Adaptive management for reintroductions: Updating a wolf recovery model for Yellowstone National Park

Ecological models have their greatest potential for conservation in the context of adaptive management, but there are few examples where mode's have been updated as new data have become available. We update a predator-prey model built to anticipate the consequences of wolf (Canis lupus) reintroduction into Yellowstone National Park, with new data accumulated since wolves were released in 1995. Observed response to wolf recovery allows us to evaluate our ability to predict system dynamics and thereby address concerns about future impacts of predation on elk (Cervus canadensis) numbers and harvest by hunters. Structural assumptions of the model include a dynamic carrying capacity for elk, K-elk(t), varying as a function of winter severity and summer forage production. During severe winters the aerial extent of Yellowstone's Northern Range gets smaller resulting in density-dependent migration of elk outside the park where they are subject to hunter harvest. The updated model predicts that hunter harvest of elk will cause a decline of mean herd size relative to that expected without harvest. Wolf predation results in a further 21% reduction in elk herd size with wolf consumption averaging approximately 1035 elk annually. Elk harvest is reduced by wolves yet annual hunter harvest was sustained at an average of 1089 elk. Predation and hunter harvest is density-dependent providing a stabilizing influence that reduces the risk of severe elk population decline. Using simulation models in adaptive management of the wolf-ungulate system in Yellowstone reinforces agency management policies, especially the density-dependent harvest quota for elk. (C) 2005 Elsevier B.V. All rights reserved.