NATIONAL PARK SERVICE Wolf Ecology and Prey Relationships on Isle Royale

In biological communities, homeostatic mechanisms evolve which increase the efficiency of energy use in the ecosystem (Watt 1970). Larger numbers of species in any given system contribute to more complex energy pathways and thus promote stability (MacArthur 1955).

Islands may have only a limited number of species, and because of this inherent simplicity fewer stabilizing mechanisms exist and fluctuations in populations are more likely to occur. Because dispersal is limited, island populations may reach higher densities than mainland populations of the same species. Conclusions drawn from ecological studies on islands must always recognize these unique characteristics.

Moose Population Regulation

During the present wolf-moose studies on Isle Royale (1970-74) it was clear that food supply played a major role in moose population dynamics, as evidenced by reduced calf production and increased signs of malnutrition compared to a decade earlier, when calf production was high and there was no evidence of limitation by food supply. It appeared at that time that intensive wolf predation was "... maintaining the moose population below the level at which food would restrict it" (Mech 1966:167). This initial interpretation has now been revised.

In spite of imprecise population estimates, there were indications of an increase in moose numbers during the 1960s. Certain environmental factors, influential during the same period, partly account for changing relationships in the browse-moose-wolf complex. Burns which formerly contained abundant moose browse became less important to moose as forest succession continued. Since the early 1960s significant utilization of these burns has been limited to winters of little snow and unrestricted mobility for moose; in recent years. (1969-74) high moose densities had to be supported in more mature forest types. Recent severe winters have emphasized the influence of environmental factors on condition and, ultimately, on population levels of moose.

In response to changes in moose vulnerability, wolf predation began to increase in 1969, especially on calves and young adults. Kill rates generally were high, carcasses were underutilized, and a second pack of wolves became firmly established on the island for the first time since these studies began.

Although Pimlott (1967) suggested that highly developed intrinsic forms of population regulation have not evolved among North American ungulates because of intensive predation throughout their history, moose do exhibit important responses to environmental influences which introduce a degree of intrinsic control to population levels. Since moose are not highly preferred prey of wolves when smaller ungulates are available (Murie 1944; Pimlott et al. 1969; Mech and Frenzel 1971), buffer species could have reduced the potential effect of predation on moose during their evolutionary development and therefore resulted in moose effective intrinsic controls. Social behavior which effectively disperses a moose population may prevent high densities when the food base is inadequate. Agonistic behavior apparently spaces moose efficiently throughout their habitat and contributes to their solitary social organization (Houston 1974), although moose may congregate where food is abundant (Geist 1974). Furthermore, sexual maturity in adult moose may be delayed by at least a year in response to a reduced food intake in the first year of life, and calf production and early survival are related to nutrition (Pimlott 1959; Markgren 1969). Thus, the growth potential of a moose population is adjusted to the nutritional level, allowing a rapid increase when densities relative to available food are low, and exerting a stabilizing influence as food supply becomes more limiting.

Irruptions of moose populations, followed by catastrophic losses to malnutrition, are rather rare. The case which is most often cited is that of Isle Royale in the early 1930s. However, the insular nature of this ecosystem, the lack of predation, and an abundant source of food that was highly sensitive to moose browsing (evidenced by the elimination of Taxus canadensis [American yew] as a shrub species) presented a unique set of ecological circumstances. Some naturally regulated moose populations, in the absence of predation, have stabilized without heavy losses (Houston 1968; LeResche 1970), supporting the hypothesis that moose have evolved some effective intrinsic controls.

The effect of climatic conditions such as snow cover (an extrinsic factor) may be determined largely by population density. Snow of a given depth, for example, may introduce more nutritional stress to Isle Royale moose than to mainland populations because of higher densities on the island. In Alaska the influence of snow on moose varied with moose densities and habitat quality (Bishop and Rausch 1974). Snow conditions are an important element in the environmental framework within which moose and wolves interact.

There appear to be no adequately documented cases of wolf predation imposing a long-term limit on ungulate populations independent of environmental influences. However, wolf predation is probably the most efficient natural mortality factor capable of altering prey population levels in response to food and climatic variations (Fig. 109). Wolf predation is a major culling force that has shaped survival patterns, behavior, and physical characteristics of prey throughout their evolutionary history, resulting in prey that are well-adapted to their environment. On Isle Royale, as predation intensified with increased prey vulnerability, the degree of control exerted on the moose population probably increased—especially as prereproductive moose became vulnerable. In such a situation wolf predation is not compensatory, as Errington (1946) defined it, since other mortality factors could not fully substitute for predation. Wolves hasten the death of moose that would die eventually of other causes, but the time element is critically important. Intensive culling ensures that limited food resources are utilized by the most vigorous members of the prey population.

Fig. 109. Wolf predation is essential to the healthy maintenance of a high moose population on Isle Royale.

The role of predation may be clarified if we consider a hypothetical question: what would happen if wolves were removed from Isle Royale? Malnutrition would take over as the principal direct and indirect source of mortality, affecting mainly calves and old moose in late winter and early spring. In the absence of predation, moose weakened by malnutrition may survive for months, competing with healthy moose for limited browse. Calves would be born weaker and smaller than usual, and neonatal mortality probably would increase. Thereafter calf mortality would be low, with the degree of loss to malnutrition in late winter and spring determined by winter weather. Over a series of mild winters the moose population would increase because of low adult mortality and relatively high calf survival. With the return of usual winter conditions mortality would greatly increase, and severe winters probably would cause large losses. Long-term population levels of moose might well be lower without predation because of wasteful browse utilization. Thus wolves are essential to the maintenance of a healthy, high-density moose population on Isle Royale.

Although wolves did not hold the population on Isle Royale below the level at which food supply and environmental conditions affected the welfare of moose, wolf predation increased during the 1970s when moose, particularly calves and young adults, became unusually vulnerable. As long as wolves and moose coexist on Isle Royale, wolf predation will remain the principal mortality factor operating on the moose herd. However, the degree of control exerted by predation will be determined largely by environmental influences (principally food supply) that act on moose reproduction as well as vulnerability to regulate numbers of moose.

Wolf Population Adjustments

The recent partitioning of Isle Royale into two pack territories is a significant departure from the basic pattern of the 1960s. The remarkable stability shown during the 1960s demonstrated that effective natural controls were operative. The primary regulatory influence was believed to be food supply or social controls, or some combination. Until recently wolf densities on Isle Royale and in other regions were not found to exceed one wolf per 25 km² (Pimlott 1967), leading to the inference that this was a maximum density beyond which wolves would not increase, perhaps because of social intolerance (Huffaker 1970). High pup mortality in other wolf populations has been regarded as important in reducing their growth potential, and the finding of an apparently malnourished pup on Isle Royale in 1964 suggests that food shortages during the pup-rearing season were of regulatory significance (Jordan et al. 1967). However, Wolfe and Allen (1973) found that the wolf population, at a low level in the late 1960s, did not increase immediately following a winter when moose vulnerability wa. high and surplus killing was evident; they proposed that social factors might be preventing expansion.

The emergence of a second pack of wolves on Isle Royale and the expansion of their population reflects an increased food supply stemming from greater vulnerability of moose and a higher beaver population. Thus, the level of food resources available to Isle Royale wolves is proposed as the principal underlying variable which determines the number of stable packs and, ultimately, wolf population levels.

In some cases "food supply" for predators may be correlated directly with prey densities, and predator populations will fluctuate in direct response to changes in prey levels. This is especially true of predators that are not as selective as wolves, such as lions preying on certain ungulates species (Schaller 1972) and coyotes preying on jackrabbits (Clark 1972). More selective predators, such as Isle Royale wolves, are particularly apt to respond to changes in numbers of vulnerable moose, rather than prey densities per se. In this context the relationship of moose welfare to environmental influences is critical in understanding wolf population responses on Isle Royale.

At the pack level, food plays an obvious role in regulating group size, since wolf-pack size is related to the size of their prey. Packs feeding on Isle Royale moose have numbered between 15 and 20 wolves, those feeding on elk (Cervus canadensis) and mule deer (Odocoileus hemionus) contained between 6 and 14 members (Carbyn 1974a, b), and wolf packs feeding only on deer usually numbered 7 or less (Mech 1973; Van Ballenberghe 1972). Food level can affect wolf populations directly, particularly through pup mortality and survival (Van Ballenberghe and Mech 1975, Seal et al. 1975). In addition, the complex social organization found within and between wolf packs may provide for subtle influences on physiology and behavior that may be of regulatory importance.

Territoriality among wolf packs apparently is a spacing mechanism which adjusts wolf densities to their food level. This explains the recent reorganization of the Isle Royale wolf population. In response to increased food resources, the single resident pack ranged in winter over only half of the island in 1971, and the second pack became established the following year on the remaining half. Without a shrinkage in the original pack's territory, a second pack could not have materialized. Similarly the increased size of the winter territory of the East Pack in the first 3 years of its existence was correlated with a decrease in the amount of food available to each pack member in winter. This led to considerable spatial overlap and, in at least one case, direct mortality from inter-pack aggression.

Social relationships within the pack may also be sensitive to food supply and thus influence pack size. Social stress, particularly for subordinate wolves, probably increases when food is short. The physiology of individual wolves may be affected by social interaction—perhaps affecting their reproductive capabilities. Social restraints seem to reduce the incidence of breeding among adults in a pack. Dispersal of young wolves may sometimes have significance. At present we do not really understand all these possible mechanisms nor their relative effectiveness.

Environmental conditions certainly influence predator populations—sometimes directly. In northern areas climatic factors may contribute to regulation of predator numbers by increasing the energy costs of body maintenance and normal activities. Pup survival may sometimes be affected (Clark 1971).

Natural regulation of a predator population appears to result from a complex of factors: food supply as the underlying principal variable, mediated by social and/or environmental influences. The nature and relative influence of these factors probably will vary considerably depending on the predator and prey species involved and the environmental conditions which render each interaction unique.