Rapid, adaptive monitoring of a declining species

Passive acoustic monitoring (PAM) has proven effective as a means of monitoring species at broad spatial scales, but implementing a monitoring effort with limited information may require an iterative approach to survey design. We illustrate this challenge using the Southern California, USA population of the California Spotted Owl (Strix occidentalis occidentalis), which has declined >50 % over the past 30 years and faces multiple ongoing threats. Monitoring goals were: assessing the owl's distribution and locating individuals to facilitate regulatory compliance. Using a preexisting PAM design developed for this species, we deployed >200 recording units across the region, then used machine learning and manual verification to identify owl vocalizations. We also used known owl territory locations and remote sensing data to create a map of core owl habitat (nesting/roosting areas). We fit a set of occupancy models and found that: occupancy was low (0.20–0.298) and that both occupancy and detection were positively related to core habitat. To refine our design, we conceptualized the focal species' space use in terms of spatial home range, acoustic home range, and territory, interpreted our results and newly available movement data with this lens to determine the appropriate survey grid resolution. The new design should increase detection and baseline occupancy, improving statistical power and better meeting monitoring goals. Rapidly adapting monitoring programs to suit the target species and its home ecosystem may be necessary to effectively inform conservation action.