An eight year field study (1966-1974) in the Denali region of Alaska (lat.,63°N, long.150°W) focused primarily on two adjacent wolf packs in a largely unexploited ecosystem, to obtain information on their social dynamics and interactions with moose, sheep, and caribou. The "Savage" pack used a home range of about 600 square miles, and the "Toklat" pack about 1,000 square miles; both were territorial to essentially the full extent of their ranges. There was frequent ritualistic scent marking by both packs (usually the alpha male and female), especially where their ranges intersected, and this appeared to be mainly for territorial maintenance. Normal late winter pack sizes were 12-15 for the Savage pack, and about the same or a little larger for the Toklat pack. Historical records suggest little change in spatial patterns and pack size for these and surrounding packs for at least 30-40 years. Moose density averaged about .70 animals per square mile in the Savage range, and about .26 in the Toklat range, with no indications of major fluctuations for at least 30-40 years, even after severe winters. Sheep densities have fluctuated widely since at least the 1920s, with exponential increases to peaks followed by relatively sharp declines accelerated by severe winters. Normal highs and lows were roughly 3.3 and 1.3 sheep per square mile in the Savage range, and 1.0 and 0.2 in the Toklat range. There were predictable summer-winter shifts in distribution within each range for both species, keyed mainly to snow depths. Survivorship data suggest patterns typical of ungulates, including high first-year mortality. Adult sex ratios indicate a preponderance of females. Caribou of the Denali Herd annually migrated over an area of up to 6,000 square miles. They were present in the Savage and Toklat ranges – moreso in the latter – for much of each summer, but largely absent each winter. The herd decreased from 20,000-30,000 caribou prior to the 1940s to about 1,500 by 1972, due mainly to emigration. The Savage pack was a kin-selected extended family group, with a strong, relatively stable dominance hierarchy and extra-hierarchial allegiances. Hales generally were dominant over females, except when a female was older. Many of the high ranks were determined by an age of 2-3 months. Unlike dominance, roles were situation specific; leadership was usually assumed by a high ranking wolf, most often the beta male. The alpha male asserted himself in a central role during only a relative few, probably "critical" activities, for example scent marking, courtship and mating, and encounters with aliens. The rule in the Savage pack appeared to be only one litter per year, produced by the alpha pair. Other males (including the beta) were usually prevented from mating by dominance actions of a higher ranking male(s). It appeared that at least some of the females were inhibited by failure to enter normal estrus, related to the presence of a higher ranking female and/or absence of a higher ranking male. The Toklat pack had a less stable, multifamily social structure during much of the study and in the past. This probably resulted from pronounced summer-winter changes in prey density (much greater change than in the Savage range), due to seasonal presence of caribou. Coincident with the latest decline of the caribou herd to a very small size, the Toklat pack underwent a change toward a Savage-type (single family) social structure, and some of the wolves departed to form a new ("Wonder Lake") pack still using a portion of the Toklat range. During summer the wolves based their activities at long established homesites (dens and rendezvous sites); nine sites were found for the Savage pack, and 16 for the Toklat-Wonder Lake packs. Annual selection of sites and patterns of switching between them (resulting in switching of predation impact) were related to prey availability, but also to human disturbance, probably training for the pups, and possibly other factors. Homesites were not used in winter; the young wolves now traveled with the pack and began to acquire major hunting skills, progressing through three phases of learning: (1) hesitance and fear (5 to 7-8 months of age), (2) overreaction (8 to 10-11), and (3) effective participation with guidance (10-11 to 23-30). Sinter splitting occurred in both packs, but with a basic difference. In the Savage pack it occurred mainly at large pack sizes and separated inexperienced from experienced pack members, leading to increased mortality and/or emigration and a reduction back to normal pack size. In the Toklat pack winter splitting occurred at normal sizes and produced a more even distribution of experienced wolves in the subunits, apparently enabling the pack to better cope with its past major summer to winter decrease in prey density. Both packs periodically trespassed deep into each other's territory during winter. Virtually all trespasses appeared to be intentional and involved much aggressive behavior; food was often obtained, but calculations suggest only some of the trespasses were motivated by food shortage. Activity budget analyses indicate greater proportions of time spent hunting in winter than in summer, and in winter more time spent hunting and more overall variability for the Toklat pack than for the Savage pack; the winter between-pack difference is attributed mainly to the lower, more variable prey density of the Toklat range. Hunting activities involved a high level of intrapack cooperation, including deployment of individuals at different locations relative to the prey prior to beginning a chase, maneuvers to force prey into difficult escape terrain, and the use of a decoy. The primary summer prey was moose {mostly young) for the savage pack, and caribou (mostly adults) for the Toklat pack, with periodic prey switching in response to spatio-temporal changes in availability of the primary species. In winter both packs relied mainly on moose and to a lesser extent sheep and caribou, including many scavenged carcasses. Winter predation was highly selective of young and old prey animals. Both the summer and winter data suggest a high percentage of all wolf kills replaced deaths about to occur from other causes, a result to be expected at normal prey densities but not likely when prey density is lowered substantially (e.g., through heavy human harvests). Mean food "adequacy" per wolf per day was calculated for both packs each winter, based on observed rates of consumption and requirement estimates corrected for winter-to-winter travel variations. There was probably prolonged food shortage during one of five winters for the Savage pack, and in two of five winters for the Toklat pack. A computer simulation model of the Denali wolf-moose-sheep interaction was developed from field data on 1) environmental conditions (snowfall), 2) moose and sheep reproduction and non-wolf mortality, 3) wolf reproduction, mortality, and social behavior, and 4) wolf predation and scavenging in relation to prey density and other factors, replacement mortality, and potential adjustments in wolf territory size. The analysis suggests a form of coevolution between the wolves and prey: under natural conditions there is a close relationship between predation impact and prey productivity, but the system still has enough resilience to absorb the added impacts of severe winters (at least as experienced since 1925). It is shown how heavy moose harvests can trigger long range declines, lowering densities to the point where wolf predation and severe winters exert a proportionately greater' impact and lead to further decline, even after the human harvesting is stopped., Examples of this sequence are seen in a number of declines currently underway in the North, along with the resulting contentious programs of wolf control. A methodology is shown for largely avoiding such problems, by using field data on spatial distribution of the ungulates and wolf kills to determine sustainable harvests. Caribou dynamics and management present a different problem because of dispersal-related coupling among all seven of the major Alaska-Yukon caribou herds, including the Denali Herd. Historical records suggest there was a more-or-less "cyclic" pattern of peaks and prolonged low densities for each herd. It is concluded that wolf predation had virtually nothing to do with causing a decline from high density, but that it was the primary factor which held a herd down once dispersal reduced it from a food-limited peak. The prolonged low enabled range (food) conditions to recover, whereupon the herd built up to a subsequent peak following dispersal from another herd which had just peaked. At present the Alaska-Yukon caribou herds have approached a condition of synchrony at low density -due to heavy harvests it is argued, and a range of harvest restriction-wolf control recovery options is discussed. It is recommended that once recovery is achieved, human exploitation should proceed by pulse-harvesting rather than according to the traditional approaches of sustained yield management. / Science, Faculty of / Zoology, Department of / Graduate

Although cooperative behavior is generally acknowledged to occur among wolves, there is a lack of systematically collected data confirming the extent of development. The objectives of this study were to collect long-term, detailed observations documenting the role of social structure, seasonal influences, and individual participation in wolf pack cooperative activities. Individual cooperative strategies were associated with age, sex, and social positions and critically compared with results of similar studies. Emphasis was placed on quantifying group and dyadic relationships, focusing on reproductive strategies and dominance structure. Additional data were collected on denning behavior, maternal care behavior, scent marking, and spontaneous individual and group howling.