资源环境科技发展态势分析平台

The conservation and management of species-at-risk requires periodically collecting information about their distributions and abundances. A comprehensive monitoring plan should, in addition to monitoring the population itself, also assess the status of habitat elements that are key factors in species survival. Places where animals seek safe and secure places to rest are such key habitat elements. We used previously published models to predict resting habitat for fishers (Pekania pennanti) throughout much of their range in California. Unique to this work is that the two models (northwestern California and southern Sierra Nevada) were developed using, as predictors, variables directly from a national plot-based forest inventory program called Forest Inventory and Analysis (FIA). Using these models, relative resting habitat suitability can be estimated at each geographically relevant plot in the FIA system every time the plot is resampled. We applied these predictive models to data collected at 3 time periods over an approximately 20 year period to evaluate the trend in predicted fisher resting habitat. None of the 8 national forests, 4 in the northwestern California region and 4 in the southern Sierra Nevada region nor either of these 2 regions as a whole exhibited trends in predicted resting habitat suitability that were significantly increasing or decreasing. Predicted resting habitat suitability tended to be lower on private land than public land, in both regions. As expected, plots that were disturbed by fire exhibited a decrease in resting habitat suitability but, surprisingly, the few plots within harvest unit boundaries had indistinguishable values before and approximately 7 years after the harvest. Using FIA data for future assessments of habitat value will avoid the significant cost incurred when the data need to be collected repeatedly using different data and a field protocol that may vary. We anticipate that the FIA program will continue to be the preeminent plot-based vegetation survey in the United States, and the data to run the resting habitat models will be available every 10 years. Moreover, access to routinely updated plot-based data provides the only way we can envision sampling something as fine-scaled as resting habitat over thousands of square miles of potentially suitable habitat. We hope our example encourages others to parlay the FIA data into a predictive model of fine-scale habitat features that is relevant to other species. Demonstrating the utility of our models should also encourage managers to use the predictions to evaluate the status of fisher habitat in California.