Ronan Hart, Craig Thompson, Jody Tucker, Sarah Sawyer, Stephanie Eyes, Saba Saberi, Zhiqiang Yang, Gavin M Jones

Forest disturbances are a natural ecological process, but climate and land-use change are altering disturbance regimes at an unprecedented rate, posing significant threats to biological communities and species of concern. Our aim was to develop an automated habitat monitoring system for the Southern Sierra Nevada Distinct Population Segment of fisher (Pekania pennanti) in California, USA to investigate long-term habitat trends and the effects of a recent megadrought and numerous megafires on fisher habitat. We used detections of female fishers (n=330) from a standardized monitoring program to develop a dynamic species distribution model using the random forest algorithm in the Google Earth Engine environment. We found that female fisher habitat remained relatively stable from 1985–2011 but declined by nearly half (48%) between 2012 and 2022, corresponding with a period of widespread forest mortality from drought and wildfire. The majority of fisher habitat loss occurred within wildfire perimeters (65%), where declines in habitat quality were associated with moderate- and high-severity fire. Female fisher habitat was more likely to burn at moderate- and high-severity than was expected by chance. Our findings emphasize the urgent conservation needs of this distinct population segment of fishers, highlighting the threat posed by novel disturbance regimes. Our results demonstrate the importance of monitoring for understanding species status, as the status of fisher habitat across the entire southern Sierra Nevada range following recent disturbances was not known. More broadly, our implementation of a cloud-based automated habitat monitoring system shows the necessity of up-to-date habitat information to apply conservation measures in rapidly changing environments and the potential for using habitat monitoring systems to investigate ecological questions of basic and applied relevance (e.g., wildfire-habitat relationships).