Ecological correlates of regional variation in life history of the moose Alces alces: comment - response to B.-E. Saether, Journal of Wildlife Management, vol. 49, p. 977, 1985

Ecology, July, 1998 by Michel Crete

Saether et al. (1996) compared some life-history characteristics of four Norwegian moose (Alces alces) populations occupying ranges of varying quality to test the food-limitation hypothesis and concluded (p. 1499) that "a stable high-density equilibrium between moose and their food resources is unlikely to occur." However, their data set may not justify such a conclusion because they neglected to consider the role of a major component in the dynamics of the moose populations studied: human harvests. On the basis of the simple models proposed by Caughley (1976) for plant-herbivore interactions and regulation of large-ungulate populations, Saether et al. (1996) cannot use population dynamics of moose in Norway to reject the hypothesis of regulation by competition for food.

To alleviate any semantic confusion, it is worth defining four concepts that are central to the understanding of Caughley's (1976) approach to population dynamics of large herbivores: K carrying capacity (KCC; Macnab 1985), limiting factors, regulating factors (Messier 1991) and sustained yield (SY). KCC represents "the equilibrium reached between herbivores and their food supply" after dampened oscillations (Macnab 1985:404). Limiting factors refer to "any processes that quantifiably affect population growth and are responsible for year-to-year changes in the rate of population growth" (Messier 1991:378). Regulating factors designate "any density-dependent processes that ultimately keep populations within normal density ranges" (Messier 1991:378). Thus regulating factors represent a subset of limiting factors, characterized by negative-feedback mechanisms that depress population growth as animal abundance increases. Finally, SY equals the annual surplus of births over deaths observed when an ungulate population below KCC is increasing; if these animals are harvested, the population remains stable below KCC.

Sinclair and Arcese (1995) described three alternative hypotheses related to regulation of large herbivores: the predator-regulation hypothesis, the predation-sensitive food hypothesis, and the surplus (or food-limitation) hypothesis. In the first case, density-dependent predation causes herbivores to stabilize at low density (relative to KCC) with access to ample quality forage. Moose populations preyed upon by wolves (Canis lupus) and black or brown bears (Ursus americanus, U. arctos) support this hypothesis (Messier and Crete 1985, Crete 1987, Gasaway et al. 1992). According to the second hypothesis, herbivore numbers stabilize at a density lower than KCC because predators remove some vulnerable herbivores that would otherwise survive. This model might apply to Isle Royale moose where only wolves prey on moose (McLaren and Peterson 1994). In the third case, the food-limitation hypothesis supposes that competition for forage causes density to stabilize at KCC, predators having no influence on herbivore density. This hypothesis should apply to Fennoscandian moose because wolves and brown bears have been reduced to insignificant numbers during the current century (Cederlund and Markgren 1987). Given the food-limitation hypothesis, one should still expect density to fluctuate around KCC, as observed for moose on Isle Royale (McLaren and Peterson 1994) or wildebeest (Connochaetes taurinus) in the Serengeti (Sinclair and Arcese 1995), due to the combined effects of limiting factors (e.g., winter harshness) and forest dynamics (e.g., forest fires).

In order to test the food-limitation hypothesis of population regulation, Saether et al. (1996) had to study moose populations in the proximity of KCC because, in large mammals, regulation generally operates at densities approaching KCC (Fowler 1981). However they provided no figures on either absolute moose density, or density with respect to KCC. Crete (1989) estimated KCC in an area of deep snow of eastern Quebec to exceed 2 moose/[km.sup.2]; the estimate varied between 3.6 and 6 animals/[km.sup.2] in southwestern Quebec due to greater forage production. Very few data have been published on Fennoscandian moose density, and none to my knowledge for unharvested populations. Bergstrom and Vikberg (1992) reported that the density increased from 1.3 to 5.7 moose/[km.sup.2] in a forested enclosure of central Sweden before being reduced by hunting. Most likely, KCC must approach 10 animals/[km.sup.2] in very productive areas of Fennoscandia, particularly because of the limited snow cover (Saether et al. 1996).

The four moose populations studied were harvested annually at a rate of 0.33-0.50 animal/[km.sup.2] (Hjeljord et al. 1994). With the information provided, it is impossible to compute which proportion of the population this SY represented, but such yields are comparatively high (Crete 1987). Most likely, annual harvests have kept densities much below KCC, particularly for the Alpine population that occupied a poor range (Saether et al. 1996). Not surprisingly, Saether et al. (1996) found no evidence of regulation driven by competition for forage.