Cryptic wide‐ranging movements lead to upwardly biased occupancy in a territorial species

Abstract

Abstract Occupancy modelling is useful for inferring population status and dynamics when occupancy reflects the presence of established individuals or populations, such as residents detected at breeding sites. However, if occupancy is assumed to reflect the presence of established residents but reflects transient movements from nonresidents at truly unoccupied sites, inferences about population status will be overly optimistic. In population monitoring, detections arising from dispersing or wide‐ranging individuals could be considered “false positives” because they do not reflect true contributions to local occupancy dynamics. The spotted owl Strix occidentalis is one of the most studied endangered species in the world and motivated the development of occupancy models to monitor populations. Because spotted owls are site‐faithful, it is assumed that detections within known breeding areas represent occupancy of established residents. We evaluated this assumption by (a) characterizing the extent to which GPS‐marked owls used multiple nest/roost areas, and (b) using detection/nondetection data to estimate the effect of false positive detections generated by wide‐ranging movements on occupancy. Thirty‐one of 36 GPS‐marked owls (86%) used nest/roost areas other than their own, and 11 owls (30%) used five or more. In occupancy analyses, 8% of all detections were confirmed to be false positives of colour‐marked wide‐ranging individuals and 20% of all detections were classified as “ambiguous” and, therefore, potential false positives. On average, failing to account for false positive detections upwardly biased occupancy by a factor of 1.29 (95% CI: 1.23–1.34). However, in the year following a large, severe fire that affected 30 of 84 owl nest/roost areas in our study area, failing to account for false positives upwardly biased occupancy by a factor of 1.65 (95% CI: 1.17–2.12). Synthesis and applications. If unrecognized or unaccounted for, “false positive” detections of wide‐ranging individuals outside their predicted territories can generate upward biases in occupancy and potentially mask effects of ecological disturbances. We recommend researchers minimize the potential for movement‐related false positive detections by developing study designs that reduce the frequency of such detections, excluding false positive detections using field data when possible, and explicitly modelling false positive errors.