Bias in little owl population estimates using playback techniques during surveys

Abstract

To test the efficiency of playback methods to survey little owl (Athene noctua) populations we carried out two studies: (1) we recorded the replies of radio–tagged little owls to calls in a small area; (2) we recorded call broadcasts to estimate the effectiveness of the method to detect the presence of little owl. In the first study, we detected an average of 8.12 owls in the 30′ survey period, a number that is close to the real population; we also detected significant little owl movements from the initial location (before the playback) to the next locations during the survey period. However, we only detected an average of 2.25 and 5.37 little owls in the first 5′ and 10′, respectively, of the survey time. In the second study, we detected 137 little owl territories in 105 positive sample units. The occupation rate was 0.35, the estimated occupancy was 0.393, and the probability of detection was 0.439. The estimated cumulative probability of detection suggests that a minimum of four sampling times would be needed in an extensive survey to detect 95% of the areas occupied by little owls.