U.S. DEPARTMENT OF THE INTERIOR BUREAU OF LAND MANAGEMENT BOISE DISTRICT IDAHO SNAKE RIVER BIRDS OF PREY RESEARCH PROJECT ANNUAL REPORT 1977 052 3807 As the Nation's principal conservation agency, the Department of the Interior has responsibility for most of our nationally owned public lands and natural resources. This includes fostering the wisest use of our land and water resources, protecting our fish and wildlife, preserving the environmental and cultural values of our national parks and historical places, and providing for the enjoyment of life through outdoor recreation. The Department assesses our energy and mineral resources and works to assure that their development is in the best interests of all our people. The Department also has a major responsibility for American Indian reservation communities and for people who live in Island Territories under U.S. administration. SNAKE RIVER BIRDS OF PREY RESEARCH PROJECT ANNUAL REPORT 1977 NOT -FOR PUBLICATION The data presented herein are preliminary and may be inconclusive. Permission to publish or cite any of these materials is therefore withheld pending specific authoriza- tion of the Boise District, BLM, and the specific Principal Investigator . Michael N. Kochert Project Leader D. Dean Bibles Boise District Manager William F. Mathews Idaho State Director ACKNOWLEDGMENTS The Bureau of Land Management wishes to thank those agencies and individuals who assisted with the project. Appreciation is extended to the Idaho Fish and Game Department, the U.S. Fish and Wildlife Service and U.S. Public Health Service. Professor John Oakley assisted us in the field as part of a sabbatical arrangement with Mira Costa Community College. The University of Montana provided a student intern to assist us in our field work. Dr. George Innis, Utah State University, provided technical modeling assistance. Thanks is extended to the Idaho Cooperative Wildlife Research Unit for furnishing the original background data and nest site locations; Mr. Morlan Nelson; and to those individuals, who volunteered their services. Special appreciation is extended to Mr. E.T. Evans for providing the area for the Melba field camp and other assistance. Ruth Dodge deserves special thanks for an excellent job of typing the manuscripts. SNAKE RIVER BIRDS OF PREY NATURAL AREA STAFF William F. Mathews D. Dean Bibles Alan B. Tripp Michael N. Kochert Albert R. Bammann Karen Steenhof Robert D. Rheiner Douglas A. Smithey COOPERATING Idaho Fish and Game Department U.S. Fish and Wildlife Service Idaho State Director Boise District Manager Chief, Division of Resources Boise District Research Project Leader Assistant Research Biologist Analytical Biologist Bruneau Resource Area Manage Management Biologist AGENCIES U.S. Public Health Service PREFACE The BLN became actively involved in raptor habitat preservation and management on October 12, 1971 when Secretary of the Interior, Rogers C.B. Morton, signed Public Land Order Number 5133 forming the Snake River Birds of Prey Natural Area (Natural Area) . This withdrawal preserved a 53 km (33 mile) portion of the Snake River Canyon for the protection of nesting habitat for this raptor population. Work in the Natural Area commenced when the BLM hired a research biologist in June 1972 to determine management and research needs for the area. A management biologist was hired in October 1973 to develop and implement the management alternatives outlined by research. The Official Snake River Birds of Prey Study Area (BPSA) was established in April 1974 while funding for the project came in the summer and field work commenced in the winter of that year. A moratorium on most land development on public lands within the Study Area boundaries was declared in 1975 pending the outcome of the Research Project. In September 1977, Secretary of the Interior, Cecil D. Andrus, instructed the Boise District to expand the study area boundary based upon current research information. The District identified 290,000 acres of public land to be added to the 1974 study area. Actions that might conflict with or jeopardize raptors or prey species in the expanded area are suspended until the completion of the research project. The Secretary also instructed the BLM to accelerate their research effort and complete the research project ahead of schedule. The completion date was moved ahead from January 1, 1980 to June 1, 1979. The Snake River Birds of Prey Research Project was designed when the basic research revealed that information on habitat requirements, land tenure, predator-prey relationships and life requirements of the raptors was needed to insure long-term manage- ment and stability of this population. The objectives of the Project are as follows: 1. Ascertain the necessary requirements in terms of habitat, successional stage, food, and space needed to sustain a breeding pair of the major raptors and rear their young. 2. Predict the effects of habitat alteration to the composi- tion and demography of the prey base and ultimately on reproductive performance and population dynamics of the major raptors. 3. Develop a long-term management model for the Natural Area and BPSA. The Project now consists of ten integrated studies (nine university investigations and a BLM in-house study) designed to investigate all components of the ecosystem. Research efforts are supervised and directed by the BLM's Project Leader. This report presents results of the third year of this four year integrated research project. The information gathered from each of the component studies will be incorporated into a descrip- tive ecosystems model to describe the interrelationship of the raptors (and mammalian predators) to prey and prey to habitat. This information will provide the basis of establishing a new permanent boundary for the Snake River Birds of Prey Natural Area and will allow managers to estimate the impact of land use changes. A final report will be submitted to the Secretary of the Interior by June 1 , 1979 . TABLE OF CONTENTS Study Title Investigators Page I REPRODUCTIVE PERFORM- ANCE, FOOD HABITS, AND POPULATION DYNAMICS OF RAPTORS. Michael N. Kochert (Project Leader), Karen Steenhof and Albert R. Baumann BLM 1 III ACTIVITY, HUNTING PATTERNS, TERRITORI- ALITY, AND SOCIAL INTERACTIONS OF BIRDS OF PREY. Dr. Thomas C. Dunstan, James Harper, and Ken Phipps Western Illinois University 40 IVA DENSITY AND SPECIES COMPOSITION OF SMALL MAMMALS AND BIRDS. Dr. Michael L. Wolfe and Larry Oftedahl Utah State University 54 IVA RAPTOR UTILIZATION OF DESERT HABITATS. Dr. Michael Wolfe and Karen Steenhof Utah State University 82 IVAS ECOLOGY OF REPTILES IN THE SNAKE RIVER BIRDS OF PREY NATURAL AREA. Dr. Donald R. Johnson and Lowell Diller Dept, of Biology University of Idaho 102 IVB DYNAMICS OF NON- AVIAN PREDATION UPON THE RAPTOR PREY BASE (BADGER SEGMENT) . Dr. Maurice Hornocker and John P. Messick Idaho Cooperative Wildlife Research Unit University of Idaho 122 IVB DYNAMICS OF NON- AVIAN PREDATION UPON THE RAPTOR PREY BASE (COYOTE SEGMENT) . Dr. Maurice Hornocker and Steve Cherry Idaho Cooperative Wildlife Research Unit University of Idaho 130 V POPULATION ECOLOGY AND HABITAT REQUIRE- MENTS OF TOWNSEND GROUND SQUIRRELS. Dr. Donald R. Johnson and Graham Smith Dept, of Biological Science University of Idaho 142 VII FEEDING ACTIVITY AND BEHAVIOR OF PRAIRIE FALCONS . Dr. Steven R. Peterson and Gayle Sitter Dept, of Wildlife Resources University of Idaho 164 VIIIA BIOENERGETICS OF NESTING GOLDEN EAGLES. Michael W. Collopy University of Michigan 178 SPECIAL PLAGUE IN THE SNAKE REPORT RIVER BIRDS OF PREY STUDY AREA Contribution from Studies IVB and V 186 Study I surveys located 57^ pairs of raptors and ravens in the BPSA. Study I efforts were expanded in 1977 to include observations of behavior and feeding rates at selected raven, red-tailed havk. , and golden eagle sites. (BLM photo by Steenhof). 1 STUDY I: Reproductive performance, food habits and population dynamics of raptors in the Snake River Birds of Prey Natural Area. INVESTIGATORS: Michael N. Kochert, Principal Investigator. Albert R. Bammann, Assistant Research Biologist. Karen Steenhof, Analytical Biologist. John H. Doremus, Research Technician. Michael DeLate, Research Technician. John Oakley, Visiting Professor. Tom Hamer, Student Intern. PROJECT SUPPORT: BLM "in house" study. OBJECTIVES: 1. To annually determine absolute nesting density and reproductive performance of the major nesting raptors. 2. To annually ascertain quantitative food habits of the major nesting raptors and to attempt to estimate absolute food intake for select species. 3. To ascertain seasonal species composition, population densities, age compositions and mortality rates of raptors within the study area. ANNUAL SUMMARY Nesting surveys located 574 occupied raptor and raven sites within the BPSA. Prairie falcons were the most abundant species, comprising 205 of the pairs. The BPNA hosted more birds of prey than any other river stretch of similar length studied. Ledges and cavities were the preferred types of nesting structures. Based on several productivity parameters, we estimated that 21 eagles, 442 prairie falcons, 73 red-tailed hawks, and 321 ravens were raised in the BPSA in 1977. Blind observations suggested that the number of nest visits by adult birds was 2.9 per day for eagles, 10.7 per day for red-tailed hawks, and 64.2 per day for ravens. We banded 313 raptors and ravens and wing marked 107. Band returns and wing mark sightings have provided insight into raptor movements. Three prairie falcons have been recovered more than 1000 km from the place of banding. Most eagles and ravens have been recovered within 200 km of the marking location. At least three eagles, two prairie falcons, one raven, and one red-tailed hawk marked as nestlings, have returned to the area to breed. Eagles and ravens remained on breeding sites in autumn. The number of eagles counted on aerial transects in southwestern Idaho was lower than in previous years, but the fall survey indicated a slight upward trend. METHODS Nesting Density end Site Occupancy _ Q n con jyscenaric searches for breeding raptors began in January and iuec throughout the breeding season. We searched all potential :g cliffs within the Birds of Prey Stud}7 Area (BPSA) , with enphasis on the Birds of Prey Natural Area (BPNA) (Fig. 1) . ~.~eyed portions of the BPSA omitted in previous years, re's "1971, golden eagle study area comprising sections of lake River and its tributaries outside the BPSA from Marsing liss, Idaho "Fig. 2 ) was surveyed several times during the .ling season. Surveys were conducted on foot and by boat. WTe .ted occupied sites on the basis of territorial activity, tship , brood-rearing activity, presence of young, or conspicuous .d sign fresh white wash or new nesting material) . Most site itioms from Swan Falls to Halverson Lake were provided by y VII personnel. We considered a "traditional site" as an area of cliff or a group of nests used in this or in previous years by a mated pair of raptors. A site was called "vacant" if no adults were seen during the breeding season, or if an adult with no site tenacity was observed. A ''one bird site'' was a site where we repeatedly observed a lone adult bird but no breeding attempt. An "occupied site" was any site where we observed reproductive related activity by one or more adult raptors (a "pair") at any tine during the breeding season. Presence of a mated pair of birds or field sign indicating occupancy7 by a pair was sufficient to establish occu- pancy for red-tailed hawk and golden eagle sites. Nest Site Selection We described nest sites of breeding pairs of raptors when the enact location of a breeding attempt was identified. We categorized cliff sites as: a "cavity" if it was enclosed on five sides and was as deep or deeper than either wide or tall; an alcove if it was an indentation in the cliff face either wider or taller than deep and not open at the top; an "open book" if it was a vertical indentation in the cliff with sharp sides and was taller than either wide or deep, and open at the top; a "vertical crack" if it was a vertical fissure in the cliff and as long and as deep as wide so as to form a chimney, and either open or closed above; a "horizontal crack" if it was a horizontal fissure in the cliff, completely overhung, often between rock strata, and usually as long and as deep as tall; a "ledge" if it was a horizontal extension or plane from the cliff face, and usually exposed and open above; a "pinnacle" if it was the top of an isolated rock or outcropping of cliff. 3 '££) Location of the Snake Liven Binds of Pney Study Anea (BPSA) showing the Natunal Anea ( BPilA ) and the 19TT expansion study anea. k s Fig. 2. Location of the Snake River Birds of Prey Study Area and other areas surveyed in 1976. 5 Aspect, distance from nest to cliff top, and distance from nest to ground were recorded at a limited number of sites. Reproductive Performance Most potentially active nest sites were entered and examined for eggs, eggshell fragments, or other signs of reproductive activity. Nests were not visited during inclement weather or immediately after hatching to prevent nest desertion or temperature stress to eggs or young. Pairs that occupied sites but showed no evidence of laying an egg (after repeated observations) were categorized as "non-breeding. A "breeding attempt" was confirmed if an occupied site contained an incubating adult, eggs, young, or any field sign that indicated eggs were laid. A "successful nesting attempt" was a breeding attempt that produced one or more young that reached fledging age (i.e. the age at which most young are capable of flight). Active golden eagle (Aquila chrvsaetos) and (Suteo spp.) nests discovered after young had fledged were considered successful if the nest platform was well worn with sufficient prey remains and fecal matter, and if no dead young birds were found within a 50 m radius of the nest. We considered renesting attempts as separate new attempts in calculating productivity and chronology data . Because of the large number of nesting raptor pairs, it was impossible to collect complete information at all sites. Conse- quently, sample sizes for each productivity parameter differ. We tried to keep sample sizes for clutch size as low as possible. Percent occupancy, percent breeding, and nest success per pair were tabulated only from traditional sites that were selected for productivity study prior to the breeding season. Nest success per attempt (including number hatched per attempt and number fledged per attempt) were tabulated from the preselected sites and additional sites found before or during the incubation phase. Sites discovered later were most likely successful and would bias our results against non-breeders and breeding attempts that failed early. We estimated the total number of young fledged in the BPSA using parameters derived from BPSA pairs using the following formula: Number of Number % of % of Breeding Number Fledged Young = of x Pairs x Attempts x Per Successful Produced Pairs Breeding Successful Attempt 6 Brood size at hatching and number of hatchlings per attempt were assessed only when broods were less than 10 days old. Other parameters were tabulated from all nests for which appropriate information was available. Egg losses were ascertained from tallies of eggs that dis- appeared from the nest prior to hatching and those that failed to hatch long after the expected time of hatching. When we could not determine if eggs disappeared before or after hatching, we considered them as "missing units." Nestling losses were assessed from direct counts of dead young in the nest and tallies of young that disappeared from the nest prior to fledging. Young found dead out of the nest were categorized as "nestling" if 25 to 50 percent of the primary vane was still in sheath. Dead young were categorized "post-f ledge" if primary vanes were less than 25 percent in sheath or if field sign indicated the young died more than 50 m from the nest. To ascertain post-fledging mortality, we conducted intensive searches for dead prairie falcons, red-tailed hawks, and ravens above and below certain nest sites. Our searches included all known perches of post-fledging young and extended several hundred meters from the nest. Our estimate of falcon post-f ledge mortality rate is probably complete. We may have underestimated raven and red-tailed hawk mortality because young wander from the nest before actually dispersing. Our searches would not have located dead young more than 250 m from the site. We collected dead nestlings and unhatched eggs to determine cause of death. Decomposed birds found in the field were only examined for obvious causes of mortality and destroyed. Fresh carcasses and unhatched eggs were frozen for later examination as in past years. Most specimens have not yet been analyzed. The cause of mortality for missing nestlings often could not be determined unless there was evidence of predation. Factors Influencing Productivity The land use status at each nest was defined by ascertaining the amount of cultivated land within a 5 km radius of the nest. Nests with less than 10 percent of the land cultivated were classified as "range." Those with 10 to 50 percent cultivated were classified as "intermediate," and those with more than 50 percent cultivated were classified as "agricultural." The effect of researcher activity on raptors and ravens was determined by comparing success rates at (1) nests that were entered during incubation, (2) nests that were entered one to two days after hatching, (3) nests that were entered when young were 2 to 10 days old, and (4) nests that were observed only from 7 distances of 100 m or more and where researchers elicited no physical or vocal response from young or adults prior to banding. Breeding Chronology Weight and external morphology (eye, beak, toenail, and or skin color and feather development) of known age birds were used to establish approximate aging criteria for each species. Some nests were inadvertently entered at hatching, and therefore, we obtained some exact hatch dates. We back dated 45 days from known and estimated hatch dates for golden eagles (Beecham 1970) and 30 days for all other species (Bent 1938, Smith and Murphy 1973) to estimate laying dates. Laying dates were assigned only when the age of young was 10 days old or less. Fledging dates were recorded only when time of fledging could be estimated within four days. Dispersal dates were assigned only when we searched the site twice within two weeks of the last observation of the bird. Average laying, hatching, fledging, and dispersal dates were calculated using Julian dates. Food Habits We made systematic (every four days) collections of prey from 7 golden eagle nests, 19 prairie falcon nests, 12 red-tailed hawk nests, 2 ferruginous hawk nests, and 8 raven nests. The nests were selected and stratified according to the amount of land cultivated within a 5 km radius around the nest. Prey remains are being analyzed by Mr. Jackson Whitman of the University of Idaho. Parental Investment and Feeding Rates Golden eagles. We observed eagles in the Con Shea eyrie from a rock and canvas blind, approximately 80 m from the nest. We observed the nest on seven full days and seven partial days when the young was 20 to 23 days old, 41 to 43 days old and 57 to 59 days old. Only full days of observation were used in the analysis. We observed post-fledging activity at the Rapids eyrie on two days . Ravens . We observed two raven nest sites on ten days from dawn to dusk and on four partial days. Nest A (Spring Draw) was observed on seven full days, and Nest B (Sun Goddess) was observed on three full and four partial days. Blinds were constructed of rock and canvas and were located 40 m from Nest A and 80 m from Nest B. Six different observers recorded observations. Both nests were located on cliffs. Three young hatched on 17 April 1977 at Nest B, and four young hatched on 20 April 1977 at Nest A. Our observations began at Site A when nestlings were 13 days old and terminated at fledging. Site B was watched only when young were 23 to 35 days of age. 8 At Site A, both adults were observed during incubation. We assumed that the bird incubating most of the time was the female (Harlow 1922, Bent 1946). She could be distinguished from her mate because the male had several primaries missing. Because of differences in observer note taking, all 14 days could not be used in the analysis. We used 10 observation days in calculating total visits per day, 6 observation days in calculating visit duration, 7 observation days in calculating the relative roles of the sexes, and 14 observation days in calculating number of young fed per visit. Rea-tailed hawks. We observed red-tailed hawks at the Cabin Draw eyrie from a canvas blind approximately 30 m from the nest. We observed the eyrie on three full days and two partial days when the four young were 19 to 23 days of age. We also watched the Cabin Draw eyrie intermittently 10 days later. We observed activity of young and adults during the post-fledging period at two other eyries in the study area on 11 days. Banding and Marking Studies Nestling raptors and ravens (Corvus corax) were banded with U.S. Fish and Wildlife Service leg bands. All birds banded by personnel in Study I, III, and VII were under the BLM Master Station Permit, and data are pooled. We measured the foot pad on all banded golden eagles (Aquila chrysaetos) , red-tailed hawks (Buteo j amaicensis) , ravens, and prairie falcons (Falco mexicanus) . We placed patagial wing markers on eagles, red-tailed hawks, ravens and a few prairie falcons. Markers were constructed from Herculite 80 (Herculite Protective Fabrics, 1107 Broadway, New York 10010). Prairie falcon, red-tailed hawk and raven markers were modified considerably from those described by Kochert et al. 1975. The shape of wraparound markers for hawks and ravens was modified from the shape of the original eagle marker (Fig. 3). We did not use wraparound markers for prairie falcons this year because of reports that they caused abrasion. Instead, we used a small herculite patch attached to a teflon ribbon (Bally Ribbon Mills, Bally, Pennsylvania) which is wrapped around the wing in a manner similar to the other markers (Fig. 3). The remiges and/or retrices of some individual raptors and ravens were painted with an acrylic base fabric paint (Vogart) for specific temporary identification. Colored leg bands (Ward 1976) were placed on some adult and juvenile prairie falcons. Sightings of marked birds were reported by study personnel and others in response to requests for information sent out in previous years. 9 GOLDEN EAGLE Fig. 3. Shapes and relative sizes of wing eagles, red-tailed hawks , ravens, markers placed on golden and prairie falcons in 1977- 10 Fall and Winter Raptor Populations Golden eagle aerial transects were flown in January and October over 17,920 km^ (7,000 mi“) of the Snake River floodplain (Fig. 4) following the procedures of Boeker and Bolen (1972) and Kochert et al. (1975). RESULTS Nesting Density and Site Occupancy A total of 574 occupied raven and raptor sites were located within the BPSA in 1977 (Table 1) . Prairie falcons were the most abundant species again this year, comprising 205 of the pairs. Ravens comprised 120 of the pairs and red-tailed hawks accounted for 63 of the occupied sites. We counted 43 kestrel (Falco sparverius) pairs, 40 barn owl (Tyto alba) pairs, and 30 golden eagle pairs in the BPSA. Other species breeding in the BPSA in lesser numbers were marsh hawks (Circus cyaneus) , ferruginous hawks (Buteo regalis) , screech owls (Otus asio) , burrowing owls (Speotyto cunicularia) , great horned owls (Bubo virginianus) , and long-eared owls (Asio otus) . At least one Swainson's hawk (Buteo swainsoni) and one turkey vulture (Cathartes aura) also nested within the BPSA. Again this year, the BPNA hosted more nesting raptors than any similar stretch of the river studied. The cliffs along the 50 km stretch of river between Halverson Lake and Black Butte had nearly twice as many raptors as any other 50 km stretch. Birds of prey were most dense between Halverson Lake and Wild Horse Butte. Raptor density upstream from Hammett was considerably lower than in the BPSA. Golden eagle numbers in the BPSA declined slightly from 1976 because two traditional eagle sites in the BPSA were vacant in 1977. Prairie falcon pairs counted in the BPSA have increased steadily since 1975. Despite the increase, seven preselected traditional prairie falcon sites were vacant this year. The number of ferruginous hawk pairs located in 1977 was the same as the number located in 1976, despite the fact that four traditional ferruginous hawk sites were vacant. Both ravens and red-tailed hawks showed sharp increases since 1976. Only two traditional red-tailed hawk and three traditional raven sites were vacant in 1977. Increases in observed numbers of kestrels and barn owls since 1975 are probably a result of more intensive searches this year . Nest Site Selection Breeding ravens and raptors nested in trees, poles, powerline structures, artificial platforms, burrows, roadside pits, on the ir Fig. k. Location of the golden eagle aerial survey with transect lines numbered. Table 1. Number of occupied raven and raptor sites located in the BPNA and the original BPSA, 1975-77 12 uO o co CO CO CNI rH o --o CO 00 H O H d OJ o 3= - d o 0) OO d d W ! 1 d d d o rH d d > X co CD •H i — 1 •H x o X £ pd d •H CU •H d d QJ 00 CO 6 u QJ £ Po d •H d QJ -u d d X d u QJ 1 O QJ o O •H rd i 4— ) d CO ■H d d QJ OO d Hi £ 0) d i — i x) CO u d d QJ d d d d d £ a d o QJ QJ QJ d d d O o d d O CO Pi o X pi S CO O ea CO HI PQ H O 13 ground and in seven types of cliff configurations (Table 2). More than 92 percent of the breeding attempts were on cliffs. Forty percent were on cliff ledges, and 27 percent were in cliff cavities. Ravens were the most versatile in site selection, nesting in six types of cliff structures and three types of non-cliff structures. Alcoves were used most frequently by ravens. Golden eagles and red-tailed hawks used ledges more than other types of nest sites. Prairie falcons, barn owls, and great horned owls seemed to prefer cavities for nesting. Ferruginous hawks nested on both the cliff and the ground, but no ferruginous hawks nested on the artificial platforms in the study area. One pair of ferruginous hawks nested in an artificial structure outside the study area (Smithey, personal communication) . Most nests that were measured were selected because of their accessibility, and our results are probably biased in favor of smaller cliffs. Two golden eagle cliff nests used in 1977 averaged 9.5 m above the ground and 6.5 m below the cliff top. Eleven prairie falcon eyries were an average of 8.7 m above the ground and 5.1 m below the cliff top. Average height of eight active red-tailed hawk nests was 8.1 m, and average distance to cliff top was 5.3 m. Ravens and ferruginous hawks seemed to select nests on smaller cliffs. Average heights of the two species' nests were 5.9 m and 5.7 m respectively. The seven raven nests were an average of 3 m from the cliff top, and the three ferruginous hawk nests were an average of only 2 m below the cliff top. Three barn owl nests averaged 2.3 m below the cliff top and 4.3 m above the ground. Great horned owl nests had the highest average height of all species: 14.7 m. These three nests, however, averaged only 1.3 m below the cliff top. Reproductive Performance Golden eagles. Eighty-four percent of 51 pairs that occupied sites were known to lay eggs. Non-breeding pairs comprised 16 percent of the population in 1977. Average clutch size of golden eagles was 1.9 (Table 3). Hatchability at sites where both clutch size and number hatched was known was 62 percent. Six cases of egg losses were recorded. One disappeared from the nest prior to the expected hatching date, and five failed to hatch for as yet undetermined reasons. The number of hatchlings per breeding attempt was 1.5 (Table 3). Among nests that hatched young, the mean brood size at hatching was 1.8. Nestling survivorship at 17 nests was 58 percent. Sixteen cases of nestling losses were recorded. Three young died of disease, and one young died of heat prostration. Seven young Open Vertical Horizontal Powerline Artificial Roadside lb Total 80 32 8 13 22 120 Table 3. Reproductive performance of raptors and ravens in the BPSA and surrounding areas, 1977 (sample size in parenthesis). 15 co a u rH O PM p •H E 4-4 4-i (1) CO 5-4 4-4 CO 6-5 6-5 6-5 6-5 6-5 O 4-4 Cl) rH ON NO a c CJ r-> NO CO o a 5-1 4-4 P P-I o co p rH o co P •H 5-4 4-4 4-4 -H CO 5-4 cd CO 6^ 6-5 6-5 6-5 6-5 O Ph a) CNJ CO NO co Ph o LO UO CO NO O 4-1 o 5-4 O p P-4 CO 1 — 1 P 4-4 X CO 4-4 u cu co Ph CD oo a) e x a CD /•— s nD /-n UO /— N £ a) o 4-4 . r\ • co . CM • \ • CO P rH P 4-i i — 1 CM co rH 6-5 /-N 6-5 6-5 6-5 4-4 > 00 r-. CO 00 LT) 00 CO ,—s On CO 5-4 uO i — | NO CO LO rH CM LO NO 00 Cl) P v—' £5 CO PC CD 4-» P oo x cd •H 03 O ,P 00 /-V O /-n O ^-n NO 00 5-i o a) o • • CM • o • /— N N CD 5-4 N 4-> i — 1 i — 1 PQ -H cd ■ — ✓ 'w' c CO PC 4-» CO Ph 00 E p CD •H 4-4 m ^ r>* /-v I 1 \ CM o i — 1 4-1 • t — i • r-- • O 5-4 CD 0) cd o PM P4 Pm P^ 16 died of unknown causes in or near the nest, and five disappeared from the nest prior to the expected time of fledging. Fifty-four percent of preselected eagle pairs produced young. Seventy-one percent of all breeding pairs found early in the season produced fledging young. Of all eggs known to have been laid, 47 percent developed into fledging age young. Eagles fledged 0.9 young per pair, 1.2 young per attempt, and 1.7 young per successful attempt (Table 3). We observed the first three chick brood at fledging in this population since 1971. Eagles were estimated to fledge 21 young in the BPSA this year. Percent breeding, percent of breeding attempts successful, and number fledged per successful attempt were all lower in the BPSA than in other areas studied. Three young golden eagles were known to have died during the post-fledging period. One was shot and two died of unknown causes. Prairie falcons. Eighty-seven percent of 23 preselected pairs occupying sites were known to lay eggs. Non-breeding pairs comprised 13 percent of the population in 1977. Average clutch size of prairie falcons was 4.7 (Table 3). Hatchability at sites where both clutch size and number hatched was known was 87 percent. Twenty-three cases of egg losses were recorded. Five eggs disappeared from the nest prior to expected time of hatching. One failed to hatch because of early embryonic mortality, and 17 failed to hatch because of as yet undetermined causes . The number of hatchlings per prairie falcon breeding attempt was 3.7 (Table 3). Among nests that hatched young, the mean brood size at hatching was 4.0. Nestling survivorship at 38 nests was 63 percent. Sixty cases of nestling losses were recorded. Five young were lost to predators, and two nestlings died of heat prostration. Four nestling deaths were attributed to abandonment, and one was attributed to cannibalism. Ten young died in the nest from undetermined causes, and 38 disappeared from the nest prior to the expected time of fledging. Fifty-two percent of preselected prairie falcon pairs fledged young. Sixty-seven percent of all breeding pairs found early in the season produced fledging young. Of all eggs known to have been laid, 52 percent developed into fledging age young. Prairie falcons fledged 2.0 young per pair, 2.4 young per attempt, and 3.7 young per successful attempt (Table 3). Prairie falcons in the BPSA fledged an estimated 442 young this year. Six young prairie falcons were known to have died during the post- fledging period. Three were killed by predators, and three died of unknown causes. The post-fledging mortality rate at 15 intensively studied nests was 11 percent. IT Red-tailed hawks. Eighty-seven percent of 31 preselected pairs occupying sites were known to lay eggs. Non-breeding pairs comprised 13 percent of the population in 1977. Average clutch size of red-tailed hawks was 4. 7 (Table 3) . Hatchability at sites where both clutch size and number hatched was known was 60 percent. Fourteen cases of egg losses were recorded. Seven were lost to predation. Two eggs were destroyed by adverse weather. Two disappeared from the nest, and three failed to hatch because of causes as yet unidentified. The number of hatchling red-tailed hawks per breeding attempt was 2.1 (Table 3). Among nests that hatched young, the mean brood size at hatching was 3.0. Nestling survivorship at 18 nests was 55 percent. Thirty-two cases of nestling losses were recorded. Five losses were attributed to human disturbance, and thirteen young died of heat prostration. Four young red-tailed hawks disappeared from the nest, and 10 died of unknown causes. Forty-eight percent of preselected pairs fledged young. Forty-nine percent of all breeding pairs found early in the season fledged young. Of all eggs known to have been laid, 25 percent developed into fledging age young. Red-tailed hawks fledged 1.2 young per pair, 1.2 young per attempt, and 2.6 young per successful attempt (Table 3). We estimate that red-tailed hawks in the BPSA fledged 73 young in 1977. The post-fledging survival rate at five intensively studied sites was 100 percent. Ferruginous hawks. Of 16 traditional sites occupied by ferruginous hawks, one was occupied by a single, unmated bird and one was occupied by a non-breeding pair. Twelve pairs were known to lay eggs, and breeding was not ascertained at two occupied sites . Average clutch size of ferruginous hawks was 3.9 (Table 3). Hatchability at sites where both clutch size and number hatched was known was 44 percent. Nine cases of egg losses were recorded. Six eggs disappeared from the nest prior to the expected time of hatching. Three failed to hatch for unidentified reasons. The number of hatchlings per breeding attempt was 2.2 (Table 3). Among nests that hatched young, the mean brood size at hatching was 2.6. Nestling survivorship at five nests was 23 percent. Ten cases of nestling losses were recorded; all of these young dis- appeared from the nest prior to the expected time of fledging. 18 Thirty-six percent of preselected pairs and 36 percent of all breeding pairs found early in the season fledged young. Of all eggs known to have been laid, 11 percent developed into fledging age young. Ferruginous hawks fledged 0.3 young per pair, 0.3 young per attempt, and 1.5 young per successful attempt (Table 3). Owls . One great horned owl nest site had a clutch size of two. We knew of at least one great horned owl egg that failed to hatch. Of 21 great horned owl breeding attempts in which the outcome was known, 19 were successful and 2 were unsuccessful. We were more likely to find successful pairs than unsuccessful ones since our intensive searches began late in the horned owl nesting season. Four of the successful nestings fledged an average of 3 young each. Two barn owl nests had an average of 5.5 young shortly after hatching. All of these young survived to fledge. In all, seven successful barn owl nestings were confirmed. One of the four long-eared owl sites located was known to be successful, and five of the seven screech owl pairs fledged young. At least one screech owl died during the post-fledging period. Eight successful burrowing owl attempts were confirmed. One nestling burrowing owl was shot to death. Other raptors. Seven successful kestrel nest sites were located. One fledged two young. We found one kestrel egg that failed to hatch for unknown reasons, and we discovered one dead nestling. Three successful marsh hawk sites were located. One of these was known to fledge three young. We found one unhatched marsh hawk egg. The single turkey vulture nest found had two young at the time of hatching. Both nestlings survived to fledging. Ravens. Ninety-five percent of 22 preselected pairs occupying sites were known to lay eggs. Non-breeding pairs comprised at least five percent of the population in 1977. Average clutch size of ravens was 5.3 (Table 3). Hatchability at sites where both clutch size and number hatched was known was 54 percent. Nineteen cases of egg losses were recorded. Eleven failed to hatch because of as yet undetermined causes. Eight disappeared from the nest prior to the expected hatching date. The number of hatchlings per breeding attempt was 3.0 (Table 2). Among nests that hatched young, the mean brood size at hatching 19 was 4.8. Nestling survivorship at 8 nests was 69 percent. Sixteen cases of nestling losses were recorded. One young raven died after the death of one of its parents. Another died as a result of human disturbance. Thirteen nestlings disappeared from the nest prior to the expected time of fledging, and one died in the nest of unknown causes. Sixty-three percent of preselected pairs fledged young. Seventy-four percent of all breeding pairs found early in the season produced fledging young. Of all eggs known to have been laid, 33 percent developed into fledging age young. Ravens fledged 2.1 young per pair, 2.4 young per attempt, and 3.7 young per successful attempt (Table 3). Ravens in the BPSA fledged an estimated 321 young in 1977. Two young ravens died during the post-fledging period of unknown causes. One young raven starved to death during the post-fledging period. The post-fledging mortality rate at ten intensively studied sites was six percent. Factors Influencing Productivity Land use. Eighty-three percent of the "range" eagle nests were successful, 53 percent of the "intermediate" nests were successful, and only 50 percent of the "agricultural" nests were successful. Sample size was small, however, and a Chi-square test showed the results were not significantly different (p>.05). We expect more definitive results when we pool several years of data. Seventy percent of prairie falcons in range areas were success- ful and 69 percent in intermediate areas were successful. The status of lands within a 5 km radius of prairie falcon nests may not be as critical to reproduction since prairie falcons hunt farther than 5 km from their nests (Dunstan et al. 1976) . Red-tailed hawks were successful approximately 50 percent of the time in all land use situations. Ravens were 73 and 77 percent successful in range and intermediate situations respectively. The single raven nest in an agricultural area was unsuccessful. All three ferruginous hawks in range areas were unsuccessful, and three of the six in intermediate types were unsuccessful. Timing of researcher visits. The number of young golden eagles fledged per breeding attempt was highest at sites that were not visited until young were of banding age (Table 4) . Number fledged per attempt was lowest at nests visited one to two days after hatching, and it was also low at nests visited during incu- bation (Table 4) . Nests visited when young were two to ten days old produced fewer fledglings than nests visited after young were 10 days old (Table 4) . A Kruskal Wallis test showed that none of the differences were statistically significant (.10 CQ • CO • i — i rH CNl 5-i i — 1 QJ •H > 4-J QJ d 2 do 20 • CN co CN! 00 qj 5-4 TO O > — ! g o TO CO 0) 4-1 CO •H C CO •H O t> i— I 4- > C CO CO 5- i -d •H H p4 TO 0J 4-1 •H CO > 4- ) CO 5- i •H 2 TO i — 1 o CO Q TO QJ TO 4-1 i — I •H c CO •H CO > On cO 4-1 Q CO 5-4 CNl *[— 1 1 2 r — I O -~N • m o ^ 00 r-N • LO CN| w d o TO -H QJ 4-1 4-) CO •H JZ CO d •H O > e i— i 4-1 CO 00 V4 d •H -H p4 5-4 d Q CT\ S N • O O i-H CNl /—N • CO CNl CN) CO /" N • o O CNl Nt O Pi d cO o cO dO QJ u do i — 1 1 — 1 CO OO CO T) d cO 2 QJ o w i — 1 d QJ •H •rl co •H cO 00 QJ 5-i 4-1 d n TO ■rH 1 5-4 QJ i — 1 CO TO 5-4 > o 5-1 QJ QJ cO a Cd Pi P4 Pi 21 Red-tailed hawk productivity showed the same pattern with the fewest young produced per attempt at sites visited one to two days after hatching (Table 4) . Red-tailed hawk sites visited during incubation also had lower production than nests visited later in the brood-rearing period. The differences in productivity were significant, according to a Kruskal Wallis test (p<.05). An independent test comparing all red-tailed hawk nests visited before young were two days old vs all nests visited later or not at all showed that the first group's productivity was significantly lower (p< . 05) . There was no apparent relationship between the timing of researcher visits and prairie falcon productivity. Small sample size and low reproductive success for ferruginous hawks prevented meaningful generalization (Table 4) . Similarly, sample size for ravens visited during brood-rearing was too small to draw conclu- sions (Table 4) . The productivity of visited nests shown in Table 4 may be inflated because of the inclusion of food habits study nests and manipulated nests. These sites sometimes received special efforts to protect young. Some food habits study sites were selected because the adults had been accustomed to investigators in past years. Young were treated for disease at three golden eagle nests and one prairie falcon nest. Shade devices were placed in five eagle nests, six red-tailed hawk nests, two prairie falcon scrapes, and one ferruginous hawk nest. Food habits study nests, manipulated nests, and nests with shade devices all produced a higher average number of young per attempt than the pooled sample. Breeding Chronology Golden eagles were first observed in courtship activities on 9 February. Newly decorated nests were seen as early as 14 January. Eagles started laying eggs on 12 February and continued to lay until 19 April. Mean laying date was 6 March (Fig. 5). Mean hatching date was 17 April but some eagle clutches hatched as late as 17 May (Fig. 5). Young eagles fledged from 1 June to 18 July. The mean fledging date was 19 June (Fig. 5). Prairie falcons were observed patrolling cliff territories as early as 2 February. Laying dates extended from 19 March to 28 April and averaged on 6 April (Fig. 5). Most falcons hatched between 24 April and 15 May, with 4 May as the mean hatching date (Fig. 5). Mean falcon fledging date was 10 June, but falcons fledged from 26 May to 4 July (Fig. 5). Average date of dispersal of young from nine sites was 7 July. One brood dispersed from the natal site as late as 21 July. Red-tailed hawks were involved in territorial activity in February. The first newly decorated nest was noted on 4 March. Prairie 22 D c o u ■a _a> 5 *U 0) 0£ J o X m D O c c 0) > 0 O' * LL. o 3 < Z U-) 3 Q£ . LO Q. ^ < C *=. a CO * o & Fig. 5. Breeding chronology of raptors and ravens in the BPSA and adjacent areas, 1977- Laying dates are designated by "L," hatching dates by "H," fledging dates by "F , " and dispersal of young dates by "D." Vertical lines indicate means; horiz- ontal lines show ranges ; and rectangles illustrate one standard deviation on eac side of the mean. 23 Red-tailed hawks laid eggs from 22 March to 24 April. Mean laying date was 4 April (Fig. 5). Average red-tailed hawk hatching date was 1 May. Young hatched as early as 4 April and as late as 24 May (Fig. 5) . Fledging dates ranged from 22 May to 9 July, with the average being 10 June (Fig. 5) . Average date of dispersal of young at five nests was 29 July. Some young stayed at their natal site until 13 August (Fig. 5) . Ferruginous hawks arrived in the area in February, and the first newly decorated nest was seen on 21 March. Four ferruginous hawk pairs laid eggs from 1 April to 20 April (Fig. 5) . Five broods hatched from 1 May to 20 May, with the average hatch date falling on 13 May (Fig. 5) . Fledge dates were known for only three sites. Mean fledging date was 25 June, but young fledged as early as 11 June and as late as 9 July (Fig. 5) . We were able to estimate hatching dates for three barn owl broods. Average date of hatching was 30 April (Fig. 5). One great horned owl brood was estimated to hatch on 15 March. One barn owl brood fledged on 8 June; a great horned owl brood fledged on 3 May. We observed ravens involved in territorial activity in February. Nest building was first observed on 2 March. The first known clutch was laid on 8 March. Mean laying date was 30 March, and some ravens laid eggs as late as 10 May (Fig. 5) . Raven broods hatched from 7 April to 17 May. Average hatching date was 26 April. Young ravens fledged from nests as early as 20 May and as late as 18 June, with an average fledge date of 3 June (Fig. 5). Raven young dispersed from territories somewhat earlier than the falconi- forms. Average dispersal date was 23 June. Some ravens departed from natal sites as early as 4 June and some left as late as 30 July. Management Experiments Fostering. Three separate fostering experiments were carried out in 1977. The first effort was in cooperation with the Peregrine Fund. Five young prairie falcons were removed from their BPNA eyrie on 8 May 1977 to make room for three peregrine (Falco peregrinus) falcon chicks that were flown in from the Fort Collins facility. Details of the successful cross-fostering reintroduction program are in Konkel 1977. Four of the five "orphan" prairie falcons were placed in a prairie falcon nest with one young. All young died in the nest, 10 to 14 days after the introduction, apparently due to great horned owl predation. The fifth young was placed in a nest with two other nestlings. All three of these nestlings successfully fledged. 2b The second fostering attempt involved a young golden eagle with a diseased eye. In late May, a nestling was found in the Castle Butte eyrie with a growth on its left eye. The nestling was taken to Dr. John Lee, DVM of Meridian, Idaho. Dr. Lee removed both the growth and the eyeball and recommended that the bird be returned to the wild. We chose not to return the eagle to the Castle Butte eyrie because the nest is "hot" with a south facing exposure. We were also uncertain if the parents would resume feeding the young after its three day absence. We decided instead to put the eagle in the Glenns Ferry eyrie where there was already one chick, approximately 25 days younger than the Castle Butte eyas. This would allow the foster bird more time to fledge and would not subject him to aggressive acts from a bird his size. Both the foster young and the original young fledged from the Glenns Ferry eyrie in June. Parental Investment and Feeding Rates Golden eagles. Adults visited the Con Shea eyrie an average of 2.9 times per day during the observation period (Table 5). Parents fed young during 50 percent of the visits; the young fed itself when it was 57 and 58 days old (Table 5) . Prey was delivered once a day except on 30 May when no prey deliveries were recorded (Table 5). All prey brought to the nest were lagomorphs ; most were identified as black-tailed jackrabbits (Lepus calif ornicus) . Prey deliveries were recorded at nearly all times of day from 0743 to 2145 hours. We were unable to detect any consistent pattern in the time of day that prey was delivered. Average visit duration was 62.8 minutes. Visits averaged 65 minutes when young were three to four weeks old, 78 minutes when young were 6 to 7 weeks old, and only 12 minutes when young were 8.5 to 9 weeks old. Our tentative identifications of sex suggested that the female visited the nest approximately three times as often as the male. Only the female fed young and rearranged the nest, but both adults delivered prey. Ravens . Adult ravens visited the nests an average of 64.2 times per day (Table 6). Adults delivered food to young during 79 percent (509 of 642) of the visits and cleaned the nest during 34 percent (219 of 642) of the visits. Adult ravens entered the nests to feed young an average of 50.9 times per day. The number of feeding visits per day ranged from 39 to 72, and there was no apparent relation between feeding frequency and time of day or stage of brood-rearing. Table 5. Parental Investment in relation to age of young golden eagles (full observation days only). 25 a: 03 03 CD CD •H d 5-4 p CD 2 > 03 <4-J *H Q p o rH rH 1 1 rH o rH rH i — i 00 CD d • d Q CD o OJ P P 2 e 03 P P 5-i P CD CD O'- CN CO P T — l CM • 2 d CM 4-1 > 2 03 03 o3 cd 03 G 03 P P P 'TJ p TD P p 00 r— 1 co i — i Csl CO P m Csl -> 2 Po 2 r r d 00 03 s 03 cti 03 d d 03 S 2 S 2 S p p d CD 04 rH 04 o rH i — 1 Csl CO CO i — i i — i c Table 6. Daily number of nest visits and feeding frequency in relation to age of common raven nestlings in Idaho. 26 "X /-v CO 4-1 kO • uo O 03 LiO 1 — 1 bO 0 00 • Cb] 03 rH CO rH , rC 03 *H 03 •H 4-J •H > 4-J > • • <3 a "O a3 4-> >4-1 O U c CD CD U CD •H CD QJ -a co O (3 n3 4-J 4-1 (3 4-J cd 4-J U Q) t — 1 CD bO •H •H > •H £ •H U S-> c 5-4 • rH 0) CD CD •H CD CD 3 (D a) 4-J a3 O -43 •H •H 1 — 1 •H CD •H O &. C rO •H > > (3 > a3 > O O e CD CD bO 03 £ O a) 3 •H • • (3 a 4-1 4-1 4H 4H 0 co c > PQ 03 •H O O CD O ■u O bO •rt 4-4 0 bO cd CD C 4-1 1 — 1 a) U (3 rH T3 u U £ u (3 U •H 3 bO (3 4-J O 0) (3 0) (3 C (3 T3 1 — I t3 03 PL •H a i — 1 rQ 03 rO O co cd S-4 CO M PM £ a 0 a (3 a O 4-4 a) C3 •X -X ■K 3 0 0 3 X M O >> > (3 •X ■vC 53 53 4-1 4-J -£) H -D C L4-4 -3< CD 03 03 03 •H 4-J u 03 CL 27 An average of 1.5 young were fed at each feeding visit. Site A ravens fed an average of 1.42 of their four young at each feeding visit, but Site B ravens fed an average of 1.3 of their three young at each feeding visit. The number of young fed per visit at Site A declined through the nesting season (Table 6) . Males and females fed similar numbers of young during each visit. On 18 and 30 May when individual young could be distinguished, it was apparent that some young were fed more often than others. However, a one degree of freedom Chi-square test showed that the differences in proportion were not greater than what would be expected by chance (p>.05). Brooding occurred in only 42 of the observed visits (Table 6) . Brooding was observed when young were 15 days old but not after young were 23 days old (Table 6) . On 6 of the 10 observation days, there was a bimodal peak in the number of nest visits. Nest visits were most numerous from dawn to 1100 and 1600 to dark. Visits tended to be less frequent between 1100 and 1600 hours (Table 7). The average visit duration was 1.9 minutes. Approximately 48 percent of the nest visits (152 of 314) were completed within 15 seconds. The total time in the nest spent by adults averaged 100.3 minutes for each day of the brood-rearing period (Table 6). Overall nest attentiveness declined generally during the brood-rearing period (Table 6) , but parental investment seemed to increase when young started to fledge. The number, of nest visits ranged from 102 per day when nestlings were 13 days old to 36 per day when nestlings were 35 days old. The decline in visit frequency can be attributed primarily to a decrease in the number of visits solely for brooding or nest maintenance. The number of feeding visits was more constant throughout the nesting period (Table 6) . The total amount of time per day spent by adults on the nest ranged from 422 minutes when the young were 13 days old to 8 minutes when young were 20 days of age (Table 6) . The adults spent 30 minutes on the nest on the day that young started to fledge. This behavioral change may account for the abrupt increase of parental investment. Adults visited the nest with four young an average of 71 times per day while the three young nest averaged only 49 visits per day. A Chi-square test with one degree of freedom showed this difference to be statistically significant (p<.05), but the difference was probably an artifact of our sampling schedule. The nest with three young was not observed early in the brood-rearing period when nest visits tend to be more frequent. There was no significant difference (p>.05) when we compared only observation days when young were between 22 and 36 days of age. 28 Table 7. Number of nest visits by adult ravens in relation to time of day Site A Dawn - 1100 - 1100 1600 1600 - Dark Total 3 May 36 26 40 102 4 May 30 23 36 89 5 May 29 28 33 90 17 May 25 24 26 75 18 May 15 8 18 41 29 May 17 20 16 53 30 May 19 15 11 45 Site B 10 May 17 7 24 48 11 May 28 21 14 63 22 May 14 11 11 36 230 183 229 642 x = 23.0 x = 18.3 x - 22.9 x = 64.: (sd = 7.6) (sd = 7.6) (sd - 10.6) (sd = 23 29 Only the female brooded the young. The male made the same number of feeding visits as the female during brooding (Table 8) . After brooding and before fledging, the female made most of the feeding visits. Food deliveries by the male, however, increased at the time of fledging (Table 8) . The female did most of the nest maintenance throughout, the nesting season. Of 183 nest maintenance visits in which sex was identified, 131 (72 percent) were made by the female (Table 8). Most nest maintenance by the male consisted merely of carrying off fecal sacs, casts, or discarded food. Additionally, the female probed and fluffed the nest. The female at Site A made more visits to the nest than the male on each observation day until the final day when one young fledged and three young branched. On the final observation day, the male made almost all of the nest visits. On 4 , 17, 18 and 29 May when sexes of adults were distinguished and visit durations recorded, the mean duration of the Site A female's visits exceeded the mean duration of the male's visits. T-tests showed that the difference was significant only on 17 May. We were unable to quantify the amount of food delivered to raven nestlings. Adults regurgitated food to young and most food could not be identified. Observers identified three confirmed and five possible deliveries of Townsend ground squirrels (Spermophilus townsendi) . Adults delivered remains of desert horned lizards (Phrynosoma platyrhinos) three times, side-blotched lizards (Uta stansburiana) once, unidentified lizards six times, and snakes 15 times. Observers identified grasshoppers in 11 visits and unspeci- fied insects in three visits. Corn was delivered twice, deer mice (Peromyscus spp) were brought four times, and unidentified rodents were listed twice. At least four eggs were brought to the nest including one chukar (Alector is graeca) egg and one black-billed magpie (Pica pica) egg. Most prey could be described only by texture and color. Red meat was delivered in 72 visits. Our blind observations indicated that collection of prey remains is a poor method of assessing food habits. We watched adult ravens removing both fecal sacs and cast pellets from the nests. The high frequency of red meat carrion in the diet makes if difficult to quantify predation rates by conventional means. The results suggest that the amount of parental investment varies according to several factors. The female apparently spends more time and energy on nest attentiveness than does the male. Both adults reduce nest attentiveness as the brood-rearing season advances. Time of day also influences the amount of nest atten- tiveness with less energy invested during midday. These variations 30 Table 8. Type of nest duties performed by both adult ravens during three stages of brood rearing at Site A. 3-5 17-18 29-30 May May May Male Female Male Female Male Female Total visits per day. 37.3 55.7 13.5 26.5 19.5 11.0 Feeding visits per day. 31.3 31.3 12.5 26.5 18.0 9.0 Nest maintenance visits per day. 15.0 36.0 0.5 9.0 3.0 2.0 Brooding visits per day. 0.0 14.0 0.0 0.0 0.0 0.0 Average number of visits dawn-1100 . 12.3 17.7 5.0 10.0 5.5 2.0 Average number of visits 1100-1600. 11.0 14.7 3.0 7.5 9.0 5.0 Average number of visits 1600-dark. 13.3 23.0 7.0 10.0 5.0 5.0 Average visit duration (seconds) . 177.3 503.8 11.6 40.7 14.1 60.8 31 should be noted by future researchers as they attempt to develop time activity budgets for ravens and other passerines. Red-tailed hawks. Our observations at the Cabin Draw eyrie suggested that adult red-tailed hawks delivered prey to the nest an average of six times per day (Table 9). The pair delivered four complete kangaroo rats (Dipodomys spp.), one partial kangaroo rat and a whipsnake (Masticophis taeniatus) on one day. Two kangaroo rats, three snakes, a lizard and a cottontail rabbit (Sylvilagus spp.) were delivered on the second day. On six occasions both members of the pair arrived at the nest simultaneously. One bird (presumably the male) delivered prey and left after only a few seconds. The other (presumably the female) fed the young on the nest for several minutes. An average of 3.5 of the 4 young were fed at each feeding visit (Table 9) . An adult shaded young during 7 of the 32 visits (22 percent) , and the nest material was rearranged once. The average visit duration was 14.5 minutes. Adults spent a combined average of 154 minutes per day on the nest (Table 9) . Average number of nest visits per day was 10.7. Of 18 prey deliveries, 11 occurred between 1000 and 1800, three occurred between 0600 and 1000 and four occurred between 1800 and 2200. During our observations on 4 June, the four young jumped out of the nest prematurely when the heat of the nest became unbearable. Three of the young died of impact injuries. We were able to rescue the fourth young and place him in a more shaded portion of the cliff. The adults continued to feed the remaining young at his new location, and the young survived to fledge. Banding and Marking Studies We banded 313 raptors and ravens in 1977, bringing the total number banded in the Kochert (1972) Study Area since 1966 to 1527 (Table 10) . This includes 200 golden eagles and 336 prairie falcons banded by Idaho Cooperative Wildlife Research Unit personnel from 1968 to 1972 and 39 eagles banded by U.S. Fish and Wildlife Service personnel from 1966 to 1968. 107 raptors and ravens were wing marked in 1977, bringing the total number wing marked since 1970 to 507 (Table 10) . This total includes 93 golden eagles wing marked by the Idaho Cooperative Wildlife Research Unit in 1970 and 1971. A total of 64 band returns and 151 wing mark sightings of raptors and ravens marked in the area of study have been reported to this office (Fig. 6). Of the band returns, 26 were in the official BPSA, 12 were in adjacent areas of study, and only 13 were more than 50 km from the banding area. Seventy-three of the wing mark sightings were in the BPSA, 19 were in adjacent areas studied by Research Project, and 22 were more than 50 km from the banding area. 32 Table 9. Parental activity at a red-tailed hawk nest when nestlings were 20 to 22 days old. 24 May 25 May 26 May Total visits to nest. 11 11 10 Total prey deliveries. 7 7 4 Number of visits in which young were fed. 4 6 4 Average number of young fed per feeding visit. 3.3 3.3 4.0 Number of visits in which nest was maintained. 0 1 0 Number of visits in which shading or brooding occurred. 0 3 4 Total "tandem" visits.* 2 2 2 Total minutes spent by adults on nest (daylight only) . 54.7 238.8 168. 3 Average visit duration (minutes) . 5.0 21.7 16.8 *Visits in which both the male and female landed on the nest at approximately the same time. 33 03 X CO cu xl qj X 5-i CO O E r- I O-N 6o i — i d •H 0) £ O C i — I *H CO co r-- to XXX i — ! i — I I I o H I X I I — I I X r"- o m w d o ■H > qj 5-i (X x d cO x X CT> X OJ X X d cO i — I X kO CO 00 0) X LO CO 1 — 1 o ("X X 1 — 1 cO c X •H o CO H X o X PM i — I i — i X x CM U~) CO cO n- cO x a> d X X ■H d CO CO CL O x — x *H X X CM CM in cm x d co m > pc; co w ^ c 0) o x X E i — I 5-i t: cO d O X H O * Includes birds banded and marked by Study III, Study VII, the Idaho Cooperative Wildlife Research Unit and the U.S. Fish and Wildlife Service. 3b Prairie falcons. Reports received so far indicate that prairie falcons travel farther than the other species studied. Three prairie falcons were recovered more than 1000 km from the place of banding. A yearling banded in 1971 was recovered near Ness City, Kansas in 1972 (Fig. 6). A falcon banded as a nestling in 1972 was recovered four months later in Mexico (Fig. 6) . Early in 1978, we received a report of a two year old wing marked female prairie falcon that died of heart rupture near Phoenix, Arizona (Fig. 6). Other falcons banded as nestlings in the BPSA and adjacent locations have been recovered near Ballantine, Montana, Pocatello, Idaho, and Aurora, Utah. Wing marked prairie falcons have been seen near Moscow, Idaho and in southwestern Montana (Fig. 6). At least three prairie falcons banded as nestlings in the BPSA have returned to occupy breeding territories in the area. Two females banded in 1972 by Verland Ogden of the Idaho Cooperative Wildlife Research Unit were trapped at BPSA eyries and fitted with radios in 1976. One of the birds was known to have bred in 1975. Both falcons successfully fledged young in 1976, but failed to return in 1977. One of the birds was known to have died in 1976. In 1977, a male prairie falcon, wing marked in 1976, occupied a site outside the BPSA that had been vacant since 1971. Although the bird was mated with an adult female, they made no breeding attempt . An adult female prairie falcon banded in January 1976 by Dr. Charles Henny, U.S. Fish and Wildlife Service, was re-trapped in spring 1977 by Study III personnel on the same territory. The falcon was radioed in 1977, and it successfully raised young. Golden eagles. Wing marked golden eagles have been sighted as far as 600 km from the place of banding. The most distant sightings were in south central Utah in winter 1975 (Fig. 6). Wing marked eagles have also been sighted in southwestern Montana, south central Washington, east central Oregon and northwest Nevada (Fig. 6) . All golden eagle band returns have been within 200 km of the banding locations. Four eagles marked as nestlings have returned to the area to occupy breeding territories. Two males marked in 1971 occupied traditional sites in 1975. The male at the Con Shea site did not attempt to breed in 1975. It was seen copulating in 1976, but no eggs were laid. This marked bird was last seen in late summer 1976. Blind observations in 1977 confirmed that there was a new unmarked male in 1977. The marked male at the Indian Cove site made an unsuccessful breeding attempt in 1975 and raised one young in 1976. It was trapped and radioed in 1977. The 1977 attempt produced two young that died prior to fledging. The marked male was still on territory at the end of 1977. This bird set up a breeding territory 6.8 km from its natal nest. 35 Band Returns Winq-mark Sightings Golden Eagle Prairie Falcon Raven Fig. 6. Band return and wing mark sightings of raptors and ravens handed as nestlings in the BPSA and adjacent areas of study (Map shows only those recoveries 50 km or more from the area of study). 36 In 1977, a male marked in either 1970 or 1971, bred at the Pole 369 eyrie. The male and its unmarked mate successfully raised one young in 1977 and have remained on the territory all year . A female marked as a nestling in 1970 began to occupy a territory in fall 1977 after the death of the resident female (see section on Fall and Winter Populations) . It was still on territory at the time of writing. Red-tailed hawks. All red-tailed hawk band return and wing mark sightings to date have been within 50 km of the banding locations . At least one red-tailed hawk marked as a nestling has returned to the area to breed. A female marked as a nestling in the BPSA in 1975 was observed calling territory in 1976 at the "Deadman" site which is in the Kochert (1972) Study Area but outside the BPSA. She was in immature plumage and no other bird was seen. In 1977, she was mated with an adult male and raised three chicks at the same site. j Ravens . A raven recovered in eastern Oregon was the only raven band return farther than 50 km from the banding location. Wing marked ravens were sighted in Nevada and central Idaho. In 1977, a two year old female wing marked raven laid two eggs in a BPSA tree nest. The breeding attempt was unsuccessful and the marked bird has not been seen subsequently. Fall and Winter Raptor Populations The number of eagles counted on the aerial transects in February 1977 was the lowest recorded on mid-winter counts since the survey began in 1973 (Table 11) . The percent of immatures seen was also lower than any previous mid-winter count. Despite the decrease in eagles, the number of other raptors seen was higher than any previous survey. The low eagle numbers may resulted from the drought. Most of the higher elevations lacked snow during most of the winter of 1976-77. It appears many of the winter migrants stayed at high altitudes and never arrived on the survey area. The October aerial transect survey showed an increase in eagle numbers since the fall 1976 count. There were still, however, no confirmed sightings of immature in the fall count. Our observacions in the fall showed that both ravens and golden eagles occupied breeding sites in the BPSA during the fall months . Table 11. Results of aerial transect sampling on 17,920 knT (7000 mi ) of the Snake River floodplain, 1972-77. 37 0) 4-1 o CO cu • 4-1 o o o o sO CO 5-i U a) o -C 4-) 4-1 cx O cc5 Pi sO LO| as CN oo cn NfOOsOOssOCslCNCsl i — I i — ! cNcsicoasasNi-com OOcONt-rHsOLnOssO 04 •H e OCsl o g t 1 CO a) 1 — I 00 ' cc! • w CN Csl CO LO nt sO 00 sO co Csl CN LO Csl i — i CO T 1 Csl 5 1 'w' Cs| CN O o • C o • CO ' CO CO sO CO Csl co 04 CO LO| o Dunstan et al. falcons were monitored by fuU-dfvT H hawks - aad prairie days. Full-day samples began prior tT o" apPTxl“ately every 8 sunrise and continued untif “ft™ 0 1 t"“ before °fficlal The non-breeding adults at the n t6S past official sunset, full-day samples approximately ever" ^days'" ™re ™n*tored on breeding adult was monitored irregularly with f n a* * ^ approximately every 12 to ?n a 8 T y th fuH~day samples Shorter duration ^ da- of more than one bird or pair of n-fi-u aiVlduals- At some sites, A 4-wheel drive International g f ”3S monitored simultaneously, antenna mast for use Tt” LIT T^Y^Ta 3 r0tati"S used in 1976 was again used in 1977 rf wheel.drive D°dge pickup made using techniques described by Dunstan" YY (jgYY Monitoring Non-instrumented Birds were “f168 ““a ada° studied. Individuals various stages of ^“a««istics (color variations, site or other family members YiLY"" aasoclatlon wlth a nesting in golden eagles. Data were ' gathered "at "the181” W3S . distinguishable instrumented birds. 8 d h same lntervals as for Instrumentation of Subjects r.vSr:L:;s;r s™. ■■ Subjects that returned with transmitters f l . St3n 1976). recaptured. Data on 78 e n • ters functronmg were not u urc u. oara on zo subiects studied in 1077 Table 1. studied m 1977 are presented in Data Analysis 45 Table 1. Data on 28 subjects studied in 1977. Species and I.D. Band Instru- Last location code Sex Age number mented contact Golden Eagle Coyote Butte CB F Imm 629-02639 Coyote Butte CB M Ad 629-02649 4 Feb 7 Apr Beecham BC F Ad 629-02643 26 Mar 31 Dec Beecham BFa M Ad 629-02644 27 Mar 29 Mar Beecham BC M Ad 31 Dec Cabin Draw CD F Ad 629-02640 8 Feb 20 Nov Cabin Floater CF F Ad 629-02642 15 Feb 10 Aug Cabin Draw CD M Ad 31 Dec Black Butte BB M Ad 31 Dec Black Butte BB F Ad 31 Dec Feedlot FL M Ad 629-02641 10 Feb 12 May^ Feedlot FL F Ad 629-02602 12 Mayc 18 Jul1 Indian Cove CV M Ad 599-05339d 7 Apr 31 Dec Indian Cove CV F Ad 31 Dec Red-tailed Hawk Halverson I H M Ad 987-23223 3 May 13 Dec1 Halverson I H F Ad 877-32805 3 May 31 Dec Cabin Draw C M Ad 987-23225 20 May 26 Jul( Cabin Draw C F Ad 877-37806 12 May 5 Aug' Black Butte B M Ad 987-16304 13 Junc 3 Aug Black Butte B F Ad 19 May 2 Aug Indian Cove V M Ad 987-23224 18 May 7 Jun Indian Cove V F Ad 23 Jul Prairie Falcon Trail T M Ad 666-84006 27 Apr 14 Jul Trail T F Ad 987-23221 27 Apr 27 Jun Camera C M Ad 836-77006 28 Apr 27 May Camera C F Ad 987-23222 28 Apr 19 Jun Indian Cove V M Ad 816-24701 17 May 23 Jul Indian Cove V F Ad 987-16306 11 May 22 Jun a Nonresident b Abandoned site after 31 March; signal later heard outside of BPNA boundaries. c Banded and instrumented in 1976. d Band and wing marker applied in 1971 when a nestling by personnel from Coop. Wildl. Res. Unit, Univ. of Idaho, e Later found dead. f Study I, personal communication. k6 data were grouped into periods (incubation, brood-rearing, fledging) to express the changing of range sizes and daily rhythms with progression of the breeding season. RESULTS Status of Data Analysis for 1977 Season The information presented in this report is only part of the total data gathered during the 1977 field season. At the present, we are completing field work on wintering subjects, grouping data from 1975 and 1976, and writing the Final Report for Study III. Therefore, in agreement with the Contracting Officer and the Contracting Officer's Authorized Representative, we have included only the 1977 data analyzed to date. Additional information from 1977 will be presented in the first Quarterly Report of 1978, and in the Final Report. Golden Eagle Data Instrumentation. Eight eagles (seven adults and one immature) were live trapped, instrumented with radio transmitters and/or banded in 1977 (Table 1) . Three of the adults and the immature did not breed within the study area. The other four adults were resident birds. Although our goal was to instrument breeding adults, there was no way of determining which eagles were floaters at the time of capture. All four radioed birds had mates, but the pairs at Cabin Draw and Feedlot did not breed. Horizontal Space Use. Horizontal space use was determined for eagles breeding at the Beecham, Cabin Draw, Black Butte, Feedlot, and Indian Cove sites. The configurations of the ranges for the pairs are shown in Figs. 1 and 2. The sizes of the ranges of individual birds and/or the ranges of the pairs are presented in Table 2. The Beecham (BC) female had the largest home range during the breeding season of 1977 (Table 2) . This bird used both the north and south side of the canyon and flew and hunted over an area of 62.6 km“ (Fig. 1). An adult male thought to be the Beecham female's mate was captured and instrumented in the same location as the female on the following morning (Table 1) . But within two days this male disappeared. The female had been incubating or at least in the incubating posture until both the male and the female spent two nights roosting on power line poles away from the nest just prior to capture on 26 March. By 18 April, the female was observed copulating with an unmarked male. The female was consistently observed with a second eagle throughout the breeding season. The Cabin Draw (CD) eagles did not breed in 1977 but had a combined range of 19.8 km^ during the spring and summer (Table 2). Table 2. Monitoring and range data for 1977 breeding season. Species and Days Individual range during breeding Breeding range of location Sex monitored3 season (km^) pair (km^) Golden Eagle B e e cham F 8 62 . 6b — Cabin Draw M 10 13 .0b 19 . 8 Cabin Draw F 10 17 .6b 19 . 8 Black Butte M 5 12.2 12.2 Black Butte F 5 11. A 12.2 Feedlo t M 10 18 .0C 18.0 Feedlo t F 10 18 . 0C 18.0 Indian Cove M 1A 5 . 2 5 . A Indian Cove F 1A A . 7 5. A Red-tailed Hawk Halverson I M 15 10 .9 10.9 Halverson I F 15 2 . 1 10.9 Cabin Draw M 10 20 . 3 22 . A Cabin Draw F 10 16 . 3 22. A Black Butte M 11 19 . 2 19 . 3 ’Black Butte F 5 7 . 1 19 . 3 Indian Cove M 10 5 9 . 5 d 59 . 5 Prairie Falcon Trail M 11 73.7 100 . 6 Trail F 10 69 . 2 100.6 Camera M A 89 . 7 95 . 9 Camera F 5 A 9 . 7 95 . 5 Indian Cove M 10 1 3 A . 9 155 . 3 Indian Cove F 10 30 . 2e 155 . 3 a Does not include partial days. b Non-breeding residents. c Nest building and courtship period only. ^ Only male pair instrumented. e Only 2 flights were tracked out of the canyon in 17 hours of observation before this bird left the BPNA. 48 2 The female had an individual range of 17.6 km , and the adult male (presumably the mate) had an individual range of 13.0 km^ . An adult male was regularly seen with, and courting, the female. These ranges are larger than the 12 km^ range for the pair at this site in 1976 when two young were produced. Although there were no young to tend in 1977, the pair again spent most of the time within the canyon, rather than out north on the flats or on pole lines. The difference in range size may be due to increased efficiency of data gathering from the instrumented female in 1977, as compared to our 1976 data based only on sight observations. This pair used both the north and south sides of the river (Fig. 1). The range of the pair of eagles at Black Butte (BB) was 12.2 km" (Table 2). The range in 1976 was 21 km^ . The difference was because no observations were made of these eagles using the southeastern portion of their 1976 range (Fig. 1). The 1977 range of the pair is also based on observations of a different female than in 1976. The 1977 female tended to remain closer to the nest than did the 1976 female. If we compare the range most often used in 1977 with that^most often used in 1976, the area is about the same (12 to 13 km“). This pair used both the north and south side of the river and shared Jackass Butte with the eagles at the Cabin Draw site. Two young were raised in 1977. The range of the pair of eagles at the Feedlot site during the early portion of the breeding season was about 18 km" (Table 2) . This is the same size as the post-breeding season range of this pair in 1976. The sizes of the individual ranges during this period were also the same. The Feedlot pair courted and built a new nest close to the nest used in 1976. However, at the expected egg laying time, the pair deserted the site. Both eagles were last seen using the site on 31 March, but were absent on 2 April and after. Both eagles were detected via radio signal during the summer but were outside of the BPNA. The reason the site was deserted is not clear. But, the nest draw was walked regularly by two different coyote trappers who trapped the draw and the talus areas. Immediately following the desertion of the site by the Feedlot eagles, a different pair of eagles moved in, and the female of the pair was seen on the nest constructed by the Feedlot birds. The nest was abandoned thereafter, and eggs were found in the nest cup (Study I personal communication). During this period, personnel from Study IVA sampled in the area and were unaware of the presence of the active nest site. One instrumented offspring from the 1976 clutch at this site was seen in the area of Jackass Butte on several occasions in July. The total range used by the Indian Cove (CV) pair during the breeding season was 5.4 km^ , which is the smallest eagle breeding range documented in our studies. The individual range of the female (not instrumented) was 4.7 km", and that of the male was 5.2 km^ . The movements and activities of this pair were easily monitored because of their small range and the presence of adequate roads and high points for tracking. The two nestlings at this site died, presumably from heat prostration (Study I data) at about two and eight weeks of age. The boundaries of the ranges have remained the same through the time of the writing of this report. This pair used both the north and south side of the river (Fig. 2). Hunting at the Indian Cove Site. This pair used a hunting technique not previously described in our reports. The male and female commonly soared and glided together at heights of 2-7 m above the terrain, and 2-10 m apart. When prey was seen, both eagles dove at it, and if the prey was missed by one, it was pursued by the other. If the prey hid among the rocks or vegetation, both eagles landed and walked among the rocks while peering into cavities. If the prey was not found within 5 minutes, the eagles abandoned the search. Similar cases of cooperative hunting were observed for the Black Butte and Feedlot eagles. Most of the hunting was done along the slopes and in gullies within the canyon. While within their range, these eagles had line of sight to the nest and young from a distance of 4.1 km. The configuration of the range for the pair was related to preferred habitat for hunting and perching. The Indian Cove eagles rarely hunted over the islands or along the river shore. They also did not often hunt over the flats above the canyon rim. In 14 days of sampling, 18 kills were seen, and 28 unsuccessful attempts were recorded. The most common prey were rabbits (Sylvilagus sp.) which lived among the rocks on the slopes. Red-tailed Hawk Data Instrumentation. Seven adult hawks (three breeding pairs and the male of a fourth pair) were captured and instrumented after their eggs hatched (Table 1) . Trapping techniques were previously described (Dunstan and Harper 1975; Dunstan et al. 1976). Horizontal Space use. Horizontal space use was determined for all subjects, and shapes of the breeding ranges are shown in Figs. 1 and 2. The sizes of the ranges of individual hawks and/or ranges of the pairs are presented in Table 2. 2 The Halverson I (H) pair had a breeding range of 10.9 km . This was the size of the range most often used and does not include two flights of 15 and 34 km made by the male to the northeast late in the breeding season. The individual ranges of the male and female were 10.9 and 2.1 knW respectively. The area used within the canyon by the male was slightly larger than that used by the female. Both birds hunted the north and south sides of the river, but usually remained within line of sight of the nest and the young. Outside of the canyon, the male frequently made flights up 50 to 5.4 km to the north. The female was not seen flying more than 200 m north of the canyon until after the young fledged. During the fledging period she flew a maximum distance of 1.1 km north of the north rim. This pair raised four young. O The Cabin Draw (C) pair had a breeding range of 22.4 km . Th^ individual ranges of the male and female were 20.3 and 16.3 km respectively. Neither bird was seen flying south of the river. The female spent most of the breeding season within Cabin Draw and along the adjacent slopes. For most of the breeding season the male ranged to the north out of the canyon more often than did the female. Only toward the end of the fledging period did the female spend much time out of the canyon and away from the nest. This pair raised one of four young in the indicated range (Fig- I)- ^ The Lower Black Butte (B) pair had a breeding range of 19.3 km (Fig. 1). Jhe individual ranges of the male and female were 19.2 and 7.1 kin respectively. Both the male and female hunted north and south of the river. The female restricted her movements within line of sight of the nest and young, while the male ranged relatively farther. This nest was exposed to direct sun during the afternoon, and at three weeks of age all four young jumped from the nest and died. The male hawk at the Indian Cove site had an individual range of 59.5 km^ (Fig. 2). The female, which was not instrumented, ranged close to the nest and was usually within line of sight. Her range was much smaller than that of the male based on visual observations. The male hunted to the north, south, east, and west of the nest, but spent most of his time south and across the river from the nest (Fig. 2). The two young of this pair died at three to four weeks of age during a week of high ambient temperature. Hunting at the Halverson 1 Site. The Halverson pair used hunting techniques not described in our previous reports. When catching cliff swallows (Petrochelidon pyrrhonata) that nested in bulb-like mud nests along the cliffs by the hawk nest. Both the male and female used a "plunge" method of flight for taking flying swallows. Typically, the hawk perched on the cliff 2 to 6 m above swallow nests and plunged down with wings partially folded. Swallow milling around in front of the nests were grabbed while flying. If the hawk missed, it perched again, waited, and plunged. Often four to eight plunges were made before prey was captured. Both adults also used a "rake and grab" method for catching swallows inside of the nests. The hawks flew around in front of a colony, and as swallows flew into the nests, the hawks flew to a nest from which a swallow peered and raked and grabbed the nest and the swallow inside. This technique was used as the hawk flew along the cliff, and was not part of the plunge method. 51 On 16 and 29 occasions prey was carried to the young by the female and male respectively. Food deliveries were made to the young at the nest by both the male and the female. The male did not remain at the nest for more than five seconds. The female fed the young. Fourteen feeding sessions by the female for young three to five weeks old, averaged eight minutes in duration and ranged from 2-3 minutes. The maximum number of feeding sessions in a day was six. At this site the male was not seen transferring prey to the female for delivery to the nest or young, nor was the reverse situation noted. Prey was abundant enough at this site to raise four young. Cooperative Effort with Red-tailed Hawks. Study III personnel coordinated tracking activities with the food habit studies of Study I at the Cabin Draw site. Study I personnel provided information on identification of prey brought to the nest by a specific parent. Personnel from Study III reported which mate had killed the prey. Because males at both the Cabin Draw and the Lower Black Butte sites passed prey to females for delivery to the nest, a cooperative effort between those persons observing kills and those persons watching deliveries was necessary. How common this behavior is at other red-tailed hawk nests in the BPNA is unknown. This and similar behavior is documented for golden eagles in the BPNA. Prairie Falcon Data Instrumentation. Six adult falcons (three breeding pairs) were captured and instrumented (Table 1) . The adults were captured during the incubation period or when nestlings were less than one week old. Horizontal Space Use. Horizontal space use for all three pairs was determined. The sizes of the ranges of individual birds and the ranges of pairs are presented in Table 2. The Trail (T) pair had a breeding range of 100.6 kirT (Fig. 1) The individual ranges of the male and female were 73.7 and 69.2 km^ respectively. This pair hunted both in the canyon and to the north, but did not hunt south of the river (Fig. 1) . Preliminary analysis of the data indicate that this pair hunted farther from the canyon in 1977 than did the pair present in 1976. Flights were made to specific areas where Townsend ground squirrels (Spermophilus townsendi) were relatively more abundant (Smith, pers. comm.). Areas of greater squirrel densities were near irrigated farms. The 1976 range was determined by tracking the resident female that dispersed and later died on migration. The 1977 ranges were determined by tracking a new female and her mate. This pair raised four of five young. 52 O The Camera (C) pair had a breeding range of 95.9 km (Fig. 1). The individual ranges of the male and female were 89.7 and 49.7 km~ respectively. This pair hunted north of the river exclusively. The male was found dead close to the scrape on 28 May, and the female abandoned the four young and the nesting effort during the following week. Preliminary data indicate that this male spent more time hunting, and made longer flights in 1977 than in 1976. The range of the male doubled in 1977. As with the Trail pair, the Camera pair also used specific areas of greater squirrel density for hunting. The male at this site died and the female abandoned the brood (Table 1) . 2 The Indian Cove (V) pair had a range of 155.3 km (Fig. 2). The individual ranges of the male and female were 134.9 and 30.2 km- respectively. In contrast to other pairs of falcons studied, the Indian Cove male hunted on both the north and south sides of the river. The female was only tracked on two flights (3 km NE and 18 km N) in 17 hours of monitoring on 4 days before she abandoned the site and disappeared. Therefore, the 30 km2 range is minimal. The male made repeated hunting flights to the south side of the river to distances 14 km from the nest. He also made a flight 16 km to the northwest. Four of the five young at this site died at about three to four weeks of age during a week of high ambient temperature. The remaining fledgling was tended by the male. Cooperative Effort with Falcons. Study III personnel coordi- nated tracking activities with the food habit studies of Study VII personnel at the Trail and Camera sites. Study VII personnel provided information on identification of prey killed and brought to the scrapes for flights tracked. They also provided data on short flights made within the canyon while Study III personnel followed long flights away from the canyon. ACKNOWLEDGMENTS We wish to acknowledge the assistance of the following persons during the 1977 field season. Bill personnel provided assistance with contractual issues and needed data from Study I. Brad James and Gayle Sitter provided data on food habits and assistance during falcon trapping. Roger Olson and Scott Sawby assisted with the trapping of study subjects. Todd Woodcock provided electronic assistance with data analysis. Max Ollieu and the Boise National Forest provided electronic instruments for range map analysis. 53 LITERATURE CITED Brown, J.L. 1969. Territorial behavior and population regulation in birds. Wilson Bull. 81 (3) : 293-329 . Dunstan, T.C. and J.F. Harper. 1975. Activity, hunting patterns, territoriality, and social interactions of birds of prey in the Snake River Birds of Prey Natural Area, Idaho. Pages 51-96 in Snake River Birds of Prey Research Proj . Annu. Rep. 1975. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 193 p. Dunstan, T.C., J.F. Harper and K.B. Phipps. 1976. Activity, hunting patterns, territoriality, and social interactions of birds of prey in the Snake River Birds of Prey Natural Area, Idaho. Pages 63-130 in Snake River Birds of Prey Research Proj. Annu. Rep. 1976. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 240 p. 54 Larry Oftedahl, Research Associate, removes a cottontail rabbit from a trap. Lagomorphs continued to be the primary focus of the four year prey base study (BLM photo by Smith). 55 STUDY IV-A: Density and species composition of Small Mammals and Birds on the Snake River Birds of Prey Natural Area, Idaho. CONTRACTOR: Ecology Center, Utah State University, Logan, Utah 84322 INVESTIGATORS: Michael Wolfe, Principal Investigator Larry 0. Oftedahl, Chief Research Technician Pat Olmstead, Biological Aide Jim Johansen, Biological Aide Anita Garcia, Biological Aide PROJECT SUPPORT: U. S. Department of The Interior Bureau of Land Management, Contract No. 52500-CT5-1003 OBJECTIVES: 1. Stratify the study area into major vegetation types and land- use patterns from existing aerial photographs and other means. 2. Sub-stratify the major vegetation types into basic land-use types, and sample the major prey species by acceptable statis- tical and censusing techniques. 3. Specifically, sample for prey species in those areas identified by radio telemetry efforts (Study III) to be. intensively hunted by golden eagles, prairie falcons, and red-tailed hawks. Aspects of vegetation composition, cover, physiography, and any other factors will be studied to determine the importance of prey vulnerability to predation. 4. Assess and compare prey densities in the different habitat and land uses upon prey densities. 5. Ascertain seasonal and yearly changes in prey species compo- sition and densitites. 6. Calculate annual prey population densities in terms of bio- mass within each strata of the total study area and in areas intensively hunted by raptors. 56 ANNUAL SUMMARY Jackrabbits (Lepus calif ornicus) were censused in major vegetation types in May. Highest densities occurred in the shadscale-winterf at fAtriplex conf ertif olia-Cerratoides lanata) association and big sagebrush (Artemisia tridentata) . Jackrabbit flushing transects were run in spring and fall with the cooperation of Study I personnel. Cottontails (Sylvilagus nuttali) were live-trapped throughout the year at 35 locations, including known golden eagle (Aquila chrysaetos) hunting areas Indices of pheasant (Phasianus colchicus) abundance, obtained in all three years of the current study, were highest in areas of old agricul- ture. Summer snap- trapping of small rodents in 14 habitat types revealed deer mice (Peromyscus maniculatus) and kangaroo rats (Dipodomys ordi) as the most abundant species. Intensive spring trapping of a single marmot (Marmota f laviventris) colony produced an estimated density of 13.5 animals/ha. Attempts to census chukars (Alectoris graeca) by the use of a recorded rally call were marginally successful. An August census yielded an estimated linear pigeon (Columba livia) density of 0.05/m of canyon cliff face. Passerine bird transects were walked in spring and fall. This report contains results of the third field season of a continuing prey-base study. Where appropriate, data from the previous two years of the study have been included in the analysis. For further clarification of the methods employed and more detailed description of the 1975 and 1976 data see Wolfe and Montan (1976) and Wolfe et al.(1977) METHODS Blacktailed Jackrabbit Jackrabbits were censused from 20 May through 10 June using a modified Area-Estimate method (Flinders and Hansen 1973). Four tran- sect routes were established on the study area to census the more common vegetation types (Figures 1-4) . Less common vegetation types were also sampled sporadically within the transects. The vegetation on Transect 1 comprised big sagebrush, mixed sagebrush and winterfat, sagebrush/grass, and an area of farm ecotones. Big sagebrush and sagebrush/winterf at were the dominant vegetation types on Transect 2, while shadscale and shadscale/winterf at were predominant on Transect 3. Greasewood (Sarcobatus vermiculatus ) comprised the major vegetation type on Transect 4. Each jackrabbit transect was censused on three consecutive nights with the aid of a hand-held spotlight (100 watt, 12 volt) by the same observer in the back of a pickup truck. Counting began at 2230 and continued for an average of 3.5 hours each night. The vehicle was driven 16 km/hr. and the observer watched both sides of the transect. When an animal was observed, species, vegetation type, time, sighting distance and perpendicular distance from the pickup were recorded. Effective transect width was established at 50 m on both sides of the pickup creating a 100 m wide transect. In each vegetation type sampled a density correction factor was derived to obtain absolute density estimates. This correction factor accounted for sighting variability due to vegetation composition and structure differences. 57 Fig. 1. Jackrabbit spotlight Transect #1, 1977* Each square is 1.6 km on a side. ; r< 58 \ V 36 ■RW X 12 : > 12 y / 2& V 21 ..J* ) $ Bf / \ Fig. 2. Jackrabbit Spotlight Transect # 2, 1977- Each square is 1.6 km on a side. 59 Jackrahtit Spotlight Transect #3, 1977. Each square is 1.6 km on a side. Fig. 3. 6o Guffey tutte WE Lt — -*■ -^301 9 MURPHY 26 Murphy / MURPHY / MUNICIPAL WATER TANK/ V\ Fig. h. Jackrabbit Spotlight Transect #k , 1911. (Remaining portion is on the next page). Each square is 1.6 km on a side. 6l Fig. 4 (Continued). Second portion of Transect #4. 62 Many vegetation types and mixes were sampled. Nine of the types sampled were considered to have sufficient transect length to be reliable enough for testing. These subjective judgements were substantiated by evaluation of the variance estimates for the samples obtained from respective vegetation types. An analysis of variance test was used to test for mean population differences among vegetation types. The Newman-Keuls range test (Hicks 1973) was used to make aposteriori tests to establish differences between mean population densities in different vegetation types or transects. Statistical differences were established at the 95 percent level of probability. In May and June, 50 jackrabbits were collected at night. Eye lenses were removed and preserved in 10 percent formalin. In November the lenses were dried and weighed. The weights were then converted to age groups based on the criteria of Tiemeier and Plenert (1964) . Flushing transects, previously established by the Bureau of Land Management in spring and fall, were walked with the cooperation of Study I Personnel. Cottontail Rabbit Cottontails were live-trapped in 35 different areas throughout the spring, summer, and late fall. Three specific areas with relatively high cottontail densities were trapped every season to record any detectable intra-annual population changes and among the three years of trapping. Many sites trapped in 1977 were attempts to confirm very low population numbers. Some sites trapped in July and August were known golden eagle hunting areas. These trapping efforts were conducted to test whether there were significant prey base differences in these hunting areas compared to sites of comparable vegetation elsewhere on the study area. The trapping results (animals caught/100 trap nights) were subjected to a two-way analysis of variance with the main factors being habitat type and season within year. The three habitat types considered over the three years of the study were: (1) canyon riparian; (2) canyon talus; and (3) canyon rim sites. Where significant differences (p <_ 0.05) existed among treatments means, Tukeys ' Honestly Significant Difference test (Tukey 1949) was used to test for possible differences between individual treatment means. "T-tests" were used to determine if signifi- cant differences in indices of cottontail abundance existed between the golden eagle hunting area sites and the non-hunting sites as defined from Study III. 63 Pheasants Indices of pheasant abundance were obtained by the crow-count method (Kimball 1949). Crow-count routes were driven between 14 and 28 April. Each route was begun one-half hour before sunrise and counts were made every 1.6 km for a duration of two minutes. The routes used were the same as in the previous two years of the study (Wolfe and Montan 1975, Wolfe et al. 1976) in an attempt to determine possible differences in pheasant abundance between areas of old and new agriculture. Old agricultural areas were defined as areas which had older farms and fence row vegetation present. New agriculture areas were defined as modern "clean" farming areas with single fields typically covering 200 ha or more and with little or no vegetative cover present along fence rows. Additionally, older agricultural areas commonly contain irrigation ditches and associated cover, whereas irrigation in the newer agricul- tural areas is typically accomplished by means of overhead sprinkler systems. Four of the routes were located in old agriculture, three in new agriculture, and one in a new agriculture area with a large riparian area through its center, referred to hereafter as the new agriculture- riparian area. Two measures of pheasant abundance were obtained: number of calls/ station and number of birds seen/kilometer. In 1975 only the latter index was obtained. A one-way analysis of variance was used to test observed differ- ences in the number of calls /station for 1976 and 1977 data and the number of birds seen/kilometer for the 1975, 1976, and 1977 samples. Where significant differences (p 0.05) existed among treatments, linear combination tests on the means (Hicks 1973) were used to establish differences between treatment means or groups of means. Rodents In July small rodents were snap-trapped in 14 different habitat types. Snap- trap lines of 25 traps each were run for two consecutive nights at each site. Applicable vegetation and snap-trap data from the 1975 and 1976 field seasons were included in the analysis. A one-way analysis of variance was used to test observed differences in indices of rodent abundance between vegetation types by species with data pooled for the three years. The Newman Keuls Range Test (Hicks 1973) and Tukey’s Honestly Significant Difference test were used to establish aposteriori differences between vegetation types by rodent species . A two-way analysis of variance test for each year was performed to test for index differences among vegetation types with data pooled for each species, and to test for index differences among species with 64 data pooled by vegetation type. The Newman Keuls Range test and Tukey’s Honestly Difference test were used to establish differences between species and between vegetation types among years. Marmots were live trapped between 22 February and 16 March along a 0.25-km section of cliff at Evans Ranch, Melba. Traps were baited with lettuce and checked in early morning and afternoon. Live-trapping provided a density estimate at a "typical" colony. Marmots were also live- trapped while trapping cottontails. These captures were considered incidental, since the traps were not ideally located for marmot trapping. Bushy tail woodrats (Neotoma cinerea) were sampled under the same sampling regime and method as cottontail rabbits. A two-way analysis of variance was performed on the number of animals caught-100 trap nights with the two main factors being habitat type and season within year for the three years of the study. The three habitat types con- sidered in analysis were: (1) canyon rim; (2) canyon riparian: and (3) canyon talus. Ghukar partridge Chukar partridge distribution was determined during August by floating a section of the Snake River from the Swan Falls dam down- stream to Evans Ranch and driving or walking transects in other areas. Recorded chukar rally calls were played at 1.6 km intervals. Each transect was run on two consecutive days. The number of responses during a 10-minute period is theoretically correlated with flock size (Oeklaus 1976 ) . Pigeons From 16 to 18 August a visual census of pigeons was conducted along the south cliff of the Snake River Canyon beginning 2 km downstream of Swan Falls dam and continuing to the Halverson Lakes area. Total transect length was 6.8 km. Along this portion of the Snake River the cliff rises vertically 80-120 m. A transect along the base of the cliff was selected and driven one hour after sunrise. Every 0.4 km a firearm was discharged, causing the pigeon flocks to fly off of the cliff face. A visual estimate was made of the flock size and photographs were taken to determine the accuracy of visual estimates. 65 Visual estimates of flock sizes were compared to the photographic counts of the same flocks and a linear regression model (Snedecor and Cochran 1967) was fit to this data to correct for visual estimate errors. The average number of pigeons seen each day was then calculated and a corrected density estimate was made by substitution into the linear regression model. Passerines From 15 April through 25 May, 13 passerine transects were sampled. Transects were square in shape, and 0.4 km on a side. They were selected in representative vegetation types. Passerines were also censused in conjunction with collection of data on the B.L.M. jack- rabbit transects in spring and fall. These transects were identical in shape and size to the primary passerine transects. Passerines were grouped into three size classes: horned larks (Eremophila alpestris) , meadowlarks (Sturnella neglecta) , and sparrows. Other incidental species not easily classed were noted separately. Vegetation A plant survey of the Snake River Birds of Prey study area north of the Snake River was conducted during April and May. Vegetative typing was aided by use of color aerial photographs taken in July 1976. Plant densities (coverage) and composition were determined for each plant type by means of ocular reconnaisance (B.L.M. Manual 4412.11 A) as the primary method. The accuracy of this technique was validated on approximately every fifth transect by use of the line intercept method. RESULTS Blacktailed Jackrabbits There were significantly more jackrabbits (F = 4.12, df = 8/17, p = 0.008) in the shadscale-winterf at complex and big sagebrush than in any other vegetation types. This is only considering the areas underlined in Table 1. Other areas show higher estimates, but they also have very short transect lengths, and as noted previously these estimates are not considered reliable. This is illustrated by the differences between density estimates for the sagebrush-winterf at vegetation type in Transects 1, 2, and 3 (Table 1) . The density estimate in Transect 3, which was 0.8 km long, is more than 10 times greater than the average of the densities in Transects 1 and 2, which were 10.5 and 11.7 km in length, respectively. Only subjective estimates of densities can be obtained from transect lengths less than 4.0 km. 66 Table 1. Jackrabbit density estimates, obtained by the area- estimate method on the Snake River BPNA, May 1977. Vegetation Type Transect Length (km) Mean Jackrabbit Density (Numb er /kin ) Transect 1 37.7 Sagebrush 10.7 20.2 Sagebrush-Winter fat Sagebrush-Grass 10.5 7.6 Patches 4.4 13.6 "Farm/Mixed" Veg. 9.2 2.8 Winterfat 1.6 0.0 Transect 2 28.6 Sagebrush-Winter fat 11.6 12.6 Sagebrush Sagebrush-Grass 4.0 31.3 Patches 1.8 36.7 Sagebrush-Budsage Sagebrush-Buds age/ 1.9 58.7 Burn 2.5 2.6 Grass (Burn) 1.9 17.6 Transect 3 28.1 Shadscale Shadscale- 14.6 19.2 Winterfat 7.4 36.7 Sagebrush-Winter fat 0.8 105.7 Winterfat 2.8 2.7 Sagebrush 1.7 12.0 Transect 4 12.3 Dense Greasewood Greasewood/Dry 6.4 0.0 Creek 3.1 41.4 Greasewood- Budsage 2.8 169.7 Only those vegetation types (and associated data) underlined were included in statistical analysis; see text for explanation. The results of B.L.M. flushing transects (Table 2) indicate an increasing jackrabbit population. In 1976 the fall estimate was 48.6 compared to 57.9 j ackrabbit s/knr in 1977. Results from this technique have shown that there may be serious problems with these estimates (Wolfe et al. 1976) and with the method itself at low population densities. Lens weight data indicate 66 percent of the jackrabbits collected were young-of-the-year (Age Class I) , 32 percent were yearlings (Age Class II) , and two percent were two years old or more (Age Class III) (Tiemeier and Plenert 1964). The overall sex ratio was 1.1 males: 1.0 females. The sex ratio among juveniles was 1.4: 1.0 (n=33) , while yearlings showed a sex ratio of 0.8: 1.0 (n=17); only one female adult was collected. The observed sex ratios among juveniles and yearlings do not represent significant departures from unity and are probably attributable to sampling error. Thirty percent of the yearling and adult females had embryos past half term (x = 2.3). Many other yearling and adult females were lactating. None of the juvenile females were pregnant or lactating. Cottontail rabbits Most of the live-trapping sites located near rocky areas contained cottontails (Table 3) . Sites located in native vegetation without rocky sites or lush cover nearby supported few animals. In three of four instances cottontails showed no apparent increase in ecotones between new agriculture and native vegetation. The only case where an increase of the cottontail index occurred was in a sagebrush/f arm ecotone . There was no significant difference among seasons and years (F = 17.14, df = 2/16, p>0.1) but there was a significant difference (F = 1.52, df = 8/16, p<0.05) among the three vegetation types tested. The canyon talus and canyon riparian sites both had significantly higher indices of abundance than the canyon rim. These three sites were near Swan Falls. Canyon talus sites in golden eagle hunting areas did not have significantly different indices of abundance (t = 1.07, df = 4, p>0.05) than canyon talus elsewhere. There was also no difference between eagle hunting areas and non-hunting areas in habitat composed of shadscale and big sagebrush with rocky areas. Pheasants In April 1977 the mean number of calls/station was 7.03 in old agriculture areas compared to 0.94 in new agriculture areas; the 68 Table 2. Jackrabbit population indices obtained during BLM jackrabbit transect sampling. Snake River Birds of Prey natural area, 1972-1977. Year Season Number of Transects Number Jackrabbits Observed X Index (N) 2 Jackrabbits /km 1972 Fall 26 9 43.8 62.6 1973 Fall 36 6 19.0 27.2 1974 Spring 34 4 12.3 17.6 1974 Fall 59 12 31.2 44 . o 1975 Spring 60 7 12.5 17.9 1975 Fall 60 11 19.4 27.7 1976 Spring 61 21 57.7 82.5 1976 Fall 61 18 34.0 48.6 1977 Spring 59 24 30.6 43.9 1977 Fall 60 22 40.5 57.9 Table 3. cottontail, marmot, and bushytail woodrat live-trapping results 1977. explanation see Appendix 1.) (For animal code Location Description Month of Number Caught/100 Trap Nights Trapping S .n. N.c. M.f . Other (89.60,32.40) Sagebrush-Winterf at Feb . _ _ _ _ (92.80,24.30) Rock Cliff/Farm Feb. - 19.0 8.7 - (88.70,34.05) Sagebrush-Winterf at /Farm Feb. - - - - (93,45,29.15) Sagebrush Feb . 0.9 - - 3.7 L.c. (93,05,29.40) Sagebrush/Farm Feb. 4.2 - - 2.1 L.c. (89, 55, 30. 'J 5) Canyon Rim (1) Mar . 3.0 2.0 1.0 - (87.25,32.10) Canyon Talus Mar. 4.6 - 6.4 - (91.00,30.25) Canyon Riparian Mar . 8.3 8.3 - - (94.75,26.30) Grass Mar . - - - - (95.35,26.40) Grass/Farra Mar. - - - - (94.60,28.20) Winter fat Mar . - - - - (94.60,28.02) Winterfat/Farm Mar . - - - - (92.40,32.55) Trio Butte (Rocky Hill) Mar . 3.1 2.3 - 2.3 L.c. (77.50,43.70) Shadscale Mar . - - - - (76.90,44.70) Shad seal e-Win terf at May - - - 0.5 L.c, (75.50,46.55) Canyon Rim (2) May 8.3 1.2 - 1.2 L.c, (74.80,34.50) Greasevood May - - - 0.6 L.c, (95.60,20.40) Greasewood/Farm May 1.9 - - - (90.60,29.50) Canyon Rim (4) June 3.3 1.7 - - (89.55,30.95) Canyon Rim (1) June 6.7 1.7 - - (87.30,32.40) Canyon Rim (5) June 3.3 1.7 - 0.8 S.p (87.25,32.10) Canyon Talus June 12.1 - 3.0 - (91.00,30.25) Canyon Riparian June 14.0 4.0 2.0 1.0 E.d (93.40,29.05) Rocky Hill-Sagebrush July 2.0 - - 4.0 L.c (94.50,29.35) Rocky Hill-Sagebrush July 4.0 1.0 - - (94.35,29.60) Rocky Hill-Sagebrush July 3.0 - - 2.0 L.c. (91.00,30.50) Sagebrush (Sandy)/Farm July - - - - (91.80,26.80) Canyon Rim (H.L.) July 2.0 10.0 - - (80.10,42.80) Rocky Hill-Shadscale July 3.0 - - 1.0 L.c. (79.85,43.10) Rocky Hill-Shadscale July 3.0 - - 5.0 L.c, (78.20,42.60) Rocky Hill-Shadscale July 2.0 - - 4.0 L.c. (77.50,42.00) Dry Creek Bed-Rocky July 3.0 - - 2.0 L.c. (77.50,41.25) Canyon Rim (Desolate) July - 4.0 - 2.0 L.c. (67.00,74.50) Hilly Talus Aug. 2.0 6.0 - - (66.95,74.50) Canyon Rim (I.C.) Aug. Aug. 13.0 7.0 ~ 1.0 E.d. 1.0 M.m. (66.40,75.50) Talus (1) Aug. 7.0 5.0 - - (66.70,75.20) Talus (2) Aug. 7.0 2.0 - - (67.05,75.25) Talus (3) Aug. 7.0 5.0 - - (89.55,30.95) Canyon Rim (1) Dec . 1.1 3.4 - 4.0 N.l. (91.00,30.25) Canyon Rim Dec . 11.4 6.8 - - (87.25,32.10) Canyon Talus Dec. 12.5 - - - TO number of birds seen/km was 0.35 and 0.013 in old and new agriculture, respectively (Table 4). Analysis of 1975. 1976, and 1977 crow-count data revealed significant differences between treatments (Table 5) . The data generally show much higher pheasant responses in old agricultural practice areas than in new agricultural practice areas. Rodents Snap-trap indices for 1975 and 1976, which were obtained in similar vegetation types as those in 1977 are included in the results for purposes of comparison (Table 6) . No significant differences in snap-trap indices were detected among vegetation types for each of the rodent species trapped except the Ord T s kangaroo rat (p = 0.006) and the deer mouse (p = 0.05). The Ord's kangaroo rat index at the big sagebrush farm ecotone site was significantly higher than all other sites sampled. Although not significant, all other trapping locations in ecotones between farms and native vegetation had higher indices than any other sites, closely followed by the grass-burn site. The sites with high Ord’s kangaroo rat indices of abundance had a road and/or open areas and loose soil nearby. No significant site differences could be found for the deer mouse. Highest indices occurred in the riparian sites closely followed by the talus slope site, greasewood , and big sagebrush winterfat mix. In every year of trapping there were significant differences between species (p < 0.001). Deer mice had higher snap-trap indices than any other species. In 1976 and 1977 Ord’s kangaroo rat had the second highest index. Ord’s kangaroo rat trapping sites were not sampled as intensively in 1975 as in 1976 and 1977. This probably accounts for observed lower abundance in 1975. Analysis of marmot live-trapping results yielded a Schnabel population estimate of 13.5 marmots/ha at the Evans Ranch cliff site. The sex ratio was 0.86 males: 1.0 females. Marmots were also captured on talus slopes, along the canyon rim, and in riparian areas near rocky outcroppings. These latter sites were cottontail live- trapping locations (Table 3). The 1976 data showed that marmot colonies were found exclusively on or near rocky areas, especially near alfalfa fields. All colonies located in 1976 were near irrigated fields or other sources of succulent vegetation. Bushytail woodrats (Table 3) were only captured along the canyon rim, on canyon talus slopes, and in canyon riparian areas. There were no significant differences in indices of abundance among years (F = 8.04, df = 8/16, p>0.40) or among the three sites (F = 3.51, df=2/16, p > 0.20) . 71 Table 4. Pheasant abundance indicies on the Snake River Birds of Prey study area, April, 1977. Route Length (km ) No. Stations No. Calls/ Station No. Seen No. Seen/ km Old Agriculture Areas Melba 1 & 2 37. ul 21 11.18 8 0.22 Grand View 16.09 10 3.90 11 0.68 Chattin Flats 16.09 10 8.33 5 0.31 Bruneau 31.56 17 4.71 Mean: 7.03 6 Mean: 0.19 0.35 New Agriculture Areas Sinker 18.91 13 1.92 0 0.00 Flying H 23.17 14 0.57 1 0.04 Mountain Home 9.65 6 0.33 Mean: 0.94 0 Mean: 0.00 0.013 New Agriculture - Riparian Area Castle 9.33 6 18.00 3 0.32 72 Table 5. Linear combination F-values and significance for pheasant crow-count data on the Snake River Birds of Prey study area. Significance less than or equal to 0.05 indicates significant differences between the treatments being tested. Number of Calls per Station (Spring 1976, 1977) Comparison F-Value (df=l/10) Significance (P) old Agriculture x New Agriculture 22.50 <0.005 Old Agriculture x New Agriculture- Riparian 9.18 <0.05 New Agriculture x New Agriculture- Riparian 8.33 <0.05 1976 x 1977 14.14 <0.005 New Agriculture-rUparian, 1976 x New Agriculture-Riparian, 1977 25.93 <0.005 Old Agriculture, 1976 x Old Agricul- ture 1977 6.56 <0.05 New Agriculture, 1976 x New Agricul- ture 1977 0.06 >0.25 Number of Pheasants seen per km (Spring 1975, 1976, 1977) Comparison F-Value (df=l/14) Significance (P) Old Agriculture x New Agriculture 32.43 <0.005 Old Agriculture x New Agriculture- Riparian 43.23 <0.005 New Agriculture x New Agriculture- Riparian 1975 x 1976 0.00 2.97 >0.01 1975 x 1977 0.96 >0.01 1976 x 1977 6.85 <0.05 73 Table 6. Rodent snap-trapping results for July. See Appendix 1 for explanation of species codes. Trap Index (Number Caught/100 Trap Nights) Vegetation Year P .m. P.p. D.o. 0.1. P.c. E.m. S.t. Sagebrush 1975 8.6 1.1 _ _ _ _ 1976 17.6 1.1 - - - 2.2 2.2 1977 12.8 2.1 2.1 4.3 - - - Sagebrush- 1976 29.3 4.9 2.4 - - - 4.7 Winterfat 1977 28.3 - 4.4 - - Winterf at 1975 5.0 1.0 - 1.0 - - - 1976 19.2 1.1 2.1 2.1 - - 3.1 1977 4.2 - - 6.3 - - - Burn-Grass 1975 37.6 - 1.3 - - - - 1976 16.3 1.3 17.5 - - - - 1977 6.8 - 13.6 - - - - Grass 1975 4.4 - - - - - - 1976 14.6 - 2.4 - - - - 1977 4.2 - - - - - - Sagebrush- 1976 7.0 - 25.6 - - - - Winterfat/ 1977 6.7 - - - - - - Farm Winterfat/ 1976 16.2 - 16.2 - - - 14.0 Farm 1977 4.9 - 22.0 - - Riparian 1975 17.4 - - - 1.2 - - 1976 31.7 - - - - - - 1977 46.3 - - - - - “ Shadscale 1975 11.1 - - 5.6 - - - 1976 23.3 - - - - - - 1977 8.0 - - - - - Shadscale- 1976 12.2 - - 2.0 - - - Winterfat 1977 9.8 2.4 7.3 - - - - Talus 1975 40.5 - - - 10.8 - - 1976 35.3 11.8 - - - - - 1977 18.4 2.0 — - - — Sagebrush/ 1977 20.5 - 23.1 - 2.6 - - Farm Rim 1977 8.7 - - - - - - Greasewood 1977 31.3 4.2 - - - - - (N) Chukar Partridge The rally-call method was only partly successful. Responses to recorded calls were highly variable in each area and from day to day. Some observed flocks did not respond to the call and others began calling before the taped call was played. A map of flock locations and census routes is on file at the Birds of Prey Office. Pigeons Three days of visual sampling in August produced a mean number of 230.33 pigeons in the 6.84 km census route with a 95 percent confidenc interval of (165.07-295.59). Using the linear regression model: Y = 2.-02 + 1.51 x R2 = .9957 yields an estimated total population of 349 pigeons or 51 pigeons/km of canyon cliff. A second degree term in the regression model (Nie et al. 1970) contributed nothing (p>0.25) to the increase of the correlation coefficient. Passerines The results of the passerine transect surveys are inconclusive (Table 7). Low numbers of sightings on each transect produced low, inaccurate results. Horned larks were more abundant near open fields, grassy sites and other areas with low growing vegetation. Sparrows were found in areas with some structural cover present. Vegetation The completed vegetation map is in the Birds of Prey office in Boise. The original 7.5 minute map has been transferred to a 1.6 cm to 1 km (1 inch to 1 mile) map. DISCUSSION Blacktailed Jackrabbit In the first year of use of the Area-Estimate method, estimates of jackrabbit abundance in more common vegetation types were obtained. Spotlight censusing may be subject to serious drawbacks since rabbit activity is dependent on many factors . However, results usually agreed with past information subjective observations. In the coming year attempts will be made to obtain uniform samples in different vegetation types allowing more reliable conclusions. 75 Table 7. Results of counts on Passerine transects on the Birds of Prey study area 1977. See Appendix 1 for explanation of species code. is umber seen/km Vegetation Season Number of "Sparrow" Horned Meadow Other Transects Lark Lark Sagebrush Spring 7 l.b 1.3 2.6 0.0 Fall 5 1.0 2.4 0.1 0 . o A. g . Sagebrush- Spring 5 1.0 0.9 0.4 0.0 Winterf at Fall 5 0.0 2.4 0.0 0.0 Winterf at Spring 4 0.2 5.1 0.0 0.5 Swallow Fall 1 0.0 3.8 0.0 0.0 Sagebrush/ Spring 3 2.3 0.2 0.6 1.2 A. s. Farm Ecotone Fall 2 0.6 2.6 0.0 0.0 Shadscale Spring 6 0.9 l.b 1 . 5 0.2 C.c. Fall 4 0.0 5.2 0.2 0.2 C.c. Greasewood Spring 2 0.0 2.2 0.3 0.0 Fall Shadscale Spring 3 0.6 0.2 0.0 ( 0.2 A. s . Winterf at Fall 1 0.0 0.0 0.0 5 0.4 C.c. vJJ . 6 Swallow 0.6 R. s . Grass Spring 4 0.0 6 . b 0.6 [0.2 C.c. Fall 1 0.6 5.0 0.0 (j 0.3 N. a. 0.0 Greasewood Spring 1 0.3 0.0 0.0 0.9 P.c. Fall 5 4.6 F0.1 P.c. ) 0 .6 P . p . \J2 . 7 P.p.i. Sagebrush Spring 3 0.4 1.0 3.3 0.2 L.e. Grass Fall 3 0.2 4.4 2.5 ',0.6 C.c. Patches i_0.8 P.p.i. Jackrabbits were much more abundant in shadscale, the shadscale- winterfat association; and in big sagebrush than in other vegetation types. Since jackrabbits do not usually utilize burrows for escape, they require heavy canopy cover with a relatively open understory for protection from avian predators. They avoid areas of extremely dense vegetation as shown by the results in greasewood, where no jackrabbits were sighted in three nights of spotlighting. This greasewood area was very dense and did not have an open understory for jackrabbit maneuverability. In contrast, an area of very tall greasewood near budsage (Artemisia spinescens) with an open understory had potentially the highest density of jackrabbits. This transect route was very short, resulting in an estimate which is probably unreliable. Jackrabbits also avoid open, short vegetation. These areas do not provide sufficient cover for protection during the day. At night jackrabbits will utilize these open areas for feeding if they are located near suitable cover. There was no indication that jackrabbits occur in higher densities in ecotones between farms and native vegetation. Westoby and Wagner (1973) concluded that there was no obvious tendency for jackrabbit densities to be higher in the immediate vicinity of a cultivated field than at distances 500 to 900 m from it. They also found 70 percent of the forage removal by jackrabbits from a crested wheatgrass (Agropyron desetorum) field was concentrated in a band 300 m wide around its edge. These findings agree with results in this study. During August, jackrabbits were observed at night in extremely large numbers along the edge of an alfalfa field adjacent to big sagebrush. Thus, it appears that the establishment of a large cultivated field in otherwise good jackrabbit cover effectively eliminates that much potential jackrabbit habitat entirely. The animals will utilize the edges of the cultivated field for foraging at night, but the field does not offer suitable cover to satisfy jackrabbit habitat requirements. Cottontail rabbits Cottontails were most abundant in or near the canyon where they find cover in rock crevices. In the riparian and talus areas their numbers are much higher because of abundant food and cover. Dense greasewood also provides some cover for cottontails. Areas containing some rabbits, but more local in abundance, were rocky areas in the desert. These "islands" provide sufficient burrow sites for cover. Areas trapped without heavy cover or rocky areas nearby did not contain many cottontails. Cottontails require heavy cover or other good burrow site conditions. These requirements are met in areas of old agriculture with fence-row and ditch-bank vegetation. New agricultural areas do not support this kind of cover and consequently do not support very high cottontail populations. Any type of agricultural practice near suitable cover sites, such as rocky areas, will support cottontails. 77 There were no differences in the relative abundance of cottontails between the northwest and southeast portions of the study area. There were also no differences between known golden eagle hunting areas and other areas of comparable habitat type. These findings indicate that the cottontails in the study area require good burrow or cover sites to survive and, given this requirement, maintain fairly uniform populations within habitat types. Pheasants Assuming that the number of calls/station and population density are correlated (Kimball 1949) it is obvious that modern agricultural practices involving large blocks of farming have a negative affect on pheasant populations in comparison to the greater pheasant densities observed in old agricultural areas which have fence-row vegetations present for use as nesting cover and food. Incidental observations in native vegetation such as big sagebrush, shadscale and winterfat indicated extremely low pheasant densities in these areas. Maintenance of the study area in its present condition will not produce an increase in pheasant numbers. If clean farming is permitted with its large, bare tracts of crop areas there will not be a signifi- cant increase unless the fields are near good cover such as big sage- brush. One way to allow farming and increase pheasant numbers signifi- cantly in the study area would be to allow "old" type farms with rela- tively small fields of crops, separation of crops by fence-row vegetation, and irrigation using gravity or ditch irrigation. Modern pivotal irri- gation systems leave large areas of bare open ground and little cover at the edges. Rodents Deer mice were more abundant than any other rodent during every year of snap-trapping in July. The only other species locally abundant was Ord's kangaroo rat, which was confined to areas which have incurred pertur- bations such as roads and farming. In many cases, these types of disturbances result in suitable burrowing conditions. All other snap- trapped species showed no significant differences in relative abundance between sites and showed low populations throughout the area. Townsend ground squirrels (Spermophilus townsendi) have begun estivation by July and snap-trapping does not reveal true trends or densities. Most rodents species snap-trapped are primarily nocturnal or crepus- cular and thus supply little available prey for diurnal raptors. The only diurnal species, other than the Townsend ground squirrel, are the whitetail antelope squirrel. 78 (Ammospermophilus leucurus) and the least chipmunk (Eutamius minimus) . Least chipmunks are restricted to big sagebrush, and are thus largely unavailable to raptors. Whitetail antelope squirrels are restricted to the south side of the Snake River. Although the marmot colony at the Evans Ranch is subject to shooting pressure, its population seems to remain fairly stable from year to year. The 1976 density was 15.9 animals/ha, compared to 13.5 animals/ha in 1977. This fairly stable population exhibited by an established colony is not surprising since marmots typically maintain stable densities by dispersal of yearlings to balance the annual recruitment of young to the population (Armitage 1973). Captures of marmots incidental to cottontail trapping efforts occurred exclusively in or near rocky areas. Talus slopes, boulders on the canyon floor, and lava outcroppings near alfalfa fields supported colonies. Many of the captures involved possible satellite sites (Svendsen 1974). These are sites which have much smaller densities than normal colonies due to topographic diversity; they may be considered microhabitats, with the most limiting resource being burrow sites. Marmot colonies located in 1976 were adjacent to irrigated fields or other succulent vegetation. This need for succulent vegetation may indicate one of the limiting factors for marmots in the study area. Many otherwise suitable habitat areas in big sagebrush, on the canyon rim, or in other dry and rocky areas are not inhabited by marmots. There is a high potential for increase of marmot colonies in these areas if farming were permitted near these rocky areas. Trapping intensity was not sufficient to demonstrate whether the other marmot colonies found differed in density from the Evans Ranch colony. Further trapping of additional colonies may be helpful. If we assume the Evans Ranch colony was not limited by burrows or food supply, which is reasonable, we may say that the density of this colony is typical of a healthy colony in the Birds of Prey study area. Bushytail woodrats are confined to areas along canyon rims or within the canyon itself in the riparian and talus areas. Since these rodents are also nocturnal they do not represent a significant prey item and are therefore of minor concern. Their habitat is secure from most types of encroachment. Other rodent species observed, but not systematically trapped, were pocket gophers (Thomomys spp.), beavers (Castor canadensis) and muskrat (Ondatra zibethica ) . Pocket gophers occurred in areas of adequate soil moisture and with a soil texture friable enough to allow tunnelling. Irrigated fields, irrigation ditchbanks, alluvial soils of the canyon bottom and side streams, and certain northfacing slopes in Con Shea Basin and on Sinker Butte harbored these animals. Beaver and muskrat were restricted to aquatic and riparian habitats along the Snake River and associated streams and marshes. 79 Chukar partridge Data collected in August during the past two years indicate that chukar partridge area are primarily restricted to the main canyon and talus slopes and tributary canyons with steeply rising walls and close proximity to water. The rally-call method was not successful in esti- mating densities. Other studies have shown it is not possible to accurately census chukar density through the use of call counts (Oeklaus 1976, Christensen 1970). The birds are highly mobile; the terrain that consitutes their habitat is very rough and most water supplies are continuous. Other methods for obtaining indices of abundance are being sought. Pigeons During August pigeons are grouped into large flocks permitting easier counts of nearly all pigeons residing in the cliff area. The density estimate of 51 pigeons/km of cliff during August was made on only one section of vertical cliff along the Snake River. Extra- polation of this estimate to other areas along the river may not be valid. Additional information will be collected next season using similar methods. Passerines The 1975 results (Wolfe and Montan 1975) indicated a higher passerine biomass/ha than rodents (excluding Towsend ground squirrels) . Due to the low number of birds seen and the problems of correcting for differential sighting distances in various vegetation types, this finding could not be replicated in 1977. In areas of local abundance, the diurnal activity of the passerines may make them a potentially important prey source. However, their high mobility probably precludes effective utilization of this resource by the raptors. LITERATURE CITED Armitage, K. B. 1973. Population change and social behavior follow- ing colonization by the yellow-bellied marmot. J. Mammal. 54:842-854. Christensen, G. C. 1970. The chukar partridge. Biol. Bull. No. 4, Nevada Dept, of Fish and Game. 82 pp. Flinders, J. T., and R. M. Hansen. 1973. Abundance and dispersion of leporids within a shortgrass ecosystem. J. Mammal. 54 : 287-291 . Hicks, C. R. 1973. Fundamental concepts in the design of experiments. Hold, Rinehart and Winston, Inc. New York. 349 pp. 80 Kimball, J. W. 1949. The crowing count pheasant census. J. Wildl. Manage. 13 (1) : 101-120 . Nie, N. H. , C. H. Hull, J. G. Jenkins, K. Steinbrenner , and D. H. Bent 1970. Statistical package for the social sciences. McGraw-Hill, Inc., New York. 675 pp. Oeklaus, W. F. 1976. Chuckar partridge dispersion along the middle and lower Snake and Columbia Rivers. M. S. Thesis. Univ. of Idaho, Moscow. 56 pp. Snedecor, G. W. , and W. G. Cochran. 1967. Statistical methods. Iowa State University Press, Ames, Iowa. 593 pp. Svendsen, G. E. 1974. Behavioral and environmental factors in the spatial distribution and population dynamics of a yellow-bellied marmot population. Ecology 55:760-771. Tiemeier, 0. W. , and M. L. Plenert. 1964. A comparison of three methods for determining the age of black-tailed jackrabbits. J. Mammal. 45:409-416. Tukey, J. W. 1949. Comparing individual means in the analysis of variance. Biometrics. 5:99-114. Westoby, M. , and F. H. Wagner. 1973. Use of a crested wheatgrass seeding by black-tailed jackrabbits. J. Range Manage. 26(5) : 349-352 . Wolfe, M. L. , and J. R. Montan. 1975. Density and species compo- sition of birds and mammals. Snake River Birds of Prey Natural Area, Idaho. Pages 98-126 in_ Snake River Birds of Prey Res. Proj . Annu. Rep. 1975. U. S. Bur. of Land Manage., Boise, Idaho. 193 pp Wolfe, M. L. , J. R. Montan, Jr., K. Montan, J. Wassink, M. Wakefield, and B. James. 1976. Density and species composition of small mammals and birds in the Snake River Birds of Prey Natural Area, Idaho. Pages 133-162 Ln Snake River Birds of Prey Research Proj. Annu. Rep. U. S. Dept, of Inter., Bur. of Land Manage., Boise, Idaho. 240 pp. 81 Appendix 1. Common and scientific names and abbreviations of animals found in tables of this report. Species Code Common Name Scientific Name P . m. Deer mouse Peromyscus maniculatus P.p. Great Basin pocket mouse Perognathus parvus D.o. Ord’s kangaroo rat Dipodomys ordi 0.1. Grasshopper mouse Onychomys leucogaster P.c. Canyon mouse Peromyscus crinitus E . m. Least chipmunk Eutamius minimus S. t. Townsend ground squirrel Spermophilus townsendi S.n. Mountain cottontail Sylilagus nuttalli N. c. Bushytail woodrat Neotoma cinerea M. f . Yellowbellied marmot Marmota flaviventris L.c. Blacktailed jackrabbit Lepus californicus E.d. Porcupine Erethizon dorsatum S.p. Spotted skunk Spilogale putorius M. m. Striped skunk Mephitis mephitis N.l. Desert woodrat Neotoma lepida P . p . i . Black-billed magpie Pica pica A.g. Chukar Alectoris graeca C.c. Common raven Corvus corax A. s . Water pipit Anthus spinoletta R.s. Golden-crowned kinglet Regulus satrapa P.c. Ring-necked pheasant Phasianus colchicus N. a . Long-billed curlew Numenius americanus L. e. Northern shrike Lanius excubitor P.p. Gray partridge Perdix perdix 82 A pickup truck with to record frequency types (BLM photo by an observation platform mounted in back was used of raptor hunting forays in different habitat Steenhof ) . 83 STUDY IV-A: Raptor utilization of desert habitats. CONTRACTOR: Ecology Center, Utah State University Logan, Utah 84322. INVESTIGATORS: Michael Wolfe, Principal Investigator. Karen Steenhof, Research Associate. PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1003 . OBJECTIVES: 1. To compare the relative importance of habitat types to each raptor species and to the collective raptor population. 2. To determine if raptors partition the desert habitat spatially or temporally to avoid competition. ANNUAL SUMMARY A short term study of raptor habitat utilization was undertaken upon termination of a prey vulnerability study. Habitat use was assessed during systematic watches at seven sites containing representative habitat at varying distances from the canyon rim. Ravens were the most numerous of the 10 species observed. There was insufficient evidence to indicate that raptors were partitioning the habitat temporally or spatially. Perch availability was the apparent factor causing restricted habitat use by some species. The results of the study indicate that habitats 15 km from the canyon are just as important to falconiform raptors as habitats adjoining the canyon rim. INTRODUCTION The Snake River Birds of Prey Natural Area is the site of one of the densest populations of breeding raptors in the world. The high density apparently occurs because of the abundance of prey and nesting sites. Previous studies (Kochert et al. 1976) have shown that the diets of raptor species in the area overlap and that potential competition for food exists between diurnal raptor species. It is possible that spatial or temporal separation of hunting areas facilitates coexistence of raptor species in the area. Spatial separation could result from: (1) a preference by some raptor species for specific habitat associations or (2) a preference by individual raptors for certain zones of space or iThis short term investigation was initiated after termination of an unsuccessful study of factors influencing prey vulnerability. 84 territories (Edington and Edington 1972) . The Bureau of Land Management must know which desert habitats are most important as raptor hunting areas. In addition, it is necessary to know if some species have specialized foraging requirements. Other studies have suggested that raptor species have special- ized habitat preferences. Prairie falcons (Falco mexicanus) , for example, may prefer to forage where vegetation is short (Enderson 1964) . Although red-tailed hawks (But eo j amaicensis) and rough- legged hawks (Buteo lagopus) show no differences in field type preference, they may partition areas according to types of perches available (Schnell 1968) . Perch type preference varies consider- ably among raptor species (Marion and Ryder 1975) . This study on raptor habitat utilization grew out of the efforts of a study on factors regulating Townsend ground squirrel vulnerability to prairie falcon predation. The vulnerability study was terminated due to poor preliminary results (i.e. inadequate sample sizes and poor reproductive success of ground squirrels resulting from drought conditions, Wolfe and Steenhof 1977). The procedures and methods for this study were slight modifications of those initially designed for the vulnerability study. METHODS Seven sites in the Snake River Birds of Prey Study Area were selected for observation of raptor utilization (Fig. 1). They were subjectively chosen to provide good visibility of representative desert habitat types. All study sites were located within the hunting range of prairie falcons, as determined by Study III (Dunstan et al. 1976) . The following sites were studied. 1. A burned area approximately 2 km from the canyon, characterized by sparse perennial grasses and scattered big sagebrush (Artemisia tridentata) ; some cultivated land was visible from the observation point, and both utility poles and large powerline structures were within the viewing area. 2. A stand of low (20 cm or less) winterfat (Cerratoides lanata) , approximately 1.5 km from the canyon rim; both utility poles and large powerline structures were within the viewing area. 3. An ecotone area of mixed vegetation and land use, 2 km from the canyon on the edge of dense sage and mixed sagebrush/ winterfat association; several large powerline structures were within the viewing area, and some cultivated land was visible. In general, this stand was characterized by diversity of vegetation^ 4. A sagebrush/winterf at mosaic viewed from a barren butte top approximately 6 km from the canyon rim; no powerlines or cultivated land were within the viewing area. 85 Fig. 1. Location of sampling areas in the Snake River Birds of Prey Study Area. 86 5. A homogeneous stand of tall and dense sagebrush approximately 7 km from the canyon rim; no powerlines or cultivated land were within the viewing area. 6. A stand of small and sparse sagebrush approximately 11 km from the canyon rim and on the edge of a large grassy area; no powerlines or cultivated land were within the viewing area but the land was heavily grazed. 7. An open grassy area approximately 15 km from the canyon rim; grass was less than 8 cm in height. There were many rock outcroppings but no man-made structures. Raptors were observed from a tower platform mounted in the back of a pickup truck or from within the truck depending on weather conditions. The observations were facilitated by the use of a 15-60X spotting scope and 7 X 50 binoculars. A stopwatch was used to record the duration of raptor utilization of each habitat site . Observations were made in two series. One series extended from 7 to 23 April, and the second extended from 6 to 27 May. During each series, every site was observed for 2.75 hours in each of four time periods; 0 to 4 hours after sunrise, 4 to 8 hours after sunrise, 8 to 12 hours after sunrise, and 12 to 16 hours after sunrise. The exact time that each site was sampled was determined by convenience and scheduling was selected neither randomly nor systematically. No site was ever observed consecu- tively, and at least one day separated observations at a given site. If rain interrupted the observation period for more than 15 minutes, the observations were completed at a later date. Raptors were recorded as soon as they entered the area of view around the site, and were observed until they left the field of vision. It is likely that the larger raptors were visible at greater distances than the smaller raptors. In addition, visibility varied with weather conditions and differing topography of the sites. Attempts were made to distinguish foraging raptors in order to more precisely delineate hunting activities. Subjective evaluation of flight pattern, head position, and occurrence of capture attempts was used to classify each bird as foraging or non-foraging. It is likely that several errors were made in the classification of foraging activity. RESULTS In general, all carnivorous birds except ravens (Corvus corax) were scarce in the desert habitats surveyed. Within this functional group ravens were the most common birds, followed by prairie falcons (Table 1) . Kestrels (Falco sparverius) , marsh hawks (Circus cyaneus) , rough-legged hawks, golden eagles (Aquila Table 1 . Number of systematic observations of diurnal raptors, Snake River Birds of prey study area, 1977. Observation Period 7-23 April 6-27 May Total Raven 201 195 396 Prairie Falcon 26 33 59 Marsh Hawk 25 12 37 American Kestrel 15 9 24 Rough-legged Hawk 23 0 23 Golden Eagle 11 8 19 2 4 _6 303 261 564 Red-tailed Hawk 88 chrysaetos) , and red-tailed hawks were seen frequently. Turkey vultures (Cathartes aura) , an accipiter (Accipiter spp) , and a ferruginous hawk were each observed on one occasion. The number of birds observed was similar in both series except that rough-legged hawks were not observed in May. Observations of marsh hawks were also somewhat fewer in May. Wintering rough- legged hawks move to northern breeding areas in late April (Kochert pers. comm.), and it is possible that some marsh hawks also migrated from the Snake River Birds of Prey Area in early spring. Spatial Use The burned area near the canyon and the sparse sagebrush stand 11 km from the river were the most important raptor habitats in terms of the number of birds observed (Table 2) . The winterfat/ sagebrush mosaic 6 km from the river and the dense sagebrush 1 km from the river had the lowest number of raptor sightings. The number of raptors observed appeared to decrease with increasing distance from the canyon. Foraging raptors used the sparse sagebrush and the grass area 11 to 15 km from the canyon most frequently (Table 3) . The sites nearer the river were used less frequently by foraging raptors. Raptors used the burn and the edge situation for longer periods than the other habitats (Table 4) . Both these sites include cropland, and the apparent preferences may result from the prolonged foraging by ravens in cultivated fields. The collective raptor population also foraged for longer periods at these two sites, and most of the differences can be attributed to increased raven use (Table 5). Raven. Ravens were abundant in all habitats and they were also observed foraging at all sites (Tables 2 and 3) . The burn had the highest number of raven sightings although ravens foraged more often in sparse sagebrush. Ravens nested on powerline structures just outside of the viewing areas of Sites 1, 2, and 3. These sites probably included some of the nesting territories and much of the raven activity observed at these three sites was social interaction. This could account for the higher number of obser- vations and longer duration of use at the sites near the canyon. Foraging use of habitats appeared to be more evenly distributed, although as previously noted, ravens foraged for prolonged periods of the croplands near the canyon. Prairie Falcon. Prairie falcons used all seven sites with similar intensity. The frequency of foraging by falcons was also similar at the seven sites. The winterfat and burn sites were used most frequently and the dense sagebrush site appeared to be the least important for both foraging and general activities. 89 Table 2. Frequency of raptor Observations in relation to site, 7 April to 27 May, 1977. Site Number and Discription 1 2 3 4 5 6 7 Burn Winter fat - Edge Butte Mix Sage- brush Sparse sage- brush Grass Raven 91 59 76 36 40 81 23 Prairie Falcon 10 14 8 8 6 9 4 Marsh Hawk — 2 5 3 6 8 13 American Kestrel 2 8 4 3 4 — Rough-legged Hawk — — 1 1 6 5 10 Golden Eagle 6 5 2 2 1 2 1 Red-tailed Hawk 1 — 3 — — 1 1 Total 110 88 99 53 63 106 55 (Excluding Ravens) 19 29 23 17 23 25 32 90 Table 3 • Frequency of observations of foraging raptors in relation to site, 7 April to 27 May, 1977. 1 2 3 4 5 6 7 Burn Winter- Edge Butte Sage Sparse Grass fat Mix Sage Raven Prairie Falcon Marsh Hawk American Kestrel Rough-legged Hawk Golden Eagle Red-tailed Hawk Total (Excluding Ravens) 12 5 0 1 0 1 _0 19 7 7 5 1 6 0 0 _0 19 12 8 3 3 4 0 0 _L 19 11 12 3 0 1 0 1 _0 17 5 7 1 3 0 0 0 _0 11 4 16 3 6 0 1 1 _JL 28 12 3 3 7 2 4 1 _1 21 18 91 Table Duration of systematic raptor observations, in relation to site. Total Minutes Observed 1 Burn 2 Winter- fat 3 Edge 4 Butte Mix 5 Sage 6 Sparse Sage 7 Grass Raven 225.5 125.0 '570.5 155.0 108.5 202.0 42. U Prairie Falcon 25.5 37.5 13.5 19.5 26.5 27.0 28.0 Marsh Hawk — 6.0 22.0 4.5 20.5 22.5 55.0 American Kestrel 46.0 85.0 20.0 5.0 4.0 — 2.0 Rough-legged Hawk — — 3.0 2.0 16.0 13.5 85.0 Golden Eagle 34.5 16.0 20.0 27.0 1.0 6.5 1.0 Red-tailed Hawk 3.0 — 160.0 — — 8.0 4.0 334.5 269.5 809.0 213.0 176.5 279.5 217.0 92 Table 5. Duration of systematic observations of foraging raptors in relation to site. Total Minutes Observed 1 Burn 2 Winter- fat 3 Edge 4 Butte Mix 5 Sage 6 Sparse Sage 7 Grass Raven 82.0 25.0 49.0 80.0 29.0 25.0 6.5 Prairie Falcon 13.0 24.0 6.0 9.5 1.0 20.5 26.0 Marsh Hawk — 3 . 0 18.5 — 8.0 20.0 44 . 0 American Kestrel 45.0 70.0 20.0 4.0 — — 1.5 Rough-legged Hawk — — — — — 1.5 18.5 Golden Eagle 6.0 — — 2.0 — 1.0 1.0 Red-tailed Hawk — — 120.0 — — 8.0 4.0 146.0 122.0 213.5 95.5 38.5 76.0 101.5 93 Marsh Hawk. Marsh hawks used all habitats except the burn area. The number of marsh hawks observed appeared to increase with increasing distance from the canyon, with the grass area 15 km from the river receiving most frequent use. Marsh hawks also foraged most frequently at the sites farthest from the river. Marsh hawks appeared to avoid areas with many pole perches available. No marsh hawks were observed foraging on the Horse Butte winter fat /sage- brush mix (Tables 4 and 5) . Duration of marsh hawk use reflected the same patterns. American Kestrel. Kestrels were observed at all sites except the sparse sagebrush stand (Table 2). Most kestrels foraged in the sites nearest the river (Table 3) . The winterfat site was most important to kestrels for both foraging and overall activities. Kestrels did not forage in the dense sagebrush during the sampling period. Both the frequency and duration (Tables 4 and 5), of kestrel observations were highest at the sites which include power- lines. Wires and poles may attract kestrels although the birds are also more conspicuous on powerlines. Proximity to nest sites may also have caused kestrels to use Sites 1, 2 and 3. The mid-desert is probably not as important to this species as the more developed sites near the canyon. Rough-Legged Hawk. Rough-legged hawks used all sites except the burn and the winterfat stand. They were seen most frequently in the three sites farthest from the canyon where poles were not available and they foraged exclusively in the grass and sparse sagebrush, more than 10 km from the river. Duration of rough- legged hawk observations reflected the same pattern. Golden Eagle. Eagles used all seven sites, but foraged at only four. Eagles used the burn area most frequently. Much of the activity on this site was merely travel. On two occasions, an eagle passed over the burn site carrying prey it had captured farther out in the desert. Red-Tailed Hawk. Red-tailed hawks were observed at four of the seven sites (Table 2), and they foraged at three sites. Most sightings were at the edge situation and the duration of obser- vations was also much higher at Site 3. This site was apparently within the hunting range of a nesting pair from the canyon. Most observations at Site 3 were of hawks perched on powerline structures. The relatively few observations of hawks far from the river suggest that red-tailed hawks depend primarily on habitat near the canyon. Time of Day Raptors were observed in the desert during all time periods, but no red-tailed hawks were seen in the desert 12 to 16 hours after sunrise (Table 6). In general other raptors used the desert most frequently 12 to 16 hours after sunrise. Ravens and rough- Table 6 . Frequency of systematic raptor observations in relation to time of day, 7 April to 27 May, 1977. 1 0-4 Hours After Sunrise 2 4-8 Hours After Sunrise 3 8-12 Hours After Sunrise 4 12-16 Hours After Sunrise Raven 78 128 73 117 Prairie Falcon 17 10 12 20 Marsh Hawk 5 10 6 16 American Kestrel 10 3 4 7 Rough-legged 6 8 3 6 Hawk Golden Eagle 7 3 5 4 Red-tailed 2 1 3 — Hawk 125 163 106 170 95 legged hawks were most frequent in late morning. Eagles and falcons used the desert most often in early morning. Red-tailed hawks were most common in the desert in early afternoon. The foraging pattern was different (Table 7). Most kestrels and red-tailed hawks were observed foraging in early morning. Most ravens and marsh hawks foraged in late afternoon. Prairie falcons and eagles foraged most during the early afternoon. Rough-legged hawks foraged throughout the morning. Duration of raptor observations in relation to time of day are shown in Tables 8 and 9. Perch Use During the study, each case of perch use was recorded (Table 10) . Again, ravens were the most flexible and used virtually all types of perches available on the observation sites. Prairie falcons were also versatile, except that they did not perch on wires. Kestrels, on the other hand, perched on wire more frequently than other perch types. Utility poles and powerline structures were also important to kestrels, although kestrels also used natural perches. Marsh hawks and rough-legged hawks used only natural perches. This may be related to the lack of many man-made structures in areas away from the canyon. Red-tailed hawks and golden eagles, however, perched on artificial structures as often or more often than on natural structures. DISCUSSIONS AND CONCLUSIONS The results suggest that habitats 15 km from the canyon are just as important to falconiform raptors as habitat adjoining the canyon. The birds observed at great distances from the canyon may or may not have been breeding raptors. In general, raptors appeared to use sparse, open areas more often than dense stands. There was little evidence for spatial partitioning by birds of prey. All species of raptors used several habitat types. It is possible that in late spring, the migrant rough-legged hawks and marsh hawks hunt the areas away from the canyon to avoid the breeding raptors. Red-tailed hawks seem to concentrate their activities in a small area around the nest site. Other raptors use these same areas, however, and aggression between ravens and red-tailed hawks was observed in the desert. Perch availability was the only apparent factor that may have caused restricted habitat use by some species. Kestrels, for example, were observed primarily in areas with powerline wires. Red-tailed hawks also may prefer areas with powerline poles (Schnell 1968). Species that do not depend on conspicuous perches. 96 Table 7 . Frequency of systematic obserbations of foraging raptors in relation to time of day. 1 2 3 4 0-4 Hours 4-S Hours 8-12 Hours 12-16 Hours After Sunrise After Sunrise After Sunrise After Sunrise Raven 13 16 11 20 Prairie Falcon 6 4 7 6 Marsh Hawk 3 5 5 7 American Kestrel 8 2 2 2 Rough-legged 2 2 1 0 Hawk Golden Eagle 1 0 2 1 Red-tailed 2 Q 1 0 Hawk 35 29 29 36 97 Table 8 . Total duration of systematic raptor observations in relation to time of day. Total Minutes Observed 1 0-4 Hours After Sunrise 2 4-8 Hours After Sunrise 3 8-12 Hours After Sunrise 4 12-16 Hours After Sunrise Raven 492. u min. 359.5 min. 220.0 min. 357.0 min. Prairie Falcon 34.5 29. u 78.o 36.0 Marsh Hawk 21.0 19.0 14.0 76.5 American Kestrel 74.0 19.0 6.0 63.0 Rough-legged Hawk 70.0 24.5 8.5 16.5 Golden Eagle 21.5 22.0 28.0 34.5 Red-tailed 124.0 37.0 14.0 0.0 Hawk 837 . 0 510.0 368.5 583.5 98 Table 9 . Duration of systematic observations of foraging raptors, in relation to time of day. Total Minutes Observed 1 2 3 4 0-4 Hours After Sunrise 4-8 Hours After Sunrise 8-12 Hours After Sunrise 12-16 Hours After Sunrise Raven 81.0 121.5 44.0 49.5 Prairie Falcon 17.5 5.o 66.5 10.5 Marsh Hawk 10.0 14.5 9.0 60. o American Kestrel 72.5 14.0 5.0 49.0 Rough-legged Hawk 10.0 7.5 2.5 — Golden Eagle 2.0 — 7.0 1.0 Red-tailed 124.0 — 8.0 — Hawk 317.0 163.0 142.0 170.5 99 Table 10. Observations of perch use by diurnal raptors, Snake River BPNA, 1977. Ground Sagebrush Rocks Wire Powerline Structure Utility Pole NRC Pole Raven 75 10 2 3 36 5 1 Prairie Falcon 4 1 1 - 3 1 - American Kestrel 1 2 1 6 4 5 - Marsh Hawk 7 4 - - - - - Rough-legged 3 1 2 - - - - Hawk Red-tailed 1 1 _ — 2 — _ Hawk Golden Hawk - - 1 - 3 - - Unidentified — 1 - - - - - Accipter Unidentified - 1 — — — - - Buteo Unidentified _ _ 91 21 8 9 48 11 1 100 on the other hand, concentrate their activity in areas lacking perches. The presence of powerline structures, wires, and utility poles may enhance the value of desert habitat for some species (kestrels and red-tailed hawks) at the expense of other species (marsh hawks and rough-legged hawks) . There was little evidence to support temporal partitioning since most raptors used the desert throughout the day. Preferred foraging times show some minor separation. Kestrels and red- tailed hawks, two species least likely to compete, shared preferred foraging times. Similarly, prairie falcons and golden eagles hunted most often during the same time periods. This finding may be an artifact of the small number of raptors observed. In fact, it seemed that several species of raptors would appear at the observation site almost simultaneously, while no raptors would be observed for several hours. Competition between raptor species for prey does not appear to be prevented by differential habitat selection. Except when different types or sizes of prey are taken, raptor species are probably directly competing for food in the Snake River Birds of Prey Study Area. LITERATURE CITED Dunstan, T.C., J.F. Harper, and K.B. Phipps. 1976. Activity, hunting patterns, territoriality, and social interactions of birds of prey in the Snake River Birds of Prey Natural Area, Idaho. Pages 63 to 130 in Snake River Birds of Prey Research Proj . Annu. Rep. U.S. Dept, of the Inter., Bur. of Land Manage Boise, ID. 240 pp. Edington, J.M., and M.A. Edington. 1972. Spatial patterns and habitat partition in the breeding birds of an upland wood. J. Anim. Ecol. 41:331-357. Enderson, J.H. 1964. A study of the prairie falcon in the Central Rocky Mountain Region. Auk 81 (3):332-352. Kochert, M.N., A.R. Bammann, J.H. Doremus, M. DeLate, and J. Wyatt. 1976. Reproductive performance, food habits, and population dynamics of raptors in the Snake River Birds of Prey Natural Area. Pages 1 to 61 ini Snake River Birds of Prey Research Proj Annu. Rep. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 240 pp. Marion, W.R., and R.A. Ryder. 1975. Perch-site preferences of four diurnal raptors in northeastern Colorado. Condor 77 (3) : 350-352 . 101 Schnell, G.D. 1968. Differential habitat utilization by wintering rough-legged and red-tailed hawks. Condor 70 (4) :373— 377. Wolfe, M.L. and K. Steenhof. 1977. Factors influencing vulnerability of Townsend ground squirrels to prairie falcon predation. Snake River Birds of Prey Research Proj . Quart. Rep. 3 (2) : 21-22 . 102 Lowell Diller, Research Associate, has found that the uncommon "Long-nosed Snake" is more abundant in the BPSA than previously thought (BLM photo by Diller). 103 STUDY IVA-S : Ecology of reptiles in the Snake River Birds of Prey Natural Area. SUBCONTRACTOR: Department of Biological Sciences University of Idaho. INVESTIGATORS Donald R. Johnson, Major- Professor Lowell Diller, Researc1 'sistant Ted Daehnke, Field As >.t PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1003 (to Utah State University) . OBJECTIVES: 1. Determine the species composition and biomass of lizards and snakes in specific habitat types within the study area. 2. Determine significant population parameters for these reptile populations including productivity, sex and age ratios, survivorship, and mortality factors. 3. Assess the availability of reptiles as prey for raptorial birds . 4. Determine the impact of snakes on prey populations utilized by raptorial birds. 5. Determine the effects of habitat alteration on reptile densities . ANNUAL SUMMARY Observational censusing of lizard densities, completed in 1977, showed that shadscale (Atriplex conf ert if olia) , talus and canyon rim habitats contained the highest densities. Western whiptail lizards (Cnemidophorus tigris) were found to have the greatest potential as a prey species. Emphasis continued to be placed on assessing density, population parameters, growth rates, food habits and prey consumption of gopher (Pituophis melanoleucus ) and western rattlesnakes (Crotalus vir idis) in the BPNA. Rattle- snakes were shown to be the most numerous snake, with the densities as high as 11 per hectare in certain localities. Gopher and whipsnakes (Mast icophis taeniatus ) were next in abundance. Preliminary data from a new drift fence census indicate estimates on relative snake abundance may need to be revised. Age structure data indicate that both gopher and rattlesnakes are long lived with four identifiable age classes. Older age classes for both species also appear to contain more male than female snakes. Reproductive data are still preliminary for gopher snakes, but rattlesnakes were shown to have a low reproductive rate. Adult rattlesnakes fed almost exclusively on Townsend ground squirrels (Spermophilus townsendi) in 1976, but a reduction in the availability of ground squirrels in 1977, as a result of the drought, caused a change in their food habits. Gopher snakes consume a variety of prey items with no dependence on any one prey species. Consumption rates calculated from feeding experiments and growth rates indicate that gopher and rattlesnakes consume about two times their own body weight of prey each year. METHODS Lizard densities continued to be censused in 1977 employing an observational strip-census technique described in the 1975 annual report. Transects on the south side of the Snake River, which were not adequately censused during the previous summers, were censused in 1977. In 1977 the mark-release studies of snakes were continued, and in addition, drift fences were erected to census snake densities in different habitat types. The drift fences consist of 100-foot (30.5 m) sections of galvanized tin 20 inches (51 cm) high with two "snake traps" located 25 feet (7.6 m) from each end. The "snake trap" is a box constructed out of 1/8 inch (3 mm) hardware cloth, which is four feet (1.2 m) long by two feet (0.6 m) wide and one foot (0.3 m) high with funnels leading into the trap from both ends. The traps are located in breaks in the "fence" so that snakes attempting to get around the barrier from either side will be funneled into the trap. Each trap is shaded so that trapped snakes will not over-heat in the sun. Captured snakes are marked; the appropriate measurements are made; and snakes are released 20- 30 m from the drift fence. Snake food habit studies by palpation and scat analysis were continued as described by Johnson and Diller (1976). Also, the snake feeding studies begun in 1976 were continued in 1977. RESULTS AND DISCUSSION Body Weights of BPNA Reptiles Cumulative data on body weight measurements for 1975, 1976 and 1977 are listed on Table 1. Since virtually every snake encountered is weighed, the N value for each snake species also reflects its relative abundance, allowing for certain biases due to activity patterns and detect ibility. The N values for lizard species in no way reflect their relative abundance, since lizards were captured for measurements in a subjective manner. 105 Table 1. Mean body weights of snakes and lizards collected in the BPNA. All Age Classes Excluding Young Species n x (g) SD n X (g) SD Lizards Western whiptail ( Cnemidophorus tigris) 13 15.35 6.96 11 17.47 5.10 Collared (Crotaphylus collaris ) IT 26.48 14.58 12 34.63 7.64 Leopard (Crotaphytus wislizenii) 30 22.63 7.43 24 25.53 4.90 Desert horned (Phrynosoma platyrhinos) TT 17-81 9.08 42 23.62 6.12 Western fence (Sceloporus occidentalis ) 29 16.75 4.77 25 18.03 3.29 Side-blotched (Uta stansburiana ) 37 3 • 66 0.8l — — — Snakes : Western rattlesnake (Crotalus viridis) 212 426.37 207.87 192 469.69 183.38 Gopher (Pituophis melanoleucus ) 143 221.78 150.53 127 250.52 i4o. 06 Striped whipsnake (Masticophis taeniatus) 81 106.00 42.22 79 109.05 4o.36 Western ground (Sonora semiannulata 15 8.27 2.78 — — — Night (Hypsiglena torquata) 13 13.95 5.81 — — — Racer (Coluber constrictor) 9 68.19 38.11 — — — Western terrestrial garter (Thamnophis elegans ) 5 106.30 61.22 — — — Long-nosed (Rhinocheilus lecontei) 4 54.45 41.88 Lizard Density and Biomass The final compilation of observational census data are listed in Table 2. The data indicate that the habitat with the greatest lizard biomass is talus slope, but the diversity of lizard species there was low. The canyon rim and shadscale habitats respectively gave the next two highest biomass estimates, and they also had the greatest diversity of lizard species. Considering the relative proportion of each to the total area of the BPNA, shadscale habitats probably contain the greatest total biomass of lizards. Greasewood (Sarcobatus vermiculatus) and riparian habitats produced moderate biomass estimates with winterfat (Cerratoides lanata) , big sagebrush (Artemisia tridentata) , and grass respectively giving the lowest estimates. The one consistent characteristic is that habitats with high lizard densities have relatively low percent vegetative cover with many areas of open ground or rocks, while areas of low lizard densities have high percent vegetative cover and little or no open ground . Relative Importance of Lizard Species in the BPNA Side-blotched lizards (Uta stansbur iana) are the most abundant lizards in the BPNA (Table 3) . They are also one of the most widespread, but because of their small size probably are not an important prey species. The lizard with the greatest potential as a prey species is probably the western whiptail, since it is the most ubiquitous, and also has a relatively high mean biomass (g/ha) . Western fence lizards (Sceloporus occidentalis) have the highest mean biomass values of any lizard, but their importance is reduced by the fact -that they occur in a few restricted habitats. Leopard lizards (Crotaphytus wislizenii) are also a potentially important prey species, because they are rather widespread and have a large body size. Snake Densities Table 4 summarizes all recorded observations on the three major snake species in the BPNA. Many of the observations are fortuitous encounters by various personnel of the BPNA research project, and the rest are observations made by members of this study. No attempt has been made to remove biases resulting from unequal time spent in different habitat types, or the det ectibility of the different species of snake. Thus, these data cannot be used to calculate relative density estimates, but they do reveal certain trends. Rattlesnakes are the most likely to be encountered in most habitat types followed by gopher and whipsnakes. The canyon rim probably has the highest density of snakes, even though the values for this habitat type are biased from unequal time being spent in this habitat. Also, rattlesnakes and whipsnakes show a strong preference for rocky substrates, while gopher snakes seem to prefer loess . 107 Table 2. Indices of lizard density and biomass. C.t. = Western Whiptail (Cnemidophorus tigris); C.w. = Leopard Lizard (Crotaphytus wislizenii) ; C.c. = Collared Lizard (Crotaphytus collaris); U.s. = Side-blotched Lizard (Uta stansburiana) ; P.p. = Desert Horned Lizard (Phrynosoma platyrhinos ) ; S.o. = Western Fence Lizard (Scelopc-us occidentalis) . Habitat Type (grid coordi- nates) Species Density (n/ha) Biomass (g/ha) Transect Total Mean Habit Biomass Winterfat (a) C.t. 0.7 10. 3 (328875) C . w. 0. 7 15.2 U.s. 3. 3 12.3 37.7 Winterfat (b) C.t. 1.0 14.9 (314896) C.w. 1.0 21.9 U.s. 15.3 56.0 92.8 65.3 Big Sage (a) C . w. 1.8 41. 9 (320924) U.s. 2.8 10.2 52.0 Big Sage (b) C.t. 1.0 15.8 (297932) P.p. 1.0 18. 3 34.2 Big Sage (c) C.t. 2.1 31.8 (308868) P.p. 2.5 9. 1 U.s. 1.2 4.5 45.4 43.9 Shadscale (a) C.t. 1.2 19.2 (371825) C . w. 0.8 18.8 P.p. 0.6 9.8 U.s. 13.8 50.6 98.4 Shadscale (b) C.t. 6.8 103.8 (385796) C . w. 0.5 10.9 P.p. 1.9 34.4 U.s. 17.7 64.9 213.9 Shadscale (c) C.t. 2.5 38.1 (303780) C . w. 1.9 42.1 P.p. 2.5 44.2 U.s. 9.9 36.4 160.7 157.7 Canyon rim (a) C.t. 2.8 43.0 (348834) S.o. 2.4 40.7 U.s. 12.6 46.2 129.9 Canyon rim (b) C.c. 1.5 39.7 (348834) C.t. 2.2 34.5 P.p. 1.0 17.8 S.o. 4.0 67.0 U.s. 6.5 23.8 182.9 Canyon rim (c) C.c. 0.6 27.0 (323873) C.t. 1.8 15.6 C . w. 1.2 26.7 S.o. 7.6 126.6 U.s. 8.2 30.1 226.1 179.6 Talus slope (a) S.o. 11.4 190. 3 (278921) U.s. 11.9 43.6 233.8 Talus slope (b) S.o. 8.5 142.9 (307904) U.s. 8.7 28.4 171.2 202.5 Riparian (a) C.t. 6.4 98.8 98.8 (370795) Riparian (b) C.t. 7.1 108.7 (300913) S.o. 6.2 22.7 131.3 115.1 Greasewood (a) C.t. 2.3 35.5 (224897) U.s. 2.8 10.2 C.w. 0.9 19.7 65.2 Greasewood (b) C.t. 5.7 87.2 (343752) C.w. 2.8 64.3 P.p. 0.9 16.9 168.4 116.9 Grass 0 0 0 0 0 (Sand Creek) Table 3. Relative Importance of Lizard Species in the BPN'A. 106 oj x CQ CQ o3 c O •H PQ ft cd CD o3 X -p CQ ft s CD -ft H CO Eft -ft LCA po CO VO i — I i — i CO CV! PO O ON ft t — t~ — PO LT\ PC i — | CD ft CD x £ CQ ft O 0) CQ ft cc3 ft ft 0) X ft g o ft ft — ft o CQ CD ft b3 ft a3 — a3 -p X CD X ft CD O CD •H X ft ft X O CD ft Es ft ft O o X X ft 1 — 1 ft X 0 ft pq ft ft ft ft 0) i CD cc3 ft 33 -p CD ft ft CD t — i CQ X CQ O CQ i — 1 (D •H CD (D CD O 03 :s ft pq o ft ft CD CQ CD ft ft CD ft CQ CD ft X cc3 > ft eft CD s 109 Table h. Numbers of snakes encountered in relation to habitat. Vegetation/ Habitat Substrate Snake Species Habitat Totals Western rattlesnake Gopher snake Striped vhipsnake Winterfat loess 7 17 8 33 rocky 1 1 Big Sage loess 20 38 1+ 62 rocky 2 2 3 6 Shadscale loess 6 7 1 ll+ rocky 2 1 3 Winterfat- loess 12 6 2 21 Big Sage rocky 19 1 6 26 Grass loess 5 13 1+ 21+ rocky 56 12 6 71+ Canyon Rim rocky 102 11 1+7 ll+8 Talus rocky 1+ 11 6 21 Riparian mixed 12 37 6 51+ Greasewcod sand/loess 2 3 1 6 Substrate Totals rocky loess other 186 50 Ik 38 86 1+0 68 19 7 292 155 61 161+ Species Totals 250 91+ 508 110 Rattlesnakes were the only snakes recaptured frequently enough to estimate their density. After repeated searches in an area, we reached a point where all rattlesnakes captured were previously marked, and thus total number marked should equal population size in that area. In this manner, density estimates of 6.6, 5.6 and 4.5 snakes per hectare were obtained for Trio Butte, north rim of "Tom Draw" and east rim of "Tick City Basin" respectively. Other areas censused by mark-release have not been adequately sampled to determine density estimates yet. A removal census of rattlesnakes in 1976 along the rim between Dedication Point and Dedication Site yielded 27 snakes in 7 succes- sive searches over a 9-day period; a density estimate of 11.0 snakes per hectare. During the remainder of that field season, eight additional snakes were found in this area, but were not included in the density estimate because of possible immigration into the area. During the 1977 field season, only three additional rattlesnakes were found in the censused area indicating that the snakes had been effectively removed the previous summer. Results of the drift fence census are summarized in Table 5. These results represent the experimental stage of this census procedure in the BPNA, and part of the peak activity period of snakes was not sampled. The results from big sage are especially biased since the peak activity period occurred before trapping was initiated there. However, it is interesting to compare relative abundance of different snake species from this sampling procedure with the N values from Table 1 and number of snakes observed in Table 4. Whipsnakes which are third in abundance as indicated in Tables 1 and 4 are most abundant according to the drift fence census. Also night snakes (Hypsiglena torquata (which are noc- turnal and seem to be very rare) and gopher snakes are second in abundance. It is apparent that the drift fence census is biased in favor of more mobile species, while visual search is biased towards conspicuous diurnal species. Note that the three habitats which were sampled over the same approximate time period had similar snake totals, but the species composition varied considerably. Western Rattlesnake Life History Data Data from 1975, 1976 and 1977 on size classes, age structure and sex ratios are given in Table 6 and Fig. 1. Although distinct size classes are obscured by highly variable growth of snakes throughout the field season, there appear to be four identifiable size classes of males and females. The first includes snakes that had been born the previous September and would be close to one year old. The second and third size classes consist of snakes nearly two and three years old respectively. The adult snakes include individuals four years old or older (possibly as old as 15 to 2C years, Fitch 1949). Note that there were more snakes found in each successive size class for both males and females. It is Table 5* Snake Drift Fence Census. Ill CO I — I <3 -p o EH Lf\ 0\ 0\ ft- H CM CM i — I CM -=t ft- CD bO <3 CO bfl •rH PQ CM ft- o LT\ LT\ O cn ft S3 CD <3 i — I *H ft u £ <3 3 ft CO -H ft CO ft c3 _(_3 ' •H CM ft c3 ft 0) CO c3 ft CO 3 rH (3 Eh ft- CM LTN ft" ft- I L/"\ i — I CM ft ft OO O i — I VO I — I ft o\ ft CO o •iH ft S3 o ft S3 <3 o LT\ CM LT\ LT\ CO c— I — I o CO 0) •rH o CD ft CO CD ft <3 sq CO CD O <3 (D ft ft CD 0? ft ft S3 CD (3 ft CO •H S3 CD ft i — 1 CO ft bO Sh CD i — 1 CD (3 CO •rH CD CD ft CD i — 1 S3 1 — 1 S3 ft ft Gj PQ ft CD CO (3 1 3 ft S3 1 ft ft ft ft S3 o CO > ct5 <3 ft O c3 CO o O P3 S3 S3 EH ft ft i — 1 CO pl Eh •H S3 ft i — 1 S3 o CD CO ft •H bO CD •H ft O ft is CO •H ft CQ CD ft S3 ft CD ft 3 ft (3 CO i — 1 oo la O X CO CO CA PP 1 X H CM CO CM OO h— UA . . . • CO CO X CM oo X Q OD OO X" o -P CD 1 — 1 bO £ CD i — 1 -P oo A- X UA PI . . , cu IX MO 00 X i — 1 t> CO 1 — 1 t— i — 1 i OO MO A— o -p 1 — 1 pi o Pi LT\ 00 o UA co Xl i — 1 i — 1 oo CA rd ■d , — .. i — l i — l d 1 O O i — l - — ' O Op P p 1 — 1 X X p O X CM oo P - X X CO CO A I i — 1 CD -P ' — i — 1 r— 1 •H Pi CO bO Pi Oj -p P CD cc3 r— 1 3 t> X Pi o pi Pi d X X CD < Ii6 o CO P (U -p cd £ •H -p c CD o p •H -p bZ P (D PI -p p cd > I -p p o S C cd CO P bO o -p co •H p co co cd i — l a CL) N •H CO CD cd P co P CD ft ft O O C\J fab •H VO -=f C\J i — 1 i — I i — I O CO VD i — I CVJ 117 Few data have been accumulated on gopher snake reproduction to date. Three captive snakes laid clutches of six, six and eight, but none of the eggs hatched. All of these females appeared to be at least four years old, indicating that gopher snakes probably reach sexual maturity in four years. There is no reason to believe that female gopher snakes reproduce on alternate years like rattlesnakes. Snake Food Habits Food habits for rattlesnakes and gopher snakes are indicated in Tables 8 and 9 respectively. It is of interest to note the extent to which rattlesnakes prey on Townsend ground squirrels. In 1976, when ground squirrels were available, adult rattlesnakes fed on them almost exclusively. In addition to the data shown in Table 8, 26 field collected snake scats all contained Townsend ground squirrel hair. It is assumed that almost all of these scats came from rattlesnakes due to their size and proximity to dense rattlesnake populations. In comparison, gopher snakes do not show a clear prey selection, and probably take their prey in relation to its availability. Note the difference in food habits between 1976 and 1977. As pointed out by Study V report, few Townsend ground squirrels were available by mid-May of 1977 because of early torpor. Thus the primary prey species for rattlesnakes was eliminated in 1977, and this was reflected in their food habits. In general, they fed less frequently. In 1976, 15.2 percent of all field captured rattlesnakes had palpable prey items in their stomachs, while in 1977 only 6.6 percent had palpable prey items. Also, rattlesnakes began taking deer mice (Peromyscus sp.) more frequently in 1977. The food habits of gopher snakes did not change drastically in 1977, since they do not depend heavily on Townsend ground squirrels. Gopher and rattlesnakes were found to take similar sized prey items, although rattlesnakes are probably capable of ingesting larger prey. The size of Townsend ground squirrels palpated from snakes ranged from 70 to 200 grams, while cottontail rabbits (Sylvilagus sp.) ranged from 115 to 190 grams. Snakes occasionally were found to ingest prey equal to their own weight, but most prey items were one half or less the snake’s weight. Snake Growth Efficiency and Prey Consumption The growth efficiencies given in Table 10 are calculated in terms of increase in snake weight in relation to whole weight of prey consumed. Rattlesnakes have significantly higher growth efficiency than that of gopher snakes. It has been noted that rattlesnakes are much slower in digesting prey in captivity, which may result in their having a higher assimilation efficiency and thus higher growth efficiency. This possibility will be investigated further during the next field season. 'able 8. Western rattlesnake food habits, number of individuals as determined, by palpation and scat analysis. 118 1 1 — 1 cd -p CVi OO CO CM 1 ! CM r — i O on Eh E- e— ON CM VC CM 1 — i i — 1 1 1 1 i VC E- O CM CM CM < — 1 1 — ! CM 1 — 1 ON Ci i — i ft •H X) ft cd [ — 1 ft 0 t — I O 0 o3 CO 0 xs P P P p P p P 0 P a? P •H 1 — 1 S 1 N O p •H P •H O ft Ct3 P CT I — 1 ft co ft 0 co 0 P ft P 1 — 1 ft ft O 0

> CO CO CO c3 > co 0 ft P - a3 bO 0 Cj p P O 1 — 1 ft P ft p ft ft 0 S 1 — 1 ct5 O p CO cij P ft 1 — 1 0 P p P P a? 0 p oc3 [3 0 P 0 ft P CO 0 0 0 P P P P 0 P Eh Q S ft O < ft 9- Gopher snake food habits , number of individuals as determined by palpati scat analysis. Prey Item 197 6 1977 Total 119 £ £ £ O 00 VO VO — V CvJ i — | i — | j — | | — j , — j OA CvJ ir\ | cm H I I I H ~ 1" i I -V I I i — I i — | i — | O i — I P £ Eh P •H CO P £ P i — 1 0 £ CD P £ £ P £ bD £ i — 1 •H Vi bD 0 •H £ £ 0 P £ c£ £ -P CO N vi vi £ p •H £ £ O vi £ i — 1 •H •H P £ £ p rO P> £ vi o O CD O CD vi vi o CD £ i — 1 O Pi 0 0 CD CO bO O £ a •H •H CO CO £ S> £ p P p £ - O V* bO O •iH •H o 1 — 1 S £ CD £ P i — i p P 1 — 1 CD £ £ P £ £ £ CD CO § P £ 1 0 0 £ P CO £ P M CD vi vi CD P £ ,3 £ vi CO Vi •H •H CD £ O O O £ £ •H £ £ Q W EH s O CO P D 120 Table 10. Consumption rattlesnakes rates and growth ef ficiencies of captive gopher and Body Weight (g) Feeding Prey Consurap- Growth Species Start Finish Period Consumed( g) tion Efficiency ( days ) g/day Gopher snake 62 71 51 43 0.84 0.21 122 175 97 204 2.10 0.26 21 4 253 44 107 2.43 0.36 154 184 65 260 4.00 0.11 319 346 137 205 1.50 0.13 1^5 168 106 225 2.12 0.10 58 6l 62 36 0.58 0.08 75 131 170 263 1.55 0.21 243 255 57 76 1.33 0.16 180 194 79 89 1.13 0.16 240 268 96 114 1.19 0.25 330 357 90 245 2.72 0.11 312 358 88 222 2.52 0.21 301 326 45 90 2.00 0.28 72 129 120 250 2.08 0.23 167 176 15 24 1.60 0.37 136 175 58 148 2.55 0.26 178 193 39 110 2.82 0.14 X = 0.20 western Rattlesnake 154 220 117 273 2.33 0.24 100 126 30 85 2.83 0.31 208 232 32 59 1.84 0.4l 47 100 121 146 1.21 0.36 15 37 122 69 0.57 0.32 15 38 165 87 0.53 0.2 6 14 50 165 109 0. 66 0.33 197 256 69 197 2.86 0.30 117 164 6l 112 1.84 0.42 133 169 31 82 2.64 0.44 152 159 26 29 1.11 0.24 198 264 66 204 3.09 0.32 258 276 19 149 7.84 0.12 380 389 21 58 2.76 0.15 X = 0.32 121 By extrapolating these growth efficiencies to wild gopher and rattlesnakes, their prey consumption can be calculated based on the mean increase in body weight between successive age classes (Tables 6 and 7). Thus we estimate that young rattlesnakes (1 year olds-sexes combined) consume 411 g of prey during one summer, while juvenile males consume 513 g and juvenile females consume 323 g. The mean increase in weight between subadult and adult age classes cannot be used to calculate prey consumption, since the adult age class contains several cohorts. The estimated prey consumption for young and juvenile gopher snakes is 186 and 296 g respectively. Although they are still preliminary, these data compare favorably with scanty data in the literature where estimates of annual consumption rates range from 2 (Fitch 1948) to 4.5 times the snake’s own weight (Klauber 1972). LITERATURE CITED Fitch, H.S. 1948. Ecology of the California ground squirrel on grazing lands. Am. Midi. Nat. 39(3): 513-596. . 1949. Study of snake populations in Central California. Am. Midi. Nat. 41(3): 513-579. Fitch, H.S. and B. Glading. 1947. A field study of a rattlesnake population. Calif. Fish and Game. 33(2): 103-123. Johnson, D.R. and L. Diller. 1976. Ecology of reptiles in the Snake River Birds of Prey Natural Area. Pages 165-179 in Snake River Birds of Prey Res. Proj . Annu. Rep., U.S. Dept. Inter., Bur. of Land Manage., Boise, ID. 240 pp . Klauber, L.M. 1972. Rattlesnakes: their habits, life history and influence on mankind. 2nd Ed. Univ. of Calif. Press. 2 Volumes. | f 122 Researchers have captured and marked 219 "badgers since the beginning of the study, indicating that the BPSA hosts one of the densest badger populations in the country (BLM photo by Dunstan). 123 STUDY IV-B (BADGER SEGMENT): Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey Natural Area, Idaho. CONTRACTOR: Idaho Cooperative Wildlife Research Unit, University of Idaho, Moscow, Idaho 83843 INVESTIGATORS: PROJECT SUPPORT: Maurice G. Hornocker, Principal Investigator. John P. Messick, Research Associate. Randall M. Olson, Research Aid. Ronald P. Olson, Research Aid. E. Martin Cobb, Research Aid. Jackson S. Whitman, Research Aid. U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1004 . OBJECTIVES: 1. Attempt to ascertain the density, sex and age structure, and other population parameters of badgers (Taxidea taxus ) . 2. Investigate habitat use and describe the social organi- zation of the badger population. 3. Gather information on the food habits of badgers as an aid to estimating the impact of this predator. 4. Assist with the coyote (Canis iatrans ) segment of Study IV-B by providing previously collected data, aiding with the interpretation of this data, and reviewing the final report . ANNUAL SUMMARY Field work was completed by September. Data gathered since January 1975 are now being analyzed in preparation for writing the final report. During 1977, 65 badgers were captured 133 times. Thirty-seven were males, 28 were females, and 35 badgers were recognized as young-of-the-year . Since the beginning of the study, 219 badgers have been captured 367 times. Carcass collections for age, reproductive, and food habits information total 204. Eighty-nine of these were obtained in 1977. Twenty-six badgers were equipped with radio transmitters this year, although 10 individuals provided the most complete data on movement, activity, and dispersal. 124 INTRODUCTION Field work was completed by September, although some samples may be obtained from trappers during the 1977-78 season. The Research Associate returned to the University of British Columbia, Vancouver, Canada, to analyze data and simultaneously complete his thesis and final report to the BLM. With data analysis now in progress, this report is intentionally brief and incomplete. The final report should be finished late in 1978. METHODS Messick et al. (1975) and Hornocker et al. (1976) describe the field techniques. Methods currently in use on teeth, reproductive tracts, and other biological material will be given in the final report . RESULTS Badgers Captured or Collected From February 1975 to the end of field work in September 1977, 219 badgers were captured 367 times. Details appear in Table 1. These were animals intended for marking and release. Additional badgers were trapped specifically for post-mortem examination, although most carcasses were obtained from trappers, found as road kills, or collected in other ways. The efforts have yielded 204 badger carcasses. Eighty-nine of these were collected during 1977. Age Structure Ages of 70 badgers are given in Table 2. Interpretation of age from counts of cementum annuli were to the nearest 0.5 year, but for brevity, ages were pooled for the arbitrary categories in Table 2. Annuli are probably formed during the winter. Thus, samples should be stratified according to the season of collection in order to slightly improve the accuracy when animals are aged to the nearest half year. This was not done for the data in Table 2. However, this approach will be tested during further study, and as additional teeth are sectioned. Radio- tracking Twenty-six badgers were fitted with radio transmitters during 1977. Ten animals provided substantial data, but the other badgers did not retain their collars for more than a few days or weeks. Table 1. Numbers and classification of badgers captured during 19 75 , 1976, and 1977, and near the Snake River Birds of Prey Natural Area, Idaho. 125 i soanpdnoo^ sojnpdn3 soanpdnoo^; saanpdn^ 8t v ^ cd ^'O o H >* O CO +J T 1 D 03 03 3* M S 1 CD O o >> £ 4-1 NO 20 17 43 CD CD 4-1 — oo On o a OJ i — I 1 — 1 LO < — 1 3 CO T3 C & 35 20 20 75 LO kO r^ n3 ch 0- r^ 4J o ON a\ O r— 1 i — I i — 1 H g^? total (34) (26) (20) (20) (60) (40) (54) (46) captures) 126 Table 2. Age of 70 badgers collected during 1975 and 1976 near the Snake River Birds of Prey Natural Area, Idaho, and adjacent areas. Age based on preliminary counts of cementum annuli to the nearest 0.5 year. Data pooled for arbitrary age classes below. Details in text. Pooled age classes in years Sex o i — 1 V 1 1. 5-3.0 3. 5-5.0 15.5 Total dV 11 11 5 2 29 ?? 19 12 6 4 41 Total 30 23 11 6 70 (% of total) (43%) (33%) (16%) (9%) 127 On 13 February, adult female No. 152 finally emerged from the den that she had occupied continuously since 3 December 1976. During those 72 days underground she emerged from this winter den only one time. I continued to monitor No. 152 until she lost her radio transmitter on 20 February. She apparently did not return to her winter den between 13 and 20 February. No. 96, a pregnant female, was fitted with a radio transmitter on 7 March. This animal was first trapped on 9 May 1976 and showed evidence of lactation at that time. No. 96 began using one den on 11 March, although she usually emerged every day. She changed dens on 4 April 1977, and R.P. Olson heard noises from her litter on 5 April. This is the pattern of den occupancy characteristic of badgers raising young. Her litter of one female and two males was captured and marked, and one of the young males was instrumented. No. 96 was refitted with a radio containing a fresh battery on 31 May. She shed this collar on 20 June. While none of the radio-tracking data has been analyzed in detail, movements and activity of adults and young generally conformed to the patterns previously reported (Messick et al. 1975; Hornocker et al. 1976). Three instrumented young were killed by other predators. Young-of-the-year male No. 180 was instrumented on 30 May. He was recaptured on 1 July in a farming area, approximately 16 km from the original capture location. No. 180 had lost a rear leg to haying operations, but the injury was healing, and no infection was evident. He was fed, treated, and released the next day, only to be killed by a farmer on 9 July. Weight lo$s between 1 and 9 July equaled 0.2 kg. Female young-of-the-year No. 199 was also killed in July, either by a dog, coyote, or possible another badger. No. 185, a young male, weighed 4.5 kg when he was released on 5 June. He occupied a new dig on 6 June, nearly 0.3 km from the release site. On 15 June, his transmitter was recovered from the vicinity of the nest of the Walter's Ferry golden eagles (Aquila chrysaetos ) across the Snake River and over 12 km from the location on 6 June. No badger remains were found, but analysis of the nest contents has not been completed. A wood rat (Neotoma sp.) had carried the radio into some rocks. The only plausible explanation is that No. 185 was killed by an eagle. Feeding Studies and Predation Townsend ground squirrels ( Spermophilus townsendi) on the intensive study area did not reproduce this spring. This lack of readily available food and certain other problems limited the feeding studies in 3 m by 4.5 m dirt-filled pens. Badgers did 128 not habituate to confinement. They cached food and probably expended more energy digging and attempting to escape than they would have if free-roaming. Extensive growth data have been obtained from No. 161, a female hand-raised since she was only a few weeks old. As pointed out earlier in this report, carcass collections for food habits and reproductive information have continued. Other Activities G. Smith (Study V) and J. Messick met in December to review and discuss antibody titers to plague (Yersinia pestis) found in badger blood samples collected throughout this study. This survey was conducted in cooperation with Study V. Numerous presentations on the badger project were given to civic and professional groups. Coyote Segment An amendment to this contract became official during 1977, and S. Cherry was hired to take over and intensify the coyote segment. J. Messick provided Cherry with previously collected coyote and bobcat (Lynx ruf us ) samples. The coyote segment is processing these samples, but Messick will assist in the interpre- tation of the data and will review the final coyote report. LITERATURE CITED Messick, J.P., M.G. Hornocker, I. McT. Cowan, and J. Whitman. 1975. Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey Natural Area. Pages 136-154 in Snake River Birds of Prey Research Project Annu. Rep., 1975. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 193 pp. Hornocker, M.G., J.P. Messick, J.S. Whitman, and G.W. Smith. 1976. Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey Natural Area. Pages 181-201 Mr Snake River Birds of Prey Research Project Annu. Rep., 1976. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 240 pp. 129 13 0 Steve Cherry, Research Associate, initiated an investigation of coyote density, food habits , and population trends in the BPSA. (Photo by Bammann) . 131 STUDY IV-B (COYOTE SEGMENT) Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey Natural Area, Idaho. CONTRACTOR: Idaho Cooperative Wildlife Research Unit, University of Idaho, Moscow, Idaho 83843. INVESTIGATORS : Maurice G. Hornocker, Principal Investigator . Steve Cherry, Research Associate. John P. Messick, Research Associate, Jackson S. Whitman, Research Aid. PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1004 . OBJECTIVES: 1. Collect information on the seasonal food habits of coyotes (Canis latrans ) and bobcats (Lynx ruf us ) . 2. Collect information on relative abundance, reproduction, and mortality /survival rates of coyotes and bobcats. 3. Incorporate food habits and population data with bioenergetics information from the literature to estimate predation intensity of coyotes and bobcats. ANNUAL SUMMARY A total of 945 coyote scats were collected during the year. When these scats are analyzed there will be food habits information from every season of the year with summer and fall being the best represented seasons. Ages were determined for 241 coyotes based on the number of cementum annuli in their canines or premolars. A total of 151 coyotes were chosen from this collection for the estimation of mortality rates. Forty of these 151 were from the winter of 1975-1976, 50 were from the winter of 1976-1977, and the remaining 61 were collected in September of 1977. The Chapman- Robson technique was used to estimate mortality. Adult survival was generally around 70 percent during all three periods. The total survival was estimated at around 60 percent during the winter of 1976-1977. The total mortality rate could not be calculated from the winter 1975-1976 and fall 1977 collections because of statistical constraints concerning the juvenile age class. A total of 60 coyote reproductive tracts were analyzed during the year. The springs of 1975, 1976, and 1977 were rep- resented in this sample. The average number of placental scars per coyote was 6.5 in the spring of 1975, 6.0 in the spring of 1976, and 6.67 in the spring of 1977. Several indices of 132 abundance were calculated. The tentative conclusion concerning these indices is that no clear-cut pattern of coyote population fluctuations since 1972 can be discerned. Analysis is continuing. Work also continued on the construction of an ecological energetics picture of the coyote via an extensive literature review. The bobcat population on the study area is very low according to all available evidence. Their very scarcity implies that their predation intensity on the raptor prey base is negligible. METHODS Coyote Food Habits A total of 285 km of scat lines were established on unimproved dirt roads throughout the study area. All but 39 km of these lines were located north of the river and no scat lines were located east of Grand View, Idaho. These lines encompassed the areas surrounding Initial Point, Coyote Butte, Christmas Mountain, Big Foot Butte, Castle Butte, Sinker Butte Farms, Sand Creek, the canyon bottom between Swan Falls and Melba, and the canyon rim between Melba and the Grand View-Mountain Home Highway. The scat lines were run twice a month during May, June, and July and once a month thereafter until November 1977 when scat collections ceased. Prior to May 1977, scat lines had not been set out and scat were collected whenever they were encountered while traveling through the study area. Other project personnel also collected any scat they found in the course of their field work. Stomachs and colon samples were collected from all coyote carcasses. J. Messick collected most of these. In addition, Messick had collected 46 coyote scat in 1975 and 1976. The scat are being analyzed by J. Whitman of the University of Idaho, and the stomachs are being analyzed by Mr. S. Hoffman of Utah State University. Coyote Population Parameters Collection of coyote carcasses continued during the year. In September 1977, 61 coyotes were collected by aerial gunning. During the winter and spring, carcasses were collected primarily from trappers. During the summer, road kills provided the bulk of the collections. J. Messick collected most of the carcasses during his three years as Research Associate on Study IV-B. A canine, pulled for aging purposes, was collected from all of the carcasses except the 61 carcasses from the September 1977 aerial gunning. A premolar was collected from these animals. Reproductive tracts were collected from female carcasses. 133 All of the teeth collected during the duration of the study were decalcified and aged. Teeth were decalcified in a five percent nitric acid solution. Teeth were checked every four to six hours to insure that decalcification had not proceeded too far. The nitric acid solution was changed every 12 hours. Canines required 12 to 72 hours for decalcification, and premolars required 4 to 8 hours. The teeth were rinsed in a cold water bath for 48 to 72 hours following decalcification and then stored in 70 percent ethyl alcohol until sectioning. Prior to sectioning all teeth were again rinsed in a cold water bath for 24 hours. Teeth were sectioned in a cryostat; several sections from each tooth were placed on a slide and stained with Paragon stain. The number of annuli was determined immediately after staining. Linhart and Knowlton (1967) stated that the age of the coyote was determined by totalling the number of annuli and then adding one to that total to obtain the correct age. This was because coyotes do not lay down an annulus until their second year. Knudsen (1976) suggested a more complicated procedure based on time of carcass collection, but his arguments were not deemed convincing. The methods of Linhart and Knowlton (1967) were followed here. Since coyotes do not lay down an annulus until their second year, neither juveniles nor yearlings will have cementum annuli present in their teeth. Another method must be used to separate juveniles from yearlings. All canine teeth were cut in half at the gum line and the ratio of pulp cavity length to total tooth length of the tooth cross-section was determined. According to Knudsen (1976) , the animal should be considered a pup if this ratio is greater than 0.32 and an adult if the ratio is less than or equal to 0.32. Separating juveniles from yearlings was more subjective in the case of premolars. The technique used was suggested by Knudsen (personal communication) . It involved separation of these two age classes based on the thickness of the cementum and the distance the cementum extended up the side of the tooth . Survival estimates were calculated from the resulting age structures using the Chapman-Robson method (Robson and Chapman 1961) . Reproductive tracts were stored in either AFA or Bouin's Solution. Uteri were examined for the presence of placental scars. Ovaries were sectioned and examined for evidence of follicular activity, corpora albicantia, or corpora lutea depending on date of collection. The U.S. Fish and Wildlife Service has been conducting a western U.S. predator survey every fall since 1972 using a technique described in Linhart and Knowlton (1975). Briefly, the technique involves the establishment of a 23.7 km line with 50 scent stations set 0.5 km apart on the line. The scent stations consist of a 13b small plastic capsule filled with a synthetic volatile fatty acid mixture (prior to 1976 a commercially produced fermented egg attractant was used) placed in the middle of a 1 m diameter circle of sifted earth. These stations are placed alternately on the left and right side of the road. The survey lines are run for five consecutive days and the scent stations maintained as required. A "coyote visit" occurs when one or more identifiable coyote tracks occur inside the circle of sifted earth. The index of relative abundance is calculated as follows: Total coyote visits x 1000 = index Total operative station nights For more information of the details of the survey and the methodology, see Linhart and Knowlton (1975). Two survey lines have been run on or near the study area since fall 1972. One of these is located near Christmas Mountain on the study area, and the other is about 19 km east of Mayfield near Willow Creek. Since fall 1974, two other lines have been run in the vicinity of Glenns Ferry. The results from these lines represent an index of coyote abundance in the general vicinity of the study area. Every fall and winter since 1972, personnel of Study I have conducted an aerial survey of golden eagles (Aquila chrysaetos) in the study area. All coyotes seen during these flights are also recorded. The average number of coyotes seen per transect was computed and these figures represent another index of abundance. The third index of abundance was calculated from the number of coyotes seen by personnel of Study 1 between 1 April and 1 July in 1971, 1975, 1976, and 1977. These researchers made regular visits to the same raptor nests every four days during these periods. All coyotes seen were recorded in field notes, and the average number of coyotes seen per man-day constitutes an index of abundance . In an effort to determine the distribution of coyotes on the study area, 110 scent post stations were established on the study area in September 1977. S. Cherry set out 90 of these and J. Messick was responsible for 20. These 110 stations were set out in lines 4.4 km long with 10 stations per line. They were run only one day on the advice of Mr. Robert Roughton, U.S. Fish and Wildlife Service, Logan, Utah. Coyote Bioenergetics Work proceeded on the development of an energetics profile of the coyote via an extensive literature review. 135 Bobcat Segment No known bobcat scats were collected. Five bobcat colon samples were mailed to J. Whitman for analysis. The total number of bobcat carcasses collected totals less than 20. This small size precludes fulfillment of the objectives of this study with respect to the bobcat. No aging was attempted and no reproductive tracts were analyzed. RESULTS Coyote Food Habits A total of 945 coyote scats were collected during the year. These scat are being analyzed at the present time. Results are not available as of this writing. Coyote Population Parameters The results of the analysis of 33 coyote reproductive tracts collected between January 1975 and April 1977 are presented in Table 1. The results from the analysis of the 27 reproductive tracts collected in September 1977 are presented in Table 2. The data in these two tables can be separated into three groups. Those animals collected between 26 February 1975 and 25 January 1976 represent animals from the spring 1975 breeding population (Table 1) . Coyotes 28, 30, and 43 in this sample were collected from around Emmett, Weiser, and Marsing and have not been included in this group. Those animals collected between 30 Jpne 1976 and 1 January 1977 represent animals from the spring 1976 breeding population (Table 1). The animals collected between 2 February 1977 and 21 March 1977 (Table 1) and those collected in September 1977 (Table 2) can be considered a sample of the spring 1977 breeding population . Only two animals from the spring 1975 breeding population sample had placental scars. The average number of placental scars per female was 6.5. Five out of nine animals in this collection showed some evidence of reproductive activity. Animals 91 and 107 had ovulated and would have probably given birth had they survived. Four coyotes in the spring 1976 breeding population sample had placental scars with an average of 6.0 scars per female. Six of 11 animals showed evidence of reproductive activity. The collection of coyotes representing the spring 1977 breeding population had 12 animals with visible placental scars. The average number of scars per female was 6.67. If coyote number 18 (Table 2) is excluded, the average is 7.0. Animals 918, 920, and 944 had ovulated and may have produced litters had they survived. Eighteen out of 35 of the spring 1977 breeding population sample showed some evidence of reproductive activity. 136 Table 1. Results of analysis of 33 coyote reproductive tracts collected between January, 1975 and April, 1977. (CL-corpora lutea; CA-corpora albicantia; PS-placental scars) . Coyote Processing CL CA PS Comments Number Date 28 1/27/75 - 0 0 30 1/27/75 - 0 0 43 2/7/75 - - - 91 2/26/75 5 - - 107 3/7/75 6 - - 273 9/6/75 - 0 0 294 12/1/75 - 0 0 295 12/1/75 - 7 7 296 12/4/75 - - 0 304 1/16/76 5 - 6 305 1/25/76 - 0 0 306 1/25/76 - 0 0 524 6/30/76 - 0 0 730 8/3/76 - - - 791 10/5/76 - 3 3 792 10/5/76 - 0 0 814 10/23/76 - 10 7 328 11/11/76 - 14 9 846 11/25/76 - 0 0 848 11/25/76 - 0 0 853 11/25/76 - 0 0 869 12/11/76 - 2 0 878 12/25/76 - 4 0 883 1/1/77 - 11 5 901 2/4/77 - - - 918 2/28/77 5 - — 920 2/28/77 2 - - 921 2/28/77 - - - 930 3/12/77 0 - - 942 3/21/77 0 — — 943 3/21/77 0 - - 944 3/21/77 6 - — 945 3/21/77 7 - 7 Evidence of follicular activity Implantation sites not visible Implantation sites not visible Evidence of follicular activity PS faint. One ovary missing Decomposed and unreadable One ovary unreadable PS faint, at least 5 Evidence of follicular activity Only one ovary ovulated One ovary missing Decomposed and unreadable Implantation sites not visible Uterine tissue present in cervix and vagina. Animal either aborted in trap or parturition had occurred. 137 Table 2. Results of analysis of 27 coyote reproductive tracts collected during September 1977. (CA — corpora albicantia; PS — placental scars). Coyote number CA PS Comments 2 2 0 4 0 0 6 8 6 7 0 0 9 11 8 14 12 9 17 13 10 18 8 3 Minimum PS count. One extensive area of discoloration in uterus probably site of overlapping PS. 19 4 7 Not that unusual. CA begin to fade during course of year. Ovaries were triple checked to verify CA number. 24 0 0 25 10 3 Very faint PS not counted. 26 0 0 27 0 0 35 0 0 36 0 0 37 5 7 42 14 5 43 3 8 One ovary missing. 44 0 0 45 0 0 47 0 0 51 0 0 56 8 7 57 0 0 58 0 0 59 4 0 60 0 0 138 Out of 241 coyote teeth that were aged, 151 were chosen for use in the estimation of survival rates. The Chapman-Robson technique was chosen to estimate survival rates (Robson and Chapman 1961), and the results are presented in Table 3. Animals collected during the winter months (October to March) were chosen because of larger sample sizes. The winter 1975-1976 sample could not be broken down like the winter 1976-1977 sample because most of these animals came from one large coyote graveyard and date of death was unknown. In many cases, the juvenile age class had to be discarded because they were not properly represented in the age structure according to the results of a statistical test found in Robson and Chapman (1961). Adult survival was generally around 0.70. Total survival (adult and juvenile survival combined) was approximately 0.60 in the winter 1976-1977 sample. The results of the calculation of relative indices of abundance offer no clear-cut pattern as to what the coyote population has been doing since 1972. No conclusions can be drawn at this time. Results are presented in Tables 4, 5, and 6. The results of the area-wide survey conducted in September were disappointing. Only two coyote visits were recorded giving an index of 18.5. The visits were recorded on lines near Castle Butte and Christmas Mountain. Coyote Bioenergetics No results of this part of the study can be presented at this time . PLANS FOR NEXT QUARTER A final report will be prepared. LITERATURE CITED Knudsen, J.J. 1976. Demographic analysis of a Utah-Idaho coyote population. MS Thesis, Utah State University, Logan. 195 pp. Linhart, S.B. and F.F. Knowlton. 1967. Determining age of coyotes by tooth cementum layers. J. Wildl. Manage. 31 (2) : 362-365 . , 1975. Determining the relative abundance of coyotes by scent station lines. Wildl. Soc. Bull. 3 (3) : 119-124. Robson, D.S. and D.G. Chapman. 1961. Catch curves and mortality rates. Trans. Am. Fish. Soc. 90 (2) : 181-189 . Table 3. Survival rates of coyotes on the Snake River Birds of Prey Natural Area as estimated by the Chapman-Robson method. Numbers in parentheses represent sample sizes. Total survival represents a survival rate estimated with the juvenile age class included in the age structure. See text for an explanation of why a total survival rate could not be estimated for all age structures. 139 G > 5-i 0) QJ 4-1 P c 03 M pci G • a > G 5-4 g G 03 CO •H ip \ — 1 G 03 o 4-J u O H 6"S m CS > 5-i CO G i — I 4-J cU G P ptf o H ! 03 > P g g P 4-1 G 03 i — r cz xf CO 00 G . a > G p a) 3 G3 CO •H IH 4-J G i— I O G U 03 < B-S m as V sO m rx 7 CO fx sO i — I o3 G Pd 03 < sO m rx 03 O •H 5-i 03 rx rx as CM CO Csl co co I I I I I O O O O O ix ix rx ix rx CM Csl i — l i — i O G 4J G •H S I in rx as sO co oo m Csl Csl fx oo oo v/ V 5/ n 1 — 1 1 1 CM 1 1 as I SO n n e • • I r — 1 G3 V sj p G QJ 1 V QJ G P xf xf as 4-J C G G p CO n o QJ G O S O 03 CO n G3 G o Pm G CO CO n n • • • 1 — 1 ip G e e e e CO 5-4 o • °H 1 1 oo n Os o G •H 0) 00 hJ oo xf rx n QJ •H p > r— 1 1 — 1 aj rG 03 -> •H G t3 00 G G Xf IX o n Csl 00 o G G P G 00 I 1 xf CM oo i — 1 Csl Csl n P •H • U °H i — I CO Cs] % / x / CO c 03 pj 1 || I G 0) b n T 1 00 QJ G P o • Cs) a> as rG 5-4 G 1 — 1 Po sO n n P P 1 — 1 p e e » QJ CO •H p P rG s G G as CO 1 1 i — 1 G P T3 Pm G rx 1 1 ^d- co n SO QJ QJ p •H i — 1 1 1 Csl C C P G 03 s •H •H G C 03 G G c G QJ S G G CJ T3 1 1 O'! rx CO as as so G G o 00 CO • 5-4 P p CO CO oo 00 rx co o 5-4 QJ G p O 03 • QJ P G G Pm G n n o n n O QJ CO 03 O G G \ / V V \J V V a QJ 3 1 G • *H o o o co CO o <3" o CO rx O G QJ 00 s co CO so as SO n 00 CM o so 1 1 G ip 03 T3 > ' 1 rx SO rx rx so rx o3 •H QJ •H G G u rG P *H CO 03 G G 1 1 p •H 03 P o fp C 03 QJ G G G n SO n n n o CO • a O G G n o CM o sO oo oo /"""S QJ CO G s •H i — 1 i — 1 1 1 CO n rx n o a • G rx SI 1 1 l — 1 i — 1 1 — 1 Csl r— 1 •H S T3 03 1 — 1 x / > — s 'x— ' Xw^ x^ 'G G G C QJ G 6 03 o xf o Xf rx o H rG rG p G n CO m O so 1 — 1 p G 03 C O'- so ix rx so rx •H •H p Csl CO n sO IX • « 5-i p G rx rx rx IX rx rx xf rG O rG G as as as as as as P a • >4 1 — 1 i — 1 1 — 1 1 — 1 i — 1 i — 1 QJ 03 G XI 1 1 QJ 03 p 3 1 ix rx ix rx sO r-a G QJ G G rx rx rx rx rx rx G •H P G •H \ "X. s-i as H e Ph G ip l40 Table 5. Average number of coyotes seen per transect during the aerial survey of golden eagles on and near the study area. Year Avg. No. coyotes/Transect Fall 1972 0.80 Winter 1973 0.95 Fall 1973 0.00 Winter 1974 0. 70 Fall 1974 0.40 Winter 1975 0.80 Fall 1975 0.45 Winter 1976 0.30 Fall 1976 0.10 Winter 1977 0. 10 Table 6. Number of coyotes per observer-day as determined from ana- lysis of field notes covering the period 1 April to 1 July for 1971, 1975, 1976, and 1977. Year Coyotes/Observer-day 1971 0. 155 1975 0.090 1976 0.166 1977 0. 153 l4l 142 Most Townsend ground squirrels failed to breed during the drought of 1977, and most estivated a month earlier than usual. Lack of green vegetation is blamed for the breeding failure. (BLM photo by Melquist). STUDY V: Population ecology and habitat requirements of Townsend ground squirrels. CONTRACTOR: Department of Biological Sciences, University of Idaho, Moscow, Idaho 838H3 INVESTIGATORS: Donald R. Johnson, Principal Investigator Graham W. Smith, Research Associate E. Martin Cobb, Research Aid Charles E. Woodley, Research Aid PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1002 OBJECTIVES : 1. Describe population variables of ground squirrel populations including productivity, recruitment, age and sex ratios, and turnover rates. 2. Ascertain population density (a) in representative habitat types, (b) in relation to land use patterns including the effects of cultivation on density, and (c) in hunting areas of those raptor outlined by the radio-telemetry study (Study III). 3. Ascertain food habits and habitat requirements in the major vegetation types. h. Isolate factors (predation, disease, parasites, etc.) which may limit ground squirrel density and distribution. ANNUAL SUMMARY Populations of Townsend ground squirrels ( Spermophilus townsendi ) were studied from January through July 1977, the third year of a four- year investigation of squirrel ecology and habitat requirements. A wide- spread drought, especially the lack of fall precipitation, reduced the germination of cheatgrass ( Bromus tectorum ) , the primary food source for squirrels during the late winter and early spring. As a result, little green forage was available to squirrels when they emerged from torpor in mid-February. Overwinter survivorship was reduced for all sex and age groups. Only ground squirrels with access to green vegetation, such as that of irrigated alfalfa fields, reproduced. Non-reproduct ive squirrels entered torpor in late April and early May, at least a month earlier than in non-drought years. Body weight at submergence was significantly lower than in 1975 and 1976. The monitoring of plague (Yersinia pestis ) in ground squirrel and badger populations continued. Eight ground squirrels have been found positive for plague since this study began. Plague titers in the badger ( Taxidea taxus ) population will "be reported "by personnel of Study IV-B, Ground squirrel vulnerability to avian predation was determined by counts of visible squirrels at index points within the hunting areas of raptors as determined in Study III. Badger hunting attempts were deter- mined by counts of digs on intensive study sites and along 2^h transects throughout the study area. Badgers concentrated hunting efforts at sites of greatest squirrel density. Because of the localized squirrel repro- duction, predators lacked access to a vulnerable juvenile cohort avail- able during years of normal rainfall. INTRODUCTION Townsend ground squirrels are the primary prey of prairie falcons (Falco mexicanus) (Ogden and Hornocker 1977), red-tailed hawks ( But eo jamaicensis ) and ravens ( Corvus cor ax) (Kochert 1973) which nest within the Birds of Prey Natural Area (BPNA). Ground squirrels comprise a minor portion of the diet of golden eagles (Aquila chrysaetos ) (Beecham 1970; Kochert 1972 , 1973). Badgers, which occur in large numbers within the BPNA, and rattlesnakes ( Crot alus viridis ) also prey heavily upon ground squirrels (Hornocker et al. 1976; Johnson and Diller, 1976). Recognizing the importance of the Townsend ground squirrel in this ecosystem, the Bureau of Land Management has contracted a four-year study of its ecology. STUDY AREA Six intensive study sites, each located in a major vegetation type, were live trapped in 1977. Each site is 1 ha in area (100 m x 100 m) with grid stakes at 10 m intervals. These sites are: 1. Bedspring Site. This site is used yearly as a sheep camp during the winter grazing season. The site was nearly bare (9&% bare ground) in 1977 (Table la). 2. Eurotia Site. This site supports a depleted stand of winter- fat ( Cerratoides lanat-a ) which was heavily grazed during the past winter by sheep (Table la). 3. Burn Site. A range fire of 2k June 197^- destroyed the original vegetation at this site. In 1977 there was a sparse cover of Sandberg bluegrass (Poa sandbergii ) and winterfat at this location (Table lb). Comparison Site. The shrub cover at this site which consists of big sagebrush ( Artemisia tr identata) , winterfat and hop sage ( Grayia spinosa) has steadily decreased during the past three years ('Table lb), although there was little range use during the winter of 1976-77. Table la. Plant coverage (%) May 1975, 1976 and 1977 on the intensive study sit 145 c/3 (L) ft- o 1 ft c— • 1 1 . cp 1 — 1 1 i — 1 i — 1 1 OO ■ — i V V f CP , — „ ft w 0 p ft VO •H OO 00 o 0 ft- p . . • 1 . 1 — 1 cp ft 1 — 1 i — 1 VO i — 1 1 o Pi 1 — 1 cn V VO 1 oo 0 ft Q 0 ft x UP CM VO CM ft- . • 1 . CP 1 — 1 CM O ft- 1 UP 1 — 1 V VO 1 oo P •H -p o p P ftl ft- 00 1 1 o ft- ft • 1 1 • . CP c— 1 1 CM 1 — 1 o 1 — 1 V CP P hi •H ft VO -P CVJ O ft CM -p ft- o . • • • r— f CP p CM i — 1 i — 1 -zT ft- VO P 1 1 p 1 — 1 V V ft- cu ft ft v — ' UP UP UP CM oo 1 — f ft- • . . . . CP -ft 1 — 1 1 — 1 00 o 1 — 1 1 — 1 1 — 1 V CM VO P ft P P P p p O p p •H 1 — 1 1 — 1 p ■H ft o hO o C/3 ft P ft 0 (D o 0 0 Ip P 0 ft o •r— 1 P ft ft O 0 P p ft ft ft 03 P 0 p P •H P 03 p p P O S ft o P P 03 o P 03 p 0 3 CD P CD p O 0 bfl ft O 03 O 03 m ft ft P ft 03 0 P P P p ft o P p CO ft CP PQ 'Table lb. Plant coverage {%) May 1975, 1976 and 1977 on the intensive study sites. 146 t- ip- ON I — I — ' V t— I rH I CO V On ft 0 C p p ft VO C t — P On d i — I ft LTN P- On i — I i — I V i — 1 On l r\ 1 VO CO 1 CO O CO _ . « • i — ! -ft LTN | ON V CO r- On — ft 0 p p P ft ft c o vo co t— *H ON P i — i cci ft £ O o LTN t- ON — 1 OJ oj vo -^r ft v i — i 1 i — I I On V CO' LO CO ft CO ft- i — i -ft- LTN ft i — l CO | C\J V V CO CO ft" CO ON LO OJ ft -ft CO r — 1 i — I CO i — ! On V V ft P P -P ft P P g P G 0 p O P ft P 3 •rH i — 1 i — 1 ■H CCS p •rH ft p o o bO o CO ft c ft P ft 0 •H 0 c_> 0 o ft P ft P 0 ft o 1 — •( — ! CO p ft ft o CCS 0 C P -p •rH P CO P o p £ •H •H P co p p 0 ft. O £ ft p ft P1 P co o P CO 0 p p P 0 P O 0 o < ft co o co ft ft ft p ft m ft p o ft co cc3 P ft ft G P o p bO 0) P P ft 5. Artemisia Site. The vegetation at this location consists of a stand of big sagebrush with a sparse grass understory (Table lc). 6. Shadscale Site. This site is dominated by shadscale ( Atriplex conferti folia) and budsage (Art emisia spinescens ) with some horsebrush ( Tetradymia glabrata) and big sagebrush. There was no grass understory (Table lc ) . METHODS Field work began in January and live trapping resumed when ground squirrels emerged in February. The field season ended in mid-August. A minimum of twenty live traps were set at each site in the morning and checked periodically throughout the day. Apple was used as bait. Trapped squirrels were anesthetized with ether and ear-tagged using self-piercing monel tags (National Band and Trap Co., Newport, KY ) . Squirrels were permanently marked by toe-clipping. The capture loca- tion, body weight and reproductive condition (testes position and nipple visibility) were recorded. Density was determined by recording the number of home burrows occupied by adults and yearlings (suspected residents). Home burrow locations were identified from trapping results . The squirrels visible within a 360° sweep were counted periodi- cally from March through early May at index points. These points were established at locations hunted by birds of prey as determined in Study III. Plant coverage within these hunting areas is summarized in Table 2. Squirrels were collected using a .22 caliber rifle. Their weight and measurements were recorded. Reproductive tracts of females were examined for visible embryos or placental scars. The stomachs were removed and their contents examined using the techniques of Johnson and Groepper (1970). Badger and ground squirrel holes were counted on 2^h transects, each 400 m in length (area 0.2 ha). The transects were located in representative habitat types throughout the BPNA. 2 Plant coverage was estimated on 0.25 m circular plots on intensive sites (66 plots each) and at index points (20 plots each). Coverage of mosses, litter and rocks was included in the bare ground estimates. Taxa with less than one percent coverage were ignored. Total cover- age sometimes exceeds 100 percent because of stratification. Counts of badger digs on intensive study sites were recorded monthly. Table lc. Plant coverage (%) May 1975, 1976 and 1977 on the intensive study sites. Mosses, litter and rock included in bare ground. Taxa with coverage less than 1% excluded. Sums may exceed 100% because of stratification. b48 c— o 1 — 1 o\ th- 1 1 O r— . . • • 1 1 • o\ 1 — 1 CM VO 1 — 1 1 1 t — ■ — i CO o 1 — 1 o CO P oi X r r\ UJ VO LT\ LTV _p- O 1 o D— I • • • • 1 • Ov 1 CM CO 1 — 1 CM i LO i — l 1 ON b- r- 1 1 1 on • 1 1 1 • Ov OD 1 i 1 i — 1 1 — I on i — 1 V V oo •H C 0 ■rH £ (U -p p < VC i — 1 1 1 i O 1 — 1 o c— • 1 1 1 • . • OV 1 1 1 CM E— C— 1 — 1 CM VO k5 •rH cd i — 1 -P CO o o3 a5 £

. PS a •H •rH 1 — 1 P CO a3 p- s £ p. oj 0 CO o o 0 •H P CO g Pi CO -P -P p -P 0 o fccO P P P -p 0 CO P o 0 < < < Eh co pq Ph 0 P cc3 p X P a 3 w O PQ Table 2. Plant coverage (%) at index points. Taxa with coverage less than 1% omitted. Ik9 0 t— i— CO 1— 1 00 • o oo H LT\ [■ — LT\ Ph CVI 1 0\ 0 0 CQ -p OO r- VO ON P -P • O p CVI i — CVJ oo IP cq 1 on 0 p> (U O P 1 — CVJ OO fa, P » O 3 r— i oo oo cvj O PQ ! On hO P CVJ CO la cd . P *H OO I 1 VO Q P 1 ON EH Pi O P P on o -V o a . o p i — CVJ 00 I oo ffi P V On o CO i — 1 CO I-H oo CVJ cd O . P P 1 — -4- 1 1 O W V On. P P VO CSV CO C— •i — i » • . cd 1 — p- LTV 1 In — o CO 1 — On PH . . P C\J 1 — CVI 1 1 t— IP =tfc V ON o CVI p . . hp 1 1 CVJ 1 — 1 1 In- IP =fe 1 V On cd p cd cd p p p cd H 0 P P •H Vi cd P •H •H i — 1 o p p p p CO o 0 0 0 p cd P p p •H •H P p C/3 o P CO CP •rH p P pi CQ g cd O co e 0 P P 0) o cd CO p P faO CO p o p p 0 CO pq p p < ol 0 cd p p P p cd O CO pq 150 Spleens and livers of moribund and dead squirrels found on the study area, together with blood samples from badgers collected by Study IV-B personnel, were analyzed by the Public Health Service, Center for Disease Control, Fort Collins, CO, for the presence of plague organisms. RESULTS AND DISCUSSION Population Characteristics One hundred sixty-five ground squirrels were captured on the in- tensive study sites in 1977 (Table 3). All were adults or yearlings since no young were produced on these sites in 1977. Captures on the Burn, Comparison and Artemisia Sites were greater than in previous years. Following the range fire in 197^+, the Burn Site has shown a steady increase in ground squirrel density (Table U). As in the past, the Shadscale Site supported few ground squirrels (Tables 3 and 4). Squirrel density was reduced on the Eurotia and Bedsprings Sites probably as a result of heavy winter grazing by sheep and the effects of the drought. Rainfall from September through December 1976 amounted to only 37 mm (Table 5). It was 67 mm the previous year when these sites supported a vigorous growth of winterfat and cheatgrass (Table la) . Hull and Pechanec (19^7) observed that late summer and fall rains induced rapid germination of cheatgrass, a winter annual. Germina- tion was poor in the fall of 1976, and few plants survived the winter. Squirrels emerged to find a barren landscape nearly devoid of green vegetation (Tables la, lb and lc ) . The incidence of recapture increased in 1977, probably because the squirrels found bait more attractive than in previous years when green vegetation was more readily available. The sex ratio of squirrels captured in 1977 favored males (52 percent) although females outnumbered males in 1976 (Table 3). Survivorship of all sex and age groups was reduced in comparision with last year (Table 5). Females and adults again exhibited higher survivorship than males and first-year squirrels respectively. Reproduction Adult males emerged with scrotal testes. No yearling squirrels were observed in breeding condition. None of the 80 females trapped on the intensive study sites bred in 1977* In 1976, however, all adult females and 62 percent of the yearling females bred. Females living in or near irrigated alfalfa fields raised litters. Juvenile squirrels produced at these sites remained active until the first week of July, a seasonal activity pattern similar to that found throughout the study area during normal (non-drought) years. In Arizona, Reynolds and Turkowski (1972) found that dry winters delayed the breeding season and reduced the litter size of round-tailed ground squirrels ( Spermophilus tereticaudus ) , but in no instance did they observe a suspension of reproduction by an entire population. Table 3. Sex and age of ground squirrels captured on study sites, 1975-77* Percent in parentheses. - = Not sampled. Site Adults and Yearlings Juveniles Total 1975 1976 1977 1975 1976 1977 1975 1976 1977 151 LTV CVJ i — I i — I 00 CVI O O VO i — I CVI CVI CVI H CVI CO ir\ H VO VO OO i — I i — I LTV O LTV OO CO VO i — I OV i— I LT\ O H L/A^r LTV H c — co n- -V lt\ o vo ro cvj ov m cvi vo CVJ 00 I— I H H 00 J- CV1 CVI CVI tv- CO CVJ LTV LTV LTV LTV CVJ CVJ t — H CVI 00 OO LTV i — I i — I o o UA LTV 00 OO CVJ CVJ i — I i — I VO CVJ o o o o o o o o o o o o O VO -P~ -H LTV LTV CVJ CVI LTV OV OO CVJ CO OO I — I I — I -=T OO CVJ CVJ i— I vo I — I o o o CVJ OO LTV CVJ i — I VO UA I — I I — I OO I — I OO O OV OO CVJ VO I — I co i — I CV) CVI OO o o VO CVJ CVJ vo H o I — I LTV CVJ I — I I — I H 00. I — I I — i CVJ o CVI CVJ o vo CVJ H HVO VO OO I — I I — I CVJ CO LTV J- LCV O LTV CO CO VO I — I O LTV CVJ CVJ i— I o CVJ CVJ O CVJ H CVJ CVJ 00 LTV o rH O I — I I — I I — I OO t— -H LTV ov O OV vo ov LTV i — 1 OO tv- _=J- LTV OO OO OO OV OV tv- 1 — 1 CVI i — 1 i — 1 CVJ -H 1 — 1 i — 1 1 — 1 i — i 1 1 1 1 VO co C fciO O aj a> £ CD 0) CD m CD •H CD ■ — i CD CD •H r — 1 aj 1 — 1 i — 1 •H i — 1 CQ i — 1 aj H H p cd 03 *H CL> aj CD aj £ CD aj •H CD aj o CD aj CD aj Pm i — i £ -P i — 1 £ i — 1 £ a3 i — 1 £ £ i — 1 £ co i — 1 £ i — 1 i — 1 £ co aj a> o aj 0) £ aj CD PM ctf CD CD aj CD aj CD aj aj CD H g 2 £ ^ O p PP O PQ m o < CO EH CVJ i — ! w 3 ra p 0) o * Incomplete 152 Table Density (N/ha) of adult and yearling Townsend ground squirrels on the intensive study sites in 1975 , 1976 and 1977* Site 1975 1976 1977 Bedspring male female 5-8 7 - 6 3 10-17 10 - l7 1-2 Eurotia male female 7 7-9 2-5 3 7 - 9 3 -7 Burn male female 3-7 3-7 5-8 3-9 8-9 5-8 Comparison male female 7-7 3-7 5-7 3-7 5-7 3-7 Artemisia male female 2-3 3-7 6-8 7-6 Shadscale male female 0-3 0-2 153 Table 5. Rainfall monthly (millimeters) at Swan Falls Power Station in the Birds of Prey Natural Area. NA = not yet available. Month 1975 1976 1977 January 9-5 15.5 9.0 February 12.0 16. 5 9.3 March 39-3 13.3 10.3 April 29.8 25-5 5-3 May 7-5 6.5 26.0 June 11.3 27.3 30.5 July 26.8 2.0 15.0 August 8.0 25.0 11.8 Sept ember 0 27.3 18.5 October Uo.o U.O NA November 9-3 3.0 NA December 17.3 3.0 NA Total 210.8 168.9 154 Tatle 6. Age-specific survivorsh i.p of Townsend ground squirrels. Percent in parentheses. MALE FEMALE ADULT FIRST YEAR ADULT FIRST YEAR 1975- 1976 21 of 68 (31) 12 of 63 (19) 26 of 75 (35) 21 of 42 (50) 1976- 1977 16 of 68 (24) 21 of 162(13) 26 of 90 (29) 31 of 151(21) Overall 37 of 136(27) 33 of 225(15) 52 of 165(32) 52 of 193(27) 155 The relationship of rodent reproduction to the availability of green vegetation and ultimately to rainfall in desert and semi- desert areas has been pointed out by Fitch (1978:556), Reynolds (1958) and recently by Beatley (1969, 1976). Certain investigators favor the hypothesis that green vegetation serves as a source of water and vitamins to rodents during reproduction (Beatley 1968, 1976; Bradley and Mauer 1971)- Soholt (1977) has recently reported that the heaviest use of green vegetation by Merriam kangaroo rats (Dipodomys merriami ) occurred during lactation. An alternative hypothesis to explain this relationship suggests that substances present in developing plants may bring the rodents into breeding condition (e.g. Chew and Butterworth 1967). Negus and Berger (1977) induced breeding in a wild population of Microtus montanus during the winter months by distributing freshly germinated wheat seedlings in the runways of these voles. A nearby control population remained in typical winter non-breeding condition. Ad- ditional studies on montane voles support this hypothesis (Negus et al. 1977), Plant compounds may also terminate reproduction in birds and mammals. Cured forage has been found to inhibit reproduction in quail (Leopold et al. 1976) and in microtines (Berger et al. 1977)- Hence plant compounds may both induce and inhibit reproduction in certain birds and mammals. Seasonal Activity Individual squirrels were observed in December and January, but most of the population emerged during the third week of February. The first squirrel was trapped on 13 February. In contrast to previous years, there was little difference with respect to sex or age as to emergence date. Cold weather may have delayed emergence in 1977 as has been observed in Uinta ground squirrels ( Spermophilus armatus ) by Knopf and Balph (1977)* Squirrel activity above ground was confined to the warmer mid- day hours in the late winter and early spring. Activity became bi- modal with a morning and late afternoon peak as the season progressed. Fitch (1978) observed a similar seasonal change in the activity of the California ground squirrel ( Spermophilus beecheyi ) . Post-emergence weight gain (in g/day) of adult males (0.008; N = 25) and that of females (0.009; N = 21 ) did not differ signifi- cantly from that of last year. With the suspension of reproduction, temporal weight gains by females were similar to that of males. Over -winter weight loss averaged 39 percent (N = 7)* In Alberta, over- winter weight loss of Franklin ground squirrels ( Spermophilus fr ank- lini ) was 77 percent (Murie 1973). The number of active squirrels decreased in late April as adult males and females began entering torpor. Yearlings followed in early 156 May. Submergence, as determined by trapping success, was at least a month earlier than in 1976 (Fig. l). The mean -weights of adult males and females at the time of submergence were 306 g and 236 g respectively in 1975 and 1976. Adult males weighed only 260 g and adult females 197 g at the time of submergence in 1977. These differences are sig- nificant statistically. Food Habits Table 7 summarizes food habits data for 1975-77- Grass leaves and stems, particularly those of cheatgrass, are important dietary compon- ents during late winter and early spring in all habitats. We did not attempt to measure changes in diet as a result of the drought, but grass comprised only eight percent of the stomach contents of nine s squirrels collected at a winterfat site in April 1977- In spite of the drought, grass remained an important food source of squirrels collected at a sage site and throughout the Sand Creek range in 1977- Because of poor cheatgrass germination and growth, it is likely that perennial grasses were utilized at these locations. The leaves of budsage represent a spring food source for populations living in shadscale habitats. As the season advances, squirrels begin utilizing grass seeds and the seeds and vegetative parts of a variety of forbs. Insects are infrequently taken. Most of those eaten are beetles and lepidop- terous larvae. There is no difference in diet between adults and juveniles, just as Fitch. ( 19^8) found for California ground squirrels. Mortality Eight squirrels have been found positive for plague during the past three field seasons. Although there has been no plague-caused epizootic in ground squirrels, badgers continue to demonstrate elevated titers of plague antibody indicating intermittent exposure. Badger hunting attempts on the Bedspring Site were much reduced in comparison with the number occurring in 1976. Hunting attempts increased on the Burn and Comparison Sites, however (Table 8). These hunting attempts parallel squirrel population changes at these sites (Table 3). There is a strong positive regression relation- ship between counts of badger holes and ground squirrel holes through- out the study area (Table 9)- We conclude that badgers shift hunting effort to those sites with greatest squirrel density. Vulnerability to Avian Predation Ground squirrel counts and hence vulnerability to avian predation varied daily and seasonally because of changes in above-ground acti- vity (Table 10). Counts usually increased during the morning hours, remained at low levels during the warmest part of the day, and increased Number of Captures 157 Fig. 1. A comparison of the number of adult and yearling squirrels captured per week in 1976 and 1977. Table 7. Estimated volume (%) of food items in 17^ stomachs. T = trace (less than 1 percent). Samples are mostly adult squirrels (Feb-April) and juveniles (May- July). 158 -p a G s <-< S ft G N A / A / 1 1 1 ' ^ A X- A A 1 ✓ A /■ — \ ^ A r5 ft O G G 1 1 ft G ft G ft £h i — 1 ft G ft G O G -P ft ctf G CD ft (D ft G -P CD ft G G ft G - ' < s ft ft < ft < ft I'D G ft < ft ft co CO v — ^ A. — ' S — * ^ a A -• ' ' ' ' ' V X Cm V -- A ' A ^ CQ CD ■ — • G G G ft ft r— 1 ft CD 1 — 1 ft -p •H CD ft ft •H G G ft CD G CD G ft CQ O ft G Cm ED •H ft S o CO G G ft o i — 1 e CD CQ i s ■H CD G ft ft CD G CQ -P CD G G ED ED CD CQ •H ft G •H •H G ft O co CQ] CQ ft pq CQ CQ ft i— I CM 4 LACO O 00 CM OO Lf\ O CM rH On i — I i — I i — I G G g ft G G < S 1-3 Table 8. Badger hunting attempts on intensive sites, 1975-1977. One attempt may include more than one dig. - = not surveyed. 159 cd •H co 'rH £ 0 -p p < t— c— On VO in- on I — I ip- c- on VO c— o\ P-J rH S O O LTV C— ON C— t— On I — I VO ic- on I — I LP\ ip- on I — I C— IP- On i — I VO c— ON I — I UP Ip- On i — I ip- c — On O O on c — co i — l O O O O CM -V LT\ CO CM UN i — I UP CM O OO O PO VO UP PO I I I I I O O O -V rH PO CO CM -V -V r-H I IOOOOONUPVOCMOO I CM -V O CO CO VO I I I I rH CM rH CM O O UP VO CM C — CM CM O i — I CM i — I CM VO i — I o PO UP UP CM H PO UN C" — o O O O O PO H I I I I I O O CM PO H CM CM CM CM t — i — I UP CM CM VO PO UP UP O O rH O C— P- PO I I I i — I i — I o UP I I p VO O o o -V UP 1 — 1 i — 1 IP- t— CO PO rH CO ft IP- i — 1 l — 1 o\ ON UP 1 — 1 o CO On i — 1 PO i — I 0 PQ UP 1 1 -V UP rH LT\ CM UP PO CO [p- 1 — 1 1 — 1 c — i — 1 PO i — 1 1 — 1 o On 1 — 1 i — 1 ft Ip, 0) p p t>5 ft o ft CU 0 ft cd -P S CU P o ,£ cd £ r£ i — i CO ( U o s s 1 — 1 -P 0 ft o •H 0) 3 -p O (U 0 cd £ C £ £ >5 r— 1 hO ft p> > o -p O cd CU cd ft cd 3 ft CD o o 0 o § S S ‘“O < CO o Q eh ON VO UP CM PO CM VO PO CO VO CM O On VO OO CO UP C — PO CO PO i6o Table 9- Correlation coefficients (r) of mean badger hole counts with respect to ground squirrel hole count on transects. * P < .05 Site N Mean Hole Counts r Significance Badger Squirrels Big sagebrush (north end) 26 29.7 17. 4 0.63 * Big sagebrush (south end) 31 16.2 8.1 0.82 Big sage-grass (north end) 13 U3.8 23.3 0.74 * Big sage-grass (south end) 11 21.8 1 — 1 1 — 1 1 — 1 0.92 * Grass (north end) 16 1+7-8 24.6 0.71 * Grass (south end) 37 23.3 8.7 0.87 Winter fat 8 1+6.0 36.9 0.83 * Wint er f at -hop sage 6 11.3 16.5 0.79 NS Big sage-winter fat 15 1+0,1 31.3 0.85 * Shads c ale 18 17 -1 16.1 0.75 * Shads c ale-wint erf at 35 17.3 17.1+ 0.81+ * Shads cale- sage 5 21.0 16.8 0.88 * i6i Table 10. The number of Tovnsend ground squirrels seen at the index points. Site Time - Number of Squirrels HLP #1 5 April 0930 - 8 1130 - 2 1330 - 3 1530 - 12 1740 - 13 23 April 0730 - 1 0930 - 2 1130 - 1 1330 - 2 1655 - 2 1830 - 6 HLP #2 5 April 0930 - 13 1130 - 2 1330 - 3 1530 - 12 1740 - 26 23 April 0730 - 0 0930 - 0 .1130 - 1 1330 - 1 1655 - 3 1830 - 8 Cairn 26 March 0930 - 13 1130 - 15 1330 - 9 1600 - 18 22 April 0730 - 9 0900 - 18 1100 ~ 2 1300 - 1 1500 - 3 1700 - 11 1855 - 8 l4 May 0930 - 0 Grass Knoll 26 March 0950 - 24 1150 - 19 1350 - 11 1610 - l4 22 April 0800 - 28 0915 - 12 1105 - 1 1305 - 1 1505 - 5 1705 - 13 1845 - 16 l4 May 1030 - 1 Rock Outcrop 26 March 1005 - 18 1200 - 6 l4lo - 10 1620 - 22 22 April 0810 - 7 0930 - 8 1115 - 1 1315 - 1 1510 - 1 1715 - 3 1830 - 6 l4 May 1130 - 0 Dry Triangle 26 March 1015 - 30 1210 - l4 1420 - 10 1630 - 12 22 April 0825 - 10 0945 - 6 1125 - 3 1330 - 4 1520 - 2 1725 - 17 1825 - 8 l4 May 1200 - 0 Coyote 8 April 0730 - 2 0930 - 6 1300 - 5 1615 - 4 8 May 0815 - 0 1000 - 0 1250 - 0 13 May 0830 - 0 1030 - 0 1335 - 0 1530 - 0 1820 - 1 Horse 8 April 0800 - l4 1000 - 12 1320 - 5 l64o - 5 8 May 0830 - 2 1015 - 1 1310 - 0 13 May 0850 - 1 1045 - 0 1355 - 0 1600 - 0 l84o - 1 Pole 233 8 April 0810 - 7 1010 - 1 1335 - 5 8 May 0845 - 0 1020 - 1 13 May 0905 - 0 1100 - 0 1400 - 0 1615 - 0 1850 - 0 Bedspring 8 April 0820 - 23 1020 - 7 1345 - 7 8 May 0855 - 0 1042 - 0 13 May 0925 0 0 1110 - 0 l4l0 - 0 1625 - 0 1929 - 0 162 again in late afternoon. By early May, few squirrels were available to raptors because of the early entrance into torpor at all but the irrigated parts of the study area. From mid-May until July, juvenile squirrels produced in the vicinity of alfalfa fields were the only ones available as prey for raptors. Prairie falcons, red-tailed hawks and ravens were observed hunting these squirrels. LITERATUKE CITED Beatley, J. C. 19^9* Dependence of desert rodents on winter annuals and precipitation. Ecology 50:721-724. Beatley, J. C. 1976. Rainfall and fluctuating plant populations in relation to distributions and numbers of desert rodents in southern Nevada. Oecologia 24:21-42. Beecham, J. J. 1970. Nesting ecology of the golden eagle in south- western Idaho. M.S. thesis, University of Idaho, Moscow. 48 pp. Berger, P. J., E. H. Sanders, P. D. Gardner and N. C. Negus. 1977- Phenolic plant compounds functioning as reproductive inhibitors in Microtus montanus. Science 195:575-577. Bradley, W. G. and R. A. Mauer. 1971. Reproduction and food habits of Merriam's kangaroo rat, Dipodomys merriami, J. Mammal. 52: 197-507. Chew, R. M. and B. B. Butterworth. 1964. Ecology of rodents in Indian Cover (Mojave . Desert ) , Joshua Tree National Monument, California, J. Mammal. 45:203-225. Fitch, H. S. 1948. Ecology of the California ground squirrel on grazing lands. Am. Midi. Nat. 39:513-596. Hornocker, M. G. , J. P. Messick, J. S. Whitman, R. M. Olson and G. W. Smith. 1976. Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey National Area, Idaho, pp. l8l- 201 In. Snake River Birds of Prey Research Proj . Ann. Rep. 1976. U.S. Dept. Interior, Bur. Land Manage., Boise, Idaho, 240 pp. Hull, A. C. Jr., and J. F. Pechanec . 1947- Cheatgrass-a challenge to range research. J. For. 45:555-564. Johnson, D. R. and K. L. Groepper. 1970. Bioenergetics of North Plains rodents. Am. Midi. Nat. 84:537-548. Johnson, D. R. and L. Diller. 1976. Ecology of reptiles in the Snake River Birds of Prey Natural Area, pp. 165-179 In_ Snake River Birds of Prey Research Proj, Ann. Rep. 1976. U.S. Dept. Interior, Bur. Land Manage. , Boise, Idaho, 240 pp. 163 Knopf, F. L. and D. F. Balph. 1977. Annual periodicity of Uinta ground squirrels. Southwestern Nat. 22:213-224. Kochert , M. N. 1972. Population status and chemical contamination in golden eagles in southwestern Idaho. M.S. thesis, University of Idaho, Moscow. 102 pp. Kochert, M. N. 1973. Density, food habits and reproductive performance of raptors in the Snake River Birds of Prey Natural Area. U.S. Dept. Interior, Bur. Land Manage., Boise, Idaho, 54 pp. Leopold, A. S., M. Erwin, J. Oh and B. Browning. 1976. Phytoestrogens: adverse effects on reproduction in California quail. Science 191:98-100. Murie, J. 0. 1973. Populations characteristics and phenology of a Franklin ground squirrel ( Spermophilus franklini ) colony in Alberta, Canada. Am. Midi. Nat. 90:334-340. Negus, N. C. and P. J. Berger. 1977. Experimental triggering of reproduction in a natural population of Microtus montanus . Science 196:1230-1231. Negus, N. C., P. J. Berger and L. G. Forslund. 1977- Reproductive strategy of Microtus montanus. J. Mammal. 58:347-353. Ogden, V. T. and M. G. Hornocker. 1977. Nesting density and success of prairie falcons in southwestern Idaho. J. Wildl. Manage. 41:1-11. Reynolds, H. G. 1958. The ecology of the Merriam kangaroo rat (Dipodomys merriami Mearns) on the grazing lands of southern Arizona. Ecol. Monogr. 28:111-127. Reynolds, H. G. and F. Turkowski. 1972. Reproductive variations in the round-tailed ground squirrel as related to winter rainfall. J. Mammal. 53:893-989- Soholt , L. F. 1977* Consumption of herbaceous vegetation and water during reproduction and development of Merriam' s kangaroo rat, Dipodomys merriami. Am. Midi. Nat. 98:445-457. 164 Gayle Sitter, Research Associate, weighs a brood of prairie falcons in his study of growth and feeding behavior. (BLM photo by Sitter) 165 STUDY VII: Feeding activity and behavior of prairie falcons. INVESTIGATORS: Steven R. Peterson, Principal Investigator. Gayle M. Sitter, Research Assistant. Bradley W. James, Assistant. PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract No. 52500-CT5-1064 . OBJECTIVES: 1. Measure the quantity of food necessary to fledge prairie falcons in Birds of Prey Natural Area. As subsidiaries of this objective, we will: a. Compare methods of analyzing species and age of prey brought to the scrape. Automatically set time-lapse cameras. Radio-controlled time-lapse cameras. Visual observation. Pellet analysis. b. Compare the daily number of feeding visits and the quantity of food brought to and consumed at the scrape with one, two, four, or six young. c. Investigate prey-caching by prairie falcons. 2. Examine nesting density in relation to cliff relief. 3. Investigate post-fledging mortality. ANNUAL SUMMARY Food habits studies indicated an increase in number of species used as prey items. Townsend ground squirrel (Spermophilus townsendi) remained important as prey even though population levels were low. Growth rates in 1977 were significantly different from 1976. This was attributed to change in diet and weather conditions. Disease was probably the largest contributing factor in mortality of young falcons. Ectoparasites, intraspecific aggression, and predation were also indicated as mortality factors. METHODS Study Area Field work for the 1977 season was concentrated in an area between Swan Falls Dam and Halverson Lakes. This is an expansion from the 1976 study area and was needed to provide additional 166 scrapes that permitted construction of blinds close to the scrapes. The additional area is similar to the 1976 study area with a dense population of prairie falcons (Falco mexicanus) which use the adjacent desert as hunting areas (Dunstan et al. 1976). Scrape Selection Preparations for the 1977 field season were initiated on 5 February 1977 in a meeting with the Project Leader. At that time, a proposed field plan was set up to make the most efficient use of field time and still concentrate work on desired objectives. Field work started in mid-March when Brad James, Field Assistant, reported to the area. He spent the first few weeks selecting appropriate sites in which to establish blinds in close proximity to the prairie falcon scrapes. He noted possible nest sites during observation periods and while walking near nesting cliffs. We later confirmed these as nesting sites through periodic observations or through physical inspection. Sites were rated using several criteria. 1. Structure - The structure of the scrape must allow construction of a blind close enough to observe activities within the scrape. 2. Accessibility - The scrape must be accessible to permit collection of pellets, prey remains and growth data on the young. The blind must be accessible with easy access to permit movement of researchers in and out of the blind with a minimum of disturbance to the birds. 3. Distribution - Several scrapes needed to be localized to allow simultaneous study without extensive travel by the researchers. 4. Nesting Chronology - Nesting stages must be comparable to other sites being studied to standardize effects of changing weather or prey availability. 5. Disturbance - Human interference from the general public must be minimal or at least comparable to other study sites . We selected ten scrapes as initial study sites with three being proposed as blind sites. The additional sites were used as study sites for growth measurements, mortality rates, and alternate blind sites. Blinds We constructed blinds at the three intensively studied scrapes. A portable 4x4x5' (1.2x1. 2x1. 5 m) blind made of V (1.3 cm) electrical conduit with canvas attached was used until permanent blinds could be constructed. Permanent blinds were constructed out of scrap lumber and woven-wire fencing. They were camouflaged with vegetation and rocks from the immediate area. We build the permanent blinds during short work periods on warm days to minimize the effect of the disturbance on young and eggs (Fyfe and Olendorff 1976). Most of the work was done after the eggs had hatched. After each construction period, we observed the prairie falcon adults from a distance until the adults appeared to accept the structure and returned to normal activities. The prairie falcons accepted the blinds readily and returned to normal activities almost immediately after we left the area of the scrape. On only one occasion the falcons refused to settle down within 10 minutes so we removed the offending structure. Due to cliff structure, we needed to move the young slightly to allow unobstructed viewing from the blinds. We accomplished this by placing the young in a new nest cup, and filling the unviewable portion of the scrape with rocks. The young were thus in sight at all times but we still allowed considerable room to move about as they grew older. The longest move was a distance of 2-3 m and involved placing the young in a different cavity. The adults accepted the move and fed the young shortly after we left the new scrape. Seasonal Work Plan Personnel, from Study III, radio-equipped the adults from two of the intensively studied scrapes (Trail and Camera) . Both males had been banded previously (see Study III) , and the camera male had also been radio-tagged the previous year. This was the third consecutive year for intensive study of the camera scrape. The use of radios helped considerably during our periodic observations to locate and identify prairie falcons in the study area. During egg laying and early incubation, we made occasional observations at the study sites and recorded information on feeding visits and intraspecific interaction. An observation period ranged from several hours to an entire day. After the young had hatched, our observation periods were dictated by a strict sampling schedule. We observed each of the three intensively studied scrapes from dawn to dusk once every four days. Observational periods on two of the three scrapes (the radioed falcons) were correlated with Study III personnel to provide a complete picture of feeding trips. We compared our data obtained during these days with field notes from Study III to clarify movements of the adults and detail where kills were made. We entered all study scrapes every fourth day for collection of pellets and prey remains. 168 We analyzed these collections at the end of the field season and identified the remains by comparing them to specimens at the University of Idaho. Growth Information Prior to hatching, we visited all selected scrapes every two or four days to ascertain hatching dates. All visits before hatching were as short as possible with no planned disturbance between visits. When the young hatched, we marked individual young by coloring the down on the head with a felt tipped marking pen. WTe could then identify each individual in the scrape by color and could follow growth patterns through the brood-rearing period. Although the color will last up to two weeks, we remarked the young on each visit to allow quicker identification. When the young were three weeks old they were sexed using size and weight differences and banded. After hatching, we collected growth data every time we gathered pellets and prey remains, usually every fourth day. On each visit, we weighed the young by placing each individual in a sack and using a Pesola tubular scale to determine gross weight. WTe collected wing length data (measurement from wrist to tip of longest primary along outstretched wing) starting when primary pins emerged. To minimize possible field errors both wings were measured and the average figure was plotted. Mortality Any disappearance of eggs or young between our routine nest visits was recorded along with any possible causes. Throughout the nesting and fledging periods, we conducted periodic searches for dead prairie falcons along specified portions of the cliff. We collected any dead young found during these searches or during routine activities and recorded the bird’s condition, location, and possible causes of death. Age of dead young was determined by comparing the wing measurements to known aged birds. RESULTS Food Habits A total of 43 collections from 5 nest cavities showed 5 species of mammals, 16 species of birds and 6 species of lizards (Table 1) . This represented an increase in number of species in all categories when compared to 1976 data (Peterson et al. 1976:231). The larger number of prey species observed this year may be due to the decreased availability of young Townsend ground squirrels (Spermophilus towns end i) . The fact that an increased number of 169 Table 1 . Items found in pellets or remains found in prairie falcon scrapes. SCIENTIFIC NAME N Mammals Spermophilus townsendi 39 Peromysous sp. 1 Miorotus montanus 1 Sylvilagus nut tail'd 2 Dipodomys sp. 3 Unidentified 2_ Total Mammals 48 Birds Faloo mexioanus 1 Eremophila alpestris 3 Sturnella negleota 11 Petroohelidon pyrrhonota 3 Carpodaous mexioanus 5 Aleotoris graeoa 1 Columba livia 11 Pipilo ery throphthalmus 5 Chondestes grammaous 2 Stumus vulgaris 2 Zenaida macroura 1 Turdus migratorius 2 Euphagus oyanooephalus 2 Pica pica 1 Molothrus ater 1 Charadrius vociferus 1 Unidentified 10 Total Birds 62 Lizards Phrynosoma platy rhinos 2 Chemidophorous tigris 8 Crotaphytus wislizenii 1 Crotaphytus collaris 1 Sceloporus oocidentalis 4 Unidentified 2 Total Lizards 18 %(N) Estimated Total Biomass (g) 7 /o Biomass 30.5 5796 37.6 0.8 17 0. 1 0.8 60 0.4 1.6 1300 8.4 2.3 167 1.1 1.6 84 0.5 37.6 7424 48. 1 0.8 42 0.3 2.3 86 0.6 8. 6 1595 10.4 2.3 45 0.3 3.9 125 0.8 0.8 559 3.6 8.6 3630 23.6 3.9 215 1.4 1.6 40 0.3 1.6 168 1.1 0.8 134 0.9 1.6 158 1.0 1.6 136 0.9 0.8 173 1.1 0.8 44 0.3 0.8 25 0.2 7.8 495 3.2 48.6 7670 50.0 1.6 35.6 0. 2 6. 3 122.8 0.8 0.8 22.6 0. 1 0.8 26.5 0.2 3.1 67.0 0.4 1.6 36.0 0.2 14.2 310.5 1.9 Total Items Id. 128 100.4 170 prey items, both species and number, were used this year, indicates prairie falcons can adjust to a changing prey base. This may explain why studies in other areas suggest a high use of avian prey. Biomass was calculated using average weights of each species (Table 1). Generally they are the same as those used by Study I. Some weights were taken from University of Idaho specimens. A complete list of weights and sources will be furnished upon request. Biomass percentages are markedly different from those reported for 1976 (Kochert et al. 1976). Townsend ground squirrels (TGS) continued to show up in pellets and remains even though TGS population levels were low this year (see Study V) . In one, when most of these squirrels had entered torpor, Townsend ground squirrels were still a significant part of the diet (Table 2) . This suggests a preference by the prairie falcon for Townsend ground squirrels and an ability to find vulnerable individuals of the prey species. Visual Observations Between egg hatching and dispersal, we observed the intensively studied nest sites from dawn to dusk for a combined total of 30 days. On seven additional days, we observed these scrapes for partial days. This resulted in 410 hours of observation during the brood-rearing period, and 83 hours of observation after fledging. Over 200 feeding visits were recorded during these periods. A summary of feeding visits for the three intensively studied sites are presented in Table 3. The high number of unidentified prey items can be attributed to characteristics of the prey and/or actions of the young. Young prairie falcons from three to six weeks old can swallow an entire lizard or small bird in a few seconds. Since most avian prey are plucked before delivery to the scrape, identification is often impossible. Although positive identification was not possible in many of these situations, we did record the size and possible identification of all unidentified prey items. The low number of recorded feeding visits for late June does not represent a reduced effort in observations. It merely shows increased difficulty in detailing and identifying feeding visits. Since the young during this period have fledged and are scattered over a larger area, the blind was not always in the best position to detect all feeding visits or interactions between parents and young. The young are often out of sight behind rocks and feeding visits were short. Any activity that we felt represented a feeding visits but did not allow a visual sighting of prey was listed as a "possible visit" but was not included in Table 3. Table 2. Number and percent (N) of mammals, birds, and lizards in pellets and prey remains found in prairie falcon scrapes, 1977. 171 rH > -I— I cU -H d ■U X) 0) O d X H X W 0> CM CO CO vO t— 1 1 — 1 ctn CO CO — 1 LO QJ 1—1 d I 3 rH t“) 00 CM CO <04 -d- r — ao t-H C\J LO r~H co i cU X S 140 00 CO O' CO 1 — I CN] cm x CO CO <04 00 O 00 col o X — 1 03 4-1 Pj g d Pd d 03 •H Pd g 03 i •H co 03 03 d P4 £ <4-1 •H 03 s CO •H o 03 d •H > 03 03 M— i C* S CO 4-1 •H p4 CO p4 •H 03 03 d 00 4-1 03 03 03 4-1 03 P>. i — i CO p4 03 rj CO ro Pj N r~* CO 03 d 03 d 03 •H 4—1 o 4J •H •H 03 cd pJ C/3 4J p4 N •H 03 £ H O CQ 03 d3 Cd o o •H ■H £ 03 03 •H ' — ✓ H H PQ -I 5 03 03 d Ew M to S'? S'? S'? S'? S'? * Contains prey identified as "mammal" but not to species. 173 Evidence indicates that in prairie falcons there is a reduction in feeding visits after fledging. However, it is difficult to ascertain the amount of prey that the young capture and consume on their own. The fledged young spend considerable time walking, hopping, or developing their flight muscles on talus slopes and along canyon rims. Both of these habitats have a relatively high density of reptilian prey (Diller, personal communication) . We have often observed young birds in actions resembling a capture of prey. Since these "prey items" (insects or lizards) are small, consumption time is rapid and the visual angle is usually difficult. We suspect they are catching prey but are uncertain of the species or amount . Growth Studies In 1977, we had growth information on 12 known aged falcons (6 females and 6 males) from three different scrapes. We separated these by sex and plotted weight vs age using day 0 as the day the young falcon broke free of the shell and could receive food. The data from day 4 to day 23, the period of rapid growth, showed a constant linear growth pattern (r = 0.98; n = 25). When data from 1976 was reevaluated using these same criteria and compared to 1977 information, the growth rate for 1977 was visibly lower. The Id values were tested (Steel and Torie 1960:173) and found statistically significant at the p = 0.90 level. Complete data and graphs will be included in the final report. We checked several possible sources of variation and found no apparent difference in number of nest visits, breeding chronology of study nest, or clutch sizes between the two years. It was obvious that the prey base had been altered from 1976 since the Townsend ground squirrels and practically no reproduction in 1977 (see Study V report) . The falcons responded to this by taking an increased number of avian prey (Table 1) . We checked climatological data for the two years at the Swan Falls reporting station, which is located within the BPNA. During the month of May, the critical hatching and growing period, average maximum temperatures were 6°C less in 1977 than in 1976. May 1977, also had 15 days of rain (trace or more) as compared to only 6 in 1976. It is probable that inclement weather can reduce growth rates in raptors by requiring more energy for maintenance and thereby leaving less for growth. Inclement weather also can reduce availability of prey. We do not feel that the change in growth rates represents starvation. Starvation of one brood of falcons (see mortality section) was accompanied by a rapid loss of weight. Since all other falcons gained weight in a continuous and predictable fashion and survived to fledging, starvation is not apparent. Age at fledging (33 to 38 days) was comparable to 1976 fledging ages. 174 The adult falcons shifted to an apparently abundant avian prey base. The time spent hunting in 1977 may have increased from 1976 (see Study III) but the adult falcons still spent considerable portions of their day resting and preening near the scrape. Other authors (Fowler 1939; Bent 1961) in other areas have shown that prairie falcons can and do prey successfully on various birds without affecting reproduction capabilities. Number of feeding visits per day increased from 4.0 trips per day in 1976 to 5.3 in 1977. Since birds and lizards are small, it may be necessary for the adults to make more trips to feed the young the same biomass. Prey caching was evident in 13 percent of the feeding visits and again points to an abundance of prey. Mortality During the 1977 field season, we documented 16 cases of mortality in prairie falcons. We divided these cases into four age categories for easier analysis. The first category covered hatching and the early nestling period. One young died during this period when approximately four days old. On a previous visit, we noted this young falcon was weak and had its beak stuck with excrement. At that time, we removed the excrement and placed the young back with its four nest mates. On the next visit, it was missing and presumed dead. The pellet collection eight days after the young was last seen contained a small falcon beak with an egg tooth. This possible cannibalism and the caching of a day old prairie falcon chick (Peterson et al. 1976) gives further evidence to the possibility that missing eggs or young at the time of hatching are in actuality a mortality of hatched young. Ogden and Hornocker (1977) found that one-fourth of all egg losses were missing eggs. In 1977, we also documented the disappearance of two eggs but did not include this in our mortality total. The eggs disappeared from a protected cavity that was inaccessible to mammalian predators. The next category contained all cases of mortality between hatching and fledging. To rule out premature fledging, we limited this class to those young whose plumage was primarily in the downy stage (at least 60 percent). We recorded 53 percent of the mortality (eight young) in this category. Fifty percent of this mortality occurred when the young were quite large and therefore very active and mobile around the scrapes. All incidences were of birds found at the base of cliffs under eyries and suggest that the deaths occurred as a result of a fall from the nest ledge. In most bases it could not be determined if the fall was direc.ly responsible for the death or if it simply left the falcon chick more vulnerable to the effects of weather, predation or starvation. It is possible for the young to survive such a fall and still 175 be cared for by the adult falcons. One young (22 days old) was found below a scrape and was being fed by the adult birds. Two of the dead young found beneath the scrape had obviously moved or been moved from the assumed place of impact. A weasel (Mustela f renata) is suspected in one such movement. The death of four young from one nest was attributed to starvation as a result of abandonment when one of the adults was killed. The adult female refused to feed the young and after four days we removed the remaining young and placed them with other broods. One of these young was a "dwarf" young (one-half the size of its nest mates) and died before the transplant was made. The three young that were moved were listed as mortality cases since without human intervention their death was certain. Five cases of mortality occurred during the post-fledging period. Disease, predation, starvation, and interspecific aggression were represented as contributing factors in these deaths. Disease (trichomoniasis, "Frounce") was the ultimate cause of one death. Four other fledged young that were captured by hand on the ground displayed the lesions characteristics of this disease. We treated these four birds and released them, but success of our treatment is unknown as none were ever seen again. We found no definite direct cause of death in the four other post-fledging mortalities. Predation or scavenging was evident in two of the remains but death by different factors is possible. A banded juvenile prairie falcon was found dead 0.8 kilometers west of its natal territory. It was separated from its home by at least three defended falcon territories. The case of adult mortality was located with the help of a radio package the bird was wearing. Circumstantial evidence suggests the death was a result of aggression. This bird was found on a small ledge in an adjacent territory. The falcon was lying on its ventral side with wings outstretched and one foot tightly clenched. Small scars on the top of its head and blood in the mouth suggest some form of physical contact. The presence of the radio package as a contributing factor can be minimized since the falcon had carried such a package for more than one year. The death occurred in late May at a time when territorial defense is diminishing . The presence of ectoparasites namely the swallow bedbug (Oeciacus vicarius) , was not a major cause of mortality as in past years. The infestations of these bedbugs appeared delayed and did not reach large proportions until late in the brood-rearing period. One of our closely watched scrapes did produce a considerable population of bedbugs in early June. The young in this scrape 176 reacted with very frenzied movements, constant and aggressive preening and eventually premature fledging. Although no mortality occurred as a result of this, we would expect their vulnerability to predation to increase. In 1977, disease was possibly the biggest contributing factor in mortality of young prairie falcons. Even though disease may not be the direct cause of death, the starvation and inactivity associated with the "Frounce" increases chances of predation. In addition to the birds mentioned previously, we also found "Frounce" lesions in three nestlings and Study I personnel found it present in several prairie falcons. The increase in "Frounce" in prairie falcons is attributed to an increase predation on rock doves (Columba livia) . LITERATURE CITED Bent, A.C. 1961. Life histories of North American birds of prey, Part II. Dover Publications, Inc., New York. 482 pp. Dunstan, T.C., J.F. Harper, and K.B. Phipps. 1976. Activity, hunting patterns, territoriality, and social interactions of birds of prey in the Birds of Prey Natural Area. Pages 63-130 in Snake River Birds of Prey Research Proj . Annu. Rep. 1976. U.S. Dept, of the Inter., Bur. of Land Manage., Boise, ID. 240 pp. Fowler, F.H. 1939. Studies of food and growth of prairie falcons. Condor 33 (5) : 193-201 . Fyfe, R.W. and R.R. Olendorff. 1976. Minimizing the dangers of nesting studies to raptors and other sensitive species. Canadian Wildlife Service. Occas. Pap. No. 23. Kochert, M.N., A.R. Bammann, J.H. Doremus, M. DeLate, and J. Wyatt. 1976. Reproductive performance, food habits, and population dynamics of raptors in the Snake River Birds of Prey Natural Area. Pages 3-57 in Snake River Birds of Prey Research Proj. Annu. Rep. 1976. U.S. Dept, of Inter., Bur. of Land Manage., Boise, ID. 240 pp. Ogden, V.T. and M.G. Hornocker. 1977. Nesting density and success of prairie falcons in southwestern Idaho. J. Wildl. Manage. 41(1) :1-11. Peterson, S.R., G.M. Sitter, A. Berglund, and B. James. 1976. Feeding activity and behavior of prairie falcons. Pages 227-240 in Snake River Birds of Prey Research Proj . Annu. Rep. 1976. U.S. Dept, of Inter., Bur. of Land Manage., Boise, ID. 240 pp. Steel, R. and J. Torie. 1960. Principles and procedures of statistics. McGraw-Hill Book Company, Inc. New York. 481 pp. 177 178 Mike Collopy's intensive observations from blinds will provide information- on growth and feeding rates of nestling golden eagles. (BLM photo by Collopy). 179 STUDY Vila: COOPERATOR: INVESTIGATOR: Bioenergetics of Nesting Golden Eagles University of Michigan, Ann Arbor, Michigan 48104 Michael W. Collopy, Principal Investigator PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Cooperative Agreement 1-01-20, and Sigma Xi OBJECTIVES: 1. To determine the effects of nest site selection on parental behavior and reproductive success of golden eagles. 2. To determine quantitatively, the food consumption and growth rates of nestling golden eagles. 3. To assess the relationship between the rate of food delivery to the nest, and the growth and survivorship of nestling eagles . ANNUAL SUMMARY In January and February 1977, I frequently visited the Birds of Prey Natural Area (BPNA) to familiarize myself with the study area and to locate pairs of reproduct ively active golden eagles (Aquila chrysaetos) . Criteria used to assess the suitability of eagle nests for study included: proximity to accessible roads, amount of human disturbance, nest site exposure, difficulty in climbing to nest, and especially, the potential for blind construction. METHODS Two active nests were identified early enough during courtship to permit construction of rock and canvas blinds, without excessively disturbing the courting eagles. These blinds, were gradually constructed during visits in late January and early February 1977. Due to variations in nesting phenology and late discovery of some suitable nests, blind construction at several other eyries was postponed until after hatching. To avoid excessive disturbance of nesting eagles, intensive observations were not made at nests with recently constructed blinds. Two other eyries (Otis and Lower Reynolds) were selected to provide information on the behavior of incubating adults. These nests were chosen primarily because they were accessible and had a history of tolerance to human disturbance. Incubating eagles were observed from a vehicle parked in the vicinity of the nest. Observations began before dawn, as soon as the nest was visible, and lasted until both adults went to roost after sunset. 180 The behavior of adult eagles was recorded with respect to elapsed time of the act and time of day. Special emphasis was placed on recording interactions at the nest. Various environmental parameters were continuously monitored. These included: wind velocity, light intensity (reflected and incident) at the nest, ambient temperature, and relative humidity. After the eggs in the Otis eyrie hatched, the young were observed regularly at four-day intervals. During these observation periods, the nest was entered, with the assistance of the Study I personnel, and the prey remains present were identified and/or collected. The age, condition, number and weight of the young also were determined. The Lower Reynolds eyrie was visited less regularly, but the same data were collected. Because these nests had no suitable blind sites, I continued watching them only until my other nests hatched young. RESULTS A brief summary of the percent of daylight hours male and female eagles incubated eggs, and the percent of time the eggs were untended is presented in Table 1. Throughout the incubation period females were observed to sit on eggs between five and eight times as long as males. The females were always observed to incubate at night as well as the majority of the time during the day. Incubation bouts by females were consistently longer than those of the males (Table 2). Overall, they averaged 3 to 3.5 times longer on the eggs than the males, during each incubation bout. Table 3 summarizes the data collected on the weight and crop content of four eaglets hatched in the two nests intensively studied during the incubation period. Just as I began regular observations at the nests with blinds, I had a climbing accident which resulted in a broken leg. This postponed any further collection of data until next breeding season. Plans for the 1978 field season. I plan to resume research on the nesting behavior of golden eagles during the 1978 field season. Observations will begin with courtship in January 1978, and will continue through fledgling dispersal in July. Noon to noon observations will be conducted at four different nests throughout the nesting season. Previously, observations were made from dawn to dusk. However, this sampling procedure was rejected due to the potential for disturbance at the nest before dawn and after dusk. Moving from blind to blind during midday was judged less hazardous because the adults are frequently off the nest at that time. l8l Table 1 . The percent of daylight hours male and female Golden Eagles incubated their eggs in relation to the stage of the incubation period. Data was collected from two nests in or near the Snake River Birds of Prey Natural Area. Day of Incubation Female Otis Lower Reynolds Male Eggs Exposed Female Male Eggs Exposed 182 Table 2. The mean duration of incubation bouts by male and female Golden Eagles at two nests in or near the Snake River Birds of Prey Natural Area. Data is presented in minutes. Otis Eyrie Day of — Incubation Female Male Lower Reynolds Evrle Female Male food exchange, female was feeding in the area. 183 Table 3* Body weight and the proportion of the crop which is full, in four nestling Golden Eagles. Weights were obtained only for the early stages of brood rearing. Otis Eyrie Lower Reynolds Eyrie Age (days) 2 Weight: Crop Contents 3 Weight: Crop Contents Welsht ‘Contents Welght ‘ConUnts 3 Younger sibling 184 During periods of observation at the nest site, I will monitor incident and reflected light intensity at the nest, ambient temperature, and humidity, with a light meter, thermometer, and sling psychrometer , respectively . At the end of each period of blind observation, I will enter the nest and obtain data on the age, number, condition, and weight of the eaglets. I also will verify the species of prey, I observed to be delivered to the nest by checking for remains. A 35 mm SLR camera will be used to document behavioral acts and to facilitate the recognition of individual eagles. Four remote-controlled movie cameras will be installed near nests not observed from blinds to provide additional data on the frequency of nest visits, "changing of the guard" at the nest, food deliveries, and general activity around the nest. These nests will be visited periodically and data on the age, number, condition, and weight of the chicks will be recorded. In addition to this intensive blind work, I plan to conduct an extensive survey of the golden eagle nests on the Birds of Prey Natural Area. Data will be collected on nest exposure, cliff height, and position of nest on cliff. The amount of cliff overhang above the nest also will be recorded. 185 186 Personnel from Study IVB and Study V have cooperated to trace the incidence of plague in the BPSA and to estimate its role in the ecosystem (BLM photo By Messick). SPECIAL REPORT : * Plague (Yersinia pestis) in the Snake River Birds of Prey Study Area. INVESTIGATORS: Graham W. Smith, Research Associate Study V. John P. Messick, Research Associate Study IVB. Donald R. Johnson, Principal Investigator Study V. Maurice G. Hornocker, Principal Investigator Study IVB. COOPERATORS: Department of Biological Sciences, University of Idaho, Moscow, Idaho 83843. Idaho Cooperative Wildlife Research Unit, University of Idaho, Moscow, Idaho 83843. U.S. Public Health Service, Vector-borne Diseases Division, Plague Branch, P.0. Box 2087, Fort Collins, Colorado 80522. PROJECT SUPPORT: U.S. Department of the Interior, Bureau of Land Management, Contract Nos. 52500-CT5-1002 and 52500-CT5-1004 , U.S. Public Health Service, Vector-borne Diseases Division, Plague Branch. OBJECTIVES: Assess the role of plague as a mortality factor in Townsend ground squirrels ( Spermophilus townsendi) . 1. Monitor plague in Townsend ground squirrels using antibody titers in badgers (Taxidea taxus) . 2. Test dead ground squirrels and other rodents for plague. SUMMARY Two hundred ninety four badger serum samples were analyzed using the passive haemagglut inat ion test for the presence of antibodies for plague. In addition eight Townsend ground squirrels were found positive for the presence of plague by the fluorescent antibody test or by isolation of Yersinia pestis . Plague appears to be widespread within the Snake River Birds of Prey Study Area. The epizoot iology of plague is closely related to the yearly cycle of the Townsend ground squirrel. The disease appears to be quiescent during squirrel hibernation then its incidence increases after squirrels emerge and reproduce. The disease is most prevalent after the emergence of the juvenile ground squirrels, the time of year the ground squirrel population is at its maximum. After squirrel entry into estivation prevalence of the disease decreases. Plague within the Birds of Prey Study Area appears to be a stable enzootic . *This is a special contribution volunteered by the personnel from Study IVB and V. 188 INTRODUCTION Just if icat ion Plague has caused sudden and massive mortality in local populations of ground squirrels, prairie dogs, and other rodents. Lechleitner et al. (1968) reported that plague decimated several colonies of Gunnison prairie dogs (Cynomys gunnisoni) within a 12 km“ area of Saguache County, south central, Colorado. The epizootic spread through the colonies during the summer of 1964. A residual phase continued and terminated in 1966, leaving only a few surviving prairie dogs. Plague has been a recurring phenomenon among California ground squirrel (Spermophilus beecheyi) populations (Meyer 1936, Murray 1964, Evans et al. 1943). Utah populations of Townsend ground squirrels remained depressed for 18 years following an apparent plague epidemic in 1937 (Hansen 1956). Svihla (1939) reported an apparent Townsend ground squirrel die-off in southeast Washington. He suggested that plague may have been responsible. A similar epizootic in Townsend ground squirrels within the Snake River Birds of Prey Study Area (BPSA) would deprive prairie falcons (Falco mexicanus) of their primary food source during the nesting season. The role of disease in regulating animal populations is not well understood. Frequently epizootics are recognized only after they occur. This precludes a study of their etiology. We hope surveys, such as this one, will help overcome this difficulty. Plague in the United States is primarily a disease of wild rodents. Transfer to man by direct contact or by fleas is uncommon, perhaps because of the relative isolation of wild rodents from man (Mitchell 1966) . This isolation may deteriorate in places such as the Birds of Prey Study Area. The Bureau of Land Management must consider the potential for human infection, however remote. Although no recent outbreaks of plague have been documented in southwestern Idaho, this does not indicate that one will not occur (Murray 1964) . The Ecology of Plague Description and History. Plague is a highly infectious disease caused by the bacterium, Yersinia pest is . Sylvatic plague is the term for the disease in wild rodents, the most common hosts. Fleas serve as vectors of the disease among rodents and sometimes other species. Direct contact may also spread plague. Man may become involved if bitten by an infected flea or by handling infected rodents. One typical symptom is swelling of lymph nodes (buboes) in the armpits and groin, hence the name bubonic plague. If the lungs are infected (pneumonic plague) , human contacts and droplet infection will result in an explosive spread of the disease (Mitchell 1966) . 189 Most authorities agree that plague originated in Asia or Africa, but how it became established in the United States is not clear. The first human case was discovered in San Francisco in 1900. It has since been found in 15 western states as far east as Kansas, Oklahoma, and Texas. This suggested to some workers that plague arrived at western seaports via ships harboring infected rats, then spread eastward from these foci (Eskey and Haas 1940, Meyer 1947). Recent reviewers of the subject (Olsen 1970) believe, however, the apparent eastward spread is an artifact of increasingly skilled and detailed detection techniques employed since the early 1900’s. It is likely that plague was introduced to North America long before modern history by rodents which crossed a land bridge from Asia (Ecke and Johnson 1952; Pollitzer 1954). Transmission. Fleas, the usual vector, originally ingest Yersinia pestis with a blood meal from an infected host. The bacteria multiply, blocking the gut of the flea. Since blood cannot enter the stomach, the flea becomes ravished and begins a feeding frenzy. Bacteria become dislodged from the plug during these futile attempts to take a blood meal, and are regurgitated into the blood stream of the previously uninfected host. Variations of this general pattern depend on the life history features of the flea, the time between initial infection, the subsequent feeding, and the degree of blocking. The factors which trigger an epizootic are complicated and are not well understood. They include ecological and physiological characteristics of both the vector and host. Ground squirrels and other conspicuous victims of plague epidemics were thought by early workers to be the reservoirs of plague. However, the frequently nearly complete die-offs of these species would not favor the maintenance of Yersinia pestis between epizootics. It is now known that partially resistant species, such as Peromyscus spp. and Microtus spp. are the genuine reservoirs. These biological storage sites become bacteremic harboring Yersinia pestis in their bodies without succumbing to the infection (Murray 1964). Resistant fleas, and even soil may also serve as plague reservoirs (Olsen 1970) . Within the Birds of Prey Study Area, only deer mice (Peromyscus spp.) and kangaroo rats (Dipodomys spp.) occur in moderate densities in areas sympatric with the Townsend ground squirrel (Wolfe et al. 1976). It is probable one or both of these species are the reservoirs for plague in the area. Cottontail rabbits (Sylvilagus spp.), which may also be involved in the epizootiology of plague (Kartman 1960) , are very scattered, occurring only in rocky areas within the Birds of Prey Study Area (Wolfe et al. 1976). It is unlikely that they serve as the main reservoirs of the disease. 190 Diagnosis Detection of plague depends on tests broadly classified as bacteriological and serological. Bahmanyar and Cavanaugh (1976) and Olsen (1970) provide an overview of these procedures. Bacteriological Tests. Proof of plague requires isolation and identification of Yersinia pestis . Tissue (blood, spleen, liver, etc.) samples from suspected hosts are cultured in appropriated media or inoculated into guinea pigs (Cavia spp.). The bacteria are identified microscopically after specific staining, or through biochemical tests. In decayed carcasses, it may not be possible to isolate Yersinia pestis . The fluorescent antibody test (Hudson et al. 1962) will give an indication of plague when samples are in poor condition. Serological Tests. These tests rely on indirect detection by measuring antibodies to Yersinia pestis in sera of partially or completely resistant hosts. Because epizootics frequently extirpate ground squirrels and other susceptible species, bacteriological methods are useful primarily after an epizootic is underway. Serological techniques provide a means of monitoring the level of infection in the true reservoirs, and other animals which develop antibodies to the plague bacteria. Antibodies to Yersinia pestis in sera of domestic dogs were associated with seven cases of human plague in New Mexico between 1959-1966 (Archibald and Kunitz 1971) . Meyer et al. (1965) demonstrated antibodies in rodents and mongooses in Hawaii. Fitzgerald et al. (unpublished manuscript) demonstrated antibodies in sera of badgers thus laying the ground work for this study. METHODS Two to 4 ml of blood were taken with a 5 cc syringe fitted with a 3.5 cm 20 gauge needle from one of the superficial veins of the badger’s front leg. After transfer to a 7 ml vacuum blood tube (without the vacuum) , formed elements and sera were allowed to separate at room temperature. After several hours, serum was transferred using a pipette into another tube and then frozen. All syringes, tubes, and pipettes were disposable to eliminate any possibility of cross contamination. Samples were kept frozen and shipped in dry ice to the Fort Collins laboratories for analysis by the passive haemagglut inat ion test (Hudson and Kartman 1967) . Results are reported as a dilution ratio with ranges from zero to 1/8192 (greater dilutions were possible). Ratios of 1/16 and above were considered evidence of plague exposure. The larger the denominator the higher the 191 antibody titer. The numerator (always one) was eliminated in our tables. Between spring 1975 and late summer 1977, 299 badger serum samples were collected. Results from five were not obtainable because of hemolysis or insufficient quantity. Fleas were collected from badgers and from squirrels by dusting or spraying with commercial dog or cat flea killing preparation containing pyrethrins. Fleas were identified by Dr. Vernon J. Tipton, Director, Health and Environmental Studies Center, Brigham Young University, Provo, Utah 84602 and by Craig R. Baird, Area Entomologist, University of Idaho Extension Service, P.0. Box 1058, Caldwell, Idaho 83605. Fleas reported in this study were not tested for plague. Eight Townsend ground squirrels found dead or sick without evidence of physical injury were autopsied. Spleen and liver samples were shipped frozen to the plague laboratories for testing with the fluorescent antibody technique and for isolation of Yersinia pestis . RESULTS AND DISCUSSION The badger in the Snake River Birds of Prey Study Area is an excellent monitor of the presence of plague in the Townsend ground squirrel population. The ground squirrel comprises the bulk of the badger's diet, and other rodents occur only in small numbers in the diet. Hornocker et al. (1976) reported that the Townsend ground squirrel occurred in 86 percent of 231 badger scats collected within the Birds of Prey Study Area. Kangaroo rats and deer mice, the rodents most likely to be involved in the maintenance of plague in the area, each occurred in less than eight percent of the scats. Flea species collected from badgers and ground squirrels are listed in Table 1. We are uncertain if these flea species are involved in the epizoot iology of plague within the Birds of Prey Study Area. The relationship between titer strength and its duration in badgers can be deduced from a comparison of the titer strengths for badgers captured two or more times. Table 2 presents the maintenance or the decline in titer strengths and the number of days between samples. The interpretation of the data is complicated by the fact that individual badgers may have been re-infected between captures. All such obvious cases were eliminated from consideration. Note the sequences of three or more subsequent captures for Numbers 164, 87 and 110 (Table 2). The results are not conclusive but it is apparent that titer strengths decrease from an initial high of 1/8192 to zero in a span of about 250 days. It is also apparent that the titer strength and its duration are quite variable between badgers. Similar results have been reported for humans (Meyer 1964), mongoose, Herpestes auropurctatus , (Meyer et al. 1965), dogs, Canis familiaris, (Archibald and Kunitz 1971), and Norway rats, Rattus norvegicus, (Williams et al. 1977). 192 Table 1. Species of fleas collected from badgers and Townsend ground squirrels within the Snake River Birds of Prey Study Area. Flea Species Source Common Host Hoplopsyllus glacialis badger Sylvilagus spp. and Lepus spp . Opisocrostis oregonensis badger , Townsend ground squirrel Spermophilus beldingi and S. townsendi Opisocrostis tuberculatus tuberculatus Townsend ground squirrel Spermophilus townsendi Thrassis francisi barnesi badger , Townsend ground squirrel Spermophilus spp. , Ammo spermophilus spp. , and occasionally Peromyscus spp. 193 Table 2. The change in haemagglut ination antibody titer for plague in badgers captured two or more times within the Snake River Birds of Prey Study Area. Only equal or decreasing sequences and sequences less than 98 days listed. Badger First capture number Date-1 Titer-1 163 9 May 77 8192 85 24 Apr 76 8192 164 9 May 77 8192 164 29 May 77 2048 98 30 May 77 2048 47 8 May 77 1024 80 23 Apr 77 1024 20 8 May 77 1024 164 7 Jun 77 1024 77 27 Feb 76 512 20 17 Jun 75 512 205 11 Jul 77 512 173 30 May 77 512 20 15 Jul 76 512 87 29 Apr 76 256 181 1 Jun 77 256 184 4 Jun 77 256 89 1 May 76 256 87 22 Jun 76 128 204 8 Jul 77 128 20 28 Aug 75 128 182 1 Jun 77 128 110 23 May 76 64 153 22 Oct 76 64 110 3 Jul 76 64 187 9 Jun 77 32 168 13 May 77 32 178 25 May 77 32 144 31 Jul 76 8 110 22 Sep 76 8 Second capture Interval Date-2 Titer-2 (day: 17 May 77 4096 8 15 Jul 76 2048 82 29 May 77 2048 20 7 Jun 77 1024 9 20 Jul 76 32 51 29 Jul 77 1024 82 13 May 77 1024 20 3 Jun 77 512 26 4 Jul 77 512 27 15 May 76 256 78 28 Aug 75 128 72 12 Aug 77 128 32 25 Jun 77 64 26 21 Oct 76 32 98 22 Jun 76 128 54 3 Jul 77 32 32 22 Jul 77 32 48 4 Aug 76 8 95 4 Aug 76 128 43 27 Jul 77 128 19 10 Nov 75 64 74 10 Jul 77 64 39 3 Jul 76 64 41 2 Nov 76 64 11 22 Sep 76 8 81 23 Jun 77 32 14 25 May 77 16 12 15 Jun 77 0 21 12 Sep 76 8 43 23 Oct 76 0 31 The juvenile members of family groups (female with young) characteristically had lower haemagglutination titers than the parent (Table 3) . Samples for Table 3 were collected before young dispersed. Badgers in southwestern Idaho are born in late March to early April. The family group dissolves in late May or early June. Hornocker et al. (1976) give additional details on parent- young associations. Young badgers probably develop their initial antibodies after feeding on infected prey brought to the maternal den by the adult. The similarity of titer levels among litter mates suggests a common source of infection (Table 3). Rust et al. (1971) demonstrated that the timing and magnitude of antibody response was closely correlated in dogs and cats (Felis domesticus) experimentally infected with plague. Strength of complement fixing antibody titers was proportional to subcutaneous or oral dosage of plague in adult coyotes (Marchette et al. 1962). If badgers possess a similar response curve, then we might expect to find previously unexposed juveniles to develop similar titer strengths after sharing infected food. It is possible that the litter could acquire antibodies from colostrum or via placental transfer (Fitzgerald et al. unpublished manuscript) . Maternal origin of antibodies to plague was documented in Norway rats, but the prevalence of this phenomena among many other mammals is unknown (Williams et al. 1977). It is conceivable that inheritance of maternal antibodies could provide newborn badgers with immunity from plague (Williams et al. 1977). The proportion of badgers that succumb to plague infection is probably very small if not nill. Fitzgerald et al. (unpublished manuscript) thought that badgers might occasionally become bacteremic. Susceptibility to plague is highly variable in carnivores (Archibald and Kunitz 1971; Meyer et al. 1965; Rust et al. 1971a, 1971b). Although we believe plague is of little importance in the population processes of the badger, only experimental determination of their sensitivity would fully answer this question. Higher titers in the parent (Table 3) might result from larger dosages of infected prey and the additive effect of repeated exposure. Meyer et al. (1965) observed that previously infected mongooses exhibited higher antibody titers after re-exposure to plague. The epizoot iology of plague is related to the yearly cycle in the Townsend ground squirrel. Ground squirrels emerge from hibernation in February. Young of the year are born in March and emerge from natal burrows in early April. During April, the population is at its yearly high and the opportunity for plague transfer between squirrels is increased. In late May, the adult males enter estivation with the adult females estivating in early to mid-June. Juvenile squirrels estivate in late June and early 195 Table 3. Haemagglutinat ion antibody titers for plague in ten family groups (female with young or litter mates) of badgers captured March-June 1976 and 1977 within the Snake River Birds of Prey Study Area, Idaho. Year Adult female No. Titer #1 Titer #2 of young #3 #4 1976 92 1,024 265 128 64 1976 98 512 64 64 64 2,048 1976 101 512 128 64 1976 104 1,024 128 1976 not captured 128 256 1976 37 512 64 1977 37 256 neg. neg. 32a 32a 1977 96 512 128 128 128 1,024 1977 155 2,048 512 256 1977 43 256 32 32 8 16 neg. a The two young listed for adult female No. 37 in 1977 with titers of 32 may be from a different litter. 196 July. By the second week of July, few squirrels are still active (Johnson and Melquist 1975; Johnson et al. 1976). Badger titer strengths appear to follow a pattern throughout the year related to the ground squirrel yearly cycle (Table 4) . High titers are found in April through July and then titer levels decrease. Some lag in time related to the initiation and the duration of the immunological response in badgers in relation to the ground squirrel cycle is expected. Interpretation of the results must take into account differing sample sizes in the three years of collection and the fact few badgers were captured from September through February, 1975-76 and 1976-77. From September to February, badger activity was much reduced and individuals appeared to be inactive much of the time (Hornocker et al. 1976) . With decreased activity and reduced hunting effort, the chance of exposure was reduced . The ground squirrel cycle differed in 1977 from that of 1975 or 1976. A drought in the region in 1977 resulted in low production of annual grasses. Annual grasses are the primary food of the ground squirrel. The use of green forage appears to initiate reproduction (see Study V). In 1977, no squirrels in native vegetation habitats within the Birds of Prey Study Area reproduced. There was less chance for the spread of plague between ground squirrels with the reduced population. In addition, the ground squirrel population estivated in late April and early May. By mid-May, no squirrels were active. The effects of these two factors can be seen for 1977 (Table 4) by the increase in the number of negative titers in 1977 as compared to 1976. The location and dates of collections of Townsend ground squirrels which were FA (fluorescent antibody test) positive within the Birds of Prey Study Area are given in Figure 1. Two of the eight yielded cultures of Yersinia pestis . The locations of plague positive dead squirrels and the locations of badgers with HA (haemagglutination antibody) titers indicate the disease is widespread throughout the Birds of Prey Study Area north of the Snake River. Continued live trapping at locations where dead squirrels were found indicates the ground squirrel population is relatively stable and that no massive die-offs have occurred. Plague within the Birds of Prey Study Area appears to be a stable enzootic. Whether under some optimum conditions, an epizootic could occur within the ground squirrel population is unknown. There is evidence that plague maintains itself in an area over long periods of time (Olsen 1970) . LITERATURE CITED Archibald, W.S. and S.J. Kunitz. 1971. Detection of plague by testing serums of dogs on the Navajo reservation. H.S.M.H.A. Health Rep. 86:377-380. Table 4. The number of haemagglut inat ion antibody titers for plague in badgers for 1975, 1976, and 1977 within the Snake River Birds of Prey Study Area. 197 CO 4-1 O H lt) o cn O N Cl (N CO co o co 00 VD 4/T7/ 7 \ horse butte RW BM 2985X RW 5/27/75 3/12/ !v A |-3054 -J- Buttej r 29 L RW^ \\ Fig. 1. Location and date of collection of plague positive Townsend ground squirrels within the Snake River Birds of Prey Study Area. 199 Bahmanyar, M. and D.C. Cavanaugh. 1976. Plague manual. World Health Organ., Geneva. 76 pp. Ecke, D.H. and C.W. Johnson. 1952. Plague in Colorado. Part I of Plague in Colorado and Texas. U.S. Pub. Health Monogr. No. 6. 54 pp. Eskey, C.R. and V.H. Haas. 1940. Plague in the western part of the United States. U.S. Pub. Health Rep. 254:1467-1481. Evans, F.C., C.M. Wheeler, and J.R. Douglas. 1943. Sylvatic plague studies. III. An epizootic of plague among ground squirrels in Kern County, California. J. Infect. Dis. 72:68-76. Fitzgerald, J.P., B.W. Hudson, R.P. Lechleitner, A.M. Barnes, and J.D. Poland. Serum responses in badgers. Unpublished manuscript Hansen, R.M. 1956. Decline in Townsend ground squirrels in Utah. J. Mammal. 37:123-124. Hornocker, M.G. , J.P. Messick, J.S. Whitman, R.M. Olson and G.W. Smith. 1976. Dynamics of predation upon a raptor prey base in the Snake River Birds of Prey Natural Area, Idaho. Pages 181-201. in Snake River Birds of Prey Res. Proj . Annu. Rep., 1976. U.S. Dept. Inter., Bur. Land Manage., Boise, ID. 240 pp. Hudson, B.W. and L. Kartman. 1967. The use of the passive haemagglut inat ion test in epidemiologic investigation of sylvatic plague in the United States. Bull. Wildl. Dis. Assoc. 3:50-59. , S.S. Quan, and L. Kartman. 1962. Efficacy of fluorescent antibody methods for detection of Pasteurella pest is in carcasses of albino laboratory mice stored for various periods. J. Hyg. 60:443-450. Johnson, D.R. and W.E. Melquist. 1975. Population ecology and habitat requirements of Townsend ground squirrels. Pages 156-176 in. Snake River Birds of Prey Res. Proj. Annu. Rep., 1975. U.S. Dept. Inter., Bur. Land Manage., Boise, ID. 193 pp. , G.W. Smith, and R.M. Olson. 1976. Population ecology and habitat requirements of Townsend ground squirrels. Pages 203-225. in. Snake River Birds of Prey Res. Proj. Annu. Rep., 1976. U.S. Dept. Inter., Bur. Land Manage., Boise, ID. 240 pp. Kartman, L. 1960. The role of rabbits in sylvatic plague epidemiology with special attention to human cases in New Mexico and use of the fluorescent antibody technique for detection of Pasteurella pestis in field specimens. Zoonoses Res. 1:1-27. 200 Lechleitner, R.R., L. Kartman, M.I. Goldenberg, and B.W. Hudson. 1968. An epizootic of plague in Gunnison’s prairie dogs (Cynomys gunnisoni) in south-central Colorado. Ecology 49:734-743. Marchette, N.J., D.L. Lundgren, P.S. Nicholes, J.B. Bushman, and D. Vest. 1962. Studies on infectious diseases in wild animals in Utah. II. Susceptibility of wild animals to experimental plague. Zoonoses Res. 1(13) :225-250. Meyer, K.F. 1936. The sylvatic plague committee. Am. J. Public Health. 26:961-969. . 1947. Relation of diseases in lower animals to human welfare: prevention of plague in light of newer knowledge. Ann. N.Y. Acad. Sci. 48:429-467. . 1964. Serological tests for the confirmation of plague infections: a preliminary communication. Bull. World Health Org. 30:750-751. , D. McNeill, and C.M. Wheeler. 1965. Results of a preliminary serological survey of small mammal populations for plague on the island of Hawaii. Bull. World Health Org. 33:809-815. Mitchell, H.H. 1966. Plague in the United States: an assessment of its significance as a problem following a thermonuclear war. Memorandum RM-4968-TAB. Rand Corp. Santa Monica, CA. 49 pp. Murray, K.F. 1964. The evolution of pest control in California. Proc. 2nd. Vert. Pest Control Conf . , Anaheim, CA. 143-140 pp. Olsen, P.F. 1970. Sylvatic (wild rodent) plague. Chapter 19, Pages 200-213. in. Davis, J.W., L.H. Karstad, and D.O. Trainer, (eds.). Infectious diseases of wild mammals. Iowa State Univ. Press. Ames. 421 pp. Pollitzer, R. 1954. Plague. World Health Organ. Monogr. Ser. 22, Geneva. (cited by P.F. Olsen, 1970 (q.v.)) Rust, J.H., Jr., D.C. Cavanaugh, R. O'Shita, and J.D. Marshall, Jr. 1971a. The role of domestic animals in the epidemiology of plague. I. Experimental infection of dogs and cats. J. Infect Dis. 124:522-526. , B.E. Miller, M. Bahmanyar, J.D. Marshall, Jr.,. S. Purnaveja, D.C. Cavanaugh, and U.S. Tin Hla. 1971b. The role of domestic animals in the epidemiology of plague. II. Antibody to Yersinia pest is in sera of dogs and cats. J. Infect Dis. 124:527-531. ~~~~ 201 Svihla, A. 1939. Breeding habits of Townsend’s ground squirrel. Murrelet 20:6-10. Williams, J.E., G.H.G. Eisenberg, Jr., and D.C. Cavanaugh. 1977. Decline of maternal antibodies to plague in Norway rats. J. Hyg. Comb. 78:27-31. Wolfe, M. , J.R. Montan, Jr., K. Montan, J. Wassink, M. Wakefield, B. James. 1976. Density and species composition of small mammals and birds in the Snake River Birds of Prey Natural Area, Idaho. Pages 133-162. in Snake River Birds of Prey Res. Proj . Annu. Rep., 1976. U.S. Dept. Inter., Bur. Land Manage., Boise, ID. 240 pp. ☆ U.S. GOVERNMENT PRINTING OFFICE:1978-777-926 / 336