(ISSN 0892-1016) Contents Experimental Design and Data Analysis for Telemetry Projects: SUMMARY OF a WORKSHOP. Vicky J. Meretsky, Technical Editor 125 Applications and Considerations for Wildlife Telemetry. Mark R. Fuller 126 Experimental Design of Telemetry Projects. Ken Pollock 129 Analysis of Survival Data from Telemetry Projects. Christine m. Bunck . . 132 Basic Techniques for Analyzing Movement and Home-Range Data. Vicky J. Meretsky 135 Detecting and Describing the Structure of an Animal’s Home Range. Paul M. Geissler and Mark R. Fuller 138 Telemetry in Studies of Predation, Dispersal and Demography. Robert Kenward 1 39 Telemetry Techniques for the Study of Raptor Migration. William W. Cochran 142 Radio Telemetry in the Study of Raptor Habitat Selection. W. Grainger Hunt 144 Electroretinographic Responses of the Great Horned Owl {Bubo virgimanus). Steven J. Ault and Edwin W. House 147 Short Communications Snowy Owl Numbers on Twelve Queen Elizabeth Islands, Canadian High Arctic. Frank L. Miller 153 A New Method to Selectively Capture Adult Territorial Sea-Eagles. Anthony L. Hertog 157 Northern Cardinal Head Attached to the Toe of a Sharp-shinned Hawk. Thomas W. Carpenter and Arthur L. Carpenter 160 News and Reviews 160 Index to Volume 21 161 9kgfcs|c3|E4t4titE9|E4t)|s3|c4;4c$ifc4t$)ic4i$ 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 1987 by The Raptor Research Foundation, Inc. Printed in U.S.A. THE JOURNAL OF RAPTOR RESEARCH A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC. Vol. 21 Winter 1987 No. 4 J Raptor Res. 21(4):125 © 1987 The Raptor Research Foundation, Inc. EXPERIMENTAL DESIGN AND DATA ANALYSIS FOR TELEMETRY PROJECTS: SUMMARY OF A WORKSHOP Vicky J. Meretsky Technical Editor A workshop on experimental design of telemetry projects and analysis of telemetry data was held at the 1985 Raptor Research Foundation Symposium on Management of Birds of Prey held in Sacramento, California. Speakers stressed the need for careful research design, thorough knowledge of study area and telemetry hardware to be used, flexibility and luck. Design and analysis techniques for mortality, home- range, habitat, migration, predation, dispersal and demography studies were discussed. The workshop was divided into two sessions: technical aspects of design and analysis, and application of techniques in the field. Speakers addressing field applications were asked to focus on examples of actual study designs and problems. Expanded reference sections accompany individual summaries and include general, theoretic and field study references. While many of the studies discussed involve hypothesis-testing research, exploratory techniques and situations are also treated. Perhaps less fashionable than hypothesis- testing, exploratory studies are a part of scientific investigation and, as such, can benefit from thoughtful experimental design and careful data analysis. We gratefully acknowledge the efforts of R. R. Olendorff and J. M. Scott in organizing the workshop, and The Raptor Research Foundation, Inc., for providing the opportunity and funding to make it possible. Reviews by L. David Mech, Gary C. White and Jimmie R. Parrish greatly improved the manuscript. Special thanks to the Condor Research Center for serving as intermediary in the editing process. Condor Research Center, 2291A Portola Road, Ventura, CA 93003. Proceedings received 25 February 1987; accepted 20 September 1987 Editor’s Note: Pages 125-146 of this issue represent summaries of material presented at the 1985 Telemetry Workshop. 125 J Raptor Res. 2 1 (4): 1 26- 1 28 © 1987 The Raptor Research Foundation, Inc. APPLICATIONS AND CONSIDERATIONS FOR WILDLIFE TELEMETRY Mark R. Fuller This presentation is a review of radio telemetry; that is, the sending of information over some distance using radio frequencies. The technique is a form of biotelemetry, which also includes laboratory/phys- iology applications wherein signals can be trans- mitted from subject to receiver/recorder. Radio te- lemetry is a research tool. When using telemetry, it is essential to consider how it is to be used to achieve one’s objectives and what time and money will be required. Telemetry can be used by the researcher to ac- curately locate animals for further observation, to determine home range, habitat use, migration routes, activity patterns, predator-prey relationships, sur- vival, and to locate nests, roosts, etc. Transmitters that gather data on the microclimate of the animal have also been developed. In addition telemetry has been used to obtain physiological data. Examples of many of these uses were presented by Amlaner and Macdonald (1980). Transmitter attachment techniques are as diverse as the size, shape, weight and application of the transmitters. References at the end of this summary provide an introduction to the general literature on wildlife telemetry techniques, and Kenward’s paper (1985) provides a good review of raptor telemetry. Successful attachment techniques are far better doc- umented than failures. Therefore, before trying new methods and equipment, check with researchers ex- perienced with similar techniques and species. To summarize briefly, tail mounts can only be used with comparatively light-weight transmitters and are lost when feathers are molted. Backpack transmitters require suitable harness material; teflon ribbon has been useful on raptors. Glue-on transmitters have not been used often on raptors. Transmitters pow- ered by solar cells are very lightweight and can be used alone or with rechargeable batteries to provide nighttime coverage (Wischusen 1981; Kenward 1987). Solar transmitters are not compatible with attachment techniques that allow birds to preen feathers over the transmitter or with animals in- habiting dense vegetation. Before beginning a radio telemetry study, biolo- gists should determine how long the animals must be monitored. Larger (and heavier) batteries provide longer life and stronger radio signals. Consider com- promising between time and weight. Make sure that the company supplying your equipment understands your needs and has experience with similar appli- cations. Design engineers usually provide optimistic estimates of transmitter life, based on signal strength and pulses of their products. However, many vari- ables that affect equipment performance on an an- imal in the field cannot be factored into basic elec- tronics considerations. Time frame of the study will also determine ap- propriate attachment techniques. Many harness ma- terials (e.g., teflon) last for months or years. Pres- ently, few attachment methods (e.g., glue, fasteners for harness material) have been developed to reliably detach at pre-determined durations after attachment. However, David Garcelon (Institute for Wildlife Studies, P.O. Box 127, Areata, CA 95521) has had success developing a drop-off attachment for Bald Eagles ( Haliaeetus leucocephalus). Tail mounts are useful for studies not extending beyond a molt. Behavioral and energetic changes in animals car- rying radio transmitters have been incompletely doc- umented for a few species and are completely un- documented for many others. In the short term there may be a period of up to several days of reduced activity as the animal adjusts to a harness and trans- mitter. Over a longer period, brood abandonment, icing and tangling have been documented for a va- riety of mammals and birds. Some diving ducks will not feed normally with harness attachments (but see Questions section). Jim Gessaman (UME 53, Utah State University, Logan, UT 84322) and Mark Fuller (Patuxent Wildlife Research Center, USFWS, Laurel, MD 20708) are investigating energetic im- plications of additional weight and thermal effects of large transmitters (as a heat sink). A recent article by Caccamise and Hedin (1985) deals with bird size and appropriate transmitter weight. In general transmitter weight should be a smaller percentage of body weight for larger birds than for smaller birds. Transmitter weight affects potential maximum ve- locity, maximum power and endurance. As a result, escape speed, pursuit speed, payload, persistence of 126 Winter 1987 Telemetry Project Design 127 Table 1. Suppliers of telemetry equipment (compiled June 1987). Advanced Telemetry Systems, Inc. 23859 NE Highway 65 Bethel, MN 55005 (612) 434-5040 Austec Electronics, Ltd. #1006, 11025-82 Ave. Edmonton, Alberta T6G 0T1 CANADA (403) 432-1878 AVM Instrument Co., Ltd. 2368 Research Dr. Livermore, CA 94550 (415)449-2286 Bally Ribbon Mills 23 N. 7th St. Bally, PA 19503 (215)845-2211 (For teflon ribbon harness material) Beacon Products Co. 360 East 4500 South Salt Lake City, UT 84107 (801)265-1383 Biotrack Stoborough Croft Grange Rd., Stoborough Wareham, Dorset BH20 5AJ ENGLAND Wareham (09295) 2992 B & R Ingenieurgesellschaft mbH Johann-Schill-Str. 22 7806 March-Buchheim, WEST GERMANY Custom Electronics of Urbana, Inc. 2009 Silver Ct. West Urbana, IL 61801 (217)344-3460 Custom Telemetry and Consulting 185 Longview Dr. Athens, GA 30605 (404) 548-1024 Holohil Systems Ltd. RR 2 Woodlawn, Ontario CANADA K0A 3M0 (613) 832-3649 J. Stuart Enterprises P.O. Box 310 Grass Valley, CA 95945 L.L. Electronics P.O. Box 247 Mahomet, IL 61853 (215)586-2132 Lotek Engineering, Inc. 11 Younge St. S Aurora, Ontario CANADA L4G 1L8 (416)727-0181 Microwave Telemetry 610 Chestnut Ave. Towson, MD 21204 Midwest Telemetry Judy Montgomery P.O. Box 773 Urbana, IL 61801 (217)367-1904 Narco Scientific (short range-biomed) 7651 Airport Blvd. P.O. Box 12511 Houston, TX 77017 (713)644-7521 Remote Monitoring Systems P.O. Box 2155 Walla Walla, WA 99362 (509)529-1060 Scien-O-Tech Consultants, Ltd. Box 14426 NAIROBI or Box 87054 Mambasa, KENYA Telemetry Systems, Inc. 11065 N. Lake View Dr. P.O. Box 187 Mequon, WI 53092 Owner — Owen Royce (414)241-8335 Telonics 932 Impala Ave. Mesa, AZ 85204-6699 Owner — Dave Beaty (602)892-4444 Wildlife Materials, Inc. R.R. 1 Carbondale, IL 62901 Wildlife Consultant — R. E. Hawkins (618)549-6330 128 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 fat reserves, flight distances and stopover times can be affected. Added drag affects aerodynamic perfor- mance and can alter a bird’s center of gravity. C. J. Pennycuick (Department of Biology, University of Florida, Coral Gables, FL 33124) has suggested tests for transmitter effects on various flight behav- iors and mechanics, and Pennycuick and Fuller are studying some of these aspects. Given that little is known of the impacts of transmitters on animals, it might be desirable to recapture test subjects and remove transmitters, which is often difficult and time consuming. A little-known fact about radio telemetry is that one needs a license from the Federal Communica- tions Commission (FCC) to conduct telemetry stud- ies. There are restrictions on frequency, power out- put, numbers of transmitters per unit area, etc. Kolz (1983) has published an informative article on the subject, giving pertinent restrictions. The U.S. Fish and Wildlife Service Bird Banding Laboratory, Lau- rel, MD 20708, provides brief information on reg- ulations, and a list of companies that manufacture wildlife telemetry equipment (Table 1). Questions William Cochran stated that Judy Montgomery (Midwest Telemetry, see Table 1) has designed a neck collar transmitter attachment that appears not to interfere with the normal feeding activity of diving ducks. References Amlaner, C. J., Jr. and D. W. Macdonald. (Eds.). 1980. A handbook on biotelemetry and radio tracking. Pergamon Press, New York, NY. Caccamise, D. F. AND R. S. Hedin. 1985. An aerody- namic basis for selecting transmitter loads for birds. Wilson Bull. 97:306-318. Cheeseman, C. L. and R. B. Mitson (Eds.). 1982. Telemetric studies of vertebrates. Proc. Symp. Zool. Soc. London. Academic Press, New York, NY. Cochran, W. W. 1980. Wildlife telemetry. Pages 507- 520. In S. D. Schemnitz, Ed., Wildlife management techniques manual, 4th ed. Wildlife Society, Wash- ington, D.C. (lists references for attachment tech- niques) International Conference on Wildlife Biote- lemetry. Various years and places. International Symposium on Biotelemetry. Various years and places. Kenward, R. 1985. Raptor radio-tracking and telem- etry. Pages 409-420. In I. Newton and R. D. Chan- cellor, (Eds.). Conservation studies on raptors. ICBP Technical Publication No. 5. Cambridge, UK. . 1987. Wildlife radio tagging: equipment, field techniques and data analysis. Academic Press, NY. Kolz, A. L. 1983. Radio frequency assignments for wildlife telemetry: a review of the regulations. Wild 1. Soc. Bull. 1 1:56-59. Mech, L. D. 1983. Handbook of animal radio-tracking. University of Minnesota Press, Minneapolis, MN. 107 pp. Patric, E. F., G. A. Shaughnessy and G. B. Will 1982. A bibliography of wildlife telemetry and radio tracking. Department of Forest and Wildlife Man- agement, College of Resource Development, University of Rhode Island. Contribution No. 2054, Rhode Island Agricultural Experiment Station, Univ. of Rhode Is- land, RI 02881. Wischusen, E. W. 1981. Population dynamics, behav- ior, and habitat use of the Red-tailed Hawk in we- stcentral Alabama. Unpub. M.Sc. Thesis. The Uni- versity of Alabama, Dept, of Biology, P.O. Box 1927, University, AL 35486. Effects of Transmitters on Behavior: Gilmer, D. S., I. J. Ball, L. M. Cowardin and J. H. Reichmann. 1974. Effects of radio packages on wild ducks. J. Wild l. Manage. 38:243-252. Horton, G. I. and M. K. Causey. 1984. Brood aban- donment by radio-tagged American Woodcock hens. / Wild! Manage. 48:606-607. Karl, B. J. and M. N. Clout. 1987. An improved radio transmitter harness with a weak link to prevent snagging. J. Field Ornith. 58:73-77. Nenno, E. S. and W. M. Healy. 1979. Effects of radio packages on behavior of Wild Turkey hens. J. Wildl Manage. 43:760-765. Perry, M. C. 1981. Abnormal behavior of canvasbacks equipped with radio transmitters. J. Wildl. Manage. 45' 786-789. U.S. Fish and Wildlife Service, Patuxent Wildlife Re- search Center, Laurel, MD 20708. / Raptor Res. 21 (4): 129-1 31 © 1987 The Raptor Research Foundation, Inc. EXPERIMENTAL DESIGN OF TELEMETRY PROJECTS Ken Pollock In its short life radio telemetry has progressed from a “fascination” stage characterized by small studies of poor design based on unrealistic expec- tation to a stage of more sober reassessment. At present we are seeing studies of possible problems involved with telemetry, composite studies which test telemetry against other techniques and a general atmosphere of more cautious expectation. In the fu- ture we can hope to work with telemetry as a thor- oughly researched tool with known strengths and weaknesses for which accompanying texts and sound analysis techniques are widely available. Telemetry studies are generally costly and there- fore tend to be multipurpose so that many questions can be addressed from one data set. This in turn leads to tradeoffs among sample size, accuracy of location, frequency of location, etc. In developing telemetry studies four points should be addressed to help ensure that the results are mean- ingful and may be analyzed as intended. 1) Define the experimental unit. Some studies will seek to analyze bird-days of observation, others will be concerned simply with the number of birds. The former contains a degree of ambiguity as one bird followed for 20 d, 20 birds followed for one d or five birds followed for four d will all yield 20 bird-days, although there are important differences in conclusions made from each data set. Defining the experimental unit as one bird avoids ambiguity. 2) Attempt to insure that study animals are ran- domly selected from the population to which you wish to make inferences. Basically, the trapping technique should be unbiased as to age, sex, size, habitat type, etc., within the chosen population. 3) Try to estimate the replication the study will require (see pilot study in Questions). 4) Determine what, if any, type of stratification you will employ (see pilot study in Questions). Inappropriate or insufficient experimental designs can be difficult to detect or remedy, but a number of them can be found in the literature. Mortality studies are often characterized by inadequate sample size and questionable experimental units. Custom- arily, survival on any given day is assumed to be independent of survival on any other day, although this assumption is not tested. Similarly, home-range studies also suffer from inadequate sample sizes, both of animals and of animal-locations. If a study is designed to produce inferences for all age/sex classes, it must have adequate representation of each of those classes. Home-range estimates are depen- dent on the time frame of the study and on sampling intensity. If either time frame or sampling intensity is increased, estimated home-range size will also in- crease. In general researchers seem to have an in- sufficient understanding of the concept of home-range and of how the picture of home-range changes de- pending on the study framework. Hopefully, con- tinued work on this problem will lead to a better definition of home-range. As a brief experiment in study design tradeoffs, let us look first at a mortality study designed to quantify overwintering survival. In order to cover all age/sex classes you probably need 50-100 ani- mals even to consider beginning the study. The good news is that you probably only need to locate the animals once each day, or with long-lived species perhaps once each week. Even with the decreased observation intensity, such a study may be imprac- tical for many species. Now let us look at an activity study designed to quantify activity patterns between and within days Initially we have several options, three of which are listed below. Design 1: one bird followed for 40 d with 16 lo- cations/d. Design 2: 40 birds followed for one d each with 1 6 locations/d. Design 3: 10 birds followed for eight d each with 16 locations/d. The first two designs are extreme and inappropriate. Design 1 looks only at one bird, so there will be no way to estimate the variance of activity patterns among birds. If you study an abnormal bird, you could generate an array of misleading information, and if you study a normal bird you will still have no way of estimating the range of normal behavior. 129 130 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 Design 2 looks at each bird for only one day, so there will be no way to estimate the variance of activity patterns between days. You can estimate the variance between birds and between days (assuming you did not follow all 40 on the same day) together, but you cannot estimate the effect of only bird-to-bird dif- ferences or of only day-to-day differences. Design 3 is one possible compromise: an intermediate number of birds, days and locations which would permit estimates of all variances of interest. However, op- timal balance among birds, days and locations is complex and can only be determined using a good pilot study. Costs will invariably affect study design, as will time needed to switch between animals and many other nonstatistical concerns. So even for a fairly straightforward question, such as the activity pattern experiment above, determining the best design can be difficult. As the previous examples illustrate, some study designs will be incompatible with each other. We could not run a mortality study and an activity study simultaneously without going to very great expense: one requires many animals with few observations/ animal/time, the other requires few animals but more observations/animal/time. Two additional problems occur with activity studies. Missing values can require increased complexity in the analysis, so it is best to develop a regular sampling framework which is always achievable. Secondly, the interde- pendence of locations which are close in time is a statistical problem only now being addressed. Many home-range analyses assume that all animal loca- tions are independent of each other, although many study designs produce dependent locations. Pantula and Pollock (1985) presented a time-series approach to this problem. One plea here from the statisticians: please do not overvalue lots of data on few animals. Many times a good design will be too costly to achieve and the biologist will continue anyway in the hopes of gain- ing at least some useful information. While this is certainly not a waste of time, writers should ac- knowledge limitations of their results and avoid mak- ing far-reaching statements from scanty data. To date statisticians have been only occasionally in- volved in telemetry analyses. Telemetry lends itself to tailored analysis techniques due to its specific problems and approaches. More statisticians should become involved so that analyses can become effec- tive and available. Questions Re continuous monitoring of animals vs. dis- tinct locations at known time intervals: in general too much data is generated from too few animals. Often no additional information is gained by much additional observation. However, in specific cases it may certainly be appropriate to monitor continu- ously, as when the exact duration of a given activity/ movement is of interest. Re the effect on mortality estimates of bird/ day units vs. bird units: estimates of mortality made using bird/day units will be unbiased, but the es- timate of the variance will be too small. Re the gathering of lots of information/bird: if the goal is to describe the activity of one or a few animals without making inference to a population, then small numbers of animals are not problemat- ical. But if your goal is statistical inference from a sample to a population, then fewer data on more animals is better. Of course lots of data on many animals is best of all but seldom practical. Re pilot studies: the problem of adequate sample size is best addressed by a good pilot study which can provide an estimate of variability of variables of interest. Pilot studies also permit estimates of cost, time, personnel needs, etc., and can save time and money in the final study. References Bart, J. and D. S. Robson. 1982. Estimating survi- vorship when the subjects are visited periodically. Ecol- ogy 63:1078-1090. Dunn, J. E. and P. S. Gipson. 1977. Analysis of radio- telemetry data in studies of home range. Biometrics 33’ 85-101. Hurlburt, S. H. 1984. Pseudoreplication and the de- sign of ecological field experiments. Ecol. Monog. 54- 187-211. Kaplan, E. L. and P. Meier. 1958. Nonparametric estimation from incomplete observations. /. Am. Stat Assoc. 53:457-481. Pantula, S. G. and K. H. Pollock. 1985. Nested analysis of variance with autocorrelated errors. Bio- metrics 41:909-920. Pollock, K. H., S. R. Winterstein and M. J. Conroy. Winter 1987 Telemetry Project Design 131 In press. Estimation and analysis of survival distri- butions for radio-tagged animals. Biometrics. Trent, T. T. and O. J. Rongst ad. 1974. Home range and survival of Cottontail Rabbits in southwestern Wisconsin. /. Wild!. Manage. 38:459-472. White, G. C. 1983. Numerical estimation of survival rates from band-recovery and biotelemetry data. /. Wildl Manage. 47:716-728. Department of Statistics, North Carolina State Uni- versity, P.O. Box 8203, Raleigh, NC 27695-8203. J Raptor Res. 21(4):132-134 © 1987 The Raptor Research Foundation, Inc. ANALYSIS OF SURVIVAL DATA FROM TELEMETRY PROJECTS Christine M. Bunck Telemetry is an increasingly popular method for studying animal movements and habitat use. Ad- ditionally, telemetry provides a means for studying survival and causes of mortality. In this paper I will be describing some statistical techniques which can provide valid estimates of survival rates based on data from telemetry studies. Two basic study schemes are used to observe sur- vival time. In the first, observations on all animals begin at the same time. In some instances time origin will correspond to some biologically meaningful date such as average fledging date, but often time origin is simply the beginning of the study. In practice it is often impossible to mark all animals in one day, but the period of capturing and marking should be as short as possible. In the second study scheme animals enter the study periodically. For wildlife studies this scheme will probably be more common than the first. The techniques I will describe can be applied to lifetime data as recorded under the first scheme. Some, but not all, of these techniques generalize to the second scheme. The following five assumptions apply: Assumption 1) The sample is representative of the population to be studied; requires that trap- ping techniques result in random captures from the population without age bias, sex bias, etc. Assumption 2) Survival is not influenced by ra- dio-marking; if not, study will give a biased es- timate of population survival rate. Assumption 3) The fate of each animal studied is independent of the fate of any other animal studied; would not be the case for nestlings. If a predator finds the nest, all or most of the nestlings will probably die. Similarly, the fate of a young animal is closely linked to its mother’s fate in many instances. Assumption 4) Censoring [censoring occurs when an animal’s fate becomes unknown (e.g., when its transmitter fails)] is independent of fate; a cen- sored animal is just as likely to be alive as dead. Assumption 5) Exact time of death is known. Simulation studies have shown that this assump- tion can be relaxed (Heisey and Fuller 1985). Assumptions 1-3 are also required for band recovery models. Assumptions 4 and 5 are unique to tech- niques for the analysis of survival data. I have classified techniques as discrete or contin- uous models. All equations have been eliminated from this summary, but can be found in the literature cited. Discrete models are those in which survival is described as an outcome observed after some unit of time, such as a day or a week. One widely used discrete model uses an approach originally proposed by Mayfield (1961) for the study of nest success. Mayfield’s model is distinguished from other discrete models by two assumptions: Assumption 1) The probability of surviving a period is the same throughout the study (e.g., chance of surviving in any day/week is the same as in any other day/week). Assumption 2) Each time unit (trial) is indepen- dent of the next trial. Estimation of survival rate over a period of days and testing procedures have been described in papers by Johnson (1979), Hensler and Nichols (1981) and Bart and Robson (1982) concerning study of nesting success but results can be applied to data from te- lemetry studies with application to both single-origin and staggered-entry study schemes. I’d like to mention three other papers that employ discrete models for data analysis from telemetry studies. Trent and Rongstad (1974) were among the first to use a Mayfield-type approach to obtain sur- vival estimates from telemetry data. White (1983) proposed a multinomial model to estimate survival rates from telemetry data and obtained estimates and tested survival rates. Heisey and Fuller (1985) re- fined the Mayfield approach to permit calculation of survival rates which are not constant over long periods (for many species, survival rates vary be- tween seasons, etc.). Intervals were set in which sur- vival rates were nearly constant, Mayfield estimates were computed for each interval and the product of these estimates used for the entire period. Programs by White and by Heisey and Fuller are available from them. Continuous models treat survival time as a con- 132 Winter 1987 Telemetry Project Design 133 Table 2. Nonparametric tests 3 for comparison of survival times. Test and Conditions No censoring (fate of all animals is known) Mann- Whitney U Wilcoxon Rank Sum Kruskal- Wallis Savage Logrank Censoring (some fates unknown) Gehan Peto-Peto Peto-Pentice Mantel-Haenszel Logrank a For further information on individual tests, see Lee, E. T., Sta- tistical methods for survival data analysis. Lifetime Learning Publ., Belmont, CA, 1980. tinuous measure using two major approaches. A parametric approach requires that distribution of survival time values be completely specified. A non- parametric approach does not make assumptions about form of survival time distribution. For parametric approaches one of three functions must be precisely defined. Table 3. Contacts for computer software programs. Program Contact BMDP 3 BMDP Statistical Software 1964 Westwood Blvd. Suite 202 Los Angeles, CA 90025 GLIM b Numerical Algorithms Group 7 Banbury Road Oxford OX2 6NN, Britain SAS SAS Institute, Inc. Box 8000 Cary, NC 27511 SURVREG C Dr. Douglas B. Clarkson IMSL Inc. 2500 Citywest Blvd. Houston, TX 77042-3020 (713)782-6060 a Biomedical Computer Package. b General Linear Interactive Modelling. c Survival Analysis with Regression. Function 1) The probability density function which describes the expected occurrence of sur- vival time values. Function 2) Survival function which is the prob- ability of surviving longer than given periods of time. Function 3) Hazard function defines chance of dying in the next small interval, given that the bird is alive at the beginning of the interval. Given one of these functions, the others can be de- rived. Exponential distribution is commonly used in sur- vival analysis and assumes that the chance of dying does not change with age or time — essentially the Mayfield approach with a continuous model. The approach is straightforward for studies with no cen- sored animals and a defined time of origin. When censoring occurs, iterative (usually computer-cal- culated) procedures are required to obtain estimates, and staggered entries introduce further complexities into the estimation and computation process. Non- parametric approaches are applied when one is un- willing to specify a model for survival time, but it is still desirable to treat survival time as a continuous variable. Kaplan-Meier (Lee 1980), or product-limit, es- timate provides a method for estimating survival function — probability a bird survives longer than some given time. Censored animals and staggered entry schemes are permissible with this approach. The Kaplan-Meier estimate can be used descrip- tively to evaluate the assumption of independence between censoring and fate of the bird by displaying worst-case/best-case scenarios. Estimates can also be used to describe cause-specific mortality. Table 2 lists additional nonparametric tests based on linear rank statistics. In the literature there are several variations for each. All can be applied to single- origin schemes, but only the logrank test generalizes to the staggered entry scheme. Programs for parametric and nonparametric ap- proaches can be found in statistical packages BMDP (Biomedical Computer Programs) (Dixon 1983) and SAS (supplemental library; SAS Institute 1985) and in SURVREG (Survival Analysis with Regression) (Clarkson and Preston 1983). See Table 3 for con- tacts. Finally, the Cox proportional hazards model (Cox and Oates 1984) provides a semi-nonparametric con- tinuous approach which assesses the relationship be- 134 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 tween survival time and related variables such as age, sex, weight at capture and condition at capture. Cox models can be fit using BMDP, a procedure in the supplemental library of SAS, and GLIM (Gen- eral Linear Interactive Modelling) (Baker and Nel- der 1978). See Table 3 for contacts. For the study of survival with telemetry tech- niques, enough locations should be obtained to avoid censoring and obtained often enough to avoid mis- classifying a death when causes of death are being studied. For some discrete models, locations should be obtained at equal intervals. References Discrete Models: Bart, J. and D. S. Robson. 1982. Estimating survi- vorship when the subjects are visited periodically. Ecol- ogy 63:1078-1089. Heisey, D. M. and T. K. Fuller. 1985. Evaluation of survival and cause-specific mortality rates using telem- etry data. J. Wildl. Manage. 49:668-674. Hensler, G. L. 1985. Estimation and comparison of functions of daily nest survival probabilities using the Mayfield method. Pages 289-301. In B. T. T. Morgan and P. M. North, Eds. Statistics in ornithology. Springer- Verlag, New York, NY. and J. D. Nichols. 1981. The Mayfield method of estimating nesting success: a model, estimators and simulation results. Wilson Bull. 93:42-53. Johnson, D. H. 1979. Estimating nest success: the Mayfield method and an alternative. Auk 96:651-661. Mayfield, H. 1961. Nesting success calculated from exposure. Wilson Bull. 73:255-261. White, G. C. 1983. Numerical estimation of survival rates from band recovery and biotelemetry data. J. Wildl. Manage. 47:716-728. Continuous Models and Proportional Hazards Model: Baker, R. J. and J. A. Nelder. 1978. The GLIM system. Numerical Algorithm Group, Oxford, En- gland. Clarkson, D. B. and D. Preston. 1983. SURVREG a program for the interactive analysis of survival regression models. Amer. Stat. 37:174. Cox, D. R. and D. Oates. 1984. Analysis of survival data. Chapman and Hall, New York, NY. Dixon, W. J. 1983. BMDP statistical software. Uni- versity of California Press, Berkeley, CA. Holford, T. R. 1980. The analysis of rates and sur- vivorship using log-linear models. Biometrics 36:299- 305. Kalbfleisch, J. D. and R. L. Prentice. 1980. The statistical analysis of failure time data. Wiley, New York, NY. Lee, E. T. 1980. Statistical methods for survival data analysis. Lifetime Learning Publ., Belmont, CA. SAS Institute. 1985. SAS users guide. SAS Institute, Raleigh, NC. Sample Size Considerations: Peto, R., M. C. Pike, P. Armitage, N. E. Breslow, D. R. Cox, S. V. Howard, N. Mantel, K. Mc- Pherson, J. Peto and P. G. Smith. 1976. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. I. Introduction and design. British J. Cancer 34:585-612. Field Studies: Nelson, M. E. and L. D. Mech. 1986. Mortality of White-tailed Deer in northeastern Minnesota. J. Wildl Manage. 50:691-698. Trent, T. T. and O. J. Rongstad. 1974. Home range and survival of Cottontail Rabbits in southwestern Wisconsin. J. Wildl. Manage. 38:459-472. Patuxent Wildlife Research Center, U.S. Fish and Wildlife Service, Laurel, MD 20708. J Raptor Res. 21 (4): 135 — 1 37 © 1987 The Raptor Research Foundation, Inc. BASIC TECHNIQUES FOR ANALYZING MOVEMENT AND HOME-RANGE DATA Vicky J. Meretsky Circular Statistics for Movement, Migration and Dispersal Data. Data collected using radio te- lemetry may comprise daily flight lines of a mi- grating bird, overall dispersal lines of a cohort of juvenile animals from natal locations to their loca- tions at some later time or some other collection of line segments. Important qualities of data are di- rection and distance of movement, which are rep- resented as vectors. Often, one may wish to analyze vector length separately from vector direction. Move- ment distance may be related to age, sex, weight, weather, etc., while direction may be related to guid- ing lines, prevailing winds, latitude, etc. (Heintzel- man 1975). Distance (length) can be analyzed using standard statistical techniques. Direction, however, often cannot. Consider two dispersing individuals, one on a heading of 359°, the other on a heading of 1°. Intuitively, the average is 0° (i.e., due north), but standard math tells us the average is 1 80°. Likewise consider 0°, 1 20° and 240°, which divide a circle into thirds. Their average is 120°, which is nonsensical. Further, if we call those same lines 120°, 240° and 360°, the average becomes 240°, although we are still analyzing the same lines. These anomalies are char- acteristic of attempts to analyze data for which no true zero point exists. Circular statistics are designed to permit analysis of angular data. One can calculate an average angle of dispersal separately from average distance, or av- erage angle of a day’s migration path separately from average daily travel. Analogs to familiar £-Test, Chi- square Test and many other parametric and non- parametric tests exist which permit one to test for differences in direction of two populations (e.g., dif- ferences in dispersal direction of two cohorts or dif- ferences between migration paths and available guiding lines). Circular statistics can also be used to determine whether path directions are randomly dis- tributed or show a significant tendency toward a given direction. The reference section lists some ex- cellent texts dealing with circular statistics. Techniques for Analyzing Home-range Data. Home range and core area sizes are frequently cited by resource managers as a quantitative basis for resource protection. Several methods can be used to calculate home ranges or land-use patterns, and no single figure can accurately reflect an animal’s use of its surroundings. I will review some of the com- moner analyses, then focus on two recent computer- generated models of home-range/use patterns. I will not address the problems associated with collecting accurate location data with telemetry (see Grainger Hunt’s summary for references). Home-range data usually consist of a series of points collected over some period, hopefully at reg- ular intervals. The simplest method of representing observations is to plot them on a map, which may be sufficient for many uses, but often you will need or want to estimate area used by study subjects or to indicate high-use or core areas. We will assume for the following discussion that in addition to the map, you have devised some sort of a grid coordinate system so that each observation point can be indi- cated by a pair of X,Y values. For many years the standard method of estimating size of home range was the minimum convex poly- gon, constructed by connecting the outermost obser- vation points. The technique tends to be overly gen- erous including many unused areas and is sample size dependent but requires no computers. A quick approach to finding core areas is simply to locate the area of the map which contains the highest density of observations and perhaps a nest, den or roost tree as well. However, the urge to gen- erate numbers may drive you to attempt an “aver- age” location by averaging the X and Y values of your observations, often producing a location that the animal never uses. For example “average” lo- cation of a creature which hunts around the edge of a pond is very apt to be in the pond. While this example is extreme, it points to a problem with straight averages — the animal has no knowledge of your X,Y grid and no reason to order its movements around the average. Despite this fact, one widely- used technique of home-range analysis relies on the arithmetic mean location to construct a series of probability ellipses or circles. A 95% probability el- lipse would be assumed to enclose 95% of the range Just as animals are generally ignorant of arithmetic means, they are not generally disposed to move in 135 136 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 Table 4. Home-range computer programs. Fea- tures Details Source Dr. Edward O. Garton Fish and Wildlife Dept. Univ. of Idaho Moscow, Idaho 83843 Ms. Laura Beery Conservation and Research Center Nat’l Zoological Park Smithsonian Institution Front Royal, VA 22630 Ms. Blair Jones Dept, of Fisheries and Wildlife Science Virginia Polytechnic Institute Blacksburg, VA 24061 Use Home Range: mainframe FORTRAN, IV or 77, Calcomp plotter McPAAL: TURBO PASCAL for IBM-PC, -XT, compatibles, Epson or IBM graphics printers, in future for Summagraphics and H P plot- ters TELEM: for IBM 370, IBM- PC, -AT. Tektronix and H P plotters Graphics Polygons, ellipses, weighted ellipses, har- monic mean, Fourier transforms, tests for distribution types, tests for independence of ob- servations Convex, concave polygons, ellipses, harmonic mean Fourier transform Convex, concave polygons, polygms, ellipses Cost $250.00 includes updates $15.00 includes updates $15.00 ellipses or circles; thus, we will dispense with further discussions of probability ellipses on the grounds that they are unrealistic and that better techniques exist. To return to the problem of centers and cores there are many different measures of the “center” of a distribution of points (e.g., arithmetic mean, geo- metric mean, mode, median, harmonic mean, etc.) each with different properties. The harmonic mean is the basis for one of the newer home-range tech- niques. A harmonic mean, which can be defined for any point on a map, not just the center, is basically a measure of the average distance from a given point to all observation points. A point that is near the center of a dense cluster of observations will have a small average distance to observations, while a point on the perimeter will have a larger average distance and hence a larger harmonic mean. The point with the smallest harmonic mean is the point that is as close as possible to as many obser- vations as possible and is called the harmonic mean center. Note the harmonic mean center must always occur in an area of high use, never in an area the animal never uses, since the harmonic mean center is defined by animal locations. Since you can measure a harmonic mean anywhere on a map, you can cal- culate a harmonic mean for all intersections of an X,Y grid (or every quadrant of each square, or what- ever) and use the values to create isolines (contours) (Dixon and Chapman 1980). Points that are the same average distance from observations will have the same harmonic mean; peaks will occur at high-use areas and valleys in low-use areas. By plotting isolines over observation points, one can calculate area of the isoline that encloses 50% of locations, which can be considered an estimate of 50% use-area. Similarly a 95% or 99% isoline can be considered an estimate of home-range. One assumption about mathematical properties of data has been made: each observation is independent of all others. The independence assumption is a ma- jor problem that we are ignoring for lack of time (but see the summary of Ken Pollock’s discussion and references hereafter). The animal defines the shape of the map and we have made no assumptions about shape or distribution of home-range. Anderson (1982) takes a different approach to the home-range problem. Think of the X,Y grid over- laying the map as a checkerboard. Every time an observation occurs in a square, put a checker in that square. Soon, we have a stacked-up checkerboard with big stacks in core areas and no checkers where the animal has not been seen. We have created a three-dimensional histogram, with frequency of oc- currence graphed as the third dimension. A measure of use can be obtained by looking at volume of check- ers over land area of interest. We could leave it here, at the checker stage, but the results are rough, and do not permit inferences about presence of travel Winter 1987 Telemetry Project Design 137 ridors (which are notoriously difficult for obtaining data) between high-use areas, etc. Anderson’s (1982) technique uses the method of Fourier transforms to smooth out the peaks and valleys a bit, put saddles between close peaks and valleys between distant peaks. By looking at volume under the landscape, one can define use areas by the isoline that encloses a given percentage of the volume — perhaps 10% use area for cores or 90% use area for an estimate of total home-range size. My summaries of home-range techniques are meant to be brief introductions and mention none of the shortcomings or underlying mathematics. I feel the techniques are promising and worthy of atten- tion, but not perfect; researchers should read the supporting papers thoroughly before employing pro- grams. Both Dixon and Chapman’s (1980) har- monic mean technique and Anderson’s (1982) Fou- rier technique are published with references for programs to generate area figures. However, several individuals and institutions have computer packages that execute these and other techniques of home- range analysis. Table 4 lists some available packages, addresses and abilities. References Circular Statistics: Batschelet, E. 1981. Circular statistics in biology. Ac- ademic Press, New York, NY. 371 pp. Mardia, K. V. 1972. Statistics of directional data. Ac- ademic Press, New York, NY. 357 pp. Home-range Analysis: Anderson, D. J. 1982. The home range: a new non- parametric estimation technique. Ecology 63:103-112. Bekoff, M. and L, D. Mech. 1984. Simulation analysis of space use: home range estimates, variability, and sample size. Behavioral Research Methods, Instruments , and Computers 16:32-37. Dixon, K. R. and J. A. Chapman. 1980. Harmonic mean measure of animal activity areas. Ecology 61: 1040-1044. Don, B. A. C. and K. Rennolls. 1983. A home range model incorporating biological attraction points, f. Anim. Ecol. 52:69-81. Dunn, J. E. and P. S. Gipson. 1977. Analysis of ra- diotelemetry data in studies of home range. Biometrics 33:85-101. MacDonald, D. W., F. G. Ball and N. G. Hough. 1980. The evaluation of home range size and config- uration using radio tracking data. Pages 405-424. In C. J. Amlaner, Jr. and D. W. MacDonald, Eds. A handbook on biotelemetry and radiotracking. Perga- mon Press, Oxford, England. Samuel, M. D. and E. O. Garton. 1987. Incorporating activity time in harmonic home range analysis. J. Wildl Manage. 51:254-257. Independence of Observations: Pantula, S. G. and K. H. Pollock. 1985. Nested analysis of variance with autocorrelated errors. Bio- metrics. 41:909-920. Swihart, R. K. and N. A. Slade. 1985. Influence of sampling interval on estimates of home-range size. /. Wildl. Manage. 49:1019-1025. and . 1985. Testing for independence of observations in animal movements. Ecology 66:1176- 1184. Field Studies of Movement and Home Range: Behrends, P., M. Daly and M. I. Wilson. 1986. Range use patterns and spatial relationships of Merriam’s Kangaroo Rats ( Dipodomys merriami). Behaviour 96. 187-209. GaUTHreaux, S. A. 1980. Animal migration, orienta- tion, and navigation. Academic Press, NY. HEINTZELMAN, D. S. 1975. Autumn hawk flights: the migrations in eastern North America. Rutgers Uni- versity Press, New Brunswick, NJ. Jones, E. N. and L. J. Sherman. 1983. A comparison of Meadow Vole home ranges derived from grid trap- ping and radio telemetry. /. Wildl. Manage. 47:558- 561. Schmidt-Koenig, K. and W. T. Keeton. 1978. Ani- mal migration, navigation and homing. Springer-Ver- lag, NY. Schroder, G. D. 1979. Foraging behavior and home range utilization of the Bannertail Kangaroo Rat ( Di- podomys spectabilis). Ecology 64:657-665. Teyssedre, A. 1986. Radio-tracking of pigeons previ- ously exposed to random oscillating magnetic fields. Behaviour 96:265-276. Condor Research Center, 2291 A Portola Road, Ven- tura, CA 93003. / Raptor Res. 2 1(4): 138 © 1987 The Raptor Research Foundation, Inc. DETECTING AND DESCRIBING THE STRUCTURE OF AN ANIMAL’S HOME RANGE Paul H. Geissler and Mark R. Fuller This presentation is an oral version of Geissler and Fuller (1985). We suggest use of casement displays (Chambers et al. 1983) to represent animal use of home range over daily and seasonal time scales. In addition we suggest a clustering technique for use in detecting patterns in structure of home range. Casement display is a graphical technique that permits more than two dimensions to be displayed in two dimensions; we use five dimensions (east, north, time of day, season of year and frequency). While casement displays have not been widely used in home-range studies, they permit faster and clearer un- derstanding of temporal and spatial patterns of home-range use than simpler mapping techniques currently in use. Casement displays require some familiarity before use becomes easy; some helpful references are listed at the end. Clustering techniques provide a quantitative basis for detecting patterns within a home range. Previous techniques have allowed quantitative assessment only of home range spatial patterns. The combination of clustering techniques with casement displays permits quantitative investigation of temporal as well as spatial patterns. References Chambers, J. M., W. S. Cleveland, B. Kleiner and P. A. Jukey. 1983. Graphical methods for data analysis. Duxbury Press, Boston. 395 pp. Geissler, P. H. and M. R. Fuller. 1985. Detecting and displaying the structure of an animal’s home range. Proc Am. Stat. Assoc., Stat. Computing Sect.: 378-383. U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center, Laurel, MD 20708 138 /. Raptor Res. 21 (4):1 39-141 © 1987 The Raptor Research Foundation, Inc. TELEMETRY IN STUDIES OF PREDATION, DISPERSAL AND DEMOGRAPHY Robert Kenward General structure of a radio-telemetry study can be divided into four stages: Stage 1 — acquiring equipment Stage 2 — capturing and marking animals Stage 3 — developing field techniques Stage 4 — analysis Each stage should be considered before starting a project, as each affects the others. Many projects fail from lack of adequate field techniques. Reliable equipment and transmitter attachment are essential (see Mark Fuller’s summary in these proceedings). Care taken in selecting equipment and in hiring personnel will yield better data. Personnel comfort, especially at night or under extreme conditions, im- proves resulting data and lengthens lifespan of field workers. Be sure you can capture enough of your animals and that you can follow them adequately. A good road net or dependable aircraft to decrease travel time is necessary when following many ani- mals. A large travel budget will also be needed. A pilot study is very helpful, especially to estimate costs: it is painfully easy to underestimate travel costs, forcing curtailment or premature termination of the study when funds run out. The role of experience in telemetry studies cannot be overestimated. It is obvious that new personnel must be given time to learn to track animals, but it is less obvious that tracking large numbers of animals for dispersal and survival studies requires different skills from those used in studies of individual be- havior or predation. Telemetry studies tend to evolve as they continue, and there must be time for this evolution. Data collection is seldom satisfactory until the second or third study season. Flexibility of study design is advisable as you gain experience in field techniques and refine your study questions. Analyze your data as you progress, and discuss results, so that changes in data gathering can be made as new ideas and questions arise. The final ingredient, luck, is unfortunately stochastic and difficult to obtain on short notice. But it does help. Predation Studies. Predator-prey interactions can be studied by marking either the predator or the prey. If behavior and predation of one predator is of interest, or if many fresh kills must be examined (e.g., for selection effects), marking the predator is indicated. If a guild of predators is of interest then it may be best to mark prey. Recording each kill adequately will be more difficult, and kills may be attributed to a scavenger rather than the real pred- ator. In studying Goshawk (Accipiter gentilis) predation on Wood Pigeons (Columba palumbus) it was easiest to mark the Goshawks. Following individual hawks gave data on time spent in various habitats, detailed movement along search paths, types of prey taken and kill rate. Analyzing fresh kills for selection ef- fects showed that Goshawks tended to take pigeons with below average weight, except when prey were taken completely by surprise (Kenward 1976). After discovering that Goshawks remained for long periods at or near large kills, and learning radio signal cues which indicated a kill, several hawks could be monitored concurrently when studying pre- dation on pheasants ( Phasianus colchicus ) in Sweden; most kills of 250 g or larger could be recorded by checking each hawk at one hr intervals. Fewer be- havioral data were obtained than when following individual hawks but larger samples of kills and of predation rates were obtained from different hawks. Radio transmitters were also used to estimate hawk density, since the number of transmittered hawks in the area was known, and each hawk seen could be checked for the presence of a radio much more easily than for a visual marker. Combining average kill rates with hawk numbers gave an estimate of pre- dation impact on censused Pheasant populations in several different areas. Diet differences between hawk sexes, and differential prey selection were also stud- ied, most recently with the aid of a radio tag that indicates when a hawk is feeding (Kenward et al. 1981a, b.). Sociality and Range Use. When recording ranges to estimate habitat use or sociality, minimizing num- ber of fixes required to estimate range size or struc- ture saves much work. If range size is plotted after each consecutive fix for a number of individuals, the area will first show a rapid observation-based in- crease as the animal is recorded throughout its nor- 139 140 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 mal range. A plot then reaches “sampling satura- tion,” after which recorded range size increases rep- resent continuing small increases in area covered by an individual animal. For grid-cell-based anal- yses it often takes 200-300 fixes to reach sampling saturation, but studies of three species [Goshawk, Badger ( Meles meles) and Grey Squirrel ( Sciurus carolinensis )] have shown saturation for outline ranges (convex polygons) in 30-40 fixes. Convex polygon areas can be dramatically in- creased by occasional excursions outside normal core areas. Nonparameteric clustering or isoline tech- niques give a better fit to fixes and make fewer assumptions about spatial distribution than earlier bivariate parametric techniques for estimating prob- abilistic core areas (e.g., probability circles and el- lipses) (see Vicky Meretsky’s summary in these pro- ceedings). Core areas can be found by inspecting for a drop in variance coefficient as size decreases along a multi-range utilization distribution; the core in- cludes percentage of fixes which give greater simi- larity between individual ranges than at larger range sizes. A sampling regime of three daytime locations plus a roost gives saturation in 10 d for Goshawk max- imum convex polygon winter ranges. Contrary to popular opinion, hawks showed little territoriality in winter and gathered with their core areas over- lapping at Pheasant farms and other sites with local prey abundance. Dispersal and Demography. For dispersal and mortality studies, it is essential that transmitters are reliable and do not adversely affect their carriers. After five years of predation studies, Goshawk trans- mitters had evolved to 1500 mA Li/CuO z cells pow- ering single-stage transmitters for 9-12 mo. Tail- mounted tags gave 3-5 km working ranges across flat ground and 10-20 km from high vantage points or aircraft. Recapture rates and weights had been similar between radio-tagged and banded birds (Kenward 1978). Unfortunately, tags could not be tail-mounted on nestlings. Nestlings were equipped with leg-tags and caught 10-20 d post-fledging for tail-tagging, all of which involved development of satisfactory leg tags (which were very prone to an- tenna breakage) and capture techniques near the nest (Kenward 1985). My study was done on the 30 000 km 2 Baltic island of Gotland, which has little emigration or immigra- tion of hawks. Location and survival of 30 juvenile hawks and 20-30 adults tagged each year were checked mainly at night, because live hawks were then in trees and thus unlikely to be overlooked due to poor signals while feeding on the ground. When several hawks had been lost, the island was searched from the air. Dispersal date was linked to hawk sex and local food abundance. Hawks tended to remain the least time around nests where and when food was scarce, especially the males, but remained longer at such nests than elsewhere when artificially fed. Dispers- ing juveniles quite frequently parasitized other fledged broods, especially in prey-rich areas and in years when prey elsewhere was scarce. Mortality was surprisingly low the first autumn, and much less than suggested by banding studies, with a peak of juvenile and adult mortality at the end of winter. Movements have been linked to local abundance of rabbits, an important prey on the island (Kenward, Marestrom and Karlbom, in prep.). References Work on dispersal and demography is not yet published Kenward, R. E. 1976. The effect of predation by Gos- hawks, Accipiter gentihs, on Woodpigeon, Columba pa- lumbus , populations. D. Phil, thesis, University of Ox- ford. . 1978. Radio transmitters tail-mounted on hawks. Ornis Scand. 9:220-223. . 1982. Goshawk hunting behavior, and range size as a function of habitat availability. J. Anim. Ecol 51:69-80. . 1985. Raptor radio tracking and radio telem- etry. Int. Council Bird Pres. Tech. Bull. 5:409-420. . 1987. Wildlife radio tagging: equipment, field techniques and data analysis. Academic Press, New York. , G. J. M. Hirons and F. Ziesemer. 1981a. Devices for telemetering the behavior of free-living birds. Symposia of the Zoological Society of London 49:129-137. , V. Marcstrom and M. Karlbom. 1981b. Goshawk winter ecology in Swedish pheasant habitats. J. Wildl. Manage. 45:397-408. Other Studies: Boutin, S. and G. J. Krebs. 1986. Estimating survival rates of Snowshoe Hares. /. Wildl. Manage. 50:592- 594. Boutin, S., C. J. Krebs, A. R. E. Sinclair and J. N. M. Smith. 1986. Proximate causes of losses in a Snow- shoe Hare population. Can. J. Zool. 64:606-610. Cumming, H. G. and D. B. Beange. 1987. Dispersion Winter 1987 Telemetry Project Design 141 and movements of Woodland Caribou near Lake Nip- igon, Ontario. /. Wildl. Manage. 51:69-79. Gillingham, P. and F. L. Bunnell. 1985. Reliability of motion-sensitive radio collars for estimating activity of Black-tailed Deer. /. Wildl. Manage. 49:951-958. Wooley, C. M. and E. J. Crateau. 1985. Movement, microhabitat, exploitation and management of Gulf of Mexico Sturgeon, Apalachicola River, Florida. N. Am J. Fish Mgt. 5:590-605. Furzebrook Research Station, Institute of Terrestrial Ecology, Wareham, Dorset, BH20 5AS, England. J Raptor Res. 21(4):142-143 © 1987 The Raptor Research Foundation, Inc. TELEMETRY TECHNIQUES FOR THE STUDY OF RAPTOR MIGRATION William W. Cochran Radio telemetry makes it possible to observe mi- grant raptors in some detail over distance and time intervals. Such observations yield inferences stronger than those generated from fleeting glimpses of pass- ing migrants or lines connecting banding and recov- ery sites. Most studies that use telemetry are conducted in areas much smaller than those over which birds range during migration. Methods for such area-related studies are applicable to studies of migrants wher- ever migrants linger [see Holthuijzen and Oosterhuis (1985), or Hunt’s summary herein]. Kenward’s comments concerning planning, equipment, person- nel, luck, etc., are applicable to serious studies that employ telemetry, including the study of raptor mi- gration. I will restrict my discussion to the use of telemetry in the study of long-distance migrants. One approach for studying migration is to observe one or a flock of individuals intensively. An airplane facilitates tracking because of its speed and freedom of movement (Gilmer et al. 1981; Mech 1983). Also, at higher altitudes reception is increased to 150 km or more for more powerful transmitters. Unfortu- nately, air tracking provides little opportunity for visual observation without risk of disturbing the sub- ject. Furthermore, small-scale movements such as short hunting flights and signal variations that sig- nify eating, hunting flight, roosting, climbing, de- scending, etc., are difficult or impossible to document from an airplane. If a rough plot of a bird’s migra- tory route is all that is desired, a plane is a good choice; if care is taken to be in the air at the right time, anyone familiar with ground tracking can suc- ceed immediately. Where road networks permit and collection of behavioral data is important, by far the best tracking conveyance is a suitably equipped automobile. Fre- quent, and often hours-long, periods of visual ob- servation of a migrant raptor as it perches, hunts, feeds, roosts or migrates provide welcome breaks in the monotony of electronic observation. By listening to the signal while watching the subject it is possible to learn to associate signal variations with particular activities. The easiest to associate are the steady sig- nals from a perched or gliding raptor and cyclic signal fade of a raptor circling in an isolated thermal. Automobile tracking requires constant route plan- ning and replanning. River crossings and large met- ropolitan areas can be particularly frustrating. Planned temporary loss of contact is often required in detouring away from the migrant, for instance to cross a river at a bridge. Unplanned loss of contact occurs frequently when the bird comes down. Trans- mitter range for a soaring raptor is typically 50 km, but in a tree or on the ground range will drop to about 10 km and one km, respectively. Fortunately, search area for a nonflying raptor does not expand with time, and by keeping a running log of azimuth an observer always knows which way to go to close distance. However, routing mistakes, especially when winds aid a migrant, will result in trackers being left hopelessly behind and will require luck or the use of a (rental) airplane to reestablish contact. An automobile and airplane form an ideal com- bination for tracking, complementing each other in their strengths and weaknesses. Cost of using both is considerably less than the summed cost of either alone (e.g., air time is not required to monitor a stopped or slowly moving migrant). Another advan- tage concerns what Kenward calls “lifespan of field workers”; by alternating duties between plane and automobile, observers can extend this lifespan that otherwise (in my experience of living in a vehicle) is about two wks. A great variety of data may be collected while observing migrants. At the least, one is interested in food sources, daily rate and direction of travel, how the day is budgeted between hunting, perching and migrating, and how all these are affected by weather, habitat and topography or vary with age, sex and geographic region. Unfortunately, the migratory seasons provide time for observations of only a few individuals at most and many seasons may be nec- essary to acquire a data base that addresses such interests in a statistically sound way. Small samples can, however, be revealing. For instance, a popular field guide (Robbins et al. 1983) comments that the Peregrine ( Falco peregrinus) “rarely soars.” The first Peregrine I tracked (for two full days in fall 1973) flew a total of 11.9 hr, 82% 142 Winter 1987 Telemetry Project Design 143 of which was interthermal soaring (Cochran 1975). “Rarely” and 82% are quite incompatible yet it is likely that “rarely” was based on a large sample of observations. So, was this bird a freak? A later in- crease in sample size to 143 hr and 6 birds changed the outcome very little (84% soaring) (Cochran, un- pub. data). One is sometimes left wondering which is better, a few high quality observations or a large sample of biased observations? The slow rate of data acquisition inherent in study- ing individuals may be avoided in part by focusing on particular questions. For instance, having learned from the study of a few inland migrant Peregrine Falcons that individual migratory flight direction does not vary much from day to day, it is now rea- sonable to determine directional distribution of in- land migrants by one- or partial-day tracks from a trapping area and to make inferences about sources and destinations. Several one-day tracks from a trap- ping site can be documented from one airplane in a single day; therefore, a large sample can be obtained in one season. Such a study would have little mean- ing at Assateague Island, MD, where individual Peregrines migrate south for several hundred km along the Atlantic coast before diverging on individ- ual courses. Thus, the value of preliminary studies of individuals is to establish a suitable context within which question-oriented studies involving many birds can be pursued with confidence. Radio tags may be used as super bird bands for the purpose of obtaining occasional or specific lo- cations. For example many tags could be attached at various trapping sites during migration and later located during one or two air searches of winter or summer range. Careful preparations should be made for such studies; transmitters and attachments must be reliable for the required time. Air searches are most efficient when reception range is maximum. Therefore, and because range is very limited when birds are on the ground, a knowledge of time of day and kinds of weather favoring perched and soaring behavior is of great value in planning air searches. Cost effectiveness of such studies improves remark- ably for large numbers of tags and can be enhanced by planning supplementary studies such as enroute tracking of a few birds, more intensive study of some individuals after relocation or gathering specific data on environs of the whole sample. Fuller (1985) used a satellite to locate a Bald Eagle ( Haliaeetus leucocephalus) over a six month period. The present limitation of satellite use, solely the result of having to use satellites designed for other missions, is that 100 g radio packages must be used, 10-100 times heavier than conventional transmitters for birds. References Cochran, W. W. 1972. A few days of the fall migration of a Sharp-shinned Hawk. Hawk Chalk 11:39-44. . 1972. Long distance tracking of birds. NASA SP 262:39-59. . 1975. Following a migrating Peregrine Falcon from Wisconsin to Mexico. Hawk Chalk 14:25-37. . 1980. Wildlife telemetry. Pages 507-520. In S. D. Schemnitz, Ed. Wildlife management techniques manual. The Wildlife Society, Washington, D.C. 686 pp. . 1985. Ocean migration of Peregrine Falcons is the adult male pelagic? Proc. Hawk Migration Conj. for HMANA\225-2?>1 . and C. G. Kjos. 1985. Wind drift and migration of thrushes: a telemetry study. III. Nat. Hist. Survey Bull No. 33. Fuller, M. 1985. Where eagles fly. Science 85 6:8. Gilmer, D. S., L. M. Cowardin, R. L. Duval, L. M. Mechlin, C. W. Shaiffer and V. B. Kuechle. 1981. Procedures for the use of aircraft in wildlife biotelemetry studies. U.S. Fish and Wildl. Serv. Res. Pub. 140. 19 pp. Holthuijzen, A. M. A. and L. Oosterhuis. 1985. Im- plications for migration counts from telemetry studies of Sharp-shinned Hawks ( Accipiter striatus) at Cape May Point, New Jersey. Proc. Hawk Migration Conf for HMANA.505-M2. Mech, L. D. 1983. Handbook of animal radio-tracking. University of Minnesota Press, Minneapolis, MN. 107 pp. Robbins, C. S., B. Bruun and H. S. Zim. 1983. Birds of North America. Golden Press, NY. 348 pp. Illinois Natural History Survey, 607 E. Peabody Drive, Champaign, IL 68120. J Raptor Res. 21 (4):144-146 © 1987 The Raptor Research Foundation, Inc. RADIO TELEMETRY IN THE STUDY OF RAPTOR HABITAT SELECTION W. Grainger Hunt Studies of raptor habitat selection using only vi- sual observation must encounter the problem of dif- ferential visibility and penetrability among habitat types. The chief virtue of radio telemetry in habitat studies is elimination of bias of detectability differ- ences since a radio-tagged bird is equally detectable in all habitats. Telemetry also permits night obser- vations which are generally impossible with other techniques. In each of three studies I will discuss, basic tech- niques are the same: tag as many birds as feasible in order to approach a good sample size (see com- ments in Pollock’s summary), use a study area large enough to accommodate the movements of all tagged birds and determine position as closely as possible, generally once or twice/d if possible. A powerful transmitter is invaluable in such studies, particularly if study animals are not predictable in their locations. Habitat must be mapped, or a previously completed map must be obtained. To explain observed pref- erences, raptor activity in favored habitat(s) is ob- served visually and aided by telemetry. There are two basic means of acquiring habitat data. Frequent surveys designed to locate all indi- viduals provide information for all birds under sim- ilar conditions. Following individuals for set periods can give more detailed information but introduces variability caused by different conditions during the observation periods. Both methods are useful, and in most cases, study design will suggest one over the other. The standard method used to determine habitat preference is to calculate amount of time or number of occurrences of birds in each habitat type compared to availability of each habitat type. Habitat avail- ability is measured by calculating total area of each type in the study area. Along a river or other linear habitat, total linear distance can be used rather than area. A Chi-square test or rank correlation test can be used to determine if raptors use habitat types in a different proportion to actual areas available (i.e., if raptor presence is distributed nonrandomly among habitat types). Several recent publications concern- ing the methodological and statistical problems of availability/preference data are listed at the end of this summary. Overlays are often used to assign habitat or other values after a location has been made. Location is plotted on the habitat map and habitat type assigned to that observation. In large heterogeneous study areas it is safer to assign a habitat type at time of observation to avoid the possibility of mismapping or slight inaccuracies in assessing location of the bird. Remember that edge is often a meaningful habitat type. When testing for nonrandom use of habitat one must be aware of possible seasonal shifts. Lump- ing data from different seasons or years may obscure seasonal or yearly preferences. References address- ing the problem of accurate location of signals, and effects of habitat on triangulation accuracy are listed following this summary. Habitat preferences can be demonstrated fairly readily using techniques outlined above. Often hab- itat preference studies function as pilot studies sug- gesting what further research is appropriate to de- termine cause of preference. Prey availability, prey preference, availability of special features (nest sites, necessary microhabitat features) or other factors may need to be measured. The common result of data collection to explain habitat preference is a welter of data from many variables potentially capable of explaining observed preferences. Multivariate data analyses may not al- ways be necessary. Simple nonparametric correla- tion and Chi-square analyses often reveal major re- lationships more cheaply and quickly. Univariate and bivariate tests should always precede and may render more complicated analyses unnecessary. Sound management guidelines and simple relationships are generally more useful than long equations for many variables of which only one or two are significant. The following three examples demonstrate the study of habitat selection using radio telemetry. Migrating Peregrine Falcons ( Falco peregrinus) were studied at Padre Island, Texas, on two con- secutive winters. Habitat types had been mapped previously and were easily identified from a plane. Twenty-seven female falcons were located 2x/d 144 Winter 1987 Telemetry Project Design 145 (weather permitting). Preferred habitat was differ- ent in each year; thus, the pooled sample showed no selection. Yearly rainfall differences accounted for change in habitat preference; prey availability seemed to be the factor most responsible (Hunt et ah 1980b, 1981). Wintering Bald Eagles ( Haliaeetus leucocephalus) were studied on the Skagit River in Washington to determine the impact of planned dam construction. Food availability seemed to explain most habitat preference. Studies of salmon ( Salmo sp.) availability and eagle feeding habits showed the area was at carrying capacity for Bald Eagles, and dam place- ment would reduce the local population as eagles emigrated into other areas. Possible alternate use areas were determined by following eagles during a period when a flood rendered the previously used area undesirable (Hunt et al. 1980a; Hunt and Johnson 1981). Bald Eagles were also studied on the Pit River in northern California. Seven subadults showed defi- nite seasonal movement patterns between the Pit River and a large lake to the north. To study use of the Pit River itself, 33 individuals (juveniles, sub- adults and nesting adults) were tracked from ground and air. River habitat was mapped in 0.1 km sec- tions. Values for prey availability, public use and other variables were assigned to each section. Eagles showed a definite preference for pools (as opposed to riffles, runs and pocket water). Further studies to determine what microhabitat variables affected eagle use of pools were conducted from blinds located near pools. In another part of the study distribution of Bald Eagles along the river proved to be related to prey biomass and prey-size variability along the river (BioSystems Analysis, Inc. and U. Cal. Davis 1985). References BioSystems Analysis, Inc. and University of Cali- fornia at Davis, Department of Wildlife and Fisheries Biology. 1985. The Pit 3, 4, and 5 Project Bald Eagle and Fish Study, Final Report. Prepared for Pacific Gas and Electric Co., Department of En- gineering Research. Hunt, W. G. and B. S. Johnson. 1981. Impacts of a proposed Copper Creek dam on Bald Eagles: second winter study. BioSystems Analysis, Inc. Environmental Report for Seattle City Light, City of Seattle, Wash- ington. , , J. B. Bulger and C. G. Thelander. 1980a. Impacts of a proposed Copper Creek dam on Bald Eagles. BioSystems Analysis, Inc. Environmental Report for Seattle City Light, City of Seattle, Wash- ington. Hunt, W. G., F. P. Ward and B. S. Johnson. 1981. Peregrine Falcon migration: continuing studies. Report to Department of Defense. Hunt, W. G., F. P. Ward, C. M. Anderson and G. P. Vose. 1980b. A study of the spring passage of Per- egrine Falcons at Padre Island, Texas, using radio telemetry. Report prepared for the U. S. Fish and wildlife Service and the National Park Service. Published results of the studies are in preparation. Sources of Error in Triangulation: Garrott, R. C., G. C. White, R. M. Bartmann and D. L. WEYBRIGHT. 1986. Reflected signal bias in biotelemetry triangulation systems. /. Wildl. Manage 50:747-752. Hupp, J. W. and R. T. Ratti. 1983. A test of radio telemetry: triangulation accuracy in heterogeneous en- vironments. Proc. Int. Wildl. Biotelem. Conf. 4:31-46. Lenth, R. V. 1981. On finding the source of a signal. Technometrics 23:149-154. Springer, J. T. 1979. Some sources of bias and sampling error in radio triangulation. J. Wildl. Manage. 43:926- 935. White, G. C. and R. A. Garrott. 1986. Effects of biotelemetry triangulation error on detecting habitat selection. J. Wildl. Manage. 50:509-513. Preference /Availability: Alldredge, J. R. and J. T. Ratti. 1986. Comparison of some statistical techniques for analysis of resource selection. J. Wildl. Manage. 50:157-165. Chesson, J. 1983. The estimation and analysis of pref- erence and its relationship to foraging models. Ecology 64:1297-1304. Heisey, D. M. 1985. Analyzing selection experiments with log-linear models. Ecology 66:1744-1748. Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluation of resource preference. Ecology 61:65-71. Neu, C. W., C. R. Byers and J. M. Peek. 1974. A technique for analysis of utilization-availability data /. Wildl. Manage. 38:541-545. Other Habitat Studies: Carson, R. G. and J. M. Peek. 1 987. Mule deer habitat selection patterns in northcentral Washington. /. Wildl. Manage. 51:46-51. Litvaitis, J. A., J. A. Sherburne and J. A. Bissonette. 1986. Bobcat habitat use and home range size in re- lation to prey density. /. Wildl. Manage. 50:110-117 146 Vicky J. Meretsky (Technical Editor) Vol. 21, No. 4 Miller, M. L. and B. W. Menzel. 1986. Movements, homing and home range of muskellunge Esox musqui- nongy, in West Okoboji Lake, Iowa. Environ. Biol. Fishes 16:243-255. Ryder, T. J. and L. L. Irwin. 1987. Winter habitat relations of pronghorns in southcentral Wyoming. J. Wildl. Manage. 51:79-85. Trivelpiece, W. Z., J. L. Bengtson, S. G. Trivelpiece and N. J. Volkmann. 1986. Foraging behavior of gentoo and chinstrap penguins as determined by new radiotelemetry techniques. Auk 103:777-781. Walton, R. and B. J. Trowbridge. 1983. The use of radio tracking in studying the foraging behavior of the Indian Flying Fox {Pteropus giganteus). J. Zool. 201: 575-579. BioSystems Analysis, Inc., 303 Potrero Street, Suite 29- 301, Santa Cruz, CA 95060. Concluding page of 1985 Telemetry Workshop Proceedings. Ed. J Raptor Res. 21(4):147-152 © 1987 The Raptor Research Foundation, Inc. ELECTRORETINOGRAPHIC RESPONSES OF THE GREAT HORNED OWL {Bubo virginianus) Steven J. Ault and Edwin W. House Abstract. — Electroretinograms were recorded from four Great Horned Owls {Bubo virginianus). Two procedures, dark adaptation and flicker stimuli, were used to assess the contributions of rod and cone systems to electroretinograms. Recordings obtained after dark adaptation demonstrated scotopic (rod- generated) components. B - waves were broad and rounded and had a fairly long latency. When high intensity single-flash stimuli were used, b - waves had shorter latencies, and prominent a-waves were present, indicative of the addition of photopic (cone-generated) activity. Photopic activity was more clearly demonstrated with flicker ERGs. Scotopic fusion frequency was approximately 16 Hz. Photopic fusion frequencies were in the range of 35-45 Hz. The Great Horned Owl retina functions optimally during low luminance levels at night. However, the presence of a functional photopic system allows this owl to also function in brighter luminances of day. The avian retina has been described by some as the ultimate in retinal organization (Walls 1942; Polyak 1957). Avian retinae, as with those of mam- mals, contain receptors for dim light (rods) and re- ceptors for bright light and colors (cones). In general retinae of diurnal birds are dominated by cones while nocturnal birds possess retinae with a large number of rods and few cones (Walls 1942; Duke-Elder 1958). In owls which as a group are typically noc- turnal, vision has retained the same importance as for their diurnal relatives. Owl retinae have been examined histologically (Bornshein and Tansley 1961; Hocking and Mitchell 1961; Oehme 1961; Fite 1973; Yew et al. 1977; Bowmaker and Martin 1978) and found to have high numbers of rods as would be expected for nocturnal animals. However, Fite (1973) and Oehme (1961) stress that in spite of the predominance of rods owl retinae contain ap- proximately seven to eight percent cones, even in the most nocturnal species. Relative contribution of the cone component to retinal function has not been studied. A preliminary study (Ault 1984) suggested that the Great Horned Owl retina produced ERGs that were qualitatively similar to those of other nocturnal vertebrates and dominated by scotopic (rod-gener- ated) components, but under appropriate stimulus conditions some photopic (cone-generated) compo- nents were also present. However, the sample size of the study (N = 1) was too small to reliably eval- uate ERG responses of the species. Martin (1982) suggested that eyes of owls and humans function similarly over a wide range of naturally occurring luminance levels as a result of their optics and struc- ture. If so, then likely the Great Horned Owl ERG, as in the human, should reveal a duplicity of function possessing both scotopic and photopic components. Functional duplicity is also expected in light of the morphological confirmation of both rods and cones within the Great Horned Owl retina (Oehme 1961; Fite 1973). The objectives of this study were to re- cord and quantitatively analyze ERGs from several Great Horned Owls in order to provide insight into the relative contributions of rods and cones to ERG response. Methods Subjects. A total of eight retinae from four injured and unreleasable Great Horned Owls provided data for this study. Owls were sedated with 10 mg/kg acepromazine maleate (Ayerst Laboratories) administered intramuscu- larly and anesthetized with 80 mg/kg ketamine HC1 (Ke- taject®, Bristol Laboratories) administered intramuscu- larly. Anesthesia generally causes only slight reduction in amplitude of ERG components (Armington 1974). Eyes were examined ophthalmoscopically prior to testing and did not possess any significant ocular lesions. Subjects re- covered fully after approximately 24 hr with no apparent after effects. Apparatus and Procedures. Light source for dark ad- aptation tests was a Kodak Carousel projector with a 300 watt bulb. Single-flash stimuli were achieved by alter- nating opaque filters with empty slots in the carousel. Starting with an opaque filter, the carousel was rotated through an empty slot to the next opaque filter producing an intense flash of light of approximately 200 msec du- ration. Light was channeled through a slide holder and focused onto a 3 mm dia fiber optic light guide which was inserted into a black box containing the anesthetized owl. The light guide was brought to within a few millimeters of the cornea. Care was taken to insure that the light guide was aligned as closely as possible with the optical axis of the eye. Maximum luminance (i.e., intensity or brightness) 147 148 Steven J. Ault and Edwin W. House Vol. 21, No. 4 A TIME (min) C 0.5 1.0 2.0 3.0 5.0 10.0 15.0 20.0 200ji\/| 200 msec A T A T B L0G| 0 RELATIVE INTENSITY I WITH BLUE FILTER AT AT Figure 1. A) Representative Great Horned Owl elec- troretinograms during dark adaptation. Owl was pre-adapted to constant light of 1.9 cd/ cm 2 for five min. Time indicates minutes after pre-adaptation light was shut off and dark adaptation began. Up arrowhead indicates stimulus on; down arrowhead indicates stim- ulus off. Stimulus: 1.9 cd/cm 2 attenuated with a 1 log unit neutral density filter and Wratten #47 blue filter. Note rise in 6- wave amplitude as time progresses. B) Representative Great Horned Owl electroretinograms produced by single-flash stimuli without (left) and with (right) blue filter. Up and down arrowheads indicate stimulus as above. Relative intensity of five indicates maximum luminance of 1.9 cd/cm 2 . Note b - wave increase with increase in intensity. ^4-waves (arrows) appear at high in- tensities. of the light source measured by photometer at the cornea was approximately 1.9 cd/cm 2 and is roughly equivalent to the brightness of a clear sky at noon. Signals from the electrodes were channeled into a Grass 7P1-A preamplifier and tracings were recorded on a Grass Model 7 oscillo- graph. Light source for flicker tests was a Grass Model PS33 photic stimulator interfaced with the fiber optic system described above. Maximum luminance measured at the cornea was approximately 0.45 cd/cm 2 . The cornea was desensitized by topical application of Lidocaine HC1 (Wyeth Laboratories) and a metal-plated mylar electrode was placed on the cornea (Chase et al. 1976). A needle reference electrode was inserted into skin of the ear canal immediately posterior to the eye and a needle ground electrode was inserted into skin of the wing. Pupil size was monitored throughout the recording session and re- mained sufficiently dilated to allow a maximum amount of light to enter the eye. We recorded the responses of the retinae to both dark adaptation and flicker tests. The dark adaptation test was used to assess scotopic recovery of the retinae following exposure to bright light. In the dark adaptation test the retina was pre-adapted to constant light of 1.9 cd/cm 2 for five min. After five min, pre-adaptation light was shut off. Single-flash stimuli attenuated with a 1.0 neutral density filter and a Wratten #47 blue filter were delivered at various intervals to the retina. After owls were fully dark adapted (approximately 45 min), single-flash stimuli of increasing intensities (removal of neutral density filters), both with and without a blue filter, were delivered in succession. The second test utilized flickering stimuli of various intensities and flicker frequencies to determine cutoff point between scotopic and photopic systems. Maximum lu- minance measured at the cornea was approximately 0.45 cd/cm 2 . Various neutral density filters, but no color filters, were used in the procedure. Results During dark adaptation, mean b -wave amplitude increased rapidly from an average of 7.8 at the beginning of dark adaptation to an average of 88.6 fxV at five min into dark adaptation. After the first five min, 6-wave amplitudes increased at a slower rate and eventually reached an average amplitude of 120.3 juV at approximately 20 min. Representa- tive dark adaptation ERGs from one owl are shown in Figure 1 A. 5-waves were broad and rounded with Figure 2. Representative Great Horned Owl flicker electroretinograms at various frequencies. Relative intensity of two indicates maximum luminance of 0.45 cd/cm 2 . Stimulus tracings are shown below each frequency label. Note one-to-one correspondence of ERG response with two Hz stimulus at all intensities (thin arrows). As intensity and/or flicker frequency is increased, one-to-one response is reduced and eventually lost or “fused” (open arrow, for example). Note also one-to-one response at high intensity and high flicker frequency (thick arrow). LOG | 0 RELATIVE INTENSITY -A _ — A A — - A — Aajuuwivl/ulaa ~ , A *A ■A- — tA — - y\/WWWL/v_y\_y\_y^_y^ 5 Hz -^WVAAAAJVUl 2Hz A 200*1 vj 200msec 10 Hz 15 Hz tl 20 Hz 25 Hz TV 30 Hz \ -Sv~v— ^ 35 Hz 150 Steven J. Ault and Edwin W. House Vol. 21, No. 4 LOG 10 RELATIVE INTENSITY Figure 3. Critical fusion frequencies (CFF) plotted as a function of stimulus intensity. CFF was de- termined as the frequency at which there was no longer a one-to-one correspondence with the stimulus. Closed squares indicate the av- erage CFF for each stimulus intensity; X in- dicates individual samples. A significant dif- ference occurs between average CFF obtained at relative intensity of zero and average CFF values for relative intensities of one and two. Average CFF for relative intensities of one and two were not significantly different from each other (Dunnett’s t -Test; P < 0.01 for all be- tween group comparisons). Relative intensity of two represents maximum stimulus lumi- nance of 1.9 cd/cm 2 . a fairly long latency (average latency = 113.2 msec) and a-waves were either very reduced or absent. Figure IB shows representative ERGs produced with single-flash stimuli of increasing intensity, both with and without a blue filter, at the end of dark adaptation. In the case without a blue filter 6-wave amplitudes increased linearly from an average of 77.2 /uV at lowest stimulus intensity to an average of 305.8 jiW at highest stimulus intensity. The same pattern was evident when a blue filter was inserted. 5-wave amplitudes also increased linearly from an average of 63.3 /uV at lowest stimulus intensity to an average of 254.5 jiV at highest stimulus intensity. Average 6 -wave latency decreased with increasing intensity both with and without a blue filter. In the case without a blue filter average latency ranged from 110 msec at lowest stimulus intensity to an average of 71 msec at highest stimulus intensity. In the case where the blue filter was inserted average latency ranged from 155 msec at lowest stimulus intensity to an average of 94 msec at highest stimulus intensity. ^4-waves also became more prominent as stimulus intensity was increased. Figure 2 shows representative flicker ERGs from one owl. At low light intensities and low flicker frequencies ERG waveforms were evident when re- sponding to stimuli on a one-to-one basis. An even- tual loss of the one-to-one response occurred as intensities and flicker frequencies increased. Critical fusion frequencies (CFF) for each stimulus intensity were determined for all owls (Fig. 3). Analysis with Dunnett’s f-Test showed that there was a significant difference between the average CFF obtained at 0.0 log units intensity (16.0 ± 3.7 S.E.) and average CFF values for 1.0 log units intensity (29.2 ± 1.5 S.E.) and 2.0 log units intensity (31.9 ± 2.1 S.E.). However, average CFF values for 1.0 and 2.0 log units intensity were not significantly different ( P < 0.01 for all between group comparisons). Interestingly, the initial flicker ERG waveform changes shape as intensity and frequency are in- creased. Increase in 6 -wave amplitude and decrease in b - wave latency occurs, and a -waves also become more prominent. Discussion Shape of the ERG waveform depends upon rel- ative contributions of scotopic and photopic signals being propagated to inner retinal layers. In duplex retinae such as in the human, for example, the (6- wave is often composed of two components (61 and 62), with different latencies; the short latency 61 component corresponds with photopic activity while the longer latency 62 component corresponds with scotopic activity (Brunette 1969). The 61 component can be isolated with the use of longer wavelength (i.e., red) stimuli while 62 components can be iso- lated with shorter wavelength (i.e., blue) stimuli. In the Great Horned Owl recovery during dark ad- aptation was dominated by scotopic processes man- ifested by slow rising 6 -waves which were fairly broad and had relatively long latencies. Since stim- ulus parameters used in dark adaptation generally elicit scotopic activity primarily, a photopic 6 1 com- ponent was not seen in this study during dark ad- Winter 1987 ERGs of the Great Horned Owl 151 aptation. With different pre-adaptation and stimu- lus parameters a duplex retina ordinarily will demonstrate an early and transient photopic recovery indicated by 6-waves with shorter latencies (domi- nated by the 61 component) followed by scotopic recovery. Higher intensity single-flash stimuli without a blue filter produced ERGs with more prominent a- and 6-waves. 5-waves were also narrower and had a shorter latency, suggesting the presence of a pho- topic component which was contributing to overall response. Higher intensity stimuli were presumably able to initiate a cone response. Change in amplitude and latency of the 6-wave and appearance of the a- wave as stimulus intensity is increased were similar to results obtained from cat (Brown 1968; Niemeyer 1976), rabbit (Ikeda 1966) and horse (Wouters et al. 1980) retinae. At low stimulus intensities the 6- wave is broad and rounded and there is no a-wave, indicating primary activity produced by the scotopic system. At higher stimulus intensities the 6 -wave increases in amplitude and becomes steeper and more pointed. Also, a -waves become more prominent, in- dicating addition of a photopic component that con- tributed to overall response. In animals that have an essentially pure cone retina ERGs show an ex- tremely high amplitude a-wave and very short la- tency 6-wave composed almost exclusively of the 61 component (Tansley et al. 1961). Addition of a blue filter also increased average latency of the 6-wave, suggesting that the photopic contribution was “fil- tered” out and the major contribution to the wave- form was from the scotopic system. Results from the flicker procedure more convinc- ingly demonstrate the existence of a cone component in the retina of these owls. Large differences in CFF from low to high intensities is an indication of a shift from scotopic to photopic functions. A one-to-one response seen at low intensities and low flicker fre- quencies was produced primarily by scotopic activ- ity. Rods were following the individual flicker, hav- ing not yet exceeded their critical fusion frequency. Loss or fusion of the one-to-one response occurred at fairly low flicker frequency. At higher intensities the one-to-one response fused at significantly higher frequencies, an indication of a switch to photopic activity since cones possess a higher critical fusion frequency than rods (Armington 1974; Fishman 1975). Such results are quite common in animals known to have mixed retinae, including humans (Ar- mington 1974). The ERG results confirm the expectation that the Great Horned Owl retina possesses a significant sco- topic component. In addition a photopic component is present but is only evident with proper stimulus parameters. These results suggest that the Great Horned Owl retina is composed primarily of rods but also contains some cones as has been confirmed anatomically by previous light-microscopic obser- vations (Oehme 1961; Fite 1973; Ault 1984). The Great Horned Owl retina is dominated by scotopic processes which certainly impart an in- creased sensitivity to low light levels typically en- countered. However, the owl is often active during the day and the few cones present may help mediate vision in more intense illumination of daylight hours. In a detailed study of optics and visual performance of the Tawny Owl ( Strix aluco) Martin (1982) sug- gested that the resolving power of the eye of this owl and the pigeon ( Columba sp.) were in fact very sim- ilar at photopic and mesopic luminances. The Tawny Owl has superior acuity to the pigeon at lower lu- minance levels, and although photopic acuity of both species are quite similar, acuity declines much faster in lower luminances for the pigeon than the owl (Martin 1982). Briefly stated, the pigeon by virtue of its optics and eye structure cannot function in lower luminance levels, while the owl by virtue of its optics and eye structure can function over a wide range of luminance from scotopic to photopic. Additionally, specific neuroanatomical arrange- ments of photoreceptors contribute to spatial reso- lution and visual acuity in the Great Horned Owl. Foveal rods, which have a lower convergence ratio, give increased ability for point-to-point resolution at higher luminances, while non-foveal rods which have higher convergence ratios and higher absolute sen- sitivity may be serving this function at lower lumi- nances (Fite 1973). These observations, coupled with duplicity in ret- inal functioning revealed in this study, suggest that the Great Horned Owl is not only an effective noc- turnal predator but is able to expand activity into the “diurnal realm” if needed. Acknowledgments We thank the Idaho Department of Fish and Game and the Liberty Wildlife Rehabilitation Foundation, Scottsdale, Arizona, for donating owls used in this study. D. J. Creel and A. G. Leventhal provided helpful com- ments on the manuscript. This study was supported m part by a Grant-in- Aid of Research to SJA from Sigma Xi, the Scientific Research Society. 152 Steven J. Ault and Edwin W. House Vol. 21, No. 4 Literature Cited Armington, J. C. 1974. The electroretinogram. Aca- demic Press, New York. Ault, S. J. 1984. Electroretinograms and retinal struc- ture of the Eastern Screech Owl ( Otus asio) and Great Horned Owl {Bubo virgimanus). Raptor Res, 18(2):62- 66 . Bornshein, H. and K. Tansley. 1961. Elektroretino- gramm und Netzhautstruktur der Sumpfohreule (Asio flammeus). Experientia 17:185-187. Bowmaker, J. K. and G. R. Martin. 1978. Visual pigments and colour vision in a nocturnal bird, Strix aluco. Vision Res. 18:1125-1130. Brown, K. T. 1968. The electroretinogram: its com- ponents and their origins. Vision Res. 8:633-677. Brunette, J. R. 1969. The human ERG during dark adaptation. Arch. Ophthal. 82:491-494. Chase, W. W., N. E. Fradkin and S. Tsuda. 1976. A new electrode for electroretinography. Am. J. Opt. Phys- iol. Optics 53:668-671. Duke-Elder, S. 1958. System of ophthalmology. Vol. 1 : the eye in evolution. C. V. Mosby, St. Louis. 843 pp. Fishman, G. A. 1975. The electroretinogram and elec- tro-oculogram in retinal and choroidal disease. Amer. Acad. Ophthal. Otolaryng., Rochester, MN. 45 pp. Fite, K. 1973. Anatomical and behavioral correlates of visual acuity in the Great Horned Owl. Vision Res. 13: 219-230. Hocking, B. and B. L. Mitchell. 1961. Owl vision. Ibis 103a:284-288. Ikeda, H. 1966. Electroretinograms in experimental an- imals. Pages 27-40. In O. Graham-Jones, Ed. Aspects of comparative ophthalmology. Pergamon Press, Ox- ford. Martin, G. R. 1982. An owl’s eye: schematic optics and visual performance in Strix aluco L. /. Comp. Phys- iol. 145:341-349. Niemeyer, G. 1976. Retinal physiology in the perfused cat eye. Pages 158-172. In F. Zetter and R. Weiler, Eds. Neural principles of vision. Springer- Verlag, Ber- lin. Oehme, H. 1961. Vergleichend-histologische Unter- suchungen an der Retina von Eulen. Die Zoologischen Jahrbuecher, Abt. 2, der Anatomie und Ontogenie 79:439- 478. Polyak, S. 1957. The vertebrate visual system. Uni- versity of Chicago Press, Chicago. Tansley, K., R. M. Copenhaver and R. D. Gunkel 1961. Some aspects of the electroretinographic re- sponse of the American Red Squirrel {Tamiosciurus hudsonicus loquax). J. Cell. Comp. Physiol. 57:11-19. Walls, G. L. 1942. The vertebrate eye. Cranbrook Institute of Science, Bloomfield Hills, MI. Bulletin No. 19. Wouters, L., A. deMoor and Y. Moens. 1980. Rod and cone components in the electroretinogram of the horse. Zbl. Vet. Med. A 27:330-338. Yew, D. T., H. H. Woo and D. B. Meyer. 1977. Further studies of the morphology of the owl’s retina. Acta Anatomica 99:166-168. Department of Biological Sciences, Idaho State Uni- versity, Pocatello, ID 83209, USA. Present address of first author: Department of Anatomy, University of Utah School of Medicine, Salt Lake City, UT 84132, USA. Received 20 June 1987; accepted 25 September 1987 /. Raptor Res. 2 1 (4) : 1 53 - 1 57 © 1987 The Raptor Research Foundation, Inc. Short Communications Snowy Owl Numbers on Twelve Queen Elizabeth Islands, Canadian High Arctic Frank L. Miller Predominantly white plumage and relatively large body size, form and characteristic flight pattern, facilitate aerial surveys of the Snowy Owl ( Nyctea scandiaca ) on tundra islands of the Canadian High Arctic. Snowy Owls were recorded on nine central (July 1985) and three western (July 1986) Queen Elizabeth Islands, Northwest Terri- tories, during Canadian Wildlife Service aerial surveys of Peary Caribou ( Rangifer tarandus pearyi ) and Muskoxen ( Ovibos moschatus). Bathurst, Alexander, Marc, Massey, Vanier, Cameron, Helena, Lougheed and Edmund Walker islands were sur- veyed by air between 10 and 25 July 1985 (Fig. 1A). Prince Patrick Island, Eglinton Island and Emerald Isle were surveyed by air between 4 and 13 July 1986 (Fig. IB). Survey design consisted of systematically spaced, north- south orientated line transects at 6.4-km intervals. A Bell- 206B “Jet Ranger” turbo-helicopter was flown approxi- mately 90 m above ground level (agl) and at an airspeed of approximately 160 km/hr. A four-person survey crew was used: pilot, navigator and two rear-seat observers. All four crew members spotted Snowy Owls. Rear-seat observers recorded the angle of depression from the hor- izontal plane to the position of each owl observed using a hand-held clinometer to calculate each owl’s right angle horizontal distance from the helicopter. All clinometer readings were taken when the observer in the helicopter was at a right angle to the point where an owl first was seen. Snowy Owls already in flight or flushed during ob- servations were classed as “flushed,” and owls that re- mained perched were classed as “perched.” Strip transect boundary width was set at 6°: all readings of >6° were designated as “on transect” and all readings of <5° were designated as “off transect.” Right angle horizontal distance from the helicopter to a Snowy Owl at a reading of 6° was 857 m (90 m agl/0.105, tangent of 6°). On this basis, “on transect” owls were within a strip transect 1.714 km wide, which yields an overall coverage of 27.3% of 20 855 km 2 in 1985 and 27.9% of 17 930 km 2 in 1986. Only “on transect” owl numbers were used in population and mean density estimates. Calculation of population estimates and mean densities and their asso- ciated statistics were done following procedures for anal- ysis of data arising from systematic transect surveys (Kingsley and Smith 1981). 1985 Survey In July 1985, 314 Snowy Owls were counted on nine islands (Fig. 1A) of which 81.2% were on transect (Table 1). Based on these numbers, I estimated that there were 932 Snowy Owls at a mean density of 45 owls/1000 km 2 on the entire survey area (Table 2). Eighty-seven percent of the owls counted on transect were either in flight or took flight during observations. Owls appeared to flush in response to the oncoming survey helicopter. The remainder of owls counted remained perched during observations. No record of “flushed” ver- sus “perched” owls was kept for the 59 Snowy Owls seen off transect. Relatively low numbers of Snowy Owls counted on all islands except Bathurst prevent “among-island” or “with- in-island” statistical examination of densities and distri- butions on the overall survey area. Densities of Snowy Owls were highest on lie Marc, Massey Island and north- western Bathurst Island (Tables 1 and 2). On a collective basis, numbers of Snowy Owls counted on Bathurst Island, compared to numbers counted on the other eight islands, was proportional to the two landmasses involved. Bathurst Island comprised 77.2% of the total landmass of the nine- island survey area in July 1985, and 84.7% of Snowy Owls counted on transect and 19.1% of the owls counted off transect were counted on Bathurst. Snowy Owls were rather evenly distributed over large areas of Bathurst Island. However, the number of Snowy Owls seen “on transect” on northwestern Bathurst Island (Stratum I, Fig. 1A) was significantly greater ( P < 0.05, X 2 = 6.22, df ~ 2) compared on a relative landmass basis to numbers of Snowy Owls seen “on transect” on north- eastern Bathurst (Stratum II, Fig. 1A) and southern Bath- urst (Stratum III, Fig. 1A) (Table 1). Collared Lemmings ( Dicrostonyx torquatus) were abun- dant over most, if not all, of Bathurst Island in July 1985 Lemmings were constantly underfoot in our field tent camp on central Bathurst and were seen at fuel caches on other parts of the island. Most burrows observed showed fresh signs of excavation. We also saw four Arctic Fox ( Alopex lagopus ) dens with pups on Bathurst Island, a positive sign of high lemming availability. 1986 Survey Only three Snowy Owls were counted during the 1986 survey, although the landmass surveyed was 86% as large as that surveyed in 1985 (Fig. 1A, B). All three owls were on Prince Patrick Island: two were on transect and one 153 154 Short Communications Vol. 21, No. 4 104 ° 102 ° 96 ° Figure 1. Location of Canadian High Arctic Islands where numbers of Snowy Owls were obtained by aerial survey. (A) nine central Queen Elizabeth Islands (11 survey strata), July 1985; and (B) three western Queen Elizabeth Islands (five survey strata), July 1986. Winter 1987 Short Communications 155 156 Short Communications Vol. 21, No. 4 Table 1 . Numbers and distributions of Snowy Owls seen during aerial survey of nine central Queen Eliz- abeth Islands, Canadian High Arctic, North- west Territories, July 1985. Island (Survey Stratum) On Tran- sect Off Tran- sect Total Bathurst, NW (I) 71 23 94 Bathurst, NE (II) 80 10 90 Bathurst, S (HI) 65 14 79 Alexander (IV) 7 4 11 Marc (V) 1 1 Massey (VI) 8 6 14 Vanier (VII) 5 1 6 Cameron (VIII) 13 1 14 Helena (IX) 2 2 Lougheed (X) 3 3 Edmund Walker(XI) 0 Strata I— III 216 47 263 Strata IV-XI 39 12 51 Strata I-XI 255 59 314 was off transect. On this basis, I estimated only seven owls to be on the entire survey area and calculated a density of only 0.4 owls/1000 km 2 in July 1986. In July 1986 the survey crew searched for lemmings or fresh signs in the base-camp area at Mould Bay, Prince Patrick Island, and in other areas each time the helicopter landed for refueling. No lemmings or fresh burrows were seen anywhere that was searched on the three-island com- plex (Fig. IB). In addition five Arctic Fox dens were found, but none were active. Tener (1963) counted only 13 Snowy Owls during an aerial survey of 7.8% of Bathurst, Alexander, Massey, Vanier and Cameron islands between 19 June and 7 July 1961. Assuming 13 Snowy Owls were counted “on tran- sect,” an extrapolated estimate would yield 166 owls total on five islands. On this basis the number of Snowy Owls on the five islands was at least 5.5 x greater in 1985 than in 1961. On 23 July 1961 Tener (1963) counted 10 Snowy Owls on Prince Patrick (4.2% coverage), six owls on Eglin- ton (5.9% coverage), and four owls on Emerald (9.2% coverage) on 24 July 1961. Assuming all 20 owls counted in July 1961 were on transect, the resultant estimate would be 384 owls versus only seven owls estimated in July 1986. Magnitudes of annual variation in Snowy Owl numbers is best illustrated with data from Eglinton Island. Miller et al. (1975) counted five, 56 and 27 Snowy Owls and estimated 20, 111 and 54 owls on Eglinton in summers of 1972, 1973 and 1974, respectively. Yet not a single owl was counted in summer 1986. The importance of lemmings to Snowy Owls is well Table 2. Mean densities and numbers of Snowy Owls on 1 1 survey strata of nine central Queen Elizabeth Islands, Northwest Territories, July 1985, obtained by aerial survey. Area Survey Stratum (Island) Stratum Size (km 2 ) Area Surveyed . (km 2 ) Owls/1 000 KM 2 Population Estimates Mean 3 95% C.I. b Estimate 3 95% C.I. b I (Bathurst, NW) 4080 1113 64 55-73 260 224-297 II (Bathurst, NE) 6650 1794 45 36-53 297 238-355 III (Bathurst, S) 5360 1478 44 28-60 236 151-321 IV (Alexander) 490 129 54 25-84 27 12-41 V (Marc) 56 15 67 0-719 4 0-40 VI (Massey) 440 122 66 29-102 29 13-45 VII (Vanier) 1130 303 16 5-28 19 6-31 VIII (Cameron) 1060 293 44 20-69 47 21-73 IX (Helena) 220 90 22 9-54 5 0-12 X (Lougheed) 1300 352 8 2-19 11 0-24 XI (Edmund Walker) 69 17 I-III 16 090 4386 49 43-56 793 689-896 IV-XI 4765 1320 30 23-36 141 109-173 I-XI 20 855 5705 45 40-50 932 827-1038 a The discrepancy of three owls (935) obtained by the summation of 11 individual strata (I-XI) versus the single calculation of 932 owls for strata I-XI is a rounding error. Also, use of whole number values for mean densities prevents exact recalculation of estimates from tabular material: e.g., (survey area x mean density = estimate) therefore, (20 855 x 45/1000 = 938) but when actual mean density of (0,0447 owls/km 2 ) is used (20 855 x 0.0447 = 932 ). Note that recalculation of any mean density values requires that the tabular value first be reduced to a unit of one (km 2 ). b Negative values reported as zero (0). Winter 1987 Short Communications 157 known. One obvious difference between the survey areas was the high numbers of lemmings present in July 1985 and the extremely low number apparently present (none counted) in July 1986. Likely, high numbers and wide- spread distribution of lemmings accounted for the com- monness of Snowy Owls throughout much of the survey area in 1985 compared to 1961. However, lemming pop- ulations can be asynchronous on adjacent islands. For example in summer 1958 lemmings were abundant (ap- prox. y 50 m 2 ), and so were Snowy Owls on Prince of Wales Island, while no lemmings or owls could be found on Somerset Island 40 km away (T. W. Barry, pers. comm.). Absence of lemmings on the July 1986 survey area could account for the difference in Snowy Owl numbers in 1961 vs. 1986. Unfortunately, I have no knowledge of what proportion of the surveyed area in each year was suitable nesting habitat or what proportion of the owls seen were associated with nests. In general plant cover is relatively sparse on the western half of Prince Patrick Island, the northern tip of Eglinton Island, and the southern end of Lougheed Island compared to the remainder of the areas surveyed in 1985 and 1986. Acknowledgments The survey was supported by the Canadian Wildlife Service (CWS), Environment Canada, and Polar Conti- nental Shelf Project (PCSP), Energy, Mines and Re- sources Canada. I am grateful to G. D. Hobson, Director, PCSP, for continued support of my research. I thank J. R. W. McGillis (1985-86) and A. Westcott (1986), CWS, for their assistance as observers; and D. M. White, pilot, and B. G. Wight, engineer, Quasar Aviation Ltd., for their assistance as navigator-spotters on the 1985 survey; I thank S. J. Barry, CWS, for statistical assistance and S. M. Popowich, CWS, for drafting the figures. T. W. Barry, R. G. W. Edwards, and G. L. Holroyd, CWS, critically read the manuscript and provided helpful comments. Literature Cited Kingsley, M. C. S. and G. E. J. Smith. 1981. Analysis of data arising from systematic transect surveys. Pages 40-48. In F. L. Miller and A. Gunn, Eds., and S. R Hieb, Publications Ed. Northwest Section, The Wild- life Society: proceedings symposium on census and in- ventory methods for population and habitats. Published by the Forest, Wildlife and Range Experiment Station, University of Idaho, Moscow, Idaho, Contribution No 217. 220 pp. Miller, F. L., R. H. Russell and A. Gunn. 1975. Distribution and numbers of Snowy Owls on Melville, Eglinton, and Byam Martin islands, Northwest Ter- ritories, Canada. Raptor Res. 9:60-64. Tener, J. S. 1963. Queen Elizabeth Islands game sur- vey, 1961. Can. Wild l. Serv. Occas. Pap. No. 4. 50 pp Canadian Wildlife Service, Western and Northern Re- gion, 2nd Floor, 4999-98 Avenue, Edmonton, Al- berta T6B 2X3, Canada. Received 5 January 1987; accepted 15 September 1987 J Raptor Res. 21(4):157-159 © 1987 The Raptor Research Foundation, Inc. A New Method to Selectively Capture Adult Territorial Sea-Eagles Anthony L. Hertog Adult eagles are difficult to capture in their territory or breeding range. In northern Australia adult territorial White-bellied Sea-eagles ( Haliaeetus leucogaster) were often attracted to capture sites but usually perched and watched from nearby. However, some came to bait but only after non-target birds had disturbed the trap. Therefore, a new, manually-operated, single-noose system was developed and compared with trapping success of three conventional methods (i.e., cage traps, cannon netting and eagle-trig- gered multi-noose systems). The new capture system re- quires a concealed hide (e.g., a camouflaged vehicle) lo- cated 200 m from the bait. One operator remains at the hide while another prepares the capture system. A capture site (approx. 2 m 2 ) clear of debris and vegetation is chosen well within an eagle’s territory and in view of the hide. Bait (normal fish prey) is aligned such that the head is facing away from the hide and secured with two 300 x 10 mm steel pegs (Fig. 1). Alignment is important because eagles usually grasp the bait lengthwise with both feet, and the noose when sprung easily snares the eagle’s legs from the side; otherwise the noose may slide up the back of the eagle. Vegetation is cleared next to the hide, and one end of a 5 m length of 10 mm surgical tubing with a loop tied at each end is pegged to the ground next to the hide entrance. The other end is stretched and pegged beyond the hide (Fig. 1). A fishing reel (120 mm dia) bolted to flat steel (300 x 50 x 8 mm thick) is placed on the ground next to the tubing at the farthest point from the bait (Fig. la). The reel held 250 m of 18 kg monofilament line and 158 Short Communications Vol. 21, No. 4 plan view of capture site; (c) enlarged side view of capture site to illustrate the angle of line pull; (d, e) enlarged views of trigger mechanism. 2.5 m of 18 kg black, plastic-coated, trace wire to form a pre-prepared noose using a running slip knot (Jenkins, M. A., N, Am. Bird Band. 4(3):108-109, 1979). Trace wire is multi-stranded and lays flat on the ground com- pared to monofilament line and is thus less likely to be disturbed by non-target animals. The noose is carefully placed around the bait so that the free end is 300 mm in front of the head of the bait and held at two corners with small clumps of soil (Fig. lb). Sides of the noose are placed 50 mm from and parallel to the bait, and a running slip knot is placed flat on the ground 300 mm behind the tail of the bait directly in line with the hide. A small forked branch, approximately 18 mm dia x 750 mm length, is pushed into the ground 2 m from the center of the noose area, and the monofilament line lays through the fork. Fork height is adjusted so that the closest point of an imaginary line from the fork to the free end of the noose is about 50 mm above the front of the bait (Fig. lc), which prevents the noose from snagging on the bait. Line is lightly held close to the base of the branch with small pieces of twig as is the remainder of the noose to minimize distur- bance by non-target animals. Combined weight of twigs does not exceed that of soil clumps; otherwise the line will pull along the ground and become entangled with the bait. A steel hoop was pegged across the main line about 5 m from the bait to prevent a captured bird from rising with the line attached. At the hide the monofilament line is tied behind the knot in the bait end of the tubing using a clove hitch knot, and the peg holding the tubing is lifted slightly and turned 180° to allow the tubing to slip off easily when pulled upwards (Fig. Id, e). The system is operated from the hide using a piece of 4 mm fencing wire bent 90° at one end and looped to form a handle at the other. The bent end is placed under the tubing just behind the front peg so that when pulled up the tubing is released from the peg which also pulls the line and causes the noose to quickly tighten around an eagle’s legs. Attempts to escape further tightens the noose and injury is minimized by the elasticity of the tubing. A captured bird is easily subdued with a hand-held catching net. When the system is ready, the person at the bait end moves to a concealed position well away from both the trap site and the hide in an attempt to deceive eagles that the area has been vacated; radio contact is maintained with the operator in the hide. Other birds, especially Black Kites (Milvus migrans ) and Whistling Kites ( Haliastur sphenu- rus ), often gather and alight on or near the bait often causing a target eagle to attempt to pick up the bait or to scatter other birds. In either case the eagle usually returns Winter 1987 Short Communications 159 Table 1. Comparative capture success for White-bellied Sea-eagles using four capture methods. Method No. Trap-Days No. Eagles Captured % SuCCESS a Total Within Terri- tory 1 ’ Total 0 Target Eagles ' 1 All Eagles Target Eagles Within Terri- Total TORY Manually-operated single-noose system 13 6 5 4 0.38 0.31 0.67 Eagle-triggered multi-noose system e 17 9 1 1 0.06 0.06 0.11 Cannon net 1 19 16 14 1 0.74 0.05 0.06 Cage trap® 56 15 4 0 0.07 0 0 a No. per trap-day. b Excludes trap-days when eagles not seen or sites disturbed by non-target animals. c Juveniles, target and transient adults. d Adults which maintained a fixed year-long territory. e Modeled after Wegner, W. A., /. Wildl. Manage. 45(l):248-250, 1980. f See Addy, C. E., U.S. Fish and Wildl. Serv., Laurel, MD, 164 pp., 1956. e Cage traps (3 x 2 x 2 m high) were positioned for three months and baited for an average of four days. Together such a trapping attempt constituted one trap-day. quickly and lands on or near the bait. Once an eagle is standing on the bait and feeding, the operator waits until the eagle lifts its head and only then triggers the system. In this study target eagles were adults which maintained fixed year-long territories rather than transient adults or juveniles. Significantly more target eagles were captured with the manually-operated noose system than with the eagle-triggered noose system (P < 0.05), cannon netting ( P < 0.01) or cage traps (P < 0.01) (Fisher Exact Prob- ability Test, Table 1). Capture success for all eagles (i.e., target, adult transient, juvenile) using the manually-op- erated noose system was also significantly greater than that using the eagle-triggered noose system (P < 0.05) and cage traps (P < 0.01), but not for cannon netting. Baited cage traps are commonly used to capture birds (Day et al., Wildlife management techniques manual, 4th Ed. The Wildl. Soc., Washington, DC, 1980) and need little modification for raptors. In northern Australia cage traps have been used successfully in capturing Black Kites and Whistling Kites (A. Hertog, unpubl.; J. Estbergs, pers. comm.), but not White-bellied Sea-eagles. In 56 trap- days in areas where eagles were known to frequent, only four were captured (seven percent success) and none were target eagles. Cannon netting was very successful for ju- venile and adult transient eagles with a 74% capture suc- cess in 19 trap-days. Only one target eagle was captured in 16 trap-days (six percent success). Although attracted to the vicinity of the trap site, target eagles tended to be wary of the net which was difficult to conceal. An eagle-triggered noose system was set 17 times but only one target eagle was captured (six percent success). Failure was due to disturbance to the noose by eagles (N = 3) or capture/disturbance by Whistling Kites (N = 5). Even when those disturbance data were excluded from results capture success was still poor (11%). In 13 trap-days five eagles were captured (38% success) using my new, manually-operated capture system, and two were missed because the trap was triggered prematurely by the operator. Other failures were due to the absence of adults at the trap site (N = 3) and disturbance at the trap site by mammals and reptiles. Excluding these data, cap- ture success was 67% for four target eagles. Apart from being superior to other conventional trapping methods, my system has the advantages of being inexpensive, quickly set up, and easily concealed. In addition birds can be selectively captured (i.e., specific sex, age, status, species) thus making the technique useful in studies with other raptors. Acknowledgments This study was part of a larger study of vertebrate predators in a tropical ecosystem at Kapalga (L. Corbett, unpubl.). I thank L. Corbett for stimulating discussions, especially during early frustrations when developing the trapping system and for critically reviewing the manu- script; also J. Randall for his assistance and patience dur- ing trapping sessions. D. Baker-Gabb, M. Lonsdale, J. Taylor, and M. Ridpath provided useful comments on early drafts as did A. Harmata on the final draft. This paper is CSIRO Tropical Ecosystems Research Centre Contribution no. 560. CSIRO, Division of Wildlife and Ecology, Tropical Ecosystems Research Centre, PMB 44 Winnellie, N.T. 5789, Australia. Received 10 December 1986; accepted 5 October 1987 160 Short Communications Vol. 21 , No. 4 J Raptor Res. 21(4):160 © 1987 The Raptor Research Foundation, Inc. Northern Cardinal Head Attached to the Toe of a Sharp-Shinned Hawk Thomas W. Carpenter and Arthur L. Carpenter On 16 May 1986 while using mist nets to capture and band migrant hawks at Whitefish Point, Chippewa County, Michigan, we captured an immature male Sharp- shinned Hawk ( Accipiter striatus ) with the head of a Northern Cardinal ( Cardinalis cardinalis ) attached to its left hallux (hind toe). The cardinal had apparently clamped its beak tightly on the hawk’s toe. Both the head and toe broke off together when we untangled the hawk. The head was very necrotid and evidently had been attached to the hawk’s toe for a considerable length of time. The hawk was in average condition (wt = 101.2 g). Average weight of 184 immature male Sharp-shinned Hawks we banded at Whitefish Point during the springs of 1985-87 was 100.2 g (range = 75.7-125.9 g; Carpenter and Carpenter, pers. obs.). Unusual injuries have been previously reported for the Sharp-shinned (Kelley and Kelley, Wilson Bull. 81:209-210, 1969; Evans, Auk 94. 585-586, 1977) but none of the type just described. Acknowledgments Edward H. Burtt, Jr., Jimmie R. Parrish and three anonymous reviewers made many helpful suggestions on an earlier version of the manuscript. This note is a con- tribution of the Whitefish Point Bird Observatory. 3646 S. John Hix, Wayne, MI 48184. Received 25 June 1987; accepted 20 September 1987 Non-releasable Animal Placement Program (NAPP) — The Non-releasable Animal Placement Program is a com- puter-based information system for non-releasable species of mammals, birds, reptiles and amphibians. The purpose of the program is to maintain a clearinghouse for wildlife professionals who deal with non-releasable, permanently injured or imprinted animals and placement of surplus animals into various wildlife programs throughout North America. A monthly report is published listing all available and requested animals. Subscription fee for the NAPP is $10.00/yr. There is no charge to list or request an animal. For additional information contact NAPP, Animal Rehabilitation Center, Inc., 449 Edgefield Lane, Midlothian, Texas 76065. J, Raptor Res. 21(4):161-176 © 1987 The Raptor Research Foundation, Inc. THE JOURNAL OF RAPTOR RESEARCH INDEX TO VOLUME 21 Compiled by Jimmie R. Parrish and F. Ruth Beck The Officers and Board of Directors of the Foundation have approved the production of a 20-year index of the Foundations publications. Subsequent yearly indexes will serve as a continuum to the 20-year volume with additional updates provided as necessary. Scientific names follow the A.O.U. checklist where appro- priate for avian scientific names, and all others follow the most current sources available. A. a. acadicus, 45 A. americana, 73 A. crecca, 73 abandoned fields, 57 abandonment, 53, 57 abdominal operations, 22 Abies concolor, 39 Abies grandis, 11 Abies lasio-carpa, 98 Accipiter gentilis, 22, 144 Accipiter nisus, 4-1 Accipiter striatus, 165 Acepromazine Maleate, 152 active nest, 27 activities and habitat utilization, 58 activity, 45 activity during the wintering phase, 68 acuity, 156 acupuncture, 24 adaptation, 51 adaptive breeding strategy, 79 adult mortality, 49, 51 adult pair, 38 adult survival, 44 Aegolius funereus, 98 Aegolius funereus richardsonii, 45 aerial and ground censuses, 88 aerial attacks, 73 aerial photographs, 57 aerie, 79 (see eyrie) aerodynamic performance, 133 aetiology, 21 Africa, 76 African predatory birds, 46 age and growth, 79 aged, 58 Agelaius phoeniceus, 62 aggression, 8, 59, 60, 122 aggressive, 60 aggressive behavior, 28, 30, 67 agnostic interactions, 58 agonistic encounters, 118 agonistic interactions, 59 agricultural practices, 16, 17 air tracking, 147 Airola, Daniel A., 121 airplane tracking, 147 Alaska, 114 aldrin, 47 alfalfa, 57 Allen Press, Inc., 2 Alopex lagopus, 158 alternate nest, 27 alternate perch sites, 69 ambient temp, 35, 36, 37 American Bison, 47 American Coot, 61 American Crow, 68 American Elm, 14 American Kestrel, 14, 24, 71, 70, 126 American Swallow-Tailed Kite, 124 American Sycamore, 14 American Wigeon, 73 amphibians, 72 Amphotericin B, 22 anaesthetic techniques, 22 analysis of prey remains and pellets, 81 analysis of survival data, 137 161 162 Index to Volume 21 Vol. 21, No. 4 analyzing movement, 140 Andersen Award, 1, 83 anesthesia, 50, 51, 152 angular data, 140 animal locations, 134 annual mortality rate, 55 annual survival, 44 antenna breakage, 145 Anus acuta, 73 Applegate, Roger D., 68 apricot, 12 aquatic habitat types, 29 Aquila chrysaetos, 37, 46, 67, 68, 80, 85, 103, 118 Aquila chrysaetos japonica, 79 Arctic Circle, 111 Arctic Fox, 158, 161 Arctic Ground Squirrel, 80 Arctic Hare, 80 Ardea herodias, 61 Argentina, 40 arithmetic mean location, 140 Arizona, 16, 35, 45 Aroclor 1248, 49 Arragonian Pyrenees and Cantabrian Mountains, 75 Artemisia spp., 86 Artemisia tridentata, 57 arthropods, 96, 126 artificial incubation, 55, 71 artificial insemination, 24, 55, 70, 71 artificial nest site, 117 artificial nests, 116 Arundinaria gigantea, 124 Arvicola terrestns, 12 Asia otus, 100 aspect, 33, 36, 37 aspen, 86 Aspergillosis, 24, 54 assemblage of breeding raptors, 46 Atriplex confertifolia, 57 attacking intruders, 60 atypical incubation, 33 auditory cues, 63 Ault, Steven J., 152 Australia, 162 Australia cage traps, 164 automobile tracking, 147 availability of carrion, 85 availability of nest sites, 37 availability of prey, 80 avian breeding biology, 79 avian prey selection, 4 avian retina, 152 axial tomography, 24 azimuth, 147 backpack transmitters, 131 bacterial infection, 25 Badger, 145 baited cage traps, 1 64 Bal-chatri trap, 15, 57 Bald Eagle, 22, 27, 30, 45, 68, 81, 118, 131, 148, 150 Baldcypress trees, 27, 30 Baldcypress/tupelo-gum forest type, 27 band recovery, 44 band recovery models, 137 banded, 23, 51, 55 barren soils, 57 Barrows, Cameron W., 95 Bateleur, 48 Beck, Thomas W, 116 beetles, 124 behavior, 8, 10, 11, 27, 57, 58, 85, 98, 103 feeding, 125 hunting, 32 behavioral variability, 35 Belthoff, James R., 80 Berger, Daniel D., 68 Beta vulgaris, 57 betalights, 98 Bighorn Sheep, 85 Bildstein, K., 1 biological council, 24 biological diversity, 47 biological status, 40 biomass, 4, 63 biomass fed, 80 Biotelemetry, 131 biotic factors, 27 Bird Banding Laboratory, 133 Bird, D,, 1 bird-days, 134 birds, 81 Bison bison , 47 Black Kites, 23, 163, 164 Black Locust, 14 Black Oak, 14 Black-tailed Jackrabbit, 118 blind, 51 body mass, 79 body size, 35 body weight, 18 “Boke of St. Albans,” 21 Winter 1987 Index to Volume 21 163 BOPA, 32, 63 Boreal Owls, 45, 98 Bos taurus , 85 Botta Pocket Gophers, 97 Boyce, D. A., 35 Brachyramphus craveri, 4 Branta canadensis , 61, 121 breeding, 58 activities, 57 and death rates, 47 area, 27, 29 behavior, 57, 119 biology, 8 chronologies, 39 cycle, 8 density, 112 dispersal, 44 Northern Harriers, 57 season, 8, 45 sites, 47 Britain, 21, 22, 24, 47 British Columbia, 33 British Isles, 3, 111 Bromus tectorum, 57 brood abandonment, 131 size, 80 rearing, 32 brooding, 58, 59, 60, 61, 116 brooding behavior, 34 broods, 53 Brown Bullhead, 81 Bryce Canyon National Park, 38 Bubo virginianus, 22, 57, 74, 103, 125, 152 Bulgaria, 8 Bull, Evelyn L., 77 Bulrush, 57 bumble foot, 51 Bunck, Christine M., 137 Bureau of Land Management, 57 Buteo buteo, 12 Buteo jamaicensis, 16, 37, 46, 60, 67, 103, 118 Buteo lagopus, 16 Buteo realis, 16, 46, 118 Buteo rufinus, 8 Buteo swainsom, 16 C.I.T.E.S. classification, 40 cached, 52 caching, 15 cage traps, 162, 164 Calcarius lapponicus, 112 calcium carbonate, 51 calcium phosphate, 51 California, 3, 16, 17, 33, 35, 95, 121, 122, 150 California Condor, 47 call variation, 45 calling activity, 45 calls of an individual eagle, 45 Calocedrus decurrens, 116, 122 Caloplaca sp. ,111 Canada, 24, 40, 44, 51, 113 Canada Goose, 61, 121 Canadian Arctic, 80 Canadian Council for Animal Care, 24 Canadian Wildlife Service, 40 Canis latrans, 61, 85 cannibalism, 32 cannon nets, 89 netting, 162, 164 Canyonlands National Park, 39 Cape Vulture, 79 captive breeding, 21, 25, 72 captive breeding and reintroduction project, 74 captive propagation, 46, 47, 74 capture, 86, 162 of Prey, 63 rate, 17 success, 15, 18, 164 captured, 57 carcass necropsy, 21 cardiac hemorrhage, 51 Cardinalis cardinalis, 165 Carduelis flammea, 112 Carpenter, Arthur L., 165 Carpenter, Thomas W., 165 Carpinus orientalis, 8, 12 carrion, 68, 69, 72, 89, 91 Cary a ovata, 80 Carya spp., 14 casement display, 143 Cat, 156 Cathartes aura, 67 cavities, 80 Cely, John E., 124 census, 15, 75 center of gravity, 133 cereal crops, 16 “chatter” calls, 45 cheatgrass, 57 chemotherapy, 22 Cheney, Carl D., 103 Chi-square (x 2 ), 61, 149 chlorinated hydrocarbon insecticides, 25 circular statistics, 140 164 Index to Volume 21 Vol. 21, No. 4 Circus cyaneus, 32, 57, 72 Citellus citellus, 12 Clark, R., 1 Clethrionomys gapperi, 45 Clevenger, Anthony P., 33 cliff aspect, 112 face area, 36 height, 36, 113 Width, 36 Cliff Swallows, 32 climate, 8, 23, 35, 57 climatological, 75 clinically, 23 clinicopathological, 24 cloud cover, 45 clustering techniques, 143 clutch size, 12, 49, 53, 55 clutches, 117 second 49, 54, 55 Cnemidophorus tigris, 57, 62 Coast Moles, 97 Cochran, William W., 68, 147 coefficients, 51 Colaptes auratus, 11 Collared Lemmings, 158 Collopy, M., 1 colony management, 49 color band numbers, 81 Colorado, 39, 44, 45, 85 Colorado Plateau, 67 Colorado River Canyon, 39 colored plastic, 58 colour phase, 40 Columba palumbus, 144 Columba sp., 156 Commentary, 1, 2, 40 commercial diet, 51 Common Barn-Owl, 3, 25, 49, 55, 74 Common Buzzard, 12 Common Merganser, 119 Common Moorhen, 72 Common Pear Common Raven, 37, 67, 86, 90, 91, 113, 1 Common Vole, 1 1 comparative capture success, 164 compass directions, 58 competition, 122 components, 152 computer software programs, 138 conglomerate cliffs, 37 Connecticut, 50 conservation, 22, 24, 25, 46 conspecifics, 46 contaminant studies, 49 contour-hugging, 17 convex polygon, areas, 145 Cooper, John E., 21 copulation, 58, 70 core area, 140 cornea, 152, 153 Corvus brachyrynchos, 68 Corvus corax, 37, 67, 80, 86, 113, 119 Corvus frugilegus, 119 Corvus monedula , 1 1 cost, 47 cost effectiveness, 148 Costa Rica, 17 Coturnix coturnix, 71 Coturnix Quail, 71 courtship, 28, 30, 33, 58, 59, 80 period, 45 rituals, 70 cover, 16 Cox Proportional Hazards Model, 138 Coyote, 61, 85 Craveri Murrelets, 4 Crawford, Walter C., 74 crayfish, 81 critical fusion frequency, 155, 156 cropland, 14 crops, 17 cross fostering, 24 cryoprotectant, 72 cryotherapy, 24 cultivated fields, 57 cultivated habitats, 57 cyclic signal fade, 147 Cynomys spp., 103 cypress, 126 daily activity, 16 energy budget, 15, 18 energy expenditure (DEE), 79 rate and direction of travel, 147 temp, 18 Dali’s Sheep, 85 dark adaptation, 152, 153 day roosts, 46 daylength, 35 dbh (diameter at breast height), 77, 81 DDE, 47, 49 DDT, 47 in human milk, 47 dead embryos, 53 Winter 1987 Index to Volume 21 165 Deer Mouse, 45, 62, 77 defect, 51 defend nests, 122 defended winter territories, 126 defense, 61 defensive posture, 11 deforestation, 46 delayed molt, 126 demography, 144 Denmark, 40 density, 16, 80, 85, 112 dependent, 80 depredation, 85 rate, 86 detectability differences, 149 detecting and describing the structure of home range, 143 development, 81 of flight ability, 79 of young, 8 dho-gaza, 57 diagnosis, 22 diagnostic aids, 21 procedures, 24 dialyzed, 70 Dicrostonyx torquatus, 158 dieldrin, 47 diet, 3, 49, 51, 55, 144 overlap, 46 shifts, 95 dietary preferences, 4 differential prey selection, 144 differential use of trees and cliffs at nest sites, 46 differential visibility, 149 digestion, 15 dimethyl sulfoxide, 72 dimethylacetamide, 72 Dtospros virginiana, 14 Dipodomys sp., 62 directional distribution, 148 discriminant function analysis, 45 disease, 86 dispersal, 46, 144 and demography, 145 distance, 81 distribution, 14, 85 distribution, status, conservation of raptors in Mad- agascar, 46 distributions of Snowy Owl, 161 disturbances, 57 disturbed, 55 diurnal activity, 45 diving ducks, 131 docking, 87 percentages, 86, 88, 89 domestic calves, 85 domestic sheep, 85 domestication, 47 dominant position, 118 Donazar, Jose A., 75 double clutching, 24 Douglas fir, 39, 46, 77 “dueting” (mutual calling), 79 Dugoni, Joseph A., 27 Duke, G., 1 Duley, Peter A., 32 Dumetella carolinensis , 1 Duncan, James, R., 125 Dunnett’s £-Test, 155 duplex retina, 156 Dusky-Footed Woodrat, 95, 97 Dytiscus sp., 124 eagle activity, 28, 30 depredation, 87 depredation rate, 91 harassment and frightening techniques, 87 eagle-triggered noose system, 164 Eakle, Wade L., 45 East Africa, 21 Eastern Cottontail, 81 Eastern hornbeam, 12 Eastern red cedar, 14, 80 Eastern Screech-Owl, 49, 80 ecological, 79 ecology, 79, 80, 100 economic losses, 85 edaphic Factors, 27 education materials, 48 egg abandonment, 53 dimension, 12 disappearance, 53 failure, 49 necropsy, 21 size, 8 -laying, 32 eggs, 49 hatched, 54, 59 that failed, 53 eggshell thickness, 24, 49 Elanoides forficatus, 124 electroretinograms, 152, 153 Eleonora’s Falcon, 47 emaciation, 51, 73 166 Index to Volume 21 Vol. 21, No. 4 embryo death, 49 development, 24, 53 necropsy, 21 embryos, 47 emergency food supply, 32 emigration, 145 endangered species, 40 endemic raptors, 47 endoscopy, 24 endrin, 49 energetic changes, 131 cost, 15 energy consumption, 15 expenditure, 15 input, 5 requirements, 79 England, 16, 113 environmental contamination, 25, 55 EPA, 103 Eqalungmiut Nunaat, 111 escape speed, 131 Eucalyptus spp., 126 Eurasian Kestrel, 16 Europe, 4, 25, 76 European Starling, 17 European Susliks, 12 Eutamias speciosus, 96 euthanasia, 22, 51 Eutriorchis astur, 47 ophthalmoscopical examination, 152 experimental design, 134 exponential distribution, 138 extinction, 47 extirpation, 48 eyrie, 33, 34, 36, 67, 111 (see aerie) aspect, 37 elevation, 113 height, 36 facilities, 56 facilities and maintenance, 49 Falco columbarius, 25, 38, 39 Falco concolor, 47 Falco eleonorae, 47 Falco mexicanus, 32, 35, 44, 46, 67, 70 Falco newtoni newtoni, 47 Falco per egrinus, 32, 33, 38, 42, 61, 67, 70, 111, 147 149 Falco punctatus, 25 Falco richardsonii , 38 Falco rusticolous, 40, 71, 80, 111 Falco sparverius, 14, 24, 70, 126 Falco tinnunculus, 16 falconers, 40, IQ, 12 Falconry, 21, 46, 47 fall migration, 68 Family Apodidae, 32 fat reserves, 133 feathering, 8 Federal Communications Commission, 133 Fedynich, Alan M., 72 feeding 28, 59, 60, 81 behaviors, 32 ecology, 98 efficiency, 5 flocks, 124 platform, 50 fertility, 70, 71, 72 Ferruginous Hawk, 16, 46, 118 fiber optic system, 153 fidelity, 44, 47 field observations, 57 field research funding, 7 Finland, 40, 116 Flammulated Owl, 45, 120 fledged Young, 37, 46, 54 fledging of young, 33 period, 12 success, 53 fledglings, 51 flicker stimuli, 152 tests, 153 flight, 28 flight distances, 133 patterns, 59 Florida, 16, 126 Fluoride, 49 food abundance, 80 choices, 103 compaction, 54 consumption, 79 deliveries, 33 habits, 46, 80, 101 niche, 98 preference, 68 requirements, 79 resource, 79 sources, 147 supplement, 23 transfers, 60, 70 foraging, 14, 28, 29, 30, 46 areas, 46 behavior, 98, 100 Winter 1987 Index to Volume 21 167 characteristics, 80 efficiency, 64, 107 habitat, 14, 74, 98 patterns, 63 quality, 126 site Selection, 16 sorties, 59 strategy, 14, 103 forest habitats, 101 Foundation for Field Research, 7 Fourier transformation, 142 foveal rods, 156 fractured leg, 54 French and Catalonian Pyrenees, 75 frequency and number of prey items, 59 “frounce,” 21 Fulica americana, 61 Fuller, Mark R., 131, 143 funding, 47 G. passerinum , 11 Galapagos Islands, 3 Gallinula chloropus, 72 Gardner, Kirk, 118 Geissler, Paul H., 143 genetic diversity, 49 gentamicin, 22 geometric mean, 141 Georgia, 17 Ghost Crabs, 100 Glaucidium gnoma, 11 Glaucomys sabrinus , 95, 96 Gleditsia triacanthos, 14 Glue, David E., 3 glue-on transmitters, 131 glycerol, 70, 72 Godfrey, Ralph D., 72 Golden Eagle, 37, 46, 67, 68, 80, 85, 103, 118 depredation on lambs of domestic sheep, 85 migration, 69 Goshawk, 144, 145 Gotland, 145 Grand fir, 77 granite, 37 grassland, 14, 17, 39 gravel pits, 57 Greasewood, 57 Great Basin Region, 57 Great Blue Heron, 61 Great Gray Owl, 116 Great Horned Owl, 22, 57, 74, 103, 125, 152 Green Lizard, 11 Green- winged Teal, 73 Greenland, 40, 111, 114 Peregrine Falcon Survey Area, 113 White-fronted Goose Study, 111 Grey Squirrel, 145 grid-cell-based analyses, 145 Griffon Vulture, 75, 76 ground squirrel, 103 growth, 79 of young, 8 gular flutter, 117 Gymnogyps californianus, 47 Gyps coprotheres, 79 Gyps fulvus, 75, 76 Gyrfalcon, 40, 71, 80, 111, 113, 114 habitat, 39, 28, 39, 57, 100 analyses, 57 availability, 14, 16 map, 149 mesic, 63 physiography, 18 preference, 62 preference studies, 149 selection, 27, 45, 149 type, 14, 18, 29, 61, 149 use, 44, 45, 57, 61, 100, 126, 131 use-females, 62 use-males, 61 utilization, 57 hacked back, 23 haematocrit, 23 Hager, Joan C., 32 Haines, Susan L., 81 Hakusan Range, 79 Haliaeetus leucocephalus, 22, 27, 45, 68, 81, 118, 131, 148, 150 Haliaeetus leucogaster, 162 Haliaeetus vociferoides, 47 Haliastur sphenurus, 163 Halogeton glomeratus, 57 hand-rearing, 24 harassment, 85, 89 Hardaswick, Victor, 70 harmonic mean, 141 harrier behavior, 57 Harris, James O., 27 harvestable surplus, 44 hatchability, 70, 71 hatching date, 117 168 Index to Volume 21 Vol. 21, No. 4 interval, 8, 9 of young, 8 periods of F. c. richardsonii, 39 success, 53 weight, 8 Hawk Mountain Research Award, 82 hawking dragonflies, 32 Hays, Larry L., 67 Hayward, Gregory D., 98 heat caused stress, 37, 117 Heather Vole, 77 height, 36 and aspect of nest cliffs, 44 helicopter, 86 Hen Harrier, 62 Henjum, Mark G., 77 Henke, Robert J., 74 herbicide, 50 Hertog, Anthony L., 162 heterozygosity, 51 Hickory, 14 Hill, James M,, 3 Hirundo pyrrhonta, 32 Hohmann, Janet E., 77 Holt, Denver W., 120 Holthuijzen, Anthonie M. A., 32 home range, 14, 57, 61, 63, 64, 81, 131 of juveniles, 79 data, 140 studies, 134 homeopathic remedies, 24 Honey Locust, 14 horse, 156 House, Edwin W., 152 House Mouse, 126 House Sparrow, 15 hovering, 15, 73 Howard, R., 1 Hoy, Judy A., 120 Hudson Bay, 113 human activities, 32 disturbance, 11, 14, 16, 86 humidity, 71 Hunt, W. Grainger, 1, 149 hunting, 58, 59, 60, 116 activity, 57 behavior, 58, 73 by female, 60 efficiency, 14, 17 flights, 58 method, 14, 15, 17, 18, 81 sites, 16 strategy, 17, 18 success, 14, 17, 81, 98 hygiene, 25 Hylocichla sp., 7 Iberian Peninsula, 75, 86 Iceland, 40 Ictalurus nebulosus, 81 Ictinia mississippiensis, 124 Idaho, 32, 39, 57, 88, 98 Ikeda, Yoshihide, 79 illegal trade, 40 Illinois, 68, 69 immigration, 75, 145 implantation of cannulae, 24 imprinted, 70 inbreeding, 49, 51 Incense Cedar, 116, 122 increase in Griffon Vulture population, 75 increased severity of spring weather, 80 incubating behavior, 71, 75 incubation, 11, 32, 33, 39, 57, 58, 59, 60, 61, 67 duties, 35 male, 12, 33, 34 period, 116 rates, 34 schedules, 34 sharing, 33 individual identification, 45 infertile, 53 infertility, 49 infrared photographs, 27 insect foraging, 125 prey, 32 insectivory, 32 “insect hawking,” 124 insects, 72, 77 insemination, 70 Instructions for Contributors, 41 inter- and intraspecific competition, 68 Interactions, 59, 60, 81, 122 International Symposium on the Diseases of Birds of Prey, 21 international trade, 40 internest distance, 80 inter- specific aggression, 114 confrontations, 118 encounters, 59, 118 interthermal soaring, 148 Winter 1987 Index to Volume 21 169 intraspecific encounters, 61 invertebrates, 17 Ireland, 4 iris color, 9 irrigation, 16, 57 isoline techniques, 141, 145 Italy, 3 Jackdaw, 11 jackrabbit, 85, 87, 89 Japan Alps, 79 Japanese Golden Eagle, 79 Jeffrey Pine, 122 Johnston, David W., 3 Juniperus osteosperma, 39 Juniperus virgimana, 14, 80 juvenile survival rates, 44 juveniles, 80 juxtaposition, 46 Kangaroo Rat, 62 Kangerlussuaq, 111 kelthane, 49 Kentucky, 17, 80 Kenward, Robert, 144 Kenya, 23 ketamine HC1, 152 kleptoparasitism, 32, 119 Komen, Joris laboratory investigations, 22 laboratory mice, 51 Lacerta viridis, 11 Lagopus mutus, 80 Lake Michigan, 69 lamb mortality, 87 production, 86 seasons, 85 Lane, Patricia A., 125 laparoscopy, 49, 50, 52, 55 Lapland Longspur, 112 laser surgery, 24 Lasiurus cinereus, 96 latency, 105 latitude, 35 latitudinal Trends, 3, 4 Laurel Oak, 124 lava beds, 57 law enforcement, 40 LD 50 , 103, 109 leg-hold traps, 86 lemmings, 158, 161, 162 leporids, 46 Lepus arcticus, 80 Lepus californicus, 118 Lepus spp., 85, 87 Leslie Brown Memorial Fund, 1, 83 Lesser Mole Rat, 12 licensing requirements, 100 lichen, 111 lidocaine HC1, 153 life history parameters, 47 limestone, 37 Lincer, J. L., 1 lipids, 79 lithium chloride, 105 live-oaks, 28 livestock, 16 depredation, 88 loafing, 59, 60 Loblolly Pine, 124 Lockhart Method, 89 Logrank Test, 138 London, 21 Long-eared Owl, 100 Long-legged Buzzard, 8 Louisiana, 27, 29 MacLaren, Patricia A., 46 macrohabitat, 45 Madagascar Kestrel, 47 Madagascar Sea Eagle, 47 Madagascar Serpent Eagle, 47 Maine, 81 malaise, 105 males, 60 defense of area, 59 mammalian faunas on islands, 5 mammalian predators, 49 prey, 3, 32 management, 37 Manao, 25 Manitoba, 116 Mann-Whitney U- Tests, 87 manually-operated single-noose System, 163 Marmota flaviventris , 87 marsh, 28, 29, 30 Martha’s Vineyard, 3 Martin, John W., 57 Maryland, 49, 81, 148 Mason Neck State Park, 81 Massachusetts, 3 170 Index to Volume 21 Vol. 21, No. 4 Matchett, Marc R., 85 Mauritius Conservation Project, 25 Mauritius Kestrel, 25 Mayfield Estimates, 137, 138 mean height of nests, 46 shape of nests, 36 medieval cures, 21 median and modal age, 44 Medicago sativa, 57 medical treatment, 21 Meretsky, Vicky J., 140 Mergus merganser, 119 Merlin, 25, 38 mesopic luminances, 156 Mexico, 16 Michigan, 17, 165 microbiology, 24 microclimate, 131 microenvironmental, 36 microhabitat, 100, 150 microlamberts of light, 101 Microtus arvalis, 1 1 Microtus californicus, 96 Microtus montanus, 57 Microtus ochrogaster , 17 Microtus pennsylvanicus , 1 Microtus, sp., 10, 57 Microtus spp., 77 Microtus subarvalis, 11 Middle East, 24 midwest, 74 Ptarmigan, 80 migration, 131, 147 Miller, Frank L., 158 Milvus migrans, 163 Milvus migrans govinda, 24 Milvus migrans parasitus, 23 mine tailings, 57 minimum oral dosage, 104 minimum polygon, 58 Minnesota, 63, 68, 119 Mississippi Kite, 124 Mississippi River, 27, 69 Missouri, 14, 74 Missouri Department of Conservation, 74 Mojave Desert, 35 molted, 131 monoculture farming, 18 Montana, 39, 85, 86, 88, 89, 120 Montane Vole, 57 moon phase, 45 Moore, Jeremy, 111 morbidity, 24 morphology of eggs, 25 of young, 8 mortality, 24, 47, 48, 51, 55, 76, 98 studies, 134, 145 motor function, 104 Mountain Goat, 85 mounted owl, 57 movement patterns, 44, 45 Mule Deer, 61, 85 Mus mu s cuius, 126 Muskoxen, 158 Mustela nivalis, 11 Nagsuggtoq, 111 Natal Bird Club, 46 natal dispersal, 44 National Audubon Society, 86 National Conservation Award, 165 National Wildlife Rehabilitators Association, 83 natural history, 21, 46 Nebraska, 39 Nebraska Brand Birds of Prey Diet, 51 Necropsy, 22, 25, 51, 86, 87 Neotoma fuscipes, 95, 96 nest, 36, 61, 75, 131 adornment, 12 area Fidelity, 123 aspects, 46 box, 50, 53 building, 58, 59 cavity, 77 height, 36, 123 microenvironment, 36 occupation, 27 placement, 44 rebuilding, 30 shape or volume, 36 site, 8, 30, 35, 37, 46, 74 site characterization, 28 site competition, 122, 123 site competition between Ospreys and Canada Geese, 121 site selection, 36, 123 substrates, 46 success, 37 temp., 36 variables, 36 nesting, 74 behavior and chronology, 116 Winter 1987 Index to Volume 21 171 chronology, 58, 116 distribution, 14 failures, 59 habitat, 27 or pair-bonded harriers, 57 season, 16 success, 74 nestlings, 32 deaths, 54 fledged, 49 losses, 49 plumage color, 8 removal, 44 net gun capture technique, 89 neuroanatomical, 156 Neurotrichus gibbsi, 96 New Brunswick, 62 New Mexico, 33, 85, 88 New York, 70, 81 Newton, Ian, 25 niche overlap, 100 night vision goggles, 98, 100 nocturnal mammals, 98 vertebrates, 152 non-foveal rods, 156 Non-releasable Animal Placement Program, 165 non-reproductive studies, 51 nonparametric clustering, 145 tests, 138 nontarget wildlife, 103 noose-halo, 57 North America, 22, 25, 77, 116 north latitude, 4, 34 Northeast Raptor Management Symposium and Workshop, 165 Northern Cardinal, 165 Northern Flicker, 77 Northern Flying Squirrel, 95, 97, 98 Northern Goshawk, 22, 24 Northern Harrier, 32, 57, 62, 72, 119 Northern Pintail, 73 Northern Pygmy-Owl, 77 northwest England, 112 Northwest Territories, 80, 158 Norway, 40 Nuclear Magnetic Resonance, 24 nursing, 21, 22, 23 nutritional requirements, 56 nutritional reserve, 79 Nuttall’s Cottontail, 62 Nyctea scandiaca, 158 Nyssa aquatica, 27 Nyssa sylvatica var. biflora, 124 O’Gara, Bart W., 85 observation of nesting pairs, 29, 30 ocular lesions, 152 Ocypode quadrata, 100 Odocoileus hemionus, 61, 85 Odocoileus virginianus , 69 Oenanthe oenanthe, 112 Ohio, 49, 53, 54, 55 old field, 14, 16 omental, 73 Ontario, 40, 55 ophthalmological operations, 22 Optimal Foraging Theory, 5, 64, 80 Order Odonata, 32 Oreamnos americanus, 85 Oregon, 77, 88, 97 organochlorine, 47 orthopaedic, 22, 23 oscillograph, 153 Osprey, 119, 121 Otus asio, 49, 80 Otus flammeolus, 45, 120 Overcup Oak, 124 overwintering survival, 134 Ovibos moschatus, 158 oviduct, 70 Ovis aries, 85 Ovis canadensis, 85 Ovis dalli, 85 owl weight, 49 pairs, yearling, 55 Palmer, David A., 45 Pampas grass, 57 Pandion haliaetus, 119, 121 Parallel Strip Transects, 87 parasites, 23 Pariah Kite, 24 Parks, John E., 70 Parrish, Jimmie R., 1, 2, 40, 48 Passer domesticus, 15 passerines, 7, 62, 72 pathogens, 23 pathology, 24, 25 Patuxent Wildlife Research Center, 49 payload, 131 Peary Caribou, 158 “pecking bruises,” 86 172 Index to Volume 21 Vol. 21, No. 4 pectoral muscles, 79 Pekliuka, 8 pellets, 95, 125 analysis, 3, 12, 98, 101, 126 collection, 100 and prey remains, 77 Pennsylvania, 68 perch hunting, 15, 18 perching, 27, 28, 29, 30, 58, 81 Peregrine Falcon, 32, 33, 38, 47, 61, 67, 70, 111, 114, 147, 148, 149 eyrie, 33 nest defence, 67 Peregrine Fund, Inc., 70 Peromyscus leucopus, 1 Peromyscus maniculatus, 45, 62, 77, 96 Persimmon, 14 pesticides, 21, 46, 47 Phasianus colchicus, 68, 144 pheasant, 144, 145 Phenacomys intermedins, 11 Phenacomys longicaudis , 95, 96 Philippine Eagle, 25 phosphore, 98 photic stimulator, 153 photoperiod, 34 photopic activity, 152 fusion frequencies, 152 recovery, 156 photoreceptors, 156 Phragmites communis, 57 physical habitat variables, 44 physiognomy, 57 physiographic characteristics, 37 physiological data, 131 physiology, 46 , Picea engelmannii- Abies lasiocarpa, 45 pigeon, 156 pilot study, 144 Pinus edulis, 39 Pinus elliotii, 126 Pinus jeffreyi, 122 Pinus nigra, 8 Pinus ponderosa, 39, 46, 121 Pinus taeda, 124 Pinyon Pine, 39 piracy, 32 pit traps, 89 Pithecophaga jefferyi, 25 Plantanus occidentals , 14 plasma protein, 23 Plectrophenax nivalis, 112 plumage, 58, 99 Pneumonia, 54 Pocket Gopher, 103, 116 Poehlmann, Ruthe J., 103 Poland, 3 polarimeter, 36 political values, 85 Pollock, Ken, 134 Ponderosa Pine, 39, 46, 121 forest, 45 Poole, K. G., 80 population densities, 5 dynamics, 44, 75 increase, 75 stability, 98 turnover rate, 47 Populus tremuloides, 86 post-breeding season, 45 post-brooding, 58, 59, 60, 61 post-fledging behavior, 80 Period, 79 post-mortem examination, 21 Potential Foraging Area (PFA), 79 potential human disturbance, 81 potholes, 37, 44 Prairie Dog, 103 Prairie Falcon, 32, 35, 44, 46, 67, 70, 71 Prairie Vole, 17 prairie-parkland biomes, 39 preamplifier, 153 precipitation, 34, 45 predation, 18, 72, 74, 96, 101, 126, 144 preferences, 97 severity, 86 predator control technique, 85 predator-killed animals, 86 predator-prey interactions, 131, 144 predators, 81 preening, 15, 18, 58 preferential predation, 95 prelaying, 80 prevention, 23 prey abundance, 17 availability, 57, 97, 100, 119, 126, 149 biomass, 150 capture, 16 capture rate, 16 characteristics, 126 deliveries, 58, 59, 60, 79 density, 63, 85 Winter 1987 Index to Volume 21 173 frequency, 96 items, 12, 58, 62, 89 populations, 45 preference, 149 remains, 88 selection, 3 size variation, 95 vulnerability, 17, 63 probability ellipses or circles, 140 Procyon lotor, 69 productivity, 37, 112 propagation, 49 Prunus armeniaca, 12 Pseudotsuga menziesn, 39, 46, 77 public attitudes, 21 Puerto Rico, 17 pursuit speed, 131 Pyrus communis , 12 Q. lyrata, 124 Queen Elizabeth Islands, 158 Quercus alba, 14 Quercus laurifolia, 124 Quercus rubra, 14 Quercus virginiana, 28 rabbit, 86, 156 Raccoon, 69 Radiation Control Officer, 100 radio instrumentation, 89, 90 telemetry, 45, 58, 98, 100 tracking, 61, 91, transmitter, 58, 86 triangulated positions, 58 instrumentation, 90 radio-tagging, 59 and monitoring, 80 Golden Eagle, 68 radioactive gas, 100 radiography, 24 radiotelemetric, 63 radiotelemetric and visual monitoring, 57 radiotherapy, 24 Raim, Arlo J., 68 ranching practices, 85 range, 45 lambing, 86 size, 144 Rangifer tarandus pearyi, 158 rank correlation, 149 raptor anatomy, 21 behavior, 103 biologists, 21, 24 care, 21 conservation, 47 diseases, 21, 25 organization Directory, 1 rehabilitation, 21 temperate zone, 47 Raptor Research, 2 Raptor Research News, 2 Raptor Rehabilitation and Propagation Project, Inc., 74 rate of growth, 79 Rattus norvegicus, 1 raven, 80, 114 recapture rates, 145 reclassification, 40 Red Tree Voles, 95 Red-backed Voles, 45 Red-tailed Hawk, 16, 37, 46, 60, 67, 103, 118 Red-winged Blackbird, 62 Redpoll, 112 rehabilitation, 22, 23, 25 reintroduction, 47 relative humidity, 55 release, 21, 22, 23 relief schedule, 34 reproduction, 37, 49 data, 50 failure, 122, 123 history, 54 success, 44, 49, 54, 75 reptiles, 72 resource partitioning, 46 resting, 58 retina functions, 152 retinae, 156 reversal learning, 107 reviewers, 44 Rhode Island, 3 Richardson’s Ground Squirrel, 87 Ring-necked Pheasant, 68 ringing recoveries, 47, 76 riparian, 57 habitat, 61, 62, 63 Robima pseudoacacia, 14 Rock Ptarmigan, 80 rodent vulnerability, 16 rodenticide, 103 rodent, 17 Rook, 119 174 Index to Volume 21 Vol. 21, No. 4 roost locations, 69 site selection, 45 sites, 80 trees, 81 roosting, 79, 81, 99, 116, 131 Rosa sp., 12 Rough-legged Hawk, 16 Runde, Douglas E., 44 Russian Thistle, 57 S. aluco, 100 Sabine, Neil, 118 sagebrush, 57, 86 sagebrush steppe habitat, 32 Sailer, James E., 38 Salix sp., 8, 57 Salmo sp., 150 salmon, 150 Salmonella mycobacterium , 23 Salsola kali, 57 sampling intensity, 134 sandstone cliffs, 37 Sarcobatus vermiculatus , 57 satellite use, 148 Saw- Whet Owl, 45 Scapanus latimanus , 96 Scapanus orarius, 97 scapular, 73 scarecrow, 85, 86, 87, 90 scavengers, 32 scavenging, 72 Scirpus sp., 57 Sciurids, 46 Sciurus carolinensis , 145 scotopic fusion frequency, 152 scotopic recovery, 156 seabirds, 24 search strategies, 68 second clutches, 49, 54, 55 secondary poisoning, 103 segregation of sexes, 126 Septicemia, 54 settle plates, 23 sex determination, 51 ratio, 50 sexual dimorphism, 55 maturity, 55 segregation, 126 Shadscale, 57 Shagbark Hickory, 80 Sharp-shinned Hawk, 165 sheep depredation, 88 shelter box, 50 Sheppey Isle, 3 shrew, 77 shrub-steppe habitat, 57, 61, 62 Sibling Vole, 11 siblings, 51 Sierra Nevada, 116 signal variations, 147 single-flash stimuli, 152, 153 site tenacity, 30 Slash Pine, 126 Slivnitza, 8 Smallwood, J. A., 1 Smallwood, John A., 127 Smith, Randall A., 116 Smith, Scott A., 32 smoke tests, 23 Snake River Birds of Prey Area, 32, 57 Snow Bunting, 112 Snowy Owl, 158 soaring, 15, 28, 30, 58, 59, 60 sonagraphic identification, 45 song posts, 46 Sooty Falcon, 47 Sorex spp., 77 Sorex trowbndgei, 96 South Africa, 46, 79 South Africa Agricultural Union, 48 South Carolina, 3, 81, 124 Spain, 3, 75, 76 Spalax leucodon, 12 Spanish Moss, 124 Spanish Ornithological Society, 75 Spanish Pyrenees, 75 Sparrowhawk, 47 sparrows, 77 species richness, 5 sperm fertility, 71 Spermophilus parryii, 80 Spermophilus richardsonii, 87 Spermophilus spp., 103 Sphyrapicus thyroideus, 77 Spotted Owl, 95 spring diets, 95 Spruce-fir, 45 St. Chrysostom, 24 standard metabolic rate, 15 Steenhof, K., 1 Stephen R. Tully Memorial Grant, 1, 82 stick nest, 37 Winter 1987 Index to Volume 21 175 still hunting, 17 Stinging Nettle, 57 stretching, 15 strike characteristics, 15 strip transect, 158 Strix aluco, 156 Strix nebulosa, 116 Strix occidentalism 95 Strix uralensis, 100 Strychnine, 103 Students £-Tests, 87 Sturnus vulgaris, 17 Subalpine Fir, 98 subcutaneous fat, 73 hemorrhage, 86 Sugar Beet, 57 surgery, 22 surgical treatment, 21 survival, 44, 131, 138 breeding adults, 44 Swainson’s Hawk, 16 swamp, 28, 29, 30 Swamp Tupelo, 124 Sweden, 40, 116, 117, 144 swifts, 32 Switch Cane, 124 Sylvilagus bachmani, 96 Sylvilagus floridanus, 81 Sylvilagus nuttallii, 62 sympatric species, 100 taiga, 39 tail mounts, 131 tail-band count, 38 tail-chasing, 17 tail-mounted tags, 145 talon wounds, 86 Tamarisk, 57 Tamarix pentandra, 57 taste aversion, 103, 105, 109 Tawny Owl, 100, 156 Taxidea taxus , 145 Taxodium distichum, 27, 126 techniques, 57 teflon ribbon, 131 telemetry, 23, 131, 144 supplies, 132 temp, 34, 45 Terathopius ecaudatus , 48 territories, 59, 111, 119 Texas, 16, 72, 85 “The Kettle,” 1 therapeutic aids, 21 procedures, 24 therapy, 21, 22 thermal enertia, 36 thermoregulation, 15 thiamine, 51 Thomomys bottae, 96, 97, 116 Thomomys spp., 103 Tillandsia usneoides, 124 time-lapse photography, 33, 117 time-series approach, 135 Toland, Brian R., 14 topography, 46, 59, 85, 86, 147 Tordon, 50 toxicology, 25 translocation, 85, 86 transmitter, 131 weight, 131 powered by solar cells, 131 Transvaal, 79 trapped, 85 trapping, 88 and translocation, 89 trauma, 49, 51 treatment, 22 triangulation, 100 accuracy, 149 Trichomoniasis, 25 Triticum aestivum , 57 Tritium, 98 trophic diversity, 4 tropical and sub-tropical, 47 Turkey Vulture, 67 turnover, 44 turtles, 81 two-way radios, 58 Tyto alba, 3, 25, 49, 74 U.S. Army, 119 U.S. Fish & Wildlife Serv., 15, 68, 86 U.S. Geological Survey, 14 U.S. Geological Survey Topographical maps, 27 Ulmus americana, 14 ultrasound, 24 umbilical hemorrhage, 54 unattended, 34 United States, 22, 70 unoccupied nest, 27 unsuccessful nests, 37 Upper Sonoran, 57 176 Index to Volume 21 Vol. 21, No. 4 Ural Owls, 100 Urtica sp., 57 USDA Forest Service, 77 USFWS, 74 USFWS Bands, 58 USGS Topographic maps, 57 Utah, 38, 118, 131 Utah Juniper, 39 Vander Wall, Stephen B., 103 variance coefficient, 145 variance estimates, 87 Vatev, Iliya Ts., 8 vegetation density, 14, 17 height, 14, 17 structure, 100 vegetative cover, 14, 85 ventriculus, 54 vertebrate prey, 8, 17 “Veterinary Aspects of Captive Birds of Prey,” 21 Vionate, 51 Virginia, 81 visceral, 73 visceral gout, 51 vocalizations, 12, 70 vocalized, 67 vole, 57, 62, 77 vulnerability, 57, 80 vultures, 48 Wasatch Mountains, 39 Washington, 63, 150 water habitats, 57 Water Rat, 12 waterfowl, 72 waterfowl traps, 73 weasel, 1 1 weather, 30, 85 weight- specific doses, 106 weighted, 58 weight, 52, 145 and length of primary feathers, 58 loss, 99 weights of birds, 52 Western Raptor Management Symposium and Workshop, 82 Western Whiptail, 57, 62, 63 wheat, 57 Wheatear, 112 Whistling Kite, 163, 164 White, Clayton M., 1, 40 White Fir, 39 White Oak, 14 White-bellied Sea-eagle, 162, 164 White-tailed Deer, 69 Wiemeyer, Stanley N., 49 Wildlife and Countryside Act of 1981, 21 Wildlife Collectibles newspaper, 40 wildlife rehabilitators, 22 William C. Andersen Award, 83 Williams, L., 1 Williamson’s Sapsucker, 77 willow, 57 wind, 45 wing length, 79 winter habitats, 16 mortality, 74 territoriality, 126 Wisconsin, 68, 69 Wood, Kristin W., 32 Wood Pigeon, 144 Wood Rat, 4 woodland, 17 Wright, Philip L., 120 Wyoming 36, 39, 44, 85 Wyoming Ground Squirrel, 46 xeric, 63 Yagi or Adcock Antenna, 58 Yellowbelly Marmot, 87 Yosemite National Park, 117 Yukon River, 33 Z statistic, 86 Zimbabwe, 47 Zion National Park, 67 “Zoo and Wild Animal Medicine,” 21 zoological gardens, 48 Zwank, Phillip J., 27