MODELING DISPERSAL WITH CAPTURE–RECAPTURE DATA: DISENTANGLING DECISIONS OF LEAVING AND SETTLEMENT

Increasing habitat fragmentation has led population ecologists to consider dispersal as a key process for population functioning and persistence. However, due to logistic and methodological problems, dispersal rates within fragmented populations have rarely been estimated, leaving many unanswered questions about the role of dispersal and the factors affecting dispersal. Recently developed multistrata capture–recapture models allow the estimation of dispersal probabilities from the analysis of recaptures or recoveries of marked individuals. Despite their undoubted efficiency at estimating dispersal rates, their contribution to the investigation of dispersal processes at the individual level is presently limited by a lack of flexibility in their parameterization. In this paper, a new parameterization of these models is presented in which the probability of leaving the site of origin and the probability that an emigrant settles on a given recipient site are modeled separately. This parameterization is then used to address the influence of local perturbations on site fidelity and settlement decisions of emigrants in a subdivided population of Black-headed Gulls, Larus ridibundus. The parameterization introduced here accurately describes dispersal probabilities whenever they result from a two-step decision-making process (i.e., when choice of a settlement site only occurs after a decision has been made to leave the current site). Furthermore, this parameterization permits investigators to disentangle hypotheses about these two components of dispersal behavior. Corresponding Editor: P. Legendre.