Susan Townsend;; Elizabeth Edson, Matthew McKown, Abram B. Fleishman, Jeff Schlueter, Zoe Gustafson, Tammy Lim

Portions of the San Francisco Bay Area burned in the summer of 2020. To understand how wildlife communities were affected by the wildfires, we monitored burned and unburned sites using acoustic sensors and cameras. Forty paired sensors, audiomoths and camera traps, were set at 0.5 km intervals in East Bay Regional Parks and Contra Costa Water District Los Vaqueros Reservoir beginning in October 2020. Our goal was to compare diversity (species richness), detection rates (intensity of use), and occupancy estimation (abundance) between the burned and unburned sites and latency to recovery using the unburned area as a comparison. We used a cloud-based acoustic processing pipeline to run an opensource bird call detection model (BirdNET). Wildlife Insights, camera trap data management and image recognition software, was used to process camera data. Image recognition output was confirmed by trained observers. Both audio and image data were used to derive effort, detection rates and single season occupancy estimates using a R-Markdown script developed for the Marin Wildlife Picture Index Project. We have tools and approaches that can be easily replicated; these protocols and R-Markdown scripts are available for use and are based on internationally accepted metrics for measuring trends in biodiversity. Understanding diversity, intensity of use and abundance as well as documenting rare species all provide the most basic, but very important information about the health of our ecosystems, and perhaps more importantly, how they may be changing.

Wildfire and Disturbance Response  InPerson Presentation