HARVARD UNIVERSITY Library of the Museum of Comparative Zoology »EAT BASIN NATURALIST MEMaR number 8 Brigham Young University 1« The Black-footed Ferret MCZ LIBRARY AUG 2 1986 HARVARD UNIVERSITY izDSot^ i^y f^c-rt-r^R^ GREAT BASIN NATURALIST Editor. Stephen L. Wood, Department of Zoology, 290 Life Science Museum, Brigham Young University, Provo, Utah 84602. Editorial Board. Kimball T. Harper, Chairman, Botany; James R. Barnes, Zoology; Hal L. Black, Zoology; Stanley L. Welsh, Botany; Clayton M. White, Zoology. All are at Brigham Young University, Provo, Utah 84602. Ex Officio Editorial Board Members. Bruce N. Smith, Dean, College of Biological and Agricultural Sciences; Norman A. Darais, University Editor, University Publications. Subject Area Associate Editors. Dr. Noel H. Holmgren, New York Botanical Garden, Bronx, New York 10458 (Plant Taxonomy). Dr. James A. MacMahon, Utah State University, Department of Biology, UMC 53, Logan, Utah 84322 (Vertebrate Zoology). Dr. G. Wayne Minshall, Department of Biology, Idaho State University, Pocatello, Idaho 83201 (Aquatic Biology). Dr. Ned K. Johnson, Museum of Vertebrate Zoology and Department of Zoology, University of California, Berkeley, California 94720 (Ornithology). Dr. E. Philip Pister, Associate Fishery Biologist, California Department of Fish and Game, 407 West Line Street, Bishop, California 93514 (Fish Biology). Dr. Wayne N. Mathis, Chairman, Department of Entomology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560 (Entomology). Dr. Theodore W. Weaver III, Department of Botany, Montana State University, Boze- man, Montana 59715 (Plant Ecology). The Great Basin Naturalist was founded in 1939 and has been published from one to four times a year since then by Brigham Young University. Previously unpublished manuscripts in English of fewer than 100 printed pages in length and pertaining to the biological natural history of western North America are accepted. Western North America is considered to be west of the Mississippi River from Alaska to Panama. The Great Basin Naturalist Memoirs was established in 1976 for scholarly works in biological natural history longer than can be accom- modated in the parent publication. The Memoirs appears irregularly and bears no geographical restriction in subject matter. Manuscripts are subject to the approval of the editor. Subscriptions. The annual subscription to the Great Basin Naturalist for private individuals is $16; for institutions, $24 (outside the United States, $18 and $26); and for student subscrip- tions, $10. The price of single issues is $6 each. All back issues are in print and are available for sale. All matters pertaining to subscriptions, back issues, or other business should be directed to Brigham Young University, Great Basin Naturalist , 290 Life Science Museum, Provo, Utah 84602. The Great Basin Naturalist Memoirs may be purchased from the same office at the rate indicated on the inside of the back cover of either journal. Scholarly Exchanges. Libraries or other organizations interested in obtaining either journal through a continuing exchange of scholarly publications should contact the Brigham Young University Exchange Librarian, Harold B. Lee Library, Provo, Utah 84602. Manuscripts. See Notice to Contributors on the inside back cover. .5-S6 15(X) 2.-?9.54 ISSN 017-3614 5REAT BASIN NATURALIST MEMOF The Black-footed Ferret -•"•-Au'lv/V? ilTDSONi ilV fH'^-HT&R, CONTENTS Introduction. Archie Carr III 1 Technical introduction. Tim W. Clark 8 Paleobiology, biogeography, and systematics of the black-footed ferret, Mustela nigripes (Audubon and Bachman), 1851. Elaine Anderson, Steven C. Forrest, Tim W. Clark, and Louise Richardson 11 Historic status of black-footed ferret habitat in Montana. Dennis L. Flath and Tim W. Clark 63 Description and history of the Meeteetse black-footed ferret environment. Tim W. Clark, Steven C. Forrest, Louise Richardson, Denise E. Casey, and Thomas M. Campbell III 72 Vegetation inventory of current and historic black-footed ferret habitat in Wyoming. Ellen I. Collins and Robert W. Lichvar 85 Comparison of capture-recapture and visual count indices of prairie dog densities in black-footed ferret habitat. Kathleen A. Fagerstone and Dean E. Biggins 94 Habitat suitability index model for the black-footed ferret: a method to locate transplant sites. B. R. Houston, Tim W. Clark, and S. C. Minta 99 Descriptive ethology and activity patterns of black-footed ferrets. Tim W. Clark, Louise Richardson, Steven C. Forrest, Denise E. Casey, and Thomas M. Campbell III ... . 115 Activity ofradio-tagged black-footed ferrets. DeanE. Biggins, MaxH. Schroeder, Steven C. Forrest, and Louise Richardson 135 Fecal bile acids of black-footed ferrets. Mark K. Johnson, Tim W. Clark, Max H. Schroeder, and Louise Richardson 141 Estimating genetic variation in the black-footed ferret — a first attempt. C. WiUiam Kilpatrick, Steven C. Forrest, and Tim W. Clark 145 Determining minimum population size for recovery of the black-footed ferret. Craig R. Groves and Tim W. Clark 150 Some guidelines for management of the black-footed ferret. Tim W. Clark 160 Black-footed ferret recovery: a discussion of some options and considerations. Louise Richardson, Tim W. Clark, Steven C. Forrest, and Thomas M. Campbell III 169 Annotated bibliography of the black-footed ferret. Denise E. Casey, John DuWaldt, and Tim W. Clark 1^^ t^! Great Basin Naturalist Memoirs The Black-footed Ferret No. 8 Brigham Young University, Provo, Utah 1986 INTRODUCTION Archie Carr III' Abstract. — Context for the Meeteetse, Wyoming, black-footed ferret studies and recovery efforts, reported in this volume, is presented. This is the second draft of this manuscript. My first draft was ready in the early summer of 1985. It conveyed a sense of confidence about the survival prospects of the black-footed fer- ret, Mustela nigripes . By the fall of the year at press time, circumstances had changed so dramatically that draft 1 became obsolete and the editor asked for a rewrite. I submit draft 2 with dismay and no sense of what the future holds. As for the ferrets, 6 are in captivity in a place called Sybille Canyon in Wyoming; per- haps 10 are still out there near Meeteetse, in a sprawling prairie dog colony in the Big Horn Basin of northwestern Wyoming; canine dis- temper has swept the ferrets; plague has been confirmed in the prairie dogs; and options for management have dwindled like Custer's Last Stand. As a result, I write from unsure footing. The series of papers contained herein were in- tended to report on the natural history and management criteria of a critically endan- gered species, the black-footed ferret, and to highlight data that might contribute to its re- covery. The research, and publication here of the results, seemed to be two steps in an orderly, modern, even scientific, recovery process. This introduction addresses that "process," endeavoring to record the perplex- ing political events that occurred concur- rently with the field research following the ferret's rediscovery in 1981; events that by late 1985 have left an unnecessarily critical and unpromising survival outlook for the spe- cies; events that may have made an epitaph of this monograph. The ferret is a legal animal. By virtue of the U.S. Endangered Species Act the ferret en- joys something akin to "standing." It cannot be legally abused or ignored. That is the beauty, the novelty, of the act. Thus, when a species enjoying that novel standing is seen to decline, one must question the fitness of the act itself. That is not to say that one can take the ESA for granted as a tool for species sur- vival. On the contrary, this act, like other federal laws, has always taught a strict lesson in civics: democracy is what you make of it. No major social advance since World War II has intrigued me more than the Endangered Species Act of 1966. I was young when the act was passed. My generation was idealistic in those days. The act restored the conservation- ist's faith in America as a rational society and in government as a body capable of respond- ing to the will of the people. The Endangered Species Act, coming as it did after the orgy of materialism during the previous decade, af- firmed that the general public was ready to acknowledge that, aside from tangible wealth, prosperity meant preserving the natural her- itage of the country. Government would be the willing vehicle of this philosophical renais- sance. 'Wildlife Conservation International, New York Zoological Swiety, Bronx, New York, 10460. Great Basin Naturalist Memoirs No. 8 The act has matured since then, and so have I. When the act was passed I supposed that species were saved and that I could devote my own energies to other matters. In the inter- vening years, years of tremendous social chal- lenges in the U.S., ranging from civil rights to Vietnam, I gradually gained insight on a con- cept I came to call "ghost government." The reality is that no matter how fine a govern- mental system may be, the citizen should ex- pect little from it if he or she is unwilling to be involved in its operation — forever. Ghost gov- ernance. As in ghost writing. A governmental system is only as successful as the ghost gov- ernment of citizens that watches it and cod- dles it and intimidates it and pats it on the back. It would seem to be an inefficient pro- cess, but it works, and the crucial thing about a democracy is that it permits this ghost gov- ernment, this band of sometimes unruly in- terest groups, to participate freely in the stew- ardship of society — and black-footed ferrets. Plainly my thesis is that a congressional mandate to a government agency has been an inadequate incentive to save the ferret. The mandate to save this animal is clearly there in the Endangered Species Act. But implement- ing that mandate has required a frustrating but persistent give-and-take between citizens and the mandated agencies. Ghosting. If the hand-holding role of the citizenry was at all unusual in 1981 in mobilizing govern- ment to respond to the ferret's plight, it was made necessary, in part, by the unusual times of the ferret. One must recall that immedi- ately before the ferret's rediscovery came the appointment of James Watt as Secretary of the Interior. The anti-act posture of this secretary demoralized federal officers and damaged budgets to such an extent that no real govern- ment rally to help the ferret ever arose. The ferret was given no significant "priority. " The ghost government responded in two ways: first, it tried to foster its own rally, and I'll discuss that more below. Secondly, it "dumped " James Watt. Democracy is what you make of it. With the exception of James Watt, the indi- viduals involved in the early acts of the ferret's saga were more captivating than the agencies. That's perhaps the way it is most often with "issues " in American life. Individuals make things happen. One figure who intrigues me in particular is Mr. Jack Turnell, manager of the Pitchfork Ranch, where most of the rem- nants of the only known ferret colony existed. I have yet to meet Mr. Turnell, but I think he is a national hero. He made certain decisions shortly after the ferrets were found that, frankly, assured them a chance for recovery. My guess is that Turnell made these decisions out of some personal conviction about wildlife and the West and humanity's obligations to the world we live in. His direct interest in the ferret was crucial because, as I have said, governmental authorities were not moving with alacrity at the time. So, Mr. Turnell fasci- nates me. Someday perhaps I'll have the chance to talk to him about those early days, and the sacrifices to his land and privacy that he perceived to be concomitant with helping a federally protected species. There are few people in history who have had the opportu- nity and the power to unilaterally decide to save a species. A second prominent figure in modern ferret lore is Dr. Tim Clark who, it is fair to say, owes much of his notoriety to Jack Turnell. One of the crucial decisions Turnell made was to agree to let Tim Clark look for and study the ferrets on the ranch. I have had the privilege of knowing Dr. Clark. He is not a boastful fellow, but surely he deserves no less a title than "Mr. Ferret. " In support of the claim I need only refer to the authorship of many of the manuscripts that follow. Tim's interests know no bounds when it comes to black- footed ferrets. He has done the hard field research; he has publicized the plight of the ferret; he has lobbied; and he has raised money. In keeping with his predilection for holistic research, Tim also took an academic interest in the dynamics of American conservation, as represented by the unfolding story of the fer- ret. My own organization. Wildlife Conserva- tion International, was scrutinized in this re- gard. Wildlife Conservation International (WiCI), the division of the New York Zoological Soci- ety (NYZS) that concerns itself with field con- servation, is chiefly devoted to work outside the United States, mostly in the tropics. This emphasis is based on the observation that the United States is amply endowed with conser- vation agencies and conservation money, es- 1986 Carr: Introduction Fig. 1. This 1906 photograph is beheved to be the first one taken of a black-footed ferret (New York Zoological Society photo). pecially when compared to the species-rich countries of the developing world. In the United States, an endangered species might expect attention from layers of interested parties: the federal government, state govern- ment, private nongovernmental organizations including universities, and, of course, indi- viduals. Such infrastructure is rarely present in the Third World, and so WiCI concentrates its efforts there, conducting and supporting research on the biology of endangered spe- cies. We call it conservation biology, and, at any given moment, we will have 30 or so projects underway. However, since the founding of NYZS in 1895, the society has never entirely divorced itself from species conservation in the United States. In the early part of this century, the vociferous contributions of William Horna- day, the first director of NYZS, to shaping the U.S. Fish and Wildlife Service, its refuge sys- tem, and the early laws entrusted to it suc- ceeded in leaving a permanent NYZS imprint on American wildlife conservation. The soci- ety also takes considerable pride in having played a central role in restoring the Ameri- can bison to the western plains between 1905 and 1919. In conjunction with the federal gov- ernment, remnant groups of bison were gath- ered at the society's Bronx Zoo, in New York. Stocks from the combined herd were sent by rail to such protected areas as the Wichita Mountains Wildlife Refuge. Among the studies sponsored by NYZS in this country is one of particular relevance to the present monograph. It is Carl Koford's work with prairie dogs, appearing in 1958 as "Prairie Dogs, White Faces and Blue Grama" in the journal Wildlife Monographs . Koford's was the first major technical paper to show a tie between prairie dog eradication and ferret decline. It was a deadly tie indeed. The society was helpful to Koford's pre- scient research and now finds itself back in the West, again with the ferrets, this time pro- moting science appropriate to recovery. De- spite our current commitment to conservation biology abroad, the society's affection for wildlife of the West is clear. In fact the attach- ment is symbolized in the logo of NYZS, a bust of the bighorn sheep. In the fall of 1981 the U. S. Fish and Wildlife Service's Endangered Species Technical Bul- letin announced the discovery of a black- footed ferret colony near Meeteetse, Wyo- ming. It was electrifying news, and a host of American conservation groups perked up, looking for ways to lend a hand. Just about every one had accepted the USFWS bitter decision three years prior to consider the fer- Great Basin Naturalist Memoirs No. ret extinct. Everyone, that is, except for a very few individuals who kept searching throughout those bleak years. A shaggy ranch dog turned things around. It's true, the dog killed the only ferret anyone had seen in years, but the single specimen, probably a wayfaring yearling from the colony, was tangible evidence that a whole species still lived. I confess, I don't mind defending the dog. I met him once. His name is Shep. He is very tractable, and blithely unconcerned with the hoopla stirred up by his routine vigilance. Following the Bulletin's report of the ferret find, Wildlife Conservation International made the decision to become involved in the species' recovery. We were influenced by three considerations: 1. Without doubt — and without apology — we saw public-relations value in taking a lead- ership role in the potential restoration of this highly publicized American species. A good job with the little ferret would help us in the chronic task of raising funds for other species. The ferret might have become a mini-panda, valuable to our image making. And so, we dominated the private funding picture from 1982 to 1985. 2. Secondly, there was the political situa- tion, to which I have alluded before. In the view of most conservationists in late 1981, the Endangered Species Act was in jeopardy, and consequently the ability of the federal govern- ment to respond constructively to the ferret find was predicted to be limited. The times were chaotic for wildlife conservation. Al- ready that autumn I had joined a letter-writ- ing campaign to halt dismantling of USFWS Cooperative Wildlife Research Units at uni- versities all over the country. The Endan- gered Species Office budget had been slashed. The secretary of interior had de- clared that his department would list no more endangered species, just as it would gazette no new national parks. It was a tough time to arise from the dust of extinction, and we at WiCI felt that if we didn't make a move to help the ferret, the little beast might actually slip back into oblivion. It's expected savior. Uncle Sam, was hobbled by an anomalous secretary of the interior. 3) Our third motivation for entering ferret history was a practical one. After reading the first reports that the ferret colony might con- sist of a couple of dozen breeding animals, we were very certain that captive breeding and establishment of new colonies would be rec- ommended. That form of animal management has attained a high degree of sophistication at the Bronx Zoo, the sister organization to WiCI in the New York Zoological Society. The cadre of NYZS people involved in captive breeding of wildlife, from curators to veterinarians, is large and skilled, and we planned to make it clear to all concerned that we were ready to contribute when the time came. With these circumstances in mind, we set about to find an outlet for our good will, tal- ent, and cash. At precisely the same moment, one Dr. Tim Clark began inquiring of possible WiCI interest in granting support for his fer- ret studies. His field work — counts, feeding behavior, reproduction studies — were pre- cisely the type of biology favored by WiCI, and his commitment to working in conjunc- tion with the complex federal-state mecha- nism reassured us that our sponsorship would go toward an influential project. We began work with Tim Clark and his Biota Research and Consulting, Inc. in 1982. Largely through Dr. Clark's initiatives, the ferrets attracted the attention of numerous other conservation organizations, most of whom assisted Clark's project directly. These included the World Wildlife Fund— U. S. , the National Geographic Society, the National Wildlife Federation, and the Charles A. Lind- bergh Fund. Aside from contributing cash, several of these prominent nongovernmental organizations assumed lobbying tasks in Washington in support of the ferret. But Dr. Clark's first support — given even before the ferrets were discovered, and sustained, one presumes, out of blind faith that some animals must have remained somewhere in the vast- ness of the West — came from the little-known Wildlife Preservation Trust International (WPTI). WPTI is an American-based offshoot of the Jersey Wildlife Preservation Trust, an institution given prominence by Gerald Dur- rell, director of the famous Jersev Zoo in Eng- land. A discernible recovery program began to take shape in Wyoming. The one known colony was secured, thanks in large part to the unusual cooperation of the owners of the only 1986 Carr: Introduction inhabited ferret land. Research was begun promptly and was pursued with vigor, to the extent that, as the papers contained herein and others reveal, we quickly learned the size of the single colony, its demographics, and that possibly "surplus" youngsters were avail- able every fall as potential candidates for cap- tive breeding or translocation. We learned how to search for ferrets, and, tragically, that over tremendous areas of potential habitat there were no more ferrets. The American conservation community rallied effectively to underwrite the bulk of the research to the tune, cumulatively, of over $550,000, accord- ing to a recent manuscript by Tim Clark. In the final analysis, cooperation in the field be- tween government and nongovernmental agencies was satisfactory. To help enhance this atmosphere of cooperation, the Black- footed Ferret Advisory Team (BFAT) was put together, a sort of clearing house for the grow- ing interest in black-footed ferrets. I became optimistic. I thought I sensed a surge of enthusiasm among ferret people, a threshold of determination that, once crossed, would overwhelm whatever obsta- cles might be thrown up by the Watt adminis- tration. In April 1982 I saw a ferret and was inspired even more. I flew out to Cody, Wyoming, with Jim Doherty, the seasoned curator of mammals for the New York Zoological Soci- ety. I was anxious for Jim to accompany me because already we were certain that captive breeding of ferrets would become a recovery priority, and Jim could represent the society's expertise in this field. We drove south to Meeteetse and joined Tim Clark and his research associates Tom Campbell, Louise Richardson, and Steve For- rest. They were the principal figures in the field program and they introduced us to the research. Later I wrote about the outing in our newsletter, the Ferret, first published shortly after WiCI joined forces with Tim and his colleagues: After a day with Tim Clark, exploring the prairie dog colony where the ferrets clung to their tenuous fu- ture, . . . Jim Doherty and I joined ferret biologist Tom Campbell for a unique adventure. Driving in a pickup truck along a graded road near the prairie dogs in the dead of night, we saw a black-footed ferret. We were lucky. Only nine individuals had been found by spotlighting since Clark and Campbell had begun their surveys back before Christmas. Our ferret came bounding across the prairie in its odd, accelerated inch-worm gait and wound up in a prairie dog den twenty feet from the right fender of the truck. We feasted on the view for many excited min- utes. The ferret was as high strung and energetic a creature as I had ever seen. It fairly crackled with nervous im- pulses, first digging, then stretching to stare, then cir- cling the den, then looping back in. I was moved by the idea that if we humans would give the ferret half a chance, that purposeful dynamo would surely do the rest. It was a nice sentiment at the time. It seems naive now, because that "half a chance" was never granted. Time passed. The field work continued. Searches for ferrets were begun in other states. Litters of ferrets were recorded at Meeteetse. Data were published. Letters were written. No progress was made toward captive breeding during 1982 and 1983. Finally, at the request of the nongovern- mental conservation community, a meeting was called by the Wyoming Game and Fish Department in the spring of 1984 in Cheyenne. Jim Doherty and I were invited to participate. The meeting would include field biologists, veterinarians, and administrators representing federal, state, and private agen- cies, essentially the extended network of peo- ple responsible for the survival of the black- footed ferret. Sure enough, captive management became the focus of the meeting just as soon as the in- troductory material was set aside. Tim Clark and his colleagues presented enough demographic data to suggest that the Meeteetse ferret colony was stable or even growing. Arriving at compara- ble figures from year to year is difficult because census methods were evolving and improving as time went by, but the best published estimates for all years, based on early August counts of adults and young, are as follows: 1982 61 (incomplete survey) 1983 88 After this meeting: 1984 129 1985 58 The 1983 figure and the abundance of youngsters every fall relative to the number of adults, were strong indications that ferrets Great Basin Naturalist Memoirs No. 8 could be captured without jeopardizing the Meeteetse colony. Thinking back to that large gathering in Cheyenne, I recall a universal consensus that establishment of one or more captive colonies was of utmost urgency. The chief justifications were (a) to provide a strate- gic cushion in the event a disease — an epi- zootic — struck the little Meeteetse popula- tion and (b) to provide, in the course of time, the stock for recolonization of suitable ferret habitat. It was sound, if belated, reasoning. The only dissention came in deciding how to doit. As early as 1981 the U.S. Fish and Wildlife Service had granted Wyoming Game and Fish Department "lead agency" status for ferret recovery, a legal courtesy permitted under the Endangered Species Act. Thus, Wyoming Game and Fish had begun to organize activi- ties, helping foster BFAT, convening the Cheyenne meeting, and generally assuming responsibility for major decisions. Assump- tion of leadership by a state agency in this manner had precedent elsewhere; and, in cases where the federally protected species is limited in distribution, it seems a logical way to implement the act. Provided the surrogate agency responds to the federal mandate, the process is viable. At Cheyenne we began to see the hang-up on captive breeding as an element in the sur- vival process. State officials, while concurring with the captive propagation tactic, an- nounced firmly that no ferrets would leave Wyoming to achieve this purpose. Simulta- neously they declared that their own Sybille Canyon Wildlife Research Unit was unsatis- factory as a captive breeding facility, an ironic viewpoint as things turned out; and they con- cluded that federal and/or private agencies should pay for the cost of building and staffing a proper facility in Wyoming. In view of the availability of well-equipped, well-staffed, well-funded facilities in several locations around the U.S., this pronounce- ment by the lead agency for ferret recovery was met with consternation by both federal and private nonprofit organization represen- tatives. The 1984 capture season (September- October, when young of the year are weaned and dispersing) came and went, but the Cheyenne impasse prevailed despite the probabilistic certainty of the consequences. In May of 1985 a decision was made by state and federal officers to attempt to capture fer- rets in October, provided the scheduled sum- mer counts showed an acceptable but un- specified surplus. Sybille Canyon was agreed upon as a holding facility, but no specific breeding facility was identified. Almost con- currently with the meeting, plague was re- ported among the white- tailed prairie dogs of Meeteetse, the prey base of the ferrets. To everyone's relief, the mustelids, evidently, were immune to plague, but there loomed the possibility of starvation for ferrets if the prairie dog die-back was too severe. As it turned out, the plague episode served chiefly as an un- nerving object lesson of the principles of epi- zootic disease, principles that were familiar to most of us from the beginning. During the period June-October 1985, the principles were applying themselves with mortal vigor. The July-August count gave strong indications that something was amiss, but no real credence was given the declining population figures until 22 October. By that time supplementary surveys in September had arrived at a count of 31 ferrets, one month after the August estimate of 58, and by Octo- ber 9 only 13 ferrets were seen in the field. Six ferrets had been captured by early Octo- ber and brought to the Sybille Canyon Wildlife Research Unit. On 22 October one of these animals was reported dead and the cause was diagnosed as canine distemper. Wyoming Game and Fish acknowledged that the disease was "probably the worst event that could have occurred in the ferret population." Immediately a capture team was sent to the field to capture as many of the threatened remaining ferrets as possible. Six were brought in by the following week when the capture term was withdrawn before capturing all the ferrets. Biologists departing the scene after the emergency exercise guessed that fewer than 10 remained in the Meeteetse pop- ulation. Their significance to the future of the species must be regarded as negligible for the time being. Their numbers are few; they are scattered over a vast terrain; distemper is pre- sumably still among them; and the Wyoming winter is coming on. Now, after additional deaths in the captive group, six ferrets remain. The Sybille Six. There is no cushion. For a while the best of 1986 Carr: Introduction American wildlife science might have gov- erned the future of this species. Now luck is the guiding force. We need luck with the Sybille Six, that they might multiply; and we need luck out on the prairies, that some stal- wart surveyor might chance upon yet another last colony of black-footed ferrets. The black-footed ferret once enjoyed a range about as extensive as any that North America can offer, encompassing all of what we call the Great Plains and beyond. The little mustelid was the incidental victim of one of the most diligent vertebrate pest control exer- cises in history: the attempt to eradicate prairie dogs for the alleged benefit of livestock grazing. The assault changed prairie dog dis- tribution dramatically. In the process it wiped out ferrets from Canada to Mexico — except for the few discovered near Meeteetse. History should record that rational people stepped forward when the Meeteetse colony was found. Among them were the authors of the papers that follow, people who assumed that they worked within a rational system, far different from the cavalier times that brought the ferret so near extinction in the first place. But that system, ultimately based in the U.S. Endangered Species Act, has failed the ferret. It has converted a tense but hopeful outlook for the species into a crisis. The system be- came impotent as decision makers locked themselves into years of indecision as to the venue for captive propagation of ferrets. Altogether, the species has not fared well in its ecological partnership with modern man. But in every such sad story there is a lesson. The ferret story may contain two. Following its first decline, we people reviewed our use of pesticides, our fanatical reaction to agricul- tural "pests," our obligation to public lands; and our general management of Great Plains land, whether private or public. I believe the message of the ferret's second decrement is that the U.S. Endangered Species Act may no longer be the safety net for American wildlife that Congress intended it to be. TECHNICAL INTRODUCTION Tim W. Clark' Abstract. — The contents of this vokime and their relationship to ferret conservation and recovery are discussed. The critically endangered black-footed fer- ret (Mustela nigripes) has been an enigma ever since its scientific discovery in 1851 by John James Audubon and John Bachman. In 1877 Dr. Elliott Coues of the Smithsonian Institution reported that the ferret was com- mon to the plains of the West and associated with prairie dogs (Cynomys sp.). Collection records show that, until the first decades of this century, ferrets were distributed over about 40 million ha in 12 states and 2 Canadian provinces. By the late 1940s, no ferrets could be located for study, ostensibly because of a precipitous decline in population size and dis- tribution, from habitat loss (the poisoning of prairie dogs), and perhaps other factors. The ferret was considered extinct or nearly so when in 1964 a small population was discov- ered in South Dakota and the species came under study for the first time — 113 years after its scientific discovery. Three chapters stand out in ferret study and conservation: (1) From the 1830s to 1964, dur- ing which time specimens were occasionally collected and a few natural history observa- tions recorded; (2) from 1964 to 1981, when the Mellette County, South Dakota, ferret population was discovered and studied for 11 years (ca 90 different ferrets were observed, including 11 litters) and it dwindled to extinc- tion. Attempts to breed a few ferrets in captiv- ity came too late, and no other populations were discovered despite surveys. Many peo- ple feared the ferret was extinct; (3) from 1981 to date, when the Meeteetse, Wyoming, fer- ret population was discovered and studied (ca 129 different ferrets were seen, including 25 litters in 1984). In part, this period closes with this volume and the many other conservation biology papers resulting from the Meeteetse studies (see Casey et al. in this monograph). All these papers describe key aspects of the Meeteetse ferret population, habitat require- ments, and means of managing and recover- ing the species. We hope this third chapter will usher in a fourth chapter — full recovery of the species to secure, viable populations scattered over portions of its former range. Individually and collectively, the 14 original contributions to this monograph, plus the in- troductory remarks by Dr. Archie Carr III, provide a much fuller understanding of the spe- cies and the foundation needed for full species recovery. Other study results on the Meeteetse ferrets have been published elsewhere, and they, too, add significantly to our understanding of the ferret and its conservation needs. To be sure, many details about ferret behavior and ecology remain to be learned, but they can wait until the ferret is more common and can accom- modate rigorous scientific scrutiny in laboratory and field. These 14 papers describe numerous aspects of ferret biology, management, and re- covery direction — all for the first time. First, Anderson et al. examine the ferret's fossil record as well as recent distribution and systematics. Pleistocene and Holocene faunas (n = 21) show ferret remains. Ferret distribu- tion based on 412 specimens in 68 museums from Arizona, Colorado, Kansas, Montana, Nebraska, New Mexico, North Dakota, Okla- homa, South Dakota, Texas, Utah, Wyoming, and Canada are summarized. Comparisons of Pleistocene with Recent specimens show no significant differences in size or morphology. Analysis suggests no consistent morphometric variation exists between ferrets found in asso- ciation with different prairie dog species. 'Department of Biological Sciences. Idaho State Universil^ 83001. Fucatello. Idaho 8;32()9, and Biota Research and ( :onsnltiMH. Inc. , Box 2705. Jackson. Wyoming 1986 Clark: Technical Introduction The second section comprises five contri- butions dealing with ferret habitat — historic habitat, the status and characteristics of the Meeteetse area, and methods for locating and measuring potential habitat for reintroduc- tions. The Flath and Clark paper gives a de- scription of ferret habitat — prairie dogs — prior to large-scale alteration of the landscape by early Montana settlers. Their paper de- scribes prairie dog distributions between 1908 and 1914, just prior to the 1915 U.S. Biological Survey efforts to destroy the prairie dog. It shows extensive prairie dog colonies, which today have been nearly eliminated (90+%). It is clear that habitat loss is the single most significant factor in ferret endan- germent. The next paper by Clark et al. gives a description and history of the Meeteetse ferret environment. It shows that ferrets have occurred in the region for at least 100 years. Currently ferrets occupy about 2,995 ha of white-tailed prairie dog (C. leucurus ) colonies that are owned in equal portions by private, state, and federal interests. Many abandoned prairie dog colonies in the immediate area, scattered over large cattle ranches, along with the currently live colonies, total about 8,400 ha. It is believed that the extensive 1930s prairie dog poisoning programs destroyed many of these. The next paper, by Collins and Lichvar, describes vegetation on selected portions of Meeteetse ferret habitat and com- pares it with vegetation on prairie dog con- lonies elsewhere in Wyoming that historically provided ferret habitat. The authors conclude that all sites measured were previously dis- turbed by heavy livestock grazing or other factors and that vegetation is not a useful at- tribute to define ferret habitat or to locate transplant sites. Fagerstone and Biggins describe prairie dog populations at Meeteetse serving as prey for ferrets and present a method to census prairie dogs as a means to locate ferret trans- plant sites. The last paper about ferret habitat by Houston et al. describes a habitat model — a habitat suitability index — useful in locating and comparing transplant sites. It suggests that year-round ferret requirements can be met in prairie dog colonies providing that: (1) prairie dog colonies are large enough, (2) bur- rows are numerous enough, and (3) adequate numbers of prairie dogs and alternate prey exist. Five variables are defined and a method to compare prairie dog colony complexes to each other and to Meeteetse is presented. The third group of papers address ferret behavior, activity patterns, and methods to locate additional ferrets. The Clark et al. pa- per on descriptive ethology and activity pat- terns describes an initial ethogram based on observations of 237 ferrets on 441 occasions (208 hrs). Ferrets were active at extremely cold temperatures (-39 C), in rain, snow, and winds to 50 kph. The next paper by Biggins et al. details activity patterns, based on radio tagging, of an adult male and a juvenile female in the fall. Both animals were primarily noc- turnal. Peak activity was in early morning hours. The female averaged 1.9 hrs per night above ground (moving 76% and stationary 24%). The last paper in this section by John- son et al. examines the use of thin-lay chro- matography to identify scats to species origin. Twenty known ferret scats were compared with 72 unknown scats. This method was not useful, and analysis with gas-liquid chro- matography may prove more definitive. The two papers in the fourth group discuss the genetic viability of the Meeteetse ferrets and minimum viable population sizes. Kil- patrick et al. found no genetic variation in three proteins examined from saliva samples from 22 ferrets. Comparative data is so limited that it is currently impossible to provide a meaningful interpretation of the lack of ge- netic variation, but it is similar to results from other carnivore studies and populations that have undergone genetic "bottlenecks. " The Croves and Clark paper examines five basic methods of determining the minimum viable population size needed for ferrets: (1) experi- ments, (2) biogeographic patterns, (3) theoret- ical models, (4) simulation models, and (5) genetic considerations. The genetic examina- tion proved most useful, resulting in a mini- mum viable population estimate of about 200 ferrets for maintenance of short-term fitness. In the fifth section, two papers deal with management and recovery of ferrets. Clark gives management guidelines for the Mee- teetse ferrets, describing a series of needed monitoring and protection actions. Compara- tive data is listed for these actions as well as the public support and organizational ar- 10 Great Basin Naturalist Memoirs No. 8 rangements needed for successful overall management and recovery of the species. Richardson et al. present a framework for re- covery planning. Three options for increasing ferret numbers are listed and discussed: (1) increase available habitat for the Meeteetse ferrets, (2) find more wild ferrets, and (3) di- rectly manipulate ferrets through transloca- tion and/or captive rearing. The captive rear- ing/translocation option for species recovery is strongly recommended. The final paper in this monograph by Casey et al. lists 351 annotated references on the ferret. These serve as a solid background for understanding the species and the history of the species' study and conservation efforts. Collectively these papers, combined with results of the South Dakota studies and com- panion papers from the Meeteetse studies published elsewhere, provide much new in- formation on the ferret, and, importantly. they outline management needed to conserve the Meeteetse ferrets and essential actions to recover the species. The vital information needed to conserve and recover the ferret is largely in place — now commitment and ac- tion by governmental agencies is called for. Finally, I personally want to extend my sin- cere thanks to all the authors in this monograph, the 40+ reviewers, and the conservation organi- zations that financially contributed to its produc- tion. Dr. Steve Wood, editor of the Great Basin Naturalist, deserves special recognition for his professional management and editing of this vol- ume. And, without the full cooperation and friendship of the Meeteetse area ranchers and the conservation community, this volume would have been impossible. My sincere thanks to all. Funding that made publication of this Memoir possible was provided by Wildlife Preservation Trust International and the new York Zoological Society. PALEOBIOLOGY, BIOGEOGRAPHY, AND SYSTEMATICS OF THE BLACK-FOOTED FERRET, MUSTELA NIGRIPES (AUDUBON AND BACHMAN), 1851 Elaine Anderson', Steven C. Forrest", Tim W. Clark," and Louise Richardson" Abstract. — Extensive literature review and 48 mammal collections containing recent specimens of the endangered black-footed ferret {Mustela nigripes) are used to characterize historic distribution of the species. Specimens (n = 120) were measured from eight collections to characterize black-footed ferret morphology and variation. Twenty-one Pleistocene and Holocene faunas in North America show ferrets dating to 100,000 yr B.P. Recent specimens (n = 412) indicate close association with the prairie dog {Cynomys spp. ) and suggest ferrets may have been less rare than previously thought. At least 103 (25%) of all specimens were taken by federal predator and rodent control agents, and males outnumber females in collections 2.04:1. Average and extreme measurement for external, cranial, and postcra- nial dimensions are tabulated. Ferrets show a high degree of sexual dimoqihism, with discriminant analysis correctly classifying 95% of all specimens to sex. Ferrets also exhibit north-south clinal variation in size, but they do not appear to exhibit variation based on species of Cynomys associate. The taxonomic relationship among ferrets and close relatives is described. The black-footed ferret (Mustela nigripes) is a medium-sized musteline that is listed as endangered throughout its former range and currently receives full protection under the U.S. Endangered Species Act of 1973 (16 use 1531 et. seq.). Endemic to North Amer- ica, black-footed ferrets formerly occupied an extensive range from the Great Plains of Canada to intermontane regions of the inte- rior Rocky Mountains and southwestern United States. The species is currently known from only one population restricted to an ap- pro.ximately 150 sq km area in northwestern Wyoming (Fig. 1). Decline of the black-footed ferret over the last 50 years is attributed to the often systematic eradication of its principal prey and associate, the prairie dog {Cijnomys spp. ), which is often viewed as an agricultural pest throughout the West. Prairie dogs are semifossorial colonial rodents (Sciuridae) that offer an abundant source of prey and burrows for ferret shelter. Because black-footed ferrets are primarily nocturnal and spend much of their time un- derground, they seldom were observed in the wild by naturalists until recent technologies, specifically the high-intensity portable spot- light, made observation possible. Few details of the species biology were known until a small population in Mellette County, South ^IS^IT^^ Fig. 1. Historic range of the black-footed ferret (shaded area) compared with the current known range (dot). Dakota, was studied from 1964 to 1974. Prior to that time information on distribution and '730 Magnolia Street, Denver, Colorado 80220. ^Department of Biological Sciences, Idaho State University, Pocatello, Idaho S3209, and Biota Research and Consulting Inc. , Box 2705, Jackson, Wyoming 83001. 11 12 Great Basin Naturalist Memoirs No. 8 specimens of ferrets were collected sporadi- cally by commercial trappers, museum collec- tors, or federal and state rodent and predator control agents of the U.S. Fish and Wildlife Service (formerly the Biological Survey [BSC] and Bureau of Sport, Fisheries, and Wildlife [BSFW]). Specimens are therefore few and scattered among many collections. Records of M. nigripes specimens and sight reports have been compiled for some states, but no comprehensive record of black-footed ferret distribution based on specimens exists other than Hall (1981). Some authors have included measurements from limited sam- ples, but no systematic analysis based on a large sample has been made. The present study is based on a comprehensive examina- tion and analysis of black-footed ferret re- mains and literature and describes the pale- obiology, distribution, and skeletal mor- phometry of M. nigripes. Materials and Methods Sixty-eight mammal collections were con- tacted and 48 of them reported having M. nigripes in their collections. Of these, eight collections were examined and measured. Collection data were supplemented by a thor- ough hterature review. Evidence of ferrets was confirmed either by the presence in mu- seums of specimens (skins, skeletal material) of M. nigripes or by observations of ferrets in hand reported in the literature by biologists familiar with the species. Some literature re- ports, therefore, include live-captured or killed animals that were not collected or pre- served as museum specimens. Sight reports or secondary sources, however authentic, were not included. Collections containing black-footed ferrets are listed below. Asterisks denote collections from which specimens were measured. AMNH — American Museum of Natural History, New York* ANSP — Academy of Natural Sciences, Philadelphia AUG — Augustana College, Sioux Falls, South Dakota BMS — Buffalo Museum of Science, Buffalo, New York BNP — Badlands National Park, Interior, South Dakota BSC — Biological Services Collection, Fort Collins, Colo- rado* CDOW — Colorado Division of Wildlife, Denver CMNH — Carnegie Mu.seum of Natural History, Pitts- burg CSU — Colorado State University, Fort C'ollins CU — Cornell University Division of Biological Sciences, Ithaca, New York DMNH — Denver Museum of Natural History, Denver, Colorado* FMNH — Field Museum of Natural History, Chicago HM — Hastings Museum, Hastings, Nebraska ISU — Iowa State University, Ames KSU — Kansas State University, Manhattan KUMNH — University of Kansas, Lawrence* MCZ — Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts* MDFWP— Montana Department of Fish, Wildlife and Parks, Bozeman* MHM — Minnilusa Pioneer Historical Museum, Rapid City, South Dakota MSU — Montana State University, Bozeman NDSHS — North Dakota State Historical Society Mu- seum, Bismarck NGFP — Nebraska Game, Fish, and Parks, Lincoln NMC — National Museum of Natural Sciences, Ottawa, Ontario NSCM — Northwestern State College, Alva, Oklahoma NYZ — New York Zoological Society, Bronx, New York NZP — National Zoological Park, Washington, D.C. OSLf — Oklahoma State University, Stillwater OU — University of Oklahoma, Norman PAT — Patuxent Wildlife Research Center, Laurel, Maryland ROM — Royal Ontario Museum, Toronto SDNHM — San Diego Natural History Museum, San Diego, California SNMH — Saskatchewan Museum of Natural History, Regina SYR — State University of New York, Syracuse SZCM — State Zoological Collection, Munich, German Federal Republic UCB — University of California, Berkeley UCM — University of Colorado Museum, Boulder* UMMZ — University of Michigan Museum of Zoology, Ann Arbor UMMNH — James Ford Bell Museum of Natural His- tory, University of Minnesota, Minneapolis UND — University of North Dakota, Grand Forks UNSM — University of Nebraska State Museum, Lincoln USD — University of South Dakota, Department of Zool- ogy, Vermillion USNM — United States National Museum, Washington, DC* UW — University of Wyoming, Laramie. UWZM — University of Wisconsin Zoological Museum, Madison WGF — Wyoming Game and Fish Department, Cheyenne WHO— W^ H. Over Museum, University of South Da- kota, Vermillion YPM — Peabody Museum, Yale University, New Haven, Connecticut ZSP — Zoological Society of Philadelphia Record localities are listed in Table 6 as they appeared on specimen labels or in the litera- ture, with any comments or clarifying notes included in the text or remarks. Specimen label data were organized by collection date 1986 LC-M2 Anderson et al.: Biogeography and Systematics A 13 LM LM' LM' LM|tr-^LM| WM,tr i,\ WM|tr p |tr yM t- Wl3-l3 POP WC-C Wp4.p4 POP Wp4-p4 Fig. 2. Skull and mandible of black-footed ferret (Ad. S , Baca County, Colorado. DMNH 2248) showing measure- ments taken. A, Lateral view of skull. B, Lateral view of mandible. C, Occlusal view of ?■*. D, Occlusal view of M'. E, Occlusal view of Mj. F, Dorsal view of skull. G, Ventral view of skull. For symbols see Materials and Methods. and state or province of collection. Localities (where known) were plotted on maps using dark circles for precise locations and open circles where location was known only to county. Specimens with one exception were mea- sured by E.A. These included 120 recent skulls, (72 of known sex), 17 skeletons, and 55 fossil (Pleistocene to Holocene) specimens. In addition, 19 skulls and one skeleton of the Siberian polecat (M. eversmanni) , a possible Asiatic conspecific of the black-footed ferret, were also measured. Data on external mea- surements were taken directly from skin tags 14 Great Basin Naturalist Memoirs No. 8 and were supplemented by field measure- ments of live-caught known adult ferrets from the extant population at Meeteetse, Wyo- ming, from 1982 to 1984. Measurements of skeletal material were made with vernier calipers to the nearest 1/10 mm. Material was separated by state, species of prairie dog in the locality collected (from the literature), and sex, if known. Figure 2 shows points between which cranial measure- ments were taken (after Anderson 1970). Cra- nial measurements taken included: 1. Condylobasal length (CBL). The least distance from a line connecting the posteriormost parts of the oc- cipital condyles to the anteriormost parts of the premaxillae. 2. Basilar length of Hensel (BL). Least distance from a line connecting the anterior border of the foramen magnum to the posterior margin of the first upper incisors. 3. Rostral breadth (WC-C). Width across the rostrum above the canines. 4. Bimolar breadth (WP^-P^). Greatest width across the hind cheek teeth measured at the posterior margin of P^ and the anterior margin of M . 5. Interorbital breadth (INB). Least distance across the frontal bones at the fronto-maxillary suture. 6. Postorbital breadth (POP). Greatest width across the postorbital processes. 7. Postorbital constriction breadth (POC). Least width across the frontal bones behind the postorbital processes. 8. Mastoid breadth (MW). Greatest width across the mastoid processes perpendicular to the long axis of the skull. 9. Mandible length (LJAW). Total length from the sym- physis at the alveolus of Ii to the most distant edge of the condyle. 10. Mandible height (H). From the lower border to the tip of the coronoid process. 11. 12. Ramus depth (DP3.4, DM1.2). Depth of the jaw between P3.4 and M1.2 measured from the level of the alveoli to the lower border. 13. Maxillary tooth row length (LC-M'). Least distance from the anterior border of the canine at the alveo- lus to the posterior border of M ' at the alveolus. 14. Mandibular tooth row length (LC-M2). Least distance from the anterior border of the canine at the alveo- lus to the posterior border of Mj at the alveolus. 15. Incisor breadth (Wl^-I^). Least width from the buccal side of right I^ to the buccal side of left I^. 16. Canine length (LC). The least distance between the anterior and posterior edges of the canine at the level of the alveolus. 17. Canine breadth (WC). Transverse width of the canine at the level of the alveolus. 18. 19. Premolar length (LP^ LP^). Least distance from the anterior to the posterior edges of the premo- lars measured on the buccal side in the plane of the tooth row. 20. Premolar breadth (WP^). Transverse width of P^ mea- sured at the center of the cusp. 21. Breadth of protocone of P'* (WP*pc). Greatest trans- verse width from the buccal border of the tooth to the edge of the protocone. 22. Upper molar breadth (WM'). Greatest transverse width M'. 23. Upper molar length (LM^nner). Greatest anterior- posterior length of the inner lobe of M . 24. Length of M, (LMj). Greatest anterior-posterior length of Ml measured on the lingual side. 2.5. Trignoid length of Mi (LMitr). From the posterior edge of the protoconid to the anterior edge of the tooth. 26, 27. Breadths of Mi (WMi tr, WMi tal). Greatest width of the trigonid measured across the protoconid- metaconid; greatest width of the talonid measured across the hypoconid-entoconid. 28. Palatal breadth at canines (PB C-C). Width of palate between canines. 29. Basioccipital breadth (B OB). Breadth of basioccipital taken at midpoint between bullae. Postcranial measurements taken included: Humerus Total length (TL). Greatest distance from the greater tuberosity to the medial epicondyle. Proximal breadth (PB). Greatest width across the greater and lesser tuberosities. Least shaft breadth (LSB). Least diameter of shaft. Distal breadth (DB). Greatest width across the medial and lateral epicondyles. Ulna Total length (TL). Greatest distance from the top of the olecranon to the styloid process. Breadth olecranon process (B 01 Pr). Maximum width of the olecranon. Radius Total length (TL). Greatest distance between the head and the styloid process. Proximal breadth (PB). Maximum breadth across the head. Distal breadth (DP). Maximum width of the distal end. Femur Total length (TL). Greatest distance from the head to the medial epicondyle. Proximal breadth (PB). Maximum breadth between the greater trochanter and the head. Least shaft breadth (LSB). Least diameter of the shaft. Distal breadth (DB). Greatest distance across the condyles. Tibia Total length (TL). Greatest distance between the lat- eral condyle and the medial malleolus. Proximal breadth (PB). Maximum breadth between the medial and lateral condyles. Distal breadth (DB). Maximum width across the distal end. Fibula Length (L). Total length between the lateral condyle and the lateral malleolus. Calcaneum Length (L). Total length between the calcaneal atuberositv and the cuboid facet. 1986 Anderson et al.: Biogeography and Systematics 15 Astragalus Length (L). Greatest perpendicular length of the bone. Baculum Length (L). Greatest length of the bone from the proxi- mal end to the base of the curve. Specimens were placed in tentative age classes by the following criteria: (1) juvenile: cranial sutures open, deciduous dentition present, but permanent teeth beginning to erupt, epiphyses of long bones not fused; (2) young adult: internasal, nasomaxillary, ba- sisphenoid, and basioccipital sutures fused but not obliterated, permanent dentition fully erupted except for upper canines, teeth un- worn or only slightly worn, epiphyses of long bones fused, but sutures still visible; (3) adult: all cranial sutures obliterated, permanent dentition fully erupted, well-developed sagit- tal crest especially on males, epiphyseal su- tures obliterated. Statistical Methods Analyses were performed on a Hewlett Packard 3000 computer using the Statistical Package for the Social Sciences (SPSS), in- cluding Discriminant Analysis and One-way Analysis of Variance (ANOVA). Linear dis- criminant analysis was performed between sexes using standardized measurements and confined to specimens of known sex. Juveniles were ommitted from the analysis to avoid allo- metric variation. Ranges, means, and stan- dard deviations were calculated for both sexes of M. nigripes . Scattergrams and frequency diagrams were used to describe relationships between fossil and recent material and inter- species comparisons. A second linear discrim- inant analysis was performed on standardized cranial measurements of male and female fer- rets to identify groups based on geographic variation and subgenera of prairie dog associ- ate. One-way analysis of variance (ANOVA) was used to explore further variation of indi- vidual variables with regard to geographic clinal variation. Description Mustela nigripes (Audubon & Bachman) Putorius nigripes Audubon & Bachman 1851: 297. Type locality Ft. Laramie, Goshen Co., Wyoming. Mustela nigripes Miller 1912: 102. First use of binomial. No subspecies are recognized. M. nigripes Recent ^ I ' I =f M. nigripes Pleistocene/ Holocene a ■ I ■ I ' I ■ I ' I ' I 6.05 6.65 7.25 7.85 8.45 9.05 6.35 6.95 7.55 8.15 8.85 9.35 Fig. 3. Frequency histogram of length of Mi for Re- cent, Pleistocene, and Holocene specimens. Diagnosis. — Mustela nigripes is a mink-sized mustelid weighing 645-1125 g. Upper parts yellowish buff, occasionally whitish, espe- cially on the face and venter; feet black; black mask across the eyes, particularly well de- fined in young animals; tail black tipped. Skull is relatively short and broad; mastoid process is notably angular (Hillman and Clark 1980). Closely resembles M. eversmanni, the steppe ferret of Eurasia. Differs from M. putorius, the European ferret, and M. vison , the Amer- ican mink, in being light colored with black markings; the latter two species are uniformly dark colored, and M. p. furo, the domestic ferret, is uniformly light colored, often al- binistic. Morphometry Data on external measurements for Recent material were taken directly from skin tags and are supplemented with field measure- ments of live-caught juvenile and adult ferrets from Meeteetse from 1982 to 1984. Average ( ± S. D.) and extreme external measurements 16 Great Basin Naturalist Memoirs No. WM, talonid 4.0 3.5 3.0 2.5 2.0 1.8 D DD D 1 D D D ■ ■ D D D D ■ D D D nn _ DD DDBBlf ■ D DD DD ■■ ■■■ ■D BSD ■ ■ D DDD ■ ■ O D DD D DB O OOOQ# DDB ■ O •0(^# i 0^ A ■ k2^ lo^ D ■ O ^OOOO <9( 9n^ D ■ A # OQ(fOO 30,000 yrs.B. P. 1 1 "Citellus" beds Nebraska, Lincoln Co. 28,000-30,000 yrs. B.P. 1 3 Cottonwood Canyon Nebraska, Lincoln Co. 28,000-30,000 yrs. B.P. 1 1 Smith Falls Nebraska, Cherry Co. Wisconsin 1 1 Harlan Co. Dam Site Nebraska, Harlan Co. Wisconsin 1 2 January Cave Alberta, Canada 23, 100-33,500 yrs. B.P. 2 7 Little Box Elder Cave Wyoming, Converse Co. 9,000- > 30,000 yrs . B.P. 15 40 Chimney Rock Animal Trap Colorado, Larimer Co. 11,980 ± 180 1 1 Burnet Cave New Mexico, Eddy Co. 11,170 ± 360 1 1 Jaguar Cave Idaho, Lemhi Co. 10,370 ± 350 2 12 Little Canyon Creek Cave Wyoming, Washakie Co. 10,170 ± 250 1 1 Orr Cave Montana, Beaverhead Co. Late Pleistocene 1 1 Old Crow River, Yukon Territory, Canada Late Pleistocene 1 Location 65 Isleta Cave New Mexico, Bernalillo Co. Late Pleistocene/ early Holocene 2 6 Red Willow Nebraska, Red Willow Co. Late Pleistocene/ early Holocene 1 1 Moonshiner Cave Idaho, Bingham Co. Early Holocene 1 2 Atlatl Cave New Mexico, San Juan Co. 2,000-3,000 yrs. B P. 1 1 Ashislepha Shelter New Mexico, San Juan Co. Archaic 1 1 Upper Plum Creek Colorado, Las Animas Co. 570 ± 50-1,050 ± 80 AD 2 4 Mustela eversinanni beringiae Fairbanks Alaska, near Fairbanks Late Pleistocene 2 3 clusters of prairie dog colonies appear neces- sary to support populations. The lack of such colonies in highly developed prairie lands is suspected as the principle cause of ferret en- dangerment, although possible catastrophic losses of prey base due to sylvatic plague in prairie dogs has also been discussed (Hubbard and Schmitt 1984). Distribution Pleistocene and Paleo-Indian Distribution Ferrets have been identified from 21 Pleis- tocene and Holocene faunas in North America (Table 4). Two ferret species, Mustela evers- manni from Fairbanks, Alaska (Anderson 1977) and M. nigripes are recognized. Six oc- currences of M. nigripes are outside the his- toric range (Fig. 9) of this species. The earliest occurrence of M. nigripes is uncertain, but the species has probably been present in North America since the Sangamo- nian about 100,000 years ago. The specimen from the Cudahy fauna, an isolated left M^ (University of Michigan Museum of Verte- brate Paleontology #38341) was originally identified as Mustela ci vison by Getz (1960), who noted slight differences between it and the comparative material. Later Hibbard (1970) referred the specimen without com- ment to M. nigripes and Corner (1977) fol- lowed this designation. Examination of the tooth showed the presence of an incipient metaconid, a characteristic of mink but not ferret. Measurements of the tooth of the two species are not diagnostic. The age of the Cud- ahy fauna is Irvingtonian (Type 0, Pearlette Ash, 600,000 yrs B.P.), and the habitat was marshy with permanent, slow-moving streams; numerous species of acjuatic and semiaquatic animals have been identified. Mink have been identified from other Irvingtonian faunas, fer- rets have not. Thus, the identity of this tooth remains questionable. We follow Getz's (1960) 1986 Anderson etal.: BiogeographyandSystematics 23 Prey species Remarks References No Cynomys, many spp. rodents ? Id. Originally identified as M. vison, see text Cynomys leucurus No Cynomys, many spp. rodents No Cynomys, many spp. rodents No Cynomys, rodents abundant Cynomys sp. No Cynomys, many spp. rodents Cynomys leucurus Cynomys leucurus Cynomys sp. Cynomys ludovicianus No Cynomys, many spp. rodents No Cynomys, many spp. rodents No Cynomys No Cynomys Cynomys gunnisoni Cynomys ludovicianus No Cynomys, many spp. rodents Cynomys gunnisoni 104 Cynomys ludovicianus No Cynomys, many spp. rodents, lagomorphs Grassland, warmer than today Originally identified as M. vison Age originally thought to be Sangamonian May be "Citellus" zone in part; open prairie Steppe Articulated skull and mandible; open prairie ? Id. Juvenile; deciduous dentition Outside historic range Outside historic range Outside historic range, cool grassland Carnivore trap; outside historic range Specimen burned 1 specimen burned, 2 found in bone cache Cool grassland. Only record of M. e. in North America Getz I960, Hibbard 1970 L. Martin, pers. comm. Stalker et al. 1982 Slaughter 1966 Dreeszen 1970 Corner, pers. comm. R. G. Corner, pers. comm. Voorhies and Corner, in press R. G. Corner, pers. comm. J. Burns, pers. comm. Anderson 1968, 1974 Hager 1972 Schultz and Howard 1935 Kurten and Anderson 1972 D. Walker, pers. comm. Cuilday and Adam 1967 C. R. Harington, pers. comm. Harris and Findley 1964 Corner 1977, pers. comm. White et al. 1984 J. Hubbard, pers. comm. W. Gillespie, ms. and pers. comm. Anderson, ms. Anderson 1977 designation of M. ci vison. The next earliest record of ferret may be late Illinoian/early Sangamonian (Adams/Clay counties, Nebra- ska, exact age uncertain; the specimens from the "Citellus" beds were originally thought to be the same age but are now regarded as late Wisconsinan in age) or Sangamonian (Medi- cine Hat), about 100,000 yrs B.P. By the late Wisconsin/early Holocene (15,000-8,000 B.P.), ferrets ranged across the Great Plains west to Montana (Orr Cave) and Idaho (Jaguar, Moonshiner caves) and even as far north as Yukon Territory (Old Crow). At most sites only a few bones representing one indi- vidual have been found, but at Little Box- Elder Cave at least 40 specimens and 15 indi- viduals (based on left mandibles) have been identified. The site, in the foothills of the Laramie Mountains, contains a large number of prey animals including Cynomys cf leucu- rus (n = 77-l-). Prairie dogs have been found in 10 of the faunas containing ferrets. The other sites did not contain prairie dogs, but various rodent and lagomorph species were abun- dant. At two archeological sites, Atlatl Cave and Upper Plum Creek Rockshelter, burned ferret bones were found, indicating their pos- sible use by Paleoindians. During the late Pleistocene the steppe fer- ret, M. eversmanni , ranged east to Beringia, the vast unglaciated land mass that extended from northeastern Siberia to western Alaska. Its remains have been found in deposits near Fairbanks, Alaska (Anderson 1973, 1977). The specimens, a partial skull and two mandibles, are characterized by large size, broad facial region, massive postorbital processes, pro- nounced postorbital constriction, crowded tooth row, and enlarged canines. Measure- ments exceeded those of M. eversmanni mich- noi, the largest extant subspecies. Anderson (1977) described the material as a new subspe- 24 Great Basin Naturalist Memoirs No. 8 .^.\ic\^^ ^ H^ Mustela eversmanni A Mustelo nigripes • Questionable record ■ Historic range of / -, M. nignpes Sr- /" y- \ \ — ~-..„^\ •! (^ '- N \ ""^^^^ f 7 W* — - / L^ /^ •/ y — / ^ V y s' (*• — 1 ; I ~L.^ 1000 KM t^; ■^w \ r J Fig. 9. Distribution of black-footed ferrets in Pleis- tocene, early Holocene, and archeological faunas com- pared with its historic range (1851-1920). cies, M. e. beringiae, and noted that it is the first and only record of the steppe ferret in North America. In the Old World ferrets are recognized in middle Pleistocene faunas in central Europe, but whether the fragmentary remains are of M. putorius or M. eversmanni is unknown. By the late Pleistocene remains of M. putorius are common in Eurasian cave faunas. Mustela eversmanni has been reported from late Pleis- tocene/Holocene faunas in Siberia, Crimea, and the Russian plains and from Holocene faunas in the Caucausus and central Asia (Vereshchagin and Baryshnikov 1984). Early Historic Record The first possible report of M. nigripes by a European may be attributed to Don Juan de Onate, a Spanish explorer of what is now the southwestern U.S. in 1599. It is a land (New Mexico Territory) abounding in flesh of buffalo, goats with hideous horns, and turkeys; and in Mohoce there is game of all kinds. There are many wild and ferocious beasts, lions, bears, wolves, tigers, penicas, ferrets, porcupines, and other animals, whose hides they tan and use. (Bolton 1916: 217; italics ours). Both domestic (M. putorius furo) and wild European polecats (M. p. putorius) were common in Europe at that time, so Ofiate could have correctly identified the North American counterpart. It is possible he could have confused ferrets with other Mustela that have no comparable Old World counterparts, particularly the "bridled" weasels (color morphs of M. frenata : M. /. arizonensis and M. /. neomexicana) of the southwest. Since Mohoce (also "Moqui," the center of the Hopi nation, in the vicinity of present-day Walupi, Arizona) is within known black-footed ferret range, it appears equally likely that what Oiiate described was in fact a black-footed ferret. Several ethnographically known tribes were familiar with and used black-footed ferret skins in ceremonial dressings (Henderson et al. 1969, Clark 1975). Tribes with knowledge of black-footed ferrets ranged from the Navaho of northern Arizona and New Mexico to the Hidatsas of the upper Missouri River Basin (Table 5). Ferrets were also reported in the records of the American Fur Company from 1835 to 1839 (Johnson 1969). Pratte, Chouteau, and Company of St. Louis listed pelts of 86 black- footed ferrets taken during this period. Trap- pers were familiar with mustelids ("weasels" are listed separately from ferrets in this tally) and probably accurately identified the species long before it was scientifically named and described by Audubon and Bachman (1851). It was Alexander Culbertson, a trapper, who first brought the species to the attention of Audubon. Known as the "French Fur Com- pany," Pratte, Chouteau, and Company be- came the Western Department of the Ameri- can Fur Company, concentrating most of their effort in the upper Missouri River basin. They operated for some time out of Ft. Kiowa near the junction of the White and Missouri rivers in present-day South Dakota (Morgan 1953). Their license was for "Sioux Country, " which encompassed parts of present-day South Dakota, Montana, and Wyoming. 1986 Anderson et al. : Biogeography and Systematics 25 Table 5. Black-footed ferret specimens associated with ethnographically known Indian tribes. Ethnographic No. of tribe Tribal name for BFFs Specimen type specimens Disposition Citations Blackfoot p chiefs headdress 1 ? Homolka 1964 Cheyenne ? chiefs headdress ? ? Henderson et al. 1969 Crow ? medicine pouch 4 Chief Plenty Coups Museum, Pryor, Montana medicine pouch 2 Plains Indian Museum, Cody, Wyoming skin 2 Plains Indian Museum, Cody, Wyoming skin 1 Colter Bay Indian Museum, Grand Teton National Park, Wyoming Hidatsas "Tahu akukahak napish" Bailey 1926 Mandan "Nazi" Bailey 1926 Navajo "dlo'ii liz-hinii" "pelts" "several" Fortenbery 1971, Halloran 1964 Pawnee "ground dogs" in a mythical story; speaks of itself as "staying hid all the time" Grinnell 1895, 1896 Sioux "pispiza etopta sapa," skins, sacred 2 St. Francis, Henderson et al. 1969 "black-faced tribal objects South Dakota prairie dog" Recent Distribution John James Audubon and John Bachman (1851) named and described Mustela nigripes from a specimen collected near Fort Laramie, in what is now Goshen County, Wyoming. This specimen was either lost or destroyed, and subsequently naturalists questioned whether the species actually existed (Gray 1865). Elliot Coues (1874:1), curator of the Smithsonian mammal collection, published a plea in the American Sportsman for additional specimens likely to be found "out on the plains in the prairie dog towns." Coues was rewarded with several specimens and ac- counts of black-footed ferrets, which he sub- sequently described (Coues 1877). Table 6 gives a comprehensive listing of each known ferret specimen by state. Black- footed ferrets have been found with three spe- cies of Cynomys: C. ludovicianus (black- tailed prairie dog), C. leucurus (white-tailed prairie dog), and C. gunnisoni (Gunnison's prairie dog). Since M. nigripes distribution and abundance is highly dependent on prairie dog distribution and abundance, we include discussion of the past and present range of Cynomys . Because genetic "bottlenecks " occur when species numbers are low and may be criti- cal for species survival (Soule 1980), an estimate of habitat available to ferrets based on prairie dog distribution at their lowest point is also given where known. Arizona Four specimens of M. nigripes are known for Arizona (Table 6, Fig. 10). The last specimen was collected in Coconino County in 1931. Two of these specimens from the U.S. National Mu- seum were described by Young and Halloran (1952). Historically two species of prairie dogs, C. lu- dovicianus Sind C. gunnisoni, inhabited Arizona. The ferrets were all collected within the range of C. gunnisoni. Cynomys ludovicianus was proha- bly extirpated in Arizona as early as 1932 (Alex- ander 1932) and no longer exists in the state (Cockrum 1960). Current distribution of C. gun- nisoni is also greatly reduced, although relict populations of sufficient size to support black- footed ferrets may persist in the northeastern cor- ner of the state. The historic, as well as present area occupied by prairie dogs is unknown. 26 Great Basin Naturalist Memoirs Table 6. Recent black-footed ferret specimen accounts by state, 1851-1984. No. 8 Skel- Year Date Disposition County Site Sex eton Crania Skin Arizona 1917 Jan 19 USNM 228233 Apache Springerville, 27 km NE M X X 1929 Jan USNM 248973 Coconino Williams, N Red Lake M X X 1931 Oct UCB 55213 Coconino Winona, 19 km W F 1931 Nov UCB 55212 Coconino Gov't, prairie near parks — Colorado ca 1876 _ Unknown (n = 3) — "Vicinity of Denver" — ca 1877 — Unknown Larimer Valley of the Cache La Poudre — X 1878 Apr AMNH 24412 El Paso — — 1887 Feb MCZ B4184 Larimer — M X X 1887 Jan 6 ANSP8640 Larimer — F X X 1887 _ ANSP8641 Larimer _ 1888 Apr DMNH653 Grand Middle Park M X 1900 Aprl UWZ 11776 El Paso Colorado Springs M X X 1904 _ Unknown El Paso Clyde Station — 1905 Jan 16 UCM 10658 Teller Divide Station M X X 1905 Apr 14 UCM 10659 Baca N of Springfield F X X 1905 Sep 23 UCM 10660 El Paso Lake Moraine F X X 1909 Jan 2 UCM-W59 Larimer Laramie R. , 19 km S of Wyoming border M X ca 1910 — Private Collection Rio Blanco Meeker, 2 km — ca 1910 - Private Collection Rio Blanco Meeker, 2 km - 1910 Mar UCM-W232 Weld Cornish, 13 km E X 1912 May 5 DMNH257 Denver Denver, Park Hill (F) X 1913 CSU Larimer _ X 1914 Maris DMNH 1208 Adams Barr F X X 1914 Dec 16 DMNH 1558 Adams Simpson M X X 1914 Dec 16 DMNH 1559 Adams Simpson F X 1915 Mar 30 AMNH 41994 Adams Simpson F X 1915 Oct 31 DMNH 5792 Jefferson Semper M X X 1916 Feb 21 DMNH 1684 Adams Simpson M X X 1916 Feb 21 DMNH 1883 Adams Simpson F X X 1919 Nov USNM 234118 Saguache Del Norte, 24 km NW M X X 1922 _ USNM 265540 Weld E of Greeley — X 1923 May 13 DMNH 1987 Weld Grover, 8 km S M X X 1923 May 13 DMNH 6726 Weld Grover, 8 km S F X 1924 Feb DMNH 2024 Baca Furnace Canyon M X X 1924 Feb DMNH 2247 Baca Furnace Canyon (M) X X 1924 Feb DMNH 2248 Baca Furnace Canyon (M) X 1926 Nov USNM 247073 Park Hartsel, llkmS F X X 1928 Feb 9 DMNH 2371 Baca Furnace Canyon (M) X 1930 Feb 11 DMNH 4322 Adams Denver, 16 km E F X 1934 Aug UCM-W493 Montezuma Mancos F X 1935 Jan 17 UCB 66019 Yuma Wray M X X 1935 Dec 6 UCB 70209 Yuma Wray M X X 1937 Nov 7 DMNH 3206 Weld Greasewood — 1939 Sep 16 DMNH 3644 Denver Denver, 1st and Holly M X 1939 Aug DMNH 3703 Denver Denver M 1940 Aug 21 UCB 95039 Moffat Craig, 35 km N M 1941 Jan CMNH 19392 Moffat Morapos Creek, 32 km SW Craig — X X 1941 Oct 16 UCB 96904 La Plata Durango — X 1941 Oct 16 UCB 96905 La Plata Durango — X 1941 Dec 21 CMNH 20627 Moffat Craig, 8 km W M X X 1942 Jan CMNH 20628 Moffat Craig F X X 1943 Apr 18 DMNH 5199 Chaffee Buena Vista M X 1946 Feb 10 AMNH 140397 Costilla Ft. Garland-Buck Mountain M X X X 1951-52 Winter Destroyed Bent Las Animas _ 1952 Sep Destroyed Weld Dearfield, 8 km E — found 1977 CDOW 230 Logan TION R45W S21 — X Kansas cal877 _ Unknown Wallace Ft. Wallace — 1883 Nov 20 CU 4971 Trego — — X 1884 Oct 10 USNM 188450 Trego — M X X 1986 Anderson et al. : Biogeography and Systematics 27 Other X-mount Citation W. S. Carlos (A. M. Alexander) O. Wright (A. M. Alexander) Meas- ured by us Remarks X Gunnison's prairie dogs Gunnison's prairie dogs Gunnison's prairie dogs Gunnison's prairie dogs X-mount Mrs. M. H. Maxwell Coues 1877 Dr. Law (through F. V. Hayden) Coues 1877 S. N. Rhoads C. K. Worthan (donated by S. N. Rhoads) C. E. Aiken C. E. Aiken E. R. Warren E. R. Warren G. DeLong R. S. Bull R. S. Bull E. R. Warren C. Deardorff E. Sutton W. W. Davidson W. W. Davidson J. B. Burns W. D. Holhster A. H. Burns A. H. Burns R. J. Niedrach S. O. Singer S. O. Singer S. O. Singer S. O. Singer D. Spencer O. W. Shirley O. W. Shirley R. Dietrich A. E. Borell W, Dicus F. Barnes F. Barnes W. Dicus W. Dicus R. C. Prater L. E. Miller Gary 1911 Warren 1910 Felger 1910; Warren 1910 Felger 1910; Warren 1910 White-tailed prairie dogs 2,806 m elevation; Gunnison's prairie dogs 3,126 m elevation; dead in lake (origin unk.) White-tailed prairie dogs R. S. Bull Collection, Meeker Hotel; white- tailed prairie dogs R. S. Bull Collection, Meeker Hotel; white- tailed prairie dogs Standing mount Standing mount Gunnison's prairie dogs Gunnison's prairie dogs Cahalane 1954 Cahalane 1954 Bissel 1979 X Gunnison's prairie dogs X Road kill Road kill White-tailed prairie dogs White-tailed prairie dogs Gunnison's prairie dogs Gunnison's prairie dogs White-tailed prairie dogs White-tailed prairie dogs X Gunnison's prairie dogs X Gunnison's prairie dogs Drowned in ditch. Road kill X-mount L. H. Kerrick A. B. Baker Not listed in Ghoate et al. 1982. 28 Table 6 continued. Great Basin Naturalist Memoirs No. 8 Disposition County Skel- Sex eton Crania Skin 1885 Oct 20 USNM 15471/22427 Trego — 1886 Apr 3 USNM 15470/22311 Trego — 1886 Nov 20 USNM 188451 Trego — 1887 Mar 31 USNM 188452 Trego — 1887 Apr 3 USNM 188453 Trego — 1887 Apr 5 USNM 188454 Trego — 1887 Apr 5 USNM 188455 Trego — 1887 Oct 17 AMNH 1203/1928 Trego _ ca 1888 — USNM 12299/22929 Wallace Ft. Wallace 1888 _ USNM 188458 Trego _ 1889 Apr 15 USNM 188456 Trego — 1889 Apr USNM 188457 Trego — 1889 Nov USNM 22537/30064 Trego — 1890 May 8 USNM 22538/30065 Gove _ 1890 June 2 USNM 22539/30066 Gove — 1891 Jan USNM 25358/32771 Trego Banner 1891 Feb 7 USNM 83992 Trego Banner 1891 Feb 4 USNM 83994 Trego Banner ca 1891 — USNM 19262/35376 Trego — ca 1891 _ USNM 19263/35016 Trego _ 1891 — USNM 19294/35017 Trego — 1891 — USNM 19295/35018 Trego — 1891 _ USNM 34977 Trego _ 1891 — USNM 35011 Trego — 1891 Mar 6 USNM 83993 Trego Banner 1891 _ USNM 19538 Trego _ 1896 Nov 24 KUMNH 1487 Kingman Kingman 1901 Nov 2 USNM 110772 Logan Oakley 1904 Dec UCM 895 Saline _ 1905 — MCZ 42723 Trego Wakeeney 1909 — to USNM (no record) (NZP 7494) Wallace — 1910 — USNM 199737 (NZP 7802) Wallace — 1910 — Destroyed (NZP 7804) Wallace — 1910 — Unknown (NZP 7803) Wallace — 1910 — London Zoo (NZP 7805) Wallace — 1914 May SDNHM6720 Decator _ 1914 Sep KUMNH 134415 Trego Banner 1930 Fall KUMNH 10177 Lincoln Lucas 1933 Oct KUMNH 11077 Hamilton Coolidge 1935 Jan 24 MCZ 43727-KU 10973 Hamilton Coolidge 1935 Jan KUMNH 12119 Hamilton Coolidge 1939 — CM 21391 Jewell Ionia 1944 Oct HM 25099 Smith Soflnvale, Neb. 1957 Dec 31 KSU Sheridan Studley found 1978 — M HP 15569 Gove Healey, 13 km NW Montana ca 1877 — Unknown _ "Milk River- 1892 Aug ANSP8041 Cascade Great Falls 1910 Jan USNM 155475 Dawson Glendive 1915 SCZ MUNICH Garfield Jordan 1915 _ FMNH 25621 Garfield Jordan 1915 — FMNH 25622 Garfield Jordan 1915 — FMNH 25623 Garfield Jordan 1916 May MSU 369 Custer _ 1916 Sep 23 UCB 25709 Garfield Jordan, 6 km S 1916 Sep USNM 224450 Custer Kimball 1916 Sep 23 AMNH 40078 Garfield Jordan, 6 km S 1919 May USNM 232400 Rosebud NW Calabar, Whitetail Creek 1920 Jan USNM 239138 Teton Choteau 1920 Apr USNM 234970 Powder River Broadus 1920 Apr USNM 234971 Powder River Broadus F X X M X X M X X F X X M X X M X M X X M X X (M) X X M X X F X X F X X F X X M X X M X X F X X M X M X M X X M X X M X X M X X M X X F X X X F X X M X M X X M X M (M) X X F F X M XX M XX MX XX MX X MX XX M M M XX — X (M) X X M X — X M X — X MX X MX XX M XX M X M XX (M) X X F XX F XX 1986 Anderson et al.: Biogeography and System atics 29 Other CoUecter A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker L. H. Kerrick A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker A. B. Baker C. A. Hawkes L. W. Purington L. W. Purington L. W. Purington A. B. Baker W. H. Osgood EH. Herrick (L. H. Kerrick?) H. Byxbe Meas- ured by us Remarks NZP specimen; (no accession card) from NZP toUSNM22May BSC BSC NZP (no accession card), NZP (no accession card), NZP (no accession card), NZP (no accession card), NZP (no accession card), NZP (no accession card). rec'd. 19FebbyUSNM rec'd. IQFebbyUSNM rec'd. 24 Feb by USNM rec'd. 24 Feb by USNM rec'd. 19 Feb by USNM rec'd. 19 Feb by USNM X X BSC NZP rec'd. 3 Apr 1909; died Nov 2 1911 X NZP rec'd. 17 Jun 1910; died 2 Jul 1915 NZP rec'd. 17 Jun 1910; died 26 Nov 1913 NZP rec'd. 17 Jun 1910 NZP rec'd. 17 June 1910; exchanged 2 Feb 1911 X-mount R. Kellogg O. Conrad D. Conard (O. Conrad?) Choate and Fleharty 1975 Taylor 1961 Boggess et al. 1980 X X Permanent loan to MCZ 4 Jan 1949 Missing "lona"; body mount C. Cavilieer R. Williams (donated by S. N. Rhoads) X-mandible L. L. Walters Coues 1877 P. Youngman, pers. comm. From FMNH Parker & Wells L. L. Walters L. L. Walters Skin missing Apr 1980 X X X ADC Reports X X X 30 Gri ■AT Basin N ATURALiST Memoirs N 0.8 Table 6 continued. Skel- Year Date Disposition County Site Sex eton Crania Skin 1920 May USNM 234972 Powder River Broadus F X X 1920 May USNM 234973 Powder River Broadus M X X 1920 Oct Unknown Rosebud Ashland — 1923 Sep USNM 243818 Rosebud Birney M X X 1923 Sep USNM 243819 Rosebud Bimey F X X 1923 Oct USNM 243820 Powder River Ashland, E F X X 1923 Nov Unknown Rosebud Lee — 1923 Nov USNM 243909 Bighorn St. Xavier F X X 1923 Nov USNM 243910 Bighorn St. Xavier (M) X X 1923 Nov Unknown Rosebud Ashland — 1923 Nov Unknown Rosebud Ashland _ rec'd 1923 USNM X 23272 Bighorn Crow Agency — X 1923 Dec Unknown Powder River Camps Pass — 1923 Dec Unknown Phillips Phillips — 1924 Jan Unknown Phillips Regina — 1924 Sep Unknown Choteau Geraldine — 1928 Aug Unknown Prairie Terry — 1928 Aug Unknown Prairie Terry — 1935 Sep UCB 78134 Carter — M X 1942 Jan USNM 288288 Fergus Harlowtown, 48 km N _ X 1944 Oct KU 14411 Carter _ M X X X 1948 Sep Destroyed Golden Valley Lavina, 8 km S M 1949 Mar MSU 370 Yellowstone Billings, lekmSE M X X 1952 — Unknown Rosebud Ingomar M X 1953 Nov USNM 287322 Carter Alzada, llkmN M X found 1983 BSC Blaine Ft. Belknap Reservation T30N R25E S17 _ X found 1984 Jan MDFWP2344 Carter Ekalaka (M) found 1984 Jan MDFWP Carter Ekalaka — — USNM 13113/21976 Bighorn Ft. Custer (M) X X Nebraska ca 1877 _ USNM 14580 _ _ (M) X X 1890s _ Private Collection Frontier Curtis _ 1890s — Private Collection Frontier Curtis ) 1890s — Unknown Lancaster Lincoln — ca 1890s _ UNZM 2333 Box Butte Marsland.SkmSE X 1917 Sep AM NH 42567 Sioux Agate M X 1919 May Brookings 1989 Frontier Maywood — 1927 _ HM 10038a Buffalo Gibbon 1934 Mario AMNH 121610 Webster Rosemont M X X 1938 Jan 26 HM 18041 Clay Glenvil — 1938 _ Unknown Custer Anselmo F 1939 Apr 16 HM 19074 Furnas Cambridge — 1946 May 6 UNZM 3323 Banner Gering, 14 km S F X 1947 — Destroyed Knox — — 1949 Mar 16 NGFP Phelps Overton, S. Platte River — _ _ Private Collection Garden Oshkosh — — Brookings 10038b Buffalo Gibbon — _ _ Unknown Buffalo Kearney _ _ _ Unknown Private Coll . Hamilton Harvard, N — — Private Collection Buffalo Kearnev — — — Unknown Custer Arnold ' — — _ USNM 12387 Lincoln N Platte X — — USNM 12409 Dawes Spotted Tail Agency X New Mexico found 1899 Jun USNM Chaves Roswell — 1915 Mar 18 YPM 1969 Catron 1918 Mayl USNM 228789 McKinley 1918 Oct 25 USNM 231363 Cibola Reserve, 24 km N San Mateo, 16 km NE Bluewater, 3 km N M XX M XX M X 1986 Anderson ETAL.iBioGEOGRAPHY AND Systematics 31 Collecter Meas- ured i)v us Remarks X X X X X X ADC Reports D. L. Flath, pers. comm. ADC Reports D. L. Flath, pers. comm. ADC Reports D. L. Flath, pers. comm. D. L. Flath, pers. comm. X ADC Reports ADC Reports M. W. Jelhson D. L. Flath, pers. comm. D. L. Flath, pers. comm. D. L. Flath, pers. comm. D. L. Flath, pers. comm. D. L. Flath, pers. comm. D. L. Flath, pers. comm. X ADC Reports ADC Reports ADC Reports ADC Reports ADC Reports ADC Reports Crabb & Watson Crabb & Watson Crabb & Watson 1950 Cahalanel954(#10) Hoffman etal. 1969:597 Road kill C. Knowles S. Forrest X-mandible T. M. Campbell III A. Thompson C. J. Pfeifer X-mount J. Shields H. Turner Stahnke R. Block W. Townsley O. Blevins Coues 1877 Fichter & Jones 1953 (#4) Fichter& Jones 1953 (#4) Fichter & Jones 19.53 (#1) Fichter & Jones 1953 (#3) Fichter & Jones 1953 (#9) Fichter & Jones 1953 (#10) Fichter & Jones 1953 (#14) Fichter & Jones 1953 (#11) Velich 1961 Fichter & Jones 1953 (#13) Fichter & Jones 1953 (#17) R. Block, pers. comm. Fichter & Jones 1953 (#18) Fichter & Jones 1953 (#15) Fichter & Jones 1953 (#12) Fichter & Jones 1953 (#5) Fichter & Jones 1953 (#6) Fichter & Jones 1953 (#7) Fichter & Jones 1953 (#16) Fichter & Jones 1953 (#2) Rees Heaton Collection; present disposition unk. Rees Heaton Collection; present disposition unk. To HM; to Hastings College; present disposition unknown To HM; to Hastings College; present disposition unknown To HM; to Hastings College; present disposition unknown Trapped Road kill To HM; to Hastings College; present disposition unknown Kearnev Public School X-mandible V. Bailey (BSC) J. S. Ligon (BSC) J. S. Ligon (BSC) C. P. Musgrave (BSC) Bailev 1932 Formerly confused as Santa Rosa, Guadalupe Co. , 1903, now believed to be misstated and is Roswell; BSC Gunnison's prairie dogs; skin measured Gunnison's prairie dogs Gunnison's prairie dogs 32 Table 6 continued. Great Basin Naturalist Memoirs No. Year Date Disposition County Site Skel- Sex eton Crania Skin 1918 Mar 22 USNM 230773 1925 Nov 14 YPM 1970 1929 Apr 7 ANSP 14509 1929 Dec 8 BSC 1210 1930 Aug 13 1934 Oct 20 1937 — KU 7146 USNM 251453 Unknown Catron Bernalli( Lincoln Colfax McKinley Magdalena, 75 mi SW Albuquerque, 12th St. Picacho, 5 km S Moreno Valley, Aqua Fria Santa Fe Santa Fe, 13 km SW McKinley Gallup Cibola El Moro National Monument Mexican Springs X X X X X X X X X X Destroyed Lovington, 17 km N North Dakota 1912 _ Unknown Morton Ft. Rice — 1913 Jun20 USNM 201945 Dunn Quinion between Killdeer & Medora F 1915 _ Unknown Mercer Stanton 1927 Aug NDSHS4173 Golden Valley Beach — 1933 Winter NDSHS5159 Slope Marmarth — 1951 Mar 5 NDSHS Hettinger Mott, 2kmS — 1954 _ NDSHS 13063 — — F 1954 Dec UM 103451 Sioux Morristown, S. Dak., lOkmN M found 1980 UND Billings SE — Oklahoma 1923 Sep USNM 243787 Cimarron — — 1924 Jul OKSU 9266 Texas Adams, 13 km SE F 1928 Jul 25 OU2211 Cleveland Norman, 2 km E — 1932 Winter Destroyed Texas 22 km S Kansas line — — — NSCM 858 Woods Hopeton — South Dakota 1889 — Unknown Shannon Pine Ridge Agency — early 1900s — MHM Pennington — M early 1900s — MHM Pennington — F 1905 Spring Unknown Hand T109N R70W S26, Bailey — 1905 Spring Unknown Hand 1905 Spring NYZP-AMNH 22894 Hand 1905 Spring NYZP Hand T109N R70W S26, Bailey T109N R70W S26, Bailey T109N R70W S26, Bailey X X X 1905 Winter Unknown Bennett Across line from Merriman, Neb. 1913 — WHO J23 Pennington Box Elder 1915 Aug USNM 209150 Mellette White River 1920 Oct 4 Destroyed Jackson Interior 1921 Feb Destroyed Pennington Scenic 1922 Mar Unknown Custer Wind Cave National Park 1923 Sep 16 USNM 243799 Shannon Pine Ridge 1923 Novl USNM 243990 Harding Covert 1924 — Unknown (n = 3) — — 1924 Sep Unknown _ — 1924 Oct Unknown (n == 6) — — 1925 — Unknown (n = 6) — — 1925 Mar Unknown (n -- 2) — — 1925 Jul Unknown — — 1925 Aug Unknown _ _ 1925 Sep Unknown _ _ 1925 Nov USNM 241014 Shannon Pine Ridge 1925 Dec 24 AMNH 70590 Pennington Scenic 1926 — Unknown — — 1927 — Unknown (n = 2) — — 1927 Mar Unknown (n = 2) — — 1927 Apr Unknown (n ^ 2) — — X X F X MX X 1986 Anderson et al.: Biogeography and Systematics 33 Other CoUecter Citation Meas- ured by us Remarks J. S. Felkner J. S. Ligon (BSC) W. Huber Aldous (BSC) Aldous 1940 X T. E. White X M. E. Musgrave (BSC) X J. Brewer Fortenbery 1971; "probable " Hubbard & Schmidt 1984 X W. E. Fair Halloran 1964; "highly probable" Hubbard & Schmidt 1984 X-mount J. Richardson "probable" Hubbard & Schmidt 1984 H. Eaton Bailey 1926 S. G. Jewett X X-mount Kellogg Bailey 1926 X-mount H. L. Rice X-mount J. H. Cramer X-mount X-mount X A. Freidt R. Crooke Gunnison's prairie dogs Gunnison's prairie dogs; skin measured Skin measured Skin measured; Gunnison's prairie dogs; kept captive 5 months Skin measured; Gunnison's prairie dogs Gunnison's prairie dogs Skin made but lost; drowned in pools; Gunnison's prairie dogs Road kill; fluid specimen made but lost Mount made but subsequendy destroyed To Ag College, Fargo BSC Received 13 Jun 1935 X-mount F. Barkley X-mount A. B. Baker X-mount H. Behrens X-mount H. Behrens X-mount X-mount X-mount X-mount R. A. Ward B. Darymple B. Darymple D. P. Stearns L. Knowles R. E. Lemley Hibbard 1934 Henderson etal. 1969 (#1) Moon 1905; Henderson et al. 1969 (#6) Henderson et al. 1969 (#6) Henderson et al. 1969 Henderson et al. 1969 (#6) Henderson et al. 1969 (#7) Henderson et al. 1969 (#10) Henderson et al. 1969 (#13) Henderson etal. 1969 (#14) Lovaas 1973 Linder et al. 1972 A DC records H. Behrens Collection H. Behrens Collection Sold alive Sold alive; probably sold to NYZP, rec'd 2 BFFs Oct 1905; further at least one of these became AMNH 22894, 1 Jun 1906 ADC records, trapped. ADC records, trapped ADC records, trapped Trapped by ADC and killed Captured by D. P. Stearns, BSC, Sep 1923 (Linder et al. 1972); to NZP (cat 1 1 . 281) 19 Sep; died 4 Nov 1925; to USNM Linder et al. 1972 Linder et al. 1972 Trapped ADC Linder et al. 1972 Trapped ADC Linder etal. 1972 Linder etal. 1972 Poisoned Linder et al. 1972 Trapped Linder et al. 1972 Trapped Linder et al. 1972 Trapped X Linder etal. 1972 X Linder etal. 1972 Linder et al. 1972 Trapped Linder et al. 1972 Trapped 34 Gf lEAT Basin N ATURALisT Memoirs N [0.8 Table 6 continued. Skel- Year Date Disposition County Site Sex eton Crania Skin 1927 Aug Unknown _ 1927 Sep AUG Pennington Rapid City, 29 km S F 1927 Winter WHOJ155 Pennington Conata — 1928 Feb Unknown — — — 1928 Mar Unknown (n = 5) — — — 1928 May Unknown — — — 1928 Aug 13 SDNHM 17538 — — F X 1928 Aug 13 SDNHM 17539 — — M X 1928 Aug 13 SDNHM 17540 — — M X X 1929 May 10 SDNHM 17537 — — M X 1931 Dec 31 UNZM 4451 Custer Hermosa M X X 1946 Nov ISU 33434 Lyman — M X 1950 Dec 8 USNM 285877 Dewey Isabel, 19 km S M X 1952 Oct 23 UMMNH3667 Perkins Zeona, N F X 1953 Dec USNM 287371 Pennington Conata Basin F X X X 1953 Aug USNM (no record) Haakon TIN R24E F 1953 Aug Unknown Haakon T1NR24E M 1953 Aug Released WCNP Haakon TIN R24E M 1953 Aug Released WCNP Haakon TIN R24E M 1953 Aug Released WCNP Haakon TIN R24E F 1954 Mar Unknown Stanley Midland, 24 km N M 1956 Summer Private Collection Lake Madison — X 1958 Jan Private Collection Ziebach Faith, 8 km SE M X 1958 Summer Destroyed Lyman Reliance, 3 km W M 1959 _ Destroyed Mellette White River, 16 km S M 1959 Fall SDGFP Sully Agar, 19 km W M 1960 Summer DMNH Washabaugh T41N R35W M 1960 Oct 22 USNM 348132 Sully Onida, 14 km W M 1961 Aug SDSUUO Lyman Reliance, 5 km N M 1963 Dec Destroyed Mellette T40N R30W, 24 km SW White River _ 1964 Sep SDSU 149 Tripp T40N R78W — X 1964 Oct 7 BNP Washabaugh T41N R37W, Wanblee M 1965 Aug SDSU 186 Mellette T40N R32W M X 1965 Sep 29 SDSU 187 Haakon T2N R19E F X 1965 Oct 10 SDSU 190 Jackson T2S R20E, 14 km W Kadoka M X 1966 Sep USNM 289498 Mellette White River F X 1966 Mar KUMNH 121795 Todd T38N R26W M X 1967 Apr 29 SDSU 212 Bennett T37N R39W F X 1967 May 16 SDSU 215 Jones T3S R29E F X 1967 Sep 12 SDSU 224 Mellette White River, 5 km W M 1971 Fall PAT Mellette — F 1971 Fall PAT Mellette — F 1971 Fall PAT Mellette — F 1971 Fall PAT Mellette — F 1971 Fall Private Collection Mellette — M 1971 Fall PAT Mellette _ M 1972 PAT Mellette — F 1973 PAT Mellette — M 1973 PAT Mellette _ F — — USNM 122620 Hughes Pierre (M) X Texas 1882 — Unknown Taylor Abilene — 1885 USNM 15018 Cooke Gainesville — X 1886 _ ROM 19-11-1-47 _ Rio Grande _ X X 1886 _ USNM 188459 Cooke _ M X found 1894 May USNM 65061 Childress Childress (M) X 1901 — ANSP 11842 Baylor Seymour F X X prior to 1902 Destroyed Crane Grand Falls 1902 Summer Unknown Lipscomb Lipscomb — X 1905 — Baylor Seymour 1986 Anderson et al.; Biogeography and Systematics 35 Other Meas- ured by us Remarks H. Behrens T. Bennett Linder et al . 1972 Trapped Linder et al .1972 Trapped Linder et al .1972 Trapped Linder et al . 1972 Trapped Linder et a . 1972 ADC capture specimen in tally from 1928 Linder et a . 1972 ADC capture specimen in tally from 1928 Linder et a .1972 ADC capture specimen in tally from 1928 Linder et a .1972 ADC capture specimen in tally from 1929 X F. M. Dille A. Lester R. Block X ADC X-mount A. Hinds X Garst 1954 G. Barnes Carcass frozen to USNM; no record. ADC G. Barnes Garst 1954 Died in captivity, skull only saved. ADC G. Barnes Garst 1954 ADC G. Barnes Garst 1954 Captive until Dec 1953. ADC G. Barnes Garst 1954 B. A. Nelson Henderson et al. 1969 (#56) Henderson et al. 1969 (#68) D. Capp Henderson et al. 1969 (#82) R. F. Wahlin Henderson et al. 1969 (#83) Road kill T. Johnson Henderson et al. 1969 (#90) Shot X-mount O. VonWald X-mount W. Allen Henderson et al. 1969 (#103) Through Glen Titus X-mount D. Badger & T. Lockwood X-mount C. F. Anderson Progulske 1969 Died in captivity Henderson et al. 1969 R. Adrian observed. Shot G. Johnson Shot X-mount Henderson et al. 1969 (# 155) Viscera at SDSU. Shot O. Huber Trapped R. Henderson Killed bv ranch dog R. Henderson Road kill X Carcass to SDSU. Shot W. Abbot Henderson et al. 1969 (#196) Listed in private collection in Henderson et al. 1969. Killed by dogs X-carcass J. Milk Road kill X-carcass D. Richardson Road kill J. Krogman Road kill X-carcass Juvenile, died of distemper vaccine X-carcass Juvenile, died of distemper vaccine X-carcass Juvenile, died of distemper vaccine X-carcass Carpenter, pers. comm. Juvenile, died 1971 of distemper vaccine X-carcass Carpenter, pers. comm. Juvenile, captive 6 years, died 1978; to Meeteetse Bank Apr 1982 X-carcass Carpenter, pers. comm. Captured as juvenile, captive 4 years, died 1976 X-carcass Carpenter, pers. comm. Captured as juvenile, died Oct 1978 X-carcass Carpenter, pers. comm. Adult, died Apr 1979 X-carcass Carpenter, pers. comm. Adult, died Jan 1979 F. J. Thompson G. H. Ragsdage Coues 1882 True 1885 Captured live; held at Cincinnati Zoo Bailey 1905 Bailey 1905 Captive Philadelphia Zoo 27 Apr 1901-2 Sep 1903; accession to ANSP 5 Jan 1904; Zool. Soc. Philadelphia Zool. Soc. Phila., received 14 Aug 1905, died 27 Nov 1905; accession ANSP 1905 36 Table 6 continued. Great Basin Naturalist Memoirs No. 8 Year Date Disposition County Site Skel- Sex eton Crania Skir 1905 Private Collection 1933 Dec UCM 5263 1934 Feb UCM 5287 1934 Dec 21 UM 76971 Baylor Lubbock Lubbock Lubbock Seymour Lubbock Lubbock Slide, 5 km SW M X M X M X 1937 Apr 21 UCB 77840 San Juan Blanding, 3 km S M Vyoming 1851 USNM Goshen Ft. Laramie X cal877 Unknown Laramie Cheyenne Depot — 1883 Dec USNM 13996/21066 Laramie Cheyenne, 19 km on Duck Creek — X 1895 May USNM 717.50 Weston Newcastle F X 1910 May USNM 168741 Weston Newcastle (M) X 1911 Spring USNM 180719 Crook Beulah M X 1911 Oct USNM 180718 Johnson Clear Creek, above B ig Red (Ucross) — X 1916 Apr USNM 211513 Niobrara Manville M X 1916-1928 Unknown (n = 10) — — — 1917 Sep USNM 227703 Converse Douglas M X 1924 Oct USNM 245641 Albanv Laramie, 8 km W M X X ca 1930 — Private Coll. (n - 2) Park — X 1935 Private Collection Sheridan Leiter, 22 km N M X 1939 Nov Private Collection Albany Eagle Park, W of Laramie Peak _ X cal946 — Unknown Sweetwater Wamsutter and Rock Springs — X 1950 Destroyed Albany Laramie, 2 km W 1955 — UW Albany Laramie — 1981 Sep 26 BSC 7934 Park Meeteetse M X X X 1982 Mar BSC (Biota #11) Park Meeteetse M X X X 1982 Spring BSC (Biota #1) Park T48NR102WS17 (F) X 1982 Winter BSC 10481 Park Meeteetse M X X X 1983 Aug WGF (Biota #16) Park Meeteetse — X X X 1983 Winter WGF (Biota #14) Park T48N R102W S18 M 1983 Oct BSC Park T48N R102W S4 F 1983 Dec USFWS Park T48N R102S S7 M X 1984 Winter WGF Park T48N R102W S8 F 1984 Sep BSC Park M X 1984 Sep BSC Park F found 1978 BSC Carbon T22N R77W S31, S Medicine Bow (M) X found 1978 Aug 15 BSC 4059 Uinta T16NR117WS7 (F) X found 1979 WGF Converse T41N R70W S32, Rosecrans X found 1979 BSC 4442 Unita T16NR118WS12 (F) X found 1979 BSC 4548 Carbon T23NR80WS18 (F) X found 1979 Sep 5 BSC 4547 Carbon T23NR81WS2. 13 ki n NE Hanna X found 1979 Sep 11 BSC 4441 Unita T16NR118WS1 _ X found 1979 BSC 4.342 (>arbon T23N R84W S34 (F) X found 1981 Aug 27 BSC 7558 Sweetwater T22N R93W S33, 19 1. :m N Wamsutter (F) X found 1982 Spring BSC (Biota # 10) Park Meeteetse (M) X X X found 1982 Mar 15 BSC (Biota #4) Park T48N R102W S8 (F) found 1982 Apr 20 BSC (Biota #5) Park T48N R102W S7 (M) X found 1982 Apr 23 BSC (Biota #6) Park T48NR102WS8 — X found 1982 Jun9 BSC (Biota #7) Park T49N R102W S31 (F) X found 1982 Jun 26 BSC (Biota #9) Park T48N R103W S2 (M) X found 1982 Aug 9 BSC (Biota #3) Park T48NR102WS7 (F) X found 1982 Sep 22 BSC (Biota #14) Park T48N R103W S2 (F) X found 1982 — WGF Park Meeteetse M X X X found 1982 — WGF Park Meeteetse M X X X found 1983 WGF (Biota #1.5) Park T48N R102W S7 _ found 1984 Apr 8 BSC Park Meeteetse found 1984 Spring WGF Park Meeteetse X 1986 Anderson etal.: BiogeographyandSystematics 37 Other Meas- ured by us Remarks D. Spencer D. Spencer D. White from ZSP. received 14 Aug 1905, died 28 Feb 1906 junnison s prairie dogs A. Culbertson Capt. J. Gillis J. Mason and C. Ruby F. Bond S. E. Piper udubon & Bachman 1851 Coues 1877 Destoyed by 1872 (Coues 1872). X-carcass X-carcass X-carcass X-carcass X-partial crania X-mandibles X-mandibles X-mandibles US Biological Survey P. Muchmore R. W. Fautin R. W. Fautin L. Hogg J. Renner L. Richardson T. W. Clark T. W. Clark M. Karl D. E. Biggins D. E. Biggins J. Hasbrouck D. E. Biggins V. Semonsen T. M. Campbell III J. Bridges V. Jameson D. Higgins S. Martin S. Martin L. Richardson T. W. Clark J. Grenier T. W, Clark S. C. Forrest L. Richardson L. Richardson L. Lee T. Thome T. Thome T. W. Clark T. Taylor B. Phillips Day & Nelson 1928 Clark 1975 Clark 1975 Clark 1975 Clark 1975 Clark and Campbell 1981 Martin & Schroeder 1978 Martin & Schroeder 1979 Martin & Schroeder 1979 Martin & Schroeder 1979 Martin & Schroeder 1979 Martin & Schroeder 1979 Killed 10 in predator trapping White-tailed prairie dogs Road kill; formerly in WGF collection; white- tailed prairie dogs Road kill Stolen; white-tailed prairie dogs X Hogg Ranch kill; white-tailed prairie dogs Road kill; white-tailed prairie dogs X White-tailed prairie dogs X Starved in burrow; white-tailed prairie dogs Juvenile; white-tailed prairie dogs Young of year; white-tailed prairie dogs X Killed by predator; young of year; white-tailed prairie dogs Adult, killed by predator; white-tailed prairie dogs Adult; white-tailed prairie dogs Partial cranium; killed by predator; juvenile; white-tailed prairie dogs Killed by predator; juvenile; white-tailed prairie dogs X Adult; white-tailed prairie dogs 1/2 skull X White-tailed prairie dogs X White-tailed prairie dogs 1/2 skull; white-tailed prairie dogs X White-tailed prairie dogs X White-tailed prairie dogs X White-tailed prairie dogs Full head (eagle); white-tailed prairie dogs X Adult; white-tailed prairie dogs X Adult; white-tailed prairie dogs X Subaduit; white-tailed prairie dogs X Adult; white-tailed prairie dogs X Subaduit; white-tailed prairie dogs X Adult; white-tailed prairie dogs X Adult; white-tailed prairie dogs Trap kill; white-tailed prairie dogs Trap kill; white-tailed prairie dogs White-tailed prairie dogs X White-tailed prairie dogs White-tailed prairie dogs 38 Great Basin Naturalist Memoirs No. Table 6 continued. Disposition Skel- Sex eton Crania Skin Saskatchewan 1924 Sep 30 SMNH 1588 — 1930 Feb NMC 11693 — 1931 Dec 20 SMNH 3183 — 1932 — NMC from SMNH — 2965 1932 Decl NMC 11703 — 1932 Dec 21 NMC 11700 — 1932 Dec 28 NMC 11752 — 1933 Jan NMC 11744 — 1933 Jan 6 SMNH 3168 — 1933 Apr SMNH 3186 — 1934 Nov 23 NMC 12682 — 1935 Nov 20 NMC 14078 — 1935 Nov NMC 14095 — 1935 Dec 5 NMC 14079 — 1935 Dec 4 SMNH 3656 — 1935 Dec SMNH 3657 — 1937 Dec 7 NMC 24235 — _ _ ROM 33-5-23-2 — — — SMNH 11441 — — — SMNH 11442 — Alberta 1901 May FMNH8207 — Additional Reports 1888 AMNH 2546 1903 AMNH 22820-NYZ 02699 1928 NYZ 02701 1928 NYZ 02700 1920s Syracuse (n -^ 3) 1934 ZSP 1934 ZSP prior to 1862 May MCZ 14947 BSC 4282 BSC 4283 USNM 35087 USNM 35088 — BMS FMNH AMNH 35041 ca 1877 USNM 11932 Missouri 1876 USNM 21965 Regina, 6 km SE Shaunavon Gergovia, S35 T2 R24 Big Beaver, S5 T2 R24 Frontier Shaunavon, 32 km SE Climax, S33T3R18 Shaunavon Climax Expanse, 8 km N Shaunavon Senate Wood Mountain South Fork, 19 km N Keeler, S22 T19 R29 Hazlet Climax, llkmN Maple Creek Gleichen Licks's River (F) M M M X (F) X X X X X X Colorado Fifty-four specimens of ferrets are listed from Colorado (Table 6, Fig. 11), including 47 specimens in museums and an additional 7 verified specimens whose present disposi- tions are unknown. Armstrong (1972) exam- ined 30 specimens from Colorado and listed an additional 27 records, many of which were sight records only. The earliest known verified specimen is AMNH 24412, collected in 1878 in El Paso County. Coues (1877), however, mentioned several accounts of black-footed ferrets from Colorado and had at least two occasions to examine specimens from there. One was a specimen in "defective" condition shot in "the valley of the Cache La Poudre River, near the northern border of Colorado" (Larimer 1986 Anderson etal.: BiogeographyandSystematics 39 Meas- ured by us Remarks C. Pickett H. F. Hughes H. F. Hughes J. Prochazka C. Guiguet W. Klvm H. F. Hughes X-mount X-mount F. Nevada C. B. Spangler Not in collection- Missing Missing Partial skull; "out of range." F. J. Thompson J. W. Munyon NYZ specimen (no ace. card); to AMNH 20 June 1888 Collected prior to Sep 1903; died Aug 1905; skin to AMNH 5 Aug 1905 Collected prior to 10 Apr 1928; died 6 Jul 1928; skin to AMNH (no record); gift of William J. Brunner Collected prior to 10 Apr 1928; died 8 Jul 1928; skin to AMNH (no record) Three specimens; no data, skins Received 1 Jun 1934; Urban J. Jones, Laureldale, Penn; died Jan 1939 Received 1 June 1934, Urban J. Jones, Laurel- dale, Penn.; died Jan 1939 In alcohol; lost No data No data National Zoo specimen (no ace. card) to USNM National Zoo specimen (no ace. card) to USNM; skeleton only No data No data; probably belongs to skull in SCZ, Munich No data Platte River, not in current records X "Out of range. County) presented to the USNM sometime between 1872 and 1877 by Dr. V. F. Hayden. This specimen is no longer in the USNM col- lection. Hayden told Coues that another fer- ret was kept in captivity for some time at Greeley. Coues also examined the collection of the pioneer naturalist Mrs. M. A. Maxwell of Boulder and verified several specimens taken "in the vicinity of Denver" at a centen- nial exhibition in Washington, D.C., in 1876 (Coues 1877). The disposition of Mrs. Maxwell's collection is unknown. Both Hayden and Maxwell "represented the spe- cies as being not at all rare." The most recent preserved specimen was obtained in Costilla County in 1946. Cahalane (1954) listed one specimen from Weld County in 1952, which was verified but subsequently 40 Great Basin Naturalist Memoirs No. Fig. 10. Black-footed ferret specimens from Arizona. Prairie dog distribution (shaded) after Cockrum (1960). destroyed. Eight of 10 of the most recent specimens (1940-1952) were collected west of the Front Range. One mandible was found in prairie dog colony searches in Logan County in 1977 (Bissell 1979), but, like many speci- mens found ejected from burrows by prairie dog digging activity, it may have been under- ground for an undetermined length of time before being brought to the surface. A record for Sedgewick County listed by Armstrong (1972) was found to be a sight record only and is not listed. Two specimens were found above 2800 m. One of these (UCM 10658) was found in asso- ciation with C. gunnisoni in Teller County at 2800 m. The other specimen (UCM 10660) was found drowned in Lake Moraine, eleva- tion 3125 m in El Paso County, far from any prairie dog colony. This specimen and an- other from Grand County (DMNH 653) were the only two specimens from Colorado not directly associated with prairie dogs and may have represented dispersing individuals. Three species of prairie dogs occur in Colo- rado: C. ludovicianus, C. leuctirus, and C. gunnisoni. Burnett (1918) estimated that the three combined species occupied 5,665,720 ha in the state in 1918. The area now occupied by prairie dogs in the state is unknown, but it is greatly reduced. Gilbert (1977) identified 10,843 ha of C. leucurus colonies in Rio Blanco and Moffat counties in 1977 and Bissell (1979) estimated 21,500 ha for 9 of 26 counties in C. ludovicianus range in the state in 1978. No estimate of C. gunnisoni distribution is available. Over 247,230 ha of C. gunnisoni - occupied colonies disappeared from 1945 to 1947 during epizootics of sylvatic plague (Armstrong 1972). Kansas Occurrence of M. nigripes in Kansas was reviewed by Choate et al. (1982). We list eight additional records, including one specimen from Decatur County and one specimen in the CU collection dated 1883 (Table 6). Addi- tional literature records include a mounted specimen from Wallace County examined by Coues (1877) supplied by L. H. Kerrick. About 1888 another ferret from Wallace County that had resided in the National Zo- ological Park was given by Kerrick to the USNM (12299/22929). These are obviously different specimens, but whether Kerrick was associated with NZP or was the collector of the Wallace County animals is unknown. The dis- position of several other animals residing at the NZP from 1905 to 1915 is also indicated in Table 6. Forty-eight specimens are known for the state (Fig. 12). Of 18 ferrets in Table 6 collected from 1877 to 1890, 15 were collected by A. B. Baker. Several museum labels listing Baker as the collecter also indicate the specimen was col- lected under the auspices of the BSC, but it is not known whether Baker was employed by BSC. Recent specimens include one collected by hand in 1957 in Sheridan County (Taylor 1961) and a skull and mandible of unknown age found on a prairie dog town in Gove County in 1978 (Boggess et al. 1980). Ferrets and prairie dogs historically occu- pied most of Kansas west of the Flint Hills (Fig. 12). However, prairie dogs that occu- pied an estimated 809,390 ha in Kansas in 1903 were reduced to some 14,570 ha (98% reduction) by 1973 (Choate et al. 1982). Choate et al. (1982) feel that ". . .the outlook is poor that the black-footed ferret will continue to occur in Kansas, if, indeed, any remain here now." 1986 Anderson ETAL.:BioGEOGRAPHY AND Systematics 41 Fig. COLORADO 11. Black-footed ferret specimens from Colorado. Prairie dog distribution (shaded) after Armstrong (1972). n Oi -TT - 1 1* T 7 • \ -T O-i -Or^tri- 5-0« i 2 i 0-2 ^e • W L ) ■~l #-3 C. ludovidanus r- 1 y ' • _. \ Fig. n. BlacHo„.edfe„a.,pedme„sf,o. Kansas. P„.ie dog d..ribuHo„ (shaded) after Choa.e e. al. (1982). 42 Great Basin Naturalist Memoirs No. Fig. 13. Black-footed ferret specimens from Montana. Prairie dog distribution (shaded) after Hall (1981). Montana Specimens of the black-footed ferret from Montana have not been described. Forty-four specimens are known from the state (Table 6, Fig. 13). Coues (1877) reported the earliest specimen (now lost) from the "Milk River." The most recent specimen was taken in Carter County in 1953. Thirty two (73%) of these ferrets come from seven counties in the south- eastern part of the state. An undated speci- men (USNM 13113/21976) lists the collection location near "Ft. Custer." Ft. Custer (in Bighorn County) was activated in 1877 and decommissioned in 1898, so it is assumed the specimen dates from that period. Prairie dogs (C. ludovicianus except for a small intrusion of C. leucurus in southern Carbon County [Flath 1979]), occupy the eastern two-thirds of the state except the three extreme northeast counties (Daniels, Roosevelt, Sheridan) north of the Missouri River (Hall 1981). Historic distribution of prairie dogs in the Burlington Northern Rail- road right-of-way showed extensive contigu- ous areas (Flath and Clark 1986). Federal pro- grams poisoned 2,832,860 ha of prairie dog and ground squirrel habitat in Montana in 1920 alone (Bell 1921). Vigorous prairie dog control efforts continued on a statewide basis until the 1950s, and in some counties areas of prairie dogs were reduced substantially (U.S. Bur. Land Mgmt. 1982). There is no estimate of the current total area occupied by prairie dogs in the state. In 1984 and T. M. Campbell and SCF found two separate remains, a black-footed ferret skull and a mandible (MDFWP 2344a and 2344b) on a prairie dog colony in Carter County, where ferrets reportedly had been observed in 1977 (Jobman and Anderson 1981). From the condition of the remains and the recent occupancy history by prairie dogs in the area, it was estimated that they were no more than 10 years old, supporting the 1977 sighting. Repeated searches in the area failed to produce other evidence or observations of living animals. Nebraska Black-footed ferrets from Nebraska were recorded by Fichter and Jones (1953) and Jones (1964). We list an additional six speci- mens, for a total of 23 from the state (Table 6, Fig. 14). The most recent specimen was a road kill from Dawson County in 1949. Additional information on two specimens is also available. A specimen mentioned by Coues (1877) and identified (USNM 14580) as coming from Nebraska has no date but should be about the time of the Coues report of 1877. 1986 Anderson ETAL: BioGEOGRAPHY AND Systematics 43 . ra — - i ^ -- ^ =1 T I> —HX-^ C. ludovicianus Vt 1 :j~~. \ _, • • • h \^ • ' --n NEBRASKA •« • /^ • ^, r » \ A • • > Fig. 14. Black-footed ferret specimens from Nebraska. Prairie dog distribution (shaded) after Jones (1964). The whereabouts of a second specimen (Fichter and Jones 1953 #10) was unknown, but it is recorded correctly in Jones (1964) as AMNH 121610. We include all of the Fichter and Jones (1953) list except number 8, from Fremont, Dodge County, which was a sec- ondary report. Prairie dogs (C. ludovicianus) probably were restricted historically to the "hard lands" described in Fichter and Jones (1953), which excludes much of the north central Sand Hills region. No estimate of their historic abun- dance is available, but they probably were found in great numbers along the many tribu- taries of the Platte and Niobrara rivers. Lock (1973) estimated only 6070 ha of prairie dog colonies remained statewide in 1971. New Mexico Status and history of the black-footed ferret in New Mexico are described in detail by Hubbard and Schmitt (1984). We include three records of "unsubstantiated" specimens in our list that are treated separately by them and described as "probable" or "highly proba- ble. ' Because existence of these specimens is documented elsewhere, we include them here but concur with Hubbard and Schmitt (1984) that some question exists as to their validity. We also agree that a ferret mandible noted in Bailey (1926) as being found in Santa Fig. 15. Black-footed ferret specimens from New Mex- ico. Prairie dog distribution (shaded) after Hubbard and Schmitt (1984). Rosa, Guadalupe County, in 1903 is the mandible catalogued in the USNM from Roswell, Chaves County, 1899 by Bailey. We list 10 extant specimens and three dis- puted specimens (Table 6, Fig. 15). The last verified specimen was taken in 1934 in 44 Great Basin Naturalist Memoirs No. NORTH DAKOTA Fig. 16. Black-footed ferret specimens from North Dakota. Prairie dog distribution (shaded) after Hall (1981). McKinley County. Hubbard and Schmitt (1954) described the substantial role of BSC trappers in the collection of ferret specimens in the state. Cynomys ludovicianus is found in the south- ern and eastern parts of the state, and C. gunnisoni is found at higher elevations in the northwest. Prairie dog area in the state de- clined from an estimated 4,856,333 ha in 1919 to less than 202,350 ha in 1979-1981 (Hub- bard and Schmitt 1984). Hubbard and Schmitt "assume the ferret is still a member of the state's fauna and that it could occur any- where that prairie dogs occur. " North Dakota No account of ferret specimens for North Dakota is available other than Bailey (1926). We located nine specimens, all collected west of the Missouri River (Table 6, Fig. 16). Re- cent specimens include one found in 1954 in Sioux County and a skull found in 1980 in southeastern Billings County. Teddy Roosevelt described ferrets found near his ranch in western North Dakota in the late 1800s as "that rather rare weasel-like ani- mal ... I have known one to fairly depopulate a prairie-dog town, it being the arch-foe of these httle rodents" (Seton 1929: 571). Little is known of former prairie dog (C. ludovicianus) distribution, although there were likely prairie dogs found east and north of the Missouri River. In 1920, 2,428,166 ha were treated with poisons for prairie dogs and ground squirrels in North Dakota (Bell 1921). Grondahl (1973) estimated only 2740 ha of prairie dogs remained by 1973, all west of the river. Seabloom et al. (1980:) ". . . regard sightings (of black-footed ferrets) as repre- senting transients rather than a viable resi- dent population " and cite the paucity of prairie dogs remaining in the southwestern part of the state. Oklahoma Lewis and Hassein (1973) listed recent fer- ret specimens and sightings for Oklahoma. Only four specimens are known, with one additional literature reference (Table 6, Fig. 17). A specimen was collected in Cleveland County in 1928, and Hibbard (1934) reported a ferret taken in Texas County in 1932. Cyno- mys ludovicianus probably occupied "mil- lions ' of hectares in Oklahoma at the turn of the century, including one colony 35 km long in tall grass prairie between Kingfisher Creek and El Reno (Lewis and Hassein 1973), but only 3845 ha remained in 1968 (Tyler 1968). 1986 Anderson etal.: Bioceocraphy and Systematics 45 O-i OKLAHOMA Fig. 17. Black-footed ferret specimens from Oklahoma. Prairie dog distribution (shaded) after Hall (1981). Black-footed ferrets were considered extir- pated in Oklahoma as of September 1980 by the U.S. Fish and Wildlife Service, Albu- querque, New Mexico (Jobman and Anderson 1981). South Dakota A detailed description of ferret distribution and occurrence is available for South Dakota, where ferrets were studied in Mellette and adjacent counties from 1964 to 1974. Hender- son et al. (1969) described ferret specimens and sight reports for South Dakota from 1889 to 1967, and additional records were dis- cussed in Linder et al. 1972. Table 6 includes an additional 15 specimens not in those ac- counts. Ninety-nine specimens are reported, with 57 specimens destroyed or of unknown disposition (Fig. 18). Additional notes were also made for the fol- lowing specimens: Moon (1905) noted a "pair sold alive" (Hen- derson et al. 1969: #6). The New York Zoolog- ical Park listed two arrivals of M. nigripes in October 1905, but no accession card was made to verify this transaction. A specimen that came from NYZP to AMNH (22894) on 1 June 1906 with no data was undoubtedly one of these animals. Disposition of the second ani- mal is unknown. We therefore list AMNH 22894 as coming from this source. D. P. Stearns, ESC, captured one ferret alive near Pine Ridge, Shannon County, on 16 September 1923. This is undoubtedly the fer- ret trapped by BSC in Pine Ridge September 1923 reported by Linder et al. (1972). This animal was sent to the NZP (11281), where it lived until 4 November 1925. It was subse- quently catalogued into the USNM (243799). Linder et al. (1972) listed 43 ferrets taken by BSC from 1924 to 1929. Table 6 lists 8 known specimens from that period. Four specimens in the SDMNH were taken in South Dakota during this period and correspond to the 3 specimens taken in 1928 not identified by month in the Linder et al. (1972) list and the 1 specimen from 1929. Therefore we have de- ducted them from the Linder et al. (1972) tally for those years. The remaining 4 specimens from that period may also have been collected by BSC, but insufficient data are available on the collectors to verify this. Both tallies are therefore included. Rose (1973) briefly discussed the history of prairie dogs in South Dakota. Tovras 24-32 km long were common in major drainages. H. R. Wells estimated 710,935 ha of prairie dogs in the state in 1923 (Linder et al. 1972). In 1968 BSFW estimated 24,281 ha in the state, a reduction of 96% (Rose 1973). Linder et al. (1972) presented data showing 405,000 ha were poisoned by various government agen- 46 Great Basin Natur\list Memoirs No. 8 h % ^_ - I [^ - «*. — . __ •-4 C. ludovicianus r^ 1— b— _ / P^ ?*'*' Ci * ^ • 1 • 9 ^- -^ \ ' — 1 — — 1 — H 1 •-3 • • \;> ^1 ^ ■^ J~M.i SOUTH DAKOTA Fig. 18. Black-footed ferret specimens from South Dakota. Prairie dog distribution (shaded) after Hall (1981). cies between 1932 and 1939. Three counties within the Pine Ridge Indian Reservation (Shannon, Jackson, Bennet) had recovered prairie dog populations occupying in excess of 120,000 ha in 1984 (R. Crete, personal com- munication). Texas The distribution of ferrets in Texas has not been described. We have established 13 veri- fied records for the state and have located nine extant specimens (Table 6, Fig. 19). Specimen USNM 15018 is similar in all respects to a specimen described by True (1895) and is listed as such. Four specimens were taken for zoos. Two specimens from Gainesville are slightly out of current range, but may have been within historic range, or Gainesville may have been chosen by the collecter as the nearest identifiable landmark. The occur- rence of ferrets in trans-Pecos Texas has been questioned (Schmidly 1977), even though it is highly likely they occurred there. None of these specimens expand the known range in the state. Bailey (1905) estimated that prairie dogs (C. ludovicianus) occupied 233,100 sq km and Fig. 19. Black-footed ferret specimens from Texas. Prairie dog distribution (shaded) after Cheatheam (1977). noted one town in the Panhandle of 6,475,111 ha (400 X 160 km). A statewide survey in 1976 showed 36,432 ha of prairie dogs, which were nowhere in great densitv (Cheatheam 1977). The U.S. Fish and Wildlife Service, Albu- (juercjue. New Mexico, considers the black- footed ferret extirpated in Texas (Jobman and Anderson 1981). 1986 Anderson et al.: Biogeography and Systematics 47 Fig. 20. Black-footed ferret specimens from Utah. Prairie dog distribution (shaded) after Durrant (1952). Utah Only one specimen is known for Utah (Dur- rant 1952), found in 1937 south of Blanding, San Juan County (Table 6, Fig. 20). Three species of prairie dogs are found in Utah: C. leucurus, C. gunnisoni, and the endemic Utah prairie dog, C. parvidens. Cynomys parvidens is geographically disjunct and there is no evidence to suggest that M. nigripes has ever occurred with this species. Wyoming Black-footed ferret reports from Wyoming have been discussed in Clark (1980) and Clark and Campbell (1981), including an additional 126 sight records not listed here. In all, 60 ferret remains are known from 1851 to 1984, and 24 of these come from the Meeteetse area where the known population is currently un- der study (Table 6, Fig. 21). Five ferrets hsted in Clark and Campbell (1981) were actually from South Dakota (Garst 1954). Ferrets WYOMING Fig. 21 . Black-footed ferret specimens from Wyoming. Prairie dog distribution (shaded) after T. W. Clark, personal communication. 48 Great Basin Naturalist Memoirs No. ^ C. ludovicianus ^>> MOHT.tU.DAK. SASKATCHEWAN Fig. 22. Black-footed ferret specimens from Saskatchewan. Estimated extent of prairie dogs shown by dotted Hne. range farther west in the state than previously reported by Hall (1981). Ferrets occurred throughout Wyoming, ex- cept the mountainous northwestern corner, in association with C. ludovicianus in the east and C. leucuriis in the west. Between 1915 and 1923, 1,120,290 ha were poisoned for prairie dogs (Martley 1954). An additional 445,080 ha were poisoned fiom 1923 to 1928 in Niobrara, Weston, and Campbell comities only, including one colony 160 km long from Indian Creek to Campbell County line (Day and Nelson 1929). Cheyenne, Wyoming, was built on the site of a large old colony (Day and Nelson 1929), where a ferret specimen was collected in 1877 (Coues 1877). Fragmentary records of prairie dog poisoning show that prairie dogs have been reduced by at least 75% since 1915 (Clark 1973). Clark et al. (1985) estimated that about 6,000 prairie dog colonies (ca 90,000 ha) still exist in Wyoming, but most are small and contain low densities of prairie dogs. Saskatchewan Twenty-one specimens were located in one U.S. and four Canadian museums (Table 6). All of these specimens were collected in southern Saskatchewan with the exception of FMNH 8207, from Gleichen, Alberta (not mapped). Gleichen is several hundred kilo- meters out of present prairie dog range and is also disjunct from the next closest record of black-footed ferret in Saskatchewan. Because we have no other evidence to support ferret occmrence or recent prairie dog occurrence at that latitude at this time, we regard this record as spurious. It is possible the skin was picked up in fur shipments from another loca- tion and subse(juently sold to FMNH. Prairie dogs were not reportcxl from Canada until" 1927 (Soper 1938, 1944, 1946) and then only in the vicinit)' of Climax and \'al Marie in extreme southwestern Saskatchewan. Ferret specimens were taken from 1924 to 1937 over a greater geographical area (Fig. 22). Prairie dogs may have been distributed at low densi- 1986 Anderson et al.: Bkk.eocraphy and Systematics 49 ties or were expanding throughout southern Saskatchewan and Alberta at that time and were not recorded in l)iological surveys. Ground-dwelhng rodents that might provide ferret habitat (with the exception of Sper- mophilus richardsonii) are absent in the area of ferret specimen distribution. Woodchuck {Marmota monax) and Frankhn's ground squirrel (S. franklinii) are typically found at the eastern range ofCijnomys in the continen- tal U.S. and are found much farther north in Canada than the known distribution of prairie dogs (Hall 1981). Rather than imply an alter- nate habitat for the black-footed ferret in Canada, the distribution of ferret specimens more likely suggests the former range of Cynomys. The fossil history of Cymmiys in Alberta goes back at least one million years. At Medicine Hat, Cynomys spp. has been found in Wisconsin-age deposits and C. leucunis has been identified in the Sangamonian and mid- dle Wisconsinan faunas (Stalker et al. 1982). Cynomys leiicurus has been at found at Janu- ary Cave (late Wisconsin, J. Burns personal communication), and C. ludovicianus was recognized in the Hand Hills fauna (Storer 1975), although this identification has been questioned (J. Burns, personal communica- tion). Cynomys has not been reported from any Pleistocene fauna in Saskatchewan. It is possible that intensive agriculture in the Prairie Provinces eliminated prairie dogs in many places before they could be recorded. Prairie dogs totaled only 503 ha in 1971 and are currently found only near Val Marie (Ker- win and Scheelhaase 1971). Ferrets are con- sidered extirpated in Canada by the Commit- tee on the Status of Endangered Wildlife in Canada, 1978 (Thornback and Jenkins 1982). Additional reports Sixteen additional specimens are catalogued in museums with little or no identifying data (Table 6). Some of these may be the speci- mens that are "unknowns" from other loca- tions. Some dates of acquisition can be guessed from catalogue numbers, but this is not reliable. MCZ 14947, labeled as received in 1862, was collected by F. J. Thompson, who was the collector of record for Abilene, Taylor County, Texas, in 1882 (Coues 1882). This specimen was lost and may have been misla- beled in the MCZ collection. Five of the spec- imens in this group were collected for zoos. Along with the Canadian evidence, additional reports outside the range of Cynomys are specimen USNM 21965 listed "from" Licks River, Missouri, and a note by Ames (1874) listing "P. nigripes", with no evidence, in the fauna of Minnesota. However, these reports are far from potential range as determined by prairie dog distribution, and we conclude they are erroneously placed as originating in these locations. In the case of the Missouri account, for example, the specimen could have been taken on the Kansas plains and subsequently ascribed by the collector to his home location. No place name for Lick's River in Missouri could be found. Summary We list 412 specimens in Table 6. The cur- rent deposition is known for 310 of them. The largest number of state records (99) and extant specimens (50) are from South Dakota. Twenty-one specimens are noted from Can- ada. Only 6 specimens were collected outside of known prairie dog range, although the asso- ciation of some of the Canadian specimens is uncertain. Of the 412 records, at least 103 (25%) were taken by federal predator and ro- dent control agents. The number taken by museum collectors is unknown but probably is also significant. At least 41 animals (10%) were captured alive and held by individuals or zoos. Specimens collected by year are given in Figure 23 (n = 318). The highest collection fig- ures date from the 1920s. This peak corre- sponds to the period in which the BSFW was entering numerous agreements with state ex- tension services in the West to control prairie dogs and carrying out large-scale poisoning campaigns (Day and Nelson 1929, Linder et al. 1972, Hubbard and Schmitt 1984). Be- cause ferrets never have been of economic value, many specimens that were taken up to this time probably were destroyed and never reported. Elsewhere, changing land use sig- nificantly reduced potential ferret habitat and contributed to ferret decline. In several east- ern prairie states (Kansas, Nebraska, Okla- homa, Texas), 65% of all specimens collected date prior to 1910. The early demise of ferrets in these states is probably directly attrib- 50 Great Basin Naturalist Memoirs No. 1880 1910 1920 1930 1940 1950 I960 1970 1980 UJ tiU ? 16 a. en 12 L. O 8 or UJ 4 ^ 3 T 1 1 1 1 1 1 1 1 1 1880 1890 1900 1910 1920 1930 1940 1950 I960 1970 Fig. 23. Collection history of black-footed ferret .specimen.s by county and by year (1880-1980). 1980 utable to the expansion of population and culti- vation into areas formerly occupied by prairie dogs. In the 1950s disappearance of large areas occu- pied by prairie dogs stimulated interest in locat- ing all possible remaining ferrets, but despite the more detailed accounting of sight and speci- men reports (e.g., Cahalane 1954), specimen re- ports continued to decline, including the num- ber of counties reporting specimens (Fig. 17). By the 1970s the number of known populations had dwindled to one, although in retrospect the pop- ulation at Meeteetse was certainly extant as well as some individuals in Carter County, Montana. At the present time only the Meeteetse popula- tion is known. Plotting locality records of ferrets with the ranges of three species of prairie dogs shows 83.0% are from C. ludovicianus range, 11.2% are from C. leucurus range, and 5.8% are from C. ^unnisoni range. Our estimates of prairie dog abundance show that 41,900,000 ha of rangeland may have been occupied by all species of prairie dogs in the early part of the 190()s. Nelson (1919) estimated 40,469,500 ha. Current areas occu- pied in all the western states and provinces is unknown but is greatly reduced, perhaps by as much as 90%. Presently, ferrets are consid- ered extirpated by U.S. and Canadian wildlife officials in Canada, Oklahoma, and Texas. Be- cause of low prairie dog numbers, the likeli- hood of persistence of black-footed ferrets in Arizona, Kansas, Nebraska, and North Da- kota is also poor. Ferrets may persist in the remaining states of its former range, but they are probably restricted to small, isolated pop- ulations. Specimen collection by month is plotted in Figure 24. Seasonal changes in collection re- turns likely reflect phases in ferret life history and trapping efforts. Since trapping for most furbearers reaches its peak in midwinter, high returns are to be expected, particularly where ferrets are caught accidentally in sets for other animals. It is interesting to note that the peak month of collection is October, the time at which most newly independent young ferrets are dis- persing (D. E. Biggins, personal conununication cited in Forrest et al.. Black-footed ferret habi- tat, 1985). The lowest specimen count occurs in June, when females with young remain for long periods imderground. Several authors have commented on the bias toward males in capture data for mus- 1986 Anderson etal.: Bioceocraphy andSystematics 51 20 - AUG SEP OCT NOV DEC Fig. 24. Collection of black-footed ferret specimens (n = 234) by month based on all records (1851-1984). telids (King 1975). The 200 ferret specimens of known sex in Table 3 (137 males and 67 females) show a sex ratio of 2.04M:1F. Since sex ratios at birth are 1:1 (Forrest et al., Life history characteristics , 1985), it seems likely that this collection bias is similar to trap biases seen for other mustelids, and is not a result of a skewed adult sex ratio. Trap biases in mustelids are a result of males having larger activity areas and longer movements (and therefore more encounters with traps or haz- ards) and being less trap-shy (King 1975, Pow- ell 1979). Black-footed ferret males have larger activity areas (Biggins et al. 1985, Richardson et al. in preparation), which fur- ther supports this theory. MORPHOMETRIC VARIATION Sexual Dimorphism Adult females averaged 93% of male body length for both museum- and field-measured groups and were 68% of males in body weight. Skull length for females averages 93% of that of males based on CBL. Five variables were chosen by stepwise maximizing of Wilks' lambda for cranial mea- surements as the best discriminators of sex: CBL, LC, POC, INB, and WM'. The results of the cranial discriminant analysis produced excellent discrimination between classes (Fig. 25). Coefficients for known specimens not Males «r i\ Jul Females DISCRIMINANT SCORE Fig. 25. Histogram of discriminant scores trom a dis- criminant analysis between sexes of black-footed ferrets. used in classification indicated that only 2.0% of males and 8.3% of females were misidenti- fied on this basis, or that grouped cases were correctly classified 95.9% of the time. Be- cause in many cases only mandibles may be found (particularly with fossil material), a sec- ond analysis using only mandibular variables 52 Great Basin Naturalist Memoirs No. 8 ^WYOMING- SOUTH DAKOTA-NEBRASKA £ / ^^ \ — — ' / ^ ^\\i 1 1 NEW MEXICO-TEXAS \ f , . ] \ ^ i '-^^ / V-'ISONTANA- i- ^^V,^__^ / NORTH DAKOTA COLORADO - KANSAS / CANONICAL DISCRIMINANT FUNCTION Fig. 26. Convex polygons containing the values for specimens of male M. nigripes from four localities (north- south) for their first two discriminant axes. MONTANA- NORTH DAKOTA \. ^Cr^^^^^^^ 1 \ • \ /^ 1 \ \ / ^"^"^-^ \ \ COLORADO-\ "-^T r-\. \ \ KANSAS \ y^ ^^ r~«»^ \ "--/^A WYOMING -SOUTH _^ DAKOTA- NEBRASKA "~~'''~-*^-.~- NEW MEXICO-ARIZONA- TEXA? CANONICAL DISCRIMINANT FUNCTION Fig. 27. Convex polygons containing the values for specimens of female M. nigripes from four localities (north-south) for their first two discriminant axes. was made. Four mandibular variables (LJAW, DP3.4, WMjtal, and LMjtr) were chosen by maximization of Wilks' lambda. Mandible measures are not as good as cranial measures as discriminators of sex, with 8.2% of males and 16.7% of females correctly classified (cor- rect classification 89.4% of the time), but they can be used when crania are not available or cannot be classified. To assign sex to crania and mandibles a deci- sion is made based on the following equations derived from the discriminant analysis. For crania: A = 19.954(CBL) + 14.129(INB) + 0.373(LC) + 24.790 (POC) - 29.706(WM') - 877.213. B = 18.936(CBL) + I1.681(INB) - 5.634(LC) + 23.099 (POC) - 23.600(WM') - 758.034. For idibk 93.278(WM,tal) A = 10.279(LJAW) + 13.849(DP3 + 65.965(LM ,tr) - 591.915. B = 9.435(LJAW) + 11.416(DP3.4) + 85.110(WM,tal) + 63.392(LM,tr) - 502.990. If A > B, then the skull is from a male. If B > A, then the skull is from a female. If the absolute difference between A and B is greater than 2.80, then P > .05 that the skull has been correctly classified. If A ^ B or the difference between A and B is less than 0.50, then the probability of correct classification is less than 60%, and no determination can be made as to sex. Geographic Variation Linear discriminant analysis was performed on 29 cranial measurements of 50 males and 31 females from four geographic regions that correspond roughly to four latitudinal gradi- ents arranged from north to south (Fig. 1). These regions were: Montana-North Dakota; Wyoming- South Dakota-Nebraska; Colo- rado-Kansas; New Mexico-Texas-Arizona. The characters best separating males by re- gion were: CBL, DF3.,, LP', WM\ and PBC- C. Characters best separating females by re- gion were: CBL, DP3.,, LP', WM\ PBC-C, and WP^pc. Discriminant scores and group centroids for the first two discriminant axes are shown in Figure 26 (males) and Figure 27 (females). There is evidence in this analysis of a north-south cline for both sexes, although overlap between clinal groups can be seen (Table 7). The extreme southern region (New Mexico-Arizona-Texas), overlaps the Colo- rado-Kansas region for males and appears sep- arated on the axis for canonical function 2 for females. The southern group is included for completeness despite the obvious violation of multivariate assumptions caused by ex- tremely small sample sizes. The present ori- entation of centroids is little affected by this region because of these small sample sizes. This partially explains high misclassification for both males and females in this group. Out- comes of discriminant classification in Table 7 1986 Anderson ETAL.; Biogeography and System atics 53 Table 7. Discriminant classification of male and female black-footed ferrets from four localities showing number of members from each location correctly classified. Number of cases Predicted group membership Actual group 1 2 3 4 Females: 72.73% of "grouped'cases correctly classified 1 (Montana, North Dakota) 8 6(75.0%) 2(25.0%) 2 (Wyoming. South Dakota, 7 1 (14.3%) 5(71.4%) 1 (14.3%) Nebraska) 3 (Colorado, Kansas) 16 3(18.8%) 11(68.8%) 2(12.5%) 4 (New Mexico, Arizona, 2 2(100.0%) Texas) Males: 56.86% of "grouped" cases correctly cl issified 1 (Montana, North Dakota) 8 7(87.5%) 1 (12.5%) 2 (Wyoming, South Dakota) 8 1 (12.5%) 4(50.0%) 1 (12.5%) 2(25.0%) 3 (Colorado, Kansas) 31 2(6.5%) 5(16.1%) 17 (54.8%) 7 (22.6%) 4 (New Mexico, Arizona, 4 2(50.0%) 1 (25.0%) 1 (25.0%) Texas) show higher overlap of nearer chnal groups and httle or no overlap as clinal groups be- come farther apart. Further analysis reveals that the source of this variation is primarily found in differences in size from north to south and not in changes in relationships of variables to each other. This is indicated by Figure 28 for CBL, which shows significantly larger measurements be- tween northern and southern groups for both males and females (males: F = 4.3, 44 df, P = ,04; females: F = 5. 1, 25 df, P = .03). ANOVA between the two northernmost and two southernmost groups showed significant dif- ferences in 17 of the 29 variables tested, with larger measurements from the northern group. Prey Species Variation Discriminant analysis was also used to test for differences between ferret specimens as- sociated with different prairie dog species. The subgenus Leucocrossuromys includes C. gunnisoni, C. leucurus, and C. parvidens. The subgenus Cynomys includes C. liidovi- cianus and C. mexicanus (Mexican prairie dog). Leucocrossuromys is considered more like ancestral Spermophilus than Cynomys; shows a less interactive social organization, organized around clans; and has a short white- tipped tail, a less massive skull, and smaller and less expanded cheek teeth (Clark 1973a). Subgenus Cynomys has a longer black-tipped tail, distinct reddish-cinnamon pelage in sum- mer, and a more complex social organization than Leucocrossuromys, organized around Males \ l\ — \ — 1 II — 1 — 1 H 1 Females MONTANA WYOMING COLORADO N.MEXICO N.DAKOTA S.DAKOTA KANSAS ARIZONA NEBRASKA TEXAS Fig. 28. Latitudinal differences in CBL for male and female black-footed ferrets showing a north-south cline in coteries (King 1955, Hoogland 1981). The cur- rent distributions of the two subgenera show an elevational and longitudinal cline, since the white-tailed species is found at higher elevations along the western portion of prairie dog range. Since the subgenera occur at dif- 54 Great Basin Naturalist Memoirs No. A A Group centroids DISCRIMINANT SCORE Fig. 29. Histogram of discriminant scores from a discriminant analysis between M. nigripes specimens taken from black-tailed prairie dog range (dark shading) and white-tailed prairie dog range (light shading). nignpes eversmonni putorius frenata erminea nivalis vison Fig. 30. Single linkage dendrogram using generalized distances between species centroids based on a consensus ofMustela males and females (after Youngman 1982). ferent densities, have different behavior pat- terns, and are geographically separated, it might be expected that ferret differentiation may have evolved with each subgenus of prairie dog, which could be reflected in mor- phological differences. For example, prairie dogs show similar external sizes and dimen- sions among species (Hall 1981) but may differ along similar latitudinal gradients. Size differ- ences could be reflected in the size of burrow openings used and weight of the animal, which could in turn affect the size or confor- mation of ferrets found with them. Because of north-south clinal variation and sexual variation among ferrets, comparisons were made only between male ferrets from the range of black-tailed prairie dog (n=^5()) and ferrets from the range of white-tailed prairie dog (n = 15) subgenera from the two regions closest to the geographic center of ferret range (Wyoming-South Dakota-Ne- braska and Colorado-Kansas). Although six variables (INB, WBC, WMjtr, LC-M\ WC, and WMjtal) were chosen by stepwise maxi- mizing of Wilk's lambda, which discriminated between white-tailed and black-tailed prey groups, only 53.3% of the white-tailed group were placed correctly in that category, indi- cating a high degree of overlap between groups (Fig. 29). This analysis suggests that no morphometric variation in black-footed fer- rets occurs based on the species of prairie dog they are found to associate with. However, other differences may exist that involve eco- logical or behavioral characteristics that could taxonomically separate these groups but are not reflected in morphometric analyses. Ferrets and Their Relatives The genus Miistela includes weasels (sub- genus Mustela), mink (subgenera Lutreola and Vison, see Youngman 1982), ferrets and polecats, (subgenus Putorius, European workers often use Putorius as a generic name), and South American weasels (subgenus Grammogale). "Ferret" and "polecat" are interchangeable common names, though polecat is generally used for the Old World species. Based on single linkage dendrograms derived from morphometric variables, Youngman (1982) suggested that the polecats M. putorius, M. everstiuinni , and M. ni- gripes, form a natural group distinct from the weasels and M. vison. Figurt^ 30 shows the phylogenetic relationships of some of the spe- cies in this group. These highly efficient small carnivores range in size from the tiny least weasel (M. nivalis rixosa , wt 38-63 gm), the smallest living carnivore, to the Siberian or steppe polecat (M. eversmanni, wt to 2050 1986 Anderson et al. : Biogeography and Systematics 55 Fig. 31. Photograph.s of A/, nitiripcs (A), M. visou (B), M. eversmanni (C), and M. ptitorius furo (D). gm). All of them have a long lithe body, short legs, a long low braincase, and short powerful jaws equipped with elongated bladelike car- nassials (P , Mi), sharp canines, and three premolars in each jaw half (two in the lower jaw of the South American weasel, M. africana). Primarily Holarctic in distribution, weasels and ferrets are terrestrial and mink semiaquatic. About 15 extant species are rec- ognized. Of these, only four, M. nigripes, M. evers- manni, M. putorius, and M. vison, concern us here. Table 8 compares the four species and Figure 31 illustrates them. Although mink and ferrets differ markedly from each other in appearance, their skulls and teeth are similar and can be confused. Table 9 shows some differences between them. The domestic ferret (M. putorius furo) was bred in captivity as early as the fourth century B.C. for use in controHing rodents and driving rabbits from their burrows (Nowak and Par- adiso 1983). It is also kept as a pet. Leonardo da Vinci's famous painting "The Lady with a Weasel" actually depicts the domestic fer- ret (Kowalski 1976). Its distribution is now worldwide in captivity. Coloration is gener- ally pale yellow or whitish (often albino) with no black or dark markings. Escaped domestic ferrets have been mistaken for black-footed 56 Great Basin Naturalist Memoirs No. 8 Table 8. Comparisons between Mustcla nigripes, M. eversmanni, M. ptitoriits, and M M. nigripes M. eversmanni M. ptitorius M. vison Geologic range Early late Mid-Pleistocene- Mid-Pleistocene- Mid-Pleistocene- Pleistocene-Recent Recent Recent Recent Geographic range Formerly S Canada, Steppes of Eurasia, Europe E to Ural North America except Great Plains to NW S to central Asia, Mountains arid areas. Introduced Texas, SW U.S. NE China in Europe Habitat Prairies, mountain basins, semiarid grasslands Open grasslands Open forests, meadows, clearings Along streams, marshes External characters Upper parts yellowish Yellowish to pale Dark brown to black. Rich dark brown, white buff. Feet black. brown. Dark feet. belly dark, silvery chin patch, tail slightly black mask across dark mask across between eyes and bushy eyes, tail tip black eyes, terminal 1/.3 tail dark ears, tail entirelv dark Size 6 TL 490-615, T S TL 450-740, T 6 TL 465-650, T 3 TL 510-,570, T 180- 107-148, Wt 915- 80-18.3, Wt to 2050 g. 115-190, Wt. 500- 2.30, Wt 680-1360 g. 1125 g. 9TL479- 9 TL 360-700, T 70- 1.500 g. 9TL.37,5- 9 TL 4.30-560, T 130- 565, T 109-141. Wt 180, Wt to 1.3.50 g. 465, T 8.5- 125, Wt 200, Wt. 565- 1089 g. 645-850 g. to 1360 g. Food Cynornys, rodents, Pikas, susliks, voles. Mice, toads and Aquatic mammals. lagomorphs hamsters, marmots frogs, birds birds, frogs, fish, crayfish Habits Mostly nocturnal, Nocturnal. Live in Nocturnal, solitary. Nocturnal, solitary. solitary. Closely rodent burrows. Often found around Den along streams associated with Avoids contacts with barns, dwellings Cynomys man Reproduction Gestation period 42- Gestation period 38- Gestation period 40- Gestation period 40-91 45 days. 3-5 young 41 days. 4-9 young 43 days. 4-6 young days. Short delayed born May-June born April- May born May-June implantation. 2-6 young born April- May Remarks Endangered species. Striking resemblance Fin- valuable (fitch). Fur valuable. Raised Closely related to toM. nigripes. Subspecies M.p. on fur farms M. eversmanni. Hunting of M. e. prohibited in Siberia /wro domesticated, used in hunting and as pets Table 9. Comparisons between Mustela nigripes I eversmanni and M. visun (after Anderson 1977). Variant M. nigripes /eversmanni M. vison Palate Basiocciput Basicranium Auditory bullae Auditory meatus Mastoid bullae Infraorbital foramen Jugal Frontals Canines, upper and lower p3 P^ M' Mandible Inferior margin of jaw at angh Lower premolars M, M, Wide between canines Narrow Well-developed tube extending from foramen ovale to anterior margin of auditory bulla More inflated External ojiening large Inflated Small Wide Roimded KeiativcK large Short, broad Hclativi'ly short protocoue Inner lobe not ex])anded RelativcK short and thick Broad, flattened Relatively short, broad Metaconid absent, talonid narro\ Relatively small Narrow between canines Wide No tube. Area between foramen ovale and auditor) bulla flat Less inflated Hxternal opening small Not inflated Large Narrow Flattened Relatively small Long, narrov\ Relatively long protocone Iimer loi)e expanded R<^latively long and slender Narrower, less flattened Relatively long, slender incipient metaconid. talonid bro; Relatively large 1986 Anderson etal: Bioceography and Systematics 57 ferrets (Choate et al. 1982), but they are en- tirely different in appearance (Fig. 31). Polecats probably arose in Europe in the Villafranchian (3-4 mil yrs B.P.). The earliest known species, Stromers polecat (M. stro- meri), ranged from the late Villafranchian to the middle Pleistocene, when it was replaced by the modern species. Though smaller in size, Stromers polecat was closely allied to the European polecat (A/, putorius) and was probably ancestral to both the European pole- cat and the steppe polecat (M. eversmanni). These two polecats have been considered con- specific by some workers, but studies by Rus- sian mammalogists (Stroganov 1962) have shown them to be distinct, well-defined spe- cies that differ in size, coloration, and habitat. Although their ranges overlap in Hungary, Romania, and southern European Russia, they are nowhere truly sympatric, being sepa- rated by different habitat preferences (Corbet 1966). Hybrids occur only under exceptional circumstances. Unlike M. nigripes, the steppe polecat is not closely associated with any one species of rodent and feeds on susliks {Spermophilus spp.), marmots, hamsters and voles; in winter, pikas (Ochotona spp.) are a major food source in some areas. Rodent bur- rows, especially those of susliks, are often ex- propriated by polecats for shelter and dens, though they may dig their own. Miistela ev- ersmanni is valued as an exterminator of ro- dents and for its fur, which is, however, of lower quality than that of M. putorius. Al- though M. eversmanni is not considered to be endangered, hunting the animal in Siberia is prohibited. Mustela eversmanni and A/, nigripes are closely related, and their possible con- specificity has been noted by several workers (see Youngman 1982 for references). Al- though their size and coloration are similar, and analysis shows only slight differences in cranial and dental measurements (Figs. 20, 21, 22), Anderson (1977) considers them sepa- rate entities. That the two species are closely related cannot be doubted, but until detailed comparative and statistical studies are made on the large collections of Mustela evers- manni in Soviet institutions, these data are compared with the information already compiled on Mustela ni- gripes, and behavioral and chromosomal studies are un- dertaken on both species, I regard them as distinct. Detailed studies are still lacking for M. ev- ersmanni, and so far there have not been any studies on genetic variation between the two species, so the question of Af . eversmanni and M. nigripes conspecificity remains unre- solved. Another taxonomic problem in the ferrets is the recognition of subspecies. No subspecies of A/, nigripes have ever been named, and our studies do not show any taxonomically signifi- cant geographic variations between samples. Two or perhaps three subspecies of A/, putori- ous are recognized based on slight differences in size and color. Seventeen subspecies of Af. eversmanni have been described, eight of them from Siberia. Strogonov (1962:370) said, "The Siberian polecat shows more geographi- cal variation than the European polecat, this being manifested in changes in fur structure and in dimensions of body, skull and claws." Whether all of these subspecies are valid or merely represent oversplitting is unknown. Of the three species of polecats, M. evers- nmnni has by far the largest geographic range, extending from Hungary to far eastern Asia across the broad band of steppes, forest steppes, and semideserts between 50° and 60° N latitude. The historic range of M. nigripes included the Great Plains and mountain valleys. This was a relatively homogeneous environment without major geographic barriers. However, Endler (1977) points out that there is no evi- dence that allopatry is necessary for differenti- ation. Gradation within a continuous range (parapatry) is very common, as is pointed out by the north-south differentiation demon- strated for Af. nigripes in this paper. Addi- tional specimens from the northern and southern extremes of the range would proba- bly demonstrate more strongly this clinal vari- ation. Whether geographic isolation, for ex- ample, in South Park, Colorado (USNM 247073), would eventually have resulted in distinct subspecies will, of course, never be known. Ferrets entered North America from Siberia, spread across Beringia, and then ad- vanced southward through icefree corridors to the Great Plains. Kalela (cited in Kurten 1957) noted that between 1880 and 1940, M. putorius extended its range in Finland from the Karelian Isthmus north to central Os- 58 Great Basin Naturalist Memoirs No. 8 trobothnia and west to the Gulf of Bothnia at a rate of 7.5 km annually or 750 kni/century. This rate is probably applicable for ferrets spreading across Siberia into the New World in the Pleis- tocene, when conditions were favorable. Discussion Our evidence supports the contention of others (e.g., Linder et al. 1972, Hubbard and Schmitt 1984) that black-footed ferrets were probably common historically. We have lo- cated physical remains or verified reports of ferrets from 128 of 513 counties (25%) within the historic range o(Cynomys . A conservative estimate is that 41,000,000 ha of western grasslands were occupied by prairie dogs in the early part of this century. Using the For- rest et al. {Life history characteristics, 1985) population density estimate of one ferret per 40-60 ha, habitat may have been available in the past to support as many as 500,000- 1,000,000 black-footed ferrets, if this habitat were fully occupied by ferrets. Although the Canadian specimens cast some doubt on the nearly obligate association between ferrets and prairie dogs, it is almost certain that alternate habitats do not provide adequate resources to support ferrets in the long term. If ferrets were living in habitats other than prairie dog colonies in Canada, then they should still be extant there; yet the last specimen was taken in 1937, about the time remnant prairie dogs in Canada were being eliminated by expansion of agriculture. Geographic variation in a species has impli- cations for any recovery program involving reintroduction of animals into areas where they have been extirpated. It would not be prudent to attempt such reintroductions us- ing animals that differ greatly from those that originally occurred in the reintroduction area. However, with black-footed ferrets there seems to be little habitat-related variation, and reintroductions should prove successful in any geographic area with any prairie dog species serving as prey, provided sufficient habitat still remains to support the ferrets and their prey. With regard to clinal or other geo- graphic variation, our analyses suggest that a case can be made for morphometric variation within this species, although the usefulness of this argument seems limited to the case where numerous populations are competing for pro- tection (Schonewald-Cox et al. 1985), which is not the case for this species. The possibility that the steppe ferret and the black-footed ferret are representatives of a single holarctic species exploiting similar eco- logical niches in the New World and Old has been suggested. This in no way diminishes the unique position the black-footed ferret holds in the prairie ecosystems of this conti- nent. It does suggest that options that might draw on M. eversmanni to assist in recovery efforts for the endangered M. nigripes should be further explored. Acknowledgments Many individuals and organizations made this study possible. We are particularly grate- ful to organizations that support our field work at Meeteetse and allowed us time to complete this manuscript: the Wildlife Preservation Trust International, New York Zoological So- cietv-Wildlife Conservation International, and'the World Wildlife Fund— U.S. In addi- tion to the many curators of collections who so kindly responded to our inquiries, several were exceptionally helpful: Carron Meaney, Betsy Webb of the Denver Museum of Natu- ral History; Phil Youngman of the National Museum of Natural Sciences, Ottawa; Nor- man Slade of the University of Kansas; Charles Smart of the Academy of Natural Sci- ences, Philadelphia; Horace F. Quick of the University of Colorado; R. George Corner of the University of Nebraska State Museum; D. E. Flath of the Montana Department of Fish, Wildlife, and Parks; W. G. Gillespie of the University of Arizona; and David Baron of the Saskatchewan Museum of Natural History. We would also like to thank the curators who allowed FA to measure collections in their care. Mark Shumar of Idaho State University, provided invaluable assistance with SPSS statistical procedures. Denise Casey provided art and technical figures. The manuscript was criticalK' reviewed by John L. Paradiso, Da\id M. Armstrong, and Steven Minta. We greatl) thank all of these organizations and individuals. LiTEiuTURE Cited Ai.DOl's, S. E. 1940. Notes on a black-foott-d ferret raised in eaptivity. J. Mammal. 21:23-26. 1986 Anderson et al.: Bioceography and Systematics 59 Alexander. A. M. 1932. Control, not extermination of Cijnoinys ludovicianus arizonensis. J. Mammal. 13:302. Ames, A, E. 1874. Mammalia of Minnesota. Minnesota Acad. Bull. 1:69. Anderson, E. 1968. Fauna of the Little Box Elder Cave, Converse County, Wyoming: the Carnivores. Univ. Colorado Stud. Ser. Earth Sci. 6:1-59. 1970. Quaternary evolution of the genus Martes (Carnivora, Mustelidae). Acta Zool. Fennica 130:1-132. 1973. Ferret from the Pleistocene of central Alaska. J. Mammal. 54:778-779. 1974. A survey of the late Pleistocene and Holocene fauna of Wyoming. In M. Wilson, ed.. Applied geology and archaeology: the Holocene history of Wyoming. Geol. Surv. Wyoming Rept. Inv. 10:78-87. 1977. Pleistocene Mustelidae (Mammalia, Car- nivora) from Fairbanks, Alaska. Bull. Mus. Comp. Zool. 148:1-21. In preparation. The fauna of Upper Plum Creek Canyon rockshelter (5LA2158), Las Animas County, Colorado. Armstrong, D. M. 1972. Distribution of mammals in Colorado. Monogr. Univ. Kansas Museum of Nat. Hist. 3:1-415. Audubon. J J , and J Bachman. 1851. The viviparous quadrupeds of North America. V. G. Audubon, New York. Vol. 2. 334 pp. Bailey, V. 1905. Biological survey of Texas. North Ameri- can Fauna 25:1-226. 1926. Biological survey of North Dakota. North American Fauna 49:171-173. 1932. Mammals of New Mexico. North American Fauna 53:1-412. Bell, W. R. 1921. Death to the rodents. USDA Yearbook 1920:421-438. Biggins, D E , M. H. Schroeder, S C. Forrest, and L. Richardson. 1985. Movements and habitat rela- tions of radio-tagged black-footed ferrets. Pages 11.1-11.17 in S. Anderson and D. Inkley, eds.. Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. BISSELL. S. J. 1979. Black-footed ferret verification and habitat inventory, June 17, 1977-September 30, 1978. Unpublished report, Colorado Div. Wildl., Denver. 15 pp. BoGGESS, E. K.. F R. Henderson, and J R. Choate. 1980. A black-footed ferret from Kansas. J. Mam- mal. 61:571. Bolton, H E. 1916. Spanish exploration in the South- west, 1542-1706. Charles Scribner & Sons, New York. 217 pp. Burnett, W L 1918. Rodents of Colorado in their economic relation. Office of State Entom. Circ. 25:1-31 Burt, W H 1960. BaculaofNorth American mammals. Misc. Publ. Univ. Michigan Mus. Zool. 113:1-75. Cahalane, V. H 19.54. Status of the black-footed ferret. J. Mammal. 35:418-424. Carpenter, J W, andC N. Hillman. 1978. Husbandry, re- production and veterinary care of captive ferrets. Amer. Assoc. Zoo Vet. Ann. Proc, Knoxville, Tenn. 1978:36-47. Gary, M. 1911. A biological survey of Colorado. North American Fauna 33:1-256. Cheatheam, L K 1977. Density and distribution of black-tailed prairie dogs in Texas. Texas J. Sci. 29:33-40. Choate, J R , E. K. Bocgess, and F R Henderson. 1982. History and status of the black-footed ferret in Kansas. Trans. Kansas Acad. Sci. 85:121-132. Choate, J. R.. and E. D. FLEHARTi', 1975. Synopsis of native. Recent mammals of Ellis Co., Kansas. Oc- cas. Papers Mus. Texas Tech. Univ. 37:1-80. Clark, T. W. 1973a. A field study of the ecology and ethology of the white-tailed prairie dog (C. leucu- rus) with a model oiCynomys evolution. Unpub- lished dissertation. University of Wisconsin, Madison. 215 pp. 1973b. Prairie dogs and black-footed ferrets in Wyoming. Pages 88-101 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State Uni- versity, Brookings. 1975. Some relationships between prairie dogs, black-footed ferrets, paleo-Indians, and ethno- graphically known tribes. Plains Anthrop. 20:71-74. 1980. A listing of reports on black-footed ferrets in Wyoming (1851-1977). Northwest Sci. 54:47-54. 1986. Some guidelines for management of the black-footed ferret. Great Basin Nat. Mem. 8: 160-168. Clark, T. W., and T. M. Campbell. 1981. Additional black-footed ferret {Mustela nigripes) reports from Wyoming. Great Basin Nat. 41:360-361. Clark, T W., L. Richardson, S. G Forrest. T. M Camp- bell III, D. E. Casey, and K A Fagerstone. 1985. Black-footed ferret prey base. Pages 7.1-7.14 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. CoCKRUM, E. L. 1960. The recent mammals of Arizona, their taxonomy and distribution. University of Ari- zona Press, Tucson. 276 pp. Corbet. G. B. 1966. The terrestrial mammals of western Europe. London, G. T. FoulisandCo. Ltd. 264 pp. Corner, R. G. 1977. A late Pleistocene-Holocene verte- brate fauna from Red Willow County, Nebraska. Trans. Nebraska Acad. Sci. 4:77-93. CouES, E. 1874. Wanted. American Sportsman 5(9): 1, Nov. 28, 1874. 1877. Fur-bearing animals: a monograph of North American Mustelidae. Dept. of Interior. USGS Misc. Publ. 8:149-153. 1882. The black-footed ferret in Texas. Amer. Nat. 16:1009. Crabb. W O.. and G W Watson. 1950. Black-footed ferret in Montana. J. Mammal. 31:99. D.-KY, AM.. AND A P. Nelson. 1929. Wild life conserva- tion and control in Wyoming under the leadership of the United States Biological Survey. Joint Publ. U.S. Biol. Surv., Wyoming State Game & Fish Dept., Wyoming State Ext. Serv., Wyoming Dept. Agric, Wyoming State Hist. Res. and Publ. Div. 26 pp. 60 Great Basin Naturalist Memoirs No. 8 Dreeszen, v. H. 1970. The stratigraphic framework of Pleistocene glacial and periglacial deposits in the central plains. Pages 9-22 in W. Dort and J. K. Jones, eds.. Pleistocene and Recent environments of the Central Great Plains. University of Kansas Press, Lawrence. Durrani, S. D. 1952. Mammals of Utah. University of Kansas, Lawrence, Publ. Mas. Nat. Hist. 6: 1-549. Endler.J. A. 1977. Geographic variation, speciation, and clines. Monogr. Population Biology, Princeton University Press 10:1-246. Fighter, E.. and J. K. Jones. 1953. The occurrence of the black-footed ferret in Nebraska. J. Mammal. 34:385-389. Flath, D. L. 1979. Status of the white-tailed prairie dog in Montana. Proc. Montana Acad. Sci. 38:63-67. Flath, D. L., andT W. Clark 1986. Historic status of black-footed ferret habitat in Montana. Great Basin Nat. Mem. 8:63-71. Forrest, S. C, T. W. Clark, D. E. Biggins, L. Richard- son, K. A. Fagerstone, and T M Campbell 111. 1985. Life history characteristics of the genus Mustela, with special reference to the black-footed ferret, M. nigripes . Pages 23. 1-23. 14 in S. Ander- son and D. Inkley, eds.. Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Forrest, S C , T W Clark, L Richardson, T M Camp- bell III. 1985. Black-footed ferret habitat: some management and reintroduction considerations. Wyoming Bur. Land Mgmt. Wildl. Tech. Bull. No. 2, Cheyenne. 49 pp. Forrest, S. C, L. Richardson. T. W. Clark, and T M. Campbell 111. 1985. Litter survey and mark/re- capture of black-footed ferrets at Meeteetse, July-October 1984. Unpublished report to Wyo- ming Game and Fish, Wildlife Preservation Trust, New York Zoological Society and World Wildlife Fund. 12 pp. Fortenbery, D K 1971. Unpublished report to Div. Wildl. Serv. Region II, Albuquerque, N. Me.x. Garst. W E 1954. Black-footed ferret in South Dakota. J. Mammal. 35:594. Getz. L. L 1960. Middle Pleistocene carnivores from southvyestern Kansas. J. Mammal. 41:361-365. Gilbert, G. L 1977. Inventory of potential black-footed ferret habitat in the White River Resource Area, Colorado. U.S. Bur. Land Mgmt. Coop. Educ. Publ. No. 1, Craig Dist., Colo. 21 pp. Gillespie. W. B. In preparation. Vertebrate remains from Atlatl Cave, Chaco Canyon, New Mexico. Gray, J. E. 1865. Revision of the genera and species of Mustelidae contained in the British Museum. Proc. oftheZool. Soc. London 1865:110. Grinnell, G B 1895. The story of the Indian. D. Apple- ton Co., New York. pp. 1896. Range ofthe black-footed ferret. Forest and Stream 47:84. Grondahl, C R 1973. Status ofthe black-tailed prairie dog and black-footed ferret in North Dakota. Pages 44-47 tnR. L. LinderandC. N. Hillnian, eds., Proc. Black- footed Ferret and Prairie Dog Workshop, South Da- kota State University, Brookings. Guilday, J E., and E K. Adam. 1967. Small mammal remains from Jaguar Cave, Lemhi County, Idaho. Tebiwa 10:26-36. Hager, M R 1972. A late Wisconsin-Recent vertebrate fauna from Chimney Rock, Animal Trap, Larimer County, Colorado. Contrib. Geol. 11:63-71. Hall. E R 1981. The mammals of North America, 2d ed. J. Wiley & Sons, New York. Vol. 2:601-1181. Halloran. a. F. 1964. The mammals of Navajoland. Navajo Tribal Museum, Window Rock, Arizona. 23 pp. H.\rris, a. H . and J S Findley. 1964. Pleistocene-Re- cent fauna ofthe Isleta caves, Bernalillo County, New Mexico. Amer. J. Sci. 262:114-120. Hassien, F D 1976. A search for black-footed ferrets in the Oklahoma Panhandle and adjacent area and an ecological study of black-tailed prairie dogs in Texas County, Oklahoma. Unpublished thesis, Oklahoma State University, Stillwater. Ill pp. Henderson. F R . P. F Springer, and R Adrian. 1969. The black-footed ferret in South Dakota. South Dakota Dept. Game, Fish, Parks, Tech. Bull. No. 4:1-37. Hibbard, C. W. 1934. The occurrence oi Erethizon epix- anthum bruneri and Mustela nigripes in Kansas. J. Mammal. 15:70-71. 1970. Pleistocene mammalian local faunas from the Great Plains and central lowland provinces of the United States. Pages 395-433 in W. Dort and J. K. Jones, Jr. , eds. , Pleistocene and Recent envi- ronments ofthe Central Great Plains, University of Kansas Press, Lawrence. Hillman, C. N., and T. W. Clark. 1980. Mustela ni- gripes. Mammalian Species 126:1-3. Hoffmann, R. S., P L. Wright, and F. E. Newby. 1969. The distribution of some mammals in Montana. I. Mammals other than bats. J. Mammal. 50:579- 604. HOMOLKA, C. L. 1964. Our rarest mammal? Audubon 66:244-246. Hoogland. J. L. 1981. The evolution of white-tailed and black-tailed prairie dogs (Sciuridai': ('yiioDUjs leu- curtis and C. ludovicianus). Ecology 62:252-272. Hubbard, J, P., and C, G. Schmitt. 1984, The black- footed ferret in New Mexico. Unpublished report to New Mexico Bur. Land Mgmt., Santa Fe. 118 pp. Jobman. W. G.. and M. E. Anderson. 1981. Potential present range ofthe black-footed ferret as of Janu- ary 1, 1981. Unpublished report, U.S. Fish and Wildlife Serv., Pierre, South Dakota. 65 pp. Johnson, D R 1969. Returns ofthe American Fur Com- pany, 1835-1839. J. Manuual. 50:836-839. Jones. J K , Jr 1964. Distribution and taxonomy of mam- mals of Nebraska. Universitv of Kansas Publ. Mus. Nat. Hist. 16:1-356. Kkuwin, L, andC. G. Scheelhaase. 1971. Present status of the black-tailed prairie dog in Saskatchewan. Blue Jay 29:35-37. Kinc;, C M 1975. The si'x ratio of trapped weasels (Mustela nivalis). Mammal Review 5:1-6. King, J 1955. Social behavior, social organization and population d\ iiamics in a black-tailed prairie dog town in the iilack Hills of South Dakota. C(mtrib. Lab. Vert. Biol. L'niversity of Michigan 67:1-123. 1986 Anderson et al. : Biogeocrapiiy and Systematics 61 KowALSKl, K. 1976. Mammals, an outline of theriology. Translated and published for the Smithsonian In- stitution and the National Science Foundation, Washington, D. C. 617 pp. KURTEN. B 1957. Mammal migrations, Cenozoic stratig- raphy and the age of Peking Man and the australo- pithecines. J. Paleontol. 31:215-257. KuRTEN. B., AND E. ANDERSON 1972. The sediments and fauna of Jaguar Cave: the fauna. Tebiwa 15:21-45. Lewis. J C. 1973. Additional records of black-footed fer- rets in Oklahoma. Southwest Nat. 18:350. Lewis, J C , and F Hassein 1973. Status of prairie dog and surveys for black-footed ferrets in Oklahoma. Pages 60-75 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State University, Brook- ings. LiNDER, R. L , R. B Dahlgren, andC N Hillman 1972. Black-footed ferret-prairie dog interrelationships. Pages 22-37 in Symposium on Rare and Endan- gered Wildlife S.W. U.S., 22-23 September 1972, Albuquerque, New Mexico, New Mexico Dept. Game and Fish, Santa Fe. Lock, R A 1973. Status of the black-footed ferret and black-tailed prairie dog in Nebraska. Pages 44-47 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State University, Brookings. LOVAAS, A L. 1973. Prairie dogs and black-footed ferrets in the National Parks. Pages 139-148 in R. L. Linder and C. N. Hillman, eds., Proc. Black- footed Ferret and Prairie Dog Workshop, South Dakota State University, Brookings. Martin, S. J 1983. Additional records of black-footed ferrets in Wyoming. Southwest. Nat. 28:95. Martley, H. J. 1954. History of organized, cooperative predator and rodent control (1915-1953). Unpub- lished report, U.S. Fish Wildl. Serv., Casper, Wyoming. 32 pp. Miller, G S , Jr. 1912. List of North American land mammals in the U.S. National Museum, 1911. Bull. U.S. Nad. Mus. 79:1-455. Morgan, D. L. 1953. Jedediah Smith and the opening of the West. University of Nebraska Press, Lincoln. 458 pp. Moon, G F 1905. What have I got? Trappers World 1:4. Nelson, E. W. 1919. Annual report of Chief of Bureau of Biological Survey. Pages 275-298 m USDA Ann. Rept. Dept. Agric. for year ended June 1919. NowAK. R. M., AND J. L. Paradiso. 1983. Walker's mam- mals of the world. 4th ed. Johns Hopkins Press, Baltimore, Maryland. Vol. 2:992-993. Powell, R. A. 1979. Mustelid spacing patterns: variation on a theme by Mustela. Zeit. Tierpsychol. 50; 153-165. Progulske, D R 1969. Observations of a penned, wild- captured black-footed ferret. J. Mammal. 50: 619-621. Rose, B J 1973. History of prairie dogs in South Dakota. Pages 76-77 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State University, Brook- ings. Ruprecht, a L. 1978. Dentition variations in the com- mon polecat in Poland. ActaTheriol. 23:239-245. SCHMIDLY, D J. 1977. The mammals of Trans-Pecos Texas. Texas A&M University Press, College Sta- tion. 225 pp. SCIIONEWALD-COX, C. M , J, W. BaYLESS. AND J. SCHONEWALD. 1985. Cranial morphometry of Pacific Coast Elk (Cervus elaphus). J. Mammal. 66:63-74. SCHULTZ, C B , AND E B HOWARD 1935. The fauna of Burnet Cave, Guadalupe Mountains, New Mex- ico. Proc. Acad. Nat. Sci. Philadelphia 87: 273-298. Seabloom.R W , M G McKenna, and R. D.Crawford. 1980. Recent records of mammals from southwest- ern North Dakota. Prairie Nat. 12:119-123. Seton, E T 1929. Lives of game animals. Pages 275-298 in Vol. 4, Pt. 1. Doubleday, Doran Co., Garden City, New York. Sheets, R. G., R. L. Linder, and R B. Dahlgren. 1972. Food habits of two litters of black-footed ferrets in South Dakota. Amer. Midi. Nat. 87:249-251. Slaughter, B H. 1966. The Moore Pit local fauna, Pleis- tocene of Texas. J. Paleontol. 40:78-91. SOPER, J. D. 1938. Discovery, habitat and distribution of the black-tailed prairie dog in western Canada. J. Mammal. 19:290-300. 1944. Further data on the black-tailed prairie dog in western Canada. J. Mammal. 25:47-48. 1946. Mammals ofthe northern Great Plains along the International Boundary in Canada. J. Mam- mal. 27:127-153. Soule, M. 1980. Thresholds for survival: maintaining fitness and evolutionary potential. Pages 151-169 in M. Soule and B. Wilcox, eds.. Conservation biology: an evolutionary ecological perspective. Sinauer Associates, Sunderland, Massachusetts. Stains, H. G. 1976. Calcanea of members ofthe Musteli- dae. Parti, Mustelinae. Bull. So. California Acad. of Sciences 75:237-248. Stalker, A MacS, C. S Churcher, and R. S. Hill. 1982. Ice age deposits and animals from the southwest- ern part ofthe Great Plains of Canada. Misc. Rept. 31, Geol. Surv. Canada, Wall Chart. Storer, J. E 1975. Pleistocene prairie dog in south-cen- tral Alberta. Blue Jay 33:247. Stroganov, S U 1962. Carnivorous mammals of Siberia. Israeli Program for Scientific Translation, Jerusalem, 1969. 522 pp. Taylor, D 1961. Notes on a recent collection of a black- footed ferret. Trans. Kansas Acad. Sci. 64:41. Thornbach, J , AND M. JENKINS. 1982. Pages 349-351 in The lUCN Mammal Red Data Book, Part 1. Gresham Press, Surrey, United Kingdom. Thorne, E T, M. H Schroeder, S C. Forrest. T. M. Campbell III, L Richardson, D. E. Biggins, L. R. Hanebury, D. Belitsky, and E. S. Williams. 1985. Capture, immobifization and care of black- footed ferrets for research. Pages 9.1-9.8 in S. Anderson and D. Inkley, eds.. Black-footed Fer- ret Workshop Proc, Laramie, Wyoming, Sep- tember 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. True, F. W 1885. A black-footed ferret fi-om Texas. Amer. Nat. 19:720. 62 Great Basin Naturalist Memoirs No. Tyler, J D 1968. Distribution and vertebrate associates of the black-tailed prairie dog in Oklahoma. Un- published dissertation. University of Oklahoma, Norman. 85 pp. U.S. Bureau Land Management. 1982. Black-tailed prairie dog control/management in Phillips Re- source Area. Kept. Bur. Land Mgt., Lewistown Dist., Lewistown, Montana. 63 pp. Velich, R. 1961. Notes on mammals from Nebraska and southwestern Iowa. J. Mammal. 42:92-94. Vereshchagin, N K., and G F Baryshnikov 1984. Quaternary mammalian extinctions in northern Eurasia. Pages 483-516 in P. S. Martin and R. G. Klein, eds., Quaternary extinctions — a prehistor- ic revolution. University of Arizona Press, Tucson. VooRHiES. M R , and R G Corner. In press.. Small mammals with boreal affinities in late Pleistocene (Rancholabrean) deposits of eastern and central Nebraska. Trans. Tertiary-Quaternary '83. Warren, E R 1910. The mammals of Colorado. G. P. Putnam and Sons, New York. 300 pp. White. J A , H G McDonald, E Anderson, and J M. SoiSET. 1984. Lava blisters as carnivore traps. Spec. Publ. Carnegie Mus. Nat. Hist. 8:241-256. Young, S. P., and A. F. Halloran, 1952. Arizona speci- mens of the black-footed ferret. J. Mammal. 33:251. YOUNGMAN, P M 1982. Distribution and systematics of the European mink, Mustela hitreola Linnaeus 1761. ActaZool. Fennica 166:1-48. HISTORIC STATUS OF BLACK-FOOTED FERRET HABITAT IN MONTANA Dennis L. Flath' and Tim W. Clark^ Abstract. — Black-footed ferrets (Mustela nigripes) use prairie dogs {Cynomys spp.) for food and their burrows for shelter. Thus, prairie dog colonies are essential ferret habitat. Prairie dog control, which resulted in permanent loss of ferret habitat, is considered the primary reason for the ferret's endangered status today. Northern Pacific Railroad (presently Burlington Northern) lands were surveyed 1908-1914, just prior to the onset of widespread prairie dog control. In Montana the surveyed area included a belt about 483 km long and 192 km wide, from the Montana-North Dakota border westward to Livingston. In all, 6,661 sections (11.8%) of 22 counties were surveyed and 1,662 of these sections (24. 9%) contained at least some prairie dogs. Prairie dog colonies (N = 1 , 985) occupied all or part of 5, 186, 16 ha (40ac) parcels and totaled a minimum of 47,568 ha, with a mean colony size of 24.5 ha (2.8% of the landscape in colonies). Two township-wide belt transect samples — ^T4N and R45E — showed colonies were clumped in distribution. Two areas with large complexes of colonies are illustrated, and each area exceeded an estimated 15,000-1- ha. The Tongue River-Otter Creek area had at least 20 complexes, with a mean intercomplex distance of 3.4 km; and the Powder River-O'Fallon Creek area had at least 33 complexes, with a mean intercomplex distance of 2.9 km. Historic land uses were similar to today's uses — grazing and a few crops. Historic prairie dog areas in Montana occupied an estimated 5,953 sq km. An estimated 90-1-% reduction in prairie dogs has occurred since 1914, largely if not totally due to poisoning. The elimination, fragmentation, and greatly reduced size of ferret habitat has undoubtedly contributed to the endangered status of ferrets. A few areas in Montana appear to contain enough prairie dogs to potentially harbor ferret populations. These areas could serve as reintroduction sites for ferrets, as well as examples of complex prairie dog ecosystems. Black-footed ferret habitat consists of biotic and abiotic components of prairie dog colonies (Coues 1877, Forrest et al. 1985). In addition to black-footed ferrets, prairie dog colonies host many vertebrate and invertebrate spe- cies, some in dependent relationships with prairie dogs, such as the black-footed ferret (Clark et al. 1982). This inter-relationship of plant and animal life centered on prairie dog colonies is often called the "prairie dog ecosystem" (Bureau of Land Management 1980). The ecology of prairie habitats in North America has been significantly altered over the past century because of the activities of man. Prairie dog ecosystems have been greatly affected by extensive poisoning over the last 100 years. As a result, many of these ecosystems were drastically reduced or totally eliminated. Many species dependent on prairie dogs have also suffered. Unfortu- nately, few data exist on actual prairie dog distribution and abundance prior to post-1915 poisoning campaigns by the Biological Survey and various states. The prairie dog ecosystems of today are perhaps the most notable exam- ples of relict, insular ecological relationships. This is biologically significant because of the large numbers of associated species involved. An understanding of historic data is essen- tial for efficient management of such relict ecological relationships. This paper describes black- tailed prairie dog (C. ludovicianus) status in Montana from 1908 through 1914 and compares it with current knowledge about Montana prairie dogs and black-footed ferret habitat requirements as described in the liter- ature. The historic extent and configuration of prairie dog colonies that determined black- footed ferret population sizes, densities and viability has not been previously described in the scientific literature. Methods Data were derived from Northern Pacific Railway land surveys for 190&-1914. The study area was a belt up to 192 km wide and 483 km long, bisected by the Trans-Montana track, which entered Montana at Wibaux, ex- tended west to Glendive on the Yellowstone 'Montana Department of Fish, Wildlife and Parks, Box 5, Montana State University, Bozeman, Montana 59717-0001. ^Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209 and Biota Research and Consulting, Inc. , Box 2705, Jackson, Wyoming 83001. 63 64 Great Basin Naturalist Memoirs No. 8 River, then paralleled the river to Livingston. The Northern Pacific Railway, which formerly owned the Trans-Montana track, merged with other railroads to form the Burlington North- ern, Inc., Railway in 1968. The Agricultural Resources Department, Burlington North- ern, Inc., Miles City, Montana, provided the original survey data. In Montana, the Northern Pacific was granted 20 odd-numbered sections of land per 1.6 km (1 mi) of track as inducement to link East to West by rail. Initially lands were se- lected from a zone 32 km (20 mi) on either side of the track. However, because some sections were previously appropriated to homestead- ers or other occupants, the government set a 96 km (60 mi) limit on each side of the track from which to select other sections. Three large exclusions within the 192 km strip were made for Northern Cheyenne and Crow In- dian reservations and for high mountainous country. As a land grant railroad, the North- ern Pacific partially funded track construction through disposition of some lands granted by the federal government. Before Northern Pacific sold or leased land grant parcels, range examinations were made to map them and determine their present and potential land uses and existing natural re- sources, including timber, grass, and water. Prior to 1908, the United States General Land Office had completed land surveys to mark section and quarter corners associated with the Montana Principal Meridian and Standard Parallel. As a result, the railroad land examin- ers accurately mapped topography, drain- ages, flatlands, timbered areas, coal outcrops, and other resource characteristics that influ- enced land value. Because prairie dogs were considered a menace that destroyed range- land forage and crops, prairie dog colonies were also mapped. The locations and extent of prairie dog colonies were indicated on original maps by writing "DOGS" or "DOG TOWN" across the occupied area, proportional to the size of the colony. For large colonies, letters appeared bold and widely spaced, and in some cases actual colony sizes were estimated. Often comments were included on prairie dog graz- ing effects or the spatial extent of colonies. Surveyed lands were mapped and color coded by estimated land use potential — grazing. cropland, and woodlands, which indicated to- pography and vegetation on which colonies were located. Herman Liebinger (personal communication to Wieland, 1979), land agent of eastern Montana land for the Northern Pacific Railway in the 1930s, was sure that all prairie dog colonies were recorded on all lands examined. We designed data sheets to record the oc- currence of prairie dogs from the original land assessment journals that recorded the sec- tions [2.56 sq km (640 acres)] containing prairie dogs. For each section we recorded a "hit" or a "miss" for prairie dog occurrence. Hits and misses from the data sheets were color coded and plotted on mylar overlays of 1:250,000 uses topographic maps. These overlays demonstrated the clustering of prairie dog colony distribution. For those sec- tions with prairie dogs, an estimate was made from the maps as to how many 16 ha (40 acre) tracts (16 per section) contained prairie dogs. Furthermore, a given 16 ha tract could have anywhere from a few holes to an entire 16 ha of prairie dogs. We used the midpoint of 8 ha per tract to estimate colony sizes when actual sizes were not given. This gave us an estimate for actual size of prairie dog colonies. Since only odd-numbered sections were surveyed by the railroad, the method constituted a sampling procedure that amounted to a maximum 50% in those townships with complete coverage. Frequently prairie dog colonies extended an unknown distance beyond the boundary of the sample section. Results The surveyed area encompasses a large portion of eastern Montana including parts of 22 counties (Fig. 1). The most prevalent land form is rolling sedimentary plains. Erosion coulees, river breaks, badlands, and intrusive mountain ranges are found throughout the area. Precipitation generally ranges from 30.5 to 40.6 cm per year, resulting in a shrub-grass steppe ecosystem. Upland sites support ex- tensive stands of sagebrush {Artemesia triden- tata ) and in some cases juniper (Juniperus sp.) or pine {Pinus sp.) woodland. Black-tailed prairie dogs occupy the eastern two-thirds of Montana, or about 220,000 sq km (Hall 1981). Although prairie dog numbers 1986 Flath, Clark: Black-Footed Ferret Habitat in Montana 65 Fig. 1. Range of the black-tailed prairie dog (shaded area) and collection sites of black-footed ferrets (dots) Montana. Specimens exist for solid circles. have been greatly reduced, the extent of their overall range has changed little since the early surveys. The belt transect study area included about half of the total prairie dog range, with detailed land examinations of 17,052 sq km, or about 7.8% of the total Montana range of the species. This broad area includes steep ter- rain, shrubby vegetation, waterways, and in- trusive mountain ranges that are not prairie dog habitat. The actual area occupied by prairie dogs throughout this large region was and is limited to relatively level areas, vege- tated with herbaceous plants and few shrubs. General George A. Custer's field journal on his travels to the Little Bighorn River in sum- mer 1876 noted several extensive prairie dog colonies along Rosebud Greek (Fulton 1982). The railroad survey journals also describe many prairie dog colonies in this area. Colony sizes varied: some entries noted only a "few holes" per section surveyed, whereas others stated that a colony was large and extended over adjacent sections (e.g., T16N R45E S21 and to the southwest). We estimated the largest single colony at 9,328+ ha (23,040+ ac) near Beaver Greek and Sweeney Creek south of Hathaway (T4N R44E). Based on many such entries, our assessment of prairie dog distribution should be considered a mini- mum estimate. Prairie dog distribution based on the railroad surveys is given by county in Table 1. In the 22 surveyed counties, 6,661 sections within 759 townships were examined, representing 11.8% of the total area of these counties. Of the 6,661 sections, 1,662 (24.9%) were partially or totally occupied by prairie dogs. The largest area was in Rosebud County: of 1,025 sections examined, 397 contained some prairie dogs (38.7%). Mc- Cone, Park, Richland, Wheatland, and Wibaux counties all showed less than 10% of the sampled sections occupied by prairie dogs. Big Horn, Carter, Golden Valley, Musselshell, Powder River, and Treasure counties all showed some prairie dogs in more than 40% of the sampled sections. Based on the frequency distribution of colonies by sections, it seems that prairie dogs were relatively abundant and widespread, exist- ing in single large colonies or in large groupings of smaller colonies. The survey located 1,985 prairie dog colonies or one colony per 3.3 sections (Table 2). These colonies occupied all or part of 5,186 16 ha par- cels, totaled a minimum of 475 sq km, and aver- aged 24.5 ha. Prairie dogs occupied a minimum 2.8% of the landscape. 66 Great Basin Naturalist Memoirs No. Table 1. Black-tailed prairie dog distribution by county in Montana (1908-1914). Square Townships Sections Percent Sections with prairie Percent 1976 land use^ County kilometers surveyed surveyed county dogs (%) Rangelands Crops Woodlands Other Big Horn 1,946 7 27 0.5 18(66.0) 84 9 6 1 Carter 1,305 13 146 4.4 59(40.4) 89 7 3 Custer 1,475 90 864 22.9 320(37.0) 92 6 1 Dawson 927 46 481 20.3 28 ( 5.8) 72 25 3 Fallon 634 40 321 19.8 51(15.9) 76 22 1 Fergus 1,695 3 7 0.2 0( 0.0) 64 20 15 Garfield 1,754 66 810 18.0 109(13.4) 93 4 2 Golden Valley 458 13 164 14.0 69 (42. 1) 83 10 5 McCone 1,026 45 607 23.1 8( 1.3) 70 28 2 Musselshell 731 49 386 20.6 186(48.2) 71 6 21 Park 1,041 2 13 0.4 0( 0.0) 49 7 42 Petroleum 645 15 81 4.9 50(61.7) 93 4 2 Powder River 1,284 31 311 9.4 132(42.4) 84 5 10 Prairie 677 49 389 22.4 39(10.0) 88 11 1 Richland 813 21 197 9.5 12 ( 6.1) 64 33 3 Rosebud 1,961 127 1,025 20.4 397(38.7) 90 4 5 Stillwater 700 8 26 1.4 4(15.4) 59 20 19 Sweet Grass 725 20 71 3.8 14(19.7) 33 18 47 Treasure 381 16 90 9.2 37(41.1) 86 9 4 Wheatland 554 23 225 15.8 5( 2.2) 87 6 6 Wibaux 347 25 192 21.6 15 ( 7.8) 65 32 1 Yellowstone 1,025 50 228 8J 109(47.8) 74 18 4 Totals 22, 105 759 6,661 11.8 1662(24.9) 78 12 9 Ross and Hunter (1976) Table 2. Number of prairie dog colonics, number of 16 ha plots occupied, total estimated area occupied by prairie dogs, and mean colony size based on Northern Pacific Railway surveys (1908-1914). No. prairie No. of 16 ha (40 ac) Estimated Mean Survey dog colonies plots occupied hectares occupied colony year encountered by prairie dogs by prairie dogs size (ha) 1908 174 743 10,031 57.6 1909 243 539 4,656 19.2 1910 600 1,735 14,812 24.7 1911 271 547 4,934 18.2 1912 56 105 672 12.0 1913 238 574 5,000 21.0 1914 403 943 7,464 18.5 Totals 1,985 5,186 47,568 24.5 To determine the clustering of prairie dog colony distribution, we sampled two belt tran- sects through the study area. The east-west distribution of prairie dogs was sampled along a township-wide (9.6 km) belt transect (T4N) beginning at the Montana-North Dakota bor- der (R61E) and running west to R15E (442 km). The south-north distribution of prairie dogs was also sampled using R45E from T5S north to T21N (240 km). Although sample sizes varied by township, 31 of the 47 town- ships (66%) along T4N contained prairie dogs and 14 of the 28 townships (50%) along R45E contained prairie dogs. Prairie dog colonies showed a markedly clumped distribution. We defined a prairie dog colony complex as two or more colonies, regardless of size, in adjacent sections. Some complexes covered more than 36 contiguous sections (9,216+ ha). Maps of the two largest complexes in the sam- pled area are illustrated in Figure 3. Not all the area in Figure 3 was surveyed, but, of that portion surveyed, extensive prairie dog colony complexes appeared closely associated with river and stream courses. Many com- plexes extended 16 km or more. Along the Tongue River and Otter Creek, at least 20 complexes totaled an estimated 15,000+ ha. Mean intercomplex distance was about 3.4 km (range 1-7). Along the Powder River and O'Fallon Creek, at least 33 complexes occu- pied a very large area (estimated 20,000+ ha). Mean intercomplex distance was about 2.9 km (range 1-4). 1986 Flath, Clark; Black-Footed Ferret Habitat in Montana 1 Park 2 Wheatland 3 Sweet Grass 4 Fergus 5 Golden Valley 6 Stillwater 7 Petroleum 8 Musselshell 9 Yellowstone 10 Big Horn 11 Garfield 12 Rosebud 13 Treasure 14 McCone 15 Prairie 16 Custer 17 Powder River 18 Richland 19 Dawson 20 Wibaux 21 Fallon 22 Carter Fig. 2. Location of the belt transect study area and counties surveyed. Current land use in the survey area (Table 1) is for grazing livestock (78%), crop production (12%), and woodlands (9%). Other human uses (e.g., roads) represent only 1%. Land uses dur- ing 1908-1914 were probably similar to today's uses except for differences in fire suppression and cropping techniques. For example, in the past, more fires probably reduced woodlands and shrublands, thereby increasing availability of herbs and grasses for livestock use. Crop pro- duction has also changed because much crop- land today is extensively irrigated with technol- ogy unavailable in the early days. Furthermore, most homesteading within the study area took place from 1915 to 1917, just after the period we examined. Range conditions, which can affect prairie dog colonization and establishment, were not definitive by modern standards for the period 1908-1914 but were generally por- trayed as conducive to increasing prairie dog populations. We estimated changes in prairie dog distri- bution between 1908-1914 and today. The study area crossed a large portion of Montana and included about 92, 736 sq km (over 40%) of the broad range of the black-tailed prairie dog in the state. Within the study area, 17,052 sq km were sampled. They contained an esti- mated minimum of 475 sq km of prairie dogs during 1908-1914. Assuming prairie dogs were distributed throughout the 220,000 sq km of known Montana prairie dog range like the distribution in the survey area, then at least 6, 160 sq km of prairie dogs existed in the state at that time. However, deletion of sev- eral major intrusive mountain ranges from this calculation results in a historic estimate of 5,953 sq km of prairie dogs. Surveys from 1980 to date suggest about 506 sq km of prairie dogs, a 90-1-% reduction. The drastic change in the status of the prairie dog ecosystem is also apparent when 68 Great Basin Naturalist Memoirs No. 8 Fig. 3A. Prairie dog colony complexes in Powder River-O'Fallon Creek area, southeastern Montana (ca 1908-1914). specific townships are compared between 1908-1914 and today. For example, one large prairie dog colony (at least 30 sections and 7,680 ha) in T4N R44E in Rosebud County 1908-1914 consisted in 1978 of only two small colonies totaling about 120 ha, only 2% of its original size. A sampling of five other town- ships for which specific data existed showed at least a 90+% reduction in prairie dog acreage. Discussion Black-tailed prairie dog colonies, with their many associated and in some cases highly de- pendent invertebrate and vertebrate species, formerly occupied large areas of eastern Mon- tana. Black-footed ferrets today are consid- ered the rarest and most endangered mammal in Montana. Most known specimens (N 44) were collected between 1915 and 1953 from 15 counties in eastern Montana (Fig. 2). Re- viewed by Anderson et al. (1986), these records indicate that black-footed ferrets were widely distributed in Montana. The minimum estimated historic prairie dog range of 5,953 sq km scattered in suitable habitat over about 220,000 sq km represented a population dis- tribution similar to that reported for other states during the 1908-1914 period (e.g.. Nel- son 1919, Seton 1929). Our estimates of prairie dog distribution in early Montana more closely approximate prairie dog distri- bution in presettlement times than prairie dog distributions seen today. Indeed, we esti- mate that current prairie dog distribution rep- resents a remnant of probably 10% of the for- mer pattern. The greatly reduced extent of the prairie dog ecosystem resulted from poisoning cam- paigns that began in an organized way in 1915 under the Biological Survey and later under the U.S. Fish and Wildlife Service. No specific data were available on the annual ex- tent of poisoning from 1915 to date for the 1986 Flath, Clark: Black-Footed Ferret Habitat in Montana 69 Fig. 3B. Prairie dog colony complex in Tongue River-Otter Creek area, southeastern Montana (ca 1908-1914). study area. However, a chronological record of poisoning exists for Phillips County, just north of our study area. If Phillips County data are representative of the annual poison- ing efforts for other parts of Montana, and there is every reason to believe that it is (e.g. , Nelson 1919, Campbell and Clark 1981, Hub- bard and Schmidt 1984), then a general chronology of reduction of the prairie dog ecosystem can be established. The systematic poisoning program in Phillips County began in 1917. Over the next 22 years, 15,411 sq km of Richardson's ground squirrels {Spennophihis richardsonii) were poisoned with 168,486 kg of strychnine- soaked grain, and 69,652 ha of prairie dogs were poisoned with 34,109 kg of poison grain (Bureau of Land Management 1982). Because ground squirrels and prairie dogs often exist sympatrically, it is not possible in many cases to determine the target species for poisoning. Prairie dogs were poisoned on 27,530 ha in 1931, on 15,789 ha in 1932, and on 25,911 ha in 1933. Some of this effort was undoubtedly a second or third followup effort, but the extent of repeated poisoning of the same areas is unknown. By the end of 1933, reports men- tioned that very few prairie dogs were left in the county. Limited poisoning continued un- til 1939, when it was felt the species was elimi- nated from the county. Various low-level poi- soning efforts have continued irregularly to the present. With the demise of prairie dogs went reduc- tions in numerous other species, and the black-footed ferret serves as a dramatic exam- ple. If the black-footed ferret occurred at den- sities seen today in the Meeteetse, Wyoming, black-footed ferret area (1 black-footed ferret/ 57 ha) (Forrest et al. 1985), then at least 150,000+ individuals existed from 1908 to 1914 within the Montana prairie dog range. Direct elimination of habitat in some areas and a significant reduction and fragmentation 70 Great Basin Naturalist Memoirs No. 8 of habitat in other areas contributed directly to the reduction, or demise, of the black- footed ferrets. The sample areas for 1908-1914 reported in this paper along the Tongue and Powder rivers and Otter, Pump- kin, and O'Fallon creeks, for example, showed numerous prairie dog colony com- plexes 1.6-11.2 km apart (mean about 1.9 km). These historical Montana prairie dog ar- eas can be compared with the existing Mee- teetse, Wyoming, black-footed ferret/prairie dog complex, which is composed of 37 colonies totaling 2,995 ha (Forrest et al. 1985). Identification of this single complex recog- nizes that the size and distribution of black- footed ferret habitat islands is critical for the continued existence of black- footed ferrets. Forrest et al. (1985) defined a "prairie dog complex" as a group of prairie dog colonies distributed so that individual black-footed fer- rets (and their genetic material) can migrate among them commonly and frequently. Within the Meeteetse complex, mean inter- colony distance is .92 km (range .13-3.70 km) and the mean black-footed ferret intercolony distance movement was 2.5 km. (5.7 maxi- mum). Early Montana prairie dog distribu- tions for 1908-1914 clearly fit the prairie dog complex definition of Forrest et al. (1985). Because of this, the historical Montana prairie dog situation undoubtedly served as high- quality black-footed ferret habitat. This con- clusion is further supported by our under- standing of black-footed ferret habitat requirements in South Dakota (Henderson et al. 1969, Hillman et al. 1979, Hillman and Clark 1980). The early Montana situation rep- resented a habitat setting in which black- footed ferrets evolved among the complex in- terrelationships of species and environmental interactions of the prairie dog ecosystem. The black-footed ferret's energetics, dispersal be- havior, predation avoidance, and litter pro- duction, for example, as seen in the Mee- teetse black-footed ferrets, seem well suited to a universe filled with numerous, large, closely spaced, and stable prairie dog colonies like those in south central Montana from 1908 to 1914 and probably earlier. A few areas in Montana and in other states may still contain sufficiently large prairie dog complexes to support a black-footed ferret population and serve as examples of complex. interactive prairie dog ecosystems. These ar- eas can be compared to the existing Meeteet- see black-footed ferret habitat (prairie dog complex) as described in Forrest et al. (1985). Their value for recovery can be assessed by using a comparative black-footed ferret habi- tat model such as that described by Houston et al. (1986). Prairie dog areas in Montana and elsewhere should be protected, as suggested by Hubbard and Schmidt (1984), as prairie dog refuges. Black-footed ferrets should be reintroduced into appropriate prairie dog refuges once they are described, management agreements are secured, and black-footed fer- rets are available for release. Acknowledgments We thank Jim Bishop of Burlington North- ern Railroad, Inc., for making original survey records available. Ron Wieland assisted with transcribing journal accounts onto data sheets. Denise Casey drew the figures and reviewed the manuscript. Finally, we ac- knowledge Ron Crete for his critical review. We were supported by Montana Department of Fish, Wildlife and Parks. Clark was also supported by the New York Zoological Soci- ety, Wildlife Preservation Trust Interna- tional, World Wildlife Fund— U.S., the Chicago Zoological Society, and others. Literature Cited Anderson, E., S. Forrest. T Clark, and L Richardson. 1986. Paleobiology, biogeography, and systemat- ics of the black-footed ferret, Mustela nigripes (Audubon and Bachman), 185L Great Basin Nat.. Mem. 8;ll-62. Bureau of Land Management. 1980. Habitat manage- ment plan- prairie dog ecosystems (draft). Bureau of Land Management. Billings, Montana. 61 pp. 1982. Black-tailed prairie dog control/manage- ment in Phillips Resource Area. BLM Program- matic Envir. Assessment. Lewistown District, Malta, Montana. 40 pp. and appendices. Campbell, T M III, and T. W Clark 1981. Colony characteristics and vertebrate associates of white- tailed and black-tailed prairie dogs in Wyoming. Amer. Midi. Nat. 105:269-276. Clark. T. W., T M Campbell, D Socha, and D Casey. 1982. Prairie dog colony attributes and associated vertebrate species. Great Basin Nat. 42(4): 572-582. CouES, E 1977. Fur-bearing animals; monograph of North American Mustelidae. U.S. Geological Sur- vey of the Territories, Misc. Pub. No. 8, U.S. GPO, Washington, D.C. 1986 Flath, Clark: Black-Footed Ferret Habitat in Montana 71 Forrest, S. C, T. W. Clark, L. Richardson, andT M. Camp- bell III. 1985. Black-footed ferret habitat: some man- agement and reintroduction considerations. Wyo- ming Bur. Land Manage. Wildl. Tech. Bui. 2. 33 pp. and appendices. Fulton, D. 1982. Failure on the plains. Big Sky Books, Mon- tana State University, Bozeman. 234 pp. Hall, E. R 1981. Mammals of North America. Edition 2. John Wiley and Sons, N.Y. 2 vols., 1,181 pp. Henderson, R F., R F Springer, and R Adrian 1969. The black-footed ferret in South Dakota. South Dakota Dept. Game, Fish, and Parks. Tech. Bull. 4:1-37. HiLLMAN. C. N, andT W. Clark 1980. Mustek nigripes. Amer. Soc. Mammal. Mammalian Species Acct. 126:1-3. Hillman, C. N , R L. LiNDER, and R B. Dahlgren 1979. Prairie dog distributions in areas inhabited by black- footed ferrets. Amer. Midi. Nat. 102:185-187. Houston, B R., T. W Clark, and S. C. Minta. 1986. Habitat suitability index model for the black- footed ferret: a method to locate transplant sites. Great Basin Nat. Mem. 8:99-114. Hubbard, J P , and C. G. Schmidt 1984. The black- footed ferret in New Mexico. Unpublished report, Bur. Land Manage, and New Mexico Game and Fish. 118 pp. Nelson, E W 1919. Annual report of Chief of Bureau of Biological Sui-vey. Pages 275-298. In Annual Rept. Dept. Agric. for Year ended June 1919. Ross, R L and H E Hunter 1976. Climax vegetation of Montana based on soils and climate. USDA Soil Cons. Serv. Bozeman, Montana. 64 pp. Seton, E.T. 1929. Lives of game animals. Double Doran and Co. , Garden City, New York. 949 pp. DESCRIPTION AND HISTORY OF THE MEETEETSE BLACK-FOOTED FERRET ENVIRONMENT Tim W. Clark', Steven C. Forrest", Louise Richardson", Denise E. Casey", and Thomas M. Campbell III^ Abstract. — The black-footed ferret {Mustela nigripes) occupied area lies in the western Big Horn Basin, Park County, Wyoming. Cody, a nearby town, shows a record high temperature of40. 5 C and a low of -43.3 C, with 173 days each year below C. Area geology is dominated by Absaroka volcanics. Soils are shallow (0.5 m) and underlain by unconsolidated gravels; well-drained, medium-textured clay-loams (ca I m in depth); or clays derived from shale parent materials. Vegetation is characterized by a wheatgrass-needlegrass shrubsteppe type {AgropyronlStipal Artemisia). Prior to white settlement, the area hosted a diverse large mammal community. First white settlement began 1878-1885, with establishment of several area ranches. Predator and prairie dog {Cynomys leucurus) poisoning began about 1884. Heavy livestock grazing of public ranges followed the demise of bison (Bison bison) by 1890, which likely was conducive to a continuation of an ungulate-range relationship favoring prairie dog habitat. Ferret specimens from Crow Indian inhabitants of the region date to 1880s and two specimens from Park County date from the 1920s-1930s. Today ferrets are found on white-tailed prairie dog colonies (a "complex") totaling ca 2,995 ha. The areas occupied by these colonies are equally owned by private, state, and federal interests. Evidence shows many abandoned prairie dog colonies which, along with the current ones, total about 8,400 ha. Many of them may have been active simultaneously prior to poisoning in the 1930s. This paper summarizes some physical and biological characteristics of the Meeteetse, Wyoming, black-footed ferret (BFF) environ- ment, serves as a general description of the region, and provides a partial description of BFF habitat. It focuses on land uses, past and present, including prairie dog poisoning pro- grams. Methods A general description of the western half of Wyoming's Big Horn Basin (the general BFF study area) was obtained from numerous site visits between October 1981 and March 1985. Extensive conversations with ranchers, histo- rians, anthropologists, state and federal wildlife managers, and literature reviews pro- vided further understanding of the area. Prairie dog colonies in the general study area, which potentially serve as BFF habitat (Lin- der et al. 1972, Hubbard and Schmidt 1984, Anderson et al. 1985, Forrest et al. 1985), were located by air and ground surveys and interviews with landowners. Summer spot- lighting surveys and winter snow-tracking surveys determined the distribution of BFF- occupied prairie dog colonies. An intensive study area was delineated within the larger study area from these data. All prairie dog colonies were mapped on 1:62500 USCS to- pographic quads. BFF-occupied colonies were mapped on 1:4800 base maps we pre- pared to detail site features. Historical infor- mation was obtained from the literature and interviews with area ranchers and participants in prairie dog poisoning programs. "Dead" prairie dog colonies were identified by the presence of unused, revegetated prairie dog mounds as described by Clark (1970). The Environment The Meeteetse study area is named after a small community in the Big Horn Basin of northwestern Wyoming (Fig. 1). The larger extensive study area includes most of the western half of the Basin (8,000 sq km) includ- ing parts of Park, Hot Springs, Big Horn, and Washakie counties. The Basin is enclosed by mountains on the west, south, and east and is open to the north. The Shoshone, Greybull, and Bighorn rivers drain the Basin. The smaller intensive study area containing the BFFs is also shown in Figure 1. Portions of the larger study area have been described by the Department of Biological Sciences. Idaho State University. Pocatello, Idaho 83209. ^Biota Research and Consulting, Inc., Jackson, Wyoming 83001. 72 1986 Clark et al.; Meeteetse Environment 73 MONTANA CO 30mi 50 km OWL CREEK MOUNTAINS Larger Study Area Intensive Study Areo Fig. 1. Location of the larger and intensive black-footed ferret study areas in the Big Horn Basin of northwestern Wyoming. U.S. Forest Service (1982) and the U.S. Bu- reau of Land Management (1982). The inten- sive study area and ferret use of that area were described by Forrest et al. (1985) and mod- eled by Houston et al. (1986). Climate Climatographs for Cody, Wyoming, located north of the intensive study area, and Ther- mopolis, Wyoming, at the south end, are shown in Figure 2. The record high tempera- ture for Cody is 40.5 C and the record low is -43.3 C, with 173 days each year below C based on 1960 U.S. Department of Com- merce records. The Thermopolis record high is 41. 1 C and the low is -41. 1 C with 194 days below C. Winds are estimated to average 13-16 kmph at Cody, with lower velocity winds at Thermopolis. The Meeteetse area is generally snow-free, because of wind action. Occasional accumulations, generally 10 cm or less, may occur for several days at a time. 74 Great Basin Naturalist Memoirs No. 8 THERMOPOLIS, WY CODY.WY 2 3 4 PRECIPITATION (CM) Fig. 2. Climatographs for Cody and Thermopolus, Wyoming (1931-1960). 2 3 4 PftECIPITATION (CM) Prevailing winter winds are westerly. Mean annual isohyets for the Big Horn Basin are shown in Figure 3. Geology Area geology is dominated by Absaroka vol- canics that compose most of the Absaroka Range and Carter Mountains immediately west of the study area. Surface geology is de- scribed by Pierce (1978) and shown in Figure 4. Most prairie dogs are associated with Cody shales or unconsolidated sediment. Prairie dog association with shale-derived soils has been identified in other studies (Stromberg 1975, Knowles 1982). Shale parent materials may provide clayey soils that are structurally more stable for burrow construction. Known geological structures for oil and gas in the region are shown in Figure 5. Oil and gas potential for much of the intensive study area is rated high (U.S. Forest Service 1982). Following discovery of the BFFs, 41 mineral leaseholders were notified in March 1982 of possible changes in lease status by the U.S. Bureau of Land Management. The affected area included about a 1 km buffer zone around the intensive study area. Soils Soils are shallow (0.5 m and underlain by unconsolidated gravels); well drained, with medium-textured clay-loams (ca 1 m in depth); or clays derived from shale parent materials. Additional soil descriptions are given by Collins and Lichvar (1986). Vegetation Vegetation is characterized by Kuchler's (1964) description of wheatgrass-needlegrass shrubsteppe type {AgropijronlStipal Artemi- sia). Vegetation of the intensive study area is dominated by Koeleria cristata, Agropyron spicatum, A. smithii, and mixed shrub (largely Artemisia tridentata) as described by Collins and Lichvar (1986). Vegetation has been heav- ily grazed by cattle, horses, and sheep for about 100 years. Prairie Dogs Current prairie dog distribution within the intensive study area is shown in Figure 6. The 37 colonies shown total 2,995 ha and contain about 125,000 prairie dog burrow entrances. BFF occupancy has been noted in 23 of these colonies. Prairie dog burrow openings aver- age 41.7 per ha, and prairie dog densities reach 9 per ha (Clark et al. 1985). BFF use of these colonies is described by Forrest et al. (1985). Surface and subsurface ownership is pre- sented in Table 1. Surface ownership is about equally divided among state (31.0%), federal (33.4%), and private (35.6%) entities. Subsur- face ownership is 57% federal, 31% state, and 1986 Clark et al.: Meeteetse Environment 75 Larger Study Area Intensive Study Area Fig. o. The mean annual isohyets for the Big Horn black-footed ferret study areas and major towns. 12% private. Seven ranches contain BFFs, with about two-thirds of the total prairie dog colony area on one ranch. The other six ranches each have l%-9% of the total BFF area. Comparisons of the Meeteetse area with eight other prairie dog study areas are shown in Table 2. Ten variables are contrasted among these areas. The Meeteetse BFF/ prairie dog site falls within ranges for these variables, except that it shows a greater mean I of northwestern Wyoming showing the larger and intensive burrow opening density and lower inter- colony distance than the other areas. Unfortu- nately, data are not complete in all cases for comparative purposes. Land Use History The Big Horn Basin was opened to white settlement in the mid-1870s. Previously the area was used as hunting and wintering ground by Mountain Crow Indians, whose major impact was likely restricted to occa- Fig. 4. Geological map of the intensive black-footed ferret study area showing prairie dog colonies in relation to underlying geology. Kp — Cody Shale (Upper Cretaceous) — Upper part buff sandy shale and thinly laminated buff sandstone. Lower part dark-gray, thin-bedded marine shale. Thickness 500-1000 m. Qp — Pediment deposits — Thin veneer of poorly rounded to subangular surficial material deposited on smooth, gently sloping erosion surfaces cut in bedrock. Q, — Terrace gravel — Unconsolidated deposits of gravel, sand, cobbles, and silt. Kn,^, — Mesaverde Formation (Upper Cretaceous) — Interbedded light gray sandstone and gray shale in upper part; lower part massive lightbuff, ledge-forming sandstone containing thin, lenticular coal beds. Kf — Frontier Formation (Upper Cretaceous) — Thick lenticular gray sandstone, gray shale, brown carbonaceous shale and bentonite. Torchlight sandstone. Kn, — Meeteetse Formation. Q3 — Alluvium — Unconsolidated deposits of silt, sand, gravel, and cobbles along stream valley and at or near present stream level. Includes alluvial fans and glacial outwash. Q^ — Colluvium — Heterogeneous deposits of rock detritus. Qi — Landslide deposits — Heterogeneous deposits of rock debris emplaced by mass movement. T„ — Wapiti Formation (Eocene) — Dark-brown andesitic breccia, tuff, volcanic sandstone, siltstone, and conglomer- ate; lava flows and flow breccias; dark to medium-brown pyrozene andesite; sparse horneblende. Includes predominantly volcanic sandstone and siltstone of the Pitchfork formation of the Sunlight Croup of the Absaroka Volcanic Supergroup in upper Greybull River area. Thickness 1000-1500 m. sional ground fires (Edgar and Turnell 1978). During 1878-1885 several area ranches were established, most notably the Pitchfork Ranch founded by Otto Franc. The Pitchfork Ranch grazed about 15,000 head of cattle on various ranges throughout the Basin by 1884, encom- passing virtually all the lands in the intensive BFF study area (Edgar and Turnell 1978). The Pitchfork Ranch incorporated surrounding ranch properties in the period 1903-1922, en- compassed 100,000 ha, and grazed 12,000- 20,000 head of cattle and 60,000 head of sheep. As D. Healy (in Killough 1977) points out, estimates of range use during the open range period are difficult to assess, and little experi- ence concerning the productivity of fenced allotments was available prior to the passage of the Taylor Grazing Act in 1934. Carrying capacity estimates were likely too optimistic, resulting in heavy overuse of public range. Killough (1977) states that this probably oc- curred throughout the Bighorn Basin. By the 1930s the Pitchfork was grazing about 20,000 sheep and 5,000 cattle on 28,000 ha of deeded land, 44,000 ha of leased land, and 24,000 ha 1986 Clark ETAL.. Meeteetse Environment 77 RI04W 103 102 101 100 99 R98 W 97 \ Heart Mountain coor ^Shoshone North %Shoiihone If) o Gypsum quarry A Half MoonI Hunt* Ferguson Ranch! Oregon Basin West% Sprini Creek SoutM Willow Draw Rose Creek^ )rawmj^ Rawh burbear Sheep Peintt J % Sunshine Nortfr WOregon Basin I Oregon Basin de tchfork" • Meeteetse OMEtTEETSE .Little Buffalo Basin 1X\ ^Gooseberry KM 10 -J Intensive Ferret Study Area Fig. 5. Map of oil and gas fields and mining regions for the western Big Horn Basin in relation to the intensive black-footed ferret study area. of permitted land (Turnell 1982, personal communication). The LU Ranch on Grass Creek, south of Meeteetse, which had poorer range conditions, controlled a comparable 100,000 ha and grazed 1,500 cattle and 15,000 sheep (Killough 1977). Oil activities, beginning in the 1950s, in- cluded seismic testing for underlying geologi- cal structures and a concomitant increase in primitive roads to maintain wells, pipelines, and support facilities. In recent years seismic activity has increased within the intensive 78 Great Basin Naturalist Memoirs No. 8 Fig. 6. Prairie dog colonies in the Meeteetse intensive black-footed ferret study ; study area. In 1981 prior to discovery of the BFF population, four wells were drilled on the Rose Creek Field directly on the largest BFF concentration (Fig. 5). In addition, two subsurface pipelines and one pumping station are located in BFF-occupied prairie dog colonies. Fauna Prior to white settlement, the area probably hosted a diverse faunal assemblage dominated by grazing ungulates and their various preda- tors. Killough (1977) found bison skulls ex- posed as deep as 3.6 m in gullies in the Big Horn Basin, showing long-term use of the area by bison and suggesting that periods of overgrazing and erosion followed by range restoration (gully healing) may have been common within recent history. White trap- pers recorded large bison herds in the upper Greybull drainage as late as 1878. James White, who worked in the Greybull River country, secured 2,000 hides in 1880 alone (Edgar and Turnell 1978). The last locally known native bison was killed along Mee- teetse Rim in 1892 (Edgar and Turnell 1978). Although bison skeletal remains were inten- sively scavenged for fertilizer markets throughout the West, numerous bones and horn sheaths can still be found in this area today. Otto Franc recorded the presence of liter- ally thousands of bighorn sheep {Ovis canadensis) wintering along the Greybull in 1880, and Archibald Rogers noted on a hunt- ing trip to the -TL Ranch (later Pitchfork) in 1893 a band of 250 bull elk {Cervus ela- phus){cited in Egdar and Turnell 1978). By 1890 Franc noted a drastic decline in big game numbers from a decade of unchecked ex- ploitation by sport, skin, and market hunters. Big game numbers continued to decline in the area through the early 1900s, although pronghorn (Antilocapra amehcana) were protected on Pitchfork Ranch lands and occa- sionally transplanted elsewhere. Big game lunnbers have experienced an apparent re- covery and have continued to increase since a low during the 1940s (J. Lawrence and J. Tur- nell, personal commimication). Grizzly bears {Ursus arctos) were apparently quite com- mon, as records of bear encounters abound (e.g., Seton 1899). By 1894, gray wolves {Canis lupus) began depredations on live- 1986 Clark ETAL.: Meeteetse Environment 79 Table 1 . Size, surface, and subsurface ownership patterns of prairie dog colonies in the intensive black-footed study area. Colony Colony size Land Status Si State arface ownership Federal Private Subsurface ownership Colony State Federal Private number name (ha)^ % % % % % % 1 Long Hollow Corrals 196.5 75 25 75 5 20 2 Long Hollow West 26.5 65 35 100 3 Long Hollow South 2.5 85 15 85 15 4 Lot 58 12.5 100 100 5 Section 30 1.5 100 100 6 Rush Creek Basin 24.0 70 30 70 30 7 Rush Creek B 3.0 100 100 8 Rush Creek C 8.5 100 100 9 Rush Creek D 5.0 100 100 10 BLMIO 49.5 100 100 11 BLM14 20.5 100 100 12 Little Rawhide West 0.5 100 100 13 BLM13 183.0 20 80 20 80 14 Rawhide West 0.5 100 100 15 Westbrook Draw 2.5 100 100 16 Lower BLM 50.5 — 85 15 48 52 17 Little Rawhide #1 1.0 100 100 18 Little Rawhide #2 2.0 — 100 100 19 Little Rawhide #3 6.0 100 100 20 Little Rawhide Basins 9.0 60 40 60 40 21 Rawhide 102.0 7 3 90 7 3 90 22 Pump station 230.0 45 55 65 35 23 Tonopah 31.5 100 100 24 Thomas 31.5 100 100 25 EAST East Core 738.5 47 53 100 25 WEST West Core 568.5 80 11 9 80 14 6 26 91 Town 97.5 75 25 75 8 16 27 Fence 0.5 100 100 28 Rose Creek 158.0 75 25 75 25 29 Island 21.0 100 100 30 Pickett Creek 211.0 5 85 10 5 85 10 31 Graveyard 51.5 60 40 60 40 32 Hogg 30.5 100 100 33 Rose Creek West 9.5 100 100 34 Spring Creek Basin A 69.5 — 100 — 100 35 Spring Creek Basin B 3.5 100 100 36 Spring Creek Basin C 11.0 — 100 — 100 37 Spring Creek Basin D 245 — 100 — 100 Total 2995.0 Stock, possibly as a result of decimated large mammal populations. Wolf predation con- cerned Franc until his death in 1902 (Edgar and Turnell 1978), and Mrs. H. C. Larsen of Wood River also noted that wolves were nu- merous from the late 1800s until the 1920s (Diem 1973). The historic relationship between prairie dogs and bison can only be surmised, al- though the role of bison in grassland ecosys- tems has been debated for some time (Larson 1940). Nevertheless, there is Uttle doubt that there was a reciprocal ecological relationship between bison and prairie dogs, each tending to maintain the shortgrasses interspersed with patches of forbs, ideal habitat for each other (Koford 1958). Osborn and Allen (1949) and King (1955) also noted that bison tended to concentrate on prairie dog colonies, their ac- tivities apparently creating an environment that favored prairie dogs. These bison activi- ties included "over grazing," trampling soil, wallowing, and defecating and urinating. In the late 1850s, Mead (1899) noted that prairie dogs disappeared from parts of Kansas shortly after the bison and concluded that bison were Great Basin Naturalist Memoirs No. 8 Table 2. Some physical characteristics of white-tailed prairie dog colony complexes near Meeteetse, Wyoming, compared to other prairie dog complexes in Wyoming and Utah. Characteristics Meeteetse Vernal UT Cisco UT Polecat Bencl I Cumberland South South Totals WY(this WY (Campbell (Clark et (Clark et NVV \VT (Clar k Flats S\V Central Central or study) & Clark 1981) al. 1982) al. 1982) 1977) Wyoming' VVyomin g^ Wyoming^ Means Study area size (km^) 333 336 1886 298 128 256 556 483 534.5 Number of colonies 37 25 18 15 4 63 164 81 388 Total area of colonies (ha) 299.5 1085 3584 566 .3020 3969 4008 4298 23,388 Percent of study area 9.0 3.2 1.9 1.9 2.3 15.5 7.2 8.9 6.2 Number of ha of dogs/ 100 km^ 899 320 190 190 2359 1550 720 890 884.6 Number of colonies/100 km" '■ 11.1 7.4 1.0 5.0 .3.1 24.6 29.5 16.8 11.8 Colony size (ha): Mean 80.9 43 199 38 755 63 24 53 163.2 SD 217.2 46.4 249 37 — — — — — Range 0.5-1307 2-184 0.2-958 9-121 120-1400 0.4-2000 0.8-510 — — Number of burrow openings examined 125,000 24,620 76,579 8,993 6,775 168,761 105,497 129,969 630,049 Burrow openings/ha: Mean 41.7 25.1 30.8 19.8 2.1 43 4 30 26.5 SD 62.1 26.4 37.6 10.2 — — — — — Range 13.7-290.7 9-129 5.1-160 2.3-40.5 — 4.2-130 0.8-41 — — Intercolony distance (km): Mean 0.9 1.5 4.4 5.5 — — — — — SD 0.78 1.0 — — — — — — — Range 0.06-3.1 0.4-3.6 0.8-11.3 0.8-11.3 — — — — — 'Clark et al. 1982; Martin and Schroeder 1979, 1980. ^Martin and Schroeder 1979. ^Martin and Schroeder 1980. necessary for prairie dog existence because they compacted soils and created communi- ties of forbs. Clark (1973) presented a model hypothesiz- ing the interrelationship of bison and prairie dogs, in which both animals functioned in a reciprocal manner ecologically to increase grassland productivity beyond what each spe- cies could contribute individually. More re- cently Bonhan and Lerwick (1976), O'Mellia et al. (1980), Uresk and Bjugstad (1980), and Coppock et al. (1983) have studied prairie dog ecology and, in some instances, prairie dog- bison relationships. Their results support the earlier observations and models described above. When large numbers of domestic live- stock replaced the bison at the turn of the century, they may have continued to provide an environment conducive to prairie dog oc- cupancy of the Meeteetse area. Historic Ferret Habitat Because BFFs retjuire prairie dogs as part of their habitat, a review of historic prairie dog status and poisoning programs in the region pro- vides data on historic BFF habitat. BFFs have most likely occupied the Meeteetse area and Big Horn Basin since pristine times. Crow Indians of the region collected BFFs as medicine objects in the mid- to late 1880s (Clark 1975). Clark (1977) listed 20 BFF reports from the Big Horn Basin from 1889 to 1977. Two of these reports originated in Park County, in the vicinity of the intensive study area. Ed Larson and Frank Smith, oldtime residents of Meeteetse, report trapping or knowing of trapped BFFs in the 1920s- 1930s from the Meeteetse Creek area. Cal Todd, former manager of the Pitchfork Ranch, and his wife, Margo, reported that their dog killed a BFF in 1962 on the headquarters groimds. The corpse was described to George Reesy, Wyoming Game and Fish Department, who did not investigate it. The skin was retained for about 6 yrs and later lost. In September 1981 a male BFF was killed hv a dog on the John Hogg Ranch (Clark and Campbell 1981). Subse- (jiiently, a nearby BFF population was located by Doug Brown, a cowboy, which lead to the present study. 1986 Clark etal.: Meeteetse Environment 81 '*::♦•. '••J ^ ACTIVE COLONIES HISTORIC COLONIES Ml 3 I r^-r— ^ H KM 4 ) Fig. 7. Location of currently active and historic "dead" prairie dog colonies, near Meeteetse, Wyoming, 1984. 82 Great Basin Naturalist Memoirs No. 8 Prairie dog control programs within the in- tensive study area began in the 1880s (Edgar and Turnell 1978) and continued sporadically until the mid- 1930s. From 1923 to 1928 ro- dent control expenditures for Park County totalled $7,476, the third highest county ex- penditure in the state. During the same pe- riod more than 500,000 ha of prairie dogs were eradicated in Niobrara, Weston, and Camp- bell counties (Day and Nelson 1929), so con- siderable control activity was occurring. Over a five-year period during the mid- to late 1930s, large and well-organized poisoning programs were conducted throughout most, if not all, of the intensive and general study areas (B. Sells, E. Larson, F. Smith, personal communication). All of Meeteetse, Rush, Rawhide, and Spring creeks, and parts of the Greybull and Wood River drainages were poi- soned, as well as much of the area north of Meeteetse Creek up to Cody. These federal programs poisoned only a portion of the total area during any one year, and, even though "kill rates" are undocumented, 50%-100% kills of prairie dogs were often obtained else- where (e.g., Tietjen 1976). The entire prairie dog complex was appar- ently never all poisoned in a single year. Be- ginning in the 1940s and continuing through the 1960s, limited poisoning was carried out on specific colonies or areas (B. Rosan, J. Win- ninger, D. Winninger, J. Turnell, J. Hogg, personal communication). This suggests that the 1930s campaigns were effective in elimi- nating prairie dogs, leaving few to be poi- soned later (M. Todd, personal communica- tion). This pattern has been seen in other areas where adequate data exist, such as Phillips County, Montana (Bureau of Land Management 1982) and in eastern Wyoming (Clark 1973, Campbell and Clark 1981). Since 1970 only a few small areas have been poi- soned (J. Hogg, J. Winninger, J. Turnell, A. Thomas, B. Gould, personal communication). Since "dead" prairie dog colonies may retain their identity for 60+ years (Clark and Camp- bell, unpublished data), we used areas for- merly occupied to get a maximum upper size (after Clark 1970) of the prepoisoning colony complex. We assumed that all the "dead" prairie dog colonies seen today were simulta- neously active prior to the large 1930s poison- ing campaigns, plus those colonies currently active. The estimated total of 8,400 ha of colonies may have been present prior to the 1930s (Fig. 7). If one assumes that these 8,400 ha of prairie dog colonies could support one BFF for each 50 ha (Forrest et al. 1985), then a maximum estimate of the pre-1930s BFF pop- ulation in the pre-1930s can be obtained. As- suming all else is equal between today's ob- served BFF population size, density, and reproductive rate, the pre-1930 BFF habitat could have supported as many as 168 adult BFFs. Discussion The overall climate, geology, soils, vegeta- tion, and land use history of the Meeteetse area, which today contains the world's only known BFF population, is similar to much of Wyoming and other areas throughout the West where prairie dogs are found. BFFs probably always inhabited the Meeteetse re- gion. Their persistence there today is proba- bly due to: (1) a historical abundance of prairie dog habitat, (2) prairie dog control programs that left active colonies or parts of colonies unpoisoned during any one year, and (3) ab- sence of catastrophic diseases (sylvatic plague, distemper). Prairie dogs were apparently kept at low levels after the 1930s, and this "bottle- neck" for BFFs persisted for some years, per- haps having genetic consequences for the BFFs of today (Pettus 1985, Kilpatrick et al. 1986). Acknowledgments This study was generously supported by The Nature Consevancy-Big Sky Field Of- fice, the New York Zoological Society (Wildlife Conservation International), and the Wildlife Preservation Trust International, Inc. B. Edgar, D. Flath, and B. Miller pro- vided critical review of the manuscript. We appreciate the contribution of Meeteetse ranchers who protect ferret habitat and those people and agencies who have made our stud- ies possible. Literature Cited Anderson, E., S. C. Forrest. T. W. Clark, and L. Richardson. 1986. Paleobiology, biogeography, and systematics of the black-footed ferret, Mustela ni^ripcs (Audubon and Bachman), 1851. Great Basin Nat. Mem. 8:11-62. 1986 Clark et al,: Meeteetse Environment 83 BONHAM, C. D., AND A Lerwick. 1976. Vegetation changes induced by prairie dogs on shortgrass range. J. Range Mgt. 29:217-220. Campbell, T M III, and T. W. Clark 1981. Colony characteristics and vertebrate associates of white- tailed and black-tailed prairie dogs in Wyoming. Amer. Midi. Nat. 105:269-276. Clark, T. W. 1970. Revegetation patterns of white-tailed prairie dog burrow mounds. Wyoming Range Mgt. 280:8-12. 1973. A field study of the ecology and ethology of the white-tailed prairie dog (Cynomys leucurus): with a model of Cynomys evolution. Unpublished dissertation, Universitv of Wisconsin, Madison. 215 pp. 1975. Some relationships between prairie dogs, black-footed ferrets, Paleo-Indians and ethno- graphically known tribes. Plains Anthrop. 20-67:71-74. 1977. Ecology and ethology of the white-tailed prairie dog (Cynomys leucurus). Milwaukee Pub. Mus. Publ. in Biol, and Geol. No. 3. 97 pp. Clark, T W . andT M Campbell III. 1981. Additional black-footed ferret [Mustela nigripes) reports from Wyoming. Great Basin Nat. 41:461^64. Clark, T W , T M Campbell III, D C Socha, and D E Casey. 1982. Prairie dog colony attributes and associated vertebrate species. Creat Basin Nat. 42:572-582. Clark, T. W., L. Richardson, S. C. Forrest, T. M. Camp- bell III, D. E. Casey, and K A. Fagerstone. 1985. Black-footed ferret prey base. Pages 7.1-7.14 in S. Anderson and D. Inkley, eds.. Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ. , Cheyenne. Collins, E. I , and R. W. Lichvar. 1986. Vegetation in- ventory of current and historic black-footed ferret habitat in Wyoming. Great Basin Nat. Mem. 8:85- 93. CoppocK, D L J K Detling.J. E Ellis, and MI Dyer 1983. Plant-herbivore interactions in a North American mixed-grass prairie: I. Effects of black- tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plant species diversity. Oecologia 56:1-9. Day, a. M., and A. P. Nelson. 1929. Wild life conserva- tion and control in Wyoming under the leadership of the United States Biological Survey. Coop. Publ. U.S. Biol. Surv., Wyoming State Game and Fish Dept., Wyoming State Ext. Serv., State Dept. Agric, Wyoming State Hist. Res. Publ. 36 pp. Diem, K. L. 1973. Wildlife: Kirwin Environmental Study. Unpublished report to ROMCOE, Denver, Colo- rado. 85 pp. Edgar, B , and J. Turnell. 1978. Brand of a legend. Stockade Publ. , Cody, Wyoming. 244 pp. Forrest, S. C, T. W. Clark, L. Richardson, and T. M. Campbell III 1985. Black-footed ferret habitat: some management and reintroduction consider- ations. Wyoming Bur. Land Mgmt. Wildl. Tech. Bull. No. 2., Cheyenne. 49 pp. Houston. BR.TW Clark, andSCMinta 1986. Habi- tat suitability index model for the black-footed ferret: a method to locate transplant sites. Great Basin Nat. Mem. 8:99-114. Hubbard. J P., and C. G. Schmidt 1984. The black- footed ferret in New Mexico. Unpublished report, Bur. Land Manage, and New Mexico Game and Fish Dept. 118 pp. KiLLOUGH, T. R. 1977. Land use and environmental analy- sis for the LU Sheep Company. Unpublished re- port, 112 pp. KiLPATRiCK, C. W., S. C. Forrest. andT. W. Clark. 1986. Estimating genetic variation in the black-footed ferret-a first attempt. Great Basin Nat. Mem. 8:145-149. King. J A. 1955. Social behavior, social organization, and population dynamics in a black-tailed prairie dog colony in the Black Hills of South Dakota. Univer- sity of Michigan Contrib. Lab. Vert. Biol. 67:1-123. Knovvles, C. J. 1982. Habitat affinity, populations, and control of black-tailed prairie dogs on the Charles M. Russell National Wildlife Refuge. Unpub- lished dissertation. University of Montana, Mis- soula. 171 pp. KOFORD, C. B. 1958. Prairie dogs, white faces, and blue grama. Wildl. Monogr. No. 3. 78 pp. KUCHLER, A. W. 1964. Potential natural vegetation of the conterminous United States. Amer. Geographical Soc, New York, Spec. Pub. No. 36. 117 pp. Larson, F 1940. The role of bison in maintaining the short grass plains. Ecology 21:113-121. LiNDER, R. L.,R B Dahlgren,andC.N. Hillman. 1972. Black-footed ferret-prairie dog relationships. In Symp. on Rare and Endangered Wildlife in S. W. United States, September 22-23, Albuquerque, New Mexico, New Mexico Dept. Game and Fish, Sante Fe. Martin, S. J., and M. H. Schroeder. 1979. Black-footed ferret surveys on seven coal occurrence areas in southwestern and southcentral Wyoming, June 8 to September 25, 1978. Final Rept., Wyoming StateOffice, BLM.37pp. 1980. Black-footed ferret surveys on seven coal occurrence areas in Wyoming, February-Septem- ber, 1979. Final Rept., Wyoming State Office, BLM.39pp. Mead, J. R 1899. Some natural history notes of 1899. Trans. Kansas Acad. Sci. 16:280-281. O Mellia, M. E., F. L. Knoph, and J. C Lewis 1982. Some consequences of competition between prairie dogs and beef cattle. J. Range Mgt. 35:580-585. Osborn, B , and p. Allen 1949. Vegetation of an aban- doned prairie dog town in tallgrass prairie. Ecol- ogy 30:322-332. Pettus, D. 1985. Genetics of small populations. Pages 22.1-22.11 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Pierce, W. G 1978. Geological map of the Cody 1 by 2 quadrangle northwestern Wyoming. U.S. Geol. Surv., Denver, Colorado. 84 Seton E T 1899. The biography ofa grizzly. Grosset and Dunlap Publ. , New York. 167 pp. STROMBERG, M. R. 1975. Habitat relationships of the black-tailed prairie dog {Cynomys ludovtcianus): vegetation, soils, comparative burrow structure and spatial patterns. Unpublished thesis. Univer- sity of Wisconsin, Madison. 175 pp. Teitten H P 1976. Zinc phosphatae-its development as a control agent for black-tailed prairie dogs. USDI, U.S. Fish and Wildlife Serv., Spec. Sci. Rep. Wildl. 195. Washington, D.C. 14 pp. Great Basin Naturalist Memoirs No. 8 U S Bureau of Land Management. 1982. Grass Creek grazing draft EIS. U.S. Dept. Inter., Worland, Wyoming. 2 vol. U S Forest Service. 1982. Oil and gas exploration and leasing within the Washakie Wilderness draft EIS. U.S. Dept. Agric, USES, Shoshone Natl. For., Cody, Wyoming. 206 pp. Uresk D W.andA.J.Bjugstad 1981. Effects of prairie dogs and cattle on vegetation of the northern high plains. South Dakota Stockgrowers 36:27-28. VEGETATION INVENTORY OF CURRENT AND HISTORIC BLACK-FOOTED FERRET HABITAT IN WYOMING Ellen I. Collins' and Robert W. Lichvar Abstract — A vegetation inventory of current and historic black- footed ferret (Mustela nighpea, BFF) habitat was completed in the 1983 growing season. White-tail prairie dog {Cynotnys leucunis) colonies near Meeteetse, Wyoming, provide the only known BFF habitat. Four other prairie dog complexes located in Wyoming, documented historic BFF habitat, were also inventoried. Prairie dog burrows occur in two of eight vegetation types present in the current BFF habitat study area. They are junegrass {Koeleria cristata) and sagebrush/junegrass {Artemisia tridentatal K. cristata; named for dominant species). In historic BFF habitat, prairie dog burrows occur in six vegetation types: birdfoot sagewort/western wheatgrass (Artemisia petadifidalAgropyron smithii), alkali sagebrush (Artemisia longiloba)/ mixed grass, Gardner saltbush (Atriplex gardneri)! mixed grass, and thickspike wheatgrass-threadleaf sedge (Agropyron dasijstachyum-Carexfilifolia), mixed shrub/mixed grass, and Gardner saltbush. Similarities among all five complexes are plant heights generally less than 66 cm, level to gently rolling topography, and severe disturbance due to historical grazing, vegetation manipulation, and other human-related activities. Of the vegetation parameters measured, only plant height appears to be important to white-tail prairie dog distribution. White-tail prairie dog colonies do not appear to depend on a particular vegetation type; consequently, vegetation type alone should not be used to identify BFF habitat. A comparative vegetation inventory includ- ing cover, shrub density, and plant height measurements with qualitative field observa- tions of current and historic BFF habitat was completed in summer 1983. This study was conducted as part of a comprehensive pro- gram designed to identify parameters impor- tant to the selection of potential BFF reloca- tion sites. This paper presents results of the study. Table 1. Location of five study sites in Wyoming. White-tail prairie dog comple Countv Legal location Current ferret habitat 1. Meeteetse complex Core Colony 91 Colony Pickett Creek Colony Thomas Colony Lower Bench Colony Tonopah Colony Pump Station Colonv BLM-13 Colony Island Colony Graveyard Colony Timber Creek Colony^ (abandoned) Historic ferret habitat 2. Larsen s Ranch Complex 3. Wasmer Flats Complex 4. Gillies Draw Complex 5. Home Flats Complex Park T48N R102W; T48N R103W Park T48N R102W; T48V2N R102W Park T48N R103W Park T48N R102W Park T48N R102W; T49N R102W Park T48N R102W Park T48N R102W Park T49N R102W Park T48N R103W Park T48N R103W Park T48N R102W; T47N R102W Sweetwater T22N R93W Uinta T16N R118W Park T47N R98W; ' r48N R98W Carbon T22N R78W; ' T21N R78W id abandoned in 1960s. Prairie dog colony names follow Forrest et al. (198.5). ^No evidence of BFF occupancy 1981-1984; probably poisoned ; '624 Rene Lane, Fort Collins. Colorado 80524. ^1216 West 31st Street, Cheyenne, Wyoming 82001. 85 86 Great Basin Naturalist Memoirs No. Table 2. Summary comparison of important descriptive parameters for the five study sites. Prairie dog Vegetation Total Shrub Plant Elevation Topography complex type cover (%) density (stems/m' height ') (cm) (m) Current ferret habitat 2134-2256 RoUing hills 1. Meeteetse Complex and flat benches Core Colony Junegrass 73 0.3 13-46 Level to SW- facing; 0°-3° slope Sagebrush/ 67 1.3 21-62 Junegrass 91 Colony Junegrass 63 1.3 15-31 Level to gently sloping; 0°-6° slope Pickett Creek Junegrass 74 0.4 15-31 Level to gently Colony sloping S -facing; 0°-2° slope Thomas Colony Junegrass 65 1.5 31-46 SE-facing; gentle slope Lower Bench Colony Junegrass 73 0.6 31-46 SW-facing; 0°-2° slope Tonopah Colony Junegrass 73 1.3 18-62 E-facing; gentle slope Pump Station Colony Junegrass 72 1.3 16-62 Level to gently sloping valley BLM-13 Colony Junegrass 83 0.4 31-62 Nearly level Island Colony Junegrass 71 0.4 15-51 Nearly level Graveyard Colony Junegrass 71 0.1 15-62 Nearly level Timber Creek Colony Junegrass 73 0.4 31-51 Nearly level (Abandoned) Historic ferret habitat 2. Larsen's Ranch Birdfoot 39 2.5 5-31 2042 Level; 0°- 10° slope Complex sage/grass 3. Wasmer Flats Alkali 53 1.3 (live) 5-62 2095 Level to gently sloping Complex sagebrush 0.7 (dead) 4. Gillies Draw Gardner 58 0.8 8-92 1652-1808 Highly dissected; series Complex saltbush of parallel draws and ridges 5. Home Flats Complex 2042 Extensive, level bench Wheatgrass- 66 1.1 26-41 sedge Mixed shrub/ 59 1.7 31-36 grass Gardner 56 1.6 10-41 saltbush 'Becker and Alyea, 1964a ^Becker and Alyea, 1964b G = Grazed PL = Pipeline PW = Powerline OW= Oil well TL = Transmission line Methods Ten of 33 white-tail prairie dog colonies uti- lized by BFF's and an abandoned colony near Meeteetse, Wyoming (Forrest et al. 1985), were arbitrarily selected for study (Table 1). Four historic BFF habitat sites (Larsen's Ranch, Wasmer Flats, Gillies Draw, Home Flats; Table 1) were selected for comparison with the current BFF habitat site at Mee- teetse. Historic habitat was identified based on the presence of BFF skeletal remains or sign (Martin and Schroeder 1979, 1980, Clark 1980, Clark and Campbell 1981). All five sites were similarly inventoried. Vegetation type units were delineated on 7.5 minute USGS topographic maps during field reconnaissance and named based on aspect dominance. Sampled types were later re- named for the dominant-co-dominant spe- cies. Thirty randomly selected plots per vegeta- tion type containing prairie dog burrows were sampled in each of the 11 mapped Meeteetse colonies. Colony boundaries within the four historic BFF habitat sites were not delineated for the authors, and no attempt to do so was 1986 Collins, Lichvar:Vegetation Inventory 87 Mean monthly temp, extremes (°F)' High Low Mean annual precipitation (cm)^ Disturbance'^ Additional vegetation types present not containing burrows 45 21 43.9 G;PL;OW;TL G;PL G G Mixed shrub; mixed grass; riparian; greasewood; rabbitbrush; wet meadow G G G G;PL Wet meadow G; PL; PW G G G 47 7 16.7 G Sagebrush 43 4 27.0 Sprayed; G Western wheatgrass playa; sagebrush 45 21 43.9 G;TL Sagebrush; greasewood 67 21 26.7 G;PL;TL Sagebrush made in the field. Rather, vegetation types within the arbitrarily designated study sites were mapped as previously described, and only those types containing burrows were sampled. A minimum 30 plots per vegetation type containing prairie dog burrows was sam- pled at each of these four complexes, with plots randomly distributed throughout sev- eral colonies within each complex. Additional plots were sampled when required to meet sample size adequacy as defined by the Wyo- ming Department of Environmental Quality (1979). A 0.5 m X 2.0 m frame was used to delineate plot boundaries. Percent bare ground, litter and rock, and cover by species were determined by ocular estimate. Shrub density was sampled using a4.0mx4.0m plot located at one corner of the 1 m^ plot. Number of stems and height measurements were recorded for each shrub species rooted in this plot. Floristic inventories were made of all five complexes. Additional qualitative observa- tions, including past and current disturbance, degree and source of disturbance, soils, to- pography, relative abundance and distribu- 88 Great Basin Naturalist Memoirs No. 8 Table 3. Summary cover and shrub density data for all five prairie dog complexes. Cover' Density Prairie dog complex x graminoid xforb x shrub xtpc* x bare X litter X shrubs/m^ Current ferret habitat 1. Meeteetse Complex Core Colony Sagebrush/ Junegrass 39.7 16.2 11.0 66.8 18.7 14.8 1.3 Junegrass 48.2 24.1 1.1 73.4 9.7 17,0 0.3 91 Colony 45.0 9.1 9.1 63.1 29.7 7.6 1.3 Pickett Creek Colony 48.7 23.3 1.8 73.7 10.3 15.7 0.4 Thomas Colony 46.0 13.1 6.0 65.1 31.7 5.4 1.5 Lower Bench Colony 42.8 29.9 1.3 74.1 22.0 6.1 0.6 Tonopah Colony 51.6 12.5 8.6 72.7 21.6 6.0 1.3 Pump Station Colony 44.4 18.3 9.7 72.3 25.8 4.9 1.3 BLM-L3 Colony 55.7 25.7 1,9 83.3 10.2 9.2 0.4 Island Colony 46.2 21.2 3.4 70.7 21.8 9.9 0.4 Graveyard Colony 47.3 23.7 71.0 18.9 11.6 0.1 Junegrass type mean 47.6 20.1 4.3 71.9 20.2 9.3 0.9 Timber Creek Colony 45.5 25.4 2.2 73.1 19.0 10.1 0.4 (abandoned) Historic ferret habitat 2. Larsen's Ranch Complex 14.4 6.7 17.4 38.5 58.3 5.0 2.5 3. Wasmer Flats Complex 26.0 12.1 14.8 52.9 .32.9 16.0 1.3 (live) 4. Gillies Draw Complex 34.3 3.3 20.3 57.8 33.4 9.0 0.7 (dead) 5. Home Flats Complex Grass 45.0 10.9 9.6 65.5 23.6 11.0 1.1 Shrub/Grass 24.1 12.3 22.9 59.4 35.6 5.4 1.7 Saltbush 13.0 17.5 25.9 56.3 39.8 4.3 1.6 Percent aerial cover *Total plant cover tion of prairie dog burrows, and community structure, were also recorded. Results Current BFF Habitat Meeteetse. — This 3887 ha site is located about 10 km northwest of Meeteetse, Wyo- ming. It is characterized by extensive, gently rolling, grassy benches ranging in elevation from 2134 m to 2256 m. Soils are medium textured, contain less than 50% coarse frag- ments, and are either deep (128-154 cm) or underlain by soft or unconsolidated geologic material (Meyer 1983, personal communica- tion) (Table 2). Primary land use is domestic livestock grazing. Eight vegetation types occur in the BFF-oc- cupied study site. They are greasewood (Sar- cohatus venniculatus), wet meadow, rabbit- brush {Chrysothamnus nauseosus), mixed shrub, riparian, mixed grass, junegrass (Koe- leria cristata), and sagebrush/junegrass {Artemisia tridentatal K. cristata). Only tlie- latter two types contain prairie dog burrows. Ten of the 11 colonies he entirely within the junegrass type, whereas one. Core Colony, is vegetated by the junegrass and sagebrush/ junegrass types (Tables 2, 3). Mean total plant cover for the BFF-utilized junegrass type ranges from 63 % to 83 % aerial cover, with grasses the dominant species. The BFF-uti- lized sagebrush/junegrass type has a mean to- tal plant cover of 67%, with sagebrush the dominant shrub (Tables 3, 4). Plant heights range from 13 to 62 cm in these two types. The Timber Creek Colony was probably poisoned and abandoned 10 to 20 years ago (Clark 1985, personal communication). Plant cover and shrub density values for the junegrass type in which this colony lies are all within corre- sponding value ranges of the BFF-occupied prairie dog colonies. Species composition and plant heights are also similar (Tables 2, 3, 4). This ma>' indicate that, unlike blacktail prairie dogs (C ludovicianus) (Bonhain and Lerwick 1976), white-tails do not significantly alter vegetation of their colony. However, it was not the purpose of this study to determine such effects. 1986 Collins, LichvariVegetation Inventory The junegrass-dominated grasslands may he a result of historical heavy grazing by livestock of native bluebunch-western wheatgrass {Agropyron spicatum-A. smithii) communi- ties. Historic photographs of the Meeteetse study area at Buffalo Bill Museum in Cody, Wyoming, indicate historic heavy cattle use of the area during various seasons. Changes in species composition similar to those indicated at Meeteetse have been documented in Ore- gon and Washington (Baker 1983, personal communication). Historic Ferret Habitat The four historic BFF habitat sites are dis- cussed separately. Larsen's Ranch. — This 259 ha study site is located about 18 km north of Wamsutter, Wy- oming (Table 1). Topography is level to gently sloping uplands with 0° to 10° slope. Mean elevation is about 2042 m (Table 2). Soils are predominantly of the Tresano-Sandbranch, nonalkaline subsoil-Sagecreek complex (Hol- brook 1983, personal communication). Two vegetation types were present, bird- foot sagewort/western wheatgrass and sage- brush {Artemisia tridentata). Burrows oc- curred only in the former, which was characterized by 58% bare ground and 39% total plant cover (Tables 3, 4). Dominant spe- cies are those for which the type was named. Plant heights range up to 31 cm. The area appeared highly disturbed by cat- tle at the time of sampling. It was trampled and vegetation was often grazed to within 6 cm of the soil. Not far from where a BFF skull was found are an abandoned barn and corral, water troughs, and salt licks around which cattle congregate. The resulting disturbance, however, did not appear to limit prairie dog use of the area. Wasmer Flats. — This site is located in a north-south trending valley approximately 24 km northwest of Evanston, Wyoming (Table 1). It encompasses about 6323 ha and is char- acterized by gently to moderately sloping to- pography (Table 2). Elevation is about 2095 m. Soils are characteristically very shallow to deep sandy, clayey, gravelly, cobbly, saline- alkaline soils on sandstone, shale, tuff, and conglomerate (United States Department of Agriculture Soil Conservation Service 1981). Primary land use is rangeland. The site is predominantly a complex mosaic of vegetation in which no single species clearly dominates (based on cover). Alkali sagebrush is the shrub with the greatest den- sity (Tables 2, 3) and provides the type with its overall aspect. Total plant cover is 53% (Table 3), with grasses providing nearly half this value. Species providing the most cover in- clude botdebrush squirreltail (Sitanion hys- trix), western wheatgrass, Sandberg blue- grass {Poa secunda), and alkali sagebrush (Table 4). Plant heights range from 5 to 62 cm. Prairie dog burrows are scattered along the valley bottom and on lower slopes where shrubs occur as scattered individuals or in small clumps. They do not occur in areas that appear to be periodically flooded, such as the western wheatgrass playa at the north end of the study site or in taller vegetation, such as the alkali sagebrush stands on upper slopes and the sagebrush {Artemisia tridentata) community at the northern end of the study site. The entire valley has been sprayed, as is evidenced by the high density of dead alkali sagebrush sampled (Table 4). Spraying ap- pears to have been most successful along the valley bottom and lower slopes, where the greatest number of dead shrubs and lowest live shrub densities were observed. The site is also crossed by several dirt roads and is paral- leled by a major highway. These roads allow easy access by humans, who were twice ob- served shooting prairie dogs from the road- side during the sampling. Gillies Draw. — This site is located ap- proximately 40 km southeast of Meeteetse, Wyoming (Table 1). It is an area of highly dissected topography, characterized by a se- ries of nearly parallel ridges separated by steep-sided valleys. Elevation ranges from 1652 to 1808 m (fable 2). The approximately 12,646 ha study site generally includes Gillies Draw itself and adjacent ridges on the east and west. Soils are typically Cadoma, Cushman, and Ulm soil types (Meyer 1983, personal communication). Primary land use is domes- tic cattle and sheep grazing. Vegetation forms a complex pattern. The valley bottom, lower slopes, and isolated level areas support a Gardner saltbush/mixed grass vegetation type. Sagebrush {Artetnisia triden- tata) dominates side-slopes and ridges and 90 Great Basin Naturalist Memoirs Table 4. Dominant plant species' sampled in each of the five prairie dog complexes. No. Meeteetse Complex" Pickett Lower Pump BLM Species Core' 91 Creek Thomas Bench Tonopah Station 13 Island Grasses Agropyron dasystachyum 7,8 12 6 4 5 10 4 12 Agropyron smithii 5 Agropyron spicatiim 6 8 Agropyron trachycaulum Bouteloua gracilis Bromus tectorum Carexfilifolia 8 4 Koeleria cristate 10,21 18 16 13 10 13 11 18 7 Oryzopsis hymenoides Poa secunda 6,11 8 9 6 6 17 18 10 19 Sitanion hystrix Stipa comata 4,0 6 5 15 6 19 5 FORBS Artemisia frigida 6, 12 4 11 3 8 4 6 4 7 Astragalus adsurgens 3 3 Astragalus grayi 2 3 2 Astragalus spatulatus Descurainia pinnata Oxytropus deflexa 0,3 3 3 4 3 Phlox hoodii 4,4 2 4 3 5 2 3 8 5 Ranunculus testiculatus Sphaeralcea coccinea 2 Vicia americana 2 4 Shrubs Artemisia longiloba Artemisia petadifida Artemisia tridentata 9,0 3 Atriplex gardneri 2 2 6 6 Chrysothamnus nauseosus 0,3 1 1 Chrysothamnus viscidijlorus Eurotia lanata 3 2 Gutierrezia sarothrae 0, .3 2 1 1 Opuntia polyacantha 0, .3 1 Dominance based on cover; only dominant species listed. ^Current BFF habitat. ^Historic BFF habitat. ■"Value listed first is for A. tridentata/ K. cristata vegetation type; value listed second is for K. cristata vegetation type. also occurs as stringers along drainages and as small inclusions within the saltbush type. Greasewood (Sarcobatus vermiculatus) iorms dense stands primarily along the main drainage. Prairie dogs utilize only the salt- bush type, where plant heights range from 8 to 92 cm. Total plant cover is 58% (Table 3). Dominant species are Gardner saltbush and a mixture of grasses (Table 4). HORNE Flats. — This site is approximately 7 km south of Medicine Bow, Wyoming (Table 1) at an elevation of 2088 m (Table 2). It is an extensive flat bench whose topography and aspect are very similar to those of the Mee- teetse study site. Soils have not been mapped. The study site encompasses about 14,227 ha, although similar habitat and prairie dog colonies are extremely extensive outside the study site boundaries as well. Primary land use is livestock grazing, and areas near water troughs and salt licks are most heavily im- pacted. Four vegetation types, thickspike wheat- grass-threadleaf sedge {Agropyron dasys- tachyum-Carex filifolia), mixed shrub/mixed grass, Gardner saltbush, and sagebrush (Artemisia tridentata) occur in the study site. Prairie dogs utilize the former three types. Burrows are evenly spaced in the grass type. They are unevenly distributed in the two 1986 Collins, Lichvar:Vegetation Inventory 91 Larsen's Wasmer Gillies Timber Junegrass Ranch Flats Draw Graveyard Creek Average Complex^ Complex^ Compk Home Flats Complex^ Grass- Shrub/ Sedge Grass Salt bush shrub types and appear to be most dense in the Gardner saltbush type. The thickspike wheatgrass-threadleaf sedge type is the most extensive type in the study site. Total plant cover (Table 3) is 66%, most of which is provided by graminoid species. Plant heights range from 26 to 41 cm. Broom snake- weed (Chrysothamnus viscidiflorus) is the predominant shrub. The second most exten- sive vegetation type is the mixed shrub/mixed grass type, characterized by vegetation 31 to 36 cm tall. Birdfoot sagewort, thickspike wheatgrass, blue grama (Bouteloua gracilis), and sagebrush are the predominant species. Total plant cover is 60% (Table 3), with grasses and shrubs contributing nearly equal cover (Table 4). The Gardner saltbush type occurs in small depressions. Total plant cover is 56%, nearly half of which is provided by shrubs (Table 3). Important species include Hood's phlox (Phlox hoodii) and birdfoot sagewort. Plant heights do not exceed 41 cm. Discussion Comparisons of measurements and qualita- tive observations reveal similarities among the five study sites. 1. Vegetation. At all five sites, vegetation types in which prairie dog burrows occur are 92 Great Basin Naturalist Memoirs No. generally characterized by vegetation of low stature (< 92 cm). This figure represents the tallest plants observed, which usually occur as scattered individuals or clumps. Overall as- pect of these vegetation types is one of shorter plants. At the Meeteetse site (current BFF habitat), maximum plant height is 62 cm. The tallest plants are represented by scattered sagebrush or isolated patches of species such as needle and thread (Stipa comata), green needlegrass (S. viridula), Sandberg bluegrass (Poa secunda), and tansymustard (Descur- ainia pinnata). Vegetation of the historic BFF habitat sites is generally shorter than that of the Meeteetse site. The tallest plants ob- served were scattered tansymustard (92 cm) in Gillies Draw. At all sites mat-forming shrubs, such as birdfoot sagewort and Gard- ner saltbush, do not exceed 26 cm. Where inclusions of tall shrubs occur, prairie dog burrows are located only on their peripheries. 2. Burrow Distribution . When a prairie dog colony lies entirely within one extensive vege- tation type, such as 10 of the Meeteetse colonies do, burrows are widely distributed and are about equidistant from one another. When two or more vegetation types occur as a mosaic and only one type is suitable prairie dog habitat, such as at Gillies Draw and Larsen's Ranch, burrows are closer together and distances between them vary. Similar patterns in burrow distribution were ob- served at all five study sites. 3. Topography. All prairie dog colonies sampled occur on level to moderately sloping sites with various exposures. When steep slopes occur, prairie dog burrows are located only in valley bottoms and on lower side slopes, even if a suitable vegetation type fol- lows the steeper slope. This is exemplified best at Gillies Draw, where the Gardner salt- bush type occurs over a variety of slopes but contains burrows only where the topography is level or moderately sloping. At Meeteetse a few small, isolated hills exist that support the junegrass vegetation type and are surrounded by burrows at their bases, but burrows are absent from the slopes. 4. Disturbance . All sites were disturbed by human-related activities. The Meeteetse site has been variously grazed by domestic live- stock for over 100 years. It is crossed by pipe- hnes, transmission lines, fences, and roads. It supports several oil wells and associated struc- tures, corrals, salt licks, and other ranch struc- tures. The four historic BFF habitat sites have been similarly disturbed. As they all occur on public lands, are relatively accessible, and are presently grazed, it is not unreasonable to assume that they, too, have been used as cat- tle and sheep rangeland for many years. All are crossed by one or more of the previously mentioned man-made structures. The Larsen's Ranch site is located in a large, active oil and gas field. In addition, the entire valley in which Wasmer Flats is located was sprayed, presumably to reduce shrub cover and in- crease grass cover several years ago. The site is also easily accessible to humans, who were twice observed by the authors to stop and shoot at prairie dogs from the roadside. In conclusion, it is obvious that white-tail prairie dogs are not dependent upon a partic- ular vegetation type to meet habitat require- ments. They appear to survive equally well in grass, shrub, or shrub/grass types where pre- dominant plant heights do not exceed about 62 cm. They also appear to be extremely toler- ant of high degrees of disturbance for long periods of time. Gonsequently vegetation cannot be considered an important parameter in identifying suitable BFF habitat. Acknowledgments The authors thank the Wyoming Bureau of Land Management and the Wyoming Coop- erative Fishery and Wildlife Research Unit for funding this study. Jack Turnell, John Hogg, Art Thomas, Jack Winninger, Bill Gould, and Rick Weitbrook are thanked for allowing access to the Meeteetse study site. Dennis Knight, Robert Dorn, and Tim Clark are thanked for reviewing this manuscript. Literature Cited Bkcker, C F . AND J D Alvka 1964a. Temperature probabilities in Wyoming. University of Wyoming Agric. Expt. Sta. Bull. 415. 1964b. Precipitation probabilities in Wyoming. University of Wvoming Agric. Expt. Sta. Bull. 416. BoNHAM. D. C AND A Lekwick. 1976. Vegetation changes induced by prairie dogs on shortgrass range. J. Range Manage. 27:221-225. Cl.ARK. T W 1980. A listing of reports of black-footed ferrets in Wvoming (1851-1977). Northwest Sci. 54:47-54. 1986 Collins, Lichvar:Vegetation Inventory 93 Clark, T W , andT. M. Campbell IIL 198L Additional black-footed ferret (Mustela nignpes) reports from Wyoming. Great Basin Nat. 4L360-36L Forrest, S C, T. W Clark, L Richardson, and T. M. Campbell IIL 1985. Black-footed ferret habitat: Some management and reintroduction consider- ations. Wyo. Bur. Land Mgt. Wildl. Tech. Bull. No. 2, Cheyenne. 49 pp. United States Department of Agriculture, Soil Con- servation Service. 1981. General soil map, Uinta County, Wyoming. Wyoming Department of Environmental Quality. 1979. Guideline No. 2. Vegetation. Cheyenne, Wyoming. Martin, S. J., and M. H. Schroeder. 1979. Black-footed ferret surveys on seven coal occurrence areas in southwestern and south central Wyoming, June 8 to September 25, 1978. Final report, Wyoming State Office BLM. 37 pp. 1980. Black-footed ferret surveys on seven coal occurrence areas in Wyoming, February-Septem- ber, 1969. Final report, Wyoming State Office BLM. 39 pp. COMPARISON OF CAPTURE-RECAPTURE AND VISUAL COUNT INDICES OF PRAIRIE DOG DENSITIES IN BLACK-FOOTED FERRET HABITAT Kathleen A. Fagerstone' and Dean E. Biggins" Abstract. — Black-footed ferrets {Mtistela nigripes) are dependent on prairie dogs (Cijnomys spp.) for food and on their burrows for shelter and rearing young. A stable prairie dog population may therefore be the most important factor determining the survival of ferrets. A rapid method of determining prairie dog density would be useful for assessing prairie dog density in colonies currently occupied by ferrets and for selecting prairie dog colonies in other areas for ferret translocation. This study showed that visual counts can provide a rapid density estimate. Visual counts of white-tailed prairie dogs {Cijnomys leucurus)were significantly correlated (r = 0.95) with mark-recapture population density estimates on two study areas near Meeteetse, Wyoming. Suggestions are given for use of visual counts. Recovery of the endangered black-footed ferret will involve the careful management of the only known population near Meeteetse, Wyoming, as well as captive breeding and translocation. Both ferret preservation and population recovery are dependent on the presence of prairie dog colonies. Ferrets have been most frequently observed in or near prairie dog colonies (Cahalane 1954, Hender- son et al. 1969), and their original distribution probably corresponded closely to the range of the black-tailed (Cijnomys ludovicianus) and white-tailed prairie dogs (Hall 1981). The black-footed ferret relies on the prairie dog for approximately 90% of its diet (Henderson et al. 1969, T. M. Campbell personal communi- cation) and on prairie dog burrows for shelter and rearing young. Prairie dog populations declined dramatically during the last century because of loss of habitat and poisoning. From an estimated 283 million ha occupied in the late 1800s (Merriam 1902), prairie dog colonies declined to less than 0.6 million ha by 1971 (Cain et al. 1971). The decline of the black-footed ferret during the last century is probably linked to the reduction in prairie dog populations. A model using growth rates of Siberian pole- cats to simulate those of black-footed ferrets estimated the annual prey requirement of the black-footed ferret to be 214 black-tailed prairie dogs (Stromberg et al. 1983) . They assumed an intrinsic rate of growth of 1.5 for prairie dog populations and calculated the prairie dog population size required to sup- port a ferret at 766. Because white-tailed prairie dogs are larger, their model predicted the annual prey requirement to be 186 ani- mals and the required population size to be 666. In telemetric studies, a radio-tagged black-footed ferret preferred areas of dense prairie dog burrows within its home range (Biggins et al. 1985), and we postulate that high prairie dog densities are important to ferrets. A rapid method of determining prairie dog population density needs to be developed that can be used to assess the prairie dog popula- tions at Meeteetse and that would allow us to monitor prairie dog populations at frequent intervals for potential problems, such as plague outbreaks or effects of oil develop- ment. A rapid density estimation procedure also could be used to assess prairie dog popu- lations in colonies being considered for ferret translocation. Prairie dog population numbers have been estimated using a variety of methods. Mark- recapture is a reliable method for estimating the density of prairie dogs because these ani- mals have relatively small home ranges and are readily trapped. However, mark-recap- ture is labor intensive and can be done only on relatively few plots; it is therefore impractical for estimating animal density over large areas. Closing burrows and counting the number 'U.S. Fish and Wildlife Service, Denver Wildlife Research Center, Bldg. 16, Denver Federal Center, P.O. Box 2.5266, Dei ^U.S. Fish and Wildlife Service, Denver Wildlife Research Center. P.O. Box 916, Sheridan, Wyoming 82801. Reference to trade names does not imply endorsement by the federal government. 94 1986 Fagerstone, Biggins: Density 95 reopened after 1 or 2 days is a method fre- quently used in conjunction with control pro- grams, where pretreatment and posttreat- ment counts are compared to determine the effectiveness of rodenticide applications to prairie dog populations (Tietjen 1976). The method provides an index to prairie dog activ- ity that may have little correlation with actual population trends (Knowles 1982); results can be variable with this technique because one prairie dog can reopen more than one burrow. Visual counts may provide a quick method of measuring prairie dog density; prairie dogs are well suited for visual counts because of their large size, their diurnal activity patterns, and their tendency to live in social colonies. Visual counts were used by Knowles (1982) to estimate black-tailed prairie dog numbers, but their precision was not assessed for white- tailed prairie dogs. This study evaluated the use of visual counts to monitor white-tailed prairie dog densities by comparing visual counts with mark-recapture data. Study Area The study was conducted 30 km southwest of Meeteetse, in Park County, Wyoming. White-tailed prairie dogs occur in colonies on about 3000 ha (Clark et al. 1984) throughout this area. We studied two colonies located between 2280 and 2380 m in elevation on short- to midgrass rangeland. Methods Mark-Recapture Prairie dog populations were censused by mark-recapture during May and July 1984 and May 1985. A 360 x 360 m trapping grid was established on each of the two study colonies using 169 National^ live traps (48 x 15 x 15 cm) located at 30 m intervals. The grid was subdi- vided into nine 120 x 120 m subplots. Before each trapping period, the traps were wired open and baited with flaked oats for a two-day familiarization period. During the subse- quent five-day trapping period, the traps were baited with oats and checked during the morning; they were closed at midday to avoid prairie dog mortality caused by heat stress. The trapped prairie dogs were aged (juvenile or adult), sexed, ear-tagged with monel No. 1 fingerling fish tags, and released at the point of capture. Population estimates for each of the trapping periods were computed using the computer program CAPTURE (White et al. 1978). Otis et al. (1978) have provided a detailed reference on the theory behind pro- gram CAPTURE. Visual Counts Prairie dogs on the study area were ob- served prior to the initiation of this study. They exhibited a bimodal activity pattern with peak numbers aboveground between 0700 and 1000 hours and with a second but lower peak between 1500 and 1800 hours. This bi- modal activity pattern is similar to that ob- served by Tileston and Lechleitner (1966) and Clark (1977) for white-tailed prairie dogs and by Althen (1975) for black-tailed prairie dogs. Visual counts were therefore conducted dur- ing the peak activity period in the morning on four consecutive days following the trapping period. During May 1984 prairie dogs were counted from portable 3-m-high towers erected in the center of each 120 x 120 m subplot. Counts from the center of each sub- plot proved labor intensive, so during July 1984 and May 1985 prairie dogs were counted from two locations outside the entire 360 x 360 m grid; observers were located a minimum of 30 m from the grid to minimize disturbance of animals. Two observers counted the grid fi-om each location. Prairie dogs on each 120 x 120 m plot were counted during a four-minute period by scanning the plot with binoculars and a 15X spotting scope. Three counts were made daily of each plot during a two or three- hour period. Plots were counted in the same sequence and at synchronized times by ob- servers at both locations. Prairie dogs that were located on the borders between two plots were counted if they were on the north and east edges and not counted if on the south and west edges. Statistics Simple linear correlation coefficients were computed (1) between the highest total count of individual prairie dogs over the entire 360 x 360 m grid and the population density gener- ated by program CAPTURE for the corre- sponding five-day period and (2) between the highest single count of individual prairie dogs 96 Great Basin Naturalist Memoirs No. 8 != 60 S 20 25 30 35 40 45 50 MAXIMUM VISUAL COUNTS Fig. 1. Prairie dog population estimates on 360 by 360 m grids (x-axis) plotted against maximum visual counts on the same areas (y-axis). The simple linear regression equation is: y = 15.56 + 0.28x. per 120 X 120 m plot and the number of prairie dogs trapped on that plot during the corre- sponding five-day trapping period (insuffi- cient numbers of prairie dogs were trapped on each 120 x 120 m plot to generate a satisfactory population density). The variation associated with location, ob- server, day, and trial (three counts per day) was determined using a procedure on SAS (SAS 1985) that estimates variance compo- nents (PROG VARGOMP). Results There was a high correlation between the population densities estimated by GAPTURE and the highest number of animals counted visually across the entire 360 x 360 m grid during the corresponding period (r = 0.95, P = 0.004, Fig. 1). The simple linear regression equation is: y = 15.56 + 0.28x, where y is the maximum visual count and x is the population density. Population density correlated better with visual counts than the total number of animals trapped (r = 0.84). Also, the maxi- mum number counted provided a better cor- relation than the average of a series of counts (r=0.74). There was a lower correlation between the highest count and number trapped per 120 x 120 m sub-plot (r = 0.69); when analyzed separately by time period the correlation was highest during May 1984 (r = 0.86) and lower during July 1984 and May 1985 (r = 0.70 and 0.61, respectively). Visual counts on small ar- eas may therefore not be as representative of actual densities as counts on larger areas. Variance component estimation revealed that trials (counts per day) accounted for the most variation in the data (Table 1). This was expected because counts were begun in the morning as prairie dogs emerged from bur- rows and were continued until prairie dogs became less active above ground in midmorn- ing. During any day, counts were normally low at first, increased to the maximum count, then decreased. Location accounted for a large portion of the variation in the data on area 1 but only a small portion of the variation on area 2. Location was important on study area 1 because tall grass grew on a portion of the study area between the time the area was chosen and the time when visual counts were begun. The grass made counting prairie dogs on part of the plot difficult from one of the two locations. When trials were removed from the analysis and only the maximum count by each ob- server per day was used, location still ac- counted for a large portion of the variation in the data on area 1. Day variation was small for area 1 (only one-third of location variation) but was comparatively large for area 2. Observer variation was negligible, but a large variance component existed for observer-day interac- tion. This would indicate that variability was present between observers over the four-day period but that observers had no consistent bias toward high or low counts. Discussion Visual counts appear to provide a useful in- dex to prairie dog population densities that can be used to monitor prairie dog popula- tions at Meeteetse and to assess ferret reloca- tion sites. Mark-recapture is a reliable 1986 Facerstone, Biggins. Density 97 Table 1. Components of variance for prairie dog visual counts of two study areas near Meeteetse, Wyoming. On each study area, counts were conducted over a four-day period from two locations by two obser\'ers at each location. The magnitude of the variance indicates the relative influence of each item in the model to the overall variation. Area l' Area 2' Source df Area 1 Area 2 maximum count maximum count Location (L) 1 11.34 -1.54- 20.77 -2.64 Observer (0) in L 2 2.73 -0.44 -0.54 -3.17 Dav (D) 3 -1.49 3.60 6.70 13.92 DxL 3 21.75 4.81 -15.94 -3.58 O .X D in L 6 -4.38 -2.41 44.62 14.67 Trial in O x D 32 33.96 23.77 Trials were removed from thi.s analysis; the maximum count per observer per clay was analyzed. ^Negative variances are usually considered to be zero. method for estimating the abundance of white-tailed prairie dogs because these ani- mals are readily trapped and have relatively small home ranges. The estimates of density using mark-recapture and visual counts were independent of one another in this study and produced comparable results. Therefore, un- less both methods were equally biased, the accuracy of visual counts was probably good. Knowles (1982) found that visual counts of black-tailed prairie dogs provided a good cor- relation between maximum counts and popu- lation levels (r = 0.942). Size of areas counted influences the reliabil- ity of visual counts. Visual counts on small areas of 1 to 1.5 ha did not correlate as well with numbers of animals trapped on those areas because of the increased edge effect on the smaller areas or because prairie dog home ranges exceeded the area counted and move- ment occurred between areas of locally high and low densities. Size of areas counted should therefore be 10 ha or greater if possi- ble. Because maximum counts provided better correlations with prairie dog population den- sities than the mean of a series of counts, counts should be made during peak activity periods in early morning or late afternoon. Trials (counts per day) accounted for the most variation in the data because counts were made from the time prairie dogs emerged early in the morning until they disappeared as temperatures increased. Because of this varia- tion, it is important that visual counts be made over a two- or three-hour period during peak prairie dog activity so a count of maximum numbers above ground is included. Because prairie dog activity is suppressed by high and low temperatures (e.g., below 10 C or above 27 C) and strong winds (Davis 1966, Althen 1974), counts should be conducted when weather conditions are moderate. Seasonal variation also exists in the percent- age of prairie dogs active above ground (Tile- ston and Lechleitner 1966, Clark 1977). Indi- vidual prairie dogs are active for only four to five months during the year. Adults emerge from hibernation in early spring and become inactive in midsummer, whereas juveniles ap- pear above ground in early June and remain active until late October or November. The total population therefore attains its maxi- mum size during late June and early July when adults and juveniles are above ground at the same time. Surveys during this period would therefore most accurately reflect total population levels. Substantial variation occurred among loca- tions when observers at one of the locations on area 1 were not able to see portions of the plot well because of intervening tall grass. Loca- tions of observers for conducting counts should therefore be chosen carefully so the entire area to be counted is readily visible. In areas of flat topography the roof of a vehicle provided a good location from which to make visual counts. The variance component for days was small for area 1, indicating that counts made on one day might be sufficient. However, the vari- ance component was larger for area 2. A zero variance component existed for observers in this study, indicating that different individu- als could count on different days and an accu- rate estimate of population density could still be obtained. However, because a large com- ponent was associated with observer-day in- 98 Great Basin Naturalist Memoirs No. 8 teraction and because a large day component was present on one area, we recommend that counts be made over several days by the same observer. Although visual counts can be a precise method of estimating prairie dog populations, they should be used with caution. Precision is based upon their repeatability. Therefore, the observer, location, and time of day should remain constant between one count and the next whenever possible. The area to be counted should be predetermined and its boundaries well marked so that prairie dogs outside the area will not be counted. System- atic scans of an area for predetermined time periods can minimize the possibility of count- ing animals more than once; the only animals counted twice are those that move across the study area during the scan. If conducted fol- lowing the guidelines suggested, visual counts can be a valuable technique for esti- mating prairie dog densities. Acknowledgments We greatly appreciate the statistical advice of R. Engeman (Denver Wildlife Center statistician). G. H. Matschke and P. L. Heg- dal kindly reviewed the manuscript. Literature Cited Althen, C L 1975. The interaction of circadian rhythm and thermal stress in controUing activity in the black-tailed prairie dog. Unpublished disserta- tion, University of Colorado, Boulder. 168 pp. Biggins, D. E., M Schroeder, S Forrest, and L Richardson 1985. Movements and habitat rela- tionships of radio-tagged black-footed ferrets. Pages 11.1-11.17 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Cahalane, V H 1954. Status of the black-footed ferret. J. Mammal. 35:418-424. CainS A ,J A Kadlec. D L Allen.R A Cooley, M.G. HoRNOCKER, A S. Leopold, and F. H. Wagner. 1971. Predator control— 1971. Report to the Council on Environmental Quality and the De- partment of the Interior by the Advisory Commit- tee on Predator Control. Inst, for Environ. Qual- ity, University of Michigan, Ann Arbor. 207 pp. Clark, T W 1977. Ecology and ethology of the white- tailed prairie dog (Ci/no/rii/.s/ewcurus). Publ. Biol- ogy and Geology No. 3, Milwaukee Public Mu- seum. 97 pp. Clark, T W , L Richardson, D Casey, T M Campbell, III, andS C Forrest 1984. Seasonality of black- footed ferret diggings and prairie dog burrow plugging. J. Wildl. Manage. 48:1441-1444. Davis, A H 1966. Winter activity of the black-tailed prairie dog in north-central Colorado. Unpub- lished thesis, Colorado State University, Fort Collins, 45 pp. Hall, E R 1981. The mammals of North America, 2d ed. Ronald Press, New York. 1181 pp. Henderson, F. R., P F. Springer, and R. Adrian. 1969. The black-footed ferret in South Dakota. South Dakota Dept, Game, Fish and Parks Bull. 37 pp. Knowles, C J 1982. Habitat affinity, populations, and control of black-tailed prairie dogs on the Charles M. Russell National Wildlife Refuge. Unpub- lished dissertation. University of Montana, Mis- soula. 171 pp. Merriam, C H 1902. The prairie dog of the Great Plains. Pages 257-270 in L. C. Everard, ed.. Yearbook U.S. Dept. Agric, Washington, D.C, Otis, D L , K P Burnham, G C White, and D R An- derson 1978, Statistical inference from capture data on closed animal populations. Wildl. Monogr. 62:1-135, SAS. 1985. SAS Users guide: statistics, version 5 edition. SAS Institute Inc., Cary, North Carolina. 956 pp. Stromberg, M R , R L Rayburn, andT W. Clark. 1983. Black-footed ferret prey requirements: an energy balance estimate, J. Wildl. Manage. 47:67-73, TiETjEN, H P 1976, Zinc phosphide — its development as a control agent for black-tailed prairie dogs. USDl, Fish and Wildl, Ser, Spec. Sci. Rept. Wildl. No. 195. 14 pp. TiLESTON , J V , AND R R Lechleitner 1966, Some com- parisons of the black-tailed and white-tailed prairie dogs in north-central Colorado. Amer. Midi. Nat, 75:292-316, White, G C , K P Burnham, D. L. Otis, and D. R. An- derson, 1978, User's manual for program CAP- TURE, Utah State University Press, Logan, Utah. 40 pp. HABITAT SUITABILITY' INDEX MODEL FOR THE BLACK-FOOTED FERRET A METHOD TO LOCATE TRANSPLANT SITES B. R. Houston', Tim W. Clark', and S. C Minta" Abstract.— A Habitat Suitability Index Model (HSI), following the U.S. Fish and Wildlife Service HSI Model Series, is described for the black-footed ferret. The literature on which the model is based is reviewed, and model assumptions and structure are discussed. A realistic model is specified with variables and their functions that embody the critical spatial and resource heterogeneity characteristic of the broad geographic environment ferrets occupy. It assumes that ferrets can meet year-round habitat recjuirements within prairie dog colonies providing: (1) prairie dog colonies are large enough, (2) burrows are numerous enough, and (3) adequate numbers of prairie dogs and alternate prey are available. Five habitat variables are identified: VI is the frequency distribution of colony sizes, V2 is the total area of colonies, V3 is burrow opening density, V4 is intercolony distance, and V5 is prairie dog density. Variables are compensatory. As more data become available and our understanding of ferrets expands, the basic model design can readily incorporate improvements without radical restructuring. Habitat models are an attempt to describe and quantify an animal's essential habitat require- ments or "life requisites" and are therefore a useful tool in habitat evaluation. The Habitat Suitability Index (HSI) Model Series, developed by the U.S. Fish and Wildlife Service (USFWS), provides habitat descriptions for several species. These models are useful for assessment of im- pacts on wildlife and habitat management (US- FWS 1980a, b) and may prove especially valu- able in endangered species management, where determination of habitat quality and suitability is often critical for management and continuation of the species. HSI "models should be viewed as hypotheses of species-habitat relationships ra- ther than statements of proven cause and effect relationships" (Schamberger et al. 1982:1). This paper applies the HSI Model format to the Meeteetse, Wyoming environment of the black-footed ferret (Mustela nigripes ; BFF) as generally described by Clark et al. (Descrip- tion and history A^^Q) and more specifically by Forrest et al. (1985)(Fig. 1). Applications and uses of the model are: (1) to compare other areas to BFF habitat at Meeteetse, (2) to use those comparisons to select areas to be searched for BFFs, and (3) to select suitable areas for transplant sites. Our use of the HSI format closely follows the USFWS (1981) and parallels applications by Allen (1982a, b, 1983, 1984) for other species. Our use of the HSI model for BFFs incorpo- rates several recent improvements on the roles of ecological models: (1) We stress model reality of a single species more than focus upon model precision or generality (see Levins 1966, Rosen 1978, Kaiser 1979, Pielou 1981). (2) Few highly measurable variables dictate the HSI, and, although some are colin- ear, together they contain high explanatory power, at the same time allowing comprehen- sible results and simplified sensitivity analy- sis. This reflects the growing consensus that there is no apparent relation between model complexity and predictive utility in any field of forecasting (e.g., Ascher 1978, K. E. F. Watt personal communication). (3) Our model uses nonlinear representations of variables, rather than linear, because those more accu- rately express the dynamic nature of biologi- cal responses and realistic species-habitat re- lations (Whittaker 1975, Green 1979, West- man 1980, Johnson 1981, Meents et al. 1983) . Nonlinearity permits us to mimic more realis- tic biological processes that involve thresh- olds and limits and the smoothed transitions between them (HoUing 1985, J. R. Krebs per- sonal communication). (4) The model vari- ables and their functions embody the critical 'Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209. ^Department of Wildlife and Fisheries Biology, University of California, Davis, California 95616. 99 100 Great Basin Naturalist Memoirs No. 8 [^fim ,/4., 5aa#* ^^ ' ' *'^" Fig. 1. Ph()t()tirai)lis ()ll(laik-fboted ferret habitat (prairie dog colonies and prairie dogs) aiidtt rift predation. Photos by Tim Clark. A. White-tailed prairie dog colony occupied by ferrets. B. Black-footed ferret at prairie dog burrow. 1986 TTOUSION KTAL.: HaBITAT SUITABILITY 101 CJ. Two wliite-tailed prairie clogs. *sii?it*" ■^R«. **< 5^ ry D D. Black-footed ferret with prairie dog prey. 102 Great Basin Naturalist Memoirs No. 8 importance of spatial and resource heterogene- ity. The structural simplicity of the BFF-prairie dog {Cynomijs spp.) community promotes a de- sign where all variables directly assess spatial patchiness and resource variability, consider- ations that have pivotal impact on population dynamics and population viability (reviews in Steele 1974, Wiens 1974, Southwood 1977, Shugart 1981). The outcome of the above four features is only a slight increase in model complexity traded for a dramatic increase in ecological reality. Perhaps of equal benefit is the ease of model validation. As more data become available and our under- standing of BFFs expands, the basic model de- sign can readily incorporate improvements with- out radical restructuring. Data sets already completed and cited below could likely be reevaluated with future model versions. This HSI application for the BFF draws on Clark et al. {Description and history, 1986) and Forrest et al. (1985), who describe the Mee- teetse, Wyoming, BFF study area (1981-1985) and its use by BFFs as well as all the data from the Mellette County, South Dakota, BFF study (1964-1974). Because of the localized nature and hmited size of these two study areas, this HSI model will likely require updating if BFFs are found in other areas in different ecological set- tings. In the meantime, this HSI model can serve as a useful tool in BFF recovery planning to evaluate proposed transplant/relocation sites. Background Requests for evaluation of BFF habitat have been frequently mentioned in the literature. The Black-footed Ferret Recovery Team (1978) requested research to define compo- nents of a prairie dog colony necessary to sup- port BFFs. The BFF Recovery Plan also notes the need to establish ideal habitat sites for successful introduction of transplanted BFFs (see Linder et al. 1972) . The South Dakota BFF and Prairie Dog Workshop in 1973 sug- gested several BFF management needs, in- cluding a definition of habitat (Hillman and Linder 1973, Stuart 1973, Erickson 1973). Others have discussed the need for BFF pre- serves and habitat descriptions (Clark 1976, 1984, 1986). Flath and Clark (1986) described historic prairie dog distributions in Montana for the period 1908-1914. This early Montana situation probably represented a habitat setting in which BFFs evolved among the complex in- terrelationships of species and environmental interactions of the prairie dog ecosystem. Hillman et al. (1979) described prairie dog dis- tribution in the area occupied by BFFs in South Dakota. Their description was widely used by management agencies as a guide to the number and spacing of prairie dog colonies to be left after prairie dog eradication programs. Clark et al. {Description and history, 1986) provided a descriptive and historical overview of the Meeteetse BFF environment. Forrest et al. (1985) noted that BFFs are restricted to a prairie dog complex — a group of prairie dog colonies distributed so that individual BFFs can migrate among them commonly and frequently. The 37 colonies of the Meeteetse complex (total size 2995 ha) were described and their occupation history by BFFs noted. The average density of adult BFFs was 1 BFF/56.6 ha. Burrow open- ings, based on literature reviews, are correlated with the number of prairie dogs present (r = 0.71). High burrow densities are desirable for BFFs in that they provide added protection from predators and shelter from the elements. Colonies greater than 100 ha supported more than two resident adult BFFs, whereas colonies from 12.5 ha to 102.0 ha supported only one BFF throughout the year. BFFs traveled among the colonies, but to an unknown extent. BFFs may use burrows at low densities and colonies of small size in travels between larger colonies. BFFs moving between colonies have a greater chance of finding another colony if the colonies are large and close together. Several bibliographies of BFFs (Harvey 1970, Snow 1972, Hillman and Clark 1980, Casey et al. 1986) and of prairie dogs (Clark 1971, in prepara- tion, Hassien 1973) exist. These also serve as background for this HSI model. General infor- mation on BFFs is sunnnarized in the bibliogra- phies listed above, in primary sources from South Dakota studies (e.g., Hillman 1968, Hen- derson et al. 1969, Fortenbery 1972), and, more recently, from Meeteetse, Wyoming (e.g., Clark et al.. Description and history, 1986; Clark et al.. Descriptive ethology, 1986; Camp- bell et al. 1985, Richardson et al. 1985; Forrest etal. 1985, Biggins et al. 1985). 1986 Houston etal.: Habitat Suitability 103 Habitat Use Information Overview A member of the family Mustelidae, the BFF is the only ferret native to North America (Hall 1981) and is perhaps the rarest and most endangered mammal species on this conti- nent (Cahalane 1954, Hillman and Clark 1980). BFFs are solitary except during breed- ing and maternal care of young and are pri- marily nocturnal. They prey on prairie dogs, whose burrows they also use for cover and litter rearing. Food The BFF relies on prairie dogs as its primary food source, although other prey, both live and dead, are taken in considerably lesser amounts (Hillman 1968, Henderson et al. 1969, Sheets and Under 1969, Sheets et al. 1972, Clark et al. 1985). Sheets et al. (1972) found 91% of 82 BFF scats from South Dakota contained prairie dog remains, and Campbell et al. (unpublished data) found 87% of 86 BFF scats from Meeteetse contained prairie dog remains. Prairie dogs, on this basis, compose the major BFF food. Stromberg et al. (1983) generated a preda- tor-prey model of metabolizable energy re- quirements that estimated: (1) annual prey requirements for one reproductive female BFF and her litter of four and (2) prairie dog population sizes needed per BFF. Powell et al. (in press) estimated BFF winter energy expenditure (about 104 kcal/day) and prey re- quirements (about 20 prairie dogs from De- cember through March) at Meeteetse. A lac- tating female with four young are predicted to need six times the winter estimate, or about one prairie dog per day in summer. Water BFFs apparently satisfy water requirements through prey consumption and have never been observed in the wild drinking free wa- ter. Henderson et al. (1969) reported that captive BFFs drank water irregularly. L. Richardson (unpublished data) watched a BFF eating snow at Meeteetse. Cover Cover for BFFs is provided by prairie dog burrows, which are used for predator avoid- ance and thermal cover throughout the year (Clark et al. 1985, Richardson et al. in press) Any prairie dog burrow is assumed to be suffi- cient to satisfy BFF cover requirements. Higher biurow densities provide greater cover. Reproduction Reproductive habitat re(|uirements for BFFs are assumed to be identical to food and cover requirements described above because all BFF activities are associated with prairie dog burrow systems (Clark et al. Descriptive ethology, 1986; Richardson et al. in press; Forrest et al. 1985) . Large, mounded, multi- entranced burrows may be important for litter rearing because of their presumed extensive tunnel network. Interspersion A picture of BFF home range patterns is emerging from research efforts at Meeteetse. A single adult male's range may encompass home ranges of several females, which show much smaller ranges (Richardson et al. un- published data). Females remain with their litters until late summer, when young become independent (Henderson et al. 1969, Clark et al. Descriptive ethology, 1986). BFFs appear to have a typical mustelid spacing pattern de- scribed by Powell (1979), Forrest et al. (1985), and Richardson et al. (in press). More infor- mation is needed on BFF home ranges and movements, dispersal of young or adults, and inter- and intrasexual interactions. Interspersion characteristics of BFFs repre- sent a two-dimensional management consid- eration — individual and populational. Indi- vidual interspersion patterns are better known than populational interspersion pat- terns required for minimum population sizes. A resident female snow-tracked from Decem- ber through March used 16.0 ha and was over- lapped by a resident male that used 136.6 ha (Forrest et al. 1985). Studies of radio-collared BFFs show a young female used 12.6 ha in October and November (Biggins et al. 1985). Population interspersion is dependent on the size, configuration, and intercolony distance of prairie dog colonies making up the com- plex. Data show that, if colonies are too small and intercolony distances are too large, then BFF populations cannot sustain themselves. 104 Great Basin Naturalist Memoirs No. 8 The search for food (energetics) becomes pro- hibitive, avoidance of predators becomes dif- ficult or impossible, and adequate thermal cover is rare or nonexistent, all reducing both individual and population survival. Special Considerations Successful management of BFFs depends on maintaining adequate numbers and areas of prairie dog colonies. Minimum viable pop- ulation (MVP) sizes and area requirements for BFFs were addressed by Groves and Clark (1986). Additional estimates of these variables are undei-way by Shaffer et al. (in prepara- tion), who are modeling effects of both demo- graphic and environmental stochasticty on BFF populations of varying sizes. The MVP represents a threshold below which popula- tions are not self-sustaining. Populations may persist for a long time below the MVP, but probably at a loss of adaptability and a high susceptibility to local extinction. Groves and Clark (1986) noted that the genetic method of determining MVP for the Meeteetse BFFs estimated that about 200 animals are needed for maintenance of short-term fitness. The es- timated 200 animals needed is about four times the number of breeding adults esti- mated to currently exist at Meeteetse (Clark 1986). Poisoning and shooting of prairie dogs should be prohibited from areas where BFFs occur as well as from other selected portions of prairie dog range. Hubbard and Schmitt (1984) suggested a "refugia" concept of man- aging prairie dogs in which relatively large areas are omitted from poisoning and other disturbance. They suggested that refugia be large enough to support a BFF MVP and based such area estimates on the Stromberg et al. (1983) predator-prey model. Clark (1986) outlined a series of management guidelines for BFFs. Differences in black-tailed (C. Itidovi- cianus) and white-tailed prairie dog colonies have been noted (Tileston and Lechleitner 1966, Campbell and Clark 1981, Clark et al. 1982). Black-tailed colonies often show great- er prairie dog and burrow opening densities — two important variables of BFF habitat. Satis- fying habitat recjuirements for BFFs on white-tailed colonies as described in our HSI model is assumed also to satisfy hai)itat re- quirements on black-tailed and Gunnison's (C. gunnisoni) prairie dog colonies. Application of Habitat Suitabilit\' Model Model Apphcability Geographic area. — Although this model was developed on data from the only two BFF populations ever studied, it should apply throughout the historic range of the BFF until additional BFF populations in different eco- logical settings are found, studied, and results show it does not apply. Even though a single prairie dog colony cannot support a BFF MVP (unless it is extremely large), it can potentially support one or more individuals. Therefore, any prairie dog colony should be considered potential BFF habitat. Historic and current land use patterns affect the quality of BFF habitat. A constellation of prairie dog colonies, described by Clark et al. (Descrip- tion and history, 1986) and Forrest et al. (1985) as a prairie dog "complex, ' is needed to support a BFF MVP. S.50 33 183.0 257 34 196.5 617 200 35 • 211.0 658 2100 36 230.0 671 2200 37 1,307.0 2242 2300 J Total area 2990 5496 6800 Total colonies 37 29 4 All colonies <1 ha are entered a.s 1 ha. Since colony sizes are often unique numbers or are entered that way as data, then Nj is com- pletely eliminated from the calculations (see example). However, if each colony area is not estimated for some reason, then they can be grouped into intervals such as 0-5 ha, 5-10 ha, etc., in which case the midpoint can be used (i.e., 2.5, 7.5). Structure of V2-V5 Variables 2, 3, 4, and 5 are intrinsically non- linear and are each derived from the differen- tial equation dy/dt = ay" + by + c and simpli- fied to the logistic form of Y = (1 + ke^7^ The logistic form is particularly suitable in describing these variables because it depicts two asymptotic limits (at and 1 adjustable toward infinity) and contains an inflection point around which the most rapid rate changes occur. For example, intercolony dis- tance (V4) reflects the ability of BFFs to inter- cept life requisites upon leaving one colony for another. If a straight-line 10 km is as much as BFFs might move in a night, then that value is the inflection point around which crit- ical and therefore extreme shifts in the suit- ability index (SI) occur. Of course, BFFs eas- ily move from to 5 km and SI values change little within that range. Likewise, once a BFF is well past the "point-of-no-return," say 15-20 km, SI value shifts are also small. An- other view is that the chance of intercepting another colony along a radius extending from its home colony is a quadratic function of dis- tance moved modified by the actual mobility and energetic characteristics of the BFF. V2:f(x) = (l + 20e"'^"T' V3: f(x) = (1 + 15 e" *T' V4:f(x)= 1-(1 +70e"''7 V5:f(x)-(1 + 200e 'T' X ha x burrows/ha xkm X prairie dogs/ha Examples of HSI Calculations Table 3 contains colony sizes for the Mee- teetse complex (Area I), an actual prairie dog complex elsewhere (Area II), and a hypotheti- cal area (Area III). Maps of these three areas follow (Fig. 3). Before computing VI with this data, it is important to understand that al- though Area I and Area II have different dis- tributions of absolute colony size, the distri- butions are quite similar in colony size relative to their total colony area. It is V2 that will accoimt for the almost double total area of Area II. VI: First, calculate S N. (i' - i). i = 2 1986 Houston etal.: Habitat Suitability Fig. 3. Maps of prairie dogs complexes used in examples of calculations of HSI. Area I = Meeteetse, Wyoming, Area II = another actual complex. Area III = hypothetical complex. Since each colony has a different area, N drops Note that we do not include areas of size 1; 1 out and we add the following series for Table 3 —1 = 0. Second, calculate which is really only the total colony area ma- nipulated as in the first calculation: showing colony sizes for each area: Area I: (1.5^-1.5) + (2^ - 2) + ...+ (230--230) + (1307'- 1307) =1,936,862 Area II; (3^-3) + (4^-4) + . . . + (671--671) + (2242--2242) = 6,473,430 Area III: (200--200) + (2100--2100) + (2200--2200) (2300- -2300) = 14,573,200 (J NiXi V- SNi 114 Great Basin Naturalist Memoirs No. 8 Area I: 2990^ - 2990 = 8,937, 110 Area II: 5496^ - 5496 = 30,200,520 Area III: 6800' - 6800 = 46,233,200 We arrive at VI by dividing the first by the second calculations for each area: VI = P(AB) = .217 for Area I, .214for Areall, and .315for Area III. Notice the influence of Area I and Us largest colony on the outcome of the first cal- culation. Area I: 1307' - 1307 = 1,706,942 and for Area II: 2242' - 2242 = 5,024,322. If we were to split Area Is 1307 ha colony into two separate colonies, it would decrease the value of VI to .131. How far apart these colonies would be is accounted for by a simultaneous increase or decrease of V4. For instance, note how the small 200 ha colony in Area III is a stepping stone between the three larger colonies. Its critical position is reflected by a lower mean intercolony distance and there- fore a higher value for V4. Graphs of the variable equations and the above values follow. V2 Total area of colonies in = 0.958 0.8 /n 0.876 0.6 / 0.4 #1 = 424 0.2 n n 2500 5000 7500 10000 TOTAL AREA OF COLONIES (HA) V3 Burrow opening density (mean number of burrow openings/ ha of colony) ■ m = 99|,^--« 0.8 /^ 0.6 mi-- 0.67] 0.4 #11 = 0.385 0.2 X 25 50 75 100 NUMBER OF OPENINGS/HA Completing the other values we obtain: Variable Area I Area Area III VI V2 (ha) V3 (burrows/ha) V4 (km) HSI = V5 (dogs/ha) .217 .424(2990) .671 (57) .980 (.92) .590 .214(5) .214 .876(5496) .385(37.3) .980(1) .613 .155(4.5) HSI for: Area I = (2 x .217 x .424 x .671 x Area II = (2 x .214 x .876 X .385 x Area 111= (2 X .315 x .958 x .992 x Variable VI Distribution of colony .315 .9,58(6800) .992(125) .969(2) .873 .987 (12) = .590 = .613 = .873 P(AB|N„ i) = , 2 ^ fr) Using the above equation for VI, area I = .217, Area II .214, and Area III- .315. V4 Intercolony distance (mean distance between colonies) I and H" 0,980 5 10 15 20 INTERCOLONY DISTANCE (KM) V5 Prairie dog density (mean number of prairie dogs/ha of colony) PRAIRIE DOGS /HA DESCRIPTIVE ETHOLOGY AND ACTIVITY PAITERNS OF BLACK-FOOTED FERRETS Tim W. Clark', Louise Richardson', Steven C. Forrest', Denise E. Casey', and Thomas M. Campbell III' Abstract. — Aspects of the aboveground ethology and activity patterns of" the black-footed ferret {Mustela nigripes) are described for a population in northwestern Wyoming as a first step in building a descriptive ethogram and quantification of activity patterns. We observed at least 237 individual ferrets for 208 hr on 441 occasions from 2 December 1981 through 25 September 1984. Maintenance behaviors (locomotion, alert, grooming and sunning, defecation and urination, digging, and predation) and social behavior (reproduction, ontogeny, maternal, play, agonistic) are described as well as some ferret-human interactions. Ferret vocalizations are subjectively described. We located ferrets during most months, including winter, but found that they were easiest to locate in summer. Ferrets were active at -38 C, in snow, in rain, and in winds to 50 kph. The black-footed ferret (BFF) is one of the least well known of all the endangered mam- mals in the United States despite 11 years (1964-1974) of intensive and extensive re- search in South Dakota (Erickson 1973, Hill- man and Linder 1973). Data are lacking on many aspects of BFF behavior and activity patterns. It is essential that the general behav- ior patterns of any animal first be qualitatively described in an "ethogram" to provide the basis for more specific, quantitative behav- ioral studies (Scott 1956, Klopfer and Hailman 1967, Lehner 1979). This paper provides an initial description toward a BFF ethogram and gives results of nocturnal observations of surface activity for the Meeteetse, Wyoming, BFFs. Behavioral descriptions are "func- tional" (Candland 1974) and definitions are operational (Sustare 1975). Methods Behavioral descriptions are based on 208 hr of direct observation of at least 237 individual BFFs on 441 occasions between 2 December 1981 and 25 September 1984. We observed maternal, play, and predatory behavior at 10 m or less, sometimes for over 1 hr per obser- vation. Daytime observations were generally made with the unaided eye, but a spotting scope and binoculars were sometimes used. Nighttime observations were made with the aid of hand-held or truck roof-mounted spot- lights following methods outlined by Clark et al. {Handbook of methods, 1984). The time and duration of each observation, description of behavior, and weather conditions were recorded, and photographs were taken when possible. Because BFFs are nocturnal, secre- tive, solitary, and active above ground briefly and irregularly and because they inhabit an environment of grass and shrubs, it is very difficult to observe and collect a complete pic- ture of their ethology. Some BFF behavior (e.g., locomotor, predatory) was in part in- ferred from 243 BFF snow-tracking records collected over three winters 1981-1984 (Richardson et al. 1985 and unpublished data). Our behavioral descriptions were facili- tated by earlier behavioral observations (by TWC) of steppe ferrets (M. eversmanni , 32 hr) and European ferrets (M. putorius, 123 hr), as well as by ethological studies on other species. Where appropriate, we compare our observations with the literature on BFFs and other mustelids. Results We describe individual maintenance, in- traspecific social, and interspecific behavior patterns (the three major behavioral cate- gories often recognized; e.g., Balph and Stokes 1963), as well as BFF vocalizations. Photographs of some of these behaviors and BFF signs are in the Appendix. 'Department of Biological Sciences, Idaho State University. Pocatello, Idaho 83209, and Biota Research and Consulting, Inc., Jackson, Wyoming 83001. 115 116 Great Basin Naturalist Memoirs No. 8 G °^<^ Fig. 1. Some black-footed ferret motor patterns and body postures: A, Walking. B, Bounding. C, In-burrow alert posture. D, All-fours alert posture. E, Upright alert posture. F, Digging, G, Play. Maintenance Behavior Maintenance behavior is performed by an animal in the normal course of its daily activi- ties and is critical to its survival. Locomotion. — BFFs either walked or bounded (Fig. 1). Walking is a forward pro- gression in the typical quadruped manner — a cross-wise stepping movement. Forward movement of the left front leg was followed by the right hind leg, then the right front leg was followed by the left hind leg. The head was usually held above the torso but was occasion- ally lowered as if to sniff" the ground. The tail was usually held off" the ground, at a variable downward angle from the torso. BFF^s walked about 2% of the distance traveled per winter night, typically near prairie dog burrow en- trances. Bounding is a leaping run or gallop in which both hind feet and then both front feet are alter- nately set before one another, with the hind feet set fairly accurately in the twin tracks of the front feet. Travel between prairie dog (Cynomys leu- curus) holes and long distance movements were in this gait. BFFs often traveled in vegetation- free areas such as cattle and game trails, roads, snow-fflled gullies, and windblown hill crests. Relatively straight line movements of 75 m were common. Hillman (1968), Henderson et al. (1969), and Fortenbery (1972) described BFF movements between prairie dog burrow openings as "run- ning." They did not describe walking or bound- ing locomotion; however, photographs of BFF tracks in snow in Fortenbery (1972) were the bounding type. 1986 Clark et al. : Ethology and Activity Patterns 117 Alert behavior. — Alertness composed a high percentage of BFF behavior and was the only activity that frequently interrupted all others. Alertness was characterized by: (1) in- burrow alert posture ("periscope"), in which only a BFF's head, part of the head, or upper torso was visible; (2) down alert posture, aboveground alertness in which all four feet were on the ground; and (3) upright alert pos- ture, in which the BFF stood on its hind feet, balancing with its tail and hind legs, with its forelegs off the ground (Fig. 1). An immobile body was the common element of the differ- ent alert postures. The in-burrow alert posture or periscope was by far the most common alert posture. The down alert often occurred between bursts of locomotion, especially if the BFF was hunt- ing in tall vegetation with the prairie dogs active nearby. The upright alert posture was less frequently observed under similar cir- cumstance and was of very short duration. Alert postures have not been described for BFFs. However, Fortenbery (1972) noted that BFFs may look out of prairie dog bur- rows, with only their heads showing (our in- burrow alert). The limited descriptions and photos in Henderson et al. (1969:7,11) and Fortenbery (1972) suggest that the BFFs in Wyoming and South Dakota have similar repertoires of alert postures. Grooming and Sunning. — BFFs scratch, mouth, and bite at their fur. These activities are functionally related to dressing the pelage, cleaning the body surface, and remov- ing parasites (Eisenberg 1968). Scratching (n = 8) consisted of perpendicular movements of one hind leg directed at various points on the body. Mouthing movements (n = 4) are complex and variable and consisted of "biting" fur on the tail, legs, and ventral and lateral areas of the torso. Grooming of fur was evi- denced by BFF hairs found in BFF scats. Washing or licking were not seen. Ticks were relatively common behind the ears, on the upper neck, and under the chin of adult BFFs. BFFs bit at flies that flew near their faces. We also observed BFFs yawn while sunning, where the head is thrown back, mouth opened full gap, and eyes closed. Henderson et al. (1969) noted that an adult female BFF scratched a scab on her head with her hind paw and that young and adult BFFs seemed bothered by external parasites (ticks, fleas, and flies) and frequently scratched themselves. However, motor patterns were not described. Henderson (personal commu- nication 1983) observed BFFs in South Da- kota yawn. Sunning consisted of lying sternally station- ary on prairie dog mounds in sunlight. We observed this three times in midsummer be- tween 0800 and 1100 hrs Henderson et al. (1969) noted that adult BFFs often basked in the warm, midmorning sun for several hours on prairie dog mounds during the young care period (July-August), fall, and spring. Progulske (in Henderson et al. 1969:7) re- ported sunning behavior in a captive adult male BFF. Henderson (personal communica- tion 1983) observed BFFs basking in the sun in the snow. Defecation and urination . — About 75 scats of possible BFF origin and an additional 15 of known BFF origin were found from Decem- ber 1981 to January 1984 and are shown in Clark et al. (Handbook of methods, 1984). Of scats of probable BFF origin, two were found on top of each of two badger (Taxidea taxus) scats, several on BFF diggings, five beside a frozen BFF corpse in February 1982, and nearly all others near prairie dog burrow openings. Urinations (n = 114) along snow- track routes were generally located near bur- row entrances but did occur in midroute (Richardson et al. unpublished data). Henderson et al. (1969) noted that scats and urinations were deposited separately, usually near a burrow mound, but the salient feature of BFF scats is that they are seldom found (Hillman 1968, Henderson et al. 1969, Fortenbery 1972). Hillman (1968) assumed and Henderson et al. (1969) suspected that BFFs defecate under- ground. Droppings of a captive adult male BFF were deposited in one corner of the pen during summer and in the burrow box during winter (Progulske 1969). Sheets and Linder (1969) re- covered BFF scats from prairie dog burrows they excavated by machine. Digging behavior. — BFFs excavate subsoil from prairie dog burrows and deposit it in a distinctive manner (Hillman 1968, Hender- son et al. 1969, Hillman and Linder 1973, Hillman and Clark 1980, Clark et al. Season- ality of black-footed ferret diggings, 1984) (Appendix). We watched BFFs dig on nine 118 Great Basin Naturalist Memoirs No. y> 20 18 14 - 10 1 1 r -i — I — I — I — I — r J FMAMJ JASONDJ FMAMJ J ASONDJ FMAMJ 1982 1983 1984 Fig. 2. Total number of ferret diggings observed on four 4-ha (16 ha) plots within the Meeteetse black-footed ferret habitat. occasions, including the formation of two "diggings" or "trenches, " as these structures were usually labeled by earlier observers. We prefer the word "diggings," since the subsoil is piled on the ground outside a prairie dog burrow and not dug into the soil surface as implied by the term "trench." When digging in a prairie dog burrow, BFFs back out of the tunnel with loosened subsoil held against their chests by their front feet. They drag the material further from the entrance with each trip (Fig. 1). The subsoil is sometimes pushed under the BFFs body, which may then arch forward, with the hind feet kicking soil further backward (pictured in Henderson et al. 1969: 15). BFFs also dig furrows in snow several centimeters deep, but we have not observed how these are made. We observed results of BFF digging mostly during winter, when white-tailed prairie dogs were hibernating, but evidence of BFF dig- ging was noted during all other seasons (Clark et al. Seasonality of black-footed ferret dig- gings, 1984; Clark et al. Handbook of meth- ods, 1984) (Appendix). The freciuency of oc- currence and density of BFF diggings were seasonally marked on four, four-ha plots (Fig. 2). BFF diggings may be related to food acqui- sition. Seasonal peaks in diggings that could be identified as BFF occurred January-March and dropped to near zero by May each year. Both peaked in January, based on samples taken from January through December 1982, as described by Clark et al. Seasonality of black-footed ferret diggings, 1984; Clark et al. Handbook of methods , 1984) at about 4% and 2.5/ha, respectively, then dropped to near zero in April and remained very low until October, when they began to increase. South Dakota researchers agree that winter is the best time to look for BFF diggings: Hillman (1968) reported seeing BFF diggings in snow. Fortenbery (1972) noted that BFF diggings made during winter may persist for a long time. Henderson et al. (1969) observed more diggings in winter and in areas with small prairie dog populations. The excavated mate- rial may have been previously excavated by prairie dogs and subse(}ucntly brought to the surface by BFFs. The function of digging snow trenches is imknown. Hillman (1968) and Henderson et al. (1969) concluded that no 1986 Clark ETAL: Ethology AND Activity Patterns 119 other mustelid that visits prairie dog colonies digs or leaves subsoil deposited in a manner like BFFs, but other mustelids and prairie dogs do excavate subsoil. An adult female BFF on 8 August 1983 moved eight stones (seven about 2.5 cm in diameter and one about 12 cm long, 5 cm wide, and 2 cm thick) from the burrow mound into her burrow over 22 mins. Each stone was individually moved, in the cases of the seven small stones, with the mouth, and the single large stone was dragged with the forelegs. The function of this activity is unknown. Predatory behavior. — BFFs presumably obtain prey mostly at night below ground in- side prairie dog burrows. Our snow tracking indicated that, in addition to taking prairie dogs, BFFs also take small rodents (Per- omysciis manictdatus) , and lagomorphs. Dur- ing daylight in summer we saw BFFs kill prairie dogs and drag them to other holes on five occasions. During summer, one BFF leaped 0.7 m onto an adult prairie dog emerg- ing from a hole and bit the back of the prairie dog's head. The BFF and prairie dog fell down inside the hole during the struggle. Two min- utes later, the BFF emerged holding a prairie dog by the throat. The kill was dragged 10 m to a hole containing at least part of the female BFFs litter. Nine prairie dogs were on the surface within 40 m just prior to the kill. An- other BFF ran 10 m to, and descended down, a hole that a prairie dog had just descended. The upper body of the prairie dog emerged from the hole but apparently was dragged back down by the BFF biting its posterior. Two minutes later the BFF emerged dragging a dead prairie dog by its throat. On two occa- sions, a BFF ran up to a prairie dog burrow opening, stopped with its body head first halfway down the hole, and waited motionless about 4 mins. At this time, the BFF dove into the tunnel, and prolonged high-pitched prairie dog "screams " and BFF "growls " em- anated from the tunnel. On both occasions, the BFF emerged with a dead prairie dog within 5 mins. In all the above cases of preda- tion, the prairie dog prey had a bloody throat and no other observable wounds. On one occasion, a BFF bounded through the tall grass and shrubs and flushed out a ground squirrel (Spermophdus armatus). Within three additional bounds the BFF leaped on the back of the fleeing squirrel and seized it with a bite to the base of the skull. The BFF then descended a nearby ground squirrel burrow carrying the dead squirrel. Another BFF dragged two juvenile prairie dogs, one at a time, near us and dropped one. The killing bite appeared to be between the shoulder blades. Another time, a BFF ran toward a prairie dog 5 m away above ground but did not enter the hole the prairie dog retreated down. BFFs are active in winter, exploring various burrows along their movements. Once a BFF enters a burrow, presumably it locates and captures prey by sound and smell. It some- times takes prey above ground away from bur- rows. BFFs may remain below ground for several days in the same hole (Richardson et al., unpublished data) . Prey were apparently often consumed below ground in burrows where kills were made. BFFs dragged prairie dog carcasses to another hole. One such kill found along a BFF drag exhibited punctures and hemorrhaging in the neck area behind the head (Appendix) . In winter BFFs do not use any one burrow as a long-term nest burrow and may use some burrows as "cache " bur- rows (Richardson et al., unpublished data). Our observations of BFF "hunting" behav- ior and those described by Hillman (1968), Henderson et al. (1969), Hillman and Linder (1973), and Fortenbery (1972) indicate that the BFF is a "searcher" predator (Alcock 1975). Our observations and those by Hillman (1968), Henderson et al. (1969), and Pro- gulske (1969) in South Dakota are similar and suggest that killing behavior is stereotyped. BFFs kifl both young and adult prairie dogs (Hiflman 1968).'Progulske (1969) observed a prairie dog bite an adult male BFF on the face. The facial cuts on BFFs we saw could have been inflicted similarly. Social Behavior Social behavior refers to the interaction of two or more conspecifics. Interaction means that the animals are mutually influencing one another through some form of communication system (Eisenberg 1968). Reproductive behavior. — We did not ob- serve this, but snow tracking suggested that breeding activity began in mid-February and continued through March (as calculated from 120 Great Basin Naturalist Memoirs No. 8 the timing of litter emergence, estimated pre- emergence occupancy (45 days), and known gestation of 42-45 days (Hillman and Carpen- ter 1983). The initiation of reproductive activ- ity in spring is further supported by the fact that an aduh male BFF road-killed in early March near our main study area showed tes- ticular mitotic activity but no spermatozoa, indicating that spermatogensis was just begin- ning (Thorne 1982, personal communication). Also in February and March, we noted BFF movements tended to increase as did activity area sizes and marking (Richardson et al., un- published data) (Appendix). In South Dakota the exact timing of mating was unknown (Henderson et al. 1969), but captive BFFs bred in March and early April (Hillman and Carpenter 1983). Breeding behavior in cap- tivity is described by Hillman and Carpenter (in Hillman and Clark 1980). Ontogeny of young and maternal behav- ior. — The duration of time that young BFFs remain in the natal burrow before emerging above ground is unknown but is estimated at about 45 days. On 28 June 1982 a female moved a three-kit litter about 20 m from one hole to another. She carried one kit at a time in her mouth in three trips totaling 15 mins. The young were quite small (est. 200 g). In mid-July young in 1 1 litters appeared half- to three-fourths grown (est. 400-500 g). Nothing is known of BFF development between birth and first appearance above ground. Mother BFFs may interact in a variety of ways with their young. In July, shortly after young began appearing above ground, mother BFFs commonly pulled young BFFs out of a burrow with her teeth and dragged them by their napes to other holes. On 11 July 1984 we watched a female with four half- grown young at a burrow. Generally the young crawled on their bellies (eyes barely reflecting our spotlight) in an area around the female while she remained standing alert watching our spotlight. At times they would all go down a nearby prairie dog hole, but reappeared three times. One time, she stood on all four feet exceptionally still while the young crawled all over her especially at her belly (nursing ?) and this lasted about three minutes. Until late July, while probably still nursing, females "coaxed" up to four young out of a burrow and led them single file in "train behavior" (also noted by Henderson et al. 1969) across the prairie to a new site. On four occasions in July females brought dead prairie dogs to their young. On several occa- sions when a litter was above ground, the mother vocalized, after which all young rapidly descended into the burrow. When young are older, from late July to early Au- gust, litter mates are often seen separated — either in separate holes or one traveling with the female. On our approach she typically brought the group together by retrieving a lone juvenile or bringing the juveniles with her to the other juveniles and then keeping them all down, while she watched us. Play . — Young BFFs in play were very quick with a variety of flexible, elastic body move- ments (Fig. 1). They played at night and in daylight. Play was the most often observed social behavior and was common in late July and early August. We categorized the types of play (1) object play, (2) autoplay, and (3) social play (even though the first two types are nonsocial, they are included here for com- pleteness of play descriptions). In object play young BFFs exhibited close orientation and visual, oral, or olfactory inspection or manipu- lation of physical objects. One young BFF repeatedly "attacked" marker flags by jump- ing at them, front legs extended and mouth open. In autoplay young BFFs moved forward and backward, with legs sometimes down to- gether, back arched, chasing their own tails while turning their bodies around and around, rolling over on the ground, and changing position by "snapping" their bodies into the air at split-second intervals. In social play two or more young BFFs en- gaged in approach-withdrawal (noncontact) or rough-and-tumble (contact) play, with the re- cipient of the play initiation either avoided or joined (Fig. 1). In approach-v\'ithdrawal play they constantly alternated distance between themselves as they chased and bounded for- ward and back-ward. The role of the pursuer and pursued were frequently interchanged within a single session. In these encounters mouths were sometimes open, the head was Ik^IcI from above to below the height of the shoulders, the tail was often extended with hairs erect, and the back was arched high (Fig. 1). No vocalizations were heard, possibly be- 1986 Clark et al.: Ethology and Activity Patterns 121 cause we were too distant to hear them. This form of play occasionally followed or preceded rough-and-tumble play, in which young BFFs bit and tumbled with their interlocked bodies rolling about. Play activity occurred on and off" burrow mounds and lasted up to 20 mins. Young BFFs also exhibited a "stiff-legged dance" form of play in which they alternated approach-withdrawal among themselves and once toward a human. The limbs were alter- nately stamped against the ground, first one or both front feet and then one or both hind feet. Hillman (1968) noted that young BFFs played above ground, running in and out of burrows in pursuit of one another. They bit and pulled at each other, humped their backs, and ran on their toes, often turning in circles, attempting to bite their tails. Henderson et al. (1969) noted one young BFF executed a midair somersault similar to what we ob- served. Scent-marking. — Scent-marks have been observed in the snow (Richardson et al., un- published data), but marking behavior is rarely observed (Appendix). One adult male was observed marking near a burrow opening in July 1984. On a grass substrate, he dropped his pelvic region on the ground with his tail extended straight behind him, the tip about 12 cm off" the ground. He moved foi-ward about 20 cm, mainly using his forelegs while dragging and wiggling his pelvic region against the substrate. He then scraped back- ward with his hind feet about four times over this area. This behavior was repeated twice more in two different grassy areas. After that, he moved to a small bush over which he ex- tended his whole body, such that the neck, abdomen, and pelvis were rubbing into and wiggling through the bush. After a vigorous rubbing, he moved his forelegs and abdomen off" the bush, leaving his pelvic region on the bush and rubbing into it another 3 sec. He circled around and repeated this marking be- havior on the same bush two more times. The BFF may also have urinated on the bush dur- ing the rubbing procedure. Each marked area exuded a strong musk odor. This behavior may have been in response to the observer standing 5 m away. After the marking behav- ior, the BFF moved away slowly to explore new burrows. Agonistic Behavior. — Agonistic behavior (conflict between two animals; Scott 1962) was not observed by us or the South Dakota re- searchers. Human-Ferret Interactions Ferret responses to human activities varied. When spotlighted BFFs generally oriented toward the light and vehicle, at least momen- tarily. Some moved to or stayed in prairie dog burrows, some retreated into burrows for ex- tended periods, and some continued their ac- tivities. When BFF heads were visible out of a hole, their eyes were often turned away from the light, perhaps avoiding the direct beam. Family groups were shy in mid-July but tended to be less shy later. Juveniles seemed to be more shy in our presence than adults. One BFF spotlighted at 75 m appeared to direct an "agoniso the observer standing 5 m way. After the marking behavior, the BFF lowly to explore new burrows, ea. This behavior was repeated twice more in different grassy areas. After that, he moved to approached to within 2 m of us in our vehicle and on one occasion walked under the truck. When approached on foot, BFF responses again varied. During daylight they typically retreated to burrows, observed us for short periods from the hole, and then descended the burrow. However, the distance from us at which BFFs retreat to burrows varied from 10 to 100 m. Again, juveniles appeared more shy. During daylight young BFFs popped their heads out of burrows, apparently ob- serving passers-by within 100 m, but re- treated if the observer approached directly. One adult female was followed at 10 m on hunting forays on 14 occasions with no appar- ent alteration in her behavior because of our presence. At night several individuals, both juveniles and adults, were quietly and slowly approached in the spotlight beam or with ffashlights to within 5 m. BFFs were wary of us, stayed near holes, and "hissed" at us but overall seemed curious about our activities. Whereas some BFFs later retreated to bur- rows, others moved slowly between burrows. BFF response to spothghting disturbances was briefly evaluated by Campbell et al. (1985). 122 Great Basin Naturalist Memoirs No. 18 ■ 16 ■ A ,» 12 ■ 10 - 8- 1983 / '^''- 6 - ^ A T''*\I \ 4- 2- n /jW ^ § 32 - Si 26- 1984 5 24 - z 22 - i \ 20- 18 n.4, ..« ,V' ■; 16- 14- A' 12- 10- 8- J» ■■■"' /v \* 6- 4- 2- ■3IOO 55no o^nn TIME (MOT.) Fig. 3. Observation frequencies (by 0.5 hr intervals) for ferret litters (solid line) and single individuals and litters (dash line) on full night surveys (1982), partial night surveys (1983) (observer nights = 41), and late night surveys (1984) (observer nights = 108). Vocalizations BFFs emit several sounds heard by us and reported in previous studies. Based on the source and the context, we classified calls as: (1) threat, (2) defense, (3) greeting, (4) mating, and (5) sounds by young. We heard calls we labeled "bark, " "huff-hiss, " "growls, " "ungh, " and "chattering-bark." Hillman (1968) heard a "hiss," "snarl," and "bark." Henderson et al. (1969) heard an "ungh" (adult ferret to her young) and a "noise" (among young playing in a burrow). Progulske (1969) heard a "chatter" and a "low hiss." Finally, Hillman and Car- penter (1983) noted a "whimpering" (in copu- lation). Because contextual definitions are in- complete for each call, we did not assign each call to a "functional" category. Activity Patterns Activity was defined as any appearance of BFFs above ground. The way in which a pop- ulation structures its activity patterns reflects its survival strategy (Orians 1961). We readily located BFFs during most months, including winter, but, like others (Hillman 1968, Hen- derson et al. 1969, Fortenbery 1972), we found that the animals themselves are easiest to locate during summer. Hillman (1968) found BFFs most active in late evening (1900-2400) and early morning (0200-0600). Frequency of BFF observation from spotlight searches conducted during July-August from 1982-1984 are shown in Figure 3. Snowtrack evidence and direct observation showed that BFFs were active after sunrise in winter. In July BFFs were observed to be active until 1200. We watched one BFF and her litter of four for 14 consecutive mornings. She became active each morning around 0830, 2 hr after sunrise, and hunted for an hr or more. After 0930 in July, BFFs made sur- face appearances for short durations off and on until 1200 and rarely in the afternoon. Also, the same BFFs were often visible at the same hrs and in the same locations. BFFs were active above ground in tempera- tures of -38 C, during snow and rainstorms, and in winds up to 50 kph. Richardson et al. (unpublished data) concluded that tempera- ture did affect BFF movements. Activity pat- terns of radio-tagged BFFs were described by Biggins et al. (1985). Discussion Although the Wyoming BFF population and the one formerly in South Dakota are associated with different prairie dog species, live in different biogeographic areas, and are under different climatic regimes, much of their gross behavior and activity patterns are similar. Even though a systematic description of BFF behavior was not previously available, our categories allow for inclusion of BFF ob- servations from South Dakota. The BFF is difficult to observe, generally being nocturnal and appearing above ground at irregular in- tervals and for irregular durations. Thus, our 1986 Clark et al. : Ethology and Activity Patterns 123 ethogram and activity data are incomplete, but BFF behavior apparently is similar to re- lated species and much BFF behavior is prob- ably homologous to other species oi Mtistela . Where behavioral data are currently lacking for BFFs (e.g., reproductive, agonistic, and ontogenetic behaviors), the most complete lit- erature on related mustelids can be used to suggest BFF behavior patterns until observa- tional data for the rare BFF become available. Furthermore, a comparative behavioral ap- proach, as discussed by Eibl-Eibesfeldt (1970) and previously elucidated by Remane ( 1952), allows identification of homologous behavior patterns if they occur in a large number of closely related species. Steppe ferrets live in large ground squirrel (suslik; Spennophilus spp.) colonies, similar to prairie dog colonies, but, in contrast to BFFs, distinctive deposition of excavated subsoil by steppe ferrets is not mentioned in the literature even though steppe ferrets do dig out susliks in winter (Stroganov 1969). Feces and urine are deposited by BFFs as waste products but may also serve in "scent marking" (Macdonald 1980). Currently it is impossible to distinguish between feces and urine as elimination products or scent marks (see Wells and Bekoff 1981). Steppe ferrets and BFFs hunt similarly (this study, Hillman 1968, Henderson et al. 1969). Killing methods of M. frenata, M. erminea, M. rixosa, M. vison, and M. ptitoriiis are basi- cally similar (Iwen 1958). Predatory behavior of M. nivalis, M. erminea, and M. putorius is similar, especially for the two weasels (Gos- sow 1970). The killing procedure for M. ni- valis is generally very rapid, ranging from 10 to 60 sec (Heidt 1970). Ewer (1973) character- ized all mustelids as solitary, opportunistic predators whose hunting behaviors include a "random search" foraging pattern and a neck bite for killing. The lack of "aggressive" behavior by the male BFF during copulation was unlike that for M. eversmanni and M. putorius (Hillman and Carpenter 1983). Other mustelid species display a copulatory pattern similar to that described for BFFs (e.g. , Wright 1948, for M. frenata; Hartman 1964, for M. nivalis; and Rowe-Rowe 1977, for Ictonyx striatus and Poecilogale albinucha). The timing of BFF reproductive activity, as suggested by our ob- servations, corresponds well with observa- tions of Henderson et al. (1969) and Hillman and Carpenter (1983) for the BFF and is simi- lar to the seasonality of reproductive activity for M. putorius (Walton 1976, Danilov and Rusakov 1969). BFF growth curves, unknown at present, may be estimated based on limited data for the steppe ferret (Sviridenko 1935), our lim- ited observations, data from the South Dakota ferret studies (Henderson et al. 1969), and from other mustelid studies (e.g.. East and Lockie 1964, 1965). Young of several mustelid species perform certain behaviors in play that probably serve them in predatory and other behaviors as adults (Gassow 1970). Play by P. albinucha and 7. striatus was mainly aggressive, involv- ing actions typical of adult fighting, prey cap- ture, and killing (Rowe-Rowe 1977). The ag- gressive play of M. putorius appears similar to our observations and those of BFF in South Dakota. Mustela putorius young exhibited all three types of play (Poole 1970) that we de- scribed for M. nigripes . BFF play probably contains many of the motor components ex- hibited by adults in agonistic and reproduc- tive behaviors, but because of the secretive, solitary, nocturnal habits of free-living BFFs, these motor patterns in an adult context are virtually impossible to observe. The stiff-legged dance we described for young BFFs corresponds to similar behavior in M. nivalis (Heidt 1970) and in Maries spp. (Schmidt 1943), in which cases the behavior pattern was thought to be agonistic. Agonistic l3ehavior has been extensively studied in M. erminea (Erlinge 1977), M. putorius, and M. furo (Poole 1966, 1967, 1972a, b, 1973, 1974) and provides descriptions that may be similar to BFFs. Our listing of vocalizations for BFFs and those from South Dakota are generally comparable to vocalizations for other muste- hds (Gossow 1970, Huff and Price 1968, Svendsen 1976, Goethe 1974, Channing and Rowe-Rowe 1977, Belan et al. 1978). Daily and seasonal activity for the Wyoming BFFs varied somewhat from that for South Dakota BFFs. Time allocation by a species reflects differences in habitat and social orga- nization (e.g. , Greenlaw 1969, Post and Baulu 1978). Our data on tracks, scats, and activity patterns have implications for conducting 124 Great Basin Naturalist Memoirs No. 8 BFF surveys, which are discussed elsewhere (Clark etal. Handbook of methods, 1984). We sought to minimize our direct contact with BFFs to reduce research impacts. Even though our data adds to an understanding of BFFs, much yet remains to be learned, in- cluding more complete behavioral descrip- tions and quantification of our ethogram. But, as noted by Marler (1968), the building up of descriptions is itself a quantitative process and an essential first step in revealing the behavior and ecology of a species. Acknowledgments Our studies were generously supported by the New York Zoological Society, the Wildlife Preservation Trust International, Inc., the National Geographic Society, the World Wildlife Fund— U.S., Charles A. Lindbergh Fund, the National Wildlife Federation, The Nature Conservancy, the U.S. National Academy of Sciences, Defenders of Wildlife, Sigma Xi, the Lander and Casper Wyoming Audubon clubs, and the Humane Society of the U.S. Several ranchers kindly made their ranches open to our work. William J. Bar- more, Jr., Mike Wells, Jon Jensen, Archie Carr 111, Craig Groves, Phil Lehner, Conrad Hillman, and Bob Henderson provided criti- cal review of the manuscript. A very sincere thank you to all the people and organizations that made our work possible. Appendix Photographs of black-footed ferrets illustrating various behaviors and ferret signs. A. Adult female ferret and young male in alert (by Tim Clark). 1986 Clark et al.: Ethology and Activity Patferns 125 Ferret walking (by Doug Brown). C. Adult female in alert (by Tim Clark). 126 Great Basin Naturalist Memoirs No. 8 D. Adult female with prairie dog she just killed (by Tim Clark). ^■ a U^^^^M^Ji--J. E. Adult female in alert (ui)U uJ i\ I iin Clark). 1986 Clark ETAL: Ethology AND Activity Patterns 127 m W.MW'^'' \^\y^^'^ S F. Adult female emerging from burrow with just-killed prairie dog (by Tim Clark). fm G. Adult female with prey — prairie dog (by Tim Clark). 128 Great Basin NATnuLisT Memoirs No. 8 H. Adult female with prey — prairie dog (not a throat bite) (by Tim Clark). I. Juvenile tenet alert mi piainc cioi; hole (b\ I mi ( .lark) 1986 •■..•i' r Clark et al.: Ethology and Activity Patterns 129 J. Adult ferret hunting prairie dogs (by Tim Clark). K. Ferret dragging dead prairie dog back to her natal burrow containing some ofher young (by Tim Clark). 130 Great Basin Naturalist Memoirs No. 8 L. Alert ferret (by Tim Clark) M ^*-< M -l^v %-% M. Ferret snow marking. Track.s in photo indicate a ferret scraped or scratched throngh the snow into the substrate in a circular area (foreground al)()ut 2.5 cm diameter), made a trough in the snow with its l)ody, and rulibed its body over and through the small shrub in backgrouiui. (Considered a scent-marking liehaxior (b\- Tim (Mark). 1986 Clark et al. : Ethol()(;y an d Activity Patterns 131 N N. Ferret snow marking and dirt scrape. Tracks indicate that a ferret entered the prairie dog burrow (hole diameter 10 cm), excavated a small amount of subsoil onto the snow (dirt scrape), and moved 0.3 m away from the burrow where it scraped or scratched through the snow within a roughly circular area, probably a scent-marking behavior. Note ferret tracks exiting upper left (by Louise Richardson). o 0. a type of ferret digging. A ferret excavated subsoil from within a prairie dog burrow. Ferrets pull dirt out of the burrow holding it against their chests with their forepaws as they move backward, depositing the dirt in a linear fashion away from the burrow opening and sometimes making a distinctive trough or "trench" within the excavated subsoil. Digging length is about L5 m (by Tim Clark). 132 GreatBasin Naturalist Memoirs No. 8 P p. A ferret rubbed its body over and through the shrub in foreground (shrub about 22 cm high and 30 cm wide), with ferret tracks evident around the base of the shrub. Behind the shrub and to the right is a patch where the ferret scraped or scratched through the snow into the substrate (snow marking about 20 cm in diameter). Both markings are probably scent marking (by Louise Richardson). Q f Q. A ferret kill drag. Ferret entered burrow (dark area in foreground about 50 cm in diameter) from right (note dual print tracks), apparently killed a prairie dog in the bur- row, and dragged it out and away from the burrow with tracks exiting to the left. The trough like depression (about 18 cm wide) was from the prairie dog's body being drug in the snow by the ferret, whose tracks are seen along the left side of the slide marks (by Tim Clark). 1986 Clark et al.: Ethology and Activity Pattkrns 133 Literature Cited Alcock, J. 1975. Animal heliavior: An evolutionary ap- proach. Sinauer Assoc, Inc. Sunderland, Mass. 547 pp. Balph, D F , and a W Stokes 1963. On the etholoj^y of a population of Uinta ground squirrels. Aincr. Midi. Nat. 69:10f>-126. Belan, I , P N Lehnek. andT W Clark 1978. VocaHza- tions of the American pine marten, Maries ameri- cana. J. Mammal. 59:871-874. Biggins, D., M. H. Schroeder, S. C. Forre.st, and L. Richardson. 1985. Movements of radio-tagged black-footed ferrets. Pages 11.1-11.17 in S. An- derson and D. Inkley, eds., Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Campbell, T M III, D. Biggins, S. C Forrest, andT W Clark 1985. Spotlighting as a method to locate and study black-footed ferrets. Pages 24. 1-24.7 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc. , Laramie, Wyoming, Sep- tember 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Candland, D, K. 1974. The physiological ethogram. Bull. Psychonomic Soc. 18:264-265. Channing, A., and D T Rowe-Rowe 1977. Vocalizations of South African mustelines. Z. Tierpsvchol. 44:283-293. Clark, T. W, T. M. Campbell, M H Sc:hroeder, and L. Richardson. 1984. Handbook of methods for lo- cating black-footed ferrets. Wyoming BLM Tech. Bull., Cheyenne. No. 1. 55 pp. Clark, T. W., L. Richardson, D Casey. T M Campbell. and S. C. Forrest. 1984. Seasonality of black- footed ferret diggings and prairie dog burrow plugging. J. Wildl. Manage. 48: 1441-1444. Danilov, P. I.andO. S. Rusakov. 1969. Special aspects of the ecology of the polecat {Mustela ptitoriiis) in the northwestern region of the European USSR. Zool. Zh. 48:138,3-1394. East, K, and J D. Lockie. 1964. Observations on afamily of weasels (M. nivalis) bred in captivity. Proc. Zool. Soc, London. 143:.359-363. 1965. Further observations on weasels (M. nivalis) and stoats (M. erminea) born in captivity. Proc. Zool. Soc, London. 147:234-238. Eibl-Eibesfeldt, I. 1970. Ethology: the biology of be- havior. Rinehart & Winston, New York. 530 pp. ElSENBERG, J. F. 1968. Behavior patterns. Pages 451-495 in J. A. King, ed.. Biology of Pe7-07n(/.sci/s (Roden- tia). Amer. Soc. Mammal. Spec. Publ. No. 2. Erickson, R. C. 1973. Some black-footed ferret research needs. Pages 153-164 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State Uni- versity, Brookings. Erlinge, S. 1977. Agonistic behavior and dominance in stoats {Mustela erminea L.). Z. Tierpsvchol. 44: 37,5-,388. Ewer, R. F. 1973. The carnivores. Weidenfeld and Nicholson, London. 494 pp. Fortenberv, D. K 1972. CJharacteristics of the black- footed ferret. USDl Res. Publ. No. 109. 8 pp. C;oetiie, F 1964. Das verhalten dei musteliden Mandb. Zool. Berl. 8:1-80. Ciossow, II. 1970. Vergleichend verhaltensstudien und marderartigen I. uber lautausserungen und zum beuteverhalten. Z. Tierpsychol. 27:405-^80. C;heenlaw. J S. 1969. The importance of food in the breeding systems of the rufous-sided towhee, Pip- ilo erythrophtalmus (L.). Unpublished disserta- tion, Rutgers University. 169 pp. Hartman, L. 1964. The behavior and breeding of captive weasels {Mustela nivalis L.). New Zealand J. Sci. 7:147-156. Heidt. G a 1970. The least weasel Mustek nivalis Lin- naeus: development biology in comparison with other North American Mustelida. Michigan State University Publ. Biol. Sci. 4:229-282. Henderson. F R., P F Springer, and R. Adrian. 1969. The black-footed ferret in South Dakota. South Dakota Dept. Game, Fish, and Parks, Pierre. 37 pp. Hillman, C N 1968. Field observations of black-footed ferrets in South Dakota. Trans. N. Amer. Wildl. and Nat. Res. Conf 33:346-349. Hillman, C N , and R. L Linder. 1973. The black-footed ferret. Pages 10-23 in R. L. Linder and C. N. Hillman, eds., Proc. Black-footed Ferret and Prairie Dog Workshop, South Dakota State Uni- versity, Brookings. Hillman, C. N., and J. W. Carpenter 1983. Breeding biology and behavior of captive black-footed fer- rets, Mustela niaripes. Int. Zoo Yrbk. 23:186-191. HillmanC. N, andT W Clark. 1980. Mustela nigripes. Mammal. Species No. 126. 4 pp. Huff. J N.. and E. O. Price 1968. Vocalizations of the least weasel, Mustela nivalis. J. Mammal. 49:548-550. IvvEN, F 1958. Killing methods of mustelids. Unpub- lished thesis, University of Wisconsin, Madison. 39 pp. Klopper, p. H , and J P Hailman 1967. An introduction to animal behavior. Prentice-Hall, Inc., Engle- wood Cliffs, New Jersey. 297 pp. Lehner. p. N. 1979. Handbook of ethological methods. Garland Press, New York, New York. 403 pp. MacDonald, D, W. 1980. Patterns of scent marking with urine and faeces amongst carnivore communities. Symp. Zool. Soc, London, 45:107-139. Marler, p. 1968. Observation and description of behav- ior. In E. O. Price and A. W. Stokes, eds., Animal behavior in laboratory and field. W. H. Freeman and Co., San Francisco, California. 130 pp. Orians, G. H. 1961. The ecology of blackbird {Agelaius) social systems. Ecol. Mongr. 31(3):285-312. Poole, T. B. 1966. Aggressive play in polecats. Symp. Zool. Soc, London 18:23-44. 1967. Aspects of aggressive behavior in polecats. Z. Tierpsychol. 24:351-369. 1970. Polecats. HMSO, London, Forestry Comm. Forest Record 76:1-17. 1972a. Some behavioral differences between the European polecat, Mustela putorius, the ferret M. ftiro, and their hybrids. J. Zool., London 166: 2.5-35. 134 Great Basin Naturalist Memoirs No. 1972b. Diadic interactions between pairs of male polecats {Mtistelafuro and M. ftiro X M. ptitorius hybrids) under standarided environmental condi- tions during the breeding season. Z. Tierpsychol. 30:4.5-58. 1973. The aggressive behavior of individual male polecats (Mustela putorius, M. fiiro, and hybrids) towards familiar and unfamiliar opponents. J. Zool, London 170:395-^14. 1974. Detailed analysis of fighting in polecats (Mtistelidae) using cine film. J. Zool., London 173:369-.393. Post, W., and J. Baulu. 1978. Time budgets of Macaca miilatta. Primates 19:12.5-140. Progulske, D. R. 1969. Observations of a penned, wild- captured black-footed ferret. J. Mammal. 50: 619-620. Remane, a. 1952. Der Grundlagen das naturlichen Sys- tems der vergleichenden anatomic und phylo- genetik. Leipzig. 364 pp. Richardson, L., T. W Clark, S C Forrest, and T. M Campbell III. 1985. Snowtracking as a method to search for and study the black-footed ferret. Pages 25.1-25.11 in S. Anderson and D. Inkley, eds.. Black-footed Ferret Workshop Proc, Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Rowe-Rowe, D. T. 1977. Reproduction and post-natal development of South African Mustelines (Car- nivora: Mustelidae). Zoologica Africana 13: 103-114. Schmidt, F. 1943. Natural history of the spruce and stone martens: with comparative observations on their nearest relatives, especially the Siberian sable and the American pine marten. Publ. Inst, fur Jagd- kunde der Univ., Gottingen. 268 pp. Scott, J P. 1956. The analysis of social organization in animals. Ecology 37:213— 221. 1962. Introduction to animal behavior. In E. S. E. Hafez, ed.. The behavior of domestic animals. Williams and Wilkins Co. Baltimore, Maryland. 619 pp. Sheets, R. G , and R L Linder 1969. Food habits of the black-footed ferret {Mustela nigripes in South Da- kota. Proc. South Dakota Acad. Sci. 48:58-61. Strocanov, S. U. 1969. Carnivorous mammals of Siberia. Acad. Sci. USST, Siberia Branch, Biol. Inst. 381-394. SUSTARE, D. 1975. Classification of behavioral units. Pa- per presented in Symposium on Quantitative methods in ethology. Anim. Behav. Soc. Meeting 23 May 1975, Wilmington, North Carolina. Svendsen, G. E. 1976. Vocalizations of the long-tailed weasel (Mustela frenata). ]. Mammal. 57: 398-399. SviRiDENKO, P. 1935. The light ferret (Putorius evers- manni) and its economic importance. USSR Publ. Walton, K. C 1976. The reproductive cycle in the male polecat Putorius putorius. J. Zool. London 180: 49f^503. Wells, M C , and M Bekoff. 1981. An observational study of scent-marking in coyotes Canis latrans. Anim. Behav. 29:332-350. Wright. P L. 1948. Breeding habits of captive long-tailed weasels (Mustela frenata). Amer. Midi. Nat. 39: 338-344. ACTIVITY OF RADIO-TAGGED BLACK-FOOTED FERRETS Dean E. Biggins', Max H. Schroeder', StovtMi C. Fori-fst", and Louise Richardson^ Abstract. — Activity of two radio-tagged hlack-footed ferrets {Mttstehi nigripes) was investigated during October- November 1981 (an adult male monitored for 16 days), and during August-November 1982 (a young female monitored for 101 days). Aboveground activity of the male averaged 2.95 hr/night, 15% of the total time monitored. From 22 September to 5 November, aboveground activity of the female averaged 1.9 hours; 26% of the time she was stationary and 74% of the time she was moving. During August the juvenile female emerged at least once on 9.3% of the nights. She was least active in November. Both animals were primarily nocturnal (although daylight activity was not uncommon), and timing of nightly activity was similar, peaking from 0100 to 0359. The discovery of a population of black- footed ferrets near Meeteetse, Wyoming, in 1981 (Schroeder and Martin 1982), provided an opportunity to investigate the behavior of this rare animal. We collected activity data on an adult male ferret radio-monitored during fall 1981, and a juvenile female ferret radio- monitored during late summer and fall of 1982. Spatial aspects of the activity of these two ferrets were summarized by Biggins et al. (1985); this paper addresses temporal ele- ments of activity. The timing of ferret activity and attendant causes has intrinsic value; however, the topic is also critical to refining spotlighting as a tool in ferret research and management (Campbell et al. 1985). Spotlighting success can be im- proved by knowing the time of year and time of night when ferrets are most active. Our small sample provides the first quantitative assessment of wild black-footed ferret time- activity patterns using radio telemetry. Study Area and Methods Black-footed ferrets occur on a 3000 ha com- plex of white-tailed prairie dog (Cynomys leii- curus) towns near Meeteetse, Wyoming. The site is a short-grass prairie at elevations rang- ing from 2000 to 2300 m. Vegetation and other site characteristics are described by Collins and Lichvar (1986) and Clark et al. {Descrip- tion and history , 1986). Ferrets were captured in Sheets' (1972) cylindrical trap and immobilized with ke- tamine hydrochloride. Animals were fitted with a 15-g transmitter collar and allowed to recover fully from anesthesia before release. Descriptions have been given of trapping and handling procedures (Thorne et al. 1985) and of development of the transmitter packages (Fagerstone et al. 1985). Ferrets were radio- tracked from 30 October to 14 November 1981 (the male) and from 13 August to 30 November 1982 (the female). Telemetric monitoring was continuous during 16 Octo- ber— 5 November 1982; for other periods monitoring was mostly during the hours of darkness. Radio-tracking in 1981 consisted primarily of simple signal-following with hand-held Yagi antennas. We recorded time the male ferret spent above and below ground, but we did not attempt to separate aboveground activity into "moving" or "sta- tionary" categories. Hand-held antennas were again employed in 1982, but most radio- tracking involved triangulation from pairs of mobile tracking stations (Biggins et al. 1985). The refined techniques and equipment al- lowed three types of signal status (involving strength and constancy of direction) to be tele- metrically correlated with different activities: (1) changes in bearing indicated that the ani- mal was moving aboveground, (2) consistent bearings with audible signal indicated that an animal was near or on the surface but rela- tively stationary, and (3) sudden loss of signal usually occurred when the animal went un- derground. We were able to describe daily 'U.S. Fish and Wildlife Service, Denver Wildlife Research Center, P. O. Box 916, Sheridan, Wyoming 82801. -BIOTA Research and Consulting, Inc., Box 2705, Jackson, Wyoming 83001, and Department of Biological Sciences, Idaho State University, Pocatello. Idaho 83209. 135 136 Great Basin Naturalist Memoirs No. 8 and seasonal changes in activity of the female from the 101 days of monitoring in 1982; how- ever, the limited detail and short (16-day) radio-tracking period made the data on the male best suited for only general comparisons with the female. To avoid semantic confusion, we developed the following definitions; period = a seasonal time category, measured in days, resulting from subdivision of the year. interval = a time category, measured in minutes or hours, resulting from subdivision of the day. bout of activity = a session of ferret activity occurring aboveground (or mostly aboveground), lasting >20 min and separated from other such bouts by > Ihr. Data were summarized by tabulating pres- ence or absence of received signal in time intervals. For overall seasonal analysis of ac- tivity data on the female, a day was divided into 48 0.5-hr intervals. For each interval, the following questions were asked and posi- tive responses recorded; (1) Was the animal telemetrically monitored during this period? (2)Was the signal absent during any part of the period? (3) Was the signal present during any part of the period? (4) Was the animal moving during any part of the period? (5) Was the animal stationary for at least part of the time? With this approach, it was possible (but un- common) to have each category of activity present in a single monitoring period. There- fore, the resulting frequency data is not addi- tive between categories; e.g., the total num- ber of intervals with audible signal is usually not the sum of intervals containing movement and intervals containing stationary activity. Relative importance of stationary and move- ment activity for the female ferret was deter- mined by contingency table analysis using four rows of seasonal periods and two columns that represented the number of 0.5-hr peri- ods in which the number of days with move- ment exceeded the number of days with sta- tionary activity and vice versa. Only the 21 periods from 1930 to 0600 (roughly sunset to sunrise) were included because the sample size of monitoring periods during other times of the day was too small for some seasonal periods (> 10 days of monitoring were deemed necessary). The seasonal emergence times of the female were compared by splitting nights (sunset to sunrise) into equal quarters to tally emer- gences. For a seasonal period, length of quar- ters was the average amount of time between sunset and sunrise during that period divided by four, quarter length being longer later in the season. Only emergences for bouts of ac- tivity as defined above were considered. Activity for the female and male from 30 October through 13 November was compared using total minutes of aboveground activity within 3-hr intervals and for each night. When gaps in monitoring occurred in either data set, corresponding time periods were deleted from both sets. This procedure allows a comparison of two animals monitored for exactly the same time periods but during two different years. Standard Chi-square tests for goodness-of-fit, Chi-square tests of indepen- dence (contingency table analysis), and t-tests were used to evaluate statistical significance of relationships, with the rejection level estab- lished at P =0.05. Times are indicated on the basis of the 24-hour clock and Mountain Stan- dard Time (M ST). Results General Activity Patterns of the Female Ferret During 23 September-5 November, 39 bouts of activity by the female ferret were monitored in entirety (from first appearance of radio signal to final disappearance). The average length of a bout was 1.9 hr, and an 11.7-hr bout on 27 September was the longest. During that bout, movements oc- curred only during the first and last hoius and were separated by 9.7 hr of stationary time during daylight. Stationary time ranging from 0.1 to 9.7 hr often preceded movement (20 occasions) or followed movement (13 occa- sions). Movement composed 74% of 1,569 min of activity sampled 30 October- 13 November. Daily and Seasonal Activity Patterns of the Female Ferret From August through mid-September, the female ferret was in transition from "a social and dependent young animal to a relatively solitary and independent individual' (Biggins et al. 1985). In October and November her behavior mav have differed from that of adult 1986 Biggins et al.: Radio-tagged Fehrpzts 137 80 - 13 August -12 September Period 60- /\ i'/^ ^> ;\ 40 ■ n 1 \ \ 20 - Jr\_j [y \^ \ M\ 1 V 1 ' ' A 1 ' ' ' 1 ' ' ' 1 'A' ' 1 ' ' ' 1 ' ' 2 O s. 22 September- 5 October Period ( ^ -\ r'l 1 \ \ / \ .\ ,--, ^.J ^ 1 ' ^ \ / \t\r \ / 1 ' k ' 1 ' ' ' 1 ' ' ' 1 ' h ' 1 ' ' ' 1 ' CO -J cr 2 80- 60- 16 October-5 November Period H 2 40- •^-"nZ.^A ^ O 1— 20- /AV, A \ 2 tij / ^^ / aA^ K^ -^^-y\ LU 1 14 ' ' 1 1 1 1 A' r ' '' . 1 , , Fig. 1. Activity of a female black-footed ferret during 4 seasonal periods. 138 Great Basin Naturalist Memoirs No. 8 females with established home ranges and presumably better hunting skills. The female was most active in the 13 Au- gust- 12 September and 22 September-5 Oc- tober periods, with > 2/3 of the days in five 30-min intervals containing aboveground ac- tivity. In the 6-30 November period, no 30-min interval had aboveground activity for more than 37% of the days monitored. Peaks in overall aboveground activity occurred within the four intervals from 0230 to 0429 in all four seasonal periods. The lull in all types of activity from about noon to sunset was similar to behavior of South Dakota ferrets (Hillman 1968). However, the female was active at least once during each hourly interval of day and night at some time during the study. (Activity does not appear in the noon to sunset interval on any graph because of the sample size re- striction mentioned.) Most daylight activity of the female occurred within the five hours fol- lowing sunrise. Similar behavior was ob- served in unmarked ferrets in the Meeteetse area (Clark et al. Descriptive ethology, 1986) and was noted in South Dakota (Hillman 1968). Morning activity was especially fre- quent during the 13 August- 12 September period (Fig. 1). The female was active from 0830 to 0859 on half the days monitored dur- ing this period. The morning peak remained in the 16 October-5 November period but was delayed, perhaps due to later sunrise. From 14 to 28 August the female (and her litter-mates) could be characterized as active and predictable. She had at least one bout of activity on 14 of 15 days (93%). She had a second bout on 8 days (53%) and a third bout on 5 days (33%,). On 11 of the 15 nights, she emerged between 1910 and 2005, shortly af- ter sunset. On 6 of the 15 nights, she emerged between 0050 and 0210 for either the first or second bout, and all five of the third l^outs were in the interval 0616-0730 (at least 0.5 In- after sunrise). Over 80% (22 of 27) of all emer- gences occurred within the three intervals listed above. The prominent bimodal peaks of night activity during the period when the fe- male was part of a litter (Fig. 1) progressively changed to a more uniform distribution of activity by November. Hillman (1968) and Clark et al. (Descriptive ethology, 1986) also found a bimodal distribution in ferret activity, but timing diilered. (vomparison with these data is difficult, because Hillman's (1968) summary covered the entire period from April through November and the information of Clark et al. {Descriptive ethology, 1986) covered July-August. As implied by the seasonal depictions of activity (Fig. 1), emergence times for bouts of activity were not equally distributed through the night. When nights were split into quar- ters, significant departures from equal num- bers of emergences each quarter were noted in two of the four seasonal periods (Chi-square goodness-of-fit, d.f = 3; 13 August- 12 Sep- tember, X' = 13.45, P = 0.004; 22 Septem- ber-5 October, X' = 5.78, P = 0.123; 16 October-5 November, X' = 2.429, P = 0.488; 6-30 November, X' = 11.35, P = 0.010). The female emerged more than expected in the first and third quarters during the 13 Au- gust- 12 September periods, and in the 6-30 November period she emerged more in the third (|uarter than in the other three quarters combined (13 of 23 times). Seasonal changes in proportions of emergences in each quarter of the night were also significant (4 season by 4 quarter contingency table, X" = 23.71, P = 0.005). Relative amounts of stationary and moving types of activity changed with seasonal pro- gression and maturity of the female. She tended to make short movements or no move- ment late in the summer (13 August- 12 Sep- tember)(Fig. 1). This phenomenon again ap- peared in late fall (6-30 November), but at that time of year all types of activity were infrequent. The shift in importance of move- ment versus stationary time is reflected by the seasonal change in frecjuency of each in the 21 0.5-hr time intervals from 1930 to 0600. Movement activity peaked dining the 16 Oc- tober-5 November period when fre(|uency of movement exceeded the trc(}uency of station- ary activity for 18 of 21 time intervals. The decrease in movements during the next pe- riod was dramatic; only 4 of 21 intervals had higher movement frecjuencies. The overall seasonal change in relative amounts of move- ment and stationary activity was highly signifi- cant (4 season bv 2 acti\ itN contingencv table, X' 37.68, P< 0.0001). ' Ferrets appeared to be relatively inactive during 6-30 November 1982. Few observa- tions of ferrets were made during spotlight surveys, and little ferret sign (diggings or 1986 BiGCINS KTAL.: RaDIO-TACX.ED FeHRETS 139 n 1300- 1559 1600 1859 2200- 0059 0100- 0359 0400- 0659 0700- 0959 1000- 1259 TIME (M.S.T.) Fig. 2. Total aboveground activity of a female (No. 536) and a male (No. 620) black-footed ferret, 30 October-13 November. tracks on snow) could be found. The radio- tagged female did not emerge during hours of darkness for 5 consecutive nights in mid- November. Comparison of the Male and Female Ferrets From 30 October to 13 November, daily activity patterns of the two ferrets were simi- lar (Fig. 2). Neither animal was active from 1300 to 1559, both animals had a small amount of activity near sunset followed by decreased activity from 1900 to 2159, and both animals reached peak activity from 0100 to 0359. Dur- ing 14,318 min of monitoring on each animal, the signals from the female and male were audible for 1,569 min (11%) and 2,125 min (15%), respectively. Average amounts of total time spent aboveground nightly were 2. 10 hr for the female (range 0-5.79 hr) and 2.95 hr for the male (range 0-4.84 hr). These figures and the patterns illustrated in Figure 2 sug- gested that the adult male was more active than the young female, but we could not de- tect a significant difference in average nightly activity (t = 1.044, P = 0.308). Durations of nightly activity of the female were mostly short; half were < 0.75 hours, with no activity on 3 of the 12 nights. There were 3 nights with > 5 hours of activity. In contrast, the male was never active for > 4.84 hours and was com- pletely inactive for only 1 night; he accumu- lated 3.24-4.84 hours of activity on 7 of the 12 nights monitored. Discussion Few general conclusions can be derived from the comparisons between these two ani- mals, because data came from a different year for each animal, sexes and ages were different, and 12 nights is a small sample. However, we can hypothesize that male ferrets are more active than females. This hypothesis is consis- tent with comparisons of spatial activity of these two animals; the area of activity of the male was more than twice as large as that of the female (Biggins etal. 1985). Males of other small mustelid species also use larger areas 140 Great Basin Naturalist Memoirs No. 8 than females, based on information about stoats {Mustela erminea) (Erlinge 1977, Simms 1979), feral domestic ferrets (M. fiiro) (Moors and Lavers 1981), and weasels (M. nivalis and M. ermine a )(Lock\e 1966). Our preliminary study has provided de- tailed information on only one animal during four months, and on a second animal for a much briefer period. The descriptive statis- tics were compiled to emphasize some behav- iors detected in this species. We do not know whether these examples typify ferret activity in general, although our data support certain observations of others (Hillman 1968, Clark et al. Descriptive ethology, 1986). Abundance and activity of prey, breeding activity, and weather may influence ferret activity. Richardson (personal communication) found a positive correlation between temperature and movements of snow-tracked ferrets and found increased movements during the breeding season. Seasonal changes observed in the radio- tagged female ferret imply that procedures used to locate ferrets (e.g., spotlighting) may not be generalized throughout the year. Our data suggest that the best time of night to conduct spotlight searches for ferrets from August through mid-October is from 0200 (MST) until dawn. This agrees with informa- tion collected by Clark et al. (Descriptive ethology, 1986) on activity of ferret litter groups in summer. Future analyses of more recent telemetric data may help identify causes of seasonal changes in activity. Acknowledgments The study could not have been conducted without cooperation and assistance from many sources. Instrumental in the success were the following radio-trackers from BIOTA Re- search and Consulting (B) and the U.S. Fish and Wildlife Service (F): T. Campbell (B), D. Casey (B), C. Halvorson (F), D. Hammer (F), L. Hanebury (F), J. Hasbrouck (B), D. Hig- gins (F), A. Jenkins (F), S. Karl (B), D. Lan- ning(F), L. Lee (B), S. Martin (F), B. Riddle (F), and S. Woodis (F). Thanks for additional assistance are due to J. Turnell (Pitchfork Ranch), J. Ross and A. Abbott (U.S. Forest Service), and T. Thorne (Wyoming Game and Fish Department). Literature Cited Biggins. D. E., M. Schroeder, S. Forrest, and L. Richardson. 1985. Movements and habitat rela- tionships of radio-tagged black-footed ferrets. Pages 11.1-11.17 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc., Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Pub!., Cheyenne. Ca.mpbell. T. M. Ill, D Biggins. S Forrest. andT. W. Clark 1985. Spotlighting as a method to locate and study black-footed ferrets. Pages 24. 1-24.7 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc. , Laramie, Wyoming, Sep- tember 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Clark, T.W.S.C. Forrest, L. Richardson, D. E Casey, AND T M Campbell III. 1986. Description and history of the Meeteetse black-footed ferret envi- ronment. Great Basin Nat. Mem. 8:72-84. Clark, T W , L. Richardson, S C. Forrest, D. E. Casey, andT M Campbell 111 1986. Descriptive ethol- ogy and activity patterns of black-footed ferrets. Great Basin Nat. Mem. 8:115-134. Collins, E. I., and R. W. Lichvar. 1986. Vegetation in- ventory of current and historic black-footed ferret habitat in Wyoming. Great Basin Nat. Mem. 8:85-93. Fagerstone, K a, D E Biggins, and T. M. Campbell 111. 1985. Marking and radio-tagging of black- footed ferrets (Mustela nigripes). Pages 10.1- 10.10 in S. Anderson and D. Inkley, eds., Black- footed Ferret Workshop Proc, Laramie, Wyo- ming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Erlinge, S. 1977. Spacing strategy in stoat, Mustela er- minea . Oikos 28:32-42. Hillman, C. N. 1968. Field observations of black-footed ferrets in South Dakota. Trans. N. Amer. Wild, and Natural Res. Conf 33:433-443. Lockie. J D. 1966. Territory in small carnivores. Symp. Zool. Soc. London 18:143-165. Moors, P. J., and R. B. Layers 1981. Movements and home range of ferrets (Mustela furo) at Pukepuke Lagoon, New Zealand. New Zealand J. Zool. 8:413-423. Schroeder, M. H. AND S J Marun 1982. Search for the black-footed ferret succeeds. W\()ming Wildlife 46(7):8-9. Sheets, R. G. 1972. A trap for capturing black-footed ferrets. American Midland Naturalist 88:461-462. SIMMS, D. A. 1979. Studies of an ermine population in southern Ontario. Canadian J. Zool. 57:824-832. TiioRNE, E T , M H Schroeder, S, C. Forrest, T. M. Campbell 111, L. Rktiardson, D Biggins, L. R. Hanebury, D Belitsky, and E S Williams. 1985. ('apture, immobilization, and care of black- footed ferrets for research. Pages 9.1-9.8 in S. Anderson and D. Inkley, eds., Black-footed Fer- ret Workshoii Proc., Laramie, Wyoming, Sep- tember 18-19, 1984. Wvoming Game and Fish i'ubl.,Chcv(MUie. FECAL BILE ACIDS OF BLACK-FOOTED FERRETS Mark K. Johnson', Tim W. Clark^ Max H. Schroeder', and Louise Richardson^ Abstract. — Fecal bile acid characteristics have been used to identify scats to species of origin. Fecal bile acids in scats from 20 known black-footed ferrets {Mustela nigripes), 7 other known small carnivores, and 72 of unknown origin were analyzed to determine if this procedure could be used as a tool to verify ferret presence in an area. Seventeen ferret scats were suitable for analysis and had a mean fecal bile acid index of 156 ± 9. This was significantly different from mean indices for the other carnivores; however, substantial overlap among confidence intervals occurred for badgers, kit foxes, and especially long-tailed weasels. We conclude this method is not useful for making positive identifications of individual ferret scats and suggest that we may be able to definitively identify individual scats with reasonable confidence by using gas-liquid chromatography. A major research goal of the Meeteetse, Wyoming, black-footed ferret (Mustela ni- gripes) (BFF) studies is development of sur- vey techniques (Clark 1984). From 1981 to 1984, 92 scats, 20 of known BFF origin and 72 of unknown origin but similar in size, shape, and color to known BFF scats, were collected (BFF scats pictured on p. 20 in Clark et al. Handbook of methods, 1984). Fecal bile acid analyses have been used to identify scats (Ma- jor et al. 1980, Johnson et al. 1984). Analysis may be performed by thin-layer (TLC) or gas- liquid (GLC) chromatography (Johnson et al. 1984). Although the latter method is more quantitative, it is also much more time con- suming and expensive than TLC and requires additional training. Costs for routine manage- ment applications would probably be pro- hibitive for most government fish and wildlife agencies, especially if analyses are needed for a large collection of scats. TLC can be per- formed in less time, and several samples can be analyzed at the same time. Initial equip- ment expense for TLC is about 20% of cost for GLC. The purpose of this study was to deter- mine if thin-layer chromatographic analyses of fecal bile acids could be used as a means to positively identify scats from BFFs and thereby provide a new tool to determine BFF presence in an area. Methods Scats from 20 BFFs were obtained from live-trapped specimens; they were collected along tracks of ferrets in snow (Clark et al. Handbook of methods, 1984; Clark et al. Sea- sonality of black-footed ferret diggings, 1984) or after field observers saw animals defecate. Another 72 scats each were collected from uncertain identity from the same area where field personnel collected the known BFF scats. To cover the range of size of the uniden- tified scats, 5 or 10 known scats each were collected from seven additional carnivore spe- cies that may frequent prairie dog (Cynomys sp.) colonies (Clark et al. 1982) 1979-1984: badgers (Taxidea taxus), long-tailed weasels {Mustela frenata), mink (M. vison), kit fox (Vulpes macrotis), striped skunk {Mephitis mephitis), red fox (Vulpes vulpes), and gray fox (Urocyon cinereoargenteus). All scats were analyzed according to the thin-layer chromatographic method (TLC) described by Major et al. (1980). Visualization of steroid bands on TLC plates was accom- phshed by spraying with 8-hydroxy- 1,3,6- pyrenetrisulfonic acid trisodium salt (5 mg in 100 ml of methanol). This reagent was used in lieu of that used by Major et al. (1980) because it does not destroy the steroids. After visual- ization, locations of all steroid bands were recorded relative to the solvent front (rf).Only bands that occurred between rf values of 15% and 75% of the solvent front were consid- ion. Lx>uisiana State University Agricultural Center. Baton Rouge. School of Forestry, Wildlife, and Fisheries, Louisiana Agricultural Experiment Station. I Louisiana 70803. ^Biota Research and Consulting, Inc.. Box 2705, Jackson, Wyoming, 8,3001, and Department of Biology. Idaho State University, Pocatello, Idaho, 83209. ^U.S. Fish and Wildlife Service, Denver Wildlife Research Center, 1300 Blue Spruce Drive, Fort Collins, Colorado 80524. 141 142 Great Basin Naturalist Memoirs No. 8 ered to be fecal bile acids. Cholic and lithocholic acids are present in most species scats and usually travel at least 15% and 75% of the solvent front, respectively, using this technique (Major et al. 1980). Bile acids can be eluted from this silica gel and used for friture GLC analyses. Because there is variation in fecal bile acid concentration, we categorized scats that had less than three distinct bile acid bands as unidentifiable. This decision was justified be- cause scats from no species previously de- scribed other than mountain lion (Felis con- color) have had fewer than three detectable fecal bile acids on TLC plates (Major et al. 1980, Johnson et al. 1981, Johnson and Aldred 1981, Johnson et al. 1984). Fresh scats from known specimens do not usually produce low quantities offecal bile acids. However, weath- ered scats may, because bile acids are highly soluble in water. An average fecal bile acid index was obtained for each species by sum- ming rf values for all bands in each scat and averaging among scats. Statistical analyses were performed by comparison of mean in- dices among species. Data herein are means and standard errors. Results Comparison of Known Scats The number of bile acid bands and index means varied among the eight carnivore spe- cies. Of 20 BFF scats, three contained fewer than three bile acid bands, seven had three bands, three had four bands, six had five bands, and one had six bands on TLC plates. The mean fecal bile acid index for BFF scats with three or more bands was 156 + 9 (Table 1). For the other seven carnivores, the number of TLC bands ranged from three to seven. Striped skunk, gray fox, and red fox scats never produced more than three bands; index means were 96 ± 1, 93 ± 3, 94 ± 2, respec- tively, and were significantly smaller than the BFF index (P < .05). The 80% confidence intervals for these species compared to those for BFFs suggested that their scats would probably not be confused with BFF .scats by TLC analysis (Table 1). The mean index for mink scats (78 ± 4), which had three or four bands, was also significantly smaller than the mean for BFFs (P < .05), and 80% confidence intervals did not overlap. Mean fecal bile acid indices for badgers (179 ± 22), long-tailed weasels (197 ± 16), and kit fox (243 ± 17) were significantly larger than the mean for BFF scats at the 0.05 level of probability for type I error. However, there was substantial overlap among confidence in- tervals between each of these species and BFF scats. Regarding this overlap, less than 1% of BFF scats having an index less than 153 would be confused with kit fox scats, and less than 5% would be confused with long-tailed weasel scats. For BFF scats with an index less than 163, less than 10% would be confused with long- tailed weasel scats. Variation in fecal bile acid indices from badger scats was so large that they could not be distinguished from BFF indices. Badger scats can often be differ- entiated by size from BFF scats; however, size overlap does sometimes occur, which is a problem for visual analysis. Ten (50%) of the BFF scats produced at least three bands on TLC plates and produced in- dices less than 163; nine (45%) of these were less than 153. Only four (20%) BFF scats pro- duced fecal bile acid indices that were less than 163 and within the 99% confidence inter- val for BFFs. Three of these scats produced indices that were less than 153. About 35% of the BFF indices were too large to be distinguished from indices from kit fox or long-tailed weasel scats using TLC anal- ysis. We estimate a 15% probability of identi- fying a BFF scat with 99% confidence that it is not from a kit fox. A 15% probability exists of identifying that a BFF scat is not from a long- tailed weasel with 95% confidence; or only a 20% probability of identification with 90% confidence. In addition to the problem of misclassifying BFF scats, long-tailed weasel scats can be misclassified as BFF scats. No kit fox scats produced indices within the 80% or 99% con- fidence intervals for BFF indices, so this would be an improbable source of error. How- ever, 40% of long-tailed weasel scats pro- duced indices that were within the 99% and 80% confidence intervals for BFF scats, thus creating a significant problem with this analy- sis. The other 60% (three of five) long-tailed 1986 Johnson et al.: Fecal Bile Acids 143 Table 1. Mean fecal bile acid indices and confidence intervals for scats from seven carnivore species with at least three detected steroid bands between rf 15-70 on 20-cm silica gel G TLC plates. N Mean ± SE Confidence intervals Species 99% 95% 90% 80% Black-footed ferret 17 156 ± 9 129-183 137-176 140-173 144-169 Badger 5 179 ± 22 69-274 109-233 123-219 131-211 Long-tailed weasel 5 197 ± 16 124-271 153-242 163-232 173-222 Kit fox 5 242 ± 17 166-321 196-290 207-280 218-269 Mink 10 78 ± 4 65- 91 69- 87 71- 85 72- 84 Striped skunk 10 96 ± 1 93- 99 94- 98 94- 98 95- 97 Gray fox 10 93 ± 3 83-103 86-100 88- 98 89- 97 Red fox 10 94 ± 2 87-101 89- 99 90- 98 91- 97 weasel scats produced indices larger than those within the 99% confidence interval for BFFs. Identifying Unknown Scats For the 72 unidentified scats from areas known to have BFFs, 14 produced indices less than 163, and 8 of these were within the 99% confidence intervals for BFFs and long-tailed weasels. Only 2 of the 8 scats produced in- dices below 153. We are confident that these scats were from BFFs because they were much smaller than most badger scats, were associated with apparent BFF sign, and had indices (both 138) outside the 99% confidence intervals for species other than long-tailed weasels. They were probably not from long- tailed weasels because the smallest weasel in- dex we obtained was 158. We recognize that this is the weakest point in our data because of the small number of known long-tailed weasel scats examined. There were 12 other scats that produced indices within the 99% confidence interval for BFFs, but these indices were higher than the mean and could not be distinguished from indices for kit fox, long-tailed weasel, or bad- ger. This analysis suggests that to be reason- ably certain that BFFs are present in any area, some of the scats must produce fecal bile acid indices between 129 and 153. Only 15% (3 of 20) of bona fide BFF scats produced indices in this range. Therefore, there is a relatively large probability of not detecting BFF pres- ence when few scats are examined. Although we estimated that about 36% (26/ 72) of the unidentified scats were likely BFF scats, only about 3% (two/72) of these scats were within the 95% confidence interval for BFFs. Assuming that no long-tailed weasel scats will produce a fecal bile acid index less than 153, then a rough estimate for the num- ber of the 72 scats that actually were BFF scats is 13 (2 -0.15). Discussion Although sample sizes were relatively small, this analysis suggests that the TLC method is not useful for making positive iden- tifications of individual BFF scats. Because of variation among fecal bile acid indices for BFFs and other carnivores, an individual scat could not be positively identified even though mean fecal bile acid indices were significantly different among the species. Differentiation of long-tailed weasel, small badger, and BFF scats is the most significant problem identified by our analysis. Because about 15% of known BFF scats seemed to produce index values distinctly dif- ferent from those produced by other carni- vores, it may be possible to establish a high probability of BFF presence using indices from a relatively large collection of scats. However, whether or not this can be done depends on obtaining a better description of variation in indices produced by scats of other species, particularly long-tailed weasels and to some extent badgers. More known scats must be analyzed to improve our data base. The differences found in fecal bile index means suggests that a more definitive type of analysis might provide a bona fide technique that can be used to identify individual scats with reasonable confidence. Recently John- son et al. (1984) demonstrated that gas-liquid chromatography (GLC) provided a more definitive distinction between fecal bile acids of bobcat (Felis rufus) and mountain lion than 144 Great Basin Naturalist Memoirs No. 8 TLC. We suggest that additional effort be ex- pended to obtain gas-liquid chromatographs as a method for identifying BFF scats. Although GLC costs are very high compared to TLC, new developments in equipment and methods will probably reduce costs substantially in the fu- ture. Acknowledgments T. P. O'Farrell and S. Minta provided scats from kit fox and badgers, respectively. We would like to thank R. E. Noble and P. J. Zwank for their critical review of this paper. Field col- lection of scats was supported by Wildlife Preservation Trust International, Inc., New York Zoological Society (Wildlife Conservation International), World Wildlife Fund— U.S., and the National Geographic Society. Literature Cited Clark, T. W. 1984. Strategies in endangered species con- servation: a research view of the ongoing black- footed ferret conservation program. Pages 145-154 in Symposium on Issues in Technology and Management of Impacted Western Wildlife, Steamboat Springs, Colorado, November 1982. Clark, T W , T M Campbell III. D Socha, and D. Casey 1982. Prairie dog colony attributes and as- sociated vertebrate species. Great Basin Nat. 42:572-582. Clark, T W ,T M Campbell III, M H Schroeder, and L Richardson 1984. Handbook of methods for locating black-footed ferrets. Wyoming Tech. Bull. No. 1.55 pp. Clark, T. W., L. Richardson, D Casey, T M Campbell III, and S C Forrest 1984. Seasonality of black footed ferret diggings and prairie dog plugging. J Wildl. Manage. 48:1441-1444. Johnson, M K., and D R Aldred 1981, Feces, bile acids and furbearers. Pages 1143-1150 in Proc Worldwide Furbearer Conference, Appalachian Environmental Lab,, Frostburg, Maryland, Vol II, Johnson, M K , D R Aldred, E W Clinite, and M J Kutilek 1981. Biochemical identification of bob- cat scats. Pages 92-96 in Proc. Bobcat Res. Conf., Nat. Wildl. Fed. Sci. and Tech, Ser, 6. Johnson. M K , R C Belden, and D R Aldred. 1984, Differentiating mountain lion and bobcat scats. J, Wildl. Manage. 48:239-244. Major, M , M K Johnson, W S Davis, and T F. Kel- logg 1980. Identifying scats by recovery of bile acids. J. Wildl. Manage. 44:290-293. ESTIMATING GENETIC VARIATION IN THE BLACK-FOOTED FERRET— A FIRST ATTEMPT C. William Kilpatrick', Steven C. Forrest^ and Tim W. Clark^ Abstract. — No genetic variation was observed for three proteins examined from samples of saliva from 22 black-footed ferrets (Mustela nigripes). The comparable data concerning levels of genetic variation in other taxa at these loci are too inconclusive to provide a meaningful interpretation of the observed absence of genetic variation. The absence of genetic variation observed in the black-footed ferret population is compatible with the reported levels of genetic variation in terrestrial carnivores and populations that have undergone bottlenecks. Suggestions for additional studies using different approaches both to increase the number of loci that are used to determine the level of genetic variability in the black-footed ferret and to provide a more meaningful comparative data base are provided. The importance of genetics in the manage- ment and conservation of endangered species has been recently discussed (Soule and Wilcox 1980, Frankel and Soule 1981, Schonewald-Cox et al. 1983). Although the primary objective of conservation and man- agement is the continued reproduction of the species, maintenance of genetic variability has also been identified as a high priority (Benirschke 1977, Chesser et al. 1980). With- out maintenance of genetic variability, the species may have an increased probability of extinction in future variable environments (Wright 1951). The objective of this study was to obtain an estimate of the level of genetic variability present in a population of the endangered black-footed ferret (BFF) near Meeteetse, Wyoming. Salivary samples were easily taken and did not aflPect the survivorship of individu- als sampled. Comparison of the genetic varia- tion observed in the Meeteetse population with reported values in the literature on other species were made to determine the potential effect of the recent history of population size (bottlenecks) and isolation. Methods Salivary samples were collected from immo- bilized animals in the field during 1982 and 1983 by swabbing the oral and buccal cavities with a cotton swab or a small piece of gauze. Samples were frozen and shipped on dry ice to the University of Vermont for analysis of elec- trophoretic variation of salivary proteins. Salivary proteins were washed from the cot- ton or gauze with 1-2 ml of distilled water. The residual protein solution was removed from the cotton or gauze by centrifugation at 600-800 rpm. A corner of the cotton or gauze was held outside a 15-ml screw cap centrifuge tube before placing the cap on the tube to separate the cotton or gauze from the liquid during centrifugation. Samples were frozen at -75 C until analysis. Prior to electrophoresis, salivary samples were concentrated by the use of acrylamide sticks (Curtain 1964, Balakrishnan and Ashton 1974). Salivary amylase (AMY) was examined by the methods of Aquadro and Patton (1980), except the sample was increased to 25 |xl per slot and the gel with the starch overlay was incubated overnight at 37 C before staining. Salivary esterase (EST-S) was examined by the methods of Tan (1976) except n-propanol was deleted from the stain. The method of Tan and Teng (1979) was usable for superoxide dismutase (SOD). Better results were ob- tained with a 10% acrylamide gel using the 8.9 tris-borate-EDTA buffer system of Coyne and Felton (1977) and with a stain of 100 ml of the SOD incubation bufiPer (Tan and Teng 1979), 30 mg MTT, 30 mg nitro blue tetrazolium, and 2 mg phenazine methosulfate. In addi- tion, the methods of Tan and Ashton (1976a) Department of Zoology, University of Vermont, Burlington, Vermont 05405. ^Biota Research and Consulting, Inc., Box 2705, Jackson, Wyoming 83001. ^Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209. 145 146 Great Basin Naturalist Memoirs No. 8 Table 1. Genetic variability (heterozygosity) among terrestrial mammals in which two or more of the proteins examined from the saliva of black-footed ferrets have been reported. References in parentheses are cited below; data for Herpestes auropunctatus ire unpublished (D. B. Hoagland, personal communication). Taxa AMY EST SOD H=' H** Homo sapiens Peromyscus maniculatus Peromyscus leucopus Mus musculus Herpestes auropunctatus Mustela nigripes h 0.138(1) 0.495(4) 0.009(5) 0.214 0.067(12) 0.315(2) 0.000(6,7) 0.158 0.122(6,7,13,14,15) 0. 185(2) 0.241(8) 0.213 0.081(7,8,16,17) 0.048(3) 0.107(9,10) 0.052 0.098(9,10) 0.000 0,000 0.000 0.000 0.037 0.000 0.000 0.000 0.020(11) 0.000 ? References: (1) Merritt et al. 1973. (2) Aquadro and Palton 1980, (3) Nielsen and Sick 197.5, (4) Tan 1976, (5) Beckman and Pakarinen 1973, (6) Avise et al. 1979. (7) Loudenslager 1978, (8) Price and Kennedy 1980. (9) Selander and Yang 1969, ( 10) Selander et al. 1969. ( 1 1 ) Selander 1976, (12) Harris and Hopkinson 1972, (13) Dubach 1975, (14) Aquadro and Kilpatrick 1981, (15) Smith 1981, (16) Selander et al. 1975, (17) Browne 1977. h_- mean heterozygosity of locus. H* - mean heterozygosity of species based on loci examined in the black-footed ferret. H** - mean heterozygosity of species. for salivary acid phosphatase, Tan and Ashton (1976b) for hexose-6-phosphate dehydroge- nase, and Tan and Teng (1979) for lactate de- hydrogenase, and saliva oxidase were at- tempted. No reactions were observed in the BFF salivary samples with these methods. Results No electrophoretic variation was observed among salivary samples from 22 BFFs at the loci for salivary amylase, salivary esterase, or superoxide dismutase. Based on this very small sample of loci (n = 3), the mean propor- tion of loci polymorphic (P) was 0.000 and the mean heterozygosity (H) was 0.000. Discussion Before the absence of genetic variability ob- served in the proteins from saliva of the BFF may be interj^reted, some comparisons are needed. Most estimates of genetic variation are based on proteins from blood, liver, kid- ney, and heart or other muscle; the levels of genetic variability present in loci for salivary proteins are not well known. Saliva has been examined in very few taxa, and, in most cases, only salivary amylase has been analyzed. Su- peroxide dismutase, which was examined from saliva from the BFF, is also expressed in the more conventional tissue sources of proteins for electrophoretic analysis and is the only locus of the three analyzed that has been examined in a large number of taxa. The extent of genetic variation at the three loci (AMY-1, EST-S, SOD-1) examined in the BFF was estimated in other taxa by the meth- ods of Lewontin and Hubby (1966) and Nei (1978). Few taxa could be found in the litera- ture for which two or more of the loci had been examined for electrophoretic variation (Table 1). Only the locus for superoxide dismutase has been examined in a sufficient number of taxa of mammals to yield a reasonable esti- mate of the average amount of genetic varia- tion present. Selander (1976) reported that SOD demonstrated a low mean heterozygos- ity (H = 0.020) among 26 species of rodents. Although the loci for nonspecific esterases are generally considered highly variable, sali- vary esterase variation (Table 1) has only been reported in humans (Tan 1976). This locus apparently has not been examined in other taxa, although it is highly variable in humans. Salivary amylase has typically not been in- cluded among the proteins examined in sur- veys of genetic variation in mammals. The taxa of mammals for which estimates of het- erozygosity are available or from which esti- mates can be calculated (Merritt et al. 1973, Nielsen and Sick 1975, Aquadro and Patton 1980) appear to represent tiixa in which this k)cus is highly variable (Table 1). Considerable variation in levels of genetic heterozygosity at the three loci was observed, ranging from high genetic heterozygosity in Peromyscus to no heterozygosity in Herpestes (Table 1). The estimates of average het- erozvgositv calculated from three loci (AMY- 1, EST-S, SOD-l) (Table 1) are greater than the reported heterozygosity calculated from a greater number of loci in man and Peromyscus but lower in Mus and Herpestes (Table 1). 1986 KiLPATRiCKETAL.: Genetic Variation 147 Table 2. Genetic variation among terrestrial carni Sample size Mean proporl tion of loci Taxa Individuals Loci Polymorphic per population Heterozygous per individual Reference Canidae Canis lupus 12 53 0.113 0.030 Fisher etal. 1976 Canis latrans 6 53 0.132 0.050 Fisher et al. 1976 Canis familiaris dingo 6 53 0.057 0.006 Fisher et al. 1976 Vulpes viilpes 282 21 0.000 0.000 Simonsen 1982 Ursidae Ursus americanus 56 6 0.000 0.000 Manlove et al. 1980 35 14 0.077 0.013 Manlove et al. 1980 64 15 0.133 0.015 Manlove et al. 1980 52 17 0.176 0.031 Manlove etal. 1980 52* 33 0.121 0.016 Manlove et al. 1980 X = 0.097 X - 0.015 Manlove etal. 1980 Ursus maritimus 52 13 0.000 0.000 Allendorfetal. 1979 Procyonidae Procyon lotor 451 24 0.133 0.021 Dew and Kennedy 1980 Mustelidae Mustela erminea 39 21 0.000 0.000 Simonsen 1982 Mustela nivalis 13 21 0.000 0.000 Simonsen 1982 Mustela putorius 24 21 0.000 0.000 Simonsen 1982 Martes martes 2 21 0.000 0.000 Simonsen 1982 Maries foina 121 21 0.000 0.000 Simonsen 1982 Meles meles 5 21 0.000 0.000 Simonsen 1982 Felidae Felis catus 56 55 0.220 0.069 O'Brien 1980 Aciononijx jubatus 50 47 0.000 0.000 O'Brien etal. 1983 Herpestidae Herpestes auropunctatus 45 29 0.241 X = 0.062 0.037 X = 0.014 unpublished data *Same population as above, including analysis of 16 additional loci. If the data for humans and Peromijscus are typical, these three loci are highly variable and tend to give a higher estimation of genetic variation than estimates based on a larger number of loci. This would suggest that a great deal of genetic variation has been lost from the Meeteetse population of BFFs. However, the estimates of genetic variation in these taxa are based, for the most part, on reports of genetic variation at a single locus and not on surveys of a number of loci. This would appear to result in a biased data set, since loci observed to be monomorphic (in- variable) are not reported unless they are part of a survey of loci. If the data for the small Indian mongoose, Herpestes auropunctatus, (Table 1) are typi- cal (or typical for carnivores), these loci demonstrate httle or no genetic variation. This would suggest that these loci provide little information concerning the total levels of genetic variation present in the Meeteetse BFF population. The comparative data avail able concerning the genetic variation at these three loci (Table 1) are inconclusive for providing a meaningful interpretation of the observed absence of genetic variation at these loci in the BFF. Although it is important to continue at- tempts to determine the existence of genetic variability that could be managed in the Mee- teetse BFF population, the absence of genetic variation observed thus far may be the result of a recent bottleneck or may be typical for carnivores. Some mammals that have passed through severe bottlenecks demonstrate an absence of or a very low level of heterozygos- ity (Bonnell and Selander 1974, Ryman et al. 1977, O'Brien et al. 1983). However, the North American bison {Bison bison), which has also gone through a bottleneck, has a mean heterozygosity of 0.023, and a small Indian mongoose population, which was derived from a few individuals introduced to St. Croix in 1884, presently has a level of heterozygosity of 0.037. The effect of the bot- 148 Great Basin Naturalist Memoirs No. 8 tleneck on the level of genetic variability is dependent upon the rate at which the popula- tion recovers from the reduced population size (Smith 1981) and not on the bottleneck alone. Pettus (1985) suggested that carnivores and perhaps other species of large mammals are employing the Mullerian strategy and would be expected to exhibit little genetic variation. Carnivores appear to have somewhat lower levels of genetic variation (Table 2), with a mean heterozygosity of 0.014 for 16 species as compared to a mean heterozygosity for 46 species of mammals of 0.036 (Nevo 1978). No genetic variation has been observed in any of the six taxa of the family Mustelidae (Table 2) that have been examined. Unfortunately, these have been examined by only one labora- tory (Simonsen 1982), and the sample sizes of some taxa were very small. Although the mean heterozygosity ob- served in carnivores is below the mean value of other mammalian taxa, several species demonstrate levels of heterozygosity typical for mammals (Table 2). Those taxa that demonstrated the highest levels of genetic variation among carnivores (Table 2), Felis catus and Canis latrans, are those with esti- mates based on the largest number of loci. The effect of examining a small number of loci is clearly seen in the estimates of genetic vari- ation in the American black bear {Ursus amer- icanus), as pointed out by Manlove et al. (1980). The level of genetic variability in dif- ferent populations increases with the number of loci examined (Table 2). Future work should include continued re- search to provide an estimate of genetic varia- tion in the Meeteetse population of BFF based on a larger number of loci. This re- search could include an examination of addi- tional loci from nontraditional sources such as saliva (Tan and Teng 1979), urine (Hayakawa et al. 1983), and feces (Scribner and Warren 1984). Examination of blood samples (hemolysate and serum), however, would al- low detection of genetic variation at 30 to 40 loci. By including loci for proteins from nontradi- tional sources, other surveys of genetic varia- tion in mammal taxa could provide a better understanding of levels of genetic variation present at these loci. Other surveys of genetic variation in carnivores, especially within the mustelids, including loci for proteins from tra- ditional and nontraditional sources, would provide a better data base from which to de- termine what portion of the total genetic vari- ability could be expected to be identified from salivary samples. Acknowledgments This work was supported in part by Institu- tional Research Grants PH 07125/18 and PH 07125/30 from the University of Vermont. Special thanks are extended to Robin Kleiman for her help in the development of techniques for the examination of salivary proteins and to Tom Thorne, Wyoming Game and Fish De- partment, for assistance in collecting samples. D. B. Hoagland of the University of Vermont provided unpublished data on genetic varia- tion of the small Indian mongoose on St. Croix. For critical reading of the manuscript, we thank Michael L. Kennedy and Earl G. Zimmerman. Field work was supported by the New York Zoological Society — Wildlife Conservation International and the Wildlife Preservation Trust International, Inc. Literature Cited Allendorf, F W , F B Christiansen, T Dobson. W. F. Eanes. and O Frydenberg 1979. Elec- trophoretic variation in large mammals. I. The polar bear, Thalarctos rnaritirnus . Heriditus 91: 19-22. Aquadro, C F , AND C W Kilpatrick 1981. Morpho- logical and biochemical variation and differentia- tion in insular deer mice, Peromyscus manicula- tus. Pages 214-230, in M. H. Smith and J. Joule, eds., Mammalian Population Genetics. Univer- sity of Georgia Press, Athens. 380 pp. Aquadro. C F , and J C Patton 1980. Salivary amylase variation in Peromyscus: use in species identifica- tion. J. Mammal. 61: 703-707. AvisE, J C , M H Smith, and R K Selander 1979. Biochemical polymorphism and systematics in the genus Peromyscus . VTI. Geographic differentia- tion in members of the truei and municulatus spe- cies group. J. Mammal. 60:177-192 Bai.akrishnan, C R , and G C Ashton 1974. Polymor- phisms of human salivarv proteins. Amer. J. Hum. Genet. 26: 145-153. Beckman, G.. and A. Pakarinen. 1973. Superoxide dis- mutase: a population study. Hum. Hered. 23:34&-351. Benirschke, K. 1977. Genetic management. Int. Zoo Yearbook 17: 50-60. 1986 KiLPATRICK ET AL. : GENETIC VARIATION 149 BoNNELL. M L , AND R K. Selander. 1974. Elephant seals: genetic variation and near extinction. Science 184:908-909. Browne, R. A. 1977. Genetic variation in insular and main- land populations of Peromyscus leucopus. Amer. Mid. Nat. 97: 1-9. Chesser. R K., M H Smith, and 1 L Brisbin, Jr. 1980. Man- agement and maintenance of genetic variability in endangered species. Int. Zoo Yearbook 20: 146-1.54. Coyne, J A , and A A Felton. 1977. Genie heterogene- ity at two alcohol dehydrogenase loci in Drosophila pseudoobscura and Drosophila per- similis . Genetics 87:285-304. Curtain, C. 1964. Concentrating protein solutions. Na- ture 203; 1380. Dew, R. D., and M L Kennedy 1980. Genie variation in raccoons, Procyon lotor . J. Mammal. 61:697-702. Dubach, J. M. 1975. Genetic variation in altitudinally allopatric populations of the deer mouse (Per- omyscus maniculatus). Unpublished thesis. Uni- versity of Colorado, Denver. 55 pp. Fisher, R. A., W. Putt, and E Hackel 1976, An investi- gation of the products of 53 gene loci in three species of wild Canidae: Canis lupus, Canis la- trans, and Canis familiaris . Biochem. Genet. 14:963-974. Frankel, O. H., and M E Soule 1981. Conservation and evolution. Cambridge University Press, Cam- bridge. 327 pp. Harris, H , and D A Hopkinson 1972. Average het- erozygosity in man. J. Human Genet. 36:9-20. Hayakawa, J., H Naido, andT. Koizumi 1983. Compo- nents of major urinary proteins (MUP s) in the mouse: sex, stain, and subspecies differences. J. Hered. 74:453-456. Lewontin, R. C, and J. L. Hubby 1966. A molecular approach to the study of genie heterozygosity in natural populations. II. Amounts of variation and degree of heterozygosity in natural populations of Drosophila pseudoobscura . Genetics 54:595-609. LoudenslagER, E. J 1978. Variation in the genetic struc- ture of Peromyscus populations. I. Genetic het- erozygosity — its relationship to adaptive diver- gence. Biochem. Genet. 16: 1165-1179. Manlove, M. N., R Baccus, M R Pelton, M H Smith, AND D Gruber 1980. Biochemical variation in the black bear. Pages 37-^1 in G. J. Martinka and K. L. McArthur, eds.. Bears — their biology and management. Bear Biol. Assoc. Conf 3:1-375. Merritt, A D , M. L. Rivas, D Bixler, and R Newell 1973. Salivary and pancreatic amylase: elec- trophoretic characterization and genetic studies. Amer. J. Hum. Genet. 25:510-522. Nei, M. 1978, Estimation of average heterozygosity and genetic distance from a small number of individu- als. Genetics 89:583-590. Nevo, E. 1978. Genetic variation in natural populations: pat- terns and theory. Theoretical Pop. Biol. 13: 121-177. Nielsen, J. T , and K Sick 1975. Genetic polymorphism of amylase isoenzymes in feral populations of the house mouse. Hereditas 79:279-286. OBrien. S J 1980. The extent and character of biochemical genetic variation in the domestic cat. J. Hered. 71:2-8. OBrien, S J , D E VVildt, D Goldman, C R Merril, AND M Bush 1983. The cheetah is depauperate in genetic variation. Science 221:459-462. PE'nus, D. 1985. Genetics of small populations. Pages 22.1-22.11 in S. Anderson and D. Inkley, eds., Black-footed Ferret Workshop Proc., Laramie, Wyoming, September 18-19, 1984. Wyoming Game and Fish Publ., Cheyenne. Price. P. K., and M L Kennedy 1980. Genie relation- ships in the white-footed mouse, Peromyscus leu- copus, and the cotton mouse, Peromyscus gossypi- nus. Amer. Midi. Nat. 103:73-82. Ryman, N. G , G Beckman, G Brun-Petersen, and C. Reuterwall. 1977. Variability of red cell en- zymes and genetic implications of management policies in Scandinavian moose (Alces alces). Hereditas 85: 157-162. Schonewald-Cox, C M , S M, Chambers, B Mac- Bryde, and W. L Thomas 1983, Genetics and conservation, Benhamin/Gummins, Menlo Park, California. Scribner, K T , and R J Warren. 1984. Electrophoretic identification of white-tailed and mule deer feces : a preliminary assessment. J. Wildl. Manage. 48:656-658. Selander, R. K 1976, Genie variation in natural popula- tions. Pages 21-45 in F, Ayala, ed.. Molecular evolution. Sinauer Assoc. Inc., Sunderland, Mas- sachusetts. 275 pp. Selander. R K., W G Hunt, and S Y Yang. 1969. Protein polymorphism and genetic heterozygosity in two European subspecies of the house mouse. Evolution 23:379-390. Selander, R. K., D W Kaufman, R J Baker, and S L Williams 1975, Genie and chromosomal differ- entiation in pocket gophers of the Geomys bursar- ius group. Evolution 28:557-564. Selander, R. K., and S Y Yang 1969. Protein polymor- phism and genetic heterozygosity in a wild popula- tion of the house mouse {Mus musculus). Genetics 63:653-667. SiMONSEN, V 1982. Electrophoretic variation in large mammals. II. The red fox Vulpes vulpes, the stoat Mustela erminea, the weasel Mustela putorius, the pine marten Martes martes , the beech marten Martesfoina , and the badger Meles meles . Hered- itas 96:299-305. Smith, M F. 1981. Relationships between genetic vari- ability and niche dimensions among coexisting species of Peromi/scus. J. Mammal. 62:273-285. SouLE, M. E., AND B A Wilcox 1980. Conservation biol- ogy: an evolutionary-ecological perspective. Sin- auer Assoc, Inc., Sunderland, Mass. Tan, S. G. 1976. Human saliva esterases: genetic studies. Hum. Hered. 26:207-216. Tan, S G , and G C Ashton 1976a, Saliva acid phos- phatases: genetic studies. Hum, Hered, 26:81-89. 1976b. An autosomal glucose-6-phosphate dehy- drogenase (hexose-6-phosphate dehydrogenase) polymorphism in human saliva. Hum. Hered. 26:113-123. TanS G.,andY S.Teng 1979. Human saliva as a source of biochemical genetic markers. I. Techniques. Hum. Hered. 29: 69-76. Wright, S. 1951, The genetical structure of populations. Ann. Eugen. 15:323-354. DETERMINING MINIMUM POPULATION SIZE FOR RECOVERY OF THE BLACK-FOOTED FERRET Craifj; R. Groves' ~ and Tim W. Clark' Abstract. — A minimum viable population (MVP) size is estimated for the critically endangered black-footed ferret by examining five basic methods: experiments, biogeographic jiatterns, theoretical models, simulation models, and genetic considerations. Each method is evaluated for its appiical)ilit\ to the ferret and endangered species in general with two criteria in mind: (1) potential research impacts to target species and (2) the value of scientific accuracy and precision in relation to short-term conservation needs. For the black-footed ferret, the genetic method proved to be the most useful, resulting in an MVP estimate of about 200 ferrets for maintenance of short-term fitness. For most endangered species, a combination of the simulation and genetic methods will probably yield the best estimate of MVP size. MVP estimates have direct implications for future research, management, and recovery efforts of endangered species. Following the 1981 discovery of the critically endangered black-footed ferret (Miistela ni- gripes) in northwestern Wyoming, the only pop- ulation of the species known, two primary goals of a ferret study were established: (1) conserva- tion of the population and (2) recovery of the species (Clark 1984a). Estimation of population parameters such as distribution, birth rate, death rate, and immigration/emigration was a major objective. In addition to providing valu- able life history data for the ferret, estimating these parameters would enable us to address a key question: What is the minimum ferret popu- lation size necessary for the species to persist (i.e., avoid extinction)? Because of the apparent small size of the Wyoming population in 1981, the answer was paramount for successful conser- vation and, ultimately, species recovery. The minimum population size concept for spe- cies conservation was recently introduced by Shaffer (1981), with potentially significant impli- cations for, endangered species programs. Al- though many endangered species studies have been conducted, few have tried to estimate min- imum population sizes (Shaffer 1978, 1981, La- cava and Hughes 1984, Salwasser et al. 1984). Although US DA Forest Service regulations re- quire that minimum viable populations of en- dangered species be maintained on Forest Ser- vice lands, these population numbers are not estimated but are instead routinely established as those levels specified in recovery plans (Lacava and Hughes 1984). In this paper, we estimate minimum ferret population numbers and, as a corollary, minimum area require- ments. Our purpose is to establish guidelines for future ferret research and management and to provide impetus, direction, and encouragement to other endangered species programs to use the concept of minimum population size. Concept OF Minimum Viable Population Size The notion of a required minimum population size for species conservation was first embodied by Shaffer (1981) in his concept of minimum viable population (MVP) size. He defined the MVP for a species as the smallest isolated popu- lation having a 99% chance of remaining extant for 1000 years despite foreseeable effects of four types of stochastic events: (1) demographic stochasticity, or the chance events in the sur- vival and reproductive fitness of a finite number of individuals; (2) environmental stochasticity, or perturbations due to habitat parameters, competitors, predators, and disease; (3) natural catastrophes, such as randomly occurring floods, fires, droughts, etc.; and (4) genetic perturba- tions residting from changes in gene frequency. Shaffer (1981) stressed the tentative nature ofhis definition and emphasized the importance of defining MVP with explicit but flc^xible criteria such as time frame and survival probability. For Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209. Present address: The Nature Conservancy, 4696 Overland Road, Suite 51H. Boise, Idaho 8.370.5. 150 1986 Groves, Clark: Minimum Population 151 example, the survival probability could be set at 95% or auy other level, aud the time frame could be shortened or lengthened as appropriate for planning needs. Shaffer (1981) proposed five basic methods for determining MVP size: experiments, biogeo- graphic patterns, theoretical models, simulation models, and genetic considerations. In the next section, we elaborate on the application and util- ity of each method for estimating MVP size for ferrets in particular and endangered species in general. The usefulness of each method is prag- matically evaluated relative to: (1) acceptable levels of potential research impacts on target species and (2) the real value of scientific accu- racy and precision in relation to short-term con- servation needs. In endangered species recov- ery work, these two concerns must be addressed. The MVP concept is a rapidly evolving one. In late 1984, the Forest Service and other agencies sponsored a workshop to provide a state-of-the- art review of the MVP issue (M. L. Shaffer, personal communication). A major purpose of the workshop was to consider revisions of Forest Service procedures for estimating MVP size. The primary focus of the workshop was a discus- sion of the relative importance of demographic, environmental, and genetic variability in deter- mining minimum population numbers as well as the development of new simulation and analyti- cal models for estimating MVP size. Proceedings of this workshop, which will be published in 1986, may provide additional methods not dis- cussed in this paper and/or may suggest revi- sions of methods outlined and discussed below for estimating MVP sizes of endangered species. Determining MVP Size Experiments This method estabhshes isolated populations of species and monitors their population dynam- ics through time. If censusing can be accom- plished without stressful manipulation of indi- viduals (as can occur in a capture/recapture program), this approach requires only low levels of research activities. The only necessary data are population numbers and the area inhabited over a specific time period (i.e., time frame in MVP definition). Determining MVP via experimental methods brings up three concerns. Although methodo- logically simple, this approach is not plausible for many species, particularly endangered ones. First, for many endangered species, there exists no possibility of studying several isolated popu- lations. In the case of the ferret, there is only one known population, and consequently there is no way to measure variability in persistence of dif- ferent-sized, isolated populations. Second, time is not an unlimited resource in most endangered species studies. The years necessary to monitor persistence of populations for MVP estimates are simply not available for most endangered species. Lastly, many wildlife species cannot be censused sufficiently without some type of mark/ recapture program. These techniques can have negative research impacts (e.g., trap mortality) that some endangered species projects may not be willing to risk, particularly in the early stages. Because endangered species programs in gen- eral possess high levels of uncertainty, there may be hesitancy to increase uncertainty to even greater levels (Clark 1984b). Such has been the case in the Wyoming ferret project, where one research team chose initially not to employ mark/recapture and radio telemetry procedures until other methods proved the population large enough to withstand some stress (Clark 1981). The experimental method could prove useful when extensive population data are available prior to a species becoming endangered. For example, mountain goat (Oreamnos ameri- canus) and bighorn sheep {Ovis canadensis) populations have been reintroduced to several sites in the western United States in recent years (Rideout 1978, Wishart 1978). If wildlife re- searchers monitor the dynamics of these intro- duced populations, the data should eventually allow them to estimate MVP sizes for these spe- cies. Such information would be useful not only for future translocations of these species' popula- tions but also for management if, for example, an isolated population showed a decline in num- bers. Biogeographic Patterns By studying the distribution patterns of spe- cies occupying patchy or insular habitats, an esti- mate of MVP size and minimum area require- ments can be obtained. However, populations examined should be in equilibrium and their approximate period of isolation known (Shaffer 1981). Under these conditions, researchers can estimate the smallest area inhabited by a species 152 Great Basin Naturalist Memoirs No. 8 Fig. 1. Distribution of 37 white-tailed prairie dog colonies on the Wyoming black-footed ferret study area, 1981-1984. and the percentage of patches of a particular size that a species occupies. The smallest population that has persisted over an extensive time period can then provide a first estimate of MVP size. Because the ferret occupies a patchy habitat (i.e., prairie dog colonies with relatively discrete boundaries. Fig. 1), this approach can be ap- phed, at least in part, to the Wyoming popula- tion. Old time trappers have reported catching ferrets in the 1920s and 1930s on the Wyoming study area, thus indicating that ferrets may have occupied the site for at least 50 years. Extinction and recolonization, although unlikely, could have occurred during this period. Population censuses over the last four years (Clark 1986) suggest that the Wyoming ferret population is not in equilibrium but is increasing. As a result, more sophisticated census techniques (e.g., mark/recapture) are being used, leading to more precise and accurate population data. In the fu- ture, researchers should thus be able to deter- mine with some degree of confidence whether the Wyoming population is reaching an equi- librium density. If the Wyoming ferret population stabilizes in numbers, then data from this population will provide the first field assessment of MVP size and minimum area requirements for the species. However, due to the time required to obtain these data, the biogeographical method will not likely be useful for short-term management de- cisions concerning MVP size. Nevertheless, useful information on minimum area require- ments can be gleaned from the biogeographical approach. In the short run, researchers will need to de- cide what size of an area is necessary for ferret reintroduction or translocation for recovery planning (Forrest et al. 1985; Houston et al. 1986). A recommendation for an area larger than the Wyoming site is warranted for two reasons. First, two habitat patches (e.g., Wyoming and South Dakota) of equal size may not be the same in habitat quality. For example, Wyoming fer- rets occur on white-tailed prairie dog {Cynomys leucurus) colonies, whereas a South Dakota population of ferrets occupied black-tailed prairie dog (C. ludovicianus) colonies (Hillman and Clark 1980). White-tails usually form small, sparsely populated colonies; black-tails form large, densely populated colonies (Hoogland 1981). Consequently, a habitat patch that is large enough in one part of the range of the species might be insufficient in another due to differ- 1986 Groves, Clark: Minimum Population 153 ences in habitat quality. Second, nothing is known about the frequency with which ferret populations go extinct on habitat patches of vari- ous sizes. Therefore, it would be inadvisable to rely on the size of the Wyoming study area as an exemplary model of area requirements for black- footed ferrets. Although the biogeographical approach can- not be used for endangered species with con- tiguous distributions, it is applicable for species with insular or patchy distributions. Its greatest use should be for endangered species that oc- cupy patches of different size and similar quality and for which data on population numbers and isolation periods are available. For example, the Shoshone sculpin {Cottus greenei) is restricted to different-sized spring systems along a 45-km stretch of the Snake River in south central Idaho (Wallace et al. 1984). Relative and absolute den- sities of the sculpin in many of these spring sys- tems were determined during a status survey in 1980-1981 (J. Griffith, personal communica- tion). By monitoring the persistence of these spring populations of this short-lived species over the next several years, researchers could estimate MVP size and minimum area require- ments for this species. Theoretical Models Although there are several theoretical models that predict extinction probabilities and times for a population, most are too complex for the simple data bases of endangered species. The theory of island biogeography (MacArthur and Wilson 1967), however, has received consider- able attention in the conservation field. Though originally developed in relation to true oceanic islands, this theory has been widely applied to continental habitat "islands" or patches (e.g.. Brown 1971). The theory's greatest potential ap- plication has been in the design of nature pre- serves for maintenance of species diversity (for review, see Margules et al. 1982). Much less attention has been devoted to those aspects of the theory that predict the distribution of a sin- gle, insular species (Smith 1974, Fritz 1979). Because the Wyoming ferret population is lo- cated on a large habitat "island" (i.e. , prairie dog colony complex in Fig. 1), we employed the Ty model of island biogeography to predict proba- bilities of successful colonization and times to extinction for the ferret. Regarding the Wyo- ming prairie dog colony complex as a single is- land is supported by three lines of evidence: (1) several prairie dog colonies at the Wyoming site contained at one time or another only one adult ferret, suggesting that interbreeding of ferrets among prairie dog colonies is occurring, (2) movement data reveal that ferrets range an aver- age of 2.5 km between colonies, with the maxi- mum intercolony movement being 5.7 km (For- rest et al. 1985), and (3) aerial and ground surveys indicate that potential ferret habitat de- clines significantly beyond the boundaries of this single colony complex, strongly suggesting that these ferrets represent a distinct, isolated popu- lation. Data required in the Ti^ model are carrying capacity (K), per capita birth rates (\ ), and per capita death rates (|x ). Direct observations of ferret litters in Wyoming indicated an average htter size of 3.4, or 1.7 female young per adult female (Forrest et al., in manuscript). We as- sumed that the sex ratio approximated unity, natality did not vary with age, and all females bred. No data on juvenile mortality were avail- able for ferrets; data fi-om other mustelids indi- cated a medium (60%) to high (80%) rate of juve- nile mortality (King 1980, M. nivalis ; King 1983, M. erminea; K. C. Walton personal communica- tion, M. putorius). We used a juvenile mortality rate of 0.7, resulting in a X of 0.5. As with juve- nile mortality rates, no data on adult mortality rates were available for ferrets. Data from other mustelids indicated adult mortality ranged from 15%-25% for Maries pennanti (Kelly 1977), M. americana (Strickland et al. 1982), and Mustela erminea (Stroganov 1937), although King (1980) reported mortality rates of 80%-90% for M. ni- valis . We varied adult mortality ( |Ji) for ferrets from 0.2 to 0.4. The probability of ferret populations of size n reaching a size where the probability of extinc- tion is nearly zero was calculated by P ~ 1 - (fJL/X)" where |x and X are the per capita death and birth rates, respectively (MacArthur and Wilson 1967, Richter-Dyn and Goel 1972). Two results emerged from this analysis (Table 1). First, as |x/X decreases, the number of female ferrets needed to colonize an area successfully de- creases for a given probability. Second, as the number of breeding females increases, a higher )x/X can still be tolerated with a high probability of successful colonization. These results have a practical application in the reintroduction of 154 Great Basin Naturalist Memoirs No. Table 1. Probability of a female black-footed ferret population of size n successfully colonizing and reaching carrying capacity. For a birth rate (X) of 0.5 and a death rate (|x) varied from 0.2 to 0.4, the resulting jjl/X ranges from 0.4 to 0.8. Table 2. Times to extinction (T|,) for black-footed fer- ret populations with a carrying capacity (K) of 40 and 50, and birth (\)/ death ((x) rates varied as shown. See text for details. Population size (n) Death rate/birth rate ratio (fJi/X) 0.9 0.8 0.7 0.6 0.5 0.4 5 0.41 0.67 0.83 0.92 0.97 0.99 10 0.65 0.89 0.97 0.99 15 0.79 0.96 0.99 20 0.88 0.99 30 0.96 40 0.98 50 0.99 black-footed ferrets. For example, if |Ji/X is con- servatively estimated to be 0.9, these data sug- gest that at least 15 females will be needed for a reintroduction that has an 80% chance of suc- ceeding. As more accurate birth and death rates are estimated for the Wyoming ferrets, this type of analysis will become even more useful for assessing the number of animals needed for translocations. We also calculated extinction times for an es- tablished population of ferrets (Table 2). These values (TJ were estimated by the MacArthur and Wilson (1967) equations Ti, ~ T, where \ - JJL Ti = 2 (X/|ji)' ■ (l/i\) i 1 As the death rate to birth rate ratio ((x/X) in- creases, the time to extinction (TJ decreases. For a given |jl/X, T,. increases as the carrying capacity (K) increases. Assuming that the carry- ing capacity of the Wyoming ferret population is approximately 40 adults (Forrest et al. 1985) and that |jl/X equals 0.8, the time to extinction for this population as predicted by the T,. model would be about 1,000 years. Data from Table 1 show that this population of 40 (20 females, 20 males) with a death rate/birth rate ratio of 0.8 would have about a 99% chance of successful coloniza- tion and conversely less than a 1% prol)al)ility of extinction. Thus, a population of 40 adult ferrets should be a reasonable hrst estimate of MVP size as predicted by the l\ model. If we assume that there is a constant average density of ferrets across the study area of 1/50 ha (Forrest et al. 1985), this MVP estimate would result in a mini- mum area re(|uirement of 2,000 ha. In actuality, M-ZX Tk (years) 40 0.5 0.2 0.4 1.0 X 10'= 0.5 0.3 0.6 3.0 X 10' 0.5 0.4 0.8 1.0 X 10' 0.5 0.2 0.4 8.0 X 10" 0.5 0.3 0.6 5.0 X 10' 0.5 0.4 0.8 5.0 X 10' because prairie dog densities varv in this habitat (Clark et al. 1985), the 2,000 ha figure is low. For endangered species with insular or patchy distributions, the theory of island biogeography can be a useful tool for estimating MVP size if data on birth and death rates are available. The T|^ model is a mathematically simple one that assumes limited exponential growth, either birth or death rates that vary linearly with den- sity, no age- specific variance of birth and death rates, and a constant environment. Neverthe- less, the Ti^ model has at least two serious draw- backs: (1) it ignores inbreeding depression (see genetic considerations), which may be impor- tant in small populations, and (2) violation of the model assumptions decreases the time to extinc- tion (T|,), thus producing what may be overly optimistic T,^ values (Shaffer and Samson 1985). For the ferret, there is good evidence that not all females breed and that natality varies with age (Forrest et al. , in manuscript), thereby violating two assumptions of the model. The degree to which these \'iolations may decrease extinction times is unknown. Other researchers have criticized all aspects of island biogeography theory, the T|^ model in- cluded, claiming it remains essentially unsub- stantiated (Gilbert 1980, Margules et al. 1982). Despite these criticisms, several studies have shown good agreement between model predic- tions and held observations. Crowell (1973) re- ported good accordance between observed and predicted times to extinction for mice intro- duced to islands. Similarly, Smith (1974, 1980) and Fritz (1979) predicted times to extinction that appeared to explain satisfactorily local dis- tributions of pikas {Ocliotomi ))rinceps) and spruce grouse {Canachitcs canadensis). We sug- gest that the T,^ model of island i)i()geography remains a useful one tliat should be used in con- junction with other methods to estimate MVP size and area re(iuirements. 1986 Groves, Clark: Minimum Pofiilation l55 Simulation Models Computer simulations may be the most useful method to estimate MVP size and minimum area requirements. Except for theoretical models like the T,. of island biogeography, computer models provide the only method in which a probability value for survival can be attached to the MVP estimate. In addition, these models are not confined to the mathematical assumptions of analytical models (e.g., density dependent na- tality or mortality rates, lack of age structure) and provide a flexible mechanism for assessing the sensitivity of MVP estimates to changes in cer- tain population parameters. Shaffer (1978) employed the simulation ap- proach to estimate MVP size and minimum area requirements for the grizzly bear (Ursiis arctos) in Yellowstone National Park. This simulation evaluated the effects of both demographic and environmental stochasticity and also indicated which population parameters were most likely to affect changes in survival probability. Watts and Conley (1981) used both stochastic and deter- ministic models to predict survival and extinc- tion probabilities for a remnant population of bighorn sheep in the southwestern United States. Their simulations evaluated the effects of demographic but not environmental stochasti- city. Although an estimate of MVP size was not an explicit result of their eflfort, it could have been obtained from their simulations. The major disadvantage of the simulation ap- proach is the extensive population data it re- quires. Minimum data requirements are the mean and variance of age-specific and sex- specific natality/mortality rates, age structure, sex ratio, and the relationship of these variables to density (Shaffer 1981). For most species, ob- taining these data requires an extensive and in- tensive mark/recapture and/or radio telemetry effort over several years. In his grizzly bear sim- ulation, Shaffer (1978) used the 12-year data base of Craighead et al. (1974). Such extensive data for any wildlife species are the exception, not the rule. However, for some species in which data are lacking, it should be feasible to substi- tute data from a closely related (congeneric) spe- cies into the simulation. The black-footed ferret is a case in point. Al- though some natality and sex ratio data are avail- able (Forrest et al., in manuscript), mortality data are not. With ferret data on natality and steppe polecat (M. eversmanni) data on mortal- ity, Shaffer (personal conmiunication) simulated some preliminary estimates of MVP size for the ferret. Work is now in progress to refine these simulations, which will take into account both demographic and environmental stochasticity. As the Wyoming ferret study proceeds, more age- and sex-specific natality/mortality data may become available to incorporate into the simula- tion. For short-term management needs, how- ever, a MVP estimate by simulation using data from a closely related species should suffice for the black-footed ferret. Although not directly useful in estimating MVP size, simulation models based on bioen- ergetics are useful in estimating minimum area requirements. Two such models exist for the ferret. One model estimates gestation, lactation, and growth energy requirements for one female ferret and her young (Stromberg et al. 1983); the other estimates energy requirements based on experimental feeding studies of steppe polecats and observed activity patterns of the Wyoming ferrets (Powell etal. 1985). Both models indicate the number of prairie dogs required to support a given population of ferrets. Combined with esti- mates of MVP size, data from these models are helpful in estimating minimum area require- ments for the ferret, assuming that prey (prairie dog) densities can be measured and that there is a close correspondence between prey density and availability. Computer simulations, though realistic in some aspects, cannot incorporate all the behav- ioral and ecological adaptations of every species (Watts and Conley 1981). They should not be used to predict actual population parameters but instead should indicate a range of possibilities for those parameters (Fowler 1981). In discussing the role of computer simulations in wildlife sci- ence, Romesburg (1981) viewed these models as comprehensive tools for integrating knowledge, common sense, hunches, and opinions. It is in this role that we feel computer simulations can be an aid to estimating MVP size and minimum area requirements for endangered species. Shaf- fer's grizzly bear simulation is presently being used by wildlife managers in Montana and Wyo- ming in this capacity. In addition, an adaptation of his simulation is currently being made avail- able to the US DA Forest Service for use in the management of vertebrate species in their forest planning process (Shaffer, personal communica- tion). 156 Great Basin Naturalist Memoirs No. Genetic Considerations Increasing attention is being given to estimat- ing minimum population numbers from a ge- netic standpoint (Soule and Wilcox 1980, Frankel and Soule 1981, Schonewald-Cox et al. 1983, Lacava and Hughes 1984, Lehmkuhl 1984, Salwasser et al. 1984). The critical ques- tion is what population thresholds are necessary to maintain short-term and long-term (evolu- tionary) fitness. Based on the extensive experi- ence of animal breeders, Franklin (1980) and Soule (1980) determined that the maximum al- lowable rate of inbreeding necessary to avoid short-term effects of inbreeding depression is 1% per generation. This rule of thumb translates to a genetically efiective size of 50 for mainte- nance of short-term fitness. The concept of a genetically effective popula- tion (NJ is important. Such a population is de- fined as one in which all individuals mate ran- domly, the sex ratio approaches unity, variance in family size is zero, and generations do not overlap. Obviously, few if any vertebrate popu- lations meet these criteria. Thus, a genetically eSective size of 50 translates into a larger actual (census) number of breeding adults for most spe- cies. Several mathematical expressions have been derived for determining the effects of variance in progeny, unequal sex ratios, overlapping gener- ations, and population fluctuations on the genet- ically effective population size (e.g. , Kimura and Crow 1963, Emigh and Pollak 1979). Lehmkuhl (1984) developed a procedure, based on the above formulae, for adjusting the genetically ef- fective population of 50 to arrive at the census number of breeding animals. With Lehmkuhl's procedure, we estimated the number of breed- ing ferrets (N) necessary to maintain a geneti- cally effective population (N^) of 50 (see the Ap- pendix for calculations). Results of this exercise indicated that a MVP of 214 breeding ferrets is necessary for maintenance of short-term genetic fitness. This MVP estimate is five times the known number of adults in the Wyoming popu- lation (Forrest et al., in manuscript). Minimum area requirements for this MVP estimate would be about 10,700 ha of white-tailed prairie dog habitat. A question arises as to what constitutes short- term versus long-term genetic fitness. This question can be satisfactorily answered with the following equation f=l (i-^)' 2N, where / equals the inbreeding coefficient, N^ equals the genetically effective population size, and t equals generation time. Animal breeders have noted a significant reduction in fecundity when/approaches 0.5-0.6 (Soule 1980). Setting Np at 50 and/at 0.6 in the above equation results in t equal to about 90 generations. If the genera- tion time for the ferret is 1 year (time from birth of a female kit to birth of her first litter), it would take about 90 years for an effective population of 50 ferrets to reach an inbreeding coefficient that could cause extinction solely on genetic grounds. If the N^ for the Wyoming ferret popu- lation is now substantially less than 50 animals, this short-term genetic threshold could be reached in considerably fewer than 90 years. The amount of time the population has already been isolated may be several decades. Franklin (1980) suggested that a genetically effective population of 500 animals is necessary for long-term genetic variation required by the evolutionary process. As with short-term fitness, a Np of 500 translates into a much larger number of breeding adults when dealing with real, as opposed to ideal, populations. As an alternative to this rule of thumb approach, Frankel and Soule (1981) suggested monitoring genetic varia- tion in target species by electrophoretic tech- niques that estimate percentage polymorphisms and percentage heterozygosity in a population. Data from natural populations suggest that rela- tively heterozygous individuals have greater vi- ability and fecundity than individuals with lower percentages of heterozygosity and polymor- phism (Soule 1980). For many species, particularly endangered ones, electrophoretic techniques can be imprac- tical for several reasons. First, they require po- tentially injurious tissue sampling. Second, small sample sizes of most studies can verify the presence but not the absence of rare alleles and preclude the use of statistics to test departures from Hardy-Weinburg frequencies. Third, many higher vertebrates lack polymorphic ge- netic markers that these techniques can detect. Nevertheless, if these techniques become more efficient and automated in the future, they could be a valuable management tool for working with threatened and endangered species. If test re- sults indicated a reduction in genetic variability, then restorative steps, such as increasing Np, 1986 Groves, Clark: Minimum Population 157 could be taken (Frankel and Soule 1981). Base- line data on genetic variation in the Wyoming ferret population are now being gathered (Kil- patricketal., 1986). Although both empirical data and theory have contributed to developing the genetic approach to estimating MVP size, this method is not with- out its shortcomings. Both the "50 and 500 rules" which are being advanced are based on simple, analytical models of population genetics that do not account for either age structure or environ- mental stochasticity. Previous work indicates that the exclusion of age structure can dramati- cally decrease MVP size (Shaffer 1978) and in so doing optimistically deflate the MVP estimates. Discussion and Conclusion At the outset of this paper, we indicated that the five methods for estimating MVP size would be evaluated in relation to research impacts on target species and the value of accurate and pre- cise data when time is a critical factor. In the black-footed ferret study, we adopted the prag- matic philosophy of Frankel and Soule (1981) that conservationists must often employ rough approximations of critical parameters instead of waiting for precise data that may never be ob- tained. Assuming that this philosophy is sound for most endangered species studies, in practice it leads us to conclude that the experimental and simulation methods for estimating MVP size have serious limitations. The former requires a lengthy study period, and the latter an extensive set of population data. Neither time nor exten- sive population data are goods in great supply for most endangered species programs. As previ- ously pointed out, though, it should be possible at times to use less than "perfect" data in the simulation approach to produce a meaningful MVP estimate. Although Shaffer (1981) underscored the im- portance of defining MVP size with explicit but flexible criteria (i.e., time frame and survival probabihty), only two of the five methods exam- ined (Tj^ and simulation) yield results that in- clude both of these criteria. Are we then to dis- miss the other approaches as meaningless exercises? We think not. When applicable, both the experimental and biogeographical methods provide empirical data for estimating MVP size, although no survival probability can be attached to their estimates. A prudent path for any con- servation biologist should be to use both empiri- cal as well as theoretical approaches such as is- land biogeography and genetic considerations. Each of the methods for estimating MVP size relies on several assumptions. Which method(s) then are most appropriate? Looking at this same question from another angle, one might ask to which type of perturbation is the target species or population most susceptible. For most endan- gered species, it is unlikely that the latter ques- tion will be answerable; it is likely that the target species will be subject to some combination of demographic, environmental, and genetic per- turbations. Therefore, those methods that take into account these perturbations should proba- bly carry the most weight. Obviously, the avail- able data will also dictate to some extent which methods can be used. For the majority of endangered species studies, empirical data necessary for estimat- ing MVP size with the experimental and bio- geographical methods will not be available. If population data on natality and mortality are obtainable, then a combination of the simula- tion and genetic approaches may yield the best estimate of MVP size. The simulation method can account for both demographic and environmental perturbations, and the ge- netic method can account for demographic stochasticity as it relates to genetic drift. On the other hand, theoretical methods like the T^ model of island biogeography only account for demographic perturbations. A combina- tion of the simulation and genetic approaches will likely produce a range of MVP estimates. Results of the recent Forest Service workshop indicated that environmental and demo- graphic variability may be more important than genetic variability in setting the lower limit to population viability (M. L. Shaffer, personal communication). One possible im- plication of these results is that the final MVP estimate should be biased toward that esti- mate produced by the simulation method. However, management constraints will also be a primary factor in establishing MVP size somewhere within the range of estimates. When no population data are available and empirical approaches have also been ruled out, then genetic considerations must take priority in estimating MVP size. This is the case for the ferret. Until more data are obtained for simula- tions, we must rely on our estimate of about 200 158 Great Basin Naturalist Memoirs No. animals by the genetics method to serve as the MVP estimate for black-footed ferrets. In the final analysis, this number may be reduced as a result of compromising between estimates fi-om the simulation and genetics methods. The preceding discussion on genetic ap- proaches is relevant only to maintenance of short-term fitness. For fitness on a long-term scale, we must try to establish several ferret populations, each of which maintains a short- term MVP size. Although such goals may now appear beyond the vision of what resource managers can do to conserve the ferrets in the near term, we must never lose sight of the fact that large, viable populations of ferrets and other endangered species are necessary for the recovery and conservation of these species fi-om an evolutionary standpoint. Acknowledgments We thank the following groups for financial support in our Wyoming ferret study: World Wildlife Fund — U.S., Wildlife Preservation Trust International, New York Zoological So- ciety's Wildlife Conservation International, National Geographic Society, Charles A. Lindbergh Fund, National Wildlife Federa- tion, The Nature Conservancy, the Joseph Henry Fund of the National Academy of Sci- ences, Defenders of Wildlife, Humane Soci- ety of the United States, Sigma Xi, the Fre- mont County (Wyoming) Audubon Club, and the Bethesda-Chevy Chase Chapter of the Izaak Walton League of America. S. Forrest, C. Fowler, R. Miller, M. Shaffer, M. Soule, and J. Terborgh provided constructive criti- cisms of earlier drafts. Discussions with M. Shaffer improved the manuscript consider- ably. Appendix Calculation of actual adult census number (N) for black-footed ferrets through adjust- ment of genetically effective population size (NJ (Lehmkuhl 1984). Step 1 : Adjustment for variation in progeny number. N = (2N, V^ + 1)/K' where Np ^ effective population size of 50, V^ ^ variance in number of young, K = mean number of young per female. When lacking data on V;,, multiply 1.4 times N^. For the black-footed ferret: 50 X 1.4 = 70. Step 2 : Adjustment for unequal se.x ratio. Nm = (Ne + [male:female ratio X Nj)/4 where Np, = number of males. Nf = N^ X femaleimale ratio N = N„ + Nf N/50 = increase in N^ N = increase x 70 (from Step 1) For the black-footed ferret: (sex ratio data from 1984 census — Forrest et al., in manuscript) N^ = (50 + 18/25(50))/4 = 21.5 Nf=21.5 X 25/18 = 29.9 N = 29.9 + 21.5 = 51.4 51.4/50 = 1.02 1.02x70 = 71.4 Step 3 : Adjustment for overlapping generation. Double the census number from Step 2. For the black-footed ferret: 2 X 71.4 = 142.8 Step 4 : Adjustment for population fluctuations. Ratio between empirical high and low population cen- suses indicates factor with which to increase Step 3 result for MVP estimate. For black-footed ferret: 43/28 =1.5 1.5 X 142.8 = 214 = MVP (1983 census = 28 adults, 1984 = 43, Forrest et al. , in manuscript) Literature Cited Brown. J. H. 1971. Mammals on niountaintops: nonequi- librium insular biogeography. Amer. Nat. 105:467- 478. Clark. T. W. 1981. A proposal for a research grant in support of the Meeteetse black-footed ferret conservation studies; for the period 15 Feb. 1982-14 Feb. 1983 (12 months). 85 pp. 1984a. Strategies in endangered species conservation: a research view of the ongoing black-footed ferret c