Decoding Owl Calls: Refining Occupancy Inference From Passive Acoustic Monitoring

Many territorial species use vocalizations as a primary form of territory defense, and the areas actively defended do not necessarily correspond with an individual's home range. As passive acoustic monitoring becomes a primary population assessment method for soniferous species, often in combination with occupancy modeling, effective conservation will require more detailed information on species-specific space use to refine interpretation. Northern spotted owls (Strix occidentalis caurina) actively vocalize during the breeding season, but interpreting acoustic data is complicated by variable detectability-particularly for females-and interference from competitive barred owls (Strix varia). Using a dense network of autonomous recording units deployed 0-3 km from known northern spotted owl activity centers, we quantified vocal activity by sex, reproductive status, landscape features, and intensity of barred owl vocalizations. Additionally, we compared detections from overlapping regional monitoring sites to refine our understanding of detectability. Male territorial calls were detected more frequently and consistently than female territorial calls. Female calls were infrequent and restricted to the activity center and immediately adjacent areas, especially when nesting. Vocal space use areas were similar in size but smaller than published home ranges, reinforcing that territorial calls represent high-use areas, not full spatial use. We propose a detection-based spectrum of weeks with detection for inferring occupancy that accounts for calling rate, caller sex, and project objectives. Stricter thresholds can minimize false positives in population assessments, while inclusive thresholds reduce false negatives when used to determine habitat protection. Our results support nuanced, objective-based thresholds for interpreting northern spotted owl detections from passive acoustic monitoring. This approach balances accuracy with conservation risk tolerance, elucidating that acoustically inferred territory does not reflect total landscape used. By clarifying vocal behavior, this study advances the application of passive acoustic monitoring for habitat management and occupancy estimation amid intense interspecies pressures and ongoing landscape change.