Robust and reliable estimates of demographic parameters are essential to understand population dynamics. Natal dispersal is a common process in monitored populations and can cause underestimations of survival and dispersal due to permanent emigration. Here, we present a multistate Bayesian capture-mark-recapture approach based on a joint estimation of natal dispersal kernel and detection probabilities to address biases in survival, dispersal, and related demographic parameters when dispersal information is limited. We implement this approach to long-term data of a threatened population: the Bonelli's eagle in Catalonia (SW Europe). To assess the method's performance, we compare demographic estimates structured by sex, age, and breeding status in cases of limited versus large data scales, with those of classical models where dispersal and detection probabilities are estimated separately. Results show substantial corrections of demographic estimates. Natal dispersal and permanent emigration probabilities were larger in females, and consequently, female non-breeder survival showed larger differences between separate and joint estimation models. Moreover, our results suggest that estimates are sensitive to the choice of the dispersal kernel, fat-tailed kernels providing larger values in cases of data limitation. This study provides a general multistate framework to model demographic parameters while correcting permanent emigration biases caused by natal dispersal.
© 2023. The Author(s).