Résumé
An adequate knowledge of the growth rate of a population often is needed in conservation biology and population management. In long-lived species, annual adult survival is the demographic parameter that has the strongest influence on population growth rate. Adult survival often is estimated by capture-recapture methods under the restrictive assumption that all individuals in a given group have the same survival and recapture probabilities. Violation of this assumption, i.e. heterogeneity among individuals, tends to bias survival estimates. In particular, heterogeneous capture probabilities independent of survival probabilities tend to negatively bias survival estimates. However, a cautious use of capture-recapture methods allows recognition of the problem and an accurate estimation of survival. We estimated adult survival in a population of Black-headed Gulls (Larus ridibundus) breeding in central France based on resightings of banded birds. The estimated survival was lower in the year after the first resighting than afterwards. We did not find any substantial biological explanation for this result (in particular, it was difficult to connect it with the existence of prospecting individuals). However, heterogeneity in the resighting probability, which is very likely in this population, could explain why apparent survival seemed lower in the year immediately after the first resighting. The higher value of the survival estimate (0.90) when capture-rate heterogeneity is accounted for is discussed relative to the growth regime of the population and habitat instability.