Heterogeneity in detection probability along the breeding season in Black-legged Kittiwakes: implications for sampling design

Abstract

In wild animal population studies, capture heterogeneity is likely to be prevalent and can reduce the accuracy of vital rate estimates. Here, we test how individual detection probabilities vary through the breeding season in a population of a cliff-nesting colonial seabird, the Black-legged Kittiwake (Rissa tridactyla). Specifically, we expected detection probability to be affected by changes in brooding behavior and nest attendance associated with the breeding phenology and the local breeding performance of individuals. As predicted, we found that strong heterogeneities in detection probability can occur in relation to the breeding performance of individuals, the breeding performance of their neighbors, and the timing of surveys. Detection probability is highest and most homogeneous at the beginning of the breeding season. Later in the season, it is lower and can vary dramatically among groups of breeding individuals. A simulation approach was used to assess the implications of these results for the performance (bias and precision) of different study designs. Clearly, investing sampling effort early in the season is an efficient way to improve the accuracy of parameter estimates in this species. Our findings stress the importance of establishing study designs that take into account the population and behavioral ecology of the focal species.