Raptor Research A Quarterly Publication of The Raptor Research Foundation, Inc. Volume 20, Number 1, Spuing 1986 {ISSN QQ9±9Q59) Contents Population Ecology of the Harris' Hawk in Arizona. Wayne H. Whaley ........ . 1 Crvop reservation of Peregrine Falcon Semen and Post-Thaw Dialysis to Remove Glycerol. John E. ftirb, Wiltord R. Hcck and Vidor Hnrrtaswick .. . + . „ + + . . , . . „ , . . 16 Characteristics or Cliffs and Nest Sites Used by Breeding Prairie Falcons, d.e. Rundcand s.H. Anderson 21 Development of Hunting and Self-sufficiency in Juvenile Red-tailed 1 Iawks (Buleu jajruucemu), -Sara Jane Jobs isun 29 Post Fledging Behavior of Ferruginous Hawks in North Dakota. PauJ M. Konrad and Dwid S. Gilmer FtttH H ,, tti t 35 Erratum . • 38 Short Communications Ohscrvadon^ of Nesting Northern Pygmy-Owls. Denver W. Hole and William D. Norton . 39 An Unusual Incident with (hr Bald Eagle [tfatuuvtm tevcttfpknJui), Jerry Olsen . 4 I I enipwal Fliittuabom of Rough-legged HiLvks During Carrion Abundance, James W. Waison. 42 Dcrafries of Rcd-iaiTeri Hawk Non i" Aspen Stands in the Entrance Baiio, Colorado. Mike McGovern jihl Juhn M. McNumcy . . r x + 43 Thesis Abstracts . . t , , , p , + , + 1 45 Dissertation Abstracts 47 Reviewers fob Raptor Researrk, L9H5 46 News and Reviews 46 News and Reviews . H 9 ? | 5 (i The Raptor Research Foundation, Inc. Provo* Utah THE RAPTOR RESEARCH FOUNDATION, INC. (Founded 1966) OFFICERS PRESIDENT: Jeffrey L. Lincer, Office of the Scientific Advisor, 2086 Main Street, Sarasota, Florida 33577 VICE-PRESIDENT : Richard Clark, York College of Pennsylvania, Country Club Road, York, Pennsylvania 17405 SECRETARY: James D Fraser, Virginia Polytechnic Institute and State University, Cheatam Hall, Blacksburg, Virginia 24061 TREASURER: Jim Fitzpatrick, Carpenter Nature Center, 12805 St. Croix Trail, Hastings, Minnesota 55033 BOARD OF DIRECTORS EASTERN DIRECTOR: James A. Mosher, Appalachian Environmental Laboratory, University of Maryland, Frostburg State College Campus, Gunter Hall, Frostburg, Maryland 21532 CENTRAL DIRECTOR: Patrick T. Redig, Department of Veterinary Biology, 295 Animal Science/Veterinary Medicine Building, University of Minnesota, St. Paul, Minnesota 55108 MOUNTAIN 8c PACIFIC DIRECTOR: A1 Harmata, Department of Biology, Montana State University, Bozeman, Montana 59717 EAST CANADA DIRECTOR: David M. Bird, Macdonald Raptor Research Centre, Macdonald Campus of McGill University, 21,111 Lakeshore Road, Ste. Anne de Bellevue, Quebec H9X ICO WEST CANADA DIRECTOR: R. Wayne Nelson, 4218 -63rd Street, Camrose, Alberta T4V 2W2 INTERNATIONAL DIRECTOR: Martin Bottcher, Postfach 2164, Steinfelder Strass 11, 5372 SCHLEIDEN, Federal Republic of Germany, GERMANY DIRECTOR AT LARGE # 1 : Michael Collopy, University of Florida, School of Forest Resources and Conservation, 118 Newins-Ziegler Hall, Gainesville, Florida 32601 DIRECTOR AT LARGE #2: Tom Dunstan, Department of Biological Sciences, Western Illinois University, Macomb, Illinois 61455 DIRECTOR AT LARGE #3: Richard P. Howard, U.S. Fish and Wildlife Service, 4696 Overland Road, Room 566, Boise, Idaho 83705 * * * sH * * $ * * * * * $ sic * * * * Persons interested in predatory birds are invited to join the Raptor Research Foundation, Inc. Dues are $15 per year in the U.S., $17 per year outside the U.S., $13 per year for U.S. students, and $15 per year for students outside the U.S. Add $2 to dues if membership is received after 15 February. The Foundation’s journal Raptor Research is distributed quarterly to all current members. Subscription price to institutions and nonmembers is the same as regular member- ship. Single copies and back issues are available from the Treasurer. A Contributing Membership is $25, a Sustaining Membership is $100, and a Life Membership is $500. All contributions to The Raptor Research Foundation, Inc., are tax-deductible. Send requests for information concerning membership, subscriptions, special publications, or change of address to the Treasurer. Other communications may be routed through the appropriate Officer or Board member. All inquiries concerning the journal should be addressed to Clayton M. White, Editor, Raptor Research, Department of Zoology, 161 WIDB, Brigham Young University, Provo, Utah 84602, U.S.A. ******************* Published quarterly by The Raptor Research Foundation, Inc. Business Office: Jim Fitzpatrick, Carpenter Nature Center, 12805 St. Croix Trail, Hastings, Minnesota 55033, U.S.A. Raptor Research A Quarterly Publication of The Raptor Research Foundation, Inc. Volume 20, Number 1, Spring 1986 (ISSN 0099-9059) Contents Population Ecology of the Harris’ Hawk in Arizona. Wayne H. Whaley 1 Cryopreservation of Peregrine Falcon Semen and Post-Thaw Dialysis to Remove Glycerol. John E. Parks, Willard R. Heck and Victor Hardaswick 16 Characteristics of Cliffs and Nest Sites Used by Breeding Prairie Falcons, d.e. Runde and S.H. Anderson 21 Development of Hunting and Self-sufficiency in Juvenile Red-tailed Hawks (Buteo jamaicensis). Sara Jane Johnson 29 Post Fledging Behavior of Ferruginous Hawks in North Dakota. Paul M. Konrad and David S. Gilmer 35 Erratum 38 Short Communications Observations of Nesting Northern Pygmy-Owls. Denver W. Holt and William D. Norton 39 An Unusual Incident with the Bald Eagle ( Haliaeetus leucocephalus). Jerry Olsen 41 Temporal Fluctuations of Rough-legged Hawks During Carrion Abundance. James W. Watson 42 Densities of Red-tailed Hawk Nests in Aspen Stands in the Piceance Basin, Colorado. Mike McGovern and John M. McNurney 43 Thesis Abstracts 45 Dissertation Abstracts 47 Reviewers for Raptor Research, 1985 48 News and Reviews 48 News and Reviews 149,150 The Raptor Research Foundation, Inc. Provo, Utah HWiMl RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. Vol. 20 Spring 1986 No. 1 POPULATION ECOLOGY OF THE HARRIS’ HAWK IN ARIZONA Wayne H. Whaley Abstract - The Harris’ Hawk (Parabuteo unicinctus) was studied in Arizona during 1976-1977. Where 2 separate populations once resided in Arizona, 1 now remains and occupies 3,880 km 2 of the Arizona Upland subdivision, Sonoran Desert. The former population along the Colorado River is extirpated. Nests were built or old nests repaired from January to August and eggs were laid from mid-January to mid-August. Subsequent clutches were documented in 50 nesting ranges. Occasionally second clutches were laid before young of first broods were fledged. Fledging dates ranged from April to October. At 9 locations active nests were 0.8 km apart and at 2 locations nests were 0.5 km apart. Within 2 study areas nesting density was 2.5 km 2 /active nest. In 1977 Harris’ Hawks reoccupied 91% of the nesting ranges used in 1976. More than 2 adults were observed at 46% of 227 nesting ranges. Productivity for 396 nesting attempts averaged 3. 16 eggs/clutch and 1 .62 young fledged/nesting attempt. Seventy-four percent of the nesting attempts were successful. Food consisted mainly of rabbits and ground squirrels. Mortality occurred mainly during the egg laying and incubation period. Early nest failures resulted in second broods, but successful pairs also had second broods. Habitat loss is the major cause of decline of the population followed by excessive human disturbance. Research on raptors has become increasingly im- portant as a result of the marked decline in several species over the past 3 decades (Arnold 1954; Cot- tarn et al. 1961; Berger, Sindelar and Gamble 1 969; Peterson 1969; Sprunt 1969; Henny and Wight 1972). In the southwest, particularly in Arizona, the ranges of some species of raptors extend marginally into the United States, and here population studies are usually most informative. Most of these border species have received little investigation. For exam- ple, the Aplomado Falcon ( Falco femoralis ) showed signs of decline as early as 1890 (Phillips, Marshall and Monson 1964). Its range and status in Arizona was poorly documented (Phillips et al. 1964) until the study of Hector (1975). Where it was once thought to be locally fairly common, it is extinct (Hector 1975). Because of habitat destruction, Gray Hawk ( Buteo nitidus) populations have declined during the last century (Richard Glinski pers. comm.). The Harris’ Hawk (Parabuteo unicinctus superior ) is another relatively unstudied southwestern raptor. Studies by Hensley (1959) and Mader (1975a, 1975b, 1977) are the only major works on this species in Arizona. Because of sudden decline and apparent extinction in southern California (R. Guy McCaskie pers. comm.), and with the recent (late 1960) increased use for falconry, knowledge of the Harris’ Hawk’s status in Arizona is critical. The primary purpose of this study was to estab- lish base-line data on nesting distribution and abundance of the Harris’ Hawk in Arizona and to compare these data with the historic record to de- termine nesting success of the present population. Study Area and Methods Arizona falls into a southwestern bi-seasonal climatic pattern of winter precipitation, spring dry period, summer precipitation, and fall dry period. The spring dry period (May - June) has higher temp and the greatest influence on the plant and animal com- munity (Lowe 1976). Sellers (1960) divided the state into homogeneous sections with respect to climate, topography, and vegetation. From the south- west section (the area of importance in this study) to the plateau section there are extreme changes in climatic conditions. The plateau section has an average annual temp 20 to 25° F lower and annual precipitation 38 to 5 1 cm higher than the flat deserts of the southwest section. The southwest section contains the lowest, hot- test and driest areas of the state (Sellers 1960). Of the 6 life-zones in Arizona, only the Lower Sonoran, con- taining portions of the Sonoran, Mojave and Chihuahuan Desert (Fig. 1), was important to this study. The Sonoran Desert has 2 subdivisions in Arizona — the Lower Colorado and Arizona Up- land subdivisions (Fig. 1). The Lower Colorado subdivision (ele- 1 Raptor Research 20 (1): 1-15 2 Wayne H, Whaley Vol. 20, No. 1 Figure 1 . Map of the historical and present distribution of the Harris’ Hawk in Arizona in relation to vegetational zones. vation 30-90 m) is primarily a creosote bush-bursage ( Larrea - Franseria) community and includes the hottest and most arid re- gions in Arizona. This subdivision is characterized by sandy and gravelly plains and mesas, sand dunes, lava flows, silty valleys, salt basins, rocky hills, and desert pavement. The Arizona Upland subdivision (elevation 150-1220 m) is typified by a paloverde- saguaro cactus ( Cericidium-Cereus ) association. It attains greatest development on the rocky soils of desert mountain slopes and on the coarse soils of upper bajadas that flank mountain ranges. The Arizona Upland subdivision has far greater numbers of plant species than the Lower Colorado subdivision. The Mojave Desert (elevation 240-1580 m) scarcely reaches into northwestern Arizona. The higher aridity, lack of summer rains and longer periods of below freezing winter temp cause this desert to be poorer in plant and animal life than the Sonoran Desert (Lowe and Brown 1973). The Chihuahuan Desert enters Arizona in a small southeastern section of the state, where it lies mostly above an elevation of 1070 m. This relatively complex desert is essentially shrubby with many grasses, several small species of cacti, and few desert trees. Field work was conducted in 1976 and 1977 from January through October of each year. The Harris’ Hawk in Arizona has never been observed nesting outside the Lower Sonoran Life- zone, so I eliminated the northern half of the state from the survey. After a broad general search of the Lower Colorado sub- division, I determined that the Arizona Upland subdivision was the most important area to nesting Harris’ Hawks. Hills, windmill towers, and other elevated points were used to locate nests. In flat country I used an 8.5 m extension ladder mounted on the bed and cab of a pickup truck which allowed visual clearance above the vegetation. All nest sites were plotted on USGS 1:250,000 topographic Spring 1986 Harris’ Hawk in Arizona 3 quadrangles and series maps. Throughout the population con- tinuum there were local concentration points and from 2 of these populous areas I attempted to find all active nests to determine total nesting density. The mean distance between nests was deter- mined by measuring the distance from each nest and its nearest neighbor using the distance between any 2 nests only once. By halving the mean distance, a value r (radius) was obtained and used to determine average nesting range size by the formula A =■ r 2 under the assumption that nesting ranges were circular. I used only those nesting ranges that were active in a given year. Each nest was visited a minimum of twice in order to determine number of eggs laid and number of young fledged (in 1976 nests were visited 3 to 5 times). When a clutch appeared small or incom- plete, a later visit was scheduled in order to obtain data for full clutch size. Young were aged by comparison with color photo- graphs of known age birds taken at 5 d intervals. Results Distribution and Habitat Characteristics Past Populations. — Historically, there were 2 localized populations of Harris’ Hawk in Arizona. One population was resident in western Arizona in Lower Colorado habitat along the Colorado River from Yuma northward to Topock (Fig. 1). The Lower Colorado population was extirpated by 1969. A second population approximately 280 km eastward in Arizona Upland habitat (Fig. 1) re- mains today. The 2 localities are separated by a dry, barren expanse of the Lower Colorado Desert. The earliest record for Harris’ Hawk in Arizona was reported near Topock by Kennedy (1859). The Lower Colorado population likely originated from Baja California, extending its range into the United States along suitable segments of the Colorado River (Fig. 1). Along the river, nesting sites were near small lakes, lagoons and swamps in flooded mesquite and in willow and cottonwood trees (Wiley 1916, 1917; Bancroft 1920; Rowley 1936). Prey items included the Purple Gallinule (Porphyrula martinica ), Sora ( Porzana Carolina), Common Teal (Anas cyanoptera), and Northern Flicker ( Colaptes auratus) (Miller 1925, 1930). Gale Monson (pers. comm.; past manager of Havasu Natl. Wildl. Refuge) included the American Coot ( Fulica americanus ), Muskrats (Andatra zibethicus), Cotton Rats ( Sigmodon hispidus), and Abert’s Towhee (Pipilo aberti) as food found in nests. Harris’ Hawks were frequently common at places along the river near Havasu National Wildlife Refuge with a resident average of 30 individual/y from 1947-59. As many as 50 individuals were often reported, and nests were sometimes less than 0.8 km apart (U.S.D.I. Fish and Wildlife Service 1952). Generally, 2 young/nest were reported with only 2 parent birds caring for them. Nests were often built on top of old Great Blue Heron (Ardea herodias) nests and were commonly only 3 to 5 m above water (Gale Monson pers. comm.). A population decline started in the late 1950’s and by 1969 the Harris’ Hawk disap- peared from the refuge. A wild population has not since been observed along the Colorado River. A concurrent decline and extinction was noted at Im- perial National Wildlife Refuge to the south. An extensive list of sources concerning Harris’ Hawks along the Colorado River is included in Whaley (1979, Appendices A and B). The Present Population (1976-1977). — The present Harris’ Hawk population in Arizona oc- cupies 3,880 km 2 in Arizona Upland habitat with elevations ranging from 396 to 1,036 meters (X = 701 m). Lowe and Brown (1973) delineate prime Arizona upland habitat as the region “. . . east then north of a line drawn from Ajo to Tucson to Flor- ence Junction, then northwest to Wickenburg” and the Harris’ Hawk followed this distribution closely (see Fig. 1). Nearly all nests were placed in paloverde-saguaro cactus habitat or in the more local narrow strips of blue paloverde-ironwood (Circidium floridum-Olneya) habitat of the large ar- royos. Three exceptions were nests placed in large cottonwood trees in riparian communities that were juxtapose to the aforementioned habitat. No nests were found, nor birds seen, in riparian com- munities along rivers that were not associated with the Arizona Upland subdivision (Fig. 1). Nest Site Characteristics Saguaro cactus was the preferred plant species used for nest sites (Table 1). Five active nests and 42 old nests were placed on electrical transmission tow- ers along a 13.7 km section of 110 kV powerline crossing paloverde-saguaro cactus habitat. Another nest was placed on a tower along a 345 kV electrical transmission line crossing excellent Harris’ Hawk habitat. Nest height ranged from 2.3 m [foothill paloverde ( C . microphyllum ) tree] to 21.3 m (electri- cal transmission tower). The average height for nests in saguaro cacti, blue paloverde trees, and foothill paloverde trees was 5.8 m, 6.3 m and 4.4 m, respectively. Seventy-two percent of all nests were placed in mature saguaro ss 4.9 m tall with substan- tial arms. 4 Wayne H. Whaley Vol. 20, No. 1 Table 1. Harris’ Hawk nest sites in 1976 and 1977 in Arizona. Nest Support Number of Nests Percentage or Total Saguaro Cactus 230 75.2 Foothill Paloverde 37 12.1 Blue Paloverde 23 7.5 Electrical Tower 6 2.0 Cottonwood 3 1.0 Ironwood 3 1.0 Mesquite 2 .6 Pine 1 .3 Palm 1 .3 Totals 306 100.0 Breeding Season Phenology Courtship. — Courtship behavior of Harris Hawks is typical of most raptors (Brown and Ama- don 1968), but on 18 February 1977 I noted a very unusual “group courtship” display involving 8 adults. The adults flying at an altitude of 150-180 m, continually engaged in soaring, tail chasing, and stooping, accompanied by much vocalizing. The long vertical stoops, which often involved all 8 hawks, were followed by tail chasing and eventual return to their former altitude where the event was repeated. Similar behavior has been reported for Eleonora’s Falcon (Falco eleonorae) when near the breeding cliffs (Brown and Amadon 1968) but is apparently rare in falconiforms. The observed be- havior lasted approximately 45 min, when the adults departed in 3 directions. Later, an active nest was found in each of the 3 areas where the respec- tive groups appeared to have flown. At 2 of these nests, 3 adults were present. Copulations occurred over a 6-month period from 28 January to 26 July. Each copulation bout lasted from 15 to 40 sec. (X = 24 sec, N = 23). Courtship behavior did not always precede mating; on several occasions I witnessed copulations in which a male flew directly to a female and copulated without obvious display by either sex. More than 1 adult males were observed at many of the nest sites ; 2 males were recorded at 41% and 3 at 5% of 227 active nesting ranges. On 11 February 1977, I ob- served 7 copulations in 2 h, involving a female and both males at a territory near Florence Junction. Polyandrous mating behavior has been previously reported for this species (Mader 1975a). Nest Construction. — Harris’ Hawks build or repair nests from January to August. Often one of several alternate nests is repaired or pairs may use the same nest several times in succession. Several nests may be repaired and 1 chosen for use. In 1 territory 8 nests were located, one of which had been repaired and was ready for eggs on 10 Feb- ruary 1977 (copulation observed on 28 January 1977). Materials are continually added to nests while eggs and nestlings are present. Timing of Incubation, Hatching, Fledging. — Using a 35-d incubation period and a 45-d nestling period (Mader 1975b), I determined the time spans for beginning of incubation, for hatching and for fledging of young (Fig. 2). Eggs were laid from mid-January to mid-August. Fifty percent of all first clutches (N = 284) were laid between 20 Feb- ruary and 22 March (Fig. 2). A portion of clutches laid during and after April were second clutches for the given year (Fig. 2). Double or triple clutches occurred 61 times in- volving 50 (21.6%) of the nests studied; 39 (63.9%) followed a successful first attempt, and 22 (36.1%) followed an unsuccessful first nesting attempt (X 2 = P < .05). Of those following a successful brood, the time interval between first and second clutches averaged 106 d. Within 8 nesting ranges the short interval between clutches (X = 75 d) indicated that eggs were laid before the first young had fledged. This was confirmed within 2 nesting ranges. Within one, 2 eggs were laid in the same nest with a 23 to 25 d old chick. In Texas the time interval from fledg- ing of the first attempt to completion of the next clutch was 28 d (range 7-59 d) for 6 second attempts (Brannon 1980). January egg laying was documented only once during the study period (eggs laid mid-January 1976). On 12 February 1977 another nest contain- ing 3 eggs was found which may have been started in January, but egg-laying or incubation date could not be determined since the eggs never hatched. Whenever pairs commenced laying of the first clutch early in the year, 3 clutches/breeding season were laid during both 1976 and 1977. Hatching dates for 284 clutches spanned 8 months (mode = April) and fledging dates spanned Spring 1986 Harris’ Hawk in Arizona 5 7 mo (mode = May) (Fig. 2). The latest fledging date was near 28 October. An extremely late fledg- ing date of 8 November has been reported for Har- ris’ Hawks in Texas (Brannon 1980). Nesting Density and Territorial Fidelity Nesting density of Harris’ Hawks in Arizona was 1 nest/2.5 km 2 in 1977 for both Study Areas A andB (Fig. 3). The mean distance between nests in both study areas was 1.8 km. At 9 locations active nests were only 0.8 km apart, and at 2 locations active nests were just 0.5 km apart. As a result of the observed close nesting patterns, I expect that ter- ritories overlap. Territorial attachment seems strongly de- veloped. Eighty-four percent of 123 active ter- ritories occupied in 1976 were reoccupied and ac- tive in 1977. Another 9 territories occupied in 1976 had adults present in 1977 but active nests were not found. On this basis, Harris’ Hawks reoccupied 91% of 123 nesting territories used in 1976. Mader (1982) found that the Savanna Hawk ( Buteogallus meridionalis ) also has strong territorial attachment from year to year. Eight nests used by Harris’ Hawks in 1976 were used by Great Horned Owl (Bubo virginianus ) (N = 7) and Red-tailed Hawk (Buteo jamaicensis) in 1977 and appeared to have little influence on Harris’ Hawks reoccupancy of old territory. The Red- tailed Hawk usurpation resulted in the Harris’ Hawks locating a new nest 160 m from their 1976 site. Productivity I recorded a total of 396 nesting attempts (in- cluding second and third clutches) involving 306 nests within 231 active nesting ranges. In order to compare productivity with that from other studies (Mader 1975b; Griffin 1975; Griffin 1976; Bran- non 1980), a successful nest was one in which a nestling reached the age of at least 28 d. Of 319 nesting attempts, 72% were successful. Seventy- seven nests had incomplete data. Seventy-five nests failed during the egg laying and incubation period. Fifteen nests failed during the nestling period (2 were man-caused, 2 were caused by inclement weather, and cause of failure could not be deter- mined for 11). Mean clutch size in 1976 was 3.04 (N = 67) and 3.22 (N = 95) in 1977 with a combined mean of 3. 16 (N = 162) (clutch size/number of clutches: 1/4, 2/29, 3/73, 4/51, 5/5) for the study period. Number Figure 2. Number of nests in relation to (A) month when incubation begins, (B) month of hatching and (C) month of fledging for 284 Harris’ Hawk nesting attempts in Arizona in 1976 and 1977. 6 Wayne H. Whaley Vol. 20, No. 1 Figure 3. Proximity of Harris’ Hawk nests in Study Area A (230 km 2 ), containing 39 nesting ranges with 51 active nests, and in Study Area B (135 km 2 ), containing 26 nesting ranges with 38 active nests. Large circles containing > 1 symbol indicate 1 nesting range and are not intended to delineate the size of a range. of fledglings/nest ranged from 1 to 4 (number fledglings/number broods: 1/59, 2/75, 3/71, 4/24) and averaged 1.62 young/all attempts (Table 2). I calculated productivity using the more com- plete data of 71 nesting attempts of 1976 and com- pared these values with productivity obtained when using the less complete, but larger sample size, of the foregoing. Using only the 1976 data, “hatching” and fledging success were calculated (Table 3). Values obtained for “mean clutch size” and “young fledged/all attempts” were similar to the corres- ponding values from the larger sample. There was a 51% loss in productivity between egg laying and fledging, most resulting from eggs failing to hatch (41%, Table 3). Egg loss was recorded for 89 nests, 42% due to nest failure (loss of complete clutches) and 58% due to partial hatching of clutches. A Texas population had a 41% productivity loss (Brannon 1 980) . At least 612 young fledged during the study period. Two hundred twelve nestlings were banded in 1977. Sex ratio of nestlings was 1.2 c7cf: 1.0 $ 2. Harris’ Hawks in sub-adult plumage were found to breed occasionally. Ten sub-adults within 6 ter- Spring Harris’ Hawk in Arizona 7 Table 2. Harris’ Hawk nesting success in Arizona, 1976 and 1977. Total Number of Nesting Attempts Number of Successful Nests 1 Number of Young Young Fledged Per all Attempts Young Fledged Per Successful Attempts 1976 143 102 (71%) 229 1.60 2.25 1977 Total nests 176 127 (72%) 289 1.64 2.28 and young 319 229 (72%) 518 1.62 2.26 1 A nest was considered successful if a chick was raised to an age of at least 28 days. Only nests where adequate fledging data was obtained appear in the table. ritories succeeded in raising 7 young. In each case the female was a sub-adult. In 3 territories all hawks were sub-adults. Mortality Factors Hatching success (59%, Table 3) indicates a high egg loss. Laying of infertile eggs, nest abandonment and destruction by predators contributed to re- duced hatching success. Infertility appeared to be the most common cause (e.g., one pair laid 4 eggs in both 1976 and 1977; incubation times were 65 and 73 d, respectively, with no hatching and no embryo development in both years). Forty-seven young died or disappeared from nests during the 2-y period (7 young were taken by man, 5 were killed by man, 5 died of disease, 2 died when a cactus con- taining a nest fell, and 1 died when it fell from its nest). Cause of death of 27 young could not be determined. Other mortality factors for adults and fledglings were electrocutions (8 cases reported), and accidental trapping of Harris’ Hawks by coyote trappers. In 1976 a resident of Owl Head Ranch near Tuc- son stated that he found a dead Harris’ Hawk and other birds floating in the metal livestock water tanks on his ranch. In July 1978 a dead juvenile female Harris’ Hawk was found floating in a large livestock water tank (Larry Livingston pers. comm.) not far from where it was banded as a nestling in May 1977. Drowning deaths may be a significant cause of mortality, especially during the dry period (May-June) when livestock tanks are the only avail- able source for drinking and bathing. Drownings in water tanks have been documented for the Prairie Falcon ( Falco mexicanus ) (Enderson 1964), Ameri- can Kestrel (Falco sparverius) (Craig and Powers 1976) and Ferruginous Hawk (Buteo regalis) (Clayton White pers. comm.). Cholla cactus (Opuntia sp.) may also cause mor- tality in Harris’ Hawks. On 26 May 1976 I found a recently fledged Harris’ Hawk partially im- mobilized (could not fly and could hardly walk) on the ground near its nest. Cholla cactus joints were stuck on its neck and between its legs. On another occasion I chased down and caught an older fledgling that was having difficulty flying and Table 3. Summary of productivity data for 71 Harris’ Hawk nesting attempts in Arizona in 1976. Results Range Mean clutch size 3.07 1 - 5 Young fledged per all attempts 1.63 0-4 Young fledged per successful attempt 2.18 1 -4 Percent hatching success 59 41% productivity loss (eggs failed to hatch) Percent fledging success 90 10% productivity loss (young failed to fledge) Percent successful nesting attempts 75 Percent unsuccessful nesting attempts 25 8 Wayne H. Whaley Vol. 20, No. 1 keeping its balance when landing. The bird was obviously weakened. I found 2 cholla cactus joints clasped in its feet such that it could not release its grip. Survival of a raptor in such condition is ques- tionable since food would be difficult or impossible to obtain. Table 4. Prey items observed at nests of Harris’ Hawks in Arizona in 1976 and 1977. Number of Percentage Species Items of Total Mammals: Cottontail Rabbit Sylvilagus audubonii 144 22.4 Harris’ Ground Squirrel Citellus harrisii 79 12.3 White-throated Woodrat Neotoma albigula 76 11.8 Black-tailed Jackrabbit Lepus calif ornicus 9 1.4 Round-tailed Ground Squirrel Citellus tereticaudus 5 0.8 Pocket Gopher Thomomys bottae 2 0.3 Merriam’s Kangaroo Rat Dipodomys merriami 1 0.2 Banner-tailed Kangaroo Rat D. spectabilis 1 0.2 Arizona Pocket Mouse Perognathus amplus 1 0.2 Unidentified Lagomorphs 65 10.1 Unidentified Kangaroo Rats 4 0.6 Unidentified Ground Squirrels 3 0.5 Unidentified Mice 2 0.3 Total Mammals 392 61.1 Birds: Gambel’s Quail Callipepla gambelii 56 8.7 Cactus Wren Campylorhynchus brunneicapillus 42 6.5 Screech Owl Otus kennicottii 14 2.2 Northern Flicker Colaptes auratus 13 2.0 (Table 4 continued) Spring 1986 Harris’ Hawk in Arizona (Table 4 continued) Number of Percentage Species Items of Total Elf Owl Micrathene whitneyi 8 1.2 Mourning Dove Zenaida macroura 6 0.9 Curve-billed Thrasher Toxostoma curvirostre 4 0.6 Gila Woodpecker Melanerpes uropygialis 3 0.5 Road Runner Geococcyx calif ornianus 2 0.3 American Kestrel Falco sparverius 1 0.2 Cooper’s Hawk Accipiter cooperii 1 0.2 Unidentified Thrashers 12 1.9 Unidentified Doves 5 0.8 Unidentified Birds 12 1.9 Total Birds 179 27.9 Reptiles: Desert Spiny Lizard Sceloporus magister 63 9.8 Regal Horned Lizard Phrynosoma solare 2 0.3 Zebra-tailed Lizard Callisaurus draconoides 1 0.2 Western Whiptail Lizard Cnemidophorous tigris 1 0.2 Leopard Lizard Crotaphytus wislizeni 1 0.2 Unidentified Reptiles 2 0.3 Total Reptiles 70 11.0 Total Number of Prey Items 641 100.0 10 Wayne H. Whaley Vol. 20, No. 1 Food Habits A quantitative analysis of food habits in Harris’ Hawks was not possible during this study because of time restraints imposed by the large number of nests being observed. Often the only sign of occur- rence of an avian prey species was the presence of feathers which denoted little or nothing of the number of individuals taken. Thus I could only indicate that a particular species was found. Since certain body parts of larger mammals (i.e., hind feet of Cottontail Rabbits [Sylvilagus audubonii] and tails of Harris’ Ground Squirrels [Citellus harrisii ]) were not eaten and remained in the nest for long periods of time, a more accurate figure for the number of these species taken could be obtained. The food data presented represents a qualitative study of the entire population, which may provide some idea of the total gamut of prey species taken throughout the Harris’ Hawk population in Arizona. Harris’ Hawks are quite catholic in feeding habits. Twenty-five vertebrate species were re- corded (Table 4), of which 61.1% were mammals. Brannon (1980) found 65.6% of Harris’ Hawk prev in Texas to be mammals. Cottontail Rabbit appear to be important in the diet, carcasses at times cov- ering the entire tops of nests. At one nest, pairs of hind feet indicated that at least 22 lagomorphs had been taken. Harris’ Ground Squirrels, White- throated Woodrats ( Neotoma albigula), Gambel’s Quail ( Callipepla gambelii ), and Cactus Wrens ( Cam - pylorhynchus brunneicapillus ) also appear to be im- portant in Harris’ Hawk diets in Arizona. Four raptors, Screech Owl (Otis kennicottii), Elf Owl (Micrathene whitneyi). Coopers’ Hawk (Accipiter cooperii ), and American Kestrel (Falco sparverius) were taken. Discussion Phenology of the Breeding Season Harris’ Hawks in Arizona may lay eggs as early as January (Ellis and Whaley 1979) and sometimes fledge young as late as December (Radke and Klimosewski 1977) which gives them the distinction of having the longest known “breeding season” (as defined by Moreau 1950) of any temperate North American falconiform. In Sonora, Mexico (27 N°), eggs are even laid in November (Ellis and Whaley 1979), suggesting that there are continuous breed- ers in some southern parts of the Sonora Desert. The terminology “continuous breeder” follows Immelmann’s (1971) definition that within a population eggs are laid during every month of the year. This is not the case (as currently known) in the Arizona population since no egg laying has been recorded for the period October-December. Lack (1968) and Immelmann (1971) suggested that, since the reproductive period is the most rigorous and critical period of a bird’s annual cycle, it is imperative that it be scheduled at a time when young can most profitably be raised with a minimum of energy expenditure on the part of the adults. A good example, Eleonora’s Falcon of the Mediterranean region, delays breeding until Au- gust when it can feed its young on the numerous passerines in fall migration (Walter 1979). Eleon- ora’s Falcons feed mainly on insects during the re- mainder of the year. Harris’ Hawks seem to follow a similar pattern, as they breed during periods of great prey abun- dance. Breeding starts quite early in Arizona. Incu- bation begins in February and March and hatching occurs primarily in April (Fig. 2). Prey species also exhibit long breeding seasons in Arizona. Harris’ Ground Squirrel breeds from late February to May with its conception period running from 31 March to late May (Neal 1965). On this basis, Harris’ Ground Squirrels are being born when the majority of Harris’ Hawk eggs are hatching. Cottontail Rab- bits breed year-round in Arizona with peaks for conception starting in April (peak hatch time for Harris’ Hawks) and running through July (Hungerford, Lowe and Madsen 1973). The White-throated Woodrat has successive litters from January to August (Vorhies and Taylor 1940), which spans the time during which most Harris’ Hawks are breeding. Thus, local abundance and long breeding seasons of prey species may be an important facilitatory factor for an extended breeding season of Harris’ Hawks in southern Arizona. Several other prey species that have long breed- ing seasons are Gambel’s Quail (March to Sep- tember, Bent 1932), Cactus Wrens (raise 3 broods from February to August, Anderson and Anderson 1960), Curve-billed Thrasher ( Toxostoma curvirostre) (February to July, Smith 1971), and Mourning Dove ( Zenaida macroura ) (March to September, Brown 1967). The Harris’ Hawk’s catholic prey habits on species with long reproductive periods may have contributed to the development of an Spring 1986 Harris’ Hawk in Arizona 11 extended breeding season and increased produc- tivity through double brooding. When the environment permits a species to raise > 1 brood/yr, breeding begins as early as possible even though the timing of the first attempt may be less favorable (Lack 1968). A species using this strategy would tend to produce the maximum number of offspring under the prevailing cir- cumstances (Lack 1954). Harris’ Hawks may follow this strategy, as second and third breeding attempts appear to be influenced by an early first attempt (Fig. 2). The outcome (i.e., success or failure) of the first breeding attempt appears to be less important than timing. Of females that laid double and triple clutches, 97% of the first breeding attempts were started before April, the majority beginning in Feb- ruary. The only January laying represented the first of 3 clutches produced in 1976 by a trio of adults which fledged 5 young from two of the three attempts. Without subsequent nesting, their pro- ductivity for 1976 would have been just 2 young. Within 6 other territories high productivity was attained through extra breeding attempts. Adults in 4 territories produced 6 and those in 2 territories produced 7 young/yr. Most of the clutches laid from June to August were second and third at- tempts (Fig. 2). An early laying pair of Harris’ Hawks in Texas produced 7 young during 3 suc- cessful nesting attempts in one season (Brannon 1980). Triple clutches in 1 season have been re- ported 3 times previously for Harris’ Hawks (Mader 1977; Whaley 1979), but in all cases only 2 of the 3 attempts were successful. Harris’ Hawks begin egg laying earlier in Arizona than in Texas (see Brannon 1980), where clutches are started in March and April. Productivity Clutch Size. — Mean clutch size for Harris’ Hawk populations in this study was 3. 16 (N = 162) with 4 and 5 eggs/clutch regularly laid. Four eggs/clutch are not uncommon for Harris’ Hawks in Texas (Griffin 1976) and Arizona (this study) whereas along the Colorado River 2 eggs/clutch are the rule (Bendire 1892). Mader (1975b) obtained an aver- age of 2.96 eggs/clutch (N = 50) for Arizona, and Griffin (1976) calculated an average of 2.85 eggs/ clutch (N = 20) for western Texas. South Texas populations appear to have a smaller clutch size of 2.33 eggs/clutch (N = 24) (Brannon 1980). Some females in my study were exceptional egg layers; 3 females each laid 16 eggs during the 2-yr period. One of these laid 4 clutches of 4 eggs, all but one of which was in the same nest. Fledging Rate. — Productivity of Harris’ Hawks in Texas is lower than for Arizona and New Mexico populations. Mader (1975) found that 1.60 young fledged/all attempts (N — 50) and 2.35 fledged/suc- cessful attempts (N = 34), which compares well with my results for Arizona (Table 2). Griffin (1975) reported similar results for New Mexico, with 1.59 (N = 17) and 2.45 (N — 11) young/attempt and young/successful attempt, respectively. Griffin (1976) reported values of 0.83 (N = 18) and 1.87 (N = 8) young/attempt and young/successful attempt, respectively, for western Texas. Brannon (1980) reported similar results of 1.37 (N = 24) and 2.06 (N = 16) for young/attempt and young/successful attempt for west Texas populations. Lower pro- ductivity in Texas populations is likely due to grea- ter fluctuations in prey abundance (Griffin 1976) and less diversity in prey species. Texas populations are more nomadic, as year to year shifts in breeding distributions occur in accordance with rainfall pat- terns and prey abundance. Dispersal of Young and Nest Helping Harris’ Hawks in Arizona are nonmigratory. Early reports of large migratory flocks (250 to 500) (Chambers 1921 and 1924; Allan Phillips pers. comm.) are unreliable, and likely the result of mis- identification since Harris’ Hawks in Arizona have a long breeding season which affords no time for migration. Harris’ Hawks occasionally wander during winter, but generally remain near or within their nesting ranges year round. Families often re- mained together. On 2 occasions I saw winter groups of 7 and 8 hawks comprised of adults and color-marked juveniles. Wilder (1916) also noted large winter groups of 10 to 20 hawks along the Colorado River in December. Juveniles do not exhibit a strong tendency to disperse from the natal area. On 10 November 1977 two color-marked Harris’ Hawks were trapped less than 0.4 km from where they fledged ca. 173 days earlier (Rich Glinski pers. comm.). On 9 occasions within those 1977 nesting ranges where nesting occurred twice in 1 year, color-marked juveniles of earlier nesting attempts were allowed within the immediate vicinity of, and often on the nest containing eggs or young of subsequent nest- 12 Wayne H. Whaley Vol. 20, No. 1 ing attempts. Those juveniles often seemed as con- cerned about my presence as the adults. On 22 occasions in 1976 unmarked juveniles which were likely still in natal ranges were sighted near active nests. On 24 September 1977, while preparing to band 3 young from a second nesting (the first brood of 3 young fledged ca 120 d earlier), a color-marked female from the first brood brought prey to the nest. Also, on 3 occasions in 1979 Brannon (1980) observed prey deliveries by juveniles to the nests of their parents’ second breeding attempts. On 12 May 1977 two juveniles that had fledged 2 or 3 weeks earlier were observed incubating 3 eggs of their parents’ second clutch. Nest-helping by juveniles fledged earlier the same season has not been previously reported in raptors. The helper system of the Harris’ Hawk may prove to be similar to that of the Florida Scrub Jay ( Aphelacoma coerulescens) (Woolfenden 1975). Young of Galapagos Hawk (Buteo galapagoensis) also remain near the natal area for several months after fledg- ing and occasionally retain close ties with their pa- rents (de Vries 1973). On 1 occasion de Vries wit- nessed a 5-month-old juvenile still begging food from the male parent while the female was in- cubating eggs of a second nesting attempt. Harris’ Hawks do not appear to disperse great distances from their natal area over long periods of time. Based on 13 band returns, sightings and cap- ture-releases, juveniles fp* 5 months fledged) traveled an average distance of 1 5 miles (range < 5 miles for a male during 3 years and 30 miles for a female during 6 months). On 20 May 1977 I banded a male nestling near Florence Junction. The same male was trapped in its nesting territory nearly 8 yrs later on 12 January 1985 (Jim Dawson pers. comm.) 35 miles south of where it had fledged. Past Populations of the Colorado River The Harris’ Hawk was a common resident along the Colorado River for at least 134 years. The population did not range far from the riparian community (U.S.D.I. Fish and Wildlife Service 1950); thus, events that may have affected this habitat may also have impacted the Harris’ Hawk populations in that region. Extensive flooding along the Colorado River in the early 1900’s evoked construction of several dams in the 1930’s and 1940’s. With flooding controlled, agricultural ac- tivities along the river increased. By the 1930’s and 1940’s, salt cedar (Tamarix pentandra) spread over large areas of the river and began to compete with the cottonwood community which was rejuvenating from past floods. Dredging operations began along the river during the 1940’s and may have been a source of disturbance. Previously inaccessible areas of the swamp were now accessible by motor pow- ered craft so subsequently there was increased use of the river for recreational activities during the 1950’s (Gale Monson pers. comm.). At Havasu Na- tional Wildlife Refuge, where Harris’ Hawks often nested low over the water in drowned out mesquite trees, increased recreational activities may have been a major factor during the nesting periods. Fishermen were seen destroying nests, probably under the mistaken impression of destroying Cor- morant (Phalacrocorax sp.) nests (Gale Monson pers. comm.). Harris’ Hawks were occasionally harassed by duck hunters who often referred to them as, “the big black hawk that catches ducks” (Miller 1925). Monson reported that Harris’ Hawk numbers were sometimes slightly depleted by waterfowl hunters in the open areas above Topock (U.S.D.I. Fish and Wildlife Service 1949). A combination of the above factors likely had an impact on Harris’ Hawk populations along the Col- orado River. Drastic habitat alteration and in- creased recreational activities perhaps yielded the greatest impact. The Present Population Qualitatively, the most obvious characteristic of Harris’ Hawk habitat in Arizona is the presence of healthy stands of paloverde-saguaro desert scrub of the Arizona Upland subdivision. The present dis- tribution of Harris’ Hawks in Arizona strictly fol- lows the Arizona Upland subdivision of the Sono- ran Desert with no nests occurring outside this habitat. Thus, this hawk habituates the most struc- turally complex habitats within the Sonoran Desert, where prey density and diversity is the greatest. The nesting distribution of the Harris’ Hawk in Arizona has shrunk since the early 1900’s due in part to the extirpation of populations along the Colorado, Gila, Santa Cruz, and San Pedro Rivers. Along the Gila and Santa Cruz habitat has been severely altered (Dawson 1921; Rea 1977). In order to determine susceptability of the present popula- Spring 1986 Harris’ Hawk in Arizona 13 Table 5. Land ownership status for 306 nest sites and 396 nesting attempts for Harris’ Hawk in Arizona in 1976 and 1977. Land Ownership Number of Active Nests Percentage of Total Number of Nesting Attempts Percentage of Total State Trust Land 140 45.8 184 46.5 Patented Land 73 23.9 84 21.2 BLM Land 40 13.0 57 14.4 Forest Service 18 5.9 24 6.0 Indian Reservation 18 5.9 23 5.8 Patented or State 7 2.3 10 2.5 BLM or State 5 1.6 8 2.0 Regional Park 3 .9 3 .8 U.S. Park 2 .7 3 .8 Totals 306 100.0 396 100.0 tion to possible habitat encroachment, I deter- mined land ownership status for 306 active nest sites (Table 5). A large proportion of nests are on patented land, and therefore are more vulnerable to the effects of man’s present and future activities. Since the Harris’ Hawk appears to be habitat re- stricted in its nesting, the retention of both state and federal lands with healthy stands of paloverde- saguaro will be crucial for its future welfare in Arizona. Urban sprawl near several Harris’ Hawk populations may result in continued loss of nesting range in Arizona. Harris’ Hawk numbers near Tuc- son and Cave Creek have declined recently as a result of urban development. The Cave Creek population is currently threatened due to urban sprawl and may soon be decimated (Jim Dawson pers. comm.). Other habitats that are threatened by urban development are in areas near Rio Verde, Apache Junction, and Florence. Acknowledgments Research was conducted as part of the M.S. degree require- ments at the Arizona Cooperative Wildlife Research Unit (AC- WRU) of the University of Arizona. The ACWRU is sponsored jointly by the University of Arizona, the U.S. Fish and Wildlife Service, the Arizona Game and Fish Department, and the Wildlife Management Institute. I thank Lyle K. Sowls, leader of the Arizona Coopertive Wildlife Research Unit, for his guidance throughout this project. I also thank other members of my committee, David H. Ellis and Stephen M. Russell, for their guidance. I express appreciation to my field assistants, Jerry Roberts, Greg Depner and Kevin Coates, for their determination when things got rough. I thank the Havasu National Wildlife Refuge and Imperial National Wildlife Refuge for access to their files. Several Arizona falconers and private individuals provided field assistance, for which I express appreciation. Eleanor Radke provided some color-marking mate- rial and banding assistance. I am especially indebted to my wife, Connie, for her patience during my extended absences while gathering field data. Literature Cited Anderson, A.H. and A. Anderson. 1960. Life histories of the Cactus Wren. pt. 3: the nesting cycle. Condor 62:351-369. Arnold, L.W. 1954. The Golden Eagle and its economic status. U.S. Fish and Wildl. Serv., Circ. No. 27. 35 pp. Bancroft, G. 1920. The Harris’ Hawk as a breeder in California. Condor 22:156. Bendire, C. 1892. Life histories of North American birds. 2 vols. U.S. Natl. Mus. Spec. Bull. No. 1. 446 pp. Bent, A.C. 1932. Life histories of North American Gal- linaceous birds. U.S. Gov. Print. Office. 490 pp. Berger, D.D., C.R. Sindelar and K.E. Gam- ble. 1969. The status of breeding Peregrines in the eastern United States. Pages 165-173 in J. J. Hickey, ed. Peregrine Falcon populations: their biology and decline. Univ. of Wis. Press, Madison. 596 pp. 14 Wayne H. Whaley Vol. 20, No. 1 Brannon, J.D. 1980. The reproductive ecology of a Texas Harris’ Hawk (Parabuteo unicinctus harrisi) population. M.S. Thesis, Univ. of Texas, Austin. 56 pp. Brown, L. and D. Amadon. 1968. Eagles, hawks and falcons of the world. McGraw-Hill Book Co. New York, N.Y. 945 pp. Brown, R.L. 1967. The extent ofbreeding by immature Mourning Doves ( Zenaidura macroura marginella ) in southern Arizona. M.S. Thesis. Univ. of Ariz., Tucson. 56 pp. Chambers, W.L. 1921. A flight of Harris’ Hawks. Condor 23:65. 1924. Another flight of Harris’ Hawks. Condor 26:75. Cottam, C., D.A. Munro, R.H. Pouch, H.A. Hochbaum, R.A. McCabe, I.N. Gabrielson. 1961. Report to the American Ornithologist’s Union by the committee on bird protection. Auk 79:463-478. Craig, Y.H. and L.R. Powers. 1976. Raptor mortality due to drowning in a livestock watering tank. Condor 78:412. Dawson, W.L. 1921. The season of 1917./. Mus. Comp. Ool. Santa Barbara, California. 2 (1/2): 27-36. De Vries, T. 1973. The Galapagos Hawk: an ecogeo- graphical study with special reference to its sys- tematic position. M.S. Thesis. Free Univ. of Amster- dam. 108 pp. Ellis, D.H. and W.H. Whaley. 1979. Two winter - breeding records for the Harris’ Hawk. Auk 96:413. Enderson, J.H. 1964. A study of the Prairie Falcon in the Central Rocky Mountain Region. Auk 81:332-352. Griffin, C.R. 1976. A preliminary comparison of Texas and Arizona Harris’ Hawks (Parabuteo unicinctus) populations. Raptor Res. 10:50-54. Griffin, R. 1975. Reproduction data on sixteen Harris’ Hawk nests in southeastern New Mexico. (Compiled during summer of 1975). MS available from: En- dangered Species Program, N.M. Dept, of Game and Fish, Santa Fe. Hector, D.P. 1975. The habitat, diet and foraging be- havior of the Apolmado Falcon (Falco femoralis). M.S. Thesis, Oklahoma State Univ., Stillwater. 189 pp. Henny, C.J. and H.M. Wight. 1972. Population ecology and environmental pollution: Red-tailed and Cooper’s hawks. Pages 229-250 in Population ecology of migratory birds: a symposium. U.S. Dept, of Interior Wildl. Res. Report No. 2. 278 pp. Hensley, M. M. 1959. Notes on the nesting of selected species of birds of the Sonoran Desert. Wilson Bull. 71:86-92. Hungerford, C.R., C.H. Lowe and R.L. Mad- sen. 1973. Population studies of the desert cottontail (Sylvilagus andubonii) and black-tailed jackrabbit (Lepus californicus) in the Sonoran Desert. Desert Biome, U.S. Intern. Biol. Program, Res. Mem. 73-20. 15 pp. Immelmann, K. 1971. Ecological aspects of periodic re- production. Pages 341-389 in D.S. Farner and J.R. King, eds. Avian Biology. Vol. 1. Academic Press. New York, N.Y. 586 pp. Kennerly, C.B.R. 1859. Report of birds collected on the route. Pages 19-35. Zoological Report No. 3 in Reports of exploration and surveys for a railroad route from the Mississippi River to the Pacific Ocean, Part 6. Route near the thirty-fifth parallel, explored by Lt. A.W. Whipple, War Dept. Wash. D.C., vol. 10. Lack, D. 1954. The natural regulation of animal num- bers. Oxford Univ. Press (Clarendon) London and New York. 343 pp. 1968. Ecological adaptations for breeding in birds. Methuen and Co. LTD, London. 409 pp. Lowe, C.H. 1976. The vertebrates of Arizona. Univ. of Ariz. Press, Tucson. 270 pp. Lowe, C.H. and D.E. Brown. 1973. The natural vege- tation of Arizona. A.R.I.S. Coop. Pub. No. 2. 53 pp. Mader, W.J. 1975a. Extra adults at Harris’ Hawk nests. Condor 77:482-485. 1975b. Biology of the Harris’ Hawk in southern Arizona. Living Bird 14:59-85. 1977. Harris’ Hawks lay three clutches of eggs in one year. Auk. 94:370-371. . 1982. Ecology and breeding habits of the Savanna Hawk in the Llanos of Venezuela. Condor 84:261-271. Miller, L. 1925. Food of the Harris’ Hawk. Condor 27:71-72. 1930. Further notes on the Harris’ Hawk. Condor 32:210-211. Moreau, R.E. 1950. The breeding seasons of African birds. 1. Land birds. Ibis 92:223-267. Neal, B.J. 1965. Reproductive habits of round- tailed and Harris’ antelope ground squirrels. J. Mammal. 46:200-206. Peterson, R.T. 1969. Population trends in Ospreys in the northwestern United States. Pages 333-337 in J.J. Hickey, ed. Peregrine Falcon populations: their biol- ogy and decline. Univ. of Wis. Press, Madison. 296 pp. Phillips, A., J. Marshall and G. Monson. 1964. The birds of Arizona. Univ. of Ariz. Press, Tucson. 212 pp. Radke, E.L. and J. Klimosewski. 1977. Late fledging date for Harris’ Hawk. Wilson Bull. 89:469-470. Rea, A.M. 1977. Historical changes in the avifauna of the Gila River Indian Reservation, central Arizona. Ph.D. Thesis. Univ. of Ariz., Tucson. 346 pp. Rowley, J.S. 1936. Notes on some nests found in eastern Riverside County, California. Condor 38:219. Sellers, W.D. 1960. The climate of Arizona in Arizona Climate. Univ. of Ariz. Press, Tucson. 616 pp. Smith, E.L. 1971. The effects of heat and aridity on reproductive success of the Curve-billed Thrasher. Ph.D. Thesis. Univ. of Ariz., Tucson. 55 pp. Spring 1986 Harris’ Hawk in Arizona 15 Sprunt, A. 1969. Population trends of the Bald Eagle in North America. Pages 347-351 in J.J. Hickey, ed. Peregrine Falcon populations: their biology and de- cline. Univ. of Wis. Press, Madison. 596 pp. U.S.D.I. Fish and Wildlife Service. 1949. September-December Quarterly Narrative Report, Havasu National Wildlife Refuge. Located at: Refuge Hdqtrs., Topock, Arizona. U.S.D.I. Fish and Wildlife Service, 1950. January- April Quarterly Narrative Report, Imperial National Wildlife Refuge. Located at: Refuge Hdqtrs., Mar- tinez Lake, Arizona. U.S.D.I. Fish and Wildlife Service. 1952. January- April Quarterly Narrative Report, Havasu National Wildlife Refuge. Located at: Refuge Hdqtrs., To- pock, Arizona. Vorhies, C.T. and W.P. Taylor. 1940. Life history and ecology of the white-throated wood rat (Neotoma al- bigula albigula ) Hartley, in relation to grazing in Arizona. Univ. of Ariz. Agr. Exp. Sta. Tech. Bull. No. 86:455-529. Walter, H. 1979. Eleonora’s falcon. Univ. Chicago Press. Whaley, W.H. 1979. The ecology and status of the Har- ris’ Hawk ( Parabuteounicinctus ) in Arizona. M.S. Thesis Univ. of Ariz., Tucson. 119 pp. Wilder, H.E. 1916. Some distributional notes on California birds. Condor 18:127-128. Wiley, L. 1916. Bird notes from Palo Verde, Imperial County, California. Condor 18:230-231. 1917. Nesting of the Harris’ Hawk in southeastern California. Condor 19:142. Woolfenden, G.E. 1975. Florida Scrub Jay helpers at the nest. Auk 92:1-15. Arizona Cooperative Wildlife Research Unit, 214 Biological Sciences Bldg., University of Arizona, Tucson, A Z 85721. Present address: Dept, of Zoology, 574 WIDE, Brigham Young University, Provo, UT 84602. Received 25 April 1985: Accepted 15 August 1985 CRYOPRESERVATION OF PEREGRINE FALCON SEMEN AND POST-THAW DIALYSIS TO REMOVE GLYCEROL John E. Parks, Willard R. Heck and Victor Hardaswick ABSTRACT — Peregrine Falcon (Falco peregrinus) semen was found to have a mean ejaculate volume, sperm concentra- tion and initial sperm motility of 95 pi, 47 X 10 8 sperm/ml and 70%, respectively. When frozen in a medium containing 0.3, 0.9 or 1 .48 M glycerol, post-thaw sperm motility was 29, 47 and 54%. Because of the contraceptive effect of glycerol, a dialysis procedure was developed to remove the cryoprotectant from post-thaw semen. Percent motility during post- thaw incubation was greater in samples from which glycerol had been removed by dialysis than in controls (P < 0.05) . Of 6 eggs from a single 9 American Kestrel (Falco sparverius) obtained after insemination with frozen-thawed, dialyzed peregrine semen, 2 were fertile and survived to pip. One interspecific hybrid was hatched and raised successfully. Captive breeding has been used effectively for the conservation of birds of prey for many years. Artificial insemination has been a useful technique in captive breeding programs since the early 1970’s (Weaver 1983). However, there are many situations in which the efficiency of a breeding program may be reduced because semen is not available at the time or place where it is needed. The ability to freeze raptor semen would facilitate captive breeding under these and other circumstances. The ability to freeze semen also would permit banking germ plasm from rare or endangered species. At present, a limiting factor in the use of frozen semen in domestic avian species is the inhibition of fertility by the cryoprotectant in the medium (Brown and Graham 1971; Lake and Stewart 1978; Sexton 1979; Lake et ab 1980; Lake et al. 1981; Graham et al. 1982). In order to overcome this problem, cryoprotectants such as glycerol or di- methylsulfoxide (DMSO) must be removed post- thaw (Lake and Stewart 1978; Lake et al. 1981; Graham et al. 1982) or new, less problematic cryo- protectants must be identified. The purpose of this study was to develop proce- dures for processing and freezing raptor semen, using the Peregrine Falcon ( Falco peregrinus) as a semen source; and to develop a procedure for re- moving the cryoprotectant glycerol from thawed semen without further loss of sperm viability. Materials and Methods Semen Collection and Handling. — Semen from 3 adult male Peregrine Falcons was collected up to 2 times/d (Boyd and Schwartz 1983) over a period of approximately 2 m. On each occasion, ejaculates from 1 - 3 J d 1 were pooled. Semen was diluted 1:3 (v/v) at 20°C in 12 X 55 mm vials containing Lake’s freezing diluent (Lake and Stewart 1978), placed directly into an ice water bath (0 - 2°C) and transported to the laboratory (see Table 1). All remaining steps up to freezing and during the thawing process were carried out at 4°C. Semen Evaluation. — Sperm motility was assessed microscopi- cally by estimating the percentage of sperm moving progressively forward (percent motility). Unfixed smears prepared from sam- ples diluted in freezing diluent were placed on a slide warmer at 37°C for 30 sec immediately prior to evaluation. Percent motility was estimated in several microscopic fields to the nearest 5% using a phase contrast microscope at a total magnification of 400x. Sperm concentration of the semen diluted 1:3 was determined with a hemacytometer after an additional dilution of 1 : 1 (v/v) in fixative (4% glutaraldehyde). Duplicate counts of each prepara- tion were averaged for use in calculating sperm concentration of the original ejaculate. Semen Freezing and Storage. — Aliquots of approximately 50 pi of cooled, diluted semen were placed into 0.25 ml French straws (IMV) for freezing. Diluent was aspirated into the straw ahead of the semen with a small air space separating the 2 liquids. Filling the straws in this way served to seal the polyvinylchloride (PVC) plug at the end of the straw without loss of semen and also prevented straws from floating when placed in liquid nitrogen. Straws were then sealed and loaded into a Planer R204 freezer (2°C). Semen was frozen in nitrogen vapor at 6°C/min to -180°C and then plunged into liquid nitrogen (Brock et al. 1984). One straw from each freezing procedure was then thawed in water (4°C) for evalu- ation of percent motility. Straws were wiped dry and the semen- containing portion was emptied into precooled tubes (6 X 50 mm). Remaining straws were stored for 1 - 2 months prior to thawing. Initially, semen was frozen in diluent containing 0.3, 0.9 or 1.48 M glycerol. Percent motility was estimated on aliquots prior to freezing and immediately post-thaw. Based on these initial trials using different levels of glycerol in the freezing medium (Table 2), 1.48 M glycerol was selected for routine use in subsequent ex- periments. Method for the Dialysis of Diluent and Diluted Semen. — Dialysis to remove glycerol from the freezing medium and thawed semen was carried out using semi-micro dialysis tubing (2.55 mm diameter, molecular weight cutoff of 12,000 - 14,000, Spectra/ Por). Tubing was washed thoroughly in twice distilled water and stored wet at 4°C prior to use. All subsequent steps in the dialysis procedures were carried out at 4°C. Tubing was tied and cut into lengths of 8-10 mm from the tied end, filled with the freezing diluent (1.48 M glycerol) and equilibrated for 10-20 min in the same diluent. Diluent was then completely removed from the tubing and 50 pi aliquots of fresh freezing diluent or of thawed semen were pipetted into the tubing using a fire-polished 200 pi capillary pipet and a capillary suction apparatus (Clay-Adams). The tubing was then closed with a Spectra/Por closure and dialyzed with stirring against 500 volumes of Lake’s thawing medium (Lake and Stewart 1978). 16 Raptor Research 20 ( 1 ): 15-20 Spring 1986 Freezing Falgon Semen 17 Table 1. Characteristics and pre-freezing treatment of Peregrine Falcon semen. Volume for Collection Period 3 Oi) Initial Sperm Interval Between Motility Concentration Collection and CoolincT (%) (x 10 6 /ml) (min) Individual Pooled Ejaculates Ejaculates X 95.4 89.0 S.D. 51.7 38.5 Range 27-208 50-127 n 23 3 All b Ejaculates 94.6 70.1 49.7 6.9 27-208 60-85 26 25 47.4 18.4 16.1 6.8 29-81 12-45 15 24 a Semen was collected between approximately 0830 - 0930 H and 1630 - 1730 H. b Volumes for pooled ejaculates included 2 samples (partial or whole ejaculates) from 2 individuals and 1 sample from 3 individuals. Volumes are included for individual birds from 3 collection periods in which ejaculates were pooled prior to freezing. c Intervals were timed from collection of the last ejaculate for pooled samples. Freezing was begun within approximately 15 to 30 min after initial dilution and cooling. Estimation of the Efficiency of Glycerol Removal by Dialysis. — Removal of glycerol from the freezing medium was measured by supplementing the medium with [2- 3 H]-glycerol (New England Nuclear, 200 at a level of 2 X 10 5 dpm/50 fi\. Appear- ance of radioactive glycerol in the dialysate relative to the initial amount placed in the dialysis tubing was used to calculate the rate and extent of glycerol removal. Preliminary trials (n = 2) indicated that > 99% of the glycerol in the freezing medium was removed after 30 min of dialysis. To establish the time-course relationship of glycerol removal, dialysis of freezing medium containing 1 .48 M glycerol was carried out for 2 h against thawing medium containing no glycerol (n = 5). Input samples; 0.2 ml aliquots of the dialysate taken at 0, 0.25, 0.5, 1.0, 2.0, 5.0, 15, 30, 60 and 120 min of dialysis; and residual material in the dialysis tubing were analyzed for 3 H content (glycerol) by liquid scintillation spectrometry. Evaluation of the Effect of Dialysis on Sperm Motility. — Because glycerol was removed so rapidly by the dialysis proce- dure, damage to sperm due to osmotic effects was considered a potential problem. Therefore, dialysis conditions were established to remove the glycerol more gradually. Material to be dialyzed was transferred at 15 min intervals to thawing solutions containing glycerol decreasing in equimolar increments (1.1, 0.74 and 0.37M and no glycerol). It was assumed that the rate of glycerol equilib- ration (and thus removal of glycerol from the thawed semen) in these steps was approximately equivalent to that observed in the one-step procedure, and the extent of total glycerol removal was calculated on this basis. An experiment was designed to assess the effect of the step-wise dialysis procedure on falcon sperm motility during post-thaw, post-dialysis incubation. A split-ejaculate technique was used in which all treatments within each experi- ment were imposed on aliquots of the same ejaculate. At the Table 2. The effect of glycerol level on pre-freeze and post-thaw motility of falcon sperm. Values are percentages. Percent Motility Pre- Freeze Post-Thaw Glycerol Level 0.3 M 0.9 M 1.48 M 0.3 M 0.9 M 1.48 M X 75 66 70 29 47 54 S.D. 4.1 4.9 7.4 11.0 2.6 5.8 Range 70-80 60-75 60-85 20-45 45-50 45-65 n 4 6 15 4 6 15 18 Parks, et al. Vol. 20, No. 1 Table 3. Post-thaw motility of falcon sperm after dialysis to remove glycerol (n = 4). a Values are percentages. Dialysate Hours of Incubation After Dialysis 0 0.5 1.0 1.5 4.0 X* 1.48 M Glycerol 41 26 24 25 13 26 Four-step Procedure (1.1 M to Glycerol-Free) 43 33 35 31 20 32 a Post-thaw, pre-dialysis motility for these samples was 55±7%. * P < 0.05 completion of dialysis, semen was emptied into tubes (6 X 50 mm) at 4°C. Percent sperm motility was determined immediately and after 30, 60, 90 and 240 min of post-dialysis incubation (38°). Artificial Insemination. — Female peregrines were not availa- ble for testing the fertility of post-thaw, dialyzed semen. One unpaired 9 American Kestrel (Falco sparverius) was inseminated with approximately 40 to 50 /ul of thawed semen, dialyzed by the step-wise procedure. Six single inseminations were made within 4 h after oviposition and the first egg laid after each insemination was artifkally incubated. Thawed samples were maintained at 4°C until the oviduct was everted for insemination. The semen was then transferred to an insemination syringe and deposited into the oviduct (Weaver 1983). Total time between thawing and insemi- nation including dialysis was approximately 90 min. Statistical Analysis. — Means and standard deviations were calculated for semen characteristics and for motility estimates on semen diluted and frozen in different levels of glycerol. The effects of dialysis on post-thaw motility were analyzed by analysis of variance after arcsin transformation of the percentage data. Results Semen characteristics and information related to initial handling of semen are presented in Table 1. Semen was not scored on appearance, but only a low to moderate level of contamination by extrane- ous cell types and other debris was observed in the ejaculates used for freezing. Initial experiments in which glycerol was the only variable tested (Table 2) indicated that 1.48 M glycerol provided greater protection during freezing than 0.9 M or 0.3 M glycerol based on post-thaw sperm motility. How- ever, glycerol levels were tested on separate semen collections so comparisons were not made on a statistical basis. The response to freezing, when us- ing 0.3 M glycerol, was consistently poor; but the difference in post-thaw motility between 0.9 M and 1.48 M diluents was small, especially when ex- pressed as the difference between pre-freeze and post-thaw motility (13 vs 16%). Rate and extent of glycerol removal during a single step dialysis of samples are presented graphi- cally in Fig. 1. Approximately 90% of the glycerol was removed by 15 min of dialysis, and after 30 min glycerol had been completely removed. Based on this rate of equilibration, a sample frozen in 1.48 M glycerol and dialyzed by the step-wise procedure was considered to contain less than 30 mM glycerol post-dialysis. containing 1 .48 M glycerol and the appearance of radioactivity in the dialysate was measured. Bars represent standard deviation for each time point (n = 5). Spring 1986 Freezing Falcon Semen 19 A comparison of post-thaw motility for sperm dialyzed by the step-wise procedure or directly against the freezing medium (1.48 M glycerol) is presented in Table 3. It is apparent that sperm survived the dialysis procedure with fair to good motility. During post-dialysis incubation motility declined with both treatments (P< 0.01). Although no significant time X treatment interaction was found, percent motility remained sufficiently higher after glycerol removal to demonstrate an advantage over the control treatment (P < 0.05). A consistent difference in the motility pattern was also observed between the dialysis treatments. Sperm from which glycerol had been removed exhibited a greater velocity, and motility was more progressive with less amplitude in the flagellar motion than with sperm in glycerol. This difference was not qualified but was readily apparent to other observers. Sperm in suspensions after glycerol removal also seemed more resistant to dessication on slides prepared for microscopic examination than those remaining in high glycerol medium, based on maintenance of motility. Of 6 American Kestrel eggs potentially fertilized by frozen, thawed and dialyzed Peregrine Falcon semen, 2 eggs were fertile and developed to pip. One of these young died at pip while the other interspecific hybrid hatched and was raised success- fully. Discussion At present, there is very little detailed informa- tion on semen characteristics of raptorial species. This study provides such information on ejaculate volume, sperm concentration and percent motile sperm for semen from the Peregrine Falcon. Val- ues for these characteristics have also been reported for the American Kestrel (Bird and Lague 1977). A comparison of the semen characteristics for these 2 species indicates an 8-fold greater ejaculate volume for the peregrine which approximates the differ- ence in body weight between it and the kestrel. Sperm concentration is over 30% greater for the peregrine. The much greater total sperm/ejaculate for the peregrine may be a necessary adaptation for ensuring adequate sperm numbers at the site of fertilization in this larger species. Percent motile sperm appears to be slightly higher for the pere- grine than the kestrel, but this may be due to dif- ferences in conditions under which sperm were examined. This type of information on semen parameters is necessary for making the most effec- tive use of artificial insemination in the species of interest, and for effective processing of semen for cryopreservation. A knowledge of semen charac- teristics may also serve as a basis for comparison when examining the effects of environment or en- vironmental contamination on reproduction (Bird and Lague 1977). In the present study, we found that peregrine semen freezes well in Lake’s diluent. It appears that a broad range in glycerol level might be acceptable, but more definitive work is required to establish the optimum glycerol concentration. Brock et al. ( 1 984) reported excellent post-thaw motility for kestrel semen frozen in Lake’s diluent, but fertility of the semen was < 5%. The requirement to remove glycerol and other cryoprotectants from post-thaw semen in order to obtain acceptable fertility has been established in domestic avian species (Brown and Graham 1971; Lake and Stewart op. cit . ; Lake et al. 1980; Lake et al. 1981; Graham et al. op. cit.). This also may be true for falcon semen. Removal of glycerol from post-thaw cock semen by dilution and centrifugation greatly improves fertility (Lake and Stewart, op. cit.), but this approach is not practical when working with microliter quantities of falcon semen. Graham et al. (op. cit.) reported that the level of cryoprotectant (DMSO and ethylene glycol used in combination) necessary to maintain vigorous post-thaw motility of turkey semen depressed fer- tility. Use of dialysis to remove the cryoprotectant significantly improved fertility; although dialysis time, dialysate composition and pH, and se- men-to-dialysate ratio all influenced the level of fertility observed. Dialysis can be adapted for use with the small semen volumes associated with rap- torial species and is a milder approach for removing cryoprotectant. Lake et al. (1980) demonstrated that in order to minimize its inhibitory effect on fertility in the Domestic Chicken (Gallus spp.), glycerol must be reduced to a level below 1% (0.11 M) in diluted semen. It is apparent from the present study that under the appropriate conditions glycerol can be reduced below the level of 1% within 30 min by dialysis. By controlling the sample volume/dialysate ratio or by adjusting the level of glycerol in the dialysate, rate of glycerol removal can be regulated to minimize the post-thaw to insemination interval necessary to remove cryoprotectant while main- taining optimal sperm viability. In this study, man- 20 Parks et al. Vol. 20, No. 1 ipulations required to transfer and dialyze the mi- croliter volumes of frozen semen were carried out with only a small reduction in motility. Maintenance of post-thaw motility was slightly, but significantly improved with glycerol removal. The relevance of greater velocity in these samples is not readily appa- rent. However, differences observed in motility of sperm after glycerol removal may translate into enhanced sperm survival in the more favorable en- vironment of the female reproductive tract. The results of this study are based on a limited number of observations, leaving many questions regarding cryopreservation of falcon semen un- answered. However, several important points can be drawn from these results. Peregrine semen can be frozen using glycerol as a cryoprotectant with good post-thaw sperm motility, and the glycerol can be rapidly removed from post-thaw semen by dialysis without substantial loss of sperm motility. The techniques used in these procedures are sim- ple, relatively inexpensive, and can be adapted for practical application. Finally, the development of 2 kestrel eggs in a clutch of 6 suggests that post-thaw dialysis is potentially useful for successful breeding with frozen falcon semen. Use of homologous species for insemination may provide a more useful measure of fertility. Refinement of these proce- dures and use of additional females to test fertility will help to establish whether post-thaw glycerol removal will make the use of frozen semen a practi- cal approach to captive breeding of falcons and other birds of prey. Acknowledgments This project was supported in part by The Peregrine Fund, Inc., and the American Wildlife Research Foundation, Inc. Literature Cited Bird, D.M. and P.C. Lague. 1977. Semen production in the American Kestrel. Canadian J. Zool. 55:1351-1358. Boyd, L.L. and C.H. Schwartz. 1983. Training im- printed semen donors. In: Falcon Propagation - A Manual on Captive Breeding (J.D. Weaver and T.J. Cade, Eds.) pp. 24-31, The Peregrine Fund, Inc., Ithaca, New York. Brock, K., D.M. Bird and G. Ansah. 1983. Cryogenic preservation of spermatozoa of the American Kestrel Falco sparverius. Internat. Zoo Yearbook 12:67-71. Brown, K.I. and E.F. Graham. 1971. Effect of some cryophylactic agents on turkey spermatozoa. Poultry Sci. 50:832-835. Graham, E.F., D.S. Nelson and M.K.L. Schmehl. 1982. Development of extender and techniques for frozen turkey semen. 1 . Development. Poultry Sci. 61:550-557. Lake, P.E., R.B. Buckland and O. Ravie. 1980. Effect of glycerol on the viability of fowl spermatozoa - impli- cations for its use in freezing semen. Cryoletters 1:299- -304. Lake, P.E., O. Ravie and J. McAdam. 1981. Preserva- tion of fowl semen in liquid nitrogen: application to breeding programs. Brit. Poult. Sci. 22:71-77. Lake, P.E. and J.M. Stewart. 1978. Preservation of fowl semen in liquid nitrogen - an improved method. Brit. Poult. Sci. 19:187-194. Sexton, T.J. 1979. Cytotoxicity of DMSO as related to components of a turkey semen extender. Poultry Sci. 58:1024-1030. Weaver, J.D. 1983. Artificial insemination. In: Falcon Propagation - A Manual on Captive Breeding, (J.D. Weaver and T.J. Cade, Eds.) pp. 19-23, The Peregrine Fund, Inc., Ithaca, New York. Department of Animal Science, 201 Morrison Hall, Cornell Univ., Ithaca, NY 14853. Address of second and third au- thors: The Peregrine Fund, Inc., Sapsucker Woods Road, Ithaca, NY 14853. Received 25 March 1985; Accepted 15 July 1985 CHARACTERISTICS OF CLIFFS AND NEST SITES USED BY BREEDING PRAIRIE FALCONS D.E. Runde and S.H. Anderson Abstract - Data from over 400 Prairie Falcon ( Falco mexicanus) eyries in 8 states show a consistent pattern of eyrie placement relative to height and aspect of nest cliffs. Mean eyrie height averaged 63% of mean cliff height. Mean exposure of eyries and nest cliffs tended to be southerly with no significant difference between eyrie and cliff exposures. Potholes were the most frequently used nest sites. Patterns of occupancy and nest success in Wyoming were statistically independent of physical habitat variables measured. However, eyries on tall exposed buttes consistently had high failure rates that appeared weather related. Attributes of Wyoming eyries are presented and suggestions made for creating new eyries for Prairie Falcons as a management tool. We summarized and compared data on nest cliffs and eyries used by the Prairie Falcon (Falco mexicanus) in the western United States. Included are data available in the literature and original data from eyries active between 1982 and 1984 in Wyoming. Our objective was to document the vari- ation in nest cliffs and eyries used by these falcons and to identify consistent patterns of use. In our study, we examined the association between physi- cal nesting habitat and patterns of eyrie occupancy and nest success. We discuss attributes of Prairie Falcon nest cliffs and eyries, and suggest guidelines for creating new falcon eyries. Study Area and Methods Data from 8 states were considered: Washington {Decker 1931), Wyoming-Colorado (Enderson 1964), Montana (Leedy 1972), Utah (Porter and White 1973), New Mexico (Platt 1974), Oregon (Denton 1975), Idaho (Ogden 1973; Ogden and Hor- nocker 1977), and Colorado (Williams 1981). We collected data from eyries and associated cliffs in southern Wyoming known to have been active between 1982 and 1984. We define eyrie as the area in and immediately around the nest scrape. Nest cliff was the rock formation above and below the eyrie. A single nesting territory may contain several separate nest cliffs and eyries that are occupied and defended in different years. We climbed to eyries by ladder or rope; nest cliff measurements were taken from below. Variables measured at eyries centered on the nest scrape, if recognizable, or on the approximate center of the eyrie. These included roof (or overhang) height, width, length, and percent coverage of nest scrape by roof or overhang. We measured area of both entrance and floor with a transparent 25 cm 2 grid. Slope of both floor and ceiling were measured with a clinometer. Eyrie exposure was measured by placing a compass in the nest scrape and taking readings along a horizontal plane to both outside edges. From these azimuths (corrected for declination) we calcu- lated mean and extent of horizontal exposure. Finally, eyries were classed as either potholes, horizontal ledges or shelves, vertical cracks or crevices, and stick nests. Nest cliff measurements focused on that portion of the cliff directly above and below the eyrie. Variables included cliff and eyrie height, inclination of the toe slope below cliff (determined by clinometer), and aspect perpendicular to the cliff. Verticality of each cliff was estimated with a large protractor and was either vertical (85-95°), overhung (<80°) or underhung ( >100°). Cliffs were classed as: canyon walls, buttes or mountains, rimrock or ridges, and isolated rock outcrops. Standard statistics are presented for most variables. Statistics from other studies, if not reported, were calculated from fre- quency tables or raw data. We analyzed our data using one-way ANOVA and Chi-square (\ 2 ) to examine relationships between physical variables and histories of nest success and occupancy within and among years. In among-year analyses, we only in- cluded sites with complete 3-yr histories. Data suitable for circular statistical analysis (Batschelet 1981; Zar 1984) were available for Montana, Utah, New Mexico, Idaho, Colorado and Wyoming, Grouped data were analyzed as suggested by Batschelet (1981); 954 confidence intervals were estimated by interpolation (Batschelet (1981: Fig. 5.2 1). We used used Rayleigh’s test to determine statistical significance of mean exposures to infer nonrandom orientations (Zar 1984). Concent- ration (r), or length of the mean vector as calculated by circular methods, ranges from 0 to 1 and is affected by variation in circular data, sample size, and grouping. Values of r near 1 indicate data points closely concentrated about the mean angle. Results Eyries. — Dimensions of eyries from the present study were summarized (Table 1), and comparable to Williams (1981) and Porter and White (1973). Williams (1981) reported eyrie lengths and widths which averaged 78.9 cm (SE = 7.63) and 81.1 cm (SE = 9.52), respectively. Both means were signific- antly less than measurements from Wyoming (t- Test; P<0.01). Roof heights of Colorado eyries (Williams 1981) averaged 83.2 cm (SE = 12.04), and were significantly greater (t-Test; P < 0.01) than we found. The Colorado data yielded an average floor area of 6821 cm 2 (SE = 1198.22). Area within 8 eyries averaged 15,000 cm 2 in Utah (Porter and White 1973). Our estimate of area lies between these, but statistical comparisons could not be made. Floors of most (80%) Wyoming eyries sloped downwards towards the front at 5-10°. Conversely, most (84%) eyrie roofs or overhangs sloped to the 21 Raptor Research 20(l):21-28 22 Runde and Anderson Vol. 20, No. 1 Table 1. Attributes of Prairie Falcon eyries in southern Wyoming, 1982 to 1984. Variable Mean SE Range N Height (cm) 47.9 3.62 11-193 68 Width (cm) 91.2 7.45 18.313 70 Length (cm) 135.4 8.82 43-400+ 71 Floor (cm 2 ) 9325 769.83 1600-29275 70 Entrance (cm 2 ) 5375 975.12 875-53500 56 Floor Slope (°) 7.4 0.78 -10-35 67 Roof Slope (°) -12.3 2.23 -65-45 56 rear at a wide range of angles (Table 1). All eyries we examined, except for 2 (3%), had overhanging rock or ceilings protecting the nest scrape. Percent coverage of nest scrapes was frequently 100%, and averaged 93%. Across their range, Prairie Falcons nested most frequently in potholes or on cliff ledges. Crevices and stick nests were used less frequently (Table 2). Nest cliffs. — Heights of eyries and nest cliffs varied widely. Relative to mean cliff heights, mean eyrie heights varied from 60-70% (Table 3). Mean heights of eyries from 8 studies were highly corre- lated with mean heights of cliffs (R 2 = 0.99; P < 0.001), and averaged 63% of mean cliff height. Minimum cliff and eyrie heights reported in these studies were 2.1 m and 0.8 m, respectively. Signifi- Table 2. Types of eyries used by nesting prairie falcons. State Pothole 1 n (%) Crevice 2 n (%) Ledge 3 Sticknest n (%) n <%> Reference WY 29 (41) 17 (24) 18 (25) 7 (10) This study CO 2 (14) 9 (64) 3 (21) -- - Williams ( 1 98 1) 4 ID 76 (60) 24 (19) 26 (21) Ogden and Hornocker (1977) OR 15 (42) 14 (39) 7 (19) - - Denton (1975) 4 NM 8 (39) 4 (19) 9 (43) Platt (1974) UT 26 (36) 7 (10) 23 (32) (16) 22 Porter and White (1973) MT 18 (37) 18 (37) 9 (18) 4 ( 8) Leedy (1972) CO-WY 20 (56) 16 (44) Enderson (1964) WA 6 (43) 1 (7) 7 (50) Decker (1931) 200 (45) 65 (15) 105 (24) 69 (16) 'Includes sites listed as caves, holes and cavities, includes sites listed as vertical cracks. 3 Includes sites listed as horizontal shelves. Includes a few sites with sticknests present. 4 Data not reported but sticknests used. Spring 1986 Prairie Falcon Nest Characteristics 23 Table 3. Mean heights of Prairie Falcon eyries and nest cliffs. Cliff Height (m) Eyrie Height (m) Percent 1 Height Reference Mean (n) Range Mean (n) Range 14.6 (71) 4.3- 34.6 9.8 (71) 2.9-30.6 67 This study 53.7 (14) 15 - 140 32.1 (14) 10 -90 60 Williams (1981) 36.9 (49) 3.7- 122+ 22.9 (41) 2.1-61 + 62 Denton (1975) 2 14.0 (21) 6-35 8.6 (21) 3 - 30 61 Platt (1974) 30.4 (126) 2.3 - 122+ 19.8 (126) 2.4- 122+ 65 Ogden (1973) 2 31.0 (44) 2.1 - 154.4 19.6 (51) 0.8 - 137.2 63 Porter and White (1973) 38.1 (57) 9.2- 92+ 24.4 (57) 3.1- 76.2 64 Leedy (1972) 3 15.8 (36) 7.7- 38.7 11.1 (36) 70 Enderson (1964) 29.3 (418) 2.1 - 154.4 18.5 (417) 0.8 - 137.2 63 Grand Mean Tercent height = Mean eyrie height x 100% Mean cliff height “Values biased low as some very tall eyries and cliffs were excluded. “Approximate values. cantly more Wyoming eyries were in the upper one-half of the cliffs than in the lower one-half (x 2 ; P < 0.005). But the upper and middle one-third of cliffs were used equally. Inclination of toe slopes ranged from 0-36°. Most (57%) cliffs were vertical (85-95°), though many (34%) sloped back into the hill ( > 100°). Only 9% were overhung (< 80°). We found nest cliffs fairly evenly distributed among volcanic buttes or mountains (34%), sandstone canyons (30%) and isolated rock out- crops (25%). Few occurred on low rimrock (11%). In Oregon, Denton (1975) reported eyries on bluffs (38%), ridges (29%), canyons (24%) and outcrops (9%). Exposure of eyries and cliffs. — Exposure data from 6 of the 8 western states are shown in Figure 1 (Appendix). Except for Utah and New Mexico, as- pects of both cliffs and eyries was reported. In no case did the distributions of eyrie and cliff aspects differ significantly (x 2 ; P >10). When 95% confi- fidence intervals could be calculated, intervals for eyrie exposure and cliff aspect overlapped exten- sively (Appendix; Fig. 3) Overall, mean eyrie as- pects averaged 161.2° (r = 0.24); mean cliff aspects averaged 158.3° (r = 0.19). Due to large disparities in sample sizes, these grand means are biased to- ward our data from Wyoming. Aspects of cliffs and eyries in Wyoming ranged from 0 - 360° (Fig. 2). We lumped eyrie exposures into 2 groups (northeast and southwest facing) and found that 70.2% faced southwest; close to the per- centage (69.4%) of eyries that faced southwest in Utah (Porter and White 1973; Fig. 16). Extent of horizontal exposure of eyries we examined (Fig. 3) ranged from 5 - 160°, and averaged 54 ± 12° (r = 0.83). Williams (1981) reported “angles of view” ranging from 30 - 170° and averaging 122 ± 3° (r = 0.74). Neither ANOVA nor x 2 tests revealed any sig- nificant (P >0.05) relationships among physical traits, including exposures, of Wyoming eyries and histories of nest occupancy and success. This held true both within and among years. 24 Runde and Anderson Vol. 20, No. 1 360 ° 360 ° Figure 1 . Mean aspect of Prairie Falcon eyries (A) or nest cliffs (B) in Montana, Colorado, Wyoming, Utah, Oregon and New Mexico. Vector lengths are proportional to degree of concentration (r) about mean aspect (Appendix). 360 ° 360 ° Figure 2. Frequency histograms of eyrie exposure (A) and cliff aspect (B) for Prairie Falcons in southern Wyoming, 1982-1984. Spring 1986 Prairie Falcon Nest Characteristics 25 B. Figure 3. Means and 95% confidence intervals for (A) extent of exposure and (B) cliff aspect (broken line) and eyrie exposure (solid line), for Prairie Falcons in southern Wyoming, 1982-1984. Discussion Compared to other western raptors, Prairie Fal- cons have narrow nest site requirements. They nest almost exclusively on rock cliffs (cf., MacLaren et al. 1984), Nest cliffs used vary widely in height, loca- tion and aspect. Yet in the studies examined, Prairie Falcons demonstrated some consistent patterns in their use of eyries. Mean eyrie and cliff heights were strongly correlated and distributions of eyrie and cliff aspects were similar. We do not infer habitat selectivity because data on availability of potential eyries and nest cliffs have neither been collected by us nor presented in the studies we cite. Eyries. — Potholes, reported most frequently, provide excellent shelter from both inclement weather and direct sunlight, but are not always available. In our experience, potholes occur most frequently in sandstone, while ledges and vertical cracks are more common in granite, basalt and con- glomerate cliffs. Vertical cracks and crevices also provide excellent shelter and shade, but may be subjected to chilling drafts. Nests in crevices were reported infrequently in the literature but were common in our study area. Nest cliffs. — Mean eyrie heights, as percentage of mean cliff heights, were surprisingly consistent and indicate extensive use of nest sites 60 - 70% up cliff faces. Further study is needed to explain this narrow range of eyrie heights relative to cliff heights. By nesting in the upper reaches of vertical cliffs, Prairie Falcons should be safe from mammalian predators. Such a location also provides a com- manding view of the territory and of potential prey or predators. Depending on aspect and prevailing winds, updrafts along cliffs may aid the falcons when leaving the nest to hunt or defend territory. Convec- tive updrafts may form more frequently on south- facing cliffs as they radiate more heat than cliffs with other aspects. Exposure. — Prairie Falcons have long been known to nest on cliffs with southerly aspects (e.g., Dawson 1913). Sixty-one and 42% of the eyries studied by Enderson (1964) and Ogden (1973) faced south. But their data, as reported, could not be included in our circular statistics analyses. With 2 exceptions (Platt 1974; Denton 1975), the studies examined indicated southerly mean aspects (Fig. 1). Although cliff and eyrie exposure varied widely in 26 Runde and Anderson Vol. 20, No. 1 most studies, Rayleigh’s test indicated significant nonrandomness in most cases (Appendix). Mean exposures were often only weakly significant (P < 0. 10) due to wide ranges of exposures and because most data from the literature were grouped into broad exposure classes. Aspects of nest cliffs in New Mexico and eyries in Oregon were random (Rayleigh’s test; P > 0.10). Excluding these, mean cliff aspect ranged from 144 - 288°, and mean as- pect of eyries ranged from 143 - 226°. Prairie Fal- cons do use north-facing cliffs and eyries (Figs. 1 and 2). In fact, Tyler (1923) reported that most Prairie Falcon eyries in southern California had northerly aspects and none were southerly. He at- tributed this to a scarcity of south-facing cliffs and an abundance of north-facing cliffs. Our results demonstrate a tendency for Prairie Falcons to use southerly aspects, but do not demonstrate any benefit in terms of nest success. Eyrie configuration and location on the cliff may override efforts of cliff aspect (Ellis 1982). Data on availability of potential cliffs and eyries with differing aspects are needed to conclude that falcons actively select south-facing cliffs. The microclimate within an eyrie is affected by its exposure and cliff aspect; southerly exposures likely help to moderate cold temperatures during incubation and brooding. Solar heating of nest scrapes may lessen the energy requirements of in- cubating and brooding falcons. The majority (97%) of eyries we examined had overhead protection. Most eyries examined by Enderson (1964), Feedy (1972), Platt (1974), Ogden and Hornocker (1977) and Williams (1981) were similarly sheltered. Overhead shade may be important during midday since nestlings exposed to direct sun when temper- atures are near 32° C may die (Nelson 1969). Such temperatures are unusual during the nestling period in our study area, but use of shaded eyries may increase nestling survival at lower elevations and latitudes. We believe shelter from late spring storms to be more important than protection from heat stress in our Wyoming study area. Such storms were as- sociated with a high rate (64%) of nesting failure on buttes in 1983. A spring snowstorm this same year was followed by desertion of all 5 active nests on buttes in a small area of northeastern Wyoming (J. Squires pers. comm.). In our study area, territories on tall exposed buttes consistently suffered higher rates of nest failure (43 - 64%) than those located on lower sheltered canyons, rims and outcrops (18 - 38%). However, x 2 t ests failed to confirm a signifi- cant association (P > 0. 10) between nest success and cliff location. In our study, neither cliff nor eyrie aspect ap- peared to determine patterns of eyrie occupancy or nest success (see also Ogden and Hornocker 1977). Eyrie, but not cliff, aspect apparently affected nest success in Colorado (Williams 1981) where all 5 southwest-facing eyries failed in 1980. Williams at- tributed this to local weather patterns. Biotic factors likely play a larger role than physi- cal habitat in determining occupancy and nest suc- cess in Prairie Falcons. Though not confirmed by statistical analyses (x 2 ; P>0.10), nesting Great- horned Owls (Bubo virginianus ) were often as- sociated with unoccupied and unsuccessful falcon territories in our study area. Although we did not quantify accessibility of eyries to mammalian pre- dators, this may be an important factor (Ogden and Hornocker 1977). Abundance of prey clearly af- fects Prairie Falcon productivity and nest success. Townsend’s Ground Squirrel ( Spermophilus townsendi ) densities in Idaho accounted for over 98% of the annual variation in Prairie Falcon pro- ductivity (USD1-BLM 1979). Management implications. — Mitigation for loss of suitable nesting cliffs due to energy, urban, or recreational development may be needed in some areas. An obvious solution is the creation of new nest sites. Such sites were rapidly colonized in California (Boyce et al. 1980) and Alberta (Fyfe and Armbruster 1977). Because the size of the non- breeding population of Prairie Falcons may be as high as 25% of the breeding population (Runde in prep.) substantial population increases seem possi- ble, but will depend upon the local availability of prey and foraging habitat. The data presented here suggest locating new nest sites on south-facing cliffs about two-thirds up the cliff. Area of the floor should be at least 7,000 cm 2 , and slope gently (5-10°) to the front. Overhead cover should be present, and horizontal exposure should be about 54° (Fig. 3A). If eyries are to be created on mine high walls as part of habitat recla- mation or mitigation, their legal, hydrological, and physical constraints must first be addressed (Tessman 1984). The data in Table 3 suggest that cliffs, or highwalls, selected for eyrie creation be at least 2.1 m tall. We suggest that newly created eyries be clustered Spring 1986 Prairie Falcon Nest Characteristics 27 together to mimic nesting territories with several alternate nest sites. These new territories could be spaced approximately 1-2 km apart unless visual barriers are present to separate them. In this situa- tion, we have seen successful nests as close as 150 m. Our data show that in a 3-yr period over one-half of the eyries were only occupied 1 yr, and 23% were occupied 2 yrs. Similar findings were reported by Enderson (1964) and Ogden (1973). We suggest that 2 or 3 nest sites be provided per cliff. The ability to use alternate eyries within a nesting terri- tory may help to reduce direct competition for nest sites with owls, and may decrease the frequency of parasitic infestations. Acknowledgments Tricia MacLaren helped locate and collect data from the Wyoming eyries. Richard Guenzel programmed the circular statistics calculations. James Enderson, Grainger Hunt, Larry Ir- win, James Mosher and Karen Steenhof reviewed earlier versions of this paper. This study was supported by the BLM, the Wyoming Cooperative Fish and Wildlife Research Unit, USFWS, National Wildlife Federation, and the late Robert Runde. Literture Cited Batschelet, E. 1981. Circular statistics in biology. Academic Press. New York. 366 pp. Boyce, D. A., Jr., L. Fisher, W.E. Lehman, B. Hipp and J. Peterson. 1980. Prairie falcons nest on an artificial ledge. Raptor Res. 14:46-50. Dawson, W.L. 1913. The nesting of the prairie falcon in San Luis Obispo County. Condor 15:56-62. Decker, F.R. 1931. Prairie falcon’s nest. The Oologist 48:96. Denton, S.J. 1975. Status of prairie falcons breeding in Oregon. M.S. Thesis. Oregon State Univ., Corvallis. 58 pp. Ellis, D.H. 1982. The peregrine falcon in Arizona: Habitat utilization and management recommendations. Inst. Raptor Stud. Res. Pap. No. 1. Enderson, J.H. 1964. A study of the prairie falcon in the central Rocky Mountain region. A uk 81:332-352. Fyfe, R.W. and H.I. Armbruster. 1977. Raptor man- agement in Canada. Pages 282-293. In R.D. Chancel- lor (Ed.). Proc. of the World Conference on Birds of Prey. Vienna, 1975. ICBP. 442 pp. Leedy, R.R. 1972. The status of the prairie falcon in western Montana: special emphasis on possible ef- fects of chlorinated hydrocarbon insecticides. M.S. Thesis. Univ. of Montana, Missoula. 96 pp. MacLaren, P.A., D.E. Runde and S.H. Ander- son. 1984. A record of tree-nesting prairie falcons in Wyoming. Condor 86:487. Nelson, M.W. 1969. Status of the peregrine falcon in the Northwest. Chp. 4 In J.J. Hickey (Ed.). Peregrine Falcon Populations: Their biology and decline. Univ. of Wisconsin Press, Madison. 596 pp. Ogden, V.T. 1973. Nesting density and reproductive success of prairie falcons in southwestern Idaho. M.S. Thesis. Univ. of Idaho, Moscow. 39 pp. Ogden, V.T. and M.G. Hornocker. 1977. Nesting density and reproductive success of prairie falcons in southwestern Idaho./. Wildl. Manage. 41:1-11. Platt, S.W. 1974. Breeding status and distribution of the prairie falcon in northern New Mexico. M.S. Thesis. Oklahoma State Univ., Stillwater. 68 pp. Porter, R.D. and C.M. White. 1973. The peregrine falcon in Utah, emphasizing ecology and competition with the prairie falcon. Brigham Young Univ. Sci. Bull. 28:1-74. Tessman, S. 1984. Habitat reclamation procedures for surface mines in Wyoming. Pages 185-195. In R.D. Comer et al. (Eds.). Proc. of the Symposium: Issues and Technology in the Management of Impacted Western Wildlife. Steamboat Springs, Colo. Nov. 15-17, 1982. Thorne Ecol. Inst. Tech. Pub. 14. Boulder, Colo. 250 pp. Tyler, J.G. 1923. Observations on the habits of the prairie falcon. Condor 25:95-97. U.S.D.I. - B.L.M. 1979. Snake River Birds of Prey Spe- cial Research Report to the Secretary of Interior. Boise, Idaho. 142 pp. Williams, R.N. 1981. Breeding ecology of prairie fal- cons at high elevations in central Colorado. M.S. Thesis. Brigham Young Univ. Provo, Utah. 44 pp. Zar, J.H. 1984. Biostatistical Analysis. Second Edition. Prentice Hall. Englewood Cliffs, NJ. 736 pp. Wyoming Cooperative Fisheries and Wildlife Research Unit, Department of Zoology and Physiology, University of Wyoming, Box 3166, University Station, Laramie, Wyom- ing 82071. Received 1 May 1985; Accepted 20 September 1985 28 Runde and Anderson Vol. 20, No. 1 Appendix: Mean aspect of Prairie Falcon eyries and Prairie Falcon nest cliffs. State Mean Aspect 1 r Value Significance Level 2 N Reference Eyries Wyoming 180°±31° 0.41 P<0.001 67 Present study Colorado 200°±60° 0.44 P<0.10 14 Williams (1981) Oregon 255° 0.18 P>0.10 36 Denton (1975) Utah 226°±34° 0.36 P<0.01 49 Porter and White (1973) Montana 143°±33° 0.34 P<0.001 49 Leedy (1972) Cliffs Wyoming 191±23° 0.40 P<0.001 71 Present study Colorado 200 p ±6(F 0.44 P<0.10 15 Williams (1981) Oregon 288°±51° 0.28 P<0.10 36 Denton (1975) New Mexico 353° 0.07 P>0.10 21 Platt (1974) Montana 144°±35° 0.34 P<0.01 45 Leedy (1972) ‘With 95% confidence interval, when calculable. 2 Rayleigh’s test. DEVELOPMENT OF HUNTING AND SELF-SUFFICIENCY IN JUVENILE RED-TAILED HAWKS (Buteo jamaicensis) Sara Jane J ohnson ABSTRACT - Forty-eight juvenile Red-tailed Hawks (Buteo jamaicensis) were observed for 2 mo following fledging. Their flight activity and capture rate of vertebrate prey were quantified as a means of describing development of self-suffi- ciency. As juveniles aged, increasing amounts of time were spent in hunting versus nonhunting activities, and versatility of hunting methods increased. Capture of vertebrate prey began 42 d after fledging, but parents continued to provide food at least to 53 d past fledging. Development of self-sufficiency was indicated to be a gradual process whereby juveniles progressively capture more and more of their own food while parental food provision declines. During 2 yrs of field study on post-fledging be- havior of the Red-tailed Hawk (Buteo jamaicensis) (Johnson 1973), I rarely saw juveniles capture ver- tebrate prey. This limited hunting ability of juveniles during the early post-fledging period was consistent with both Amadon’s ( 1 964) and Ashmole and Tovar’s (1968) assumptions that the long post- fledging period in many species of birds is an adaptation to enhance survival of juveniles while they acquire specialized feeding techniques. Juvenile foraging behavior in a variety of bird species has been quantified to indicate progressive improvements in foraging efficiency through in- creases in hunting versus nonhunting activities (Buckley and Buckley 1974; Davis 1975), as well as progressive development of increasingly complex capture techniques (Dunn 1972; Smith 1973; Buckley and Buckley 1974; Feare 1975; Davies 1976; Davies and Green 1976). For raptors, de- velopment of juvenile hunting skills in the field has been defined in descriptive rather than quantitative terms, and only for several species such as the Swainson’s Hawk (Buteo swainsoni ) (Fitzner 1979) and Peregrine Falcon (Falco peregrinus) (Sherrod 1983). The objective of this study was to describe the development of self-sufficiency of the juvenile Red-tailed Hawk in the early post-fledging period by quantifying flight types when juveniles re- mained localized in the vicinity of the parental ter- ritory. Methods I conducted my study during the 1971-1973 breeding seasons in the Gallatin Valley, Gallatin Co., Montana. For details of the site seejohnson (1973). The Gallatin Valley is comprised of a mixture of pasture and dryland farming. Deciduous trees generally occur only along rivers and creeks. A total of 48 juveniles were observed, 21 from 8 nests in 1971, 20 from 8 nests in 1972, and 7 from 2 nests in 1973. All juveniles were color-marked on ventral wing surfaces with nontoxic enamel spray paint. No abnormal wear and tear was observed on marked feathers. Radio transmitters were placed on 10 and 6 juveniles in 1972 and 1973, respectively. Transmitters weighed up to 41 g, including the harness, frequency range 150-151 mhz, and trans- mitted approximately 0.3 km at ground level to 25-3 1 km from the air. Transmitters were monitored with a 12-channel AVM porta- ble receiver and four-element Yagi antenna. A double-layered polyethylene harness held the transmitter on the bird’s back be- tween the wings with the antenna extending parallel to the tail. A dissolvable gut suture attachment gradually deteriorated and al- lowed the harness to eventually fall off the bird. The earliest any harness fell off was 33 d following placement on the bird. The influence of the transmitter on the hawk’s flight behavior was unknown; some influence may have been possible. Frequency, but not duration, of 7 flight types was measured during hour-long observation periods throughout each day be- tween 0900 H and 1900 H. Measurements were initiated on an opportunistic basis when located birds became active. Due to diffi- culty in locating untagged birds, most flight measurements were recorded on radio-tagged individuals. The 7 flight types quan- tified included 1) perch/perch-direct flights between 2 elevated perches, 2) perch/quarter-indirect flights between elevated perches during which birds engaged in random quartering flights within 3-15 m of the ground, 3) perch/ground-direct flights from an elevated perch to the ground, 4) quarter/ground-indirect flights to the ground during quartering flights, 5) ground/ ground-flights which were initiated and terminated on the ground, 6) perch/adult-approaches to the parent birds and 7) perch/soar-initiation of soaring flights. I considered 3 of the above flights as hunting activity: perch/ quarter, quarter/ground and perch/ground. Quartering flight is a common prey-search method for buteos (Wakeley 1974), and Red-tailed Hawk flights to the ground generally occur during prey-capture attempts. In some instances, the 3 flight types I have identified as hunting activity may not have actually involved hunting activity. However, there is no means of separating these out, and I do not believe they contribute any significant problem to data collection. The 4 remaining flight types were defined as general movement activity (perch/perch, perch/soar), harassment of parent birds for food (perch/adult) and play and/or capture of invertebrates (ground/ground). I did not consider soaring flight a juvenile hunting activity; during the 3 field seasons, I never observed soaring juveniles attempting to capture ground-level prey. I did observe soaring juveniles grab air-borne invertebrates with their feet. For purposes of analysis, I combined the hourly observation samples into 5 age classes. Age classes were initially designated at 10-day intervals, since notable range expansions of juveniles oc- occurred at approximately 20 and 30 d past fledging (Johnson 1973). However, because the number of monitored individuals 29 Raptor Research 20 (l):29-34 Table 1. Mean (± S.D.) number of moves per hour by age class for 7 types of flights by juvenile Red-tailed Hawks. The sample is based on a total of 48 known-aged juveniles observed during 3 breeding seasons. Percentages (in parentheses) include only the non-soaring flight activities. 30 Sara Jane Johnson Vol. 20, No. 1 a o o H S a « £ g 2 « g Cm 3 < q < ^ CM w Q i-i ' — ' Cm © to co © © © CM q d — CO , CM cd , cd +1 j +1 1 +i i +1 i +1 CM 00 co 00 © CO CO © i— 1 CM csi © d © © in CM © CO © © OO d , d , © , © ! j +1 +1 +i : +1 o © CO oo CO © d d © © © © © CO ly q S' <=> © © ! i | | +1 d i" c. +! © © CM © © © O CM © o © oo © ! 1 3±0. (0.7) 6±0. (4.4) 7±0. (4.7) © +! r- o CM © d o © © © CO © © CM © 6±0. (1-4) © © © © © : ! +1 d © o +1 o © © CM d d © © © to ’sH © © CO o ^ cq ^ rq © © © d m d in d CM d +1 +1 00 + 1 05 +1 cd + 1 00 05 -1 © Cm © © © ' © o o 1—1 1—1 © CO © © © ao © ^ © j 1 J 1 2±0 (8.7) + 1 © +1 05 OO ~ OO i — 1 co © © o © © © CO CO CO © © 05 ^ CM ^ — 1 cm CM © d © CM 00 CM + 1 CM +1 d +1 in +1 d + 1 cr> ' — i r- co © © Tf © CM © © CM CO CM CM d o 6 o bn 10 cm, 0 if < 10 cm Rough-legged Hawks have been described as “microtine specialists” that can shift to other prey when voles are unavailable (White and Cade 1971). Snow depths, particularly above 10 cm, likely precluded small mammal use by hawks and they moved to roads to take advantage of increased carrion. Road-killed rabbits, which were the primary source of carrion, created a phenomenon similar to that for Bald Eagles ( Haliaeetus leucocephalus) attracted to fish killed by powerplant operations (Ingram 1965) or fluctuations in waterrelease rates at dams (Steenhof 1976). In both situations, a human-caused superabundance of carrion served to attract physiologically stressed raptors during winter. Such conditions probably “short-stopped” some Rough-legged Hawks migrating through the Snake River Plain during the early and latter part of the survey period, and accounted for a portion of the winter hawk population. Hawk counts were known to reflect the presence of transmitter-equipped resident hawks which shifted foraging ter- ritories to highways during inclement weather in winter 1982-83 (Watson 1984). Although this species is known to take a variety of avian and mammalian prey (see review by Sherrod, 1978), prior accounts of carrion consumption (Weller 1964; Schnell 1967) are qualitative with carrion forming an insignificant portion of the total diet During winter 1982-83, leporids comprised 48.6% of prey num- bers and 70.1% of biomass of prey consumed by Rough-legged Hawks, whereas voles comprised 41.3% and 9.6%, respectively (Watson 1984). The reported variability in the consumption of carrion and movement of hawks to roads in cold weather (Schnell 1968; Bildstein 1978; Fleming 1981; Klein and Mason 1981), is evidently linked to winter range character. When snow cover reduces rodent availability, Rough-legged Hawks have the option of switching to alternate prey such as carrion, if it is available, or relocating to other areas. Acknowledgments This research was a contribution of the INEL Radioecology and Ecology Program, funded by the Office of Health and Environ- mental Research, Department of Energy in cooperation with the Fish and Wildlife Program, Department of Biology, Montana State University. Published as Journal Series No. 1729, Montana Agricultural Experiment Station. Thanks are extended to R.L Eng, J.W. Grier, R.P. Howard and O.D. Markham for reviewing drafts of this note and to H. Lee, K. McGarigal and R.A. Watson for assistance. Literature Cited Bildstein, K.L. 1978. Behavioral ecology of Red-tailed Hawks (Buteo jamaicensis). Rough-legged Hawks (B lagopus), Northern Harriers ( Circus cyaneus), American Kestrels (Falco sparverius) and other raptorial birds wintering in south central Ohio. Ph.D. Diss., Ohio State Univ., Columbus. 364 pp. Craig, T.H. 1978. A car survey of raptors in southeast- ern Idaho 1974-76. Raptor Res. 12:40-45. Fleming, T.L. 1981. A 3-winter raptor survey of the Columbia Basin, Washington/Oregon. U.S. Army Corps of Engineers, U.S. Bureau of Reclamation, U.S Bureau of Land Management, WA. 143 pp. Spring 1986 Short Communications 43 Harniss, R.O. and N.E. West. 1973. Vegetation pat- terns of the National Reactor Testing Station, south- eastern Idaho. Northwest Sci. 47:30-43. Ingram, T.N. 1965. Wintering bald eagles at Gutten- burg, Iowa - Cassville, Wisconsin, 1964-65. Iowa Bird Life 35:66-78. Klein, R.J. and D.R. Mason. 1981. Change in raptor hunting behavior following heavy snowfall. Raptor Res. 15:121. Montgomery, D.C. and E.A. Peck. 1982. Introduction to linear regression analysis. John Wiley and Sons, Inc., New York, NY. 504 pp. Schnell, G.D. 1967. Population fluctuations, spatial distributions, and food habits of Rough-legged Hawks in Illinois. Kansas Ornith. Soc. Bull. 18:21-28. 1968. Differential habitat utilization by wintering Rough-legged and Red-tailed Hawks. Condor 70:373-377. Sherrod, S.K. 1978. Diets of North American fal- coniformes. Raptor Res. 12:49-121. Steenhof, K. 1976. The ecology of wintering bald eagles in southeastern South Dakota. M.S. Thesis. Univ. of Missouri, Columbia. 148 pp. Stoddart, L.C. 1983. Relative abundance of coyotes, lagomorphs and rodents on the Idaho National En- gineering Laboratory, pp. 268-277. In 1983 Progress Report, Idaho National Engineering Laboraory Radioecology and Ecology Programs, ID- 12098. Nat. Tech. Inf. Serv., Springfield, VA. Thurow, T.L., C.M. White, R.P. Howard and J.F. Sul- livan. 1980. Raptor ecology of Raft River Valley, Idaho. EGG-2054. Idaho Falls, ID. 45 pp. Watson, J.W. 1984. Rough-legged Hawk winter ecol- ogy in southeastern Idaho. M.S. Thesis, Montana St. Univ., Bozeman. 101 pp. Weller, M.W. 1964. Habitat utilization of two species of buteos wintering in central Iowa. Iowa Bird Life 34:58-62. White, C.M. and T.J. Cade. 1971. Cliff-nesting raptors and ravens along the Colville River in Arctic Alaska. Living Bird 10:107-150. Zar, J.H. 1974. Biostatistical analysis. Prentice-Hall, Inc., Englewood Cliffs, NJ. 620 pp. Department of Biology, Montana State University, Bozeman, MT 59717. Present address: RR 1 Box 860- A, Warrenton, Oregon 97146. Densities of Red-tailed Hawk Nests in Aspen Stands in the Piceance Basin, Colorado Mike McGovern and John M. McNurney This note describes dissimilar nesting densities of the Red-tailed Hawk {Buteo jamaicensis) in 2 areas in Colorado. Although Red- tailed Hawks nest in a variety of habitats (Knight et al. 1982; Smith and Murphy 1982) hawks were observed nesting only in aspen ( Populus tremuloides) trees. Between 21 June and 1 July 1983, Red-tailed Hawk nests were surveyed in pure stands of aspen in 2 areas (designated A and B) in the Piceance Basin, Garfield County, Colorado. Areas A and B are approximately 38 and 28 km, respectively, north of Debeque, Mesa County, Colorado, at elevations between approximately 2400 and 2500 m. Both areas have similar types of vegetation. Area A was 28.7 km 2 in size and contained 24 aspen stands cover- ing 3.1 km 2 . Area B was 14.0 km 2 in size and contained 17 aspen stands covering 2.8 km 2 . The remainder of the areas consisted primarily of shrubs (1 to 3 m in height) including mountain mahogany ( Cercocarpus montanus), serviceberry (Amelanchier utahensis), Gambel oak ( Quercus gambelii), big sagebrush ( Artemesia tridentata) and others, with occasional areas composed of annual grasses. Surveys of all aspen stands in both areas were done by helicopter (approximately 40% of the survey) or on foot. For those stands surveyed on foot, transects were walked at 50-m intervals follow- ing the elevational contours of each stand until all trees were examined. Nests were deemed occupied if young were seen in the nest, if the nest was recently decorated by greenery, or if a nest was defended by an adult hawk. Locations of occupied nests were marked on 7.5-min topographical maps. Nearest neighbor analyses (Clark and Evans 1954) were conducted to determine if hawk nests were spaced randomly throughout each area. Density of occupied nests was one/5.74 km 2 on area A, and mean distance between nests was 2.23 (± 0.46 S.D.) km. Mean distance between nests in area B was 0.68 (± 0.33 S.D.) km, with 1 breeding pair/2.00 km 2 . Mean density of nests and distances between nests on areas A and B were comparable with data found in the litera- ture (Table 1). However, mean distance between nests on area B (0.68 km) was lower than all values reported (Table 1). In area A, nearest neighbor analysis indicated that occupied nests tended toward uniform distribution and were significantly different from random (R = 1.84; c:= 3-60, P < 0.01). In area B, occupied nests were not significantly different from random dis- tribution (R = 1.19; C = 0.93, P < 0.10). The percentage of area covered by aspen on area A (11%) was less than that of area B (20%). In addition, there were more trees within the aspen stands on area. A that were small (3-5 m high) (R.W. Beck and Associates 1983a, 1983b) and apparently ill-suited for Red-tailed Hawk nest sites. Therefore, available nesting habitat in the vicinity of oc- cupied nests in area A may have been more limited than in area B. Indeed, there was a mean of 0.60 km 2 (range = 0.0 1-1.14 km 2 ) of suitable nesting habitat (trees > 5m in height) within a 1-km radius of the nests in area A. In area B there was a mean of 1.04 44 Short Communications Vol. 20, No. 1 Table 1. Nesting density and distance between Red-tailed Hawk nests. Area/Breeding Pair (Km 2 ) Distance Between Nests (Km) Source 7.5 1.79 McInvaille & Keith 1974 2.13 2.05 1.90 1.60 2.40 - 5.60 Knight etal. 1982 6.2 1.50 Springer & Kirkley 1978 - 3.30 Smith & Murphy 1973 - 0.84 Wiley 1975 - 6.40 5.1 1.76 Hagar 1957 - 2.08 Seidensticker & Reynolds 1971 2.6 — Gates 1972 1.3 1.76 Fitch et al. 1946 8.8 — Orians & Kuhlman 1956 6.2 — 6.9 ... Luttich et al. 1971 24.9 ... Corman 1973 in Springer & Kirkley 1978 7.9 — Johnson 1975 7.7 ± 6.8 2.5 ± 1.58 X ± S.D. of reported values 5.74 2.23 This study, Area A mean (N = 5) 2.00 0.68 This study, Area B mean (N = 6) km 2 (range = 0.62 - 1.48 km 2 ) of suitable habitat within a 1-km radius of nests in that area. Although dissimilar, the means were not statistically different (Student’s t test; 0.05 < p < 0.10). Newton (1976) suggested that Red-tailed Hawk densities are determined by availability of nesting sites and food. Thus, an explanation for the lower density of nests and the greater dis- tances between nests on area A may have been availability of nesting sites. However, the apparently high density of nests and low mean distance between nests on area B remains unexplained. Acknowledgments We thank Dr. Clayton M. White of Brigham Young University for his helpful comments on this note. Literature Cited Clark, Philip J. and Francis C. Evans. 1954. Distance to nearest neighbor as a measure of spatial relation- ships in populations. Ecology 35:445-453. Fitch, H., S. Swenson and D.F. Tillotson. 1946. Be- havior and food habits of the Red-tailed Hawk. Condor 48:205-237. Gates, John M. 1972. Red-tailed Hawk populations and ecology in east-central Wisconsin. Wilson Bull 84:421-433. Hagar, Donald C., Jr. 1957. Nesting populations of Red-tailed Hawks and Horned Owls in central New York State. Wilson Bull. 69:263-272. Johnson, Sara Jane. 1975. Productivity of the Red- tailed Hawk in southwestern Montana. Auk 92:732- 736. Knight, Richard L., Dwight G. Smith and Albert Erickson. 1982. Nesting raptors along the Columbia River in north-central Washington. The Murrelet 63:2-8. Luttich, Stuart N., Lloyd B. Keith and J.D. Stephen- son, 1971. Population dynamics of the Red-tailed Hawk (Buteo jamaicensis ) at Rochester, Alberta. Auk 88:75-87. McInvaille, William B., Jr. and Lloyd B. Keith. 1974. Predator-prey relations and breeding biology of the Great Horned Owl and Red-tailed Hawk in central Alberta. The Canad. Field-Natural. 88:1-20. Spring 1986 Short Communications 45 Newton, I. 1976. Population limitation in diurnal rap- tors. The Canad. Field-Natural. 90:274-300. Orians, Gordon and Frank Kuhlman. 1956. Red- tailed Hawk and Homed Owl populations in Wiscon- sin. Condor 58:371-385. R.W. Beck and Associates. 1983a. Environmental baseline description for Getty Oil Company’s Oil Shale Project - Garfield County, Colorado. R.W. Beck and Associates, Denver, Colorado. 1983b. Environmental baseline de- scription for Cities Service Oil Shale Project - Garfield County, Colorado. R.W. Beck and Associates, Denver, Colorado. Seidensticker, John C. IV and Harry V. Reynolds III. 1971. The nesting, reproductive performance, and chlorinated hydrocarbon residues in the Red- tailed Hawk and Great Horned Owl in south-central Montana. Wilson Bull. 83:408-418. Smith, Dwight G. and Joseph R. Murphy. 1 973. Breeding ecology of raptors in the eastern Great Basin of Utah. Brig. Young Univ. Sci. Bull. 18:1-76. Smith, Dwight G. and Joseph R. Murphy. 1982. Nest site selection in raptor communities of the eastern Great Basin Desert. Great Basin Natural. 42:395-404. Springer, Mark A. and John S. Kirkley. 1978. Inter- and intraspecific interactions between Red-tailed Hawks and Great Horned Owls in central Ohio. Ohio J. of Sci. 78:323-328. Wiley, James W. 1975. Three adult Red-tailed Hawks tending a nest. Condor 77:480-482. R.W. Beck and Assoc., 660 Bannock St., Denver, CO 80204. Received 15 April 1985; Accepted 1 July 1985 Thesis Abstracts The Sharp-Shinned Hawk (Accipiter striatus Vieillot ) in Interior Alaska Breeding ecology of the Sharp-shinned Hawk (Accipiter striatus) was studied at 19 nests in interior Alaska from 1978 to 1981 . Hawks nested in conifers in dense, young stands of mixed deciduous and coniferous trees. Sharp-shins primarily ate small birds, apparently hunted the most productive habitats and captured prey in proportion to availability. Growth and food requirements of 4 captive-reared nestlings were monitored to supplement data on wild young. A typical family required an estimated 13,620 g of prey during the breeding season. In comparison to other studies, Sharp-shinned Hawks in Alaska 1 ) reoccupied old nest areas more frequently, 2) occupied smaller home ranges, 3) nested in greater densities, 4) completed breeding cycles more quickly, 5) laid more eggs and 6) hatched and fledged more young. In future studies, which are important because of the sharp-shin’s extensive range and susceptibility to pollution and habitat destruction, Alaskan birds could serve as standards of comparison. — Clarke, Ronald Gordon. 1984. M.S. Thesis, University of Alaska, College, Alaska. Characterization of Nesting Habitat of Goshawks (Accipiter gentilis) in Northwestern California Habitat use of nesting Goshawks (Accipiter gentilis) was studied during 3 breeding seasons in Six Rivers National Forest, Humboldt and Trinity Counties, California. Habitat characteristics of the nesting areas were examined on 4 levels: community patterns, nest stand, nest site and the nest and nest tree, for 10 nests. Nest stands typically were dense single-storied stands of young Douglas-fir (Pseudotsuga menziesii) with scattered hardwood components and large mature trees and a park-like understory. Locations varied in slope and elevation, but consistently faced northeast. Nest sites typically were small stands of dense mature trees within the nest stands. Tree density and canopy closure were less in nest sites than in the surrounding nest stands. Nests generally were constructed of sticks, were adjacent to the stem, and were below or within the lower quarter of the canopy on the downslope side of a Douglas-fir. Distance to the nearest water source and human disturbance ranged widely. Potentially suitable foraging and alternate nesting areas averaged 41m and 30 m respectively from the nest tree. — Hall, Patricia A. 1984. M.S. Thesis, Humboldt State University, Areata, California. 46 Thesis Abstracts Vol. 20, No. 1 Distribution and Density of the Four Common Passerines in West Greenland Quantitative analyses of songbird distribution, influence of boulders and depressions of numbers near Peregrine Falcon ( Falco peregrinus) eyries have not been previously reported for inland west Greenland. In the present study passerine distribution, habitat utilization, density and response to nesting Peregrine Falcons were determined by conducting line transects near eyries and on the open tundra. Passerine distribution is strongly influenced by habitat and presence of boulders. A marked depression in passerine numbers was recorded within 400 m of active peregrine eyries. Densities were estimated as 0.23 - 0.38 birds/ha. Density estimates are lower than those reported from several Arctic areas. — Meese, Robert J. 1984. M.S. Thesis, Brigham Young University, Provo, Utah. A Phenology of Wintering Bald Eagles in the Chilkat Valley, Alaska Communal winter feeding and roosting Bald Eagles, ( Haliaeetus leucocephalus alascanus) , between September and March, 1977-78 and 1978-79. During this period 1 10 censuses were conducted. Twenty-one environmental, population and habitat use variables were quantified; these data were analyzed using bivariate and univariate statistical procedures to ascertain the effects of environmental variation on the population dynamics and habitat use of Bald Eagles wintering in an undistrubed communal roost. Immigration of Bald Eagles into the Chilkat Valley began the first week of October. The major influxes of eagles occurred between the second week of October and the second week of November. A peak number of 2,578 eagles was recorded in 1977 on 8 November and in 1978 a peak number of 2,254 was recorded on 24 December. Emigration from the area was completed in the last week of March in 1978 and the third week of February in 1 979. The percentage of juvenile eagles in the population decreased from the second week of October to March. Eagles were distributed throughout the valley during autumn. In the second week of November they concentrated in the ice-free area where spawned-out salmon were available. Intense feeding activity was significantly correlated with above-freezing temperatures that accompanied severe wind chill conditions froze the salmon carcasses. During these periods the eagles would abandon the feeding areas and utilization of trees for shelter increased significantly. The number of eagles observed decreased by approximately 30% when cold, clear weather persisted, but increased with the return ofwarmer, overcast weather. — Waste, Stephen McIntosh. 1985. M.Sc. Thesis, Humboldt State University, Areata, California. 86 pp. Behavior and Habitat Use of Breeding Northern Harriers in Southwestern Idaho Radiotelemetric and visual monitoring of 4 breeding Northern Harrier ( Circus cyaneus) pairs in predominantly sagebrush (Artemesia sp.) habitat of the Snake River Bird of Prey Study Area, Idaho, indicated that harriers used riparian and cultivated habitats disproportionately. Males were observed in an apparent habitat and prey shift, changing from Meadow Voles (Microtus pennsylvanicus) in alfalfa {Medicago sativa) fields as growth approached 46 cm, to Whip-tailed Lizards ( Cnemidophorous tigris) in open sagebrush habitat. Home ranges of males were estimated at 15.7 km 2 , while those of females were estimated at 1. 13 km 2 . Males were far ranging and were observed at distances of 9.5 km from the nest. Male hunting activities were highest in the second week post-hatching. Most of the time both males and females were observed resting or preening less than 0.5 km from the nest. — Martin, John W. 1984. M.S. Thesis, Brigham Young University, Provo, Utah. Raptor Inventory and Ferruginous Hawk Breeding Biology in Southeastern Oregon Raptor inventories were conducted in southeastern Oregon in 1979 and 1980 on extensive study areas, and on randomly selected 10.4-km 2 study units. Overall raptor densities ranged from 10-23 pair/100 km 2 . Estimates of Golden Eagle (Aquila chrysaetos ) densities are 2-4 and 4-5 pair/ 100 km 2 , respectively. An important nesting area for Ferruginous Hawk (Buteo lagopus) and Prairie Falcon was discovered near Vale, Oregon. In 1980, 32 nesting pairs of Ferruginous Hawks were located on a 312 km 2 study area. Clutch size averaged 3.9, and 3.2 young fledged/ nesting attempt. This is one of the densest and most productive populations of Ferruginous Hawks ever reported. Ferruginous Hawks nested on the ground, on outcrops and on cliffs. They preyed upon Townsend Ground Squirrel (Spermophilus townsendii) almost exclusively. Ground squirrel distribution is related to soil characteristics. Soils with shallow duripans or clay appear to be unfavorable ground squirrel habitats. Vegetation parameters account for 28 percent of the observed variability in ground squirrel hole counts along transects. Crested wheatgrass seedings are occupied by ground squirrels and are compatible with Ferruginous Hawk management in the study area. The over-riding influence of soil type on ground squirrel distribution suggests that soil maps may be an effective way of locating areas with a high potential for raptor nesting concentrations. A survey of Oregon wildlife biologists determined that 100 active Ferruginous Hawk nests have been identified in Oregon. — Lardy, Michael Edward. 1980. M.S. Thesis, University of Idaho, Moscow. Spring 1986 Thesis Abstract 47 The Peregrine Falcon (Falco peregrinus) in Southern Brazil: Aspects of Winter Ecology in an Urban Environment The study was undertaken in an urban environment of Porto Alegre, Brazil (30 02" S and 51 13" W), during 2 austral summers (December 1978 to March 1979 -January to March 1980). Seven falcons (4 <$