Use of mixture models to characterize site quality in philopatric animals: a case study with Bonelli's eagle

The French population of the Bonelli’s eagle Aquila fasciata has suffered a sharp decline over the last decades, resulting in the implementation of several conservation actions since 1990. In parallel, a capture–recapture (CR) program has been conducted to monitor the population. Field practitioners suspected that certain breeding sites were responsible for most of the adult deaths and were interested in confirming this hypothesis in order to focus their actions on these particular sites. Our study sought to determine whether there was indeed a quantifiable difference in the quality of the eagle’s breeding sites. Since Bonelli’s eagles are highly faithful to their breeding site, we tested whether the turnover of breeding individuals observed at different sites resulted from a mixture of two different turnover rates. We also used the recent multi-event CR modeling approach to test whether or not we could detect adult survival heterogeneity in the population. Our study revealed that there was a difference in breeding-site quality a decade ago, before large efforts were made to insulate high-voltage power lines, but failed to detect a difference in breeding-site quality more recently. No survival heterogeneity was detected. These counterintuitive results could be due to a lack of statistical power or because the conservation actions performed over the last 20 years cloud the issue. When no a priori information is available on habitat quality differences, we advocate for the use of a mixture modeling approach since it is consistent and fairly general.