Spatial application of a predictive wildlife occurrence model to assess alternative forest management scenarios in northern Arizona

Abstract

Wildlife species of conservation concern can present forest managers with a particular challenge when habitat needs appear to be in contrast with other management objectives, particularly fuel reduction to reduce wildfire risk. Proposed actions can be opposed by stakeholders, delaying management activities until a resolution is met. In the southwestern USA, the primary goal of forest management is to reduce the risk of severe wildfire through forest restoration treatments. The USDA Forest Service has designated the northern goshawk (Accipiter gentilis) a management indicator species in this region. However, it has been difficult to achieve a common understanding of goshawk habitat needs among forest stakeholders. We combined two separate and complementary modeling approaches – a statistically based occurrence model and alternative forest treatment models – to yield predictions about forest management effects on the goshawk in ponderosa pine-dominated forests (Pinus ponderosa) on the Kaibab Plateau, Arizona. Forest treatment models were developed based on USDA Forest Service recommendations for goshawk habitat and post-treatment data from ecological restoration experiments, both of which were also components of forest treatment guidance from the Kaibab Forest Health Focus collaborative planning effort. All treatment alternatives resulted in a 22–26% reduction in estimated goshawk occurrence, but the declines were not uniform across the study area, varied by forest type, and were not as large as the effects of recent and severe wildfire (44% reduction in occurrence). Considering the controversial history of forest management with respect to the goshawk, it is prudent to interpret results from this study in the context of tradeoffs between wildfire risk reduction and wildlife habitat quality that can be effectively evaluated through science-based collaborative assessment and planning. While developed for a specific, high-profile species in the southwestern USA, the approach is applicable to many other species whose occurrence has been monitored over multiple years.