Modeling and predicting Mexican spotted owl occurrence in southcentral Colorado using passive acoustic monitoring

Abstract

Survey efforts for wildlife are often limited by time‐intensive and expensive methods. Biologists in Colorado, USA, have monitored the threatened Mexican spotted owl ( Strix occidentalis lucida ) using call‐playback and visual observations for decades. These methods are time consuming and expensive, so our objective was to find a more efficient method using passive acoustic data to quantitatively monitor the Mexican spotted owl population and distribution, and to identify drivers influencing their occurrence. We employed passive acoustic monitoring across the most actively used critical habitat unit by Mexican spotted owls in Colorado to model and predict their occurrence. We collected acoustic samples from 43 survey cells during the 2022 Mexican spotted owl breeding season. We sampled each survey cell for 7 consecutive nights using 3 autonomous recording units. We successfully detected ≥1 Mexican spotted owl at 9 survey cells. Using occupancy analysis, we estimated an occupancy probability of 0.0005 and a detection probability of 0.40 for Mexican spotted owls across our study area. Our models predicted increasing owl occupancy probability in areas with steep slopes and greater cover of evergreen forest but decreasing occupancy probability in areas with greater cover of deciduous open tree canopy. The occupancy prediction model and sampling method presented in this study have applications for future long‐term monitoring to aid in the conservation of Mexican spotted owls and other aurally distinguishable species.