Trophic interactions in two northern wolf-ungulate systems

Abstract

Four population models for northern ungulates are formulated based on specific regulatory mechanisms: FOOD model, PREDATION-FOOD one-state model, PREDATION-FOOD two-state model and PREDATION model. In this paper I contrast the dynamics of the North American moose (Alces alces), a non-migrant ungulate, with the dynamics of the barren-ground caribou (Rangifer tarandus), a migrant ungulate. Predation by the grey wolf (Canis lupus) and by bears (Ursus arctos and U. americanus) is considered. Moose densities in North America are generally below 0.5 animals km-*2. In most areas, density-dependent food exploitation seems to operate at densities greater than 1.0 moose km-*2. Results from 27 studies were used to assess the functional and numerical responses of wolves to changing moose density. Per capita killing rate and wolf density were related (P = 0.01) to moose density following a Type I1 response. Wolf predation rate proved to be density dependent from 0 to 0.65 moose km", and inversely density dependent at higher moose densities. An empirical model based on these results suggests that moose would stabilise at high densities in the absence of predators (FOOD model) or in the presence of wolves only (PREDATION-FOOD one-state model). If moose productivity level is diminished through deteriorating habitat quality, or if bear predation is added, then a low-density equilibrium (0.2-0.4 moose km*2) is predicted (PREDATION model). The demography of the George River caribou herd in northern QukbecILabrador is reviewed for the period 1955-93. The herd increased from 5000 animals in 1955 to 644000 (c. 1.5 animals km-*2) in 1984, following an exponential growth rate (r) of 0.16. The herd reached a plateau of c. 725 000 animals in early 1990. A decline is predicted for the near future because of severe habitat degradation. In the 1980s, nutritional stress has been documented, especially in summer when caribou used traditional ranges depleted of key forage resources such as lichens and tundra shrubs. Food competition is viewed as the dominant regulatory factor (FOOD model), although the nature of caribou-habitat interactions may generate longterm fluctuations in numbers because of a time-delay effect. Wolf predation does not appear to be an important mortality factor capable of regulating the George River herd. The above case studies can be interpreted within the context of the 'food chain dynamics' theory. For the non-migrant ungulate (moose), a three-link system is proposed where predators hold down grazers, which in turn use only a small fraction of annual forage production. For the migrant ungulate (caribou), a two-link system is proposed where high-density grazers exert a strong regulatory control on plant productivity and where predators have minimal effect on grazers.