Wolves in Yellowstone, Jackson Hole, and the North Fork of the Shoshone River: Simulating Ungulate Consequences of Wolf Revovery

Abstract

The gray wolf (Canis lupus) was extirpated from Yellowstone National Park by U.S. Government personnel during 1914-1926. Since then, occasional reports of wolves in Yellowstone National Park have been recorded (Weaver 1978), but no recent records exist of wolves breeding in the park. In recent years, public attitudes towards predators have changed such that predators are more commonly viewed as an integral component of natural ecosystems (see e.g., Mech 1970, Despain et al. 1986, Dunlap 1988). An increasing proportion of the American public desires that wolves be reestablished in Yellowstone National Park (McNaught 1987, Bath 1991). ln 1987, the U.S. Fish and Wildlife Service approved a Recovery Plan for the Northern Rocky Mountain wolf (U.S. Fish & Wildlife Service 1987). Before proceeding with wolf recovery, however, Congress appropriated funds in 1988 and 1989 and directed that studies be conducted by the U.S. Fish and Wildlife Service and the National Park Service to determine the effects of wolf recovery on ungulate populations. Boyce (1990) developed a predator-prey model for ungulate populations in Yellowstone National Park as a part of this Congressional charge to determine the probable outcome of wolf recovery. Our purpose is to expand upon the simulation model of Boyce (1990) to predict the probable consequences of wolf reintroduction in Yellowstone National Park to ungulate populations in Jackson Hole and along the North Fork of the Shoshone River. As in the previous model, this model allows the user to choose among several likely management scenarios. By manipulating alternatives, the user of the model can explore the consequences of management actions. In particular, it is essential to be able to anticipate if wolves will be culled if they leave the parks, if poaching can be controlled within the park, and if hunting for bison and elk will continue in the Yellowstone River valley north of Gardiner, Montana. Any such model must incorporate the natural variability in the environment, because the vagaries of climate can have enormous effects on ecological processes. Therefore, the model is a stochastic one, i.e., it contains random variation in climatic variables. Such stochastic model structure is important because it helps to educate the user that it is impossible to predict precisely the consequences of wolf recovery. It is not the purpose of this effort to offer recommendations for whether wolf recovery should take place, but rather to provide resource managers with an additional tool which will assist them in making that decision.