RAPTOR RESEARCH Volume 7 Number 2 Summer 1973 Raptor Research Foundation. Jnc, Vermillion, South Dakota, L'.S.A- RAPTOR RESEARCH Volume 7, Number 2, Pages 25-92 Summer 1973 CONTENTS SCIENTIFIC PAPERS Predatory Efficiency of American Kestrels Wintering in Northwestern California— Michael W. Collopy 25 Foraging Behavior of Ospreys in Northwestern California— Meyer L. Ueoka and James R. Koplin 32 Differential Habitat Use by Sexes of American Kestrels Wintering in Northern Califomia-James R. Koplin 39 Post-fledging Activity of the Red-tailed Hawk-Sara Jane Johnson ... 43 Spatio-temporal Relationships between Breeding Red-tailed Hawks and Great Horned Owls in South Dakota— Thomas C. Dunstan and Byron E. Harrell 49 REPOR TS, RE VIEWS , AND OPINIONS Report: Proceeding of the Conference on Raptor Conservation Techniques, Fort Collins, Colorado, 22-24 March 1973. Part 1. Introduction - Byron E. Harrell 55 Report: Proceedings of the Conference on Raptor Conservation Techniques, Fort Collins, Colorado, 22-24 March 1973. Part 2. Raptor Ecology Session— edited by Byron E. Harrell 62 Publication Notice : New publications from Brigham Young University . 70 NOTES, NEWS, AND QUERIES Publication Date of Spring Issue, 1 973 72 New Membership Mailing Address 72 Publication of Raptor Research Abstracts 72 Raptor Telemetry Research Survey Report ............ 72 Breeding Project Information Exchange - • . 72 Veterinary Aspects of Captive Birds of Prey . . . - 72 RAPTOR RESEARCH Published by: RAPTOR RESEARCH FOUNDATION, INC. c/o Department of Biology University of South* Dakota Vermillion, South Dakota 57069 U.S.A. Editor: Richard R. Olendorff 3317 Olympus Drive Bremerton, Washington 98310 U.S.A. Associate Editors: Byron E. Harrell Department of Biology University of South Dakota Vermillion, South Dakota 57069 U.S.A RAPTOR RESEARCH is published quarterly in Spring, Summer, Fall, and Win- ter issues and occasional Supplements. The contents are usually divided into three sections. The first section is SCIENTIFIC PAPERS for reports of original research or theoretical analyses. These papers will be given careful editorial and referee scrutiny. A second section, REPORTS , REVIEWS, AND OPINION, will include secondary material, translations of material originally published else- where, reports of work still in progress, reports on meetings, often in some de- tail, book reviews, and other similar items. This material will be edited for accu- racy but will not receive the critical review given the Scientific Papers. Because of the preliminary or secondary nature of the material in this section the Edi- tors recommend that this material be cited in other papers only with great care or in a very general way and especially with specific preliminary or conference material only after consultation with the source of that information. Papers which express a personal opinion or letters to the Editor will be included in this section. NOTES, NEWS, AND QUERIES is used for notices of information or events, requests for information, news items either specially prepared or re- printed from other sources, and similar small items. This journal began publication as RAPTOR RESEARCH NEWS with Volume 1 in 1 967 as a quarterly in typewritten mimeographed form on an 8Vi” by 11” page size. Volumes 2 and 3 in 1968 and 1969 were offset printed but continued the same frequency, page size, and standard typewriter type. An analytical in- dex for Volumes 1-3 was published. Volumes 4 and 5 in 1970 and 1971 were published six times a year in offset printing, SW by 8V2” page size, and with IBM Composer typefaces; an analytical index for Volumes 4-5 is in preparation. In 1 972, Volume 6, the name of the journal was changed to reflect the broader scope to RAPTOR RESEARCH. Currently the journal is published quarterly by offset printing with 6%” by 9Vi” page size and IBM Composer typefaces and annual analytical indexes. For membership and publication costs see inside back cover. Donald V. Hunter, Jr. Centerville, South Dakota 57014 U.S.A. PREDATORY EFFICIENCY OF AMERICAN KESTRELS WINTERING IN NORTHWESTERN CALIFORNIA* by Michael W. Collopy School of Natural Resources California State University Humboldt, Areata, California 95521 ABSTRACT. A study of foraging behavior of the American Kestrel ( Falco spar- verius) wintering in Humboldt County, California was undertaken from Octo- ber 1972 to February 1973. An analysis of predatory efficiency and prey selec- tion is presented on the basis of numbers of species taken relative to the num- ber of capture attempts, and the relative frequency with which they occur in the Kestrel’s diet. A total of 498 capture attempts was observed; of these, 233 (46.8%) were successful, 221 (44.4%) were unsuccessful, and 44 (8.8%) were undetermined. Of the 233 successful captures 199 (85.4%) were invertebrates (beetles, grasshoppers, earth worms, butterflies and other insects) and 34 (14.6%) were vertebrates (mice, shrews and birds). The American Kestrel ( Falco sparverius ) is one of the most common raptors in the New World, and has been the subject of many studies. Cade (1955) studied winter territoriality among Kestrels in southern California and suggested that territoriality serves to secure an adequate food supply through the winter. Roest (1957) described different aspects of breeding behavior, hunting meth- ods and social interactions among Kestrels. Willoughby and Cade (1964) de- scribed the breeding behavior of captive Kestrels. Enderson (1960) reported on movements in a resident Kestrel population in east-central Illinois. The lack of studies concerning prey selection and predatory efficiency among raptors, in general, and the American Kestrel, in particular, prompted this study of foraging behavior. Observations were made from October 1972 to February 1973 in the Areata Bottoms east of Areata, Humboldt County, California. The results reported here are part of a larger study to relate predatory efficiency, prey selection and activity budget, as observed in the field, to metabolic rates measured in the laboratory. The project is ongoing; more field data will be col- lected. I am indebted to my advisor, J. R. Koplin, for his guidance and assistance throughout this study. I am also grateful to Howard Levenson and Larry Norris for censusing small mammals during the fall and winter months of 1972 and 1973. *This paper was presented at the Conference on Raptor Conservation Tech- niques in Fort Collins, Colorado, 22-24 March 1973. 25 Raptor Research 7(2): 25-31, 1973 26 RAPTOR RESEARCH Vol. 7, No. 2 Study Area The Areata Bottoms consist of extensive pasture land used for grazing cattle. It is an ecological unit bordered on the south by Humboldt Bay, on the west by the Pacific Ocean, on the north by the Mad River, and on the east by the city of Areata and the surrounding Redwood ( Sequoia sempervirens ) forests. Vege- tation in the Areata Bottoms consists predominantly of grasses, forbs, shrubs and few trees. Materials and Methods Observations of Kestrels were made during all daylight hours including as many full days as possible. Kestrels habitually used fence posts, power poles and power lines as perches; observations were readily obtainable. A spotting scope, binoculars and stop watch were used to observe and time the hunting activities of Kestrels under observation. Rodent censuses in areas of hunting activities were made with use of Sher- man live traps and the rate of capture technique (Davis 1963). Prey Selection Comparative analysis of the hunting efficiencies of male and female Kestrels was not possible because of the small number of males observed. The male to female ratio of Kestrels wintering in the Areata Bottoms was one to nine (Kop- lin 1973). Observations of the males present were difficult to obtain; for this paper, data on both sexes were combined. Observations were made on 24 American Kestrels, with intensive observa- tions on six. Dives which resulted in the capture of prey or which resulted in landing on the ground were considered “completed dives.” Over the five-month period, 498 completed dives were recorded; of these, 233 were successful. Identification of food items showed that 34 were verte- brates and 199 were invertebrates. The relative frequency of each is shown in Table 1 . Identifiable invertebrates included five grasshoppers, nine beetles, nine earthworms, and two butterflies. Identifiable vertebrates included six Califor- nia Meadow Voles ( Microtus calif ornicus), three Western Harvest Mice ( Reith - rodontomys megalotis), and 18 Vagrant Shrews ( Sorex vagrans ). Two small birds and five small mammals could not be identified. Considered on a seasonal basis, the data show that the number of inverte- brate prey decreased and the number of vertebrate prey increased during the winter (Figure 1 ). The capture of vertebrates reached a peak during December, the coldest month in Humboldt County history. In subsequent months, the fre- quency of invertebrates taken increased with a corresponding decrease in the number of vertebrates taken. Censuses of insect populations were not conducted. No quantitative evidence exists to determine whether the Kestrels selected rodents and shrews in prefer- ence to available insects, or fed primarily on rodents and shrews in the absence of insects. However, during December, unusually large numbers of Killdeer ( Charadrius vociferus ) moved into the study area from the north and from the Summer 1973 Collopy— Predatory Efficiency of Kestrels 27 Table 1 . Prey species captures by American Kestrels wintering in Humboldt County, California. Prey Species Number Captured Percent of 233 Prey Captured Invertebrates grasshoppers 5 2.2 beetles 9 3.9 earthworms 9 3.9 butterflies 2 0.9 unidentified insects 174 74.7 total invertebrates 199 85.6 Vertebrates Micro tus 6 2.6 Reithrodontomys 3 1.3 Sorex 18 7.7 unidentified mammals 5 2.2 small birds 2 0.9 total vertebrates 34 14.7 Total Prey Captured 233 99.3 surrounding foothills. A great many of these terrestrial insect-eating birds sub- sequently starved to death. This was one indication that Kestrels were feeding primarily on rodents and shrews in the absence of insects. An attempt was made to assess variation in the availability of rodents and shrews. Because of the small numbers captured by Kestrels and the lack of population estimates from trapping data, small birds and Western Harvest Mice were not included in the analyses. Through the fall and winter months, esti- mates of prey densities were obtained in two areas where Kestrels commonly hunted. During the census periods, there was no significant difference between the two areas in the relative numbers of prey species trapped (X^=0.72; p=0.25- 0.50, 1 d.f.). There was a decline in prey densities in excess of 50 percent between the fall and winter censuses. However, in each area the difference between the relative number of prey species trapped in the winter, and in the fall, was not statistic- ally significant (X^=3. 19; p=0. 10-0.05, 1 d.f.). Thus, it was possible to combine the data from both areas for estimates of densities of Micro tus and Sorex populations. A Chi-square test was made of the 28 RAPTOR RESEARCH Vol. 7, No. 2 Figure 1. Numbers of vertebrates and invertebrates captured monthly by winter- ing American Kestrels. relative frequency of Microtus and Sorex captured by the Kestrels, and the rela- tive numbers trapped. This test was made to determine if selection for a particu- lar prey species was occurring. The Kestrels captured Microtus and Sorex in the same relative proportions as trapped (X^= 1.18; p=0.025-0.500, 1 d.f.). This in- dicates no measurable selection for either of the two major vertebrate prey spe- cies. A final comparison was made between the fall and winter rodent and shrew densities, and the fall and winter rodent and shrew captures by the Kestrels. The increase in the incidence of captures in the winter and the measured de- creases in the densities were significantly different (X2=9.28; p=0.005, 1 d.f.). This supports the postulate that Kestrels exploited the rodent and shrew popu- lations in the absence of insect prey. Summer 1973 Collopy- Predatory Efficiency of Kestrels 29 Hu n ti ng Effi ci ency I observed 498 capture attempts; of these, 233 (47 percent) were successful, 221 (44 percent) were unsuccessful and 44 (9 percent) were undetermined (Fig- ure 2). The inability to assess the success of a dive resulted from a Kestrel div- ing into a gulley, behind a bush or behind other visual obstructions. Eliminating such observations, the Kestrels captured prey during 5 1 percent of their capture attempts. Kestrels most commonly hunted from a perch; however, they also hovered while hunting. Of 95 completed dives from a hovering position, 22 (i.e. 23 per- cent) were successful, 69 (73 percent) were unsuccessful, and 4 (4 percent) were undetermined. Of the 403 completed dives from a perch, 211 (52 percent) were successful, 152 (38 percent) were unsuccessful and 40 (10 percent) were undetermined. These data demonstrate that hunting from a perch was a more 500 400 CO LLI > o Q 300 5 o u 200 lOO | | successful unsuccessful undetermined 52% 38% 47% 44% PERCH HOVER TOTAL Figure 2. The success of hunting methods used by wintering American Kestrels. 30 RAPTOR RESEARCH Vol. 7, No. 2 efficient means of capturing prey. I attempted to differentiate vertebrate hunting attempts from invertebrate hunting attempts. This separation was based on observed differences in hunting methods. Fast, flex-gliding dives, were classified as “vertebrate dives”, and soft, fluttering dives were classified as “invertebrate dives.” Of those dives in which the success was determined, 36 of 145 (25 percent) of the “vertebrate dives” were successful, while 198 of 309 (64 percent) of the “invertebrate dives” were successful. These results reflect the greater mobility of the rodent and shrew prey. Considering the diverse diet of the Kestrel, the possible existence of prey- specific hunting behavior should be investigated. The evolution of specialized hunting methods might be a contributing factor in the widespread success of this species. Discussion Several recent studies provide additional information on the hunting success of Kestrels. In Costa Rica, Jenkins (1970) recorded a hunting efficiency of 39 percent by a male Kestrel. Sparrowe (1972) observed 54 capture attempts by 20 Kestrels in Michigan; of these, 33 percent were successful. The differences in predatory efficiency between these studies and mine could be attributed to small sample size. However, these differences could also reflect differences in habitats, or differences in availability of prey species, or a combination of both. Rudebeck (1951) recorded an over- all hunting efficiency of eight percent among migrating European Sparrow Hawks ( Accipiter nisus ), Peregrine Falcons (. Falco peregrinus), Merlins (F. columbarius ) and White-tailed Sea Eagles (Hali- aeetus albicilla). This is considerably lower than the 51 percent hunting success recorded among American Kestrels wintering in the Areata Bottoms, indicating the advantage of maintaining a territory. Increased hunting efficiency could re- sult from familiarity of winter residents with locations of available food re- sources. Future observations on hunting efficiency of migrating Kestrels in the Areata Bottoms should indicate whether or not this is the case. Literature Cited Cade, T. J. 1955. Experiments on winter territoriality of the American Kestrel, ( Falco sparverius). Wilson Bull. 67:5-17. Davis, D. E. 1963. Estimating the numbers of game populations: In Wildlife In- vestigational Techniques, pp. 89-1 18. Enderson, J. H. 1960. A populations study of the Sparrow Hawk in east-central Illinois. Wilson Bull. 72:222-231. Jenkins, R. E. 1970. Food habits of wintering Sparrow Hawks in Costa Rica. Wilson Bull. 82:97-98. Koplin, J. R. 1973. Differential habitat use by sexes of American Kestrels win- tering in northern California. Raptor Res. 7:39-42. Sparrowe, R. D. 1972. Prey-catching behavior in the Sparrow Hawk. J. Wildlife Manage. 36:297-308. Summer 1973 Coll opy— Predatory Efficiency of Kestrels 31 Roest, A. I. 1957. Notes on the American Sparrow Hawk. Auk 74:1-19. Rudebeck, G. 1951. The choice of prey and modes of hunting of predatory birds with special reference to their selective effect. Oikos 3 (part 2): 200- 231. Willoughby, E. J. and T. J. Cade. 1964. Breeding behavior of the American Kestrel (Sparrow Hawk). Living Bird 3:75-96. FORAGING BEHAVIOR OF OSPREYS IN NOR1HWESTERN CALIFORNIA* by Meyer L. Ueoka and James R. Koplin School of Natural Resources California State University Humboldt, Areata, California 95521 ABSTRACT. Fishing success of Ospreys ( Pandion haliaetus ) at South Hum- boldt Bay, California, was determined during spring and summer, 1972. Five hundred twenty-three of the 639 observed fishing efforts (82 percent) were successful. Fifty-six percent of the fishing efforts required only one dive. The mean length of time spent foraging for each successful fishing effort was 1 1.8 minutes. The highest fishing success occurred during the fledging period, and lowest during incubation, 86 and 79 percent, respectively. Successful fishing efforts were highest on the outgoing tide early in the breeding season. Fishing success on outgoing tides decreased as the season progressed, and increased (59 percent) on the incoming tide. Successful fishing efforts were the same (43 per- cent) on both incoming and outgoing tides during the brooding period. Skill and swiftness in capturing prey animals are survival requisites for avian predators occupying terminal positions of food chains. The present study was made to determine the success of Ospreys in securing fish during the breeding season. The feeding habits of Ospreys were virtually unknown until fairly recently. Lambert (1943) observed Osprey strikes and found their efficiency to be 89 percent. His observations did not take into account the number of dives re- quired per individual Osprey to capture prey. However, as a result of more re- cent studies, we have a better understanding of the fishing success of Ospreys in different areas of the western United States. Over a three-year period, MacCarter (1972) found Ospreys at Flathead Lake, Montana, to be 83 percent successful in 158 fishing efforts. At Eagle Lake, Lassen County, California, Garber (1972) found that Ospreys were 80 percent successful in 25 fishing efforts. Ospreys fishing the surf at Usal Creek, northern Mendocino County, California, were 86 percent successful in 1 16 fishing efforts (French 1972). Sixty-three percent of fishing efforts at Flathead Lake resulted in capture of prey on the first dive. At Eagle Lake and Usal Creek, successful fishing efforts on the first dive were 52 and 67 percent, respectively. The foraging behavior of Ospreys at South Humboldt Bay, California was de- *This paper was presented at the Conference on Raptor Conservation Tech- niques in Fort Collins, Colorado, 22-24 March 1973. 32 Raptor Research 7 (2): 3 2-38, 1973 Summer 1973 Ueoka and Koplin— Foraging of Ospreys 33 termined during spring and summer, 1972. This is a preliminary report of an ongoing study and is based on more than 600 observed fishing efforts. We are grateful to Jon French, Craig Bradshaw, David LaConte and Paul Col- bert, present and former students at California State University, Humboldt, who assisted in field observations. Study Area Humboldt Bay is located on the coastal region of Humboldt County, Cali- fornia, approximately 435 kilometers (270 miles) north of San Francisco. It is 22.5 kilometers (14 miles) long and from 0.8 to 6.4 kilometers (0.5 to 4.0 miles) wide, and lies in a southwesterly to northeasterly direction. Being con- nected to the ocean, it is subject to tidal changes. The surface area of Humboldt Bay fluctuates from 62.7 sq. kilometers (24.2 sq. miles) at mean high tide to 33.2 sq. kilometers (12.8 sq. miles) at mean low tide. The Bay is composed of three different parts with respect to fluctuating surface areas. South Bay has 18.4 sq. kilometers (7.1 sq. miles) of surface area at mean high tide, and 7.8 sq. kilometers (3.0 sq. miles) at mean low tide; En- trance Bay has 7.5 sq. kilometers (2.9 sq. miles) at both tides; and North Bay has 36.8 sq. kilometers (14.2 sq. miles) at mean high tide, and 17.9 sq. kilo- meters (6.9 sq. miles) at mean low tide (Skeesick 1963). Preliminary observations at Humboldt Bay revealed that Ospreys were pri- marily fishing South Bay. Therefore, observations were restricted to the south- ern portion of Humboldt Bay. Methods A 20x spotting scope, binoculars and stopwatch were used to obtain infor- mation on the fishing activities of Ospreys. Fishing success was measured by re- cording the total number of fishing efforts— total dives made by one Osprey while under observation. Measurements of predatory efficiency were obtained by recording the total number of dives per fishing effort resulting in capture of prey. Each fishing effort was timed from the moment an Osprey first was sighted until it either captured fish, or was lost to view. Observations were made from dawn to dusk for two days each week from a vantage point where a major portion of South Bay could be viewed. Tide con- ditions were noted during each day of observations. Fishing Success and Predatory Efficiency A total of 639 fishing efforts was observed at South Humboldt Bay (Table 1). Fifty-six percent of fishing efforts resulted in capture of prey on the first dive. However, on one occasion an Osprey required 13 dives to capture a fish. Over-all, Ospreys were 82 percent successful in their fishing efforts. Of the 116 unsuccessful fishing efforts (18 percent), only 21 (3 percent) involved Os- preys that flew inland without securing fish. The remaining 95 (15 percent) un- successful fishing efforts involved birds lost to view. 34 RAPTOR RESEARCH Vol. 7, No. 2 Table 1 . Relative fishing success of Ospreys fishing South Humboldt Bay, Usal Creek, Eagle Lake, California; and Flathead Lake, Montana. Modified from Koplin et al (in press). No. of dives S. Humboldt Usal Creek Eagle Lake Flathead Lake per successful Bay (% of 639 (% of 116 (% of 25 (% of 158 fishing effort fishing efforts) fishing efforts) fishing efforts) fishing efforts) 1 56 67 52 63 2 18 15 16 12 3 6 3 8 7 4 1 1 4 1 5 0.6 - — — 6 0.2 - - - 13 0.2 - - - Total % Successful fishing efforts 82 86 80 83 The 639 fishing efforts were stratified into five phases of the breeding cycle: pre-incubation, incubation, brooding, fledging and post-fledging (Figure 1). This was done to determine whether fishing success of Ospreys changed through the breeding season. At Nova Scotia, Lambert (1943) found a decrease in the cap- ture of fish from May through August. Fishing success varied through the sea- son at South Bay. The highest success was during the fledging period and the lowest during incubation, 86 and 79 percent, respectively. Time Spent Foraging Mean length of time spent foraging from first sighting until the capture of prey for the 523 successful fishing efforts was 11.8 minutes. However, in terms of different phases of the breeding season, it appears as if average foraging time was inversely related to food demand (Figure 2). That is, as the demand for fish by the growing young increased, average foraging time decreased. Consequently, when demand was highest during brooding, average foraging time was lowest. When young began fledging and a continuous supply of fish was no longer es- sential, average foraging time again increased through the late fledging and post- fledging periods. Effects of Tidal Change Fishing success was highest (59 percent) on the outgoing tide early in the breeding season. Success in relation to season declined on the outgoing tide and increased on the incoming tide (Figure 3). Successful fishing efforts were the Summer 1973 Ueoka and Koplin— Foraging of Ospreys 35 0 J— , , , Preincubation Incubation Brooding • i Fledging Postfledging Figure 1 . Successful fishing efforts for each period of the breeding season. same (43 percent) on both incoming and outgoing tides during the brooding period. We are at a loss to explain these findings, but suspect differential move- ments of fish in and out of the Bay on a seasonal basis. Prey Species Sixty-three percent of the fish taken by Ospreys were surfperch (Embiotoci- 36 RAPTOR RESEARCH VoE 7, No. 2 Figure 2. Mean length of time spent foraging for successful fishing efforts for each period of the breeding season. dae). Of the fish caught, three percent were anchovies (Engraulidae), two per- cent silversides (Atherinidae), two percent herrings (Clupidae), one percent scul- pins (Cottidae), and 29 percent were unidentifiable. In addition, the remains of nine surfperch, two silversides and one sculpin were found beneath a perch used by Ospreys for feeding at South Bay. The lengths of 21 1 fish caught by Ospreys were estimated to be 18 to 23 cm (7-9 inches). Seven fish were less than 10 cm (4 inches) long, and only one was greater than 38 cm (15 inches) long. Discussion Mech (1966) stated that predators miss far more prey than they catch. How- ever, as indicated in this and other studies, Ospreys are an exception. They are successful predators. Ospreys are capable of capturing two fish, one in each foot, on a single dive as observed by French (1972) at Usal Creek. One of us (M. L. U.) also witnessed a double catch at Usal Creek. On several occasions, when Ospreys dropped the fish they were carrying, they quickly secured a second. Ospreys were seen to consume a small fish com- Summer 1973 Ueoka and Koplin-Foraging of Ospreys 37 Figure 3. Successful fishing efforts for each period of the breeding season in relation to the tide cycle. pletely and begin fishing again, capturing a second, often larger fish before heading inland to the nest site. As indicated by the average time spent foraging and the number of success- ful fishing efforts requiring one dive, Ospreys have little difficulty in locating and capturing prey. When only the 523 successful fishing efforts are considered 69 percent were successful on the first dive; 523 of the 834 total dives (62 per- cent) resulted in capture of fish, indicating that Ospreys catch fish more often than they miss. 38 RAPTOR RESEARCH Vol. 7, No. 2 Literature Cited French, J. M. 1972. Distribution abundance, and breeding status of Ospreys in northwestern California. M.S. Thesis, Calif. State Univ., Humboldt, Areata, California. 58 pp. Garber, D. P. 1972. Breeding ecology of Ospreys in Lassen and Plumas Coun- ties, California. M.S. Thesis, Calif. State Univ., Humboldt, Areata, Califor- nia. 59 pp. Koplin, J. R., D. S. MacCarter, D. P. Garber, and D. L. MacCarter. In press. Food resources and fledgling productivity of California and Montana Os- preys. Proc. N. Amer. Osprey Res. Conf. Lambert, G. 1943. Predation efficiency of the Ospreys. Can. Field-Natur. 57: 87-88. MacCarter, D. S. 1972. Food habits of Ospreys at Flathead Lake, Montana. M.S. Thesis, Calif. State Univ., Humboldt, Areata, California. 80 pp. Mech, L. D. 1966. The wolves of Isle Royale. Fauna of the Nat. Parks of the U. S., Fauna Ser. 7. Skeesick, D. G. 1963. A study of some physical-chemical characteristics of Humboldt Bay. M.S. Thesis, Calif. State Univ., Humboldt, Areata, Cali- fornia. 148 pp. DIFFERENTIAL HABITAT USE BY SEXES OF AMERICAN KESTRELS WINTERING IN NORTHERN CALIFORNIA* by James R. Koplin School of Natural Resources California State University Humboldt, Areata, California 95521 ABSTRACT. Differential habitat use by sexes of American Kestrels ( Falco spar- verius ) wintering in Humboldt, Mendocino, Lake, Colusa and Yolo Counties, California, was discovered and quantified by roadside counts conducted during the winters of 1970-71 through 1972-73. A prevalence of females was discover- ed in expansive pastures and fields devoid of or with few trees, and a prevalence of males was discovered in orchards and small pastures and fields surrounded by trees. However, Chi-square tests of homogeneity and frequency indicated a complex transition in habitat use. Relative abundance of Kestrels, measured by linear distance between 340 successive observations, differed significantly only between the Central Valley and the other areas sampled. Kestrels averaged 1 15 individuals per 100 miles (161 km; 135 observations) in the Central Valley and 34 birds per 100 miles (161 km; 205 observations) elsewhere. The differential use of habitats by wintering Kestrels is interpreted as character displacement. Its relevance to the principle of competitive exclusion is discussed. Population studies of raptors wintering in the Areata Bottoms near Areata, Humboldt County, California, in the fall of 1970 revealed an unexpectedly high proportion of female Kestrels {Falco sparverius) in censuses. Six to twenty times as many females as males were counted through the winter. Roadside counts of Kestrels wintering in other areas in coastal Humboldt County also revealed high numbers of females. Since these findings were contrary to those reported by Roest (1957), who found that males constituted approximately 60% of late summer, fall and winter populations of Kestrels over a wide geo- graphic area in the United States, I extended the study to include other geo- graphic areas in northern California. Methods I conducted roadside counts of Kestrels between late November and mid- February in Humboldt, Mendocino, Lake, Colusa and Yolo Counties. Habitats censused included open areas relatively devoid of trees, areas vegetated pre- dominantly by trees, and areas transitional between these two major habitats. *This paper was presented at the Conference on Raptor Conservation Tech- niques in Fort Collins, Colorado, 22-24 March 1973. 39 Raptor Research 7(2): 39-42, 1973 40 RAPTOR RESEARCH VoE 7, No. 2 Open habitats included pastures in the vicinity of Humboldt Bay and agri- cultural fields in the Central Valley. Wooded areas censused were in the Coast Range. Transitional habitats censused were immediately east of Humboldt Bay and riparian habitats along the Sacramento River in the Central Valley. Kestrels were observed mainly in agricultural areas in the Coast Range. So few were observed in redwood ( Sequoia sempervirens ) and Douglas fir ( Pseudo - tsuga menziesii ) forests on the west slope and in pine-oak-chapparal woodlands on the east slope that they were not counted in these habitats. Thus, surveys in the Coast Range were restricted to agricultural areas along the Mad, Elk, Van Duzen, Eel and Mattole Rivers, and Jacoby and Salmon Creeks on the west slope; and to agricultural areas along Cache Creek and Clear Lake on the east slope. Each Kestrel observed was examined carefully with the aid of 10-power bi- noculars or a 15-power spotting scope to determine its sex. Only those birds in which sex was positively identified were recorded. Attempts to age Kestrels in the field were unsuccessful. Distance between successive observations of 340 birds was recorded to the nearest 0.1 mile (0.16 km) on automobile odometers as a means of measuring density. Results and Discussion A total of 670 Kestrels was sexed in all habitats surveyed during the winters 1970-71 through 1972-73 (Table 1). Chi-square tests of homogeneity and fre- Table 1. Sex ratios of Kestrels wintering in open, wooded, and transitional habitats in northern California. Geographic area and habitat No. fern ales No. males Sex ratios* females:males Coastal Humboldt County open pasturelands 151 16 9.4: l a transitional habitats 192 55 3.5: 1^ Central Valley open farmlands 73 19 3.8: l b transitional habitats 26 21 1.2:l c ** Coast Range wooded pastures and orchards 35 85 1:2.3 d * Ratios followed by different letters are significantly different from each other (p 4 0.05); ratios followed by the same letter are insignificantly different from each other (p = 0.75). * insignificantly different from a 1 : 1 sex ratio (0.75 < p < 0.50); all others are significantly different from a 1 : 1 sex ratio (p 4. 0.05). Summer 1 973 Koplin— Habitat Use of Kestrels 41 quency indicated significant differences in sex ratios among Kestrels in pastures on the coast, riparian habitats in the Central Valley, and wooded habitats in the Coast Range. Sex ratios in transitional habitats on the coast and in agricultural fields in the Central Valley were statistically similar, but differed significantly from the other three habitat types. The null hypothesis that sex ratios were 1 : 1 was rejected for all but Kestrels in riparian habitat in the Central Valley. Analysis of variance of distance between successive observations showed only two differences: average densities of 34 Kestrels per 100 miles (161 km; 95% confidence estimate of mean = 34 ± 11 Kestrels per 100 miles— 161 km) in coastal and Coast Range habitats, and 115 Kestrels per 100 miles (161 km; 95% confidence estimate of mean = 115+25 Kestrels per 100 miles- 161 km) in Central Valley habitats. This indicated a difference in relative suitability of these areas for Kestrels, including differences in suitability of coast transitional and Central Valley open habitats in which there was no significant difference in sex ratios. These results clearly indicate habitat segregation between sexes of Kestrels wintering in northern California. Males predominate in wooded, agricultural habitats and females predominate in open pastures and fields. The open and transitional habitats differed latitudinally as well as climatically. Therefore, it is not possible to interpret the decrease in the relative proportion of females in open and semi-wooded habitats between coastal Humboldt County and the Central Valley. More extensive surveys are required to determine if sexes of Kestrels are segregated latitudinally (i.e., geographically) as well as on the basis of habitat. Habitat segregation by sexes of wintering Kestrels is interpreted here as char- acter displacement (Brown and Wilson 1956) reducing intersexual competition for food resources. Habitat segregation by sexes is but one of several methods by which intersexual competition for food may be alleviated. In his general re- view of the phenomenon, Selander (1966) discussed several alternatives includ- ing one other that has been documented for Falconiformes: the selection of prey differing in size by sexes of accipiters (Hoglund 1964, Storer 1966) and Peregrine Falcons ( Falco peregrinus ) (Cade 1960). Accipiters and Peregrine Fal- cons exhibit pronounced sexual dimorphism in body size; the size of prey selected by these birds is directly proportional to their own body size. Presum- ably, Kestrels also fit this pattern. However, even though the sexes of Kestrels differ in body size, they are much less dimorphic than are the sexes of accipit- ers and Peregrine Falcons (Selander 1966:139). Therefore, the sexes of the rela- tively monomorphic Kestrels would be expected to select prey of similar size in different habitats, whereas sexes of the more dimorphic accipiters and Peregrine Falcon would be expected to select prey of different sizes in the same habitat. In support of this postulate, Cade (1960:243) reported that he has unpublished data indicating that sexes of Kestrels select prey of similar size and Collopy (pers. comm.) reports no evidence of differential size selection of prey by sexes 42 RAPTOR RESEARCH Vol. 7, No. 2 of Kestrels. Thus, habitat segregation of Kestrels by sexes may be an adaptation necessary to alleviate intersexual competition for food. Literature Cited Brown, W. L. and E. O. Wilson. 1956. Character displacement. Syst. Zool 5:49- 64. Cade, T. J. 1960. Ecology of the Peregrine and Gyrfalcon populations in Alas- ka. Univ. Calif. Publ. Zool 63:151-290. Hoglund, N. M. 1964. Uber die Ernahrung des Habichts (Accipiter gentilis Lin.) in Schweden. Viltrevy 2:271-328. Roest, A. I. 1957. Notes on the American Sparrow Hawk. Auk 71:1-19. Selander, R. K. 1966. Sexual dimorphism and differential niche utilization in birds. Condor 68:113-151. Storer, R. W. 1966. Sexual dimorphism and food habits in three North Ameri- can accipiters. Auk 83:423-436. POST-FLEDGING ACTIVITY OF THE RED-TAILED HAWK* by Sara Jane Johnson Department of Zoology and Entomology Montana State University Bozeman, Montana 59715 ABSTRACT. During a two-year study of post-fledging activity of Red-tailed Hawks ( Buteo jamaicensis ) in the Gallatin Valley, Montana, in 1971 and 1972, data were collected on mortality, length of time the fledglings remain within the home territory, and characteristics of their movements out from, and pos- sibly, back to their home territory. Mortality was found to be low during the first 18-25 days after fledging, when the young were quite inactive. Mortality also may be low until the young begin leaving the home territory. The length of time the young remained associated with the adults ranged from 30 to 70 days after fledging for all fledglings observed. For nest-mates, time spans be- tween departures ranged from zero to 31 days after fledging. Movements of the fledglings out from and back to the home territory were variable. Some young made no movements out from the home territory until they left permanently, while others left and returned up to five times. The length of time the young remained associated with the adults, and thus within the home territory, did not determine the number of movements it made into the areas surrounding the home territoiy, before it left it permanently. A study was conducted during the summers of 1971 and 1972 on the post- fledging activities of Red-tailed Hawks {Buteo jamaicensis). The study area was located in the Gallatin Valley, Gallatin County, in southwestern Montana. The objectives of the study were to gather information on mortality, length of time the fledglings remained within the adult territory, their movements both within and outside of the adult territory, and the development of hunting and social behaviors. In this paper hunting and social behavior are not considered. Methods The study area included approximately 145 sq. km (56 sq. miles) and, during each summer, contained a maximum of 18 pairs of nesting Red-tailed Hawks. In 1971, eight successful nests were observed soon after the young fledged, but after they became more active, fledglings from only three nests were observed. All young were color-marked for identification. In 1972, eight nests were ob- served from the time the young fledged until they could no longer be located. *This paper was presented at the Conference on Raptor Conservation Tech- niques in Fort Collins, Colorado, 22-24 March 1973. 43 Raptor Research 7(2):43-48. 1973 44 RAPTOR RESEARCH Vol. 7, No. 2 All young were color-marked and nine were radio-tagged. Color-marking involved spraying the undersides of the wings and tail with non-toxic spray paint. The transmitters, purchased from the AVM Instrument Company and weighing approximately 20 g, were attached to the fledglings by a harness. Life of the transmitters was approximately three months. The har- ness was attached with dissolvable gut-suture, which would in time wear away and cause the transmitter to fall off the bird. These transmitters were placed on nine fledglings from seven nests at seven to eight weeks of age. All of the radio- tagged young were then located three times daily during the first and second day of a three-day cycle; just the young from a single nest were observed throughout the third day. Mortality Results on mortality of fledglings from the time they left the nest until they could no longer be located are given in Table 1. After the young fledged, they remained within a limited area, a post-nesting area, until from 18 to 25 days Table 1. Length of time fledglings observed after fledging when no mortality Nest occurred. Year Time spent Number of Days AF* when observed in post- young in all nestmates 61 1972 nesting area 18 the nest 4 known alive 41 7 1971 25 3 54 34 1971 3 33 7 1972 25 3 32 34 1972; 1971 • 23 3 47 11 24 2 33 9 1972 17 2 31 11 1972 21 2 42 33 1972 2 34 62 1972 20 2 39 8 1972 18 2 18 2 24 1971 23 3 23 2 25 1971 18 2 26 2 26 1971 22 3 27 2 15 1971 3 25 2 36 1971 2 23 Idays after fledging. ^Observations were discontinued at that age due to difficulty in locating the fledglings. Summer 1973 Johnson— Post-fledging Activity of Red-tailed Hawks 45 after fledging as measured from 13 nests during the two-year study. During this post-nesting period, the young attempted no hunting, remaining completely de- pendent upon the adults for food. The tendency of the adults to bring food less and less often as the post-nesting period progressed, and the increased develop- ment of the young, were probably two important factors causing the young to leave the post-nesting area and to move out where the adults were hunting. A total of 41 young from 16 nests were observed at least up to the end of the post-nesting period. During this time no mortality was observed. The earliest age when observations were discontinued on young from a nest was 18 days after fledging; thus, no mortality was observed for 41 young up to 18 days after fled- ging. Of 10 nests containing 26 young (all observed throughout the summer un- til the young could no longer be located), the latest date that all young were observed, and thus known to be alive, was 31 days after fledging. For these 26 young, no mortality had occurred at least through the first 31 days after fled- ging. The average age of these 26 young when last observed was 43 days after fledging. For fledgling Red-tailed Hawks in the Gallatin Valley it appears that mortality is very low for the first 17 to 25 days after fledging, during the post- nesting period, and, possibly, mortality is low until the young begin leaving the home territory. Length of Time Young Remain Home The ages in days after fledging when the 26 fledglings from the 10 nests which were observed until the fledglings could no longer be located, are given in Table 2. As only nine of the 26 young were radio-tagged, the reason for the Table 2. Age after fledging when fledglings last observed in the home territory. Nest Year Number Observed of young in nest 61 1972 4 34 1971 3 7 1971 3 7 1972 3 34 1972 3 11 1971 2 9 1972 2 11 1972 2 33 1972 2 62 1972 2 Days after fledging fledglings last observed 1st 2nd 3rd 4th 39 r 41 42 70 r 31 42 42 53 53 53 30 r 37 61 r 46 46 5 l r 33 47 31 r 31* 42 r 43* 34 r 35* 38 5 l r r Fledgling was radio-tagged. * Fledgling was not checked after that age. 46 RAPTOR RESEARCH Vol. 7, No. 2 disappearance of the remaining 17 young from the home territory was not known. This may have been due to their death or their leaving the home terri- tories. The ages of disappearance fell within the age range at which radio-tagged young left, which was from 30 to 70 days after fledging. The age at which the first nestmate from a nest left the home territory permanently ranged from 30 to 53 days after fledging. The first day that all nestmates from one nest were permanently gone from the home territory ranged from 47 to 70 days. Con- cerning the permanent departure of nestmates, only two nests contained more than one radio-tagged young, enabling a definite determination of the age both nestmates left the home territory. For these nests, time spans between depar- tures of nestmates were 31 days in both cases. When the ages of permanent de- parture of all 26 young are considered, there appears to be no pattern for de- parture of nestmates. In three nests in which a single radio-tagged young left first, the nestmates were not further observed. In the remaining seven nests, departure of nestmates was staggered (at least five days between departures) for three nests containing two, two, and three young. It was simultaneous (leav- ing on the same day or within several days later) for one nest containing three young. In the remaining three nests, containing three, three, and four young, departures of nestmates were both simultaneous and staggered. Thus, perman- ent departure of fledglings, whether nestmates or not, from the home territory appears to be quite variable, with a 40-day range between all departures, and a 31 -day range between departures of nestmates. Movement of Fledglings Outside of the Flome Territory Eight of the nine radio-tagged fledglings, and three color-marked, untagged fledglings were located at least once away from the home territory. A total of 23 movements away from the home territory were recorded; these are given in Table 3. In 91.3% of these movements, the direction traveled was south, west, east, and a combination of these three directions. Only 8.7% of the movements involved a northerly direction. Of these 23 recorded movements, 17 of them included the return of fledg- lings back to the home territory. Movements out where the fledglings later re- turned were generally of two patterns. Eighty-nine percent of the movements out were of short duration, where the fledgling returned either on the same day or within two days. These movements (88.9% of the total) ranged from 1.6 to over 8 kilometers (1 to over 5 miles) and averaged 3.06 kilometers (1.9 miles). The second type of movement out (1 1.1% of the total movements) was of long- er duration, five to seven days. Distances traveled from the home territory were 28.4 and 35.4 kilometers (17 and 22 miles) in the two cases observed. The two fledglings involved in these two longer movements out from the home territory were older than when either they or the other radio-tagged young made shorter movements out. The amount of movement out from and back to the home territory was vari- able for the nine radio-tagged young. Three young did not leave once until they left permanently. Three other young left and returned once before leaving per- manently. Three remaining young left and returned four, four, and five times Summer 1973 Johnson— Post- fledging Activity of Red-tailed Hawks 47 Table 3. Description of movements of young out of the home territory, includ- ing age in days after fledging when they left, and the direction and maxi- mum distance traveled from the home territory. Unradioed young which were occasionally observed away from the home territory are included. 60 8 | = Jl| H OJ -t— « <3J 8 | r-> G O +-» ^ <+-'■ — GO CD -t— “ 1 3 O ’’B ^ CD ^ Ti