| Summary: | The American marten (Martes americana) has experienced an apparent loss of occupied range over the last 75 years in northeastern California. Extinction selectivity, or relative vulnerability, is non-random and individual traits make some species more extinction-prone than others. Martens possess many life history traits that promote species risk including specialized habitat preference, restricted distribution, low fecundity and high trophic level. Studies conducted elsewhere in North America have shown martens are sensitive to forest fragmentation and to thresholds in landscape characteristics. I used field surveys and Geographic Information System (GIS) data to identify the landscape-scale habitat associations of martens and develop a model that predicts their occurrence. The study area included five national forests (Klamath, Shasta-Trinity, Lassen, Plumas, Tahoe) and Lassen Volcanic National Park. This region encompasses 27,700 km2 of largely mountainous terrain. Systematic surveys of forest carnivores were conducted by U.S. Forest Service personnel at 184 sample units using track plate stations and remotely-triggered cameras. Marten detections were clustered in three distinct areas, occurring at 10.8% of sample units. I investigated marten habitat selection using circular plots created at three spatial extents: 3, 20 and 80 km2. An information-theoretic method was used to rank 89 a priori candidate models. Multivariate models were constructed using combinations of environmental variables hypothesized to be important to marten ecology and management. Predictor variables included elevation, stream density, land ownership, road density, nearby marten detections and landscape metrics of forest vegetation. Forests closely associated with marten reproduction, the most important aspect of their life history, were selectively chosen based on type, size class and canopy closure using the California Wildlife-Habitat Relationships (CWHR) system. I created a composite vegetation index (CVI) by combining these forests ...
|
|---|