RAPTOR RESEARCH Volume 13 Nmiuher 1 Spring 1979 Riipiur Rtsourch Foundation, Inc. Provo, Utah, U,5.A, RAPTOR RESEARCH Published Quarterly by the Raptor Research Foundation, Inc. Editor Dr. Clayton M. White, Dept, of Zoology, 161 WIDB, Brigham Young University, Provo, Utah 84602 Editorial Staff Dr. Frederick N. Hamerstrom, Jr. (Principal Referee) Dr. Byron E. Harrell (Editor of Special Publications) Dr. Joseph R. Murphy (Printing Coordinator) The Raptor Research Foundation, Inc., welcomes original articles and short notes concerning both diurnal and nocturnal birds of prey. Send all papers and notes for publication and all books for review to the Editor. Most longer articles (20 or more typeset pages) will be considered for publication in Raptor Research Reports , a special series for lengthy and significant contributions containing new knowledge about birds or new interpretations of existing knowledge (e.g., review articles). However, authors who pay page costs (cur- rently $20.00 per page) will expedite publication of their papers, including lengthy articles, by ensuring their inclusion in the earliest possible issue of Raptor Research . Such papers will be in addition to the usual, planned size of Raptor Research whenever feasible. SUGGESTIONS TO CONTRIBUTORS: Submit all manuscripts in duplicate, typewritten, double spaced (all parts), on one side of 8V2 x 1 1 inch paper, with at least 1 inch margins all around. Drawings should be done in India ink and lettered by lettering guide or the equivalent, if possible. Photographs should be on glossy paper. Avoid footnotes. Provide an abstract for all papers more than four double-spaced typed pages in length, not to exceed 5 percent of the total length of the paper. Keep tables at a minimum, and do not duplicate material in either the text or graphs. For advice concerning format refer to the Council of Biological Editors’ Style Manual for Biological Journals or to previous issues of Raptor Research. Proofs will be sent to senior authors only. Major changes in proofs will be charged to the authors. Reprints should be ordered when proofs are returned. BREEDING RESPONSES OF RAPTORS TO JACKRABBIT DENSITY IN THE EASTERN GREAT BASIN DESERT OF UTAH by Dwight G. Smith Department of Biology Southern Connecticut State College New Haven, Connecticut 06515 and Joseph R. Murphy Department of Zoology Brigham Young University Provo, Utah 84602 Abstract The relationships between a collective breeding raptor population of the Golden Eagle (Aquila chrysaetos), the Great Homed Owl ( Bubo virginianus), the Ferruginous Hawk ( Buteo regalis), and the Red-tailed Hawk ( Buteo jamaicensis) and its main prey base, the black-tailed jackrabbit, is examined. By biomass, the jackrabbit is the single most important food item in the diet of each of the raptor species. Populations of each raptor species varied synchronously with jackrabbit abundance. Excepting Red- tailed Hawks, reproduction and operative breeding season mortalities were influenced by jackrabbit abundance. The role of the central Utah environment in dictating pat- terns of raptor response to its staple prey base is considered. Introduction Solomon (1949) proposed that predators respond to prey fluctuations in two dis- tinctive ways, functionally and numerically. The functional response involves a change in food consumed by the individual predator. The numerical response involves an increase in predator density as prey increases, usually through recruitment by im- migration and increased predator reproduction. In this paper we examine the relationships between the breeding population dy- namics of a collective raptor population consisting of the Golden Eagle (Aquila chry- saetos ), the Great Horned Owl (Bubo virginianus), the Ferruginous Hawk (Buteo re- galis ), and the Red-tailed Hawk (Buteo jamaicensis) and abundance of their main prey base, the black-tailed jackrabbit (Lepus californicus ). The study was conducted from November 1966 through August 1970. Some addi- tional observations were also made during the late spring and summer months of 1971 and 1972. Data similar to that which we give here are presented by Woffinden and Murphy (1977) for 1973 and 1974 for the same study area. The exact nature of the response to changing prey densities remains unclear for many raptor species. Nicholson (1930) and Pitelka et al. (1955 a, b) have shown that tundra-inhabiting raptors exhibit strong numerical and functional responses to a cyclic and limited prey base. Responses of temperate raptor populations may be less dra- 1 Raptor Research 13(1): 1-14 2 RAPTOR RESEARCH Vol. 13, No. 1 matic. In Alberta, Mclnvaille and Keith (1974) reported that Great Horned Owls, but not Red-tailed Hawks, showed changes in breeding season density and food habits oc- curring synchronously with snowshoe hare (. Lepus americanus ) populations. Although Craighead and Craighead (1956) concluded that raptor population densities in Mich- igan tended to remain relatively stationary, the situation elsewhere is largely un- known. Study Area We began long term raptor studies on a 7,700 km 2 portion of the eastern Great Basin Desert in winter 1966. Initial studies in 1966-1968 focused primarily on Gold- en Eagles, Great Horned Owls, and Ferruginous Hawks (Murphy et al. 1969). Addi- tional information on these and the Red-tailed Hawk was presented in a four-year collective raptor study on a smaller, 207 km 2 intensive study area (Smith and Murphy 1973). Herein we analyze raptor-jackrabbit relationships on a 1,170 km 2 portion of the original study area. Topographically, the area is characterized by broad, flat alkaline valleys separated by high, north-south running ridges and hills. Valley elevations range from 1,460 to 1,620 m, and maximum elevations of the ridges and hills range from 1,830 m to up- wards of 2,440 m. Climatically the area is characterized as a northern cold desert (Shelford 1963). An- nual precipitation averages 38 cm, and mean monthly temperatures range from -5°C in January to 24°C in July with wide daily and seasonal variations. Two distinctive vegetation associations occur, the mountain desert shrub commu- nity and the dwarf conifer community. The desert shrub community is present over the lower elevations and on the valley floors. It consists of mixtures of shrubs, herbs, and grasses, several of which may, under certain edaphic soil conditions, form large communities. Two predominant shrub communities thus formed include big sagebrush (Artemisia tridentata ) on the better drained soils, and greasewood ( Sarcobatus ver- miculatus ) on the poorly drained valley floors. The well-drained slopes and hills sup- port a dwarf conifer community of Utah juniper (, Juniperus osteosperma ) and pinyon pine (. Finns monophylla) which occur in stands of widely varying density. This part of the Great Basin has a history of chronic livestock over-grazing which dates from the late 1800s. This circumstance coupled with extensive state and federal predator-control programs, has produced conditions which allowed jackrabbit popu- lations to drastically increase periodically to the point at which they are considered a major range pest species (Stoddard et al. 1975). Methods Jackrabbit Populations . We determined jackrabbit abundance by (1) transect counts, (2) road kill counts, and (3) queries of local ranchers and stockmen. Permanent linear transects, each 805 m in length, were marked off through six large desert scrub communities^ selected at random in the study area. Vegetation of the transects included typical predominant vegetation of the study area. At biweekly intervals from mid-December through March all transects were walked by two ob- servers spaced 20 m apart. Counts derived from transects are indications of relative abundance rather than absolute population counts and reflect the basic weakness of the transect method (Southwood 1966). They do, however, provide adequate indices of population change (Gross et al. 1974) and were the only practicable means of esti- Spring 1979 Smith & Murphy— Breeding Responses 3 mating jackrabbit population abundance over our large study area. Each year we recorded winter monthly tallies of the number of local kills observed on nearby highways. Such a tally is a rough but useful index, especially when used in conjunction with other methods. Each year we asked resident ranchers and stockmen and itinerant hunters to esti- mate and compare jackrabbit abundance with previous years. This questioning pro- vided subjective information which may at the very least indicate high and low years of jackrabbit abundance. Raptor Populations. From November 1966 through August 1970 we visited the study area at least once and usually two or more times per week to obtain informa- tion on raptor population dynamics. The 207 km 2 intensive study area was divided into subsections, each approximately 1.6 km 2 , to facilitate survey. These were system- atically searched in a rotating sequence for raptor nesting pairs, nonnesting pairs, and individuals at biweekly intervals beginning in mid-December and lasting through mid- May of each year. Foot searches were supplemented by aerial surveys, but the latter were of limited usefulness because of minimum speed and altitude requirements. Status of nesting pairs, nonnesting pairs, and single birds on the intensive study area was checked at least twice weekly during the nesting season and once weekly on the 1,170 km 2 study area. Throughout the four breeding seasons the single exceptional time interval between successive study area visits was 13 days in April 1970. Raptor Food Habits. Food habits of each raptor species were determined by weekly tabulations of prey items and analysis of pellets gathered from the nest site. Prey re- mains were usually left on the nest, after being marked by toe clipping to avoid du- plication in tabulation. Results Jackrabbit Populations. Jackrabbit numbers were low in 1967 and 1970, inter- mediate in 1968, and high in 1969 (fig. 1). The average number of individuals per transect count each year were: 1967, 3.3 + 0.7 (range 0-7 individuals); 1968, 9.03 + 1.2 (ranges 3-17); 1969, 13.8 + 1.5 (ranges 6-31); and 1970, 3.6+ 1.0 (ranges 0-9). Winter road kill counts in terms of individuals per mile of road travelled for each of the study years are: 0.8 in 1967; 1.4 in 1968; 1.9 in 1969; and 1.1 in 1970. Subjective observations by local stockmen, ranchers, and sportsmen familiar with the study area also suggest that jackrabbit populations peaked in 1969 and declined in 1970. Our own data plus casual observations indicate that jackrabbit abundance peaked in 1969 and declined in 1970. In 1969, for example, we commonly were able to observe concentrations of 50-75 individuals feeding in small open fields of 3-5 acres during the late evening hours after sunset. While none of these results provide conclusive evidence of jackrabbit population fluctuations, they do, considered together, strongly suggest that populations of 1968 and 1969 were considerably larger than populations of 1967 and 1970. Gross et al. (1974) conducted a separate study of jackrabbit populations approximately 200 km north of our study area. Their estimates of jackrabbit densities as determined by transects and drive counts, show a similar population low in 1967 followed by in- creasing population densities in 1968 and 1969. They did not find a corresponding decline in 1970 but did observe that synchrony of intermountain valley populations were in some instances 1-2 years out of phase with one another. 4 RAPTOR RESEARCH Vol. 13, No. 1 Avian Predator Food Habits. We have previously described the food habits of each of the avian predators (Murphy et al. 1969, Smith and Murphy 1973, and in press) and herein present only brief summaries (table 1). On a weight basis the black-tailed jackrabbit was the single most important prey item of each raptor species, contrib- uting from a minimum of 81.6% to a maximum of 94.1% of the prey biomass con- p-h td © © © n o> cq pf n ^ ©n w o o N I S lO © ^ H c-i id d c-i © in (M i— i en ® ® oo ^ ^ OOOO-ddd^ o co n in ® 00 00 CO Tt; r oo nq in in d r co no -h t © co t> o d i> in © i i—i t~- l> CD CN CO H rH jg co ni ni m o o co os m ■ d S i § ” 05 i'S U £ O J* f B T3 §a -E O *3 kJ cnO « H O £ ° § d & S 'Si'S 3,1 g 3 5 II §£#5 6 Be? dS | % ffi eg, " e- s « s g s & § ^ - fl >S os t, ■© ,2 'w> I £ |.j= ^ Q-iS 2 S' S -E « o b ^ cnQ OWEh d S 03 3 tuoS c -S ^ is 5.2 £ -p as V ^ *b 1) O ^3 ^ #0 « rt E-I s!§ Jj S. 1 4 S b,| % ju fir * Sr P Spring 1979 Smith & Murphy— Breeding Responses 5 sumed. This finding supports the contention of Clark (1972) that jackrabbits form a major part of the food base of the northern cold desert ecosystem. On a percent frequency basis, however, some differences in prey species occur- rence are evident between high prey years (1968 and 1969) and low years (1967 and 1970). Chi-square tests of the frequency of occurrence of each of the prey categories reveal significant statistical differences between years (P<0.01) in Ferruginous Hawk and Red tailed Hawk food but not for that of Great Homed Owls or Golden Eagles, although the two latter raptor species exhibit some diet differences. These differences in food between high- and low-jackrabbit years suggest a shift to alternate or buffer prey species. The observed degree of shift to alternate species was, however, not commensurate with the decline in jackrabbit abundance, and this prey species still comprised 82% or more of the food biomass of each raptor species even in low-jack- rabbit years. We believe that the abundance of jackrabbits is thus an important limiting factor which largely determines annual changes in the size and reproductive performance of the collective raptor breeding population. Avian Predator Populations. A summary of yearly raptor population trends is pres- ented in table 2. The annual breeding-season populations of the four species were ap- proximately, in order of most to least common: Buteo regalis, Bubo virginianus, Aquila chrysaetos, Buteo jamaicensis. Each year pairs of another large raptor, the Swainson’s Hawk (Buteo swainsoni), were consistently present on the study area, but were too few in number to permit detailed analysis. The correlation between the raptor population and jackrabbit abundance is readily evident and may be categorized as the first component of numerical response. Total numbers of raptors in low-jackrabbit years varied from 18-49% below populations of high-jackrabbit years. The magnitude of change is illustrated by 1967 (a low year) and 1969 (a peak year) populations, in which almost twice as many raptors were found on the study area. Populations of Golden Eagles and Great Horned Owls, both permanent residents, showed comparatively less yearly variation than the migratory Ferruginous Hawk and Red-tailed Hawk populations. Thus, after an initial increase from 1967 to 1968, Gold- en Eagle numbers varied only by 3 and Great Horned Owl numbers by 4 during the three years, 1968-1970. In contrast Ferruginous Hawk numbers changed considerably over the four-year period, with the 1969 population about one-third larger than 1968 and about twice as large as 1967 and 1970 populations. The observed recruitment of Ferruginous Hawks is probably to some extent a function of its migratory status. Studies by Craighead (1959), Naumou (1948), Galushin (1974), and others have noted the ability of several nomadic avian predator species to concentrate rapidly in areas of prey abundance. Pitelka et al. (1955 a, b) reported large variations in the highly mobile raptor populations of the Arctic tundra, and Linkola and Myllymaki (1969) observed invasions of certain raptor species during high-rodent years in Finland. Comparatively, the strongly territorial and permanently resident Tawny Owl (Strix aluco) was noted by Southern (1970) to maintain relatively stable numbers from year to year on his Wytham Wood Study Area in England. Breeding Fraction of the Avian Predator Population. The tendency of pairs to breed or refrain from breeding may be considered as a part of the second component of nu- 8 RAPTOR RESEARCH Vol. 13, No. 1 Table 2. Raptor population trends in central Utah, 1967-1970. Raptor Species 1967 1968 1969 1970 Golden Eagle No. pairs 9 14 14 13 No. breeding pairs 7 13 13 11 No. single birds 1 1 1 0 Total no. 19 29 29 26 Great Horned Owl No. pairs 7 16 16 13 No. breeding pairs 6 14 16 10 No. single birds 2 1 1 3 Total no. 16 33 33 29 Ferruginous Hawk No. pairs 19 31 38 20 No. breeding pairs 15 28 34 13 No. single birds 5 1 2 4 Total no. 43 63 78 44 Red-tailed Hawk No. pairs 8 12 13 12 No. breeding pairs 5 10 12 11 No. single birds 1 0 1 0 Total no. 17 24 27 24 Totals (All Species) No. pairs 43 73 81 58 No. breeding pairs 33 65 76 45 No. single birds 9 4 5 7 Total no. 95 149 167 123 merical response, which inludes all strategies directed towards increasing predator population growth. We classified a pair as breeding if a nest was constructed and at- tended. For the collective raptor population, the ratio of breeding pairs to total pairs varied synchronously with jackrabbit abundance. Percentage ratios (pairs breeding to total pairs present) were 76.7% and 77.6% for the low-jackrabbit years of 1967 and 1970 respectively, 89.0% for the intermediate year 1968, and 93.8% for the peak-jack- rabbit year of 1969. Only one species did not follow this pattern: the number of Red- tailed Hawk breeding pairs in 1970 was one larger than in 1968, although lower than in 1969. In studies elsewhere, Hamerstrom (1969), Pitelka et al. (1955 a, b), and Rusch et al. Spring 1979 Smith & Murphy— Breeding Responses 7 (1972) all reported increased percentages of breeding pairs of raptors during high- prey years, which contrasted sharply with a low intensity of nesting efforts in low- prey years. Southern (1970) found that the percentage of Tawny Owl pairs which each year attempted to breed varied from zero in unfavorable years to 87% in high- rodent years. He further observed that the most common way in which Tawny Owl pairs reacted to low prey was by refraining altogether from breeding. Pairs which have constructed and attended a nest may lay their eggs or, con- versely, refrain from depositing eggs and subsequently desert the nest. We did find a slightly higher incidence of nest desertion in low prey years, but unfortunately the question of possible human interference, not uncommon in our study area, renders analysis of this relationship unclear. It is quite possible, however, that pairs attempt- ing to breed in times of low prey are more intolerant of disturbance. In favorable years breeding pairs appeared to be far more tolerant of our activities in the vicinity of the nest site (cf. Woffinden 1975 for data on this aspect). Clutch Size. Mean clutch size of Golden Eagles, Great Horned Owls, and Ferru- ginous Hawks, but not Red-tailed Hawks, varied in synchrony with rodent density (table 3). Clutches of Golden Eagles varied from 1.9 in low years to 2.2 in peak years. The 1968 and 1969 clutches of Great Horned Owls and Ferruginous Hawks tended to contain a minimum of one additional egg as compared to 1967 and 1970 clutches (t — 2.58; 2.75 respectively, P<0.05 for both), but not significantly different from one another (t = 1.26; P>0.05). In addition, mean clutch size of 1967 and 1970 did not differ significantly (l = 1.15; P>0.05). Similarly Ferruginous Hawk clutches of 1968 and 1969 were significantly larger than 1967 and 1970 clutches (t = 2.56; 2.92 respectively; P<0.05) but not different from one another (t = 0. 1 1; P>0.05). Again, clutches of 1967 and 1970 did not differ significantly (t = 0. 1 1 ; P>0.05). Mclnvaille and Keith (1974) found a significant increase in Great Horned Owl clutch size during years of high snowshoe hare abundance, and Houston (1971) deter- mined that the largest brood sizes of this nocturnal avian predator coincided with years in which its prey was most abundant. Unlike the other raptor species, clutch size of Red-tailed Hawks consistently in- creased each year, from a low of 2.5 in 1967 to a high of 3.3 in 1970. In contrast, Mclnvaille and Keith (1974) found that peak clutch size of Red-tailed Hawks near Alberta did occur simultaneously with highest annual snowshoe hare abundance. For three of the four species the combination of increased numbers of breeding pairs and large clutches resulted in a greater total productivity of eggs in years of jackrabbit abundance. Correlation tests reveal a significant relationship between prey abundance and total egg production of Great Horned Owls (r = 0.97; t = 5.24; P<0.05) and Ferruginous Hawks (r=1.98; t = 6.64; P<0.05). Total egg production of Golden Eagles and Red-tailed Hawks was slightly greater in 1968 and 1969, but the relationships are not significant (P>0.05 for both species). Fledged Brood Rates. Reproductive success as measured by fledged brood rates var- ied considerably among the raptor species. A clear correlation between total number of fledged young, nesting success (percentage of young fledged per eggs laid), and jackrabbit abundance can be observed only for Ferruginous Hawk populations. Great Homed Owl fledged brood rates were considerably higher in favorable years, but percentage breeding success was lower in those years. To complicate the problem, the fledged brood production of Red-tailed Hawks was generally higher in favorable years, despite the increased clutch size in 1970, which was offset by increased morta- 8 RAPTOR RESEARCH Vol. 13, No. 1 lity of young in the nest. A major factor contributing to greatly fluctuating mortality rates from hatching to fledging is the comparatively high level of human disturbance affecting many avian predator pairs on our study area. Golden Eagles, the largest and most conspicuous of the diurnal raptors, were frequently the object of such activities as egg collecting, photography of young, deliberate nest destruction, and attempts by hunters to kill the adults. These disturbances, of course, render any analysis of yearly breeding success unrealistic unless their effect on the behavior of the birds can be re- solved. Table 3. Reproduction of large raptors in central Utah, 1967-1970. Raptor Species 1967 1968 1969 1970 Golden Eagle No. clutches produced 7 13 13 11 Av. clutch size 1.9 2.0 2.2 1.9 Total eggs produced 13 26 27 21 No. nests hatching young 5 12 10 10 No. young hatched/nest 1.2 1.2 1.2 1.8 Total young hatched 1 6 (46.2) 15 (57.7) 11 (40.7) 18 (85.7) No. nests fledging young 5 11 9 10 No. young fledged/nest 0.8 0.9 1.0 1.7 Total young fledged 2 4 (30.8) 10 (38.5) 9 (33.3) 17 (81.0) Great Horned Owl No. clutches produced 6 13 15 10 Av. clutch size 2.0 2.9* 3.3** 2.4 Total eggs produced 12 38 50 23 No. nests hatching young 5 12 14 8 No. young hatched/ nest 1.8 2.5 2.7 2.3 Total young hatched 9 (75.0) 30 (78.9) 38 (76.0) 19 (82.6) No. nests fledging young 5 11 13 8 No. young fledged/ nest 1.8 1.9 2.4 1.8 Total young fledged 9 (75.0) 21 (55.3) 31 (62.0) 15 (65.2) Ferruginous Hawk No. clutches produced 14 28 34 12 Av. clutch size 2.5 3 . 7 ** 3.8** 2.9 Total eggs produced 36 104 129 35 No. nests hatching young 11 25 32 9 No. young hatched/ nest 1.3 2.4 3.1 1.4 Total young hatched 15 (41.7) 60 (57.7) 99 (76.7) 17 (48.6) No. nests fledging young 10 25 31 7 Total young fledged/nest 1.2 2.2 2.9 1.4 Total young fledged 12 (33.3) 55 (52.9) 90 (69.8) 10 (28.6) Spring 1979 Smith & Murphy— Breeding Responses 9 Red-tailed Hawk No. clutches produced 5 10 12 11 Av. clutch size 2.5 2.8 3.1 3.3 Total eggs produced 13 28 37 36 No. nests hatching young 5 9 11 11 No. young hatched/ nest 1.8 2.3 2.5 2.3 Total young hatched 9 (69.2) 21 (75.0) 28 (80.0) 26 (72.2) No. nests fledging young 5 9 10 9 No. young fledged/nest 1.5 1.6 2.2 1.5 Total young fledged 8 (61.9) 15 (53.6) 22 (59.5) 14 (38.9) ‘Figures in parentheses are percentages of total eggs hatched to total eggs laid. “Figures in parentheses are percentages of total young fledged per total eggs laid. "Significant at PClevel. '"Significant at the P<0.01 level. Discussion Breeding Mortality and Regulation of Avian Predator Populations. If we arbitrarily allow that the maximum reproductive potential of the collective raptor population (excepting Red-tailed Hawks) was achieved during the peak prey year of 1969, then we may comparatively measure the limiting effects of declining prey populations on raptor reproduction. Several factors decreased breeding productivity during unfavor- able years (or conversely, increased mortality). Among these were: (1) failure of pairs to nest, (2) failure of pairs to achieve maximum clutch size, and (3) failure to hatch brood, and fledge a maximum number of young. These mortality factors are related in sequential fashion. Thus, for a breeding avian predator population to have a pro- ductive year there must be a certain minimum prey density which provides the stim- ulus and physiological ability to breed. Further increases in prey density may be re- flected in increased productivity, up to the point at which other resources may become limiting. Conversely, lower prey densities inhibit breeding and increase the probability of mortality at all breeding stages. Southern (1970) has shown that the relationships between breeding mortalities and prey abundance may be examined by key factor analysis. This technique, developed by Varley and Gradwell (1960) and critically discussed by Ito (1972), permits in- spection of the contribution of each mortality (ki) to the total mortality (K) acting on the population. We confined our key factor analysis to the sequence of mortality operating on the following phases of the breeding population: (1) kl = logarithm of maximum eggs producible by maximum-sized breeding popu- lation-logarithm of the number of eggs produced by the actual-sized breed- ing population. The maximum breeding season population of each avian predator species was at- tained in 1969, and kl measures the mortality increase in other years when the re- spective species did not achieve the 1969 breeding population level. In effect, kl is a measure of the mortality produced by a pair’s decision to breed or forgo breeding, which we believe, will be a function of available food resources. (2) k2 = logarithm of eggs produced with all breeding pairs achieving maximum aver- age clutch size— logarithm of actual number of eggs produced by breeding pairs. 10 RAPTOR RESEARCH Vol. 13, No. 1 The yearly average clutch size is measured against the maximum yearly average clutch size achieved. Three species attained a maximum in 1969: Golden Eagle, 2.2; Great Horned Owl, 3.3; Ferruginous Hawk, 3.8. Red-tailed Hawk maximum average clutchsize of 3.3 occurred in 1970; it was 3.1 in 1969. (3) k3 = logarithm of eggs actually laid— logarithm of eggs which subsequently do hatch This key factor measures the mortality caused by the proportion of eggs laid which do not hatch, regardless of specific cause. (4) k4 = logarithm of the number of young which hatch— logarithm of the number of young which fledge. This mortality factor measures the impact of loss of young which have actually hatched but for whatever reason do not fledge. (5) K = kl+k2 + k3 + k4 K will thus represent the absolute impact of the four stages of mortality which oc- cur during the nesting sequence. Results of key factor analysis for each raptor species are presented in figure 2. Al- though key factors are occasionally subjected to statistical treatment, Eberhardt (1970), Ito (1972), and Kuno (1973) have all suggested defects of regression analysis which may distort and invalidate results. We support their conclusions and restricted our analysis to graphical inspection, similar to that of Varley and Gradwell (1960). Inspection suggests that the two most important mortality factors operative on Great Horned Owl populations are K 1, the failure of a maximum number of pairs to nest, and k 2, the failure of nesting pairs to achieve maximum clutch size. The graph of k4, the mortality of hatched young, diverges synchronously from K and suggests that k 1 and k2 mortalities are at least partly compensated by increased survival of young in poor prey years. Mortality factors k 1 through k3 are equally additive in impact on Ferruginous Hawk reproduction. Factor k4 also follows K, albeit weakly. The close graphical cor- relation of kl-k4 mortality factors with K suggests that Ferruginous Hawk reproduc- tion is strongly influenced in a density-dependent manner by jackrabbit abundance. Graphs of key factors operative on Golden Eagle populations are less readily inter- preted. Although failure of maximum pairs to nest plus achieve maximum clutch size was contributory, k3, the failure of eggs to hatch, was the single most important mor- tality factor. Unfortunately, the previously discussed high level of human disturbance of Golden Eagle nests may have contributed to the importance of this mortality fac- tor. Graphs of Red-tailed Hawk k 1, k2, and k4 factors partially paralleled K, suggesting that Red-tailed Hawks may have partially adjusted their reproduction to jackrabbit abundance in density-dependence fashion. The k3 mortality partially diverges in com- pensatory fashion. The overall relationships between Red-tailed Hawk reproduction and jackrabbit abundance, however, remain unclear. Red-tailed Hawk food habits during low prey years did show a higher utilization of alternate prey species as com- pared to the other large raptors (table 1), and it is possible that this usage of other prey was fully commensurate with jackrabbit decline. Nature of the Predator-Prey Relationship. The nature of the relationship between the large avian predators and their stable prey base in central Utah warrants further consideration. Luttich et al. (1970) have suggested that regional raptor populations Spring 1979 Smith & Murphy— Breeding Responses 11 may exhibit different adjustment patterns to a fluctuating prey base. They note that the staple prey base of the North is limited to just a few species of highly cyclic mi- crotine rodents that Arctic raptors tend to concentrate upon and effectively exploit. They compare this response to temperate-zone raptors, which maintain notably sta- tionary local populations, despite fluctuating prey abundance, and instead vary repro- duction with prey abundance. The large raptor populations of central Utah appear to combine elements of both response patterns. Thus, populations of each raptor species increased in high-jackrab- bit years and decreased in low-jackrabbit years, suggesting that even the permanently resident populations of Golden Eagle and Great Horned Owls are in fact able to shift into an area of prey abundance. In central Utah, this ability to concentrate upon and exploit local prey abundance is most dramatically illustrated by the migratory Ferru- ginous Hawk populations. However, unlike northern raptor populations, which may entirely desert a locality in low-prey years, breeding populations of each large raptor species were present even during unfavorable prey years in central Utah. In these in- stances, the response patterns of the nesting raptors are similar to the reproductive responses revealed by other studies of temperate raptor populations, with each raptor species increasing reproductive efforts in favorble prey years. The response patterns of the central Utah breeding raptor populations may reflect the unique nature of the disturbed habitat. The dual environmental impact of over- grazing and predator control programs have produced a comparatively simplistic food web, characterized by the superabundance of a single prey species, the black- tailed jackrabbit, which exhibits strong fluctuations in density. This results in a local desert environment which in some respects resembles the northern tundra ecosystem, and which may in turn produce the raptor breeding responses discussed in this paper. Summary In the years 1966-1971 we studied the relationships between a collective breeding raptor population of Golden Eagles, Great Horned Owls, Ferruginous Hawks, and Red-tailed Hawks and its staple prey base, the black-tailed jackrabbit, in central Utah. Densities of jackrabbits fluctuate widely in this area. Each raptor species responded functionally to these jackrabbit fluctuations, consuming proportionally greater amounts of this prey species in years of higher abundance. Each raptor species also responded numerically to jackrabbit fluctuations. Two forms of numerical responses were observed: increased concentrations of rap- tors in favorable years, and increased reproductive efforts of breeding raptor popu- lations, which also varied synchronously with jackrabbit abundance. Key factor analy- sis of raptor species densities suggest two major reproductive strategies, the tendency of pairs to breed in favorable years and refrain from breeding in low prey years, and a variation of clutch size, with maximum clutch size produced synchronously with jackrabbit density highs. The relationship of the raptor species populations with their chief prey base is re- vealed by response pattens similar in respects to both North and temperate zone rap- tor populations, and may reflect the disturbed nature of the central Utah habitat. Acknowledgments We wish to express our sincere appreciation to Clyde C. Edwards, David H. Ellis, 12 RAPTOR RESEARCH Vol. 13, No. 1 Franz J. Camenzind, J. Bradford Weston, W. Bruce Arnell, and Charles R. Wilson for valuable field assistance. Howard S. Gershman prepared figures 1 and 2 of this paper, for which we express thanks. Financial support for this study was provided by research funds from a predoctoral fellowship awarded the senior author, from the Department of Zoology, Brigham Young University, and a grant from the National Audubon Society. Literature Cited Clark, F. W. 1972. Influence of jackrabbit density on coyote population change. J. Wildl. Mgmt. 36:343-356. Craighead, J. J. 1959. Predation by hawks, owls and gulls. Pages 35-42 in The Ore- gon meadow mouse irruption of 1957-1958. Oregon State College Federal Co- operative Extension Service Publ., pp. 1-88. Craighead, J. J., and F. C. Craighead. 1956. Hawks, owls, and wildlife. The Stackpole Company. Harrisburg, Pa. 443 pp. Eberhardt, L. L. 1970. Correlation, regression, and density dependence. Ecology 51:306-310. Galushin, V. M. 1974. Synchronous fluctuations in populations of some raptors and their prey. Ibis 116:127-134. Gross, J. E., L. C. Stoddart, and F. H. Wagner. 1974. Demographic analysis of a northern Utah jackrabbit population. Wildl. Mongr. 40:1-68. Hamerstrom, F. 1969. A harrier population study. Pages 367-384. In J. J. Hickey, ed., Peregrine Falcon populations, their biology and decline, Univ. Wisconsin Press, Madison, Milwaukee, and London. Houston, C. S. 1971. Brood size of the Great Horned Owl in Saskatchewan. Bird- Banding 42:103-105. Ito, Y. 1972. On the methods for determining density-dependence by means of regres- sion. Oecologia 10:347-372. Kuno, E. 1973. Statistical characteristics of the density-independent population fluc- tuation and the evaluation of density-dependence and regulation in animal pop- ulations. Res. Popul. Ecol. 15:99-120. Linkola, P. A., and A. Myllymaki. 1969. Der Einfluss der Kleinsauger-fluctuationen auf das Bruten einiger kleisaugefressende Vogel im sudlichem Hame, Mittelfinn- land 1952-1966. Orn. Fenn. 46:45-78. Luttich, S. N., D. H. Rusch, E. C. Meslow, and L. B. Keith. 1970. Ecology of Red- tailed Hawk predation in Alberta. Ecology 51:190-203. Mclnvaille, W. B., and L. B. Keith. 1974. Predator-prey relations and breeding biolo- gy of the Great Horned Owl and Red-tailed Hawk in Central Alberta. Canad. Field. Natur. 88:1-20. Murphy, J. R., F. J. Camenzind, D. G. Smith, and J. B. Weston. 1969. Nesting ecolo- gy of raptorial birds in central Utah. Brigham Young Univ. Sci. Bull. 10(4): 1-36. Naumou, N. D. 1948. Sketches of the comparative ecology of mouselike rodents. Izd. AN SSR. M. Nicholson, E. M. 1930. Field notes on Greenland birds. Ibis 6:280-313, 395-428. Pitelka, F. A., P. Q. Tomich and G. W. Treichel. 1955a. Ecological relations of jae- gers and owls as lemming predators near Barrow, Alaska. Ecol. Monogr. 25:85-117. Spring 1979 Smith & Murphy— Breeding Responses 13 Pitelka, F. A., P. Q. Tomich, and G. W. Treichel, 1955b. Breeding behavior of jae- gers and owls near Barrow, Alaska. Condor 57:3-18. Rusch, D. H., E. C. Meslow, L. B. Keith, and P. D. Doerr. 1972. Response of Great Horned Owl populations to changing prey densities. J. Wildl. Mgmt. 36:282-296. Shelford, V. E. 1963. The ecology of North America. Univ. Illinois Press. Urbana, Il- linois. Smith, D. G., and J. R. Murphy. 1973. Breeding ecology of raptors in the eastern Great Basin of Utah. Brigham Young Univ. Sci. Bull. 18(3): 1-76. Smith, D. G., and J. R. Murphy. Biology of the Ferruginous Hawk in central Utah. Sociobiology, in press. Solomon, M. E. 1949. The natural control of animal populations. J. Anim. Ecol. 18:1-35. Southern, H. N. 1970. The natural control of a population of Tawny Owls ( Strix aluco ). J. Zool., London 162:197-285. Southwood, T. R. E. 1966. Ecological methods. Methuen, London. Stoddard, L. A., A. D. Smith, and T. W. Box. 1975. Range management. 3rd ed. New York: McGraw-Hill. Varley, G. C., and G. R. Gradwell. 1960. Key factors in population studies. J. Anim. Ecol. 29:399-401. Woffinden, N. D. 1975. Ecology of the ferruginous hawk ( Buteo regalis) in central Utah: Population dynamics and nest site selection. Ph.D. dissertation, Brigham Young Univ., Provo, Utah, 102pp. Woffinden, N. D., and J. R. Murphy. 1977. Population dynamics of the Ferruginous Hawk during prey decline. Great Basin Nat. 37:411-425. 14 RAPTOR RESEARCH Vol. 13, No. 1 1967 1968 1969 1970 Figure 1. Transect counts of jackrabbit abundance on the central Utah study area, 1967-1970. Months presented for each year are November (N), January (J) and March (M). Vertical lines represent one stan- dard deviation above and below average count. Figure 2. Graphs of key factor analysis for reproduction data of each large raptor species population on the intensive study area in central Utah, 1967-1970. Individual key factors as presented in text. THE EGGS OF THE BLACK HAWK-EAGLE by Lloyd F. Kiff Western Foundation of Vertebrate Zoology 1100 Clendon Avenue Los Angeles, California 90024 There appears to be no published account of the eggs of the Black Hawk-Eagle (. Spizaetus tyrannus). Therefore, it seems worthwhile to place on record the descrip- tions of four infertile eggs recently laid by two females of this species in the Los An- geles Zoo. The birds were obtained by the zoo in approximately 1968 and were prob- ably captured in Ecuador (Frank Todd pers. comm.). The eggs are now in the collection of the Western Foundation of Vertebrate Zoology (WFVZ). On 5 February 1978 one female laid an egg (WFVZ 42,004) which is white with heavy scallop-shaped markings and splotches of reddish-brown on the large end and a dense sprinkling of fine reddish-brown streaks and dots over the remainder of the sur- face. The egg is short oval in shape (Preston in Palmer, Handbook of North American birds, vol. 1, Yale Univ. Press, New Haven, 1962), and not glossy, and it measures 57.42 X 47.18 mm. The empty dry shell weight is 5.222 g. The same bird laid another egg on 10 March 1978, or 34 days after the first egg. The second egg (WFVZ 42,005) does not differ markedly from the first in color, shape, or texture, and it measures 58.06 X 48.06 mm. It weighed 73.522 g on 11 March before preparation, and the empty dry shell weight is 5.699 g. The other female laid single eggs on 7 February and 10 March 1978, or 32 days apart. The 7 February egg (WFVZ 42,007) is white with heavy smudges of reddish- brown towards the small end and sparse scrawls and dots of reddish-brown over the rest of the surface. The egg is subelliptical (Preston in Palmer op. cit.), and not glossy, and it bears two small calcareous tubercles on the large end. It measures 62.28 X 48.86 mm with an empty dry shell weight of 7.356 g. The 10 March egg (WFVZ 42,006) laid by this female is similar in most details to her first egg, except that the pattern of superficial markings is reversed with the heavy reddish-brown smudges being located nearer the large end than the small end. The egg measures 63.44 X 48.65 mm, and had a whole weight of 81.551 g on 11 March. The empty dry shell weight is 6.050 g. The laying schedule of these captive birds suggests a clutch size of one, as is the case for several Old World species of Spizaetus (Brown and Amadon, Hawks, Eagles, and Falcons of the World, vol. 2, New York, McGraw-Hill, 1968). However, a Black Hawk-Eagle nest observed by Smith (1970, Condor 72:247-248) in Panama in 1965 and 1968 contained two young in both years. I am very grateful to Michael Cunningham of the Los Angeles Zoo for making these eggs available for study, to Ed Harrison for preparation of the specimens, and to Michael Morrison for his assistance in various ways. 15 Raptor Research 13(1): 15 PRAIRIE FALCON CARRIES STICK TO NEST by John C. Barber Division of Birds Room E-607 NHB Smithsonian Institution Washington, D.C. 20560 On 27 March 1976, I watched a male Prairie Falcon ( Falco mexicanus ) carry a stick in its talons during a 91-meter (300-foot) flight from a cliff and across a canyon to a nest scrape on a ledge in Aravaipa Canyon, Graham County, 16 km (10 mi.) west of Klondyke, Arizona, The nest scrape was on a cliff facing due east, in a small pothole under an overhang, approximately 12 meters (40 feet) from the top of the cliff and 46 meters (150 feet) from its bottom. The stick was approximately 20 cm (8") long and 13 mm (Vz") in diameter, and was deposited on the nest ledge. The male flew off the ledge without the stick about 15 seconds after its arrival. The male was silent during the flight to and from the nest site. The female was perched above and to the south of the scrape and did not react with any movements or vocalizations to the male’s behavior. Judging by the timing of two other nests in this canyon, the incident occurred two or three days before the first egg would have been laid. The scrape was deserted be- fore June 1, for unknown reasons. I could not recover the stick because of the degree of overhang and the steepness of the cliff. The literature does not contain any previous record of such behavior. Nest-building in the larger falcons has not been recorded in the field with the exception of several undocumented cases (cf. Brown and Amadon, 1968, Hawks, Eagles, and Falcons of the World, McGraw-Hill, New York, pp. 839 and 842). Amadon (pers. comm.) has reaffirmed his earlier statement (p. 839 above) that the larger falcons probably do not build their own nests. The idea that the male may have flown with the stick in reac- tion to me is unlikely as I was hidden during the entire flight. Other Falconiformes sometimes bring sticks to the nest while the female is incubating, instead of bringing food, as reported by Schnell for the Goshawk ( Condor 60:382) and Barber and Schnell for the Mexican Black Hawk (unpublished observations), but at this time the female was not incubating. This incident appears to be a previously unreported type of pair-bonding behavior for the Prairie Falcon, but whether or not it was unique to this individual is uncer- tain. The publication of this note may stimulate further discussion. Dean Amadon, M. Ralph Browning, Frank L. Beebe, and Richard R. Olendorff made helpful comments on this note. Their help is gratefully acknowledged. 16 Raptor Research 13(1): 16 COURTSHIP OF COMMON CARACARAS IN COSTA RICA Lawrence Kilham Department of Microbiology Dartmouth Medical School Hanover, New Hampshire Between 8 and 16 January, 1978, I studied the courtship of a pair of Common Cara- caras ( Caracara cheriway) building a nest, concealed by leaves and ca 35 m up in the top of a tree (unidentified) that grew in front of the Palo Verde station of the Organiza- tion of Tropical Studies in Guanacaste, Costa Rica. I made observations with 8 X 30 binoculars, spending 4 to 5 h per day for 9 days, aided by my wife, Jane Kilham. Al- though I could not tell the sexes apart at a distance, I could distinguish the male (identi- fied at times of copulation) from the female when the two were perched close to one another (a) by his being smaller and (b) having less barring on his white breast. Behavior either not previously undescribed (Bent 1938, Brown and Amadon 1968) or described in minimal detail is presented here. Foraging. The nest tree and all of the main perching trees of the caracaras were at the edge of a tropical dry forest bordering a marsh teeming with waterfowl. As far as I could determine all foraging was done over the marsh and mostly by the male. Of 13 items brought in, 11 were definitely, and 2 probably, birds. All of these were partially plucked or dismembered by the time I saw them. Two of the heads recovered seemed to be those of Blue-winged Teal (Anas discors ). I could not tell whether the prey was cap- tured alive or found dead. The diet of birds was apparently unusual. I have encountered two reports of caracaras catching live birds, one by Layne et al. (1977) for Common Caracaras catching Cattle Egrets ( Bulbulcus ibis) and one by Myers (1978) on the catch- ing of Southern Lapwings ( Vanellus chilensis) by the Crested Caracara ( Polyborus plan- cus ). A neighboring pair of caracaras inhabiting a dry cattle pasture were seen feeding on carrion (once an iguana and once a tree porcupine) in the manner described by Glaze- ner (1964) for Common Caracaras in Texas. Transfer of prey. The female caracara began making “wuck” vocalizations on 15 Jan- uary when the male, carrying the carcass of a bird in his beak, came from the marsh to the tree where she was perching. The female walked along the large limb where the two perched and took half the prey from him. Both birds then fed by holding the prey down with one foot. When the female finished, she took the remains from her mate. After a few minutes the male walked over and took part of this back again. When the female had finished her second portion, she again took the male’s portion. The male yielded his portion to her three times without overt signs of being disturbed. On four of the five occasions when I saw the female feeding, it was when she took the prey from the male in this manner. On the fifth occasion the male flew with a portion of a carcass to the large, nearly horizontal limb where his mate was standing and immediately walked over to give the prey to her. Then, as she stood with head lowered and the prey in her beak, he mounted in an incomplete copulation. Taking prey remains to the nest. On seven occasions the caracaras brought prey re- mains, which consisted of nothing but bones and tendons, to the nest. On five of these 17 Raptor Research 13(1): 17-19 18 RAPTOR RESEARCH Vol. 13, No. 1 the bird bringing the debris waited 13 to 20 min before its partner left, whereupon it deposited the material. Presumably the remains were being used to line the nest, for the caracaras were also bringing in sticks at this time. I could not reach the nest because the branches at the very top of the tree, where the nest was located, were too small for climbing. Perching and allopreening. The caracaras, especially the female, spent much of each day perching in trees overlooking the marsh. I saw the pair allopreen 6 times. The events on 16 January were representative. The female had finished feeding. Shortly thereafter the male finished, and he approched the female and perched within a few inches of her. When she put her head down, he nibbled at the feathers of her crown. After both birds had preened individually for a few minutes, the male presented the side of his neck and she nibbled at it. More individual preening ended when she again put her head down for him to preen. The male had thus preened the female twice to her once, a consistency shown in other sessions. The two caracaras, when resting and preen- ing, often perched within 30 cm of each other. Copulatory behavior. I witnessed four incomplete copulations— two on the limb where the pair fed, one on a small tree top in the marsh, and one (not well seen) on the nest. In three of the four the male alighted on the back of the female without preliminaries, staying on for about 4 seconds. In the fourth episode the male first presented his mate with food, then mounted while she held the prey in her beak, as described above. Vocalizations. The only vocalizations heard were low, single “wuck” or “g-wuck” notes. The notes were given when one caracara was approaching the other, either with prey, or with debris when one partner was on the nest. Strong winds in the area made study of vocalizations difficult. I never heard the cry, described by Bent (1938) and Slud (1964) that is given from some high perch with a backward toss of the head. Conflict. One of the pair pursued an intruding caracara on two occasions. When the intruder flew near the nest, the two caracaras with claws outstretched, beat their wings as they tried to grapple with each other in the air. In summary, the birds did little flying other than to forage over the marsh. No display flights were observed. The birds often perched close to one another and allopreened, usually after the male had brought prey to the female and both birds had fed. There were no preliminaries to incomplete copulations in three instances but presentation of prey by the male preceded a fourth. The female assumed no consistent precopulatory pose. Acknowledgments I am obliged to James N. Layne and to David H. Ellis for reading and commenting on my observations. Literature Cited Bent, A. C. 1938. Life histories of North American birds of prey. U. S. Natl. Mus. Bull. 170. Brown, L. and Amadon, D. 1968. Eagles, hawks and falcons of the world. McGraw-Hill, New York. Glazener, W. C. 1964. Note on the feeding habits of the Caracara in South Texas. Con- dor 66:162. Spring 1979 Kilham— Courtship of Caracaras 19 Layne, J. N., Lohrer, F. E. and Winegarner, C. E. 1977. Bird and mammal predators on the Cattle Egret in Florida. Florida Field Nat. 5:1-4. Myers, J. P. 1978. One deleterious effect of mobbing in the Southern Lapwing. Auk 95:419-420. Slud, P. 1964. The birds of Costa Rica. Bull. American Mus. Natl Hist. 128:1-430. AN OBSERVATION OF THE AERIAL COURTSHIP OF THE RED-TAILED HAWK by Mark Andrew Springer Department of Zoology Miami University Oxford, Ohio 45056 On April 5, 1976, while in an observation blind 35 feet from a Red-tailed Hawk nest at Alum Creek Reservoir in Delaware County, I observed the aerial courtship of the Red- tailed Hawk ( Buteo jamaicensis). As the pair flew into the thermals, both hawks participated in dives, barrel rolls, and ascents. While coming out of indepen- dent dives, the birds were observed to make contact with their talons. During the three-second contacts, the birds spiraled downward and then separated. Typical dives and ascents followed each of the two encounters. As the female began to return to the nest, the smaller male approached her from above. With legs and tail extended downward, he dropped down and grasped the back of the female. She responded by raising her tail. The copulation lasted about two seconds after which both birds re- turned to the nest tree where normal copulatory behavior commenced. Conner (1974, Bird Banding. Summer Vol., p. 269) noted that the male of a pair of Red-tailed Hawks made aerial contact with its mate, but the female did not respond by lifting her tail. Raptor Research 13(1): 19 CAUTION ON USING PRODUCTIVITY OR AGE RATIOS ALONE FOR POPULATION INFERENCES by James W. Grier, Zoology Department North Dakota State University Fargo, North Dakota 58105 One frequently sees published statements or hears that a population is declining be- cause reproduction is decreasing or because age ratios are changing. Examples are easy to find; I deliberately have not singled out any specific instances here. The con- clusion “declining population” might be true, but it also might not be. Such a con- clusion does not necessarily follow from the observed reproductive or age-ratio changes unless other characteristics of the population, such as time-specific mortality rates, remain constant. The assumption that other population characteristics remain constant is rarely stated and may be very dangerous to the interpretation. The points I wish to make are already knovUi to population biologists (see, e.g., C. J. Krebs, Ecology, Harper and Row, 1978, ch. 9-11; and D. B. Mertz, in: J. H. Con- nell et al. (eds.), Readings in Ecology and Ecological Genetics, Harper and Row, 1970, pp. 4-17). Caughley (/. Wildl. Manage. 38:557-562, 1974) described the problem well: “Age ratios unsupported by other information seem to be statistics in search of an application.” But the principles are not intuitive or obvious. Enough researchers, including raptor workers, are routinely trapped by them that they need to be empha- sized. To illustrate these principles I have chosen hypothetical but realistic examples us- ing standard life-table or life-equation calculations (see Krebs, 1978, ch. 10). The computations were facilitated with a BASIC computer program (which can be used on most computers, including many small office or home models). The program is of general utility and is provided in the Appendix. Any potential users are cautioned, however, that life tables themselves can be misleading; they involve several assump- tions, such as age stability and constant population characteristics. Life tables are use- ful for modeling, as I have done here, but they should not be used for estimating sur- vival rates from band recoveries, a common practice in the past (see note in the program listing). The program given in the Appendix additionally assumes a constant age-specific mortality rate for all birds over one year of age and death before they reach their physiological “old-age” under normal ecological conditions. For example proving that reproductive information alone can be misleading, con- sider the hypothetical situation in table 1. In population B the rate of reproduction is only one-half that of population A, but population B is growing at a rate of 14 per- cent per year while population A is decreasing at a rate of 12 percent per year! If you find that hard to believe, study the table and perhaps perform the calculations youself. The reason for the seeming paradox is that higher survival (lower mortality) more than compensated for the lower reproduction in population B. Table 2 illustrates the problem with age ratios. Again, one of the hypothetical pop- ulations is declining while the other is increasing, but the age ratios are identical! A researcher watching these two populations, perhaps Peregrine Falcons, passing on mi- gration could not distinguish between them on the basis of age ratios. The increasing 20 Raptor Research 13(l):20-24 Spring 1979 Grier— Caution on Population Inferences 21 population does so because higher reproduction is producing more first-year birds and because lowered mortality is resulting in more older birds. But the ratio of first- year to older birds has remained the same. The point is that age ratios are difficult, if not often impossible, to interpret. It is like, in a case of simple division, trying to de- termine the dividend and the divisor when only the quotient is known. Caughley (1974) provides more examples. Age ratios in the form of “catch curves” are occasionally useful in fisheries, but even they are beset with problems. They often have the advantage of strong age- classes to help serve as markers in the population. Usually when age ratios are re- liable, one does not know it unless other data are also available for comparison. But then those other data are sufficient for what one wants to know, and the age ratios are unnecessary. Population trends can be estimated by comparing age-specific reproduction and mortality (i.e., the life-table approach) but only when both reproduction and morta- lity information are available or if a constant time-specific mortality rate can be as- sumed. The only good, clear data that I have seen in an example where reproductive changes are in fact reflected in population changes involves Paul Spitzer’s Osprey data for the northeastern United States (personal communication and presentation at the 1978 RRF Annual Meeting). In addition to the presence of good data on both re- production and the actual total size of the breeding population over a period of years, the Osprey example involves a relatively unique (for raptors) situation: a small but fairly rapidly growing population where intraspecific competition and related mortality probably have not resumed significantly. Hence the mortality rate is prob- ably at a relatively low, constant “background” level. In the case of Peregrine Falcons in the eastern United States (see J. J. Hickey, ed.), Peregrine Falcon Populations, Univ. of Wisconsin Press), reproduction dropped and the falcons disappeared. But the specific relationships betwen reproduction, mortality, time, and actual population changes are far from clear, and we are left with only a rough picture based on hindsight. Reproductive changes might be serious, as they ap- parently were for Peregrines, and we could lose a population by waiting. Or changes in reproduction might not be significant and we can make fools of ourselves and stretch credibility by rushing in and sounding alarms that are not needed. Hindsight can be painful either way. To avoid the problem of not knowing whether estimated reproductive changes are serious and having to wait for hindsight, the solution is to invest more time and effort (and money) to obtain better data for both reproduction and mortality. Trying to make inferences when you have only half the picture can be tricky. As a further safeguard for one’s inferences, it helps to have, if possible, other measures of the population, such as proper sampling surveys and/or mark-recapture programs with adequate proportions of recaptures. In summary, although age ratios and reproductive information occasionally might be useful if used alone, it is risky unless one has additional information (as in P. Spit- zer’s Osprey case) for a backup. Perhaps in the future after more clear examples con- sistently show that assumptions like constant time-specific mortality are reasonable, we can use age ratios and reproductive data alone with more confidence. But until then I urge that we proceed with caution. Acknowledgments I thank D. H. Johnson and H. PostovR for comments on earlier drafts of this ar- 22 RAPTOR RESEARCH Vol. 13, No. 1 tide and P. Spitzer for discussions on his Osprey data. Table 1. Hypothetical example of lowered reproduction in an increasing population. Population 1 Population 2 Average number young per successful nest 6 2 Average number young per adult 1.5 0.5 Age begin breeding 2 2 Proportion of adult females successfully breeding 50% 50% Average number of daughters per successful female 3 1 First-year mortality 70% 35% Annual mortality for older birds 50% 10% ANNUAL RATE OF POPULATION CHANGE -24% + 14% Table 2. Age ratios in two hypothetical Peregrine Falcon populations. Life Table Characteristics Declining Population Increasing Population GIVEN First-year mortality 70% 50% Annual mortality for older birds 25% 20% Age begin breeding 3 3 Breeding success rate for adult females 60% 60% Average number of daughters per successful female 1.15 1.25 CALCULATED Average number of young per adult 0.69 0.75 Maximum age (for an initial cohort of 1,000) 18 24 Annual rate of population change -12% + 2% Age ratios: first year (immature plumage) 32% 31% over first year (“adult” plumage) 68% 69% APPENDIX 10 PRINT 20 PRINT " 30 PRINT 40 PRINT '* 50 PRINT '• 60 PRINT *** AVIAN LIFE TABLE ANALYSIS ***'• BY JAMES GRIER" ZOOLOGY DEPT.. NDSU. FARGO 58105” 70 PRINT BO PRINT "DO YOU WANT INTRODUCTORY INFORMATION (1=YES» 2=N0)"J 90 INPUT AO 100 IF A 0 = 2 THEN 310 CORRECTION: Table 1, page 22, Vol. 13 (1) 1979 Grier — Caution on Population Inferences Table I. Hypothetical example of lowered reproduction in an increasing population Population A Population B Average number young per successful nest 4 2 Average number young per adult 1.0 0.5 Age begin breeding 2 2 Proportion of adult females successfully breeding 50% 50% Average number of daughters per successful female 2 1 First year motality 65% 35% Annual mortality for older birds 35% 10% ANNUAL RATE OF POPULATION CHANCE -12% + 14% Table 2. Age ratios in two hypothetical Peregrine Falcon populations, Life Table Declining Increasing Characteristics Population Population GIVEN First-year mortality 70% 50% Annual mortality for older birds 25% 20% Age of first breeding 3 3 Breeding success rate for adult females 60% 60% Average number of daughters per successful female 1.15 1.35 CALCULATED Average number of young per adult 0.69 0.81 Annual rate of population change -12% + 3.5% Age ratios: first year (immature plumage) 32% 32% over first year (“adult” plumage) 68% 68% Spring 1979 Grier— Caution on Population Inferences 23 110 PRINT 120 PRINT 130 PRINT 140 PRINT 150 PRINT 160 PRINT 170 PRINT 130 PRINT 190 PRINT 200 PRINT 210 PRINT 220 PRINT 230 PRINT 240 PRINT 250 PRINT 260 PRINT 2 70 PRINT 280 REM 290 REM 300 REM 310 PRINT 320 PRINT 330 INPUT 340 PRINT 350 PRINT 360 PRINT 370 INPUT 330 PRINT 390 PRINT "THIS PROGRAM CALCULATES POPULATION GROWTH RATES AND" "A NUMBER OF OTHER POPULATION CHAR ACT ER I S T I CS FOR" "BIRDS ASSUMING A CONSTANT MORTALITY RATE FOR " "INDIVIDUALS OVER ONE YEAR OF AGE AND THAT THE BIRDS" "DIE BEFORE REACHING THEIR PHYSIOLOGICAL LIMITS," •THAT IS, FEW BIROS DIE IN ThE WILD FROM "OLD AGE."* "FOR A REFERENCE FOR SOME GF THE OTHER ASSUMPTIONS." "SYMBOLS AND MATHEMATICS, SEE I C. J. KREBS." " 1978 . ECOLOGY. HARPER AND ROW, PUSL. N.Y. CHAPTER 10." "CAUTION: LIFE TABLES ARE USEFUL FOR MODELING BUT THclR" " INTERPRETATION CAN BE TRICKY AND THEY SHOULD" *• NOT BE USED FOR ESTIMATING SURVIVAL RATES " ” ScE: BROWNIE, ANDERSON, BURNHAM, AND ROBSON," " 1978. U.S. DEPT. OF INTERIOR, FISH & WILDLIFE" " SERVICE, RESOURCE PU8L, NO. 131." NOTE TO PERSONS READING ThE PROGRAM: SOME GF THE VARIABLES USED INTERNALLY (NOT PRINTED) MAY NOT BE THE SAME AS IN THE Tt_XT. E.G., R1 = "R(01" AND R2="L ITTLE R". "WHAT SPECIES ARE YOU WORKING WITH"; SS "NOTE: INDICATE ALL RATES OR PERCENTAGES AS PERCENTAGES ." "WHAT IS YOUR ESTIMATE OF The FIRST YEAR MORTALITY" "rate pgr ";s$;"S"; I "WHAT IS YOUR GUESS FUR ThE ANNUAL MORTALITY" "RATE OF GLOER BIRDS"; 400 4 1 0 420 4 30 440 4 50 460 4 70 4 30 490 500 510 520 530 540 550 560 570 5 30 590 600 6 10 620 630 640 650 660 670 630 690 700 710 720 730 740 750 760 770 7 80 790 800 8 10 820 3 30 840 850 860 870 680 INPUT □ PRINT "AT WHAT AGE DO YOU BELIEVE ";S$;"S NORMALLY" PRINT "BEGIN BREED I NG" 1 INPUT 3 PRINT !"WHAT IS YOUR ESTIMATE OF THE PERCENTAGE OF ADULT" PRINT "FEMALES THAT PRODUCE FLEDGLINGS EACH YEAR"; INPUT F PRINT "WHAT IS THE ANNUAL AVERAGE NUMBER OF FEMALE YOUNG" PRINT "RAISED BY SUCCESSFUL FEMALES (CAUTION: THINK" PRINT "CAREFULLY THIS IS USUALLY CGTAINED BY DIVIDING THE" PRINT "AVERAGE TOTAL NUMBER OF YCUNG PER FEMALE BY 2)"J INPUT Y PR I NT DIM X(75J,L(75),D(75),Q(75),S( 75).M(75),W(75) ,A(7 5) MAT X=ZER(75) MAT L=ZER(75) MAT D=ZER(75) MAT Q=ZER(75) MAT S=ZER(75) MAT W=Z£R<75) MAT M— Z E R ( 7 5) FOR P=1 TO 75 LET X(P)= P-1 NEXT P LET L ( 1 )=1000 LET L(2) = INT( 10 00* ( l-< I /I 00) ) ) FOR P = 2 TO 75 LET L(P + l )=INT(L(P)*( l-( 0/1 00) ) ) IFL(P+1 )=0THEN700 NEXT P FOR P=1 TO 75 LET D(P )=L(PJ-L (P + 1 ) IFD(Pi— 0THEN740 NEXT P FOR P=1 TO 75 IFL< P) = 0THEN780 LET Q(P)=D(P)/L(P) NEXT P FUR P=1 TO 75 LET 3(P)=1-Q(P) I FL ( P ) — 0 THE N 820 NEXT P FOR P=1 TO 75 IFX(P) " , "L ( XIM(X ) *• 1070 PRINT " 1080 FOR P=1 TO 75 1090 PRINTUSING1030.XIP) ,L(P),D(PJ,Q(P)*S(P).M(PI 1100 IFLI PI=0THEN1 120 1110 NEXT P 1120 PRINT 1130 LET T=0 1140 MAT A=ZER(75) 1150 FOR P=1 TO 75 1160 LET AlP)=(EXP(t — R2 )*X(P) )J*M(P) 1170 LET T=T+A(P) l 160 NEXT P 1190 PRINT 1200 PRINT "TABLE 2. ESTIMATE OF INNATE CAPACITY FOR INCREASE-" 1210 PRINT 12 20 PRINT " X" * " E ♦ * C — L I T ' 1230 PRINT 12 40 FOR P=i TO 7 5 1250 PRINT X (P). A ( P ) 1 260 I F L( P ) — 0THEN1280 1270 NEXT P 1280 PRINT 1290 PRINT " •< , n 1300 PRINT " TOTAL". T 13 10 PRINT 1320 PRINT " THE ESTIMATE 1330 PRINT « < " L I TTLE R") 1340 PRINT ' t"BIG R" ) OF 1350 PRINT is*;R2;». or a finite annual rate* ;exp(R2 ) ; • . • 1360 PRINT "ARE YOU SATISFIED WITH THIS ESTIMATE (1=YES* 2=N0) M 1370 INPUT A 1380 IF A=2 THEN 1640 1390 MAT W=ZER(75) 1400 FDR P=1 TO 75 1410 LET W (P )=( ( EXP ( R2 ) 1 - < P— 1420 IFL( PJ=CTHEN1440 1430 NEXT P 1440 LET T=0 1450 FDR P-1 TO 75 . ) ) i * I L ( P J/ 10 0 0 ) 1460 LET T=T+W(P) 1470 NEXT P 1400 PRINT 1490 PRINT "TABLE 3- PERCENTAGE OF POPULATION IN EACH AGE CLASS." 1500 PRINT " (THE STAELt AGE DISTRIBUTION)" 1510 PRINT 1520 PRINT " X" • "PERCENTAGE IN CLASS" 1530 PRINT 1540 FOR P=1 TO 75 1550 PRINT X(P), 100*( W(P)/T) 1560 IFL(P)=OTHEN16lO 1570 NEXT P 1580 PRINT "DO YOU WANT TO RUN ANOTHER LIFt TABLE ANALYSIS 1590 INPUT A 9 1600 IF A9=l THEN 310 1610 PRINT 1620 PRINT "SEE YA" 1630 STOP 1640 PRINT "WHAT INSTANTANEOUS RATE 00 YCO WANT TO TRY THIS 1650 INPUT R2 1660 GOTOl 130 1670 END ( 1 =Y ES » 2 = N0) T IME" ; Spring 1979 Abstracts of Theses ABSTRACTS OF THESES AND DISSERTATIONS 25 THE BIOLOGY OF THE NEW ZEALAND FALCON (Falco novaeseelandiae Gmelin 1788). Morphological differences of 232 New Zealand Falcons were investigated by mea- suring nine parameters. Three allopatric forms occupying different habitats are recog- nized. These are small, dark forest-living “Bush Falcons,” large, pale “Eastern Fal- cons” living in predominantly open montane areas, and intermediate “Southern Falcons” inhabiting Fiordland, New Zealand, and the Auckland Islands. Seventeen adaptive features of New Zealand Falcons were compared with those of congeners to investigate possible taxonomic relationships. The preliminary opinion is reached that F. novaeseelandiae is most closely related to F. deiroleucos, F. rufigularis, and F. femoralis. Eight factors which might be adaptive toward sexual dimorphism are briefly discussed, and the hypothesis was reached that the strongest selection pressure toward sexual dimorphism operates through the attacking ability of juvenile raptors, more so in those species which specialize in the attacking aspect of hunting and less so in those which specialize in the searching aspect. The diet of Eastern Falcons was deduced by analysis of 932 pellets and 661 prey remnants, representing about 1,434 prey individuals of about 32 species captured by 30 pairs of falcons. Analyses of 21 stomachs and 68 attacks on prey by wild falcons are documented. About 80% of prey individuals (61% by biomass) were small passer- ines, particularly the Yellowhammer (Emberiza citrinella), Greenfinch ( Chloris chloris). Skylark (Alauda arvensis ), New Zealand Pipit (Anthus novaeseelandiae ), and Black- bird ( Turdus merula ). About 3.2% of individuals (38% by biomass) were introduced mammals, mainly Brown Hares ( Lepus europaeus) and European Rabbits ( Oryctolagus cuniculus ). Skinks ( Leiolopisma sp.) and insects (16.4% of individuals) comprised only 0.8% by biomass. The dietary values of preys and the food consumption of two New Zealand Falcons were measured experimentally and used to calculate the prey re- quirements of a breeding pair of falcons. Data on 280 rangle stones from 16 pairs of wild New Zealand Falcons are given. The hunting strategies of raptors are briefly reviewed and 7 search and 4 attack behaviors recognized and described. Thirty-eight hunts by wild New Zealand Falcons and 194 hunts by trained or semitrained (at hack) falcons are categorized. Instances of nest-robbing and food storing are described. Methods used to find 144 nest areas in 5 study areas totalling 7,800 km 2 in South Island, New Zealand, and the known history of each site are given. Ninety-four nest- ing territories in area A and 21 in area C averaged 3.80 and 3.95 km apart. Orienta- tions and heights of 44 roosts of 3 types, and orientations and descriptions of 42 nests of 4 types are given. Nests were found on cliff ledges, ground ledges, under logs in forest, and among epiphytes in trees. Tape recordings and sonograms of 6 calls are supplied and 19 postures or behaviors are described and illustrated. Courtship and stages of the reproductive cycle are outlined. Mean clutch size was 2.68 (mode = 3, n = 25), mean egg size was 48.7 X 36.7 mm (n = 69), mean brood size for 32 nesting attempts was 1.88 (range 0-4), for 23 successful nestings, 2.61. Incubation was about 30-32 days, recycling time about 16 days. Data on weight gains, tarsus and rectrix growth, and developmental stages of chicks are supplied. The methods and results of 26 RAPTOR RESEARCH Vol. 13, No. 1 a successful captive breeding project are given. Organochlorine levels in pectoral muscles of 12 New Zealand Falcons and 14 other New Zealand birds of prey are tabulated. Five juvenile falcons contained a mean of 2.57 mg total DDT per kg wet weight of muscle, 5 adults 13.74 mg/kg. Shell-thin- ning since 1948 was 0-3.3% in 15 shells. The known status of New Zealand Falcons in all parts of New Zealand is given and summarized in two distribution maps. About 3,100-3,200 pairs of Eastern Falcons, 450-850 pairs of Bush Falcons and 140-270 pairs of Southern Falcons may exist. Probably 3,000-4,500 pairs of New Zealand Fal- cons is a realistic estimate. Fox, Nicholas C. 1977. The biology of the New Zealand Falcon ( Falco novaesee- landiae Gmelin 1788). Ph.D. thesis. University of Canterbury, Christchurch, New Zealand. 421pp. ECOLOGY OF THE WHITE-TAILED KITE IN SAN DIEGO COUNTY White-tailed Kites were studied from April 1975 to March 1978 in San Diego, Cal- ifornia. The ecology and behavior of both breeding and nonbreeding kites were exam- ined. Data were obtained on nest sites, nesting success, diet, foraging activity, prey den- sity, and communal roosting behavior. Of 26 nests studied, 20 contained eggs (mean of 4.0 eggs per clutch) and 17 fledged young (mean of 1.9 fledglings per clutch). Four pairs produced a second clutch during the 1977 nesting season. Territoriality existed to varying degrees. Intraspecific territoriality was uncommon yet readily visible. Two pairs of kites were evicted from their nesting territories by pairs nesting nearby. Interspecific territoriality was most commonly seen against Red- tailed and Marsh hawks. Other bird species were attacked less frequently. In 2886 pellets analyzed, 3266 prey animals were represented. Of these 2759 (84.5%) were Microtus californicus, 344 (10.2%) Reithrodontomys megalotis, 143 (4.4%) Mus musculus and 30 (0.9%) other organisms. Only five birds and no reptiles or in- vertebrates were found in the pellets. With use of the runway analysis technique, the relative Microtus population den- sities were determined in the kite’s hunting areas for each site. A stepwise multiple regression anaylsis was used to relate these variables (nest data, diet, and prey density) to number of eggs, number of hatchlings, number of fledgl- ings, and nesting success of each nest site. The number of eggs was related to the mean number of active Microtus runways (prey density) in the kites’ hunting areas and to the height of the nest in the ( R 2 = 52.0%). Number of hatchlings were not significantly related to any of the variables. The number of fledglings, and nesting success, were related to the percent of Micro- tus in the kite’s diet (R 2 = 26.9% and R 2 = 41.1% respectively). From these results it was determined that the number of eggs is related to the prey density, and the number of fledglings and nesting success are related to the percent of Microtus in the diet. Spring 1979 Abstracts of Theses 27 Communal roosting sites were observed at four locations in San Diego County. A roost at Sorrento Valley was observed at least once per week from early October, 1977 to late March, 1978. Abundant rainfall flooded the valley for most of Decem- ber, 1977 and February, 1978, decimating most of the Microtus population. As the prey density in the valley decreased, so did the number of kites using the valley for hunting purposes and for communal roosting. I found that the kites entered and left the roost in response to conditions of visibility, which was closely related to light in- tensity except under foggy conditions. Wright, Bruce Albert. 1978. Ecology of the White-tailed Kite in San Diego County. M.S. thesis. San Diego State Univ. San Diego, Calif. 60 pp. Present Address: Rt. 4 Box 4617-1, Juneau, Alaska 99803. ECOLOGY OF WINTERING BALD EAGLES ON THE SKAGIT RIVER, WASHINGTON Winter ecology and behavior of a Bald Eagle ( Haliaeetus leucocephalus ) wintering population were studied in the winters of 1973-74 and 1974-75. Analysis was under- taken of the distribution of the eagles along the river in relationship to the distribu- tion and abundance of the food source. Wintering Bald Eagle habitat selection in re- lation to habitat availability, distribution, and human disturbance were also described. Criteria for aging sub-adult Bald Eagles in the field were substantiated through molt research on captive eagles. Plumage aging techniques were used to determine the dif- ferential arrival and departure dates of different age classes, and behavioral relation- ships between eagles of different ages in the wintering area. The main food source of the Skagit wintering population is dead salmon ( Oncorhynchus spp.). Eagles were never observed to kill live salmon. Eagle numbers were correlated to the amount of available salmon. When most salmon carcasses were either washed away by river cur- rents or consumed by eagles, the wintering population dispersed and left the area. Eagles were concentrated in a seven mile stretch of river and were further concen- trated within this seven mile stretch at certain gravel bars where salmon carcasses and perching sites were abundant. The population begins to arrive in mid-October with adults arriving first. Most sub-adults arrive in early December. The eagle popu- lation peaks were 93 eagles in mid-January 1974 and 165 eagles in mid-February 1975. The lower population level in 1974 was influenced by a flood on 16 January which prematurely removed most salmon carcasses that year. The Bald Eagle popu- lation disperses from the Skagit area during March, and few eagles remain after 1 April. The average percentage of sub-adults in the population was 52.6%. This figure is higher than all other wintering sub-adult percentages except Shea’s 1971 figure of 54.5% in Glacier National Park, Montana. This may indicate a healthy, productive population, however, sub-adult percentages cannot be utilized to determine popu- lation productivity until much more is known about the winter distribution and habi- tat selection of different age classes. Eagle activity was affected by weather condi- tions. High winds and clear skies stimulated soaring and flying activity. Consequently, high eagle counts occurred during calm periods with low overcast skies when most 28 RAPTOR RESEARCH Vol. 13, No. 1 eagles were perching along the river. During sunny, windy weather, eagles soared in groups and did not display usual feeding and distribution patterns. The social func- tion of such group soaring is discussed. Eagles initially utilized areas on the river that were isolated from human disturbance, and only when food was depleted in these areas did the eagles use sites close to human disturbance. Of 3,322 eagle observations in 1974-75, 68.5% were on the side of the river having no road access; 19.7% were on islands in the river; and 11.8% were on the side of the river where the main road is, and most human activity occurs. Management alternatives to minimize human dis- turbance and preserve eagle wintering habitat are discussed. Servheen, Christopher W. 1975. Ecology of Wintering Bald Eagles on the Skagit River, Washington. M.S. thesis, University of Washington, Seattle. 96 pp. THE INFLUENCE OF FORCED-RENESTING ON REPRODUCTIVE PARAMETERS OF CAPTIVE AMERICAN KESTRELS Abstract From 1974 to 1977, the first clutches of 78 pairs of captive American Kestrels ( Fal - co sparverius ) were removed to induce laying of replacement clutches. This pro- cedure was termed forced-renesting. First clutches were artificially incubated and the hatchlings hand-reared to fledging age. A Maximum Likelihood Program revealed that replacement clutches had fewer eggs, longer eggs, and eggs with thicker shells than first clutches; but they did not differ in fertility, hatchability, overall growth, and fledging success of young. Clutch size, egg length, eggshell thickness, and fresh-egg weight declined seasonally. Hatch- ling weight and fresh-egg weight were highly correlated, but neither was a reliable index of growth beyond 6 days of age. Hand-rearing was associated with slower growth rates and the production of phys- ically smaller adults. Hand-reared females laid the largest clutches and the largest and heaviest eggs and were associated with higher fertility than hand-reared males. The implications of forced-renesting are discussed. Bird, David M. 1978. The influence of forced-renesting on reproductive parameters of captive American Kestrels. Ph.D. thesis, McGill University, Montreal, 111 pp. Spring 1979 Abstracts of Theses 29 BEHAVIORAL ECOLOGY OF RED-TAILED HAWKS (. BUTEO JAMAICENSIS), ROUGH-LEGGED HAWKS (B. LAGOPUS), NORTHERN HARRIERS ( CIRCUS CYANEUS ), AMERICAN KESTRELS (FALCO SPARVERIUS ), AND OTHER RAPTORIAL BIRDS WINTERING IN SOUTH CENTRAL OHIO Abstract A field study of the behavioral ecology of sympatric Red-tailed Hawks (RTH), Rough-legged Hawks (RLH), Northern Harriers (NH), American Kestrels (AK), and other raptorial birds wintering in a 73.5 km 2 tract in south central Ohio was con- ducted during the winters of 1973-74 through 1976-77. Ten species of Falconiformes and three species of Strigiformes were seen in the area during the four winters of study. RTH, RLH, NH, and AK comprised over 99 percent of the individuals seen. The size and composition of this four-species raptor community remained remarkably constant during the four winters of study, ranging from 1.22 birds/km 2 in 1976-77 to 1.40 birds/km 2 in 1975-76. Interspecific spatial overlap was greater in areas of high use than in areas of mod- erate and low use. Intraspecifically, NH showed the greatest amount of interseasonal overlap, and AK showed the least. The four species all used cropland more than expected and grazed pastures less than expected on the basis of habitat type availability. Interspecific differences in all species pairs occurred with respect to vegetation type, field size, and field slope se- lected. Activity dependent shifts in vegetation type use occurred in all four species. Intraspecific differences in vegetation type use occurred in NH. The four species differed in use of fields bordered by woodlots, tree rows, or roads. RTH, RLH, and NH avoided areas within 75 m of active farmsites whereas AK preferred them. De- pending on the field characteristic examined, niche breadths (J’) differed considerably among the four species, and when these are viewed in conjunction with changes in the degree of niche overlap («c) between species pairs for each of the field character- istics, an extremely complex pattern of niche overlap and habitat use emerges. RLH, NH, and AK all showed significant shifts in temporal activity. While RTH, RLH, and AK did not differ from one another with respect to temporal patterning, all three species did differ significatly from NH. Significant temporal shifts in the percentage of birds flying when sighted occurred in RTH and RLH, but not in NH or AK. Interspecific differences in the ratio of flying to perched birds occurred in all paired species comparisons. Only NH were observed more frequently flying than perched, and only NH pounced more frequently from flight than from a perch. The four species differed with respect to both perching height and perching substrates. The three tree perching species (RTH, RLH, and AK) partitioned that resource ac- cording to the height of the tree, their perching height, the location of the tree, and their perching location in the tree. The four species differed in the amount of flap- sailing, hover-flying, and soaring they engaged in, as well as the height and speed at which they flap-sailed. Harriers differed intraspecifically in the flight types they em- ployed and in the height and speed at which they flew. 30 RAPTOR RESEARCH Vol. 13, No. 1 There were intraspecific differences in pouncing success and hunting bout success. Juvenile harriers were less proficient hunters than adult harriers. In all species hunt- ing success depended on vegetation type hunted. In NH and AK hunting success de- pended on pounce type. Variations in temperature, relative humidity, solar radiation, precipitation, and wind velocity were accompanied by shifts in raptor activity. Numerically small mammals made up the bulk of RLH and NH diets, and in all species but RTH small mammals comprised the majority of the diet by biomass. RTH pirated more prey from other raptors than did the other three species. Adult male harriers took more avian prey than did adult female or unsexed juvenile harriers. The percent of insects in the diet of AK increased with increasing temperatures. One hundred three interspecific and 69 intraspecific encounters among RTH, RLH, NH, and AK were observed. In 20 percent of the encounters, prey robbery or carcass displacement was attempted. Both RLH and NH were victims of piracy. While overal species overlaps (oC) indicated a similar degree of niche overlap for all four species, RTH, NH, and AK all exhibited at least one overwhelming difference in their niche from the other three species while RLH did not. On the basis of the relative degree of overlap, activity and diet, rather than habitat, were the most im- portant niche dimensions, and time was least important. Weather-related changes in the behavior of the four species resulted in shifts in the diversity of niche parameters that were of the same order of magnitude as interspecific differences in the same pa- rameters. Viewed in their entirety, these data indicate considerably complex resource partitioning among the four species comprising this open-habitat raptor guild. Bildstein, K. L. 1978. Behavioral ecology of Red-tailed Hawks ( Buteo jamaicensis). Rough-legged Hawks (B. lagopus), Northern Harriers ( Circus cyaneus ), American Kes- trels ( Falco sparverius), and other raptorial birds wintering in south central Ohio. Ph.D. dissertation. The Ohio State University, Columbus. 364 pp. Present address: Keith L. Bildstein Department of Biology Winthrop College Rock Hill, South Carolina 29733 ASPECTS OF THE BIOLOGY OF THE AUSTRALASIAN HARRIER (CIRCUS AERUGINOSUS APPROXIMANS PEALE 1848) Abstract The study is based on 18 months of intensive fieldwork on the southwestern coast of the North Island of New Zealand. During this time 212 Australasian Harriers were trapped, re trapped, measured, sexed, aged, individually marked, and observed. Fort- nightly observations of the individually marked population were made over a further seven months. The Australasian Harrier and European Marsh Harrier are considered to be conspecific. Evidence is presented showing that there is no valid reason for considering Circus aeruginosus of the Pacific Islands to be a different subspecies from C. aeruginosus of Australia and New Zealand. During the breeding season ten terri- Spring 1979 Abstracts of Theses 31 tories in the 12 km 2 study area averaged 31 ha, nest sites averaged 910 m apart, pairs’ overlapping home ranges averaged 9 km 2 , and favourite hunting areas 3 km 2 . A high population density of one bird per 50 ha was calculated. A low fledging success rate of 1.8 young per successful pair and 1.1 young per nest site and two cases of po- lygyny were recorded during two breeding seasons. Territorial and courtship behav- iour, nest parameters, and the parental division of labour are described. Seasonal movements and the dispersion of age and sex classes from the study area at the end of the breeding season are described. Most (66.7%) individually marked adults re- turned after the autumn dispersal phase and established winter home ranges aver- aging 9 km 2 . The home range of an adult female in open farmland was calculated to be 14 km 2 using radiotelemetry techniques. A nonbreeding season population density of one bird per 80 ha was calculated. Communal roosting, which occurred through- out the year, is discussed. Four hundred and seventy food items were identified in the diet from pellets, prey remains, stomach contents, and field observations. In descend- ing order of numerical importance in the diet were mammals (46.4%), introduced passerines (29.0%), insects (7.6%), game birds (6.7%), birds’ eggs (4.8%), and aquatic prey (4.6%). Australasian Harriers ate significantly greater numbers of live prey than carrion annually. Adults took significantly greater numbers of agile food items than juveniles. Females ate significantly more large ( 4 200 g) and fewer agile food items than did males. Seven search techniques and five attack techniques, including some buteonine techniques, are identified and described in the Australasian Harriers’ wide range of hunting techniques. Ninety-five attacks on prey are recorded, and 15.8% of them were successful. Adults were significantly more successful hunters than juveniles. Cooperative hunting, hunting in the daily cycle, feeding behaviour at carrion, inter- specific competition for carrion, interspecific disruption of hunting, and prey escape tactics are described. From a computer analysis of hunting behaviour data it is con- cluded that adult males are more maneuverable and less conspicuous than adult fe- males and juveniles because they flew significantly lower and faster. Adult males also hunted, to a significantly greater degree, those habitats where there were greater numbers of agile prey. The hunting inexperience of juveniles was quantified. The Australasian Harrier is moderately sexually dimorphic. Current hypotheses proposed to explain the degree of sexual dimorphism in raptors and why the females of most raptor species are larger than males are critically reviewed. Baker-Gabb, David J. 1978. Aspects of the biology of the Australasian Harrier ( Cir- cus aeruginosus approximans Peale 1848). M.Sc. thesis, Massey University, Palmerston North, New Zealand. 221 pp. Present address: Zoology Department Monash University, Clayton Victoria 3168, Australia. BEHAVIORAL AND PREDATORY DYNAMICS OF AMERICAN KESTRELS WINTERING IN THE ARCATA BOTTOMS A field study of American Kestrels ( Falco sparverius) was conducted in the Areata Bottoms, Humboldt County, California, during the winters of 1972-73 and 1973-74, 32 RAPTOR RESEARCH Vol. 13, No. 1 to evaluate the influence of season, chronological hour, and environmental conditions on time and activity budgets, predatory efficiency, and prey capture. Data were collected on the behavior of individual Kestrels observed from dawn to dusk. All behavioral acts were recorded with respect to time of day and elapsed time. Special attention was given to their predatory behavior (e.g., mode of hunting, pre- datory success, prey species captured, and food consumption). In addition to intensive observation of the behavior of Kestrels, monthly road cen- suses were conducted in the Areata Bottoms to obtain density indices of wintering raptor populations. During each field season, densities of small mammal prey species were estimated by periodic trapping in areas Kestrels commonly hunted. Kestrels were 45.4 percent successful at capturing prey on 635 attempts in 1972- 73; of the prey captured, 32 (11.1 percent) were vertebrates and 257 (88.9 per- cent) were invertebrates. In 1973-74, Kestrels were 60.7 percent successful on 1,198 attempts; of the prey captured, 46 (3.8 percent) were vertebrates and 1,167 (96.2 per- cent) were invertebrates. The decrease in the percentage of vertebrates captured in 1973- 74 compared to 1972-73 corresponded with a drastic reduction in abundance of small mammals. The increase in predatory efficiency during 1973-74, reflected the greater efficiency of Kestrels in capturing invertebrates over that of capturing verte- brates. The modes of hunting used by Kestrels were perch-hunting, hover-hunting, and flight-hunting. In both years, perch-hunting was most efficient, 51.8 percent and 66.9 percent, respectively; followed by flight-hunting, 23.3 percent and 48.8 percent, re- spectively; and hover-hunting, 23.8 percent and 26.9 percent, respectively. The pre- datory efficiency of each mode of hunting was inversely related to the proportion of vertebrate prey captured. Caching behavior by American Kestrels is described. In 1972-73, 20 acts of prey caching behavior were observed; of these, 11 involved food storage and 9 involved food retrieval. Unsuccessful attempts to retrieve prey were observed 6 times in 1972-73. In 1973-74, 59 acts of prey caching behavior were observed; of these, 36 involved food storage and 23 involved food retrieval. Six unsuccessful attempts to re- trieve prey were witnessed in 1973-74. The relative proportion of prey species in- volved in caching behavior corresponded directly to the relative proportions cap- tured. In 1972-73, 16 (88.9 percent) of the prey items cached were small mammals, and 1973-74, 28 (63.6 percent) were frogs. The relevance of this study to the functional and numerical components of pre- dation; a comparative analysis of the interrelationships among predatory efficiency, hunting strategies, and taxons of prey; the relevance of time and activity budgets, body size, and foraging strategies to energy economy are discussed. Collopy, M. W, 1975. Behavioral and predatory dynamics of American Kestrels wintering in the Areata Bottoms. M.S. thesis. Humboldt State University, Areata, Cal- ifornia. xiv + 211 pp. Present address: School of Natural Resources University of Michigan Ann Arbor, MI 48104