FOOD HABITS OF ARCTIC WOLVES IN GREENLAND

FOOD HABITS OF ARCTIC WOLVES IN GREENLAND ULF MARQUARD-PETERSEN Greenland Wolf Research Project, 3701 Beeman Circle, Anchorage, AK 99507 Analysis of feces was used to detennine natural prey of arctic wolves (Canis lupus) in Nansen Land, North Greenland (83'N, 43'W), and Hold with Hope, East Greenland (74'N, 22aw). A total of 461 feces was collected during summers 1991-1995. In Nansen Land en = 110), muskox (Ovibos moschatus) was the major food of wolves with a frequency of occurrence of 79.4%. Lemmings (Dicrostonyx groenlandicus) were the next most-important group with a frequency of 19.6%. followed by arctic hares (Lepus arcticus) at 7.5%, wolves at 0.9%, and arctic foxes (A/apex lagopus) at 0.9%. Plastic and nylon rope occurred in three feces, suggesting scavenging on garbage. In Hold with Hope (n = 351), muskoxen also were the predominant prey with a frequency of 65.4%, followed by geese and other birds at 35.5%, lemmings at 10.2%, arctic hares at 2.3%, and beetles at 0.3%. Remains of garbage (paper and paint chips) occurred in one fece. Differences in availability of regional prey may best explain dissimilarity in food habits between areas, but seasonal differences in contents of feces also may be important. Key words: Canis lupus aretos, arctic wolves, food habits, Greenland, muskoxen Diet of wolves (Canis lupus) in North America has been subject to intensive study. These investigations have established that wolves kill a variety of prey. Most of the diet typically consists of large ungulates such as moose (Alces alees), caribou (Rangifer tarandus), and deer (Odoeoileus). Small mammals such as beavers (Castor canadensis), snowshoe hares (Lepus americanus), and microtines may be seasonally or locally important prey (Mech, 1991). A variety of techniques has been employed to evaluate diets of wolves. The most effective methods are direct observation of kills and examination of kill sites. Nonetheless, these procedures are not always feasible, so analysis of feces is frequently used despite inherent short-comings. First, an investigator is unable to distinguish between scavenging and predation. Second, analysis of prey species that differ markedly in size can result in overrepresentation of the proportion of smaller animals eaten relative to larger prey because of varying surface:volume ratios. Smaller prey has a higher surface: volume ratio and is covered by more hair or feathers per mass of flesh relative to larger prey. Because hair and feathers constitute the primary identifiable remains in carnivore feces, the contribution of smaller prey may be overestimated. This potential bias is minimized with the linear-regression model developed by Floyd et al. (1978) and refined by Weaver (1993). Content of feces is reported as the number of feces containing a certain prey species divided by the total number of fecal samples (frequency of occurrence). This technique is an accurate and justifiable method (Corbett, 1989), Food habits of wolves at high latitudes in the Arctic have not been subject to intensive study; only a single investigation has been published. Tener (1952) examined 85 feces of wolves collected during investigations of muskoxen (Ovibos moschatus) on Ellesmere Island, Northwest Temtories, Canada, and reported that 83.3% contained remains of arctic hares (Lepus arctieus) and only 16.7% contained remains of muskoxen. There are no empirical data published on food habits of wolves in Greenland. As part of general zoological studies, some investigators analyzed feces of wolves from Journal of Mammalogy. 79( I );236--244. 1998 236

February 1998 MARQUARD-PETERSEN-FOOD HABITS OF WOLVES 237 Nansen Land _ Amdrup Land km 500 Holdwilh Hope FIG. I.-Greenland showing study areas and wolf range (shaded area). North Greenland, but sample sizes were not identified (Dawes et al., 1986). This paper presents data on the food habits of arctic wolves in Greenland as part of a broader study of the ecology of wolves in the eastern Arctic (Marquard-Petersen, 1994, 1995). This study was undertaken to: 1) provide data on the extent of wolf predation on muskoxen, 2) assess relative abundance of the different prey species in the diet of wolves, and 3) evaluate geographic variation in the diet of arctic wolves in north and east Greenland. The population of muskoxen in northeastern Greenland has decreased significantly since the early 1980s (Boertmann and Forchhammer, 1992), when wolves moved into Greenland. Whether wolf predation played a role in this decline is not known. Based 00 results of studies of food habits of wolves in North America, I hypothesized that frequency of occurrence of ungulate (muskoxen) in feces would be highest and that of small mammals lowest. I also hypothesized that there would be significant differences in diets of wolves in Nanseo Land, North Greenland, compared with wolves in Hold with Hope, East Greenland, because of variation III availability of prey species. MATERIALS AND METHODS Study area.-prior to our initial field work in 1991, little was known about distribution and core areas of wolves in Greenland. No information was available suggesting localities where enough feces could be collected for a study of food habits of wolves. Therefore, I analyzed 136 observations of wolves made by military personnel and government scientists during 1978-1990. This analysis suggested that wolves occurred in several areas. I selected two areas, from which I assumed that a sufficient sample size of feces could be obtained: Nansen Land in North Greenland, and Hold with Hope in East Greenland (Fig. 1). The first study area was located in Siriuspasset (82 55'N, 41 30'W), a valley transecting southeastern Nansen Land from J. P. Koch Fjord to Brainard Sund, 6-7 kin wide and 24 km long. The valley consisted of clay soils derived from Cambrian siltstone. The surrounding topography was characterized by high-alpine terrain with mountain peaks to an elevation of 1,300 m. Numerous glaciers descended into the valley. Siriuspasset was considered true polar semidesert but was one of the few areas in North Greenland

238 JOURNAL OF MAMMALOGY Vol. 79, No. J with relatively luxuriant vegetation. Ninety-two species of vascular plants have been identified (Bay, 1992). There were few meteorological data for the area; annual precipitation was believed to be <50 nun with a mean annual temperature of -16 Q C (Klein and Bay, 1994). The only extant ungulate was muskoxen. Greenland caribou (Rangifer tarandus) did not occur in North and East Greenland. Collared lemmings (Dicrostonyx groenlandicus), arctic hares, and arctic foxes (Alopex Jagopus) were common. The population of lemmings in North and East Greenland was cyclic, and variation in density was likely critical in availability of prey for wolves. Little is known about population fluctuations of arctic hares and arctic foxes in northeastern Greenland. The second study area was located in Hold with Hope including Badlanddalen (73 30'N, 21 45'W), a valley 7-12 Ian wide and 21 Ian long, transecting western Hold with Hope from Loch Fyne to Mackenzie Bay. Badlanddalen consisted of quaternary marine deposits of 72-97% sand (Marquard-Petersen, 1994) with a complex of ridges, river valleys, and ponds. Part of the valley had a desert-like appearance and was subject to severe sandstonns. Vegetation cover of other areas of Badlanddalen was 25% to >50% (Bay and Boertmann, 1988). The surrounding topography was characterized by highalpine terrain with mountain peaks > 1,250 m. There were few recent meteorological data for the area. The nearest weather station was in Daneborg ca. 100 Ian to the north. which reported a mean annual temperature of -9.9 C and a mean annual precipitation of 285 mm (Klein and Bay, 1990). Yearly precipitation could be substantial depending on the extent of ice cover on the Greenland Sea. The fauna was rich and varied. Wolves, muskoxen, collared lemmings, arctic hares, and arctic faxes were common and occurred in the area year-round. Greenland caribou became extinct in this area around 1900. Avifauna consisted of 38 species; most were present from mid-may through August (Elander and Blomqvist, 1986). Both study areas were located in the Northeast Greenland National Park. Field collections.-wolf feces were conected opportunistically. In Nansen Land, most feces (n = 57) were collected at Brainard Sund in July and August 1991. Another 15 feces were collected throughout Siriuspasset. In May 1995, 12 feces were collected at Brainard Sund and 22 feces around Frigg Fjord (83 N, 34 W). Because of homogeneity of wolf habitat in North Greenland, those feces were pooled with feces from Nansen Land, with two feces from Wanning Land (81 30'N, 52 W). and two from Amdrup Land (80 0 45'N, 16 W; Fig. 1); total sample size was 110. North and east Greenland were frequented by sled dogs of the Danish military who patrolled the region. Therefore, methods were developed to differentiate between feces of dogs and wolves. This procedure consisted of an on-site inspection using knowledge of differential food habits. Military sled dogs subsisted on a diet of dried cod (Gadus morhua). Feces of dogs typically took on a clay-like appearance when compressed and contained only occasional hair. At Brainard Sund, feces of dogs were deposited at intervals along a straight line. because dogs were chained' in this fashion during rest stops. In contrast, feces of wolves had a high concentration of manunalian hair and solid bone fragments and were scattered around a site. Feces of wolves were assumed to represent the yearround diet because they were collected away from den sites and varied in age as evidenced by varying degrees of desiccation. In Hold with Hope, most feces (n = 297) were conected in July and August 1992 at two different den sites. One den was not in use that year. A total of 156 desiccated and weathered feces of undetennined age was collected and assumed to represent a denning season prior to 1992, likely 1990 when the den was occupied by at least two adults and two young (Burton, 1990). The other den was occupied by a pack of six adults and three young (Marquard-Petersen, 1994). After this den was abandoned on 28 July 1992, 141 fresh feces were collected. In July 1993, an additional 54 fresh and desiccated feces were located in the home range of the pack. These were assumed to represent the year-round diet. Hair samples of lemmings, arctic hares, wolves, and muskoxen were col1ected for later comparison with guard hairs in feces. In July 1991, hair samples of lemmings and arctic hares were collected in Nansen Land when eight juvenile lemmings and 15 adult hares were killed for an unrelated study. In July 1992, hair samples of wolves were located at the occupied den site at Hold with Hope. Hair samples of muskoxen were collected from nine carcasses believed to represent kill sites from the previous

February 1998 MARQUARD-PETERSEN-FOOD HABITS OF WOLVES 239 TABLE I.-Frequency of occurrence of prey in 451 feces of wolves from Nansen Land, North Greenland, and Hold with Hope, East Greenland, collected during summers 1991-1995. Nansen Land Hold with Hope Hold with Hope Hold with Hope year-round year-round den before 1992 1992 den Prey remains (n = 107) (n = 51) (n = 154) (n = 139) Muskoxen Birds Lemmings Arctic hares Wolves Arctic foxes Beetles Total 85 21 8 116 winter. but this could not be confirmed due to lack of fresh blood and wolf tracks. Laboratory analysis.-feces were numbered. broken apart. and a sample was taken of the hair or feathers in each fece. Samples were cleaned in acetone, mounted pennanently. and assigned the same accession number as the feces. Guard halls in each permanent mount were examined under a compound microscope and compared with reference hall samples. An identification to species was made based on coloration, length, and pigment banding. No criteria have been developed to accurately differentiate remains of muskoxen calves from those of older animals; therefore. separation of these age groups was not possible. Feathers and avian bone fragments were compared with reference material and identified to species if possible. Remains of birds other than geese were lumped as unidentified bird. Items considered accidentally ingested (gravel. plant matter) were not reported. Calculations of frequency of occurrence were based on number of occurrences of each type of prey divided by the total number of feces and were expressed as a percentage of the total sample. Statistical analysis.-analyses were based on contingency table log-likelihood G-statistic (Zar. 1984). I tested the hypothesis that there was no difference in use of prey between areas and years. I also tested the hypothesis that there was no difference between contents of feces collected throughout the range versus those collected at den sites. Feces that contained remains of more than one prey item (n = 54) were not included in statistical analyses, which assumed independence of samples (Zar. 1984). Statistical significance was at P = 0.05. 37 8 5 5 55!O5 83 61 56 15 14 2 184 154 RESULTS Nansen Land. North Greenland.-Identifiable prey occurred in 107 of 110 feces (Table 1). Prey species included muskoxen, lemmings. arctic hares, wolves, and arctic foxes. Results indicated that the major food of wolves was muskoxen with a frequency of occurrence of 79.4% (Fig. 2). Lemmings constituted the second most important food with a frequency of occurrence of 19.6%, followed by arctic hares with a frequency of occurrence of 7.5%. Hair and bone fragments of wolf occurred in one fece, and hair of arctic foxes occurred in one fece (frequency of occurrence. = 0.9%). More than one prey item occurred in 8.4% (n = 9) of the feces. Three feces contained nylon rope, remains of a plastic bag, and plastic, respectively. Three feces contained unidentifiable remains and were not included in the analysis. Use of prey in Nansen Land was skewed toward a higher occurrence of muskoxen than small mammals (G ~ 48.9. d.! ~ 1. P < 0.01). In Nansen Land. density of wolf feces at Brainard Sund was higher than density of feces collected elsewhere in Siriuspassel. Fifty-seven wolf feces of undetermined age littered the area around the Brainard Sund camp in July 1991. Hold with Hope, East Greenland.-Identifiable prey occurred in 344 of 351 feces

240 JOURNAL OF MAMMALQGY Vol. 79, No. / 80 - C60 ~ c ~ 40 ~ 20 r o f.ol!i',-l- = '--c- 1 Muskox Bird LemmlOg Arctic Hmc Wolf Arctic Fox ll~e(lc Prey Ite m Nansen Land Hold with Hope Flo. 2.- Frequency of occurrence of prey items in 107 wolf feces from Nansen Land, North Greenland. and 344 feces of wolves from Hold with Hope, East Greenland. (Table 1). Prey species identified included muskoxen, barnacle geese (Branta leucopsis), pink-footed geese CAuser brachyrhynchus), lemmings, arctic hares, and beetles (Coleoptera). Results suggested that the major food of wolves was muskoxen with a frequency of occurrence of 65.4% (Fig. 2). Geese and other birds were the second most important food (35.5% of the diet). Remains of geese in feces (n = 32) typically consisted of I) feathers with flight feathers typically only represented by the proximal 3 cm of the quill, 2) epiphyses of long. bones. and 3) mandibles. Geese feathers could not be separated to genus or species with certainty. Feathers in 96 feces ::: 40 g 30 ~ 20 10 o ~M;;:",l'.ko~, -f:"':;; mminl!'---'a~"::; ' ;:C ";;:":;;:"-,"llird Prey Item 199C 1992 FIG. 3.-Frequency of occurrence of prey items in 293 feces of wolves from two dens in Hold with Hope, East Greenland. showed no characteristics helpful in identification and were reported as unidentified bird. Based on pigment and size, I believe that many of these feathers represented remains of geese, but this could not be verified. Frequency of occurrence of lemmings in feces was 10.2% and arctic hares 2.3%. Beetles occurred in a single wolf feee (0.3%). More than one prey item was detected in 13.4% (n = 47) of the feces. Six feces contained no identifiable material and a single feee contained garbage (paper remains and blue paint chips); those feces were not included in the analysis. Proportions of prey in feces collected at the two den sites varied Httle between years (Fig. 3). Frequency of occurrence of remains of muskoxen was 68.2% for the denning season in 1990 and 59.7% for the denning season in 1992. Frequency of occurrence of bird remains was 39.6% (1990) and 40.3% (1992). Frequency of occurrence of lemming remains was 9.7% (1990) and 10.1% (1992). Frequency of occurrence of arctic hare remains was 1.3% (1990) and 0.7% (1992). More muskoxen was found in feces (G = 146.43, dj = 1, P < 0.001) than other prey items. There was no difference in contents between feces collected at the two

February 1998 MARQUARD-PETERSEN-FOOD HABITS OF WOLVES 241 t 80 60 8 ~ 40 5 8 20 o FIG. 4.-Frequency of occurrence of prey items in 51 feces of wolves from summer and winter range in Hold with Hope, East Greenland. dens (G = 1.23, d.f = 3, P > 0.05), but there was a difference in contents between feces collected at den sites and those collected on summer and winter ranges (Fig. 4) (G = 7.57, d.f = 2, P < 0.05). There was no difference in contents between feces collected year-round in Nansen Land or Hold with Hope (G = 1.68, d.f = 2, P > 0.10). DISCUSSION Most studies of food habits of wolves in North America indicate that ungulates are a primary prey (Clark, 1971; Fuller and Keith, 1980; Haugen, 1987), which is likely related to availability of prey and energetic costs of capturing and handling prey. Although capturing larger prey requires increased effort relative to small animals, more food is secured when a kill is made. When large ungulates become unavailable, a functional response of wolves causes increased reliance on smaller prey. Availability and vulnerability are important factors in selection of prey. When these factors vary bet~een areas, a difference in diets of wolves is likely to exist. There is little published information on wolf predation on muskoxen, so a quantitative comparison with my results is difficult. Tener (1952) reported that 16.7% of feces of wolves (n = 85) collected on El1esmere Island contained remains of muskoxen, but new methods (Floyd et ai., 1978; Weaver, 1993) suggested that these findings may not reflect a reliance on arcti~ hare as the primary prey. Kuyt (1972) reported that wolf predation on muskoxen in the The10n Game Sanctuary, Northwest Territories, Canada, was insignificant (1.8%), perhaps due to the availability of caribou. More recently, analysis of 98 feces collected at two dens in Ellesmere Island National Park Reserve revealed that 97.5% contained remains of muskoxen (B. Troke, in litt.). In addition to energy requirements, the most likely reason for the high frequency of occurrence of muskoxen in feces of wolves from both study areas was related to availability of prey. Unfortunately, accurate data on population sizes of muskoxen were not available. No comprehensive census has been conducted in either study area and guesses vary. Boertmann et al. (1992) reported 1-50 muskoxen in Nansen Land. I counted 56 muskoxen in central Nansen Land during wolf surveys in July August 1991. and groups of $12 muskoxen have been observed in western Nansen Land, an area I did not visit. Therefore, population size may be closer to 75 as suggested by Bennike et al. (1989) or 50-100 as reported by. Boertmann and Forchhammer (1992). Overall density of muskoxen in North Greenland was extremely low. ca. 1.7 muskox/ioo km' (Aastrup et al., 1986). In contrast, density of muskoxen in Siriuspasset was estimated at 38 muskoxen/loa km 2 in July 1991 (Klein and Bay, '1994). Nonetheless, overall density in Nansen Land is probably closer to 3.4-4.5 muskoxen/i 00 km 2, assuming the population consisted of 75-100 animals and that the area excluding fjords was ca. 2,200 km 2 Relative to ungulate herds in lower latitudes, the population of muskoxen in Nansen Land appeared extremely small, but many areas of North Greenland consisted of large plateaus essentially devoid of vegetation and mammals. The prey base for wolves in Nansen

242 JOURNAL OF MAMMALOGY Vol. 79, No.1 Land could be evaluated based on this infannation. Availability of prey to wolves in Hold with Hope was substantially higher than in N ansen Land. The population of muskoxen was estimated at 1,000-2,000 animals, including muskoxen in adjacent Hudson Land and Gauss Peninsula (Boertmann et al., 1992). In Bad1andda1en, Hold with Hope, muskoxen density in mid-july 1992 and mid-july 1993 was ca. 22.8 and 23.4/100 km 2, respectively. Although those densities were remarkably similar between years, muskoxen density in Badlanddalen may not be representative of the entire Hold with Hope region. The central northern part of the valley was almost devoid of vegetation and thus offered little forage for muskoxen. Overall muskoxen density in the Hold with Hope region was between 13.6 and 27.21 100 km 2 following the population estimate by Boertmann et a1. (1992) and assuming the region consisted of ca. 7,350 km 2 southeast of Promenadedal. Thus, muskoxen density in this southern study area was at least four times higher compared with N ansen Land, but feces from Hold with Hope suggested that wolves relied less on muskoxen for primary prey than wolves in North Greenland. This outcome may be related to the high occurrence of birds, probably predominately geese, in feces from Hold with Hope, suggesting that this abundant summer food may act as a buffer to predation of muskoxen. This hypothesis could not be tested because no feces were collected in North Greenland exclusively representing summer food habits; no dens have yet been located there. My data indicate that both study areas were capable of sustaining a sufficiently large population of muskoxen to sustain wolves, but this is uncertain given the lack of data on movements of wolves. Both areas have good habitat for muskoxen (Aastrup et ai., 1986; Bay and Boertmann, 1988), and there is mounting evidence that the areas also are important habitat for wolves. Wolves were first detected in Nansen Land in Apri11984 (Dawes et ai., 1986) and have been observed every year since, except in 1988 and 1990. Known pack sizes have ranged from two in 1995 to seven in 1984 (Dawes et ai., 1986; U. Marquard-Petersen, in litt.). In Hold with Hope, wolves were first observed in March 1980, and known pack sizes have ranged from two in 1980 to nine in 1992 (Dawes et al. 1986; Marquard-Petersen, 1994). My finding that muskoxen constituted the major prey of wolves correlates well with what is known about core areas for wolves in Greenland; wolves tend to occur where muskoxen are common. In areas where muskoxen density is low, there are correspondingly few observations of wolves. For example, Hall Land (81 43'N, 60 0 00'W) lacks muskoxen (Boertmann et ai., 1992; Boertmann and Forchharmner, 1992), and no wolves or tracks have been reported observed there since 1965 (Dawes, 1978; Dawes et ai., 1986; U. Marquard-Petersen, in litt.). In contrast, in Peary Land, where muskoxen are co~on, wolves are observed yearly in packs numbering up to eight animals. The high occurrence of avian fauna, including geese, in feces of wolves from Hold with Hope was unexpected. There are few data in the literature on the importance of geese in the diet of wolves, so a quantitative comparison with this study is difficult. Kuyt (1972) reported that geese were not an important prey of wolves on the range of barren-ground caribou in Northwest Territories, Canada, despite being common in the study area where both breeding and molting took place (0.3% of 595 wolf feces). The most likely reason why avian fauna constituted an important secondary prey during both the 1990 and the 1992 denning season probably can best be explained in tenns of availability and vulnerability (ease of capture). Aerial surveys conducted in 1988 produced estimates of 1,104-2,550 pink-footed geese and 567-1,061 barnacle geese (Bay and Boertmann, 1988). Vulnerability is likely variable depending on spe-

February 1998 MARQUARD-PETERSEN-FOOD HABITS OF WOLVES 243 des. For example, barnacle geese usually nest colonially at highly protected sites (e.g., cliff ledges), whereas nests of pinkfooted geese are on top of river banks or large hummocks (Boertmann, 1994) and may be more accessible to wolves. During molting of flight feathers, adult post-breeding pink-footed geese remain flightless for ca. 25 days in July-August, and barnacle geese become flightless for 3-4 weeks starting in early July (nonbreeders), or between mid-july and mid-august for postbreeders (Cramp and Simmons, 1977). This period corresponds to the second half of the denning season, where the nutritional demands of young wolves are great and where adult wolves must secure enough food to ensure survival of their litter. The 297 feces collected at the two dens at Hold with Hope, representing two different denning seasons, suggest that geese are highly vulnerable to wolf predation during the flightless period. This conclusion is supported by observations from the Copper River Delta, Alaska, where two wolves were observed killing three molting dusky Canada geese (Branta canadensis occidentalis) on one occasion (Stephenson and Van Ballenberghe, 1995). Lemmings appeared to be a secondary prey of wolves in Nansen Land but were of less importance to wolves in Hold with Hope. Other researchers have reported a comparable occurrence. Clark (1971) noted that wolves on Baffin Island, Northwest Territories, Canada, during summers 1966 and 1969 frequently ate lemmings (13.2% and 11.5%, respectively). Kuyt (1972) made no attempt at distinguishing lemmings from other microtines but reported an overall occurrence, including Dicrostonyx, of 12.8%. Occurrence of arctic hares was low ~n feces from both study areas. Occurrence of arctic hares in the 297 wolf feces collected at dens in Hold with Hope also was low, suggesting a consistently low dependence on arctic hares in summer. There are no reliable data on hare densities in the two areas, so a comparative analysis is difficult. Perhaps predation on arctic hares in Hold with Hope in summer is buffered by the presence of thousands of vulnerable geese. A substantial amount of wolf hair and parts of a tooth and claw from an arctic wolf occurred in a single fece from Nansen Land, suggesting conspecific aggression or scavenging of a wolf carcass. Most natural mortality among wolves in lower latitudes is conspecific (Mech, 1991), but it is not known if this also holds in the High Arctic, where wolf densities are among the lowest recorded, thereby reducing the probability of interterritorial strife. There was no significant difference in contents between feces collected year-round in both Nansen Land and Hold with Hope, suggesting that differences between the samples from the two areas are strictly seasonal. Future research needs include development of methods for differentiating remains of muskoxen calves versus adults in wolf feces, and investigation of the effects of predation by wolves on the population of muskoxen in northeastern Greenland to ascertain if predation has played a role in its decline. ACKNOWLEDGMENTS My thanks to C. Bay, Botanical Museum, University of Copenhagen, for collecting wolf feces and help with logistics that made the initial field work possible; M. Meldgaard, Danish Polar Center, for valued input during the early stages of this project; D. R. Klein, Alaska Cooperative Fish and Wildlife Research Unit, for assistance with collecting wolf feces in Nansen Land; F. Morschel, University of Alaska Fairbanks, and D. Carter, United Kingdom, for memorable companionship in Hold with Hope. I also thank D. Boertmann, Greenland Environmental Research Institute; R. Burton, Arcturus Expeditions, England; P. R. Dawes, J. D. Friderichsen, A. K. Higgins, J. S. Peel, Greenland Geological Survey; E. M. Laasonen and L. Laasonen, University of Tarnpere, Finland; P. Michelsen, Norway; B. Nielsen, Greenland Telecom; P. V. Norit, Danish Military High Command; J. V. N. Thrner, The Wildfowl and Wetlands Trust, England; and

244 JOURNAL OF MAMMALOGY Vol. 79, No.1 F. van der Stijl, Platinova AlS, Nuuk, Greenland. Funding was provided by the Greenland Wolf Research Project. I appreciated suggestions by J. E. Blake, R. T. Bowyer, E. H. Follmann, and anonymous reviewers. LITERATURE CITED AASTRUP, P., C. BAY, AND B. CHRISTENSEN. 1986. Biologiske miljl'lunders~gelser i Nordgr~nland. 1984-85. Greenland Environmental Research Institute, Copenhagen, Denmark, 113 pp. (in Danish with English summary) BAY, C. 1992. A phytogeographical study of the vascular plants of northern Greenland-north of 74 0 northern latitude. Meddelelser om Gnmiand, Bioscience, 36: 1-102. BAY, c., AND D. BOERThfANN. 1988. Biologiskark<rologisk kortla:gning af Gr~nlands ~stkyst mellern 75 N og 79 30'N. Flyrekognoscering mellem Mesters Vig (72 12'N.) og Nordmarken (78 N.). Greenland Botanical Survey and The Zoological Museum, Greenland Department, Copenhagen, Denmark, 1:1-64. (in Danish with English summary) BENNIKE, 0., A. K. HIGGINS, AND M. KELLY. 1989. Mammals of central North Greenland. Polar Record, 25:43-49. BOERlMANN, D. 1994. Annotated checklist to the birds of Greenland. Meddelelser om Gr~nland, Bioscience, 38:1-63. BOERlMANN, D., AND M. FORCHHAMMER. 1992. A review of muskox observations from North and Northeast Greenland. Greenland Environmental Research Institute, Copenhagen, Denmark, 36 pp. BOERlMANN, D., M. FORCHHAMMER, C. R. OLESEN, P. AASTRUP, AND H. THING. 1992. The Greenland muskox population status 1990. Rangifer, 12:5-12. BURTON, R. 1990. Wolves breeding in northeast Greenland. Polar Record, 26:334. CLARK, K. R. E 1971. Food habits and behavior of the tundra wolf on central Baffin Island. Ph.D. dissertation, University of Toronto, Ontario, Canada, 233 pp. CORBETT, L. K. 1989. Assessing the diet of dingoes from feces: comparison of 3 methods. The Journal of Wildlife Management, 53:343-346. CRAMP, S., AND K. E. L. SIMMONS (EDS.) 1977. The birds of the western Palearctic. Oxford University Press, Oxford, United Kingdom, 1: 1-722. DAWES, P. R. 1978. Ulve i Nordgr~nland. Gr~nland 3: 289-302. (in Danish) DAWES, P. R., M. ELANDER, AND M. ERICSON. 1986. The wolf (Canis lupus) in Greenland: a historic review and present status. Arctic, 39:119-132. ELANDER, M., AND S. BLOMQVIST. 1986. The avifauna of central northeast Greenland, 73 15'N, 74 04'W, based on a visit to Myggbukta, May-July 1979. Meddelelser om Gr~nland, Bioscience, 19:1-44. FLOYD, T. S., L. D. MECH, AND P. A. JORDAN. 1978. Relating wolf feces content to prey consumed. The Journal of Wildlife Management, 48:528-532. FULLER, T. K., AND L. B. KEITH. 1980. Wolf population dynamics and prey relationships in northeastern Alberta. The Journal of Wildlife Management, 44: 583-602. HAUGEN, S. H. 1987. Den-site behavior, summer diet and skull injuries of wolves in Alaska. M.S. thesis, University of Alaska, Fairbanks, 205 pp. KLEIN, D. R., AND C. BAY. 1990. Foraging dynamics of muskoxen in Peary Land, northern Greenland. Holarctic Ecology, 13:269-280. ---. 1994. Resource partitioning by mammalian herbivores in the High Arctic. Oecologia, 97:439-450. KuYT, E. 1972. Food habits of wolves on barrenground caribou range. Canadian Wildlife Service Report Series, 21: 1-36. MARQUARD-PETERSEN, U. 1994. Dens and summer pack size of arctic wolves in Hold with Hope, East Greenland. Polar Record, 172:46-49. ---. 1995. Status of wolves in Greenland. Pp. 55-58, in Ecology and conservation of wolves in a changing world (L. N. Carbyn, S. H. Fritts, D. Seip, eds.). Edmonton: Canadian Circumpolar Institute, Edmonton, Alberta, Canada, 620 pp. MECH, L. D. 1991. The way cif the wolf. Voyageur Press, Stillwater, Minnesota, 120 pp. STEPHENSON, T. R., AND V. VAN BALLENBERGHE. 1995. Wolf, Canis lupus, predation on dusky Canada geese, Branta canadensis occidentalis. The Canadian Field-Naturalist, 109:253-255. TENER, J. S. 1952. A preliminary study of the muskoxen of Slidre Fiord District, Fosheim Peninsula, Ellesmere Island. M.A. thesis, University of British Columbia, Vancouver, Canada, 91 pp. WEAVER, J. L. 1993. Refining the equation for interpreting prey occurrence in grey wolf feces. The Journal of Wildlife Management, 57:534-538. ZAR, J. H. 1984. Biostatistical analysis. Prentice-Hall, Eaglewood Cliffs, New Jersey, 620 pp. Submitted 28 March 1996. Accepted 26 February 1997. Associate editor was James R. Purdue.