NATIONAL PARK SERVICE Wolf Ecology and Prey Relationships on Isle Royale

Populations of wolves that are unexploited by man are rare and generally so remote that a long-term study is prohibitively expensive or impractical. National parks provide unique environments in which to observe these important predators. The insular nature of Isle Royale lends special significance to wolf-population studies, for there is relatively little opportunity for transfer between the island and the mainland.

While the wolf population has exhibited great year-to-year stability in total numbers, there have been important variations in its social organization. Since the population is rather small, significant changes in its structure can be linked circumstantially to single events such as the death of an alpha male (Jordan et al. 1967) or possible ingress of a new pack (Wolfe and Allen 1973). The basic, long-term pattern of a single, large pack and several smaller social units has recently changed. In the early 1970s the island supported two large packs that each utilized about half the area. This provided the potential for an increase in total population, and wolf numbers reached a midwinter high of 44 in 1976.

The recent development of a second major pack on the island appears to have resulted from a significant increase in moose vulnerability and a higher beaver population, an important summer prey species. This addition has caused a higher level of predation on moose in winter, especially when deep-snow conditions increased the vulnerability of calves.

Observations of wolves in winter were made either from light aircraft or from the ground using a telescope at long range. Summer observations were limited to a period of several days at one rendezvous site.

Alpha wolves were the only animals consistently identified from year to year. In addition to providing the pack with leadership, they were most active in scent-marking their environment (during winter observations) and were the most active breeders. Alpha wolves restricted the courtship activities of subordinate members of the pack, and mate preferences were demonstrated, both of which contributed to a reduction in courtship behavior and presumably of mating among subordinate wolves.

Numbers and Organization

The recent establishment of a second wolf pack on Isle Royale was a significant departure from the pattern observed in the 1960s, when the population remained remarkably constant. The history of this wolf population has illustrated the effectiveness of natural mechanisms which adjusted wolf numbers to their food base.

A Brief History

Yearly variations in the Isle Royale wolf population are detailed by Mech (1966), Jordan et al. (1967), and Wolfe and Allen (1973). These provide the basis for the following review. During the initial 11 winters of the project (1959-69), the wolf population varied between 17 and 28 (Table 5). The highest population occurred while the large pack was still in operation in 1965; the lowest was in 1969 after 2 years of social instability.

TABLE 5. Estimated number of wolves on Isle Royale in midwinter, 1959-76.

From 1959 through 1966, the population contained only one large pack. For the first 3 years, this pack traveled over the entire island in midwinter; in subsequent years, its movement usually was restricted to the southwestern two-thirds of the island.

The large pack numbered 15-17 wolves from 1959 through 1963. It reached an all-time high in 1964, when 22 wolves were seen. From 1964 to 1966, the pack remained large, at 15-20 wolves. In 1965, a pack of five appeared, believed to have separated from the large pack. This group may have persisted as a pack of four in 1966, although there was some speculation that it left the island in late winter 1965.

In 1966, the large pack initially numbered 15, but three wolves dissociated from the pack shortly after the winter study began. The alpha male, recognizable from 1964 to 1966, developed a limp and was apparently killed by other wolves in March 1966. For the remainder of the 1966 study, the largest group numbered eight wolves. The strong leadership of the alpha male was thought to be instrumental in the maintenance of the large pack, and its fragmentation was linked circumstantially to his demise.

When the 1967 winter study began, two packs (six and seven wolves) were found in the central and southwestern parts of the island. These two packs may have been remnants of the large pack since their travels overlapped considerably. Another pack of four occupied the northeastern end of the island. In February, a pack of seven, including four black wolves, was seen in Amygdaloid Channel, apparently having crossed the frozen channel from Ontario. There was evidence of violence among wolves; a wolf with a bloody head was seen running toward Canada when the "Black Pack" was first seen, and a few days later an injured wolf was observed near the lodge buildings at Rock Harbor. The packs of six and four were not relocated after the Black Pack was seen.

Between 2 and 7 February 1968, two black wolves were observed in a pack of six at the west end of the island. Another pack, numbering seven (the Big Pack) was first seen on 12 February and included one black wolf. While these packs could have been the same, Wolfe and Allen (1973) considered them distinct and suggested that the pack with two black wolves left the island via an existing ice bridge. The single black wolf in the Big Pack was probably one of the four black wolves first seen in 1967. How this wolf became integrated into a resident pack is unknown; Wolfe and Allen (1973) speculated that it could have been associated previously in some way with wolves in this pack, thus implying an additional interchange of wolves between the island and the mainland.

The Big Pack included three wolves that were recognizable from 1968 through 1970—the alpha male and female and the black wolf, a male. The alpha pair was observed mating in 1968. There was mutual courtship observed in this pair in 1969 and 1970, indicating probable mating. In all 3 years, the black male was often seen in the company of the alpha pair and seemed to enjoy special status. Consequently, he was designated the second-ranked, or beta, male. A photograph of the alpha female, taken in 1968 by D. L. Allen, revealed an unusual conformation in her left front leg. By 1972, she had developed a severe limp in this leg and was not seen the following year. The Big Pack persisted into the present study period and became known as the West Pack after the establishment of a second pack in 1972.

Annual Fluctuations, 1971-74

During the present study, the Isle Royale wolf population continued to increase from a low of 17 in 1968 to a high of 31 in 1974. In summer 1971, a second pack became established. During winter studies from 1972 through 1974, each pack occupied approximately half of the island.

1971. The Big Pack (hereafter referred to as the West Pack) was recognized in 1971 by the presence of the black male and alpha female. The black male was clearly the alpha male, replacing the large gray male that had been dominant from 1968 through 1970. Although ten members were seen twice, the pack usually numbered seven to nine wolves. On the basis of limited behavioral information, two pups were believed present in the pack. The pack ranged over the southwestern third of the island, venturing northeast as far as the middle of Siskiwit Lake.

In addition to the main pack, three duos were observed: one traveled among the peninsulas of the northeast end of the island, a second duo ranged from Moskey Basin through Chippewa Harbor to Wood Lake, and a third inhabited the shoreline of Siskiwit Bay, traveling between Houghton Point and Malone Bay.

Four single wolves also were recognized, with their respective activities centered at the southwest end, northeast end, the north shore west of Todd Harbor, and Malone Bay.

While the maximum number of wolves seen on a single day was 16, the presence of three duos and four singles was well established, and the population may be summarized as follows:

1972. Two packs (West and East) accounted for most of the island's wolves from January to March 1972. Each pack commonly numbered eight wolves in late January, but consistently numbered seven and ten, respectively, after mid-February. The ranges of these packs did not overlap—each occupied about half of the island.

In addition, a trio of wolves operated in the Malone Bay-Siskiwit Bay area, with tracks suggesting that they ranged along the shore of the island as far as Chippewa Harbor.

The East Pack had its origins within the wolf population present the previous winter, since there was no ice bridge to Canada in the interval between the winter studies of 1971 and 1972. Besides the alpha male and female there were six wolves that were uniform in size and body markings, virtually indistinguishable during observations or in photographs (Fig. 30). All six had the physical appearance of pups—presumably a litter from the alpha pair. This conclusion was supported by the fact that the alpha male was never observed chasing any of these six wolves away from the alpha female during the mating season.

Fig. 30. Alpha pair (female facing right) and six probable pups in East Pack, 1972.

In 1972 the alpha wolves in the West Pack were the same individuals as in 1971—the black male and small, gray female (Fig. 31). Although no other wolves in the pack were identifiable from 1971, the presence of pups was not confirmed. In spite of a limp, the alpha female was able to retain her dominant status, though occasionally she had trouble keeping up with the other wolves in the pack. I saw this female for the last time in May 1972, when she walked, still limping, along the shore of an inland lake with the black male; her summer coat was quite reddish. In September 1972, the black male, a smaller, reddish wolf (probably the alpha female), and three gray wolves were seen lying on an open ridge (Coley Thede, pers. comm.). The black alpha male and female apparently died between September 1972 and January 1973. The black male was then at least 6.5 years old, since he was first observed in 1967. The alpha female was also at least 6.5 years old when she died, because she mated in 1968 and had to be at least 22 months old at that time. She was probably older, since it is rather unlikely that she could have reached the position of alpha female by her second year.

Fig. 31. West Pack alpha pair (male is black), 1972. Note crooked front leg of alpha female.

On 24 February, a total of 20 wolves was observed (packs of seven and ten plus the trio). It was obvious from tracks on fresh snow that at least two single wolves were also present—one in the vicinity of Chippewa Harbor and one on the north shore near Little Todd Harbor. On 3 March, a wolf was seen following and apparently trying to remain hidden from the pack. Accordingly, the population totaled:

1973. The East and West packs, numbering 8 and 13 wolves, were again well defined. Spatial arrangements between the packs were similar, although their travels overlapped along the north shore of the island, where they visited each other's kills. On 24 and 25 February, a total of 23 wolves was seen. The final population estimate was:

The "Todd duo" was seen only three times. Judging from tracks, most of their activity was centered in the Todd Harbor area, although once they traveled from Little Todd Harbor to Lake Whittlesey. The loner, positively identified as a male, was seen only once, but tracks indicated that he ranged along the north shore from a point opposite Lake Desor to the northeast end of Amygdaloid Island, a distance of 40 km.

The leadership of the West Pack had changed completely since the previous year—a new alpha pair had replaced the black male and his limping mate. From their appearance and, especially, their behavior, four wolves in this pack were classed as pups. One of these disappeared from the pack around 20 February and was not seen again.

The East Pack numbered 12 or 13 for the entire winter study. Observations were hampered by the wolves' extreme avoidance of the study plane, probably resulting from disturbance by other aircraft earlier in the winter. The alpha pair had not changed from the previous year (Fig. 32). The number of pups was estimated from the increase in maximum pack size from 1972 to 1973—certainly a minimum figure since it assumes no mortality in the intervening year.

Fig. 32. East Pack alpha pair in 1972 (a), 1973 (b), and 1974 (c). Alpha female was the same wolf all 3 years; alpha male the same in 1972 and 1973.

1974. Both main packs increased in size from 1973, and a duo and at least one single were present. The 31 wolves observed on 17 February provided the following minimum count:

Again, each pack inhabited its respective end of the island, but movements of the East Pack into West Pack territory increased the amount of overlap. Two wolves again were active in the Todd Harbor area, quite possibly the duo of 1973.

In late January, the West Pack numbered 11 or 12. The increase in pack size from 1973 indicated the presence of at least four pups. Early in February, the pack broke into several smaller groups, and the alpha male, recognizable from 1973, was the dominant wolf in one group of four. His mate, the alpha female, was also in this group—there was some uncertainty that this was the same female as during the previous year. A single wolf was tolerated by this group near a kill, and it was probably one of the original pack members. Several days later, another group of four was seen leaving a kill in the interior of the island. Two other wolves, soon joined by a single, were observed in the Washington Harbor area, and this trio stayed together for the remainder of the study period. Since no other wolves were seen in the West Pack's range, it appeared that the West Pack had broken into units of 4, 3, 3, and 1. The 4 wolves, one group of 3, and a single wolf reunited in March, forming a pack of 8.

Winter Movements

On Isle Royale, wolf movements in winter vary from year to year, depending on snow conditions and the presence of shoreline ice. Extensive travel within a pack's range is necessary to locate vulnerable prey; such travel is lowest in years when vulnerable prey are abundant.

Travel Routes

Natural topography determines the ease of travel in different areas of the island, with the principal avenues for wolf movements consisting of chains of lakes, shorelines, old beachlines of Lake Superior, and bedrock ridges. The shorelines of the island stand out as principal hunting areas for wolves. Moose often seek conifer cover in winter, and since most of the conifer cover on the island is located in predominantly spruce-fir forests near lake level (Linn 1957), moose densities in midwinter tend to be highest along lakeshores. This creates an optimum hunting arrangement for wolves. Of a total of 325 wolf-killed moose located in winters from 1959 through 1974, 45% were within 200 m of either Lake Superior or Siskiwit Lake, a large interior lake.

The distribution of wolf-killed moose from 16 winter periods further suggests that some areas of the island produce more favorable hunting conditions than others. Kill density is obviously high in the area of North Gap (mouth of Washington Harbor), Malone Bay, Chippewa Harbor-Lake Mason, and Blake Point. All of these locations receive a high level of hunting effort in winter either because they are land masses lying between frozen lakes or bays, or because many travel routes intersect in those areas. Blake Point was hunted by a large pack in 1972 for the first time in several years, and perhaps a high proportion of vulnerable moose had been allowed to accumulate there. Other areas, notably the 1936 burn, have produced few kills in recent years, probably because of a gradual reduction in use of old burns by moose and unusually deep snow in several recent winters that restricted moose to more dense forest types.

Travel routes of the East and West packs during winters 1972-74 are shown in Figs. 33-35, along with locations of old and fresh kills. With the exception of the West Pack in 1974 (which fragmented and was impossible to track adequately), the routes shown represent continuous movements during the period of study. Variations in extent of travel and actual routes used are explained below in relation to snow and ice conditions.

Fig. 33. Travels and kills of the East and West packs, January-March 1972. (click on image for a PDF version)

Fig. 34. Travels and kills of the East and West packs, January-March 1973. (click on image for a PDF version)

Fig. 35. Travels and kills of the East and West packs, January-March 1974. (click on image for a PDF version)

EFFECT OF SHORELINE ICE

Travel around the perimeter of the island was extensive in 1972 and 1974, but quite reduced in 1973 (Figs. 33-35). In both 1972 and 1974, shelf ice was continuous around the island for most of the study period, and shorelines were used commonly by wolves (Fig. 36). In contrast, little shelf ice formed in 1973, and wolves had to travel onshore. Similarly when there was no shelf ice in 1969, wolves made extensive use of the interior even though snow was exceptionally deep (Wolfe and Allen 1973).

Fig. 36. Wolves frequently use shelf ice while traveling.

Occasionally, wolves venture onto ice that is very thin, especially if it is covered with snow. One morning in February 1974, the East Pack rested within 50 m of the edge of the shelf ice near Houghton Point. Tracks of one wolf led directly to the edge and then back to a resting place close to the other wolves. In the afternoon, the thin ice where the wolf had walked broke off and floated away.

EFFECT OF SNOW CONDITIONS

Relative to moose, wolves have a lighter foot loading (weight-load-on-track) and consequently receive greater support from snow of a given density. Weight load-on-track for five wolves in the Soviet Union ranged from 89 to 114 g/cm² (Nasimovich, 1955). In contrast, a cow and calf necropsied on Isle Royale in 1973 had foot-loadings of 488 g/cm² and 381 g/cm², respectively. Measurements by others range from 420 g/cm² to over 1000 g/cm², depending on the sex and age of the moose (Nasimovich 1955; Kelsall 1969; Kelsall and Telfer 1971).

Moose rarely receive consistent support from crusts on the surface or within the snow profile (Kelsall and Prescott 1971), and wolves often have a considerable advantage when crusts are strong enough to support them. For example, in 1972, wolves on Isle Royale appeared to be supported by a crust located 20 cm below the surface of the snow, although moose calves broke through and moved with difficulty. Crusting conditions and frequent thaws (which increase snow density) during the entire 1973 winter study allowed wolves to travel with relative ease throughout the interior of the island (Fig. 34). Similar conditions prevailed during the first half of March 1974. At such times moose usually remained in areas of conifer cover, and their movements seemed greatly restricted.

Since wolves have relatively short legs, they are greatly handicapped by deep, soft snow. Nasimovich (1955) found that wolves sank to their chests in snow of density 0.21 or less, which describes essentially all fresh-snow conditions. Thus, wolves generally travel in single file through snow, and have been observed moving into this formation in response to as little as 20-25 cm of snow along lake edges (Fig. 37). Nasimovich also found that wolves had difficulty chasing ungulate prey when snow depths exceeded 41 cm, and, with depths greater than 50-60 cm pursuit through untracked snow was almost impossible. In 1971, 41 cm of fresh snowfall on a 51-cm base precluded extensive travel by wolves in the interior of the island. Frequent fresh snow in 1972 kept depths in open areas above 75 cm, and, in spite of a crust within the snow profile, movements of wolves usually were limited to shorelines. The distribution of wolf-killed moose illustrates one effect of deep snow. When snow depth exceeded 75 cm, there was a significant increase in the number of kills located within 0.8 km of a shoreline, although part of this increase is related to changes in moose distribution.

Fig. 37. Wolves usually travel single-file through snow.

Distance Traveled by Wolf Packs

Since most pack movements involve hunting, the amount of travel should roughly reflect success and, indirectly, the relative abundance of vulnerable prey. Average distances traveled by Isle Royale packs between kills are quite variable in different years (Table 6), ranging from a low of 18.5 km/kill to a high of 54.1/kill.

TABLE 6. Travel estimates for wolf packs, 1971-74.

Highest travel per kill was shown by both East and West packs in 1973, a year when the average daily mileage was also highest for both groups. This suggests that moose vulnerability was lowest in 1973, a hypothesis supported by the fact that calves were killed least often in that year. Frequent snow crusts which made travel for wolves relatively easy contributed to the greater movements in 1973.

Minimum movement between kill (18.5 km) was registered by the West Pack in 1971, when wolves had little trouble finding vulnerable prey, especially calves, along shorelines. Even shorter distances were reported by Kolenosky (1972), who found that a pack traveled only 14.7 km between kills of deer in Ontario during 1969 when deep snow rendered deer more vulnerable and probably reduced wolf movements.

Food Economy of Isle Royale Wolves

Fundamentally, predator-prey interaction involves energy transfer from one trophic level to another, as from herbivore to carnivore. The complexity of this energy transfer is directly related to the number of species in a particular food web. On Isle Royale, the wolf, the major carnivore, is entirely dependent on moose and beaver, which are primary consumers of vegetation. This is a relatively simple system compared to the food web described by Cowan (1947) for the Rocky Mountain national parks of Canada, where wolves depended heavily on elk but also killed deer, moose, bighorn sheep, caribou, and, at certain seasons, snowshoe hare and beaver.

Isle Royale wolves prey on moose at all times of the year, while beaver are available only during the ice-free season. While quite variable from 1971 to 1974, the entire food base of the Isle Royale wolves was probably higher in the early 1970s than during the previous decade, owing to increased vulnerability of moose (at least in winter) and an increased beaver population. This is probably why the island was partitioned into two pack territories after 1971.

The Wolf as a Predator of Big Game

Food habits of wolves have been studied intensively, mainly because of human concern for the prey species, domestic or otherwise. Wolves are well adapted both physically and behaviorally for predation on large mammals, and an absence of large ungulate prey may adversely affect resident wolf populations, especially pups (see Pup Production).

Food habits of wolves seem to be most variable in tundra areas where wolves typically prey on a single ungulate species. Clark (1971) found wolves in central Baffin Island to be almost completely dependent on caribou, while Tener (1954) indicated snowshoe hares as the principal prey species for wolves on Ellsmere Island. Wolves denning in the northern Brooks Range in Alaska often ate small rodents, birds, fish, and insects, although Stephenson and Johnson (1972) believed that wolves nonetheless depended primarily on ungulates. Pimlott et al. (1969) pointed out that wolves have never been shown to thrive for a significant period on prey smaller than beaver, and most biologists agree that wolves are characteristically dependent on large mammals.

Moose, white-tailed deer, and beaver are the principal prey species of wolves in mainland areas adjacent to Lake Superior (Thompson 1952; Stenlund 1955; Pimlott et al. 1969; Mech and Frenzel 1971). Prey size and numbers determine which species are most important for the wolf. Where deer are available they are highly preferred (Pimlott et al. 1969; Mech and Frenzel 1971). Since beaver are small and, presumably, easily killed by wolves, predation on them is determined largely by availability.

Nonwinter Food Resources

Analysis of wolf scats collected in 1973 showed that Isle Royale wolves preyed on beaver to a much greater extent than a decade earlier (Fig. 38). A sample of 554 wolf scats was collected in 1973 from homesites and game trails used by both packs (Table 7). Most of these were from 1973, although a small proportion of the scats collected at the East Pack den were probably from 1972.

TABLE 7. Occurrence of prey remains in wolf scats, 1973.

Fig. 38. Scat analysis was used to determine summer wolf diet.

Beaver and moose calves together constituted 90.0% of the food items in nonwinter scats. Remains of beaver (hair and occasionally claws) were found in 75.8% of the total scat sample and made up 50.5% of 831 prey occurrences. Moose hair occurred in 69.3% of the scats and comprised 49.5% of the total food items. Of the identifiable moose remains (in scats deposited before the change in calf pelage in early August), 85.7% were from calves. Hare and bird remains were identified from only one scat and are unimportant as prey. While the 1973 blueberry crop was the best remembered by many long-time island residents, fruit was not found in any of the scats. Vegetation (mostly grass) was found in 6.1%, and unidentified seeds in 2.2% of the scats. These nonanimal items were not tallied in Table 7. Murie (1944) suggested that grass may act as a scouring agent against intestinal parasites, a hypothesis supported by his discovery of roundworms among blades of grass in some scats. An 18-inch section of tapeworm (Taenia sp.) was found in a fresh Isle Royale wolf scat containing grass, and Kuyt (1972) reported a similar finding.

INCREASED PREDATION ON BEAVER

The incidence of beaver remains in fresh scats from 1958 to 1960 was 13.1% (Mech 1966) (Fig. 39). In the following 3-year period beaver occurrence was essentially the same, 15.6% (Shelton 1966). Although there were no systematic scat collections in subsequent years, field examination of scats found incidental to other work showed no obvious changes (Jordan et al. 1967; Wolfe and Allen 1973). However, the 1973 data clearly demonstrate a significant increase in predation on beaver (t_s = 13.7, P 0.001) since 1958-63 (Table 8).

TABLE 8. Beaver occurrence in summer wolf scats, and beaver population trends.

In the decade between the scat analyses, the beaver population doubled, with the estimated number of active colonies (determined from aerial count) increasing from 140 in 1962 to 300 in 1973 (Shelton, unpubl. data). During the same period, wolf predation on beaver tripled, with percentage of beaver in wolf scats increasing from 14.4% to 50.5%.

Pimlott et al. (1969) found that the frequency of beaver in wolf scats in the Pakesley area of southern Ontario was 59.3%, compared to 7.1% in nearby Algonquin Park. A study by Hall (1971) showed that the beaver population in the Pakesley area was at least three times more dense than that of Algonquin Park. Clark (1971) pointed out that increased predation on beaver at higher densities could result from a shift in hunting effort to the more abundant beavers or simply from an increased frequency of encounters between wolf and beaver. Hall (1971) reported an increase in predation on beaver and a decrease in predation on deer in Pakesley during the 1960s, corresponding to changes in densities of these two prey species. He believed this indicated a shift in hunting effort.

On Isle Royale, known wolf trails often parallel water courses and pond edges, but we do not know whether predation on beaver is limited to chance encounters or whether purposeful hunting is important. Field observation did not indicate depression of beaver numbers around wolf homesites. Some of these were adjacent to active beaver ponds, suggesting that wolves did not spend a lot of time stalking and hunting beaver.

The relative levels of predation on moose and beaver as indicated by scats were consistent throughout spring-fall 1973 (Table 9). Shelton (1966) found a slight increase in beaver occurrence in wolf scats in the fall, when beaver are actively cutting winter stores of food and consequently are more vulnerable. Although scats from the last East Pack rendezvous did not indicate any increase, most of the scats from this area probably dated from August and early September, before intensive cutting begins.

TABLE 9. Incidence of beaver and moose remains in wolf scats from various homesites and associated trails, 1973.^a

Fig. 39. Beavers are an important prey for wolves in summer.

While it is difficult to estimate the current importance of beaver to Isle Royale wolves in terms of biomass or numbers of prey, a comparison with other studies provides a rough assessment. The highest reported occurrence of beaver in wolf scats came from studies in the Pakesley area of Ontario. Beaver remains were found in 62% of the scats examined in 1960, and beaver comprised 59% of the total food items (Pimlott et al. 1969). By 1964, the frequency of occurrence of beaver in wolf scats in Pakesley had increased to 77%, and beaver were regarded as the primary summer prey of wolves (Hall 1971; Kolenosky and Johnston 1967). The incidence of beaver in scats from Isle Royale wolves (76%) and the percentage of beaver in total food remains (51%) are second only to the reported data for the Pakesley area.

A high beaver population on the northeastern half of the island may have been an important factor allowing rapid growth of the East Pack (from 8 to 10 in 1972, to 16 in 1974). Over a 3-year period a minimum of 13 pups survived to midwinter in this pack. The general appearance in July of the 1973 pups and the rapid growth of at least two of them between observations in July and August suggest an abundant food supply. The dense beaver population probably has been an important factor ensuring high pup survival at a time when the production of moose calves, the other principal summer prey, was subnormal.

Winter Predation Patterns

Winter food habits determined from direct observations and aerial tracking showed that wolves on Isle Royale continue to subsist in winter almost entirely on moose. The snowshoe hare population was relatively low during this study, and while wolves occasionally flushed hares during observations, they never gave chase. No indications of wolf predation on hares in winter were found. Beaver were available only in rare instances when they ventured from beneath the ice to cut food. During mild weather between January and March 1973, we discovered two wolf-killed beaver. Likewise, during a thaw in March 1974, one or more beaver were killed on the Big Siskiwit River.

Although wolves rarely find active beaver in winter, they show great interest in beaver lodges and dams encountered during their travels (Fig. 40). The East Pack even dismantled a lodge in February 1973, near Harvey Lake. The wolves had killed two moose within 100 m of the lodge, and their activities while in the area for several days centered on digging out the lodge.

Fig. 40. Wolves investigating beaver lodge.

HUNTING SUCCESS

In forested regions such as Isle Royale, wolves depend heavily on their sense of smell for prey detection. Of 30 observations of wolves detecting moose from 1972 to 1974, it was possible to determine the method of prey detection 17 times. In 10 cases in which wolves caught the scent of a moose, they either approached directly upwind or turned toward their prey after crossing downwind from the moose. Mech (1966) reported that wolves seemed to sense prey 2.4 km away, underscoring their olfactory sensitivity. Wolves visually detected moose six times, and once they followed a fresh moose track to the animal.

Mech (1966, 1970) provided an extensive discussion of the results of moose-wolf encounters observed in the first three winters of the project. The basic pattern he observed has not changed significantly. Moose that stand their ground when wolves approach are not killed; all observed encounters on Isle Royale that ended in a kill occurred after the victim initially ran from wolves. For unknown reasons, vulnerable moose do not stand and face wolves when first approached. While chasing a moose, wolves apparently respond to vulnerability cues that are not obvious to aerial observers; sometimes they quit immediately, at other times the chase might last for long distances (Fig. 41).

Fig. 41. East pack in successful chase. Victim was 16-year-old cow with arthritic hip and vertebrae and was pulled down within 2 minutes. Alpha wolves indicated.

The primary point of attack is the hindquarter region of the moose, where wolves can dash in and out and stand the best chance of avoiding the quick strikes of the hooves. When wolves inflict serious wounds, they are often content to wait until the moose weakens. In February 1972, however, the East Pack, after spending most of a night close to a wounded adult near Lake Richie, abandoned the animal around daybreak. The moose continued to stand in heavy cover that morning, but by afternoon was lying on its side. This was an 8-year-old cow in apparently good condition, with abundant marrow and visceral fat reserves and pregnant with one fetus. She had deep wounds around the anal opening and had apparently lost a considerable amount of blood. In the next 5 weeks, the East Pack never returned to the carcass, but by early May the wolves had consumed it entirely.

Mech (1966) found that wolves have a low rate of hunting success, presumably because most of the moose they encounter are not vulnerable. Of 77 moose tested by wolves, 6 were killed. From 1972 to 1974, 38 moose were tested during observations, and only one was killed. A schematic representation of results of moose-wolf encounters in the two periods is presented in Fig. 42. Observations of hunts in the recent period were too few to determine changes in hunting success.

Fig. 42. Results of observed moose-wolf encounters on Isle Royale, 1959-61 (Mech 1970:210) and 1972-74 (this study).

EFFECT OF SNOW CONDITIONS

Crusts within the snow profile or on its surface provide support for wolves but interfere with moose movements. Since crusts are frequent on Isle Royale, deep snow often results in increased hunting success for wolves. This was apparent in 1969 when more kills were found than in any previous winter (Appendix L). An increased kill rate on Isle Royale was also evident in the "deep snow" winters of 1971 and 1972. In all three of these winters, the degree of carcass utilization was noticeably less, indicating higher hunting success (Wolfe and Allen 1973; Peterson and Allen 1974) (Fig. 43).

Fig. 43. Carcasses of wolf kills were often underutilized in the early 1970s.

Increased calf vulnerability due to reduced mobility in deep snow is reflected in a high kill of moose calves when snow depths exceed 75 cm (Fig. 44). Most calves are killed near shorelines, which are traveled heavily by wolves when snow is deep and shelf ice present. Calves may be so restricted that they are left in shoreline areas by their mothers who have gone elsewhere to feed. In 1972, the West Pack encountered two adults and a calf on the south shore. Both adults ran along the shore, but the calf headed inland and was pulled down by wolves within 100 m. Either the calf's mother was behaving in a highly abnormal fashion or she was not present. In 1971, we saw two calves without a mother present; one of these was killed by a single wolf (Peterson and Allen 1974).

Fig. 44. East Pack (1973) in unsuccesful chase. Note cow at rear of calf.

WINTER FOOD AVAILABILITY

Estimates of food consumption by wild wolves usually are derived by multiplying the average weight of prey by the number killed in a specific period (Mech 1970). When calculated in this manner, food availability rather than actual consumption is estimated, since utilization of carcasses varies considerably with the size of the pack, size of prey killed, and the ease with which additional prey may be taken (Fig. 45). In winters when moose are more vulnerable to wolves, the kill rate may go up, while the corresponding degree of carcass utilization declines. The calculated availability of food for Isle Royale wolves is most useful for comparisons of hunting success.

Fig. 45. Wolves feeding on carcass.

Whole weights of several Isle Royale moose (Appendix E) provided the basis for estimates of the potential food contributed by each bull, cow, and calf (assumed average whole weights of 432, 364, and 159 kg, respectively). The primary inedible portions of a moose are the stomach contents and some of the hide and skeleton. The stomach of a 400-kg bull necropsied in February weighed 65 kg, or about 16% of its body weight. Inedible stomach and intestinal contents of adults were assumed to weight 68 kg, and an additional 34kg were subtracted for portions of hide and skeleton usually left uneaten. Stomach and intestinal contents of calves were assumed to weigh half those of an adult, or 34 kg. Although wolves sometimes eat the entire skeleton and hide of calves in winter, 11 kg were subtracted for parts usually left uneaten. Thus the potential food of each bull, cow, and calf when killed by wolves is a calculated 330, 261, and 114 kg. Adults of unknown sex were assumed to contribute 295 kg. Carcasses of moose collected for necropsy were consumed by the West Pack; weights of these carcasses were included in the calculations for the West Pack on the assumption that the wolves would have otherwise killed a moose themselves (Fig. 46).

Fig. 46. Wolves bedded near kill.

From 1971 through 1973, calculated availability of food for both East and West packs varied between 6.2 and 10.0 kg/wolf/day, while in 1974 daily figures for both East and West packs dropped to 5.0 and 4.4 kg/wolf, respectively (Table 10). The drop in availability of food in 1974 stems largely from the increase in pack sizes from 1973 to 1974 and the fact that a high percentage of the kills were calves. Winter availability of food on an individual basis declined for each pack through the period of study, partially reflecting a decline in ease of prey capture from winters in 1971 and 1972 when unusually deep snow contributed to a high kill-rate.

TABLE 10. Estimates of food availability for West and East packs, 1971-74.

Food available to pack members from 1971 through 1973 on Isle Royale was greater than that indicated for the former large pack (Mech 1966). Using the moose weights given above, that pack had available 4.9, 3.8, and 5.1 kg/wolf/day in 1959, 1960, and 1961, respectively.

Food available to Isle Royale wolves is well above the minimum amount required in winter. Mech (1970) estimated the daily food requirement for a wild wolf at about 1.7 kg on the basis that active domestic dogs need about this amount. Growing wolf pups and captive adults can be maintained on this amount of food (Kuyt 1972; Mech 1970). Food availability for an Ontario wolf pack was estimated at 3.7 kg/wolf/day during one winter season (Kolenosky 1972). A Minnesota wolf pack increased after a winter with 5.8 kg/wolf/day of available food, remained the same size at 3.6 kg/wolf/day, and decreased at 3.4 and 3.0 kg/wolf/day (Mech, in press).

The food economy of loners (single wolves) and small groups is difficult to study because the extent of their movements and feeding patterns is usually unknown. Although Jordan et al. (1967) described some loners as "gaunt" and implied that most led a rather tenuous existence, this may not be the case in winters of abundant prey. For example, in 1971 a loner subsisted for several weeks on three moose carcasses in the Malone Bay area and apparently moved very little. Likewise, the Todd duo killed two moose and fed on two old kills in a 15-day period in February 1974, rarely moving out of the Todd Harbor area.

Long-term Changes in Food Resources

The appearance of two socially stable wolf packs on Isle Royale was not observed prior to 1972; this appearance, presented earlier, probably resulted from an increase in the food base of the wolf population. The beaver population increased in the 1960s, as did wolf predation on beaver during the nonwinter months. Since production of moose calves was noticeably lower in recent years than in the early 1960s, beaver assumed a position of significance by supplying food during the critical pup-rearing season.

While the moose population also appeared to increase during the 1960s, this would not in itself provide an immediate increase in prey for wolves. The food supply for wolves depends on the density of vulnerable moose rather than absolute moose densities. Thus, a moose population in the early stages of a natural decline may provide wolves with a maximum number of available prey. This was apparently the case on Isle Royale in the early 1970s.

The establishment of the East Pack has probably brought about a greater utilization of prey within this territory, where previously only loners or packs of two or three wolves lived. For example, the number of moose killed on the northeast half of the island during the winter study period increased greatly from 1971 to 1972, after the appearance of the East Pack (Fig. 47). In its first three winters of operation, the pack killed nine moose on the Blake Point peninsula, about 8.5 km² in area, during a total of 18 weeks of aerial tracking. Ground search turned up eight additional kills on this peninsula. Moose densities in midwinter in this area commonly exceed 4-6/km².

Fig. 47. Distribution of wolf kills located during winter study before (1971) and after (1972) the establishment of the East Pack. (click on image for a PDF version)

Intraspecific Behavior

Like physical characteristics, an animal's behavior has been shaped by rigorous selection pressures, resulting in behavior patterns that are closely adapted to a particular function in the ecosystem. A comparison of the red fox and wolf provides a simple illustration. While both are canids, foxes exhibit much less diversity in behavioral expression and communication than do wolves (Fox 1970). The fox, a semi-solitary creature, preys extensively on game smaller than itself and, at certain seasons, depends heavily on plant fruits and carrion. Therefore, in terms of food acquisition, there would be no advantage for young foxes to remain with their parents in a social group. The behavioral repertoire of foxes is less diverse, yet sufficient for its more solitary way of life. Cooperation among wolves in a pack, however, is essential to their ecological role as a predator on large ungulates. Consequently a complex dominance hierarchy and elaborate array of behavioral expression have evolved among wolves, allowing them to live in close association as group-hunting carnivores.

The organization of wolf populations into packs does not fully explain wolves' diverse means of expression and communication, since other group-hunting canids, notably the bush dog (Speothos venaticus) of South America and the African wild dog (Lycaon pictus), do not exhibit a similar, high level of behavioral expression (Fox 1971; Kruuk 1972). While little is known of the ecology of the bush dog, the wild dog of Africa exists year-round in a cohesive pack, with social bonds apparently maintained by highly ritualized food-begging behavior (K¨hme 1965). Individuals within a wolf pack are frequently separate, however, especially in summer when most hunting is done individually or in small groups. This led Fox (1971) and Kruuk (1972) to suggest that the well-developed means of expression among wolves is not only important in coordinating group activities and maintaining order in the pack but also provides for more effective reintegration of individuals into the group after separation.

The territorial nature of wolf packs helps maintain pack integrity and seems to apportion space among resident packs according to the availability of food. Mechanisms of territory maintenance may include scent-marking and howling and agonistic behavior during rare confrontations between packs.

In spite of extensive field studies of the wolf, many generalizations concerning behavior within and between packs are poorly documented in the wild, primarily because observations are hampered by the wolf's environment and mobility. Lengthy ground observations have been possible only at den sites in tundra regions (Murie 1944; Haber 1968; Clark 1971); aerial observations, a primary research tool, are limited in scope. Insight into the ecological significance of wolf behavior patterns and a proper appreciation of their variability can be gained only by intensive study of many packs in different ecological settings.

Social Hierarchy Within Packs

The basic social structure of wolf packs is well understood from studies of captive wolves (Schenkel 1947, 1967; Rabb et al. 1967). Behavioral interaction within a pack occurs in a framework of dominance relationships or social hierarchy. A dominant (or alpha) male and female are the central members of a pack, and the other wolves constantly reaffirm their subordinate status through postures of submission directed toward the dominant individuals. Males and females have a separate dominance ranking, and the subordinates have definite dominance relationships among themselves, although interaction is less frequent and relationships are less well defined.

Aggression is channeled into ritualized behavior patterns within the dominance framework, reducing the amount of direct conflict within the pack and promoting social order and stability. Alpha wolves provide leadership during travels of the pack, initiate many pack activities, and sometimes exert considerable social control over activities of subordinate wolves, notably their sexual behavior.

Restriction of courtship behavior among subordinates, together with well-developed mate preferences among adults, is thought to reduce the potential number of breeding pairs in a pack, often resulting in the birth of only a single litter. The whole pack participates in gathering food and caring for the young, and this contributes both to the survival of young and cohesion within the pack.

EXPRESSION OF DOMINANCE AND SUBORDINATION

Facial expression, tail position, and posture combine to indicate subtleties of mood and desire. These indicators provided the basis for determining the social position of certain wolves in the packs on Isle Royale, especially the alpha wolves (Fig. 48).

Fig. 48. Alpha female and male lead West Pack, 1972. Note raised tails of dominant wolves.

Tail position is easily seen from the air and is thus an obvious indicator of wolf status. The importance of the tail in communication probably lies in the fact that the hindquarters and anogenital region have a considerable function in olfactory and visual expression (Kleiman 1967; Schenkel 1947). Presentation of the anal region by a raised tail indicates a position of dominance, while a lowered tail (during interaction with other wolves) covering the anal region, is a component of submissive behavior. Postural changes reinforce these expressions: a dominant wolf stands erect with tail raised, while an extremely subordinate animal may pull its tail between its legs and lower its rear end to the ground (Figs. 49, 50).

Fig. 49. Alpha female (Brookfield Zoo) follows subordinate female. Note posture and head and tail position.

Fig. 50. Alpha male (East Pack, 1974) stands in dominant display next to subordinate wolf.

While the movements and positions of the ears, eyes, forehead, nose, and mouth of a wolf can be combined to produce subtle variations of expressions (Schenkel 1947), most are not observed by humans except at close range. Ears of dominant wolves are forward, while those of subordinate wolves are turned back or flattened against the head. Teeth are more exposed as the intensity of a threat increases. Wolves of high social standing often stare directly at another wolf as part of an expression of dominance or a mild threat, and subordinates respond by turning the head away and avoiding direct eye contact.

Inferior wolves constantly show submissive behavior toward dominant wolves. Schenkel (1967:324) defined submission as "an impulse and effort of the inferior towards friendly and harmonic social integration." He described two basic types of submission in wolves, "active" and "passive." During active submission,

. . . the posture is slightly crouched, the tail is low, and the ears are directed backwards and lie close to the head. . . . The inferior pushes the muzzle of the superior with his nose, licking it tenderly without any pressure. In addition, he may perform pawing movements—lifting one forepaw and moving it in the direction of the superior, or, while making little steps, tapping the floor with his forepaws alternately. Often the tail is wagging sideways, and sometimes the whole hindquarters are also swinging. (Schenkel 1967:322-323).

Active submission often is seen as an element in greeting behavior, and is the most obvious form of expression during the "group ceremony," described below. Passive submission is usually shown by an inferior wolf in response to a threat from a superior individual:

The inferior wolf lies half on his side and half on his back, exposing the ventral side of his chest and sometimes the abdomen. The latter occurs regularly as a reaction to olfactory investigation in the genital region. The ears are directed backwards and lie close to the head. The tail is more or less bent ventrally so that it passes between the thighs. The often enthusiastic and friendly activity of active submission is reduced. Sometimes the tail is wagging sideways with extremely reduced amplitude and the activity is replaced by a passive posture full of trust, devotion, and demonstrated helplessness (Schenkel 1967:323).

An important element in passive submission is "inguinal presentation," in which the wolf lying on its side raises its hind leg, thus exposing its inguinal region to the dominant wolf. Passive submission, and inguinal presentation in particular, seem to inhibit aggression in dominant wolves and thus are considered appeasement or "cut-off" gestures (Fox 1971). Many times on Isle Royale, active agonistic behavior, or even mild threats, from a dominant wolf caused subordinate individuals to fall into passive submission. The dominant wolf usually would reduce the level of its threat, and either investigate the prone wolf or simply stand over it for a minute or more. Slight movement by the inferior wolf usually brought a quick snap from the dominant. The subordinate wolf usually lay still, often with hind leg raised, until the dominant wolf walked off. Once, in the West Pack, a subordinate wolf maintained a position of inguinal presentation after the black alpha male walked away, and even rolled over and raised the other hind leg when the alpha male wandered behind him. Any other movement by the inferior male brought immediate punishment from the alpha male.

Members of a pack often congregate in a "group ceremony," a greeting centered around the alpha animals. Subordinates crowd around the dominant wolves and show exuberant active submission and much body contact. Group ceremonies were observed 34 times among Isle Royale wolves from 1972 to 1974. Most commonly, they occurred immediately after the pack arose from sleep, or when one or several members returned to the pack after a brief absence. Frequently active submission toward an alpha by one wolf brought the rest of the pack running over to join in the proceedings, and sometimes a group ceremony ensued when wolves clustered about an alpha inspecting an inferior wolf lying on the ground. Group ceremonies also were seen when a pack "regrouped" after an unsuccessful chase of a moose. Such ceremonies often terminated with threats directed toward an inferior wolf by an alpha, perhaps in response to overenthusiastic greeting behavior. Group ceremonies provide a means of reaffirming dominance relationships, probably reinforcing both the status of alpha wolves and existing social bonds. Additionally, they may provide reassurance for pack members at critical periods; for example, when the East Pack traveled outside of its normal territory in 1974, subordinate wolves constantly crowded about the alpha wolves in a group greeting.

PACK LEADERSHIP

Alpha wolves sometimes retain their dominant position for several years and may be instrumental in maintaining a stable pack. Jordan et al. (1967) recognized the alpha male in the large pack on Isle Royale from 1964 to 1966 and found that pack formation in 1966 coincided with his death. There has been relatively little turnover in the alpha positions in the West and East packs (Fig. 51). The small, gray female with the deformed left front leg held the alpha position for at least 5 years (1968-72) in the Big Pack (West Pack). The black male was associated with this female during all 5 years, apparently first as a subordinate (beta) male with special privileges allowing him to travel and rest near the alpha pair (Wolfe and Allen 1973), and finally as alpha male in 1971 and 1972. This was the only case from Isle Royale in which the previous history of an alpha animal has been known.

Fig. 51. Changes in leadership of the large packs on Isle Royale.

None of the recognizable alpha wolves on Isle Royale has been seen after a known change in its dominant status, but whether their deaths preceded or followed the change in leadership is unknown. Jordan et al. (1967) found circumstantial evidence that the alpha male in the large pack in 1966 had been killed by his associates. During the present study, three alpha wolves disappeared; all three were last seen in summer. While the alpha male in 1966 apparently was killed after he developed a limp, the alpha female in the West Pack in 1972 managed to maintain her dominant status in spite of a limp which occasionally prevented her from retaining her customary position at the front of the pack.

Winter observations on Isle Royale indicated that alpha wolves usually led the pack during its travels. Of 61 cases in which it was possible to determine whether the alpha male or female led the pack, an alpha wolf was first in line 70% (n = 43) of the time. In 33 cases the alpha female was first, the alpha male led in 6 cases, and 4 times the two dominant wolves were side by side. In many instances the alpha male showed obvious sexual interest in the alpha female and consequently followed her.

Alpha wolves, usually at the front of the pack, normally choose the direction of travel and specific travel routes. This clearly was the case during an observation of the East Pack in 1974. The alpha female led the pack through the narrows between Wood Lake and Siskiwit Lake, then lagged behind to sniff an old moose track. Other wolves then assumed the lead position until they reached the first peninsula, where they stopped and waited for the alpha female to move to the front. She immediately set the direction of travel, led the way briefly, then fell back into a position in the middle of the pack. The same procedure was followed at the next point of land.

A clear example of decision-making on the part of an alpha animal was observed in 1974 when the East Pack encountered a scent post of the West Pack. After the pack had examined the scent mark, the alpha female reversed the direction of travel and led the pack back to more familiar range.

In the absence of alpha leadership, subordinate wolves may be indecisive. Once in 1972 we observed six probable pups in the East Pack by themselves when the alpha pair had dropped back several miles. Twice a wolf stopped and watched its back-trail. When the six wolves emerged on the shore of an inland lake, they vacillated for 15 minutes, sniffing snowed-in tracks and making false starts, and finally all started off in the same direction.

Alpha wolves appear to provide leadership at critical times such as hunting, encountering novel stimuli, and perhaps when contacting neighboring packs. The position of the alpha wolves was observed in only six encounters with moose, but an alpha wolf led the pack in four of these cases. Certainly the wolves at the front of a pack would be the first to detect and chase prey.

In February 1973, when the West Pack was under observation at a moose carcass across the harbor from Windigo, the alpha male detected us. He trotted excitedly toward shore, then back to arouse other members of the pack. After a group ceremony centered around him, he led the pack into thick cover near shore.

Later in the same winter we watched the East Pack file along the ice on the north shore of Rock Harbor. When they reached open water, they moved onshore and slowly worked their way long the slippery, ice-coated shoreline, finally congregating on a small point. Rounded chunks of broken, "pancake" ice ranging up to several feet in diameter had frozen loosely together adjacent to the point. One wolf reached out with a foot and pushed on an ice chunk, withdrawing its foot quickly. Next the alpha male, together with an unidentified wolf, walked out on a large chunk of ice and stood for a few seconds. Suddenly they bolted back to shore, apparently after the ice had shifted. The alpha male then led the pack away, continuing the course onshore.

No observations were made of confrontations between different packs of wolves on Isle Royale. However, we might expect that the alpha animals would take a leading role in such a situation, much as they did when an alpha wolf led the pack in chasing a fox in two observed cases. Although the alpha male and female did not lead the East Pack during their first foray out of their territory on 15 February 1974, they were obviously key figures. As the pack ventured across Siskiwit Bay, most of these wolves probably were encountering the area for the first time, since this pack was not observed southwest of Malone Bay in the previous two winters. In addition to an unusual amount of scent-marking as they crossed to Houghton Point, the subordinate wolves were constantly clustered around the dominant pair in a sort of mobile group ceremony. Perhaps the intense, active submission directed toward the alpha pair resulted from uncertainty and excitement among subordinate wolves.

Courtship and Breeding

Largely because of complex social relationships, such as mate preferences and a dominance hierarchy, the breeding potential of a wolf pack rarely is realized. Studies of captive wolves have demonstrated that breeding within a pack usually is limited to a few animals (Rabb et al. 1967), and a similar situation has been observed in packs on Isle Royale. In studying wild wolves, it should be remembered that sexually immature pups may account for a sizable proportion of a pack, a partial explanation for limited breeding activity.

Mate preferences are recognized clearly on Isle Royale, and incidence of courtship among Isle Royale wolves indicates that mating is most likely to occur between the dominant male and female in a pack (Fig. 52, Table 11). It is also obvious that alpha wolves interfere with courtship attempts of subordinates. These topics will be presented in depth later in this section.

TABLE 11. Minimum number of breeding pairs present, 1971-74.

Fig. 52. Incidence of courtship behavior between wolves of different social rank in the East and West packs on Isle Royale. 1972-74.

The primary function of courtship behavior is to establish and maintain a pair bond. Unlike many vertebrates, male wolves play an integral role in feeding and raising the pups, and a close relationship between a male and female wolf remains important on a year-round basis. Although there are records of more than one litter born in a pack (Murie 1944; Haber 1968; Clark 1971), one litter per pack is usually the rule (Van Ballenberghe and Mech 1975) and probably a safe assumption when pack sizes are not large. Other adults in a pack help raise the pups, enhancing their chances of survival. On this basis we can predict that natural selection would favor offspring from dominant, breeding wolves that interfered with mating attempts of subordinates in a pack.

In species with well-developed threat behavior, such expressions are well hidden during courtship, since they would be detrimental to the formation of a close relationship between male and female (Eibl-Eibesfeldt 1970). Consequently, courting wolves display much greeting behavior, play soliciting, and submissive postures, all of which tend to decrease "social distance" (Fox 1971) (Fig. 53).

Fig. 53. Play behavior is an integral part of wolf courtship (Brookfield Zoo).

Courtship behavior was observed among Isle Royale wolves throughout the annual winter study periods, with the peak in sexual activity usually sometime in February. During 52 hours of aerial and ground observation in winters from 1972 through 1974, courtship behavior was recorded 71 times. One "instance" of courtship behavior consisted of a well-defined behavior or sequence of behavior, such as mounting, a mutual greeting between mates, etc. Behavior patterns which were considered as courtship in at least some contexts are described in Appendix F. Most of the courtship behavior recorded consisted of males mounting females or males examining the genital region of females (genital snuffling). Undoubtedly, these behaviors are somewhat overrepresented because they are so easily recognized. Greeting and play behavior were also commonly seen but were not recorded as courtship unless there were other indications of sexual interest.

Subtleties of behavioral expression are not seen readily from aircraft. When observing wolves whose sex, age, and relationships are unknown or poorly understood, some ambiguous behavior is difficult to classify. For example, it was not uncommon to see one wolf approach another with tail flagged, posture erect, ears and eyes forward, and the second wolf walk off in a generally submissive posture, tail tucked between its legs. This usually indicated a dominance display, but similar behavior was seen when a male tried to court an uncooperative female. Such ambiguities were resolved by carefully watching for subsequent interaction between the same individuals and their relationships to other wolves. Fortunately, most displays of dominance, submission, and courtship involved recognizable alpha wolves and were interpreted with little difficulty.

MATE PREFERENCE

Mate preference can be a powerful limitation on the amount of breeding within a pack. Clearly, if there is no mutual courtship between a male and female wolf, a mating between the two is unlikely. The breeding potential of the Brookfield wolves was reduced considerably by such "one-sided" courtships (Rabb et al. 1967).

Pair bonds between mates may be very stable from year to year, although wolves will mate with other individuals if their preferred mate is not available (Rabb et al. 1967). Wolfe and Allen (1973) indicated a stable pair bond between the alpha male and female in the Big Pack (West Pack) from 1968 through 1970. This male had disappeared by 1971, but the same female mated in 1971, and presumably in 1972, with the black male that assumed the alpha position. The East Pack provided another example of a female accepting a new mate after the probable death of the alpha male. In 1974 the new alpha male courted the incumbent alpha female, who accepted his approaches with friendly greetings, indicating probable receptiveness. In all the packs observed from 1971 through 1974, the alpha pair either mated or showed mutual courtship and was considered a bonded pair.

Studies of the Brookfield wolves (Rabb et al. 1967) showed that both males and females sometimes courted members of the opposite sex that were unreceptive, and in these cases courtship action was ignored or rebuffed with threats (Figs. 54, 55). My own observations at Brookfield indicated clear differences between the behavior of females that simply were not ready for copulation and those that were totally rejecting a male. A female that temporarily was rejecting a male responded to his advances with mild threats, or simply pulled away, and elements of greeting and play behavior were still seen between partners. This was typically the situation between alpha males and females on Isle Royale, and also a subordinate pair in the West Pack in 1972 that eventually mated. However, a female that was unreceptive to a particular male responded to his courtship attempts with obvious threats and showed little friendly behavior, except perhaps in the context of a group greeting ceremony. A subordinate male in the West Pack in 1973 frequently showed interest in a female, but she always replied with aggressive snapping, never exhibiting any friendly behavior toward the male. A mating between these wolves seemed unlikely.

Fig. 54 Female accepting a courting male (Brookfield Zoo).

Fig. 55. Female rejecting a courting male (Brookfield Zoo).

Little is known of the development of mate preferences among wild wolves, but among Brookfield wolves there were strong indications that future mate preferences are crystallized during the juvenile period (prior to sexual maturity at 22 months). Also, it appears that a young wolf generally develops a preference for the alpha wolf of the opposite sex, or at least a dominant individual (Rabb et al. 1967; Woolpy 1968). It is significant that the alpha female in the West Pack in 1971 accepted the black male as her new mate—a wolf that had enjoyed a close relationship with the alpha pair for at least 3 years.

RESTRICTION OF SEXUAL BEHAVIOR AMONG SUBORDINATE WOLVES

During three breeding seasons on Isle Royale, 69% of the observed courtship behavior (n = 71) occurred between alpha wolves. Studies of captive wolves have shown that dominant wolves restrict and, in some cases, eliminate courtship behavior and mating among subordinates (Schenkel 1947; Rabb et al. 1967; Woolpy 1968). Since the reduction of mating among subordinate adults could contribute to population regulation, it is important to try to determine the effectiveness of such restrictions among wild wolves.

Rabb et al. (1967) noted an increase in agonistic behavior between dominants and subordinates during the breeding season of Brookfield wolves (Fig. 56). Observations in February on Isle Royale indicated frequent threats to subordinate wolves by the alpha male and female. In many cases a strong assertion of dominance seemed to be stimulated directly by courtship behavior among subordinates. This was further indicated by the lack of overt threats from the alpha male in the East Pack in 1972, when the pack was believed to consist primarily of an alpha pair and their offspring; the latter would have been sexually immature in their first winter. In 1973, however, when pups of the previous year could have been sexually mature, on two occasions the alpha male chased other wolves away from the alpha female. Both instances occurred on 16 February when frequent genital sniffing by the alpha male suggested that his mate was in heat.

Fig. 56. Alpha male attacking (with inhibited bites) a subordinate male mounting a female (Brookfield Zoo).

The West Pack provided the best opportunity to record interference of alpha wolves in the sexual behavior of subordinates. The black alpha male in this pack was very possessive of his mate, the alpha female of long standing (Fig. 57). One of my first observations of this pack was from the ground at Windigo on 29 January, 1972. As they rounded Beaver Island, there was much playful sparring as the pack moved along the ice, and at one point a subordinate wolf mounted the alpha female. She eventually squirmed away and snapped at the other wolf, and this brought the black male on a run. He knocked the subordinate over with a body slam, and then mounted the alpha female himself. This was typical of his behavior when other wolves approached his mate.

Fig. 57. Black alpha male stands between alpha female (behind him) and subordinate male (head turned away), West Pack, 1972.

The most interesting interaction in 1972 spanned several days, beginning 24 February. A subordinate pair managed to stay in the West Pack and mate in the presence of the alpha pair, in spite of repeated punishment from both the alpha male and female. Identification of the subordinate pair was not always positive: the male was a thin-tailed wolf that looked like one other wolf in the pack, and the female was one of three full-tailed wolves in the pack. This complicated the interpretation of observations made at different times, but since there was never any indication of sexual interest in more than two subordinate wolves, in the following account from my field notes, it will be assumed that the thin-tailed male and the full-tailed female were consistently the same individuals.

February 24: The West Pack (seven wolves) was traveling southwest along the shoreline of the island opposite L. Halloran at 5:15 p.m., when we began observations. The alpha female was limping very noticeably and had fallen to the rear of the pack, accompanied by the alpha male. The full-tailed female assumed the lead position for most of the observation period of 45 minutes. Once she squatted and urinated on the ice, and the alpha female accompanied by her mate, walked over and inspected the urine mark. The other wolves crowded around and likewise sniffed the spot. Then the full-tailed female joined the group and, when approached by the alpha female, immediately dropped on her side in passive submission. The alpha female went back to the urine mark for an additional sniff, and the full-tailed female arose, only to be forced back down on her side by the alpha female. When the pack began to move on, the full-tailed female again assumed the lead position, the alpha pair falling to the rear. Since the alpha female was almost always in the front of this pack when it traveled, it appeared that the full-tailed female might have been punished for assuming the lead position. Perhaps the urine mark of the female indicated an estrous condition, which might have precipitated the punishment. We followed for another 30 minutes as they encountered several moose along the shore, finally killing a calf.

February 26: The pack (seven) left the kill made 2 days earlier and traveled around the western end of the island. We watched them from 4:20p.m. to 5:40p.m. as they rounded the western-most point of land and entered Washington Harbor. In the first 55 minutes of observation, the alpha female "punished" the full-tailed female four times, forcing her over on her side. During this time the thin-tailed male sniffed the genital region of the full-tailed wolf, and they engaged in a brief nose-greeting as they traveled. The wolves were spread out on the ice, with the subordinate pair usually in the front and the alpha pair near the center. At the mouth of Washington Harbor, the thin-tailed male mounted the full-tailed female briefly, causing the alpha pair and the rest of the pack to run over. After the alpha female "punished" the full-tailed female their identification became uncertain in the flurry of activity which ensued. Suddenly four subordinate wolves, including the full-tailed female, broke from the group and ran toward Card Point, 300 m to the south, pursued by the alpha pair. Three of the four wolves stopped when they reached the point, but the full-tailed female ran across the point, around its western end, and then doubled back across the mouth of Washington Harbor. The alpha pair joined the three wolves on the shore of Card Point briefly, but when the full-tailed female came into view, the alpha female took off on a run after her, accompanied by her mate, the black male. The rest of the wolves joined half-heartedly in the chase. The full-tailed female maintained a long lead on the alpha pair, however, and after running across the point on the north side of the mouth of Washington Harbor, the alpha pair and the rest of the pack slowed to walk. They continued to track the full-tailed female, who after running 4.0 km, was still trotting, with frequent glances over her shoulder, when darkness forced us to terminate observations.

February 27: Six wolves were observed near the carcass of a moose near Windigo. The alpha pair and two others walked out of view onto shore, leaving a thin-tailed wolf standing near the carcass and a full-tailed female feeding voraciously on the remaining bones, suggesting a late arrival. It seemed reasonable that this was the subordinate pair, since the thin-tailed wolf pawed and sniffed at the genital region of the full-tailed wolf after the others had left. The pack still was missing one wolf.

February 29: The pack had traveled 37 km to Houghton Point, where they killed a moose and were seen in early afternoon resting nearby. Only six wolves were seen, and only the alpha pair was identified.

March 2: In the afternoon all seven wolves of the West Pack entered Washington Harbor, on their way to the bait carcass. The full-tailed female and thin-tailed male led the pack, traveling close together. At 6:00 p.m. the wolves arrived at the carcass, which was covered with snow. The thin-tailed male and full-tailed female remained close together, often playfully chasing one another. Once, as this pair came romping toward the carcass, the male sniffed her rear end while in view of the alpha pair. The black alpha male raised his tail and stared at them, and the alpha female walked stiffly toward the full-tailed female, forced her into passive submission, and nipped at her neck as she lay on the ice. The subordinate pair walked a short distance away from the carcass and things quieted down. For approximately the next hour the wolves either rested quietly on the ice or chewed on the remaining bones.

At 7:15 p.m., apparently with no preliminaries, the thin-tailed male mounted and "tied" with the full-tailed female. The male swung one leg over the back of the female 35 seconds later, and they stood back to back. The alpha pair immediately ran over to the tied pair, who then dropped to the ice on their sides, still tied. The black male stayed for only a few seconds, nipping at the male and probably threatening, and then returned to feed on the carcass. The alpha female bit at the neck of the prostrate female many times, then stood over the pair for the next 2 minutes. Another wolf came to watch the proceedings, and the black male made another brief visit. The alpha female continued to stare at the tied female, her upright tail slowly moving back and forth. Five minutes and 50 seconds after tying, the pair uncoupled, and the alpha female responded with a brief "punishing" sequence, consisting primarily of nips directed at the still-prone female. Then the alpha female headed back to the carcass, and the subordinate pair ran together, nose-greeting each other, wagging their tails and occasionally standing on their hind legs with front limbs entwined. Three minutes later they joined the other wolves on the carcass, passing the alpha female with no overt threatening gestures from the latter.

March 3: West Pack (seven) still near the carcass, with only the full-tailed female that mated actually feeding. Her mate inspected her genital region several times with the alpha pair nearby, but stimulated no visible threat from them. When they rested, they lay together, just as the alpha pair did.

March 9: The pack was observed for the final time in late afternoon—seven wolves traveling along the north shore of the island. A thin-tailed wolf trotted close behind a full-tailed wolf, and there was no observed interaction between the alpha pair and the other wolves.

Significant in these observations was the very aggressive attitude of the alpha female toward the full-tailed female when she was courting the thin-tailed male. Having been chased from the pack, however, the subordinate female managed to reinstate herself and mate successfully in spite of her "punishment." The black male usually did not interfere with the subordinate male's courtship activities and showed brief aggression only when the subordinate pair actually tied. In this case the discouraging influence of the alpha pair was not sufficient to prevent mating of subordinates, although the length of their copulatory tie was shorter than normal.

A subordinate pair was present in the West Pack in 1973, and the alpha pair actively interfered with their courtship activities. On 6 February they were observed near a carcass at Windigo:

The subordinate (long-tailed) male continued to show an interest in the subordinate female, who rebuffed him each time. Twice the alpha male reacted to the courtship attempts of the subordinate male with threats and inhibited bites to the head and scruff, usually after the latter had dropped on his side in passive submission. Each time, however, the long-tailed male lay on the ice until the alpha had left the immediate area, then resumed courting the subordinate female. Finally both the alpha male and female approached the subordinate pair, and each threatened the subordinate of its respective sex. The alpha wolves stood over them as they assumed a position of passive submission. After inhibited biting of the prostrate male, the alpha male walked over to the subordinate female and evidently bit her—her yelp was clearly heard at our observation post 0.8 km away.

While the efforts of the alpha pair to discourage courtship in this subordinate pair were persistent, of greater importance was the subordinate female's apparently irreversible lack of interest in the advances of the male.

Before the West Pack fragmented in early February 1974, a presumably subordinate pair was observed mating while the pack rested nearby. The status of the alpha male from 1973 was not established before the mating took place, and immediately afterwards he behaved in a very subdued manner, walking at the rear of the pack with his tail down, while the mating pair led the way. This suggested a change in his status, yet he was still the alpha wolf in both the group of four in which he was later found and the pack of eight that reformed in early March.

From the above accounts, it is obvious that alpha wolves usually interfere with attempts at courtship among subordinate wolves, although I did not record a case when they were actually able to prevent mating among subordinates. Such behavior on the part of the alpha wolves may, however, discourage pair-bond formation or initial sexual interest among subordinates. In the wild, of course, a subordinate pair could leave a pack and breed with no disturbance, but in such a case reintegration into the pack might be difficult.

Our understanding of the effect of the dominance hierarchy on the formation of breeding pairs within a pack is still inadequate. In several packs, both captive and wild, the alpha male did not father the pups, or was relatively inactive sexually (Murie 1944; Rabb et al. 1967; Haber 1968). In the Brookfield pack, a male reduced his participation in courtship activities after assuming the alpha position. However, alpha wolves in both the East Pack and West Pack on Isle Royale have exhibited the most courtship behavior. Individual personalities and attributes and filial or allegiance bonds among wolves can greatly alter relationships within a pack (Rabb et al. 1967) and ultimately will limit the degree to which we can generalize about mate preference and the restriction of breeding among subordinate wolves.

IMPLICATIONS OF SOCIALLY CONTROLLED MATING

Behavioral limitations on mating, including mate preferences, may hold the productivity of wolves considerably below the theoretical maximum, and often only one litter of pups is born, even in large packs. The food-gathering abilities of the adults in the pack then contribute to the growth of a relatively small number of pups, enhancing their chances of survival.

Since packs are basically family groups, there is obviously a high potential for inbreeding in stable packs. Woolpy (1968) studied the genetic implications of social organization in wolves. He contended that the notion that inbreeding results in deleterious effects probably is of little significance when genes are naturally "preselected" for combinations of adaptive value, as they are in wolves. As a demonstration of this principle he cited a study by Scott and Fuller (1965), who inbred beagles and basenjis with no deleterious effects, after preselecting them for fertility, behavior, and body conformation. We have already seen that the parents of wolf pups in the wild are likely to be dominant wolves, already preselected for traits of leadership and physical attributes (Fox and Andrews 1972). Of course, natural selection will rapidly eliminate inferior pups born in the wild.

Woolpy (1968) further concluded that the organization of wolf populations into discrete packs, or subpopulations, was of considerable evolutionary significance. According to his hypothesis, over a period of several years of strong leadership in which most pups are born to a single pair, the expression of available genotypes (gene combinations) within a pack will be greatly reduced. Simultaneously, because of inbreeding, viable recessive gene combinations will appear more frequently. In the long run, this could result in greater variability between wolf packs. Thus, several genetic "lines" of wolves are maintained, with genetic variability partitioned "to give maximum exposure (to recessive gene combinations) at all times and to allow them to compete with each other and thus . . . provide the potential to move the population to new adaptive phenotypes" (Woolpy 1968:32). In a sense, due to wolves' social organization, evolution of the species would be accelerated, resulting in rapid adaptation to different environments. Such adaptability is evidenced by the original widespread distribution of wolves in North America and the description of 23 original subspecies on this continent (Goldman 1944). A major challenge facing us today is whether we can preserve a sufficiently large number of natural ecosystems to allow wolves and other species to achieve their own evolutionary potential.

Territoriality

In spite of extensive field studies of wolves in various parts of North America, the precise nature of spatial relationships between adjacent packs was unknown until recently. It was unclear whether wolf packs occupied exclusive, nonoverlapping territories or whether neighboring packs utilized common hunting grounds, simply avoiding each other through direct and indirect communication.

Recent studies of radio-marked wolves in many packs in Minnesota helped crystallize a concept of "land tenure" among wolf packs (Van Ballenberghe 1972; Mech 1972, 1973). Individuals within a pack utilize a common territory or "defended area" (Noble 1939), and the home range of individual wolves, defined as the area in which they travel during normal activities (Burt 1943), coincides with the territory of the pack to which they belong. The Minnesota studies revealed that packs usually occupy exclusive, nonoverlapping territories, with territory size and wolf density probably related to food supply.

In northern Minnesota wolves began to travel outside their former territory in response to a shortage of their principal prey, white-tailed deer, indicating that territories may be enlarged in response to a decreased food supply (Mech, in press). Along the Minnesota shoreline of Lake Superior, where deer densities are very high, wolf densities reached 1/14 km², with pack territories among the smallest reported for wolves (Van Ballenberghe 1972). Five resident packs totaling 40 wolves occupied an area the size of Isle Royale.

The inherent flexibility of territory size, demonstrated by these Minnesota studies, allows for great adaptability to local conditions. We would expect that a pack's territory would be no larger than necessary to obtain sufficient prey. A shrinkage of territory in response to an expanded food base allows for the establishment of additional packs, as on Isle Royale. Significantly, during the winter prior to the appearance of the East Pack on Isle Royale, the West Pack utilized only half of the island. Simultaneously, the amount of food available to Isle Royale wolves was higher than at any other time during this study (Table 10).

In 1972-74 Isle Royale had two primary packs, each occupying about half of the island. Additional duos and trios usually occupied areas along the boundary between the two large territories. Loners followed the large packs and scavenged their kills, or existed independently. As pack sizes increased from 1972 to 1974, the amount of spatial overlap between the two packs increased; there was a simultaneous decline in the calculated amount of prey available to wolves in both packs.

EXCLUSIVE TERRITORIES

Pack territories probably are no larger than the minimum size necessary to provide sufficient food and are sensitive to changes in density of vulnerable prey. This flexibility in territory size is advantageous to wolves in maximizing their hunting efficiency. Restriction of pack activity to a certain area ensures an intimate knowledge of that area (prey location and easiest travel routes) and prevents wasteful overlap in hunting efforts of two packs. A mechanism that spaces packs in relation to prey may be of greatest importance during the pup-rearing season, when pack activity is centered around the relatively immobile pups. Litters distributed so that there is plenty of food in the surrounding area would provide for rapid growth of pups.

The advantages accruing from a system of exclusive territories should apply equally well to other species of group-hunting carnivores. Indeed, the spotted hyena (Crocuta crocuta) exhibits a similar pattern of social organization and spacing, at least at high population densities (Kruuk 1972). At lower predator densities the "need" for exclusive territories would be lessened; Kruuk (1972) found that in the Serengeti, where prey are highly mobile in response to environmental factors, the hyena population was relatively low and territories were not as clearly defined as in regions of higher hyena densities. Eaton (1974) suggested that naturally low densities of cheetahs might explain the lack of exclusive territories in this species.

TERRITORY SIZE

Territory size of the East Pack increased from 1972-1974, while there was no consistent trend in the range of the West Pack (Fig. 58; Tables 12, 13). Simultaneous with an increase in the territory of the East Pack there was a consistent decline in availability of food per wolf in both packs.

TABLE 12. Numbers, movements, range, and prey availability for the Easts Pack, 1971-74.

TABLE 13. Numbers, movements, range, and prey availability for the West Pack, 1972-74.

Fig. 58. Winter territories of Isle Royale wolf packs, 1971-74.

The amount of overlap in pack territories increased from 1972-1974. In 1972, the first winter for the East Pack, approximately 9% of the island was not utilized by either main pack. The following year the two packs overlapped on 6% of the island during winter. In 1974, the amount of overlap increased to 16% of the island, primarily because of the movements of the East Pack into traditional West Pack territory. It seems reasonable that the increased amount of overlap resulted from the growth of the East Pack, with a concurrent increase in the food requirement.

EXPANSION OF EAST PACK TERRITORY

When the East Pack moved into what had been regarded as West Pack territory in 1974 its behavior and movements were of great interest (Fig. 59). The following account was edited from field notes:

Fig. 59. Movements of the East Pack into traditional West Pack territory in 1974.

Prior to February 1974, the East Pack had never been observed west of Siskiwit Lake. On 12 February, however, tracks indicated that it had traveled to Mud Lake, 2 km southwest of Siskiwit Lake. On 15 February we circled the pack as it crossed its former territorial boundary and entered Siskiwit Bay.

As the pack approached Little Siskiwit Island the alpha female squatted and marked the ice with urine. Another wolf rolled on a spot at the tip of the island after sniffing the urine: a second rubbed its neck on the same spot. A subordinate wolf urine-marked the same place, and soon all the wolves clustered about the area with their noses to the ice. The alpha male marked once, and the alpha female marked again before they left the island. Frequent scent-marking was observed after they left the island (10 urine marks in 0.50 hour). Several subordinate wolves marked in addition to the alpha wolves. Urination posture for all wolves was either stretching forward or squatting; no raised-leg urination was observed. As they proceeded, the subordinate wolves usually clustered about the alpha pair in a mobile group ceremony.

The alpha pair and two other wolves of the West Pack were resting by a kill on Francis Point, only 5 km away, across the frozen bay. Although ravens feedings on this kill saw the approaching East Pack and flew out to meet them, rough ice apparently blocked the view of the wolves.

Fifteen minutes after the pack left Little Siskiwit Island, we left to refuel. On our return in 0.50 hour the East Pack was lying on the ice in the middle of Siskiwit Bay. Bad weather allowed only 5 more minutes of observation—during this time the pack got up and began moving toward Houghton Point.

February 16: The East Pack wolves killed a moose when they reached Houghton Point and were resting on the ice at the end of the peninsula. Five West Pack wolves were at their kill only 4 km away.

February 17: Tracks showed that the East Pack traveled due west into Siskiwit Bay for about 2 km and then returned to its kill on Houghton Point. The tracks of the West Pack went round the head of Siskiwit Bay and then southeast until they met the tracks of the East Pack. The tracks then abruptly changed direction and crossed to the opposite side of the island.

February 18: Bad weather prevented any flights.

February 19: In the morning we found the East Pack lying on shelf ice on the south side of Houghton Point, about 1.5 km from the carcass of a wolf that it evidently had killed (Fig. 60). From tracks it appeared that there may have been as many as three wolves in the group that met the East Pack coming from Houghton Point. The dead wolf, a female pup, was definitely not a member of the East Pack (which was unaltered in size); we could only surmise that it was a member of the fragmented West Pack. The wolf had been deeply wounded in the throat area, and puncture wounds were frequent, especially in the rump region. None of the carcass had been consumed by the East Pack. The rump of the dead wolf had been heavily marked with urine, and three other fresh scent marks were found in the vicinity of the carcass.

Fig. 60. Author recovering carcass of female pup killed by East Pack during trespass in 1974.

Several hours later the pack resumed travel southwest along the shore. Scent-marking was frequent—at least four wolves urine-marked in 14 minutes (alpha female, two subordinates, and one of unknown status marked in either squat or stretch position). At the mouth of a small creek it encountered a well-used travel route of the West Pack. Most wolves had their noses to the ice. They quickly gathered around the base of a large spruce tree, crowding together in order to sniff its trunk, which undoubtedly had been previously marked by the West Pack (Fig. 61). Suddenly they scattered away in several directions, as if frightened. After standing briefly along the bank of ice next to shore, the alpha female led the pack back in the direction from which it had come, apparently in response to the first obvious sign of extensive West Pack activity that it encountered.

Fig. 61. East Pack (a) gathered around West Pack scent post while trespassing in 1974, (b) standing on shore after scattering from scent post.

Several important aspects of the above account should be emphasized: (1) the East Pack, although it traveled into "foreign" territory, moved into an area that had not been used recently by the West Pack, and it turned back when it encountered significant West Pack activity; (2) the East Pack outnumbered any of the fragments of the West Pack with which it had direct or indirect contact. The alpha pair and two other wolves of the West Pack avoided areas crossed by the East Pack, even areas within their own territory; (3) as the East Pack ventured into areas probably unfamiliar to it, there was an unusual amount of scent-marking, by both dominant and subordinate members of the pack.

MAINTENANCE OF TERRITORY

Wolves must rely heavily upon indirect means of communication to delineate territorial boundaries. Potentially, any behavior which advertises the presence of a pack and causes another to avoid intrusion is of territorial significance. Scent-marking, howling, direct aggression, and avoidance may all serve to maintain territory.

Scent-marking. In addition to other functions discussed later, scent-marking serves to delineate territorial boundaries. The above account of the movement of the East Pack into West Pack territory is the only reported observation of a pack encountering another pack's scent-marks. One additional instance of avoidance behavior along a territorial boundary was deduced from tracks in February 1973 (Fig. 62).

Fig. 62. Avoidance behavior of the West Pack in response to an East Pack scent post, 1973.

Peters and Mech (1975) detailed four cases, determined from tracks, in which a pack responded to foreign scent-marks by avoidance. In one instance a pack chasing a wounded deer ceased pursuit at its territorial boundary. Generally, the outermost kilometer of a pack's territory was scent-marked profusely compared to the center. They concluded that scent-marking was an extremely important part of territorial behavior that contributed to efficient spacing among wolves.

Howling. Joslin (1967) suggested that howling may be of territorial significance, since it is an effective form of long-distance communication and may convey enough information to permit identification of howling wolves. Since the interface between packs on Isle Royale is so short and territories are long and narrow, howling may be a less effective method of communication between packs than it would be elsewhere.

Responses of wolves to human imitations of wolf howls are variable. In one case in June 1973, Sheldon L. Smith (pers. comm.) was sitting on a ridge less than a mile from where human imitations of wolf howls were being broadcast. Although he did not hear the human howls, he heard at the same time seven or eight brief howls from several wolves that were passing him through adjacent thick vegetation. The wolves seemed to be responding to the human howls, but were traveling in the opposite direction.

In another case, in September 1972, we elicited a howling response from several wolves in East Pack territory. We approached the group and howled again, and soon a single wolf approached us. Its body and tail markings suggested it was the alpha male of the East Pack. When the wolf saw us it turned and ran back to the rest of the group, and the pack disappeared. Only the one wolf, quite possibly the dominant male, left the others to confront what he might have thought were foreign wolves. Joslin (1967) was often approached by one or more wolves when he howled within 200 yards of homesites. He interpreted this as active resistance toward intruders.

Direct aggression and avoidance. When adjacent packs make visual contact with each other, such as across the ice of a large lake, they must either confront or avoid each other. The response of any two packs probably would depend on their previous history of association and perhaps their numerical strength. The East Pack killed a strange wolf even though it was in unfamiliar surroundings. The outcome of such a confrontation might well have been different if the East Pack had met the entire West Pack, instead of not more than three. Lack of numerical strength may have been the reason for the previously discussed avoidance of the East Pack by four West Pack wolves in Siskiwit Bay, even though the West Pack animals were in their own territory.

SMALL PACKS AND "LONERS"

In all 3 years that both East and West packs have existed, an additional, small group of wolves has been present (Fig. 63). In 1972, a pack of two or three was seen in the Malone Bay area, ranging over to Houghton Point and possibly as far as Chippewa Harbor. In the next two winters, two wolves (the "Todd duo") traveled the north shore in the vicinity of Todd Harbor and were also seen near Intermediate Lake and Lake Whittlesey. The smaller of the pair was noticeably reddish on its lower flanks and belly. Their friendly greetings suggested a male and female pair.

Fig. 63. Travel routes and carcasses utilized by small groups of wolves on Isle Royale in midwinter, 1972-74. (click on image for a PDF version)

It is significant that in the years when two large packs "divided" the island approximately in half, the small pack usually inhabited an area either between the two packs or overlapped by each. This supports the hypothesis of Mech and Frenzel (1971:33), who believed that wolves in Minnesota were organized into breeding packs occupying exclusive territories, with "loners and other nonbreeding population units" inhabiting nonexclusive areas among the pack territories. These small groups probably survive only by their ability to avoid the large packs.

The "loners" are more difficult to locate and observe than the larger groups. Their ecology and social status relative to other wolves in the population is little known. Jordan et al. (1967) described several stages of "dissociation" of single wolves from a pack. They believed that many loners were aged and socially subordinate wolves that were gradually excluded from the pack, although in individual cases it is usually not possible to determine sex, age, or previous social relationships.

I have seen only one case in which a single wolf which was following a pack might have been a "dropout." After the West Pack declined from eight wolves to seven in mid-February 1972, a wolf was seen following it, often hesitating and apparently trying to remain hidden from the view of the pack. The pack had left the carcass of a moose near Windigo, traveled a short distance, and then lay down on the ice north of Beaver Island. The single wolf walked up on a 50-m rise on the north side of Beaver Island, then sat down at the edge of a cliff overlooking the pack and watched them intently, hidden from view by trees.

The different reactions of a pair of wolves to single wolves were recorded in 1974 (field notes):

On February 6, 1 or 2 days after the West Pack fragmented, a pair of wolves (the McGinty Duo) were found on a kill at the southwestern end of the island. One wolf was larger than the other, and the friendly greetings suggested a male-female combination. On February 13 a single wolf, tracked from a previous kill of the West Pack, joined them. The single wolf participated in greetings between the duo, although it was very submissive. Perhaps all 3 were formerly included in the 12 wolves of the West Pack.

On February 14 we found another wolf following the tracks of the third wolf around the southwestern end of the island, and it appeared about to join the newly-formed trio, who were feeding on a kill on the ice at the entrance to Washington Harbor. However, when the fourth wolf was finally seen by the trio on the kill, the duo gave immediate chase, while the other newcomer stayed at the kill and fed. The duo chased the single down Washington Harbor at top speed, then followed it overland back to the lakeshore at the southwest end of the island. After running 5 km, the duo continued to follow the single for at least 19 km before returning to the Washington Harbor area, where they were again joined by the third member.

Communication Among Wolves

VOCALIZATION

The highly social nature of wolves and the flexibility of their group structure and hunting habits probably account for the diversity in forms of vocal communication found in this species. Howling, the most widely known and most unique wolf vocalization, is of obvious significance in long-range communication. Individual wolves have distinctly different howls and seem to be capable of distinguishing differences in howls, so there is a high potential for exchange of information via howling (Theberge and Falls 1967). Other widely recognized sounds that are not often heard in the wild include the whimper, growl, and bark (Mech 1970).

Howling. Howls can be heard for several miles under certain conditions, and Joslin (1967) reported that howling could advertise the presence of wolves over a 130-km² area. In addition to possible territorial significance, howling helps to assemble individuals in a pack after they have been separated. On Isle Royale in 1973, howling also was of obvious importance in coordinating moves of a large pack between summer homesites.

Spontaneous howling of East Pack wolves was heard 62 times during approximately 383 hours spent near their rendezvous sites (homesites) in 1973. Most of the howling was heard at night (Fig. 64), when more adults were hunting and spatially separate. Such howling may help wolves coordinate hunting efforts. Pups and adults at or near a homesite often howled in response to howls of distant adults. Almost half (45%) of the howls heard near East Pack homesites included adults that howled some distance away.

Fig. 64. Frequency of howling near East Pack rendezvous sites in 1973 (4 July-6 October). (click on image for a PDF version)

Increased howling at dawn and dusk may be associated with departures and arrivals of adults at the rendezvous areas. Carbyn (1974a) recorded dawn and dusk peaks in howling and general activity at wolf rendezvous sites in Jasper National Park in Alberta. Murie (1944) described how adults assembled at the den before departure for their nightly hunt. Howling at this time accompanied generally friendly behavior, with much greeting among the adults. Group howling and greeting ceremonies often occurred together among members of the captive pack at Brookfield Zoo (Fig. 65). Group howling also is common among coyotes and jackals (Canis aureus) (Kleiman 1967).

Fig. 65. Group howl (Brookfield Zoo).

A group howl was observed at a summer rendezvous of the East Pack in July 1973 (paraphrased from field notes):

In mid-morning the only wolves in view were four pups lying motionless in thick grass in a favorite resting place. An unseen adult in the woods nearby began to howl, probably in response to a distant howl that was faintly audible to us. Three unseen adults then began to howl at the rendezvous. Another adult wolf about 1 km from the rendezvous also joined the howl. The four pups became excited and chimed in with yips and short, high-pitched howls. One stood and howled with its head slightly raised, while the other three sat and howled with their muzzles pointing at the sky. The howling lasted for about 2 minutes, after which the pups looked briefly in the direction of the howling adults and then resumed their normal activities of playing, snapping at insects, and resting. Ten minutes later all the pups ran out the west side of the rendezvous and greeted an incoming adult.

Although wolves are capable of fine auditory discrimination, they may howl in response to sounds which, to human ears, are quite distinct from actual wolf howls. At Brookfield Zoo, howling often occurs in response to sirens. Human "howling" is often an adequate substitute for prerecorded wolf howls when at tempting to stimulate howling among wolves. The common loon (Gavia immer) has a call that closely resembles a wolf howl—twice in 1973 Isle Royale wolves at summer homesites began to howl immediately after hearing loons. Once, the pups were clearly the first to respond. On two occasions I heard loons calling shortly after wolves began to howl.

Other vocalization. Only limited information was gathered on Isle Royale on other forms of vocal communication, mainly because they are inaudible at long distances. An adult whimpered when it arrived at a summer rendezvous after the rest of the pack had left. Whimpering, interspersed with occasional high-pitched yipping, was frequently heard from pups as they mobbed adults arriving at rendezvous sites. Joslin (1966) believed that whimpering was a friendly greeting, sometimes conveying a submissive attitude. Whimpering was often part of low-intensity friendly greetings at Brookfield Zoo, especially between pairs during the mating season.

Barking was heard only during group howls at rendezvous areas. Much of the pup vocalization during group howls consisted of high-pitched "yips," and adult barking sometimes accompanied these pup vocalizations, especially near the end of a howl, much as Joslin (1966) described. He considered barking to be either of a threatening or alarm nature. The "alarm bark" is short and often seemed to cut off a howling session. Joslin occasionally elicited a threatening bark by howling at close proximity to wolves at a rendezvous. In such cases, the barking was more continuous and interspersed with growling.

SCENT-MARKING

Humans, with a poor sense of smell, are ill-equipped to appreciate the importance of olfactory communication. Scent-marking helps maintain territories, contributes to pair-bond formation, provides information on social and sexual status and individual identity, and helps orient wolves in their environment (cf. Peters and Mech 1975).

In canids, elimination (urination, defecation) and rubbing of certain body areas may have scent-marking significance (Kleiman 1966). Scent-marking differs from simple elimination by its directional and repetitive nature—that is, the same object may repeatedly be scent-marked. Kleiman also suggested that this form of scent-marking developed from autonomic responses to strange or frightening situations. Initially, scent-marking could have reassured an animal entering a strange environment and may have since acquired additional signal value in territoriality and courtship (Fig. 66).

Fig. 66. Pilot D. Murray examines wolf scent post.

Wolves have at least two specialized scent glands (Mech 1970; Fox 1971). The anal gland is located on each side of the anal sphincter; presumably scent deposition takes place with each passage of feces. A tail (precaudal) gland of unknown marking function occurs on the dorsal surface of the tail near the base, under a dark patch of hair ("dorsal spot").

Urine is also of considerable scent-marking importance among wolves. In a field study based on tracking wolves in snow, Peters and Mech (1975) distinguished four types of scent-marks: (1) raised leg urination (RLU); (2) squat urination (SQU); (3) defecation (scats); and (4) scratching. They found that the RLU was the most frequent and significant type of scent-mark.

Scent-marking by Isle Royale wolves was observed only in winter, usually from the air. Scent-marking was often difficult to distinguish from normal elimination, which seemed to be most common when packs were resting near kills or just beginning to travel. In these cases I ignored defecation and urination unless clearly directed at an object.

Frequency of scent-marking. Obvious differences in frequency of scent marking occurred among Isle Royale wolves (Table 14). In all cases the packs were traveling. When the East Pack first entered "foreign" territory we observed 10 scent-marks in a half-hour of observations, compared to 2 in an equivalent length of time as the same pack reentered its own territory several days later.

TABLE 14. Scent-marking frequency in traveling wolves.

The highest level of scent-marking occurred when three wolves (McGinty duo + 1) left a kill in full view of four wolves of the West Pack (including the alpha pair) who had bedded down 1 km away after following the tracks of the three for many miles. The West Pack wolves were not watching the trio, one of which glanced in the direction of the sleeping wolves twice as we circled. This wolf made six of the seven scent-marks observed. In this case frequent urinations might have resulted from autonomic responses to fear or apprehension and might not have been actual scent-marking.

Peters and Mech (1975) found that when packs traveled within a kilometer of the edge of their territories, the frequency of RLU's was twice as high as when packs traveled in the center of their territories. Thus, an accumulation of marks characterized territorial boundaries. The strongest stimulus to scent-mark was the mark of a neighboring pack. In one case when a pack discovered fresh tracks of a neighboring pack on its territorial boundary, these researchers found 30 RLU's, 10 scratches, 2 SQU's, and 1 scat. During normal travel in winter, wolf packs left a sign every 240 m on the average, including a RLU every 450 m. At their normal rate of travel of about 8 km/hr (Mech 1966), that implies an olfactory mark about every 2 minutes, with a RLU every 3 minutes. Isle Royale wolves demonstrated comparable marking frequency (Table 14).

Indicator of sexual and social status. The alpha male and female in all packs observed from 1972 through 1974 accounted for most of the recorded instances of scent-marking (27 of 39 cases). In all cases there was active courtship between the alpha wolves; scent-marking clearly played a role in these activities at times. On four occasions the alpha female was seen urine-marking an object, and the alpha male, usually right behind her, sniffed the location and then urinated on it. Twice the alpha female marked a scent-post of the alpha male. In these cases, scent-marking should be considered part of the mechanism of pair-bond formation, as Schenkel (1947) suggested. Twice an alpha male mounted the alpha female immediately after inspecting her fresh urine-mark.

The frequency of genital inspection of females during the mating season indicates the importance of olfactory cues in sexual behavior. Experimental work with domestic dogs reviewed by Johnson (1973) documented that urine from estrous females was more attractive to males than that of anestrous females and stimulated mounting among males. Peters and Mech (1975) found that the frequency of RLU and SQU increased before and during the wolf breeding season. During the breeding season they often found a SQU and RLU together in the snow, indicating a female-male combination as described above.

We might surmise that scent-marking in the wild would be an important means of establishing initial contact between potential mates in widely dispersed populations. Peters and Mech (1975) pointed out that a lone wolf would be able to determine where potential mates lived and whether they were already paired off, since mating pairs often mark the same points. Also, territorial boundaries were marked with such clarity and frequency that a newly formed pair could easily tell whether they were in an occupied territory, along a territorial boundary, or in unoccupied space.

Limited data suggests that scent-marking is related to social status. Alpha wolves often marked when exhibiting no sexual behavior. In some cases subordinate wolves did not mark when they might have been expected to do so. For example, in 1972 the alpha female in the East Pack squatted and urinated on a ridge of ice as the pack was traveling. The spot was inspected by a subordinate wolf, who continued on its way without marking. The alpha male, however, lifted his leg and urinated on the spot after sniffing it. In 1973, four subordinate wolves of the West Pack followed the alpha pair into the woods next to shore. The alpha male led the way, marking a tree on the shoreline. The alpha female followed suit, but the other four wolves sniffed the scent post and left without adding their own scent. Mech and Frenzel (1971) recorded an instance when a wolf believed to be the alpha male was more active than the others when the pack encountered scent posts.

Peters and Mech (1975) reported that in two captive packs only high-ranking wolves raised their legs when urine-marking. Twice when they tracked wild pups, they found several SQU's, but no RLU's. While males tend to lift their legs more often than females when scent-marking (Kleiman 1966), several alpha females did this on Isle Royale. I did not observe subordinate wolves lift a hind leg while scent-marking, at least while alpha wolves were present.

Orientation and information exchange. Scent-marking in canids may also serve for orientation and information exchange. Humans naturally regard visual signals as the most important means of orientation but, for wolves, a keen sense of smell would be more valuable. Scent-marks are frequent along pack travel-routes and are especially prominent around kills or other centers of activity. About half of the scent-marks recorded by Peters and Mech (1975) were at trail junctions.

Peters (1973) found that fatty acid content of anal gland secretions differed between males and females and that all individuals differed slightly from one another. Peters and Mech (1975) showed that wolves tend to remark fresh scent-marks more often than old marks, and inferred that wolves also could discriminate between marks of different ages. Thus, by simply sniffing a scent-mark, a wolf can probably tell whether the marking wolf was a stranger, a male or female and its reproductive status, and how long ago the mark was made.

Territorial marking. Young (1944) reported that wolves became greatly excited and scent-marked frequently when encountering the introduced scent of a strange wolf. Schenkel (1947) believed that scent-marking was of territorial significance; avoidance behavior in response to another pack's scent posts along a territorial boundary occasionally has been documented (see Maintenance of Territory).

Rolling and scratching. Kleiman (1966) believes that rolling has some scent marking function (perhaps self-marking); Fox (1971) suggested that this may promote interaction with other pack members by encouraging social investigation of wolves carrying interesting odors. Wolves that are separated from the pack could transport odors on their fur and perhaps transmit information to other pack members. I have seen Isle Royale wolves roll at various places: a moose bed in the snow, the kill site of a fox, the dug-out remains of an adult moose killed 7 months earlier, and in snow next to a fresh moose kill. Also, when the East Pack reached Little Siskiwit Island on its trek into new territory, two wolves rolled on the ice after sniffing it; perhaps this was an old scent post.

Wolves often scratch with their feet after urinating and defecating (Fig. 67). Usually, only high-ranking wolves exhibit such marking (Peters and Mech 1975). The function of this behavior is not clear, but Mech (1970) noted that it would increase the visual signal value of a scent-mark. Schenkel (1947) thought that it might be a behavioral rudiment which perhaps has lost its original function. This is reasonable in light of comparative work with other carnivores. Kruuk (1972) found that male hyenas scrape only as a sexual display in courtship—probably to distribute the scent of their interdigital glands. While interdigital glands have not been described in wolves and coyotes, Fox (1971) stated that such glands are found in red foxes and perhaps existed in primitive canids.

Fig. 67. Scent post and scratch mark of alpha pair, West Pack, 1972.

Summer Ecology

The shy and elusive nature of wild wolves makes summer ecological studies difficult. Significant observations in the wild are possible at dens or other centers of activity on the tundra of Alaska and Canada, but in forested areas rarely more than a fleeting glimpse of wolves is possible from the ground (Fig. 68). Radio-tracking has provided detailed knowledge of wolf movements in summer in Ontario and Minnesota (Kolenosky and Johnston 1967; Van Ballenberghe 1972; Mech and Frenzel 1971).

Fig. 68. Rare look at wolves in summer. Second wolf running through brush in upper right.

During this study, a den used by the East Pack was found in early July 1973. The wolves had abandoned it but were located at a rendezvous area 1 km away. Subsequent movements of the pack were followed to three additional rendezvous sites until pack movements became extensive in late September. One rendezvous site of the West Pack was found in September, about a month after it had been vacated. Direct observations of wolves were possible only at one rendezvous of the East Pack in 1973.

Dens

Wolves occasionally dig out dens weeks in advance of the birth of pups, which probably takes place in late April on Isle Royale. While wolves usually dig underground dens in sandy soil, they have also used hollow logs, rock cavities, old fox dens, and beaver lodges (Mech 1970). Dens commonly are close to water, perhaps because nursing females have a high water requirement.

The whelping den used by the East Pack in 1973 was an abandoned beaver lodge, whose entrance had been exposed when a dam broke. Another abandoned lodge 10 m away and a nearby hole in a sandy bank also appeared to have been used (Fig. 69). All holes were within 20 m of water. Many old scats under leaves and other debris indicated that this den had been used before—possibly during the first 3 years of the East Pack's existence. Tracks of a wolf were followed to this den in March 1974, and some scratching was found, but the site was not used in 1974.

Fig. 69. Wolves dug out this old den site (a beaver lodge) in March.

Both lodges had a central chamber; barely large enough for an adult wolf, and many interconnecting tunnels which only pups could use. The only obvious alteration by wolves was the enlargement of at least one entrance and tunnel to the central chamber. Scattered around the den area were bones from at least six beaver, one muskrat, and one adult and one calf moose.

In 1975, the East Pack again denned in a beaver lodge. The West Pack denned in a hollow white pine (Pinus strobus) trunk that had fallen to the ground (Fig. 70). The log was 9 m long with the major opening 45 cm high and 55 cm wide. Pups frequently had used smaller openings created by decay of the wood. Three other possible whelping dens also were found—two abandoned beaver lodges and a hollow log.

Fig. 70. West Pack den in hollow log.

Wolves visited all six dens during both summer and winter. Because there are very few opportunities for wolves to dig dens on Isle Royale, they take full advantage of existing structures.

Rendezvous Areas

In temperate regions pups are usually moved from the den site in late June or early July, after the pups have been weaned (Mech 1970). Thereafter, the activities of the pack center around "rendezvous sites" (Murie 1944:40) or "loafing areas" (Young 1944:103), where the pups remain while the adults make hunting forays. A succession of rendezvous sites are used by a pack until the pups are able to accompany the adults on all their travels.

Rendezvous sites, like whelping dens, usually are near water and often are adjacent to bogs (Joslin 1967). In 1973, five rendezvous sites were found on Isle Royale, four of the East Pack and one of the West Pack (Fig. 71). All five were located by abandoned beaver ponds, with water still available nearby. Size varied from 0.4 ha to a drainage 1 km long. Most had a prominent open area where the vegetation had been matted, and holes often had been dug in nearby banks. A small den was found beneath the roots of a cedar tree at one area, and a beaver lodge had been excavated and used at another area. Both dens and rendezvous are frequently reused, with former rendezvous sites possibly serving as den sites at a later date, and vice versa. Over a 3-year period, a pack studied by Carbyn (1974a) used the same den and the same first rendezvous site each year.

Fig. 71. Aerial view of wolves at rendezvous site in October.

Rendezvous sites generally are used for shorter periods than den sites. Joslin (1967) found that packs moved an average of every 17 days, possibly influenced by frequent human howling nearby. Baffin Island packs moved to different summer dens (analogous to rendezvous sites) about every 30 days (Clark 1971); (Van Ballenberghe 1972). On Isle Royale, rendezvous areas were occupied from 11 to at least 48 days (Table 15). Wolves have been seen at rendezvous sites as late as October on Isle Royale and in northeastern Minnesota.

TABLE 15. Successive rendezvous areas of the East Pack, 1973.

OBSERVATIONS AT A MIDSUMMER RENDEZVOUS

In July 1973, we observed the East Pack at its second rendezvous site from about 200 m. The pups usually were the only wolves in sight. Adults spent much of the day in the cooler forest surrounding a central open area. Most activity was observed before 11:00 a.m. or after 5:00 p.m.

Seven pups, probably the total in the pack, were seen at this rendezvous, along with at least seven different adults. Most of the adults in the pack probably visited the rendezvous periodically.

Adults were observed arriving at the rendezvous nine times, always before 10:00 a.m. or after 5:00 p.m. Only once did two wolves enter together, indicating that most of the hunting effort by adults in summer is done by individuals or small groups. Pups often sensed the imminent arrival of adults and ran out as a group to meet them. Such an arrival was an occasion of great excitement for the pups, and they greeted the adults by yipping and jumping at their heads. Excited licking of the mouth acts as a stimulus causing regurgitation of food for the pups. On three occasions the arriving adults regurgitated food for the pups immediately, while being greeted. The pups ate such regurgitated food within a minute. Adults rarely remained in open areas for any length of time. About 11 kg of food per day would be required to feed seven pups; (based on Kuyt 1972); providing this amount is undoubtedly a demanding task.

Pup activity alternated with long periods of sleep, but even then pups frequently looked up or stood and readjusted their position. Their ears were in constant motion because of insects, mostly mosquitoes. When resting, pups often sought each other's company, even flopping down directly on another sleeping pup. Rest was interrupted with jaw-wrestling, scruff-holding, and occasional nibbling of legs and tails of nearby pups.

Many of the pup activities were group-oriented, such as play-fighting and competition for bones or sticks—appropriately termed "trophies" by Crisler (1958). Pups probably did most of the digging found at rendezvous sites. Many items were chewed extensively—moose bones, antlers (especially those in velvet), sticks, and at one site, an aluminum canteen. One evening, five pups gathered around a rotten birch log. They attacked the log in much the same way that they would later treat a moose carcass; each pup lay on its belly, chewing on its portion of the log and snapping at any encroaching sibling. They ripped enthusiastically at loose pieces of rotten wood and occasionally wandered off with a chunk for more peaceful chewing.

Six of the pups were of uniform appearance and impossible to tell apart. The seventh, called "7-up," was much lighter in color than the others. Its activities often set it apart. When first distinguished, this pup was the scapegoat during vigorous play-fighting of four pups. With tail firmly planted between its legs, "7-up" continually was the object of chases and alternately was submissive and defensive. Another time, this pup was chewing on a calf-leg bone when two others walked up and stood over "7-up" with a dominant attitude; one finally grabbed the bone. After a spirited defense, "7-up" ended up on its back, entirely submissive.

On at least two occasions, "7-up" was the only pup at the rendezvous; once it appeared that no other wolves were present. During this time a cow moose walked slowly into the open area while the pup was out of sight. She stopped and sniffed the ground thoroughly. Undoubtedly, the scent of wolf pervaded the area, and she seemed hesitant, her movements very slow. Every few steps she stopped and looked about, frequently sniffing the ground in matted places. Finally she walked down the drainage and disappeared. Almost immediately, "7-up," with nose to the ground, scampered into the opening and followed the moose briefly. It is quite possible that the pup had had the moose under surveillance but was reluctant to show itself when the moose was nearby. Crisler (1958) and Fentress (1967) reported that their captive wolves were initially afraid of large animals, even traditional prey. Considerable experience is probably necessary before pups become effective predators of ungulates as large as moose. Moose commonly exhibit no fear of wolves. They were seen several times browsing on the edge of a rendezvous. Once I watched a bull, apparently unconcerned, browsing within 100 m of some pups and adults that were howling just out of sight.

MOVEMENTS BETWEEN RENDEZVOUS SITES

Wolves move to different rendezvous sites for seldom-known reasons. The accumulation of feces and debris eventually may render dens less desirable (Young 1944; Rutter and Pimlott 1968); perhaps the same applies to rendezvous sites. In some cases wolves might move the pups to a fresh kill. At two of the five rendezvous examined in 1973, a moose-kill was found in the center of the activity area (Fig. 72). At the rendezvous that the pack reused in 1974, a fresh kill was found in almost the same location as a kill that had been made the previous year.

Fig. 72. Well-utilized carcass of adult moose killed by wolves in summer.

We watched the East Pack abandon its second rendezvous of 1973. Howling helped to coordinate the move to a new site, as shown by the following field notes:

Early in the afternoon of 28 July no wolves were visible, nor were any vocalizations heard until 5:00 p.m., when a single pup howled briefly, then whined. An hour later an adult howled, and in the next 2 hours howls of adults were heard seven times, although no wolves could be seen. Most of the howling was done by one wolf, although once, four adults howled simultaneously. No pups howled with the adults. At 9:30 p.m. an adult arrived at the rendezvous, and "7-up" ran out to greet it. The adult ignored the single pup, and as two adults howled in the adjacent woods this newcomer joined in, then trotted off toward them. The pup wandered about alone among the trampled vegetation, often with nose to the ground. The adult came back into the opening and played with "7-up" for several minutes but then left the pup alone again. At 10:00 p.m. the pup disappeared from view, and at 10:03 p.m. one pup and three to four adults howled together. A single adult howled briefly at 10:06 p.m. Observations ended at 10:10 p.m., and no more howling was heard that night. Although pups had often been absent in the past, the length of the separation and the unusual amount of adult howling suggested that some of the adults and six of the pups had abandoned the area.

The next morning observations began at 6:40 a.m. At 7:15 a.m. the same adult walked into the rendezvous area again and stood still for a minute, sniffing the air. It whimpered loudly and sat down for another minute, constantly looking around and sniffing the air, then left. An hour later the alpha female arrived and was enthusiastically greeted by "7-up." She did not regurgitate for the pup, but walked the length of the activity area, sniffing the air and looking about. She disappeared, followed a short while later by the single pup. Between 1:00 and 2:00 p.m. howling of one or two adults was heard twice near the rendezvous. In the evening both the alpha male and alpha female arrived individually and lay down in the open activity area. "7-up" joined the pair, lying down near the alpha male.

Two hours later I heard a faint howl in the distance, and the alpha male rose and howled in response. The female joined him, as did another adult out of view. They howled continuously for 2 minutes, and after rubbing muzzles in a greeting, the alpha pair left the rendezvous, heading in the direction of the distant howls. Two minutes later the pup, who had remained silent throughout the howling, followed the pair. The rendezvous was silent for the rest of the night. The area was observed the next day to verify that the rendezvous had been abandoned.

Earlier in the summer, five pups were observed en route from the first to the second rendezvous. In this case adults were howling periodically at both locations, and the five pups went to the next site by themselves. The following day, a sixth pup was still present at the first site. Two nights after initial occupancy of the second rendezvous, adults present at the first site were heard howling in response to pups and adults at the new site, indicating that several days may be necessary for complete relocation.

Pup Production

Although pup mortality is widely regarded as an important factor controlling wolf populations, information on pup production and survival on Isle Royale is very limited (Table 16). Sometimes the minimum number of pups in packs can be estimated during winter, but this is not a valid year-to-year index.

TABLE 16. Pup production on Isle Royale, 1970-73.

Pup condition may provide some indication of the extent of mortality. A dead, emaciated pup was found in 1964 (Jordan et al. 1967), suggesting that inadequate food supply early in life might be a critical factor on Isle Royale. A decrease in food supply seems to be an important reason for poor pup condition and low survival in Minnesota (Mech 1973; Van Ballenberghe and Mech 1975; Seal et al. 1975). Kuyt's (1972) data suggested lower pup survival in areas where tundra wolves relied heavily on small mammals when caribou were absent.

A visual comparison of pups on Isle Royale and in Minnesota suggests that the pups on the island were faring well. I first saw the East Pack pups in late July. Subsequently I saw four pups, weighing between 8 and 13 kg, live-trapped in northern Minnesota in late September. By comparison, the Isle Royale pups seen 2 months earlier weighed about 9-12 kg. This is within the range of weight of captive pups of the same age (Kuyt 1972), and is higher than weights of pups caught in Minnesota, where there was a food shortage (Van Ballenberghe and Mech 1975).

Two pups of the East Pack were seen about a month later. Growth in the tntervening period was obvious; weight was estimated at 16 kg. They appeared full-bodied, with well-developed coat and guard hairs. These two pups were larger and appeared heavier than the four pups caught in Minnesota a full month later.

These observations of Isle Royale pups suggest that the midsummer food supply, at least in 1973, was sufficient for normal growth and development. However, there can be great differences in pup weights even within a single litter (Van Ballenberghe and Mech 1975). There was some evidence of retarded winter pelage development among some pups in the East Pack in February 1974 (Fig. 73). Nonetheless, winter observations of this pack since 1972 indicate rapid numerical growth, suggesting high pup survival from 1971 through 1973 (three successive litters).

Fig. 73. Probable pup in East Pack, 1974. Note lack of characteristically stiff guard hairs.

Relations with Nonprey Species

In winter, wolves encounter scavengers for which moose carcasses are a principal source of food. Besides the red fox, many birds also utilize wolf-killed moose—primarily the raven, gray jay, black-capped chickadee, and an occasional eagle. Only the fox and raven will be considered here.

Red Fox

While wolves were seen chasing foxes six times in winter 1972-74, none was caught. Foxes can often run on light snow crusts where wolves break through, and they invariably outrun wolves when chased overland in snow. In the only chase seen on ice, the fox had such a long head start that it reached the shore with no trouble. In 1972, the East Pack was observed just leaving a fox it had killed on the open ice of Malone Bay. The area was matted with wolf tracks, and much hair had been pulled from the fox, though it was not eaten.

The fox's ability to outrun wolves in most snow conditions may be an important reason for its continued coexistence with wolves on Isle Royale. Coyotes, however, disappeared from the island around 1957, less than a decade after the arrival of the wolf. Foxes have thrived recently on Isle Royale, and perhaps even increased after the disappearance of coyotes. While foxes have been observed on Isle Royale since the mid-1920s, long-time island residents report that foxes were uncommon, at least relative to coyotes, before wolves became established.

Moreover, less competition for food resources exists between wolves and foxes than between wolves and coyotes. Johnson (1969) reported that snowshoe hares were the most important year-round food for Isle Royale foxes, and that at certain seasons they made extensive use of insects and fruit. Coyotes relied heavily on moose carcasses. Wolves apparently eliminated coyotes on Isle Royale (Mech 1966; Krefting 1969; Wolfe and Allen 1973), probably through direct killing and competition for food.

Wolves occasionally were indifferent to the presence of foxes. In 1973, the West Pack bedded down on the ice after feeding on a moose carcass. Soon a fox approached, cautiously staying out of sight of the wolves when possible. At the carcass, the fox chased away several ravens and woke the wolves in the process, but they merely raised their heads for a brief look.

During winter periods when foxes were unable to catch snowshoe hares because of deep snow they relied heavily on carcasses of wolf-killed moose (Fig. 74). Foxes have difficulty penetrating the thick hide of a moose—they depend on wolves not only to kill the moose but also to open it up. In winters when utilization of kills by wolves is less than usual, moose carcasses may attract a large number of foxes—as many as 10 at one time in 1972 (Appendix G).

Fig. 74. Wolf kills provide food for scavengers such as red foxes.

Raven

Ravens on Isle Royale in winter are almost entirely dependent on food indirectly provided by wolves (Fig. 75). Ravens often accompany the large packs in their travels, sitting in trees when the wolves stop to rest. Fresh kills draw ravens from miles—28 ravens were seen once on a moose carcass. Ravens also eat wolf scats, especially fresh ones with much incompletely digested meat. Similarly, they feed not only on fresh mountain ash fruit but also on fox scats that are loaded with fruit remains.

Fig. 75. Ravens are found at virtually all wolf kills.

Since ravens and wolves often feed on the same carcasses, there is much interaction. Ravens seem to tease resting wolves, swooping low over their heads, landing nearby and hopping close, further arousing the wolves. (Murie 1944; Crisler 1958; Mech 1966). Wolves, in turn, leap at ravens in the air, stalk them on the ground, and scatter them from kills.

In February 1974, Don Murray and I were circling a kill of the West Pack, with four wolves resting nearby. Suddenly a wolf made a couple of quick bounds—it had caught a raven, something Murray had not seen in 16 winters of flying on Isle Royale (Fig. 76). The wolf shook the raven vigorously in its mouth, then trotted by two other wolves, lay down on its belly and shook it again. Another wolf followed with great interest but was repulsed by a snap from the prize-holder. Finally, the wolf with the raven buried it in snow among some alders and trotted out to greet the other wolves. Next, it dug out the raven and paraded around with it in its mouth, always refusing to let the other wolves inspect it closely. After 15 minutes of this activity we left, but returned an hour later to find the wolves still playing with the raven's carcass. One wolf buried it below a shelf of ice next to shore, then stood above it while another wolf closed in on the buried trophy. When the wolf below came within 2 m, the one above leaped off the ledge and rolled the other over. A brief chase ensued, and then the whole pattern was repeated. The following day the wolves were gone, leaving the raven carcass in the snow.

Fig. 76. Wolves playing with carcass of raven they caught near a kill.