Fall and rise of a threatened raptor: Unraveling long-term population dynamics with spatially explicit integrated models.

Population dynamics are governed by the so-called four BIDE processes: birth, immigration, death, and emigration. However, most population models fail to explicitly consider all four processes, which may hinder a comprehensive understanding of how and why populations change over time. The advent of Integrated Population Models (IPMs) and recent developments in spatial mark-recapture models have enabled deeper insights into demography and dispersal. In this study, we merged both kinds of models into a spatially explicit IPM. By integrating count, reproduction, mark-recapture, and dispersal data, this framework permitted the separate modeling of all BIDE processes, which subsequently allowed (1) a fine-scale estimation of population dynamics and (2) the estimation of central population parameters and stages that have traditionally been elusive in demographic studies but are key to applied conservation, such as the long-term dynamics of floaters (sexually mature non-breeders), sink-source status, and dispersal processes. Using this approach, we carried out a fine-scale assessment of the long-term dynamics and demographic drivers of one long-lived Bonelli's eagle population from Western Europe (1986-2020). Our results illustrated a considerable population decline and subsequent recovery alongside multiple demographic insights scarcely documented to date in long-lived species. First, we reported a decrease and subsequent increase in floater numbers probably associated with parallel changes in the breeding population, hence contributing to the scarce empirical knowledge available about the role and dynamics of floaters. Second, we detected a change in average population functioning from a sink to a neutral contributor, thus shedding light on the flexibility and drivers of sink-source dynamics. Third, we underscored the central role of non-breeder survival for population recovery, suggesting that long-lived species conservation action should not only focus on adult or breeding populations, as is typically the case. Fourth, we quantified the magnitudes and variations of local and dispersal processes in the long term and discussed their potential implications in terms of management implementation. Overall, our study highlights the potential of spatially explicit IPMs to build more complete assessments of population dynamics, contribute to better-informed conservation action, and help fill knowledge gaps in ecological sciences.