(ISSN ORE) 2-11)1 6) Contents The Breeding Biology of the Broad-winged and Red-shouldered Hawks IN Western New York. Scott T. Crocoll and James W. Parker 125 Environmental Contaminants in Blood of Western Bald Eagles. Stanley N. Wiemeyer, Richard W. Frenzel, Robert G. Anthony, B. Riley McClelland and Richard L. Knight 140 Hazards to Raptors from Strychnine Poisoned Ground Squirrels. Josef K. Schmutz, Karrie A. Rose and Robert G. Johnson 147 Surveys for Wintering Birds of Prey in Southeastern Colorado: 1983—1988. David E. Andersen and Orrin J. Rongstad 152 Food Habits of Nesting Prairie Falcons in Campbell County, WYOMING. John R. Squires, Dr. Stanley H. Anderson and Robert Oakleaf 157 Observations on the Evening Departure and Activity of Wintering Short-eared Owls in New Jersey. Thomas Bosakowski 162 Sexual Differences in Timing of American Kestrel Migration at HAWK MOUNTAIN SANCTUARY, PA. Nancy G. Stotz and Laurie J. Goodrich 167 The Use of Line Transects to Evaluate the Abundance of Diurnal MAMMALIAN Prey. Joan L. Morrison and Patricia L. Kennedy 172 Use of Explosives to Enhance a Peregrine Falcon Eyrie. Joel e. Pagei 176 Short Communications Range Extension of the Barked Owl in Western Washington and First Breeding Record on the Olympic Peninsula. Devora Ukrain Sharp 179 Peregrine Falcon Takes Black-bellied Plover from Sea off Kenya. Jennifer F. M. Horne and Lester Short 181 Northern Harrier ( Circus cyaneus) Predation of Lesser Prairie-chicken ( Tympanuchus pallidicinctus ). David A. Haukos and Gerald S. Broda 182 An Instance of Carrion-feeding by the Peregrine Falcon (Falco peregrinus). Dan C. Holland 184 Thesis Abstracts 185 News and Reviews 186 ******************** Persons interested in predatory birds are invited to join The Raptor Research Foundation, Inc. Send requests for information concerning membership, subscriptions, special publications, or change of address to Jim Fitzpatrick, Treasurer, 12805 St. Croix Trail, Hastings, Minnesota 55033, U.S.A. The Journal of Raptor Research (ISSN 0892-1016) is published quarterly for $15.00 per year by The Raptor Research Foundation, Inc., 12805 St. Croix Trail, Hastings, Minnesota 55033, U.S.A. Application to mail at second class rate is pending at Hastings, Minnesota, and additional mailing office. Printed by Allen Press, Inc., Lawrence, Kansas, U.S.A. Copyright 1989 by The Raptor Research Foundation, Inc. Printed in U.S.A. THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. Vol. 23 Winter 1989 No. 4 J Raptor Res. 23(4):125-139 © 1989 The Raptor Research Foundation, Inc. THE BREEDING BIOLOGY OF BROAD-WINGED AND RED-SHOULDERED HAWKS IN WESTERN NEW YORK Scott T. Crocoll and James W. Parker Abstract. — Nesting Red-shouldered ( Buteo hneatus; RSH) and Broad-winged Hawks {Buteo platypterus) BWH) were studied in 1978-1980 in Chautauqua County, New York. Both species nested predominantly in upland maple-beech-hemlock associations. Six of 18 BWH nests were in larch ( Larix decidua)-, 7 of 18 RSH nests were in American beech ( Fagus grandifolia). BWH nested closer to woodland openings and lower in a tree than RSH although neither difference was statistically significant. RSH began nesting about a month before BWH. In general, BWH produced more young and were more successful than RSH. Both species suffered similar types of mortality, and Great Horned Owls {Bubo virginianus; GHO) were the largest cause of nest failure for both species. A comparison of our data with other studies indicated that for many criteria, BWH and RSH were less productive in western New York than elsewhere. Reasons for this were not clear for RSH, although nest predation was an important factor. Our more extensive data for BWH explain in part their lower productivity in western New York as a function of habitat, clutch size, age of nesters, predation and reuse of nest. BWH nesting success was greater for new nests than rebuilt ones; greater for adult/adult pairs compared to adult/yearling pairs, greater in deciduous-mixed woodlands compared to conifer plantations, and greater for nests more distant from woodland openings. Three-egg clutches produced more fledglings than 2-egg clutches, but 2-egg clutches produced more fledglings per egg. We interpreted the BWH as an r-selection strategist over its lifetime, compared to many larger buteos, but the species seems capable of optimizing its reproductive effort in New York in any given year by employing brood reduction and possibly clutch adjustment reproductive strategies. Although the natural history (Burns 1911), nest- ing ecology (Matray 1974), and nesting habitat (Reran 1976) of the Broad-winged Hawk {Buteo platypterus', BWH) have been studied, only a few studies included in-depth quantitative analysis of reproductive success (Rusch and Doerr 1972; Reran 1 97 6; Armstrong and Euler 1 982; Janik and Mosher 1982; Rosenfield 1984). There is documentation of the nesting density of the species in Wisconsin, Min- nesota and Alberta (Rusch and Doerr 1972; Reran 1978; Rosenfield 1984), but not in the Northeastern United States. BWH populations appear healthy, but many researchers have suggested a general de- cline in Red-shouldered Hawk {Buteo Hneatus ; RSH) populations (Bent 1937; Brown 1971; Henny et al. 1973; Mason 1980; Bednarz and Dinsmore 1981), and the species appears on several state endangered and threatened species lists (i.e., Illinois, Iowa, Michigan, Minnesota, New York, Wisconsin; Anon. 1984). Bent (1937) and Stewart (1949) suggested that BWH and RSH select different nesting habi- tats, and Titus and Mosher (1981) and Armstrong and Euler (1982) quantified differential habitat uti- lization of both species. However, preliminary field work in western New York indicated that these species were locally sympatric and could be studied simultaneously. Because the BWH’s 1-4 eggs hatch asynchro- nously (Lyons and Mosher 1987), the species ap- pears to be a simple brood reduction strategist (O’Connor 1977). However, Howe (1976) showed that the Common Grackle {Quiscalus quiscula ) can combine this strategy with a resource storage strat- egy (see O’Connor 1977) in which eggs are provi- sioned with extra lipids so that young may survive short periods of food shortage. If a food shortage is lengthy, then brood reduction will occur. We looked for the presence of a second reproductive strategy in 125 126 Scott T. Grocoll and James W. Parker Vol. 23, No. 4 the BWH by assessing clutch size differences be- tween years and marking and measuring eggs and monitoring the hatching sequence within clutches. Egg size/hatching sequence data will be the subject of a future paper. Specific objectives of our study were: To document ecological nesting densities of both species; to char- acterize and compare breeding habitats of both; and to compare quantitatively rates of reproductive suc- cess during 2 or more breeding seasons, looking es- pecially for nest success in relation to clutch size. Study Area Our study area was in central and northern Chautauqua County, New York, a characteristically flat plain along Lake Erie bordered by hilly terrain a few miles from the lake. Chautauqua County is bordered by Pennsylvania to the south, Lake Erie to the northwest, and Cattaraugus County to the east. Historically, land along the lake has been used largely for grape production. Upland areas have been farmed or logged. Some farmland has returned fallow and some was used by the Civilian Conservation Corps in the 1930s for monocultural plantings of Scotch pine (Pinus sylvestris ), red pine (P. resinosa ) and tamarac or European larch ( Larix decidua). Overstory was comprised of approximately 45-yr-old individuals of these species, and the understory was dominated by sugar maple ( Acer saccharum). Much acreage resembled or was succeeding to a typical maple-beech-hemlock (Acer-Fagus- Tsuga) as- sociation (Shelford 1963). Although most of the study area was forested by a ma- ple-beech-hemlock association or a prior sere, some acreage supported oak-hickory ( Qj terms- Carya spp.) or aspen-black cherry ( Populus spp .-Prunus serotina). In a few areas larch was the dominant species with sugar maple (A. saccharum) understory. Topography was very hilly with numerous deep ravines. Water in some form (stream, pond, swamp) was present within 0.4 km of all the study sites. The principal study site was the New York State Can- adaway Creek Wildlife Management Area (approxi- mately 875 ha) and adjoining land, a broad, deeply dis- sected upland plateau predominantly covered with deciduous forest, but also with conifer plantings. The area is managed primarily to ensure quality habitat for wildlife, especially game species, and secondarily is available for recreational use. Methods This study was conducted from 1978 through 1980. Thorough ground censuses for old stick nests were con- ducted during late winter and early spring before nesting hawks returned. All old nest sites were revisited after hawks returned, and other areas where BWH and RSH were observed were also censused. Nests being used (active) were revisited approximately once/wk until hatching to document clutch size and egg survival. Use of a pole and mirror device (Parker 1972) minimized nest disturbance. Eggs were measured and weighed approximately 1 wk prior to hatching. Nest height (measured using the pole and mirror device or carefully estimated from sightings on a person of known height standing at the tree base), distance of nest tree from a roadway or woodland opening (measured by pacing), and the number of limbs supporting a nest were also recorded. Once eggs hatched, nests were visited every 2-4 d and nestlings were marked, weighed, and measured. Culmen, tarsal length, and lengths of the sixth primary (conven- tional numbering) of the right wing and the right central rectrice were measured. Young were banded with U S Fish and Wildlife Service bands prior to fledgling. In 1979 plumages of breeding Broad-wings were used as a basis for age determination. Naphthalene crystals were placed at the base of active nest trees to deter terrestrial predators. Addled eggs and prey remains were collected at nests whenever possible, and behavior of the nesting hawks was observed. Predation was detected by observing Great Horned Owls ( Bubo mrgimanus, GHO) near nests, pres- ence of GHO feathers in nests, or at 1 nest noticing the strong odor of Striped Skunk ( Mephitis mephitis) in the nest (Striped Skunks are not known to climb trees [R. Cole pers. comm.] and GHO often prey on Striped Skunks). Nestlings were aged by observing hatching dates. Sta- tistical analyses followed Sokal and Rohlf (1969). Results Nesting Habitat. Both BWH and RSH nested in approximately the same type of upland woods, but with specific differences in nest sites. Each species used a variety of tree species, but each showed a preference (Table 1). BWH used larch ( Larix de- cidua) most often (33%), whereas RSH apparently preferred American beech ( Fagus grandifolia 39%). BWH showed a tendency to nest on slopes, but RSH chose relatively level terrain. Both species nested < 1 km from streams, ponds, or swamps. BWH seemed to nest closer to woodland openings (road cuts, field, etc.) and lower in trees than RSH, respectively: 90 ± 54 m [N = 17] versus 182 ± 164 m [N = 6]; and, 11.8 ± 3.2 m [N = 18] versus 13.4 ± 2.4 m [N — 11], although neither comparison was statistically significant (distance to woodland opening F = 2.15, P > 0.1, single classification analysis of variance (ANOVA) and nest height F = 1.85, P > 0.1, AN- OVA). Both species seemed to prefer the first sub- stantial crotch in a tree except in larch where nests were placed on a platform of horizontal branches against the trunk, and both used approximately the same number of limbs for nest support (BWH, 4.3 ± 1.1 [N = 16] and RSH, 4.2 ± 0.7 [N = 12]). Nesting Density. We could measure ecological density only on the 1200 ha of the Ganadaway Greek Wildlife Management Area and adjoining private land. In 1978, 6 RSH pairs and 5 BWH pairs were found. Additionally, we suspected the presence of 1 additional pair of each species based on sightings Winter 1989 Broad-wing and Red-shoulder Biology 127 Table 1. Species of trees used for nest sites by Broad- winged and Red-shouldered Hawks in western New York. Only active nests were counted. Tree Species Frequency of Use Red- Broad- shoul- WING DER Scotch Pine ( Pinus sylvestris ) 1 0 European Larch ( Larix decidua) 6 1 Quaking Aspen ( Populus tremuloides) 1 1 Yellow Birch ( Betula alleghaniensis) 1 2 American Beech ( Fagus grandifolia ) 0 7 Northern Red Oak ( Quercus rubra ) 2 0 Sweet Crab Apple ( Malus coronaria) 1 0 Black Cherry ( Prunus serotina ) 3 1 Sugar Maple (Acer saccharum) 2 0 Red Maple (Acer rubrum ) 0 2 Maple sp. (Acer sp.) 0 2 American Basswood (Tilia americana) 1 0 White Ash (Fraxinus americana) 0 2 Total number of tree species 9 8 Total nests 18 18 and calls of hawks. Including suspected pairs, nest- ing densities of RSH and BWH were 1 pair/171 ha and 1 pair/200 ha, respectively, of forested land. BWH and RSH more often nested adjacent to each other than to conspecifics. Nearest neighbor dis- tances were 877 ± 422 m (N = 12) for BWH/RSH, 1441 ± 331 (N = 11) for BWH/BWH, and 1271 ± 640 (N = 4) for RSH/RSH. A single active Barred Owl ( Strix varia) nest was 120 m from a RSH nest and 578 m from a BWH nest. Several other species of raptors were present dur- ing the breeding season in the Ganadaway Creek area (Crocoll and Parker 1986; Table 2) but were not carefully censused, and some nesting pairs were probably overlooked. Nevertheless, overall raptor density was conservatively estimated at 1 pair/63.2 ha. Nesting Chronology. BWH. Broad-winged Hawks were first seen on the study area on 12 April in 1978, on 23 April in 1979, and on 19 April in 1980. These were probably local breeding birds be- cause active nests were subsequently found near sightings. Migrant BWHs were also observed in the study area in April. Nest building or rebuilding began soon after occupying nesting territories. Active Table 2. Raptor species occurring in the 1200 ha Can- adaway Creek Wildlife Management Area and adjacent woodlands and probable densities in 1978. Species Occurrence Cathartes aura 1 or 2 pairs Circus cyaneus 1 pair Accipiter striatus 1 pair Buteo jamaicensis 1 pair Buteo hneatus 6 or 7 pairs Buteo platypterus 5 or 6 pairs Falco sparverius 1 pair Asio otus 1 pair Bubo virginianus 1 pair Strix varia 1 pair nests were verified by the presence of green sprigs or down, or a bird on the nest. Nests were found as early as 24 April, and partially constructed nests were found as late as 14 May, suggesting a nest construction period of approximately 3 wks. Back- dating from hatching dates, and assuming 30 days for incubation as did Matray (1974), egg laying occurred mainly during the second and third wks of May. Eleven of 12 BWH eggs hatched in the first 2 wks of June (Fig. 1). Variation in hatching date seemed unrelated to year or clutch size. Mean first egg hatch- ing dates for 1978, 1979, and 1980 were 16 June, 7 June, and 10 June, respectively. Mean first egg hatching date for the 3-yr period was 10 June. Average interval ( x ± S.D.) between any 2 eggs hatching for any size clutch was 1.87 ± 1.36 d (N = 31). Eggs hatched 1 d apart (N = 4) in all 2-egg clutches (N = 4). In 3-egg clutches the average time between any 2 eggs hatching was 2.00 ± 1.41 d (N = 24). The second egg hatched an average of 0.38 ± 0.52 d (N = 8) after the first, and the third egg hatched an average of 2.62 ± 0.74 d (N = 8) after the second. In a single 4-egg clutch, the second egg hatched <24 hr after the first, the third egg failed to hatch, and the fourth hatched 3 d after the second. First flights of nestlings, usually in response to disturbance, occurred at 31.4 d of age ( X ; N = 11, range 29-39). There was little variation in the age of first flight among individuals from broods of dif- ferent sizes. In this study fledging was defined as the ability of an individual to sustain horizontal flight and occurred when the nestlings were 5-6 wks old. 128 Scott T. Grocoll and James W. Parker Vol. 23, No. 4 A. < z <0 H CO Ul z 6 - 5- 4- 3- 2 - 1 JUNE 1 JUNE 8 JUNE 16 JUNE 24 TO 7 TO 15 TO 23 TO 30 4 3 2 1980 1979 1978 Figure 1. Hatching dates for the first egg in clutches of Broad-winged Hawks. The numbers show clutch sizes in A and years in B. Earliest fledging date was 4 July and the latest was 2 August. Most young (88.9%) fledged during the second and third wks of July. BWH were last seen on the study area on 5 September in 1978 and on 3 September in 1979. RSH. Twenty active RSH nests were found dur- ing our 3-yr study but were intensively studied only during 1979. The following schedule was construct- ed primarily from data obtained in 1979. RSH were first seen during the second and third wks of March when nest relining was also observed. Adults were sitting on nests as early as the first wk in April and may have been incubating eggs. Most eggs hatched during the second and third wks of May, which suggests a minimum incubation period of 4-5 wks in western New York. Fledging occurred during the third and fourth wks of June, with the exception of 1 nest at which a single young fledged on 11 July, suggesting a nestling period of approx- imately 5 wks. Reproductive output and nest success were ana- lyzed for each of 3 time periods. The egg period encompassed the time from laying of the first egg until its hatching, the nestling period was the time from the hatching of the first egg until the fledging of the first young, and the fledging period was the time from the fledging of the first young until mi- gratory departure. In most respects the BWH seemed more successful in 1979 than in 1978 (Table 3), but sample sizes were small and statistical significance could seldom be shown. Clutch size was significantly larger (F = 4.92, P < 0.05, ANOVA) in 1979 than 1978. Most notably, overall productivity of the fledglings per nesting attempt in 1979 was nearly twice that of 1978. Ten of 18 BWH nesting attempts suffered some mortality. Total nest loss occurred more often during the egg period in both years, but only for 1979 was the difference statistically significant (t s = 1.96, P < 0.05, arcsine test). In 1979 more nests hatching eggs survived the nestling period than in 1978 (t s = 1.71, P < 0.05). In 1979 the overall percentage of suc- cessful nests seemed higher than 1978, but the dif- ference was not statistically significant ( t s = 1.25, P ~ 0 . 10 ). Because records for some RSH nests are incom- plete, only limited generalizations are possible. RSH clutch size was larger than BWH (Table 3), but RSH fledged fewer young, which probably reflects largely the lower hatching success for RSH. A third egg which failed to hatch was suspected for two 2-egg clutches because of the presence of egg shell fragments beneath the nest early in the egg period. Therefore, data presented in Table 3 represents the minimum percentage of eggs hatched. Discounting the observed shell fragments would have resulted in a hatching percentage of 77%. Rates of nest success for RSH and BWH were similar during the egg period, but BWH nests seemed more successful dur- ing the nestling period. Success and Clutch Size. Only 1 BWH clutch each of 1 egg and 4 eggs was found. Therefore, comparisons of success (Table 4) were made only between clutches that were originally of 2 or 3 eggs. The average number of fledglings per nest was not statistically greater with larger clutch size, but a larger sample size might well establish statistical significance. Two-egg clutches produced more fledg- lings/egg hatched than did 3-egg clutches ( t = 2.45, P < 0.01), but a similar comparison for fledglings/ egg laid was not statistically significant. Percentages Winter 1989 Broad-wing and Red-shoulder Biology 129 Table 3. Productivity and nest success of Broad- winged and Red-shouldered Hawks in western New York, Sample sizes are in parentheses. Parameter Broad-winged Hawk Red- shouldered Hawk 1978 1979 Total All Years Total nests found 6 10 18 a 9 Average clutch size 2.00 ± 0.82 (4) 2.89 ± 0.60 (9) 2.60 ± 1.09 (15) a 3.00 ± 1.00 (5) Average no. of eggs hatching 1.40 ± 1.34 (5) 2.30 ± 1.25 (10) 1.88 ± 1.32 (17) a 2.00 ± 1.58 (5) Percentage of eggs hatching 90.9 (11) 88.5 (26) 83.3 (42) a 66.7 (15) Percentage of fledglings per egg laid 75.0 (8) 73.1 (26) 69.2 (39) a — Percentage of fledglings per egg hatched 85.7 (7) 82.6 (23) 84.4 (32) a — Avg. fledglings per nest 1.00 ± 1.10 (6) 1.90 ± 1.20 (10) 1.50 ± 1.20 (18) a 1.11 ± 1.17 (9) Avg. fledglings per successful nest 2.00 ± 0 (3) 2.38 ± 0.74 (8) 2.25 ± 0.62 (12) a 2.00 ± 0.71 (5) % Nests successful in: Egg period 66.7 (6) 80.0 (10) 72.2 (18) a 75.0 (8) Nestling period 75.0 (4) 100 (8) 92.3 (13) a 83.3 (6) Overall 50.0 (6) 80.0 (10) 66.7 (18) a 55.6 (9) ■' Total includes 2 nests found in 1980. of successful nests in both the egg and nestling pe- riods were very similar between 2-egg and 3-egg clutches. Success and Age of BWH, BWH yearlings were first observed paired and breeding with adults in 1979. Age of both parents was known for only 1 pair in 1978 and 1980. All breeding yearlings were males. Adult/adult pairs were more successful than adult/yearling pairs for most criteria in Table 5. However, only the percentage of eggs hatching showed a significant difference (t s = 2.31, P — 0.01) favoring the adult/yearling pairs. Clutch size of 1 mixed pair was unknown and was excluded from some of the calculations. Success and Reuse of Nests by Broad-wings. Of fifteen BWH nests of known age, 5 (33%) had been rebuilt (Table 6). Clutch size was not statis- tically different between the 2 ages of nests, but nest success appeared greater in all criteria for newly constructed nests compared to reused nests, although Table 4. Productivity of different clutch sizes 3 in the Broad-winged Hawk in western New York. Sample sizes are in parentheses. Clutch Size 1 2 3 4 Number of nests 1 5 8 1 Average No. of eggs hatching 0(1) 1.60 ± 0.89 (5) 2.62 ± 1.06 (8) 3.00 (1) Percentage of eggs hatching 0(1) 80.0 (10) 87.5 (24) 75.0 (4) Percentage of fledglings per egg laid 0(1) 80.0 (10) 66.7 (24) 75.0 (4) Percentage of fledglings per egg hatched — 100 (8) 76.2 (21) 100 (3) Average fledglings per nest 0(1) 1.60 ± 0.89 (5) 2.00 ± 1.06 (8) 3.00 (1) Average fledglings per successful nest % Nests successful in: — 2.00 ± 0 (4) 2.28 ± 0.76 (7) 3.00 (1) Egg period 0(1) 80.0 (5) 87.5 (8) 100 (1) Nestling period — 100 (4) 100 (7) 100 (1) Overall 0(1) 80 (5) 87.5 (8) 100 (1) ,l Clutch size was not determined at 2 nests and one pair did not lay eggs. 130 Scott T. Crocoll and James W. Parker Vol. 23, No. 4 Table 5. Reproductive output and nest success by age composition of pairs of breeding Broad-winged Hawks. Sample sizes in parentheses. Pair Composition Adult/ Adult Adult/Subadult Total number of nests 6 4 Average clutch size 3.00 ± 0.63 (6) 2.67 ± 0.58 (3) Average No. of eggs hatching 2.33 ± 1.21 (6) 2.00 ± 1.41 (4) Percentage of eggs hatching 77.8 (18) 100 (8) Percentage of fledglings per egg laid 72.2 (18) 75.0 (8) Percentage of fledglings per egg hatched 92.9 (14) 75.0 (8) Average fledglings per nest 2.17 ± 1.17 (6) 1.50 ± 1.39 (4) Average fledglings per successful nest % Nests successful in: 2.60 ± 0.55 (5) 2.00 ± 1.00 (3) Egg period 83.3 (6) 75.0 (4) Nestling period 100 (5) 100 (3) Overall 83.3 (6) 75.0 (4) the difference showed statistical significance only in the cases of percentage of eggs hatching (t s = 2.22, P < 0.05) and percentage of fledglings/egg laid (t s = 2.27, P < 0.05). Success and BWH Nest Sites. Fledging success of BWH was not related to the number of limbs supporting nests (F s = 0.62, P > 0.5, ANOVA) or nest elevation (F s = 0.81, P > 0.25, ANOVA) (Fig. 2A and 2B). A trend toward increased success was indicated for nests farther from woodland openings but was not statistically significant (F s = 2.1, P > 0 1, ANOVA; Fig. 2C). Essentially, BWH nested in 2 habitat types: de- ciduous mixed woodland, and small-stand conifer plantations (approximately 1 ha) contiguous with beech-maple-hemlock forest. Average clutch size and average number of fledglings/nest were similar be- tween different habitats (Table 7). However, ap- proximately 1 less egg/nest hatched from nests in conifer plantations than from those in the mixed deciduous-conifer habitat, apparently the result of greater loss of nests in plantations during the egg period (t s = 1.59, P — 0.06). A larger percentage of eggs hatching in deciduous mixed habitat as com- pared with plantations (t s = 2.72, P < 0.01) was associated with a higher percentage of fledglings/ Table 6. Nest success of new versus rebuilt Broad-winged Hawk nests in western New York. Sample sizes in parentheses. New Nests Rebuilt Nests Total number of nests 10 5 Average clutch size 2.44 ± 1.13 (9) 2.40 ± 0.39 (5) Average No. of eggs hatching 2.10 ± 1.20 (10) 1.60 ± 1.52 (5) Percentage of eggs hatching 95.4 (22) 66.7 (12) Percentage of fledglings per egg laid 86.4 (22) 50.0 (12) Percentage of fledglings per egg hatched 90.5 (21) 75.0 (8) Average fledglings per nest 1.90 ± 1.00 (10) 1.20 ± 1.30 (5) Average fledglings per successful nest % Nests successful in: 2.38 ± 0.52 (8) 2.00 ± 1.00 (3) Egg period 80.0 (10) 60.0 (5) Nestling period 100 (8) 100 (3) Overall 80.0 (10) 60.0 (3) Winter 1989 Broad-wing and Red-shoulder Biology 131 egg laid, although the only nest that failed during the nestling period was in mixed habitat. Percentage of successful nests was essentially the same during the egg and nestling periods in the mixed habitat, but a greater percentage of nests failed in the egg period than in the nestling period in conifer plan- tations ( t s = 2.28, P < 0.05). Mortality Factors. Four mortality factors re- duced BWH reproductive success. Nest desertion caused only the loss of 1 egg and 1 nest attempt, and egg failure was known to cause only the loss of 1 egg. Predation accounted for half the loss of Broad- wing nests (5 of 10), and most occurred prior to hatching. Five eggs and 4 nestlings were known to be lost to predators, and GHO were judged respon- sible at 4 nests. During the 3 yrs of the study, 4 adults were killed by predators, all at different nests, and 3 of those by GHO (see Methods). The re- maining adult died from an unknown cause. Four yearlings were recruited into the breeding popula- tion in 1979, indicating that at least 3 additional adults in the local population probably suffered some form of mortality prior to the breeding season. Star- vation was secondary to predation by causing the loss of 4 nestlings (2 in 1 nest, 1 in each of 2 others). RSH mortality showed a roughly similar pattern. Predation eliminated 3 nests (1 egg, 1 nestling known lost). Starvation impacted 2 nesting attempts (3 nest- lings at 2 nests) but resulted in no complete nest failures. It was not possible to determine if predation on Red-shoulder nests was more prevalent during the egg period than the nestling period, but Great Horned Owls were the responsible predator in all cases judging by the presence of feathers and other observations similar to failed BWH nests. Food Habits. BWH in Chautauqua County, New York were generalized predators bringing at least 14 species of vertebrates to nests (Table 8). Because invertebrate remains were rarely found in the nests, their importance in the BWH diet is unclear. Clear- ly, mammals (chipmunks, moles, shrews) and mis- cellaneous birds predominated, but toads and garter snakes were taken also. The percent composition of different vertebrates in the Broad-wing’s diet changed considerably between 1978 and 1979 with the great- est shift being to birds. Of 20 prey items found in 1978, 12 (60%) were mammals, 4 (20%) were rep- tiles, and only 2 each (10%) were birds and am- phibians. In 1979, 14 (33%) of 42 prey items were birds, mammals fell to 40% (17), while reptiles and amphibians 4 (10%) and 7 (17%), respectively, re- co Ul z o UJ (9 a Ui a z o >- a: UJ m 2 3 Z 'M' > b > B. NUMBER OF LIMBS SUPPORTING NEST DISTANCE OF NEST FROM WOODLAND OPENING(METERS) Figure 2. Production of BWH fledglings associated with A, nest height; B, number of supporting limbs, and C, distance of nest from a woodland open- ing. Each point represents the mean and the bars represent the S.D. Nest sample size ap- pears above each bar. mained uncommon. In 1978 an average of 0.35 ver- tebrate prey items were found during each nest visit (N = 57) in comparison to 0.54 prey items found per visit (N = 78) in 1979. Prey items [2 Eastern Chipmunks; 1 Shorttail Shrew; 1 Masked Shrew (Sorex cinereus )] were found at only 2 RSH nests. Discussion Nesting Characteristics and Densities of BWH and RSH. BWH. In most parts of its range, BWH associates with dense forests (Burns 191 1; Bent 1937). However, forest type is quite variable [e.g., maple- beech-hemlock in this study, oak-aspen in Wiscon- sin-Minnesota (Keran 1978), hemlock-yellow birch 132 Scott T. Crocoll and James W. Parker Vol. 23, No. 4 Table 7. Reproductive success of Broad-winged Hawks in two habitat types in western New York. Sample sizes in parentheses. Deciduous Mixed Woodlands Conifer Plantations Total number of nests 11 7 Average clutch size 2.56 ± 1.13 (9) 2.28 ± 0.76 (7) Average No. of eggs hatching 2.44 ± 1.01 (9) 1.43 ± 1.40 (7) Percentage of eggs hatching 95.6 (23) 62.5 (16) Percentage of fledglings per egg laid 73.9 (23) 62.5 (16) Percentage of fledglings per egg hatched 77.3 (22) 100 (10) Average fledglings per nest 1.54 ± 1.13 (11) 1.43 ± 1.40 (7) Average fledglings per successful nest % Nests successful in: 2.12 ± 0.64 (8) 2.50 ± 0.58 (4) Egg period 90.0 (10) 57.1 (7) Nestling period 88.9 (9) 100 (4) Overall 80.0 (10) 57.1 (7) in the Adirondack Mountains (Matray 1974)]. Only in this study has BWH been observed nesting in conifer plantations. Other studies reported nesting close to some type of forest opening (Table 9), but BWH in western New York unaccountably nested farther from forest openings than other populations. Nest heights appeared to be similar among different regions except for Wisconsin where nests were sub- stantially lower (Keran 1978). Clearly, BWH dis- plays a degree of adaptability in choice of nesting habitat, but no detailed patterns are apparent. Few comparative data exist on BWH nesting den- sities. Rusch and Doerr (1972) reported about 1 pair/23.3 km 2 , which is much less dense than the Chautauqua study area (approximately 1 pair/2 km 2 ), probably because the Alberta study site was on the western edge of the species’ range. Keran (1978) and Rosenfield (1984) reported breeding den- sities of 1 pair/5.2 km 2 and 1 pair/2.4 km 2 , re- spectively, for Minnesota and Wisconsin. In Chau- tauqua County the BWH appeared subjectively to be about the third or fourth most common raptor, behind RSH and Red-tailed Hawks ( Buteo jamai- censis ) and possibly GHO. We know of only 3 other studies that reported on densities of raptor communities: Craighead and Craighead (1956) in Michigan (1 pair/144.7 ha), Brown (1966) in Kenya (1 pair/84 ha), and U.S.D.I. (1979) in the Snake River Birds of Prey study area, Idaho (1 pair/91. 7 ha). Clearly, total raptor density in the Canadaway Creek (1 pair/63.2 ha) ranks with others as a very dense raptor community, whether or not birds forage outside the study area (as do the birds at Snake River). Breeding schedule of BWH in western New York seems to coincide most closely with Wisconsin-Min- nesota (Keran 1976) and the central Appalachians (Janik and Mosher 1982). Egg laying and hatching occurred from early May to early June and fledging from mid to late July in both the Wisconsin-Min- nesota and central Appalachian populations. RSH. Although several studies (Stewart 1949; Wiley 1975; Preston et al. 1989) indicate an asso- ciation between Red-shouldered Hawk nesting and low floodplain forests, only 1 of our 20 active RSH nests was in floodplain forest. Most were in upland hills, but all were close to some form of surface water which is in agreement with other studies. Forest habitat used by RSH in this study was most similar to that used by populations in the Waterloo region of Ontario (Sharp and Campbell 1982) and in south- western Quebec (Morris et al. 1982). The Canadian studies as well as Titus and Mosher (1987) also indicated preference for the American beech as a nest tree. A comparison of several studies (Table 10) sug- gests that RSH use nest sites with similar charac- teristics throughout much of their geographic range However, few estimates of nesting densities are available. Craighead and Craighead (1956) com- puted only crude density for an entire township, which included unsuitable cropland as well as suit- able woodland habitat. Consequently, their nesting density is not directly comparable to western New Winter 1989 Broad-wing and Red-shoulder Biology 133 York. Stewart (1949) documented a nesting density in central Maryland that was very similar to western New York even though habitats were much differ- ent. In central Maryland, RSH nested only in low river floodplains where the only other nesting raptor was the Barred Owl. Interestingly, the RSH was the most common nesting raptor in Chautauqua County during the study despite the fact that numbers appear to be declining in many parts of the range (Crocoll and Parker 1988). Most studies (Bednarz and Dinsmore 1981; Kimmel and Fredrickson 1981; Sharp and Campbell 1982) indicate that the primary cause of decline is loss of extensive forested habitat. Cur- rently, habitat does not appear to be limiting in western New York (Crocoll and Parker, pers. obs.) and may in fact be increasing due to a reversion of farm land and old timber harvest areas to woodlands. Several other studies have reported on RSH nest- ing chronology (Craighead and Craighead 1956; Henny et al. 1973; Wiley 1975; Portnoy and Dodge 1979; Kimmel and Fredrickson 1981). Chronology in Michigan (Craighead and Craighead 1956) was more similar to western New York than other areas, where some showed earlier (Henny et al. 1973; Wi- ley 1975; Kimmel and Fredrickson 1981) and some later (Portnoy and Dodge 1979) schedules. Other studies (Stewart 1949; Titus and Mosher 1981) indicated that BWH and RSH do not usually nest in close proximity. However, we found the op- posite true in western New York as did Armstrong and Euler (1982) in central Ontario. Though the general nesting habitats of each species are similar, there probably exist subtle habitat differences sim- ilar to those reported in Titus and Mosher (1981), such as distance to nearest forest opening, topogra- phy, tree density, ground cover, and percent conif- erous trees (Armstrong and Euler 1982) that help reduce potential competition for nest sites. Also, as Fuller (1979) suggested, two species may use over- lapping home ranges at different times of the day or week to reduce potential competition for space. Comparative Reproductive Success of Differ- ent BWH and RSH Populations. BWH. Table 1 1 presents reproductive data for BWH populations from 6 studies. They permit few statistical compar- isons and only one comparison among all six pop- ulations (% nests successful overall). Only for the population in western New York has the relation- ship between habitat, age of nesters, clutch size, nest reuse, productivity, and reproductive success been Table 8. Food items in 13 Broad-winged Hawk nests 1978-1980 in Chautauqua County, New York. Species N % Occurrence Mammals Blarina brevicauda 9 13.0 Condylura cristata 6 8.7 Parascalops breweri 2 3.0 Tamias striatus 10 14.5 Microtus pennsylvanicus 2 3.0 Unidentified 3 4.3 Total mammals 32 46.4 Birds Colap tes auratus 2 3.0 Cyanocitta cristata 3 4.3 Bombycilla cedrorum 1 1.4 Junco hyemalis 2 3.0 Unidentified 10 14.5 Total birds 18 26.1 Reptiles Storeria dekayi 1 1.4 S. occipitomaculata 1 1.4 Thamnophis sirtalis 5 7.2 Unidentified 2 3.0 Total reptiles 9 13.0 Amphibians Bufo americanus 7 10.1 Rana pipiens 2 3.0 Total amphibians 9 13.0 Unidentified vertebrate 1 1.4 Total vertebrates 69 100 Invertebrates Crustacea (Crayfish) 2 p Insecta (grasshoppers and caterpillars) p p reported. Some measures of productivity were roughly similar among different populations, but the western New York population with a high average clutch size had one of the lowest average numbers of young fledged per nest. This study area also seems to have the lowest percentage of successful nests, although the difference was not statistically significant. Lower success in western New York was at least partly due to the statistically lower success of nests in the egg period (G = 7.44, P < 0.025 R x C test of inde- pendence). Surprisingly, causes of mortality and nest failure have received attention in only 1 other study (Ro- 134 Scott T. Grocoll and James W. Parker Vol. 23, No. 4 Table 9. Nesting habitat characteristics of several Broad-winged Hawk populations. Sample size in parentheses. Nest Height (m) No. Support Limbs Distance to Forest Opening (m) Wisconsin-Minnesota (Reran 1978) — 3.8 ± 0.4 42 Central Appalachians (Titus and Mosher 1981) 13.7 ± 3.0 (24) — 63 ± 61 (24) Western New York (this study) 11.8 ± 3.2 (18) 4.3 ± 1.1 90 ± 54 (17) Adirondack Mountains, NY (Matray 1974) 13.3 ± 1.36 (14) — — Wisconsin (Rosenfield 1984) 8.2 ± 2.7 (72) — — Central Ontario (Armstrong and Euler 1982) 11.8 ± 2.8 (27) — 42.1 ± 30.2 (27) senfield 1984) which reported that almost twice as many eggs as nestlings were lost, concluding that major impacts occurred during the egg period, as we observed. Rosenfield (1984) also observed the same types of mortality, except starvation of nestlings was not detected. He did record a significantly higher incidence of eggs that failed to hatch (9.1%) com- pared to our western New York study (2.4%) (f s = 1 75, P — 0.04 arcsine transformation, test of equal- ity of 2 percentages) and suspected that predation was due to Raccoon ( Procyon lotor) and American Crow ( Corvus brachyrhynchos). Ours is the first study to report predation by GHO, which clearly can have a substantial impact on a local BWH population (also see Parker 1974). Several raptors are known to reuse nests (Brown and Amadon 1968). Rosenfield (1984) reported reuse of nests both in the year after and 2 yrs after con- struction. To reuse nests is probably energetically more advantageous than building anew (Weeks 1978). In addition nest reuse is most likely to occur if the nest was successful the previous year (Craig- head and Craighead 1956). In our study, however, reuse only occurred 2 yrs after construction and most often in pine plantations. In fact, reused nests in New York were less successful than new ones, nests further from woodland openings appeared more pro- ductive than nests relatively close to openings (Fig. 2), and reused nests were much closer to openings than newly constructed nests. We suggest reuse of a nest, placement nearer open areas, or both in- creases the likelihood of nests being noticed by GHO One possible pressure for nest reuse in pine plan- tations was the lack of suitable nest trees (Crocoll and Parker, pers. obs.). Although Buteo species usually start breeding at 2 or 3 yrs (Newton 1977) in adult plumage, there have been some reports of breeding subadults (Red- Table 10. Red-shouldered Hawk nest site characteristics. Sample size in parentheses. Nest Height (m) No. Supporting Limbs Distance To Forest Opening (m) Breeding Density Western New York (this study) 13.4 ± 2.6 (11) 4.2 ± 0.7 (12) 182 ± 164 (6) 1 pair/171 ha Southwest Quebec (Morris, et al. 1982) 14.0 ± 3.6 (54) 4.2 (54) — — Waterloo Region (Sharp and Campbell 1982) 14 (9) — — — Missouri (Kimmel and Fredrickson 1981) 14.4 ± 3.6 (14) — — — Iowa (Bednarz and Dinsmore 1982) 19.1 ± 4.8 (11) 3.6 ± 0.5 (11) — — Massachusetts (Portnoy and Dodge 1979) 12.0 ± 1.7 (12) — — — Central Appalachians (Titus and Mosher 1981) 13.4 ± 3.0 (10) — — Central Maryland (Stewart 1949) 15 (57) — 180 + 141 (10) 1 pair/48. 7 ha Michigan (Craighead and Craighead 1956) — — 1 pair/645 ha Central Ontario (Armstrong and Euler — 1982) 14.0 ± 3.6 (9) — 27.8 ± 18.2 (9) — Arkansas (Preston et al. 1989) — 173.3 ± 73.1 (19) — Winter 1989 Broad-wing and Red-shoulder Biology 135 Table 11. Comparison of Broad-winged Hawk reproductive output among 6 populations. Sample sizes in parentheses Central Ontario (Arm- strong and Euler 1982) Alberta, Canada (Rusch and Doerr 1972) Minnesota and Wisconsin (Reran 1976) Chautauqua County, NY (This Study) Central Appala- chians (Janik AND Mosher 1982) Wisconsin (Rosen- FIELD 1984) Total nests found 16 5 12 18 36 72 Average clutch size — 2.40 ± 0.55 (5) 2.17 ± 0.39 (12) 2.60 ± 1.09 (15) 2.7 (15) 2.36 (70) Average no. of eggs hatching — 2.40 ± 0.55 (5) 2.00 ± 0.60 (12) 1.88 ± 1.32 (17) 2.1 (29) 1.81 (70) Percentage of eggs hatching — 100 (12) 92.3 (26) 83.3 (42) — 77.0 (165) Percentage of fledglings per egg laid — 83.3 (12) 84.6 (26) 69.2 (39) — 64.8 (165) Percentage of fledglings per egg hatched — 83.3 (12) 91.7 (24) 84.4 (32) — 84.3 (127) Average fledglings per nest 1.5 2.00 (5) 1.83 ± 0.83 (12) 1.50 ± 1.20 (18) 1.74 (34) 1.53 (70) Average fledglings per successful nest 1.7 2.00 (5) 2.00 ± 0.63 (11) 2.25 ± 0.62 (12) 1.90 (31) 1.94 (55) % Nests successful: Egg period — 100 (5) 100 (12) 72.2 (18) — — Nestling period — 100 (5) 91.7 (12) 92.3 (13) — — Overall 87.5 (16) 100 (5) 91.7 (12) 66.7 (18) 86.1 (36) 78.6 (70) tailed Hawks, Luttich et al. 1971; and Red-shoul- dered Hawks, Henny et al. 1973; Wiley 1975; Apan- ius 1977). Yearling Broad-wings have not previously been reported to nest (Newton 1979). In general young breeders are not as successful as older individuals (Lack 1968; Fisher 1975; Brown 1978). Additionally, Crawford (1977) indicated that subadult female Red-winged Blackbirds ( Agelaius phoemceus ) and Yellow-headed Blackbirds ( Xan - thocephalus xanthocephalus ) nested in lower quality territories than adults. In our study, BWH pairs that included subadults were apparently less suc- cessful than adult/adult pairs (Table 5) even though the former used the habitat (mixed deciduous co- niferous forest) producing more young per nest and greater nest success (Table 7). Male BWH provide all the food during incuba- tion and the early nestling period (Matray 1974). Therefore, mixed pairs in which the male is the subadult should be most successful in years of high food abundance when inexperienced breeding males would have a better chance of obtaining sufficient food. Subadult breeders of many species start nesting later in the season than more experienced breeders [Laysan Albatross ( Diomedia immutabilis) , Fisher 1975; European Sparrowhawk ( Accipiter nisus), Newton 1976; Great Horned Owl, Mclnvaille and Keith 1974]. However, in our study mixed pairs did not breed later in the season than adult/adult pairs, and overall the population in western New York appeared to breed synchronously, as do BWH in the Adirondacks (Matray 1974). Several studies showed that nesting habitat has an effect on the nesting success of raptors (Howell et al. 1978; Newton 1976). In western New York, mixed deciduous woodlands were apparently supe- rior to conifer plantations for BWH breeding (Table 7), even though the difference in reproductive success was not large. This situation may be explained (at least in part) by a combination of several factors. First, 67% of the rebuilt nests were found in pine plantations, and reused nests were less productive than new nests. Second, a positive (though statisti- cally non-significant) relationship exists between nest success and distance of nest from a woodland opening (Fig. 2G), and nests in mixed deciduous woodlands were farther from a woodland opening than nests in pine plantations (92.6 m and 85.7 m, respectively). However, mixed pairs were in deciduous-mixed 136 Scott T. Crocoll and James W. Parker Vol. 23, No. 4 Table 12. Comparison of Red-shouldered Hawk reproductive output among different populations. Sample sizes in parentheses. Total Nests Found Average Clutch Size Average No. of Eggs Hatching Average Fledglings per Nest Average Fledglings per Success- ful Nest % Nests Success- ful California (Wiley 1975) 29 2.69 ± 0.54 (29) 2.07 ± 1.03 1.34 ± 1.14 (29) 2.05 (19) 65.5 (19) Iowa (Bednarz 1979) 8 — 2.9 (8) 2.9 (8) 3.3 ± 0.76 (8) 87.5 (8) Michigan (Craighead and Craighead 1956) 40 3.42 (40) 2.55 (40) 1.77 (40) — — Missouri (Kimmel and Fredrickson 1981) 9 — — 2.56 ± 0.53 (9) 2.56 ± 0.53 (9) 100 (9) New York (this study) 9 3.00 ± 1.00 (5) 2.00 ± 1.58 (5) 1.11 ± 1.17 (9) 2.00 ± 0.71 (5) 55.6 (9) Maryland — western (Janik and Mosher 1982) 17 3.1 (6) 2.4 (8) 1.8 (10) — 52.9 (9) Maryland — central (Henny et al. 1973) 74 — — 1.58 (74) 2.34 (50) 67.6 (74) Massachusetts (Portnoy and Dodge 1980) 9 3.33 (9) 2.67 (3) 2.00 (5) 2.50 (4) 80.0 (5) Central Ontario (Arm- strong and Euler 1982) 6 — — 1.8 2.2 83.3 (6) habitat, but were less successful than pairs of adults. Consequently, nest success for BWH is less affected by habitat at the nest than by age of nesters. Newton (1976) noted a similar situation in regard to habitat for the European Sparrowhawk and suggested that the higher success observed in the best habitat was due to considerably higher prey availability. We did not quantify prey availability in different hunting habitats. Many researchers have shown that species diver- sity and individual species densities are often greater in an ecotone than in adjacent habitats. Our obser- vations agree with Reran (1978) and Titus and Mosher (1981) that BWH hunt the forest edge, particularly near wetlands or ponds in response to higher prey density. Gates and Gysel (1978) docu- mented an increased density of open-nesting pas- serines in a field-forest ecotone even though there was a concomitant decrease in fledging success, ap- parently because of predation. In fact, as for pas- serines, BWH nesting success declined near forest edge, possibly because of the increased likelihood of predation. RSH. RSH populations in western New York appear uniformly less successful than those in other regions (Table 12), and the species should be con- sidered as threatened in New York. Weather and human disturbance can be major causes of repro- ductive failure for RSH (Craighead and Craighead 1956; Henny et al. 1973; Wiley 1975; Sharp and Campbell 1982), and Raccoons and GHO are known predator threats (Craighead and Craighead 1956; Wiley 1975; Bednarz 1979; Portnoy and Dodge 1979). However, 'only Wiley (1975) recorded GHO predation more than once. In contrast the GHO appears to be the major reason for the low success of RSH in western New York. Breeding Strategies of the Broad-winged Hawk. Newton (1977) discussed trends in breeding strategy over evolutionary time in small, short-lived raptors having large clutches, high breeding rates, and early maturity (“r-selection,” MacArthur and Wilson 1967 and many later references) and in large, long-lived raptors having small clutches, low breeding rates and late maturity (“k-selection”). Although these terms, and the theory associated with them, have undergone considerable discussion and controversy, they still can be validly used to represent two contrasting pat- terns of reproduction at opposite ends of a contin- uum. Compared to buteos, BWH in western New York tends toward an “r”-strategy as evidenced by high adult mortality, early maturity, and seeming early recruitment into the breeding population. Rel- atively high reproductive failure for BWH in some Winter 1989 Broad-wing and Red-shoulder Biology 137 (lesser-quality) habitats and in years of apparent low food supplies may or may not reinforce an r strategy for BWH reproduction, but would not select against it. O’Connor (1977) theorized that birds might evolve 1 of 3 breeding strategies by which to maximize reproductive effort within a given year: clutch ad- justment, brood reduction, or resource storage. Going further, Hirschfield and Tinkel (1975) suggested that an organism might be flexible, and to maximize contributions to future generations, adjust repro- ductive strategy from one breeding season to the next, based on environmental constraints. Indeed, Howe (1976) showed that the Common Grackle can si- multaneously utilize 2 strategies. Clearly, the BWH often shows brood reduction as do many other rap- tors (see Newton 1977 for a general review). High adult BWH mortality (Crocoll and Parker, pers. obs.) probably puts an adaptive premium on max- imum, annual reproductive output, unlike some oth- er raptors (Newton 1979; Newton et al. 1981; Fors- man et al. 1984). For the BWH it might also be advantageous to have a way to optimize reproductive effort in any year to adjust for unpredictable food variability; to move away from being an extreme r-strategist and use clutch adjustment. O’Connor (1977) predicted clutch adjustment would occur when eventual food supply for young was predictable prior to egg laying and likely to remain stable throughout the nestling period. Lack (1966) noted for the Common Swift {Apus apus) that optimal clutch size in one year might not be the optimal clutch size in another year. For our study average BWH clutch size was sig- nificantly higher in 1979 (~3) than 1978 (~2). Diet analysis indicated that food was not as plentiful in 1978 as 1979, and brood reduction only occurred in clutches of 3 and 4, never for 2. We suggest that a lower food supply in 1978 may have, in part, de- termined the lower reproductive effort by triggering use of a clutch adjustment strategy. Brood reduction would be the more significant and functioning strat- egy in years of initially higher food supplies. Al- though our study is insufficient to confirm this gen- eralization, the BWH appears to be comparatively an r-selectionist, capable of optimizing its repro- ductive effort in any given year by employing the brood reduction strategy, but with occasional less- ening of an r- selection pattern by resort to clutch adjustment. Additional long-term studies of repro- duction and prey availability could determine if BWH or other raptors regularly resort to such complex strategies, and the terms r- and k-selection may prove inadequate to describe some breeding patterns. Acknowledgments The authors would like to thank the following people who assisted with field work in this project: Bob Horan, Dave Dick, Charlie McVay, Betty Chapman, Jeff Broder, Ralph Pagano, Mike Murphy and Mike Carlton. Typing was done by Donna Lasher and figures were produced by Katherine Barnes. The Sigma Xi Scientific Research So- ciety and the Environmental Resources Center of State University College at Fredonia assisted in funding this study. Helpful comments on an earlier draft were provided by Jim Bednarz, Kim Titus and Jimmie Parrish. The senior author would like to thank “Uncle Berny” Petrusky for long ago interesting a young boy in the wonder of nature and wildlife. The senior author would especially like to thank his wife Janie. She assisted often in the field, occasionally did some typing and encouraged him, but more importantly, she sacrificed a more normal marriage to allow him to follow a dream. Literature Cited Anonymous. 1984. Status reports: state endangered and threatened raptor species. The Eyas 7(3):17-20. Apanius, Victor. 1977. Red-shouldered Hawks in ju- venile plumage nest successfully. Auk 94:585. Armstrong, Edward and David Euler. 1982. Hab- itat usage of two woodland Buteo species in central Ontario. Can. Field-Nat. 97:200-207. Bednarz, James C. 1979. I. Productivity, nest sites and habitat of Red-shouldered and Red-tailed Hawks in Iowa and II. Status, habitat utilization and manage- ment of Red-shouldered Hawks in Iowa. M.S. Thesis, Iowa State University, Ames, IA. and James J. Dinsmore. 1981. Status, habitat use and management of Red-shouldered Hawks in Iowa. J. Wildl. Manage. 45:236-241. AND . 1982. Nest sites and habitat of Red-shouldered and Red-tailed Hawks in Iowa. Wil- son Bull. 94:31-45. Bent, A. C. 1937. Life histories of North American birds of prey. Volume 1. Reprint, Dover Publications, New York, NY. Brown, Charles R. 1978. Clutch size and reproductive success of adult and subadult purple martins. South- west. Nat. 23:597-604. Brown, Leslie H. 1966. Observations on some Kenya eagles. Ibis 108:531-572. and Dean Amadon. 1968. Eagles, hawks and falcons of the world. McGraw-Hill, New York. Brown, W. H. 1971. Winter population trends in the Red-shouldered Hawk. Amer. Birds 25:813-817. Burns, Frank L. 1911. A monograph of the Broad- winged Hawk ( Buteo platypterus) . Wilson Bull. 23: 139-320. 138 Scott T. Crocoll and James W. Parker Vol. 23, No. 4 Craighead, John J. and Frank C. Craighead, Jr 1956. Hawks, owls and wildlife. The Stackpole Co., Harrisburg, PA and Wildl. Manage. Institute, Wash- ington, DC. Crawford, Richard D. 1977, Breeding biology of year- old and older female Red- winged and Yellow-headed Blackbirds. Wilson Bull. 89:73-80. Crocoll, Scott T. and James W. Parker. 1 986. Can- adaway Raptors. The Conservationist 41:36-41. AND . 1988. The life and times of the Red-shouldered Hawk. The Conservationist 43:41-45. Fisher, Harvey I. 1975. The relationship between de- ferred breeding and mortality in the Laysan Albatross. Auk 92:433-441. Forsman, E. D., E. C. Meslow and H. M. Wight. 1984. Distribution and biology of the Spotted Owl. Wildl. Monogr. 87. 64 pp. Fuller, Mark R. 1979. Spatiotemporal ecology of four sympatric raptor species. Ph.D. Dissertation, Univer- sity of Minnesota, Minneapolis, Minnesota. Gates, J. Edward and Leslie W. Gysel. 1978. Avian nest dispersion and fledging success in field-forest eco- tones. Ecology 59:871-883. Henny, C. J., F. C. Schmid, E. L. Martin and L. L. Hood. 1973. Territorial behavior, pesticides, and the population ecology of Red-shouldered Hawks in cen- tral Maryland, 1943-1971. Ecology 54:545-554. Hirshfield, Michael F. and Donald W. Tinkle. 1975. Natural selection and the evolution of repro- ductive effort. Proc. Nat. Acad. Sci. U.S.A. 72:2227- 2231. Howe, Henry F. 1976. Egg size, hatching asynchrony, sex and brood reduction in the Common Grackle. Ecol- ogy 57:1195-1207. Howell, J., B. Smith, J. B. Holt, Jr. and D. R. Osborne. 1978. Habitat structure and productivity in Red-tailed Hawks. Bird-Banding 49:162-171. Janik, Cynthia A. and James A. Mosher. 1982. Breeding biology of raptors in the central Appala- chians. Raptor Res. 16:18-24. Reran, Doug C. 1976. Nest site selection by the Broad- winged Hawk in north central Minnesota and Wis- consin, 1971-1974. M.S. Thesis, St. Cloud State Uni- versity, St. Cloud, MN. . 1978. Nest site selection by the Broad-winged Hawk in north central Minnesota and Wisconsin. Rap- tor Res. 12:15-20. Kimmel, Vicki L. and Leigh H. Fredrickson. 1981. Nesting ecology of the Red-shouldered Hawk in south- eastern Missouri. Trans. Missouri Acad. Sci. 15:21-27. Lack, David. 1966. Population studies of birds. Clar- endon Press, Oxford. . 1968. Ecological adaptations for breeding in birds. Methuen, London. Luttich, S. N., L. B. Keith and J. D. Stephenson. 1971. Population dynamics of the Red-tailed Hawk {Buteo jamaicensis) at Rochester, Alberta. Auk 88:73- 87. Lyons, David M. and James A. Mosher. 1987. Mor- phological growth, behavioral development, and pa- rental care of Broad-winged Hawks. /. Field Ornithol 58:334-344. Mac Arthur, Robert H. and Edward O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, NJ. Mason, Timothy. 1980. The Red-shouldered Hawk of the Yellow River. Iowa Bird Life 50:68-70. Matray, Paul F. 1974. Broad-winged Hawk nesting and ecology. Auk 91:307-324. McInvaille, William B. and Lloyd B. Keith. 1974. Predator-prey relations and breeding biology of the Great Horned Owl and Red-tailed Hawk in central Alberta. Can. Field-Nat. 88:1-20. Morris, Michael M. J., B. L. Penak, R. E. Lemon and D. M. Bird. 1982. Characteristics of Red-shoul- dered Hawk {Buteo lineatus ) nest sites in southwestern Quebec. Can. Field-Nat. 96:139-142. Newton, Ian. 1976, Breeding of Sparrowhawks, (Ac- cipiter nisus) in different environments. J. Anim. Ecol 45:831-849. . 1977. Breeding strategies in birds of prey. Liv- ing Bird 16:51-82. . 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. . 1986. The Sparrowhawk. T. and A. D. Poyser, Calton, England. O’Connor, Raymond J. 1977. Growth strategies in nestling passerines. Living Bird 16:209-238. Parker, James W. 1972. A mirror and pole device for examining high nests. Bird-Banding 43:216-218. . 1974. The breeding biology of the Mississippi Kite in the Great Plains. Unpublished Ph.D. Disser- tation, Univ. Kansas, Lawrence. Portnoy, John W. and Wendell E. Dodge. 1979 Red-shouldered Hawk nesting ecology and behavior Wilson Bull. 91:104-117. Preston, C. R., C. S. Harger and H. E. Harger. 1989 Habitat use and nest-site selection by Red-shouldered Hawks in Arkansas. Southwest. Nat. 34:72-78. Rosenfield, Robert N. 1 984. Nesting biology of Broad- winged Hawks in Wisconsin. Raptor Res. 18:6-9. Rusch, Donald H. and Phillip D. Doerr. 1972 Broad-winged Hawk nesting and food habits. Auk 89. 139-145. Sharp, Mirek J. and Craig A. Campbell. 1982 Breeding ecology and status of Red-shouldered Hawks {Buteo l. lineatus ) in Waterloo Region. Ont. Field Biol 36:1-10. Shelford, Victor. 1963. The ecology of North Amer- ica. University of Illinois Press, Urbana, IL. Sokal, Robert R. and F. James Rohlf. 1969. Biom- etry. W. H. Freeman and Company, San Francisco. Winter 1989 Broad-wing and Red-shoulder Biology 139 Stewart, Robert E. 1949. Ecology of a nesting Red- shouldered Hawk population. Wilson Bull. 61:26-35. Titus, Kimberly and James A. Mosher. 1981. Nest- site habitat selected by woodland hawks in the central Appalachians. Auk 98:270-281. AND . 1987. Selection of nest tree species by Red-shouldered and Broad-winged Hawks in two temperate forest regions. /. Field Ornithol. 58:274-283. U S D. I., Bur. Land Manage. 1979. Snake River Birds of Prey Special Research Report to the Secretary of the Interior. Boise District, ID. Weeks, Harmon P., Jr. 1978. Clutch size variation in the Eastern Phoebe in southern Indiana. Auk 95:656- 666 . Wiley, James W. 1975. The nesting and reproductive success of Red-tailed Hawks and Red-shouldered Hawks in Orange County, California, 1973. Condor 77:133-139. Department of Biology, State University College, Fre- donia, NY 14063. Present address of first author: Habitat Inventory Unit, New York State Depart- ment of Environmental Conservation, 700 Troy- Schenectady Road, Latham, NY 12110. Present ad- dress of second author: Aerie East, Rt. 3, Box 3110, Farmington, ME 04938. Received 14 September 1988; accepted 15 December 1989 J Raptor Res. 23(4): 140-1 46 © 1989 The Raptor Research Foundation, Inc. ENVIRONMENTAL CONTAMINANTS IN BLOOD OF WESTERN BALD EAGLES Stanley N. Wiemeyer, Richard W. Frenzel, Robert G. Anthony, B. Riley McClelland and Richard L. Knight Abstract. — Blood samples collected in 1979-81 from wintering Bald Eagles ( Hahaeetus leucocephalus ) in Oregon and northern California, residents in Oregon, migrants in Montana and residents in Washington were analyzed for lead (Pb), mercury (Hg) and organochlorines. Lead was detected infrequently (5%) and at low concentrations (<0.25 ppm) in nestlings from Oregon, more frequently (41%) and at occa- sionally elevated concentrations (>0.40 ppm) in wintering Bald Eagles in Oregon and northern California and migrants in Montana, and most frequently (56%) in nestlings from Washington but at low concen- trations (<0.40 ppm). Mercury concentrations were low (<0.70 ppm) in samples from Washington nestlings and higher in samples from Oregon and northern California birds and in Montana migrants. Adults tended to have higher concentrations of Hg than hatch year birds or nestlings. Two Bald Eagles from Montana had clearly elevated Hg concentrations (7.0 and 9.5 ppm). DDE and polychlorinated biphenyl (PCB) concentrations were generally low (most means <0.20 ppm) with adults having higher concentrations than subadults or nestlings. A few resident adult Bald Eagles from Oregon had elevated concentrations of DDE. Bald Eagle ( Hahaeetus leucocephalus) populations in the United States and Canada have been adversely impacted by environmental contaminants. DDE has been strongly implicated in reduced reproductive success (Grier 1982' Wiemeyer et al. 1984a), and birds have died of dieldrin and endrin poisoning (Kaiser et al. 1980; Reichel et al. 1984). Bald Eagles also have died of lead (Pb) poisoning (Kaiser et al. 1980; Reichel et al. 1984), primarily caused by in- gesting Pb shot from hunter-crippled and killed wa- terfowl (Pattee and Hennes 1983). Although some Bald Eagles have been exposed to mercury (Hg) contamination, most populations do not appear to have been affected (Belisle et al. 1972; Wiemeyer et al. 1984a). Environmental contaminants in Bald Eagle pop- ulations have been monitored through analysis of tissues of birds found dead (Reichel et al. 1984), eggs (Wiemeyer et al. 1984a) and blood, including plasma (Henny et al. 1981; Pattee and Hennes 1983). We collected blood samples for contaminant analyses from Bald Eagles in Oregon, northern California, Mon- tana and Washington. Our objectives were to deter- mine contaminant concentrations in the birds, to compare concentrations among areas, age classes and residency status, and relate concentrations to sources of exposure and possible effects on populations. Methods Sample Collection. Blood samples (6-12 cc) were col- lected in 1979-81 during the breeding season from resident (including 7-1 1-wk-old nestlings and adults) and subadult Bald Eagles from the Klamath Basin and Cascade Lakes areas of Oregon, and from wintering Bald Eagles from the Klamath Basin, Oregon and northern California (Frenzel 1985). Blood samples were taken with heparin- ized glass syringes (washed with detergent and rinsed with residue grade acetone) then frozen. Blood samples (ca. 10 cc) were collected from migrant Bald Eagles in Glacier National Park (GNP), Montana (McClelland et al. 1982) during October-December 1980 and October-November 1981 using disposable syringes. Samples collected in 1980 were preserved with formalin (1 part/20 parts blood; Wiemeyer et al. 1984b); those collected in 1981 were preserved by freezing. Some birds were equipped with patagial markers and radio trans- mitters to monitor movements (Young 1983). Blood samples (6 cc) were collected from 7-9-wk-old nestling Bald Eagles in San Juan Island County, Wash- ington during June 1980 with heparinized disposable sy- ringes. Samples were preserved with formalin. A sample of formaldehyde from the lot used in preserving blood samples was also submitted for chemical analysis. Blood samples (5 cc) were collected from 5 captive Bald Eagles at the Patuxent Wildlife Research Center, Laurel, Maryland on 19 October 1983 for comparative purposes, using disposable syringes. Birds had been in captivity for 1-13 yrs. Samples were frozen. Two unused syringes from the same lot used in collecting blood samples were also submitted for chemical analysis. Blood samples were placed in glass jars that had been cleaned with nitric acid and rinsed with deionized water, acetone and hexane, and capped with lids equipped with teflon liners. Bald Eagles, except for nestlings, were aged on the basis of plumage characteristics. Eagles with white heads were classified as adults. In Oregon and California, all eagles lacking a white head were classified as subadults. In Mon- tana, nonadult eagles were classified as hatch year birds or as subadults. Birds having the “dark immature” plum- 140 Winter 1989 Contaminants in Bald Eagles 141 age (with dark brown eye and completely dark brown or black bill) as described by Clark (1983) were classified as hatch year; older nonadult birds (with varying degrees of lighter plumage, eye and bill coloration) were classified as subadults. Ages of nestlings were estimated by develop- ment of plumage (Stalmaster 1987), size and nesting chro- nology. Chemical Analysis. Samples were homogenized and subsampled for various analyses. A 0.5 g aliquot was used for Pb analysis, 2.0 g for Hg and 5.0 g for organochlorines. Organochlorines were analyzed by the methods of Cro- martie et al. (1975) and Kaiser et al. (1980), except that fractions I and II were combined. Glassware was rinsed with 1 5% ethyl ether in hexane prior to use. Contaminants in each fraction were identified and quantitated by electron capture gas chromatography using a 1.5/1.95% SP-2250/ SP-2401 packed column. The lower limit of reportable residues was 0.01 ppm for pesticides and 0.05 ppm for polychlorinated biphenyls (PCBs) in Oregon and northern California samples, and 0.05 ppm for pesticides and 0.10 ppm for PCBs in Montana samples. A gas chromatograph- mass spectrometer was used to confirm the identity of contaminants in about 10% of samples containing detect- able concentrations. Samples were analyzed for Hg by cold vapor atomic absorption spectrophotometry using previously described methods (Monk 1961; Hatch and Ott 1968). The lower limit of reportable residues was 0.02 ppm. Lead was ana- lyzed by graphite furnace atomic absorption spectropho- tometry using a wet ash procedure (Hinderberger et al. 1981). The lower limit of reportable residues was 0.05 ppm. Frozen duck blood samples stored for 2 and 8 mo lost 35% of their DDE concentrations compared to those ana- lyzed fresh, whereas no DDE was lost from samples pre- served with formalin and stored for the same periods (Wie- meyer et al. 1984b). Data on the stability of PCBs under these preservation methods are not available. Freezing and formalin preservation of blood were equally suitable when dealing with Pb and Hg residues (Wiemeyer et al. 1984b). Only samples from Montana in 1980 and from Wash- ington nestlings in this study were preserved with for- malin. Samples from Oregon and California were stored frozen for 8-24 mo before chemical analysis. Samples from Mon- tana that were collected in 1980 were analyzed for organo- chlorines in August 1981 and for Pb and Hg in February 1983. Montana samples collected in 1981 were analyzed for organochlorines and metals in July and November 1982, respectively. Samples from Washington were ana- lyzed in September 1982. Samples from captive birds were analyzed within 3 wk of collection. Lead or Hg was not detected in a sample of formalin from the lot used to preserve samples from Washington. Mercury was not detected in solutions used to rinse sy- ringes from the same lot as those used in collecting blood samples from captive Bald Eagles. Statistical Analysis. Geometric means are reported throughout when >50% of samples contained a detectable concentration of a given contaminant. Samples containing nondetectable residues were assigned values equal to one- half the detection limit in computing means. T-tests were used to determine if significant differences occurred be- tween means of log transformed values. Results Lead was detected infrequently (5%) and at low concentrations in nestling Bald Eagles from Oregon (Table 1), but more frequently in Washington nest- lings (x 2 ; P < 0.001). Forty-one percent of samples from wintering Bald Eagles in Klamath Basin and also in migrants from Montana had detectable Pb concentrations. Only 3 samples from all areas con- tained >0.4 ppm Pb, 2 migrants from Montana and 1 wintering subadult from Klamath Basin. All but 1 sample contained detectable Hg con- centrations (Table 1). Concentrations ranged widely, with the lowest mean concentration found in nest- lings from W ashington and the highest in subadults and adults from Oregon and northern California. Oregon nestlings had significantly higher (P < 0.0001) concentrations than Washington nestlings. Subadults and adults tended to have higher Hg con- centrations than nestlings or hatch year birds. Con- centrations in Oregon resident adults were signifi- cantly higher (P = 0.019) than in Oregon nestlings. Blood samples from 5 captive Bald Eagles contained a mean of 0.23 ppm Hg (range 0.17-0.31 ppm). The primary organochlorines detected were DDE and PCBs (Table 2); however, DDD, trans- non- achlor and cfi-nonachlor were detected in a few sam- ples from Oregon (Frenzel 1985). DDE was de- tected in 61% of samples from Oregon nestlings and 100% of samples from subadults and resident adults. DDE was detected in 95% of samples from subadults and adults wintering in Oregon and northern Cal- ifornia. Polychlorinated biphenyls w r ere detected much less frequently than DDE in samples from Oregon and northern California; only 15% of sam- ples from nestlings and 59% of samples from sub- adults and adults contained detectable concentra- tions. Samples from resident adults in Oregon contained significantly (P < 0.0001) higher DDE and PCB concentrations than did nestlings from the same area. Resident adults from Oregon had sig- nificantly (P < 0.0001) higher DDE and PCB con- centrations than those of wintering adults from the Klamath Basin. Few samples from migrant Bald Eagles from Montana contained detectable DDE and PCB concentrations, although the low^er limits of reportable residues were higher than for Oregon and northern California samples. DDE was detected most frequently in adults. 142 WlEMEYER ET AL. VOL. 23, NO. 4 Table 1. Frequency of occurrence and concentrations (ppm wet weight) of lead (Pb) and mercury (Hg) in blood samples from western Bald Eagles. Number Pb Hg State, Status and Age Sampled With Detect- able 3 Geo- metric Mean Range Number SAMPLED b Geometric Mean Range Oregon Resident Nestling 58 3 nd c -0.22 82 d 1.2 nd-4.2 Subadult e 2 0 - — 2 3.0 2. 8-3.2 Adult 5 1 — nd-0.25 7 2.3 i oo Oregon and northern California Wintering f Subadult 4 3 0.129 nd-0.62 5 2.2 1. 6-2.7 Adult 13 4 — nd-0.25 15 2.3 1. 1-5.4 Montana Migrant Hatch year 12 4 — nd-0.23 12 1.5 0.94-3.2 Subadult 14 7 0.072 nd-1.9 14 1.8 0.89-9.5 Adult 3 1 — nd-1.9 3 - 2.0 0.85-4.5 Washington Resident Nestling 9 5 0.066 nd-0.36 9 0.23 0.075-0.65 a Lower limit of reportable residues 0.05 ppm. b All samples contained Hg except as noted in footnote d. c nd = none detected. d One sample contained no detectable Hg. e These birds were probably nomadic and not true residents even though they were sampled during the breeding season. r Data from Frenzel and Anthony (1989). Discussion Lead. Only 3 eagles in our study had recent sig- nificant exposure (>0.40 ppm Pb in blood) to Pb. Although a number of additional eagles had detect- able Pb concentrations in blood (evidence of Pb ex- posure), their exposure appeared minimal. Bald Ea- gles that were experimentally dosed with 10 pellets of No. 4 Pb shot had a mean of 0.8 ppm Pb in blood 1 d after dosage and 2.8 ppm 3 d after dosage, whereas unexposed Bald Eagles had no detectable Pb concentrations (<0.1 ppm) in their blood (Hoff- man et al. 1981). Bald Eagles wintering in Klamath Basin fed most- ly on waterfowl (Frenzel and Anthony 1989), whereas fish were predominant in the diet of resident eagles during spring and summer (Frenzel 1985). Detectable Pb concentrations were found more fre- quently in adult and subadult Bald Eagles wintering in Klamath Basin (41%) than in subadults and res- ident adults from Oregon (14%); however, the dif- ference was not significant (x 2 ; P = 0.20). Seasonal shifts in food habits and the possible entry of birds into the wintering population that could have been previously exposed to Pb may explain these differ- ences; however, larger samples would be required to examine this issue. Most Bald Eagle food items from Oregon had low concentrations of Pb in carcass, except for grebes, gulls and Belding’s Ground Squir- rels ( Spermophilus beldingi) (Frenzel 1985). The most likely source of Pb exposure to Bald Eagles is inges- tion of Pb shot from hunter-killed or crippled wa- terfowl (Pattee and Hennes 1983). Migrant Bald Eagles in GNP fed on fish (McClelland et al. 1982). Lead was detected in none Winter 1989 Contaminants in Bald Eagles 143 Table 2. Frequency of occurrence and concentrations (ppm wet weight) of organochlorines in blood samples from western Bald Eagles. State, Status, Age and Collection Year DDE PCB N NO. WITH Detect- able 3 Geo- metric Mean Range NO. WITH Detect- able 15 Geo- metric Mean Range Oregon c Resident Nestling 75 46 0.015 nd d -0.15 11 nd-0.29 Subadult e 3 3 0.12 0.06-0.20 1 — nd-0.08 Adult 8 8 0.50 0.08-1.4 8 0.25 0.05-0.71 Oregon and northern California 0 Wintering f Subadult 5 5 0.030 0.01-0.14 2 nd-0.08 Adult 16 15 0.042 nd-0.13 8 0.018 nd-0.12 Montana Migrant Hatch year 1980s 5 0 0 1981 c 6 0 — — 0 — — Subadult 1980s 5 2 nd-0.06 0 \0 00 n 5 0 — — 0 — — Adult 1980s 7 7 0.19 0.07-0.71 4 0.28 nd-0.71 1981° 3 3 0.086 0.05-0.13 0 — — d Lower limit of reportable residues 0.01 ppm for Oregon and northern California samples and 0.05 ppm for Montana samples. b Lower limit of reportable residues 0.05 ppm for Oregon and northern California samples and 0.10 ppm for Montana samples. ‘ Samples preserved by freezing. d nd = none detected. c These birds were probably nomadic and not true residents even though they were sampled during the breeding season. r Data from Frenzel and Anthony (1989). s Samples preserved with formalin. of 8 samples collected on or before 25 October, whereas 12 of 21 (57%) samples collected after that date had detectable concentrations (x 2 ; P = 0.005). Samples from subadults and adults (N = 2 and 0, respectively) were poorly represented in the earlier time period. For hatch year birds only, none of 6 samples collected on or before 25 October had de- tectable Pb concentrations, whereas 4 of 6 collected later had detectable concentrations (x 2 ; P < 0.025). Exposure to Pb shot would be expected to increase after the start of the waterfowl hunting season and may have occurred before arrival in GNP. One adult female Bald Eagle found dead in Klam- ath Basin in 1982 died of Pb poisoning; liver con- tained 27 ppm Pb (Frenzel and Anthony 1989). Lead concentrations in livers of 11 other adult and subadult Bald Eagles from Oregon and northern California that died in 1979-82 were <6 ppm; 7 were <2 ppm (Frenzel 1985). Twenty-two Bald Eagles dying in Oregon, Montana, and Washington were necropsied during 1978-81; none died of Pb poisoning (Reichel et al. 1984). Bald Eagles dosed with Pb shot all had >10 ppm Pb in liver at death (Pattee et al. 1981). Although the exposure of Bald Eagles to Pb in our study areas generally appeared low, even some nestlings were exposed. A few migrant and wintering Bald Eagles were at risk from Pb poisoning. Ad- ditional information relating known Pb concentra- tions in blood of Bald Eagles to effects on health and risk of poisoning would be helpful in interpreting data from field studies. The eventual ban on use of 144 WlEMEYER ET AL. VOL. 23, No. 4 Pb shot in waterfowl hunting in the 1991-92 hunt- ing season (U.S. Fish and Wildlife Service 1986) should greatly reduce the risk of Bald Eagles dying of Pb poisoning. Studies on exposure of wild Bald Eagles to Pb following the ban should be conducted to determine impact on populations. Mercury. Bald Eagles appear to routinely have higher concentrations of Hg in their blood than other species of birds. For example, nearly all untreated Mallards ( Anas platyrhynchos ) had <0.07 ppm in blood (Heinz 1980), whereas wild Rock Doves ( Co - lumba livid) from Mississippi had 0.005-0.012 ppm Hg in blood (Knight and Harvey 1974). Nestling Bald Eagles from Washington and captive Bald Ea- gles both had means of 0.23 ppm Hg in blood. Young Common Terns ( Sterna hirundo ) from Long Island, New York that may have been exposed to minor Hg contamination and were classified as normal with regard to feather development had 0.37 fi g/ml Hg in blood (Gochfeld 1980). Mallards fed 0.5 ppm (dry weight) methylmercury for 7 mo had blood levels of 0.5 to 0.6 ppm Hg (Heinz 1980). Wild Bald Eagles, except for Washington nest- lings, in our study had far higher Hg concentrations in blood. Fifteen eagles had >3 ppm Hg in blood (1 1 from Oregon or northern California and 4 from Montana). Two Montana eagles, both subadults, had >6 ppm in blood (7.0 and 9.5 ppm). Higher concentrations in blood samples of adults than in those from nestlings and hatch year birds correspond with the known accumulative nature of Hg. Although Bald Eagle blood samples from Oregon and northern California had higher concentrations than other species or captive Bald Eagles, there is no evidence that Hg was having an adverse impact on the population. Reproductive success appeared normal (Frenzel 1985). Eight clutches of Oregon Bald Eagle eggs from the Klamath Basin and Cas- cade Lakes regions collected during 1979-81 con- tained <0.26 ppm Hg (clutch means) (Frenzel 1985). Eight of 9 Bald Eagles found dead in the same area during 1979-82 had <3 ppm Hg in liver, with 1 bird having 8 ppm (Frenzel 1985). These concen- trations are far below those associated with effects on reproduction (0.5 to 1.5 ppm in eggs; Wiemeyer et al. 1984a) or survival (>20 ppm in liver; Finley et al. 1979). Mercury residues in prey of Oregon Bald Eagles also tended to be low (Frenzel 1985). Some Bald Eagles migrating through GNP may have been exposed to Hg contamination in Canada. Transmitter-equipped Bald Eagles migrating through GNP were tracked to summering areas in Northwest Territories (NWT) and northeastern Al- berta, Canada (Young 1983). Elevated Hg concen- trations in fish have been reported for several Can- adian areas within the migratory corridor of some Bald Eagles passing through GNP, to include: Giauque and Thompson Lakes, 100-125 km north of Great Slave Lake, NWT (Moore and Sutherland 1980); the North Saskatchewan River, near Ed- monton, Alberta (Ramamoorthy et al. 1985); and some lakes in northern Saskatchewan, especially Cumberland Lake on the Saskatchewan River (Murray 1978). A Bald Eagle found dead in Mon- tana south of GNP in March 1985 had 35 ppm (wet weight) Hg in liver, a concentration suggesting Hg poisoning. The western states and British Columbia lie in a mercuriferous belt where underlying rock contains elevated Fig levels (Jonasson and Boyle 1971). This naturally occurring source of Hg in the region may contribute to Hg in food chains and subsequently in Bald Eagles. However, this source should contribute little to Hg found in hatch year Bald Eagles mi- grating through GNP, their origin in Canada being outside the mercuriferous belt. Mercury residues in blood of hatch year Bald Eagles sampled in GNP were not correlated (P > 0.05) with date of collec- tion. Lack of information relating known exposures of Hg to concentrations in blood and effects on health of Bald Eagles prevents adequate interpretation of our data. Therefore, risk cannot be assessed. The presence of highly elevated concentrations in a few birds is cause for concern. Data on sources of ex- posure could lead to control thus reducing risk to Bald Eagles. Organochlorines. Direct comparisons of DDE concentrations between samples that were preserved by freezing and those that were preserved with for- malin should be conducted with caution (see Meth- ods). Organochlorine concentrations in plasma or se- rum of birds have been significantly correlated with concentrations in tissues, making possible the mon- itoring of contamination in wild populations without sacrifice of birds (Capen and Leiker 1979; Friend et al. 1979; Henny and Meeker 1981). The nature of the relationship between organochlorine concen- trations in whole blood and that in carcass or other tissues is unknown for Bald Eagles. Henny and Meeker (1981) predicted DDE burdens in eggs based Winter 1989 Contaminants in Bald Eagles 145 on residues in plasma of laying females for American Kestrels ( Falco sparuerius) and accipiters. Applica- tion of predictive equations to our data is question- able because of differences in residue concentrations between plasma and whole blood and possible species differences. Plasma should contain about twice the concentration in whole blood. Relative DDE con- centrations in blood from resident adult Bald Eagles from Oregon were similar to the relative DDE con- centrations in eggs (Frenzel 1985). Concentrations of DDE in eggs from the southern Oregon popu- lation were high enough to be associated with egg- shell thinning and reproductive failure for a few breeding pairs (Wiemeyer et al. 1984a; Frenzel 1985). A few resident adult Bald Eagles from Or- egon had clearly elevated DDE concentrations in blood. Henny et al. (1981) found low 2 DDT (al- most all DDE) concentrations (x = 0.06-0.14 ppm) in plasma of Bald Eagles wintering in Colorado and Missouri in 1977-78. Organochlorine concentra- tions in the environment have declined following bans and restrictions on usage which should result in continuation of reduced risks to Bald Eagles. Acknowledgments We thank M. L. Gay for analytical methods and A. J. Krynitsky, J. F. Moore and B. M. Mulhern for analysis of samples. P. T. McClelland, L. S. Young, J. C. Cren- shaw, H, A. Allen and R. H. Fevold assisted with blood sample collection in GNP. D. K. Garcelon and T. G. Grubb assisted with field work in Washington. J. D. An- derson, D. K. Edwards, F. B. Isaacs and G. S. Miller assisted with field work in Oregon and northern Califor- nia. O. H. Pattee, C. P. Rice and C. J. Henny provided helpful comments on earlier drafts of the manuscript. Literature Cited Belisle, A. A., W. L. Reichel, L. N. Locke, T. G. Lamont, B. M. Mulhern, R. M. Prouty, R. B. DeWolf AND E. Cromartie. 1972. Residues of or- ganochlorine pesticides, polychlorinated biphenyls, and mercury and autopsy data for Bald Eagles, 1969 and 1970. Pestic. Monit. J. 6:133-138. Capen, D. E. and T. J. Leiker. 1979. DDE residues in blood and other tissues of White-faced Ibis. Environ. Pollut. 19:163-171. Clark, W. S. 1983. The field identification of North American eagles. American Birds 37:822-826. Cromartie, E., W. L. Reichel, L. N. Locke, A. A. Belisle, T. E. Kaiser, T. G. Lamont, B. M. Mulhern, R. M. Prouty and D. M. Swineford. 1975. Residues of organochlorine pesticides and poly- chlorinated biphenyls and autopsy data for Bald Ea- gles, 1971-72. Pestic. Monit. J. 9:11-14. Finley, M. T., W. H. Stickel and R. E. Christensen. 1979. Mercury residues in tissues of dead and sur- viving birds fed methylmercury. Bull. Environ. Contam Toxicol. 21:105-110. Frenzel, R. W. 1985. Environmental contaminants and ecology of Bald Eagles in southcentral Oregon. Ph.D Thesis, Oregon State Univ., Corvallis, OR. 143 pp. and R. G. Anthony. 1989. Relationships of diets and environmental contaminants in wintering Bald Eagles. J. Wildl. Manage. 53:792-802. Friend, M., M. A. Haegle, D. L. Meeker, R. Hudson and C. H. Baer. 1979. Correlations between resi- dues of dichlorodiphenylethane, polychlorinated bi- phenyl, and dieldrin in the serum and tissues of Mal- lard ducks ( Anas platyrhynchos). Pages 319-326. In Animals as monitors of environmental pollutants. Na- tional Academy of Sciences, Washington, DC. Gochfeld, M. 1980. Tissue distribution of mercury in normal and abnormal young Common Terns. Marine Pollut. Bull. 11:362-377. Grier, J. W. 1982. Ban of DDT and subsequent re- covery of reproduction in Bald Eagles. Science 218 1232-1235. Hatch, W. R. and W. L. Ott. 1968. Determination of sub-microgram quantities of mercury by atomic ab- sorption spectrophotometry. Anal. Chem. 40:2085-2087 Heinz, G. H. 1980. Comparison of game-farm and wild- strain Mallard ducks in accumulation of methylmer- cury. J. Environ. Pathol. Toxicol. 3:379-386. Henny, C. J. and D. L. Meeker. 1981, An evaluation of blood plasma for monitoring DDE in birds of prey Environ. Pollut. 25A:29 1-304. , C. R. Griffin, D. W. Stahlecker, A. R. Har- mata and E. Cromartie. 1981. Low DDT residues in plasma of Bald Eagles ( Haliaeetus leucocephalus ) wintering in Colorado and Missouri. Canad. Field-Nat 95:249-252. Hinderberger, E. J., M. L. Kaiser and S. R. Koirtyohann. 1981. Furnace atomic absorption analysis of biological samples using the l’vov platform and matrix modification. Atomic Spectroscopy 2:1-7. Hoffman, D. J., O. H. Pattee, S. N. Wiemeyer and B. M. Mulhern. 1981. Effects of lead shot ingestion on 5-aminolevulinic acid dehydratase activity, hemo- globin concentration, and serum chemistry in Bald Ea- gles. /. Wildl. Dis. 17:423-431. Jonasson, I. R. and R. W. Boyle. 1971. Geochemistry of mercury. Pages 5-21. In Mercury in man’s envi- ronment. The Royal Society of Canada Symposium, Ottawa, Canada. Kaiser, T. E., W. L. Reichel, L. N. Locke, E. Cro- martie, A. J. Krynitsky, T. G. Lamont, B. M Mulhern, R. M. Prouty, C. J. Stafford and D M. Swineford. 1980. Organochlorine pesticide, PC B, and PBB residues and necropsy data for Bald Eagles from 29 states — 1975-77. Pestic. Monit. J. 13:145-149 Knight, L. A., Jr. and E. J. Harvey, Sr. 1974. Mer- 146 WlEMEYER ET AL. Vol. 23, No. 4 cury residues in the Common Pigeon ( Columba livia) from the Jackson, Mississippi, area — 1972. Pestic. Momt. J. 8:102-104. McClelland, B. R., L. S. Young, D. S. Shea, P. T. McClelland, H. L. Allen and E. B. Spettigue. 1982. The Bald Eagle concentration in Glacier Na- tional Park, Montana: origin, growth, and variation in numbers. Living Bird 19:133-155. Monk, H. E. 1961. Recommended methods of analysis of pesticide residues in food stuffs. Report by the Joint Mercury Residues Panel. Analyst 82:608-614. Moore, J. W. and D. J. Sutherland. 1980. Mercury concentrations in fish inhabiting two polluted lakes in northern Canada. Water Res. 14:903-907. Murray, A. R. 1978. An analysis of mercury contam- ination in fish from northern Saskatchewan lakes, 1969- 76. Department of Northern Saskatchewan, Fisheries Technical Report No. 2, La Ronge, Saskatchewan. 138 pp. Pattee, O. H. and S. K. Hennes. 1983. Bald Eagles and waterfowl: the lead shot connection. Trans. No. Amer. Wildl. Natr. Resources Conf. 48:230-237. , S. N. WlEMEYER, B. M. Mulhern, L. Sileo and J. W. Carpenter. 1981. Experimental lead- shot poisoning in Bald Eagles. J. Wildl. Manage. 45: 806-810. Ramamoorthy, S., J. W. Moore and L. George. 1985. Partitioning of mercury in the North Saskatchewan River. Chemosphere 14:1455-1468. Reichel, W. L., S. K. Schmeling, E. Cromartie, T. E. Kaiser, A. J. Krynitsky, T. G. Lamont, B. M. Mulhern, R. M. Prouty, C. J. Stafford and D. M. Swineford. 1984. Pesticide, PCB, and lead res- idues and necropsy data for Bald Eagles from 32 states — 1978-81. Environ. Monit. Assess. 4:395-403. Stalmaster, M. V. 1987. The Bald Eagle. Universe Books, New York, NY. 227 pp. U.S. Fish and Wildlife Service. 1 986. Migratory bird hunting; criteria and schedule for implementing non- toxic shot zones for 1987-88 and subsequent waterfowl hunting seasons. Fed. Register 51:23444-23447. Wiemeyer, S. N., T. G. Lamont, C. M. Bunck, C. R Sindelar, F. J. Gramlich, J. D. Fraser and M. A Byrd. 1984a. Organochlorine pesticide, polychlo- robiphenyl, and mercury residues in Bald Eagle eggs — 1969-79 — and their relationships to shell thinning and reproduction. Arch. Environ. Contam. Toxicol. 13:529- 549. , J. F. Moore and B. M. Mulhern. 1984b. Formalin preservation of avian blood for metal and DDE analysis. Bull. Environ. Contam. Toxicol. 33:525- 532. Young, L. S. 1983. Movements of Bald Eagles associ- ated with autumn concentrations in Glacier National Park. M.S. Thesis. Univ. of Montana, Missoula. 102 pp. U.S. Fish and Wildlife Service, Patuxent Wildlife Re- search Center, Laurel, MD 20708. Address of sec- ond and third authors: U.S. Fish and Wildlife Ser- vice, Oregon Cooperative Wildlife Research Unit, Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR 97331. Present address of second author: 51 N. Dean St., Coquille, OR 97423. Address of fourth author: School of Forestry, University of Montana, Missoula, MT 59812. Ad- dress of fifth author: State of Washington, Depart- ment of Game, 600 N. Capitol Way, Olympia, WA 98504. Present address of fifth author: Department of Fishery and Wildlife Biology, Colorado State University, Fort Collins, CO 80523. Received 20 December 1988; accepted 15 August 1989 / Raptor Res. 23(4): 1 47-1 5 1 © 1989 The Raptor Research Foundation, Inc. HAZARDS TO RAPTORS FROM STRYCHNINE POISONED GROUND SQUIRRELS Josef K. Schmutz, Karrie A. Rose and Robert G. Johnson Abstract. — To evaluate the potential for secondary poisoning of raptors from poisoning campaigns for ground squirrels, we placed eviscerated, strychnine-poisoned Richardson’s Ground Squirrels ( Spermophilus richardsonii ) into Swainson’s Hawk ( Buteo swainsont) nests and monitored nestling survival. Of 52 nestlings which consumed 67 poisoned ground squirrels, 49 nestlings survived the treatment period. We could not detect a difference in growth rates between treated and control nestlings, nor a difference in survival after the treatment period. We present evidence to suggest that raptors eviscerate ground squirrels prior to consumption. This evisceration probably has a profound influence on the survival of raptors scavenging poisoned ground squirrels. Use of poisons to control unwanted wildlife poses potential threats to predators and other non-target species (Mendenhall and Pank 1980; Townsend et al. 1981; Holler and Schafer 1982; Merson et al. 1984; Marsh et al. 1987; Hegdal and Colvin 1988). Strychnine, the active ingredient in several forms of ground squirrel poison (e.g., “Gopher Cop”® or “Strychnine Gopher-kill Liquid”®), is highly toxic to birds; LD 50 s for most species tested ranged from 2-25 mg/kg (Hudson et al. 1984). Strychnine has been implicated in the mortality of several species of raptors (Reidinger and Crabtree 1974; Cromartie et al. 1975; Kaiser et al. 1980; Redig et al. 1982; Bortolotti 1984; Reichel et al. 1984). Strychnine is used throughout prairie Canada each year to poison Richardson’s Ground Squirrels ( Spermophilus rich- ardsonii ), and usage will probably continue until alternative, biological methods of control become available (e.g. YVodzicki 1973). In this study we attempted to determine the po- tential threat to young hawks that are fed poisoned ground squirrels. Nestling hawks may be more sub- ject to secondary poisoning because they, unlike adults (e.g., Brett et al. 1976), have less opportunity to reject poisoned prey. Cheney et al. (1987) found that even some adults did not learn to avoid strychnine- laced mice in captivity. We examined the survival and growth rate of nestling Swainson’s Hawks ( Bu- teo swainsoni ) into whose nests we placed poisoned ground squirrels. We selected Swainson’s Hawks for this study because >65% of their prey biomass was comprised of ground squirrels (Schmutz et al. 1980), and because Swainson’s Hawks nested pref- erentially in cultivated areas (Schmutz 1987, 1989) where ground squirrels are poisoned. We also stud- ied the frequency with which poisoned ground squir- rels died above ground and were available to scav- enging hawks. Rather than examining the effect of known amounts of poison on hawks, we simulated field conditions in an attempt to determine the mag- nitude of the threat under current poisoning prac- tices. Study Area and Methods This work was carried out adjacent to a study area near Hanna, Alberta (Schmutz et al. 1980). The area consisted of mixed-grass prairie in a semi-arid climate. The primary land use was grazing of cattle with 14% of the land under dry-land cultivation for cereal production. To evaluate the potential impact of poisoning of ground squirrels on raptors, we offered poisoned ground squirrels to nestling Swainson’s Hawks. Ground squirrels were cap- tured locally using live traps (#102 and #103; Tomahawk Live Trap Co., Tomahawk, Wisconsin). Squirrels were held, one to a trap, in a ventilated room and offered water and approximately 200 g of poison-coated grain imme- diately after capture. “Strychnine Gopher-kill Liquid”® (Sanex Inc., Mississauga, Ontario; Registration No. 15849) containing 2% strychnine was used to coat oats according to the manufacturer’s instructions. Up to 24 hr elapsed until some squirrels died, but most died within 12 hr. After a squirrel’s death, we removed stomach, large in- testine and small intestine, and marked the squirrel with two monel metal tags. One squirrel was deposited into a Swainson’s Hawk nest (treated nest) on each of 3 consec- utive d unless the nestlings had died. In the evening of the third day we visited the nests a second time to determine the fate of deposited prey and of nestlings. We chose a 3-d period assuming that this would approximate the period during which dead squirrels would be available to hawks during actual poisoning campaigns. All prey items found in treated nests were removed during daily visits. Prey items in control nests were not removed nor was prey added. These “control nests” did not represent a true control because of a net addition of food to treated nests Sixty-nine ground squirrels were added to treated nests but only 23 prey items found were removed. Nestlings were weighed during daily visits between 9- 28 July 1988. To account for varying amounts of crop contents at the time of weighing, we subtracted 10% of a nestling’s mass if the crop was judged full and propor- 147 148 SCHMUTZ ET AL. Vol, 23, No. 4 tionately less if partially full. Age of nestlings at the start of the treatment period varied, as reflected in their body mass (J. K. Schmutz 1977). Mean age of 52 nestlings in “treated” nests was 13 d (range 2-26). Mean age of “con- trol” nestlings was 16 d (range 3-25). We compared reproductive performance among 3 groups of nests to examine the impact of poisoned food on nest- lings. Twenty-five treated nests were selected just outside the eastern edge of a study area which was part of a long- term, ecological study of prairie raptors (e.g., Schmutz et al 1980). Treated nests were interspersed with 13 “control nests.” We assumed that the effect of poison on nestlings was independent of the amount of food available to them. We also compared the number of young fledged from treated nests with the success of a large sample of nesting pairs on the adjacent study area. To determine whether Swainson’s Hawks actually scavenge dead squirrels, we placed 3 marked, unpoisoned ground squirrels within 30 m of each of 4 nests. To examine whether placement of poison in or outside a burrow affected whether squirrels die above ground or below, we distributed poison on 9 plots of approximately 0 5 ha each. Poison was placed either inside the burrow (4 plots selected at random) or, contrary to the manufac- turer’s recommendation, atop a mound adjacent to a bur- row entrance (5 plots). We returned the next day and recorded the number of dead ground squirrels outside a burrow. At this time, we plugged 25 burrow entrances with dry vegetation on each of 6 plots (3 plots with poison outside of burrows and 3 within). We recorded the number of plugged burrows that had been opened the following day and compared these proportions in relation to where poison was placed (atop or inside a burrow). We used burrows in the center of a poisoned area to minimize counting dispersing squirrels using the plugged burrows. Results Observations of Raptor Feeding Behavior. Swainson’s Hawks and other raptors may be subject to secondary poisoning if they scavenge dead or take dying squirrels. We did not visit those nests where ground squirrels were placed on the ground fre- quently enough to document the fate of all squirrels. One marked squirrel was taken to the nest by a Swainson’s Hawk. All other marked squirrels had disappeared. On 5 other occasions we have observed Swainson’s Hawks attempt to or actually retrieve road-killed Richardson’s Ground Squirrels. In ad- dition, 2 crushed ground squirrels, evidently road killed, were found in nests during a study of food habits (Schmutz et al. 1980). The impact of secondary poisoning on raptors could be much reduced if raptors rejected viscera. Some of our ground squirrels died only hrs after the consumption of poisoned bait, presumably after suf- ficient poison had been absorbed into the blood stream. Excess poisoned grain was contained in the squirrels’ stomachs. We removed stomach, and large and small intestine as the hawks did. Of 91 whole ground squirrels found in Ferruginous Hawk ( B . regalis) nests and of 20 in Swainson’s Hawk nests, 64 (70%) and 11 (55%) respectively had been evis- cerated. Our additional observations of raptors in- dicate that the viscera are actually rejected. We ob- served a Northern Harrier ( Circus cyaneus), a Ferruginous Hawk and 2 Swainson’s Hawks kill ground squirrels, pull out the viscera in 2-3 strokes and drop the viscera onto the ground. The Ferru- ginous Hawk subsequently offered the ground squir- rel to nearby fledglings without feeding itself. In 7 additional cases we observed a Northern Harrier, a fledgling Ferruginous Hawk, a Red-tailed Hawk (B. jamaicensis), and 4 Sw'ainson’s Hawks (1 fledg- ling) eat a duck, 5 ground squirrels or a mouse, with viscera lying on the ground within 1 m of the carcass. A captive Great Horned Owl ( Bubo virginianus ) re- jected viscera when the entire ground squirrel was offered as food. We found mouse viscera discarded at the entrances into the nests of Burrowing Owls (. Athene cunicularia) at least 8 times. The rejection of viscera may be related to prey size relative to raptor size because we observed a Short-eared Owl (Asio jlammeus) swallowing viscera of a mouse con- sumed in pieces. In 29 cases we were able to examine portions of viscera that were found on the ground at kill sites or had been removed from squirrels found otherwise whole in nests (Fig. 1). In these cases the liver was attached to other rejected parts and may have been discarded accidentally. Of 69 poisoned squirrels placed into treated nests, 6 were seen partially consumed 8-24 hr later. Sixty squirrels had disappeared from the nests and pre- sumably had been consumed. Three squirrels were uneaten. We believe that all poisoned ground squir- rels offered to surviving nestlings (67 squirrels) were eaten. In a previous study in which food was sup- plemented (Schmutz et al. 1980), excess food was reluctantly discarded by the parents if at all. Food began to rot and later dry up in the nests. Hawk Survival. No adult mortality was evident in this study. Two adults were seen after the treat- ment period at all nests. We examined the potential impact of poisoning on nestlings in 3 ways, by (1) recording survival during the 3 treatment days, (2) monitoring survival after the treatment period, and (3) by comparing growth. Three of a total of 58 nestlings may have died from secondary poisoning. Three poisoned ground squirrels had disappeared Winter 1989 Secondary Poisoning of Raptors 149 (0 3 0 “ E o ® 20 - k Q. O *- i OH A E a Tfll irjl u< i Jj »1 • i 1 1 n wr rS Large Small Intestine t Raptor O N. Harrier □ Red-tailed □ Swainson’s □ Ferruginous — 1' Stomach Liver Figure 1. Portions of the gastrointestinal tract removed by various raptors from a duck, a vole, a song- bird and 26 Richardson’s Ground Squirrels. in these nests and had presumably been consumed. One partially eaten squirrel remained. An additional six nestlings died in 2 nests between the first and second visits before actually consuming any of the poisoned ground squirrels. These 2 nests were hence excluded from further analysis. Three of these 6 nestlings were partially consumed, suggesting pre- dation as a possible cause of death. None of 28 nest- lings in control nests died during this 3-d period. There was no significant difference in the survival between 52 treated and 28 control nestlings during the treatment period (x 2 = 1.68, P = 0.195). All nests were visited a final time for banding. At this time no differences in brood size could be de- tected. Of the 52 nestlings in treated nests, 42 (72%) survived from the conclusion of the treatment period to the time of banding, an average of 1 2 d later (range 5-21 d). In control nests 23 of 28 (82%) nestlings survived to banding, an average of 10 d later (range 3-20; x 2 = 0.97, P = 0.325). Successful, treated pairs raised an average of 2.0, as did successful pairs which were not part of this study on the adjacent study area. Successful control pairs raised an average of 1.8 young. Nestling Body Mass. Average mass of treated and control nestlings increased throughout the period (Fig. 2). On the third day, at the end of the treatment period, there was no significant difference in the mass gained by treated nestlings compared to con- trols (Mann-Whitney U — 451, P = 0.794). Poison Placement. Potential for scavenging by raptors was not reduced by placing poison inside ground squirrel burrows. Eighteen squirrels were found at 606 burrows (3.0/100 burrows) at which poison was placed outside compared to 19 squirrels Figure 2. Growth of control and treated nestling Swain- son’s hawks which were fed strychnine-poi- soned Richardson’s ground squirrels over a three day period in southeastern Alberta. Lines represent standard error and numbers = sam- ple size. Seven nestlings in control nests were not weighed on day 3 due to time constraints. near 434 burrows (4.4/100 burrows) where poison was placed within. Whether poison was placed out- side or inside the burrow also had no significant influence on squirrel survival based on the frequency with which plugged burrows were opened. Thirty- two of 75 burrows with poison placed above ground were opened a day later, compared to 30 of 75 bur- rows with poison placed below ground (x 2 = 0.11, P = 0.740). Discussion Absence of a clear indication of reduced survival due to secondary poisoning among treated nestling Swainson’s Hawks was unexpected. There was also no evidence that nestling growth was affected by the treatment. If local environmental conditions had any bearing on the outcome of this study, the drought conditions of 1988 may have worsened the impact. We attempted to simulate actual field conditions and conclude that under those conditions, secondary poi- soning through ground squirrels is a minor threat for Swainson’s Hawks. Other buteonine hawks may be similarly unaffected since it is unlikely that Swainson’s Hawks are unique with regard to the threat of secondary poisoning. We could not control some variables in this field study. Our daily visits may have had a detrimental impact (e.g., White and Thurow 1985). We also did not know how much “clean” food was brought to nestlings by their parents, a factor which may sim- ilarly lessen the impact of actual poisoning cam- paigns. 150 SCHMUTZ ET AL. VOL. 23, No. 4 Population performance of pairs of Ferruginous and Swainson’s Hawks monitored during the last 13 years is consistent with a minor (if any) impact from secondary poisoning in the area. In 1986 strychnine use (1782 1 of “Gopher Cop” in the Han- na area alone) was greater than during any other year in the last decade (Schmutz and Hungle 1989). Landowners distributed more poison in response to an increase in ground squirrel abundance. Nesting densities on the study area of both species were also higher in 1986 than during the preceding 10 yrs. Brood size among Ferruginous Hawks was greater in 1986 than during years of low strychnine use, while brood size of Swainson’s hawks remained un- changed. Hegdal and Gatz (1977) also could not detect a detriment to raptors arising from ground squirrel poisoning. Some mortalities did occur but could not be attributed to the poisoning. Two Great Horned Owls monitored using radiotelemetry in their study, frequented the treatment area and were alive 4 mo after treatment. Evisceration of poisoned ground squirrels is with- out doubt an important factor affecting the hazard to raptors. However, even eviscerated squirrels sometimes contained poisoned bait in their cheek pouches which could pose a threat. Our conclusion of a minor impact on raptors presupposes that the gastrointestinal tract of poisoned ground squirrels is not eaten. As soon as strychnine is absorbed into the blood stream, death probably occurs quickly, allow- ing relatively little strychnine to be present in the flesh. Raptors under food stress, however, may de- vour parts or all of a poisoned gastrointestinal tract and die. Evisceration of prey prior to consumption is not unique to the raptors studied. Vatev (1987) found an eviscerated vole in a nest of Long-legged Buzzard (B. rufinus ) in Bulgaria. Poole and Boag (1988) found that Gyrfalcon ( Falco rusticolus ) in Canada’s north eviscerated Rock Ptarmigan ( Lagopus mutus) and Arctic Ground Squirrel (Spermophilus parry ii). Fal- cons frequently left the stomach and intestines of large prey on the nest site but ate small passerines and microtines entirely. Mammalian predators which may eat ground squirrel viscera face a greater danger from secondary poisoning than do hawks. Marsh et al. (1987) found, however, that Coyotes ( Cams latrans) rejected viscera from strychnine-poisoned ground squirrels 66% of the time. One coyote died in their study. On our study area, one landowner’s dog died after eating ground squirrels from a poisoned area. The potential threat to livestock arising from poison-coated grain is well recognized by landowners. Efforts to estimate the potential impact from sec- ondary poisoning to wildlife assume that poison users adhere to recommended procedure. Regulatory mea- sures to protect wildlife may have little effect if not widely adhered to. Some landowners canvassed by us deliberately used more than the recommended dose in preparing bait; others used less. The pre- vailing practice was to spread poison in a given area once and to repeat poisoning 10-14 d later, depend- ing on the degree of Richardson’s Ground Squirrel survival and immigration. In some cases poisoned bait was placed inside a tire which resulted in per- sistent exposure to squirrels and thus to raptors. The influence of these deviations from recommended practice on secondary threats to wildlife is not known. Because of their characteristic dispersal tenden- cies, especially by juveniles in July, squirrel popu- lations are resilient to periodic, local eradication (Schmutz et al. 1979). Even adults removed in May- June repopulated a 4 ha plot to near former levels in <1 month (S. M. Schmutz 1977). From our ex- perience, many landowners would attempt to ame- liorate secondary impacts if made aware of the threat to non-target wildlife. Granivorous birds apparently face a greater threat from ground squirrel poisoning than do raptors. In a study in south-central Wyo- ming Hegdal and Gatz (1977) detected strychnine residues in the gastrointestinal tract of dead birds found in the area where poison had been broadcast. Casualty species included Mallard (Anas platyrhyn- chos ), Mourning Dove ( Zenaida macroura), Horned Lark ( Eremophila alpestris), American Crow (Corpus brachyrhynchos), European Starling (Sturnus vul- garis ), Yellow-headed (Xanthocephalus xanthoceph- alus ), Red- winged (Agelaius phoeniceus ) and Brew- er’s Blackbird (Euphagus cyanocephalus) , Brown- headed Cowbird (Molothrus ater), and Savannah (Passer cuius sandwichensis) , and Vesper Sparrow (Pooecetes gramineus ). No strychnine was detected in birds that were collected by various means outside the broadcast area. Horned Larks and Mourning Doves were most vulnerable, galliforms least vul- nerable. Acknowledgments We thank R. J. P. Meschishnick, D. and G. Wood for their observations on the rejection of viscera by hawks and owls. We are grateful to E. A. Driver, D. G. Forsyth, S M. Schmutz and S. N. Wiemeyer for their helpful com- Winter 1989 Secondary Poisoning of Raptors 151 ments on earlier versions of this manuscript. This work was supported by Environment Canada, the Special Areas Board of Hanna and the University of Saskatchewan. Literature Cited BORTOLOTTI, G. R. 1984. Trap and poison mortality of Golden and Bald Eagles. J. Wildl. Manage. 48:1173- 1179. Brett, L., P. Walter, G. Hankins and J. Garcia. 1976. Prey-lithium aversion. Ill: Buteo hawks. Be- havioral Biology 17:87-98. Cheney, C. D., S. B. Vander Wall and R. J. Poehl- MANN. 1987. Effects of strychnine on the behavior of Great Horned Owls and Red-tailed Hawks. J. Rap- tor Res. 21:103-110. Cromartie, E., W. L. Reichel, L. N. Locke, A. A. Belisle, T. E. Kaiser, T. G. Lamont, B. M. Mulhern, R. M. Prouty and D. M. Swineford. 1975. Residues of organochlorine pesticides and poly- chlorinated biphenyls and autopsy data for Bald Ea- gles, 1971-1972. Pestic. Monit. J. 9:11-14. Hegdal, P. L. and T. A. Gatz. 1977. Hazards to seed- eating birds and other wildlife associated with surface strychnine baiting for Richardson’s Ground Squirrels. Unpubl. Report, U.S. Fish and Wildlife Service, Den- ver, Colorado. 84 pp. Hegdal, P. L. and B. A. Colvin. 1988. Potential haz- ard to Eastern Screech Owls and other raptors of bro- difacoum bait used for vole control in orchards. En- viron. Toxicol. Chem. 7:245-260. Holler, N. R. and E. W. Schafer, Jr. 1982. Potential secondary hazards of avitrol baits to Sharp-shinned Hawks and American Kestrels. J. Wildl. Manage. 46: 462-468. Hudson, R. H., R. K. Tucker and M. A. Haegele. 1984. Handbook of toxicity of pesticides to wildlife. U.S. Fish and Wildlife Service, Resource Publication 153. 90 pp. Kaiser, T. E., W. L. Reichel, L. N. Locke, E. Cro- martie, A. J. Krynitsky, T. G. Lamont, B. M. Mulhern, R. M. Prouty, C. J. Stafford and D. M. Swineford. 1980. Organochlorine pesticide, PCB, and PBB residues and necropsy data for Bald Eagles from 29 states — 1975-1977. Pestic. Monit. J. 13:145- 149. Marsh, R. E., R. H. Schmidt and W. E. Howard. 1987. Secondary hazards to coyotes of ground squir- rels poisoned with 1080 or strychnine. Wildl. Soc. Bull. 15:380-385. Mendenhall, V. M. and L. F. Pank. 1980. Secondary poisoning of owls by anticoagulant rodenticides. Wildl. Soc. Bull. 8:311-315. Merson, M. H., R. E. Byers and D. E. Kaukeinen. 1 984. Residues of the rodenticide brodifacoum in voles and raptors after orchard treatment. J. Wildl. Manage. 48:212-216. Poole, K. and D. A. Boag. 1988. Ecology of gyrfalcons, Falco rusticolus, in the central Canadian Arctic: diet and feeding behavior. Can. J. Zool. 66:334-344. Redig, P. T., C. M. Stowe, T. D. Arendt and D. H DUNCAN. 1982. Relay toxicity of strychnine in rap- tors in relation to a pigeon eradication program. Vet Hum. Toxicol. 24:335-336. Reichel, W. L., S. K. Schmeling, E. Cromartie, T E. Kaiser, A. J. Krynitsky, T. G. Lamont, B. M. Mulhern, R. M. Prouty, C. J. Stafford and D. M. Swineford. 1984. Pesticide, PCB, and lead res- idues and necropsy data for Bald Eagles from 32 states — 1978-1981. Environ. Monit. Assess. 4:395-403. Reidinger, R. F. Jr. and D. G. Crabtree. 1974. Or- ganochlorine residues in Golden Eagles, United States — March 1964-July 1971. Pestic. Monit. J. 8:37-43. SCHMUTZ, J. K. 1977. Relationships between three species of the genus Buteo (Aves) coexisting in the prairie-parkland ecotone. M.Sc. Thesis, University of Alberta, Edmonton. 126 pp. . 1987. The effect of agriculture on Ferruginous and Swainson’s hawks. J. Range Manage. 40:438-440. . 1989. Hawk occupancy of disturbed grasslands in relation to models of habitat selection. Condor 91 362-371. and D. J. HunGLE. 1989. Populations of Fer- ruginous and Swainson’s Hawks increase in synchrony with ground squirrels. Can. J. Zool. 67:2596-2601 , S. M. Schmutz AND D. A. Boag. 1980. Co- existence of three species of hawks ( Buteo spp.) in the prairie-parkland ecotone. Can. J. Zool. 58:1075-1089. Schmutz, S. M. 1977. Role of dispersal and mortality in the differential survival of male and female Rich- ardson’s Ground Squirrels. M.Sc. Thesis. University of Alberta, Edmonton. 96 pp. , D. A. Boag, and J. K. Schmutz. 1 979. Causes of the unequal sex ratio in populations of adult Rich- ardson’s Ground Squirrels. Can. J. Zool. 57:1849-1855 Townsend, M. G., M. R. Fletcher, E. M. Odam and P. I. Stanley. 1981. An assessment of the secondary poisoning hazard of warfarin to Tawny Owls. J. Wildl Manage. 45:242-248. Vatev, I. Ts. 1987. Notes on the breeding biology of the Long-legged Buzzard ( Buteo rufinus ) in Bulgaria J. Raptor Res. 21:8-13. White, C. M. and T. L. Thurow. 1985. Reproduction of Ferruginous Hawks exposed to controlled distur- bance. Condor 87:14-22. Wodzicki, K. 1973. Prospects for biological control of rodent populations. Bull. Wld. Hlth. Org. 48:461-467. Department of Biology, University of Saskatchewan, Saskatoon, S7N 0W0, CANADA. Received 12 April 1989; accepted 10 December 1989 /. Raptor Res. 23(4):152-156 © 1989 The Raptor Research Foundation, Inc. SURVEYS FOR WINTERING BIRDS OF PREY IN SOUTHEASTERN GOLORADO: 1983-1988 David E. Andersen and Orrin J. Rongstad Abstract. — From 1983 through 1988 we conducted 45 road surveys for wintering birds of prey on the Pihon Canyon Maneuver Site in southeastern Colorado. Surveys were conducted from January through mid-March along a 75.7 km survey route that passed through pinyon ( Pinus «?£ftifc)-juniper (Juniperus monosperma) and shortgrass prairie habitat. Twelve raptor species were sighted on these surveys, with Red-tailed Hawk (Buteo jamaicensis), American Kestrel ( Falco sparverius), Golden Eagle ( Aquila chry- saetos), and Loggerhead Shrike ( Lanius ludovicianus) together representing over 73% of all sightings. As a group, birds of prey were consistently sighted most frequently (raptors/km) along the limestone breaks/ pinyon-juniper section of the survey route (P < 0.05). Indices to overall raptor abundance varied little among winters, although relative abundance varied Road surveys have historically been used to index densities of wintering birds of prey (Nice 1934; Leo- pold 1942; Craig 1978; Fuller and Mosher 1981). These surveys have also been used to estimate density (Andersen et al. 1985), distribution (Gessaman 1982), and total population size (Craighead and Craighead 1956; Woffinden and Murphy 1977; Bildstein 1978), determine perch (Marion and Ryder 1975; Preston 1980) and habitat use (Koplin 1973; Fischer et al. 1984), and make comparisons in relative abundance among years and areas (Johnson and Enderson 1972; Bauer 1982). However, most surveys in open hab- itats have been conducted where utility poles parallel roads, which potentially bias raptor distribution (Stahlecker 1978). Additionally, many surveys have been conducted for only 1 or 2 consecutive winters in a given location. Here, we report on the results of road surveys for wintering birds of prey that were conducted for 6 consecutive winters in an area where utility poles were not present along roads. Our objectives were to index wintering populations of birds of prey and to determine which habitats in southeastern Colo- rado consistently were used by wintering raptors. Study Area and Methods Road surveys were conducted on the 1040 km 2 Pihon Canyon Maneuver Site (PCMS), in Las Animas County in southeastern Colorado. Elevation on PCMS ranged from 1300-1700 m, and topography consisted of broad, moderately sloping uplands bordered by the Purgatoire River Canyon on the east, limestone hills on the west, and a basalt hogback on the south (U.S. Dept. Army 1980). Annual precipitation averaged approximately 32 cm, fluc- tuating widely from year to year and between areas of the parcel (U.S. Dept. Army 1980). Climate was classified as mid-latitude semiarid and mean monthly temperature ranged from — 1°C in January to 23°C in July. significantly during the 6-year period (P < 0.05). Vegetation on PCMS was dominated by shortgrass prairie and pinyon {Pinus edw/h)-juniper (Juniperus mono- sperma) woodland (Costello 1954; Kendeigh 1961). Three major and distinct habitats occurred on PCMS (Fig. 1). sandstone breaks/pinyon-juniper (sandstone) habitat oc- curred along the Purgatoire River Canyon and associated side canyons, shortgrass prairie (prairie) habitat covered the central, northern, and extreme western portion of PCMS, and limestone breaks/pinyon-juniper habitat (limestone) occurred in the west and northwest portions. Blue grama (Bouteloua gracilis) , alkali sacaton ( Sporobolus airoides), galleta ( Hilaria jamesii), western wheatgrass (Agropyron smithii), walking stick cholla ( Opuntia imbri- cata), and small soapweed (Yucca glauca) dominated short- grass prairie vegetation. See Shaw and Diersing (in press) for a detailed description of habitat on PCMS. We conducted road surveys for wintering birds of prey from 1983 through 1988. Surveys were conducted every 10-14 d from January through mid-March in all winters, except in 1983 when surveys were conducted at 6-d in- tervals. We selected a 75.7 km route that included 24.6 km (32.5%) of sandstone habitat, 36.0 km (47.6%) of prai- rie, and 15.1 km (19.9%) of limestone (Fig. 1). The route was selected on the basis of distribution of existing roads that were likely to be passable in winter, and so that each major habitat was represented. Utility poles were largely absent from the PCMS and did not parallel the survey route. Survey methods followed those outlined by Andersen et al. (1985). For each raptor sighted, we recorded species, age, sex, individual description (color morph), time of sighting, perch characteristics, direction of flight, and be- havior, where applicable. Starting points were alternated between ends of the route and we began surveys approx- imately 0.5 hr after sunrise. Surveys were initiated only on days when estimated wind speed was <10 km/hr and cloud cover was <50%. Two observers participated in all surveys and speed was maintained between 25 and 40 km/ hr. Surveys were completed in 3-4 hrs. Statistical procedures follow those outlined in Ryan et al. (1976), Snedecor and Cochran (1980), and Sokal and Rohlf (1980). Multiple comparisons were made using the protected least significant difference (LSD) method (Sned- ecor and Cochran 1980:234) with an experiment-wise al- 152 Winter 1989 Winter Surveys in Colorado 153 Figure 1. Habitats and survey route for wintering birds of prey on the Pinon Canyon Maneuver Site, Colorado. Sandstone breaks, basalt breaks, and limestone breaks all supported a pinyon-juniper plant association. pha of 0.05. Surveys were repeated over the same route each winter and we were not able to randomly sample habitats on PCMS. Thus, our analyses do not test for treatment (i.e., habitat) differences (see Hurlbert 1984), only differences among locations in sighting frequency. Years were treated as independent replicates in analysis of variance (ANOVA) procedures. Results From 1983 through 1988 we conducted 45 surveys for wintering birds of prey on PCMS. Twelve species of raptorial birds were observed on these surveys (Table 1) with Red-tailed Hawk (48.8%), American Kestrel (13.5%), Golden Eagle (9.4%), and Loggerhead Shrike (9.0%) totalling over 73% of all sightings. Raptors were consistently sighted more frequently in limestone habitats than in either sandstone or prairie habitats (Table 2; 2-way AN- OVA, F = 5.43; 2,10 df; P < 0.05). Variation among winters in raptor densities on PCMS could be related to factors independent of habitat. We adjusted for annual variation in sighting frequency by subtract- ing yearly mean sighting frequency (habitats com- bined) from sighting frequency in each habitat (ad- justed sighting frequency). Adjusted sighting frequency was significantly higher in limestone hab- itats than in either sandstone or prairie habitats (1- way ANOVA, F = 8.16; 2,15 df; P < 0.005; LSD procedures), indicating that winter raptor density was also highest along the limestone section of the survey route. In the winters of 1982-83 (x 2 = 6.22, 2 df, P < 0.05) and 1984-85 (x 2 = 8.14, 2 df, P < 0.025; Table 2) birds of prey were sighted more frequently (unadjusted sighting frequency) than expected in limestone habitats and less frequently than expected in sandstone habitats, based on the proportion of the survey route that passed through each habitat. No significant differences in sighting frequency were found among habitats in the winters of 1983-84, 1985-86, 1986-87, or 1987-88 (x 2 -tests, all Ps > 0,10). Raptors (species combined) were sighted most frequently (sightings/km) in limestone habitats in every winter (N = 6) that surveys were conducted (Exact Randomization Test, P < 0.005; Table 2), indicating that raptors were consistently most abun- dant along the section of the survey route that passed through limestone habitat. Overall, wintering raptor abundance (species combined) on PCMS was similar among years (x = 14.1 km/raptor, coefficient of variation = 19.4%). However, counts of individual raptor species sighted on survey routes changed among winters (x 2 = 27.13, 15 df, P < 0.05), indicating that the abundance of different species on PCMS also differed among win- ters. 154 David E. Andersen and Orrin J. Rongstad Vol. 23, No. 4 Table 1. Birds of prey sighted on a 75.7 km winter 3 survey route on Pinon Canyon Maneuver Site, Colorado from 1983 through 1988. Number of surveys is given in parentheses. Species 82-83 (11) 83-84 (6) Winter 84-85 85-86 (8) (7) 86-87 (8) 87-88 (5) Total (45) Red-tailed Hawk {Buteo jamaicensis) 29 16 10 21 20 6 102 American Kestrel ( Falco sparverius ) 4 0 5 10 10 4 33 Golden Eagle ( Aquila chrysaetos) 6 6 3 1 6 1 23 Loggerhead Shrike ( Lanius ludovicianus ) 2 1 5 7 1 6 22 Prairie Falcon ( Falco mexicanus) 3 4 2 1 4 3 17 Ferruginous Hawk ( Buteo regalis) 5 0 0 1 3 1 10 Northern Harrier ( Circus cyaneus ) 0 0 1 0 3 1 5 Rough-legged Hawk {Buteo lagopus) 0 2 5 2 2 0 11 Merlin {Falco columbarius ) 0 0 2 1 1 4 8 Cooper’s hawk {Accipiter cooperii ) 0 1 1 1 0 2 5 Sharp-shinned Hawk {Accipiter stmatus) 0 2 0 1 1 1 5 Bald Eagle {Haliaeetus leucocephalus ) 0 0 0 0 2 0 2 Unidentified 0 1 0 0 0 0 1 Total 49 33 34 46 53 29 244 km/sighting 17.0 13.8 17.8 11.5 11.4 13.0 14.1 a Surveys were conducted in January, February, and early March. Discussion Wintering birds of prey were consistently sighted more frequently in limestone than in any other hab- itat. Sighting probabilities can differ among habitat types or years (Andersen et al. 1985), although be- cause the number of sightings/habitat/year for each species was small, we were unable to estimate de- tection functions or raptor densities. However, the probability of detecting a raptor in sandstone and prairie habitats was at least as high as in limestone habitats, based on vegetative characteristics that in- fluence detectability (U.S. Dept. Army 1980; Shaw and Diersing, in press; Millsap and LeFranc 1988). Higher sighting frequencies in limestone habitat in- dicate that densities of wintering raptors were also highest in limestone habitats. Local prey abundance (Craighead and Craighead 1956; Enderson 1964; Phelan and Robertson 1977; Newton 1979:289) and availability (Sylven 1978; Baker and Brooks 1981) influence winter raptor abundance. On PCMS, we did not have an index to winter prey abundance or availability. However, variation among years in raptor abundance was low, indicating either that prey densities were similar among years, or that factors unrelated to fluctuating prey availability influenced raptor abundance (Bild- stein 1978). As an index to abundance, distance travelled per sighting (habitats combined) ranged from 11.4 km (0.087 raptors/km) in 1983-84 to 17.8 km (0.056 raptors/km) in 1984-85. These indices indicate that winter densities of raptors on PCMS were lower than in similar areas in eastern Colorado in other years: 6.3 km/raptor (Enderson 1965), 5.8 km/rap- tor (Johnson and Enderson 1972), 8.9 km/raptor (Bauer 1982), and 5. 4-8. 7 km/raptor (Andersen 1984). Excluding shrikes (to make results compa- rable with cited studies), distance travelled/sighting on PCMS ranged from 7.7 km in 1983-84 to 20.9 in 1984-85. However, other studies were conducted where utility poles paralleled the survey route (ex- cept Andersen 1984) which likely influenced winter raptor distribution (Enderson 1964; Stahlecker 1978) by concentrating raptors along the survey route and increasing the probability of being observed. Thus, winter density on PCMS may be similar to other areas of eastern Colorado where surveys have been conducted and winter sighting frequency on PCMS may be a useful index of winter raptor density in eastern Colorado. Acknowledgments Funding was provided by the U.S. Army, Environment, Energy, and Natural Resources Division, Fort Carson, Colorado through the Colorado Field Office of the U.S. Winter 1989 Winter Surveys in Colorado 155 Table 2. Raptors sighted (by habitat) during winter road surveys conducted on Pinon Canyon Maneuver Site, Colorado from 1983 through 1988. The total survey route was 75.7 km, of which 24,6 km (32.5%) was sandstone, 36.0 km (47.6%) was prairie, and 15.1 km (19.9%) was limestone habitats. Winter Habitat No. OF Sandstone Prairie Limestone Surveys N km/Sighting N km/Sighting N km/Sighting 82-83 11 10 24.5 23 17.2 16 10.5 83-84 6 12 12.3 12 18.0 9 10.1 84-85 8 6 32.8 15 19.2 13 9.3 85-86 7 15 11.5 17 14.8 14 7.6 86-87 8 13 15.1 26 11.1 14 8.4 87-88 5 11 11.2 9 20.0 9 8.4 X 17.9 16.7 9.0 Fish and Wildlife Service and the Wisconsin Cooperative Wildlife Research Unit. Support was also provided by the Graduate School and the College of Agricultural and Life Sciences at the University of Wisconsin-Madison. W. R. Mytton, B. D. Rosenlund, T. Prior, T. L. Warren, S. R. Emmons, and A. R. Pfister helped coordinate this project on the PCMS, and W. P. Fassig, K. M. Firchow, E. M. Gese, C. Gordon, D. J. Grout, G. Hughes, L. L. Kinkel, T. R. Laurion, A. W. Stephenson, and E. H. Valentine assisted in conducting surveys. L. L. Kinkel critically read earlier drafts of this manuscript and comments by M. R. Fuller and J. A. Gessaman improved the manuscript. Literature Cited Andersen, D. E. 1984. Military training and the ecol- ogy of raptor populations at Fort Carson, Colorado. M.S. Thesis, Univ. of Wis., Madison. 142 pp. , O. J. Rongstad and W. R. Mytton. 1985. Line transect analysis of raptor abundance along roads. Wildl. Soc. Bull. 13:533-539. Baker, J. A. and R. J. Brooks. 1981. Distribution patterns of raptors in relation to density of meadow voles. Condor 83:42-47. Bauer, E. N. 1982. Winter roadside raptor surveys in El Paso County, Colorado, 1962-1979. Raptor Res. 16: 10-13. Bildstein, K. L. 1978. Behavioral ecology of Red-tailed Hawks ( Buteo jamaicensis), Rough-legged Hawks ( B . lagopus). Northern Harriers ( Circus cyaneus), Ameri- can Kestrels ( Falco sparverius ) and other raptorial birds wintering in south central Ohio. Ph.D. Thesis, Ohio State Univ., Columbus. 364 pp. COSTELLO, D. F. 1954. Vegetation zones in Colorado. Pp. iii-x. In H. D. Harrington. Manual of the plants of Colorado. Sage Books, Denver, Colo. 666 pp. Craig, T. H. 1978. A car survey of raptors in south- eastern Idaho 1974-1976. Raptor Res. 12:40-45. Craighead, J. J. and F. C. Craighead, Jr. 1956. Hawks, owls and wildlife. Stackpole Books, Harris- burg, PA and The Wildl. Manage. Inst. Dover, NY reprint, 1969. 443 pp. Enderson, J. H. 1964. A study of the Prairie Falcon in the central Rocky Mountain Region. Auk 81:332- 352. . 1965. Roadside raptor count in Colorado. Wil- son Bull. 77:82-83. Fischer, D. L., K. L. Ellis and R. J. Meese. 1984 Winter habitat selection of diurnal raptors in central Utah. Raptor Res. 18:98-102. Fuller, M. R. and J. A. Mosher. 1981. Methods of detecting and counting raptors; a review. Pages 235- 246. In D. J. Ralph and J. M. Scott, Eds. Estimating numbers of terrestrial birds. Studies in Avian Biol 6 Cooper Ornithol. Soc. 630 pp. GESSAMAN, J. A. 1982. A survey of raptors in northern Utah, 1976-79. Raptor Res. 16:4-10. Hurlbert, S. H. 1984. Pseudoreplication and the de- sign of field experiments. Ecol. Monogr. 54:187-211 Johnson, D. and J. H, Enderson. 1972. Roadside raptor census in Colorado — winter 1971-72. Wilson Bull. 84:489-490. Kendeigh, S. C. 1961. Animal ecology. Prentice-Hall, NJ. 468 pp. Koplin, J. R. 1973. Differential habitat uses by sexes of American Kestrels wintering in northern California Raptor Res. 7:39-42. Leopold, A. 1942. A raptor tally in the Northwest. Condor 44:37-38. Marion, W. R. and R. A. Ryder. 1975. Perch-site preference of four diurnal raptors in northeastern Col- orado. Condor 77:350-352. Millsap, B. A. and M. N. LeFranc, Jr. 1988. Road transect counts for raptors: how reliable are they'* J Raptor Res. 22:8-16. 156 David E. Andersen and Orrin J. Rongstad Vol. 23, No. 4 Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. 399 pp. Nice, M. M. 1934. A hawk census from Arizona to Massachusetts. Wilson Bull. 46:93-95. Phelan, F. J. S. and R. J. Robertson. 1977. Predatory responses of a raptor guild to changes in prey density. Can. J. Zool. 56:2565-2572. Preston, C. R. 1980. Differential perch site selection by color morphs of the Red-tailed Hawk {Buteo ja- maicensis). Auk 97:782-789. Ryan, T. A., Jr., B. L. Joiner and B. F. Ryan. 1976. MINITAB student handbook. Duxbury Press, North Scituate, MA. 341 pp. Shaw, R. B. and V. E. Diersing. In Press. Impact of two-years of tracked vehicular training on the vege- tation of the U. S. Army’s Pinon Canyon Maneuver Site, CO. J. Envir. Quality. Snedecor, G. W. and W. G. Cochran. 1980. Statis- tical methods 7th ed. Iowa State Univ. Press, Ames. 507 pp. Sokal, R. R. and F. J. Rohlf. 1980. Biometry 2nd ed. W. H. Freeman and Co., New York. 859 pp. Stahlecker, D. W. 1978. Effect of a new transmission line on wintering birds of prey. Condor 80:444-446. Sylven, M. 1978. Interspecific relations between sym- patrically wintering Common Buzzards, Buteo buteo and Rough-legged Buzzards, Buteo lagopus. Ornis Scand 9:197-206. U.S. Dept. Army. 1980. Draft: Environmental impact statement for training land acquisition for Fort Carson Colorado. U.S. Army Corp of Engineers, Omaha, NE. 284 pp. Woffinden, N. D. and J. R. Murphy. 1977. A road- side raptor census in the eastern Great Basin — 1973- 1974. Raptor Res. 11:62-66. Department of Wildlife Ecology, University of Wis- consin, Madison, WI 53706 (DEA and OJR) and Colorado Field Office, U.S. Fish and Wildlife Ser- vice, 730 Simms Street, No. 290, Golden, CO 80401 (DEA). Current address of first author: U.S. Fish and Wildlife Service, Minnesota Cooperative Fish and Wildlife Research Unit, Department of Fish- eries and Wildlife, University of Minnesota, St. Paul, MN 55108. Received 10 March 1989; accepted 20 December 1989 J Raptor Res. 23(4):157-161 © 1989 The Raptor Research Foundation, Inc. FOOD HABITS OF NESTING PRAIRIE FALCONS IN CAMPBELL COUNTY, WYOMING John R. Squires, Dr. Stanley H. Anderson and Robert Oakleaf Abstract. — Fifteen species of prey were utilized by nesting Prairie Falcons ( Falco mexicanus ) as deter- mined through pellet analysis. Thirteen-lined Ground Squirrels ( Spermophilus tridecemlineatus ), the most common prey, were present in 91% of the pellets, followed by Western Meadowlarks ( Sturnella neglecta) which were present in 56% of pellets. Horned Larks ( Eremophila alpestris) and Lark Buntings ( Calamospiza melanocorys ) were found in 23% and 12% of the pellets, respectively. Additional prey species were present in low frequencies (<5%). Eighty-nine percent of pellets contained both bird and mammal remains. Four percent of pellets contained only bird remains while 7% contained only mammalian remains. The Prairie Falcon ( Falco mexicanus ) is an open- country raptor that forages on a diversity of birds and small mammals (Enderson 1962; Denton 1975; Becker 1979; Boyce 1985). Diets of nesting Prairie Falcons depend on local prey abundance. For ex- ample, some nesting Prairie Falcons forage primar- ily on mammalian prey (Porter et al. 1973; Ogden and Hornocker 1977), while others forage primarily on avian prey (Marti and Braun 1975; Becker 1979; Boyce 1985). Therefore, predicting falcon diets based on studies from different areas is difficult. Our analysis of Prairie Falcon diet was part of a larger study that investigated movements and hab- itat-use patterns. Our objective was to determine important prey species used by a small population of nesting Prairie Falcons in northeastern Wyoming. Study Area and Methods The study area located in northcentral Wyoming in- cluded the Pumpkin Butte formation, which contains 5 buttes used by a small, isolated population of nesting Prai- rie Falcons (x = 6 pairs/yr). Average distance between occupied territories was 0.56 km, and prairies surrounding the buttes were either sagebrush ( Artemisia tridentata ) steppes or open grasslands. Pellets and prey remains were collected at falcon eyries from 1982 through 1985. Mammalian prey were identified by examining both internal and external hair character- istics and through comparisons with specimens collected at the study site. Hair identification procedures were con- ducted according to those outlined by Moore et al. (1974). Avian remains in pellets were identified by comparing characteristic feather, beak, and bone fragments with the University of Wyoming’s museum collection. Pellet analysis cannot be used to directly quantify the number of prey items consumed due to several biases as- sociated with the technique (Craighead and Craighead 1956). For example, adult falcons have been observed removing pellets and uneaten prey remains from their nest site (Fowler 1931; Wayre and Jolly 1958). Therefore, the percentage of prey items in the diet as determined through analysis of pellet and prey remains should be interpreted as an approximation of the falcon’s true diet. To supple- ment data gathered from pellet and prey remain analyses, foraging falcons were observed when possible. Prey Abundance Indices. During 1984 and 1985, birds and small mammals were sampled on transects that were established in areas frequently used by foraging falcons Transect sampling was used as an index of the relative abundance of prey species and to document changes in prey abundance between years. All prey-sized birds oc- curring along a 1.6 km, walked, belt-transect were count- ed. Bird transects (N = 6) were censused twice each sum- mer between July 2 and July 16, at approximately 0.5 hr after sunrise. Each small mammal transect (N = 4) consisted of 100 Sherman live-traps placed 10 m apart that were trapped for 2 successive days and nights. Traps were baited with a mixture of peanut butter and oats. Traps were checked approximately 1 hr after sunrise and 1 hr before sunset. All transects were sampled between 8 July and 24 July during both years. Transect data were compared using £-Tests [STATIS- TICAL package for the social sciences (SPSS), Nie et al. 1975; a = 0.05). The 2 replicates of the 1984 indices were averaged and then compared to the average of 1985 indices. Results Food Habits. A total of 243 pellets, plus prey remains from 9 eyries, were examined. At least 15 prey species were used by this population of Prairie Falcons (Table 1). Four species, Thirteen-lined Ground Squirrels ( Spermophilus tridecemlineatus ), Western Meadowlarks ( Sturnella neglecta ), Horned Larks {Eremophila alpestris) and Lark Buntings {Ca- lamospiza melanocorys ) were important prey species during each year of the study (Table 2). Most of the remaining prey species occurred in pellets in low' frequencies (<5%). Eighty-nine percent of the pel- lets contained both bird and mammal remains. Four percent of the pellets contained only bird remains while 7% contained only mammalian remains. Foraging Behavior. Foraging falcons used low, flushing-type flights when hunting. Typically, fal- 157 158 Squires et al. VOL. 23, No. 4 Table 1. Food items in pellets of Prairie Falcons nesting in Campbell County, Wyoming, 1982-85. Prey Species Number of Pellets Containing Prey Item Per- cent Pel- lets with Prey Item Mammals 13-lined ground squir- rel ( Spermophilus tridecemlinea- tus) 221 91 Least Chipmunk ( Eutamias mini- mus) 7 3 Leporidae spp. 8 3 Unknown mammal 2 1 Mule Deer ( Odocoileus hemionus) 1 <1 Birds Western Meadowlark ( Sturnella neglecta ) 135 56 Horned Lark ( Eremophila alpes- tris ) 55 23 Unknown passerine 45 19 Lark Bunting ( Calamospiza me- lanocorys ) 28 12 McCown’s Longspur ( Calcarius mccownii ) 13 5 Brewer’s Sparrow ( Spizella brew- eri ) 4 2 Pinyon Jay ( Gymnorhinus cyano- cephalus) 2 1 Brewer’s Blackbird ( Euphagus cy- anocephalus ) 1 <1 Northern Flicker ( Colaptes aura- tus ) remains* <1 Anseriform spp. remains <1 Mollusks Gastropoda spp. 1 <1 * Remains found in eyrie. cons would attempt to surprise prey by flying ap- proximately 6 m off the ground while traveling rap- idly down draws. Falcons using fast, contour-hugging flights have been noted by other researchers (Web- ster 1944; Dunstan et al. 1978). Foraging falcons were also observed “still hunting” from fence posts, telephone poles, and high-tension power poles. On 16 June 1984, a male falcon was observed hunting Cliff Swallows ( Hirundo pyrthonota ). He flew rapidly along a cliff face in an apparent effort to capture swallows leaving their nests. On several oc- casions, the falcon hung upside from the Cliff Swal- low nests while reaching inside, presumably to ex- tract nesting swallows or their young. The falcon was unsuccessful though persistent in capturing swallows during these observations, however which suggests that the foraging technique must at times be successful. A single case of kleptoparasitism was observed when a male falcon forced a Northern Harrier ( Cir- cus cyaneus ) to drop a mouse which the falcon re- trieved. Both the methods of hunting cliff swallows and prey piracy from a Northern Harrier were also documented by Holthuijzen et al. (1987). Prey Abundance Indices — Passerine Birds. Fifteen species of birds were observed on bird tran- sects (Table 3). Lark Buntings, Brewer’s Sparrows, Horned Larks, and Vesper Sparrows (Pooecetes gra- mineus) were present on >75% of all transects sug- gesting an even distribution on falcon foraging areas. Remaining species were observed infrequently (on <13% of the transects). The f-Test statistic was used to test the null hy- pothesis that both 1984 and 1985 bird transects had equal bird numbers. We failed to accept the null hypothesis (P = 0.019) in favor of the alternate hypothesis which suggested that significantly more prey-sized birds were present on falcon foraging areas in 1985 than in 1984. Prey Abundance Indices — Small Mammals. During 800 trap days and 800 trap nights, a total of 166 small mammals from 3 species were captured (Table 4). A total of 141 Thirteen-lined Ground Squirrels were captured accounting for 85% of total captures. This was the only species captured during day trapping while all 3 species were captured at night. However, 23 Thirteen-lined Ground Squir- rels classified as night captures were probably trapped either early morning or late evening and did not represent nocturnal activities for the species (Murie and Michener 1984). The only species captured on transect 1 located in a grassland habitat was Thirteen-lined Ground Squirrel while all 3 species were captured on the other transects located in mixed grass-sage habitat types (Table 4). The greatest number of small mam- mals were captured on transect 4 which was located in a similar habitat type as were transects 2 and 3 but in an area of very sandy soil. The uTest statistic was used to test the null hy- pothesis that the numbers of small mammals cap- Winter 1989 Prairie Falcon Food Habits 159 Table 2. Percent frequency of prey remains found in Prairie Falcon pellets according to year (1982-85), Campbell County, Wyoming. 1982 (N = 21) 1983 (N = 91) 1984 (N = 57) 1985 (N = 74) Four Yr. Wt. Ave. Mammals 13-lined Ground Squirrel ( Spermophilus tridecemlineatus) 95 95 75 97 91 Least Chipmunk ( Eutamias minimus) 5 2 7 * 3 Leporidae spp. — 7 2 — - 6 Mule Deer ( Odocoileus hemionus) — — 2 — tr** Unknown mammal 10 1 — — 1 Western Meadowlark ( Sturnella neglecta) Birds 81 66 53 38 56 Horned Lark ( Eremophila alpestris) 24 29 26 10 22 Lark Bunting ( Calamospiza melanocorys) 14 19 4 11 13 McCown’s Longspur ( Calcarius mccownii) 19 4 7 3 6 Brewer’s Sparrow ( Spizella breweri) — 4 — — 1 Brewer’s Blackbird ( Euphagus cyanocephalus) — 1 — — tr Pinyon Jay ( Gymnorhinus cyanocephalus) — — — 3 1 American Kestrel ( Falco sparverius) — — 2 — tr Unknown passerine — 10 9 43 19 Gastropoda spp. Mollusks 1 tr * Prey item not present. ** tr = trace. tured in 1984 and 1985 were equal. We failed to reject this hypothesis (P = 0.785) suggesting that no significant differences in the number of small mam- mals were present between years. Discussion Prairie Falcons at Pumpkin Butte used Thirteen- lined Ground Squirrels as their primary prey species during the nesting season. Thirteen-lined Ground Squirrel litters emerge from their burrows in mid- June (Streubel and Fitzgerald 1978) making them highly available to foraging falcons. Ninety-three percent of pellets contained some feather remains indicating these falcons frequently foraged on birds. Other studies have documented similar high use of avian prey by nesting Prairie Falcons (Marti and Braun 1975; Denton 1975; Pe- terson et al. 1977; Voilker, unpublished data cited in Sherrod 1978; Becker 1979). However, Mac- Laren (1986) found that pellets from Prairie Falcons in southeastern Wyoming were dominated by mam- malian prey (86.3%) with only 13.7% containing avian remains. Western Meadowlarks were the primary avian prey found in 56% of the pellets in our study. How- ever, meadowlarks were not the most common prey- sized bird present on the study area according to our bird transects. Prairie Falcons may actively select meadowlarks, even though other avian prey were more abundant. Other researchers have also noted a similar high incidence of meadowlark in Prairie Falcon diets (Fowler 1931; Enderson 1962; Leedy 1969; Platt 1974; Denton 1975; Becker 1979). Two pellets contained unusual prey remains for Prairie Falcons. One pellet was composed entirely of Mule Deer ( Odocoileus hemionus ) hair. There are at least 2 possible explanations for such an occur- rence. The falcon either fed on deer carrion directly or consumed the stomach of a carrion feeding bird (i.e., Pinyon Jay ( Gymnorhinus cyanocephalus ), Clark’s Nutcracker ( Nucifraga Columbiana ), Brew- er’s Blackbird ( Euphagus cyanocephalus)). The first explanation was thought more probable since the pellet did not contain any feather remains. To our knowledge, this is the first documentation of Prairie Falcons possibly consuming carrion. The second unusual pellet was composed entirely of aquatic snail shells with no trace of feather or fur 160 Squires et al. Vol. 23, No. 4 Table 3. Abundance and distribution of birds observed on 6 strip transects of 1.6 km each. Species Number Observed Percent- age of Tran- sects on Which Species Were Observed Lark Bunting ( Calamospiza melanocorys) 284 88 Brewer’s Sparrow ( Spizella breweri ) 176 83 Horned Lark ( Eremophila al- pestris) 174 71 Vesper Sparrow ( Pooecetes gra- mineus ) 133 75 Mourning Dove ( Zenaida ma- croura) 66 25 Cliff Swallow ( Petrochelidon pyrrhonota) 47 8 McCown’s Long- spur ( Calcarius mccownii) 43 13 Brewer’s Blackbird ( Euphagus cyanocephalus ) 22 13 Western Meadow- lark ( Sturnella neglecta) 14 25 Lark Sparrow ( Chondestes grammacus) 6 13 Sage Thrasher ( Oreoscoptes montanus) 2 4 Common Night- hawk ( Chordeiles minor ) 2 4 Chestnut-collared Long- spur ( Calcarius ornatus) 2 4 Red-winged Black- bird ( Agelaius phoeniceus) 2 4 Killdeer ( Charadrius vociferus ) 1 4 remains. It is difficult to envision how a Prairie Falcon would encounter snails while foraging. Pos- sibly, the falcon ate the stomach of an aquatic-feed- ing bird (Killdeer ( Charadrius vociferus ) or water- fowl) which contained the shells, but no feather remains were present in the pellet. The falcon might have selectively picked snails from aquatic vegetation while drinking or bathing. Other species of birds do consume snails with shells to help meet their calcium requirements (Krapu and Swanson 1975; Beasom Table 4. The number of small mammals captured on live-trap transects during 1984-85, Campbell County, Wyoming. Tran- sect Number (Year) Thirteen- lined Ground Squirrel Deer Mouse Grass- hopper Mouse Day Night Day Night Day Night 1-(1984) 9 0 0 0 0 0 1-(1985) 14 2 0 0 0 0 2-(1984) 8 0 0 4 0 0 2-(1985) 11 5 0 7 0 1 3-(1984) 2 0 0 0 0 0 3-(1985) 2 1 0 2 0 1 4-(1984) 39 6 0 2 0 5 4-(1985) 33 _9 0 J _ 0 2 Total 118 23 0 16 0 9 and Patte 1978; Ankney and Scott 1980). It is pos- sible that Prairie Falcons also consume snails to compensate for calcium deficiencies incurred during egg production. In summary, Prairie Falcons at Pumpkin Butte foraged on a diversity of prey that included at least 1 5 species. However, 4 species, Thirteen-lined Ground Squirrel, Western Meadowlark, Horned Lark, and Lark Bunting comprised the bulk of the diet. Most pellets (93%) contained feather remains suggesting that falcons frequently foraged on avian prey. Acknowledgments This project was funded by the Wyoming Game and Fish Department with assistance from the Wyoming Co- operative Fish and Wildlife Research Unit. J. Depper- schmidt, S. Irvine, and M. Fowler provided valuable as- sistance with field data collections. Literature Cited Ankney, C. D. and D. M. Scott. 1980. Changes in nutrient reserves and diet of breeding Brown-headed Cowbirds. Auk 97:684-696. Beasom, S. L. and O. H. Patte. 1978. Utilization of snails by Rio Grande Turkey hens. J. Wildl. Manage. 42:916-919. Becker, D. M. 1979. A survey of raptors on national forest land in Carter County, Montana. U.S. Dept, of Agriculture, Forest Service, Final Progress Rept. 1977- 79. 61 pp. Boyce, D. A., Jr. 1985. Prairie Falcon prey in the Mohave desert, California. Raptor Res. 19:128-134, Craighead, J. J. and F. C. Craighead Jr. 1956. Winter 1989 Prairie Falcon Food Habits 161 Hawks, owls and wildlife. The Stackpole Co., Har- risburg, Pa. and the Wildlife Management Institute, Washington, DC. 443 pp. Denton, S. J. 1975. Status of Prairie Falcons breeding in Oregon. M.S. Thesis, Oregon State University, OR. 58 pp. Dunstan, T. C., J. H. Harper and K. B. Phipps. 1978. Habitat use and hunting strategies of Prairie Falcons, Red-tailed hawks, and Golden Eagles. U.S.D.I. Bu- reau of Land Management contract 52500-CT5-1013. 177 pp. Enderson, J. H. 1962. Ecology of the Prairie Falcon (Falco mexicanus schlegel) in the central Rocky Moun- tain Region. Ph.D. Thesis, University of Wyoming. 87 pp. Fowler, F. H. 1931. Studies of the food and growth of the Prairie Falcon. Condor 33(5): 1 93-201 . Holthuijzen, A. M. A., P. A. Duley, J. C. Hager, S. A. Smith and K. N. Wood. 1987. Piracy, insectivory and cannibalism of Prairie Falcons ( Falco mexicanus ) nesting in southwestern Idaho. J. Raptor Res. 21(1): 32-33. Krapu, G. L. and G. A. Swanson. 1975. Some nutri- tional aspects of reproduction in prairie nesting Pin- tails. /. Wildl. Manage. 39:156-162. Leedy, R. R. 1969. The status of the Prairie Falcon in western Montana: special emphasis on possible effects of chlorinated hydrocarbon insecticides. M.S. Thesis, University of Montana, Missoula. 96 pp. MacLaren, P. A. 1986. Resource partitioning in an assemblage of breeding raptors from southeastern Wy- oming. M.S. Thesis, University of Wyoming, Laramie. 64 pp. Marti, C. P. and C. E. Braun, 1975. Use of tundra habitats by Prairie Falcons in Colorado. Condor 77(2): 213-214. Moore, T. D., L. E. Spence, G. E. Dugnolle and W. Hepworth. 1974. Identification of the dorsal guard hairs of some mammals of Wyoming. Wyoming Game and Fish Dept., Cheyenne. 177 pp. Murie, J. O. and G. R. Michener. 1984. The biology of ground-dwelling squirrels. University of Nebraska Press, Lincoln and London. 459 pp. Nie, J. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner and D. H. Bent. 1975. Statistical package for the social sciences. McGraw-Hill Book Co., New York. 675 pp. Ogden, V. T. and M. G. Hornocker. 1977. Nesting density and success of Prairie Falcons in southwestern Idaho. J. Wildl. Manage. 41:1-11. Peterson, S. R., G. M. Sitter and B. W. James. 1977. Feeding activity and behavior of Prairie Falcons. Pages 165-177. In Snake River Birds of Prey Research Prog. Annual Report 1977. U.S. Dept, of Interior, Bureau of Land Management, Boise, ID. 201 pp. Platt, S. W. 1974. Breeding status and distribution of the Prairie Falcon in northern New Mexico. M.S. thesis, Oklahoma St. Univ., Stillwater. 69 pp. Porter, R. D., C. M. White and R. J. Erwin. 1973. The Peregrine Falcon in Utah, emphasizing ecology and competition with the Prairie Falcon. Brig. Young Unw. Sci. Bull. 18:1-74. Sherrod, S. K. 1978. Diets of North American falconi- formes. Raptor Research 12(3/4):49-121. Streubel, D. P. and J. P. Fitzgerald. 1978. Sper- mophilus tridecemlineatus. Mammal. Species 103:1-5 Wayre, P. and G. F. Jolly. 1958. Notes on the breed- ing of the Iceland Gyr falcon. Brit. Birds 51:285-290 Webster, H. Jr. 1944. A survey of the Prairie Falcon in Colorado. Auk 61:609-616. Wyoming Cooperative Fish and Wildlife Research Unit, P.O. Box 3166 University Station, Laramie, WY 82071. Address of third author: Wyoming Game and Fish Dept. 260, Buena Vista, Lander, WY 82520. Received 27 February 1989; accepted 15 December 1989 J Raptor Res. 23(4):162-166 © 1989 The Raptor Research Foundation, Inc. OBSERVATIONS ON THE EVENING DEPARTURE AND ACTIVITY OF WINTERING SHORT-EARED OWLS IN NEW JERSEY Thomas Bosakowski Abstract. — Wintering Short-eared Owls ( Asio flammeus ) were primarily crepuscular and nocturnal. Roost departure occurred most frequently after sunset (83%) with most exceptions occurring on heavily overcast days. Owls usually departed singly or in tandem and engaged in a steady direct flight, presumably to a predetermined hunting area. Hunting was rarely initiated near the roost site. Night observations up to 5 hr after sunset revealed that owls hunted continuously into the night and were not merely crepuscular. Despite their reputation as an on-the-wing predator, extended periods of perch-hunting were often observed after sunset, particularly on windless nights. Active hunting from perches was evidenced by a continual series of pounces and hunting flights that were launched from the same or nearby perches. Owls responded on 3 of 5 trials to broadcasts of prerecorded Short-eared Owl calls with vocalizations and/or vigorous circling flights over the calling station. The Short-eared Owl ( Asio flammeus) is primarily nocturnal during the winter months, and hence rel- atively few attempts have been made to study activity patterns and behavior of the species on wintering grounds (e.g., Short and Drew 1962; Clark 1975; Marr and McWhirter 1982). In this paper I present information on evening roost departure, social in- teractions and hunting activities of wintering Short- eared Owls in New Jersey. Study Area and Methods The primary study site was an inland tidal marsh known as the Hackensack Meadowlands (Lyndhurst, Bergen Co.) previously described in detail (Bosakowski 1984). Briefly, the marshes are dominated by common reed ( Phragmites communis) and are surrounded by dense urban develop- ment. Several active and inactive landfill mounds are pres- ent in the marsh as well as many warehouses, rights-of- way, and occasional light industry. Weather data were obtained from the National Weather Service, Newark In- ternational Airport, located 13 km south of the study area. During the winters of 1982 and 1983, 25 hr of obser- vation were made on 22 different evenings. I usually ar- rived at least 0.5 hr before sunset to make observations at dusk at a known roosting area (Bosakowski 1986). Ob- servations were continued until at least 0.5 hr after sunset and occasionally up to 1.5 hr. From a 3 m mound, I recorded the number, location, activity, flight direction, and time that owls emerged for initial evening flight. Dur- ing late winter 1988, an additional 22 hr of observation were conducted on 7 evenings from late afternoon up to 5 hr after sunset in this same general study area, plus 5 hr at the “Sod Farms” of Pine Island (Orange Co., NY), 2 hr at Great Swamp National Wildlife Refuge (Morris Co , NJ), and 2 hr at a Hightstown roost (Monmouth Co , NJ). Evening roost arrival times of coexisting North- ern Harriers ( Circus cyaneus) in the Meadowlands site were previously reported (Bosakowski 1983). On 5 different nights, I played prerecorded Short-eared Owl “barking” calls (A Guide to Bird Songs of Eastern and Central North America, Peterson Field Guide Series Record, Houghton Mifflin Co., Boston) at known Short- eared Owl locations for at least 8 min. Taped calls were broadcast at full volume with a portable 7 watt-output cassette tape-recorder placed on the roof of a parked vehicle with observers inside. Results and Discussion Evening Departure. Short-eared Owls were highly crepuscular and nocturnal in the Meadow- lands study area. I observed evening departures on 22 occasions with departures ranging between 28 min before-24 min after sunset, although the ma- jority of owls (45 of 54 = 83.3%) emerged after sunset (Fig. 1) which is very similar to the 81.9% which Clark (1975) observed. On 3 evenings owls departed before sunset, but the sky was heavily over- cast (total sky cover rating 10 of 10, National Weath- er Service). Hendrickson and Swan (1938:585) stat- ed “the birds were observed to hunt on several occasions as early as 3 p.m. on a cloudy day and just before sunset on a clear day.” Despite 8 winters of field work in the Meadow- lands area, I have only observed 1 Short-eared Owl flying during mid-day. However, the owl was being mobbed by an American Crow ( Corvus brachyryn- chos) and may have been flushed. The owl quickly dove for cover and was never observed hunting. At the Great Swamp site I observed 2 owls emerge just after sunset while waiting at a known roost site. At the Hightstown roost 1 owl emerged just after sunset. At the Sod Farms site, I observed from 3-6 owls on 3 evenings. On the first observation day, owls were actively hunting and flying at about 90 min before sunset even though the sky was partly sunny. How- 162 Winter 1989 Short- eared Owls in New Jersey 163 o 1 owl A 2 owls □ 3 owls Y = 0.068X - 15.9 Figure 1. Short-eared Owl roost departure times relative to sunset. Data are combined from the winters 1981-2 and 1982-3. For computation, 12 December was considered day 0 with tic marks displayed at 20-day intervals ever, on 2 subsequent observation days, owls emerged just at or shortly after sunset. I observed a slight tendency for roost departures to occur closer to sunset and greater variability as the season progressed toward spring (Fig. 1). This slight change may be due to increasing day length which shortens the available time period for noc- turnal activities. Clark (1975) also found a similar tendency, but owls in his study area emerged slightly earlier relative to sunset. In either study, evening departure from winter roosts was usually quite close to sunset. These results are in close agreement with studies in other owl species which show that the initial activity or roost departure is closely correlated with time of sunset (Glass and Nielsen 1967; Smith and Murphy 1973; Fuller 1979; Wijnandts 1984). Light cycle has been clearly demonstrated as the primary synchronizer of circadian rhythms and diel cycles (reviewed by Marler and Hamilton 1966) although Tester (1987) has shown some marked seasonal plasticity in activity periods of several free- ranging animal species depending on important eco- logical/ethological events (e.g., care of young, court- ship, hut building). While the Short-eared Owl is frequently diurnal during the nesting season (Bent 1938; Pitelka et al. 1955), my studies have shown the species to be primarily nocturnal in winter which generally agrees with the observations of Clark (1975) and the Craigheads (1956). Although winter roost departure times are often closely allied with sunset, Short-eared Owls occasionally showed some plas- ticity by departing earlier on overcast days (this study) and later during inclement weather (Clark 1975). Interspecific Interactions. Avoidance of inter- ference interactions with diurnal raptors (Jaksic 1982) could be another factor affecting the emer- gence time of owls in the Meadowlands study area. The owls I observed were roosting within 100 m of a communal roost of 6-9 Northern Harriers (Bo- sakowski 1983) and usually emerged just after the last harrier had entered the roost for the night (Fig. 2). It is tempting to speculate that this nearly com- 164 Thomas Bosakowski Vol. 23, No. 4 — — last Northern Harrier arriving — O first Short —eared Owi leaving Figure 2. Minimal temporal overlap between Short-eared Owls and Northern Harriers at nearby winter roosts in 1983. plete lack of temporal overlap was intentional on the part of the owls. Clark (1975) and Watson (1977) have noted frequent agonistic encounters and piracy between harriers and Short-eared Owls. The min- imal temporal overlap observed in the Meadowlands area permitted only 2 interspecific agonistic encoun- ters [1 Northern Harrier, 1 Rough-legged Hawk ( Buteo lagopus )] and suggests a benefit of temporal avoidance. In contrast, at least 3 agonistic encounters with neighboring harriers were observed when owls hunted in the late afternoon at the Sod Farms site. Initial Activity. On 22 evenings at the Mead- owlands site, I observed a total of 54 instances of owls emerging from their ground roosting sites in the marsh. In most cases the initial flight was di- agnostic as owls would suddenly emerge from the reeds and immediately ascend to an altitude of 15- 20 m. At this point, owls would normally engage in a steady direct flight with unchanging altitude, some- times in tandem (11 times), presumably en route to a predetermined hunting site. Only occasionally did hunting begin immediately at the roost departure site. Nocturnal Activity. Kemp (1982) reported that Short-eared Owls have been seen by car headlights hunting in total darkness (no data or source given). Clark (1975) suggested that some hunting takes place at night for he was able to ‘squeak them in’ at various times of the night in areas where he knew birds often hunted. In my study, observations of nocturnal ac- tivities were aided by the use of car headlights, street- lights, flashlights, or scanning for owl silhouettes on moonlit nights or against the glow of city lights on the horizon. I made night observations up to 5 hr after sunset (2230 H) and observed that the owls hunted throughout the period on at least 7 nights. These observations help support the notion that the species is not merely crepuscular, but nocturnal as well. Short-eared Owls used a combination of flying and perch-sitting as hunting methods. On-the-wing hunting was used conspicuously more on windy nights Winter 1989 Short-eared Owls in New Jersey 165 as would be expected on the basis of flight energetics (Schnell 1967). Low coursing flights were made over the reeds, typically 2-4 m above the ground similar to that previously described in detail by Clark (1975). Only 4 observations of hover hunting were observed; 2 owls hovered about 15 m above ground adjacent to streetlights. Clark (1975) observed frequent use of hovering and suggested that this hunting strategy perhaps correlated with low prey density. Lack of suitable perch sites could also induce such behavior, for hovering is often used extensively by wintering Red-tailed Hawks (JButeo jamaicensis ) hunting in open Phragmites marsh or on top of sanitary landfills (pers. obs.). Perch hunting was used for extended periods when wind speed was near zero. Short-eared Owls chose a variety of elevated perch sites (total 33) including bare- topped telephone poles (14), telephone wires (7), saplings (4), fence posts (3), broken-off tree stubs (2), bent steel cable (1), high-tension wire (1), metal sign (1) and once on a Wood Duck (Aix sponsa) box. Often an owl would make several short flights or pounces returning frequently to the same perch. On 1 windless night, a Short-eared Owl was observed perching atop a telephone pole for at least 87 min during which time it made 4 unsuccessful pounces (the owl was still perched when I left at 2007 H). On another windless night, an owl was observed at 5 different perches during a 25 min period and then proceeded to make 7 additional hunting flights from the same telephone pole during the next 75 min period (ending at 2145 H). Clark (1975:35) stated that “Short-eared Owls accomplish an undeter- mined, but probably small, amount of hunting from a perch.” My observations, accomplished before and after dark, suggest that perch-hunting plays a more significant role than previously thought, especially after dark, and could represent the primary hunting method used during periods of low wind velocity. Social Interactions. Few agonistic interactions were observed between owls in the study areas. I observed 1 case of attempted piracy, 1 skirmish at the Sod Farms site on 20 February 1988, and 1 brief encounter, possibly courtship-related, on 10 March 1983 at the Meadowlands site. Similarly, other in- vestigators did not usually notice any significant ag- onistic interactions until March (Short and Drew 1962) or late February when breeding territories are first defended (Clark 1975). In the Sod Farms site up to 6 owls were seen hunting simultaneously be- fore dark with no apparent territoriality. Clark (1975) has noted as many as 6 owls hunting the same 20 ha field in winter. Vocalizations were rarely heard in the field (2 times). However, on 5 different nights, I broadcasted taped Short-eared Owl calls at known Short-eared Owl locations and owls responded on 3 nights. The owls typically responded within 10 s-3 min with several vigorous circling flights over the vehicle. On 2 occasions, owls also responded vocally to the tape by producing the same call — several short series of 3 barking notes (“wrak, wrak, wrak”). Owl response quickly waned in less than 2-3 min and could not be induced again despite continued or peri- odic broadcasting at the same location. On 2 unsuc- cessful nights, the vehicle was in view of a perched owl which only reacted by frequent glancing at the broadcast site. Since the owls I studied did not appear to be territorial, the vigorous response to playback may have been due to social curiosity and mate seek- ing. I am not aware of any previous reports of call playback techniques to detect Short-eared Owls, but the present results suggest that this method could be a useful management tool for population surveys and identifying Short-eared Owl habitat. Acknowledgments I thank J. Benzinger for providing the owl tape and joining me on several of the nighttime observations. I also thank S. and D. Zamos for introducing me to the Sod Farms site and Richard Kane for information on other owl locations. I am also indebted to Richard J. Clark and Dwight G. Smith who carefully reviewed and commented on several drafts of this paper. Literature Cited Bent, A. C. 1938. Life histories of North American birds of prey. U.S. Natl. Mus. Bull. 170. 482 pp. Bosakowski, T. 1983. Density and roosting habits of Northern Harriers wintering in the Hackensack Meadowlands. Rec. NJ. Birds 9:50-54. . 1984. Roost selection and behavior of the Long- eared Owl ( Asio oius) wintering in New Jersey. Raptor Res. 18:137-142. . 1986. Short-eared Owl winter roosting strate- gies. Amer. Birds 40:237-240. Clark, R. J. 1975. A field study of the Short-eared Owl, Asio flarrirrieus (Pontopiddan), in North America. Wild! Monographs 47:1-67. Craighead, J. J. and F. C. Craighead, Jr. 1956 Hawks, owls, and wildlife. Stackpole Publ. Co., Har- risburg, PA. 443 pp. Fuller, M. R. 1979. Spatiotemporal ecology of four sympatric raptors. Ph.D. Thesis, Univ. of Minnesota Glass, M. L. and T. H. Nielsen. 1967. The evening departure of the Long-eared Owl ( Asio otus) from the winter roost. Dansk. Omithol. Foren. Tids. 61:100-106 166 Thomas Bosakowski Vol. 23, No. 4 Hendrickson, G. O. and C. Swan. 1938, Winter notes on the Short-eared Owl. Ecology 19:584-588. Jaksic, F. M. 1982. Inadequacy of activity time as a niche difference: the case of diurnal and nocturnal rap- tors. Oecologia 52:171-175. Kemp, J. B. 1982. Winter roosts and habits of Long- eared and Short-eared Owls. Brit. Birds 75:334-335. Marler, P. and W. J. Hamilton, III. 1966. Mech- anisms of animal behavior. John Wiley and Sons, Inc., New York. 771 pp. Marr, T. G. and D. W. McWhirter. 1982. Differ- ential hunting success in a group of Short-eared Owls. Wilson Bull. 94:82-83. Pitelka, F. A., P. Q. Tomich and G. W. Treichel. 1955. Breeding behavior of jaegers and owls near Barrow, Alaska. Condor 57:3-18. ScHNELL, G. D. 1967. Environmental influence on the incidence of flight in the Rough-legged Hawk. Auk 84: 173-182. Short, H. L. and L. C. Drew. 1962. Observations concerning behavior, feeding and pellets of Short-eared Owls. Amer. Midi. Nat. 67:424-433. Smith, D. G. and J. R. Murphy. 1973. Breeding ecol- ogy of raptors in the eastern Great Basin of Utah. Brigham Young Univ. Biol. Series 18:1-76. Tester, J. R. 1987. Changes in daily activity rhythms of some free-ranging animals in Minnesota. Can. Field- Nat. 101:13-21. Watson, D. 1977. The hen harrier. T. and A. D. Poy- ser, Berhamsted, England. 307 pp. Wijnandts, H. 1984. Ecological energetics of the Long- eared Owl ( Asio otus). Ardea 72:1-92. Department of Biological Sciences, Rutgers Univer- sity, Newark, NJ 07102. Received 22 June 1988; accepted 15 December 1989 J Raptor Res. 23(4):1 67-171 © 1989 The Raptor Research Foundation, Inc. SEXUAL DIFFERENCES IN TIMING OF AMERICAN KESTREL MIGRATION AT HAWK MOUNTAIN SANCTUARY, PA Nancy G. Stotz and Laurie J. Goodrich Abstract. — Bimodal migration patterns occur in many raptor species but have not been conclusively documented for American Kestrels ( Falco sparverius) along their Appalachian migration route. Kestrels migrating past Hawk Mountain Sanctuary, Pennsylvania had a bimodal fall migration pattern when data were averaged over a 26-year period (1963-1988). Peaks at Hawk Mountain centered around 11 September and 2 October. Proportion of males increased slowly over the course of the fall migration; the median date for female kestrels preceded males’ by 11 days. Both males and females showed bimodal migration patterns. Potential factors resulting in such patterns include timing of molt, location of sum- mering grounds, and seasonal weather patterns. Migration patterns of numerous North American raptors exhibit a bimodal distribution. Juvenile Sharp-shinned Hawks ( Accipiter striatus), Cooper’s Hawks {A. cooperii), Northern Goshawks (A. gen- tlin'), and Northern Harriers ( Circus cyaneus) pre- cede conspecific adults in migration (Bildstein et al. 1984; Clark 1985a; Mueller and Berger 1967, 1968; Mueller et al. 1981) and female Merlins {Falco co- lumbarius) migrate earlier than males (Clark 1985b). Female American Kestrels precede males during fall migration at one site in the eastern Great Lakes region (Haugh 1972; Duncan 1985). An analysis of banding data for kestrels during fall migration east of 100° latitude suggests that a two-wave fall mi- gration pattern may extend throughout the eastern United States (Smallwood 1988). However, no ob- servational data document this pattern along specific migration routes other than the Great Lakes flyway. Although kestrels migrating past the Great Lakes and Appalachian Mountains represent 2 different populations with distinct winter ranges (Roest 1957; Clark 1975; Duncan 1985), we cannot assume mi- gration patterns are similar. Banding data alone cannot substantiate migration patterns, since data may have a bias toward birds which are more susceptible to capture (McClure 1984:323-324). Timing of banding, type of lure used, and behavioral differences between sexes (especially if timing of capture susceptibility varies between sexes) could all contribute to an apparent bimodal pattern that does not accurately reflect true migra- tion timing. Patterns seen in banding data and rep- licated with observational data become more believ- able. In this study, we first examine kestrel migration data collected at Hawk Mountain Sanctuary, Penn- sylvania to see whether a bimodal pattern for kestrels exists along Appalachian migration routes. We fol- low this with analysis of counts of known sex kestrels seen at Hawk Mountain, to see if a sexual difference in timing of migration exists. Documentation of such a pattern along the Appalachian migration route would support the hypothesis that separation of sexes must occur on breeding grounds or early in migration (Smallwood 1988). Methods Hawk Mountain Sanctuary, Pennsylvania (40°44'N, 75°50'W), is located on the Kittatinny Ridge, the southern- most line of the northern Appalachians. Geographic and topographic features concentrate tens of thousands of mi- grating raptors along this ridge each fall (Brett 1986). Migration counts have been taken at Hawk Mountain Sanctuary North Lookout (elevation 464 m) since estab- lishment in 1934. Observers record number and species of all migrating raptors seen from the lookout almost daily during the fall. We summed the number of kestrels sighted during each week of fall migration from daily counts made during fall migrations of 1963-1988. We then calculated a weekly average for the 10 most active weeks of kestrel migration, 25 August through 2 November. We omitted observations made before 1963 because of numerous gaps in the data Heavy precipitation usually halted migration and re- sulted in a daily count of zero. We included zero-count days in the calculation of weekly means with the assump- tion that overall weekly migration volume was not affected; kestrels halted by poor weather conditions probably re- sumed migration when weather conditions improved. Weekly means were used to alleviate short-term fluctua- tions in migration volume caused by the passage of weather fronts. Before 1 September, some zero-count days were caused by the absence of an observer at the lookout. Although data for the first week of the migration period may slightly underestimate migration volume, we feel that the differ- ence was minor and did not affect the overall migration 167 168 Nancy G. Stotz and Laurie J. Goodrich Vol. 23, No. 4 DATE Figure 1. Mean number of American Kestrels seen weekly during fall migration at Hawk Moun- tain Sanctuary, 1963-1988. Mid-week dates are given. pattern observed. We included a test of kurtosis (Sokal and Rohlf 1981) to verify non-normality of the data. In 1979 observers at Hawk Mountain began to record sex of kestrels passing the lookout. Sightings were tabu- lated by sex and date of observation. We searched for a difference in the timing of migration by each sex using ranked dates of observation in a Wilcoxon two-sample test (Sokal and Rohlf 1981). To test for a difference in relative numbers of males and females during 2 periods of the migration period, we used a G-test (Sokal and Rohlf 1981) for all sexed birds. G-test analysis was performed twice. First, we used 18 September as the date to divide the migration period; 18 September represented a trough in the bimodal pattern of kestrel migration. Second, we used the median date for all sexed kestrels (22 September) as a dividing date. We compared weekly totals for counts of each sex to the overall migration pattern, again limited to the ten heaviest weeks of kestrel migration. Sampling effort for sexed kestrels varied over the course of the migration pe- riod (<6%->22% of passing kestrels). We adjusted raw numbers of kestrels of each sex observed each week by multiplication with a correction factor for sampling effort, as follows: number of kestrels sexed during the week total number of kestrels seen during the week Results Mean weekly number of kestrels during fall mi- gration at Hawk Mountain peaked on 11 September and 2 October (Fig. 1) when averaged over the 26- year period (1963-1988), representing an extreme platykurtosis (g 2 = —0.796; t s = —17.304, P < 0.001) A bimodal pattern existed in many individual years at Hawk Mountain and was especially clear in 1 963- 1965, 1968, 1971, 1974, 1977, 1985 and 1987 (Fig. 2) - Relative number of males increased gradually during the 10 heaviest weeks of the migration period (N = 793), although both sexes showed a bimodal distribution in timing of migration (Fig. 3). Because of low counts for the final 2 weeks of the period, data for 20 October-2 November were lumped into a single data point for presentation. When all sexed birds were tabulated (N = 837), the G-test revealed a significant difference in the proportions of males and females around the trough date (G = 35.47, P < 0.001). On or before 18 September, females out- numbered males (males =161, females = 201); after 18 September, males outnumbered females by almost 2:1 (males = 309, females = 166). Conclusions did not differ when the median date of all sexed birds (22 September) was used as the dividing date (G = 32.58, P < 0.001). Median date of sighting for female kestrels, 14 September, differed significantly from that of males, 25 September (t s = 6.547, P < 0.001). Discussion American Kestrels migrating past Hawk Moun- tain Sanctuary (1963-1988) had a bimodal migra- tion pattern (Fig. 1). The proportion of male kestrels increased as the fall progressed, indicating a differ- ential timing of migration by each sex. Relative val- Figure 2. Total number of American Kestrels seen during each week of fall migration at Hawk Mountain Sanctuary, 1963-1988. Week numbers correspond to mid-week dates presented in Figure 1. NUMBER Winter 1989 200 -. o- 200 -| o- 200 0 - 200-j 0 - 200 0 - 200 -, 0 - 200 - 0 - 200-1 0 - 200 -, 0 -- Kestrel Migration Patterns 169 1964 1982 WEEK Vol. 23, No. 4 170 Nancy G. Stotz and Laurie J. Goodrich DATE Figure 3. Proportion of male American Kestrels (•) ob- served at Hawk Mountain Sanctuary, relative to all known sex kestrels recorded during each week (1979-1988). Adjusted numbers (see text) of male (A) and female (□) kestrels seen each week of the migration period (1979-1988). ues of these proportions over time should be unaf- fected by the apparent skew toward males in the sexed kestrels. The skew toward males does not ap- pear to represent a bias toward males in sampling for sex, as the pattern for all sex and age classes (Fig. 1) shows a corresponding larger second peak. Our data indicate that a sexual difference in migra- tion timing may be a general pattern common to all kestrels in eastern North America (Duncan 1985; Smallwood 1988). The data may also confirm that different arrival times of sexes to wintering grounds is not an artifact of a local separation of birds on or near wintering grounds, but rather a difference in time of departure from breeding grounds (Small- wood 1988). Sexual difference in migration timing may be re- lated to different roles during the breeding season. Smallwood (1988) suggested that female kestrels are able to initiate migration before males because fe- males complete their molt earlier. Males provide most of the food for their mate and developing young during the breeding season and molt later than fe- males (Willoughby and Cade 1964; Smallwood 1988). Female kestrels are slightly larger than males (Reynolds 1972; Snyder and Wiley 1976), but no ^ sexual dominance is obvious. On wintering grounds m in Florida, both territorial male and female kestrels JO o are able to exclude late-arriving members of either 2 sex (Smallwood 1988). Therefore, females may ben- r efit by migrating as early as their molt permits in “ order to establish territories in favorable habitats before males arrive on wintering grounds (Small- wood 1988). Sexual variation in migration timing o could explain differential habitat use which has been S reported for kestrels (Koplin 1973; Mills 1976; Stin- > son et al. 1981; Bohall-Wood and Collopy 1986; ™ Smallwood 1987, 1988). Differential migration timing by sex may also be related to a difference in distance traveled. Some investigators have reported that male kestrels winter further north than females (Roest 1957; Willoughby and Cade 1964; Johnson and Enderson 1972; Stin- son et al. 1981). Males compete for breeding terri- tories in the spring (Newton 1979), and may be at an advantage to winter on or near breeding terri- tories. Selection may favor males with greater ability to survive in the rigorous habitats of more northern wintering grounds (Mills 1976), and these males may delay their migration relative to females. Sexual difference in migration timing among kes- trels passing Hawk Mountain Sanctuary does not preclude the possibility that juveniles precede adults in migration and may also contribute to the observed bimodal pattern. Juveniles of many raptor species migrate before adults (Mueller and Berger 1967, 1968; Mueller et al. 1981; Bildstein et al. 1984). Analyses of banding data (Smallwood 1988) suggest that juvenile kestrels and adult females precede adult males in migration. We expect the effect of juveniles on the patterns observed at Hawk Mountain to be minimal since the migrations of juvenile raptors are concentrated along the Atlantic coast (Clark 1985b). The striking trough seen in migration counts for both sexes of kestrels passing Hawk Mountain (Fig. 3) may be related to a number of factors. The molt timing hypothesis (Smallwood 1988) described above suggests that non-breeding males, in addition to fe- males, could benefit from an early migration, and the first peak of males (Fig. 3) could represent non- breeding males that did not have their molt delayed by food provisioning. Confirmation would require matching breeding status of migrants to the timing of their migration but would not explain the second peak for females. An alternative explanation might be that the 2 Winter 1989 Kestrel Migration Patterns 171 peaks for each sex represent individuals from sep- arate populations. Verification would require knowledge of summering grounds of kestrels passing Hawk Mountain. This explanation alone does not seem adequate, because the data accordingly suggest that sex ratios of each population are different. The trough could also indicate an overriding an- nual weather pattern which consistently limits kes- trel migration during the third week of September. However, variation in annual patterns (Fig. 2) sug- gests that such a mechanism did not operate every year. A more detailed examination of weather effects at Hawk Mountain is needed. None of our expla- nations excludes any other and observed patterns probably result from a complex interaction of nu- merous factors. Acknowledgments This project was initiated during a research internship sponsored by the Hawk Mountain Sanctuary Association, which provided logistical and financial support. Archbold Expeditions later provided similar support through an internship at the Archbold Biological Station. We are in- debted to R, L. Curry, D. Moskovits and G. E. Woolfen- den for their assistance with data analyses and presenta- tion. The following people offered valuable comments on earlier versions of this manuscript: V. Apanius, J. C. Bed- narz, R. Bowman, J. R. Parrish, S. E. Senner, D. R. Smith, and 3 anonymous reviewers. This project would have been impossible without the patient hours of data collection each fall by the volunteers and staff of Hawk Mountain Sanctuary. Literature Cited Bildstein, K. L., W. S. Clark, D. L. Evans, M. Field, L. Soucy AND E. Henckel. 1984. Sex and age dif- ferences in fall migration of Northern Harriers. J. Field Ornithol. 55:143-150. Bohall-Wood, P. and Collopy, M. W. 1986. Abun- dance and habitat selections of two American Kestrel subspecies in north-central Florida. Auk 103:557-563. Brett, J. J. 1986. The mountain and the migration: a guide to Hawk Mountain. Kutztown Publishing Co., Inc., Kutztown, PA. Clark, W. S. 1975. Analysis of data. Pages 87-97. In Michael Harwood, Ed. Proc. of the North American Hawk Migration Conference, 1974. Shriver Mountain Press, Washington Depot, CT. . 1985a. The migrating Sharp-shinned Hawk at Cape May Point: banding and recovery results. Pages 137-148. In Michael Harwood, Ed. Proc. Hawk Mi- gration Conf. IV. Hawk Migration Association of North America. . 1 985b. Migrations of the Merlin along the coast of New Jersey. Raptor Res. 19:85-93. Duncan, B. W. 1985. American Kestrels banded at Hawk Cliff, Ontario 1972-1983. Ontario Bird Banding 17:35-40. Haugh, J. R. 1972. A study of hawk migration in America. Search 2:1-60. Johnson, D. and J. H. Enderson. 1972. Roadside raptor census in Colorado — Winter 1971-72. Wilson Bull. 84(4):489-490. Koplin, J. R. 1973. Differential habitat use by sexes of American Kestrels wintering in northern California Raptor Res. 7:39-42. McClure, E. 1984. Bird banding. The Boxwood Press, Pacific Grove, CA. MILLS, G. S. 1976. American Kestrel sex ratios and habitat separation. Auk 93:740-748. Mueller, H. C. and D. D. Berger. 1967. Fall mi- gration of Sharp-shinned Hawks. Wilson Bull. 79:397- 425. and . 1968. Sex ratios and measurements of migrant Goshawks. Auk 85:431-436. , and G. Allez. 1981. Age, sex, and seasonal differences in size of Cooper’s Hawks. J. Field Ornithol. 52:112-126. Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, SD. REYNOLDS, R. T. 1972. Sexual dimorphism in accipiter hawks: a new hypothesis. Condor 74:191-197. Roest, A. I. 1957. Notes on the American sparrow hawk. Auk 74(1 ): 1 — 1 9. Smallwood, J. A. 1987. Sexual segregation by habitat in American Kestrels ( Falco sparverius) wintering in south-central Florida: vegetation structure and re- sponses to differential prey availability. Condor 89:842- 849. . 1988. A mechanism of sexual segregation by habitat in American Kestrels ( Falco sparverius ) win- tering in south-central Florida. Auk 1 05(1 ):36— 46. Snyder, N. F. R. and J. W. Wiley. 1976. Sexual size dimorphism in hawks and owls of North America Ornithol. Monogr. 20:1-95. Sokal, R. R. AND F. J. Rohlf. 1981. Biometry: the principles and practice of statistics in biological re- search. W. H. Freeman and Co., San Francisco. Stinson, C. H., D. L. Crawford and J. Lauthner 1981. Sex differences in winter habitat of American Kestrels in Georgia. /. Field Ornithol. 52:29-35. Willoughby, E. J. and T. J. Cade. 1964. Breeding behavior of the American Kestrel (Sparrow Hawk). Living Bird 3:75-96. Hawk Mountain Sanctuary Association, Rt. 2, Kemp- ton, PA 19529. Present address of first author: Northern Arizona University, Department of Bio- logical Sciences, NAU Box 5640, Flagstaff, AZ 86011. Received 10 March 1989; accepted 15 December 1989 J Raptor Res. 23(4):172-175 © 1989 The Raptor Research Foundation, Inc, THE USE OF LINE TRANSECTS TO EVALUATE THE ABUNDANCE OF DIURNAL MAMMALIAN PREY Joan L. Morrison and Patricia L. Kennedy ABSTRACT. — While studying the foraging behavior of accipiters in northern New Mexico, chipmunk abundance was evaluated with a line transect technique. This is a timely, cost-effective method for providing quantitative estimates of the abundance of diurnal small mammal prey in different habitats. The limitations and general applicability of this method to raptor prey studies are discussed. Many feeding ecology studies of diurnal raptors suffer from difficulties in assessing prey abundance in foraging areas. Fitzner et al. (1977) described techniques for determining densities of raptor prey species, emphasizing procedures that would provide quantitative data for comparisons of prey exploita- tion rates and prey densities. Methods traditionally used for collecting these data are time-consuming and labor-intensive. Other problems with these methods have been widely addressed, and no unified approach toward obtaining comprehensive and sys- tematic estimates of small mammal density or pop- ulation size has been developed (Otis et al. 1978). Most density estimation methods also sample rela- tively small areas (<10 ha) and are not practical for determining animal abundance in large foraging areas used by raptors (>100 ha for many species). In addition, precise quantitative density estimates may not be necessary to answer many of the questions addressed in raptor studies. Many studies have attempted to estimate animal abundance by counting all individuals in a known area (Hayne 1949; Krebs 1966; Hirst 1969; Emlen 1971; Franzreb 1981). However, most of these meth- ods were not designed to sample chipmunks ( Tamias sp.) or other diurnal small mammals which are com- mon raptor prey. In addition, species and habitat comparisons may not be possible because the sample area sizes and the relationship between indices of abundance and absolute abundance are difficult to assess (Burnham et al. 1980). Because of these dif- ficulties, many recent papers have suggested that enumeration methods are not sound and should not be used to evaluate animal abundance (Burnham et al. 1981, 1985; Jolly and Dickson 1983; Smith and Brisbin 1984; Montgomery 1987). We have been studying the foraging behavior of a population of Cooper’s Hawks ( Accipiter cooperii ) nesting in northern New Mexico. Cooper’s Hawks feed primarily on medium-sized passerines, wood- peckers, and chipmunks (Kennedy 1985). To eval- uate foraging areas, sampling methods were needed that were suitable for sampling prey populations over large areas in a timely, cost-effective manner and would provide results appropriate for compar- isons. To determine chipmunk abundance in Cooper’s Hawk foraging areas, we modified the line transect method of Burnham et al. (1980, 1981). Burnham et al. (1980) have shown that line transect sampling is practical, relatively inexpensive, and efficient for calculating density estimates, particularly when a study area is stratified by some feature such as hab- itat. Despite the potential usefulness of line transect sampling, it has been infrequently used (Burnham et al. 1980). In this paper we describe our application of this method and evaluate its usefulness in quan- tifying the abundance of diurnal small mammals in different habitats. Study Area The study was conducted in the Jemez Mountains in north-central New Mexico. The study area and the Coo- per’s Hawk nesting habitat are described in detail in Ken- nedy (1988). Methods For comparison of prey abundance between foraging areas, 2 habitats commonly used by Cooper’s Hawks (as determined from radio-tracking data) were examined: mesa tops and canyon bottoms. Mesa tops are dominated by pinyon-juniper ( Pinus edulis-Juniperus sp.) woodland and Gambel oak (Quercus gambeln). Canyon bottoms are char- acterized by large ponderosa pine (Pinus ponderosa), scat- tered Douglas fir (Pseudotsuga menzesii), cottonwood (Pop- ulus sp.), and numerous shrub species. During 1986, prey populations in foraging areas of 5 nesting pairs of Cooper’s Hawks were sampled. Transects of varying lengths (1.61-3.22 km) were established in mesa top and canyon bottom habitats. Transect lengths were determined by the amount of homogeneous habitat [from vegetation maps (Allen 1989)] available for sampling in each foraging area. Transects were run for 3 sampling periods in 1986 which were designed to coincide with the 172 Winter 1989 Small Mammal Prey Abundance 173 late incubation/early nesting period (late May-early June), late nestling/early fledgling dependency period (late June- July), and late fledgling dependency period (August-early September). During 1988, transects varying from 2.90-3.70 km in length were established in mesa top and canyon bottom habitats near 2 Cooper’s Hawk nest sites. In 1988, tran- sects were run during the first and third sampling periods. Total transect lengths established in each habitat for 1986 and 1988 are shown in Table 1. An observer walked along each transect at a continuous pace of about 1.6 km/hr, alternating between slow walking and brief pauses to look and listen. All chipmunks seen or heard along the transect were counted. The type of detection (auditory or visual) was noted with each obser- vation. In the canyon bottom habitat, estimates of the perpendicular distance from the observer to each detection (Burnham et al. 1980) were recorded within each of four distance categories: 0-7.6 m, 7.7-15.2 m, 15.3-22.9 m, and 23.0-30.5 m. In the mesa top habitat, which had more vegetative cover, distances could not be accurately esti- mated beyond 23 m so only the first three distance cate- gories were used. Efforts were made not to count individ- uals more than once. When an observed individual fled, the escape route was monitored to ensure against dupli- cation in counts. Only 1 transect was traversed each day. Sampling guide- lines established for breeding bird transects (Emlen 1971) were followed. One observer conducted all censuses to avoid multi-observer biases (Faanes and Bystrak 1981). Transect counts were conducted only on days with no precipitation, moderate cloud cover (<50%) and low wind speeds (<1 m/sec) (Newman 1959; Robbins 1981). Tran- sects were run for approximately 2 hr and were traversed from 0800-1100 (Verner and Ritter 1986). Chipmunk densities for each sampling period within each habitat were calculated using the exponential poly- nomial estimator in program TRANSECT (Burnham et al. 1980). Computation of the Shapiro-Wilk statistic (W) indicated that the density data were a random sample from a normal distribution (W = 0.95, P = 0.74). Differences in chipmunk density between habitats were evaluated for the data using a paired f-test. All statistics were computed with the SAS Statistical Program (SAS 1985a, 1985b). Results Table 1 shows estimates of chipmunk density within each habitat calculated using the line transect method. Significantly more chipmunks were counted in the canyon bottom habitat than in the mesa top habitat during all sampling periods ( t = 3.37, P — 0 . 02 ). Discussion Our results indicate that this line transect method is suitable for evaluating chipmunk abundance with- in different habitats. However, this method has lim- itations; these and the assumptions addressed by Burnham et al. (1981) should be considered before Table 1. Estimates of chipmunk densities in mesa top and canyon bottom habitats in Cooper’s Hawk hunting areas. Sam- pling Period Habitat Mesa Top Canyon Bottom Tran- sect Length (km) Density (#/ha) (SE) Tran- sect Length (km) Density (#/ha) (SE) 1986 1 12.87 0.45 (0.34) 7.24 2.69 (0.87) 2 14.48 1.82 (0.64) 18.02 1.98 (0.47) 3 11.26 1.32 (0.61) 23.02 2.60 (0.48) 1988 1 9.33 0.79 (0.52) 10.14 4.32 (0.94) 3 9.33 0.90 (0.55) 9.66 3.14 (0.82) Average density (SE) 1.09 (0.25) 2.78 (0.37) this method is used to sample prey populations in raptor studies. Four basic assumptions in line transect sampling were recognized by Burnham et al. (1981): (1) an- imals directly on or very near to the line will always be detected; (2) there is no movement of animals in response to the observer and none are counted more than once during a given walking of the line; (3) all distance data are recorded without measurement error; and (4) sightings of different individuals are statistically independent events. We violated as- sumptions 1 and 2. However, Burnham et al. (1981) indicate that the robustness of the TRANSECT es- timators allows for moderate violations of these as- sumptions. Habitat type influences the level of survey accu- racy because more detectability problems occur in dense, heavily vegetated habitat than in open, sparse- ly vegetated habitat. The screening effect of dense vegetation can result in variable detectability of an- imals near the transect line (violation of assumption 1). Thus, we recommend the line transect method be used primarily in relatively open habitats where the observer can see clearly in all directions to catch quick movements of individuals. In addition to limited visibility in dense habitat, the observer makes more noise walking through thick vegetation. As a consequence, individuals may be frightened and move away from the center line, thus 174 Joan L. Morrison and Patricia L. Kennedy Vol. 23, No. 4 Perpendicular Distance Categories (m) Figure 1. A frequency histogram of chipmunk perpen- dicular distance data collected in the mesa top habitat during sampling period 1 in 1986. escaping detection. Undetected nonrandom move- ment as a result of lower visibility or reaction to the observer is the problem most frequently encountered in applying line transect sampling methods to wild- life populations (Burnham et al. 1981). Such non- random movement tends to increase the perpendic- ular distance of the animal from the line or cause it to be missed. If the animal is spotted, it would be at a point farther from the line than its original posi- tion; this violates assumption 2 (Burnham et al. 1980). The characteristic pattern one can expect to see in the data histogram generated by TRANSECT if evasive movement occurs is shown in Figure 1, a representative frequency histogram of the chipmunk distance data collected in this study. Tests of the robustness of various estimators to animal movement (Burnham et al. 1980) revealed that the exponential polynomial estimator is substantially more robust to movement than other estimators, thus we used it to calculate chipmunk densities. Due to the movement response of chipmunks to the observer, we do not recommend measuring the exact perpendicular distance to each individual from the transect line. Additional observer movements re- sulting from these measurements would increase evasive responses of the sample animals and intro- duce additional errors into the density estimates. Assigning individuals to distance categories during sampling eliminates this problem and the ability to take distance data as grouped greatly extends the applicability of the line transect procedure (Burn- ham et al. 1981). Density estimates can be calculated from grouped distance data and assumption 3 is not violated if there is no error in category assignment (Burnham et al. 1980). Minimizing the number of observers and training them in distance estimation in each habitat prior to sampling will improve estimator accuracy. The number of distance categories should be as large as possible to improve estimator accuracy but not so large that distance estimation errors are introduced. Although we used this line transect method only to assess chipmunk abundance, the method is suit- able for surveying other small, diurnal mammals. When we began this study, we attempted to record all species of diurnal mammals encountered on each transect. To improve consistency of the methodology and thus density estimates (Temple 1981), we rec- ommend that each survey be conducted for 1 or 2 species at a time. Obtaining simultaneous counts for calculating density estimates on gregarious ground- dwelling small mammals [chipmunks or Golden- mantled Ground Squirrels ( Spermophilus lateralis )] is not difficult. This may be more difficult for less detectable mammals such as Rock Squirrels ( S . varie- gatus) which may have a greater flushing distance, and consequently a lower probability of detection along the transect line. To obtain suitable counts for more reclusive species, these species should be sur- veyed separately. In addition, line length and/or the number of distance categories may need to be in- creased. This technique is also suitable for arboreal mam- mals such as Abert’s Squirrels ( Sciurus aberti) or Red Squirrels ( Tamiasciurus hudsonicus ). Surveying the trees and the ground simultaneously is difficult; therefore, we recommend that arboreal species be surveyed at a different time than ground-dwelling species. In summary, our results indicate that within cer- tain limitations, this line transect method can be useful for evaluating prey populations in raptor for- aging areas. Acknowledgments This study was funded by Associated Western Univer- sities Inc.; Los Alamos National Environmental Research Park, Los Alamos National Laboratory; New Mexico De- Winter 1989 Small Mammal Prey Abundance 175 partment of Game and Fish Share with Wildlife Program; and the U.S. Forest Service, Santa Fe National Forest. We are grateful for help from D. Crowe, S. Crowe, B. Edeskuty, W. Gannon, D. Hafner, D. Kimmey, D. Pa- dien, J. Sanderson, R. Smartt, and R. Warren. We also thank K. Burnham for assistance in using program TRANSECT, and J. Brown, G. Smith and an anonymous reviewer for reviewing this manuscript. Literature Cited Allen, C. D. 1989. Changes in the landscape of the Jemez Mountains, New Mexico. Ph.D. dissertation, Univ. Calif. Berkeley, CA. 346 pp. Burnham, K. P., D. R. Anderson and J. L. Laake. 1980. Estimation of density from line transect sam- pling of biological populations. Wildl. Monogr. No. 72. 202 pp. , and . 1981, Line transect esti- mation of bird population density using a Fourier se- ries. Studies in Avian Biology 6:466-482. , AND . 1985. Efficiency and bias in strip and line transect sampling. J. Wildl. Manage. 49:1012-1018. Emlen, J. T. 1971. Population densities of birds derived from transect counts. Auk 88:323-342. Faanes, C, A. AND D. Bystrak. 1981. The role of observer bias in the North American breeding bird survey. Studies in Avian Biology No. 6:353-359. Fitzner, R. E., L. E. Rogers and D. W. Uresk. 1977. Techniques useful for determining raptor prey species abundance. Raptor Res. 11:67-71. Franzreb, K. 1981. The determination of avian den- sities using the variable-strip and fixed-width transect surveying methods. Studies in Avian Biology 6:164-169. Hayne, D. W. 1949. An examination of the strip census method for estimating animal abundance. /. Wildl. Manage. 13:145-157. Hirst, S. M. 1969. Road-strip census techniques for wild ungulates in African woodland. J. Wildl. Manage. 33:40-48. Jolly, G. M. and J. M. Dickson. 1983. The problem of unequal catchability in mark-recapture estimation of small mammal populations. Can. J. Zool. 61:922- 927. Kennedy, P. L. 1985. The nesting ecology of Cooper’s Hawks and Northern Goshawks in the Jemez Moun- tains, New Mexico, 1984 results. Unpubl. Rep. to N.M. Dept. Game and Fish and U.S. Dept. Agric., For. Ser., Santa Fe Natl. For., Santa Fe, NM. 67 pp. . 1988. Habitat characteristics of Cooper’s Hawks and Northern Goshawks nesting in New Mexico. Pages 218-227. In Glinski, R. L. et al., Eds. Proceedings of the Southwest Raptor Management Symposium and Workshop. Natl. Wildl. Fed. Sci. Tech. Ser. No. 11. Natl. Wildl. Fed., Wash., DC. 395 pp. Krebs, C. J. 1966. Demographic changes in fluctuating populations of Microtus californicus. Ecol. Monogr. 36' 239-273. Montgomery, W. I. 1987. The application of capture- mark-recapture methods to the enumeration of small mammal populations. Symp. Zool. Soc. Lond. 58:25-57 Newman, D. E. 1959. Factors influencing the winter roadside count of cottontails. /. Wildl. Manage. 23:290- 294. Otis, D. L., K. P. Burnham, G. C. White and D. R. Anderson. 1978. Statistical inference from capture data on closed animal populations. Wildl. Monogr. No 62. 135 pp. Robbins, C. S. 1981. Bird activity levels related to weather. Studies in Avian Biology 6:301-310. SAS. 1985a. User’s guide: basics, Ver. 5 Ed. SAS In- stitute Inc., Cary, NC. 1290 pp. . 1985b. User’s guide: statistics Ver. 5 Ed. SAS Institute Inc., Cary, NC. 956 pp. Smith, L. M. and I. L. Brisbin, Jr. 1984. An evaluation of total trapline captures as estimates of furbearer abundance. J. Wildl. Manage. 48:1452-1455. Temple, S. A. 1981. Summarizing remarks: estimating relative abundance (Part II). Studies in Avian Biology 6 : 112 . Verner, J. and L. V. Ritter. 1986. Hourly variation in morning point counts of birds. Auk 103:117-124 Los Alamos National Laboratory, Environmental Sci- ence Group, MS J495, Los Alamos, NM 87545 (Cur- rent address: Los Alamos National Laboratory En- vironmental Protection Group Ms K.490, Los Alamos, NM 87545). Address of second author: Biology De- partment, Utah State University, Logan, Utah 84322 (Current address: Los Alamos National Laboratory, Environmental Science Group, MS J495, Los Ala- mos, NM 87545). Received 15 February 1989; accepted 15 December 1989 /. Raptor Res. 23(4):176-178 © 1989 The Raptor Research Foundation, Inc. USE OF EXPLOSIVES TO ENHANCE A PEREGRINE FALCON EYRIE Joel E. Pagel Abstract.— Explosives were used effectively to increase the size and nesting potential of a traditional Peregrine Falcon (Falco peregrinus anatum) eyrie in northern California. Ledge dimensions were increased from 15 cm deep x 31 cm wide to 41 cm deep x 152 cm wide by removal of blast loosened rock. Prior to this enhancement (1982, 1983), nesting attempts were unsuccessful due to the displacement of eyases and eggs from the small eyrie. From April 1984-June 1988, 13 Peregrine eyases (2.6/yr) have successfully fledged from the enhanced eyrie, with no further fatalities induced by ledge dimensions. Nesting sites for many raptors may be artificially produced or altered to increase reproductive poten- tial. Bohm (1977) noted that artificial nesting plat- forms could be used to excise Great Horned Owls (Bubo virginianus ) from natal areas by placing struc- tures in locations where adequate nests were lacking. Fyfe and Armbruster (1977) described the physical excavation of sandstone cliffs for the creation of ar- tificial ledges for Prairie Falcons (Falco mexicanus ) which resulted in an increase of the resident breeding population. Boyce et al. (1980) discussed an artificial nesting structure that had been erected on a historic Prairie Falcon ledge with productive results. Peregrine Falcons (F. peregrinus) show high fi- delity to established nest sites (Hickey and Anderson 1969; Ratcliffe 1980), an aspect of their nesting bi- ology which makes it practical to enhance traditional and/or alternate nest ledges. Boyce et al. (1982) reported a nesting pair of Peregrines that deserted a historical nest site for an alternate ledge enhanced through excavation. Attempts to create nesting platforms in rock faces with explosives have been documented for falcons. Becker (1981) used Kinepeck explosives to create nesting cavities for Prairie Falcons. Smith (1985) reported the use of Detaprime WG (pentaerythri- tioltetranitrate or PETN) to blast potential Prairie Falcon eyries in rock exposed by a large surface coal mine. Investigations of the nesting chronology and re- productive success of a Peregrine Falcon eyrie in California were completed in 1982 and 1983 (Ledig 1982; Lehman 1983) and revealed that eyases and eggs, respectively, had been displaced from a small nest ledge in both years. Herein, I report and de- scribe the successful use of explosives to enhance this traditional Peregrine eyrie in northern California. Materials and Methods During September-December 1983, approximately 120 person-hr were devoted to the enhancement of a Peregrine Falcon ledge in the Klamath National Forest (KNF) of northwestern California. The eyrie was located 23 m above the base of a 36 m dolomitic limestone cliff. The nesting ledge prior to en- hancement measured 15 cm deep x 31 cm wide and was created by a natural horizontal fracture. Personnel entered the existing eyrie following repro- ductive failure in 1983 to assess the potential for enlarge- ment of the eyrie (Lehman 1983). Hand tools proved ineffective to enhance the nest ledge, and explosives were believed to be necessary to enlarge the existing falcon ledge to dimensions regarded as typical (50 x 50 cm) (Ratcliffe 1980). The base of the cliff was located 1 . 1 km from the nearest road and was situated above a 60-70% slope. Consequent- ly, all materials had to be hand-carried to the site. The cliff top and base were easily accessible to climbers. Access to the ledge was gained by rapelling. A drill and explosives were hoisted to the ledge by a z-drag hauling system backed up by reverse jumars (May 1972). Climbers placed 1 piton and 4 Star climbing bolts (9.5 mm x 50 mm) with normal hangers 1-2 m above the ledge to allow greater mobility during preparation procedures. A gasoline powered Cobra percussion rock drill was used to bore 4 holes 60 cm into the rock. A 1 .9 cm rotating hardened steel drill bit was used with good results. Holes were placed 25-30 cm apart directly above the largest portion of the existing ledge. A staggered pattern of drilling facilitated a rhomboidal area of concentrated blast zone (USFS 1980) similar in principal to shaped charges (Smith 1985). High ambient air temp resulted in frequent pauses to prevent the drill bit and engine from overheating. A civilian licensed blaster was contracted to prepare and detonate the charges. The type of explosive charges to be used was decided by the blaster. Ditching dynamite (60% nitroglycerin) in 2.5 x 46 cm cartridges was used in com- bination with electrically initiated, instantaneous blasting caps; 1.5 cartridges per bore hole were used. Thirty m of electric lead line was attached to the blasting cap leg wires and strung up the cliff face to a protected location. Logging roads beneath the blast zone were blocked and unnecessary personnel vacated the site during actual blasting. The 176 Winter 1989 Peregrine Falcon Eyrie Enhancement 177 charges were detonated by touching the ends of the electric lead line to the battery pack of a hand-held 1.5 volt portable radio. Results and Discussion Following detonation of charges, the ledge was examined and excess debris was removed. Upon de- scent to the ledge we noted that 2 charges with blast- ing caps attached had not detonated. These charges were removed and examined manually. We believe that these charges did not explode due to the age of the ditching compound (estimated to be 15+ years old). Though only 2 of 4 charges (1.5 sticks/charge/ bore hole) had exploded, the ledge was enlarged to a sufficient size. Ditching dynamite is a high energy explosive agent which detonates at 5791 m/sec (Du Pont, Inc. 1977). This explosive was not the best to use in dolomitic limestone due to the proximity of the bore holes. If all charges had detonated as intended, the existing ledge may have been destroyed. As ditching dynamite progresses in age, it becomes more unstable and dan- gerous to use. Most licensed blasters are cognizant of this fact. Extreme caution is advised in determin- ing the type and condition of explosives used on similar projects to prevent elimination or destruction of ledge or eyrie. Access to the cliff was an important consideration m the application of explosives on this ledge. This facet makes wilderness or remote enhancements with explosives difficult at best. Bolts affixed to the rock above the eyrie proved crucial in securing climbers, necessary equipment and to serve as placement for suspension of the rock drill during use, or rest breaks. The bolts have been subsequently helpful to climbers entering the eyrie to band young. No rust streaks have occurred to date despite precipitation runoff down the cliff face, and the bolts are invisible to observers from the base of the cliff. Portions of the ledge were loosened to such a de- gree that 34 kg of masonry mortar was required to re-affix some blast debris to the main portion of the ledge. Dimensions of the ledge were found to have been increased to 41 cm x 152 cm. Mortar was allowed to cure, and 45 kg of sifted sand and gravel was spread on the ledge to form adequate substrate for nesting falcons. The eyrie has been examined yearly (1984-1988), with no further work required to replace or repair the mortared rock. The enhanced ledge has been occupied by Pere- grine Falcons yearly since the enhancement with 13 eyases (2.6/year) having fledged between 1984 and 1988 (Pagel 1988). Total investment in this project was $2000. Rough breakdown of these costs is as follows: $300, explosives and blaster’s services; $400, ropes, bolts, ascenders and slings; and $1300 for 120 person-hr of labor. Conclusions Explosives offer an effective means to enhance existing nest ledges of cliff dwelling raptors when hand tools are not effective. Extreme care and fore- thought should be considered at active nest sites. Any work, including preliminary examinations, should not be conducted during the nesting period. Im- provement of alternate nest ledges should be consid- ered before the time and expense of explosive en- hancement is utilized. Blasts more forceful than necessary may decrease the dimensions of the nesting platform or remove the existing nest ledge entirely. Acknowledgments The U.S. Forest Service supported and funded this proj- ect. I thank K. Baldwin, B. Bemis, P. Cilwick, P. Gar- rahan, G. Harper, B. Roberts, A. Scott, and K. Townsend for their energetic assistance and D. Post for reviewing this manuscript. Literature Cited Becker, D. M. 1981. Development of artificial nest sites for prairie falcons. Prog. Rep., Montana Coop. Wildl Res. Unit, Missoula. 8 pp. Bohm, R. T. 1977. Artificial nest platforms for raptors. Raptor Res. 11:97-99. Boyce, D. A., L. Fischer, W. E. Lehman, B. Hipp and J. Peterson. 1980. Prairie falcons nest on an arti- ficial ledge. Raptor Res. 14:46-50. , C. M. White, R. E. F. Escano and W. E Lehman. 1982. Enhancement of cliffs for nesting Peregrine Falcons. Wildl. Soc. Bull. 10:380-381. Du Pont, Inc. 1977. Blasters handbook. Du Pont Tech Serv Section, Explosives Product Div., Wilmington Fyfe, R. W. and H. I. Armbruster. 1977. Raptor research and management in Canada. Pp. 282-293. In R. D. Chancellor, Ed. World conference on birds of prey. Vienna, 1975. Int. Counc. for Bird Pres. Hickey, J. J. and D. W. Anderson. 1969. The Per- egrine Falcon; life history and population literature. PP- 4-42. In J. J. Hickey, Ed. Peregrine falcon pop- ulations; their biology and decline. Univ. Wisconsin Press, Madison. Ledig, D. 1982. Peregrine Falcons status in Klamath National Forest. (Unpubl. ms.) U.S. Fish and Wildl. Serv. 34 pp. Lehman, B. 1983. Peregrine falcon status in the Klam- 178 Joel E. Pagel Vol. 23, No. 4 ath National Forest. (Unpubl. ms.) Santa Cruz Pred- atory Bird Res. Group. 12 pp. May, W. G. 1972. Mountain search and rescue tech- niques. Rocky Mtn. Rescue Group, Inc., Boulder. PAGEL, J. E. 1988. Peregrine falcon status within Cal- ifornia portions of the Klamath, Rogue River and east- ern Six Rivers National Forest. (Unpubl. ms.) Klamath Nat. For. 22 pp. RATCLIFFE, D. A. 1980. The Peregrine Falcon. T. & A. D. Poyser, London. Smith, E. D. 1985. Construction of artificial nesting sites for prairie falcons. Wildl. Soc. Bull. 13:543-546. USFS. 1980. Blaster’s handbook. U.S. Forest Serv., Publ. No. 7109.51. Joel E. Pagel, Klamath National Forest, Happy Camp, CA 96039. Received 24 April 1989; accepted 15 December 1989 J. Raptor Res. 23(4):179-180 © 1989 The Raptor Research Foundation, Inc. Short Communications Range Extension of the Barred Owl in Western Washington and First Breeding Record on the Olympic Peninsula Devora Ukrain Sharp The Barred Owl ( Stnx varia ) has been extending its range west and south for more than 4 decades (Grant 1966; Taylor and Forsman 1976; Boxall and Stepney 1982). The first observation of a Barred Owl in Washington State was made in 1965 in the eastern edge of the state by Rogers (1966). In 1973 the first Barred Owl recorded on the west side of the Cascade Mountains was found dead in Sky- komish, King Co. (Taylor and Forsman 1976). Reichard (1974) found a pair in the northwestern Cascade Moun- tains, and since 1976 several records of Barred Owls nest- ing along the western flanks of the Cascade Mountains have been reported (Allen et al. 1985; Harrington 1985, Wash. Dept. Wildl., pers. comm.). No formal surveys have been conducted on the Olympic Peninsula. However, because Barred Owls respond readi- ly to Spotted Owl (S. occidentalis ) calls (Hamer 1987), surveys for Spotted Owls might effectively be considered Barred Owl surveys. Spotted Owl surveys have been con- 179 180 Short Communications Vol. 23, No. 4 ducted on the Olympic National Forest since 1978 but have elicited no Barred Owl responses. Similarly, Spotted Owl survey results from the Washington Department of Wildlife since 1983 have not reported Barred Owl re- sponses. Postovit (1977) surveyed Spotted Owls along 55 routes, each about 4 km in length, on the Olympic Pen- insula, including 8 drainages in Olympic National Park. No Barred Owl responses were reported. Barred Owls were first observed on the Olympic Pen- insula in 1985 during a 2 yr call survey for Spotted Owls in Olympic National Park by an interagency team with representatives from the Washington Department of Wildlife (WDW), Department of Natural Resources, and Olympic National Park (WDW 1987). Four river drain- ages were call surveyed for Spotted Owls along the west side of Olympic National Park between May and July 1985 (WDW 1987). Responses were obtained from Barred Owls at 4 locations: 2 in the Queets River valley and 2 in Bogachiel valley (Fig. 1). At the Bogachiel response sites Spotted Owls answered either simultaneously, within a half-hour, or within 500 m of the Barred Owls. In 1986 both sites were occupied by Barred Owls, and no Spotted Owls were heard, suggesting that the Spotted Owls might have moved or been displaced. One response site along the Queets Valley road was occupied by a pair of Barred Owls that produced 3 young in 1985 and 1986. Responses in Olympic National Park in 1987 (Sharp et al. 1987) were from river terrace or lowland areas. Range expansion by the Barred Owl might be a threat to the Northern Spotted Owl, whose numbers are declining due to fragmentation and loss of old-growth habitat (Mar- cot and Gardetto 1980; Forsman et al. 1984; Simberloff 1987). Barred Owls are known to occupy a greater variety of habitats, and have several adaptive breeding strategies and behavioral patterns that might cause competition with the Spotted Owl (Hamer 1987). Acknowledgments Kate Foster and Chuck Sisco discovered and identified the first Barred Owls on the west side of the Olympic Peninsula. Eric Forsman, Bruce Moorhead and L. Scott Mills reviewed the manuscript. Literature Cited Allen, H. L., T. Hamer and L. W. Brewer. 1985. Range overlap of the Spotted Owl ( Strix occidentalis caurina) and the Barred Owl ( Strix varia) in Wash- ington and implications for the future. In Raptor Re- search Foundation Symposium on the Management of Birds of Prey, Session 11: The biology, status, and management of owls. November 9-10, 1985, Sacra- mento, CA. (Abstr.) Boxall, P. C. and P. H. R. Stepney. 1982. The dis- tribution and status of the Barred Owl in Alberta. Can, Field Nat. 96:46-50. Forsman, E. D., E. C. Meslow and H. M. Wright. 1984. Distribution and biology of the Spotted Owl in Oregon. Wildlife Monog. no. 87. Grant, J. 1966. The Barred Owl in British Columbia Murrelet 47:39-45. Hamer, T. 1987. Interspecific competition between Barred and Spotted Owls in western Washington. The Raptor Research Foundation Annual Meeting, Octo- ber 28-31, 1987, Boise, Idaho. (Abstr.). Harrington-Tweit, B., and P. W. Mattacks, Jr. 1985. Northern Pacific Coast Region. Amer. Birds 39 - 953-956. Marcot, B. G. and J. Gardetto. 1980. Status of the Spotted Owl in the Six Rivers National Forest, Cali- fornia. Western Birds 11:130-146. Postovit, H. R. 1977. A survey of the Spotted Owl in northwest Washington. National Forest Products As- sociation, Washington, D.C. Reichard, T. A. 1974. Barred Owl sightings in Wash- ington. Western Birds 5:138-140. Rogers, T. H. 1966. The fall migration. Northern Rocky Mountain-intermountain region. Aud. Field Notes 20: 74. Sharp, D. U., K. H. Foster and B. Moorhead. 1987 Inventory and monitoring of Spotted and Barred Owls, Olympic National Park, 1987 progress report. Olym- pic National Park. Simberloff, D. 1987. The Spotted Owl fracas: mixing academic, applied, and political ecology. Ecology 68 766-772. Taylor, A. L. and E. D. Forsman. 1976. Recent range extensions of the Barred Owl in western North Amer- ica, including the first records for Oregon. Condor 78 560-561. Washington Department of Wildlife (WDW). 1987 A survey of Spotted Owls in western Olympic National Park. Wash. Dept, of Wildlife, Olympia. Olympic National Park, Natural Science Studies, Port Angeles, WA 98362. Present address: Everglades National Park, P.O. Box 279, Homestead, FL 33030. Received 1 May 1988; accepted 15 August 1989 Winter 1989 Short Communications 181 J Raptor Res. 23(4):181-182 © 1989 The Raptor Research Foundation, Inc. Peregrine Falcon Takes Black-bellied Plover from Sea Off Kenya Jennifer F. M. Horne and Lester Short At 0645 H on 8 January 1989 we noticed a Peregrine Falcon ( Falco peregrinus ) in flight Vi km from shore at Watamu, Kenya. The falcon was making rather slow, circling swoops between us ashore and a grassy, rocky islet about 1 km east of the former home of the late falconiform specialist Leslie Brown. We then observed a bird in the water around which the falcon flew. The bird in the sea attempted to fly and actually rose from the water for 1-2 m, then crashed down. The falcon increased its speed and swooped at the bird, which raised 1 or both wings and lowered its head at the approach of the falcon. The bird in the water appeared waterlogged and injured, but for 10 min or so it made movements (beating wings, head and neck thrust under water) to evade the Peregrine, whose stoops became more regular and in tighter circles. At 0658 H we began counting the Peregrine’s swoops, and noted that the prospective prey, which likely had been attacked by the falcon and injured before we arrived at the beach, was less active, only ducking its head and mov- ing its wings on the water. By 0707 H we had counted 75 stoops of the Peregrine, the “circles” over the bird had become smaller, more elliptical, and from one side and then the other (instead of full swoops around and back from only one direction). At 0710 H the Peregrine adroitly lifted the bird from the sea, which was moderate but not rough in a light wind. The Peregrine made a half circle, clutching the bird, about 5 m over the water, then headed momentarily toward the islet before reversing and flying back (west) toward us, with the east (inshore) wind behind it The falcon landed on a rock 100 m south of us, near the shore, and began plucking feathers from its prey. The Peregrine likely was a subadult, as it was mainly barred below with some streaking, and pale above. We approached the falcon slowly at 0711 H. One min later it carried the prey 30 m farther south to a larger rock, where it continued plucking and began to feed. We had seen feathers fall at the first site, so we now left the beach, hoping the falcon would finish its meal and leave enough remains for us to identify the rather small bird it had taken. We returned about 20 minutes later at 0735 H, the Peregrine was gone. We sought remains of the prey, finding: a) at the first site body feathers and the head of a Black-bellied Plover ( Charadrius squatarola ); and, b) at the second, intestines, some tail and body feathers and the 2 wings, attached by some flesh, bearing the diagnostic black axillary feathers of the species. We surmise that the plover, common at Watamu in the boreal winter, had been flying north or south parallel to the coast, where it was attacked by the Peregrine (likely hunting out from the islet). It must have been injured when struck, likely just over the sea (this plover usually flies close to the water), into which it fell before the Peregrine could retrieve it. The actual kills of wild Peregrines are rarely witnessed (Ratcliffe 1980:128). It is clear (Brown 1976; Brown et al. 1982) that most prey of Peregrines is taken in the air. Experienced (adult) Peregrines are adept at shepherding prospective landbird prey out over the water (although they may pursue it toward land when the prey tires and becomes more vulnerable), and water birds, inland (Hunt et al. 1975; Treleaven 1977; Sherrod 1983). Beebe (1960) studied breeding Peregrines of the north- western North American coast, and found that they preyed almost exclusively on small alcids and storm-petrels that were attacked mainly over land and the shoreline as the prey headed from ground nests toward the sea. His many observations include only one hunting episode over the sea itself. Peregrines are apt to avoid getting wet or plunging into the sea, especially after heavy prey that has fallen. Walter (1979:96), concerning the related Eleonora’s Fal- con ( Falco eleonorae ) wrote that they mainly hunt small landbirds over the sea and “avoid touching the salt water ” Prey, other than very small land birds that can be easily plucked from the surface, pose problems of the Peregrine getting wet, and attracting the attention of competitive large gulls (e.g., Larus marinusj or, in Kenya, Larus hem - prichii ) and sea eagles (in Kenya, the large, coastal African Fish-eagle, Haliaeetus vocifer ) that can enter the water and might prey on the falcon (see Rudebeck 1953:212, cases 12 and 17). Actual accounts of “plucking” from water by the Peregrines are few. Rudebeck (1953:210) noted a dead starling ( Sturnus vulgaris ) taken from the sea; Cade’s works (1960, 1982) mention only a jay ( Perisoreus canadensis ) taken from a river from which it was “floundering out” (Cade 1960:215), a longspur ( Calcarius sp.) plucked from a river by a large female Peregrine, and a grayling fish ( Thymallus arcticus) grasped by a Peregrine as it was “breaking the surface” of the water. Ratcliffe (1980:156) stated that British coastal breeding Peregrines “are mostly careful not to knock their prey into the sea.” The persistence of the Peregrine we observed is notable. Rudebeck (1953) gave no numbers, but mentioned that Peregrines rarely stooped at fleeing prey more than 50 182 Short Communications Vol. 23, No. 4 times. Although some Peregrines may be persistent in actual chases, we have found no descriptions of repeated attacks on live prey that they have knocked into the sea. It is well known that Peregrines may burst into a flock, e.g., of shorebirds, then select and pursue a single indi- vidual, but in one-on-one attacks many shorebirds dodge well and evade Peregrines (Hunt et al. 1975:121). Tre- leaven (1977) noted that young Peregrines in the autumn concentrate on snatching unsuspecting waders from the edges of flocks feeding on the shoreline. Black-bellied Plovers in coastal Kenya tend to fly alone, and not high above the sea, into which they can go as a last resort (not diving directly, as, e.g., can alcids). The plover we observed, whether it had been wounded or driven into the sea by the Peregrine, never actually went beneath the surface, nor did it defend itself. Rather, all its move- ments were to evade the falcon by lowering its head and fluttering away when the falcon stooped, and to attempt to get airborne. It did not move any distance under its own power during the course of our observations. Our observations suggest that the killing and plucking of a 1 90- 200 g (Cade 1960) Black-bellied Plover presented a major problem to the Peregrine. Its persistence, however, paid off. Acknowledgments We thank Dean Amadon and anonymous referees for helpful comments on this manuscript. Literature Cited Beebe, F. L. 1960. The marine peregrines of the North- west Pacific Coast. Condor 62:145-189. Brown, L. H. 1976. British birds of prey. Collins, Lon- don. pp. XIII + 400. , E. K. Urban and K. Newman. 1982. The birds of Africa. Vol. I. Academic Press, London. XIII + 521 pp. Cade, T. J. 1960. Ecology of the Peregrine and Gyr- falcon populations in Alaska. Univ. Calij. Publ. Zool 63:151-290. . 1982. The falcons of the world. Cornell Univ Press, Ithaca, NY. 192 pp. Hunt, W. G., R. R. Rogers and D. J. Slowe. 1975 Migratory and foraging behavior of Peregrine Falcons on the Texas Coast. Can. Field-Nat. 89:111-123. Ratcliffe, D. 1980. The Peregrine Falcon. Buteo Books, Vermillion, SD. 416 pp. Rudebeck, G. 1953. The choice of prey and modes of hunting of predatory birds with special reference to their selective effect. Oikos 3(11): 199-231. Sherrod, S. K. 1983. Behavior of fledgling Peregrines. Peregrine Fund, Ithaca, NY. XI + 202 pp. Treleaven, R. B. 1977. Peregrine. Headland Publ , Penzance, U.K. 152 pp. Walter, H. 1979. Eleonora’s Falcon. Univ. Chicago Press, Chicago, IL. XIII + 410 pp. National Museums of Kenya, Box 40658, Nairobi, Kenya, and American Museum of Natural History, New York, NY 10024-5192. Received 21 April 1989; accepted 15 December 1989 / Raptor Res. 23(4):182-183 © 1989 The Raptor Research Foundation, Inc. Northern Harrier ( Circus cyaneus) Predation of Lesser Prairie-Chicken {Tympanuchus pallidicinctus) David A. Haukos and Gerald S. Broda The Northern Harrier {Circus cyaneus) is not considered a significant predator of Lesser Prairie-Chicken {Tym- panuchus pallidicinctus) (Palmer et al. 1988). However, Campbell (1950) reported that Northern Harriers will harass lekking Lesser Prairie-Chickens by repeated lek flushes and an occasional attack. The objective of this paper is to report and describe 5 cases of Northern Harrier predation on Lesser Prairie-Chicken in west Texas. Throughout 2 spring lekking periods (1987 and 1988), we observed and recorded interactions of Northern Har- riers with Lesser Prairie-Chickens during >750 hrs of lek observation (Haukos 1988). The study area was in Coch- ran and Yoakum Counties, Texas. The area has a sandy duned topography supporting range cattle production; with dominant vegetation of sand shinnery oak {Quercus ha- vardii ), sand sagebrush {Artemisia Jilifolia), and bluestem (. Andropogon spp.) plant communities. Other raptors in the area include Golden Eagle {Aquila chrysaetos), Cooper’s Hawk {Accipiter cooperii). Red-tailed Hawk {Buteo ja- maicensis), Swainson’s Hawk {B. swainsoni), Rough-legged Hawk {B. lagopus). Ferruginous Hawk {B. regalis), and Prairie Falcon {Falco mexicanus). The number and com- position of raptors on the study site fluctuates greatly during the lekking period as raptors migrate through the area. However, harriers have been observed ground-nest- ing on the study site (pers. obser.). Winter 1989 Short Communications 183 Throughout the study, harriers approached and ha- rassed (caused lek display to cease) Lesser Prairie-Chick- ens on leks. Northern Harriers, alone (females) or in pairs (female and male), quartered leks and caused a full lek flush on 3 of 67 observed lek approaches (4%). During the morning display period (Vi hr bef ore-2 hrs after sun- rise), observed harriers quartered only leks and not areas between leks, causing partial lek flushes in 30% (20 of 67 approaches) of lek approaches. Male prairie-chickens usu- ally ceased displaying when harriers quartered over leks. However, no attacks by harriers on Lesser Prairie-Chick- ens were observed at leks. All observed harrier predation of prairie-chickens were on birds away from leks, but within 100 m of a lek. On 10 March 1988, a live juvenile female Lesser Prairie- Chicken was recovered from a female Northern Harrier. The prairie-chicken had been captured on the side of a sand dune dominated by sand shinnery oak. A male harrier was present on the ground near the site. The female harrier was attempting to control the prairie-chicken by standing on her back; however, the prairie-chicken was able to travel several m (<10) after the initial attack while the harrier w r as grasping her. We did not observe any male harrier assistance in subduing the prairie-chicken. Our presence caused the harriers to flush and we were able to retrieve the prairie-chicken. Upon examination, the prairie-chick- en had suffered no severe damage (i.e., flight capable) except a puncture wound just below the nape between the wings caused by the raptor’s talons. The area immediately surrounding the wound was free of feathers. Prairie-chick- en feathers were scattered around the capture site indi- cating a struggle or plucking by the harrier. On 23 February 1988 a female harrier was flushed from a recently killed adult male prairie-chicken. Feathers were scattered around the site. An area bare of feathers was found below the nape and the carcass had been decapitated. On 19 March 1988, a female harrier was flushed from a fresh carcass of a juvenile female prairie-chicken. Prai- rie-chicken feathers were scattered over the kill area and the carcass had been decapitated. Evidence at the site in- dicated that the carcass had been dragged from the kill site 10 m into bluestem grass cover, but whether the harrier dragged the carcass is unknown. On 23 February 1 987, an adult male harrier was flushed from a carcass, no more than 2 d old (the carcass was not near the location 2 d previous), of an adult male prairie- chicken. A scattering of feathers was present in an area approximately 7 m in diameter. The carcass had been decapitated. On 16 March 1988, a juvenile male prairie-chicken was trapped which had talon wounds along with an area free of feathers below the nape. We determined that the bird had likely survived a Northern Harrier attack and was in good condition with no obvious difficulty in flight. Examination of all birds and carcasses found with Northern Harriers had the following characteristics: (1) a talon wound on the neck which was surrounded by a small area free of feathers; (2) carcasses were decapitated; and (3) prairie-chicken feathers were scattered around the area representing either a struggle during the kill or pluck- ing by the harrier. We suggest that harriers are probably unable to kill Lesser Prairie-Chickens in flight or deliver death blows with talons. Harriers do appear, however, to be capable of pinning Lesser Prairie-Chickens to the ground and successfully decapitating them. Northern Harriers have been studied in the presence of Greater and Attwater’s Prairie-Chicken (T. cupido ) Lehmann (1941) noted that harriers repeatedly flushed and dove at Attwater’s Prairie-Chicken; no kills were re- ported. Berger et al. (1962) reported that harriers caused a full lek flush on 38% of all approaches (886) and a partial lek flush on 30% of all approaches. In comparison to this study, harriers cause a full lek flush more frequently in Greater Prairie-Chickens, but partial lek flushes are the same between the 2 studies. Berger et al. (1962) re- ported 1 observed kill of a Greater Prairie-Chicken by harriers and concluded that “harrier harrassment is or- dinarily a matter of little consequence.” In all likelihood, few Lesser Prairie-Chickens are prob- ably killed directly by Northern Harriers. However, sub- sequent mortality by survivors of harrier attacks may oc- cur. Attacked birds may be in a weakened state and subject to a higher probability of predation from other avian and mammalian predators. Acknowledgments L. M. Smith, J. F. Bergan, and M. T. Merendino provided helpful comments on drafts of the manuscript. Financial support was provided by the Caesar Kleberg Wildlife Institute and the Houston Livestock Show and Rodeo. This is manuscript T-9-582, College of Agricul- tural Sciences, Texas Tech University. Literature Cited Berger, D. D., F. Hamerstrom and F. N. Hamer- STROM. 1963. The effect of raptors on prairie chick- ens on booming grounds. /. Wildl. Manage. 27:778 791. Campbell, H. 1950. Note on the behavior of marsh hawks toward lesser prairie chickens. J. Wildl. Manage. 14:477-478. Haukos, D. A. 1988. Reproductive ecology of lesser prairie-chickens in west Texas. M.S. Thesis. Texas Tech Univ., Lubbock. 82 pp. Lehmann, V. W. 1941. Attwater’s prairie chicken, its life history and management. USDI Fish Wildl. Serv. N. Am. Fauna 57. 65 pp. Palmer, R. S., K. L. Bildstein and J. B. Gollop. 1988. Northern Harrier. Pages 261-303. In R. S. Palmer, Ed. Handbook of North American birds. Volume 4. 433 pp. Department of Range and Wildlife Management, P.O. Box 4169, Texas Tech University, Lubbock, TX 79409. Received 20 February 1989; Accepted 15 December 1989 184 Short Communications Vol. 23, No. 4 /. Raptor Res. 23(4): 184 © 1989 The Raptor Research Foundation, Inc. An Instance of Carrion- Feeding by the Peregrine Falcon ( Falco peregrinus) Dan C. Holland Diet of the Peregrine Falcon {Falco peregrinus) consists primarily of birds (see Bent 1937; Hickey 1969; Palmer 1988), but predation on mammals has also been observed (Fischer 1968; C. White, pers. comm.). Few reports de- scribe the use of carrion by these falcons (Beebe 1960; Ratcliffe 1980; C. White, pers. comm.), an instance of which I report here. Observations were made on 11-12 July 1988 at the U.S. Fish and Wildlife Service San Sim- eon field station at Piedras Blancas, San Luis Obispo County, California. Activity was observed from a distance of 114m with a 60 x Bausch and Lomb spotting scope. A single falcon in juvenile plumage was first seen perched on a telephone pole near the station access road at 0910 H on 12 June. The bird made 5 short (8-10 m height) stoops on the carcass of a juvenile California Ground Squirrel {Otospermophilus beechyi ) laying in the road. The squirrel was killed by a vehicle at approximately 1730 H on 1 1 June. At 0920 H the bird seized the carcass and dragged it to the road’s edge, and after a brief pause, moved it off the road. The bird mantled the carcass and began feeding at 0924, making a total of 47 feeding movements in a 3 min period followed by 3 min of inactivity. At 0930 activity resumed and 111 feeding movements were observed in a 25 min period. The bird ceased feeding until 1036, and made another 21 feeding movements before being fright- ened off the carcass at 1038 by a passing vehicle. The falcon had not returned to the carcass by 1108, at which time the squirrel was collected. The carcass had been moved 3 m, and the remains weighed approximately 150 g. The head was missing, and most of the pelt in the left axillary area had been removed. Extensive tissue removal was noted from the deltoideus, rhomboideus, triceps and biceps. Acknowledgments I thank C. Bird, J. Parrish, D. Pashley and C. White for reviewing this note. Literature Cited Beebe, F. L. 1960. The marine peregrines of the north- west Pacific coast. Condor 62:145-189. Bent, A. C. 1937. Life histories of North American birds of prey. Vol. II. U.S. Natl. Museum Bull. 170. FlSCHER, W. 1968. Der Wanderfalk. Neu Brehm-Bii- cherei. No. 380. Hickey, J. (Ed.). 1969. Peregrine Falcon populations: Their biology and decline. Univ. Wisconsin Press, Madison, WI. Palmer, R. S. 1988. Handbook of North American birds. Vol. V. Yale Univ. Press, New Haven, CT. Ratcliffe, D. 1980. Peregrine Falcon. Buteo Books, Vermillion, SD. Department of Biology, University of Southwestern Louisiana, Lafayette, Louisiana 70504-2451. Received 1 May 1989; accepted 15 December 1989 J. Raptor Res. 23(4):185 © 1989 The Raptor Research Foundation, Inc. Thesis Abstracts Characteristics of Breeding Habitat for Peregrine Falcons on the Colorado Plateau The endangered Peregrine Falcon ( Falco peregrinus ) has experienced large scale population declines in much of its cosmopolitan range. Although the Peregrine is one of the most widely studied birds, breeding habitat requirements have not been critically evaluated, especially in the Rocky Mountain states. Data from 65 historically and currently active eyries were analyzed to describe and quantify habitat parameters for breeding Peregrines on the Colorado Plateau. Surveys for active eyries were concentrated in national parks between 1984-1986. A total of 144 variables (e.g., topography, hydrology, vegetation, competing species, cliff characteristics, human disturbance factors, etc.) were measured and rated for each eyrie. Physiographic parameters, trends in eyrie occupancy, and reproductive success were also examined. Correlations between eyries were observed in cliff height and distance to permanent water. Computer programs (SAS) were used to analyze individual variables. This data base is useful for locating new breeding sites and for potential reintroduction of Peregrines at historical sites. This data base will be useful in future years for standardizing and correlating new information. Teresa, Sherry, Master of Arts in Geography, University of California, Los Angeles, 1988. Professor Hartmut Walter, Chair. Reversed Size Dimorphism, Determinants of Intraspecific Dominance, and Habitat Use by American Kestrels {Falco sparvenus ) Wintering in South Carolina This study examines the importance of sex, age, weight, and residency on the dominance status of wintering American Kestrels {Falco sparverius ), examines habitat segregation by sex in kestrels wintering in South Carolina and Maryland, and experimentally tests the female dominance hypothesis proposed to explain reversed size dimorphism in raptors Kestrels were randomly paired in an arena containing food (a mouse) either in aviaries at Clemson University or in the field. During 20 minute observation periods, all interactions between birds were recorded. Females tended to dominate males, and heavier birds tended to dominate lighter ones. The female dominance hypothesis was supported Although sample sizes were too small for strong conclusions, there were indications that age and residency are also important determinants of dominance. Field observations revealed no habitat segregation, either in South Carolina, or m Maryland, but did suggest strong non-random habitat use. Both sexes were found significantly more in open habitats, such as plowed fields, pastures, and meadows, and significantly less often in wooded and residential areas. Aborn, David. A. 1989. M.Sc. Thesis, Dept, of Biological Sciences, Clemson University, Clemson, SC 29634-190 3. Thesis Advisor: Dr. Sidney A. Gauthreaux, Jr. Raptors of El Imposible Forest, El Salvador, C.A. The purposes of this research were to determine raptor species composition and obtain baseline natural history data Thirty species of resident and migratory raptors were observed. Representatives of 6 species were mist-netted and banded. Data were recorded on the nesting of 13 species, and 111 hr of observations were recorded on a King Vulture {Sarcoramphus papa) nest. Four of the observed species were listed as endangered, five threatened, and four previously unrecorded in El Salvador. West, Jane Noll. 1988. M.Sc. Thesis, Department of Biology, Central Washington University, Ellensburg, WA 98926. 185 J. Raptor Res. 23(4): 186 © 1989 The Raptor Research Foundation, Inc. News and Reviews A Dictionary of Ethology by Klaus Immelmann and Colin Beer. Harvard University Press, Cambridge, Massachusetts and London, England, 1989. 336 pp. Cost, hardcover, $35.00 U.S. We have good news for fellow lexicophiles (i.e., those who compulsively browse through dictionaries), especially those with a behavioral bent. Over the last 2 decades, several volumes defining the vernacular of ethology have appeared on the market. The publication listed above (hereafter, Dictionary) is the most recent book attempting to clarify the terms and concepts of animal behavior. This volume is, in our opinion, the most useful to students of ethology. In a text that is well-organized, attractively illustrated (about 20 line and stipple drawings), and largely error-free, the authors present about 700 terms in clear concise language. They provide not only a definition, but also examples, information on the origin of the term, very often a synonym, and, even more useful, an evaluation of the current value of each term. Classifying the book as a dictionary somewhat disguises its full value. The volume includes much more information than would be expected even in an unabridged dictionary. Rather, the tome appears to be a powerful attempt to canonize the preferred terminology of ethology. Far beyond merely defining terms, it provides a commentary on which terms have come into general use, which have not, and which are so abused in the literature (e.g., instinct, motivation) that they are now of little use to the scientist. The book also provides definitions and clarifications of terms which even the serious student of ethology has probably given little thought. To illustrate the Dictionary’s value as a clarifier of appropriate uses for specific terms, let’s examine the term, “interaction”. Define the term in your own mind and state all appropriate uses as you perceive them. Next, compare your perceptions with the 3 uses presented by the authors: (1) an encounter between animals, (2) an influence of an animal on its environment or vice versa, and (3) in a statistical sense, the synergistic influence of 2 or more variables on an animal or its environment. Now, decide if you have not in fact sharpened and deepened your own perceptions of the ethological concept. With the dictionary in hand, of course, this exercise is much more effective as you examine, in this case, 13 lines of text rather than our terse summary of the 3 uses of the term. As much as we recommend the book, in our role as critics, we must state a few minor concerns. First, this volume covers a wider range of terms than would be expected for an ethological dictionary. It is surprising to see that the authors have included definitions for such basic biological terms as gamete, embryo, fossil, FSH, and many more Second, as students of raptor ecology, we take exception to the authors citing raptors as prime examples of species for which the adults commit infanticide (i.e., the authors interpret nestling neglect [passive] as a form of infanticide [active]) Third, although over 75 publications from the primary and secondary literature are cited in the text, most terms and concepts, even when lengthy, remain uncited. Perhaps this omission was essential in controlling the length of the text, but the result of this exclusion is that the book is much less useful as a means of entering the ethological literature. Finally, Beer states in the preface that the authors excluded highly technical terms (e.g., the names/titles of specific behavioral acts) to make the book more useful to the non-specialist. Fortunately, titles and definitions of a few of these specialized terms are available in Heymer’s Ethological Dictionary (Verlag Paul Parey, Berlin and Hamburg, 1977) In our opinion, however, Immelmann and Beer missed a propitious opportunity. The time is ripe for the presentation of standard terminology naming homologous behavioral acts (ethons) that are common to different taxonomic groups. Standardization is possible for some ethons at least as high as the Class level. Shaking (of the body), various feeding ethons, and many more could be linked by common terminology. A few ethons (e.g., yawning, stretching, and running as performed by different taxonomic Classes) are so obviously related that their titles have received de facto standardization even without the ethologist’s help. Many more terms could also be standardized but the authors have refused to do so on the grounds that including highly specialized terms would “overload” the text. They canonize instead such curious terms as “shake to death.” Even with this omission, we consider the Dictionary invaluable in clarifying ethological concepts and facilitating communication between ethologists and other scientists in related fields. Immelmann and Beer have provided a powerful tool for refining concepts and thereby generating original thought, the mechanism of discovery in any branch of science. David H. Ellis and Linda J. Miller, U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center, Laurel, MD 20708. 186 Winter 1989 News and Reviews 187 James R. Koplin Travel Award. In honor of Dr. James R. Koplin, the Raptor Research Foundation, Inc., established an endowment to support the travel of students to annual RRF meetings. Dr. Koplin was a professor of wildlife management at Humbolt State University, in Areata, California, and had a long-term interest in the study of raptor distribution and abundance. All students are eligible to apply for this award; however, because of Dr. Koplin’s interest in Latin American raptors, preference will be given to applicants from the neotropics. This program was initiated at the 1989 RRF meeting held in Veracruz, Mexico, and consists of a $100 award. In the future, as the endowment permits, more awards in greater amounts will be provided. Selection of the recipients of this award will be based on the quality of the research conducted, financial need, and professional potential as a raptor biologist. To be eligible for the Koplin Travel Award, a student must be a member of the Raptor Research Foundation, Inc., and the senior author on a paper to be presented at the meeting for which travel funds are requested. To apply for the 1990 award, students must submit their request to the J. R. Koplin Travel Awards Committee Chairperson by 1 September 1990. Application materials include: (1) a two- page (double-spaced) description of the research results to be presented at the meeting; (2) an itemized budget of the total cost associated with attending the meeting, accompanied by an explanation of how the expenses not covered by this award will be met; and (3) a letter of recommendation. This letter is very important in the selection process and should be written by the student’s major professor. It should contain a thorough evaluation of the applicant’s academic abilities, the significance of the research being reported, his or her contribution to this research, and the potential for future contributions by the student to the field. Send application to Michael W. Collopy, James R. Koplin Travel Award Committee, % Dept, of Wildlife and Range Sciences, 118 Newins- Ziegler Hall, 0304 IF AS, University of Florida, Gainesville, Florida 32611-0304 USA. Course Study on Birds of Prey. The Institute of Conservation and Ecology at the University of Kent at Canterbury is offering a 2-yr part-time course in the study of birds of prey leading to an advanced university diploma in raptor biology. The program is comprised of 2 parts: a 100-hr instructed course dealing with all aspects of raptor biology, and an individual study project. The course is designed for anyone with a genuine interest in raptors including enthusiastic amateurs, curators of raptor collections and conservation workers. Part I will commence in the 1989/90 academic year. For further details on the program outline, admission requirements and fees please write to: Mike Nicholls, Christ Church College, Canterbury CT1 1QU, U.K. Biology & Conservation of Small Falcons A three day Conference under the aegis of The Hawk Trust and hosted by the Durrell Institute of Conservation and Ecology at the University of Kent at Canterbury, England on 6-8 September 1991 The study of the smaller falcons is often neglected in favour of that of the larger species. The theme of the Conference will therefore be aspects of the biology and conservation of kestrels, merlins, hobbies and other small raptors, particularly, but not exclusively, in the genus Faico. Topics to be covered at the conference will include aspects of taxonomy and evolution, ecology and behaviour, health and disease, captive breeding and associated research. Further offers of papers and poster demonstrations are invited, and requests for more information should be sent to: Mike Nicholls, Christ Church College, CANTERBURY, Kent, CT1 1QU, England. Tully Memorial Research Funds Available. The Raptor Research Foundation will provide a $600 grant from the Stephen R. Tully Memorial Fund for research, management, or conservation of birds of prey. Students and serious amateurs are encouraged to apply. Send 5 copies of a succinct proposal (5 pages maximum), outlining your background, study goals and methods, anticipated budget, and other funding requests to James H. Enderson, Chairman, Tully Grant Committee, Colorado College, Colorado Springs, CO 80903. Applications must be received by 10 September 1990, and the award will be announced in October 1990. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. (Founded 1966) OUTGOING EDITOR Jimmie R. Parrish INCOMING EDITOR Josef K. Schmutz ASSOCIATE EDITORS Reed Bowman — Behavior Susan Chaplin — Anatomy and Physiology Richard J. Clark — Order Strigiformes Jeffrey L. Lincer — Environmental Chemistry and Toxicology Carl Marti — Ecology Patricia P. Rabenold — New World Vultures Patrick T. Redig — Pathology, Rehabilitation and Reintroduction Sanford R. Wilbur — Old World Vultures INTERNATIONAL CORRESPONDENT Richard J. Clark CONTENTS FOR VOLUME 23, 1989 Number 1 Commentary: Year-End Message from the President, January 1989. Gary E. Duke 1 Avulsion of the Brachial Plexus in a Great Horned Owl (. Bubo virgimanus) . Michael P. Moore, Erik Stauber and Nancy Thomas 3 A Floating Fish Snare for Capturing Bald Eagles. Steven L. Cain and John I. Hodges 10 Short Communications Unusual Leg Injury in a Nestling Bald Eagle. Richard E. Yates and B. Riley McClelland 14 Predation of Bald Eagles ( Haliaeetus leucocephalus ) on American Coots ( Fulica americana ). Bayard H. Brattstrom 16 Commentary: Effects of Strychnine on Raptors. Stanley n. wiemeyer 18 News and Reviews (Resolutions) 20 News and Reviews 22 Number 2 Neotropical Raptors and Deforestation: Notes on Diurnal Raptors at Finca El Faro, Quetzaltenango, Guatemala, jay p. Vannini 27 Diets of Breeding and Nonbreeding California Spotted Owls. James Thrailkill and Michael A. Bias 39 Food Habits of Turkey Vultures in West Texas. Christos Thomaides, Raul Valdez, William H. Reid and Ralph J. Raitt 42 Short Communications Determining Age and Sex of Nestling Gyrfalcons. K. G. Poole 45 Observations on Cooperative Mobbing of a Bald Eagle. Robert Charles Humphrey 48 Activities of a Migrant Merlin during an Island Stopover. Arlo Raim, William W. Cochran and Roger D. Applegate 49 Bald Eagle Dies from Entanglement in Fish Net. James W. Watson 52 Food Habits of Red-tailed Hawks in Boulder County, Colorado. Daniel T. Blumstein . . 53 Bald Eagle Kills Sharp-shinned Hawk. Jay R. George 55 Thesis Abstracts Reversed Size Dimorphism, Determinants of Intraspecific Dominance, and Habitat Use by American Kestrels ( Falco sparverius) Wintering in South Carolina. David A. Aborn 57 Biosystematics of the Gray Hawk, Buteo nitwvs (Latham). Brian A. Millsap 57 Dissertation Abstract Winter Ecology of Urban Merlins ( Falco columbarius). Ian G. Warkentin 58 News and Reviews 59 Number 3 Investigations of the Decline of Swainson’s Hawk Populations in CALIFORNIA. Robert W. Risebrough, Ronald W. Schlorff, Peter H. Bloom and Edward E. Littrell 63 Censusing of Diurnal Raptors in a Primary Rain Forest: Comparative Methods and Species Detectability. Jean-Marc Thioilay 72 Natural History of the American Kestrel in Venezuela. Thomas G. Balgooyen 85 Observations on Post-fledging Dependence of Kestrels ( Falco tinnunculus rupicolus ) in an Urban Environment, joris Komen and Elizabeth Myer 94 Responses of Breeding American Kestrels to Live and Mounted GREAT Horned Owls. Nicholas W. Gard, David M. Bird, Robin Densmore and Manon Hamel 99 Observations of a Zone-tailed Hawk Family During the Post Fledging Period. F. Hiraldo, M. Delibes and R. R. Estrella 103 Short Communications Eggs of Captive Crested Eagles ( Morphnus guianensis). Lloyd F. Riff, Michael P. Wallace and Nathan G. Gale 107 Serum Estradiol- 17/3 and Testosterone Levels in Great Horned Owls (Bubo virginianus) . Susan A. Mainka, George J. Haimazna and Lori M. Rogers 108 Changes in Winter Distribution of Bald Eagles Along the Colorado River in Grand Canyon, Arizona. Bryan T. Brown, Robert Mesta, Lawrence E. Stevens and John Weisheit 110 Egg Measurements from a Northern Goshawk ( Accipiter gentihs gentilis) Including One Abnormally Large Egg With Twin Embryos. S. J. Petty and D. I. K. Anderson 113 Unusual Roost Site Selection and Staging Behavior of Black-Shouldered Kites. William S. Clark and Brian K. Wheeler 116 The Bartos Trap: A New Raptor Trap. Robert Bartos, Penny Olsen and Jerry Olsen 117 Observations of Autumnal Courtship Behavior in Peregrine Falcons. A. J. Meier, R. E. Noble and P. M. McKenzie 121 Dissertation Abstracts Egg Quality, Nestling Development and Dispersal in the Sparrowhawk ( Accipiter nisus ). Ron Frumkin 123 Trends in Geographic Variation of Cooper’s Hawk and Northern Goshawk: A Multivariate Analysis. Wayne H. Whaley 123 The Biochemistry of Nearctic Peregrine Falcons. Jimmie R. Parrish 124 Number 4 The Breeding Biology of the Broad-winged and Red-shouldered Hawks in Western New York. Scott t. Crocoii and James w. Parker 125 Environmental Contaminants in Blood of Western Bald Eagles. Stanley N. Wiemeyer, Richard W. Frenzel, Robert G. Anthony, B. Riley McClelland and Richard L. Knight 140 Hazards to Raptors from Strychnine Poisoned Ground Squirrels. Josef K. Schmutz, Karrie A. Rose and Robert G. Johnson 147 Surveys for Wintering Birds of Prey in Southeastern Colorado: 1983—1988. David E. Andersen and Orrin J. Rongstad 152 Food Habits of Nesting Prairie Falcons in Campbell County, WYOMING. John R. Squires, Dr. Stanley H. Anderson and Robert Oakleaf 157 Observations on the Evening Departure and Activity of Wintering Short-eared Owls in New Jersey. Thomas Bosakowski 162 Sexual Differences in Timing of American Kestrel Migration at Hawk Mountain Sanctuary, PA. Nancy G. Stotz and Laurie j. Goodrich 167 The Use of Line Transects to Evaluate the Abundance of Diurnal Mammalian Prey. Joan L. Morrison and Patricia L. Kennedy 172 Use of Explosives to Enhance a Peregrine Falcon Eyrie. Joel E. Pagel .... 176 Short Communications Range Extension of the Barred Owl in Western Washington and First Breeding Record on the Olympic Peninsula. Devora Ukrain Sharp 179 Peregrine Falcon Takes Black-bellied Plover from Sea off Kenya. Jennifer F. M. Horne and Lester Short 181 Northern Harrier ( Circus cyaneus ) Predation of Lesser Prairie-chicken ( Tympanuchus pallidicinctus ) . David A. Haukos and Gerald S. Broda 182 An Instance of Carrion-feeding by the Peregrine Falcon ( Falco peregrinus ). Dan G. Holland 184 Thesis Abstracts 185 News and Reviews 186 THE RAPTOR RESEARCH FOUNDATION, INC. (Founded 1966 ) OFFICERS PRESIDENT: Gary E. Duke SECRETARY: James D. Fraser VICE-PRESIDENT: Richard J. Clark TREASURER: Jim Fitzpatrick BOARD OF DIRECTORS EASTERN DIRECTOR: Keith Bildstein CENTRAL DIRECTOR: Patrick T. Redig MOUNTAIN & PACIFIC DIRECTOR: W. Grainger Hunt EAST CANADA DIRECTOR: David M. Bird WEST CANADA DIRECTOR: Lynn Olifhant INTERNATIONAL DIRECTOR: Bernd Meyburg DIRECTOR AT LARGE #1: Michael Collofy DIRECTOR AT LARGE #2: Gary Duke DIRECTOR AT LARGE #3: Jeffrey L. Lincer ****** ************** EDITORIAL STAFF OUTGOING EDITOR: Jimmie R. Parrish, Department of Zoology, 159 Widtsoe Building, Brigham Young University, Provo, Utah 84602 INCOMING EDITOR: Josef K. Schmutz, Department of Biology, University of Saskatchewan, Saskatoon, SK, S7N 0W0 Canada ASSOCIATE EDITORS Reed Bowman — Behavior Susan Chaplin — Anatomy and Physiology Richard J. Clark — Order Strigiformes JEFFREY L. Lincer — Environmental Chemistry and Toxicology Carl Marti — Ecology PATRICIA P. RaBENOLD — New World Vultures PATRICK T. Redig — Pathology, Rehabilitation and Reintroduction Sanford R. Wilbur — Old World Vultures INTERNATIONAL CORRESPONDENT : Richard J. Clark, York College of Pennsylvania, Coun- try Club Road, York, Pennsylvania 17405 The Journal of Raptor Research is distributed quarterly to all current members. Original manuscripts dealing with all aspects of general ecology, natural history, management and conservation of diurnal and nocturnal predatory birds are welcomed from throughout the world, but must be written in English. Contributors should submit a typewritten original and three copies of text, tables, figures and other pertinent material to the Editor. Two original copies of photographic illustrations are required. All submissions must be typewritten double-spaced on one side of 8 Vi x 1 1-inch (21 Vi x 28 cm) good quality, bond paper. Number pages through die Literature Cited section. The cover page should contain the full title and a shortened version of the title (not to exceed 30 characters in length) to be used as a running head. Author addresses are listed at the end of the Literature Cited section. Provide an abstract for each manuscript more than 4 double-spaced typewritten pages in length. Abstracts are submitted as a separate section from the main body of the manuscript and should not exceed 5% of the length of the manuscript. Both scientific and common names of all organisms are always given where first appearing in the text and should conform to the current checklists, or equivalent references, such as the A.O.U. Checklist of North American Birds (6th ed., 1983). Authors should ensure that all text citations are listed and checked for accuracy. If five or fewer citations appear in the text, place the complete citation in the text, following these examples: (Brown and Amadon, Eagles, Hawks and Falcons of the World. McGraw-Hill, New York, 1968), or Nelson (Raptor Res. 16(4):99, 1982). Metric units should be used in all measurements. Abbreviations should conform with the Council of Biology Editors (CBE) Style Manual, 5th ed. Use the 24-hour dock (e.g., 0830 and 2030) and “conti- nental” dating (e.g., 1 January 1984). A more detailed set of instructions for contributors appeared in J. Raptor Res., VoL 21, No. 1, Spring 1987, and is available from the Editor. Send all manuscripts for consideration and books for review to the Editor.