Evaluating Management Risks Using Landscape Trajectory Analysis
Ecosystem management requires an understanding of how landscapes vary in space and time, how this variation can be affected by management decisions or stochastic events, and the potential consequences for species. Landscape trajectory analysis, coupled with a basic knowledge of species habitat selection, offers a straightforward approach to ecological risk analysis and can be used to project the effects of management decisions on species of concern. The fisher (Martes Pennant) occurs primarily in late-successional forests which, in the Sierra Nevada mountains, are susceptible to high-intensity wildfire. Understanding the effects of fuels treatments and fire on the distribution of fisher habitat is a critical conservation concern. We assumed that the more a treated landscape resembled occupied female fisher home ranges, the more likely it was to be occupied by a female and therefore the lower the risk to the population. Thus, we characterized important vegetation attributes within the home ranges of 16 female fishers and used the distribution of these attributes as a baseline against which the effects of forest management options could be compared. We used principal components analysis to identify the major axes defining occupied female fisher home ranges and these, in addition to select univariate metrics, became our reference for evaluating the effects of landscape change. We demonstrated the approach at two management units on the Sierra National Forest by simulating the effects of both no action and forest thinning, with and without an unplanned fire on vegetation characteristics over a 45-yr period. Under the no action scenario, landscapes remained similar to reference conditions for approximately 30 years until forest succession resulted in a loss of landscape heterogeneity. Comparatively, fuel treatment resulted in the reduction of certain forest elements below those found in female fisher home ranges yet little overall change in habitat suitability. Adding a wildfire to both scenarios resulted in divergence from reference conditions, though in the no action scenario the divergence was 4 times greater and the landscape did not recover within the 45 year timeframe. These examples demonstrate that combining the results of forest growth and disturbance modeling with habitat selection data may be used to quantify the potential effects of vegetation management activities on wildlife habitat.