- Forests in the western United States have changed drastically over the past 150 years given a long history of utilization (harvest and grazing) and associated fire exclusion. These actions have altered the composition and structure of these forests as well as affected ecosystem function. Current policies for federal land management have shifted to improve forest health and resiliency. However, wildlife species sensitive to forest disturbance (e.g., wildfire, timber harvest) often conflict with forest management goals and activities that reduce the amount of available intact habitat and connectivity for mesocarnivores.
Pacific marten (Martes caurina) populations have become fragmented and constricted throughout their western range due to a combination of increased wildfire, timber harvest, climate change, and historic fur trapping. Martens rely on rest structures, often live trees, snags, or logs where martens sleep, for protection from inclement weather and predation risk. Rest structures are considered a limiting habitat element; their abundance, type, and distribution has been suggested as a way to evaluate habitat quality. Our study objective was to better understand the selection of resting habitat by martens in order to help foresters and biologists better manage marten species on the landscape.
Our first objective was to understand whether marten rest structure selection was correlated with seasonal activity patterns. Pacific martens in the Sierra Nevada Mountains of California appear primarily nocturnal during winter and diurnal during summer. Diurnal rest sites (live trees, logs, snags, stumps) in both seasons have been described previously, but nocturnal summer rest sites have not. We hypothesized that all resting structures used by martens would be specialized (e.g., larger, older) compared to available structures, but that nocturnal structures would be larger in diameter than daytime sites and restricted to structures with cavity microsites. From 2009 – 2013 and 2015 – 2017, we collected radio telemetry on 37 martens (23♂, 14♀) locating 140 diurnal and 32 nocturnal rest structures. We did not detect a difference in the size of structures between diel periods; however, martens used rest structures that were much larger than the majority of measured random structures (used = 95.5 ± 31.8cm; random = 52.1 ± 25.2cm, x̅ ± SD, t = 15.1, p<0.001). The selection of rest structure and microsite types did not differentiate between diel periods. There was some evidence of increased use of cavity and subterranean microsites at night. Our results emphasize the importance of conserving a diversity of structure types and sizes likely needed for marten habitat.
Our second objective was to examine multiscale habitat relationships at Pacific marten rest structures using fine-resolution vegetation data (30m airborne LiDAR) and ground-based data. Using a moving-window framework to compare selection, we optimized 14 covariates at 12 spatial scales (30m-990m). We monitored martens from 2009-2012 and 2015-2017 (n=312 resting structures, 31 martens), and then compared used vs. randomly-sampled locations (nrand=624) in order to develop multivariate habitat selection models. Our top model included trees per acre (990 meter scale) and elevation (900m), suggesting that martens select for increased tree cover at higher elevations at the home range scale. Increased structural complexity and canopy cover surrounding rest structures (270 and 30 meters, respectively) increased probability of selection. Because martens selected locations with vegetation characteristics optimized at 30-270m, 270m may be an appropriate scale to consider for management, for instance, establishing leave islands or focal areas for restoration. We provide the first evaluation of marten habitat using LiDAR, which can be broadly and accurately extrapolated for management planning and restoration prioritization.