost in the vicinity of the nest. Their parents, year-long residents in northern California, provided food throughout the post-fledg- ing period, but juveniles occasionally found carrion, usually fish. Some eagles were more precocious than others in making their first flights and in venturing to other parts of the territories. At 14-16 wk, ju- veniles began soaring during mid-day for days or even weeks prior to departing unexpectedly. Pre- migratory soaring flights were local in nature; the eagles always returned to the territory on the same day. We made no attempts to locate the outer ranges of these pre-migratory flights. Fifteen of 17 juvenile eagles departed from their natal areas from 19 July to 22 August (x = 2 August) at an average age of about 16 wk (range 14-17). Of the remaining two birds, one had radio failure and probably departed after 3 August; the other eagle died, apparently drowning after 17 August as a re- sult of a wing injury. Using 12 wk as a standard fledging age, the interval between fledging and de- parture averaged 3.5 wk (range 2-5 wk). In contrast, post-fledging Bald Eagles marked in central Sas- katchewan (Besnard Lake) initiated “oriented mi- gratory movements,” generally to the south, seven or eight wk after fledging in mid- August at age 12 or 13 wk (Gerrard et al. 1974). Maine juveniles dispersed 5 to 10 wk (x = 7 wk), after fledging at 11-13 wk, and apparently moved southwest along the Atlantic seaboard (McCollough 1986). All five birds tracked on migration departed be- tween 1130-1330 H. Four, whose roosting places were located, traveled an average of 157 km (range 87-203 km, straight line distances between nest area and roost site) on the day of departure (Table 1). The onset of the northward migration for the Cal- ifornia juveniles was triggered by unknown causes. The apparent lack of carrion fish, previously abun- dant during the early nesting cycle, may have af- fected their attachment to natal areas. The south- ward migrations of young eagles described by Hodges et al. (1987) in southeast Alaska and Harmata et al. (1985) in central Saskatchewan were apparently influenced by the decline of foraging opportunities. Routes selected by migrating eagles (Fig. 1) showed preferred directions (Baker 1984), although physi- ography modified actual course. Harmata et al. (1985) speculated that physiography facilitated nav- igation of juvenile Bald Eagles moving along a pre- selected compass direction. Eagles 1 and 2 migrated along the crest and western slopes of the Cascade Mountains, respectively. Eagles 3, 4, and 5 moved northwest until encountering the Pacific coast which they followed northward. Occasional sightings of migrating eagles from both ground and aircraft indicated that they traveled alone, suggesting that migratory directions were genetically influenced. Interestingly, the migratory paths of sib- ling eagles 4 and 5 were far more similar than those of the other three eagles (Fig. 1). Directional or route programs might be encoded most similarly in the genes of close relatives and within demes. Geograph- ic differences in natal origin would likely demand appropriate differences in migration strategies. Overall, eagles following the coast traveled more slowly (x = 130 km/d, N = 14 d) than those fol- lowing the mountains (x = 184 km/d, N = 21 d). Slower travel may have been due to prevailing head- winds and coastal fog preventing thermal activity needed for soaring. Eagle 4, for example, contended with fog and northerly winds along the Oregon coast and averaged only 89 km per day (Table 1). Its daily travel increased to 231 km in Washington where clearer skies and more favorable winds prevailed. Harmata (1984) reported an average migration rate of 180 km/d (range 33-435) for adult Bald Eagles on northward spring migration from the San Luis Valley of south-central Colorado; no movement oc- curred on overcast days or when winds exceeded 56 km /hr prior to 0900 H. All five California eagles migrated during warm periods of the day. They departed their roost sites in late morning (typically 1000-1100 H) and usually stopped migrating about 1700-1800 H, a schedule that coincided with thermal activity. This pattern is consistent with behavior of the adult spring migrants from the San Luis Valley (Harmata 1984). Daily flights of two juvenile Bald Eagles departing their March 1992 Bald Eagle Migration 21 Table 1 . Roosting locations of five radio-tagged juvenile bald eagles during their northward migrations from California. Locations in parentheses indicate that the eagle was still soaring when radiotracking was terminated that day (times given). Distances in parentheses are 2-d totals. Date Location (N Lat. W Long.) Km Moved Eagle No. 1 (1984) 8/06 Pit River near Burney, CA (40°59' 121°45') Nest 8/07 Klamath Marsh, OR (42°52' 121°40') 208 8/08 18 km NNE Mt. Jefferson, OR (44°49' 121°42') 216 8/09 35 km SW Cle Elum, WA (47°01' 121°19') 247 8/10 10 km W Princeton, BC (49°28' 120°40') 257 8/11 20 km ESE Williams Lake, BC (52°10' 121°52') 305 8/12 11 km WSW Prince George, BC (53°54' 122°55') 199 Eagle No. 2 (1984) 7/23 Fall River Mills, CA (40°59' 121°28') Nest 7/24 24 km E Klamath Falls, OR (42°13' 121°28') 137 7/25 Hills Creek Reservoir, OR (43°36' 122°28') 175 7/26 10 km E Crabtree, OR (44°39' 122°46') 119 7/27 Mount Saint Helens, WA (46°10' 122°11') 176 7/28 (Fairfax, WA at 1711 H) (47°01' 122°01') >94 7/29 11 km SW Darrington, WA (48°12' 121°43') 134 7/30 near Princeton, BC (49°26' 120°50') 152 7/31 (N of Kamloops, BC at 1750 H) >160 Eagle No. 3 (1985) 7/30 Baum Lake, CA (40°55' 121°31') Nest 7/31 Probably near Montague, CA — 8/01 25 km SW Corvallis, OR (44°23' 123°25') (418) 8/02 12 km NNE Manzanita, OR (45°48' 123°52') 160 8/03 NE Tokeland, WA (46°46' 123°55') 108 8/04 Near Quinalt River, WA (47°24' 124°03') 72 8/05 Cape Flattery, WA (48°23' 124°43') 120 Eagle No. 4 (1986) 8/16 Eagle Lake, CA (40°33' 120°46') Nest 8/17 10 km NW Fall River Mills, CA (41°04' 121°33') 88 8/18 8 km SW Cave Junction, OR (42°08' 123°44') 218 8/19 Cape Sebastian, OR (42°19' 124°25') 57 8/20 10 km S Port Orford, OR (42°41' 124°26') 39 8/21 3 km S Coos Bay, OR (43°21' 124°24') 76 8/22 13 km SE Florence, OR (43°54' 123°59') 69 8/23 ? — 8/24 11 km S France, WA (46°28' 123°25') (291) 8/25 6 km SE Seiku, WA (48°13' 124°15') 204 8/26 16 km SE Kelsey Bay, BC (50°17' 125°48') 258 8/27 (19 km S Kingcome, BC at 1340 H) (50°47' 126°04') >60 Eagle No. 5 (1986) 8/22 Eagle Lake, CA (40°33' 120°46') Nest 8/23 6 km W Callahan, CA (41°18' 122°54') 197 8/24 5 km E Brookings, OR (42°05' 124°14') 144 8/25 5 km E Gold Beach, OR (42°25' 124°22') 40 8/26 10 km E Port Orford, OR (42°46' 124°22') 41 22 Hunt et al. Vol. 26, No. 1 Figure 1 . Roosting locations of five radio-tagged juvenile Bald Eagles during their northward migrations from Cal- ifornia. Locations in parentheses indicate that when ra- diotracking was terminated for the day the eagle was still migrating (soaring). All eagles continued to migrate be- yond the last point of detection. natal Besnard Lake in central Saskatchewan were initiated between 1100-1300 H and terminated be- tween 1400-1550 H (Harmata et al. 1985). The late morning timing for initiation of the eagles’ daily migration contrasted with an adult, north-bound Arctic Peregrine Falcon ( Falco peregrinus ) in spring which began migrating each morning before dawn (W.G. Hunt, unpubl.). The California juvenile eagles seemed to prefer wooded east-facing slopes for roosting, and mature conifers when available for perching. Eastern ex- posures were selected presumably because the sun would warm such terrain earlier the next morning and produce thermals. We estimated typical alti- tudes of migrants at 200-500 m above the terrain on clear, warm days. The altitude of migratory flights for migrating Saskatchewan juveniles reportedly ranged from 54-180 m, but weather conditions were mostly overcast (Harmata et al. 1985). We did not detect foraging activity along migration routes. Eagle 1 departed after mid-day and roosted the first night at the Klamath Marsh, Oregon (Table 1). The mountainous route chosen by this bird con- tinued along the ridge of the Cascades and passed close to the major peaks (Mt. Hood, Mt. Adams and Mt. Rainier). Five days after departure, Eagle l’s signal emanated from the heart of a significant thun- derstorm near Princeton, British Columbia, and we last detected it at Prince George, British Columbia. Eagle 2 followed the western slope of the Cascades and was last detected on a 20° magnetic bearing from Kamloops Lake moving in the direction of a major thunderstorm. Eagles 3, 4, and 5 migrated northwest to the Pa- cific Ocean and then north along the coast, although Eagle 3 reached the coastline further north than the others (Fig. 1). Eagles 3 and 4 roosted on the south shore of the Strait of Juan de Fuca, then crossed it the next morning. We lost contact with Eagle 3 in adverse weather after crossing the strait. Eagle 4 proceeded 258 km over the central mountains of Vancouver Island and reached the Strait of Georgia at Kelsey Bay, British Columbia. The next day we followed it across water to mainland British Colum- bia, but were obliged to turn back when the bird entered high mountains enshrouded in clouds. If the eagle maintained its course and speed it would have passed the Skeena River in the area of Terrace, British Columbia, the next afternoon, and by the following day it would have reached southeast Alas- ka. The region from the Skeena northwestward con- tains abundant salmon carrion from late July through October. Chum Salmon ( Oncorhynchus keta) car- casses normally peak during the first week of August and Pink Salmon ( O . gorbuscha ) during the second and third weeks of August (K. Imamura, pers. comm.). Similarly, the courses of Eagles 1 and 2 would have taken them to salmon areas. From its last detection (12 August) at Prince George, British Columbia, Eagle 1 was in the vicinity of Stuart River tributaries, a part of the Frazer River system where spawning of a race of Sockeye Salmon ( O . nerka) peaks in late July to early August (J. Adams, pers. comm.). Alternatively, Eagle 1 could have reached southeast Alaska within 2-3 d of its last radio de- tection. Results on post-fledging movements of juvenile Bald Eagles presented here are evidence of a func- March 1992 Bald Eagle Migration 23 tional migration rather than dispersal. All eagles departed within a narrow time frame, traveled alone in the same direction, displayed strong course fidel- ity, and showed consistent daily activity patterns. Timing of migration coincided with the occurrence of salmon carrion in areas reached by the eagles within 1-2 wk after leaving natal sites. The paths of the five migrant juveniles suggested two routes to the northern salmon areas. Eagles 1 and 2 followed the Cascade Range into interior British Columbia; eagles 3, 4, and 5 flew northwestward to the coast and followed it northward. The mountain migrants moved faster but must have occasionally encountered extreme turbulence (thunderstorms). The coastal migrants moved slower, but were assured of reaching the salmon areas if they could survive a possible increase of migration time imposed by weather. Ea- gles apparently traveled without foraging or resting during the day, suggesting a substantial caloric re- ward at the final destination. Acknowledgments This study was funded as part of a larger research project conducted under the direction of the Pacific Gas and Electric Company by BioSystems Analysis, Inc. We thank C. Brown, M. Glynn, J. Gruter, C. Himmelwright, P. Hunt, M. Hunt, B. Hunt, J. Linthicum, K. Teare, P. Roush, B. Gaussoin, and L. Small for assistance in the field. R. Knight, G. Keister, A. Harmata, T. Brown, and P. Kerlinger gave valuable comments on the manuscript. Literature Cited Baker, R.R. 1984. Bird navigation: the solution of a mystery? H.M. Holmes and Meier Publishers, New York. Beebe, F.L. 1974. Field studies of the Falconiformes of British Columbia. Occasional Papers of the British Columbia Provincial Museum, No. 17, Victoria, BC, Canada. Broley, C.L. 1947. Migration and nesting of Florida Bald Eagles. Wilson Bull. 59:3-20. Gerrard, P., J.M. Gerrard, D.W.A. Whitfield and W.J. Maher. 1974. Post-fledging movements of ju- venile Bald Eagles. Blue Jay 32:218-226. Harmata, A.R. 1984. Bald Eagles of the San Luis Valley, Colorado: their winter ecology and spring mi- gration. Ph.D. thesis, Montana State University, Boze- man, MT. , J.E. Toepfer and J.M. Gerrard. 1985. Fall migration of Bald Eagles produced in northern Sas- katchewan. Blue Jay 43:232-237, with addendum in Blue Jay 44:1. Hodges, J.I., E.L. Boeker and A.J. Hansen. 1987. Movements of radio-tagged Bald Eagles, Hahaeetus leucocephalus , in and from southeastern Alaska. Can Field-Nat. 101:136-140. McCollough, M.A. 1986. The post-fledging ecology and population dynamics of Bald Eagles in Maine. Ph.D. thesis, University of Maine, Orono, ME. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. Servheen, C.W. and W. English. 1979. Movements of rehabilitated Bald Eagles and proposed seasonal movement patterns of Bald Eagles in the Pacific North- west. Raptor Res. 13:79-88. Received 15 November 1990; accepted 15 November 1991 /. Raptor Res. 26(1 ):24— 26 © 1992 The Raptor Research Foundation, Inc. DETERMINING SEX OF EASTERN SCREECH-OWLS USING DISCRIMINANT FUNCTION ANALYSIS Dwight G. Smith Biology Department, Southern Connecticut State University, New Haven , CT 06515 Stanley N. Wiemeyer U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center, Laurel, MD 20708 Abstract. — Seven morphometric characteristics and weight of males and females of a captive colony of Eastern Screech-Owls ( Otus asio) were compared. Females were significantly larger than males in weight, total length, and length of tail, wing and bill. A discriminant function analysis based on weight, wing and tail length correctly identified the sex of 88% of the 77 birds. Identificacion del sexo en el Tecolote Nororiental ( Otus asio) Extracto. — Se compararon siete caracteristicas morfometricas, y el peso de tecolotes ( Otus asio ) machos y hembras de una colonia cautiva. Las hembras fueron significativamente mas grandes que los machos en peso, largo total, y largo de la cola, alas y pico. Un analisis de funcion discriminante basado en peso, alas y largo de la cola, identified correctamente el sexo en un 88% del total de 77 aves. [Traduccion de Eudoxio Paredes-Ruiz] Earhart and Johnson (1970), Snyder and Wiley (1976), Mueller (1986) and McGillivray (1987) de- scribed sexual dimorphism in owls and noted that this is less pronounced in screech-owls and some other small owls than in many of the larger owls and diurnal birds of prey. Owen (1963a, 1963b) and Marshall (1967) reported screech-owl size varia- tions in association with differences in plumage col- oration and geographical location. The lack of con- spicuous sexual dimorphism makes it difficult to accurately sex Eastern Screech-Owls {Otus asio). We examined sexual dimorphism in a captive colony of Eastern Screech-Owls and used discriminant func- tion analysis (DFA) to determine the most effective combination of characters to distinguish the sex of individuals. Materials and Methods We measured seven morphological characteristics and weight of 77 live Eastern Screech-Owls that formed the captive breeding colony at Patuxent Wildlife Research Center. Details of this colony are provided in Wiemeyer (1987). Most of the screech-owls in this colony are from parent stock originally obtained from Ohio; a few were from Maryland. The sex of each individual was determined by laparos- copy. Measurements of tarsus, ulna, tail and toe follow Baldwin et al. (1931). Total length was measured as the distance from the feathers of the top of the head to the tip of the central tail feather; this measure was taken by plac- ing the owl’s head against an angle iron fixed to a ruler. Wing length was measured as the length of the flattened wing from the tip of the longest primary to the anterior surface of the distal end of the radius. Bill length was taken with a caliper from the tip to the base at the fron- tonasal hinge. Owls were weighed to the nearest 1 g with calibrated Pesola scales. All birds were measured and weighed between 27 February and 3 March 1986. All had hatched in captivity in 1984. Stepwise discriminant function analysis was used to identify the combination of characteristics that provided the highest discrimination between sexes. Data were stan- dardized following SPSSx procedures prior to entry. All statistical analyses followed SPSSx routines (Norusis 1985). Default criteria were used throughout and Ma- halonobis was the selected method for stepwise discrimi- nant function analysis. The critical level for all tests was et — 0.05. Results and Discussion Significant differences between sexes occurred in weight, total length, length of wing and tail, and bill size (Table 1). Weight showed the most obvious difference between sexes. Average weight was 18.2% greater in females than in males, which is generally consistent with weight differences reported by Ear- hart and Johnson (1970), VanCamp and Henny (1975) and Wiemeyer (1987). Female total length averaged 4.0% greater, and tail length 6.0% greater than males. Ulna, right middle toe and tarsus lengths showed the least variation between females and males. 24 March 1992 Determining Eastern Screech-Owl Sex 25 Table 1 . Morphometric comparisons of adult Eastern Screech-Owls based on measurements of 77 owls at the Patuxent Wildlife Research Center, Laurel, Maryland. Length is given in cm, weight in g, and all other measurements in mm. Ranges given in parentheses. Character Males x ± SD (N = 38) Females x ± SD (N = 39) t P Weight 171.0 ± 12.8 (145-208) 202.1 ± 21.9 (163-263) -7.65 <0.001 Length 20.0 ± 0.8 (18-21) 20.8 ± 0.8 (19-23) -4.31 <0.001 Wing 162.9 ± 4.0 (157-171) 169.9 ± 3.9 (163-177) -7.77 <0.001 Ulna 61.8 ± 3.5 (58-71 ) a 62.5 ± 3.2 (57-68) b -0.69 0.497 Bill 22.9 ± 1.1 (21-26) 23.6 ± 0.9 (21-26) -3.04 0.003 Tail 80.4 ± 3.1 (73-88) 85.2 ± 4.4 (78-98) -5.58 <0.001 Tarsus 28.4 ± 2.6 (23-33) 27.6 ± 2.2 (24-34) 1.49 0.141 Middle toe 24.0 ± 1.7 (22-29) c 24.4 + 1.7 (21-28) b -0.79 0.438 a N — 20. b N = 22. C N = 15. The largest individual was a female with wing length of 171 mm, tail length of 83.5 mm, and weight of 263 g. The smallest was a male with a wing length of 160 mm, tail length of 81 mm and weight of 145 g. Weights of captive owls averaged slightly higher than weights of captured wild owls or specimens obtained as fresh fall, winter and early spring road kills obtained in Connecticut between 1974 and 1990 (D.G. Smith, unpubl.); 39 wild females averaged 195.8 g (SD = 21.0), 52 males averaged 165.1 g (SD = 8.4). Although the weights of the wild and captive females did not differ ( t = —1.30, P = 0.199), the weights of the wild males were significantly smaller than those of the captive males ( t = —2.64, P = 0.010). The somewhat larger weights of the captive owls may reflect a comparatively sedentary activity and consistently available food. Wing, bill and toe lengths (claw included) of captive owls also were larger than wild owls. Differences in wing length may reflect reduced feather wear of captive owls while longer toes and bills may be associated with consumption of soft food, primarily ground meat and day-old chicks (Wiemeyer 1987) or lack of wear. The Classification Function (DF). The sex of 88.3% of the captive owls at Patuxent could be de- termined from a combination of weight (WT), wing (WG) and tail (T) lengths: DF = 0.1128(WG) + 0.0884(T) + 0.0337(WT) - 32.384 If DF < 0.0 the individual is classified as a male. Thirty-four of 38 (89.5%) males and 34 of 39 (87.2%) females were correctly identified using this equation. Three of the four misclassified males were larger, heavier birds with longer wings and tails. Converse- ly, the five misclassified females had shorter wings. Although weight may provide the best single dis- criminator between male and female Eastern Screech- Owls (Mueller 1986, this study), caution is needed when using weight as a criterion in wild or captive owls. Studies of radiotransmitter-equipped individ- uals revealed wide seasonal and sometimes daily weight changes (Smith and Gilbert 1984). Henny and VanCamp (1979) and Wiemeyer (1987) also reported seasonal changes in average weights of Eastern Screech-Owls. Because of variations in weight, discrimination functions based on other criteria were also deter- mined. A function with tail length (T), total length (TL) and wing length (WG) correctly identified the sex of 80.5% of the owls: DF = 0.0766(T) + 0.3900(TL) + 0.1818(WG) - 44.55 If DF < 0.0, classify the individual as male. This function correctly identified 84.6% of females and 76.3% of the males. Again, wing length was an im- portant factor in the misclassification of both males and females. Total length may be difficult to measure, es- pecially by inexperienced workers; thus a discrim- ination function based on wing and tail lengths was determined: 26 Dwight G. Smith and Stanley N. Wiemeyer Vol. 26, No. 1 DF = 0.1 6531 (T) 4- 0.37896(WG) - 76.74288 If DF < 0.0, classify the individual as male. This function correctly identified 77.9% overall; 78.9% of the males and 76.9% of the females. For the Eastern Screech-Owl, overlap between males and females was too great to permit identifi- cation of sex by any single characteristic. If weight is excluded from the equation, combinations of 2 or 3 of the most significantly different characteristics provided correct classification rates of only 78-81 %. Because of the geographic size variations (Owen 1963a, James 1970), and probable differences be- tween captive and wild owls, the discriminant func- tions we present may not be applicable to other data sets for Eastern Screech-Owls. Nevertheless, our ap- proach should be a useful model for constructing new discriminant functions for other wild Eastern Screech-Owl populations. Acknowledgments Kathy Ellis and Janet McMillan assisted in weighing and measuring birds at Patuxent. Samuel J. Barry, Rich- ard J. Clark, Mark R. Fuller, William A. Link, Jeffrey Marks, and Carl D. Marti provided useful comments on earlier drafts of this manuscript. Literature Cited Baldwin, S.P., H.C. Oberholser and L.G. Worley. 1931. Measurements of birds. Sci. Pubi. Cleveland Mus. Nat. Hist. 2:1-165. Earhart, C.M. and N.K. Johnson. 1970. Size di- morphism and food habits of North American owls. Condor 72:251-264. Henny, C.J. and L.F. VanCamp. 1979. Annual weight cycle in wild screech owls. Auk 96:795-796. James, F.C. 1970. Geographic size variation in birds and its relationship to climate. Ecology 51:365-390. Marshall, J.T., Jr. 1967. Parallel variation in North and middle American screech owls. Monogr. West Found. Vert. Zool. 1:1-72. McGillivray, W.B. 1987. Reversed sexual dimor- phism in 10 species of northern owls. Pages 59-66 in R.W. Nero, R.J. Clark, R.J. Knapton and R.H. Hamre [Eds.], Biology and conservation of northern forest owls General Technical Report RM-142, Forest Service, U.S. Department of Agriculture, Fort Collins, CO. Mueller, H.C. 1986. The evolution of reversed sexual dimorphism in owls: an empirical analysis of possible selection factors. Wilson Bull. 98:387-406. Norusis, M.J. 1985. SPSSx advanced statistics guide McGraw Hill Book Company, New York. Owen, D.F. 1963a. Variation in North American screech owls and the subspecies concept. Systematic Zool. 12. 8-14. . 1963b. Polymorphism in the screech owl in eastern North America. Wilson Bull. 75:183-190. Smith, D.G. and R. Gilbert. 1984. Eastern Screech- Owl home range and use of suburban habitats in south- ern Connecticut. /. Field Ornithol. 55:322-329. Snyder, N.F.R. and J.W. Wiley. 1976. Sexual size dimorphism in hawks and owls of North America. Ornithol. Monogr. 20. VanCamp, L.F. and C.J. Henny. 1975. The screech owl: its life history and population ecology in northern Ohio. North Amer. Fauna, No. 71, U.S. Fish and Wildlife Service, Denver, CO. Wiemeyer, S.N. 1987. Propagation of captive Eastern Screech-Owls. /. Raptor Res. 21:49-56. Received 2 August 1991; accepted 16 November 1991 J Raptor Res. 26(l):27-32 © 1992 The Raptor Research Foundation, Inc. DIET SHIFTS OF BLACK-CHESTED EAGLES ( Geranoaetus melanoleucus) FROM NATIVE PREY TO EUROPEAN RABBITS IN CHILE Eduardo F. Pavez and Christian A. Gonzalez Union de Ornitologos de Chile, Casilla 268, Correo Central, Santiago, CHILE Jaime E. Jimenez 1 Departamento de Ecologia, Universidad Catolica de Chile, Casilla 114-D, Santiago, CHILE Abstract. — For two years we studied the feeding ecology of Black-chested Eagles ( Geranoaetus melanoleu- cus) in San Carlos de Apoquindo, central Chile. We estimated the diet of eagles using three different methods and evaluated the abundance of introduced European Rabbits ( Oryctolagus cuniculus) throughout one year. Eagles preyed in decreasing order of frequency on mammals, reptiles, insects and birds. The biomass component of insects in the diet was negligible. At the species level, the European Rabbit was the most frequently consumed prey (43.9% by number and 81.7% by biomass). The diet composition changed only slightly throughout the year. During non-breeding seasons birds and insects were not consumed. Depending on the technique used, we found differences in the estimated diet. Birds were underestimated and insects were overestimated by direct observations of prey captures and transportations. In contrast, insects were underestimated by the analysis of prey remains. At the class level, the analysis of regurgitated pellets seems to represent the diet of eagles better. European Rabbits were present throughout the year in the study site, peaking during the eagles’ breeding season, when kittens were abundant. Eagles preyed preferentially on small rabbits and apparently avoided large individuals. Eagles have increased their consumption of European Rabbits over time as did foxes in the study area. Cambio en la dieta del Aguila Chilena ( Geranoaetus melanoleucus ) de presas nativas a Conejos Europeos en Chile. Extracto, — Durante dos anos estudiamos la ecologia trofica de Aguilas Chilenas {Geranoaetus mela- noleucus) en San Carlos de Apoquindo, en Chile central. Estimamos la dieta de las aguilas usando tres metodos diferentes y evaluamos la abundancia de Conejos Europeos {Oryctolagus cuniculus) introducidos a lo largo del ano. Las aguilas depredaron en orden de frecuencia decreciente sobre mamiferos, reptiles, insectos y aves. El componente de biomasa de insectos en la dieta fue insignificante. A1 nivel de especie, el Conejo Europeo fue la presa mas frecuentemente consumida (43.9% en numero y 81.7% en biomasa). La composition de la dieta solo cambio ligeramente a lo largo del ano. Durante la estacion no reproductiva aves e insectos no fueron consumidos. Dependiendo de la tecnica usada, encontramos diferencias en la estimation de la dieta. Las aves fueron subestimadas y los insectos sobreestimados mediante observaciones directas de capturas y transposes de presas. En contraste, los insectos fueron subestimados por los analisis de restos de presas. Al nivel de clase, los analisis de egagropilas parecian representar mejor la dieta de las aguilas. Los Conejos Europeos estuvieron presentes a lo largo de todo el ano en el sitio de estudio, alcanzando maximos durante la estacion reproductiva de las aguilas, cuando los conejos pequenos eran abundantes. Las aguilas depredaron preferentemente sobre conejos pequenos y aparentemente evitaron a los adultos. Las aguilas, al igual que los zorros, han aumentado el consumo de Conejos Europeos en el tiempo en el sitio de estudio. The Black-chested Eagle, Geranoaetus melanoleu- cus, is broadly distributed in South America, from Venezuela to Tierra del Fuego (Brown and Amadon 1 Present address: Department of Wildlife and Range Sci- ences, 118 Newins-Ziegler Hall, University of Florida, Gainesville, FL 32611. 1968). It is found throughout Chile (Goodall et al. 1951), especially in open habitats, from sea level up to 2200 m (Barros 1967). Information on the feeding ecology of this eagle is scarce (Jimenez and Jaksic 1989). Most authors report only qualitative and an- ecdotal information on its diet (see review by Ji- menez and Jaksic 1990). In general, previous au- 27 28 Pavez et al. Vol. 26, No. 1 thors concur that these eagles prey in decreasing frequency on birds, rodents, young rabbits, and rep- tiles. Results differ among localities, however (Ji- menez and Jaksic 1990). This implies that the eagles show considerable plasticity, opportunistically cap- turing and consuming the most abundant prey. Quantitative studies in central Chile report that the Black-chested Eagles prey primarily on rodents, especially on Octodon degus (75.9%, 53.8%, and 43.4% by number of total prey, according to Schlatter et al. 1980, Jimenez and Jaksic 1989, 1990, respec- tively), and secondarily on European Rabbits ( Oryc - tolagus cumculus; 18.8%, 13.8%, according to Schlat- ter et al. 1980, and Jimenez and Jaksic 1989, respectively). The relatively low incidence of Eu- ropean Rabbit in the diet of the eagles, and in those of other sympatric predators, has been attributed to the “lack of behavioral adjustment” of native pred- ators to hunt “for this recently introduced prey” (Jaksic and Soriguer 1981). Here, we analyze the seasonal variation in the diet of eagles using three different techniques, and compare the results with the estimated abundance of rabbits in the field. We show that since earlier studies, eagles have markedly increased their con- sumption of introduced rabbits. Study Area and Methods We studied Black-chested Eagles in a 900-ha area lo- cated in San Carlos de Apoquindo (33°23'S 70 o 31'W), at the foothills of the Andes (1200 m elevation), 20 km east of Santiago, Chile. The study site has rugged topography, with high ridges, deep ravines, and few flat areas. The climate is of the Mediterranean type with vegetation com- posed primarily of evergreen shrubs. Additional infor- mation on the study site can be found in Jimenez and Jaksic (1989). Eagles remained in the study site the year round. Two pairs nested in 1987 and three in 1988. We studied the diet of the eagles using pellets (e.g., Marti 1987), prey remains, and observations of kills and prey carried. We analyzed 236 regurgitated pellets, and 158 prey remains collected under 5 nests and 11-15 roost- ing places every other week between December 1987 and March 1989. We recorded the capture and transport of prey by eagles during an entire day at least every other week from August 1987 to September 1988. These ob- servations were conducted from 0800-1900 H from van- tage points that provided good visibility over the entire area, using binoculars (16 x 50) and a spotting scope (20- 40 x 60). We divided the year into two periods: the non- breeding season, from March through August, and the breeding season, from September through February (see Jimenez and Jaksic 1989). Prey were identified at least to class. We assessed the abundance of European Rabbits in the area by counting all the individuals observed within a 70- m-wide and 6-km-long transect while walking between 0800-1000 H and between 1800-2000 H. We surveyed the rabbits during 27 transects between November 1987 and October 1988. Both the observed rabbits and those captured and transported by the eagles were assigned to three size classes: large, medium, and small. These size classes corresponded roughly to the weight classes deter- mined by Zunino and Yivar (1983-85; larger than 1100 g, between 800 and 1100 g, and smaller than 800 g, re- spectively). Results Diet. The 624 prey items were composed of mam- mals (78.2%), reptiles (11.4%), insects (7.8%) and birds (2.6%; Table 1). The most frequently con- sumed prey was the European Rabbit (43.9%), fol- lowed by the rodent Octodon degus (18.9%) and in turn by the snake Philodryas chamissonis (9.5%). In- sects were primarily coleopterans, which were cap- tured by juvenile eagles during their early hunting flights. By biomass, the bulk of the eagle’s diet was com- posed of mammals (95.8%), 81.7% of the total bio- mass were European Rabbits (Table 1). In decreas- ing frequency eagles preyed on reptiles (3.4%) and birds (0.8%). Insects added a small amount of bio- mass (0.02%). Diet by Season. Prey classes were similar for the four seasons of this study (Table 1). At the species level, the rank order in frequencies of prey eaten by the eagles was positively correlated between the 1987 and 1988 breeding seasons (Spearman rho = 0.882, P < 0.001), as well as with that of the 1988 non- breeding season (rho = 0.951 and 0.852, P < 0.001, respectively). The 1987 non-breeding season showed the same trend, but we did not use statistics due to the small sample. Neither European Rabbits, nor reptiles were consumed during the 1987 non-breed- ing season. Birds and insects were apparently not consumed during the non-breeding seasons (Table 1). During breeding seasons, the number of different prey taken by the eagles increased, as birds and insects were incorporated in the diet (Table 1). Influence of Technique Used. Estimates of the eagle’s diet at the class level were influenced by study technique (Table 2; x 2 = 39.39, df = 6, P < 0.005). Although we found significant differences only for total prey and for insects, some trends were evident among the other prey classes. Direct observations of capture and transport of prey by eagles apparently overestimated consumption of insects and underes- timated birds and reptiles (Table 2). On the other March 1992 Diet Shifts of Chilean Eagles 29 Table 1. The diet, in percentage of items, of Black-chested Eagles in a central Chilean site during sucessive breeding (Bree.) and non-breeding (Non-Br.) seasons. For the entire study period, the percent biomass is also shown. Subtotals for class are in parentheses. Prey Weight 1 Non-Br. 1987 Bree. 1987 Non-Br. 1988 Bree. 1988 Total by Number Biomass Mammals (100.0) (74.5) (95.2) (82.2) (78.2) (95.80) Oryctolagus cuniculus 800 0.0 47.5 47.6 39.0 43.9 81.69 Octodon degus 184 44.4 21.5 33.3 11.7 18.9 8.05 Abrocoma bennetti 231 0.0 1.1 4.8 4.7 2.4 1.29 Phyllotis darwini 62 0.0 0.5 0.0 0.5 0.5 0.07 Unidentified rodent 159 44.4 3.1 9.5 10.3 6.4 0.18 Unidentified mammal 319 11.2 0.8 0.0 16.0 6.1 4.52 Birds (0.0) (2.6) (0.0) (2.8) (2.6) (0.78) Zenaida auriculata 137 0.0 0.3 0.0 0.0 0.2 0.05 Metriopelia melanoptera 125 0.0 0.0 0.0 0.9 0.3 0.10 Unidentified bird 131 0.0 2.3 0.0 1.9 2.1 0.63 Reptiles (0.0) (12.9) (4.8) (9.9) (11.4) (3.40) Philodryas chamissonis 150 0.0 10.8 4.8 8.0 9.5 3.30 Callopistes palluma 65 0.0 0.5 0.0 0.5 0.5 0.07 Unidentified iguanid 9 0.0 1.6 0.0 1.4 1.4 0.03 Insects (0.0) (10.0) (0.0) (5.1) (7.8) (0.02) Unidentified scarabaeid 1 0.0 2.2 0.0 0.0 1.3 0.00 Unidentified coleopteran 1 0.0 1.8 0.0 4.2 2.5 0.01 Unidentified insect 1 0.0 6.0 0.0 0.9 4.0 0.01 Total prey items 2 9 381 21 213 624 100.00 Pellets 0 266 4 112 382 Prey remains 0 53 4 101 158 Transport and capture 9 62 13 0 84 1 Weights (in g) were taken from Zunino and Vivar (1983-85) for Oryctolagus, from J. E. Jimenez (unpublished data) for Zenaida, and from Jimenez andjaksic (1989) for the other taxa. Weights for unidentified prey are the average weight of the most related identified taxa, except for Iguanidae, for which we averaged the weights of the species that occur at the study site. We assumed that insects weighed 1 g - 2 The totals may be overestimated because some prey may have been counted more than once. hand, pellets and prey items yielded similar results, except for a relative underestimation of insects using prey items. Abundance of European Rabbits and Their Consumption by Eagles. The estimated abundance of rabbits in San Carlos de Apoquindo varied throughout the year. The number of rabbits observed on transects (N =56) was greater during the eagle’s breeding (74.1%) than non-breeding season (25.9% of the monthly average count). Counts peaked in December and January. A similar pattern applied to both large and small rabbits (2.51 and 1.12 vs. 1.04 and 0 daily average count for breeding and non- breeding seasons, respectively; Fig. 1). Medium-sized rabbits were more common during the non- breeding season (0.28 vs. 0.17 daily average count, for non- breeding and breeding seasons, respectively). Data for rabbits captured by eagles during years when the rabbit abundances were estimated, showed that of 21 observed captures, 68.2% were made dur- ing the breeding season and 31.8% during the non- breeding season (Fig. 1). Of the rabbits captured 57.1% were small, and all of these were taken during the breeding season. Comparing the abundances of different rabbit size classes in the field with those captured by eagles shows that eagles preyed on rabbits primarily when kittens were available (Fig. 1). Eagles did not cap- ture rabbits independently of their body sizes (x 2 = 18.27, df = 2, P < 0.001). When kittens were not 30 Pavez et al. Vol. 26, No. 1 Table 2. Diet of Black-chested Eagles in a central Chilean site based on pellets, prey items, and direct observation of transport and capture of prey. Percent of prey in the diet of eagles at the class level and for Oryctolagus cuniculvs, and statistics for the independence tests are shown. Pellets Items in Nests Items Carried Total % Number X 2 df P Mammals 75.9 84.8 76.2 78.2 488 1.15 2 >0.25 Oryctolagus cuniculus 45.5 45.6 33.3 43.9 274 2.67 2 >0.25 Birds 3.4 1.9 0.0 2.6 16 5.47 2 >0.05 Reptiles 12.3 12.7 4.8 11.4 71 4.49 2 >0.1 Insects 8.4 0.6 19.0 7.8 49 27.43 2 <0.005 Total items 382 158 84 624 39.39 6 <0.005 available (non-breeding season), eagles seemed to prey on medium-sized rabbits (Fig. 1). Discussion Our results concur with those of previous studies on the diet of eagles, by showing that the eagles prey primarily on mammals (>70% by number; see Schlatter et al. 1980, Jimenez and Jaksic 1989, 1990). We also found that reptiles were their second most frequent prey. However, unlike previous work, we found that insects were more frequent in the diet than were birds. Former studies recorded no insect prey. These results may reflect differences in prey availability, different prey (Jimenez and Jacksic 1990), or different food study techniques (but see below). Mammals were the staple prey of eagles during the two years and during breeding and non-breeding seasons. There were seasonal differences in the other prey classes. Birds and insects were prey only during breeding seasons, when reptiles also tended to be more commonly eaten. Although European Rabbits were present during the 1987 non-breeding season, they were not detected as a component of the eagles’ diet, perhaps because of the small sample. A seasonal difference in diet reported by Jimenez and Jaksic (1990) for several eagle populations, was also ob- served in this population, suggesting that eagles prey opportunistically on the most abundant prey avail- able. In other raptor studies, the results of dietary es- timation with different techniques are controversial. Whereas Collopy (1983) found no differences in dietary estimates by using different techniques, oth- ers have found differences. Unlike our findings, the use of prey remains as diet estimators has been re- ported to overestimate the vertebrate prey taken, especially for those prey for which only the bones are found, and to underestimate invertebrate prey (Errington 1932, Snyder and Wiley 1976, Simmons et al. 1991). In contrast, pellet analysis has been cc LLI m 3 2.5 NON-BREEDING SEASON 2 - 1.5 - 1 - 0.5 LARGE MEDIUM SMALL Figure 1 . Daily average number of large-, medium-, and small rabbits observed on transects ( N = 56, open bars) and captured by Black-chested Eagles ( N = 21, closed bars) during breeding and non-breeding seasons in central Chile. March 1992 Diet Shifts of Chilean Eagles 31 reported to underestimate the occurrence of verte- brate prey (Errington 1932). At the prey species level, we found some differ- ences with previous work. Our data show that the most important mammal in the eagle’s diet was the European Rabbit. Despite working in the same study site during 1984-85, Jimenez and Jaksic (1989) found that rabbits were only the second most im- portant prey item, comprising 13.8% of the Black- chested Eagle’s diet. Schlatter et al. (1980), working in a nearby site during 1973-74, also reported a low incidence of rabbits in the diet (18.8% of prey items). It appears, then, that eagles are now eating more rabbits than they used to. Interestingly, studies on the diet of foxes ( Pseu - dalopex culpaeus ) in the same study site also showed a substantial increase in rabbit consumption over time: from 19.7% in 1976 to 37.0% in 1983, and to 48.0% in 1984 (Jaksic et al. 1980, Simonetti 1986, Iriarte et al. 1989b, respectively). The increase in European Rabbits in the diet of some central Chilean predators over time may be due to an increase in the number of rabbits. Unfor- tunately there are no previous density estimates available for our study site. The habitat conditions in San Carlos de Apoquindo seem to have remained the same during the past 15 years according to our casual observations. Therefore, an increase of rabbits owing to the clearing of brush, as suggested by Fu- entes and Jaksic (1980), is probably not the case. Rabbits have been present at the site for more than 50 years according to local residents. This obser- vation refutes the hypothesis that rabbits are still arriving from elsewhere into the site. The human presence in the area seems to have increased in recent years (pers. observation, P. Ra- mirez, pers. comm.). People come to the area mainly to hunt and trap rabbits. Consequently, if humans have some influence on rabbit abundances, their in- creased activities in San Carlos de Apoquindo should have been detrimental rather than favorable for the rabbit populations. The explanation of Simonetti (1986), that the diet shift of foxes toward more con- sumption of rabbits in the study area “may relate to artificial causes, such as increased human trapping of rabbits,” does not apply to eagles, as we have never seen them capturing rabbits caught in snare traps. As Simonetti (1986) pointed out, the hypoth- esis of a dietary shift toward rabbits because the relative abundance of native rodents may have de- creased (Jaksic and Soriguer 1981), does not explain the change observed, since native rodents are now at least as abundant as before judging from rodents per trap per day encountered by Jaksic et al. (0.03; 1981), Simonetti (0.06; 1986), Iriarte et al. (0.10; 1989a), and Jimenez and Jaksic (0.14; 1989). To our knowledge, so far only Iriarte et al. (1989a) attempted to estimate the abundance of rabbits in central Chile over time. Although they were able to capture only kittens and subadults, their density es- timates throughout a year concur roughly with our estimates. As in our findings, Zunino and Vivar (1983-85) documented that European Rabbits in- creased in numbers as they reproduce between Au- gust and February. The number of rabbits taken by eagles showed the same trend as rabbit abundance along the year. The fact that eagles prey more on rabbits when they are more abundant in the field implies that Black-chested Eagles exhibit a func- tional response to this prey. The apparent avoidance of adult rabbits by eagles may be explained by their size. Adult rabbits were abundant throughout the year in our study site. They are probably too large to be captured and success- fully handled by eagles. This may explain why adult rabbits behaved as if they were immune from pre- dation (see also Jaksic et al. 1979a, 1979b, Jaksic and Ostfeld 1983). Acknowledgments We are indebted to the Sporting Club of the Catholic University of Chile, which allowed us to work on their property. We thank J.L. Besa and P. Fiegehn for helping us with the field work. Earlier versions of the manuscript were greatly improved by comments and constructive crit- icisms made by M.W. Collopy, J.A. Donazar, R. Ed- wards, P. Feinsinger, E. Inigo-Elias, F.M. Jaksic, M.N. Kochert, K.H. Redford, J. Rottmann, J.C. Torres-Mura, and P. Widen. This work was conducted while J.E.J. held a grant from the World Wildlife Fund (No. 1297), and was completed under tenure of fellowships from the Ful- bright Commission and the Program for Studies in Trop- ical Conservation of the University of Florida. Literature Cited Barros, R. 1967. Notas sobre el tiuque cordillerano y el aguila. Anales de la Academia de Ciencias Naturales de Chile 30:105-112. Brown, L. and D. Amadon. 1968. Eagles, hawks and falcons of the World. McGraw-Hill Co., New York Collopy, M.W. 1983. A comparison of direct obser- vations and collections of prey remains in determining the diet of Golden Eagles. J. Wildl. Manage. 47:360- 368. Errington, P.L. 1932. Technique of raptor food habits study. Condor 34:75-86. 32 Pavez et al. Vol. 26, No. 1 Fuentes, E.R. and F.M. Jaksic. 1980. Consideraciones teoricas para el control biologico del Conejo Europeo en Chile central. Medio Ambiente {Chile) 4:45-49. Goodall, J.D., A.W. Johnson and R.A. Philippi. 1951. Las aves de Chile. Volume II. Platt Establecimientos Graficos, Buenos Aires, Argentina. Iriarte, J. A., L.C. Contreras and F.M. Jaksi6. 1 989a. A long-term study of a small-mammal assemblage in the central Chilean matorral. J. Mammal. 70:79-87. , J.E. Jimenez, L.C. Contreras and F.M. Jaksic. 1989b. Small mammal availability and consumption by the fox Dusicyon culpaeus, in central Chilean scrub- lands. J. Mammal. 70:641-645. JaksiO, F.M. and R.C. Soriguer. 1981. Predation upon the European Rabbit {Oryctolagus cuniculus) in Med- iterranean habitats of Chile and Spain: a comparative analysis. J. Anim. Ecol. 50:269-281. AND R.S. Ostfeld. 1983. Numerical and be- havioral estimates of predation upon rabbits in Med- iterranean-type shrublands: a paradoxical case. Revista Chilena de Historia Natural 56:39-49. , E.R. Fuentes and J.L. YA$ez. 1979a. Two types of adaptation of vertebrate predators to their prey. Archivos de Biologia y Medicina Experimentales {Chile) 12:143-152. , and 1979b. Spatial distribution of the Old World rabbit {Oryctolagus cuniculus) in cen- tral Chile. J. Mammal. 60:207-209. , R.P. Schlatter and J.L. YAnez. 1980. Feed- ing ecology of central Chilean foxes, Dusicyon culpaeus and Dusicyon griseus. J. Mammal. 61:254-260. , and E.R. Fuentes. 1981. Assessing a small mammal community in central Chile. J. Mam- mal. 62:391-396. Jimenez, J.E. and F.M. Jaksic. 1989. Behavioral ecol- ogy of Grey Eagle-Buzzards, Geranoaetus melanoleucus, in central Chile. Condor 91:913-921. AND . 1990. Historia natural del aguila Geranoaetus melanoleucus: una revision. Hornero {Ar- gentina) 13:97-110. Marti, C.D. 1987. Raptor food habits studies. Pages 67-80 in B.A. Giron Pendleton, B.A. Millsap, K.W. Cline and D.M. Bird [Eds.], Raptor management tech- niques manual. National Wildlife Federation, Wash- ington, DC. Schlatter, R.P., J.L. YA51ez and F.M. Jaksic. 1980 Food-niche relationships between Chilean Eagles and Red-backed Buzzards in central Chile. Auk 97:897- 898. Snyder, N.R. and J.W. Wiley. 1976. Sexual size di- morphism in hawks and owls of North America. Or- nithological Monograph No. 20, American Ornithol- ogists’ Union, Lawrence, KS. Simmons, R.E., D.M. Avery and G. Avery. 1991. Bi- ases in diets determined from pellets and remains: cor- rection factors for a mammal and bird-eating raptor J. Raptor Res. 25:63-67. Simonetti, J.A. 1986. Human-induced dietary shift in Dusicyon culpaeus. Mammalia 50:406-408. Zunino, S. and C. Vivar. 1983-85. Ciclo reproductor de los conejos en Chile central. I. Madurez y relation sexual. Anales del Museo de Historia Natural de Val- paraiso {Chile) 16:101-110. Received 22 August 1991; accepted 5 December 1991 ]. Raptor Res, 26(l):33-35 © 1992 The Raptor Research Foundation, Inc. Short Communications Methods of Locating Great Horned Owl Nests in the Boreal Forest Christoph Rohner Department of Zoology, University of British Columbia, Vancouver, BC, Canada V6T 2 A9 Frank I. Doyle Kluane Lake Research Base, Mile 1054 Alaska Highway, Yukon Territory, Canada Y1A 3V4 The boreal forest is one of the largest and least under- stood ecosystems in North America, and is currently being logged at an alarming rate (Pruitt 1978, McLaren 1990). The owls and diurnal raptors of the boreal forest are no exception to this lack of knowledge (Duncan 1991). Great Horned Owls ( Bubo virginianus) are common throughout North America and are one of the most abundant predators in the boreal forest, but almost all research on them has been carried out in the temperate zone of the continent (Donazar et al. 1989). The northernmost population study was a project in central Alberta in an area of mainly poplar (Populus sp.) forest mixed with agricultural land (Rusch et al. 1972, Mclnvaille and Keith 1974, Adamcik et al. 1978). To our knowledge, the only published research on Great Horned Owls in the coniferous boreal forest is a diet study based on four nests found in the Yukon Territory and Alaska (Weir and Hanson 1989). There are two rea- sons for this lack of information. First, the vast area of the boreal forest biome (about 3.3 million km 2 in Canada) is largely uninhabited, making access and logistics difficult. Second, and probably more important, Great Horned Owl nests are extremely difficult to find in the dense cover of the coniferous boreal forest. As part of a collaborative project on the boreal forest ecosystem, we are studying the Great Horned Owls at Kluane Lake in the southwestern Yukon (Krebs et al. 1986). At first, we had great difficulties finding nests, but we have now developed an efficient method. We describe this method here in the hope of encouraging research on these birds in northern forests. The methods may also apply to other dense coniferous forests, such as temperate rainforests in the Pacific Northwest, where Great Horned Owls have recently gained attention as possible competi- tors and predators of Spotted Owls ( Strix occidentals; Gu- tierrez 1985, Carey et al. 1990). Step 1: Acoustic Triangulation of the Nesting Area. Great Horned Owl nests stand out in deciduous forests and can be detected over large areas with systematic searches from the air or from the ground in late winter (Rusch et al. 1972, Petersen 1979). In coniferous forests, such meth- ods are not feasible because of the dense cover. In this case, the nesting area has to be pinpointed before a visual search can begin. The roosting and calling behavior of the male are the key to the location of the nesting area. We found that four radio-tagged males regularly roosted with- in 100 m of the incubating female. Roosting males leave secretively when an observer approaches, and they are difficult to find. However, of 19 nests checked during in- cubation in March and April 1991, 10 departing males were discovered closer than 100 m from the nest (median distance 35 m). Petersen (1979) also reported that all three of his radio-tagged males used to roost within 75 m of the nests in his study area in Wisconsin. As a result of this roosting behavior, males begin their hooting close to the nest about 1 hr after sunset, and again hoot in the nesting area about 1 hr before sunrise when they settle down for the daytime. The females on the nest usually join the males with one to several hoots at the beginning and the end of each activity period. Triangulations on hooting birds dur- ing these specific times provide a preliminary location of the nest site. This then serves as a starting point for a detailed visual search of the area. Depending on the terrain and on the experience of the observers, one to several hooting sequences may be required to locate a nesting area accurately. Step 2: Visual Search for the Nest. Great Horned Owls in the Yukon breed mainly on “witches’ brooms” (fungus-induced clumps of dense foliage in White Spruce Picea glauca ), or they use old nests built by Red-tailed Hawks ( Buteo jamaicensis) or Common Ravens ( Corvus corax). The incubating female is often not visible from the ground, and the large number of witches’ brooms in our study area required careful visual inspection of almost all individual trees in a nesting area. In all 36 nests that we found by visual searches from 1988-91, we saw at least one down feather at the edge of the nest or in nearby branches. This was also true for 21 nests that were found by locating radio-tagged females and for 4 nest sites re- 33 34 Short Communications Vol. 26, No. 1 used from previous years. We therefore conclude that the presence of fresh down feathers is the best visual cue to determine whether a witches’ broom is used as a Great Horned Owl nest or not. Such feathers are particularly conspicuous when moving in a breeze, or when seen against the sunlight. The maximum size of an area that can be searched reliably depends on the local situation and on the experience of the observer. In our case, this was usually about 200 m in diameter. If a nest is not found, one can return to the first step and locate the nesting area more accurately. Discussion The method described is time-intensive but reliable. It enabled us to find 23 nests in 28 searched owl territories after 1-5 triangulations for step 1, and 0.5-5 hr searching time for step 2. After 5 hr of unsuccessful search we usually stopped the search until the time near fledging. At this stage the adults engage more aggressively in nest defense and are likely to hoot when an observer approaches close to a nest. A possible improvement to this method involves daytime playback of Great Horned Owl hoots. Males might reply from the roost, thereby revealing the nest location. We found one nest within minutes of using this method, but systematic trials using daytime playback during incubation m 1990 and 1991 showed that only 3 of 10 males respond- ed Perhaps playbacks could be applied more efficiently at particular stages of the breeding cycle and times of day (e g., dawn). A concern over any new technique is the disturbance it causes to the animals. For example, it is possible that daytime playbacks of hoots, or visits to nests disturb pairs and cause eggs or young chicks to chill (Fyfe and Olendorff 1976). In our study, we have little evidence of observer disturbance. We approached more than 50 nests of 20 different pairs during incubation or with young nestlings, and only 6 females were flushed from the nest. Two broods failed, but we suspect that the very late breeding date and food shortage explained these failures. We therefore feel that the “triangulation-and-search” technique described here causes little disturbance. Further tests of playback methods may be useful to see if their efficacy can be en- hanced, and if they can be used without disturbing the owls. Resumen. — En la foresta boreal predominan las coniferas y debido a la densa vegetacion es dificil localizar nidos de Tecolote Cornudo ( Bubo virginianus). Recomendamos un metodo que sa basa en dos etapas. Primera: El area de los nidos es determinada por una triangulacion del lugar donde los buhos ululan. Con este metodo se tiene mas exito en la madrugada y al atardecer, cuando los machos emiten su llamado cerca del nido y las hembras que estan incu- bando puede que respondan. Segunda: Una vez que el area de localization del nido es determinada, se efectua una intensa busqueda para ubicarlo. Plumones al borde del nido fueron el mejor indicio visual, y durante nuestro estudio notamos que el observador desde el suelo puede causar solo minima perturbation a las aves en el nido. Este metodo puede tambien aplicarse para otras especies de buhos, en habitats con densa vegetacion. [Traduction de Eudoxio Paredes-Ruiz] Acknowledgments We are thankful to Christoph Schmid for sharing many hours of locating owl nests, and to J.N.M. Smith, C.L. Esser, and C.J. Krebs for their help in improving the manuscript. We were supported by the National Sciences and Engineering Research Council of Canada (Collabo- rative Project Grant to C.J. Krebs), and an R.J. Thomp- son Wildlife Fellowship to Ch. Rohner. Literature Cited Adamcik, R.S., A.W. Todd and L.B. Keith. 1978. De- mographic and dietary responses of the Great Horned Owls during a Snowshoe Hare cycle. Can. Field- Nat. 92:156-166. Carey, A.B., J.R. Reid and S.P. Horton. 1990. Spot- ted Owl home range and habitat use in southern Or- egon coast ranges. /. Wildl. Manage. 54:11-17. DgnAzar, J.A., F. Hiraldo, M. Delibes and R. Rodriguez-Estrella. 1989. Comparative food habits of the Eagle Owl Bubo bubo and the Great Horned Owl Bubo virginianus in six palearctic and nearctic biomes. Ornis Scand. 20:298-306. Duncan, P. A. 1991. Status report on the Northern Hawk owl Surnia ulula. Committee on the Status of Endan- gered Wildlife in Canada, Environment Canada, Ot- tawa, Canada. Fyfe, R.W. and R.R. Olendorff. 1976. Minimizing the dangers of nesting studies to raptors and other sensitive species. Occasional Paper 23, Canadian Wild- life Service, Environment Canada, Ottawa, Canada. Gutierrez, R.J. 1985. An overview of recent research on the Spotted Owl. Pages 39-45 in R.J. Gutierrez and A.B. Carey [Eds.], Ecology and management of the Spotted Owl in the Pacific Northwest. Report PNW- 185, Forest Service, U.S. Department of Agriculture, Washington, DC. Krebs, C.J., B.S. Gilbert, S. Boutin, A.R.E. Sinclair and J.N.M. Smith. 1986. Population biology of Snowshoe Hares. I. Demography of food-supple- mented populations in the southern Yukon, 1976-84. J. Amm. Ecol. 55:963-982. McInvaille, W.B. and L.B. Keith. 1974. Predator- prey relations and breeding biology of the Great Horned Owl and Red-tailed Hawk in central Alberta. Can Field-Nat. 88:1-20. McLaren, C. 1990. Heartwood. Equinox 53:43-55. Petersen, L.R. 1979. Ecology of Great Horned Owls and Red-tailed Hawks in southeastern Wisconsin. Technical Bulletin 111, Wisconsin Department of Natural Resources, Madison, WI. PRUITT, W.O. 1978. Boreal ecology. Studies in Biology 91:1-73. March 1992 Short Communications 35 Rusch, D.H., E.C. Meslow, L.B. Keith and P.D. Doerr. 1972. Response of Great Horned Owl pop- ulations to changing prey densities. /, Wildl. Manage. 36:282-296. Weir, D. and A. Hanson. 1989. Food habits of Great Horned Owls, Bubo virginianus, in the northern taiga of the Yukon Territory and Alaska. Can. Field-Nat. 103:12-17. Received 18 June 1991; accepted 13 September 1991 J Raptor Res. 26(l):35-36 © 1992 The Raptor Research Foundation, Inc. Food Habits of the Short-eared Owl (Asw flammeus ) in Southern South America Jaime R. Rau, Marcelo C. Villagra, Marta L. Mora, David R. MartInez and Maria S. Tilleria Laboratorio de Ecologia, Instituto Profesional de Osorno, Casilla 933, Osorno, Chile Although the Short-eared Owl ( Asia flammeus ) is dis- tributed throughout the Americas, its food habits have received considerable study only in North America (Clark 1975 and references therein). Currently, there is no pub- lished quantitative information on its food habits in all of South America, except for a report made by Fulk (1976) from central Chile, who unfortunately pooled the pellets of both Short-eared and Common Barn Owls (Tyto alba). We report the prey identified in 53 pellets of the Short- eared Owl, collected May-June (Winter) 1986 and Sep- tember (early Spring) 1987-89 at Fundo Quirislahuen, Alberto Hott Siebert Airport and Isla Teja island. The first two places are pasture grasslands located at the out- skirts of the city of Osorno, southern Chile (40°34'S 73°08'W). The third site consists of marshland in the city of Valdivia (39°48'S 73°14'W). Seven of the 53 pellets contained no identifiable prey remains, 17 were weathered and 29 were fresh. Prey remains (mostly native cricetid rodents) were identified using keys for tooth rows (Reise 1973). The 29 fresh pellets measured 33.5 (SD = 1.9) by 18.1 (SD = 0.9 mm) (x ± 2 SE length times maximum width) and the mean dry weight was 3.0 ± 0.4 g. All measure- ments were lower than those reported by Holt et al. (1987), possibly because of our small sample (29 compared to 1 80 pellets). To increase the sample size, we pooled pellets from different areas. Among 46 pellets, we found 69 vertebrate prey items (96%) and only three invertebrates (4%). The diet of the Short-eared Owl by number of occurrences was as follows: 34 (47%) Olivaceus Field-mice ( Akodon oliva- ceus ), 8 (11%) Austral Greater Mice ( Auliscomys micro- pus ), 8 (11%) Long-haired Field-mice ( Akodon longipilus). 2 (3%) Long-tailed Rice Rats ( Oryzomys longicaudatus) , 2 (3%) unidentified members of Akodon, 3 (4%) Black Rats Rattus rattus, 1 (1%) Darwin’s Leaf-eared Mouse ( Phyl - lotis darwini) , 11 (16%) unidentified cricetids, 1 (1%) Gryl- lid and 2 (3%) unidentified insects. The most frequent mammalian prey of Short-eared Owls, Olivaceus Field- mice, Long-haired Field-mice and Austral Greater Mice weighed 23.9 ± 0.5, 34.7 ± 0.3, and 57.6 ± 1.9 g, re- spectively (Pearson 1983). Overall, the mean number of prey items/pellet was 1.7 ± 0.2 ( N = 29, range 1-4). The high consumption of the Olivaceus Field-mouse by Short-eared Owls was in close agreement with the fall peak of this vole-like mouse at the prairie-scrublands of San Martin, Valdivia (Murua and Gonzalez 1986), a site located approximately 192 km north by road from Osorno. Judging from previous records in North America (Clark 1975 and references therein), the food habits of the Short- eared Owl in southern South America are very similar to those in the northern hemisphere. Resumen. — Aunque el nuco (Asio flammeus) tiene una amplia distribucion, su dieta ha sido estudiada solo en America del norte. Analizamos 46 egagropilas colectadas en dos areas del sur de Chile durante invierno (1986) y primavera (1987-89). Se identificaron 72 presas, de las cuales un 96% corresponds a vertebrados. El roedor Ako- don olivaceus constituyo casi la mitad de la dieta. En Sud- america meridional, la estenofagia del nuco es similar a la documentada para el hemisferio norte. Acknowledgments We thank E. Ide, owner of the Fundo Quirislahuen and the staff of Alberto Hott Siebert Airport, for their 36 Short Communications Vol. 26, No. 1 cooperation during this study, and also J. Barriga, for collecting some owl pellets at the airport. We thank Fa- bian M. Jaksic for comments on the draft manuscript and Denver W. Holt who kindly sent us one of his articles. This work was partially funded by FONDECYT 89- 0034. Literature Cited Clark, R.J. 1975. A field study of the Short-eared Owl, Asio flammeus, in North America. Wildlife Monogr, 47: 5-67. Fulk, G.W. 1976. Owl predation and rodent mortality: a case study. Mammalia 40:423-427. Holt D.W., L.J. Lyon and R. Hale. 1987. Tech- niques for differentiating pellets of Short-eared Owls and Northern Harriers. Condor 89:929-931. Murua R. and L.A. Gonzalez. 1986. Regulation of numbers in two neotropical rodent species in southern Chile. Rev. Chil. Hist. Nat. 59:193-200. Pearson, O.P. 1983. Characteristics of a mammalian fauna from forests in Patagonia, southern Argentina. /. Mammal. 64:476-492. Reise, D. 1973. Clave para la determinacion de los craneos de marsupiales y roedores chilenos. Gayana (Zoologia) 27:1-20. Received 21 June 1991; accepted 12 October 1991 / Raptor Res. 26(l):36-37 © 1992 The Raptor Research Foundation, Inc. Golden Eagles Feeding on Fish Bryan T. Brown 1 Cooperative Park Studies Unit, P.O. Box 5614, Northern Arizona University, Flagstaff, AZ 860 1 1 Prey of Golden Eagles ( Aquila chrysaetos ) in western North America consists primarily of small mammals such as lagomorphs and rodents (Carnie 1954, Mollhagen et al. 1972, Olendorff 1976, Bloom and Hawks 1982, Eakle and Grubb 1986). Fish comprised only 0.4% of individual food items found in nests across North America (Olendorff 1976). Most authors of studies of Golden Eagle food habits reported no fish (e.g., Collopy 1983, Eakle and Grubb 1986), while others indicated that fish comprised from 0.2% (Kochert 1972) to 3.6% (Carnie 1954) of individual food items in their diet, and only for eagles nesting near lakes or rivers. Migrant Golden Eagles may feed on fish carrion, such as dead salmon ( Oncorhynchus spp.) in the northwestern United States (Palmer 1988), but most food habit studies were conducted during the nesting season when spawning fish were generally unavailable. No ob- servations of capture of live fish have been published. This study reports on the strategy and success of Golden Eagles foraging on spawning Rainbow Trout ( Oncorhynchus my~ kiss ) in Arizona. Study Area and Methods The 1 km 2 study area consisted of the most downstream 600 m of Nankoweap Creek at its confluence with the Colorado River in Grand Canyon National Park, Arizona. The creek was 1-2 m wide and about 30 cm deep, with winter flows typically 0.05-0.20 m 3 / sec (Brown et al. 1989). 1 Present address: P.O. Box 3741, Tucson, AZ 85722. The creek mouth (elevation 880 m) was located in a desert river canyon, with adjacent vertical cliffs rising over 1000 m Trout were introduced into the Colorado River follow- ing completion of Glen Canyon Dam, 112 km upstream of Nankoweap Creek, in 1963 and the subsequent trans- formation of the river to conditions suitable for a cold- water fishery. The most downstream 0.5 km of Nanko- weap Creek has supported spawning rainbow trout since the mid-1970s. Up to 1500 trout may enter the creek at peak times during spawn, which may last from November through April (Brown et al. 1989). Observations of Golden Eagle prey capture attempts and feeding activity were made from 6 February to 20 March 1990, and from 24 January to 13 March 1991. Continuous, daily observations were made from 0.5 hr before sunrise to 0.5 hr after sunset, except at times when recreational activity in the study area precluded eagle feed- ing activity (ca. 10% of daylight hours). Observations were made from an uncamouflaged observation post located 800 m west of and 100 m above the creek mouth, using 10- 45 x spotting scopes. A prey capture attempt was defined as any effort to secure prey, including both single and multiple pounces for an individual trout. Results and Discussion Golden Eagles were present on 32 (76%) of 42 d in 1990 and 39 (80%) of 49 d in 1991, with one to four individuals present on any one day. At least five individuals were recorded on 10-11 March 1990 (three adults on 10 March, two subadults on 11 March). I suspect that more than five Golden Eagles were present each year, but in- dividuals were not marked, March 1992 Short Communications 37 Golden Eagle vernal migration at this latitude began in late February (Palmer 1988), indicating that most eagles observed were wintering. Some may have been breeding residents, because copulation by a pair of adults was ob- served in early March 1991. No aggressive, territorial behavior between Golden Eagles was observed in either year, further suggesting that most were wintering or mi- grant eagles. Eighty-three Golden Eagle prey capture attempts on rainbow trout were observed during the study period: 59 (71%) on live trout, 13 (16%) on carrion trout, and 11 (13%) on trout of unknown status. Seventy-two (87%) prey capture attempts resulted in the acquisition of prey (48 live trout, 13 carrion trout, and 11 of unknown status). Forty-eight (81%) of 59 prey capture attempts involving live trout were successful. Most live fish captured appeared to be healthy, with few exhausted, spawned-out trout taken as prey. Most prey capture attempts involving live trout were in the shallow water of Nankoweap Creek, and only one W'as in the Colorado River in an isolated pool at the river’s edge created by dropping water releases from Glen Canyon Dam. Fifty-one (61%) prey capture attempts were by adults, 17 (21%) by subadults, and 15 (18%) by eagles of unknown age. Several attack methods were observed. Most eagles jumped or walked onto trout from the shore (36, 58%), others made flights of 3-325 m onto trout (25, 40%), or reached out from shore with their beaks to capture trout (1, 2%). Attack perches used by eagles ( N = 55) included streamside boulders (26, 47%), gravel bars along the shore (15, 27%), the stream bed (8, 15%), cliffs (3, 5%), riparian trees (2, 4%) and talus slopes (1, 2%). Mean distance from attack perch to foraging location was 15.9 m (SD — 54.6 m, range 0.1-325 m, N = 49). Fish were eaten at the capture site (14, 26%), dragged to shore before consump- tion (38, 70%), carried to shore before consumption (1, 2%), or carried to a distant feeding perch or out of sight before consumption (1, 2%). Consumption of 35 fish was monitored. Most fish were consumed within the study area (17, 49%), some partially consumed fish were abandoned or lost by being dropped into the river or other inaccessible site (11, 31%), while other fish were carried out of sight (7, 20%). Golden Eagles successively consumed two trout on at least three occasions. For example, on 23 February 1990, an adult captured and consumed two trout from the creek within 28 min between 0717-0745 H. Each trout appeared to be ap- proximately 30-35 cm long and probably weighed about 700 g (Carothers and Minckley 1981). The opportunistic nature of Golden Eagle foraging (Palmer 1988) suggests that potential prey items will be taken when they are available. These findings indicate that fish occasionally constitute a substantial proportion of the diet of some wintering and migrant Golden Eagles. Resumen. — Aguilas Reales ( Aquila chrysaetos) se alimen- taron de truchas de la especie Oncorhynchus my kiss que desovaban en Nankoweap Creek a lo largo del Rio Col- orado en el Gran Canon, Arizona, durante los inviernos 1989-90 y 1990-91. La mayoria de los intentos de captura (59 de 83, 71%) fueron con truchas vivas; y de estos in- tentos, la mayoria (48 de 59, 81%) tuvo exito. La mayor parte de las aguilas saltaron o caminaron hacia la trucha desde la orilla (36, 58%), otras hicieron vuelos de 3-325 m hacia estos peces (25, 40%), o se estiraron desde la orilla para capturar la presa con el pico (1, 2%). El pescado puede, ocasionalmente, constituir una porcion substancial de la dieta de algunas Aguilas Reales en el proceso de migration, o en invierno. [Traduction de Eudoxio Paredes-Ruiz] Acknowledgments I thank E. Baldwin, P. Becker, B. Casler, B. Dye, C. Hallett, N. Nahstoll, T. Yates and other volunteers for the prolonged periods they spent in the field assisting this study. C.R. Blem, P.H. Bloom, P.T. Redig, C. van Riper, D.L. Wegner and two anonymous reviewers made useful comments on an early draft of this manuscript. Funding for this study was provided by the Glen Canyon Envi- ronmental Studies, U.S. Bureau of Reclamation, as part of a project to determine the influence of Glen Canyon Dam on the foraging ecology of wintering Bald Eagles along the Colorado River. The Bureau of Reclamation and National Park Service provided administrative sup- port and research permits. Literature Cited Bloom, P.H. and S.J. Hawks. 1982. Food habits of nesting Golden Eagles in northeast California and northwest Nevada. Raptor Res. 16:110-115. Brown, B.T., R. Mesta, L.E. Stevens and J. Weisheit 1989. Changes in winter distribution of Bald Eagles along the Colorado River in Grand Canyon, Arizona. J. Raptor Res. 23:110-113. Garnie, S.K. 1954. Food habits of nesting Golden Ea- gles in the coast ranges of California. Condor 56:3-12 Carothers, S.W. and C.O. Minckley. 1981. A survey of the aquatic flora and fauna of the Grand Canyon Final report, U.S. Bureau of Reclamation, Boulder City, NV. COLLOPY, M.W. 1983. A comparison of direct obser- vations and collections of prey remains in determining the diet of Golden Eagles. J. Wildl. Manage. 47:360- 368. Eakle, W.L. and T.G. Grubb. 1986. Prey remains from Golden Eagle nests in central Arizona. Western Birds 17:87-89. Kochert, M.N. 1972. Population status and chemical contamination in Golden Eagles in southwestern Ida- ho. M.S. thesis, University of Idaho, Moscow, ID. Mollhagen, T.R., R.W. Wiley and R.L. Packard. 1972. Prey remains in Golden Eagle nests: Texas and New Mexico. J. Wildl. Manage. 36:784-792. Olendorff, R.R. 1976. The food habits of North Amer- ican Golden Eagles. Amer. Midland Nat. 95:231-236 Palmer, R.S. [Ed.]. 1988. Handbook of North Amer- ican birds. Vol. 5. Yale University Press, New Haven, CT. Received 20 September 1991; accepted 3 December 1991 38 Short Communications Vol. 26, No. 1 J Raptor Res. 26(l):38-39 © 1992 The Raptor Research Foundation, Inc. Greater Yellow-headed Vulture {Cathartes melambrotus ) Locates Food by Olfaction Gary R. Graves Department of Vertebrate Zoology , National Museum of Natural History, Smithsonian Institution, Washington, DC 20560 Turkey Vultures ( Cathartes aura) and Lesser Yellow- headed Vultures ( Cathartes burrovianus ) have acute senses of smell (Stager 1964, Houston 1986, 1988). Although the olfactory capacities of the Greater Yellow-headed Vulture (C. melambrotus ) are unknown, they are thought to be similar to those of its congeners (Houston 1988). Here I report observations that strongly suggest the use of smell by this species to locate carrion. Collectively, these data indicate that acute olfaction is a shared derived character of Cathartes within the Cathartidae, as reflected by the large olfactory lobe (Bangs 1964, pers. observation). I made daily observations of vulture abundance and behavior on the east bank of the Rio Xingu (3°39'S 52°22'W), 52 km SSW of Altamira, Para, Brazil (Graves and Zusi 1990), from 14 August to 29 September 1986. Evidence of olfaction in the Greater Yellow-headed Vul- ture was observed in two instances. In the first instance, I flushed an adult vulture from the carcass of a White-faced Tree Rat ( Echimys chrysurus) on the ground in a dense stand of bamboo in terra firme forest several km from the river. The startled vulture clumsily zigzagged on foot through the bamboo culms, took flight from the trail cut through the bamboo, and flew to an emergent tree. The carcass, which appeared to be less than a day old, was almost completely consumed; only the skin from the rat’s head and back and the attached tail re- mained. Adult E. chrysurus weigh from 415-890 g (Em- mons and Feer 1990). Because the canopy of the bamboo Table 1. Relative abundance of vultures on the lower Rio Xingu, Para, Brazil, from 14 August to 29 September 1986. Species Number of Days Observed Largest Daily Total Total Greater Yellow-headed Vulture 31 6 101 Turkey Vulture 7 1 7 Black Vulture 6 5 18 King Vulture 5 4 8 (6-8 m) was exceedingly dense, the carcass could not be observed from the air. Presumably, the vulture landed on the trail and approached the carcass, some 10 m from the trail edge, on foot. On another occasion, a pair of Greater Yellow-headed Vultures discovered a cache of day-old flesh from a large museum specimen of the Brazilian Porcupine ( Coendou prehensilis) that had been dumped 200 m from camp at the end of a poorly marked trail. The flesh had been partly covered with leaf litter and was further obscured from view by the canopy of tall (30 m) seasonally flooded forest. The cache was difficult for me to visually locate when I returned to the area, and in all likelihood, could not have been observed by vultures flying above the canopy. In both cases, Greater Yellow-headed Vultures presumably de- tected the carrion by scent alone. Turkey Vultures were relatively rare near camp and were never observed at car- rion in the forest. Cathartes spp. and Black Vultures ( Coragyps atratus) appear to partially segregate by habitat along the lower Rio Xingu. Away from towns, agricultural clearings, and rivers, the Greater Yellow-headed Vulture was the prev- alent vulture of pristine forest, outnumbering other species by an order of magnitude (Table 1). Black Vultures, which have poorly developed olfactory senses and rely on vision to locate food, scavenged beached fish on sandbars near camp and were not observed away from the river in un- broken forest. The few King Vultures (Sarcoramphus papa) recorded during the study were observed soaring at great heights near the braided channel of the river, Resumen. — Presentamos evidencias que sugieren que los buitres de la especie Cathartes melambrotus usan el olfato para localizar sus alimentos. En la floresta de espesura continua, a lo largo del bajo Rio Xingu, Para, Brasil, esta especie excede en numero a otras especies de buitres en una magnitud significativa. [Traduccion de Eudoxio Paredes-Ruiz] Acknowledgments Field work was facilitated and supported by the Aca- demia Brasileira de Ciencias, through a grant from Elec- tronorte administered by Paulo Vanzolini. Transportation March 1992 Short Communications 39 to and from Brazil was provided by the Smithsonian’s Neotropical Lowland Research Program. I thank Mike Carleton, Louise Emmons, Linda Gor- don, and Richard Zusi for assistance on the Xingu, Van- zolini and Bea Ribeiro for providing logistical support on the Xingu and in Sao Paulo, and Floyd Hayes, David Houston and Patricia Rabenold for comments on the manuscript. Literature Cited Bangs, B.G. 1964. The nasal organs of the Black and Turkey vultures: a comparative study of the cathartid species Coragyps atratus atratus and Cathartes aura sep- tentrionalis (with notes on Cathartes aura jalklandica, Pseudogyps bengalensis and Neophron percnopterus). J. Morph. 115:153-184. Emmons, L.H. and F. Feer. 1990. Neotropical rain- forest mammals: a field guide. University of Chicago Press, Chicago, IL. Graves, G.R. and R.L. Zusi. 1990. Avian body weights from the lower Rio Xingu, Brazil. Bull. Br. Ornithol. Club 110:20-25. Houston, D.C. 1986. Scavenging efficiency of Turkey Vultures in tropical forest. Condor 88:318-323. . 1988. Competition for food between Neotrop- ical vultures in forest. Ibis 130:402-417. Stager, K.E. 1964. The role of olfaction in food location by the Turkey Vulture. Los Angeles County Mus. Contr. Set. 81:1-63. Received 11 October 1991; accepted 3 December 1991 / Raptor Res. 26(1):40 © 1992 The Raptor Research Foundation, Inc. Letters Intraspecific Nest Defense by Prairie Falcons A west facing cliff in Sunol Regional Park, Sunol, California, supported an active Prairie Falcon ( Falco mexicanus ) eyrie in 1991. Nesting at this eyrie has been recorded for the preceding six years by the author and for many years prior by local falconers and birdwatchers. Historically, this site was occupied by Peregrine Falcons ( Falco peregrinus). I recorded observations of a resident pair of adult Prairie Falcons nesting at this site on 7 March 1991, as they defended their eyrie from an intruding female Prairie Falcon, not identified as to age. The intruding bird approached the cliff from the southwest. The resident male and female stooped at the intruding falcon repeatedly. She responded by rolling upside-down, presenting talons in defense, and by stooping at the resident female. The intruder was pursued and harassed by the resident male as she continued to make aerial advances toward the cliff face. Vocalizations were made by more than one bird, but particularly by the resident female. The resident female appeared heavy and sluggish. Her undertail coverts were drooping and she may have been carrying an egg. This presumption matched with egg laying dates later deduced from hatching dates. She flew to a ledge near the nest cavity. Even though aggressively pursued by the resident male, the intruding female flew to the cliff and landed near the resident female. Both birds were facing each other and were screaming. Their bodies were crouched down in a posture similar to that described as the “horizontal threat display” by D.A. Ratcliffe (1980, The Peregrine Falcon, Buteo Books, Vermillion, SD). The intruder took a few steps toward the resident female when the latter advanced and while still vocalizing, placed her head under the chest of the intruder and, lifting upwards, forced her to take flight. During these events, the male was still flying near the cliff. When the intruding female was again airborne, the male pursued her vigorously and within 5 sec, had turned her flight away from the cliff and to the southwest. The male pursued her in a tail chase for 5-6 sec, covering a distance of about 200 m. The intruding female flew from the area, never changing her direction from southwest even after the male had ended his pursuit. Having returned to the cliff, the male made several flights past the ledge where the resident female had been sitting He then perched above the nest cavity. Shortly thereafter, the female left her perch for another where her solicitation to the male quickly led to a brief copulation. Immediately following, the female flew to the nest cavity and, after scraping the ground briefly and peering over her shoulder, lay down with her head facing west. During the subsequent observation period, from 1115 to 1310 H, there were no further interactions involving this pair. I wish to thank Hans Peeters and Kevin Shea for their review and editing of this manuscript. — Joseph E. DiDonato, East Bay Regional Park District, 2950 Peralta Oaks Court, Oakland, CA 94605. J. Raptor Res. 26(1):40-41 © 1992 The Raptor Research Foundation, Inc. Nestling Red-tailed Hawk in Occupied Bald Eagle Nest On 1 June 1990 we approached a Bald Eagle ( Haliaeetus leucocephalus ) nest, located north of Merrit, Michigan, to band and draw blood from a nestling eagle for contaminant analysis. Two nestling eagles were visible from the ground. Nest height was 13.7 m in a 21.0 m Quaking Aspen ( Populus tremuloid.es') with a DBH of 57.1 cm. J.B. Holt climbed to the nest and discovered a nestling Red-tailed Hawk {Buteo jamaicensis) in addition to the two nestling eagles. We aged one of the nestling eagles as approximately 7.5 wk old, using an eighth primary length of 175 mm and equations previously derived for nestling growth rates (G.R. Bortolotti, 1984, /. Field Ornithol. 55:467-481). The nestling hawk was downy and approximately 3.5 wk old. It was in excellent condition, with no apparent signs of abuse by the adult or nestling eagles. 40 March 1992 Letters 41 The nestling hawk was removed from the nest and taken to the Small Animal Veterinary Clinic at Michigan State University. There it was temporarily cared for and then placed into a local Red-tailed Hawk nest. The presence of a nestling hawk in a Bald Eagle nest is unexplained, but at least two possible scenarios may apply. One possibility is brood parasitism by the parent hawks. Although brood parasitism has been documented in five avian families: Anatidae, Cuculidae, Indicatoridae, Icteridae, and Ploceidae, (J.C. Welty, 1975, The Life of Birds, W.B. Saunders Co., New York) it has never been recorded in the family Accipitridae. For brood parasitism to have occurred, a female Red-tailed Hawk had to enter the eagle nest, contend with an adult Bald Eagle with newly hatched nestlings, and lay its egg. The egg then had to be incubated for approximately 30 days (F.L. Beeby, 1976, Hawks, Falcons and Falconry, Hancock Home Publisher, Ltd., New York) by either the adult eagles, the nestlings or both. This scenario is unlikely. A second explanation is non-lethal predation followed by parental care. One of the adult eagles may have captured the nestling hawk and failed to kill it during capture and transport. The nestling hawk had a full crop when we discovered it, suggesting that either the adult eagles had fed it, or that it had recently been transported to the nest and still had a full crop from its natural parents. Corroborating this is the presence of other raptor species in an analysis of Bald Eagle prey items collected at nests in the midwest (K.D. Kozie and R.K. Anderson, 1991, Archives of Environmental Contaminants and Toxicology 20:41-48; W.W. Bowerman IV and J.P. Geisy, Jr., 1991, Ecology of Bald Eagles on the Au Sable, Manistee, and Muskegon Rivers, Michigan State University, East Lansing, MI). We believe that this scenario is a more likely explanation than brood parasitism. Funding for this project was received from Consumers Power Company, Michigan Department of Natural Resources Natural Heritage Program-Nongame Wildlife Fund, U.S.D.A.-Forest Service and EARTHWATCH. Thanks goes to B.R. Richardson for assisting in the field. P.T. Stapp and L.L. Williams provided helpful comments on an earlier draft. Assistance was provided by J.A. Sikarskie at Michigan State University-Small Animal Veterinary Medicine. — Patrick R. Stefanek and William W. Bowerman IV, Department of Fisheries and Wildlife, Pesticide Research Center, Institute for Environmental Toxicology, Michigan State University, East Lansing, MI 48824, U.S.A.; Teryl G. Grubb, U.S.D.A.-Forest Service, Rocky Mountain Forest and Range Experiment Station, Forestry Sciences Lab, ASU Campus, Tempe, AZ 85287-1304, U.S.A.; and John B. Holt, Jr., 838 Johnson Street, North Andover, MA 45858, U.S.A. J Raptor Res. 26(1):41 -42 © 1992 The Raptor Research Foundation, Inc. Talon Grappling by Aplomado Falcons and by Golden Eagles On 31 January 1979, while observing raptors of seven species along a fire line in central Venezuela, I saw various chasing and stooping interactions between some of the five Aplomado Falcons ( Falco femoralis) present. One of these stoops I fortuitously photographed. A juvenile (by plumage) female (by size) made a long, shallow stoop on an adult (by plumage) male (by size) perched about 20 m above the ground in the upper canopy of a tree. The defensive response of the male was so quick that the episode was little more than a blur to the unaided eye. Fortunately, the camera (Fig. 1) captured the critical detail. From the photograph and knowledge of the birds’ positions just before and after the incident, it was clear that the male sprang into the air, thrust his feet upward, grasped the female’s foot, and, impelled by the female’s momentum, whirled at least half a revolution before the birds separated. This may be the first published account of grappling and whirling for the Aplomado Falcon. It provides detail on the mode of contact of the birds during the bout, something that is seldom determinable for any species. The second observation is an extreme example of grappling and whirling for the Golden Eagle ( Aquila chrysaetos). Grappling is common for some eagles (L. Brown and D. Amadon, 1968, Eagles, Hawks and Falcons of the World, McGraw-Hill Book Go., New York), but rare for the golden (D.H. Ellis, 1979, Wildlife Monographs No. 70; A.R Harmata, 1982, Raptor Research 16:103-109). On 18 March 1985, I observed three adult Golden Eagles flying along an elevated cliff rim in southeastern Arizona. The largest bird (presumably a female) was attended by two smaller birds (presumably males) that emphatically stooped at one another. I observed this combat from 1317-1323 H when the two males bound together and whirled at least 19 revolutions before disappearing from view beneath the forest canopy. I estimated that the two fell at least 100 m during the bout, but could not determine whether the eagles actually collided with the ground. The female swooped twice above the combatants on the ground, then drifted away from the 42 Letters Vol. 26, No. 1 Figure 1. Aplomado Falcons grappling and whirling. Adult male (above) was perched in treetop when juvenile female stooped from left. area. After 1 1 min, first one male, and then the second, flew from the touch down zone and rejoined the female. The trio then drifted north out of sight with the males again stooping at each other. A Red-tailed Hawk ( Buteo jamaicensis ) was present, soaring above and occasionally stooping at the trio, during almost the entire episode. This may be the most extended grappling and whirling bout ever reported, at least for the Golden Eagle. — David H. Ellis, U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center, Laurel, MD 20708. J. Raptor Res. 26(l):42-43 © 1992 The Raptor Research Foundation, Inc. A Previously Undescribed Falcon Flight Display It is something of a surprise for a social display to suddenly be discovered in a supposedly well-studied species. An example is the recent description of the Bowing display for the Cooper’s Hawk ( Accipiter cooperii ) (R.N. Rosenfield and J. Bielefeldt, 1991, Condor 93:191-193). The Prairie Falcon ( Falco mexicanus ) and Peregrine Falcon ( Falco peregrinus) have also received great attention during the last three decades, yet I am unaware of any published account of a “deep-flap” display I will next describe. In 1981, at three different cliffs, all in Arizona, I observed a flight display that was distinctly different from normal flapping flight. On 11 February at 1607 H, at a site on the Tohono O’Odham Indian Reservation, an adult Prairie Falcon (probably a male by size comparison with the other falcon attending the cliff) stooped upon and flushed a Red- March 1992 Letters 43 tailed Hawk (Buteo jamaicensis) from the vicinity of the 1980 Prairie Falcon eyrie. After pursuing the hawk nearly 1 km west, the falcon returned to the eyrie cliff by flying along the cliff rim while performing a flapping flight with slow, deep, exaggerated wingbeats reminiscent of the flight of a nighthawk ( Cordeiles sp.). The wings of the falcon nearly touched at the top and at the bottom of each stroke. On 24 February at 1010 H, on the East Tactical Military Range south of Gila Bend, I saw an even more elaborate performance of this same display. An adult Prairie Falcon (probably female by size) flushed from her eyrie cliff after the third pass by a pair of A- 10 military jets about 100 m from the cliff. She soared in front of the cliff for 6 min, then performed five series of the exaggerated deep-flap display. Each series consisted of one to four flaps. After the display, the falcon soared briefly, then lit on the cliff rim near the 1980 eyrie. The third observation of the deep-flap display occurred on 17 March near a traditional Peregrine Falcon eyrie in Pinal County. During the day, four falcons (two peregrines and two Prairie Falcons), two Red-tailed Hawks, and two Golden Eagles ( Aquila chrysaetos ) were observed attending the same cliff, a massive wall about 1 km long and over 150 m high. During the 7 hr 46 min observation period, many territorial interactions were observed. In one of these encounters (1157 H), one of the falcons left the cliff and stooped on a distant hawk or eagle. Then, while circling back toward the cliff, it performed one bout of the deep-flap display. The bird performing this bout was at such a great distance that, although I thought it to be one of the Prairie Falcons, it may have been one of the peregrines. In fact, later in the day (1441 H), the adult male Peregrine Falcon led its mate toward the eyrie cliff as it performed a brief bout of exaggerated flaps that was similar to the Prairie Falcon’s deep-flap display. It may be that this, and perhaps other yet undescribed social displays, have gone unobserved, or at least unreported, for so long because the courtship phase of the breeding cycle of even well-studied raptors has been under-emphasized. From my own work with large falcons over two decades, all observations of deep-flap were in a two-month period when the peculiar demands of my study required that I concentrate on the pre-laying stage. — David H. Ellis, U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center, Laurel, MD 20708. J Raptor Res. 26(1 ):43 © 1992 The Raptor Research Foundation, Inc. An Unusual Death of a Nesting Golden Eagle On 16 May 1991, while flying nesting surveys for Golden Eagles {Aquila chrysaetos) in Sheridan County, Wyoming, we observed a dead adult eagle in a nest. The nest site was located in a shallow cave of a clay bank approximately 200 m above Big Goose Creek, west of Sheridan, Wyoming, and had been active for at least 10 years. Closer inspection of the nest from the ground showed that a large rock measuring 71 cm x 43 cm x 28 cm and weighing approximately 45 kg had broken loose from the clay bank directly above the nest and fallen over the torso of the incubating bird. The dead eagle was an adult female that had been banded near the nest in February 1988. Remnants of one egg were found under her body. We estimated that the bird died in late April following a series of wet snowstorms, which contributed to loosening the rock from the bank. The rock was removed from the nest so that the site could be used again in future years. Incidents of this type are undoubtedly rare throughout the entire range of the Golden Eagle. However, mortality associated with the erosion of banks supporting nests may be common in parts of Wyoming, Colorado and Nebraska where Golden Eagles use creek banks as nest sites. — Robert L. Phillips and John L. Cummings, U.S. Department of Agriculture, P.O. Box 25266, Denver, CO 80225-0266. John D. Berry, Kiewit Mining Group, Inc., P.O. Box 3049, Sheridan, WY 82801. J Raptor Res. 26(l):43-44 © 1992 The Raptor Research Foundation, Inc. Chicks of Black Kites Attacked by Ants while Hatching In 1988 while routinely visiting nests of Black Kites ( Milvus migrans ) the following observations were made at two nests in the north of Donana National Park, South West Spain (37°N 6°W) which show that ants have some influence on the survival of chicks at hatching. In one nest, on 15 June, one egg had already hatched, and the second egg of a clutch of two was in an advanced stage of hatching with half the shell already open. The adult was not incubating 44 Letters Vol. 26, No. 1 when I arrived at the nest and the embryonic membranes were dry. The hole in the egg was pointing downward and the shell around it had collapsed inward suggesting that the egg had been slightly crushed. The whole nest was covered with ants ( Tapinoma nicerrimum) many of them going in and out of the egg. On 16 June the chick had completely broken the shell but still had not emerged, probably due to the very dry membranes strongly adhering to its body. The exposed part of the body was completely covered with ants showing clear signs of having been bitten. I freed the chick from the eggshell. On 17 June the chick was observed ridding itself of ants with its beak or by rubbing itself with nest material. Although the nest continued to be full of ants, there seemed to be fewer on the chick. This chick survived subsequently until fledging. In a second nest, on 22 June, the first egg of a clutch of two had been hatching for three days, after which the shell was only I 2 A open. It was not in the center of the nest but placed between sticks at the edge of the nest and completely infested with ants ( Crematogaster scutellaris) . A large part of the exposed area (wing, neck, breast, leg and especially the face) was nibbled to expose raw flesh. Twice I saw ants go in and out of the mouth. The egg also appeared to have been crushed and the head of the chick was slightly misshapen. I removed the chick from the egg. It was cold with a moribund appearance. The second egg was also at the edge of the nest, with the shell very scratched, possibly from rubbing against the nest sticks. When I arrived at the nest an adult flew from it, and during the nest check, I saw it circling the nest. On day two, the chick was dead and half eaten by ants. The remaining egg had fallen below the nest, and I did not see any adult near the nest. As far as I know, this is the first recorded observation of an attack by ants on chicks of raptors, at least in a temperate area where these insects are more gatherers and carrion eaters than hunters (W.M. Wheeler 1960, “Ants, their structure, development and behavior,” Columbia Univ. Press, New York). It must be of infrequent occurrence as I have only been able to observe these two instances despite following hatchings in 57 Black Kite nests in the same area over two years. Nonetheless, a high proportion of trees with active nests were infested with these ants, especially Crematogaster. The unusual weather in Donana in 1988 could have led to these ocurrences. The spring was exceptionally wet, with rainy and unsettled conditions until June. When these observations were made the days were very hot following a stormy period, and at this time much ant activity could be noted. This suggests the ants might have been more inclined to use food sources not regularly exploited. Also, an abnormal behavior of the parents or chicks on hatching could have had an influence. In the first nest the incubation may have been somewhat irregular, as indicated by very dry membranes. In the second nest the chick took more than two days to hatch, a longer period than the normal 24-48 hours (unpubl. data). Both crushing of the eggshell and the positioning of a hatching egg at the edge of the nest were only observed in these two nests. It is possible that the parents, in order to avoid the infestation, either crouched more against the nest to impede the ants’ entry or pushed the egg away from the ants. If so, these actions appeared to be of little use, or, in the instance of the crushing of the shell, were counterproductive and probably made hatching more difficult. Possibly this type of attack can only be successful at the critical time of hatching since soon after hatching the chick appeared able to defend themselves against insect attacks. I thank A.M. Jones for a translation of the first draft and F. Hiraldo and A. Coole for their helpful comments. — Javier Vinuela, Museo Nacional de Ciencias Naturales, %/Jose Gutierrez Abascal 2, 28006 Madrid, Spain. /. Raptor Res. 26(l):44-45 © 1992 The Raptor Research Foundation, Inc. Thesis Abstracts Changes in the Body Mass of American Kestrels {Falco sparverius) During the Breeding Season I monitored the changes in the body mass of adult American Kestrels ( Falco sparverius ) during the breeding season to determine: 1 ) whether changes in the mass of breeding kestrels occurred, and 2) what factors influenced these changes in mass. I first examined the relative influences of environmental and reproductive factors on the changes in body mass of breeding kestrels. I then examined whether food provisioning by adults was a physiological stress on the physical condition of adults or whether changes in mass could be related to an adaptive strategy of mass loss. I conducted my research on a breeding population of American Kestrels at Besnard Lake, Saskatchewan (55°20'N 106°00'W). In the two years of my study (1988 and 1989), 643 measures of mass were collected by live-trapping techniques from April March 1992 Thesis Abstracts 45 through July of each breeding season. I also monitored 35 nests for 877 hr to obtain measures of mass using electronic balances and data on provisioning rates of parents during the brood-rearing period. The results of my study suggest that environmental factors, such as climate and food availability, may have had a strong influence on the physical condition of adults early in the breeding season. During the brood-rearing period, females with 5-chick broods displayed trends in mass loss that may have been related to increased provisioning rates, while males did not. Parents with smaller broods did not display significant trends in mass related to their provisioning rates, possibly because of the lower stress in feeding fewer young. The stress of foraging on parents may have been further mitigated by the use of nutrient reserves and the combined effort by breeding pairs to feed young. Although foraging may have caused stress-related mass loss, I also found that the physical condition of parents may influence the subsequent change in mass and the amount of food delivered to offspring. Parents in good condition were more likely to lose mass and deliver more food to their offspring than parents in poor condition. Parents may therefore be balancing their physical condition (good or poor) within the limits of self-maintenance with their level of food provisioning throughout the breeding season. — William M. Iko. 1991. M.Sc. thesis, Department of Biology, Uni- versity of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 0W0. J Raptor Res. 26(1 ):45 © 1992 The Raptor Research Foundation, Inc. Foraging Ecology of Urban-Breeding Merlins ( Falco columbarius ) I studied the foraging ecology of breeding Merlins {Falco columbarius ) in Saskatoon, Canada, from May to July, 1987-90. I identified 1332 prey items belonging to 36 vertebrate species from prey remains collected near 65 Merlin nests. The principal prey of breeding Merlins was the House Sparrow {Passer domesticus ), which along with the Horned Lark {Eremophila alpestris ) were usually taken more frequently than expected from their relative abundance in the environment. Other potential prey species were usually taken in proportion to, or less than, expected. The proportion of adult House Sparrows in the diet decreased while that of juveniles increased significantly as the Merlin breeding season progressed. During the incubation and nestling periods, the relative abundance of male and female House Sparrows did not differ significantly from rates at which they were consumed. In the fledging period, based on relative abundances, adult House Sparrows were taken less often than expected while juveniles were eaten more often than expected by Merlins. Twenty-seven Merlins were radiotracked. Mean hunting range of resident (hatched in the city) and immigrant (presumed to have hatched outside the city) males were 6.3 ±1.3 (SE) km 2 and 33.7 ± 12.1 km 2 , respectively. Mean hunting ranges of resident and immigrant females were 6.6 ± 3.4 km 2 and 8.6 ±1.6 km 2 , respectively. Merlins that hunted exclusively in the city used all habitats in proportion to availability. Merlins that hunted both in and outside the city usually avoided hunting in agricultural areas, which were relatively low in prey abundance. During the incubation and nestling periods, males with more prey birds in their ranges had significantly smaller hunting ranges. Male Merlins with more prey birds in their ranges spent relatively less time hunting and males with larger broods spent more time hunting compared to those with smaller broods. For female Merlins, hunting range was inversely correlated with both body mass and House Sparrow abundance. — Navjot S. Sodhi. 1991. Ph.D. thesis, Department of Biology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 0W0. Present address: Department of Veterinary Anatomy, Western College of Veterinary Medicine, University of Saskatch- ewan, Saskatoon, Saskatchewan, Canada S7N 0W0. 46 Thesis Abstracts Vol. 26, No. 1 Spotted Owl Nest Site Characteristics in Mixed Conifer Forests of the Eastern Cascade Mountains, Washington This study describes habitat characteristics associated with Spotted Owl ( Strix occidentalis) nest sites in mixed-conifer forests of the east slope of the Cascade Mountains, Washington. I gathered habitat data at 85 historic nest sites; at 62 of these sites I also collected data at random sites within the same stand to be used in a comparative analysis. Nest elevations ranged from 1250-4800 feet. Most nests were on the lower third of slopes, but nests were also found on other slope positions and on bottom land. The distribution of nest site aspects was random but differed significantly from the aspect of paired comparison sites. Various types of natural disturbances, such as fire, root rot, and mistletoe infection, influenced many nest sites. In addition, 46% of the stands had been partially harvested at least 40 years prior to this study. Analyses of data from paired sites revealed a number of significant differences between nest trees and randomly selected comparison trees. Douglas-fir (. Pseudotsuga menziesii ) was used at a greater frequency (92%) than expected in comparison to other tree species in the vegetation plots. Most nest trees occupied either dominant or condominant crown positions in the canopy, and were typically alive and fully intact. However, more nest trees had broken tops than random trees. Nest trees ranged from 66 to 700 years of age, the median age was 137 years. Nest trees on south- facing slopes were significantly larger (dbh) and older than those on north-facing slopes. Cavity and broken-top nest trees were significantly larger (dbh) and older than trees that supported other nest types. The most common nest type (55.3%) used by Spotted Owls included those originally made by Northern Goshawks (Accipiter gentilis); other nests were located in mistletoe growth (24.7%), in cavities (10.6%), on broken-tops (5.9%), and on large horizontal branches (3.5%). Certain features of vegetation structure varied significantly between nest and random sites. Habitat characteristics with greater values at nest sites included height of intermediate trees, canopy height of codominant and intermediate trees, basal area of class IV snags, abundance of small (4-13 inches dbh) snags, abundance of 14-23 inch (dbh) Douglas-firs, and basal area of all Douglas-firs and of all live trees combined. Habitat characteristics with greater values at random sites included basal area of class I and II snags and basal area of a group of several less common conifer species. Differences in habitat features between nest and random sites were noted in comparisons of sites with and without evidence of timber harvest, and between sites of the same or different aspect. Two features often associated with Spotted Owl habitat, volume of downed wood (decay classes I-IV) and percent canopy cover, generally did not differ significantly between nest and random sites. A basic logistic regression model correctly classified 71 percent of nest and random sites. Four supplemental submodels developed to account for differences in aspect and logging history correctly classified between 65 and 93 percent of nest and random sites combined. The best predictor variables were canopy height of dominant trees, basal area of decay class I snags, basal area of decay class II snags, abundance of 4-13 inch (dbh) snags, and abundance of 24-33 inch (dbh) Ponderosa Pines. The five models classified sites significantly better than chance and goodness of fit analyses indicated that the models provided appropriate fit to the data. However, classification rates obtained in cross-validation indicate the models were not stable because of the high degree of variance associated with habitat features at nest sites. Inclusion of stand and/or landscape habitat characteristics will be required to improve all modelling efforts. Spotted Owls nest in a wide variety of forest stand conditions on the east slope of the Cascade Mountains. Over 50% of the nest sites are in relatively young (median age =130 years) even-aged patches or stands of trees. These results are in contrast to Spotted Owl habitat research in other regions of the Pacific Northwest. The variability of habitat features within and between regions indicates a substantial degree of flexibility in the habitat selection process The suppression of forest fires on the east slope of the Cascade Mountains since about the turn of the century has had a profound effect on forest stand dynamics. Recent consequences of fire suppression include increased fire intervals and changes in tree species composition and stand structure. Although these factors have produced certain stand conditions favorable to Spotted Owls, the legacy of fire suppression enhances the probability of widespread stand- replacement wildfires. An active program of forest management, using prescribed fire and various silvicultural tech- niques, should be developed now to reduce this risk. All such adaptive management experiments must be compatible with management guidelines that ensure the long-term survival of Spotted Owls. Joseph B. Buchanan, 1991. M.S. thesis, College of Forest Resources, University of Washington, Seattle, WA 98195. Present address: Nongame Program, Washington Department of Wildlife, 600 Capitol Way N., Olympia, WA 98501-1091 and Cascadia Research Collective, Waterstreet Building, Suite 201, 21814 West Fourth Avenue, Olympia, WA 98501. J. Raptor Res. 26(l):47-48 © 1992 The Raptor Research Foundation, Inc. News and Reviews Phylogeny and classification of birds by Charles G. Sibley and Jon Ahlquist. Yale University Press, New Haven, CT, 1991. 1008 pp. ISBN 0-300-04085-7. $100 U.S. Distribution and taxonomy of birds of the world by Charles G. Sibley and Burt L. Monroe, Jr. Yale University Press, New Haven, CT, 1991. 1136 pp. ISBN 0-300-04969-2. $125 U.S. The first of these two monographs by Charles G. Sibley and Jon Ahlquist, presents the results of long-term studies of the higher classification of the Class Aves, using the complex technique of DNA-hybridization. The book includes detailed summaries of past classifications. One finds that the Order Falconiformes as such has disappeared; all the diurnal birds of prey now reside in a vastly expanded Order Ciconiiformes where they fall between the auks and the grebes! The New World vultures are placed next to the storks proper. Anatomists have long since told us that the cathartids belong with the storks, but what about all the others? This conglomerate Ciconiiformes contains not only auks, hawks, grebes and storks but also the shorebirds, albatrosses, loons, and even the penguins! If this is indeed a natural clade, its origin, keeping one eye on the fossil record, must have been early in avian history, perhaps the Cretaceous. Subsequent radiation has been so great that probably little if anything is to be learned about falconiforms, with the possible exception of the cathartids, by comparing them with any of their existing “relatives.” Whether this new classification is the last word is a moot point; for some criticism see Gill and Sheldon (1991, Science 252:1003- 1005). At the family level the treatment is conventional and conservative: with families Accipitridae, Sagittariidae, Falconidae, Cathartidae and subfamily Pandioninae. In the second monograph Charles Sibley, with the energetic assistance of Burt M. Monroe, Jr., provides a list of the roughly 9000 species of living or recently extinct birds, showing complete ranges and some indication of habitat. Subspecies are not listed. At the genus and species level the treatment is conservative; a bit more so than I would prefer. Borderline taxa are usually given the benefit of the doubt and called species; superspecies are indicated for such allospecies, which is helpful. For example, the Gray Hawk is divided into 2 species, Asturina plagiata and A. nitida ; the American Elanus kite, is a full species E. leucurus as is the Guadalupe Caracara, Polyborus lutosus, etc. Turning to the owls, the Order Strigiformes has survived, but with the former Caprimulgiformes (e.g., nightjars) brought into it. Nothing drastic there. The barn owls are given family ranking, Tytonidae. Does the student of raptors need these 2 ponderous, expensive volumes? Various world lists of avian species are now available; though none with ranges in quite so much detail. Treatment of the falconiformes in the revised first volume of “Peters” Check-list of Birds of the World (1979, Cambridge Mus. Comp. Zool.) does include subspecies, but with briefer ranges. Volume 5 in this series 1941, which includes owls, is much out of date and there is no good world list of owls with subspecies. Hawks and Owls of the World by the reviewer, J. Bull, J.T. Marshall and B King, lists species with brief ranges (1988, Western Found. Vert. Zool., Los Angeles). All of these publications and others that might be mentioned, will differ in the treatment of some taxa. For North America, reference to the 6th edition of the Check-list of North American Birds (Washington: Am. Orn. Union), which now includes the area south through Panama, should invoke little criticism. Meanwhile, the more difficult genera of owls attract attention: C. Koenig (1991, Ekol. Vogel 13:15-76) has described 2 new species of pygmy owls ( Glaucidium ) from the Andes, while Joe Marshall and his associates continue to struggle with the puzzling scops and screech owls, Otus ( Wilson Bull. 1991:311-315). — Dean Amadon 4th World Conference on Birds of Prey and Owls. This conference will be held from 10-17 May 1992 in Berlin. Up to 31 December 1991, the Registration Fee is US$110.00 (£75.00), and thereafter US$135.00 (£90.00). For further information, apply to the World Working Group on Birds of Prey (15b Bolton Gardens, London SW5 0AL, Great Britain or Wangenheimstr. 32, 1000 Berlin 33, Germany). The Scientific Program will comprise the following paper sessions (and conveners): The Systematics and Taxonomy of Raptors: With Emphasis on Contemporary Methodology (C.M. White and A. Kemp), Population Studies: Aspects of Long-term Changes in Numbers and Distribution of Raptors and Owls (A. Kostrzewa and V. Galushin), Declining Raptor Populations: Their Biology and Conservation (B.-U. Meyburg and R.D. Chancellor), Environmental Con- taminants and Raptors (R.W. Risebrough), Biology and Conservation of the Large Falcons in the Subgenus Hierofalco (T J. Cade, W. Baumgart and C.M. White), Population Ecology of Owls (E. Korpimaki and H. Pietiainen), The Biology of Extirpated, Rare or Lesser Known Owls (R.J. Clark and H. Mikkola), Tropical Rain Forests and Raptors 48 News and Reviews Vol. 26, No. 1 (J.-M. Thiollay), Reintroductions of Eagles, Vultures and Other Raptors (J. Love and M. Terrasse) and Trapping, Marking and Radio-tagging Techniques (R. Bogel and R. Kenward). Due to the recent fundamental political changes, it is now possible to offer excursions from Berlin to the new federal states of Brandenburg and Mecklenburg-Vorpommern (in the former German Democratic Republic or “East Ger- many”) which have hitherto been virtually inaccessible to ornithologists from the West and where there are good chances to observe White-tailed Sea Eagle, Osprey, Lesser Spotted Eagle, Hen Harrier, Montagu’s Harrier, Red Kite, Peregrine Falcon, Great Bustard, Black Stork. News 1991 Leslie Brown Memorial Grant Recipient John D. Foss Born in Chicago in 1953, John D. Foss grew up in the rural Chain-of-Lakes region of northern Illinois. After high school, the lure of the southwest caused a migration to the Four Corners region where he completed his Bachelor of Science in Geology and Biology (minor) at Fort Lewis College in Durango, Colorado in 1989. John is currently working on a Master of Science degree in raptor biology under the supervision of Tom Cade at Boise State University in Boise, Idaho. His thesis project involves a river survey of habitat use by raptors and other avian species along Chile’s Rio Bio-Bio, currently threatened by a series of hydroelectric projects partially funded by the World Bank. Fabian Jaksic, RRF International Director, and Bryan Brown will be assisting John in the field. John’s academic and conservation interests include the impact of dams on the physical and biological components of ecosystems. John is an avid climber, kayaker, skier and mountain biker who has worked as a guide and naturalist in California, Idaho, Prince William Sound, Alaska and the Grand Canyon, Arizona. He has been involved with informal bird surveys at the Chilkat River Bald Eagle Reserve in Haines, Alaska and Tufted Puffin colonies in Kachemak Bay, Alaska. He has also worked as a geologist and hydrologist on watershed rehabilitation projects for the Forest Service and private consultants. A copy of “Establishing a Nest Box Program for American Kestrels Along an Interstate Highway: Recommendations Based on the Iowa Program,” by Daniel E. Varland, Ronald D. Andrews, and Bruce L. Ehresman is available free of charge. This illustrated, 8-page color report sponsored by the Iowa Department of Transportation not only describes how to start an interstate highway nest box program for kestrels but also contains a nest box plan and life history information for the American Kestrel. Write to: Iowa Department of Transportation, Office of Project Planning, Planning and Research Division, 800 Lincoln Way, Ames, IA 50010. THE RAPTOR RESEARCH FOUNDATION, INC. (Founded 1966) OFFICERS PRESIDENT: Richard J. Clark SECRETARY: Betsy Hancock VICE-PRESIDENT: Michael W. Collopy TREASURER: Jim Fitzpatrick BOARD OF DIRECTORS EASTERN DIRECTOR: Keith L. Bildstein CENTRAL DIRECTOR: Thomas Nicholls MOUNTAIN & PACIFIC DIRECTOR: Stephen W. Hoffman CANADIAN DIRECTOR: Paul C. James INTERNATIONAL DIRECTOR #1: Fabian M. Jaksi6 INTERNATIONAL DIRECTOR #2: Eduardo E. ISigo-Elias DIRECTOR AT LARGE #1: Michael W. Collopy DIRECTOR AT LARGE #2: Robert E. Kenward DIRECTOR AT LARGE #3: Jeffrey L. Lincer DIRECTOR AT LARGE #4: David M. Bird DIRECTOR AT LARGE #5: Paul F. Steblein DIRECTOR AT LARGE #6: Gary E. Duke EDITORIAL STAFF EDITOR IN CHIEF: Josef K. Schmutz, Department of Biology, University of Saskatchewan, Sas- katoon, SK., Canada, S7N 0W0 ASSOCIATE EDITORS Keith L. Bildstein Susan B. Chaplin Charles J. Henny C. Stuart Houston Robert. E. Kenward Eudoxio Paredes-Ruiz Patricia P. Rabenold Patrick T. Redig EDITOR OF RRF KETTLE: Paul F. Steblein 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, letters to the editor, thesis abstracts 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 215 by 280 mm (8 Vi x 11 in.) or standard international, white, bond paper, with 25 mm (1 in.) margins. 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. Submit the current address on a separate page placed after the literature cited section. 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 numerals. 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 (6th ed., 1983) 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 1990). Refer to a recent issue of the journal for details in format. Explicit instructions and publication policy are outlined in “Information for contributors,” /. Raptor Res., Vol. 24(1-2), which is available from the editor. 1992 ANNUAL MEETING The Raptor Research Foundation, Inc. 1992 annual meeting will be held on 11-15 November at the Red Lion Inn in Bellevue (Seattle suburb), Washington. Details about the meeting and a “call for papers” will be mailed to Foundation members in summer, and can be obtained from Mark Stalmaster, Scientific Program Chairperson, Stalmaster and Associates, 209 23rd Avenue, Milton, WA 98354; Tel. (206)922-5435. For further information about the meeting, or the associated Spotted Owl Symposium and art show, contact Lenny Young, Local Committee Chairperson, 5010 Sunset Dr. NW, Olympia, WA 98502; Tel. 0(206)753-0671 H(206)866-8039, FAX (206)586-6126. RAPTOR RESEARCH REPORTS #1, R.R. Olendorff. 1971. Falconiform Reproduction: A Review Part 1. The Pre-nestling Period. $10.00 members, $12.50 non-members. #2, F.N. Hamerstrom, B.E. Harrell and R.R. Olendorff [Editors]. 1974. Management of Raptors. Pro- ceedings of the Conference on Raptor Conservation Techniques, Fort Collins, CO, 22-24 March 1973. $10.00 members, $12.50 non-members. #3, J.R. Murphy, C.M. White and B.E. Harrell [Editors]. 1975. Population Status of Raptors. Proceedings of the Conference on Raptor Conservation Techniques, Fort Collins, CO, 22-24 March 1973. (Part 6). $10.00 members, $12.50 non-members. #4, R.R. Olendorff, A. Miller and R. Lehman [Editors]. 1981. Suggested Practices for Raptor Protection on Powerlines: State of the Art in 1981. $5.00 members, $20.00 non-members. #5, S.E. Senner, C.M. White and J.R. Parrish [Editors]. 1986. Raptor Research Conservation in the Next Fifty Years. Proceedings of a Conference held at Hawk Mountain Sanctuary, Kempton, PA, 14 October 1984. $3.50 members, $4.50 non-members. #6, D.M. Bird, and R. Bowman [Editors]. 1987. The Ancestral Kestrel. Proceedings of a Symposium on Kestrel Species, St. Louis, MO, 1 December 1983. $10.00 members, $12.50 non-members. #7, R.R. Olendorff [Editor]. 1989. The Raptor Research Foundation, Inc. Bibliographic Index (1967-1986) . $2.50 members, $5.00 non-members. #8, R.R. Olendorff, D.D. Bibles, M.T. Dean, J.R. Haugh and M.N. Kochert. 1989. Raptor Habitat Management under the U.S. Bureau of Land Management Multiple-Use Mandate. $5.00 members, $6.50 non-members. Add $2.50 for postage and handling, and $1.00 each for additional reports. BOOKS Biology and Management of Bald Eagles and Ospreys. Proceedings of the First International Symposium, Montreal, Canada. D.M. Bird [Editor]. 1983. $15.00 members, $18.00 non-members plus $5.00 shipping. JOURNAL BACK ISSUES Journal Back Issues are available. For details see page 73 of the Kettle or write: Jim Fitzpatrick, Treasurer, Raptor Research Foundation, Inc., Carpenter St. Croix Valley Nature Center, 12805 St. Croix Trail, Hastings, MN 55033.