Abstract
During migration, migratory bird species often aggregate at the same stopover sites due to geographical features that channel migratory routes (coasts, valleys) or locally abundant food resources (e.g. reedbeds, fruiting bushes). In migration ecology, however, stopover behavior is often studied on a single species, limiting the generality of inferences and their application to habitat management. If stopover decisions are similar across co‐occurring species, characterizing stopover ecology at a community level could provide more comprehensive insights. Using recent advances in modelling, we adapted a multi‐species capture‐recapture model to 1) quantify synchrony in stopover departure probabilities across species and 2) identify covariates driving this synchrony. We applied this model to three migratory songbirds commonly captured at French stopover sites: the sedge warbler Acrocephalus schoenobaenu s, reed warbler Acrocephalus scirpaeus and bluethroat Luscinia svecica . Departure decisions were largely synchronous across species, with time since arrival (TSA) as the primary synchronizing factor and weather conditions as a secondary influence. High synchrony in departure timing produced waves of migrant departures and consistent patterns in departure decisions across species, suggesting shared fuelling and timing strategies. Although this phenomenon has been documented via visual counts and radar, it has not been formally quantified at the species‐level using capture‐recapture methods. Our model is flexible and can test hypotheses regarding spatial synchrony in departure decisions. With decades of capture–recapture data available across Europe and North America, our approach offers new potential for studying stopover ecology at the community level, over large geographic and temporal scales.